The Liver: Biology and Pathobiology [6th Edition] 9781119436836

Table of contents :
Dedication

List of Contributors

Preface

Acknowledgments

Part One: Introduction

1 Organizational Principles of the Liver
Peter Nagy, Snorri S. Thorgeirsson, Joe W. Grisham

2 Embryonic Development of the Liver
Kenneth S. Zaret, Roque Bort, Stephen A. Duncan

Part Two: The Cells

Section A: Cell Biology of the Liver

3 Cytoskeletal Motors: Structure and Function in Hepatocytes
Mukesh Kumar, Arnab Gupta, Roop Mallik

4 Hepatocyte Surface Polarity
Anne Muesch, Irwin Arias

5 Primary Cilia
Carolyn M. Ott

6 Endocytosis in Liver Function and Pathology
Micah B. Schott, Barbara Schroeder, Mark A. McNiven

7 The Hepatocellular Secretory Pathway
Catherine Jackson, Mark A. McNiven

8 Mitochondrial Function, Dynamics, and Quality Control
Marc Liesa, Ilan Benador, Nathanael Miller, Orian S. Shirihai

9 Nuclear Pore Complex
Michelle A. Veronin, Joseph S. Glavy

10 Protein Maturation and Processing at the Endoplasmic Reticulum
Ramanujan S. Hegde

11 Protein Degradation and the Lysosomal System
Susmita Kaushik, Ana Maria Cuervo

12 Peroxisome Assembly, Degradation, and Disease
Rong Hua, Peter K. Kim

13 Organelle–Organelle Contacts: Origins and Functions
Uri Manor

14 Gap and Tight Junctions in Liver: Structure, Function and Pathology
John W. Murray, David C. Spray

15 Ribosome Assembly and its Role in Cell Growth and Proliferation in the Liver
Katherine I. Farley-Barnes, Susan J. Baserga

16 miRNAs and Hepatocellular Carcinoma
Yusuke Yamamoto, Isaku Kohama, Takahiro Ochiya

17 Hepatocyte Apoptosis: Mechanisms and relevance in liver diseases
Gregory J. Gores, Harmeet Malhi

Section B: The Hepatocyte

18 Copper Metabolism and the Liver\
Cynthia Abou Zeid, Ling Yi, Stephen G. Kaler

19 The Central Role of the Liver in Iron Storage and Regulation of Systemic Iron Homeostasis
Tracey A. Rouault, Victor Gordeuk, Gregory J. Anderson

20 Disorders of Bilirubin Metabolism
Namita Roy Chowdhury, Yanfeng Li, Jayanta Roy Chowdhury

21 Hepatic Lipid Droplets in Liver Function and Disease
Douglas G. Mashek, Wenqi Cui, Linshan Shang, Charles P. Najt

22 Lipoprotein Metabolism and Cholesterol Balance
Mariana Acuña-Aravena, David E. Cohen

Section C: Transporters, Bile Acids, and Cholestasis

23 Bile Acids Metabolism in Health and Disease: An update
Tiangang Li, John Y. L. Chiang

24 TGR5 (GPBAR1) in the Liver
Verena Keitel, Christoph G.W. Gertzen, Lina Spomer, Holger Gohlke, Dieter Häussinger

25 Bile Acids as Signaling Molecules
Thierry Claudel, Michael Trauner

26 Hepatic Adenosine Triphosphate-Dependent (ABC) Transporters and Their Role in Physiology
Peter L.M. Jansen

27 Basolateral Plasma Membrane Organic Anion Transporters
M. Sawkat Anwer, Allan W. Wolkoff

28 Hepatic Nuclear Receptors
Raymond E. Soccio

29 Molecular Cholestasis
Paul Gissen, Richard J. Thompson

30 Pathophysiologic Basis for Alternative Therapies for Cholestasis
Claudia D. Fuchs, Emina Halilbasic, Michael Trauner

31 Adaptive Regulation of Hepatocyte Transporters in Cholestasis
James L. Boyer

Section D: Non-Hepatocyte Cells

32 Cholangiocyte Biology and Pathobiology
Massimiliano Cadamuro, Romina Fiorotto, Mario Strazzabosco

33 Polycystic Liver Diseases: Genetics, Mechanisms, and Therapies
Tatyana Masyuk, Anatoliy Masyuk, Nicholas LaRusso

34 The Liver Sinusoidal Endothelial Cell: Basic Biology and Pathobiology
Karen K. Sørensen, Bård Smedsrød

35 Fenestrations in the Liver Sinusoidal Endothelial Cell
Victoria C Cogger, Nicholas J Hunt, David G Le Couteur

36 Stellate Cells and Fibrosis
Youngmin A. Lee, Scott L. Friedman

Part 3: Functions of the Liver

Section A: Metabolic Functions

37 Nonalcoholic Fatty Liver Disease and Insulin Resistance
Max C. Petersen, Varman T. Samuel, Kitt Falk Petersen, Gerald I. Shulman

38 AMPK: Central regulator of glucose and lipid metabolism and target of type 2 diabetes therapeutics
Daniel Garcia, Maria M. Mihaylova, Reuben J. Shaw

39 Insulin-mediated PI3K and AKT Signaling
Jeffrey E. Pessin, Hyokjoon Kwon

40 Ca2+ Signaling in the Liver
Mateus T. Guerra, M. Fatima Leite, Michael H. Nathanson

41 Clinical Genomics of NAFLD
Frank Lammert

Section B: Liver Growth and Regeneration

42 Stem Cell-Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis
Wencheng Zhang, Amanda Allen, Eliane Wauthier, Xianwen Yi, Homayoun Hani, Praveen Sethupathy, David Gerber, Vincenzo Cardinale, Guido Carpino, Juan Dominguez-Bendala, Giacomo Lanzoni, Domenico Alvaro, Eugenio Gaudio, Lola Reid

43 Developmental Morphogens and Adult Liver Repair
Mariana Machado, Anna Mae Diehl

44 Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology
David A. Shafritz, Markus Grompe

45 Liver Regeneration
George K. Michalopoulos

46 β-Catenin Signaling
Satdarshan P. S. Monga

47 Polyploidy in Liver Function, Mitochondrial Metabolism and Cancer
Evan R. Delgado, Elizabeth C. Stahl, Nairita Roy, Patrick D. Wilkinson, Andrew W. Duncan

Part 4: Pathobiology of Liver Disease

48 Hepatic Encephalopathy
Roger F. Butterworth

49 The Kidney in Liver Disease
Moshe Levi, Shogo Takahashi, Xiaoxin X. Wang, Marilyn E. Levi

50 Alpha-1 Antitrypsin Deficiency
David A. Rudnick, David H. Perlmutter

51 Pathophysiology of Portal Hypertension
Roberto J. Groszman, Yasuko Iwakiri

52 Nonalcoholic Fatty Liver Disease: Mechanisms and treatment
Yaron Rotman, Devika Kapuria

53 Alcoholic Liver Disease
Bin Gao, Xiaogang Xiang, Lorenzo Leggio, George F. Koob

54 Drug-induced Liver Injury
Lily Dara, Neil Kaplowitz

55 Oxidative Stress and Inflammation in the Liver
John J. Lemasters, Hartmut Jaeschke

56 The Role of Bile Acid-mediated Inflammation in Cholestatic Liver Injury
Shi-Ying Cai, Man Li, James L. Boyer

57 Toll-like Receptors in Liver Disease
So Yeon Kim, Ekihiro Seki

Part 5: Liver Cancer

58 Experimental Models of Liver Cancer: Genomic assessment of experimental models
Sun Young Yim, Jae-Jun Shim, Bo Hwa Sohn, Ju-Seog Lee

59 Epidemiology of Liver Cancer
Hashem B. El-Serag

60 Mutations and Genomic Alterations in Liver Cancer
Jessica Zucman-Rossi, Jean-Charles Nault

61 Treatment of Liver Cancer
Tim F. Greten

Part 6: Hepatitis

62 Molecular Biology of Hepatitis Viruses
Christoph Seeger, William S. Mason, Michael M.C. Lai

63 Immune Mechanisms of Viral Clearance and Disease Pathogenesis during Viral Hepatitis
Carlo Ferrari, Valeria Barili, Stefania Varchetta, Mario U. Mondelli

64 Clinical Implications of the Molecular Biology of Hepatitis B Virus
Timothy M. Block, Ju-Tao Guo, W. Thomas London

65 Viral Escape Mechanisms in Hepatitis C and the Clinical Consequences of Persistent Infection
Marc G. Ghany, Christopher M. Walker, Patrizia Farci

66 Tracking Hepatitis C Virus Interactions with the Hepatic Lipid Metabolism: A hitchhikerás guide to solve remaining translational research challenges in hepatitis C
Thomas Pietschmann, Gabrielle Vieyres

67 Nucleoside Antiviral Agents for HCV: What’s left to do?
Franck Amblard, Seema Mengshetti, Junxing Shi, Sijia Tao, Leda Bassit, Raymond F. Schinazi

68 Hepatitis E Virus: An emerging zoonotic virus causing acute and chronic liver disease
Xiang-Jin Meng

69 Biological Principles and Clinical Issues Underlying Liver Transplantation for Viral-Induced End-stage Liver Disease in the Era of Highly Effective Direct-Acting Antiviral Agents
Michael S. Kriss, James R. Burton, Jr., Hugo R. Rosen

70 Time for the Elimination of Hepatitis C Virus as a Global Health Threat
John W. Ward, Alan R. Hinman, Harvey J. Alter

Part 7: Horizons

H1 Genome Editing by Targeted Nucleases and the CRISPR/Cas Revolution
Shawn M. Burgess

H2 Imaging Cellular Proteins and Structures: Smaller, Brighter, and Faster
Aubrey V. Weigel, Erik Lee Snapp

H3 Liver-directed Gene Therapy
Patrik Asp, Chandan Guha, Namita Roy Chowdhury, Jayanta Roy Chowdhury

H4 Telomeres and Telomerase in Liver Generation and Cirrhosis
Sonja C. Schätzlein, K. Lenhard Rudolph

H5 Toxins and Biliary Atresia
Michael Pack, Rebecca G. Wells

H6 The Dual Role of ABC Transporters in Drug Metabolism and Resistance to Chemotherapy
Jean-Pierre Gillet, Marielle Boonen, Michel Jadot, Michael M. Gottesman

H7 Stem-cell-derived Liver Cells: From model system to therapy
Helmuth Gehart, Hans Clevers

H8 Extracellular Vesicles and Exosomes: Biology and pathobiology
Gyongyi Szabo, Fatemeh Momen-Heravi

H9 Integrated Technologies for Liver Tissue Engineering
Tiffany N. Vo, Amanda X. Chen, Quinton B. Smith, Arnav Chhabra, Sangeeta N. Bhatia

H10 Stem Cells, IPS Cells, Reprogramming: Therapy promise
Stephen A. Duncan

H11 Chromatin Regulation and Transcription Factor Cooperation in Liver Cells
Ido Goldstein

H12 Drug Interactions in the Liver
Guruprasad P. Aithal, Gerd A. Kullak-Ublick

H13 Metabolic Regulation of Hepatic Growth
Wolfram Goessling

H14 The Gut Microbiome and Liver Disease
Lexing Yu, Jasmohan S. Baja, Robert F. Schwabe

H15 Lineage Tracing: Efficient tools to determine the fate of hepatic cells in health and disease
Frédéric Lemaigre

H16 The Hepatocyte as a Household for Plasmodium Parasites
Vanessa Zuzarte-Luis, Maria M. Mota

Index

Citation preview

The Liver

Dedication

This book is dedicated to Win Arias, whose enthusiasm, insight, and scientific rigor have served as an inspiration to several generations of investigators, providing the essential foundation and

tools for building bridges between basic and clinical hepatologists as they elucidate together the mysteries of liver function in health and disease.

The Liver Biology and Pathobiology Sixth Edition Edited by Irwin M. Arias MD Emeritus Senior Scientist, National Institutes of Health, Bethesda, MD, USA Emeritus Professor of Physiology, Tufts University School of Medicine, Boston, MA, USA Emeritus Professor of Medicine, Albert Einstein College of Medicine, Bronx, NY, USA

Harvey J. Alter MD, MACP Distinguished NIH Scientist Emeritus, Department of Transfusion Medicine, National Institutes of Health, Bethesda, MD, USA

James L. Boyer MD Ensign Professor of Medicine, Department of Internal Medicine and Liver Center, Yale University School of Medicine, New Haven, CT, USA

David E. Cohen MD, PhD Vincent Astor Distinguished Professor of Medicine Chief, Division of Gastroenterology and Hepatology, Joan & Sanford I. Weill Department of Medicine, Weill Cornell Medical College Co‐Director, Center for Advanced Digestive Care, New York‐Presbyterian Hospital and Weill Cornell Medical Center, New York, NY, USA

David A. Shafritz MD Professor of Medicine, Cell Biology & Pathology, Albert Einstein College of Medicine Associate Director, Marion Bessin Liver Research Center, Bronx, NY, USA

Snorri S. Thorgeirsson MD, PhD Senior Scientist, Laboratory of Human Carcinogenesis, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA

Allan W. Wolkoff MD The Herman Lopata Chair in Liver Disease Research Professor of Medicine and Anatomy and Structural Biology Associate Chair of Medicine for Research Chief, Division of Hepatology Director, Marion Bessin Liver Research Center Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY, USA

This edition first published 2020 © 2020 John Wiley & Sons Ltd Edition History John Wiley & Sons Ltd. (5e, 2009) All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions. The right of Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff to be identified as the authors of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Office 9600 Garsington Road, Oxford, OX4 2DQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats. Limit of Liability/Disclaimer of Warranty The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting scientific method, diagnosis, or treatment by physicians for any particular patient. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Library of Congress Cataloging‐in‐Publication Data Names: Arias, Irwin M., editor. Title: The liver : biology and pathobiology / edited by Irwin M. Arias [and six others]. Other titles: Liver (Arias) Description: Sixth edition. | Hoboken, NJ : Wiley-Blackwell, 2020. | Includes bibliographical references and index. Identifiers: LCCN 2019024961 (print) | LCCN 2019024962 (ebook) | ISBN 9781119436829 (hardback) |   ISBN 9781119436836 (epub) | ISBN 9781119436843 (adobe pdf) Subjects: MESH: Liver–physiopathology Classification: LCC QP185 (print) | LCC QP185 (ebook) | NLM WI 702 | DDC 612.3/52–dc23 LC record available at https://lccn.loc.gov/2019024961 LC ebook record available at https://lccn.loc.gov/2019024962 Cover images: main image and background image courtesy of Sandor Paku; top left image courtesy of Scott Friedman; top middle images courtesy of Sandor Paku; top right image courtesy of Peter Kim Cover design by Wiley Set in 9.5/11.5pt Times LT Std by SPi Global, Pondicherry, India 10 9 8 7 6 5 4 3 2 1

ffirs.indd 4

11/30/2019 9:33:25 AM

Contents

List of Contributors

x

Preface xx Acknowledgments xxi PART ONE:  INTRODUCTION

1

  1 Organizational Principles of the Liver Peter Nagy, Snorri S. Thorgeirsson, and Joe W. Grisham

3

  2 Embryonic Development of the Liver Kenneth S. Zaret, Roque Bort, and Stephen A. Duncan

14

PART TWO:  THE CELLS

23

SECTION A:  CELL BIOLOGY OF THE LIVER

25

  3 Cytoskeletal Motors: Structure and Function in Hepatocytes Mukesh Kumar, Arnab Gupta, and Roop Mallik

27

  4 Hepatocyte Surface Polarity Anne Müsch and Irwin M. Arias

36

  5 Primary Cilia Carolyn M. Ott

50

  6 Endocytosis in Liver Function and Pathology Micah B. Schott, Barbara Schroeder, and Mark A. McNiven

62

  7 The Hepatocellular Secretory Pathway Catherine L. Jackson and Mark A. McNiven

75

  8 Mitochondrial Function, Dynamics, and Quality Control Marc Liesa, Ilan Benador, Nathanael Miller, and Orian S. Shirihai

86

  9 Nuclear Pore Complex Michelle A. Veronin and Joseph S. Glavy

94

10 Protein Maturation and Processing at the Endoplasmic Reticulum Ramanujan S. Hegde

108

vi Contents

11 Protein Degradation and the Lysosomal System Susmita Kaushik and Ana Maria Cuervo

122

12 Peroxisome Assembly, Degradation, and Disease Rong Hua and Peter K. Kim

137

13 Organelle–Organelle Contacts: Origins and Functions Uri Manor

151

14 Gap and Tight Junctions in Liver: Structure, Function, and Pathology John W. Murray and David C. Spray

160

15 Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver Katherine I. Farley‐Barnes and Susan J. Baserga

174

16 miRNAs and Hepatocellular Carcinoma Yusuke Yamamoto, Isaku Kohama, and Takahiro Ochiya

183

17 Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases Harmeet Malhi and Gregory J. Gores

195

SECTION B:  THE HEPATOCYTE

207

18 Copper Metabolism and the Liver Cynthia Abou Zeid, Ling Yi, and Stephen G. Kaler

209

19 The Central Role of the Liver in Iron Storage and Regulation of Systemic Iron Homeostasis Tracey A. Rouault, Victor R. Gordeuk, and Gregory J. Anderson

215

20 Disorders of Bilirubin Metabolism Namita Roy Chowdhury, Yanfeng Li, and Jayanta Roy Chowdhury

229

21 Hepatic Lipid Droplets in Liver Function and Disease Douglas G. Mashek, Wenqi Cui, Linshan Shang, and Charles P. Najt

245

22 Lipoprotein Metabolism and Cholesterol Balance Mariana Acuña‐Aravena and David E. Cohen

255

SECTION C:  TRANSPORTERS, BILE ACIDS, AND CHOLESTASIS

269

23 Bile Acid Metabolism in Health and Disease: An Update Tiangang Li and John Y.L. Chiang

271

24 TGR5 (GPBAR1) in the Liver Verena Keitel, Christoph G.W. Gertzen, Lina Spomer, Holger Gohlke, and Dieter Häussinger

286

25 Bile Acids as Signaling Molecules Thierry Claudel and Michael Trauner

299

26 Hepatic Adenosine Triphosphate‐Binding Cassette Transport Proteins and Their Role in Physiology Peter L.M. Jansen

313

27 Basolateral Plasma Membrane Organic Anion Transporters M. Sawkat Anwer and Allan W. Wolkoff

327

28 Hepatic Nuclear Receptors Raymond E. Soccio

337

29 Molecular Cholestasis Paul Gissen and Richard J. Thompson

351

30 Pathophysiologic Basis for Alternative Therapies for Cholestasis Claudia D. Fuchs, Emina Halilbasic, and Michael Trauner

364

Contents

vii

31 Adaptive Regulation of Hepatocyte Transporters in Cholestasis James L. Boyer

378

SECTION D:  NON‐HEPATOCYTE CELLS

391

32 Cholangiocyte Biology and Pathobiology Massimiliano Cadamuro, Romina Fiorotto, and Mario Strazzabosco

393

33 Polycystic Liver Diseases: Genetics, Mechanisms, and Therapies Tatyana Masyuk, Anatoliy Masyuk, and Nicholas LaRusso

408

34 The Liver Sinusoidal Endothelial Cell: Basic Biology and Pathobiology Karen K. Sørensen and Bård Smedsrød

422

35 Fenestrations in the Liver Sinusoidal Endothelial Cell Victoria C. Cogger, Nicholas J. Hunt, and David G. Le Couteur

435

36 Stellate Cells and Fibrosis Youngmin A. Lee and Scott L. Friedman

444

PART THREE:  FUNCTIONS OF THE LIVER

455

SECTION A:  METABOLIC FUNCTIONS

457

37 Non‐alcoholic Fatty Liver Disease and Insulin Resistance Max C. Petersen, Varman T. Samuel, Kitt Falk Petersen, and Gerald I. Shulman

459

38 AMPK: Central Regulator of Glucose and Lipid Metabolism and Target of Type 2 Diabetes Therapeutics Daniel Garcia, Maria M. Mihaylova, and Reuben J. Shaw

472

39 Insulin‐Mediated PI3K and AKT Signaling Hyokjoon Kwon and Jeffrey E. Pessin

485

40 Ca2+ Signaling in the Liver Mateus T. Guerra, M. Fatima Leite, and Michael H. Nathanson

496

41 Clinical Genomics of NAFLD Frank Lammert

509

SECTION B:  LIVER GROWTH AND REGENERATION 521 42 Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis Wencheng Zhang, Amanda Allen, Eliane Wauthier, Xianwen Yi, Homayoun Hani, Praveen Sethupathy, David Gerber, Vincenzo Cardinale, Guido Carpino, Juan Dominguez‐Bendala, Giacomo Lanzoni, Domenico Alvaro, Eugenio Gaudio, and Lola Reid

523

43 Developmental Morphogens and Adult Liver Repair Mariana Verdelho Machado and Anna Mae Diehl

539

44 Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology David A. Shafritz and Markus Grompe

550

45 Liver Regeneration George K. Michalopoulos

566

46 β‐Catenin Signaling Satdarshan P.S. Monga

585

47 Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer Evan R. Delgado, Elizabeth C. Stahl, Nairita Roy, Patrick D. Wilkinson, and Andrew W. Duncan

603

viii Contents

PART FOUR:  PATHOBIOLOGY OF LIVER DISEASE

615

48 Hepatic Encephalopathy Roger F. Butterworth

617

49 The Kidney in Liver Disease Moshe Levi, Shogo Takahashi, Xiaoxin X. Wang, and Marilyn E. Levi

630

50 α1‐Antitrypsin Deficiency David A. Rudnick and David H. Perlmutter

645

51 Pathophysiology of Portal Hypertension Yasuko Iwakiri and Roberto J. Groszmann

659

52 Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment Yaron Rotman and Devika Kapuria

670

53 Alcoholic Liver Disease Bin Gao, Xiaogang Xiang, Lorenzo Leggio, and George F. Koob

682

54 Drug‐Induced Liver Injury Lily Dara and Neil Kaplowitz

701

55 Oxidative Stress and Inflammation in the Liver John J. Lemasters and Hartmut Jaeschke

714

56 The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury Shi‐Ying Cai, Man Li, and James L. Boyer

728

57 Toll‐like Receptors in Liver Disease So Yeon Kim and Ekihiro Seki

737

PART FIVE:  LIVER CANCER

747

58 Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models Sun Young Yim, Jae‐Jun Shim, Bo Hwa Sohn, and Ju‐Seog Lee

749

59 Epidemiology of Hepatocellular Carcinoma Hashem B. El‐Serag

758

60 Mutations and Genomic Alterations in Liver Cancer Jessica Zucman‐Rossi and Jean‐Charles Nault

773

61 Treatment of Liver Cancer Tim F. Greten

782

PART SIX:  HEPATITIS

793

62 Molecular Biology of Hepatitis Viruses Christoph Seeger, William S. Mason, and Michael M.C. Lai

795

63 Immune Mechanisms of Viral Clearance and Disease Pathogenesis During Viral Hepatitis Carlo Ferrari, Valeria Barili, Stefania Varchetta, and Mario U. Mondelli

821

64 Clinical Implications of the Molecular Biology of Hepatitis B Virus Timothy M. Block, Ju‐Tao Guo, and W. Thomas London

851

65 Viral Escape Mechanisms in Hepatitis C and the Clinical Consequences of Persistent Infection Marc G. Ghany, Christopher M. Walker, and Patrizia Farci

868

66 Tracking Hepatitis C Virus Interactions with the Hepatic Lipid Metabolism: A Hitchhiker’s Guide to Solve Remaining Translational Research Challenges in Hepatitis C Gabrielle Vieyres and Thomas Pietschmann

889

Contents

ix

67 Nucleoside Antiviral Agents for HCV: What’s Left to Do? Franck Amblard, Seema Mengshetti, Junxing Shi, Sijia Tao, Leda Bassit, and Raymond F. Schinazi

906

68 Hepatitis E Virus: An Emerging Zoonotic Virus Causing Acute and Chronic Liver Disease Xiang‐Jin Meng

915

69 Biological Principles and Clinical Issues Underlying Liver Transplantation for Viral‐Induced End‐Stage Liver Disease in the Era of Highly Effective Direct‐Acting Antiviral Agents Michael S. Kriss, James R. Burton, Jr., and Hugo R. Rosen

926

70 Time for the Elimination of Hepatitis C Virus as a Global Health Threat John W. Ward, Alan R. Hinman, and Harvey J. Alter

935

PART SEVEN:  HORIZONS

953

71 Genome Editing by Targeted Nucleases and the CRISPR/Cas Revolution Shawn M. Burgess

955

72 Imaging Cellular Proteins and Structures: Smaller, Brighter, and Faster Aubrey V. Weigel and Erik Lee Snapp

965

73 Liver‐Directed Gene Therapy Patrik Asp, Chandan Guha, Namita Roy Chowdhury, and Jayanta Roy Chowdhury

979

74 Telomeres and Telomerase in Liver Generation and Cirrhosis Sonja C. Schätzlein and K. Lenhard Rudolph

992

75 Toxins and Biliary Atresia Michael Pack and Rebecca G. Wells

1000

76 The Dual Role of ABC Transporters in Drug Metabolism and Resistance to Chemotherapy Jean‐Pierre Gillet, Marielle Boonen, Michel Jadot, and Michael M. Gottesman

1007

77 Stem Cell‐Derived Liver Cells: From Model System to Therapy Helmuth Gehart and Hans Clevers

1015

78 Extracellular Vesicles and Exosomes: Biology and Pathobiology Gyongyi Szabo and Fatemeh Momen‐Heravi

1022

79 Integrated Technologies for Liver Tissue Engineering Tiffany N. Vo, Amanda X. Chen, Quinton B. Smith, Arnav Chhabra and Sangeeta N. Bhatia

1028

80 Pluripotent Stem Cells and Reprogramming: Promise for Therapy James A. Heslop and Stephen A. Duncan

1036

81 Chromatin Regulation and Transcription Factor Cooperation in Liver Cells Ido Goldstein

1043

82 Drug Interactions in the Liver Guruprasad P. Aithal and Gerd A. Kullak‐Ublick

1050

83 Metabolic Regulation of Hepatic Growth Wolfram Goessling

1058

84 The Gut Microbiome and Liver Disease Lexing Yu, Jasmohan S. Bajaj, and Robert F. Schwabe

1062

85 Lineage Tracing: Efficient Tools to Determine the Fate of Hepatic Cells in Health and Disease Frédéric Lemaigre

1069

86 The Hepatocyte as a Household for Plasmodium Parasites Vanessa Zuzarte‐Luis and Maria M. Mota

1075

Index 1081

List of Contributors

Cynthia Abou Zeid Section on Translational Neuroscience, Molecular Medicine Branch Intramural Research Program, National Institutes of Health, Bethesda, MD, USA

M. Sawkat Anwer Tufts Clinical and Translational Science Institute Tufts University Cummings School of Veterinary Medicine Department of Biomedical Sciences North Grafton, MA, USA

Mariana Acuña‐Aravena Division of Gastroenterology and Hepatology, Joan & Sanford I. Weill Department of Medicine Weill Cornell Medical College New York, NY, USA

Irwin M. Arias National Institutes of Health Bethesda, MD, USA

Guruprasad P. Aithal Nottingham Digestive Diseases Centre, School of Medicine, University of Nottingham, Nottingham, UK; NIHR Nottingham Biomedical Research Centre, Nottingham University Hospitals NHS Trust and the University of Nottingham, Nottingham, UK Amanda Allen Department of Cell Biology and Physiology University of North Carolina School of Medicine Chapel Hill, NC, USA Harvey J. Alter Department of Transfusion Medicine, Clinical Center National Institutes of Health Bethesda, MD, USA Domenico Alvaro Department of Medico‐Surgical Sciences and Biotechnologies, Sapienza University of Rome, Latina, Italy Department of Medicine and Medical Specialties Sapienza University of Rome Rome, Italy Franck Amblard Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA Gregory J. Anderson QIMR Berghofer Medical Research Institute Brisbane, Queensland, Australia

Patrik Asp Marion Bessin Liver Research Center Department of Surgery, Albert Einstein College of Medicine Bronx, NY, USA Jasmohan S. Bajaj Virginia Commonwealth University and McGuire VA Medical Center Richmond, VA, USA Valeria Barili Unit of Infectious Diseases and Hepatology Department of Medicine and Surgery, University of Parma Parma, Italy Susan J. Baserga Department of Molecular Biophysics & Biochemistry Department of Genetics Department of Therapeutic Radiology, Yale University School of Medicine New Haven, CT, USA Leda Bassit Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA Ilan Benador Department of Medicine, Division of Endocrinology and Department of Molecular and Medical Pharmacology David Geffen School of Medicine at UCLA Los Angeles, CA, USA



List of Contributors

Sangeeta N. Bhatia Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA; Harvard‐MIT Department of Health Sciences and Technology, Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Boston, MA; Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA; David H. Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA; Howard Hughes Medical Institute, Chevy Chase, MD, USA Timothy M. Block Baruch S. Blumberg Institute Doylestown, PA, USA Marielle Boonen Laboratory of Intracellular Trafficking Biology, URPhyM, NARILIS Faculty of Medicine, University of Namur Namur, Belgium Roque Bort Instituto de Investigación Sanitaria La Fe (IIS La Fe) Unidad de Hepatología experimental València, Spain James L. Boyer Department of Internal Medicine and Liver Center, Yale University School of Medicine New Haven, CT, USA Shawn M. Burgess National Human Genome Research Institute National Institutes of Health Bethesda, MD, USA James R. Burton, Jr. University of Colorado School of Medicine Aurora, CO, USA Roger F. Butterworth Department of Medicine University of Montreal Montreal, QC, Canada Massimiliano Cadamuro Department of Molecular Medicine University of Padua, Padova, Italy; International Center for Digestive Health (ICDH) University of Milan‐Bicocca, Monza, Italy Shi‐Ying Cai The Liver Center, Yale University School of Medicine New Haven, CT, USA Vincenzo Cardinale Department of Medico‐Surgical Sciences and Biotechnologies Sapienza University of Rome Latina, Italy

xi

Guido Carpino Department of Movement, Human and Health Sciences, Division of Health Sciences University of Rome “Foro Italico”; Department of Anatomical, Histological, Forensic Medicine and Orthopedics Sciences Sapienza University of Rome Rome, Italy Amanda X. Chen Department of Biological Engineering, Massachusetts Institute of Technology Cambridge, MA, USA Arnav Chhabra Harvard‐MIT Department of Health Sciences and Technology, Institute for Medical Engineering and Science, Massachusetts Institute of Technology Boston, MA, USA John Y.L. Chiang Department of Integrative Medical Sciences Northeast Ohio Medical University Rootstown, OH, USA Thierry Claudel Hans Popper Laboratory of Molecular Hepatology, Division of Gastroenterology and Hepatology Department of Internal Medicine III Medical University of Vienna Vienna, Austria Hans Clevers Oncode Institute, Hubrecht Institute, Royal Netherlands Academy of Arts and Sciences (KNAW) and University Medical Centre (UMC) Utrecht; Princess Máxima Centre for Paediatric Oncology Utrecht, The Netherlands Victoria C. Cogger Centre for Education and Research on Ageing University of Sydney and Concord RG Hospital Sydney, NSW, Australia David E. Cohen Division of Gastroenterology and Hepatology, Joan & Sanford I. Weill Department of Medicine Weill Cornell Medical College New York, NY, USA Ana Maria Cuervo Department of Developmental and Molecular Biology Institute for Aging Research, Albert Einstein College of Medicine Bronx, NY, USA Wenqi Cui Department of Biochemistry, Molecular Biology and Biophysics University of Minnesota Minneapolis, MN, USA

xii

List of Contributors

Lily Dara Research Center for Liver Disease, Department of Medicine, Division of Gastrointestinal and Liver Diseases Keck School of Medicine, University of Southern California Los Angeles, CA, USA Evan R. Delgado Department of Pathology McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center University of Pittsburgh Pittsburgh, PA, USA Anna Mae Diehl School of Medicine, Duke University Durham, NC, USA Juan Dominguez‐Bendala Department of Medicine and Medical Specialties Sapienza University of Rome Rome, Italy; Diabetes Research Institute, Miller School of Medicine University of Miami Miami, FL, USA Andrew W. Duncan Department of Pathology, McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center, University of Pittsburgh Pittsburgh, PA, USA Stephen A. Duncan Department of Regenerative Medicine and Cell Biology Medical University of South Carolina Charleston, SC, USA Hashem B. El‐Serag Department of Medicine Baylor College of Medicine and Michael E. DeBakey Veterans Affairs Medical Center Houston, TX, USA Patrizia Farci Hepatic Pathogenesis Section, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health Bethesda, MD, USA Katherine I. Farley‐Barnes Department of Molecular Biophysics & Biochemistry Yale University School of Medicine New Haven, CT, USA Carlo Ferrari Unit of Infectious Diseases and Hepatology Department of Medicine and Surgery, University of Parma Parma, Italy Romina Fiorotto International Center for Digestive Health (ICDH), University of Milan‐Bicocca Monza, Italy;

Liver Center and Section of Digestive Diseases, Department of Internal Medicine, Section of Digestive Diseases, Yale University School of Medicine New Haven, CT, USA Scott L. Friedman Division of Liver Diseases Icahn School of Medicine at Mount Sinai New York, NY, USA Claudia D. Fuchs Hans Popper Laboratory of Molecular Hepatology, Division of Gastroenterology and Hepatology Department of Internal Medicine III Medical University of Vienna Vienna, Austria Bin Gao Laboratory of Liver Diseases, National Institute on Alcohol Abuse and Alcoholism National Institutes of Health Bethesda, MD, USA Daniel Garcia Molecular and Cell Biology Laboratory The Salk Institute for Biological Studies La Jolla, CA, USA Eugenio Gaudio Department of Anatomical, Histological, Forensic Medicine and Orthopedics Sciences Sapienza University of Rome Rome, Italy Helmuth Gehart Oncode Institute, Hubrecht Institute, Royal Netherlands Academy of Arts and Sciences (KNAW) and University Medical Centre (UMC) Utrecht Utrecht, The Netherlands David Gerber Department of Surgery University of North Carolina School of Medicine Chapel Hill, NC, USA Christoph G.W. Gertzen Institute of Pharmaceutical and Medicinal Chemistry, Heinrich‐Heine University Düsseldorf, Düsseldorf, Germany Marc G. Ghany Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, MD, USA Jean‐Pierre Gillet Laboratory of Molecular Cancer Biology, URPhyM, NARILIS Faculty of Medicine, University of Namur Namur, Belgium Paul Gissen UCL Great Ormond Street Institute of Child Health and Great Ormond Street Hospital for Children, London, UK



List of Contributors

Joseph S. Glavy University of Texas at Tyler Fisch College of Pharmacy Tyler, TX, USA Wolfram Goessling Division of Gastroenterology Massachusetts General Hospital Harvard‐MIT Division of Health Sciences and Technology Harvard Medical School Boston, MA, USA Holger Gohlke Institute of Pharmaceutical and Medicinal Chemistry, Heinrich‐Heine University Düsseldorf, Düsseldorf, Germany Ido Goldstein Institute of Biochemistry, Food Science and Nutrition The Robert H. Smith Faculty of Agriculture, Food and Environment The Hebrew University of Jerusalem Rehovot, Israel Victor R. Gordeuk University of Illinois at Chicago Chicago, IL, USA Gregory J. Gores College of Medicine, Division of Gastroenterology and Hepatology Mayo Clinic Rochester, MN, USA

Chandan Guha Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY; Departments of Radiation Oncology and Pathology, Albert Einstein College of Medicine Bronx, NY, USA Ju‐Tao Guo Baruch S. Blumberg Institute Doylestown, PA, USA Arnab Gupta Department of Biological Sciences, Indian Institute of Science Education and Research Kolkata, Mohanpur, India Emina Halilbasic Hans Popper Laboratory of Molecular Hepatology, Division of Gastroenterology and Hepatology Department of Internal Medicine III Medical University of Vienna Vienna, Austria Homayoun Hani Department of Cell Biology and Physiology University of North Carolina School of Medicine Chapel Hill, NC, USA Dieter Häussinger Clinic for Gastroenterology, Hepatology and Infectious Diseases, University Hospital Düsseldorf Medical Faculty at Heinrich‐Heine‐University Düsseldorf, Germany

Michael M. Gottesman Laboratory of Cell Biology, Center for Cancer Research National Cancer Institute, National Institutes of Health Bethesda, MD, USA

Ramanujan S. Hegde MRC Laboratory of Molecular Biology Cambridge, UK

Tim F. Greten Thoracic and Gastrointestinal Malignancies Branch Center for Cancer Research, National Cancer Institute Bethesda, MD, USA

James A. Heslop Department of Regenerative Medicine and Cell Biology Medical University of South Carolina Charleston, SC, USA

Joe W. Grisham Department of Pathology and Laboratory Medicine University of North Carolina Chapel Hill, NC, USA

Alan R. Hinman Task Force for Global Health Decatur, GA, USA

Markus Grompe Papé Pediatric Research Institute Oregon Health Sciences University Portland, OR, USA Roberto J. Groszman Section of Digestive Diseases, Yale School of Medicine New Haven, CT, USA Mateus T. Guerra Departments of Medicine and Cell Biology Yale University School of Medicine New Haven, CT, USA

xiii

Rong Hua Cell Biology Program, Hospital for Sick Children Toronto, ON, Canada Nicholas J. Hunt Centre for Education and Research on Ageing University of Sydney and Concord RG Hospital Sydney, NSW, Australia Yasuko Iwakiri Section of Digestive Diseases Yale University School of Medicine New Haven, CT, USA

xiv

List of Contributors

Catherine L. Jackson Institut Jacques Monod, UMR7592 CNRS Université Paris‐ Diderot, Sorbonne Paris Cité Paris, France

Isaku Kohama Division of Molecular and Cellular Medicine, National Cancer Center Research Institute Tsukiji, Tokyo, Japan

Michel Jadot Laboratory of Physiological Chemistry, URPhyM, NARILIS, Faculty of Medicine University of Namur, Belgium

George F. Koob National Institute on Alcohol Abuse and Alcoholism and National Institute on Drug Abuse National Institutes of Health Bethesda, MD, USA

Hartmut Jaeschke Department of Pharmacology, Toxicology and Therapeutics University of Kansas Medical Center Kansas City, KS, USA Peter L.M. Jansen Department of Hepatology and Gastroenterology Academic Medical Center Amsterdam, The Netherlands; LiSyM research network University of Freiburg Freiburg, Germany Stephen G. Kaler Section on Translational Neuroscience, Molecular Medicine Branch Intramural Research Program, National Institutes of Health Bethesda, MD, USA Neil Kaplowitz Research Center for Liver Disease, Department of Medicine, Division of Gastrointestinal and Liver Diseases Keck School of Medicine, University of Southern California Los Angeles, CA, USA Devika Kapuria Liver Energy and Metabolism Section, Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, MD, USA Susmita Kaushik Department of Developmental and Molecular Biology Institute for Aging Research, Albert Einstein College of Medicine Bronx, NY, USA

Michael S. Kriss Division of Gastroenterology & Hepatology University of Colorado School of Medicine Aurora, CO, USA Gerd A. Kullak‐Ublick University Hospital Zurich and University of Zurich Zurich, Zurich, Switzerland; Mechanistic Safety, Chief Medical Office and Patient Safety, Novartis Global Drug Development Basel, Switzerland Mukesh Kumar Department of Biological Sciences, Tata Institute of Fundamental Research, Navy Nagar, Colaba, Mumbai, India Hyokjoon Kwon Department of Medicine, Division of Endocrinology, Metabolism and Nutrition Rutgers‐Robert Wood Johnson Medical School New Brunswick, NJ, USA Michael M.C. Lai China Medical University and China Medical University Hospital Taichung, Taiwan Frank Lammert Department of Medicine II Saarland University Medical Center Homburg, Germany Giacomo Lanzoni Diabetes Research Institute, Miller School of Medicine University of Miami Miami, FL, USA

Verena Keitel Clinic for Gastroenterology, Hepatology and Infectious Diseases, University Hospital Düsseldorf Medical Faculty at Heinrich‐Heine‐University Düsseldorf, Germany

Nicholas LaRusso Division of Gastroenterology and Hepatology Mayo Clinic College of Medicine Rochester, MN, USA

Peter K. Kim Cell Biology Program, The Hospital for Sick Children; Department of Biochemistry, University of Toronto Toronto, ON, Canada

David G. Le Couteur Centre for Education and Research on Ageing University of Sydney and Concord RG Hospital Sydney, NSW, Australia

So Yeon Kim Division of Digestive and Liver Diseases, Department of Medicine Cedars‐Sinai Medical Center Los Angeles, CA, USA

Ju‐Seog Lee Department of Systems Biology The University of Texas M.D. Anderson Cancer Center Houston, TX, USA



List of Contributors

Youngmin A. Lee Department of Surgery, Vanderbilt University Medical Center Nashville, TN, USA Lorenzo Leggio Section on Clinical Psychoneuroendocrinology and Neuropsychopharmacology, National Institute on Alcohol Abuse and Alcoholism and National Institute on Drug Abuse National Institutes of Health Bethesda, MD, USA M. Fatima Leite Department of Physiology and Biophysics UFMG Belo Horizonte, Brazil Frédéric Lemaigre Université catholique de Louvain, de Duve Institute Brussels, Belgium John J. Lemasters Departments of Drug Discovery and Biomedical Sciences and Biochemistry and Molecular Biology Medical University of South Carolina Charleston, SC, USA Marilyn E. Levi Department of Medicine, Division of Infectious Diseases, University of Colorado Aurora, CO, USA Moshe Levi Department of Biochemistry and Molecular and Cellular Biology Georgetown University Washington, DC, USA W. Thomas London (deceased) Formerly, Fox Chase Cancer Center Philadelphia, PA, USA Man Li The Liver Center, Yale University School of Medicine New Haven, CT, USA Tiangang Li Department of Pharmacology, Toxicology and Therapeutics University of Kansas Medical Center Kansas City, KS, USA Yanfeng Li Departments of Medicine and Genetics and Marion Bessin Liver Research Center Albert Einstein College of Medicine, Bronx, NY, USA Marc Liesa Department of Medicine, Division of Endocrinology and Department of Molecular and Medical Pharmacology David Geffen School of Medicine at UCLA Los Angeles, CA, USA

xv

Mariana Verdelho Machado Gastroenterology and Hepatology Department Hospital de Santa Maria CHLN, Lisbon; Faculty of Medicine Lisbon University Lisbon, Portugal Harmeet Malhi College of Medicine, Division of Gastroenterology and Hepatology Mayo Clinic Rochester, MN, USA Roop Mallik Department of Biological Sciences, Tata Institute of Fundamental Research, Navy Nagar, Colaba, Mumbai, India Uri Manor Waitt Advanced Biophotonics Center, Salk Institute for Biological Studies La Jolla, CA, USA Douglas G. Mashek Department of Biochemistry, Molecular Biology and Biophysics and Department of Medicine, Division of Diabetes, Endocrinology and Metabolism University of Minnesota Minneapolis, MN, USA William S. Mason Fox Chase Cancer Center Philadelphia, PA, USA Anatoliy Masyuk Division of Gastroenterology and Hepatology Mayo Clinic College of Medicine Rochester, MN, USA Tatyana Masyuk Division of Gastroenterology and Hepatology Mayo Clinic College of Medicine Rochester, MN, USA Mark A. McNiven Department of Biochemistry and Molecular Biology, Division of Gastroenterology and Hepatology, Mayo Clinic Rochester, MN, USA Xiang‐Jin Meng Department of Biomedical Sciences and Pathobiology Virginia Polytechnic Institute and State University Blacksburg, VA, USA Seema Mengshetti Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA George K. Michalopoulos Department of Pathology University of Pittsburgh School of Medicine Pittsburgh, PA, USA

xvi

List of Contributors

Maria M. Mihaylova Molecular and Cell Biology Laboratory The Salk Institute for Biological Studies La Jolla, CA, USA Nathanael Miller Department of Medicine, Division of Endocrinology and Department of Molecular and Medical Pharmacology David Geffen School of Medicine at UCLA Los Angeles, CA, USA Fatemeh Momen‐Heravi Department of Medicine University of Massachusetts Medical School Worcester, MA, USA

Liver Unit, Hôpital Jean Verdier, Hôpitaux Universitaires Paris‐Seine‐Saint‐Denis, Assistance‐Publique Hôpitaux de Paris, Bondy, France Takahiro Ochiya Division of Molecular and Cellular Medicine, National Cancer Center Research Institute Tsukiji, Tokyo Institute of Medical Science, Tokyo Medical University Shinjuku, Tokyo, Japan Janelia Research Campus, Ashburn, VA, USA Carolyn M. Ott Janelia Research Campus Ashburn, VA, USA

Mario U. Mondelli Division of Infectious Diseases and Immunology, Fondazione IRCCS Policlinico S.Matteo, Pavia; Department of Internal Medicine and Therapeutics University of Pavia, Pavia, Italy

Michael Pack Departments of Medicine (GI Division) and Cell and Developmental Biology Perelman School of Medicine, University of Pennsylvania Philadelphia, PA, USA

Satdarshan P.S. Monga Pittsburgh Liver Research Center University of Pittsburgh, School of Medicine and UPMC Pittsburgh, PA, USA

David H. Perlmutter Department of Pediatrics, Division of Gastroenterology, Hepatology, and Nutrition and Department of Developmental Biology, Washington University School of Medicine in St. Louis, St. Louis Children’s Hospital, St. Louis, MO, USA

Maria M. Mota Instituto de Medicina Molecular João Lobo Antunes, Faculty of Medicine, University of Lisbon, Lisbon, Portugal Anne Müsch Department of Developmental & Molecular Biology Albert Einstein College of Medicine Bronx, NY, USA John W. Murray Marion Bessin Liver Research Center Albert Einstein College of Medicine; Department of Anatomy and Structural Biology, Albert Einstein College of Medicine Bronx, NY, USA

Jeffrey E. Pessin Albert Einstein‐Mount Sinai Diabetes Research Center and the Fleischer Institute for Diabetes and Metabolism, Albert Einstein College of Medicine, Bronx, NY; Department of Medicine and Department of Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, NY, USA Kitt Falk Petersen Department of Internal Medicine, Section of Endocrinology Yale University School of Medicine New Haven, CT, USA

Peter Nagy First Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest, Hungary

Max C. Petersen Department of Internal Medicine, Section of Endocrinology, and Department of Molecular and Cellular Physiology Yale University School of Medicine New Haven, CT, USA

Charles P. Najt Department of Biochemistry, Molecular Biology and Biophysics University of Minnesota Minneapolis, MN, USA

Thomas Pietschmann Institute of Experimental Virology, TWINCORE, Centre for Experimental and Clinical Infection Research, Hannover; German Centre for Infection Research (DZIF) Braunschweig, Germany

Michael H. Nathanson Departments of Medicine and Cell Biology Yale University School of Medicine New Haven, CT, USA Jean‐Charles Nault Centre de Recherche des Cordeliers, Sorbonne Université, Inserm, Université de Paris, Functional Genomics of Solid Tumors Laboratory, Paris, France

Lola Reid Department of Cell Biology and Physiology Program in Molecular Biology and Biotechnology University of North Carolina School of Medicine Chapel Hill, NC, USA Hugo R. Rosen University of Southern California Keck School of Medicine Los Angeles, CA, USA



List of Contributors

Yaron Rotman Liver Energy and Metabolism Section, Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, MA, USA Tracey A. Rouault Eunice Kennedy Shriver National Institute of Child Health and Human Development National Institutes of Health Bethesda, MD, USA Nairita Roy Department of Pathology McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center University of Pittsburgh Pittsburgh, PA, USA Jayanta Roy Chowdhury Departments of Medicine and Genetics and Marion Bessin Liver Research Center Albert Einstein College of Medicine Bronx, NY, USA Namita Roy Chowdhury Departments of Medicine and Genetics and Marion Bessin Liver Research Center Albert Einstein College of Medicine Bronx, NY, USA David A. Rudnick Department of Pediatrics, Division of Gastroenterology, Hepatology, and Nutrition and Department of Developmental Biology, Washington University School of Medicine in St. Louis, St. Louis Children’s Hospital St. Louis, MO, USA K. Lenhard Rudolph Research Group on Stem Cell Aging, Leibniz Institute on Aging – Fritz Lipmann Institute (FLI) Jena, Germany Varman T. Samuel Department of Internal Medicine, Section of Endocrinology Yale University School of Medicine New Haven; Veterans Affairs Medical Center West Haven, CT, USA Sonja C. Schätzlein Spark@FLI, Leibniz Institute on Aging – Fritz Lipmann Institute (FLI) Jena, Germany Raymond F. Schinazi Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA Micah B. Schott Department of Biochemistry and Molecular Biology, Division of Gastroenterology and Hepatology, Mayo Clinic Rochester, MN, USA

xvii

Barbara Schroeder Department of Biochemistry and Molecular Biology, Division of Gastroenterology and Hepatology, Mayo Clinic Rochester, MN, USA Robert F. Schwabe Columbia University New York, NY, USA Christoph Seeger Fox Chase Cancer Center Philadelphia, PA, USA Ekihiro Seki Division of Digestive and Liver Diseases, Department of Medicine, Cedars‐Sinai Medical Center, Los Angeles, CA; Department of Biomedical Sciences, Cedars‐Sinai Medical Center, Los Angeles, CA, USA Praveen Sethupathy Department of Genetics University of North Carolina School of Medicine Chapel Hill, NC; Department of Biomedical Sciences Cornell University Ithaca, NY, USA David A. Shafritz Marion Bessin Liver Research Center Albert Einstein College of Medicine Bronx, NY, USA Linshan Shang Department of Biochemistry, Molecular Biology and Biophysics University of Minnesota Minneapolis, MN, USA Reuben J. Shaw Molecular and Cell Biology Laboratory The Salk Institute for Biological Studies La Jolla, CA, USA Junxing Shi Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA Jae‐Jun Shim Department of Internal Medicine, Kyung Hee University College of Medicine Seoul, Korea Orian S. Shirihai Department of Medicine, Division of Endocrinology and Department of Molecular and Medical Pharmacology David Geffen School of Medicine at UCLA Los Angeles, CA, USA Gerald I. Shulman Department of Internal Medicine, Section of Endocrinology Department of Molecular and Cellular Physiology and Howard Hughes Medical Institute Yale University School of Medicine New Haven, CT, USA

xviii

List of Contributors

Bård Smedsrød Vascular Biology Research Group, Department of Medical Biology UiT The Arctic University of Norway Tromsø, Norway

Gyongyi Szabo Beth Israel Deaconess Medical Center Harvard Medical School Boston, MA, USA

Quinton B. Smith Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA, USA

Shogo Takahashi Department of Biochemistry and Molecular and Cellular Biology Georgetown University Washington, DC, USA

Erik Lee Snapp Janelia Research Campus of the Howard Hughes Medical Institute Ashburn, VA, USA

Sijia Tao Laboratory of Biochemical Pharmacology, Emory University School of Medicine Atlanta, GA, USA

Raymond E. Soccio University of Pennsylvania Perelman School of Medicine, Department of Medicine Division of Endocrinology, Diabetes, and Metabolism Institute for Diabetes, Obesity, and Metabolism Philadelphia Crescenz VA Medical Center Philadelphia, PA, USA

Richard J. Thompson Institute of Liver Studies, King’s College London London, UK

Bo Hwa Sohn Department of Systems Biology The University of Texas M.D. Anderson Cancer Center Houston, TX, USA

Michael Trauner Hans Popper Laboratory of Molecular Hepatology, Division of Gastroenterology and Hepatology Department of Internal Medicine III Medical University of Vienna Vienna, Austria

Karen K. Sørensen Vascular Biology Research Group, Department of Medical Biology UiT The Arctic University of Norway Tromsø, Norway Lina Spomer Clinic for Gastroenterology, Hepatology and Infectious Diseases, University Hospital Düsseldorf Medical Faculty at Heinrich‐Heine‐University Düsseldorf, Germany David C. Spray Marion Bessin Liver Research Center, Albert Einstein College of Medicine Department of Neuroscience, Albert Einstein College of Medicine Department of Medicine, Albert Einstein College of Medicine Bronx, NY, USA Elizabeth C. Stahl Department of Pathology McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center University of Pittsburgh Pittsburgh, PA, USA Mario Strazzabosco International Center for Digestive Health (ICDH), University of Milan‐Bicocca, Monza, Italy; Liver Center and Section of Digestive Diseases, Department of Internal Medicine, Section of Digestive Diseases, Yale University School of Medicine, New Haven, CT, USA

Snorri S. Thorgeirsson Laboratory of Human Carcinogenesis, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA

Stefania Varchetta Division of Infectious Diseases and Immunology, Fondazione IRCCS Policlinico S.Matteo, Pavia; Department of Internal Medicine and Therapeutics University of Pavia, Italy Michelle A. Veronin University of Texas at Tyler Fisch College of Pharmacy Tyler, TX, USA Gabrielle Vieyres Institute of Experimental Virology, TWINCORE, Centre for Experimental and Clinical Infection Research Hannover, Germany Tiffany N. Vo Institute for Medical Engineering and Science, Massachusetts Institute of Technology Cambridge, MA, USA Christopher M. Walker Center for Vaccines and Immunity, The Research Institute at Nationwide Children’s Hospital, and College of Medicine, The Ohio State University Columbus, OH, USA Xiaoxin X. Wang Department of Biochemistry and Molecular and Cellular Biology Georgetown University Washington, DC, USA



List of Contributors

John W. Ward Task Force for Global Health, Decatur, GA; Centers for Disease Control and Prevention Atlanta, GA, USA Eliane Wauthier Department of Cell Biology and Physiology University of North Carolina School of Medicine Chapel Hill, NC, USA Aubrey V. Weigel Janelia Research Campus of the Howard Hughes Medical Institute Ashburn, VA, USA Rebecca G. Wells Departments of Medicine (GI Division) and Pathology and Laboratory Medicine, Perelman School of Medicine, and Bioengineering, School of Engineering and Applied Sciences University of Pennsylvania Philadelphia, PA, USA Patrick D. Wilkinson Department of Pathology McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center University of Pittsburgh Pittsburgh, PA, USA Allan W. Wolkoff Division of Gastroenterology and Liver Diseases Marion Bessin Liver Research Center Albert Einstein College of Medicine and Montefiore Medical Center Bronx, NY, USA Xiaogang Xiang Laboratory of Liver Diseases, National Institute on Alcohol Abuse and Alcoholism National Institutes of Health Bethesda, MD, USA; Department of Infectious Diseases, Ruijin Hospital, School of Medicine Shanghai Jiao Tong University Shanghai, China Yusuke Yamamoto Division of Molecular and Cellular Medicine, National Cancer Center Research Institute Tsukiji, Tokyo, Japan

Ling Yi Section on Translational Neuroscience, Molecular Medicine Branch Intramural Research Program, National Institutes of Health, Bethesda, MD, USA Xianwen Yi Department of Surgery University of North Carolina School of Medicine Chapel Hill, NC, USA Sun Young Yim Department of Internal Medicine, Korea University College of Medicine Seoul, Korea Lexing Yu Second Military Medical University Shanghai, China Kenneth S. Zaret Institute for Regenerative Medicine, Department of Cell and Developmental Biology Perelman School of Medicine, University of Pennsylvania Philadelphia, PA, USA Wencheng Zhang Department of Cell Biology and Physiology University of North Carolina School of Medicine Chapel Hill, NC, USA Jessica Zucman‐Rossi Centre de Recherche des Cordeliers, Sorbonne Université, Inserm, Université de Paris, Functional Genomics of Solid Tumors Laboratory, Paris, France; Hôpital Européen Georges Pompidou, Assistance Publique‐ Hôpitaux de Paris Paris, France Vanessa Zuzarte‐Luis Instituto de Medicina Molecular João Lobo Antunes Faculty of Medicine, University of Lisbon Lisbon, Portugal

xix

Preface

The pace of discoveries in basic biomedical sciences and engineering and their application to diagnosis and treatment of liver disease continues to exceed greatly the expectations expressed in the Preface to the previous editions published in 1982, 1988, 1994, 1999, and 2009. Concomitantly, the challenge addressed by this book has not changed since first appearing in the Preface to the first edition over 30 years ago: The amazing advances in fundamental biology that have occurred within the past two decades have brought hepatology and other disciplines into new, uncharted and exciting waters. The dynamic changes in biology will profoundly influence our ability to diagnose, treat and prevent liver disease. How can a student of the liver and its diseases maintain a link to these exciting advances? Most physicians lack the time to take post‐graduate courses in basic biology; most basic researchers lack an understanding of liver physiology and disease. This book strives to bridge the ever‐increasing gap between the advances in basic biology and their application to liver structure, function and disease.

Molecular biology was not the only great wave in contemporary science, nor is it surely the last. Remarkable advances in genetics and various omics are increasingly linked with dynamic super‐resolution light microscopy, which permits the study of cellular, molecular, and organ‐based physiology at nano‐levels. The expanding worlds of RNA structure and function, CRISPR‐ type gene editing, and chromatin biology coupled with single‐ cell and single‐molecule genomic analyses are facilitating discoveries with great importance in organ physiology and medicine, including personalized diagnosis and treatment. ­ Unexpected discoveries are certain to emerge from the ongoing bridge‐building between chemical and physical structural

analysis, engineered drug design, signaling networks, immune mechanisms and tolerance, the brain, and metabolic/digestive functions. Discoveries in these disciplines have already facilitated diagnosis, treatment, and improved clinical outcome of many liver diseases. Much more is undoubtedly yet to come. This sixth edition contains new chapters that present major progress that has been achieved in research laboratories and clinics around the world. All other chapters have been completely revised and updated. Following the death of our colleague Nelson Fausto, Snorri Thorgeirsson became an Associate Editor. Previous editions included a section called “Horizons,” devoted to extraordinary advances in areas of potentially major importance to the liver. Virtually all of these fields have rapidly expanded and become topics for later chapters. Sixteen new “Horizons” chapters are presented in this edition. One may safely predict that their impact on the field of hepatology will be considerable. As stated in the Preface to previous editions: The amazing advance in science proceeds at an ever‐increasing pace. The implications for students of liver disease are ­considerable. The authors and editors will have achieved our goals if the reader finds within this volume glimpses into the current state and future direction of our discipline and p­ erspectives that lead to better understanding of liver ­function and disease.

Irwin M. Arias Harvey J. Alter James L. Boyer David E. Cohen David A. Shafritz Snorri S. Thorgeirsson Allan W. Wolkoff

Acknowledgments

We thank the distinguished authors for their expertise, ­enthusiastic participation, and patience in responding to editorial suggestions. Appreciation is also extended to the staff

of Wiley Blackwell and also to the freelance project manager Gillian Whitley.

PART ONE: INTRODUCTION

1

Organizational Principles of the Liver Peter Nagy1, Snorri S. Thorgeirsson2, and Joe W. Grisham3 First Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest, Hungary Laboratory of Human Carcinogenesis, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD, USA 3 Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, NC, USA 1 2

PRINCIPLES OF LIVER STRUCTURE AND FUNCTION The liver is the largest organ of the mammalian body and has a highly versatile and complex function. Its specialized role is shown by the fact that, despite intense efforts, the activity of the liver cannot be replaced by artificial equipment. The liver par­ ticipates in the maintenance of the organism’s homeostasis as an  active, bidirectional biofilter. It is classed as bidirectional because it filters the portal blood that transports nutritional and toxic compounds from the environment through the gastrointes­ tinal tract and also filters the systemic blood (the body’s own products, e.g. bilirubin), providing the only channel of the body, the biliary system, through which non‐water‐soluble substances can be removed. It is classed as an active filter because it rapidly metabolizes most nutritional compounds and neutralizes and prepares for removal toxic exogenous (xenobiotics) and endog­ enous (worn out) materials. Because of these major functions the liver is constantly exposed to intense microbiological and antigenic stimuli which require function of the innate and adap­ tive immune systems. These diversified functions are executed by a structurally complex, multicellular tissue with a unique angioarchitecture, and by the combined and integrated activities of the participants. There are only two unique cell types in the liver – hepatocytes and biliary cells (or cholangiocytes). The hepatocytes are “the most valuable” parenchymal cells of the hepatic tissue. They do not constitute a homogeneous cell population. They are highly polarized cells (i.e. molecular specializations of the various ­surface membranes, including receptors, pumps, transport chan­ nels and carrier proteins) and their functions and to a certain extent morphology depend on their location in the parenchyma. This polarization makes the hepatocytes the logical center of the

liver. In addition, they perform the most complex metabolic tasks of the mammalian organism. The cholangiocytes form the channels that constitute the ­biliary system, which drains the parenchyma and guarantees the permanent flow of the bile, a highly toxic solution. Cholangiocytes also modify the composition of the bile and, in case of adverse conditions, can participate in repair mechanisms. These liver cells could not carry out their specific functions, of course, ­without the support of several “communal” cell types, which are  highly adapted to the special function and architecture of the liver. The endothelial cells of the parenchyma have a unique fenestrated structure and various different ­subpopulations can be  distinguished. There are several subpopulations of hepatic myofibroblasts as well. In addition to their mechanical functions the myofibroblasts can store special substances (e.g. vitamin A in stellate cells) and are a major source of growth factors and cytokines. The Kupffer cells are the resident macrophages in the liver. In addition to filtering the blood, they perform their tradi­ tional immunregulatory function. The presence of almost all subtypes of lymphocytes and dendritic cells makes the liver the largest organ of the immune system. The mesothelial cells of the Glisson capsule are, beside their mechanical function, an impor­ tant source of lymph production and can contribute to the gen­ eration of other hepatic cell types. The features of the hepatic extracellular matrix are unique. The components of the basement membrane are present around the sinusoids in an “unstructured” fashion, and cannot be detected by electron microscope, yet they can perform certain functions. Another fundamental feature of liver organization is its unique vascular pattern. Two afferent vessels supply blood to the liver: the portal vein and the hepatic artery. The blood of the  portal vein, having already “drained” the stomach, gut, ­pancreas, and spleen, is reduced in oxygen and pressure, and is

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

4

THE LIVER:  FUNCTIONAL ANATOMY OF THE LIVER

enriched in nutrients and toxic materials absorbed from the alimentary tract and in viscerally generated hormones and ­ growth factors. The arterial blood of the hepatic artery has sys­ temic levels of oxygen, pressure, and composition. The major function of the hepatic artery is to supply the peribiliary vascu­ lar plexus, the portal tract interstitium, the hepatic capsule, and the vasa vasorum of major vessels. In some species, the hepatic artery forms anastomosis with the branches of the portal vein, but even then this blood also ends up in the sinusoids. The blood of the liver is collected by one efferent draining system, the hepatic or “central” veins, which reach the systemic circulation via the inferior vena cava. The sinusoids form a very special vascular system, which is interposed between the afferent and efferent vessels. The large number and capacity of the sinusoids and the special arrangement of the supplying vessels provide a large volume of blood at a high flow rate via the large vessels with high compliance and capacity. At the same time the ­sinusoids are perfused with blood at low pressure and flow rate. These arrangements (i.e. low flow, specifically fenestrated (­perforated) endothelial cells, and the lack of the structured basement membrane) provide an especially efficient communi­ cation between the blood and hepatocytes. This is well illus­ trated by the pathological condition of liver cirrhosis, when the changes in hemodynamic condition (i.e. the “capillarization” of the sinusoids) disrupts this communication, resulting in severe dysfunction of the liver. Bile acids and their enterohepatic circulation are another good example of the cumulating functions. The bile acids are synthe­ tized in the hepatocytes by a complex biochemical process that requires 16 different enzymes, which are further modified by the gut microbiota. The primary physiological function of the bile acids is to convert lipid bilayers into micelles. This makes pos­ sible the excretion of important waste products from the blood. The bile acids also emulsify elements of the food in the gut and aid their ­absorption. In addition, bile acids act as signaling mol­ ecules, synchronizing the cooperation of the liver and gut. The different types of cells and vessels mentioned above can operate only if they are organized in a well “designed” structure. The most widely studied and analyzed morphological and func­ tional unit or module of the liver is the hepatic lobule. The popu­ larity of this structure for studies can be partly explained by the fact that lobules are outlined nicely in some species (pig, camel, bear) by connective tissue septa, and can therefore be easily ­recognized on the two‐dimensional histological sections com­ monly used in structural studies. The idealized lobule has a polygonal (usually hexagonal) shape. The terminal branch of the hepatic vein (central vein) is in the center of the lobule while the corners are occupied by the “portal triads.” The components of the triad are the interlobular bile ducts and the terminal branches of the portal vein and hepatic artery. The blood carried by these afferent vessels is distributed by the inlet venules and arteries along the virtual “vascular septa.” This vascular frame is filled up columns (or sheets in three‐dimensional space) of the hepatocytes constructed as “plates” arranged in a radial fashion. The hepatic plates are separated by the similarly distributed sinusoids. The blood runs in a centripetal direction from the ­vascular septa to the central vein. The vascular septa secure the mixing of the portal venous and arterial blood and the more‐ or‐less equal supply to the sinuses. The bile produced by the

hepatocytes runs in a centrifugal direction in the bile canalicules formed by neighboring hepatocytes and is collected by the inter­ lobular bile ducts of the portal triads. There is thus a countercur­ rent between the flow of the blood and bile at lobular level.

FUNCTIONAL ANATOMY OF THE LIVER Macroanatomy The liver is a continuous sponge‐like parenchymal mass ­penetrated by tunnels (lacunae) that contain the interdigitating networks of afferent and efferent vessels [1]. The adult human liver weighs from 1300 to 1700 g, depending on sex and body size. It is relatively small compared to other species (2% of the body weight) – in rat and mouse the liver is 4–5% of the body. In most mammalian species the liver is multilobed, the indi­ vidual lobes reflecting the distribution of the major branches of afferent and efferent blood vessels. In contrast, the human liver parenchyma is fused into a continuous parenchymal mass with two major lobes, right and left, delineated only by being supplied and drained by separate first‐ and second‐order branches of the portal and hepatic veins. Right and left lobes are topographically separated by the remnants of the embryonic umbilical vein (the falciform ligament), but this landmark does not locate the true anatomic division. Anatomically, the medial segment of the left lobe is located to the right side of the falciform ligament, ­centered on the anterior branches of the left portal vein. Interdigitation of first‐ and second‐order branches of the portal and hepatic veins produces eight macrovascular parenchymal segments centered on large portal veins and separated by large hepatic veins [2]. Hemodynamic watersheds or fissures separating afferent and efferent macrovascular segments permit the surgical resection of individual or adjacent segments. Liver transplantation and surgery has reached such a com­ plexity, however, that the traditional eight‐segment scheme is no longer sufficient. Detailed histological and imaging investiga­ tions have revealed that the number of second‐order branches given off by the left and right portal veins is much higher, and the mean of their number is 20, leading to the “1–2–20” concept of portal venous segmentation [3]. The recognition of the water­ shed septa between the variable actual segments is helped by intraoperative imaging techniques in real operative situations.

Microanatomy Normal liver function requires the unique arrangement of basic components of hepatic tissue: portal vein, hepatic artery, bile duct, hepatic vein, and hepatocytes. These form in two‐­ dimensional sheets the above‐mentioned hepatic (classical of Kiernan’s) lobules. Profiles of portal tracts and hepatic veins of various sizes are a prominent feature of liver histology [4–6]. Smaller branches of the afferent and efferent vessels (together with their stromal components) predominate in tissue sections taken from peripheral, subcapsular locations, whereas tissue sections taken from more proximal areas nearer to the hilum contain larger vascular structures  [6]. These vascular/stromal elements are contained in tunnels (lacunae) that penetrate the



1:  Organizational Principles of the Liver

Figure 1.1  Schematic drawing of the organization of blood vessels (arteries, red; portal veins, purple; central veins, blue; bile ducts, green; lumen of the biliary channels, including bile canaliculi, yellow) in two adjacent lobules of human liver. One sixth of a lobule is visible on the right and one third of a lobule is visible on the left. A terminal portal venule, arteriole, and bile ductule (canal of Hering) are present in the vascular septum between the lobules. The arteriole is connected directly to the sinusoids or enters the inlet venule. Bile is drained over the whole surface of the lobule. Arterioles and bile ductules are not present in the vascular septum in rodents. The bile is drained through canals of Hering connected to hepatocytes of the limiting plate. The arterioles anasto­ mose with the portal system at higher level as well. Courtesy of Sandor Paku, Semmelweis University, Budapest, Hungary.

parenchymal mass [4]. The hepatocytes arranged in plates fill in the space between the portal tracts and hepatic veins (Figure 1.1). The hepatic plates form brick‐like walls (muralia) of hepato­ cytes one cell (one brick) thick. The first hepatocytes of the hepatic plates form a virtual barrier between the periportal con­ nective tissue and the liver parenchyma called a limiting plate. The blood vessels and their investments of connective tissue provide the soft, spongy liver with its major structural support, or “skeleton.” Larger afferent vessels, portal veins, and hepatic arteries are contained together with bile ducts in connective ­tissue – the portal tracts – which are continuous with the mesen­ chymal components of the liver’s mesothelium‐covered surface capsule (Glisson’s capsule). Portal tracts also contain lymphatic vessels, nerves, and varying populations of other types of cells, such as macrophages, immunocytes, myofibroblasts, and pos­ sibly hematopoietic stem cells (see [7] and references therein). The collagenous investment of the efferent vessels is less robust and lacks large numbers of adventitious cells. The hepatic artery is distributed to the tissues of portal tracts, the liver capsule, and the walls of large vessels [4–6]. In portal tracts arterial branches form a capillary network (the peribiliary plexus) arborized around bile ducts [8, 9]. Efferent twigs from the peribiliary plexus empty into adjacent portal veins in rat and mouse but not in human and hamster [10]. The portal vein sup­ plies blood to the parenchymal mass through the so‐called inlet venules [9, 11]. In histological sections of mammalian liver, afferent and efferent vessels interdigitate regularly in an approximate ratio of 5–6 portal tracts for each profile of a hepatic vein, to form a pat­ tern of cross‐sections of portal tracts and hepatic veins separated by parenchyma [5, 6]. Most of the cross‐sectioned portal tracts

5

contain preterminal hepatic venules. These vessels represent the seventh‐ to tenth‐order branches from the hilar portal vein in large mammals, such as humans. These small portal tracts and hepatic (central) veins penetrate the parenchyma in nearly paral­ lel orientations about 0.5–1.0 mm apart. The portal inlet venules are very short vessels with no smooth muscle in their walls. They branch from preterminal and terminal venules at points on the circumference of the lobules at about 120 radial degrees (­triradial branching) and penetrate the parenchyma together with terminal arteriolar branches approximately perpendicular to and midway between two adjacent terminal hepatic venules [5, 6]. During their course through the parenchyma portal inlet venules break up completely into sinusoids, which are oriented more or less perpendicularly to the veins. Because they are hardly larger than sinusoids, the inlet venules are not conspicuous in humans and other mammals that lack a definite connective sheath around them. However, in adult swine their course through the paren­ chyma is clearly marked by connective tissue. Capillary‐size sinusoids occupy the smallest and most ­numerous tunnels (lacunae) in the parenchymal mass [4]. Unlike capillaries elsewhere, liver sinusoids are composed of endothe­ lial cells that are penetrated by holes (fenestrae) and lack a structured basal membrane [12], features that allow free egress of the fluid components and solutes of the perfusing blood. For example, tagged albumin has access to a space in the liver that is about 48% larger than the sinusoidal volume, in contrast to other tissues in which capillary space and albumin space are nearly the same [13]. In favorably oriented histological sections, more or less parallel, longitudinal profiles of sinusoids alternate with hepatic plates  [14]. A narrow cleft, called the space of Disse, separates sinusoids from hepatocytes located in adjacent hepatic plates [12, 15]. At their proximal (portal venous) ends, sinusoids are narrow and somewhat tortuous, whereas their mid­ dle and distal (hepatic venous) portions are larger and straighter [9, 16, 17]. Sinusoids and hepatic plates are disposed radially around the draining hepatic veins and extend directly to the ­supplying inlet venules [17]. Three‐dimensional reconstruction of the interlobular zone revealed the existence of a small vessel in this plane, the ­vascular septum, that serves as a starting pool for intralobular sinusoids. This is a hemodynamic barrier, a “watershed” between the two  neighboring lobules. This “interlobular” septum contains ­connective tissue matrix in pig, camel, bear, etc. and outlines the lobules nicely; it also exists in a rudimentary form in human liver [18]. The bile  –  the excretion product of the hepatocytes  –  is ­collected and transported in bile canaliculi, which are formed by the apical sides of two adjacent hepatocytes in the hepatic plate. The network of canaliculi is drained into the interlobular bile ducts through interface structures called canals of Hering. These are intermediary structures constructed partly by hepatocytes or cholangiocytes (Figure  1.2). Since these structures are the ­primary candidates to harbor the hepatic stem cell compartment, they are the subject of intensive investigations [19]. The distribu­ tion of canals of Hering shows variation among different spe­ cies. They are characterized by a distinct (EMA−/CD56+/CD133+) immunophenotype in humans, leave the periportal space and spread into the parenchyma along the rudimentary interlobular septa, and thus do not enter into the hepatic lobule  [18].

6

THE LIVER:  FUNCTIONAL ANATOMY OF THE LIVER

Figure 1.2  Normal human liver stained for CD10. The hepatocytes form trabeculae, 1–2 cells thick, radiating from a central vein. CD10 is expressed on the canalicular domain, indicating the polarization of the cells. Courtesy of Sandor Paku, Semmelweis University, Budapest, Hungary.

Figure 1.4  Normal human liver stained for pankeratin (green) and kera­ tin 7 (red). The dark lane in the center represents an interlobular vascular septum. The double positive (yellow) biliary ductules have several con­ nections with the limiting hepatocytes but do not enter into the lobules. Courtesy of Sandor Paku, Semmelweis University, Budapest, Hungary.

modular arrangement can improve the interpretation of lesions, especially in pathologically altered livers, but it is certainly not easy to transform the two‐dimensional observations into three‐ dimensional space.

Functional unit of the liver

Figure 1.3  Normal rat liver stained for pankeratin (green) and laminin (red), the nuclei are labeled by TOTO (blue). There is a cross‐section of a canal of Hering beside a small portal vein. The cytoplasm of the chol­ angiocytes is strongly positive for keratin and these cells are outlined by laminin (basement membrane) but basement membrane is absent at the pole where the ductile is connected to an adjacent hepatocyte. Courtesy of Sandor Paku, Semmelweis University, Budapest, Hungary.

The interlobular bile ducts are lined by a single layer of cuboidal cholangiocytes (Figure  1.3). They anastomose and unite larger septal and hilar branches. The connective tissue around the larg­ est biliary branches contain peribiliary glands which also secrete into the biliary tract (Figure 1.4). Teutsch and coworkers [20, 21] analyzed serial sections of rat and human livers to reconstruct the three‐dimensional structure of hepatic tissue. Although there were differences between the two species, the basic arrangement was similar. The reconstruc­ tion revealed primary “modules,” which constructed a more complex “secondary” module. The integration was based on a common drainage by branches of the hepatic veins and supply­ ing portal veins, and the modules were covered by continuous vascular septa. The primary modules correspond to the two‐ dimensional hepatic lobules. Quite a substantial variation in the shape and size of the modules was found, which provides ­morphogenetic plasticity to construct the whole organ. This

The concept of the primary functional unit of the liver has been the subject of debate for more than 350 years since its descrip­ tion by Wepfler in 1664 [22]. The first and most widely accepted traditional unit of the liver is “Kiernan’s lobule” [23], as described earlier. This is the efferent microvascular segment, being the smallest unit of parenchyma that is drained of blood by a single efferent (terminal hepatic or central) vein. It is quite easy to identify it, especially in species where they are outlined by connective tissue. The major criticism of the concept is that the terminal afferent vessels through the vascular septa contrib­ ute to the blood supply of adjacent lobules, and therefore the lobule cannot be a “basic functional unit.” Rappaport defined the basic unit as the compartment of the hepatic parenchyma sup­ plied with blood by a single terminal portal vein and called this unit the “liver acinus” [24]. Now we know that this unit is also supplied by a single terminal branch of the hepatic artery. The simple acinus is a parenchymal mass around a portal tract and it is drained by more hepatic venules. The acinus is subdivided into three zones, based on the distance from the portal vein. The distribution of these areas fits to the functional zonality of the hepatic parenchyma. Pathological lesions (e.g. steatosis or necrosis) also often follow this zonal pattern, which made this unit attractive. However, this zonality did not correspond per­ fectly to the distribution of enzyme activities and the hepatic modules described by Teutsch [20, 21] were also not compatible with the concept of the acinus. Matsumoto and his colleagues [6] investigated the angioarchi­ tecture of the human liver on thousands of serial sections, dis­ tinguishing conducting and parenchymal portions of the portal venous tree. A cone‐shaped parenchymal portion (primary lob­ ule) was defined which was supplied by a terminal portal venule.



1:  Organizational Principles of the Liver

The circulatory network of this unit was named the hepatic microcirculatory subunit (HMS). Ekataksin and Wake [25] showed that the afferent microvascular segment is also the small­ est unit of parenchyma drained of bile by terminal bile ducts [25], demonstrating that this hemodynamic segment is also the smallest excretory unit of parenchyma. The compartment of the hepatic parenchyma associated with an HMS contains all of the elementary structures of the hepatic tissue and may represent the elementary functional and morphological unit of the liver. It was named cholehepaton. The cholehepaton is also the smallest bile–blood unit and conforms with the principle of countercur­ rent flow. The cholehepaton has no anatomical or structural ­borders, and cannot be recognized on histological sections. It is defined by its function and mostly corresponds to Matsumoto’s primary lobules. Six of these Matsumoto’s primary lobules con­ stitute the secondary lobule, which is almost identical with the classical Kiernan’s lobule. This long and complicated detour returned us to the classical lobule, which is widely used today to analyze reactions of the hepatic tissue. However, it is worth remembering that there are other types of hepatic functional units and some workers contend that the hepatic tissue is an indivisible continuum that has no definable units [26].

Liver hemodynamics The hepatic vasculature is characterized by high capacity, high compliance, and low resistance [27]. Blood vessels comprise about 22% of the liver’s mass/volume [13] and the liver contains about 12% of the body’s total blood volume under physiological conditions [27], a sizeable fraction of which can be expelled by contraction of larger vessels by sympathetic nerve stimulation. In other words, the liver is a blood reservoir. The pressure of portal venous blood is reduced as the major afferent vessels dichotomize through the parenchyma, from about 130 mmH2O in the extrahe­ patic portal vein to about 60 mmH2O in the preterminal portal veins of the exteriorized liver of the anesthetized rat, amounting to about 60% of the total transhepatic pressure gradient [13]. A ­similar portal pressure gradient has been found in humans [27]. Blood flow through the liver amounts to about 1500–2000 mL min−1 in adult humans, about 25% of the resting cardiac output [27]. About 25% of the total liver blood flow is derived from the hepatic artery at prevailing arterial pressure and oxygenation. The portal venous blood (about 75% of total liver blood flow) arrives at the liver partially depleted of oxygen and at a reduced pressure as a consequence of having already perfused the splanchnic viscera. In aggregate, sinusoids comprise about 60% of the liver’s vascular volume, or about 13% of the total liver mass/volume [13]. A significant decrease in blood pressure occurs in sinu­ soids (about 40% of the transhepatic pressure gradient), the pressure declining to about 25 mmH2O in terminal hepatic veins of exteriorized liver of anesthetized rats [28]; the pressure gradi­ ent in the short sinusoids is especially steep. Blood pressure in the inferior vena cava approximates that in the terminal hepatic vein [27]. Consequently, although flow of blood through sinu­ soids faces little resistance, it is slow and somewhat intermittent and is assisted by negative pressure produced by respiratory expiration [27]. Possible mechanisms of regulating blood flow within sinusoids are controversial. The sinusoidal circulation is

7

regulated by a sphincter in the terminal arterioles [29]. In addi­ tion, hepatic stellate cells are the pericytes of the sinusoids and are the principal regulators of sinusoidal blood flow. They have receptors for molecular mediators regulating contraction and are closely associated with nerve endings, indicating neurogenic influence [30]. Sinusoids appear to have limited contractile abil­ ity, possibly produced by contraction of encircling stellate cells (pericytes) [29, 31]. Studies of the exteriorized liver of rodents suggest that sinusoidal flow may be regulated at both inlet and outlet levels [32], although other similar studies have not detected sphincters at either point [28]. However, sinusoidal flow is strongly affected by post‐sinusoidal resistance [27].

THE CELLS OF THE LIVER Hepatocytes Hepatocytes are commonly denoted as “parenchymal cells” and the other cells of the liver tissue matrix as “non‐parenchymal cells.” This convention is somewhat artificial since hepatocytes alone are not competent to perform all ­ essential hepatic ­functions, and several types of cells in the liver tissue matrix function as an integrated community to carry out conjointly the multiplicity of hepatic functions. Functional integration of this cellular community is accomplished by several communication mechanisms, including signaling networks involving numerous cytokines and chemokines, and by direct transfer of small mol­ ecules through gap junctions [33]. Hepatocytes, responsible for most of the synthetic and many of the metabolic functions of the liver (see Chapters 23 and 24), are large polygonal cells (averag­ ing about 25–30 μm in cross‐section and 5000–6000 μm3 in ­volume [34]). They are the most numerous cells in the liver parenchyma; the adult human liver probably contains about 1011 hepatocytes, representing about 60% of all cells in the paren­ chymal matrix and comprising about 80% of its mass/volume [15]. Hepatocytes are shaped as complex rhomboids with ­several distinct surfaces [34]. They are polarized by molecular specializations of their various surface membranes in the forms of receptors, pumps, transport c­hannels, and carrier proteins (see Chapters 5 and 6) that comprise three functionally distinct domains (see Chapter  6): the basolateral domain, the lateral domain, and the canalicular domain. • Basolateral domain: The basolateral (or sinusoidal) domain constitutes about 35% of the total hepatocyte surface facing the sinusoids. The area of this surface is greatly amplified by the folding of the plasma membrane to form innumerable micro­ villi that extend into the space of Disse [34]. This membrane is equipped for extraction of a great variety of molecules from the blood and for the simultaneous secretion into the blood of other molecules that have been modified or synthesized by hepato­ cytes. The cell matrix adhesion molecules are also present on this side of the hepatocyte. Although there is no structured basement membrane between the hepatocytes and sinusoidal endothelial cells, the lack of integrin‐linked kinase 1 (Ilk1) results in disruption of the normal structure  [35], supporting the necessity of hepatocyte–matrix interaction. About 50% of the total hepatocyte surface faces adjacent hepatocytes [34].

8

THE LIVER:  THE CELLS OF THE LIVER

• Lateral domain: The plasma membranes of these intercellular surfaces are mostly flat lateral domain, which is the least com­ plex. This surface contains intercellular adhesion complexes (tight junctions, intermediate junctions, and desmosomes) that pin together the adjacent hepatocytes, form a permeabil­ ity barrier between the peri‐sinusoidal space of Disse and bile canaliculi, and include gap junctions that allow commu­ nication between adjacent hepatocytes by transfer of small molecules. • Canalicular domain: Infoldings of the lateral surfaces create bile canaliculi, which comprise about 13% of the total hepato­ cyte surface. This is termed the canalicular domain [34]. It is also amplified by microvilli and modified for bile excretion. The canalicular surface is confined by strong junctional com­ plexes. Bile canaliculi form a belt‐like extracellular space (about 1 μm in diameter) that is continuous along the lengths of the hepatic plates, connecting at the portal ends with bile ducts. The molecular mechanisms responsible for the polarity of the canalicular domain are well characterized. Hepatocyte nuclear factor 4 (Hnf4α) is the master regulator of morpho­ logical and functional hepatocyte differentiation [36], but sev­ eral other factors, such as liver kinase B1 (Lkb1) [37], vacuolar sorting protein 33b (Vps33b) [38], and claudin‐15 [39], are also required for the polarization. As befits their numerous metabolic functions, hepatocytes contain a complex array of mitochondria (~1700 per cell on average), peroxisomes (~370 per cell), lysosomes (~250 per cell), Golgi complexes (~50 per cell), aggregates of rough and smooth endoplasmic reticulum (~15% of cell volume), and numerous microtubules/microfilaments [34]. Polyploidization is another unique feature of the hepatocytes. This is the result of defective cytokinesis, which seems to be regulated by the insulin–Akt signaling pathway [40]. The extent of ploidy increases with age. The presence of aneuploid hepato­ cytes in the normal liver is also well established [41]. The exact function of this process is still unknown but it is thought to help in adaptation to chronic injury. Hepatic plates and adjacent sinusoids form associations that are structurally similar in all parts of the liver. Various liver cells show numerical, structural, and functional heterogeneities related to their location along the afferent–efferent axis of hepatic plates and sinusoids. Among the structural differences are ploidy variations in hepatocytes; in adult mammals hepato­ cytes located at the portal ends of hepatic plates are diploid, whereas cells of higher ploidy are located further downstream [42]. Gap junctions containing connexin 26 are more numerous on portal hepatocytes, whereas junctions containing connexin 32 are distributed on hepatocytes in all parts of hepatic plates [43]. These variations in structure and cellular composi­ tion are associated with functional differences among hepato­ cytes located at different points along the afferent–efferent axis of plates and sinusoids. Rappaport divided the portal‐hepatic (afferent–efferent) lengths of hepatic plates into three arbitrary zones (termed I, II, and III). Hepatocytes located in these zones differ in their functional capabilities and susceptibilities to path­ ological damage [24], and pathways performing opposing ­functions follow an inverse distribution along the portocentral axis. This is strikingly exemplified by regional differences in

carbohydrate metabolism (gluconeogenesis and glycogen stor­ age by periportal hepatocytes; glycolysis by perihepatic vein ­hepatocytes). The ammonia and fatty acid metabolisms are also zonally distributed. Zone 1 hepatocytes are engaged in urea pro­ duction and β‐oxidation of fatty acids, while zone 3 hepatocytes remove nitrogen by glutamine synthetase and perform lipogen­ esis. The metabolism of xenobiotics is more prominent in the pericentral hepatocytes. Recent research has shown that many liver functions are dispersed heterogeneously, with dispersed functions often acting as integrated parts of coordinate meta­ bolic systems [44]. Zonation of liver functions was thought to be related to ­sinusoidal hemodynamics, which produced gradients in blood‐ borne substances available to hepatocytes and other cells of the parenchymal matrix [44]. Recent evidence gained mostly from experiments with genetically engineered mice proved that the Wnt/β‐catenin pathway is the master regulator of hepatocytic zonation, but the interaction with lymphoid enhancer factor 4 (Lef4) and HNF4α is also critical [45]. Hepatocytes and other cells located at afferent and efferent ends of hepatic plates are subjected to different microenvironmental conditions. Certain molecules are largely extracted by the first hepatocytes that encounter the perfusing blood, lowering their concentration downstream. For example, oxygen levels in the blood at affer­ ent and efferent ends of sinusoids differ greatly because oxygen is efficiently extracted by hepatocytes located at the afferent ends of hepatic plates, exposing downstream hepatocytes to relatively hypoxic conditions; the oxygen gradient alone can explain much of the heterogeneity of hepatocyte function related to position in plates [46]. Other molecules modified or produced by upstream hepatocytes are excreted into the sinu­ soidal blood and may be removed by hepatocytes located fur­ ther downstream. The complex interplay of metabolite concentration in the perfusing blood, coupled with extraction, modification, secretion, re‐extraction, and further modification, influence the metabolic events that occur in individual cells and define ­unequal parenchymal territories that produce zonal vari­ ations in different physiological capabilities and pathological susceptibilities [44]. Disruption of metabolic zonation has ­ severe consequences. Physiological turnover of hepatocytes occurs slowly. They have a lifespan of about 400 days in an adult steady‐state hepat­ ocyte population, about 0.025% of which typically are cycling [32] and the remainder rest in G0 phase. Although the hepato­ cytes are ready to re‐enter the cell cycle upon injury, this capac­ ity decreases with age. There are, however nonresolved opposing opinions about the replacement of lost hepatocytes during homeostatic conditions. The contribution of hematopoietic cells to liver maintenance was a very popular idea 20 years ago. Repopulation of bone marrow‐derived cells (macrophages, myofibroblasts) in the liver is evident in recipients of liver trans­ plants, in which these types of liver cells are replaced with cells of the host genotype [47]. In contrast, hepatocytes are not gener­ ated from bone marrow cells in significant numbers under either circumstance [47]. However, the ancient and very attractive theory of “streaming liver” keeps returning. It was originally proposed by Zajicek and colleagues [48] that periportal hepatocytes have enhanced repli­ cative potential and the progeny of these cells spread under



1:  Organizational Principles of the Liver

homeostatic condition in a portal–central direction. Although few studies applying lineage‐tracing methodology have supported this notion, most of the studies have ruled out this option. Most recently, Axin2+ hepatocytes [49] abutting hepatic veins as well as hybrid periportal hepatocytes [50] have been reported to have selective growth advantage. Based on these results a bidirectional streaming hypothesis, “somehow like ocean water entering into the delta of the Amazon River at high tide” has been proposed [51]. This issue will no doubt be resolved in the near future.

Cholangiocytes Cholangiocytes, or biliary epithelial cells, comprise much less than 1% of the total number of cells in the liver parenchyma, since most are located in bile ducts in portal tracts [52]. Only the small­ est bile ducts penetrate the parenchymal mass in the company of terminal portal veins, where they connect with bile canaliculi in hepatic plates. The points of connection of ducts with hepatic plates are defined by tubular structures called the canals of Hering, which are composed of both cholangiocytes and hepato­ cytes [53]. They are also the location of liver epithelial stem cells that can differentiate into both hepatocytes and cholangiocytes [54, 55]. Larger bile ducts contain cholangiocytes that rest on a basal membrane and vary in number and size in proportion to duct size [56]. The cholangiocytes are also polarized cells, with an apical (luminal) and a basolateral domain. Their luminal sur­ face membranes are expanded by microvilli and a single primary cilium, which is a sensor for mechanical, osmolar, and chemical stress [57]. Although they contain fewer mitochondria and sparser endoplasmic reticulum than do hepatocytes, cholangiocytes in intrahepatic bile ducts, together with the network of capillaries that surrounds them (the peribiliary plexus), form a metabolic unit that modifies the composition of canalicular bile [58]. The biliary tree is not just a draining pipe and 70–90% of the bile volume is produced by the cholangiocytes. They change the composition of the bile by secretion and absorption. The main secretory product is bicarbonate, while they absorb bile acids, glucose, and glutamate. The cholangiocytes also play an impor­ tant immunoregulatory role. They are in the first line of defense against microbial components of the biliary tract, xenobiotics, and foreign antigens. The cholangiocytes tackle these chal­ lenges by maintaining immunotolerance. They are equipped with pathogen recognition receptors (PRR), all members of the toll‐like receptors (TLR1–10), as well as related signaling mol­ ecules. Cholangiocytes produce antibacterial products (e.g. defensins, lactoferrin, lysozyme, and transport IgA) into the lumen and express MHC class I and II molecules and antigen presenting cells [59]. It is not surprising, therefore, that the bil­ iary tree is often affected by immunological disorders, such as primary sclerosing cholangitis, primary biliary cholangitis, and graft‐versus‐host disease. The cholangiocytes are slow turnover cells, but under special condition they can also participate in the regeneration of liver parenchyma (see later).

Endothelial cells Endothelial cells of sinusoids comprise about 3% of the paren­ chymal mass/volume [15] and probably number about 3 × 1010 in an adult human liver. Liver sinusoidal endothelial cells are

9

highly specialized endothelial cells with peculiar morphology and function. They are extremely thin in normal liver, 150–170 nm across, and this attenuated cytoplasm is fenestrated. The diameter of these “perforations” ranges from 50 to 200 nm, and they are clustered together in groups to form “sieve plates” [60]. The fenestrae are dynamic structures, and their actual diameter is influenced by blood pressure, composition of extracellular matrix, hormones, etc. The porosity of the sinusoidal endothe­ lial cells is polarized, and the fenestrae are more numerous in the centrilobular zone. The maintenance of the fenestration requires paracrine and autorcrine signals, which are mostly pro­ vided by hepatocytes and stellate cells [61]. Surprisingly, these endothelial cells have mostly anaerobic metabolism, providing lactate to the adjacent hepatocytes. High‐resolution in vivo microscopy has revealed swelling and contracting of these cells as a response to vasoactive sub­ stances, suggesting that they participate in the regulation of blood flow [62]. A unique functional feature of the sinusoidal endothelial cells is their high endocytic capacity. This process is mediated by all sorts of scavenger receptors, providing the main pathway for clearance of weakened molecules from circulation [63, 64] (see Chapters 26–28). Endothelial sinusoidal cells also express several types of PRR (e.g. mannose receptor several TLRs) and are able to produce inflammatory cytokines (e.g. TNF and IL‐6) as well as playing a significant role in the innate immunity. In spite of intensive studies, their role in adaptive immune reactions is still controversial [52]. The lifespan of sinusoidal endothelial cells is not known; they divide rarely and progenitor cells seem to be important in their maintenance. Two populations of progenitor cells can be distinguished [65]: the liver‐resident population is thought to be responsible for normal cell turnover, while bone marrow‐derived cells help to replenish the sinusoidal endothelial cells when necessary.

Hepatic immune cells The human liver is exposed to 1.5 L of blood every minute and a massive load of harmless dietary and commensal antigens, to which it must remain tolerant. The predominantly tolerogenic role of the hepatic immune system is well known [52], but the liver is also exposed to a variety of viruses, bacteria, parasites, and metastatic tumor cells, so needs mechanisms to override immune tolerance. In addition, the liver’s native immune sys­ tem has a major regulatory role in the repair of the liver after cell injury and loss. The liver‐centered immune system is largely segregated from the rest of the body’s immune system [66, 67]. The human liver is estimated to contain about 1010 lymphocytes of different phenotypes, located along sinusoids and in portal tracts [66]. It includes a major fraction of the body’s innate (native) immune capacity, as well as a small com­ ponent of its acquired (adaptive) immune capacity [66, 67]. The major components of the hepatic immune system are innate lymphocytes, which include a variety of T cells and non‐T cells that are able to respond rapidly to conserved ligands. These cells do not express T cell receptor (TCR) antigens. This group of immune cells includes NK cells, CD56+ T cells, natural killer T (NKT) cells, γ/δ T cells, and mucosal associated invariant T  (MAIT) cells. The liver contains multiple types of antigen

10

THE LIVER:  ONTOGENESIS OF THE LIVER

presenting cells such as hepatic myeloid dendritic cells (DCs), plasmacytoid DCs, CD11+ DCs, and NK1.1+ cytotoxic DCs [68]. In addition to the professional lymphocytes and DCs, sev­ eral populations of hepatic resident cells (e.g. Kupffer cells, sinusoidal endothelial cells, stellate cells, and cholangiocytes) are also important and active players of the liver‐centered immune system.

Macrophages Macrophages are myeloid cells that are widely distributed throughout the tissues of mammalian organisms. The liver har­ bors 80% of all macrophages of the body. In addition, it is also patrolled by blood monocytes [69]. Hepatic macrophages can be divided into two classes based on their origin [70]: resident macrophages and bone marrow‐derived macrophages. Resident macrophages of the liver, traditionally termed Kupffer cells, are established during embryonic development from the yolk sac and persist independent of blood monocytes. These cells self‐ renew during homeostatic conditions. Bone marrow‐derived blood‐borne monocytes give rise to monocyte‐derived hepatic macrophages that are more characteristic of liver injury. Kupffer cells are not optimally suited to migration, so hepatic injuries massively recruit blood monocytes to the liver. These resemble Kupffer cells phenotypically, but they remain functionally ­different. Macrophages are highly versatile cells that play a sub­ stantial role in liver homeostasis, promoting and resolving inflammatory processes and fibrosis [69]. Liver macrophages are avidly phagocytic through C3 and Fc  receptors, clearing the sinusoidal blood of relatively large particulate materials, including bacteria and weakened cells (worn‐out erythrocytes, dead or damaged hepatocytes, etc.) [63, 64]. Together with sinusoidal endothelial cells they form the organism’s major system for removing worn‐out cells and pro­ teins from perfusing blood. Activated macrophages produce many chemokines and cytokines that have a fundamental role in the implementation of the liver’s acute‐phase reaction, coordi­ nating the responses of all the parenchymal cells to injury [67].

Myofibroblasts Myofibroblasts are not present in the normal liver, but several cell types which are present physiologically can transform or be activated or transdifferentiated into the phenotype, which is the major source of extracellular matrix components and plays a basic role in pathological processes of the liver [71]. Hepatic stellate cells (HSCs) [72] are liver‐resident mesen­ chymal cells which play an important role in liver physiology and pathology. They are found in the space of Disse, between the sinusoidal endothelial cells and hepatocytes. This special location makes them able to respond to numerous kinds of injury. In addition, by encircling the sinusoids they can func­ tion as pericytes and are thought to be the most important regulator of sinusoidal diameter and blood flow. They com­ prise about 1.5% of the parenchymal volume/mass [15] and are multifunctional (see Chapters 28 and 29). The embryonic origin of stellate cells is still uncertain. Most likely they origi­ nate from the septum transversum‐derived mesothelial cells, but other options are still open. In the healthy liver they are the

largest reservoir of vitamin A, hence their former name “­fat‐ storing cells.” When the liver is injured the stellate cells trans­ differentiate into myofibroblasts and are the major producer of extracellular matrix (ECM). HSCs are important sources of cytokines and growth factors. This way they have impact on the proliferation, ­differentiation, and morphogenesis of the other hepatic cell types during liver development and regen­ eration [72]. Portal fibroblasts are spindle‐shaped mesenchymal cells in the periportal connective tissue [73]. They are distinct from HSCs in both distribution and phenotype. They do not store vitamin A but express elastin and Thy‐1. Portal fibroblasts take part in physiological ECM turnover and can contribute to the myofibroblast population in cholestatic liver injury. Bone marrow‐derived mesenchymal stem cells can also dif­ ferentiate into myofibroblasts. The contribution of bone marrow cells is well established in the fibrotic processes of kidney and lung but it seems to be quite limited in the liver. Epithelial–­ mesenchymal transition (EMT) is another recently described mechanism. Both hepatocytes and cholangiocytes can undergo such phenotypic change in tissue culture, but lineage‐tracing experiments in mice provide evidence against the role of EMT in myofibroblast ­generation in hepatic tissue in vivo [74, 75].

ONTOGENESIS OF THE LIVER The evolutionary steps that resulted in the emergence of the mammalian liver with its multiple types of functioning cells are obscure. To the extent that the ontogenesis of the liver mirrors its phylogenesis, the embryonic development of the mammalian liver suggests the way in which the aggregation of multiple types of cell into the hepatic parenchyma may have evolved (for details see Chapter 2). The development sequences of the liver in fish, birds, and mammals are similar [74–76], but endothelial cells do not appear to direct the emergence of endodermal cells from the gut during development of the liver in zebrafish [76]. Furthermore, endothelial cells with scavenger activity are in the gills and kid­ neys of cartilaginous and bony fish, and not in the liver as in all terrestrial animals [77]. The location of scavenger endothelial cells in the liver reflects a late step in the evolution of the mam­ malian liver. Nevertheless, the general pattern of expression of transcription factors and genes involved in liver development is conserved in all these species [78], suggesting a common tran­ scriptional strategy for assembling the liver. Information on when this strategy first emerged awaits further genetic analysis of gut appendages in chordate ancestors of vertebrates.

Liver parenchymal repair Three distinct processes have evolved to generate new hepato­ cytes needed to meet both increased physiological functional demands and to replace hepatocytes lost to trauma and/or toxicity. These processes comprise either a temporary reactivation of cell cycle transit in fully differentiated, mitotically quiescent hepato­ cytes, or the generation of entirely new hepatocyte lineages from adult liver stem cells (see Chapters 36 and 38).



1:  Organizational Principles of the Liver

The most direct and rapid parenchymal augmentation/ replacement processes involve the induction of hepatocyte rep­ lication in the absence of a preceding increase in hepatocyte death, often associated with increased hepatic functional demand due to physiological need [78, 79]. Hyperplasia of hepatocytes by this mechanism enlarges the parenchymal mass and increases hepatocyte functional capacity. This process is regulated by the binding of ligands to hepatocyte nuclear recep­ tors, nearly 50 of which have been identified [79, 80]. Nuclear receptors are transcription factors that, when bound to ligands, directly upregulate the combination of genes required to drive hepatocytes through the cell cycle [79, 80]. Several ligands for nuclear receptors (termed “primary hepatocyte mitogens”), including adrenal corticoids, bile acids, sex steroids, thyroid hormone, peroxisome proliferators, and 9‐cis,cis‐retinoic acid, directly stimulate the proliferation of hepatocytes and increase liver mass after binding to nuclear receptors [80]. Although it would appear that endothelial cells would be needed to support the additional hepatocytes, no documentation of coordinate endothelial cell proliferation has been presented; it is, however, possible that new endothelial cells could be derived from the bone marrow. Next in process complexity and speed of response is the replacement of the diverse liver parenchyma by the sequential proliferation of all of the component cells (hepatocytes, chol­ angiocytes, endothelial cells, macrophages, stellate cells, and immunocytes), and the merging of the new cells into a tissue that closely resembles the functional units of the undamaged liver [78]. This process, which can replace up to 70% of the parenchymal mass in mammals, is often called “liver regenera­ tion,” although this is a misnomer since in mammals the part of the liver removed surgically does not “regenerate” in the way that body parts in certain lower animal species do. Instead, the liver, after resection, is enlarged by the expansion of remaining units (lobes), in a biological process defined as “compensatory hyperplasia.” In contrast to liver repair in mammals, liver repair after partial hepatectomy in fish most intensively involves cells at the resection margin [81, 82] and may culminate in the regrowth (regeneration) of the resected tissue [81]. The cell proliferation phase of this reparative process in mammals has been subjected to intensive kinetic and regulatory analyses (see Chapter  45 and references therein). After tissue loss, residual hepatocytes are activated to proliferate within few hours. Hepatocyte proliferation begins at the portal ends of plates  [83], and successive waves of hepatocyte proliferation ultimately involve virtually all residual hepatocytes [83]. Hepatocyte replacement is followed sequentially by prolifera­ tion of sinusoidal endothelial cells and macrophages [83, 84], and the other cells of the parenchymal matrix. To the extent that it has been elucidated (see Chapter 45), regulation of hepatocyte proliferation is regulated by a complex mixture of cytokines and growth factors [85]. Most of the regulatory molecules are pro­ duced by various liver cells or are released from storage sites within the liver [85], and many are components of the acute‐ phase reaction [86] and other elements of the liver’s native immune system [87–89]. The less completely analyzed remod­ eling phase primarily involves endothelial cells and likely the other cells of the liver parenchyma. For example, proliferating hepatocytes initially form focal multicellular clumps [90, 91],

11

which are cleaved into one‐hepatocyte‐wide plates by signaling from and separation by endothelial and stellate cells [90, 91]. Eventually, the remaining lobes increase exclusively by the enlargement of preexistent hepatic lobules, contrary to the ­physiological liver growth in young animals, when new lobules are formed [92]. Although known regulatory mechanisms drive the reparative process, the mechanism that “triggers” the onset of repair after loss of liver tissue is still obscure. Since the liver vasculature must accept the entire portal blood volume, it has long been suspected that the trigger may be the massive increase in portal blood flow per unit of residual mass that follows loss of liver tissue [93]. Increased portal blood flow and pressure cause shear stress in sinusoids [94], which produces a burst of nitric oxide and prostaglandin production by sinusoidal endothelial cells, possibly providing the molecular trigger [95, 96]. Alternatively (or in concert), early activation of the nuclear receptor mecha­ nism of hepatocyte proliferation may function as a trigger [96], and it is possible that multiple alterations in the physiological status of the liver remaining after tissue loss may converge to produce a “mass action” trigger. If the hepatocytes are compromised, there are alternative mechanisms of liver regeneration. Enlargement or hypertrophy of hepatocytes can compensate for the lost parenchyma [97], but this response usually provides just a transient solution. There has been much debate about the participation of stem or progenitor cells in liver regeneration. A peculiar cell population, named after the shape of their nuclei as “oval cells,” were observed in hepatocarcinogenesis experiments in rodents. Similar cells have been described in several other species and their emer­ gence has been named “ductular reaction.” There is convincing evidence that these cells can replace the lost liver parenchyma [98] and behave as the amplification compartment of hepatic stem cells. Several candidates for hepatic stem cells are known, but most data indicate that the terminal segment of the biliary system, the canals of Hering, harbor the adult hepatic stem cells. Lineage‐tracing experiments in rats [54, 55, 57, 99, 100] and zebrafish [101] demonstrated the regeneration of hepatocytes from biliary stem cells. The application of the cre/lox lineage tracing in mice did not support this prevailing model, because no biliary cell‐derived hepatocytes were observed, although the hepatocytic origins of biliary cells and cholangiocarcinomas were demonstrated [102]. However, eventually hepatic progeni­ tor cells of biliary origin with liver repopulation capacity were shown in mice following complete blockage of hepatocyte pro­ liferation [103]. At present there seem to be general agreement [104] that both hepatocytes and cholangiocytes (or their sub­ populations) are able to behave as stem cells, and under specific conditions are capable of regenerating both epithelial cell ­compartments of the liver. The capacity of these highly differen­ tiated cells is also referred to as “plasticity” [105] but this is mostly a debate about terminology – how we should refer to a peculiar biological reaction. Initial observations indicate that these “back‐up” stem cells, which support regenerative pro­ cesses in rat and human, are organized along the branches of the portal vein [106, 107] and are regulated by elements of hepatic immunomodulation centered on the acute‐phase reaction [98, 108], similar to the organization of liver architecture during embryonic development.

12

THE LIVER:  REFERENCES

Although the subject of intensive scrutiny recently, there is no substantial evidence that hematopoietic stem cells or mesen­ chymal components of the liver are a significant source for the generation of hepatocytes or biliary epithelial cells in either humans or experimental animals [47]. This situation contrasts with the replenishment from hematopoietic sources of other cells of the liver parenchyma [47].

REFERENCES 1. Elias, H. A re‐examination of the mammalian liver, I. Parenchymal ­architecture. Am J Anat, 1949;84:311–33. 2. Couinand, C. Liver anatomy: portal (and suprahepatic) or biliary segmentation. Dig Surg, 1999;16:459–67. 3. Majno, P., Mentha, G., Toso, C. et al. Anatomy of the liver: an outline with three levels of complexity: a further step towards tailored territorial liver resections. J Hepatol, 2013;60:654–62. 4. Elias, H. A re‐examination of the mammalian liver, II. The hepatic lobule and  its relation to the vascular and biliary systems. Am J Anat, 1949;84: 379–456. 5. Rappaport, A.M., Borowy, Z.J., Lougheed, W.M. et al. Subdivision of hex­ agonal liver lobules into a structural and functional unit. Role in hepatic physiology and pathology. Anat Rec, 1954;119:11–34. 6. Matsumoto, T., Komori, R., Magara, M. et al. A study of the normal structure of the human liver, with special reference to its angioarchitecture. Jikeikai Med J, 1979;26:1–40. 7. Kotton, D.W., Fabian, A.J., and Mulligan, R.C. A novel cell population in adult liver with potent hematopoietic reconstitution activity. Blood, 2005;106:1574–80. 8. Mitra, S.K. The terminal distribution of the hepatic artery with special refer­ ence to arterio‐portal anastomosis. J Anat, 1966;100:651–63. 9. Kardon, R.H. and Kessel, R.G. Three‐dimensional organization of the hepatic microcirculation in the rodent liver as observed in the scanning elec­ tron microscopy of corrosion casts. Gastroenterology, 1980;79:72–81. 10. Yamamoto, K., Sherman, I., Phillips M.J. et al. Three‐dimensional observations of the hepatic arterial terminations in rat, hamster and human liver by scanning electron microscopy of microvascular casts. Hepatology, 1985;5:452–6. 11. Watanabe, Y., Püschel, G.P., Gardemann, A. and Jungermann, K. Presinusoidal and proximal intrasinusoidal confluence of hepatic artery and portal vein in rat liver: functional evidence by orthograde and retrograde bivascular per­ fusion. Hepatology, 1994;19:1198–207. 12. Wisse, E., DeZanger, R.B., Charels, K. et al. The liver sieve: considerations concerning the structure and function of endothelial fenestrae, the sinusoidal wall and the space of Disse. Hepatology, 1985;5:683–92. 13. Goresky, C.A. A linear method for determining the liver sinusoidal and extravascular volumes. J Physiol, 1963;204:626–40. 14. Grisham, J.W., Nopanitaya, W., Compagno, J., and Nagel, A.E.H. Scanning electron microscopy of normal rat liver: the surface structure of its cells and tissue components. Am J Anat, 1975;144:295–322. 15. Blouin, A., Bolender, R.P., and Weibel, E.R. Distribution of organelles and membranes between hepatocytes and nonhepatocytes in the rat liver paren­ chyma. A stereological study. J Cell Biol, 1977;72:441–55. 16. Miller, D.L., Zanolli, C.S., and Gumucio, J.J. Quantitative morphology of the sinusoids of the hepatic acinus. Gastroenterology, 1979;76:965–69. 17. Grisham, J.W. and Nopanitaya, W. Scanning electron microscopy of casts of hepatic microvessels, in Hepatic Circulation in Health and Disease (ed. W. Laut), Raven Press, New York, 1981, pp. 87–107. 18. Dezső, K., Paku, S., Papp, V. et al. Architectural and immunohistochemical characterization of biliary ductules in normal human liver. Stem Cells Dev, 2009;18:1417–22. 19. Theise, N.D., Saxena, R., Portmann, B.C. et al. The canals of Hering and hepatic stem cells in humans. Hepatology 1999;30:1425–33. 20. Teutsch, H.F., Schuerfeld, D., and Groezinger, E. Three‐dimensional ­reconstruction of parenchymal units in the liver of the rat. Hepatology, 1999;29:494–505. 21. Teutsch, H.F. The modular microarchitecture of human liver. Hepatology, 2005;42:317–25.

22. Bloch, E.H. The termination of hepatic arterioles and functional unit of the liver as determined by microscopy of the living organ. Ann NY Acad Sci, 1970;170:78–87 23. Kiernan, F. The anatomy and physiology of the liver. Trans R Soc Lond, 1833;123:711–70. 24. Rappaport, A.M. The microcirculatory hepatic unit. Microvasc Res, 1973;6:212–28. 25. Ekataksin, W. and Wake, K. Liver units in three dimensions: I. Organization of argyrophil connective tissue skeleton in porcine liver with particular refer­ ence to the “compound hepatic lobule.” Am J Anat, 1991;191(2):113–53. 26. Elias, H. and Sherrick, J.C. Morphology of the Liver. Academic Press, New York, 1969. 27. Lautt, W.W. and Greenway, C.V. Conceptual view of the hepatic vascular bed. Hepatology, 1987;7:952–63. 28. Nakata, K., Leong, G.F., and Brauer, R.W. Direct measurement of blood pressure in the minute vessels of the liver. Am J Physiol, 1960;199:1181–8. 29. McKuskey, R.S. Morphological mechanisms for regulating blood flow through hepatic sinusoids. Liver, 2000;20:3–7. 30. Hellerbrand, C. Hepatic stellate cells  –  the pericytes in the liver. Pflugers Arch Eur J Physiol, 2013;465:775–8. 31. Reynaert, H., Thompson, M.G., Thomas, T., and Geerts, A. Hepatic stellate cells: role in microcirculation and pathophysiology of portal hypertension. Gut, 2002;50:571–81. 32. Grisham, J.W. A morphologic study of deoxyribonucleic acid synthesis and cell proliferation in regenerating rat liver; autoradiography with thymidine H3. Cancer Res, 1962;22:842–49. 33. Kmeic, Z. Cooperation of liver cells in health and disease. Adv Anat Embryol Cell Biol, 2001;161:1–151. 34. Weibel, E.R., Stäubli, W., Gnägi, H.R., and Hess, F.A. Correlated ­morphometric and biochemical studies on the liver cell. 1. Morphometric model, stereologic methods, and normal morphometric data for rat liver. J Cell Biol, 1969;42:68–90. 35. Gkretsi, V., Apte, U., Mars W.M. et al. Liver specific ablation of integrin‐ linked kinase in mice results in abnormal histology, enhanced cell prolifera­ tion, and hepatomegaly. Hepatology 2008;48:1932–41. 36. Parviz, F., Matullo, C., Garrison, W.D. et  al. Hepatocyte nuclear factor 4 alpha controls the development of a hepatic epithelium and liver morpho­ genesis. Nat Genet, 2003;34:292–6. 37. Woods, A., Heslegrave, A.J., Muckett, P.J. et al. LKB1 is required for hepatic bile acid transport and canalicular membrane integrity in mice. Biochem J, 2011;434:49–60. 38. Cullinane, A.R., Straatman‐Iwanowska, A., Zaucker, A. et  al. Mutations in VIPAR cause an arthrogryposis, renal dysfunction and cholestasis syndrome phenotype with defects in epithelial polarization. Nat Genet, 2010;42:303–12. 39. Cheung, I.D., Bagnat, M., Ma, T.P. et al. Regulation of intrahepatic bile duct morphogenesis by Claudin 15‐like b. Dev Biol, 2012;361, 68–78. 40. Celton‐Morizur, S., Merlen, G., Conton, D. et al. The insulin/Akt pathway controls a specific cell division program that leads to generation of binucle­ ated tetraploid liver cells in rodents. J Clin Invest, 2009;122:3307–15. 41. Duncan, A.W., Hanlon Newell, A.E., Sith, L. et  al. Frequent aneuploidy among normal human hepatocytes. Gastroenterology, 2012;142:25–8. 42. Gupta, S. Hepatic polyploidy and liver growth control. Semin Cancer Biol, 2000;10:161–71. 43. Traub, O., Look, J., Dermietzel, R. et  al. Comparative characterization of 21‐kD and 26‐kD junction proteins in murine liver and cultured hepatocytes. J Cell Biol, 1989;108:1039–51. 44. Jungermann, K. and Katz, N. Functional specialization of different hepato­ cyte populations. Physiol Rev, 1989;69:708–64. 45. Gougelet, A., Torre, C., Veber, P. et al. T cell factor 4 and β‐catenin chromatin occupancies pattern zonal liver metabolism. Hepatology, 2014;59:2344–57. 46. Jungermann, K. and Kietzmann, T. Oxygen: modulator of metabolic zona­ tion and disease of the liver. Hepatology, 2000;31:255–60. 47. Thorgeirsson, S.S. and Grisham, J.W. Hematopoietic cells as hepatocyte stem cells: a critical review of the evidence. Hepatology, 2006;43:2–8. 48. Zajicek, G., Oren, R., and Weinreb, M. Jr. The streaming liver. Liver, 1985;5:293–300. 49. Wang, B., Zhao, L., Fish, M. et al. Self‐renewing diploid Axin 2+ cells fuel homeostatic renewal of the liver. Nature, 2015;524:180–5. 50. Font‐Burgada, J., Shalapour, S., Ramaswamy, S. et  al. Hybrid periportal hepatocytes regenerate the injured liver without giving rise to cancer. Cell, 2015;162:766–79.



1:  Organizational Principles of the Liver

51. Allison, M.R. Hepatocytes come out of left field. Hepatology, 2016;63: 1041.–2. 52. Jennem, C.N. and Kubes, P. Immune surveillance by the liver. Nat Immunol, 2013;14:996–1006. 53. Saxena, R. and Theise, N. Canals of Hering: recent insights and current knowledge. Semin Liver Dis, 2004;24:43–8. 54. Thorgeirsson, S.S. and Grisham, J.W. Liver stem cells, in Stem Cells (ed. C.S. Potten), Academic Press, London, 1997, pp. 233–85. 55. Kuwahara, R., Kofman, A.V., Landis, C.S. et al. The hepatic stem cell niche: identification fy label‐retaining assay. Hepatology, 2008;47:1994–2002. 56. Benedetti, A., Bassotti, C., Rapno, K. et al. A morphometric study of the epi­ thelium lining the rat intrahepatic biliary tree. J Hepatol, 1996;24:335–42. 57. Bing, Q., Masyuk, T.V., Muff, M.A. et al. Isolation and characterization of cholangiocyte primary cilia. Am J Physiol Gastrointest Liver Physiol, 2006;291:G500–9. 58. Alpini, G., Ulrich, C., Roberts, S. et al. Molecular and functional heteroge­ neity of cholangiocytes from rat liver after bile duct ligation. Am J Physiol Gastrointest Liver Physiol, 1997;272:G289–97. 59. Zhang, H. How biliary tree maintains immune tolerance? BBA Mol Basis Dis, 2018;1864:1367–73. 60. Braet, F. and Wisse, E. Structural and functional aspects of liver sinusoidal endothelial cell fenestrae: a review. Comp Hepatol, 2002;1:1–10. 61. Sorensen, K.K., Simon‐Santamaria, J., McCuskey, R.S. et al. Liver sinusoi­ dal endothelial cells. Compr Physiol, 2001;5:1751–74. 62. McCuskey, R.S. and Reilly, F.D. Hepatic microvasculature: dynamic struc­ ture and its regulation. Semin Liver Dis, 1993;13:1–12. 63. Smedsrød, B., DeBleser, P.J., Braet, F. et al. Cell biology of endothelial and Kupffer cells. Gut, 1994;35:1509–16 64. Smedsrød, B. Clearance function of scavenger endothelial cells. Comparative Hepatol, 2004;3(Suppl 1):S22. 65. Wang, L., Wang, X., Wang, L. et  al. Hepatic vascular endothelial growth factor regulates recruitment of rat liver sinusoidal endothelial cell progenitor cells. Gastroenterology, 2012;143:1555–63. 66. Doherty, D.G. and O’Farrelly, C. Innate and adaptive lymphoid cells in the human liver. Immunol Rev, 2000;174:5–20. 67. Parker, G.A. and Picut, C.A. Liver immunobiology. Toxicol Pathol, 2005;33:52–62. 68. Doherty, D.G. Immunity, tolerance and autoimmunity in the liver: a compre­ hensive review. J Autoimmun, 2016;66:60–75. 69. Ju, C. and Tacke, F. Hepatic macrophages in homeostasis and liver disease: from pathogenesis to novel therapeutic strategies. Cell Mol Immunol, 2016;13:316–27. 70. Gomez, P.E., Klapproth, K., Schulz, C. et al. Tissue‐resident macrophages originate from yolk‐sac derived erythro‐myeloid progenitors. Nature, 2015;518:547–51. 71. Lee, U.E. and Friedman, S.L. Mechanisms of hepatic fibrogenesis. Best Pract Res Clin Gastroenterol, 2011;25:195–206. 72. Yin, C., Evason, K. J., Asashina, K. et al. Hepatic stellate cells in liver devel­ opment, regeneration and cancer. J Clin Invest, 2013;123:1902–10. 73. Wells, R.G. The portal fibroblast  –  not just a poor man’s stellate cell. Gastroenterology, 2014;147:41–7. 74. Elias, H. Origin and early development of the liver in various vertebrates. Acta Hepatol, 1955;3:1–567. 75. Le Douarin, N.M. An experimental analysis of liver development. Med Biol, 1975;53:427–55. 76. Field, H.A., Ober, E.A., Roeser, T., and Stainier, D.Y.R. Formation of the digestive system in zebrafish. I. Liver morphogenesis. Dev Biol, 2003;253: 279.–90. 77. Seternes, T., Sørensen, K., and Smedsrød, B. Scavenger endothelial cells of vertebrates: a nonperipheral leukocyte system for high capacity elimination of waste macromolecules. Proc Natl Acad Sci U S A, 2002;99:7594–97. 78. Columbano, A. and Ledda‐Columbano, G.M. Mitogenesis by ligands of nuclear receptors: an attractive model for the study of molecular mechanisms implicated in liver growth. Cell Death Differ, 2003;10:S19–21. 79. Zaret, K.S. Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet, 2007;3:499–512. 80. Nagy, L. and Schwabe, J.W. Mechanism of the nuclear receptor molecular switch. Trends Biochem Sci, 2004;29:317–24.

13

81. Zhao, R. and Duncan, S.A. Embryonic development of the liver. Hepatology, 2005;41:956–67. 82. Zaret, K.S. Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet, 2007;3:499–512. 83. Sadler, K.L., Krahn, K.N., Gaur, N.A., and Ukomadu, C. Liver growth in the embryo and during liver regeneration in zebrafish requires the cell cycle regulator, uhrf1. Proc Natl Acad Sci U S A, 2007;104:1570–5. 84. Okihiro, M.S. and Hinton, D.E. Partial hepatectomy and bile duct ligation in rainbow trout (Oncorhynchus mykiss): histologic, immunohistochemical and enzyme histochemical characterization of hepatic regeneration and hyperplasia. Toxicol Pathol, 2000;28:342–335. 85. Michalopoulos, G.K. Hepatostat: liver regeneration and normal tissue maintenance. Hepatology 2017;65:1384–92. 86. Fulop, A.K., Pocsik, E., Brozik, E. et al. Hepatic regeneration induces tran­ sient acute phase reaction: systemic elevation of acute phase reactants and soluble cytokine receptors. Cell Biol Int, 2001;25:585–92. 87. Fausto, N. Involvement of innate immune system in liver regeneration and injury. J Hepatol, 2006;45,347–9. 88. Dong, Z., Wei, H., Sun, R., and Tian, Z. The role of innate cells in liver injury and regeneration. Cell Mol Immunol, 2007;4:241–52. 89. Preziasi, M.E. and Monga, S.P. Update on the mechanisms of liver regen­ eration. Semin Liver Dis, 2017;37:141–51. 90. Martinez‐Hernandez, A. and Amenta, P. The extracellular matrix in hepatic regeneration. FASEB J, 1995;9:1401–10. 91. Wack, K.E., Ross, M.A., Zegara, V. et al. Sinusoidal ultrastructure during revascularization of regenerating liver. Hepatology, 2001;33:363–78. 92. Papp, V., Dezso, K., Laszlo, V. et al. Architectural changes during regenera­ tive and ontogenic growth in the rat. Liver Transplant, 2009;18:177–83. 93. Sato, Y., Kayama, S., Tsukada, K., and Hatakeyama, K. Acute portal hyper­ tension reflecting shear stress as a trigger to liver regeneration following partial hepatectomy. Surg Today (Jpn J Surg), 1997;27:518–26. 94. Schoen, J.M., Wang, H.H., Minuk, G.Y., and Lautt, W.W. Shear stress‐ induced nitric oxide release triggers the liver regeneration cascade. Nitric Oxide, 2001;5:453–64. 95. Schoen‐Smith, J.M. and Lautt, W.W. The role of prostaglandins in trigger­ ing the liver regeneration cascade. Nitric Oxide, 2005;23:111–17. 96. Huang, W., Ma, K., Zhang, J. et al. Nuclear receptor‐dependent bile acid sign­ aling is required for normal liver regeneration. Science, 2006;312:233–6. 97. Matot, I. and Nochmansson, N. Impaired liver regeneration after hepatec­ tomy and bleeding is associated with shift from hepatocyte proliferation to hypertrophy. FASEB J, 2017;31:5283–95. 98. Thorgeirsson, S.S., Factor, V.M., and Grisham, J.W. Early activation and expansion of hepatic stem cells, in Handbook of Stem Cells, Introduction to Adult and Fetal Stem Cells, vol. 3 (eds. R. Lanza, H. Blau, D. Melton et al.), Academic Press, San Diego, 2004, pp. 497–512. 99. Evarts, R.P., Nagy, P., Marsden, E., and Thorgeirsson, S.S. A precursor‐­ product relationship exists between oval cells and hepatocytes in rat liver. Carcinogenesis, 1987;8:1737–40. 100. Paku, S, Schnur, J., Nagy, P et  al. Origin and structural evolution of the early proliferating oval cells in rat liver. Am J Pathol, 2001;158:1313–23. 101. Choi, T.Y., Ninov, N., Stainer, D.Y. et al. Extensive conversion of hepatic biliary epithelial cells to hepatocytes after near total loss of hepatocytes in zebra fish. Gastroenterology, 2014;146:776–88. 102. Schaub, J.R., Malato, Y., Gormond, C. et al. Evidence against a stem cell origin of new hepatocytes in a common mouse model of chronic liver injury. Cell Rep, 2014;8:933–939. 103. Lu, W.Y., Bird, T.G., Boulter, L. et al. Hepatic progenitor cells of biliary origin with liver repopulation capacity. Nature Cell Biol, 2015;17:971–83. 104. Michalopoulos, G.K. and Khan, Z. Liver stem cells: experimental findings and implications for human disease. Gastroenterology, 2015;149:876–82. 105. Kopp, J.L., Grompe, M., and Sander, M. Stem cells versus plasticity in liver and pancreas regeneration. Nature Cell Biol, 2016;18:238–44. 106. Dezso, K., Papp, V., Bugyik, E. et al. Structural analysis of oval‐cell‐medi­ ated liver regeneration in rat. Hepatology, 2012;56:1457–67. 107. Dezso, K., Rokusz, A., Bugyik, E. et al. Human liver regeneration in advanced cirrhosis is organized by the portal tree. J Hepatol, 2017;66:778–86. 108. Cast, A.E., Walter, T.J., and Huppert S.S. Vascular patterning sets the stage for macro and micro hepatic architecture. Dev Dynam, 2015;244:497–506.

2

Embryonic Development of the Liver Kenneth S. Zaret1, Roque Bort2, and Stephen A. Duncan3 Institute for Regenerative Medicine, Department of Cell and Developmental Biology, Perelman School of Medicine, University of Pennsylvania, Smilow Center for Translational Research, Philadelphia, PA, USA 2 Instituto de Investigación Sanitaria La Fe (IIS La Fe), Unidad de Hepatología experimental, València, Spain 3 Department of Regenerative Medicine and Cell Biology, Medical University of South Carolina, Charleston, SC, USA 1

INTRODUCTION The liver is one of the first organs to develop in the embryo and it rapidly becomes one of the largest organs in the fetus. The most essential function of the mammalian fetal liver is to provide a site for hematopoiesis. The early dependence of the fetus on its own blood cell supply makes embryonic liver growth and viability a sensitive phenotypic indicator for gene inactivation studies. From an experimental perspective, the large size and small number of cell types in the developing liver make it easy to study, and new insights have emerged from recent work on genetically modified mice, explants of embryonic tissues, pluripotent stem cell differentiation, and other vertebrate models such as zebrafish and frogs. Much is being learned about how gene function is orchestrated to control tissue morphogenesis, helping to establish liver development as a paradigm for the genesis of other gut‐derived tissues. Understanding the mechanisms that govern liver development should also provide insight into future therapies for liver diseases. Examples of such applications include activating stem cells in the adult liver, replenishing diseased livers with cells from pluripotent stem cells, transdifferentiating cells from different organs, and reconstituting proper liver morphology and function. This review will describe the early stages of liver development from the initial specification to hematopoietic cell invasion, which covers the competence of progenitor cells to become hepatocytes, the formation of the liver bud, and the early morphogenesis and differentiation of the liver. The review will also present the key effectors of hepatocyte maturation and discuss the control of liver size and regeneration in the embryo and in the adult. The reader is referred to other reviews for summaries of the mid‐ and late‐fetal stages of liver development [1–9].

ACQUISITION OF HEPATIC COMPETENCE WITHIN THE ENDODERM The liver, lung, pancreas, thyroid, and gastrointestinal tract are derived from the anterior‐ventral definitive endoderm, which is one of the three germ layers that arise during gastrulation. Initially, the endoderm is an epithelial sheet that lines the ventral surface of the embryo. Infolding of the sheet at the anterior and posterior of the embryo generates the foregut and hindgut (Figure  2.1). When these morphogenetic movements reach the middle of the embryo, the gut tube closes off. During gut tube formation, different tissues are specified along the anterior–posterior and dorsal–ventral axes of the embryo, with the liver arising from the prospective ventral endoderm domain of the foregut (Figure  2.1). Are there “pre‐ patterns” or local domains of endoderm that are competent to differentiate into the liver? LeDouarin demonstrated, using heterotypic grafts of quail embryo segments into recipient chick embryos, that only the prospective anterior‐ventral domain of endoderm had the capacity to develop into the liver [10, 11]. High‐resolution lineage tracing in mouse embryos has revealed that distinct domains within the endoderm contribute to the developing liver [12]. The majority of cells derive from two bilateral patches within the lateral endoderm, and a smaller population derives from a small clutch of cells at the ventral midline of the anterior endoderm. Morphogenic movements ensure that these populations converge during the formation of the liver bud. More recent studies with mouse embryos and more sensitive assays of gene expression revealed that while the prospective dorsal endoderm (Figure 2.1) does not normally activate liver genes or become liver, in a tissue explant assay the dorsal endoderm cells could initiate liver gene expression when isolated

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



2:  Embryonic Development of the Liver

15

endoderm cardiac mesoderm

head

–10 kb

(FGF source)

septum transversum mesenchyme (BMP source)

X

alb1–

GATA FoxA foregut

somites hindgut

ventral-anterior endoderm (suppression of Wnt allows hepatic potential)

liver bud alb1+

dorsal-posterior endoderm (Wnt signaling suppresses hepatic potential)

Figure 2.1  Parasagittal diagram of a mammalian embryo at the time of hepatic specification. Relevant tissues and signaling molecules are indicated. The time of development corresponds to about 8.25 days’ gestation in the mouse and about three weeks in the human.

from its adjacent mesodermal tissues [13]. The molecular mechanism underlying these transplantation studies has been deciphered using Xenopus (frog) embryos, and similar mechanisms likely operate in mammalian embryos. It was found that due to the anterior expression of Wnt inhibitors such as Dkk, sFrp‐1 or sFrp‐5, the function of the Wnt/β‐catenin signaling pathway is repressed upon gastrulation in the anterior endoderm [14], but it is active posteriorly, where the Wnt inhibitors are absent. Wnt downregulation in the anterior endoderm was shown to be crucial for liver and pancreas specification; in fact, posterior endoderm that was forced to activate GSK‐3β, a Wnt inhibitor, possessed hepatogenic competence [15]. Wnt signaling induces the transcription factor Vent in the endoderm, which in turn represses the homeobox gene Hex that is required in the endoderm for liver and pancreas development [16–21]. In summary, Wnt repression in the anterior endoderm allows liver and pancreas specification, whereas active Wnt signaling in the posterior endoderm suppresses those fates. These findings provide a molecular basis to help explain the patterning of the endoderm. Another approach to understanding how the endoderm gains the competence to develop into liver is to identify the regulatory transcription factors that directly enable the process. Regulatory factors that are known to be important for liver differentiation are expressed in prehepatic endoderm, prior to hepatic specification. Such factors could operate by helping to open up chromatin structure for genes that need to be transcribed during liver differentiation [22]. Transcription factors that are expressed in the prehepatic domain of ventral foregut endoderm, and later in the liver, include FOXA1, FOXA2 [23–26], GATA4, and GATA6 [27–30]. The roles of these transcription factors can be discerned by understanding how the factors function in a chromatin context. DNA binding by both FOXA and GATA factors is required for the activity of a transcriptional enhancer of alb1, the gene encoding serum albumin [31, 32]; alb1 is one of the earliest genes to be activated in hepatic development [13, 33]. An analysis of the DNA binding sites for FOXA2 and GATA [34] showed that both sites are occupied on the alb1 enhancer in the endoderm (Figure  2.2), prior to alb1 transcriptional ­activation or hepatic commitment [13, 31]. Once hepatic specification has occurred, adjacent binding sites for a variety of

C/EBP

FoxA eY FoxA NF1

Figure 2.2  Transcriptional competence factors in the endoderm. DNA binding sites for GATA and FOXA transcription factors are occupied on the alb1 gene prior to alb1 expression or hepatic commitment. During hepatic specification, other transcription factors bind DNA at adjacent sites and the albumin gene becomes active.

other transcription factors become occupied at the enhancer and alb1 becomes active (Figure 2.2). Genetic experiments have confirmed the essential role of FOXA and GATA factors in hepatogenesis. A transgenic mouse with a conditional ablation of both FOXA1 and FOXA2 in the foregut endoderm completely lacked the induction of hepatic markers of specification, such as alb1, afp, or ttr [35]. Similarly, while the liver is specified in GATA6−/− or GATA4−/− mouse embryos (though subsequent development is blocked), a double knockout of both genes in zebrafish revealed a complete failure in liver induction [36]. The role of GATA factors in controlling endodermal fate appears to be conserved in humans. Several groups have shown that depletion of GATA6 affects the differentiation of endoderm from human induced pluripotent stem cells (iPSCs) [37–39]. These findings provide genetic evidence supporting the concept that FOXA and GATA factors serve as “pioneer factors” in the endoderm, being among the first to bind a target gene in development and endow competence for the gene to be activated, under the influence of cell‐type inductive signals.

FROM DEFINITIVE TO HEPATIC ENDODERM Once the hepatogenic competence in the definitive endoderm is acquired, what causes the liver to be specified from the ventral foregut endoderm? The mesoderm secretes patterning signals that instruct the differentiation of the underlying endoderm. Several lines of research conducted in chicken and mouse demonstrated that cardiac mesoderm helps instruct the underlying ventral foregut endoderm to become hepatic endoderm (Figure  2.1). In vitro culture of mouse embryonic explants together with the use of zebrafish and Xenopus embryos have identified some of the relevant molecular signals. There are over 20 genes for fibroblast growth factors (FGFs) and 4 genes for FGF receptors [40, 41]. Each of the FGF receptors has different binding specificities for FGFs, and ­ the existence of multiple spliced isoforms of the receptors provides further complexity to the receptor–ligand relationships.

16

THE LIVER:  FROM HEPATIC ENDODERM TO LIVER BUD

Cardiogenic mesoderm expresses FGF1, FGF2, FGF8, and FGF10 in the period prior to hepatic induction in the endoderm. In an embryonic tissue explant system, purified FGF1 or FGF2 can efficiently activate liver gene expression in ventral foregut explants where cardiogenic mesoderm has been removed, and an FGF antagonist can inhibit liver gene induction in foregut explants where cardiogenic mesoderm has been retained [42]. FGF binding induces tyrosine phosphorylation activity by the cytoplasmic domain of the receptors, resulting in activation of mitogen‐activated protein kinase (MAPK) signaling pathways within cells [41, 43]. A combination of in vivo‐genetic, whole‐ embryo culture and tissue explant approaches revealed that a transient activation of the MAPK pathway by FGF signaling in foregut endoderm explants is necessary for the initiation and stabilization of the hepatic program [44]. Although the PI3K/ AKT pathway is activated in the endoderm shortly after the MAPK pathway, PI3K/AKT pathway activation appears not to be downstream of FGF signaling and the pathways do not crossregulate in the hepatic endoderm [44]. Ultimately, the action of FGF signaling must activate gene expression programs that drive hepatic fate. The differentiation of human iPSCs to a hepatic fate has allowed researchers to identify the immediate targets of FGF signaling [45]. Genes that are directly activated during the endoderm‐to‐hepatic transition include several transcription factors, growth factors, and signaling molecules. Interestingly, one of the direct targets of FGF encodes NKD1, which is a repressor of WNT signaling. When NKD1 was depleted, differentiation of the endoderm to a hepatic fate was inhibited [45]. These findings imply that FGF drives hepatic fate in part by transiently suppressing canonical WNT signaling. BMP4 is a member of the TGFβ superfamily. BMP2, BMP4, BMP5, and BMP7 are highly expressed in the septum transversum mesenchyme (STM), the latter consisting of loose mesenchyme cells that surround the cardiac and ventral endoderm domains [46–50]. The BMP receptors BMPRIA, BMPRII, and ActRIIA are expressed in the endoderm [51, 52]. The mouse foregut explant system described above was used to reveal that BMP signaling from STM cells is also crucial for hepatic gene induction in the endoderm [53]. Thus, hepatic induction requires positive signals (FGF and BMP) from two different cell sources (cardiac mesoderm and STM), indicating the importance of combinatorial signaling. The general relevance of FGFs and BMPs for hepatic induction has been highlighted by genetic experiments in zebrafish [54]. The zebrafish studies of hepatic induction, which cleverly traced the fate of individual endoderm cells [54], confirmed an initial observation made with populations of endoderm cells isolated from mouse embryos. That is, the foregut endoderm makes a cell fate choice for the liver or pancreas, under the influence of FGF and BMP signaling, as no FGF and BMP allow pancreas induction, whereas elevated levels of both signaling molecular induce a liver fate [55]. Subsequent studies revealed that the signaling network is dynamic, with intracellular FGF and BMP responses changing in the endoderm tissue as it moves past the cardiac mesoderm and STM signaling centers [56]. During this period, the hepatogenic BMP signal in endoderm is transmitted to target genes via the transcription factor SMAD4, which in turn recruits the p300 coactivator [57]. P300 is a histone acetyltransferase and its activity modulates liver gene induction in the

endoderm and thus the number of cells that become specified to liver, instead of pancreas [57]. The role of Wnt signaling during hepatic induction appears dynamic. The FGF‐BMP induction of the liver occurs when Wnt signaling is being suppressed in the foregut. But shortly after the induction of the hepatic program in the endoderm, Wnt signaling appears to be required for further outgrowth of the endoderm into a liver bud [15]. In zebrafish, expression of Wnt2b in the lateral plate mesoderm, acting through the β‐catenin canonical pathway, appears essential for liver specification in the endoderm and bud induction [58]. Differences in the development of the liver in fish versus amniotes may explain an ­earlier positive role for Wnt signaling in hepatic induction [5]. Like most organs, the liver is asymmetrically positioned with the developing embryo. Patients with heterotaxy syndrome (isomerism), in which organ asymmetry is disrupted, commonly have defects associated with liver function including biliary atresia [59]. The molecular basis for the asymmetric nature of the liver has only recently been explored and our understanding remains rudimentary. However, studies in zebrafish have found that communication between EphrinB1 and EphB3 are crucial to coordinate the movement of the hepatic endoderm with the adjacent lateral plate mesoderm [60]. EphrinB1 controls hepatoblast migration while EphB3 acts in the mesoderm to repel the hepatoblasts. By working together these proteins control the directional migration of the hepatoblasts that is required to correctly position the liver.

FROM HEPATIC ENDODERM TO LIVER BUD The progress from hepatic endoderm to liver bud has been divided into three morphogenetic stages [17]: stage I, the formation of a thickened, columnar hepatic epithelium; stage II, the formation of a pseudostratified epithelium (Figure  2.3c); and stage III, laminin breakdown and hepatic cell migration from the epithelium into the STM. Hepatoblast migration is accompanied by major remodeling of the extracellular matrix surrounding the hepatic cells [61]. During this period, the entire ventral foregut domain extends toward the midgut, bringing the liver region with it (Figure 2.3a,b). The mass of cells emerging from the endodermal epithelium and concentrating in the ­septum transversum is referred to as the liver bud, and the cells within the liver bud are referred to as hepatoblasts. Hepatoblasts will later differentiate into hepatocytes and cholangiocytes (­biliary cells). At stage I of liver bud formation in mammals, endothelial cells, not yet assembled into a vasculature, are adjacent to the hepatoblast epithelium [62]. Genetic depletion of endothelial cells at this stage led to the discovery that endothelial cells are an essential stimulus to further liver bud development: the absence of endothelial cells halted liver bud development at stage II, with hepatoblasts remaining within the limits of the basal epithelial membrane. A similar organogenic stimulatory role has been attributed to endothelial cells in the developing pancreas [63]. In the pancreas, VEGF‐A secreted by beta cells attracts endothelial cells to the developing islet [64]. Endothelial



2:  Embryonic Development of the Liver

(a)

17

(b) neural tube dorsal aorta gut tube pericardio/ peritoneal canal sinus venosus liver bud

(c) hepatic endoderm cells (pseudo-stratified epithelium) endothelial cells and septum transversum mesenchyme cells

Figure 2.3  The beginning of liver development. (a) Mouse embryo, 9 days gestation. Tail is removed; arrow indicates plane of section in (b). (b) 100× magnification of transverse section. Note the thickening of the endodermal epithelium of the gut tube at the region of the liver bud. Boxed area is magnified to 400× in (c). Arrows depict the liver bud and other landmarks of the embryo. Photographs courtesy of J. Rossi.

cells, in turn, form a basal membrane rich in laminins, collagen IV, and fibronectin required for the correct function of beta cells. The nature of the signals originating from the endothelial cells that trigger morphogenesis and cell differentiation in the emerging liver bud is not known. As already stated, the formation of the liver bud requires the degradation of the basal membrane to allow migration of hepatocytes into the STM. One of the genes controlling this step is Prox1 [61]. The Prox1−/− mouse embryonic hepatic endoderm is unable to degrade the basal epithelial membrane, resulting in the clustering of hepatoblasts into a smaller liver. It is not known whether this activity is caused by a slow degradation or production of excess basal membrane components by the embryonic hepatoblasts. Expression of Prox1 is regulated in part by the transcription factor Tbx3. As in Prox1−/− embryos, outgrowth of the liver bud is severely inhibited in Tbx3−/− embryos; however, it has been suggested that Tbx3 acts to control hepatic progenitor cell fate and proliferation through regulating expression of multiple transcription factors and cell cycle regulators, rather than acting solely through Prox1 [65, 66]. As noted above, the homeodomain factor Hex is crucial for liver development [20] and plays an essential role in transformation from stage I to stage II [17]. Hex seems to control different aspects of liver bud formation, including the adequate proliferation of hepatic endoderm cells, the formation of a pseudostratified epithelium, and the stable maintenance of the hepatic cell type [16–18]. How Hex exerts its role is not known, but it has been suggested that it represses sonic hedgehog signaling within

the ventral foregut endoderm, contributing to the exclusion of the intestinal fate [17].

BEYOND THE LIVER BUD: INTRAHEPATIC CHOLANGIOCYTE DIFFERENTIATION The complex signals that cause the emergence of the liver bud are followed closely by distinct signals required to grow the  bud into the liver organ. The hepatoblasts differentiate into hepatocytes and intrahepatic cholangiocytes at about ­embryonic day of gestation 13.5 (E13.5) in the mouse and 7  weeks in humans [67–71]. Substantial advances in our understanding of the mechanisms that control the formation of the intrahepatic ducts have been provided through the study of conditional knockout mouse embryos and the use of Cre‐mediated lineage tracing [72]. The intrahepatic bile ducts consist of cholangiocytes (biliary epithelial cells). They derive from hepatoblasts that surround branches of the portal vein to form a structure called the ductal plate. Portal vein mesenchyme secretes high levels of TGFβ which drives the differentiation of cholangiocytes and suppresses hepatocyte fate. The action of TGFβ is strictly regulated and only a single layer of hepatoblasts that surround the portal vein differentiate to cholangiocytes [73]. The molecular effectors involved in the fate decisions involve the ONECUT transcription

18

THE LIVER:  ROLE OF MESENCHYMAL CELLS IN HEPATOCYTE DIFFERENTIATION

Pre-natal liver development BMP4 FGF2 FGF8 Definitive endoderm FoxA1/2 Gata4/6

(low Wnt)

Notch Hepatic endoderm Hex Prox-1

Bipotential hepatoblasts Hex OC1(HNF-6) OC2

intrahepatic bile duct cholangiocytes HNF-6 HNF-1b Embryonic hepatocytes HNF-4 C/EBPs

Post-natal

Adult hepatocytes

Figure 2.4  Schema of steps of liver development and relevant signals and transcription factors that help mediate the steps. See text for details.

factors HNF6 (OC1) and OC2. Double homozygous OC1/2 mutant embryos present a disturbance of the TGFβ gradient across the axis of the portal vein to the hepatic parenchyma. As a result of the increase in TGFβ signaling, hepatoblasts lying distal to the portal veins express both hepatocyte and cholangiocyte markers [73] (Figure  2.4). Control of TGFβ signaling occurs at multiple levels, including regulation of the concentration of TGFβ type II receptors (TBRII) on the cell surface of periportal hepatoblasts [74]. The Hippo‐YAP signaling pathway has also been implicated in controlling the TGFβ axis by directly promoting expression of TGFβ2 and inhibiting expression of key hepatocyte transcription factors, such as HNF4 [75]. Together with TGFβ, Notch signaling is also required for ­normal biliary tract morphogenesis. Alagille syndrome, a developmental disorder characterized by a paucity of intrahepatic bile ducts (IHBD), is caused predominantly by mutations in the Jagged1 (JAG1) gene, which encodes a ligand for Notch family receptors [76]. Studies using notch loss‐of‐function mice and zebrafish models have indicated that Notch regulates bile duct abundance rather than cell fate specification [77–79]. However, overexpression of the Notch intracellular domain, which drives expression of Notch target genes, increases the expression of cholangiocyte transcription factors and represses hepatocyte transcription factors [80]. These findings imply that Notch could directly promote cholangiocyte differentiation. The mechanism through which Notch controls cholangiocyte formation is complicated by the fact that Notch interaction with its ligands Jagged 1 and Delta can have both cell‐extrinsic and cell‐intrinsic effects [81]. Although during development cholangiocytes derive from the hepatoblasts within the ductal plate, it has recently been shown that hepatocytes retain the capacity to fully transdifferentiate into cholangiocytes in the livers of adult mice that lack an intrahepatic biliary system [82]. Unlike the differentiation of cholangiocytes from hepatoblasts, transdifferentiation of adult hepatocytes occurred independently of Notch and was driven by TGFβ signaling. These findings suggest new strategies for the potential treatment of Alagille syndrome and cholestatic liver disease. On the other hand, a recent report suggests that the cells of the extrahepatobiliary system (gallbladder, the hepatic, cystic and common ducts), share a common origin with the ventral pancreas in a group of cells that are slightly caudal to the liver

bud, or within the caudal portion of the liver bud [82A]. Cell‐ tracing experiments will definitively assess this issue.

ROLE OF MESENCHYMAL CELLS IN HEPATOCYTE DIFFERENTIATION Hepatocyte differentiation and liver morphogenesis is dependent upon the cellular microenvironment. During early stages of liver bud development, as discussed above, the microenvironment is provided by endothelial and mesenchymal cells in the STM and, after E10.5 in the mouse, when liver becomes a hematopoietic organ, hematopoietic stem cells (HSCs) also contribute. The relevance of these mesenchymal cell types is highlighted by genetic loss‐of‐function studies in mice, where genes expressed in mesenchymal cells surrounding the nascent liver but not in hepatoblasts are found to be essential for correct liver formation. One of these cases, Lhx2, is a LIM‐homeobox gene expressed in the STM and mesenchymal components in the adult liver (presumably stellate cells). Lhx2−/− livers display a disrupted cellular organization with increased deposition of ECM and an altered gene expression pattern of the early hepatocytes [83]. The mesenchymal cells are also required for the proliferation of the fetal hepatocytes. When Gata4 was deleted specifically in the liver mesenchymal cells, embryos exhibited severe defects in liver growth, hepatocyte proliferation and survival, and fetal hematopoiesis [84]. Besides the early role of individual endothelial cells in promoting liver bud growth, blood vessels develop de novo within the liver bud (Figure 2.3b), forming a capillary bed that becomes interspersed within the expanding hepatoblast population [85]. These transitions establish the liver’s sinusoidal architecture, which is critical for organ function and sets the stage for the fetal liver to support hematopoiesis. Hematopoietic cells migrate to the early liver first from the yolk sac [86] and later from the aorta–gonad–mesonephros region [87]. Proper intermediate filament expression is critical for blood cell homing, as erythrocytes accumulate excessively in the fetal liver of keratin 8 mutants [88]. Embryos deficient in the heavy metal‐responsive transcription factor MTF‐1 exhibit reduced cytokeratin expression as well as enlarged sinusoids, and dissociated epithelial cells, but not anemia [89]. The primary defect in MTF‐1‐deficient embryos



2:  Embryonic Development of the Liver

appears to be in a failure to control metal homeostasis and the oxidation–reduction state in hepatocytes at mid‐gestation. Both hematopoietic and endothelial cells provide differentiating signals to the hepatoblasts, because heterogeneity in hepatic gene expression has been found to be related to the vascular architecture of the fetal liver [90]. More specifically, mouse gene inactivation studies have shown that oncostatin M signaling from hematopoietic cells to nascent hepatocytes is critical for liver growth [91]. Impaired hematopoietic cell proliferation in c‐myb mutant embryos [92] or impaired erythrocytic cell proliferation in retinoblastoma gene (Rb) mutant embryos [93, 94] also result in impaired liver growth. The failure of hematopoietic cells to migrate to the liver in β1 integrin‐null embryos is therefore also thought to contribute to the liver developmental defect in the β1 mutants [95, 96]. As the fetal liver matures and grows, a capsule forms around it from mesothelial cells. Although the role of the mesothelium during liver development has not been appreciated, a growing body of evidence supports a role in promoting fetal hepatocyte proliferation. N‐myc gene expression in the liver capsule and jumonji gene expression in the hepatic stromal cells help promote growth during mid‐gestation [97–99]. The mesothelial capsule is a rich source of growth factors that may act on the fetal hepatocytes [100]. Moreover, deletion of the zinc finger transcription factor Wt1, which is highly expressed in the liver capsule during development, inhibits proliferation of fetal hepatocytes, diminishes liver growth, and affects formation of the liver lobules [100, 101]. Although many of the genes that affect hepatoblast proliferation are probably most important during the initial transition from the liver bud to the organ stage, inactivation of these genes usually manifests itself as a hematopoietic defect well after the organ is formed. In these cases, a liver ­capsule develops and hematopoietic cells migrate to the liver, but the paucity of hepatoblasts leads to a defect in the hematopoietic environment and, consequently, embryonic lethality. Furthermore, mutations of certain liver regulatory factors yield a fetal liver growth defect due to apoptosis of the hepatoblasts. These proteins include c‐jun [102, 103], IKK2 [104], RelA [105], and XBP‐1 [106].

EMBRYOLOGIC CONTROL OF LIVER REGENERATION The liver is among the few internal organs that can rapidly regenerate after removal of tissue in the adult. Recent studies indicate that the regenerative capacity of the liver is attained as early as the hepatoblast stage in embryos. Tissue complementation experiments, where Hex−/− mouse embryonic stem cells were injected into Hex+/+ blastocysts, discovered that at the onset of liver morphogenesis (E9.0), wild‐type hepatic progenitor cells can increase their proliferation rate to compensate for the failure of Hex−/− cells in the liver bud to survive [17]. In an independent study, when two‐thirds of liver progenitors are genetically ablated between E9.5 and E13.5, the remaining cells are still able to engage in a compensatory growth, compensating to generate a normal‐sized fetal liver within 4 days [107]. The  pancreas, also an endoderm‐derived organ, is unable to

19

grow in this fashion, resulting in major losses of pancreatic mass after partial ablation of its embryonic progenitors [107]. It seems that liver, but not the pancreas, has a full regenerating potential at nearly all developmental stages and while the growth limit of the liver is set by the needed organ size, in the case of the pancreas it is set by the cell division number. It is noteworthy that even when the pancreas does not recover its original size after 50% pancreatectomy, it does recover nearly 100% of its beta cell mass after four weeks [108], suggesting different regenerating capabilities between the endocrine and exocrine compartment. Another interesting aspect is the difference between human and animals, since humans, unlike rodents, do not have a significant increase in beta cell proliferation after partial pancreatectomy [109]. Identifying the genetic programs that cause the marked difference in hepatic and pancreatic regeneration at the embryonic stage, when the cells are otherwise quite immature, could be a convenient way to reveal the basis for adult liver regeneration.

HEPATOCYTE DIFFERENTIATION Hepatocyte differentiation spans from liver specification in ventral foregut endoderm until the postnatal maturation of hepatocytes (Figure  2.4). Downstream of signaling molecules that induce liver differentiation are the transcription factors that execute the liver program, including HNF1, HNF4, HNF6, FOXA, and C/EBP [4–6]. These liver‐specific genes activate, in a cross‐regulatory fashion, each other’s promoters. By establishing both positive and negative feedback loops, the transcription factors generate stable gene regulatory networks that ensure expression of genes that are critical for liver function [72, 110]. In mice, HNF4 acts as a central regulator of hepatocyte gene expression, but is dispensable for early formation of the liver [111]. However, when its role is examined during the differentiation of human iPSCs, the requirement for HNF4 in controlling the onset of hepatocyte gene expression is even more strict [112]. Whether the strict requirement for HNF4α in human cells reflects differences between species or between the model systems remains to be established. Reverse transcription polymerase chain reaction (RT‐PCR) analyses in the mouse revealed that the expression of the primary liver‐enriched transcription factors was unchanged in HNF4−/− livers at E12, except for PXR and HNF1α [111]. However, a large number of tissue‐specific genes involved in the maturation of hepatoblasts to hepatocytes failed to be induced in HNF4−/− embryos. The extensive effects were explained by a genome‐wide chromatin analysis, whereby the location of HNF4 protein bound to promoter sequences of nearly 10 000 genes was assessed simultaneously, using microarray technology [113]. Such analysis was also performed for HNF6 and HNF1α [113]. HNF1α in human hepatocytes was bound to 1.6% of the genes on the array; HNF6 bound 1.7% of the genes; and HNF4α bound to a striking 12% of the genes on the array. The genes bound by HNF4α corresponded to ~42% of the genes bound by RNA polymerase II. That is, nearly half of the active genes tested in the liver are bound by HNF4α, and similar results were obtained by analyzing HNF4α binding to the entire genome [114]. Subsequent studies found that HNF4

20

THE LIVER:  REFERENCES

regulates genes involved in cell junction assembly and adhesion in the developing liver [115], consequently promoting epithelial maturation of the liver parenchyma [116]. The apparently limited role of relevant transcription factors such as HNF1α, in liver development, based on gene inactivation studies in animals, contrasts sharply with ectopic and overexpression studies in cultured hepatic cell lines, which suggested that HNF1α is critical for the expression of a wide variety of liver‐specific genes [117]. Such distinctions indicate how strongly the mechanisms of gene regulation are influenced by whole‐animal physiology, beyond simple notions of gene redundancy (e.g. [118, 119]). Ultimately, it is crucial to determine how inductive signaling pathways and epigenetic modifications converge on regulatory transcription factor genes to coordinately promote early liver differentiation and morphogenesis.

FUTURE AND PERSPECTIVES This review has highlighted ways that our understanding of embryonic liver development has expanded greatly in the past 25 years or so. The regulatory molecules that control liver development are now being used successfully to program hepatic cells from embryonic stem cells and other sources, including the combination of various cell types to make organoids [120–122]. Given the large number of proteins that were discovered in studies of adult livers and yet give embryonic liver phenotypes when deleted in mice, there is high confidence that many of these proteins’ functions in the adult are a recapitulation of activities in the embryo. Thus, continued investigation of embryonic liver development is certain to be instructive about the function, regeneration, and repair of the adult liver, and seems likely to provide new sources of cells and molecules to combat liver disease [123]. Further refinements in the ability to inactivate genes in the early liver and better methods of embryo tissue culture are needed to advance the analysis of liver development. Such advances may come from adaptations of methodology for other endoderm‐derived organ systems [124]. In addition, relevant new genes will emerge from studies of model organisms where genetic screens can be coupled with knowledge of genomic sequences and interrelationships of protein function. New models have also emerged that allow high‐ resolution insight into the molecular basis of cell differentiation [125]. The differentiation of human embryonic stem cells and induced pluripotent stem cells toward a hepatic fate appears to closely mimic known developmental processes [126]. Having access to a tissue culture model of liver cell differentiation can complement the study of traditional animal models. The differentiation of iPSCs in culture is relatively synchronous, which allows researchers to address the molecular changes that rapidly and dynamically occur in response to growth factor signaling [45, 127]. Moreover, tissue culture models allow researchers to apply high‐­throughput methods and genome‐wide assays to identify novel pathways and mechanisms that control cell fate [127–129]. Considering that our mechanistic understanding of liver development has emerged so recently, the prospects are bright for much deeper knowledge and new applications for liver therapies in the future.

ACKNOWLEDGMENTS Thanks to Melanie Song for help in preparing the manuscript. R.B. is supported by a grant from the Spanish Ministry of Science (SAF‐51991R). K.Z.’s research on liver development is supported by a grant from the NIH (GM36477).

REFERENCES 1. Du Bois, A.M. The embryonic liver, in The Liver, Vol. 1 (ed. C.H. Rouiller), Academic Press, New York, 1963, pp. 1–39. 2. Elias, H. Origin and early development of the liver in various vertebrates. Acta Hepat, 1955;3:1–57. 3. Severn, C.B. A morphological study of the development of the human liver. Am J Anat, 1968;133:85–108. 4. Zaret, K.S. Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet, 2002;3(7):499–512. 5. Zaret, K.S. Genetic programming of liver and pancreas progenitors: lessons for stem‐cell differentiation. Nat Rev Genet, 2008;9(5):329–40. 6. Zhao, R. and Duncan, S.A. Embryonic development of the liver. Hepatology, 2005;41(5):956–67. 7. Si‐Tayeb, K., Lemaigre, F.P., and Duncan, S.A. Organogenesis and development of the liver. Dev Cell, 2010;18(2):175–89. 8. Ober, E.A. and Lemaigre, F.P. Development of the liver: Insights into organ and tissue morphogenesis. J Hepatol, 2018;68(5):1049–62. 9. Gordillo, M., Evans, T., and Gouon‐Evans, V. Orchestrating liver development. Development, 2015;142(12):2094–108. 10. Fukuda‐Taira, S. Hepatic induction in the avian embryo: specificity of reactive endoderm and inductive mesoderm. J Embryol Exp Morphol, ­ 1981;63:111–25. 11. Le Douarin, N.M. et al. Origin of hemopoietic stem cells in embryonic bursa of Fabricius and bone marrow studied through interspecific chimeras. Proc Natl Acad Sci U S A, 1975;72(7):2701–5. 12. Tremblay, K.D. and Zaret, K.S. Distinct populations of endoderm cells ­converge to generate the embryonic liver bud and ventral foregut tissues. Dev Biol, 2005;280:87–99. 13. Gualdi, R. et  al. Hepatic specification of the gut endoderm in vitro: cell ­signaling and transcriptional control. Genes Dev, 1996;10:1670–82. 14. Schohl, A. and Fagotto, F. A role for maternal beta‐catenin in early ­mesoderm induction in Xenopus. Embo J, 2003;22(13):3303–13. 15. McLin, V.A., Rankin, S.A., and Zorn, A.M. Repression of Wnt/beta‐catenin signaling in the anterior endoderm is essential for liver and pancreas ­development. Development, 2007;134(12):2207–17. 16. Bort, R. et al. Hex homeobox gene‐dependent tissue positioning is required for organogenesis of the ventral pancreas. Development, 2004;131(4):797–806. 17. Bort, R. et al. Hex homeobox gene controls the transition of the endoderm to  a pseudostratified, cell emergent epithelium for liver bud development. Dev Biol, 2006;290(1):44–56. 18. Hunter, M.P. et al. The homeobox gene Hhex is essential for proper hepatoblast differentiation and bile duct morphogenesis. Dev Biol, 2007;308(2):355–67. 19. Keng, V.W. et al. Homeobox gene Hex is essential for onset of mouse embryonic liver development and differentiation of the monocyte lineage. Biochem Biophys Res Commun, 2000;276(3):1155–61. 20. Martinez‐Barbera, J.P. et al. The homeobox gene hex is required in definitive endodermal tissues for normal forebrain, liver and thyroid formation. Development, 2000;127(11):2433–45. 21. Wallace, K.N. et al. Zebrafish hhex regulates liver development and digestive organ chirality. Genesis, 2001;30(3):141–3. 22. Zaret, K., Developmental competence of the gut endoderm: genetic potentiation by GATA and HNF3/fork head proteins. Dev Biol, 1999;209:1–10. 23. Ang, S.L. et al. The formation and maintenance of the definitive endoderm lineage in the mouse: involvement of HNF3/forkhead proteins. Development, 1993;119(4):1301–15. 24. Monaghan, A.P. et al. Postimplantation expression patterns indicate a role for the mouse forkhead/HNF‐3α, β, and γ genes in determination of the definitive endoderm, chordamesoderm and neuroectoderm. Development, 1993;119:567–78.



2:  Embryonic Development of the Liver

25. Ruiz i Altaba, A. et  al. Sequential expression of HNF‐3α and HNF‐3β by embryonic organizing centers: the dorsal lip/node, notochord, and floor plate. Mech Dev, 1993;44:91–108. 26. Sasaki, H. and Hogan, B.L.M. Differential expression of multiple fork head related genes during gastrulation and pattern formation in the mouse embryo. Development, 1993;118:47–59. 27. Arceci, R. et  al. Mouse GATA‐4: a retinoic acid‐inducible GATA‐binding transcription factor expressed in endodermally derived tissues and heart. Mol Cell Biol, 1993;13:2235–46. 28. Gao, X. et al. Distinct functions are implicated for the GATA‐4, ‐5, and ‐6 transcription factors in the regulation of intestine epithelia cell differentiation. Mol Cell Biol, 1998;18:2901–11. 29. Laverriere, A.C. et al. GATA‐4/5/6, a subfamily of three transcription factors transcribed in developing heart and gut. J Biol Chem, 1994;269:23177–84. 30. Suzuki, E. et al. The human GATA‐6 gene: structure, chromosomal location, and regulation of expression by tissue‐specific and mitogen‐responsive signals. Genomics, 1996;38:283–90. 31. Bossard, P. and Zaret, K.S. GATA transcription factors as potentiators of gut endoderm differentiation. Development, 1998;125:4909–17. 32. Liu, J.K., DiPersio, C.M., and Zaret, K.S. Extracellular signals that regulate liver transcription factors during hepatic differentiation in vitro. Mol Cell Biol, 1991;11(2):773–84. 33. Cascio, S. and Zaret, K.S. Hepatocyte differentiation initiates during endodermal‐mesenchymal interactions prior to liver formation. Development, 1991;113:217–25. 34. Mueller, P.R. and Wold, B. In vivo footprinting of a muscle specific enhancer by ligation mediated PCR. Science, 1989;246:780–6. 35. Lee, C.S. et  al. The initiation of liver development is dependent on Foxa transcription factors. Nature, 2005;435(7044):944–7. 36. Holtzinger, A. and Evans, T. Gata4 regulates the formation of multiple organs. Development, 2005;132(17):4005–14. 37. Fisher, J.B. et al. GATA6 is essential for endoderm formation from human pluripotent stem cells. Biol Open, 2017;6(7):1084–95. 38. Tiyaboonchai, A. et al. GATA6 plays an important role in the induction of human definitive endoderm, development of the pancreas, and functionality of pancreatic beta cells. Stem Cell Rep, 2017;8(3):589–604. 39. Shi, Z.D. et al. Genome editing in hPSCs reveals GATA6 haploinsufficiency and a genetic interaction with GATA4 in human pancreatic development. Cell Stem Cell, 2017;20(5):675–688.e6. 40. McKeehan, W.L., Wang, F., and Kan, M. The heparan sulfate‐fibroblast growth factor family: diversity of structure and function. Prog Nucl Acid Res Mol Biol, 1998;59:135–76. 41. Szebenyi, G. and Fallon, J.F. Fibroblast growth factors as multifunctional signaling factors. Int Rev Cytol, 1999;185:45–106. 42. Jung, J. et al. Initiation of mammalian liver development from endoderm by fibroblasts growth factors. Science, 1999;284:1998–2003. 43. Wang, J.‐K., Gao, G., and Goldfarb, M. Fibroblast growth factor receptors have different signaling and mitogenic potentials. Mol Cell Biol, 1994;14:181–8. 44. Calmont, A. et  al. An FGF response pathway that mediates hepatic gene induction in embryonic endoderm cells. Dev Cell, 2006;11(3):339–48. 45. Twaroski, K. et al. FGF2 mediates hepatic progenitor cell formation during human pluripotent stem cell differentiation by inducing the WNT antagonist NKD1. Genes Dev, 2015;29(23):2463–74. 46. Dudley, A.M., Rougeulle, C., and Winston, F. The Spt components of SAGA facilitate TBP binding to a promoter at a post‐activator‐binding step in vivo. Genes Dev, 1999;13(22):2940–5. 47. Lyons, K.M., Pelton, R.W., and Hogan, B.L. Patterns of expression of murine Vgr‐1 and BMP‐2a RNA suggest that transforming growth factor‐ beta‐like genes coordinately regulate aspects of embryonic development. Genes Dev, 1989;3(11):1657–68. 48. Solloway, M.J. and Robertson, E.J. Early embryonic lethality in Bmp5;Bmp7 double mutant mice suggests functional redundancy within the 60A subgroup. Development, 1999;126(8):1753–68. 49. Winnier, G. et al. Bone morphongenetic protein‐4 is required for mesoderm formation and patterning in the mouse. Genes Dev, 1995;9:2105–16. 50. Zhang, H. and Bradley, A. Mice deficient for BMP2 are nonviable and have  defects in amnion/chorion and cardiac development. Development, 1996;122(10):2977–86. 51. Mishina, Y. et al. Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev, 1995;9(24):3027–37.

21

52. Roelen, B.A. et al. Differential expression of BMP receptors in early mouse development. Int J Dev Biol, 1997;41(4):541–9. 53. Rossi, J.M. et  al. Distinct mesodermal signals, including BMP’s from the septum transversum mesenchyme, are required in combination for hepatogenesis from the endoderm. Genes Dev, 2001;15:1998–2009. 54. Shin, D. et al. Bmp and Fgf signaling are essential for liver specification in zebrafish. Development, 2007;134(11):2041–50. 55. Deutsch, G. et al. A bipotential precursor population for pancreas and liver within the embryonic endoderm. Development, 2001;128:871–81. 56. Wandzioch, E. and Zaret, K.S. Dynamic signaling network for the specification of embryonic pancreas and liver progenitors. Science, 2009. 324(5935), 1707–10. 57. Xu, C.R. et al. Chromatin “prepattern” and histone modifiers in a fate choice for liver and pancreas. Science, 2011;332(6032):963–6. 58. Ober, E.A. et  al. Mesodermal Wnt2b signalling positively regulates liver specification. Nature, 2006;442(7103):688–91. 59. Lakshminarayanan, B. and Davenport, M. Biliary atresia: A comprehensive review. J Autoimmun, 2016;73:1–9. 60. Cayuso, J. et  al. EphrinB1/EphB3b coordinate bidirectional epithelial‐­ mesenchymal interactions controlling liver morphogenesis and laterality. Dev Cell, 2016;39(3):316–28. 61. Sosa‐Pineda, B., Wigle, J.T., and Oliver, G. Hepatocyte migration during liver development requires Prox1. Nat Genet, 2000;25(3):254–5. 62. Matsumoto, K. et  al. Liver organogenesis promoted by endothelial cells prior to vascular function. Science, 2001;294:559–63. 63. Lammert, E., Cleaver, O., and Melton, D. Induction of pancreatic differentiation by signals from blood vessels. Science, 2001;294(5542):564–7. 64. Nikolova, G. et  al. The vascular basement membrane: a niche for insulin gene expression and beta cell proliferation. Dev Cell, 2006;10(3):397–405. 65. Suzuki, A. et  al. Tbx3 controls the fate of hepatic progenitor cells in liver ­development by suppressing p19ARF expression. Development, 2008;135(9): 1589–95. 66. Ludtke, T.H. et  al. Tbx3 promotes liver bud expansion during mouse ­development by suppression of cholangiocyte differentiation. Hepatology, 2009;49(3):969–78. 67. Germain, L., Blouin, M.J., and Marceau, N. Biliary epithelial and hepatocytic cell lineage relationships in embryonic rat liver as determined by the differential expression of cytokeratins, α‐fetoprotein, albumin, and cell ­surface‐exposed components. Cancer Res, 1988;48(17):4909–18. 68. Shiojiri, N. Enzymo‐ and immunocytochemical analyses of the differentiation of liver cells in the prenatal mouse. J Embryol Exp Morphol, 1981;62:139–52. 69. Shiojiri, N. Analysis of differentiation of hepatocytes and bile duct cells in developing mouse liver by albumin immunofluorescence. Dev Growth Differ, 1984;26:555–61. 70. Shiojiri, N. Development and differentiation of bile ducts in the mammalian liver. Microsc Res Technol, 1997;39(4):328–35. 71. Antoniou, A. et al. Intrahepatic bile ducts develop according to a new mode of tubulogenesis regulated by the transcription factor SOX9. Gastroenterology, 2009;136(7):2325–33. 72. Gerard, C., Tys, J., and Lemaigre, F.P. Gene regulatory networks in differentiation and direct reprogramming of hepatic cells. Semin Cell Dev Biol, 2017;66:43–50. 73. Clotman, F. et al. Control of liver cell fate decision by a gradient of TGF beta signaling modulated by Onecut transcription factors. Genes Dev, 2005;19(16): 1849–54. 74. Takayama, K. et al. CCAAT/enhancer binding protein‐mediated regulation of TGFbeta receptor 2 expression determines the hepatoblast fate decision. Development, 2014;141(1):91–100. 75. Lee, D.H. et al. LATS‐YAP/TAZ controls lineage specification by regulating TGFbeta signaling and Hnf4alpha expression during liver development. Nat Commun, 2016;7:11961. 76. Oda, T. et  al. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat Genet, 1997;16(3):235–42. 77. Kodama, Y. et al. The role of notch signaling in the development of intrahepatic bile ducts. Gastroenterology, 2004;127(6):1775–86. 78. Lorent, K. et  al. Inhibition of Jagged‐mediated Notch signaling disrupts zebrafish biliary development and generates multi‐organ defects compatible with an Alagille syndrome phenocopy. Development, 2004;131(22):5753–66. 79. Lozier, J., McCright, B., and Gridley, T. Notch signaling regulates bile duct morphogenesis in mice. PLoS ONE, 2008;3(3):e1851.

22

THE LIVER:  REFERENCES

80. Tanimizu, N. and Miyajima, A. Notch signaling controls hepatoblast ­differentiation by altering the expression of liver‐enriched transcription ­factors. J Cell Sci, 2004;117(15):3165–74. 81. Kaylan, K.B. et al. Combinatorial microenvironmental regulation of liver progenitor differentiation by Notch ligands, TGFbeta, and extracellular matrix. Sci Rep, 2016;6:23490. 82. Schaub, J.R. et al. De novo formation of the biliary system by TGFbeta‐ mediated hepatocyte transdifferentiation. Nature, 2018;557(7704):247–51. 82. A. Spence, J.R. et al. Sox17 regulates organ lineage segregation of ventral foregut progenitor cells. Dev Cell, 2009;17(1):62–74. 83. Wandzioch, E. et al. Lhx2−/− mice develop liver fibrosis. Proc Natl Acad Sci U S A, 2004;101(47):16549–54. 84. Delgado, I. et  al. GATA4 loss in the septum transversum mesenchyme ­promotes liver fibrosis in mice. Hepatology, 2014;59(6):2358–70. 85. Medlock, E.S. and Haar, J.L. The liver hemopoietic environment: I. Developing hepatocytes and their role in fetal hemopoiesis. Anat Rec, 1983;207(1):31–41. 86. Johnson, G.R. and Moore, M.A. Role of stem cell migration in initiation of mouse foetal liver haemopoiesis. Nature, 1975;258:726–8. 87. Medvinsky, A. and Dzierzak, E. Definitive hematopoiesis is autonomously initiated by the AGM region. Cell, 1996;86(6):897–906. 88. Baribault, H. et  al. Mid‐gestational lethality in mice lacking keratin 8. Genes Dev, 1993;7:1191–202. 89. Günes, C. et al. Embryonic lethality and liver degeneration in mice lacking the metal‐responsive transcriptional activator MTF‐1. EMBO J, 1998;17:2846–54. 90. Gaasbeek Janzen, J.W. et al. Gene expression in derivatives of embryonic foregut during prenatal development of the rat. J Histochem Cytochem, 1988;36(10):1223–30. 91. Kamiya, A. et al. Fetal liver development requires a paracrine action of oncostatin M through the gp130 signal transducer. Embo J, 1999;18(8):2127–36. 92. Mucenski, M.L. et  al. A functional c‐myb gene is required for normal murine fetal hepatic hematopoiesis. Cell, 1991;65(4):677–89. 93. Jacks, T. et  al. Effects of an Rb mutation in the mouse. Nature, 1992;359:295–300. 94. Lee, E.Y. et al. Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature, 1992;359(6393):288–94. 95. Hirsch, E. et al. Impaired migration but not differentiation of haematopoietic stem cells in the absence of beta1 integrins. Nature, 1996;380(6570):171–5. 96. Houssaint, E., Differentiation of the mouse hepatic primordium. I. An ­analysis of tissue interactions in hepatocyte differentiation. Cell Differ, 1980;9:269–79. 97. Giroux, S. and Charron, J. Defective development of the embryonic liver in N‐myc‐deficient mice. Dev Biol, 1998;195:16–28. 98. Motoyama, J. et  al. Organogenesis of the liver, thymus and spleen is affected in jumonji mutant mice. Mech Dev, 1997;66:27–37. 99. Sawai, S. et al. Defects of embryonic organogenesis resulting from targeted disruption of the N‐myc gene in the mouse. Development, 1993;117:1445–55. 100. Onitsuka, I., Tanaka, M., and Miyajima, A. Characterization and functional analyses of hepatic mesothelial cells in mouse liver development. Gastroenterology, 2010;138(4):1525–35, 1535.e1–6. 101. Ijpenberg, A. et al. Wt1 and retinoic acid signaling are essential for stellate cell development and liver morphogenesis. Dev Biol, 2007;312(1):157–70. 102. Eferl, R. et al. Functions of c‐Jun in liver and heart development. J Cell Biol, 1999;145(5):1049–61. 103. Hilberg, F. et al. c‐Jun is essential for normal mouse development and hepatogenesis. Nature, 1993;365:179–81. 104. Li, Q. et  al. Severe liver degeneration in mice lacking the 1κB kinase 2 gene. Science, 1999;284:321–5. 105. Beg, A.A. et al. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF‐KB. Nature, 1995;376:167–70. 106. Reimold, A.M. et al. An essential role in liver development for transcription factor XBP‐1. Genes Dev, 2000;14(2):152–7.

107. Stanger, B.Z., Tanaka, A.J., and Melton, D.A. Organ size is limited by the number of embryonic progenitor cells in the pancreas but not the liver. Nature, 2007;445(7130):886–91. 108. Lee, C.S. et al. Regeneration of pancreatic islets after partial pancreatectomy in mice does not involve the reactivation of neurogenin‐3. Diabetes, 2006;55(2):269–72. 109. Menge, B.A. et  al. Partial pancreatectomy in adult humans does not ­provoke beta‐cell regeneration. Diabetes, 2008;57(1):142–9. 110. Kyrmizi, I. et  al. Plasticity and expanding complexity of the hepatic transcription factor network during liver development. Genes Dev, ­ 2006;20(16):2293–305. 111. Li, J., Ning, G., and Duncan, S.A. Mammalian hepatocyte differentiation requires the transcription factor HNF‐4alpha. Genes Dev, 2000;14(4): 464–74. 112. DeLaForest, A. et  al. HNF4A is essential for specification of hepatic progenitors from human pluripotent stem cells. Development, ­ 2011;138(19):4143–53. 113. Odom, D.T. et al. Control of pancreas and liver gene expression by HNF transcription factors. Science, 2004;303(5662):1378–81. 114. Hoffman, B.G. et  al. Locus co‐occupancy, nucleosome positioning, and H3K4me1 regulate the functionality of FOXA2‐, HNF4A‐, and PDX1‐ bound loci in islets and liver. Genome Res, 2010;20(8):1037–51. 115. Battle, M.A. et al. Hepatocyte nuclear factor 4alpha orchestrates expression of cell adhesion proteins during the epithelial transformation of the developing liver. Proc Natl Acad Sci U S A, 2006;103(22):8419–24. 116. Parviz, F. et al. Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis. Nat Genet, 2003;34(3): 292–6. 117. Tronche, F. et al. Hepatocyte nuclear factor 1(HNF1) and liver gene expression, in Liver Gene Expression (eds. F. Tronche and M. Yaniv), R.G. Landes Company, 1994, pp. 155–82. 118. Barbacci, E. et al. Variant hepatocyte nuclear factor 1 is required for visceral endoderm specification. Development, 1999;126(21):4795–805. 119. Cereghini, S. et al. Expression patterns of vHNF1 and HNF1 homeoproteins in early postimplantation embryos suggest distinct and sequential developmental roles. Development, 1992;116:783–97. 120. Gouon‐Evans, V. et  al. BMP‐4 is required for hepatic specification of mouse embryonic stem cell‐derived definitive endoderm. Nat Biotechnol, 2006;24(11):1402–11. 121. Loh, K.M. et  al. Efficient endoderm induction from human pluripotent stem cells by logically directing signals controlling lineage bifurcations. Cell Stem Cell, 2014;14(2):237–52. 122. Takebe, T. et  al. Vascularized and functional human liver from an iPSC‐ derived organ bud transplant. Nature, 2013;499(7459):481–4. 123. Goldman, O. and Gouon‐Evans, V. Human pluripotent stem cells: myths and future realities for liver cell therapy. Cell Stem Cell, 2016;18(6):703–6. 124. Wells, J.M. and Melton, D.A. Vertebrate endoderm development. Annu Rev Cell Dev Biol, 1999;15:393–410. 125. Vallier, L. Heps with pep: direct reprogramming into human hepatocytes. Cell Stem Cell, 2014;14(3):267–9. 126. Yiangou, L. et  al. Human pluripotent stem cell‐derived endoderm for modeling development and clinical applications. Cell Stem Cell, ­ 2018;22(4):485–99. 127. Bertero, A. et al. Activin/nodal signaling and NANOG orchestrate human embryonic stem cell fate decisions by controlling the H3K4me3 chromatin mark. Genes Dev, 2015;29(7):702–17. 128. Jing, R., Duncan, C.B., and Duncan, S.A. A small‐molecule screen reveals that HSP90beta promotes the conversion of induced pluripotent stem cell‐ derived endoderm to a hepatic fate and regulates HNF4A turnover. Development, 2017;144(10):1764–74. 129. Camp, J.G. et al. Multilineage communication regulates human liver bud development from pluripotency. Nature, 2017;546(7659):533–8.

PART TWO: THE CELLS

SECTION A: CELL BIOLOGY OF THE LIVER

3

Cytoskeletal Motors: Structure and Function in Hepatocytes Mukesh Kumar1, Arnab Gupta2, and Roop Mallik1 Department of Biological Sciences, Tata Institute of Fundamental Research, Navy Nagar, Colaba, Mumbai, India Department of Biological Sciences, Indian Institute of Science Education and Research Kolkata, Mohanpur, India

1 2

INTRODUCTION The hepatocyte is the major epithelial cell of the liver, making up to 70–80% of liver mass. Hepatocytes produce and secrete bile, extract specific molecules from blood, and secrete oth­ ers into blood flow. This “sieving” function makes the liver a key homeostatic organ and requires intracellular transport of a vast variety of lipids/proteins towards multiple distinctly polarized surfaces in hepatocytes. In order to do this, hepatocytes must partition as well as connect two different environments – the blood and the bile. To serve this function in the liver, hepatocytes are arranged in cords. An individual cell is polygonal and faces at least two blood sinusoids (the basal domain). A branched network of grooves between adjacent cells forms the bile canaliculus, representing the apical domain. Such a polygonal shape means that hepatocytes do not have a single basolateral‐to‐apical axis, and the transcy­ totic pathways between basolateral and apical membranes are more complex than other single apical–basolateral axis polar­ ized epithelial cells, such as enterocytes, renal epithelial cells, etc. [1]. Most newly synthesized membrane and secre­ tory proteins exiting the trans‐Golgi network (TGN) are directed via the basolateral membrane, with less TGN‐exit traffic directed towards the apical side (Figure  3.1a). Similarly, endocytosed cargos including plasma membrane (PM) fragments internalized at the basolateral (blood) side are transported to lysosomes, the apical PM, or bile via tran­ scytotic pathways. Since these multiple and distinct sorting steps of the cargos in the biosynthetic and endocytic path­ ways rely on microtubules, microfilaments (actin), and their respective motor proteins, the cytoskeletal system functions as the primary coordinator between the apical and basolateral regions in hepatocytes [2].

HEPATOCYTE POLARITY AND THE ACTIN NETWORK Although hepatocyte polarity is well‐described at an anatomical level, little is known about the molecules and the mechanisms that initiate, establish, and maintain this polarization. During mouse development, a liver bud composed of nonpolarized hepa­ toblasts is detected as early as embryonic day 9.5. Studies on mouse liver development by Feracci et al. have demonstrated that initiation of hepatocyte polarization occurs very early in develop­ ment (day 15). From embryonic day 17 onwards, hepatoblasts aggregate together resembling acini. The acini exhibit a simple polar phenotype, with their apical surfaces facing a central lumen [3]. Gradually, into the postnatal stages, the simple polarity changes to polygonal hepatic polarity, a phenomenon adeptly emulated in WIF‐B cells. This entire process of ­hepatocyte as well as WIF‐B polarization is heavily dependent on the spatiotemporal location and function of the cytoskeletal ­proteins. The anatomy of micro­ filament in the polarized hepatocyte was studied in great detail by Ishii et  al. [4]. WIF‐B cells maintain PM proteins that are restricted to either the apical (bile  canalicular) or basolateral domain, with tight junctions demarcating the apical–basolateral boundaries. Phalloidin staining reveals a thin but highly intense ringed network of actin f­ ilaments around the bile canaliculus/api­ cal membrane (BC), along with a sparse network close to the basolateral membrane (Figure 3.1a). Densely concentrated actin filaments were identified around the bile canaliculi in the form of microvillous core filaments and pericanalicular web filaments. The microvillous core filaments exhibited growth at their apical ends. In contrast, filaments of the pericanalicular web, running parallel to the cell surface, showed no fixed polarities. Interestingly, adjacent ­ filament pairs often showed opposite polarities, an alignment necessary for filament sliding. A group

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

28

THE LIVER:  HEPATOCYTE POLARITY AND THE ACTIN NETWORK

(a)

(b) Bile canaliculus Cargo binding domain

Myosin V

SAC ?

Golgi

Coiled coil domain

?

+

?

?

Nucleus

Light chain binding domain Motor domain

Basola + te endos ral ome MyoVb with ATP7B as cargo Cortical actin

Figure 3.1  Actin‐dependent trafficking in hepatocytes and the myosin‐V motor. (a) Most of the apically targeted proteins (copper transporting ATPase, ATP7B in this case) travels to the destination via basolateral endosomes. The microtubule‐based motor proteins that drive the cargo from the TGN to the basolateral endosomes and subsequently to the subapical compartment (SAC) has not been determined (shown as interrogative clause). Once the cargo‐carrying vesicles reach the SAC, myosin Vb drives its recycling between the SAC and the bile canaliculus/apical mem­ brane. The myosin Vb remains anchored to the actin microfilaments that form a dense network beneath the apical membrane. (b) The general structure of myosin V that is conserved in all the myosin V proteins (MyoVa, Vb, and Vc).

of sporadic actin f­ ilaments with no uniform polarity acts as a link between the canalicular membrane and coated vesicles [4]. Such spatial organization of actin filaments is implicated in mainte­ nance of microvilli length, contraction of the canalicular walls facilitating bile secretion, and transport of coated vesicles in the apical membrane (bile canalicular membrane). Treatment of hepatocytes with cytochalasin D, which targets actin, completely inhibited self‐assembly of rat hepatocytes into spheroids. Thus, hepatocytes require an intact actin network to self‐assemble effi­ ciently into functional tissue‐like structures [5]. Of all the actin‐related motors, most work has been done on myosin V (Figure  3.1b). Myosin Vb is most abundant in the liver as well as the WIF‐B cells [6]. It is an unconventional myo­ sin regulating vesicle trafficking by binding Rab11a [7, 8]. Myosin Vb inhibition induces failure to attain complete polarity as well as trafficking defects of various apically targeted cargos, marked by accumulation of metabolites leading to a diseased state. Understanding human genetic diseases linked to muta­ tions in myosin Vb and its effector proteins has been greatly helpful in comprehending liver polarity and polarized protein trafficking. Mutations in myosin Vb were found to cause micro­ villus inclusion disease (MVID), which is characterized by a lack of apical microvilli and intracellular structures containing microvilli in the intestine, leading to a severe form of congenital diarrhea [9]. Years before detailed mechanistic understanding of MVID phenotypes, Wakabayashi et  al., while investigating mechanisms of apical targeting, serendipitously discovered that knocking down of Rab11 before its polarization prevented cana­ licular formation in WIF‐B9 cells. Polarization could also be arrested by overexpression of the Rab11a‐GDP locked form or the myosin Vb tail dominant negative (DN) construct. In WIF‐ B9 cells, which lack bile canaliculi, apical ABC (ATP‐binding cassette) transporters colocalize with transcytotic membrane

proteins in Rab11a‐containing endosomes and did not distribute to the PM [10]. The study demonstrated that polarization of hepatocytes requires recruitment of Rab11a and myosin Vb to intracellular membranes that contain apical ABC transporters, permitting their targeting to the PM. Stable MYO5 knockdown (MY05B‐KD) in CaCo‐2 cells induced loss of microvilli, alterations in tight junction proteins, and disruption of polarized trafficking [9, 11, 12]. Due to short life expectancy, liver phenotypes usually go undetected in MVID patients [13]. However, MVID patients developed a cholestatic liver disease similar to progressive familial intrahe­ patic cholestasis (PFIC) and benign recurrent cholestasis (BRC) [14]. PFIC and BRC have been linked to mutations in ABCB11 (encoding bile salt export pump, BSEP) and ATP8B1 (encoding FIC1 protein) genes, respectively [15, 16]. Interestingly, in neither of those two genes were disease‐causing mutations ­ detected in any of these patients. Immunohistochemistry stud­ ies ­suggested mislocalization of BSEP in cytoplasmic vesicular staining as compared to controls, where the protein primarily localized at the apical membrane. This phenotype is a conse­ quence of impairment of the MY05B/RAB11A apical recycling endosome pathway in hepatocytes [14]. Utilizing polarized WIF‐B cells, it was demonstrated that the copper‐exporting ATPase ATP7B (also known as the Wilson disease protein) is a major cargo of myosin Vb in the liver. Mutations in ATP7B imparts a nontrafficking phenotype to the protein, causing a ­disorder of liver copper accumulation termed Wilson disease [6]. Interestingly, a similar copper accumulation phenotype was demonstrated in polarized WIF‐B cells when myosin Vb was acutely knocked down using a DN mutant. Detailed micro­ scopic studies revealed that in the absence of functional myosin Vb, ATP7B‐containing vesicles accumulated beneath the ­apical membrane and failed to fuse with it. This study further



3:  Cytoskeletal Motors: Structure and Function in Hepatocytes

delineated the physical and functional localizations of micro­ tubule and cortical actin with respect to apical trafficking. Apically targeted cargos exit the TGN and traffic basolaterally before getting redirected apically in vesicles utilizing the micro­ tubular network [17]. The vesicles are then transferred to the myosin Vb motor anchored at the cortical actin that is located beneath the apical membrane (Figure  3.1a). This transfer of ­cargos occurs in specialized endosomal compartments called subapical compartments (SACs) lying in close proximity to the microtubule organizing center (MTOC). Studies on the role of myosin Vb in trafficking of apically targeted proteins reveals a fascinating spatiotemporal correlation with the polarization ­status of WIF‐B cells. In completely polarized WIF‐B cells, myosin Vb is primarily localized as a ring‐like pattern around the F‐actin ring at the bile canaliculus (Figure 3.1a). However, in pre‐polarized cells, myosin Vb, as well as the myosin Vb tail DN protein, localize to a tight juxtanuclear intracellular site described as the “apical compartment” [18]. This apical compart­ ment is an intracellular membranous structure clustered near the minus ends of microtubules that are labeled with γ‐tubulin, where apically destined cargos accumulate selectively.

MICROTUBULES: HISTORY, STRUCTURE AND DYNAMICS Microtubules (MTs) are tracks for transport of intracellular cargo by microtubule‐dependent motors such as dynein and kinesin (Figure  3.2a). MTs were observed as rod‐like structures in ­spermatozoid of Sphagnum and ciliated epithelia [19, 20], and were hypothesized to power the beating movements of cilia. Improvement in structural preservation techniques [21] showed that MTs contain multiple (for example 13) subunits arranged in a circular fashion [22]. Ledbetter and Porter named the structure “microtubule,” because the protofilaments were organized around a hollow center and therefore appeared tubular. Taylor and cow­ orkers characterized a single [3H]‐colchicine binding protein with (a)

a sedimentation coefficient of 6S and an Mr of 110 000–120 000 [23]. Later, it was shown that the denaturation of the 6S protein produced α and β monomeric subunits (Mr = 55 000). These two very similar subunits were found to exist within MTs in a 1 : 1 molar ratio, forming αβ heterodimers [24]. The term “tubulin” was coined by Mohri in 1968 [25]. The ability of tubulin purified from brain homogenate to polymerize in vitro in the presence of GTP, Mg2+, and EGTA greatly improved our understanding of MT assembly and dynamics [26]. A third type of tubulin, γ‐tubulin, was discovered and was found to be associated with the minus end of MTs near the centrosome. Although γ‐tubulin is not a major component of MTs, it plays an i­ mportant role in the assembly and expression of MT arrays in eukaryotic cells [27]. MTs undergo rapid stochastic growth and shrinkage (Figure 3.2b), known as “dynamic instability” [28, 29]. This phe­ nomenon depends on various intrinsic and extrinsic factors, including the concentration of tubulin. The tubulin subunits asso­ ciate/dissociate at different rates at the plus and minus ends of protofilaments. The critical concentration is the minimum con­ centration of tubulin dimers above which there is a net growth of tubulin polymer. Nucleotides at the N‐site are not available for exchange, unlike nucleotides at the E‐site which can exchange with the solution [30]. Preferential polymerization of tubulin dimers at the β‐tubulin end defines the MT polarity. The rapidly polymerizing β‐end is called the plus end whereas the slow‐ growing end with exposed α‐tubulin is called the minus end. GTP hydrolysis within the polymer is slower than GTP addition at the plus end. Resultant accumulation of a “GTP cap” at the plus end favors dimer addition [28, 30]. As the tubulin dimer concentration decreases below a critical value, dimer addition at the plus end is reduced in comparison to GTP hydrolysis. This leads to a rapid shrinkage of MTs and is termed catastrophe. Catastrophe can be rescued by the addition of GTP‐tubulin [28]. The competition between polymerization and catastrophe can be regulated by MT‐associated proteins and MT‐severing proteins inside cells [31, 32]. Although dynamic, the MTs still collec­ tively maintain a stiff intracellular skeleton to impart shape to a cell and support intracellular cargo transport.

(c)

(d)

+

Cargo

29

IC LIC

Light chain

Tail domain

cro

tub ule

Linker Depolymerization phase Polymerization phase

β

Mi

–

(b)

α

Heavy chain

Stalk domain Head/motor domain MT binding domain

Motor domain

Dynein Kinesin

Figure 3.2  Microtubules and associated motor proteins. (a) Microtubules normally extend from the minus end of the microtubule (anchored at the microtubule organizing center (MTOC)) to the plus end (fast‐growing). (b) Microtubules consisting of αβ‐tubulin dimers (shown) remain in a dynamic mode, alternating between polymerization and depolymerization phases. The molecular structures of (c) dynein (retrograde motor) and (d) conventional kinesin (anterograde motor) are shown. IC, intermediate chain; LIC, light intermediate chain.

30

THE LIVER:  MICROTUBULE ORGANIZATION, POLARIZATION, AND CARGO TRAFFICKING IN HEPATOCYTES

MICROTUBULE MOTORS MTs also serve as tracks against which molecular motors can generate mechanical force (Figure 3.2a). Motors utilize ATP as fuel to generate force and movement for intracellular transport, cell division etc. Gibbons and Rowe were the first to report MT motor‐based ATPase activity when they purified the dynein ATPase from Tetrahymena cilia [33]. The term dynein was derived from the Greek word dyne, meaning “force,” as this motor was thought to be responsible for beating of cilia. Discovery of dynein in the flagella of sea urchin spermatozoa confirmed that dynein was responsible for ciliary and flagellar beating [34]. Although the MT motor proteins were suspected to power mitosis and organelle transport, molecular characteriza­ tion was difficult due to their low cytosolic concentration. The use of extracts from squid giant axon paved the way for studying intracellular transport in vitro [35, 36]. This was followed by the biochemical purification of two organelle‐associated cytoplas­ mic motors, namely dynein and kinesin, that usually transport cargo in opposite directions along MTs [37, 38].

Cytoplasmic dynein Dynein (Figure  3.2c) is a multi‐subunit protein complex of ~2000 kDa consisting of two heavy chains (DHC; each ~530 kDa), a variable number of light chains (LCs) and intermedi­ ate chains (ICs). The ICs bind to DHC and multiple LCs bind to ICs at separate sites [39]. DHC is a single polypeptide and an unusual member of the hexameric AAA+ superfamily (ATPase associated with various cellular activities). The two motor domains of DHC interact with MTs. Each motor domain contains six AAA+ sub‐domains, two of which (AAA1 and AAA3) can bind and hydrolyze ATP [40, 41]. Force is generated by a shift in the linker domain that joins the dynein tail with AAA1 [42]. Intermediate chains are thought to bind dynein to its many cargos, such as vesicles, Golgi body, lipid droplets, kinetochores, and mRNA [43]. Recent studies suggest that different adaptor proteins also mediate the cargo binding and motor activities of dynein. Jun N‐termi­ nal kinase interacting proteins (JIP), Rab7‐interacting lysoso­ mal proteins RILP/p150Glued and sorting nexins are well‐known adaptor proteins for dynein [39, 42, 44, 45]. Intriguingly, although the dynein motor generates a smaller force than kinesin, it can work in large teams by virtue of an in‐built gear mechanism and the ability to catch‐bond to MTs against high opposing load [46, 47]. Taking these findings further, choles­ terol‐enriched lipid rafts on an endosome/phagosome mem­ brane were shown to serve as a platform where many dynein motors can cluster together. This geometric clustering allowed a large team of dyneins to simultaneously contact a single MT, and generate a collective force that rapidly transports the phagosome towards degradative lysosomes [48, 49]. Because cytoplasmic dynein transports cargos towards the minus end of MTs, it is known as a retrograde motor [50]. The orienta­ tion of MTs differs in epithelial cells of different origins. Therefore, the dynein‐driven motion may localize cargos towards the apical or the basolateral membrane, depending on cell type. In hepatocytes, dynein likely drives cargos towards the perinuclear region [51, 52].

Kinesin Since the discovery of conventional kinesin in 1985 by Brady and Vale [36, 53], a large number of ATPases have been categorized into the kinesin superfamily [53, 54]. Kinesins are classified into 14 classes (kinesins 1–14), all these motors contain an ATP‐bind­ ing domain with significant homology [55]. Kinesins are protein complexes of approximately 380 kDa, consisting of two heavy chains (each 120 kDa) and two light chains (each 64 kDa) [56]. As shown in Figure 3.2d, the N‐terminal globular head domains of the heavy chains are the “motor domains” that bind to the MT [57] and also hydrolyze ATP to provide energy for movement [58]. The processive motion of conventional kinesins is explained by their “hand‐over‐hand” walking [59]. The flexible stalk domains help kinesin to dimerize, and the neck linker acts as a lever to facilitate steps [60]. The C‐terminal tail domain, in association with the light chain, forms the cargo‐binding domain of kinesin (Figure  3.2d). The kinesin tail domain forms complexes with adaptor proteins such as Miro/Milton, syntabulin, DENN/MADD, etc. [55]. These combinations achieve a high level of specificity in cargo recognition. Spatiotemporal regulation of cargo delivery is achieved by the phosphorylation of kinesin, Rab GTPase activity and Ca2+ signaling [55]. Most kinesins have a motor domain at the N‐terminal region (known as N‐kinesins) that drive cargos towards the plus end of the microtubule (anterograde motors). Others kinesins (known as C‐kinesins, e.g. kinesin‐14) have the motor domain at the C‐terminal, while a few kinesins have a cata­ lytic domain in the middle (M‐kinesin, e.g. kinesin‐13). C‐kinesin drives minus end‐directed motility of cargos, while M‐kinesin depolymerizes MTs [55, 61]. Similar to dynein, the net direction of kinesin‐driven cargo movement depends on the MT orientation within a particular cell type.

MICROTUBULE ORGANIZATION, POLARIZATION, AND CARGO TRAFFICKING IN HEPATOCYTES Like other epithelial cells, hepatocytes must exchange macro­ molecules between the external environment and their interior milieu. To achieve this, hepatocytes are polarized with distinct apical (or bile canalicular) and basolateral (or sinusoidal) domains segregated by tight junctions. The lateral surfaces of hepatocytes form cell–cell contacts while the basal surface inter­ acts with the extracellular matrix. Unlike typical epithelial cells, hepatocytes have a multipolar organization. Each hepatocyte shares a boundary with multiple narrow lumina, bile canaliculi, and basal domains of neighboring cells that face the endothelial lining. A hepatocyte sandwiched in this manner has to sustain two countercurrent flow systems: the synthesis/secretion of bile and uptake/secretion of blood components at the apical and basolateral surfaces, respectively [1]. The two surface domains exhibit distinct protein and lipid compositions [62], and suste­ nance of this complex polarity likely requires vesicular transport along MTs and actin inside hepatocytes. Novikoff et al. imaged the MT organization in cultured ­primary rat hepatocytes by immunofluorescence labeling of tubulin [63]. A “starburst” pattern, commonly identified as radially organized



3:  Cytoskeletal Motors: Structure and Function in Hepatocytes

31

Golgi

Bile canaliculi

sER LDs Kinesin-1

Sinusoid Microtubule Nucleus Cortical actin

Figure 3.3  Microtubule organization in hepatocytes. Microtubules extend from pericanalicular region towards the sinusoidal region. Lipid drop­ lets (LDs) are driven by conventional kinesin towards peripherally located smooth ER (sER) near the plus ends of microtubules (shown). Endocytic vesicles are driven dominantly by dynein and exocytic vesicles by kinesin motors (motors are not shown).

MTs (e.g. in melanocytes), was observed with one end of MTs focused at a centrally located centrosome, and the other ends emanating out like an umbrella to line the cortical region of the hepatocyte. In agreement with this geometry, the microtubule organizing center (MTOC) in polarized hepatocytes is localized adjacent to a perinuclear region and MTs extend from canalicu­ lar to sinusoidal region [64]. Cytoplasmic dynein is associated with, and may drive ligand‐containing endosomes towards the central centrosomal region of hepatocytes [51, 65]. A simplified view of the microtubule orientation in relation to the apical and basolateral membrane in hepatocytes is shown in Figure 3.3. As a consequence of this geometry, MT‐based motion towards the apical PM in hepatocytes may require a plus end‐directed motor (e.g. kinesin‐1). The MT orientation and directionality of motors is also relevant to lipid trafficking. For example, the smooth endoplasmic reticulum (sER) (where lipids are packaged for secretion) appears located towards the ­periphery of hepatocytes and may be supplied with lipids via kinesin‐driven transport of lipid droplets [66]. The liver is an organ of immense metabolic importance. It is involved in the production of bile acid, cholesterol, plasma pro­ tein, assembly/secretion of triglyceride‐laden very low‐density lipoprotein particles (VLDL), elimination of toxic substances, and processing of hormones and cytokines [1]. All these pro­ cesses involve directed intracellular transport between distinct subcellular compartments and membranes of a hepatocyte. The ER–Golgi network is the central pathway in protein sorting and endosomal targeting. Nascent proteins arising in the ER are transported to the trans‐Golgi network (TGN) via the Golgi cis­ ternae. The proteins sequestered at TGN are actively sorted and packaged into vesicles to be further transported to the apical membrane or to endosomal compartments [67]. MT motors are essential for the generation and transport of TGN‐derived ­transport carriers [68, 69]. Disruption of MTs results in the mis­ sorting of apical proteins [70]. A function‐blocking antibody against kinesin also inhibits exit of an apical reporter protein, NTRp75, from the TGN and leads to accumulation of small

vesicles around the Golgi [71]. Exposure of hepatocytes to col­ chicine reduces endosome–lysosome fusion, suggesting that dynein transports endocytic vesicles to the apical surface [72]. Fluorescently labeled early endocytic vesicles isolated from rat liver are driven along MTs by kinesin‐1 and KIFC2 motors [73]. In contrast to the early endocytic vesicles, dynein and KIF3A drive motion of late endosomes which do not exhibit significant fission [74].

MICROTUBULE MOTORS IN TRIGLYCERIDE SECRETION The liver can be thought of as the “lipid manager” of the body. Most of this lipid is stored away in intracellular organelles called lipid droplets (LDs). LDs are loaded with triglyceride (TG) molecules that consist of fatty acid chains esterified to glycerol. Our view that LDs are coalesced fat sitting idle in cells has changed drastically. Proteomic and imaging studies hinted that LDs interact with mitochondria, peroxisomes, and ER depending on the cell/tissue types and metabolic requirements [75–77]. This interaction of LDs with a variety of cellular com­ partments requires long‐distance transport of the LDs within cells. Live imaging reveals that LDs move in a directed manner in many cell types ranging from algae to hepatocytes [66, 78]. Cytoplasmic dynein and kinesin are physically associated with LDs to drive intracellular motion of LDs [66, 79]. The TG in LDs can be utilized to generate energy in h­ epatocytes, or can be secreted in the form of VLDL. The bulk of TG (~70%) in VLDL is derived from cytosolic LDs [80]. The earliest report suggesting that MTs are involved in lipoprotein secretion from the liver was published in 1973 [81]. Furthermore, Reaven et al. dem­ onstrated that depolymerization of microtubules drastically decreased VLDL lipidation in hepatocytes [82]. More recently, we found that LDs purified from rat liver show rapid plus end‐directed motion along MTs [83]. We also showed that kinesin‐1‐driven LD

32

THE LIVER:  CONCLUSIONS

transport supplies triglyceride to the smooth ER in hepatocytes for assembling VLDL, and that knockdown of kinesin‐1 in McA‐ RH7777 cells inhibits the characteristic peripheral localization of LDs [66, 84]. Intriguingly, kinesin‐1 is recruited to LDs in a com­ plex with ADP ribosylation factor 1 (ARF1), a small GTPase and a key regulator of lipolysis [66]. Because ARF1 induces ­membrane curvature, it is possible that the additional force from kinesin causes vesiculation from ARF1‐rich regions on the LD membrane. ARF1 and kinesin were activated on LDs in an insulin‐dependent manner, making this pathway responsive to the metabolic state of an animal. In the fed state kinesin‐1 was more active on LDs, transporting them to the peripherally localized sER in hepato­ cytes, where sER‐resident lipases could catabolize the TG to gen­ erate diacylglycerol [85]. Because it has two fatty acid chains, diacylglycerol may equilibrate across the sER membrane, where­ after it is reconverted into TG in the sER lumen and supplied to synthesize mature VLDL particles [86]. The catabolism of LD‐TG for VLDL lipidation is extremely efficient [80] and may require a synergy between kinesin, ARF1, and specific LD/sER‐associated lipids in the fed state. In con­ trast, in the fasted state, lowered insulin‐signaling reduces kine­ sin‐1 on LDs to limit sER–LD contact and tempers TG supply for VLDL lipidation [66]. Tuning kinesin‐dependent transport of LDs in hepatocytes may therefore enable the liver to protec­ tively sequester massive amounts of TG after fasting (fasting‐ induced steatosis), thus preventing lipotoxic effects of TG on peripheral tissues. The cellular mechanisms that control TG secretion from the liver, and therefore systemic TG homeostasis across daily metabolic cycles, can help understand TG imbal­ ance leading to lipodystrophies such as fatty liver and diabetes [87]. Indeed, we have seen significant changes in lipid/protein trafficking to LDs across feeding–fasting cycles [66, 88]. It is surprising that relatively little work has been done employing animal models and the simple feeding–fasting response. We also hope that disease‐relevant mutations in animal models can answer how motor proteins transport LDs within hepatocytes to control trafficking in response to metabolic states. MT‐dependent transport is also essential for hepatitis C virus (HCV) replication in host hepatocytes. HCV replicates its genome at the ER‐derived membranous web, and assembles new infectious particles using ER resident structural proteins [89]. Newly synthesized viral proteins (e.g. HCV‐core) must transfer from ER to LDs for HCV to propagate. This transfer and HCV assembly was diminished upon kinesin‐1 knockdown [66]. HCV‐core induces LD redistribution in an MT‐ and dynein‐ dependent manner [90]. HCV also cleaves the Rab‐interacting lysosomal protein to redirect Rab7‐containing vesicles from dynein‐ to kinesin‐dependent transport, and in this manner pro­ motes virion secretion [91]. MT‐dependent transport, therefore, impacts multiple aspects of the viral lifecycle in hepatocytes, and is a rich area for further scientific investigations.

of MTs within cells. The most common and evolutionarily con­ served modification on tubulin is acetylation. The ε‐amino group of lysine‐40 (K40) residue of α‐tubulin is predominantly acetylated with αTAT1 (α‐tubulin acetyltransferase‐1) [92, 93] and reversed by deacetylases: Sirt2 (sirtuin type 2) [94] or his­ tone deacetylase (HDAC6) [95]. The acetylation‐induced weak lateral interactions of protofilaments increase MT flexibility and make MTs resistant to mild cold, nocodazole, and colchicine exposure [96, 97]. Besides acetylation, the K40 site is also sus­ ceptible to trimethylation [98]. In mitotic cells, α‐tubulin of the central spindle is trimethylated at K40, and enriched at the plus ends of MTs at metaphase. Despite numerous reports on the phosphorylation of α‐ and β‐tubulin at serine, threonine, and tyrosine residues, the physiological significance of these modi­ fications remains to be understood. The unstructured C‐terminal tail (CTT) domains of α‐ and β‐tubulin are frequent targets of tyrosination, Δ2 modification (penultimate glutamate is also removed from detyrosinated tubulin) and detyrosination. α‐Tubulin undergoes tyrosination–detyrosination cycles via the activity of tubulin tyrosinase ligase and carboxypeptidase. An in vitro study suggests that PTM of tubulin regulates the MT depolymerization rate. Moreover, modifications in CTT of tubu­ lin are recognized by molecular motors that also govern the motor velocity and its processivity [99]. In vivo, detyrosinated MTs persist for hours, but tyrosinated MTs turn over in minutes [100]. The KIF2C kinesin that depolymerizes MTs has prefer­ ential activity for tyrosinated MTs both in vivo and in vitro. The binding of KIF5 kinesin to MTs is also regulated by multiple PTMs. The presence of tyrosinated tubulin also favors interac­ tion of the MTs with plus end‐tracking proteins (+TIPs) that have conserved CAP‐Gly domains [101]. Some tubulin modifi­ cation, such as glutamylation, succination, and O‐Glc‐N‐acylation have been recently identified, but their detailed physiological impact is as yet unexplored. Alcoholic liver may develop as a result of abnormal PTMs in hepatocytes. Acetylation of MTs in hepatocytes increases tubu­ lin stability [102]. Treatment of hepatoma‐derived WIF‐B cells with a drug that specifically depolymerizes dynamic MTs impairs MT‐dependent trafficking along three distinct cellular pathways [103, 104]. Tubulin is hyperacetylated in ethanol‐fed rats, which increases MT stability and might also alter intracel­ lular motility [105]. Acetylated MTs are required for adipogen­ esis, and this acetylation is controlled by AMPK‐mediated phosphorylation of αTAT1 [106]. AMPK, a master regulator of lipid homeostasis, also promotes relocation of LDs and mito­ chondria on detyrosinated MTs in VERO cells of nonhepatic origin to promote fatty acid oxidation [107]. Despite its poten­ tial physiological consequences, the effect of alcohol‐induced acetylation on LD dynamics in hepatocytes is not known.

CONCLUSIONS MICROTUBULE MODIFICATIONS AND LIVER PATHOLOGY MTs and their associated functions are tuned by the post‐trans­ lational modification (PTM) of tubulin. PTMs and various microtubule‐interacting proteins contribute to the heterogeneity

Motor‐dependent motion along actin and MTs sustains the intracellular organization of many cell types, and is particularly relevant to polarized epithelial cells such as hepatocytes. In  hepatocytes, the MTs are organized in a radial fashion ­originating from the centrosome and spreading out towards the peripheral sinusoidal regions [108, 109]. This arrangement of



3:  Cytoskeletal Motors: Structure and Function in Hepatocytes

MTs, along with the directionality of cytoplasmic dynein and kinesin motors, defines endo/exocytosis and protein trafficking in hepatocytes [63]. As an example, kinesin‐dependent lipid droplet transport maintains lipid homeostasis across metabolic transitions. Viruses (such as HCV) may also require MT‐ dependent transport for assembly in hepatocytes and secretion to the blood plasma in complex with VLDL particles [90, 110]. PTM of tubulin affects the dynamics and organization of MTs, and also motor‐driven transport on MTs [99]. PTMs regulate intracellular cargo transport and organelle–organelle contacts, and aberrant PTMs might hinder substrate delivery leading to physiological anomalies such as fatty liver disease. In conclu­ sion, the MT cytoskeleton in hepatocytes provides a rich and very disease‐relevant context for further investigations. Of par­ ticular interest in this context is the recent finding that MT motors control lipid flux from the liver across metabolic cycles [66]. How cargos are handed over between the actin and MT systems in hepatocytes is also poorly understood and warrants investigation.

REFERENCES 1. Treyer, A. and Müsch, A. Hepatocyte polarity. Compr Physiol, 2013;3(1):243–87. 2. Musch, A. Microtubule organization and function in epithelial cells. Traffic, 2004;5:1–9. 3. Feracci, H., Connolly, T.P., Margolis, R.N., and Hubbard, A.L. The estab­ lishment of hepatocyte cell surface polarity during fetal liver development. Dev Biol, 1987;123(1):73–84. 4. Ishii, M., Washioka, H., Tonosaki, A., and Toyota, T. Regional orientation of actin filaments in the pericanalicular cytoplasm of rat hepatocytes. Gastroenterology, 1991;101(6):1663–72. 5. Tzanakakis, E.S., Hansen, L.K., and Hu, W.S. The role of actin filaments and microtubules in hepatocyte spheroid self‐assembly. Cell Motil Cytoskeleton, 2001;48(3):175–89. 6. Gupta, A., Schell, M.J., Bhattacharjee, A., Lutsenko, S., and Hubbard, A.L. Myosin Vb mediates Cu+ export in polarized hepatocytes. J Cell Sci, 2016;129(6):1179–89. 7. Hales, C.M., Griner, R., Hobdy‐Henderson, K.C. et  al. Identification and Characterization of a Family of Rab11‐interacting Proteins. J Biol Chem, 2001;276(42):39067–75. 8. Lapierre, L.A., Kumar, R., Hales, C.M. et al. Myosin vb is associated with plasma membrane recycling systems. Mol Biol Cell, 2001;12(6):1843–57. 9. Müller, T., Hess, M.W., Schiefermeier, N. et  al. MY05B mutations cause microvillus inclusion disease and disrupt epithelial cell polarity. Nat Genet, 2008;40(10):1163–5. 10. Wakabayashi, Y., Dutt, P., Lippincott‐Schwartz, J., and Arias, I.M. Rab11a and myosin Vb are required for bile canalicular formation in WIF‐B9 cells. Proc Natl Acad Sci U S A, 2005;102(42):15087–92. 11. Ruemmele, F.M., Müller, T., Schiefermeier, N. et  al. Loss‐of‐function of MY05B is the main cause of microvillus inclusion disease: 15 Novel muta­ tions and a CaCo‐2 RNAi cell model. Hum Mutat, 2010;31(5):544–51. 12. Knowles, B.C., Roland, J.T., Krishnan, M. et al. Myosin Vb uncoupling from RAB8A and RAB11A elicits microvillus inclusion disease. J Clin Invest, 2014;124(7):2947–62. 13. Kravtsov, D., Mashukova, A., Forteza, R. et al. Myosin 5b loss of function leads to defects in polarized signaling: implication for microvillus inclusion disease pathogenesis and treatment. Am J Physiol Liver Physiol, 2014;307(10):G992–1001. 14. Girard, M., Lacaille, F., Verkarre, V. et al. MY05B and bile salt export pump contribute to cholestatic liver disorder in microvillous inclusion disease. Hepatology, 2014;60(1):301–10. 15. Strautnieks, S.S., Bull, L.N., Knisely, A.S. et al. A gene encoding a liver‐­ specific ABC transporter is mutated in progressive familial intrahepatic ­cholestasis. Nat Genet, 1998;20(3):233–8.

33

16. Klomp, L.W.J., Vargas, J.C., Van Mil, S.W.C. et al. Characterization of muta­ tions in ATP8B1 associated with hereditary cholestasis. Hepatology, 2004;40(1):27–38. 17. Nyasae, L.K., Schell, M.J., and Hubbard, A.L. Copper directs ATP7B to the  apical domain of hepatic cells via basolateral endosomes. Traffic, 2014;15(12):1344–65. 18. Tuma, P.L., Nyasae, L.K., and Hubbard, A.L. Nonpolarized cells selectively sort apical proteins from cell surface to a novel compartment, but lack apical retention mechanisms. Mol Biol Cell, 2002;13(10):3400–15. 19. Manton, I. and Clarke, B. An electron microscope study of the spermatozoid of sphagnum. J Exp Bot, 1952;3(3):265–75. 20. Fawcett, D.W. and Porter, K.R. A study of the fine structure of ciliated ­epithelia. J Morphol, 1954;94(2):221–81. 21. Sabitini, D.D., Bensch, K., and Barrnett, R.J. Cytochemistry and electron microscopy. The preservation of cellular ultrastructure and enzymatic ­activity by aldehyde fixation. J Cell Biol, 1963;17:19–58. 22. Ledbetter, M.C. and Porter, K.R. Morphology of microtubules of plant cell. Science, 1964;144(3620):872–4. 23. Borisy, G.G. and Taylor, E.W. The mechanism of action of colchicine. Binding of colchicine‐3H to cellular protein. J Cell Biol, 1967;34:525–33. 24. Bryan, J. and Wilson, L. Are cytoplasmic microtubules heteropolymers? Proc Natl Acad Sci U S A, 1971;68(8):1762–6. 25. Mohri, H. Amino‐acid composition of “tubulin” constituting microtubules of sperm flagella. Nature, 1968;217:1053–4. 26. Borisy, G.G., Olmsted, J.B., and Klugman. R.A. In vitro aggregation of cyto­ plasmic microtubule subunits. Proc Natl Acad Sci U S A, 1972;69(10):2890–4. 27. Oakley, B.R. γ‐Tubulin: the microtubule organizer? Trends Cell Biol, 1992;2(1):1–5. 28. Mitchison, T. and Kirschner, M. Dynamic instability of microtubule growth. Nature, 1984;312(5991):237–42. 29. Matov, A., Applegate, K., Kumar, P. et al. Analysis of microtubule dynamic instability using a plus‐end growth marker. Nat Methods, 2010;7(9):761–8. 30. Downing, K.H. and Nogales, E. Tubulin and microtubule structure. Curr Opin Cell Biol, 1998;10:16–22. 31. McNally, F.J. and Vale, R.D. Identification of katanin, an ATPase that severs and disassembles stable microtubules. Cell, 1993;75(3):419–29. 32. Dehmelt, L. and Halpain, S. The MAP2/Tau family of microtubule‐­ associated proteins. Genome Biol, 2005;6(1):204. 33. Gibbons, I.R. and Rowe, A.J. Dynein: a protein with adenosine ­triphosphatase activity from cilia. Science, 1965;149(3682):424–6. 34. Gibbons, B.H. and Gibbons, I.R. Flagellar movement and adenosine triphos­ phatase activity in sea urchin sperm extracted with triton X‐100. J Cell Biol, 1972;54(1):75–97. 35. Allen, R.D., Metuzals, J., Tasaki, I., Brady, S.T., and Gilbert, S.P. Fast axonal transport in squid giant axon. Science, 1982;218(4577):1127–9. 36. Brady, S.T., Lasek, R.J., and Allen, R.D. Fast axonal transport in extruded axoplasm from squid giant axon. Science, 1982;218(4577):1129–31. 37. Schroer, T.A., Schnapp, B.J., Reese, T.S., and Sheetz, M.P. The role of kine­ sin and other soluble factors in organelle movement along microtubules. J Cell Biol, 1988;107(5):1785–92. 38. Schnapp, B.J. and Reese, T.S. Dynein is the motor for retrograde axonal transport of organelles. Proc Natl Acad Sci U S A, 1989;86(5):1548–52. 39. Allan, V.J. Cytoplasmic dynein. Biochem Soc Trans, 2011;39(5):1169–78. 40. Gibbons, I.R., Lee‐Eiford, A., Mocz, G. et al. Photosensitized cleavage of dynein heavy chains. Cleavage at the “V1 site” by irradiation at 365 nm in the presence of ATP and vanadate. J Biol Chem, 1987;262(6):2780–6. 41. Reck‐Peterson, S.L. and Vale, R.D. Molecular dissection of the roles of nucleotide binding and hydrolysis in dynein’s AAA domains in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A, 2004;101(6):1491–5. 42. Kardon, J.R. and Vale, R.D. Regulators of the cytoplasmic dynein motor. Nat Rev Mol Cell Biol, 2009;10:854–65. 43. Vallee, R.B., Williams, J.C., Varma, D., and Barnhart, L.E. Dynein: an ancient motor protein involved in multiple modes of transport. J Neurobiol, 2004;58:189–200. 44. Johansson, M., Rocha, N., Zwart, W. et al. Activation of endosomal dynein motors by stepwise assembly of Rab7‐RILP‐p150Glued, ORP1L, and the receptor βIII spectrin. J Cell Biol, 2007;176(4):459–71. 45. Fu, M.M. and Holzbaur, E.L. Integrated regulation of motor‐driven organelle transport by scaffolding proteins. Trends Cell Biol, 2014;24:564–74. 46. Mallik, R., Carter, B.C., Lex, S.A., King, S.J., and Gross, S.P. Cytoplasmic dynein functions as a gear in response to load. Nature, 2004;427(6975): 649–52.

34

THE LIVER:  REFERENCES

47. Rai, A.K., Rai, A., Ramaiya, A.J., Jha, R., and Mallik, R. Molecular adapta­ tions allow dynein to generate large collective forces inside cells. Cell, 2013;152(1–2):172–82. 48. Anishkin, A. and Kung, C. Stiffened lipid platforms at molecular force foci. Proc Natl Acad Sci U S A, 2013;110(13):4886–92. 49. Rai, A., Pathak, D., Thakur, S. et al. Dynein clusters into lipid microdomains on phagosomes to drive rapid transport toward lysosomes. Cell, 2016;164(4):722–34. 50. Hirokawa, N. Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science, 1998;279:519–26. 51. Oda, H., Stockert, R.J., Collins, C. et  al. Interaction of the microtubule cytoskeleton with endocytic vesicles and cytoplasmic dynein in cultured rat hepatocytes. J Biol Chem, 1995;270(25):15242–9. 52. Wang, Y. Cytoplasmic dynein participates in apically targeted stimulated secretory traffic in primary rabbit lacrimal acinar epithelial cells. J Cell Sci, 2003;116(10):2051–65. 53. Vale, R.D., Reese, T.S., and Sheetz, M.P. Identification of a novel force‐ generating protein, kinesin, involved in microtubule‐based motility. Cell, 1985;42(1):39–50. 54. Brady, S.T., Lasek, R.J., and Allen, R.D. Fast axonal transport in extruded axoplasm from squid giant axon. Science 1982;218(4577):1129–31. 55. Hirokawa, N., Noda, Y., Tanaka, Y., and Niwa, S. Kinesin superfamily motor  proteins and intracellular transport. Nat Rev Mol Cell Biol, 2009;10:682–96. 56. Kuznetsov, S.A., Vaisberg, E.A., Shanina, N.A. et al. The quaternary struc­ ture of bovine brain kinesin. EMBO J, 1988;7(2):353–6. 57. Bloom, G.S., Wagner, M.C., Pfister, K.K., and Brady, S.T. Native structure and physical properties of bovine brain kinesin and identification of the ATP‐ binding subunit polypeptide. Biochemistry, 1988;27(9):3409–16. 58. Asenjo, A.B., Krohn, N., and Sosa H. Configuration of the two kinesin motor domains during ATP hydrolysis. Nat Struct Biol, 2003;10(10):836–42. 59. Yildiz, A., Tomishige, M., Vale, R.D., and Selvin, P.R. Kinesin walks hand‐ over‐hand. Science, 2004;303(5658):676 LP‐678. 60. Yun, M., Bronner, C.E., Park, C.G. et al. Rotation of the stalk/neck and one head in a new crystal structure of the kinesin motor protein, Ncd. EMBO J, 2003;22(20):5382–9. 61. Noda, Y., Sato‐Yoshitake, R., Kondo, S., Nangaku, M., and Hirokawa, N. KIF2 is a new microtubule‐based anterograde motor that transports membra­ nous organelles distinct from those carried by kinesin heavy chain or KIF3A/B. J Cell Biol, 1995;129(1):157–67. 62. Rosario, J., Sutherland, E., Zaccaro, L., and Simon, F.R. Ethinylestradiol administration selectively alters liver sinusoidal membrane lipid fluidity and protein composition. Biochemistry, 1988;27(11):3939–46. 63. Novikoff, P.M., Cammer, M., Tao, L. et al. Three‐dimensional organization of rat hepatocyte cytoskeleton: relation to the asialoglycoprotein endocytosis pathway. J Cell Sci, 1996;109(Pt 1):21–32. 64. Murray, J.W. and Wolkoff, A.W. Roles of the cytoskeleton and motor pro­ teins in endocytic sorting. Adv Drug Deliv Rev, 2003;55:1385–403. 65. Goltz, J.S., Wolkoff, A.W., Novikoff, P.M., Stockert, R.J., and Satir P. A role for microtubules in sorting endocytic vesicles in rat hepatocytes. Proc Natl Acad Sci U S A, 1992;89(15):7026–30. 66. Rai, P., Kumar, M., Sharma, G. et al. Kinesin‐dependent mechanism for con­ trolling triglyceride secretion from the liver. Proc Natl Acad Sci U S A, 2017;114(49):12958–63. 67. Allan, V.J., Thompson, H.M., and McNiven, M.A. Motoring around the Golgi. Nat Cell Biol, 2002;4:E236–42. 68. Van der Sluijs, P., Bennett, M.K., Antony, C., Simons, K., and Kreis, T.E. Binding of exocytic vesicles from MDCK cells to microtubules in vitro. J Cell Sci, 1990;95(Pt 4):545–53. 69. Fath, K.R., Trimbur, G.M., and Burgess, D.R. Molecular motors and a spec­ trin matrix associate with Golgi membranes in vitro. J Cell Biol, 1997;139(5): 1169–81. 70. Kreitzer, G., Schmoranzer, J., Low, S.H. et al. Three‐dimensional analysis of post‐Golgi carrier exocytosis in epithelial cells. Nat Cell Biol, 2003;5:126–36. 71. Kreitzer, G., Marmorstein, A., Okamoto, P., Vallee, R., and Rodriguez‐ Boulan, E. Kinesin and dynamin are required for post‐Golgi transport of a plasma‐membrane protein. Nat Cell Biol, 2000;2(2):125–7. 72. Berg, T., Kindberg, G.M., Ford, T., and Blomhoff, R. Intracellular transport of asialoglycoproteins in rat hepatocytes. Evidence for two subpopulations of lysosomes. Exp Cell Res, 1985;161(2):285–96.

73. Bananis, E., Murray, J.W., Stockert, R.J., Satir, P., and Wolkoff, A.W. Regulation of early endocytic vesicle motility and fission in a reconstituted system. J Cell Sci, 2003;116(Pt 13):2749–61. 74. Bananis, E., Nath, S., Gordon, K. et al. Microtubule‐dependent movement of late endocytic vesicles in vitro: requirements for dynein and kinesin. Mol Biol Cell, 2004;15(8):3688–97. 75. Brasaemle, D.L., Dolios, G., Shapiro, L., and Wang, R. Proteomic analysis of proteins associated with lipid droplets of basal and lipolytically stimulated 3T3‐L1 adipocytes. J Biol Chem, 2004;279(45):46835–42. 76. Yang, L., Ding, Y., Chen, Y. et al. The proteomics of lipid droplets: structure, dynamics, and functions of the organelle conserved from bacteria to humans. J Lipid Res, 2012;53(7):1245–53. 77. Gao, Q. and Goodman, J.M. The lipid droplet – a well‐connected organelle. Front Cell Dev Biol, 2015;3. 78. Guimaraes, S.C., Schuster, M., Bielska, E. et al. Peroxisomes, lipid droplets, and endoplasmic reticulum “hitchhike” on motile early endosomes. J Cell Biol, 2015;211(5):945–54. 79. Boström, P., Rutberg, M., Ericsson, J. et al. Cytosolic lipid droplets increase in size by microtubule‐dependent complex formation. Arterioscler Thromb Vasc Biol, 2005;25(9):1945–51. 80. Wiggins, D. and Gibbons, G.F. The lipolysis/esterification cycle of hepatic tria­ cylglycerol. Its role in the secretion of very‐low‐density lipoprotein and its response to hormones and sulphonylureas. Biochem J, 1992;284(Pt 2):457–62. 81. Orci, L., Marchand, Y.L.E., Singh, A., Assimacopoulos‐Jennnet, F., Rouiller, C., and Jeanrenaud, B. Role of microtubules in lipoprotein secretion by the liver. Nature, 1973;244:30. 82. Reaven, E.P. and Reaven, G.M. Evidence that microtubules play a permis­ sive role in hepatocyte very low density lipoprotein secretion. J Cell Biol, 1980;84(1):28–39. 83. Barak, P., Rai, A., Rai, P., and Mallik, R. Quantitative optical trapping on single organelles in cell extract. Nat Methods, 2013;10(1):68–70. 84. Schulze, R.J. and McNiven, M.A. Fasting inhibits the recruitment of kine­ sin‐1 to lipid droplets and stalls hepatic triglyceride secretion. Hepatology, 2019;69(1):444–6. 85. Gilham, D., Alam, M., Gao, W., Vance, D.E., and Lehner, R. Triacylglycerol hydrolase is localized to the endoplasmic reticulum by an unusual retrieval sequence where it participates in VLDL assembly without utilizing VLDL lipids as substrates. Mol Biol Cell, 2005;16(2):984–96. 86. Lankester, D.L., Brown A.M., and Zammit, V.A. Use of cytosolic triacylg­ lycerol hydrolysis products and of exogenous fatty acid for the synthesis of  triacylglycerol secreted by cultured rat hepatocytes. J Lipid Res, 1998;39(9):1889–95. 87. Anderwald, C., Bernroider, E., Krssak, M. et al. Effects of insulin treatment in type 2 diabetic patients on intracellular lipid content in liver and skeletal muscle. Diabetes, 2002;51(10):3025–32. 88. Sadh, K., Rai, P., and Mallik, R. Feeding‐fasting dependent recruitment of membrane microdomain proteins to lipid droplets purified from the liver. PLoS One, 2017;12(8). 89. Scheel, T.K.H. and Rice, C.M. Understanding the hepatitis C virus life cycle paves the way for highly effective therapies. Nat Med, 2013;19:837–49. 90. Boulant, S., Douglas, M.W., Moody, L. et al. Hepatitis C virus core protein induces lipid droplet redistribution in a microtubule‐ and dynein‐dependent manner. Traffic, 2008;9(8):1268–82. 91. Wozniak, A.L., Long, A., Jones‐Jamtgaard, K.N., and Weinman, S.A. Hepatitis C virus promotes virion secretion through cleavage of the Rab7 adaptor protein RILP. Proc Natl Acad Sci U S A, 2016;201607277. 92. LeDizet, M. and Piperno, G. Identification of an acetylation site of Chlamydomonas alpha‐tubulin. Proc Natl Acad Sci U S A, 1987;84(16):5720–4. 93. Shida, T., Cueva, J.G., Xu, Z., Goodman, M.B., and Nachury, M.V. The major alpha‐tubulin K40 acetyltransferase TAT1 promotes rapid ciliogenesis and efficient mechanosensation. Proc Natl Acad Sci U S A, 2010;107(50): 21517–22. 94. North, B.J., Marshall, B.L., Borra, M.T., Denu, J.M., and Verdin, E. The human Sir2 ortholog, SIRT2, is an NAD+‐dependent tubulin deacetylase. Mol Cell, 2003;11(2):437–44. 95. Hubbert, C., Guardiola, A., Shao, R. et  al. HDAC6 is a microtubule‐­ associated deacetylase. Nature, 2002;417(6887):455–8. 96. Xu, Z., Schaedel, L., Portran, D. et  al. Microtubules acquire resistance from  mechanical breakage through intralumenal acetylation. Science, 2017;356(6335):328–32.



3:  Cytoskeletal Motors: Structure and Function in Hepatocytes

97. LeDizet, M. and Piperno G. Cytoplasmic microtubules containing acety­ lated  ??‐tubulin in Chlamydomonas reinhardtii: spatial arrangement and properties. J Cell Biol, 1986;103(1):13–22. 98. Park, I.Y., Powell, R.T., Tripathi, D.N. et al. Dual chromatin and cytoskel­ etal remodeling by SETD2. Cell, 2016;166(4):950–62. 99. Sirajuddin, M., Rice, L.M., and Vale, R.D. Regulation of microtubule motors by tubulin isotypes and post‐translational modifications. Nat Cell Biol, 2014;16:335. 100. Kreitzer, G., Liao, G., and Gundersen, G.G. Detyrosination of tubulin regulates the interaction of intermediate filaments with microtubules in vivo via a kinesin‐dependent mechanism. Mol Biol Cell, 1999;10(April): 1105–18. 101. Peris, L., Thery, M., Fauré, J. et al. Tubulin tyrosination is a major factor affecting the recruitment of CAP‐Gly proteins at microtubule plus ends. J Cell Biol, 2006;174(6):839–49. 102. Groebner, J.L. and Tuma PL. The altered hepatic tubulin code in alcoholic liver disease. Biomolecules, 2015;5:2140–59. 103. Poüs, C., Chabin, K., Drechou, A. et al. Functional specialization of stable and dynamic microtubules in protein traffic in WIF‐B cells. J Cell Biol, 1998;142(1):153–65.

35

104. Phung‐Koskas, T., Pilon, A., Poüs, C. et  al. STAT5B‐mediated growth ­hormone signaling is organized by highly dynamic microtubules in hepatic cells. J Biol Chem, 2005;280(2):1123–31. 105. Kannarkat, G.T., Tuma, D.J., and Tuma, P.L. Microtubules are more stable and more highly acetylated in ethanol‐treated hepatic cells. J Hepatol, 2006;44(5):963–70. 106. Mackeh, R., Lorin, S., Ratier, A. et  al. Reactive oxygen species, amp‐ Activated protein kinase, and the transcription cofactor p300 regulate α‐ Tubulin acetyltransferase‐1 (αtat‐1/mec‐17)‐dependent microtubule hyperacetylation during cell stress. J Biol Chem, 2014;289(17):11816–28. 107. Herms, A., Bosch, M., Reddy, B.J.N. et al. AMPK activation promotes lipid droplet dispersion on detyrosinated microtubules to increase mitochondrial fatty acid oxidation. Nat Commun, 2015;6:7176. 108. Ihrke, G., Neufeld, E.B., Meads T. et al. WIF‐B cells: An in vitro model for studies of hepatocyte polarity. J Cell Biol, 1993;123(6 II):1761–75. 109. Meads, T. and Schroer, T.A. Polarity and nucleation of microtubules in polarized epithelial cells. Cell Motil Cytoskeleton, 1995;32(4):273–88. 110. Coller, K.E., Heaton, N.S., Berger, K.L. et al. Molecular determinants and dynamics of hepatitis C virus secretion. PLOS Pathog, 2012;8(1): e1002466.

4

Hepatocyte Surface Polarity Anne Müsch1 and Irwin M. Arias2 1 2

Department of Developmental & Molecular Biology, Albert Einstein College of Medicine, Bronx, NY, USA National Institutes of Health, Bethesda, MD, USA

INTRODUCTION With exception of erythrocytes and possibly a few others, all mammalian cells are polarized. Epithelial cell polarity, as present in the liver, gastrointestinal tract, and kidney, developed during the Precambrian Period and is essential for the formation of multicellular organs and species. Selective absorption and secretion requires cellular polarization. Hepatocyte polarization is neglected, unique, complex, and, as shown in recent studies, critical in many inheritable and acquired liver diseases. Surprisingly, only one current text in hepatology or pathology [1] mentions hepatocellular polarity. Neither the authors nor a large number of hepatologists or pathologists recall seeing a pathology report which comments on the polarization status of hepatocytes! A possible explanation is that hepatology developed in the late nineteenth century, when efforts were always made to associate pathology with clinical signs, symptoms, and outcome. The major tool in this effort was hematoxylin and eosin (H&E)‐stained sections of liver. The H&E stain does not reveal the small bile canaliculus, which is the apical polarized domain of the hepatocyte. The bile canaliculus is functionally sealed by tight junctions and, with its microvilli, constitutes ~13% of total hepatocyte plasma membrane [1]. Discovery of the bile canaliculus did not occur until transmission and scanning electron microscopy were introduced in the 1940s. Despite the availability of numerous bile canalicular proteins and associated antibodies, such as 5′ nucleotidase, cCAM 105, ABC B1, ABC B11, and others, they are rarely used in routine studies of liver histopathology. Probably because polarity is not described by pathologists, it is also neglected by clinicians. This chapter will consider mechanisms responsible for hepatocellular polarity. Canalicular network formation and hepatocyte polarity require coordinated expression of several key

elements, including extracellular matrix (ECM), adherens and tight junctions, intracellular protein trafficking machinery, cytoskeleton, and energy production. Inheritable and acquired defects in these elements that result in depolarization or failure to polarize are presented in Chapter 29.

THE UNIQUE HEPATOCYTE POLARITY PHENOTYPE The liver presents a remarkable example of how cell shape serves function. The two liver epithelial cell types – hepatocytes and bile duct cells  –  adopt radically different polarity phenotypes that serve their distinct physiological roles. Bile duct cells, which form simple conduits for bile, organize like other tubule‐ forming epithelial cells with a single luminal domain perpendicular to cell–cell adhesion domains and opposite a basal surface; they are monopolar. Hepatocytes, which mediate extensive bidirectional molecular exchange with the blood, maximize contact with blood vessels by establishing a second basal surface in place of the biliary cell’s apical surface. The hepatocyte apical domains assemble instead at cell–cell contact sites, interrupting cell–cell adhesion domains (Figure  4.1). Each ­ hepatocyte forms two or more lumina with several of its neighbors, yielding the branched luminal network known as bile canaliculi. Hepatocytes are therefore multipolar. While the ­ monopolar organization of bile duct cells is similar to that of most epithelia, the hepatocyte polarity phenotype is unique. Immunohistochemistry, electron microscopy, and gene expression analysis collectively indicate that hepatocytes form functional cell–cell adherens and tight junctions, and that they express and localize the constituents of evolutionary conserved

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



4:  Hepatocyte Surface Polarity

(a)

37

Basal

Hepatocytes

Lateral Luminal/apical

Bile ducts

(b)

(A)

(B)

HA 321

HA4

10

(C)

(D)

Sinusoidal front BC

Golgi complex

BC

High concentrations of endo/exocytotic organelles

BC

BC

Figure 4.1  (a) The two polarity phenotypes in the liver – hepatocytes and bile duct epithelia – have three distinct membrane domains: basal domains facing the extracellular matrix and underlying blood vessels; lateral domains engaged in cell–cell adhesion; and luminal domains, which face the outside world (they are situated at the cell apex in monopolar cells and hence also called apical domains). Hepatocytes have multiple luminal surfaces that interrupt cell–cell adhesion domains and most feature two basal surfaces facing the space of Disse on either side. Bile duct epithelial cells each have a single luminal and basal domain flanked by lateral surfaces. Adapted from Müsch, Curr Opin Cell Biol, 2018;54:18–23 with permission of Elsevier. (b) The architecture of the liver is unique. A: A scanning electron micrograph of a portion of a liver lobule is shown. A continuous network of bile canaliculi runs along the exposed cell surfaces of the liver plate. B: The two distinct PM domains are visualized by immunofluorescence detection of the basolateral PM protein, HA321/BEN and the apical PM protein, HA4/cell‐CAM105/ectoATPase. An electron micrograph of a hepatocyte (C) and a corresponding schematic drawing (D) highlight the “active zones” in vesicular trafficking. The major sorting organelles (TGN and endosomes) and transport vesicles are concentrated in small “clear” zones that are probably the most active in vesicle traffic. These zones are located between the Golgi and the apical PM and near the basolateral PM (the shaded regions in D). Reproduced from Braiterman and Hubbard, The Liver, 5th edn, Fig. 6.1.

38

THE LIVER:  EXPERIMENTAL SYSTEMS FOR THE STUDY OF EPITHELIAL POLARIZATION IN THE LIVER

polarity complexes, which establish epithelial surface identities, in a similar manner to that of monopolar epithelial cells. These observations make it likely that the distinct polarity phenotypes are caused by different regulation of the same key polarization mechanisms. Identification of regulatory mechanisms that define the two polarity phenotypes is only beginning to emerge. For a comprehensive listing of polarity‐associated proteins, their functions, and expression in hepatocytes, see Braiterman and Hubbard, The Liver, 5th edn, Table 6.2.

ESTABLISHMENT OF CONNECTED BILE DUCT AND BILE CANALICULAR NETWORKS DURING LIVER DEVELOPMENT The two liver epithelial cells originate from common precursors, called hepatoblasts. Hepatoblasts delaminate from a monolayered epithelial tube, the foregut, proliferate and invade the surrounding mesenchyme [2, 3] (see Chapter  2 for more details). Hepatoblasts around the portal vein first form a monolayered epithelium, the ductal plate  [4]. Two prominent features distinguish this from the adjacent hepatoblasts: a ­ laminin‐positive basement membrane and strong E‐cadherin labeling at cell–cell contact sites. Parts of the ductal plate develop into bile ducts, while the remnants have been proposed to become hepatic stem cells [5, 6]. Bile ducts develop when the ductal plate cells signal adjacent hepatoblasts to adopt a similar biliary fate. The two cell layers establish a basal lamina encircling them, and then generate a lumen between them. Hepatoblasts outside the vicinity of portal veins develop into hepatocytes. Hepatocytes acquire polarity (i.e. form bile canaliculi) only in late gestation, and the bile canalicular network continues to elongate postnatally [7, 8]. The hepatocyte polarization spurt coincides with the onset of bile acid synthesis [8–10]. Remarkably, the major bile acid taurocholate stimulates hepatocyte polarization in culture, suggesting that taurocholate serves as hepatocyte polarization trigger [11]. For bile excreted into hepatocyte bile canaliculi to reach the bile duct, the two epithelial tissues must link their respective luminal networks into a contiguous entity. This is accomplished when major bile ducts branch into smaller ductules, which connect to the bile canaliculi. Techniques that combined thick section immunofluorescence on cleared tissue with carbon ink injections into the common bile duct revealed that branching of smaller ductules from the larger ducts happened in parallel with the spurt in bile canaliculi formation and with the first appearance of bile in the intestine at E18 in the mouse [12]. Pharmacological inhibition of the hepatocyte bile acid transporter Mrp2, which pumps bile acids into the bile canaliculi, prevented bile duct branching. Bile acids that reach bile ducts from these newly formed bile canaliculi might thus constitute signals for ductal branching into narrower ductules, which then connect to the bile canaliculi. Taurocholate signaling for bile canaliculi formation involves activation of the kinase liver kinase B1 (LKB1) and its effector AMP‐regulated kinase (AMPK) [11], a signaling cascade that has also been linked to

branching morphogenesis in the lung [13]. Future work is expected to determine whether a taurocholate AMPK signaling cascades also governs bile duct branching.

EXPERIMENTAL SYSTEMS FOR THE STUDY OF EPITHELIAL POLARIZATION IN THE LIVER Due to lack of a readily available stable culture system for primary hepatocytes, much of what has been learned about hepatocyte polarization comes from polarized hepatocytic cell lines, in particular from Hep G2 and WIF‐B cells, which are derived from cancers [14–17]. Both form spherical lumina between neighboring cells but do not develop a canalicular network. The recently established Can10 line, which, like WIF‐B, is derived from a rat hepatoma, is the first hepatocytic cell line in which individual lumina connect to form canaliculi; however, it has not yet been extensively characterized [18]. HepG2 and WIF‐B cells recapitulate some critical hallmarks of the hepatocytic polarity phenotype observed in primary cultures or in vivo, such as the gradual, nonsynchronous formation of luminal surfaces at cell–cell contact sites, which contrast with the rapid and synchronous polarization of monopolar epithelial cells in culture, the asymmetric cell divisions subsequently discussed, and the indirect targeting of apical single membrane‐spanning and GPI‐ anchored proteins. HepG2 polarization is sensitive to oncostatin M, a cytokine critical for hepatocyte differentiation in vivo [19]; however, neither cell line responds to taurocholic acid, the presumptive in vivo trigger of bile canaliculi formation. This is a serious limitation to the study of signaling mechanisms leading to hepatocyte lumen formation. Also, the mechanism and regulation of ABC transporter trafficking in WIF‐B and HepG2 cells differ in several respects from that observed in hepatocytes [20]. Primary hepatocytes, isolated by liver perfusion with collagenase, re‐establish polarity and the dense canalicular network found in intact liver when cultured in collagen sandwiches. These cultures not only repolarize in a predictable manner, but they also maintain function and gene expression profiles of mature hepatocytes for up to two weeks [21–24]. This culture system, although less amenable to experimental manipulation than cell lines, provides information directly related to hepatocellular processes as confirmed by subsequent gene knockout studies subsequently described [25]. In addition to models for hepatocyte polarity, a few cell lines derived from normal biliary cells have also been established [26, 27]. These lines polarize with monopolar organization in 2D culture on permeable filter substrates where they acquire high trans‐epithelial resistance, a measure of epithelial barrier function [28]. The Normal Rat Cholangiocyte line (NRC) developed by LaRusso’s group [27] develops a primary cilium, which is a sign of biliary differentiation [29], and recapitulates the in vivo regulation and the cell surface appearance of biliary water and ion channels [30–32]. In development and during recovery from injury, hepatocytes and biliary cells evolve from common progenitors. These include stem cells and transit‐amplifying cells, called hepatoblasts [33]. The liver stem cell reservoir is believed to reside at

4:  Hepatocyte Surface Polarity

POLARIZATION MECHANISMS Polarity complexes Cell surface polarity is established when signals generated by local cues or by stochastic fluctuation become amplified through feedback mechanisms to yield a robust segregation of distinct membrane domains. Work in invertebrates combined with theoretical modeling identified a set of core signaling mechanisms that could generate cell‐autonomous polarity. They include several conserved epithelial signaling complexes operating through a combination of positive feedback and mutual antagonistic signaling to stake out distinct surface domains (reviewed in [49–51] and Figure 4.2). The transmembrane protein Crumbs, in conjunction with a Cdc42–aPKC–Par6–Par3 signaling network generates apical polarity in part by clustering and stabilizing Crumbs; a Dlg–Lgl–Scribble network promotes basolateral surface identity by preventing Crumbs accumulation and by promoting the accumulation of lateral membrane proteins, including E‐cadherin. The third complex centered around PALs/ Patj and the aPKC–Par3 proteins operates at the intersection

Crb ls

P

Pa

Cdc42

Par6

Patj

aPKC

TJ P

P

Par1

P Par3

Lgl

TJ

Crb Cdc42 Par6

T

the canal of Hering (the junction between bile ductules and hepatocytes) and in peribiliary glands and the gallbladder and is present throughout life [5, 34]. Hepatoblasts, on the other hand, are abundant during liver development but in the adult only detectable in response to liver injury. Both types of progenitors have been isolated and expanded in culture in attempts to differentiate them into the two epithelial lineages [7, 35–37]. Few of these studies have convincingly demonstrated hepatocytic or biliary polarity. In those instances where proper polarization was achieved, the progenitors gave rise to only one but not both lineages [38, 39]. Thus, in vitro recapitulation of the branching of the two liver epithelial cell types still eludes the field. A patented human cancer cell line with hepatoblast characteristics, HepaRG [40], can be polarized with either hepatocyte polarity when cultured in spheroids [41], or as monolayered hollow cysts (i.e. with bile duct polarity) in 3D matrigel culture [42]. Because of its human origin and relatively high degree of functional differentiation when cultured in spheroids, these cells have become popular for toxicology studies, although they have yet to be exploited for polarity studies. Two technical advances have opened the door to elucidating polarity mechanisms even in the intact liver. The first pertains to our ability to express or ablate proteins in hepatocytes by tail vein injection of adenoviruses [43, 44] or of even naked DNA via hydrodynamic injections [45, 46]. This “poor man’s genetics” approach allows acute in vivo manipulation of hepatocytes with relative ease when compared to the generation of knockout and transgenic animals. The second advance is in intravital imaging, which now allows live cell imaging of cellular processes in liver lobes [47]. Such analysis has recently shown that hepatocytes lacking the kinase LKB1 (discussed in more detail below) have leaky tight junctions [48]. These novel techniques add to the traditional biochemical, histochemical, and electron microscopic studies of the liver and complement dynamic live cell imaging performed in cell lines.

39

Lgl Scrib Dlg

P

Par3

T



Par1

Figure 4.2  Polarity complexes organize two distinct membrane domains (apical and basolateral) through self‐recruitment, positive feedback, and mutual inhibition. The transmembrane protein Crumbs (Crb) engages in lateral homotypic interactions via its extracellular domain and becomes phosphorylated by aPKC at its cytoplasmic domain. Both events stabilize the protein, which determines the apical domain. Crb interacts with the Pals/Patj complex, which links the apical domain to tight junctions. aPKC, which is activated by scaffolding with active Cdc42 and Par6, also directs its substrate Par3 to tight junctions and prevents basolateral surface determinants Lgl and Par1 from attaching to the apical domain. Conversely, Lgl, which genetically interacts with Scribble and Dlg, promotes Crb endocytosis from the basolateral domain and inhibits the Cdc42/Par6 complex to operate at the lateral membrane. Par1‐mediated phosphorylation of Par3 removes Par3 from the lateral domain.

between the apical and lateral complexes and in mammalian cells is critical for the formation of tight junctions. Available evidence suggests that the Par, Crumbs, and Scribble complexes are present in both monopolar epithelia and in hepatocytes. Thus, in the adult liver, immunohistochemical analysis has determined that aPKC iota and zeta and Par3 are colocalized with tight junction markers, specifically ZO‐1, occludin, and claudin‐3 at the boundary between bile canaliculi and sinusoidal membranes, suggesting that they function in apical junctional complexes as in other epithelial cells [52]. The basolateral polarity determinants Scribble and Lgl‐2 have been localized to the sinusoidal domains in the polarized hepatocytic WIF‐B cells and Scribble has been shown to mediate similar protein–protein interactions in the hepatocytic cell line WIF‐B and in monopolar (kidney‐derived) MDCK cells [53].

Cell matrix adhesion In the context of the tissue, these “polarity complexes” are likely positioned by external cues, which include diffusible signals and signals originating from cell–ECM and cell–cell adhesion. Thus, during bile duct formation signals from the portal mesenchyme mediate the deposition of a basement membrane between mesenchymal and ductal cells [54, 55]. The basement

40

THE LIVER:  POLARIZATION MECHANISMS

membrane is a two‐dimensional sheet of matrix molecules secreted by epithelial cells and the underlying connective tissue. It is organized into sheets primarily by laminin and fibrous collagen IV networks that are connected with each other via nidogen and the proteoglycan perlecan. It captures growth factors and cytokines and activates cell surface receptors, mostly integrins. ECM–cell signaling triggered by the basement membrane in the ductal cells cues apical surface formation on the membrane domain opposite the basement membrane. Consistent with a role for laminin in bile duct formation, 3D polarization of hepatoblast‐derived biliary cells into a hollow monolayered cyst in vitro required culture in laminin‐rich matrices and was dependent on the laminin receptor β1‐integrin [38, 56]. Remarkably, hepatocytes differ from other epithelial cells in that they do not assemble a basal lamina [57]. While collagen IV and laminin expression is stimulated during hepatoblast differentiation along the biliary lineage, expression of these ECM proteins is not stimulated during hepatocyte differentiation [58]. Mature hepatocytes completely lack expression of the critical basal lamina constituents laminin and nidogen [59]. Whether this unique epithelial feature contributes to the unique hepatocyte polarity phenotype has not been directly tested in vivo but several pieces of circumstantial evidence support such a scenario: (1) Fibrosis, which results in extensive matrix deposition between hepatocytes and endothelial cells results in loss of hepatocyte polarity [60, 61]; (2) lack of ECM deposition proved critical for hepatocytic polarity in an experimental model in which monopolar and hepatocytic polarity can be switched. This model relies on the kidney‐derived cell line MDCK, which exhibits monopolar organization but switches to hepatocytic organization when induced to overexpress the kinase Par1b [62]. Hepatocytic polarization upon Par1b overexpression coincided with reduced ECM deposition and focal adhesion; importantly, this hepatocytic polarity phenotype was reversed when the cells were plated on a collagen IV matrix [63]. Further investigation revealed reduced RhoA activity downstream of the lack of ECM signaling as the critical polarity cue: RhoA depletion was sufficient to induce hepatocytic polarity in MDCK cells, while pharmacological Rho activation in the hepatocytic cell line WIF‐B promoted their monopolar organization [64, 65]. These findings point to RhoA signaling downstream of cell adhesion signaling as a putative key regulator of the polarity phenotypes.

junctions are not just organizers of cell structure but serve as signaling platforms regulating gene expression, differentiation, proliferation, and morphogenesis. It is this signaling role that makes them well suited to translate an external cue (i.e. cell–cell contact) into a polarization sequence. Consistent with this idea, hepatocyte luminal domains form at sites of cell–cell contact. Similarly, in the experimental model of a single nonpolarized MDCK cell, which first establishes polarity after cell division yields two daughters, the new cell–cell contact site serves as patch for the establishment of a lumen between the two cells (Figure 4.3). Lumen formation in this experimental system ensues when endocytosed apical proteins, which constitutively cycle between endosomes and the entire (nonpolarized) plasma membrane in single cells, are specifically targeted to and accumulate at the new cell–cell contact site. In this manner, apical proteins are progressively cleared from the plasma membrane outside the cell–cell contact region and are transported to the nascent luminal domain [66, 67]. This constitutes a pathway called basolateral‐to‐apical transcytosis. Once MDCK cells have established their first luminal surface, however, new cell–cell contacts generated after additional cell divisions no longer ­support de novo lumen formation, resulting in the characteristic monopolar organization with a single luminal domain. Hepatocytes differ from monopolar cells in that each cell–cell contact may trigger a new lumen. We hypothesize that two principal mechanisms are responsible for this difference: the nature of cell–cell adhesion signaling and apical ­protein trafficking. (a)

Bas Bas

(b)

Monopolar

Bas

Lat

Bas

Ap

Ap

Multipolar

Cell–cell junctions In polarized epithelia the contacting membranes form three types of intercellular junctions: tight junctions, anchoring junctions, and gap junctions. Tight junctions provide a barrier for paracellar flow of macromolecules and solutes, thus enforcing their vectorial transport across epithelial cells; they also restrict the diffusion of membrane proteins between the apical and basolateral domains, thereby maintaining surface polarity. Anchoring junctions, which include adherens junctions and ­desmosomes, couple cytoskeletal elements to the plasma membrane, providing mechanical integrity and allowing mechanical coupling of cells. Gap junctions mediate cell–cell communication by permitting the passage of small molecular weight solutes (up to 1 kDa) directly between neighboring cells. Intercellular

Figure 4.3  Model for lumen formation in monopolar (ductal cells) versus multipolar (hepatocytes) epithelial cells. (a) Both monopolar and hepatocytic cells initiate de novo lumen formation at cell–cell adhesion domains. In monopolar MDCK cell doublets embedded in a 3D matrix the cell–cell contact domain serves as targeting patch for recycling apical membrane cargo, which prior to cell–cell contact formation recycles in a nonpolarized manner. When the apical targeting patch is sealed from the surrounding lateral membrane by tight junctions, water transport and the resulting turgor can inflate a lumen. This mechanism is also hypothesized for hepatocytes, which similarly initiate lumina at cell– cell contact sites. (b) Monopolar and multipolar epithelia differ in the formation of secondary luminal domains: Hepatocytes can form lumina at each of their cell–cell contact sites. Monopolar epithelia, by contrast, do not form additional lumina at new cell–cell contact sites. Instead, cell divisions enlarge the existing lumen.



4:  Hepatocyte Surface Polarity

As determined in MDCK cells, basolateral‐to‐apical transcytosis is the mechanism by which apical domains form de novo when new cell–cell junctions become a target site for apical‐ directed cargo. Basolateral‐to‐apical transcytosis is also the mechanism by which newly synthesized single membrane‐ spanning and glycosylphosphatidylinositol (GPI)‐anchored apical proteins are targeted to the luminal domain in hepatocytes [68]. By contrast, MDCK cells and other monopolar epithelia differ from hepatocytes in that they predominantly target newly synthesized apical proteins directly to the apical surface [69]. Taken together, these two observations suggest that transcytotic apical targeting in hepatocytes ensures that there is an abundant supply of apical proteins to create new apical lumina at each new hepatocyte contact domain. In contrast, direct apical targeting to an existing apical domain in monopolar epithelia, combined with the inherent low rate of endocytic recycling at the apical surface [70, 71] means apical cargo is unavailable for endocytic targeting to new cell–cell contact sites. This could prevent formation of new apical domains at cell–cell junctions in monopolar epithelia such as bile ducts. Cell–cell adhesion is initiated by nectins, a family of four IgG‐like, Ca2+‐independent adhesion molecules, which activate GTP exchange factors for the small GTPases Cdc42, Rac1, and Rap1. Signaling by these GTPases then promotes establishment of Ca2+‐dependent adhesion by cadherins, predominantly E‐cadherin, which in mature adherence junctions clusters in a zonula adherens adjacent to tight junctions and is linked to a circumferential actin belt [72, 73]. Hepatocytes also express N‐ cadherin [74, 75], which belongs to the same class as E‐cadherin but in other epithelial cells replaces E‐cadherin during epithelial‐to‐mesenchymal transitions (EMT) [76]. How N‐cadherin‐ expressing hepatocytes avoid the fate of EMT is not known. Junctional adhesion molecules (JAMs), of which JAM‐A is the best studied, are, together with claudins and occludin, the adhesion molecules of tight junctions. They initially mingle with nectin and E‐cadherin at cell–cell contacts in immature junctions prior to their segregation into separate adherens and tight junctions [77, 78]. Of these TJ proteins, JAM‐A is absolutely essential for lumen formation in both hepatocytic (WIF‐B) [79] and monopolar (MDCK) cell lines [80], whereas claudin composition appears to contribute to the decision to polarize with monopolar or heptocytic polarity. Claudins are the multi‐­ membrane‐spanning proteins whose extracellular loops are thought to establish the barrier for paracellular flow. The cell type‐specific combination in which its 23 members are expressed determines the “tightness” of the permeability barrier as well as its ion selectivity. Curiously, siRNA‐mediated depletion of claudin‐2 from WIF‐B cells [81] and of claudin‐3 from the related Can10 cells [82] resulted in a switch from hepatocytic to monopolar organization, suggesting specific signaling roles for claudins beyond regulating paracellular flux. E‐cadherin is considered a key trigger of epithelial polarization in monopolar cells [83]. This was established in so‐called “Ca‐switch assays,” which exploit the Ca‐dependence of E‐cadherin adhesion. In Ca‐free medium monopolar epithelial cells such as MDCK cells do not polarize even when plated at confluence (i.e. in the presence of only nectin‐mediated cell–cell adhesion). Re‐addition of Ca triggers rapid, synchronized tight junction establishment and lumen formation [84, 85]. It was unexpected,

41

therefore, that a hepatocytic cell line (HepG2) which failed to target E‐cadherin and β‐catenin to the cell surface still established functional tight junctions and bile canaliculi‐like luminal domains (albeit with delayed kinetics), suggesting that E‐cadherin is not absolutely essential for the establishment of hepatic polarity [86]. Conversely, increasing E‐cadherin levels in this cell line led to the formation of a horizontal tight junction belt characteristic of monopolar cells [87]. In MDCK cells, substitution of E‐cadherin for an adhesion‐deficient mutant promoted lumen establishment at cell–cell contact sites as in hepatocytes, at least transiently [88]. Collectively, these data suggest a model by which E‐cadherin‐ mediated adhesion signaling promotes monopolar organization and antagonizes hepatocytic polarity. Consistent with this model, immunohistochemistry of liver tissue sections shows significantly higher E‐cadherin staining of bile ducts when compared with adjacent hepatocytes [4]. This might be due to lower E‐cadherin mRNA levels in hepatocytes, as suggested by recent RNA sequencing data [89]. Even if lumen formation in hepatocyte does not require E‐cadherin adhesion, it likely depends on other cell–cell adhesion molecules. This can be concluded from cell culture studies in which mechanical cell compaction [90] or spheroid formation [91], both conditions that maximize cell–cell contacts, ­significantly stimulated lumen formation.

The taurocholate signaling mechanism Analysis of taurocholate signaling, the putative hepatocyte polarization trigger, in primary cultures established that taurocholate increased hepatocyte cAMP levels, which led to activation of EPAC, a GEF for small GTPases of the Rap family. Rap signaling in turn stimulated activity of AMP‐regulated kinase (AMPK) via its main activating kinase in the liver, LKB1 [11] (see Chapter 38 for further details). LKB1 was discovered as a tumor suppressor mutated in the human Peutz–Jegher syndrome, which is characterized by hamartomas and polyps in the gastrointestinal tract and by an elevated cancer risk. LKB1 activity depends on association with a pseudokinase and a scaffolding protein [92]. In some tissues, various growth factors activate LKB1, suggesting that hepatocyte polarity and other functions may be regulated by hormones and/or growth factors. Notably, in single intestinal cells LKB1 activation induces cell‐autonomous cell surface polarization, hinting that LKB1 might initiate signaling by the self‐organizing polarity complexes mentioned above [93]. In addition, in MDCK cells AMPK was required for tight junction formation downstream of cell–cell adhesion [94, 95]. Tight junctional proteins appear to be the main AMPK targets in its role as a polarity kinase. They include the substrate Gα‐interacting vesicle‐associated protein, (aka Girdin), which when phosphorylated by AMPK re‐enforces tight junctions under energy stress [96], and cingulin [97], the suggested AMPK substrate responsible for leaky tight junctions observed by intravital imaging in LKB1 knockout livers in vivo [48]. In addition to tight junction defects, LKB1 knockout livers also presented with defective targeting of bile acid transporters to the luminal surface, which were corrected by cAMP activation [25]. As can be expected from the polarity defects, LKB1 conditional liver knockout mice suffered cholestasis and liver injury [98].

42

THE LIVER:  POLARIZED PROTEIN TRAFFICKING

POLARIZATION REQUIRES ENERGY Hepatocyte polarization and all of its components are energy‐ dependent and have been increasingly related to two protein kinase, LKB1 and AMPK (see Chapter 38 for further details). AMPK and LKB1 contribute to hepatocyte polarization by controlling ATP synthesis and energy metabolism. AMPK, a serine threonine kinase containing a catalytic α subunit and regulatory β and γ subunits, controls energy metabolism within cells by sensing the cellular AMP to ATP ratio. Activation of AMPK by phosphorylation of the α subunit Thr172 decreases energy consumption and increases energy production during cellular stress, such as hypoxia, glucose deprivation, and ischemia, and has an important role in hepatic metabolism through effects on glucose, lipid and protein homeostasis and mitochondrial biogenesis. Long‐term effects involve regulation of the glycolytic and lipogenic pathways. In collagen sandwich cultured hepatocytes, the stress of isolation resulted in depolarization, ATP depletion, and mitochondrial fragmentation [99]. Mitochondrial fusion occurred within 2 days associated with increased ATP synthesis from oxidative phosphorylation and canalicular network formation. Subsequent AMPK activation upregulated glucose uptake, glycolysis, and a further increase in ATP. These studies reveal that, after stress, hepatocytes preferentially restore polarity even at low ATP levels, suggesting that polarity is a prime requirement for cellular activity [100].

HEPATOCYTE POLARITY IS LINKED TO THE EXPRESSION OF DIFFERENTIATION MARKERS When hepatocytes are cultured on rigid 2D matrices they rapidly lose expression of differentiation markers such as albumin or various transporters. The degree of de‐differentiation is proportional to the ECM concentration regardless of its nature. Increasing matrix concentrations also cause increasing cell spreading which is incompatible with hepatocyte polarization [101–103]. Conversely, hepatocytes cultured on soft matrices, particularly in sandwich configuration, maintain expression of differentiation markers; they also acquire polarity [21, 22]. This correlation between increased expression of markers for hepatocyte function and better polarization was also observed when hepatocytes or hepatic cell lines such as HepaRG were cultured in the absence of exogenous matrix but with extensive cell–cell contact in spheroids [91]. Its crux is cytoskeletal ­tension, which activates a mechano‐transduction cascade in which integrin signaling leads to FAK→RhoA/RhoK→ERK1/2 activation, which promotes proliferation and induces an EMT [104, 105]. Importantly, activation of RhoA/RhoK in this pathway inhibits expression of the hepatocytic master transcription factor HNF4a, which controls a hepatocyte‐specific transcriptional network that includes both functional proteins and polarity determinants [106]. Forced HNF4a expression on rigid matrix remedied loss of several, although not all, functional markers. As previously discussed, low RhoA activity is also a

prerequisite for heptocytic polarization in cell culture models. Thus, a RhoA–HNF4a axis might link hepatocyte polarity to functionality.

POLARIZED PROTEIN TRAFFICKING The vectorial activities of epithelia require that each transporter/ channel has a clearly defined apical–basal polarity, which results from sorting events in the biosynthetic and recycling routes at the trans‐Golgi network (TGN), common recycling endosomes (CRE), and apical recycling endosomes (ARE) [107] (Figures 4.4 and 4.5). Biosynthetic protein trafficking itineraries in hepatocytes have been established by combining an in vivo pulse‐chase protocol with cell fractionation [108]. This protocol, which exploits the observation that 35S‐methionine, when injected into the tail vein first reaches the liver and is mostly incorporated into newly synthesized hepatocyte proteins, is in fact the only in vivo approach to protein trafficking for any mammalian epithelium. It reveals that hepatocytes target polytopic membrane proteins, such as ABC transporters, from the TGN directly to the bile canalicular domain, some of which first accumulate at a Rab11a‐positive ARE pool, from which they then cycle to the canalicular membrane [109]. In contrast, all single membrane‐spanning and GPI‐anchored bile canalicular membrane proteins are targeted from the TGN to the basolateral plasma membrane, from where they transcytose in endosomes through the cell to the canalicular membrane [108, 110, 111]. This is different from all monopolar epithelia studied to date, which utilize the basolateral‐to‐apical transcytotic route for apical targeting to a lesser extent than hepatocytes. The most poignant difference is that unlike monopolar epithelia, hepatocytes lack a pathway that targets soluble proteins from the TGN to the bile canalicular domain via vesicular transport. All soluble cargo that passes through the secretory pathway is (predominantly) directed into the space of Disse [112, 113]. The mechanisms that segregate apical and basolateral proteins and mediate their surface domain‐specific targeting from the TGN have been largely elucidated in the epithelial model cell line MDCK. Work over three decades by many groups using apical and basolateral model proteins yielded the following model of polarized protein delivery in the direct biosynthetic pathway (Figure 4.4): Sorting machinery at the TGN recognizes apical and basolateral sorting signals and packages apical and basolateral proteins into separate TGN‐derived transport carriers [107]. Sorting signals in basolateral proteins have affinities for clathrin adaptors, which mediate their clathrin‐dependent basolateral surface targeting; in the absence of clathrin basolateral proteins still reach the plasma membrane, but they are mis‐ targeted to the apical domain [114]. Based on their interaction with different adaptors, basolateral cargoes take at least two different routes out of the TGN: If not captured by TGN‐localized clathrin adaptors such as AP1A, they are targeted to transferrin‐positive recycling endosomes where basolateral proteins interact with the endosomal clathrin adaptor AP1B for basolateral surface delivery. AP1A‐interacting cargo, by contrast, is delivered to the basolateral surface without traversing recycling compartments. While some basolateral proteins are known to



4:  Hepatocyte Surface Polarity

(a)

43 AP pathways

(b)

BL pathways

AP MyoVb Rab11a

ARE

AEE

AP1B

(i) MyoVb Rab11a

CRE BC BC TGN

ARE

AP1A AP4

TGN trans-

(ii)

TGN

BL

BL

Figure 4.4  Targeting pathways from the TGN to the apical/canalicular and basolateral surfaces in MDCK and hepatocytes. (a) Known pathways in MDCK cells. Apical (AP) pathways (green arrows): depending on their presence in detergent‐insoluble microdomains, raft‐dependent and raft‐ independent apical routes have been distinguished. Several membrane and secretory proteins have been shown to traverse the apical recycling endosome (ARE) (in a MyoVb and Rab11‐dependent manner) or the apical early endosome (AEE) before reaching the apical domain. Basolateral (BL) pathways (blue arrows): basolateral protein exiting from the TGN is clathrin mediated. Multiple pathways have been distinguished. Some basolateral proteins reach the basolateral domain without intermediate (dependent on TGN‐localized clathrin adaptors AP1A and AP4), others traverse the common recycling endosome (CRE), from where their basolateral targeting is mediated by clathrin adaptor AP1B. (b) Known and hypothetical pathways in hepatic cells. Two classes of apical proteins have been distinguished: polytopic membrane proteins, such as bile acid transporters, travel from the TGN to the apical domain directly (at least some via a Rab11 compartment); GPI‐anchored and single membrane‐spanning membrane proteins first reach the basolateral domain from the TGN. It is not resolved whether (i) apical and basolateral proteins are sorted into distinct TGN‐derived transport carriers that both reach the basolateral domain, or (ii) whether apical and basolateral proteins travel in common carriers to the basolateral domain. Since hepatocytes lack AP1B, it is unlikely that hepatocytes target their biosynthetic cargo via the CRE. Adapted from Treyer and Müsch, Compr Physiol, 2013;3:243–87 with permission of John Wiley & Sons.

enter both pathways, others are predominantly targeted via one or the other pathway [115, 116]. Apical sorting signals, on the other hand, are pleomorphic, constituted by N‐glycans, GPI anchors, and specialized transmembrane protein domains with affinity for lipid rafts. Lipid rafts are formed by sphingolipids and cholesterol, which cluster apical cargo and also establish membrane domain boundaries that segregate apical and basolateral cargo domains [117]. Apical proteins are targeted from the TGN in microtubule (MT)‐dependent tubular carriers, independently of clathrin [69]. Apical and basolateral transport carriers also differ in their fission and fusion machineries. Thus, while basolateral carrier scission depends on the activities of protein kinase D and on CtBP1‐S/BARS‐induced activation of lysophosphatidic acid acyltransferase δ [118–120], apical carriers require the scission factor dynamin 2, which functions in conjunction with branched actin filaments that assemble around the apical carriers at the scission site [121, 122]. Basolateral fusion is mediated by SNARE pairings containing VAMP3 and syntaxin 4, while apical carriers fuse via tetanus‐resistant v‐ SNAREs (VAMP7,8) that pair with syntaxin 3 [44, 123–127]. Hepatocytes can be expected to utilize some of the same targeting mechanisms, yet there also have to be differences to

account for heptocyte TGN‐to‐basolateral targeting of apical proteins that in MDCK cells are targeted to the apical surface directly. So far, only one contributing mechanism has been clearly identified: the lack in hepatocytes of myelin and lymphocyte 1 (MAL1), a proteolipid tetraspanning raft‐associated membrane protein, which in MDCK cells is critical for TGN‐ derived apical transport of proteins that depend on association with lipid rafts for their polarized targeting [128, 129]. Expression of MAL1 in the hepatocytic line WIF‐B (which ­normally lack it) promoted vectorial delivery of GPI‐apical proteins and single‐pass transmembrane proteins from the TGN to the apical surface [130]. It was also suggested (for the HepG2 cell line) that the transcytotic mechanism is regulated by the levels of surface E‐cadherin [86], although no mechanism for such regulation has been proposed. Hepatocytes also lack the basolateral cargo adaptor AP1B. In all other epithelia that naturally fail to express AP1B, AP1B‐ dependent basolateral proteins, such as transferrin receptor, localize at the apical domain [116]. This is because their endosomal sorting in both the exocytic and the recycling pathways are compromised. However, such polarity reversal is not observed in hepatocytes or hepatocytic cell lines. This implies

44

THE LIVER:  CYTOSKELETAL MICROFILAMENT AND MICROTUBULAR SYSTEMS IN TRAFFICKING

BL

LYS AEE MyoVb Rab11a

BC

ARE ?

LYS

BEE

BL proteins

SAC

BEE

AP proteins

Figure 4.5  Targeting pathways from the apical and basolateral plasma membranes in hepatocytes. Pathways for apical (AP) proteins (green arrows): Upon endocytosis from the canalicular domain proteins are sorted in the apical early endosome (AEE) into either the late endosomal/lysosomal pathway or for recycling to the apical recycling endosome (ARE). Apical proteins that have reached the basolateral surface from the TGN are internalized via clathrin‐coated vesicles (non‐raft) or in a clathrin‐independent, flotillin‐dependent manner (raft) and reach the subapical endosome (SAC) via basolateral early endosomes (BEEs). From the SAC they are targeted to the apical surface via the ARE. Basolateral (BL) resident proteins (blue arrows) undergo fast recycling from the BEE directly or via recycling endosomes. Whether they mingle with apical‐directed cargo in a common recycling endosome (CRE) as described for MDCK cells is still debated. While the SAC has been described by some as the equivalent of the MDCK CRE, others have suggested that BL and AP proteins are segregated before apical‐directed cargo reaches the SAC. Similar to the AEE, the BEE also sorts proteins for degradation to the lysosome (LYS). Adapted from Treyer and Müsch, Compr Physiol, 2013;3:243–87 with permission of John Wiley & Sons.

that hepatocytes organize their endosomal recycling itineraries differently from MDCK cells (Figure 4.5). In the MDCK model, endocytosed cargo can either undergo fast recycling to their membrane of origin from early endosomes or, alternatively, recycle from a CRE, where cargo from both surfaces mix and is sorted into distinct apical‐ or basolateral‐directed recycling carriers [131–133]. The CRE is also where basolateral‐to‐apical transyctotic cargo is segregated from basolateral recycling cargo [132, 134]. Apical carriers exiting the CRE either mature into or fuse with the AREs beneath the apical surface, while basolateral cargo returns to the surface directly [135, 136]. Studies in the hepatocytic cell lines WIF‐B and HepG2 as well as electron microscopy 3D reconstruction of the hepatocyte endosomal system have indeed suggested differences in endosome organization compared to MDCK cells. In particular, a subapical compartment (SAC) has been described, which according to different schools either mediates apical and basolateral cargo sorting equivalent to the MDCK CRE or represents a unique compartment of the basolateral‐to‐apical transcytotic pathway [16, 111, 137–139]. Regardless of the model and cell type, all

epithelial cells accumulate subapical membranes that appear to be bottlenecks for protein delivery to the apical surface. They are enriched in Rab GTPases, including Rab11a and its effector, the actin‐associated molecular motor myosin Vb, as well as the Rab11 adaptor proteins Fip1 and Fip2 [140–143]. Inhibition of Rab11a or myosin Vb indeed prevented polarization of WIF‐B cells and primary hepatocytes [144] and, when introduced into polarized cells, prompted depolarization and internalization of apical proteins [145]. These observations highlight the importance of apical membrane traffic for the establishment and maintenance of polarity. The Rab11‐positive endosomes also serve as “holding cells” for hepatocyte ABC transporters, the majority of which are maintained intracellularly rather than at the bile canalicular membrane [25, 146]. This intracellular pool is mobilized by bile acids circulating in the enterohepatic circuit, and by postprandially secreted peptide hormones, which increase cAMP production in hepatocytes to cope with the increased demand for bile acid secretion. Taurocholate and cAMP activate distinct signaling pathways to mobilize ABC transporters to the canalicular domain. Regulation of these two responses differs. Increase in cAMP concentration results in PKA‐mediated stimulation of phosphoinositide 3‐kinase but not taurocholate‐stimulated incorporation of ABCB11 into the canalicular membrane [147–150]. Bulk endocytosis from the apical surface of polarized epithelia occurs at much lower rate than endocytosis from the basolateral domain [70]. This is likely due to the extensive direct and indirect linkage of apical membrane proteins to actin filaments, which makes it hard for the endocytic machinery to deform the membrane. On the other hand, a branched actin network at the base of a nascent endocytic vesicle promotes clathrin‐mediated endocytosis when linked to the scission factor dynamin via the actin and dynamin‐binding protein cortactin [151, 152]. Apical bile acid transporters BSEP, MDR1, and MDR2 contain in their cytoplasmic domains a binding site for HCLS1‐associated protein X‐1 (HAX‐1) [153], which also binds cortactin and hence could recruit these ABC transporters into productive endocytic pits. Consistent with this hypothesis, HAX‐1 depletion increased BSEP accumulation at the apical domain [153]. Together, their facilitated endocytosis and regulated re‐insertion from an endocytic pool allows for the highly dynamic behavior of ABC transporters at the bile canalicular domain.

CYTOSKELETAL MICROFILAMENT AND MICROTUBULAR SYSTEMS IN TRAFFICKING(SEE CHAPTER 3) Proper endosomal trafficking and recycling of proteins to all plasma membrane domains requires an intact actin and microtubular cytoskeletal system. In particular, plus‐end dynamic microtubules marked by CLIP170 and EB1 mediate trafficking of secreted and canalicular proteins. Newly synthesized ABCB11, the canalicular bile acid transporter, and other canalicular ABC transporters traffic in post Golgi vesicles from the TGN along microtubules. However, dynamic microtubules do not attach to the canalicular membrane and their cargo endosomes are transferred to the pericanalicular actin system



4:  Hepatocyte Surface Polarity

(Figure 4.5). The complex mechanism for cargo transfer is not known; however, microtubules become associated with actin through a pericanalicular actin‐binding complex containing IQ Gap, APC, Hax‐1, and cortactin proteins. Live cell imaging studies reveal that selective plasma membrane localization of transporter proteins is predominantly due to the localization of specific docking proteins. In polarized WIF‐B cells, ABCB11 and ABCB1 traffic along microtubules throughout the cell but only attach to specific sites on the canalicular membrane. The docking site has been proposed to be syntaxin 3 that facilitates fusion of protein‐sorting vesicles with the inner leaflet of the canalicular membrane from which they diffuse until reaching the tight junction which limits presence to the canalicular domain. The actin‐binding protein radixin links some cargo molecules, such as ABCC2 and ABCB11, to the pericanalicular actin system. Radixin knockout mice manifest impaired ABCC2 and ABCB11 localization to the canalicular domain which becomes progressively devoid of microvilli, resulting in hepatocyte injury. Formin controls the assembly and disassembly of short actin filaments which are involved in endosomal transport. In the HepG2 hepatoma cell line INF2, CDC42 and the transmembrane protein MAL2 are required for transcytotic trafficking of canalicular membrane proteins.

HEPATOCYTE POLARITY AND CELL DIVISION Although mature polarized hepatocytes are quiescent, they can, under conditions of severe injury, re‐enter the cell cycle and proliferate [154]. Hepatocytes also actively divide during postnatal development while their bile canaliclar network matures. How proliferating epithelial cells orient their mitotic spindle and the cleavage furrow, which always forms perpendicular to the ­spindle pole axis, is of critical importance for both tissue and cellular organization [155]. Monopolar epithelial cells orient their metaphase spindle parallel to the basement membrane [156, 157]. This ensures that the cleavage furrow bisects their luminal domain, yielding two identical daughters that both remain in the epithelial plane, thereby ensuring that the epithelium remains monolayered. Critical to this outcome are cortical cues positioned at equal distance from the basal surface at opposite lateral domains. The cues consist of an evolutionarily conserved protein complex in which the α subunit of a trimeric G‐protein provides the cortical anchor and the minus end‐ directed microtubule motor dynein binds the plus ends of astral microtubules. In metaphase these complexes on opposite membrane domains capture one of each set of astral spindle microtubules, thereby aligning the metaphase spindle parallel to the basal domain [158, 159]. Unlike monopolar cells, hepatocytes rarely bisect their luminal domain in cytokinesis [160]. This preserves canalicular lumen organization and prevents the generation of acini. Remarkably, molecular analysis in the WIF‐B and HepG2 culture models suggested that the different cell division outcomes in monopolar and hepatocytic cells occur even though both cell types use similar spindle orientation mechanism, that is, both place their cortical spindle captures cues

(a)

Monopolar

45 Hepatocytic

(b)

Figure 4.6  Metaphase spindle orientation (a) and division outcome (b) in cultured monopolar and heptocytic cells. Monopolar epithelia orient their metaphase spindle parallel to their basal surface when astral microtubules bind to cortical cues (green) below the apical surface (red) (a). The cleavage furrow, which is established perpendicular to the spindle pole axis, bisects the apical surface and yields two identical daughters, each attached to the substratum (b). In hepatocytes the spindle attaches adjacent to two different lumina. The cleavage furrow therefore does not bisect the luminal domains. In cultured WIF‐B and HepG2 cells, a stereotypic spindle tilt with respect to the substrate contacting basal domain causes one daughter cells to be removed from the substratum. Whether this is relevant in vivo remains to be determined. Adapted from [64].

adjacent to their luminal surfaces [64, 161] (Figure 4.6). Lumen geometry and the presence of at least two luminal domains per hepatocyte likely cause each astral microtubule set to attach next to a different lumen, thereby preventing bisection of either lumen. By contrast, in monopolar cells, the two astral microtubule attachment sites flank the same (single) luminal domain, resulting in its bisection. In an additional difference to monopolar cells, hepatocytic cell lines can accommodate the metaphase spindle only in a quasi‐diagonal position, resulting in a stereotypic spindle tilt with respect to their basal domains, which yields a division out of the monolayer and results in the bilayering observed in HepG2 and WIF‐B cell cultures [65]. Live cell imaging of mammalian liver tissue [48, 162] should make it feasible to test the relevance of this division mode in vivo.

LIVER DISEASE AND POLARITY As detailed in Chapter 29, mutations in MYO5B encoding myosin Vb and Rab11b, Rab25, and Rab8, cause microvillus inclusion disease (MVID), in which malabsorption results from the absence of the intestinal brush border. Other patients manifest cholestasis and progressive liver disease. Mouse Rab8 conditional knockouts mimic MVID. Mutations affecting syntaxin 3, an apical membrane SNARE (family of membrane proteins that ensures fusion between opposing membranes), suggesting that myosin 5b, Rab8, and syntaxin 3 may be involved in the same

46

THE LIVER: REFERENCES

trafficking pathway. Discovery of loss‐of‐function mutations in genes encoding recycling endosome‐associated proteins such as myosin 5b in MVID, VPS33B and VIPAR in arthrogryposis, renal dysfunction, and cholestasis syndrome (ARC), also supports the importance of the RE in establishment and maintenance of hepatocyte polarity.

REFERENCES 1. Arias, I.M. et al. The Liver: Biology and Pathobiology. 5th edn (eds. I.M. Arias et al.), Wiley‐Blackwell, Chichester, 2009. 2. Si‐Tayeb, K., Lemaigre, F.P., and Duncan, S.A. Organogenesis and development of the liver. Dev Cell, 2010;18(2):175–89. 3. Gordillo, M., Evans, T., and Gouon‐Evans, V. Orchestrating liver development. Development, 2015;142(12):2094–108. 4. Antoniou, A. et al. Intrahepatic bile ducts develop according to a new mode of tubulogenesis regulated by the transcription factor SOX9. Gastroenterology, 2009;136(7):2325–33. 5. Zhang, L. et  al. The stem cell niche of human livers: symmetry between development and regeneration. Hepatology, 2008;48(5):1598–607. 6. Carpentier, R. et al. Embryonic ductal plate cells give rise to cholangiocytes, periportal hepatocytes, and adult liver progenitor cells. Gastroenterology, 2011;141(4):1432–8, 1438.e1–4. 7. Wauthier, E. et al. Hepatic stem cells and hepatoblasts: identification, isolation, and ex vivo maintenance. Methods Cell Biol, 2008;86:137–225. 8. Gallin, W.J. Development and maintenance of bile canaliculi in vitro and in vivo. Microsc Res Tech, 1997;39(5):406–12. 9. Kanamura, S., Kanai, K., and Watanabe, J. Fine structure and function of hepatocytes during development. J Electron Microsc Tech, 1990;14(2):92–105. 10. Little, J.M. et al. Taurocholate pool size and distribution in the fetal rat. J Clin Invest, 1979;63(5):1042–9. 11. Fu, D. et al. Bile acid stimulates hepatocyte polarization through a cAMP‐ Epac‐MEK‐LKB1‐AMPK pathway. Proc Natl Acad Sci U S A, 2011;108(4):1403–8. 12. Tanimizu, N. et al. Intrahepatic bile ducts are developed through formation of homogeneous continuous luminal network and its dynamic rearrangement in mice. Hepatology, 2016;64(1):175–88. 13. Lo, B. et  al. Lkb1 regulates organogenesis and early oncogenesis along  AMPK‐dependent and ‐independent pathways. J Cell Biol, 2012;199(7):1117–30. 14. Darlington, G.J., Kelly, J.H., and Buffone, G.J. Growth and hepatospecific gene expression of human hepatoma cells in a defined medium. In Vitro Cell Dev Biol, 1987;23(5):349–54. 15. Decaens, C. et al. Establishment of hepatic cell polarity in the rat hepatoma‐ human fibroblast hybrid WIF‐B9. A biphasic phenomenon going from a simple epithelial polarized phenotype to an hepatic polarized one. J Cell Sci, 1996;109(Pt 6):1623–35. 16. Ihrke, G. et  al. Apical plasma membrane proteins and endolyn‐78 travel through a subapical compartment in polarized WIF‐B hepatocytes. J Cell Biol, 1998;141(1):115–33. 17. Kelly, J.H. and Darlington, G.J. Modulation of the liver specific phenotype in the human hepatoblastoma line Hep G2. In Vitro Cell Dev Biol, 1989;25(2):217–22. 18. Peng, X. et al. How to induce non‐polarized cells of hepatic origin to express typical hepatocyte polarity: generation of new highly polarized cell models with developed and functional bile canaliculi. Cell Tissue Res, 2006;323(2):233–43. 19. van der Wouden, J.M., van IJzendoorn, S.C., and Hoekstra, D. Oncostatin M regulates membrane traffic and stimulates bile canalicular membrane biogenesis in HepG2 cells. EMBO J, 2002;21(23):6409–18. 20. Wakabayashi, Y., Lippincott‐Schwartz, J., and Arias, I.M. Intracellular trafficking of bile salt export pump (ABCB11) in polarized hepatic cells: constitutive cycling between the canalicular membrane and rab11‐positive endosomes. Mol Biol Cell, 2004;15(7):3485–96. 21. Dunn, J.C., Tompkins, R.G., and Yarmush, M.L. Long‐term in vitro function of adult hepatocytes in a collagen sandwich configuration. Biotechnol Prog, 1991;7(3):237–45.

22. Dunn, J.C. et  al. Hepatocyte function and extracellular matrix geometry: long‐term culture in a sandwich configuration. FASEB J, 1989;3(2):174–7. 23. LeCluyse, E.L., Audus, K.L., and Hochman, J.H. Formation of extensive canalicular networks by rat hepatocytes cultured in collagen‐sandwich configuration. Am J Physiol, 1994;266(6 Pt 1):C1764–74. 24. Michalopoulos, G.K. et al. Comparative analysis of mitogenic and morphogenic effects of HGF and EGF on rat and human hepatocytes maintained in collagen gels. J Cell Physiol, 1993;156(3):443–52. 25. Homolya, L. et al. LKB1/AMPK and PKA control ABCB11 trafficking and polarization in hepatocytes. PLoS One, 2014;9(3):e91921. 26. Grubman, S.A. et al. Regulation of intracellular pH by immortalized human intrahepatic biliary epithelial cell lines. Am J Physiol, 1994;266(6 Pt 1):G1060–70. 27. Vroman, B. and LaRusso, N.F. Development and characterization of polarized primary cultures of rat intrahepatic bile duct epithelial cells. Lab Invest, 1996;74(1):303–13. 28. Salter, K.D. et al. Modified culture conditions enhance expression of differentiated phenotypic properties of normal rat cholangiocytes. Lab Invest, 2000;80(11):1775–8. 29. Gradilone, S.A. et al. Cholangiocyte cilia express TRPV4 and detect changes in luminal tonicity inducing bicarbonate secretion. Proc Natl Acad Sci U S A, 2007;104(48):19138–43. 30. Lazaridis, K.N. et  al. Kinetic and molecular identification of sodium‐ dependent glucose transporter in normal rat cholangiocytes. Am J Physiol, 1997;272(5 Pt 1):G1168–74. 31. Spirli, C. et al. Functional polarity of Na+/H+ and Cl−/HCO3− exchangers in a rat cholangiocyte cell line. Am J Physiol, 1998;275(6 Pt 1):G1236–45. 32. Zsembery, A. et al. Purinergic regulation of acid/base transport in human and rat biliary epithelial cell lines. Hepatology, 1998;28(4):914–20. 33. Miyajima, A., Tanaka, M., and Itoh, T. Stem/progenitor cells in liver development, homeostasis, regeneration, and reprogramming. Cell Stem Cell, 2014;14(5):561–74. 34. Reid, L.M. Stem/progenitor cells and reprogramming (plasticity) mechanisms in liver, biliary tree, and pancreas. Hepatology, 2016;64(1):4–7. 35. Tanimizu, N. et al. Long‐term culture of hepatic progenitors derived from mouse Dlk+ hepatoblasts. J Cell Sci, 2004;117(Pt 26):6425–34. 36. Wang, Y. et al. Paracrine signals from mesenchymal cell populations govern the expansion and differentiation of human hepatic stem cells to adult liver fates. Hepatology, 2010;52(4):1443–54. 37. Huch, M. et al. Long‐term culture of genome‐stable bipotent stem cells from adult human liver. Cell, 2015;160(1–2):299–312. 38. Tanimizu, N., Miyajima, A., and Mostov, K.E. Liver progenitor cells develop cholangiocyte‐type epithelial polarity in three‐dimensional culture. Mol Biol Cell, 2007;18(4):1472–9. 39. Tanimizu, N. et al. Hepatic biliary epithelial cells acquire epithelial integrity but lose plasticity to differentiate into hepatocytes in vitro during development. J Cell Sci, 2013;126(Pt 22):5239–46. 40. Guillouzo, A. et al. The human hepatoma HepaRG cells: a highly differentiated model for studies of liver metabolism and toxicity of xenobiotics. Chem Biol Interact, 2007;168(1):66–73. 41. Gunness, P. et al. 3D organotypic cultures of human HepaRG cells: a tool for in vitro toxicity studies. Toxicol Sci, 2013;133(1):67–78. 42. Dianat, N. et  al. Generation of functional cholangiocyte‐like cells from human pluripotent stem cells and HepaRG cells. Hepatology, 2014;60(2):700–14. 43. Nakatani, T. et al. Assessment of efficiency and safety of adenovirus mediated gene transfer into normal and damaged murine livers. Gut, 2000;47(4):563–70. 44. Shayakhmetov, D.M. et al. Analysis of adenovirus sequestration in the liver, transduction of hepatic cells, and innate toxicity after injection of fiber‐modified vectors. J Virol, 2004;78(10):5368–81. 45. Zhang, G., Budker, V., and Wolff, J.A. High levels of foreign gene expression in hepatocytes after tail vein injections of naked plasmid DNA. Hum Gene Ther, 1999;10(10):1735–7. 46. Eggenhofer, E. et al. High volume naked DNA tail‐vein injection restores liver function in Fah‐knock out mice. J Gastroenterol Hepatol, 2010;25(5):1002–8. 47. Marques, P.E. et al. Imaging liver biology in vivo using conventional confocal microscopy. Nat Protoc, 2015;10(2):258–68. 48. Porat‐Shliom, N. et al. Liver kinase B1 regulates hepatocellular tight junction distribution and function in vivo. Hepatology, 2016;64(4):1317–29.



4:  Hepatocyte Surface Polarity

49. Tepass, U. The apical polarity protein network in Drosophila epithelial cells: regulation of polarity, junctions, morphogenesis, cell growth, and survival. Annu Rev Cell Dev Biol, 2012;28:655–85. 50. Thompson, B.J. Cell polarity: models and mechanisms from yeast, worms and flies. Development, 2013;140(1):13–21. 51. Rodriguez‐Boulan, E. and Macara, I.G. Organization and execution of the epithelial polarity programme. Nat Rev Mol Cell Biol, 2014;15(4):225–42. 52. Takaki, Y. et al. Dynamic changes in protein components of the tight junction during liver regeneration. Cell Tissue Res, 2001;305(3):399–409. 53. Kallay, L.M. et  al. Scribble associates with two polarity proteins, Lg12 and  Vang12, via distinct molecular domains. J Cell Biochem, 2006;99(2):647–64. 54. Terada, T. and Nakanuma, Y. Expression of tenascin, type IV collagen and laminin during human intrahepatic bile duct development and in intrahepatic cholangiocarcinoma. Histopathology, 1994;25(2):143–50. 55. Shiojiri, N. and Sugiyama, Y. Immunolocalization of extracellular matrix components and integrins during mouse liver development. Hepatology, 2004;40(2):346–55. 56. Tanimizu, N. et  al. alpha1‐ and alpha5‐containing laminins regulate the development of bile ducts via beta1 integrin signals. J Biol Chem, 2012;287(34):28586–97. 57. Schaffner, F. and Poper, H. Capillarization of hepatic sinusoids in man. Gastroenterology, 1963;44:239–42. 58. Yang, L. et al. A single‐cell transcriptomic analysis reveals precise pathways and regulatory mechanisms underlying hepatoblast differentiation. Hepatology, 2017;66(5):1387–401. 59. Martinez‐Hernandez, A. and Amenta, P.S. The hepatic extracellular matrix. I. Components and distribution in normal liver. Virchows Arch A Pathol Anat Histopathol, 1993;423(1):1–11. 60. Bataller, R. and Brenner, D.A. Liver fibrosis. J Clin Invest, 2005;115(2):209–18. 61. Wells, R.G. The role of matrix stiffness in regulating cell behavior. Hepatology, 2008;47(4):1394–400. 62. Cohen, D. et al. Mammalian PAR‐1 determines epithelial lumen polarity by organizing the microtubule cytoskeleton. J Cell Biol, 2004;164(5):717–27. 63. Cohen, D. et al. The serine/threonine kinase Par1b regulates epithelial lumen polarity via IRSp53‐mediated cell‐ECM signaling. J Cell Biol, 2011;192(3):525–40. 64. Lazaro‐Dieguez, F. et al. Par1b links lumen polarity with LGN‐NuMA positioning for distinct epithelial cell division phenotypes. J Cell Biol, 2013;203(2):251–64. 65. Lazaro‐Dieguez, F. and Musch, A. Cell‐cell adhesion accounts for the different orientation of columnar and hepatocyte cell division. J Cell Biol, 2017;216(11):3847–59. 66. Bryant, D.M. et al. A molecular network for de novo generation of the apical surface and lumen. Nat Cell Biol, 2010;12(11):1035–45. 67. Galvez‐Santisteban, M. et al. Synaptotagmin‐like proteins control the formation of a single apical membrane domain in epithelial cells. Nat Cell Biol, 2012;14(8):838–49. 68. Tuma, P. and Hubbard, A.L. Transcytosis: crossing cellular barriers. Physiol Rev, 2003;83(3):871–932. 69. Weisz, O.A. and Rodriguez‐Boulan, E. Apical trafficking in epithelial cells: signals, clusters and motors. J Cell Sci, 2009;122(Pt 23):4253–66. 70. Boulant, S. et al. Actin dynamics counteract membrane tension during clathrin‐mediated endocytosis. Nat Cell Biol, 2011;13(9):1124–31. 71. Szalinski, C.M. et al. PIP5KIbeta selectively modulates apical endocytosis in polarized renal epithelial cells. PLoS One, 2013;8(1):e53790. 72. Sato, T. et al. Regulation of the assembly and adhesion activity of E‐cadherin by nectin and afadin for the formation of adherens junctions in Madin‐Darby canine kidney cells. J Biol Chem, 2006;281(8):5288–99. 73. Takai, Y. et  al. Nectins and nectin‐like molecules: roles in cell adhesion, migration, and polarization. Cancer Sci, 2003;94(8):655–67. 74. Tsuchiya, B. et al. Differential expression of N‐cadherin and E‐cadherin in normal human tissues. Arch Histol Cytol, 2006;69(2):135–45. 75. Straub, B.K. et al. E‐N‐cadherin heterodimers define novel adherens junctions connecting endoderm‐derived cells. J Cell Biol, 2011;195(5):873–87. 76. Wheelock, M.J. et  al. Cadherin switching. J Cell Sci, 2008;121(Pt 6):727–35. 77. Rajasekaran, A.K. et al. Catenins and zonula occludens‐1 form a complex during early stages in the assembly of tight junctions. J Cell Biol, 1996;132(3):451–63.

47

78. Asakura, T. et al. Similar and differential behaviour between the nectin‐afadin‐ponsin and cadherin‐catenin systems during the formation and disruption of the polarized junctional alignment in epithelial cells. Genes Cells, 1999;4(10):573–81. 79. Braiterman, L.T. et al. JAM‐A is both essential and inhibitory to development of hepatic polarity in WIF‐B cells. Am J Physiol Gastrointest Liver Physiol, 2008;294(2):G576–88. 80. Rehder, D. et al. Junctional adhesion molecule‐a participates in the formation of apico‐basal polarity through different domains. Exp Cell Res, 2006;312(17):3389–403. 81. Son, S. et al. Knockdown of tight junction protein claudin‐2 prevents bile canalicular formation in WIF‐B9 cells. Histochem Cell Biol, 2009;131(3):411–24. 82. Grosse, B. et al. Build them up and break them down: tight junctions of cell lines expressing typical hepatocyte polarity with a varied repertoire of claudins. Tissue Barriers, 2013;1(4):e25210. 83. Takeichi, M. Cadherin cell adhesion receptors as a morphogenetic regulator. Science, 1991;251(5000):1451–5. 84. Gonzalez‐Mariscal, L., Chavez de Ramirez, B., and Cereijido, M. Tight junction formation in cultured epithelial cells (MDCK). J Membr Biol, 1985;86(2):113–25. 85. Vega‐Salas, D.E., Salas, P.J. and Rodriguez‐Boulan, E. Exocytosis of vacuolar apical compartment (VAC): a cell‐cell contact controlled mechanism for the establishment of the apical plasma membrane domain in epithelial cells. J Cell Biol, 1988;107(5):1717–28. 86. Theard, D. et al. Cell polarity development and protein trafficking in hepatocytes lacking E‐cadherin/beta‐catenin‐based adherens junctions. Mol Biol Cell, 2007;18(6):2313–21. 87. Konopka, G. et al. Junctional adhesion molecule‐A is critical for the formation of pseudocanaliculi and modulates E‐cadherin expression in hepatic cells. J Biol Chem, 2007;282(38):28137–48. 88. Cohen, D., Tian, Y., and Musch, A. Par1b promotes hepatic‐type lumen polarity in Madin Darby canine kidney cells via myosin II‐ and E‐cadherin‐ dependent signaling. Mol Biol Cell, 2007;18(6):2203–15. 89. Su, X. et al. Single‐cell RNA‐Seq analysis reveals dynamic trajectories during mouse liver development. BMC Genomics, 2017;18(1):946. 90. Wang, Y. et al. Mechanical compaction directly modulates the dynamics of bile canaliculi formation. Integr Biol (Camb), 2013;5(2):390–401. 91. Godoy, P. et al. Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non‐parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch Toxicol, 2013;87(8):1315–530. 92. Alessi, D.R., Sakamoto, K., and Bayascas, J.R. LKB1‐dependent signaling pathways. Annu Rev Biochem, 2006;75:137–63. 93. Baas, A.F. et al. Complete polarization of single intestinal epithelial cells upon activation of LKB1 by STRAD. Cell, 2004;116(3):457–66. 94. Zhang, L. et al. AMP‐activated protein kinase regulates the assembly of epithelial tight junctions. Proc Natl Acad Sci U S A, 2006;103(46):17272–7. 95. Zheng, B. and Cantley, L.C. Regulation of epithelial tight junction assembly and disassembly by AMP‐activated protein kinase. Proc Natl Acad Sci U S A, 2007;104(3):819–22. 96. Aznar, N. et al. AMP‐activated protein kinase fortifies epithelial tight junctions during energetic stress via its effector GIV/Girdin. Elife, 2016;5. 97. Yano, T. et  al. The association of microtubules with tight junctions is promoted by cingulin phosphorylation by AMPK. J Cell Biol, 2013; ­ 203(4):605–14. 98. Woods, A. et al. LKB1 is required for hepatic bile acid transport and canalicular membrane integrity in mice. Biochem J, 2011;434(1):49–60. 99. Fu, D. et al. Coordinated elevation of mitochondrial oxidative phosphorylation and autophagy help drive hepatocyte polarization. Proc Natl Acad Sci U S A, 2013;110(18):7288–93. 100. Moghe, P.V. et  al. Culture matrix configuration and composition in the maintenance of hepatocyte polarity and function. Biomaterials, 1996;17(3):373–85. 101. Mooney, D. et al. Switching from differentiation to growth in hepatocytes: control by extracellular matrix. J Cell Physiol, 1992;151(3):497–505. 102. Ezzell, R.M. et al. Effect of collagen gel configuration on the cytoskeleton in cultured rat hepatocytes. Exp Cell Res, 1993;208(2):442–52. 103. Berthiaume, F. et al. Effect of extracellular matrix topology on cell structure, function, and physiological responsiveness: hepatocytes cultured in a sandwich configuration. FASEB J, 1996;10(13):1471–84.

48

THE LIVER: REFERENCES

104. Fassett, J., Tobolt, D., and Hansen, L.K. Type I collagen structure regulates cell morphology and EGF signaling in primary rat hepatocytes through cAMP‐dependent protein kinase A. Mol Biol Cell, 2006;17(1):345–56. 105. Godoy, P. et al. Extracellular matrix modulates sensitivity of hepatocytes to fibroblastoid dedifferentiation and transforming growth factor beta‐induced apoptosis. Hepatology, 2009;49(6):2031–43. 106. Desai, S.S. et al. Physiological ranges of matrix rigidity modulate primary mouse hepatocyte function in part through hepatocyte nuclear factor 4 alpha. Hepatology, 2016;64(1):261–75. 107. Rodriguez‐Boulan, E., Kreitzer, G., and Musch, A. Organization of vesicular trafficking in epithelia. Nat Rev Mol Cell Biol, 2005;6(3):233–47. 108. Bartles, J.R. et al. Biogenesis of the rat hepatocyte plasma membrane in vivo: comparison of the pathways taken by apical and basolateral proteins using subcellular fractionation. J Cell Biol, 1987;105(3):1241–51. 109. Kipp, H. and Arias, I.M. Newly synthesized canalicular ABC transporters are directly targeted from the Golgi to the hepatocyte apical domain in rat liver. J Biol Chem, 2000;275(21):15917–25. 110. Schell, M.J. et al. 5′nucleotidase is sorted to the apical domain of hepatocytes via an indirect route. J Cell Biol, 1992;119(5):1173–82. 111. Barr, V.A., Scott, L.J., and Hubbard, A.L. Immunoadsorption of hepatic vesicles carrying newly synthesized dipeptidyl peptidase IV and polymeric IgA receptor. J Biol Chem, 1995;270(46):27834–44. 112. Saucan, L. and Palade, G.E. Differential colchicine effects on the transport of membrane and secretory proteins in rat hepatocytes in vivo: bipolar secretion of albumin. Hepatology, 1992;15(4):714–21. 113. Bastaki, M. et al. Absence of direct delivery for single transmembrane apical proteins or their “Secretory” forms in polarized hepatic cells. Mol Biol Cell, 2002;13(1):225–37. 114. Deborde, S. et al. Clathrin is a key regulator of basolateral polarity. Nature, 2008;452(7188):719–23. 115. Gravotta, D. et al. The clathrin adaptor AP‐1A mediates basolateral polarity. Dev Cell, 2012;22(4):811–23. 116. Folsch, H. Role of the epithelial cell‐specific clathrin adaptor complex AP‐1B in cell polarity. Cell Logist, 2015;5(2):e1074331. 117. Schuck, S. and Simons, K. Polarized sorting in epithelial cells: raft clustering and the biogenesis of the apical membrane. J Cell Sci, 2004;117(Pt 25):5955–64. 118. Yeaman, C. et al. Protein kinase D regulates basolateral membrane protein exit from trans‐Golgi network. Nat Cell Biol, 2004;6(2):106–12. 119. Valente, C., Luini, A., and Corda, D. Components of the CtBP1/BARS‐ dependent fission machinery. Histochem Cell Biol, 2013;140(4):407–21. 120. Pagliuso, A. et  al. Golgi membrane fission requires the CtBP1‐S/BARS‐ induced activation of lysophosphatidic acid acyltransferase delta. Nat Commun, 2016;7:12148. 121. Kreitzer, G. et al. Kinesin and dynamin are required for post‐Golgi transport of a plasma‐membrane protein. Nat Cell Biol, 2000;2(2):125–7. 122.. Salvarezza, S.B. et  al. LIM kinase 1 and cofilin regulate actin filament population required for dynamin‐dependent apical carrier fission from the trans‐Golgi network. Mol Biol Cell, 2009;20(1):438–51. 123. Breuza, L., Fransen, J., and Le Bivic, A. Transport and function of syntaxin 3 in human epithelial intestinal cells. Am J Physiol Cell Physiol, 2000;279(4):C1239–48. 124. Ikonen, E. et  al. Different requirements for NSF, SNAP, and Rab proteins  in apical and basolateral transport in MDCK cells. Cell, 1995;81(4):571–80. 125. Low, S.H. et al. Differential localization of syntaxin isoforms in polarized Madin‐Darby canine kidney cells. Mol Biol Cell, 1996;7(12):2007–18. 126. Pocard, T. et  al. Distinct v‐SNAREs regulate direct and indirect apical delivery in polarized epithelial cells. J Cell Sci, 2007;120(Pt 18):3309–20. 127. Reales, E. et al. Basolateral sorting of syntaxin 4 is dependent on its N‐terminal domain and the AP1B clathrin adaptor, and required for the epithelial cell polarity. PLoS One, 2011;6(6):e21181. 128. Puertollano, R. et al. The MAL proteolipid is necessary for normal apical transport and accurate sorting of the influenza virus hemagglutinin in Madin‐Darby canine kidney cells. J Cell Biol, 1999;145(1):141–51. 129. Cheong, K.H. et al. VIP17/MAL, a lipid raft‐associated protein, is involved in apical transport in MDCK cells. Proc Natl Acad Sci U S A, 1999;96(11):6241–8. 130. Ramnarayanan, S.P. et al. Exogenous MAL reroutes selected hepatic apical proteins into the direct pathway in WIF‐B cells. Mol Biol Cell, 2007;18(7):2707–15.

131. Sheff, D.R. et  al. The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions. J Cell Biol, 1999;145(1):123–39. 132. Wang, E. et al. Apical and basolateral endocytic pathways of MDCK cells meet in acidic common endosomes distinct from a nearly‐neutral apical recycling endosome. Traffic, 2000;1(6):480–93. 133. Maxfield, F.R. and McGraw, T.E. Endocytic recycling. Nat Rev Mol Cell Biol, 2004;5(2):121–32. 134. Odorizzi, G. et al. Apical and basolateral endosomes of MDCK cells are interconnected and contain a polarized sorting mechanism. J Cell Biol, 1996;135(1):139–52. 135. Apodaca, G., Katz, L.A., and Mostov, K.E. Receptor‐mediated transcytosis of IgA in MDCK cells is via apical recycling endosomes. J Cell Biol, 1994;125(1):67–86. 136. Barroso, M. and Sztul, E.S. Basolateral to apical transcytosis in polarized cells is indirect and involves BFA and trimeric G protein sensitive passage through the apical endosome. J Cell Biol, 1994;124(1–2):83–100. 137. Rahner, C., Stieger, B., and Landmann, L. Apical endocytosis in rat hepatocytes In situ involves clathrin, traverses a subapical compartment, and leads to lysosomes. Gastroenterology, 2000;119(6):1692–707. 138. Marbet, P. et  al. Quantitative microscopy reveals 3D organization and kinetics of endocytosis in rat hepatocytes. Microsc Res Tech, 2006;69(9):693–707. 139. Hoekstra, D., Tyteca, D., and van IJzendoorn, S.C. The subapical compartment: a traffic center in membrane polarity development. J Cell Sci, 2004;117(Pt 11):2183–92. 140. Casanova, J.E. et al. Association of Rab25 and Rab11a with the apical recycling system of polarized Madin‐Darby canine kidney cells. Mol Biol Cell, 1999;10(1):47–61. 141. Hales, C.M., Vaerman, J.P., and Goldenring, J.R. Rab11 family interacting protein 2 associates with Myosin Vb and regulates plasma membrane recycling. J Biol Chem, 2002;277(52):50415–21. 142. Jing, J. and Prekeris, R. Polarized endocytic transport: the roles of Rab11 and Rab11‐FIPs in regulating cell polarity. Histol Histopathol, 2009;24(9):1171–80. 143. Lapierre, L.A. et al. Phosphorylation of Rab11‐FIP2 regulates polarity in MDCK cells. Mol Biol Cell, 2012;23(12):2302–18. 144. Fu, D. et al. Regulation of bile canalicular network formation and maintenance by AMP‐activated protein kinase and LKB1. J Cell Sci, 2010;123(Pt 19):3294–302. 145. Wakabayashi, Y. et al. Rab11a and myosin Vb are required for bile canalicular formation in WIF‐B9 cells. Proc Natl Acad Sci U S A, 2005;102(42):15087–92. 146. Wakabayashi, Y., Kipp, H., and Arias, I.M. Transporters on demand: intracellular reservoirs and cycling of bile canalicular ABC transporters. J Biol Chem, 2006;281(38):27669–73. 147. Gatmaitan, Z.C., Nies, A.T., and Arias, I.M. Regulation and translocation of ATP‐dependent apical membrane proteins in rat liver. Am J Physiol, 1997;272(5 Pt 1):G1041–9. 148. Misra, S. et  al. The role of phosphoinositide 3‐kinase in taurocholate‐ induced trafficking of ATP‐dependent canalicular transporters in rat liver. J Biol Chem, 1998;273(41):26638–44. 149. Misra, S. et  al. Phosphoinositide 3‐kinase lipid products regulate ATP‐ dependent transport by sister of P‐glycoprotein and multidrug resistance associated protein 2 in bile canalicular membrane vesicles. Proc Natl Acad Sci U S A, 1999;96(10):5814–19. 150. Misra, S., Varticovski, L., and Arias, I.M. Mechanisms by which cAMP increases bile acid secretion in rat liver and canalicular membrane vesicles. Am J Physiol Gastrointest Liver Physiol, 2003;285(2):G316–24. 151. Cao, H. et  al. Cortactin is a component of clathrin‐coated pits and participates in receptor‐mediated endocytosis. Mol Cell Biol, ­ 2003;23(6):2162–70. 152. Smythe, E. and Ayscough, K.R. Actin regulation in endocytosis. J Cell Sci, 2006;119(Pt 22):4589–98. 153. Ortiz, D.F. et al. Identification of HAX‐1 as a protein that binds bile salt export protein and regulates its abundance in the apical membrane of Madin‐Darby canine kidney cells. J Biol Chem, 2004;279(31):32761–70. 154. Miyaoka, Y. et al. Hypertrophy and unconventional cell division of hepatocytes underlie liver regeneration. Curr Biol, 2012;22(13):1166–75. 155. Ragkousi, K. and Gibson, M.C. Cell division and the maintenance of epithelial order. J Cell Biol, 2014;207(2):181–8.



4:  Hepatocyte Surface Polarity

156. Reinsch, S. and Karsenti, E. Orientation of spindle axis and distribution of plasma membrane proteins during cell division in polarized MDCKII cells. J Cell Biol, 1994;126(6):1509–26. 157. Tuncay, H. and Ebnet, K. Cell adhesion molecule control of planar spindle orientation. Cell Mol Life Sci, 2016;73(6):1195–207. 158. McNally, F.J. Mechanisms of spindle positioning. J Cell Biol, 2013;200(2):131–40. 159. Kotak, S. and Gonczy, P. Mechanisms of spindle positioning: cortical force generators in the limelight. Curr Opin Cell Biol, 2013;25(6):741–8.

49

160. Bartles, J.R. and Hubbard, A.L. Preservation of hepatocyte plasma membrane domains during cell division in situ in regenerating rat liver. Dev Biol, 1986;118(1):286–95. 161. Slim, C.L. et  al. Par1b induces asymmetric inheritance of plasma membrane domains via LGN‐dependent mitotic spindle orientation in proliferating hepatocytes. PLoS Biol, 2013;11(12):e1001739. 162. Meyer, K. et  al. A predictive 3D multi‐scale model of biliary fluid ­dynamics in the liver lobule. Cell Syst, 2017;4(3):277–90.

5

Primary Cilia Carolyn M. Ott Janelia Research Campus, Ashburn, VA, USA

INTRODUCTION A multicellular organism has a need to coordinate multiple processes in space and time. During development, multiple cell lineages migrate and differentiate simultaneously. Individual organ function, though physically isolated, must be coordinated with the activity of other organs. Similarly, each specialized cell coordinates with neighboring cells. Coordination is achieved as individual cells respond to messages from other cells and the environment, and transmit signals about their own state and needs. Amid the cacophony of potential messages, individual cells must discern which messages are relevant. Cilia project away from the cell surface and function as a tunable sensing organelle. Because cilia are continuous with both the cellular cytoplasm and plasma membrane, specialized barriers restrict entry and exit so that cells can form and adjust the composition of cilia to sense and respond to signals. The term “cilia” has been generalized to encompass all types of ciliated structures, including flagella, and motile and non‐motile cilia. Cilia and flagella, present throughout the eukaryotic lineage, are distinct in structure and composition from prokaryotic flagella. In eukaryotes, microtubules provide both structural support and a transportation highway within cilia. Unicellular eukaryotes use cilia to process responses to stimuli from the environment, mate, and feed. In mice and other mammals, genetic ablation of primary cilia terminates development. In adults, ciliary loss leads to disease [1]. Primary cilia are typically solitary projections from the cell surface and they lack the structural components that facilitate active beating in motile cilia and flagella. However, primary cilia are not stationary, but can move with the extracellular environment or pivot due to forces from the intracellular actin cytoskeleton [2].

Primary cilia are present on a subset of cells in the liver; cholangocytes have primary cilia, while hepatocytes do not. The absence of cilia may be essential to specialization. Outside the hematopoietic lineage, all progenitor cells have primary cilia, so cilia‐less, differentiated cells have lost the ability to respond to messages in the same way, and thus may be insulated. Like hepatocytes, adipocytes and myocytes lack cilia, but are derived from ciliated progenitors. The liver is a source of signals that can be perceived by primary cilia elsewhere in the body. For example, insulin‐like growth factor‐1 (IGF‐1) is produced by the liver. The IGF‐1 receptor is expressed in diverse tissues and can localize to cilia. IGF‐1 signaling through cilia can initiate cilia resorption as cells enter the mitotic cycle [3]. Cells have many modes of perceiving the extracellular environment. What advantages do cilia provide for signal perception and transmission? Location, structure, isolation, and size make the primary cilium a unique environment.

Location Like a probe or an antenna, primary cilia extend away from the cell and are immersed in the extracellular environment. The structure is positioned to detect and report the conditions outside the cell. Cells can specify which part of the environment to monitor using the cilium. For example, polarized cells typically position cilia at the apical surface.

Structure Two structural features of primary cilia are essential to consider. First, because the cilium is a long narrow structure, it has a very high surface area‐to‐volume ratio. The plasma membrane provides a platform for detecting the extracellular environment.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



5:  Primary Cilia

Second, the polarized microtubules that provide structural support also act as highways for traffic along the length of the cilium. Transport along these microtubules can be regulated and facilitates signaling. Tubulin in the microtubules themselves may also have direct effects on proteins in the membrane [4].

Isolation Although small proteins may be able to access primary cilia, entry of larger proteins is restricted and targeting sequences or interacting partners are thought to be essential. Regulation of both the cytoplasmic and plasma membrane content of the primary cilium tunes the cilium to detect specific ligands. In addition, downstream signaling molecules within the cilium are protected from modification by enzymes in the cytoplasm. Although ions can pass between the cilium and the cytoplasm, channels localized in the ciliary plasma membrane create a unique ionic environment inside cilia that may influence protein activity [5].

Size The volume of a 5 μm cilium is approximately 0.35 μm3, which is less than 0.05% of the total eukaryotic cell volume [6]. As a result, within the cilium the concentration of a single molecule is much larger than it would be in the cytoplasm. In addition, because changes in cilia length change cilia volume, concentrations within the cilium could also be regulated by processes that affect cilia growth.

CILIUM COMPOSITION AND ORGANIZATION In the cytoplasm, most microtubules are hollow tubes created by assembly of α and β tubulin monomers assembled into 13 protofilaments. Cilia microtubules are doublets composed of a 13‐protofilament tubule (designated the A tubule) and a 10‐protofilament B tubule that closes onto the outside of three A tubule protofilaments (Figure 5.1). Ciliary microtubules receive several post‐translational modifications: acetylation, detyrosination, glutamylation, and glycylation (Figure 5.2a) [7]. The nine doublet microtubules are direct extensions of centriole microtubules. As a consequence, primary cilia are always anchored in the cytoplasm. Collectively, cilia microtubes are referred to as an axoneme. Although polarized, axoneme microtubules do not treadmill like cytoplasmic microtubules because there is no depolymerization at the minus end within the centrosome. Axoneme length is mediated by addition or removal of tubulin at the plus end at the cilium tip. Unlike flagella and motile cilia that have internal structures to maintain the radial symmetry of the microtubules along the entire length of the cilium, primary cilia microtubules can rearrange toward the distal end. Primary cilia can also narrow as individual doublets terminate (Figure 5.1) [8]. As protein synthesis does not occur in cilia, growth requires tubulin monomers to be imported into the cilium and travel to the tip. The intraflagellar transport (IFT) complex facilitates

51

directed movement along axonemal microtubules (reviewed in [9]). IFT was first discovered as movement in the flagella of the unicellular alga Chlamydomonas reinhardtii [10]. The IFT complexes form trains (approximately 200 nm long in Chlamydomonas flagella) that bind to microtubule motor proteins, cargo, such as tubulin, and cargo adaptors that bind soluble and membrane proteins. IFT complexes are large multiprotein complexes about the size of the ribosomal small subunit (16S) that can be biochemically separated into two subcomplexes, IFT‐A and IFT‐B [11]. The IFT‐B subcomplexes associate with kinesin‐2 to facilitate anterograde transport along the B tubules of the axoneme doublets. The retrograde motor, dynein, is carried to the plus end of ciliary microtubules through association with the IFT‐A subcomplex. At the tip, kinesin‐2 dissociates from the IFT complex, retrograde trains form, and IFT‐A‐associated dynein drives retrograde movement along the A tubules of the axoneme doublets (Figure 5.1) [12]. Two kinesin‐2 family members, Kif3 and Kif17, process along ciliary microtubules: Kif3 is required for ciliogenesis while Kif17 appears to participate in more specialized trafficking. The specific tubulin isotypes in cilia, as well as the multiple tubulin modifications are thought to affect IFT transport. Membrane surrounds cilia microtubules as they extend away from the centriole (also called a basal body). While some centrosomes are at the cell surface, many are recessed and have a cililary pocket: a specialized membrane domain with active exocytosis and endocytosis [13]. The base of the cilium, near the centriole, serves as the access point for both cytoplasmic and membrane components. The composition of a cilium defines the sensitivity and responsiveness of the organelle. While the ciliary membrane is continuous with the plasma membrane of the cell, access to both the ciliary membrane and cytoplasm is restricted by a boundary created by a structure located inside the cilium, just above the centrosome, called the transition zone. Mutations in transition zone proteins or other proteins responsible for creating and regulating cilia composition are responsible for many cilia‐related diseases, termed ciliopathies. Protein entry through the transition zone is gated and although small molecules can pass through the transition zone, several active processes maintain the unique environment of the cilium. Most ciliopathies are pleiotropic diseases that affect several overlapping organ systems including liver, kidney, retina, heart, and bones [1]. Several ciliopathies, including nephronophthisis (NPHP), Meckel syndrome (MKS), and Joubert syndrome, are caused by mutation of transition zone proteins [14]. While often considered a single defined structure, transition zone organization and length can vary between cell types [15]. Tissue‐specific differences in transition zones might contribute to the differences in symptoms upon loss or mutation of different transition zone components. Genetic, biochemical, and proteomic strategies identified many components of the transition zone [16–18]. By electron microscopy the transition zone is visible as “Y”‐shaped bridges with two branches attached to the ciliary membrane and a single anchor at each microtubule doublet (Figure 5.1). Subdiffraction imaging strategies have been employed to reconcile the parts list with the visible structure, and to begin to attribute functions to individual components. Three key transition zone protein complexes were named in part after the diseases caused by

52

THE LIVER:  CILIUM COMPOSITION AND ORGANIZATION

Tubulin dimer

kin

es

in

(+)

IFT-A IFT-B

Microtubule doublet

dynein

IFT-B IFT-A

kinesin

Primary cilium dynein

Y-link

Intraflagellar transport (IFT) Transition zone

Ciliary pocket

Basal body (Mother centriole)

Distal appendages 200 nm

Figure 5.1  Primary cilia architecture and intraflagellar transport. The cross‐sections show the structural changes as the microtubules extend from the centrosome (described from bottom to top). The drawings are based on EM images in [8]. At the centrosome there are nine microtubule triplets. Distal appendage proteins appear like a pinwheel emanating from the centriole microtubules. At the transition zone, Y‐links are visible. The doublet microtubules of the ciliary axoneme are extensions of two of the triplet centriolar microtubules. Toward the distal end, the axoneme doublets can rearrange and terminate. The doublets are composed of α and β tubulin dimers that assemble a closed cylinder (A tubule) and a second cylinder that closes on the outside of the first (B tubule). On the right is a representation of intraflagellar transport (IFT). The IFT complex is a large multi‐subunit complex that can be biochemically separated into the IFT‐A and IFT‐B subcomplexes. The IFT‐A subcomplex interacts with dynein and facilitates retrograde transport along the A tubule. For anterograde transport toward the plus end of the microtubules at the cilium tip, the IFT‐B subcomplex associates with kinesin‐2. IFT complexes associate with cargo in both directions. The retrograde motor, dynein, is an anterograde cargo, but the kinesin motor dissociates at the tip and diffuses back to the base. Because the time for kinesin‐2 to return to the base is length‐dependent, the concentration of kinesin‐2 has been proposed as a length regulator. Adapted from Chien 2017 and Hendel 2018 [109, 110].

mutation of the complex. MKS and NPHP complex proteins create the Y‐links while Cep290 localizes adjacent to the Y‐ links, proximal to the centrosome [19, 20]. How stable are the Y‐link structures? Mating experiments in Chlamydomonas revealed NPHP4 is static, but CEP290 can exchange between transition zones of different flagella [21, 22]. Researchers found no fluorescent recovery within 30 min after photobleaching green fluorescent protein‐tagged MKS complex proteins indicating that the proteins do not exchange with the unbleached pool [23]. These results suggests that the MKS and NPHP complexes form stable structures, while the pool of CEP290 may be more dynamic.

The structure of the transition zone cannot yet explain its function as a barrier capable of selective transport. It has been shown, however, that depletion of transition zone proteins that cause disease permit typically excluded proteins into the cilium [22]. In addition to restricting entry, the transition zone can also limit exit and may function as a staging platform because several dynamic cilia proteins have been found to accumulate above this region [24, 25]. To pass through the barrier, many proteins utilize escorts such as nuclear import effectors, the BBSome complex, and tubby family proteins (the latter two are discussed in the section later on ciliary signaling).



5:  Primary Cilia (c)

(a)

CEP164

53

FBF1

Merge

200 nm

(d)

FBF1 and SCLT1

CEP164 and SCLT1

5 m anti-acetylated tubulin

200 nm anti-detyrosinated tubulin

Hoecst

(b)

(e) DAM

Y-links

Ciliary pocket

DABs

xy xz

5 m cilia-enriched GPCR centrosome

CEP164 FBF1 SCLT1 CEP89 CEP83 CP110

T TBK2 IFT88

EHD1

Arl13B

Figure 5.2  Conventional and super‐resolution views of primary cilia. (a) Microtubules within cilia are post‐translationally modified. These cilia on Madin–Darby canine kidney (MDCK) cells have been stained with antibodies that recognize acetylated and detyrosinated tubulin and visualized using confocal microscopy. The image is a maximum intensity projection of a z‐stack. The inset images show that both acetylated and detyrosinated tubulin are found along the entire length, however the distributions of the two are not identical. (b) Cilia‐enriched GPCR localization. Live imaging of NIH3T3 fibroblast cell shows the concentration of the GPCR adjacent to the centrosome in cells expressing the GPCR, melanin concentrating hormone receptor 1 (tagged with the fluorescent protein tdTomato), and the centrosome localization sequence of pericentrin (PACT; tagged with tag‐blue fluorescent protein). The upper image is an xy projection; the narrow, lower panel is an xz projection of the same cell. (c–e) Super‐resolution imaging brings together protein localization and structure to reveal organization at the base of the cilium. (c) Two‐color axial view dSTORM images of distal appendage proteins, FBF1 and CEP164. This approach was used to calculate the relative angular positions of different distal appendage proteins (represented by symbols in the far right image). From [92]. (d) Two‐color lateral view dSTORM images of FBF1, SCLT1, and CEP164 (three representative images of each). The relative height of FBF and SCLT1 are designated by the arrowheads on the far left. From [92]. (e) Three‐dimensional computational model created by combining information about the relative axial and lateral positions of distal appendage proteins and several ciliary proteins. DAB, dorsal appendage blade; DAM, dorsal appendage matrix. From Yang 2018 https://www.nature.com/ articles/s41467‐018‐04469‐1#rightslink. Licensed under CCBY 4.0 [92].

The mechanism remains to be elucidated, but several lines of evidence suggest that entry and exit of ciliary matrix proteins is related to gating at the nuclear pore. Small soluble proteins can freely enter the cilium, but, like the nuclear pore that excludes proteins above ~50 kDa, large proteins require active transport into the primary cilium [26]. Several fluorescently labeled nucleoporin (NUP) proteins localize to the base of the primary cilium and disrupting their function alters the transport of proteins into and out of cilia [27, 28]. The NUPs found in cilia are not the membrane or inner, nuclear basket proteins, but rather NUPs that contain phenylalanine–glycine repeats that are thought to create the permeability barrier. To pass through the nuclear pore, cargo with import or export signals bind to transport adaptor proteins, such as importin β. Adaptor binding and release on the opposite side of the nuclear envelope is facilitated by a gradient of RanGTP inside the nucleus and RanGDP in the

cytoplasm. Like the nucleus, the cilium is also rich in RanGTP and the nuclear import adaptor importin β binds to ciliary localization sequences and can facilitate delivery of soluble proteins through the boundary at the base of the cilium [29]. Because primary cilia lack structures that resemble nuclear pores, it is not yet clear where NUP proteins assemble or how they create a permeability barrier. Although the full definition and diversity of ciliary lipids is not known, several lines of evidence indicate that, like the protein composition, the lipid environment of cilia is unique and actively regulated. Early studies revealed that cilia membranes are generally cholesterol rich, however, the transition zone may have less cholesterol [30, 31]. Cilia have condensed membranes and ciliogenesis requires phosphatidylinositol 4‐phosphate adaptor protein‐2 (FAPP2) [32]. One of the causes of the ciliopathy Joubert syndrome is mutation of INPP5E, a ciliary

54

THE LIVER:  CILIA‐MEDIATED SIGNALING

phosphoinositide 5‐phosphatase that localizes along the length of the cilium, where it facilitates the conversion of phosphoinositide 4,5‐bisphosphate (PI(4,5)P2) to phosphatidylinositol 4‐phosphate (PI(4)P) [33–36]. While adjacent membranes are rich in PI(4,5)P2, INPP5E localization to cilia creates a PI(4)P‐ rich membrane. Targeting of INPP5E and other farnesylated proteins to cilia is thought to be mediated by release of cargo from the prenyl‐binding protein, phosphodiesterase 6 delta subunit, mediated by Arl3 at the cilium [37]. Disrupting the transition zone permits PI(4,5)P2 phospholipids to enter ­primary cilia [38]. As mentioned earlier, the volume of the cilium is a fraction of the volume of the rest of the cell and it is membrane bound except at the base. Although ions can pass into and out of the base of the cilium, transport of ions through ciliary membrane channels helps maintain the unique ionic environment inside cilia. The resting membrane potential of cilia on cultured human retina pigment epithelial cells is ~30 mV more positive than the resting potential of the plasma membrane of the rest of the cell [39]. The concentration of Ca2+ in cilia is approximately seven times higher than in the cytoplasm of cultured human pigment epithelial cells [39]. While the cytoplasm and cytoplasmic organelles serve as a sink for Ca2+, channels in the ciliary membrane may import enough Ca2+ to maintain a concentration greater than 500 nM. Ciliary ion channels are essential components of the vision and olfaction signal transduction pathways. Several other ciliary ion channels, including polycystin 2 (PKD2) and PKD2‐like 1 (PKD2‐L1), are members of the transient receptor potential (TRP) family of cation channels. Homomeric PKD2 conducts Na+ and K+ and is impermeable to Ca2+, while PKD2‐L1 is permeable to Na+, K+, and Ca2+ [40, 41]. The ionic environment defined by the channels contributes to cell and organismal health. Mutations in PKD2 cause autosomal dominant polycystic kidney disease (ADPKD) and deletion of the gene causes embryonic lethality. Without PKD2‐L1, mice have reduced responses to hedgehog signaling and mild hedgehog‐related developmental phenotypes [39]. It is not yet clear yet which cilia proteins are influenced by the unique ionic environment. Metabolites can pass between the cilium and the cytoplasm; however, evidence suggests that active processes can sustain different metabolite concentrations in cilia relative to the cytoplasm as a whole. Several isoforms of adenylyl cyclase, which generates cyclic adenosine monophosphate (cAMP) through a cyclizing reaction of ATP, localize to primary cilia [39, 40]. cAMP concentrations in primary cilia are ~5 times higher than in the cell body and the levels can be altered through ciliary signaling pathways [42]. Ciliary Ca2+ concentration also regulates adenylyl cyclase activity, and thus could regulate ciliary signal responses.

CILIA‐MEDIATED SIGNALING Primary cilia are positioned to sample the extracellular environment near the cell. The ability to detect defined molecules and signal their presence to the rest of the cell is a feature primary cilia share with motile cilia and flagella. Sensory functions may

have co‐evolved with motility in early organisms as suggested by a study that revealed several cilia signaling pathway components in organisms that are thought to share an ancient early ancestor with modern animals [43]. Olfaction and vision are both sensory functions mediated by primary cilia [44]. Rod outer segments are modified cilia filled with membrane disks containing rhodopsin, a G protein‐ coupled receptor (GPCR). Smell is the consequence of an action potential initiated by ligand binding to olfactory GPCRs on the cilia of olfactory neurons. Stereocilia, which detect sound during hearing, are actin‐based structures and therefore are not true cilia. The hair cells that project stereocilia do have a true cilium, called a kinocilium. Although the kinocilium is not required for mechanosensation of sound, mutations that prevent cilia formation on hair cells disrupt proper stereocilia organization possibly through failures to signal polarity cues during development [45]. Sensing and signaling performed by primary cilia are essential for differentiation and tissue morphogenesis during development. Mutations that cause systemic loss of primary cilia result in embryonic lethality accompanied by patterning defects such as left–right asymmetry [46, 47]. Loss of cilia in adult animals leads to kidney cyst formation and obesity due to abnormal appetite signaling [48]. Engineered loss of cilia in defined cell types causes many additional phenotypes (see, for example, [49–51]). Some ciliary proteins may contribute to signaling in cancer [52] and signaling through primary cilia also affects response to injury [52, 53]. Primary cilia in diverse tissues do not participate in a single, stereotyped signaling pathway. Rather, the expression and localization of receptors and downstream effectors varies by both cell type and life stage. There is a growing list of genes that contribute to ciliopathies; mutation of many of which alter signal perception or transmission [1]. A key paradigm for ciliary signaling is through GPCRs. Figure 5.3 summarizes some of the potential cilia GPCR signal transduction outcomes described below. Similar to rhodopsin and olfactory GPCRs, cilia targeting of many GPCRs is so efficient that the concentration of the GPCR in the ciliary membrane far exceeds the receptor concentration anywhere else in the cell (Figure  5.2b). Sequence features on the intracellular face GPCRs contribute to ciliary targeting [53–55]. These features are recognized by tubby (Tub) family proteins including Tub and tubby‐like protein 3 (TULP3). Tub and TULP3 associate with the plasma membrane through binding PI(4,5)P2, where they capture GPCRs with cilia localization sequences. Tub and TULP3 binding to IFT‐A facilitates cilia delivery. Within the cilium there is little PI(4,5)P2, so Tub and TULP3 dissociate from the membrane and GPCR cargos are released [56]. Loss of Tub, which is predominantly expressed in the brain, prevents cilia localization of GPCRs critical for appetite regulation and results in obesity [57]. Delivery of membrane proteins, including GPCRs and ion channels, to the base of the cilium can be mediated by lateral passage from the plasma membrane, direct targeting of Golgi‐ derived vesicles or indirect trafficking through endosome‐ derived vesicles [58]. Bardet–Biedl syndrome (BBS) is a ciliopathy caused by mutation in proteins that form a complex called the BBSome [59, 60]. GPCRs are absent from cilia of



5:  Primary Cilia

55

(a)

Without ligand ? Rab8 BBSome

Tub

?

GPCR Adenylyl cyclase

PI(4)P GDP

Tub

PKA Kinase target Ion channel

Upon ligand binding

(b) Ligand

Adenylyl cyclase

PI(4)P

+

cAMP

GTP

cAMP P P

cAMP

PKA

BBSome

Possible consequences of ligand binding G Protein

ATP

Adenylyl cyclase GPCR

GTP

GDP

cilium

or

cAMP

exit with BBSome

Kinases

inactive

ACTIVE

Channels

closed

open

Figure 5.3  Generalized perspective of signaling through primary cilia. (a) Many GPCRs localize to primary cilia. Trafficking of GPCRs to primary cilia is mediated by Rab8 and the BBSome. Tub family proteins escort GPCRs into primary cilia and release them when they reach the PI(4)P‐rich membrane (orange). In the absence of ligand, GPCRs can associate with G proteins. (b) Ligand binding to GPCRs can have many potential consequences. Activated receptors can trigger replacement of Gα‐associated GDP with GTP. As a result, the GTP‐bound Gα subunit dissociates from the G protein β/γ subunits and both propagate signaling cascades. One possibility shown here is that Gα can stimulate adenylyl cyclase, which produces cAMP. cAMP can activate kinases such as PKA and could affect ion channel activity. Receptor activation also leads to BBSome‐dependent internalization of GPCRs that may reset signaling pathways. As indicated at the bottom, signaling cascades could also have the opposite effects, such as decreasing cAMP production, inactivating kinases, or closing channels.

primary cells cultured from animals lacking individual BBS proteins [61]. The BBSome complex forms a coat structure that can interact directly with both GPCRs and secretory pathway directors such as Rab8 and facilitate delivery of GPCRs to the cilium. Ciliary GPCRs bind diverse ligands including bile acids, nucleotides, neuropeptides, and proteins [62]. Upon ligand binding, the receptors facilitate replacement of Gα‐associated GDP with GTP. The trimeric G protein then dissociates from the receptor and the Gα and Gβγ subunits both initiate downstream signaling events. There are multiple isoforms of the G protein

subunits and, in general, different GPCRs are thought to recruit G proteins containing a specific subtype. However, the GPCR TGR5 associates with a different Gα isoform in ciliated versus unciliated cells [63]. Ciliary G proteins can activate different, and sometimes opposing, downstream effector pathways. For example, once liberated, stimulatory Gα subunits (Gαs) increase adenylyl cyclase activity while inhibitory Gαi subunits decrease cAMP levels. cAMP acts as a second messenger and promotes activity of downstream kinases including protein kinase A (PKA). Ligand binding leads to retrieval of activated GPCRs from the cilium [64–67]. Although additional signaling of

56

THE LIVER:  BUILDING, MAINTAINING, AND DISMANTLING CILIA

internalized GPCRs has been demonstrated in other contexts [68], no activity has been reported for GPCRs internalized from cilia. Internalization of activated receptors could help reset signaling pathways. The BBSome, which is implicated in trafficking proteins to cilia, also facilitates exit of activated GPCRs through the transition zone in complex with the small G protein Arl6 [65]. Upon receptor activation, Gαi signaling, which reduces adenylyl cyclase and thus PKA activity, causes BBSome proteins, Arl6, and retrograde IFT proteins to accumulate at the distal end where the proteins form large IFT trains, which then bind to activated GPCRs, transit to the base of the cilium and escort the GPCR through the transition zone [69]. An alternate mechanism to remove GPCRs and reset signaling pathways is through direct release of the GPCRs in cilia‐derived extracellular vesicles (EVs) [70]. Rhodopsin and the membrane disks of photoreceptor cells are processively removed from the distal end of the outer segment [71]. EVs containing primary cilia proteins have also been collected in urine [72]. Actin, which is generally not polymerized in cilia, can drive release of cilia membranes in response to increased PI(4,5)P2 when INPP5E is disabled or removed from cilia [73, 74]. Many critical signaling pathways center on primary cilia [75]. Obesity generated in ciliopathies is caused by disruptions in satiety signaling through neuropeptide‐binding GPCRs in hypothalamus cilia. The hedgehog signaling pathway is critical for developmental and can contribute to cancer. During chronic liver disease, hedgehog signaling in liver progenitor cells contributes to regeneration [76]. Most components and effectors of this complicated signaling cascade function in cilia. In the cilium, the activity of the hedgehog receptor Patched influences the ciliary localization of the GPCR Smoothened. The activity of Smoothened in the cilium affects processing of Gli transcription factors. Another GPCR, Gpr161, also localizes to the cilium and mediates hedgehog signaling [77]. Although it is clear that ligand binding to receptors in the unique ciliary environment initiates signaling cascades, how the physiological consequences of these cascades propagate and cause changes in the rest of the cell is generally less clear. An exception is the hedgehog signaling pathway: the Gli family transcription factors localize to primary cilia and can be processed and trafficked out of the cilium in response to pathway activation [78, 79]. There are many other possible outcomes in addition to transcription alterations; however, all of these are subject to the same limitation that Delling and colleagues revealed regarding Ca2+: upon exiting, soluble molecules will be significantly diluted [5]. Delivery of cilia effectors into the nucleus could be streamlined by the overlapping components of the cilia and nuclear import/export pathways. As the centrosome serves as both the basal body and the microtubule organizing center, it is possible that molecules exiting cilia have a transient local concentration high enough to transmit signals to proteins recruited to the centrosome. Association with carrier molecules or membrane vesicles could be an additional strategy. cAMP and cGMP generated by signaling cascades can stimulate ion channel activity and alter ciliary Ca2+ levels; this is a

critical part of signal transduction in olfaction and vision [80, 81]. Ciliary ion channels may have multiple effects on signaling cascades. Adenylyl cyclase 3 activity is affected by Ca2+ [82] and it has been hypothesized that changes in intracellular Ca2+ in cilia might alter downstream signaling components such as adenylyl cyclase [83]. In animal models of PKD, tissue levels of cAMP are elevated [84]. Activation of ion channels could, if the paradigm is true, exert global effects on any ciliary proteins that are sensitive to changes in ion concentrations. Such a scenario might explain why PKD2‐L1‐deficient mice have reduced responses to Shh [39]. It is not clear whether cilia act as mechanosensitive flow sensors. The prevailing paradigm that primary cilia bending communicates the presence of external fluid flow via mechanosensitive activation of calcium ion channels in the ciliary membrane has been reconsidered. Detailed studies using a fluorescent Ca2+ sensor localized to primary cilia revealed no changes in ciliary Ca2+ levels upon deflection with flow from a micropipette [5]. If cilia are mechanosensitive, the signal transduction does not appear to be mediated by changes in ciliary calcium. In cartilage, chondrocyte cilia participate in mechanotransduction, not as a mechanosensor, but as a chemical signal transducer. Receptors in chondrocyte cilia membranes bind ATP that is released by the cell body upon compression [85]. Although signal reception in cilia is compared to radio signal reception using antennae, cilia are likely capable of being more than just a passive sensor. It is possible that, like other EVs, cilia‐derived vesicles may be biologically active. Chlamydomonas releases vesicles from flagella that facilitate breakdown of the wall encapsulating the daughter cells after mating [86]. In addition to releasing particles, primary cilia may participate in signaling through direct touch. Cholangiocyte cilia within the bile duct and other cilia that project into tubules can adhere to one another [87], which might facilitate direct intracellular communication. Primary cilia in many other tissues are not in a lumen, but rather, can be enmeshed in extracellular matrix, project between layers of cells, or entwined with other extracellular processes (examples include cilia on chondrocytes, mammary myoepithelial cells, and neurons) [88, 89]. In deep tissue, it is possible that primary cilia could signal through direct contact.

BUILDING, MAINTAINING, AND DISMANTLING CILIA As cells cycle and differentiate, primary cilia are deconstructed and recreated. When maintained, cilia length can vary by an order of magnitude between cell types (from ~2 to >20 μm) but is typically uniform within a population. The presence and length of a cilium define the ability of a cell to detect and respond to extracellular cues, so the following questions are relevant to understanding the role of cilia in health and development: How do cilia form and how is biogenesis regulated? How is cilium length determined and maintained? When and how are cilia removed from the cell surface? Partial answers to these questions are discussed below.



5:  Primary Cilia

Ciliogenesis How is the critical and unique ciliary environment created? Primary cilia can grow from a centriole deep in the cytoplasm or docked at the plasma membrane (Figure  5.4a,b; reviewed in [90]). In both pathways microtubule doublets extend from the triplet microtubules of the centriole, membrane is added, then presumably differentiated as the transition zone forms. The internal pathway requires the recruitment of vesicles that anchor and fuse to form a large ciliary vesicle docked to the mother centriole. As microtubules extend from the centriole, membrane is added and the ciliary vesicle deforms to ensheath the nascent cilium. The final stage of internal ciliogenesis requires the membrane around the tip of the growing cilium to fuse with the plasma membrane. Ciliogenesis at the plasma membrane is thought to begin when a centriole docks at the plasma membrane. Membrane is added at the plasma membrane as the microtubules extend to ensheath the cilium. The membrane‐associated GTPases Rab11 and Rab8 target vesicles to nascent cilia to provide both lipids and key ciliary membrane proteins [91]. Every centriole is not capable of becoming a basal body. Structures called distal appendages or transition fibers must be added to the distal end of the ninefold symmetric centriole triplet microtubules to recruit vesicles or dock at the plasma membrane. Several components assemble to form distal appendages, and recent super‐resolution imaging efforts have revealed the organization of many key distal appendage proteins (Figure 5.2) [92]. CEP164 and SCLT1 are components of the distinctive blade structure that is visible in electron micrographs. In contrast, FBF1 localizes to the distal appendage matrix – a space that appears empty in electron micrographs. Distal appendage proteins contribute to several aspects of ciliogenesis. Studies of initiation steps in intracellular ciliogenesis [93] revealed that fusion of small vesicles recruited to distal appendages requires membrane tubulating proteins: EHD1 and EHD3. Subsequently, CEP164 in the distal appendage blade participates in both microtubule growth and membrane expansion (reviewed in [94]). CEP164 recruits tau tubulin kinase 2 (TTBK2), a protein required for a key transformation at the centrosome – removal of proteins that cap the plus ends of the centrosome microtubules. TTBK2 facilitates removal of CP110 from the distal end of the centriole. Although it must be removed prior to ciliogenesis, mice with no CP110 have fewer cilia. This suggests that CP110 may be required for early steps in centrosome remodeling and then be removed prior to microtubule extension [95]. In the internal ciliogenesis pathway CEP164 interactions with both Rab8a and its GTPase‐exchange factor Rabin 8 are thought to promote membrane docking of early ciliary membranes [96]. Targeting and transport of ciliary components is also essential for cilia formation. Mutation or deletion of IFT‐B complex proteins, which facilitates anterograde transport, disrupts cilia formation. FBF1 in the distal appendage matrix colocalizes with IFT88 and facilitates entry of IFT complexes into cilia [92, 97]. Members of the IFT‐B complex bind tubulin and may escort it through the permeability barrier where it is concentrated in the cilium [28, 98]. Interference with the formation of centriolar satellites, which are thought to function as a protein complex assembly platform, can also prevent cilia formation. Several

57

cilia proteins, including BBS4 and CEP290, associate with the key scaffolding protein pericentriolar material‐1 (PCM‐1) in centriolar satellites. Impaired autophagy can also prevent ciliogenesis because oral–facial–digital syndrome 1 proteins are not degraded but instead accumulate at centriolar satellites [99]. The two centrioles inside each cell are different and only one of them undergoes remodeling to become the basal body. During each cell division cycle, the two centrioles duplicate and then separate to form the spindle pole. Each daughter inherits a pair of centrioles, a duplicate (called the daughter) and an older original (called the mother). The centriole that becomes the basal body is always the older centriole. Cilia formation can be different between sister cells because the two mother centrioles they inherit are not equivalent. In the preceding cell cycle one of the mother centrioles had been a daughter. The sister cell that inherits the older mother centriole forms a cilium earlier, which could influence perception of extracellular developmental cues and downstream differentiation [100]. The older mother centriole may be faster because it already has distal appendages and can retain association of ciliary membranes [101]. Many polarized cells build cilia at the apical plasma membrane and several lines of evidence indicate that polarization may position and promote ciliogenesis (reviewed in [102]). Proteins that localize to the boundary between apical and basolateral membranes in polarized epithelial cells also localize to the base of the primary cilium. Although how polarity proteins influence cilia formation and maintenance is unclear, their importance has been demonstrated. Deletion, mutation, or pharmacological disruption of polarity proteins prevents ciliogenesis. PAR complex proteins  –  PAR3, PAR6  –  and an atypical PKC (aPKC) associate with the anterograde kinesin‐2 component Kif3a. The lipid ceramide influences ciliogenesis and is required for association of aPKC with other PAR proteins [103]. Cdc42, another PAR complex component, interacts with and promotes the ciliary localization of exocyst proteins. The exocyst complex facilitates targeting and tethering of post‐Golgi vesicles and can interact with Rab8 to direct trafficking in polarized cells [104]. Ciliogenesis at the plasma membrane may be facilitated in some way by the midbody remnant (Figure  5.4b). After the cytokinetic bridge is severed during cell division, part of the bundle of microtubules and membrane that made up the bridge can remain associated with the plasma membrane of one of the daughter cells. Movement of the remnant across the apical surface of the cell to where the centriole is docked precedes cilia growth. Although cilia and midbodies share many molecular components, it is not clear if any of these are directly transferred to contribute to cilia growth [105].

Maintaining or altering cilia length Cilia length is typically constant across a population but can vary across cell types. This suggests that cellular parameters such as protein expression levels or turnover rates influence the cilia length setpoint. Several factors that alter the cilia length setpoint have been identified. For example, tubulin concentrations and modifications can affect cilia length [28, 106, 107]. In addition, proteins involved in IFT are known to affect cilia length: increased expression of IFT‐B components yields longer

58

THE LIVER:  BUILDING, MAINTAINING, AND DISMANTLING CILIA

(a)

Ciliary vesicle

Distal appendages Centriole

(b)

Midbody remnant Transition zone

?

(c)

? Dissasembly

Severing

Resorption

Figure 5.4  Ciliogenesis and cilia elimination. (a) Internal ciliogenesis begins with maturation of the mother centriole. Although they are depicted as discrete stages, initial vesicle recruitment and axoneme extension may occur simultaneously. As the axoneme extends, the transition zone forms and additional membrane must be delivered. When the membrane encapsulating the cilium contacts the plasma membrane, the bilayers are thought to fuse so the membrane adjacent to the microtubules becomes continuous with the plasma membrane. The outer membrane of the ciliary vesicle may become part of the ciliary pocket membrane. (b) Ciliogenesis at the plasma membrane requires formation of distal appendages and movement of the mother centriole to the cell surface. It is not currently clear how the midbody remnant promotes cilia assembly at the cell surface. Like in the internal pathway, the transition zone forms and microtubule growth must be coupled to membrane addition to extend the nascent cilium. (c) Three ­possibly strategies that could all contribute to cilium disappearance are depicted. Microtubule disassembly may be stimulated by activation of the tubulin deacetylase HDAC6. Membrane removal via endocytosis could also contribute. Severing of ciliary membrane is known to occur and has been associated with cilium removal. A third strategy involves dissolving the barriers that separate the ciliary cytoplasm and membrane from the rest of the cell. There has been some evidence to support each possibility.

cilia [108]. IFT‐B delivery of tubulin to the tip may increase the local concentration of tubulin and drive polymerization. Kinesin‐2 diffuses back to the base of the cilium after it carries IFT trains to the tip. Because diffusion of kinesin‐2 back to the base of the cilium where it can form new anterograde trains is length dependent, kinesin‐2 concentration has been proposed as a key regulator of cilia length [109, 110]. The actin cytoskeleton can also influence cilia length [111]. Cilia length increases in cells when F‐actin is destabilized by treatment with cytochalasin D. Similar increases in cilia length have been reported by depleting Arp3, which nucleates F‐actin branching and by treatment with a microRNA that downregulates

four positive regulators of branched F‐actin [112]. Cilia elongation upon treatment with cytochalasin D is accompanied by an influx of membrane‐associated actin‐binding proteins in cilia [113]. Cdc42 and aPKC, components of the PAR complex, participate in signaling to influence F‐actin [114]. How disrupting branched F‐actin affects cilia length is not yet clear. Perhaps the actin motor myosin Va, which can bind Rab8 and a transition zone protein [115], participates in actin regulation of cilia length. Several signaling pathways and disease conditions alter cilia length. Inhibition of adenylyl cyclase by treatment with lithium causes cilia elongation [116]. In physiological contexts, adenylyl cyclase activity can be stimulated or inhibited by Gα proteins



5:  Primary Cilia

upon activation of ciliary GPCRs. It is not yet clear whether cAMP as a second messenger directly influences a length regulating pathways described above, or if it promotes cilia extension through other binding partners. Another candidate downstream of signaling pathways that may also influence cilia length is the proteasome, which associates with a transition zone protein and cleaves Gli transcription factors [117].

Cilia elimination While some organisms retain cilia or flagella, mammalian cells disassemble cilia in G1 phase of the cell cycle or at the beginning of mitosis. Cilia removal could be achieved by: (i) severing; (ii) disassembly; and (iii) resorption upon decommissioning the boundaries that segregate the cilium (Figure 5.4c). The three mechanisms could work together and there is evidence to suggest that all three processes can happen [93]. Electron microscopy of epithelial cells provided early support for the model that cilia are resorbed into the cytoplasm [118]. Stimulated release of vesicles from the tip can precipitate cilia disassembly [74]. Aurora A kinase, which is regulated by many factors and functions as a part of the cell cycle, phosphorylates both INPP5E and the deacetylase HDAC6. HDAC6 destabilization of acetylated microtubules may facilitate axoneme deconstruction. The activity of microtubule‐depolymerizing kinesins associated with the axoneme may also contribute to reducing axoneme length. It is not clear if there is a role for endocytosis in recovering membrane from cilia. Polo‐like kinase 1, a kinase regulated by the cell cycle, promotes cilia disassembly and phosphorylates a component of the transition zone, possibly disrupting the integrity of the ciliary gate (see [119]).

CONCLUSION Because of their distinct structure in electron microscopy, primary cilia were identified on many cell types throughout the twentieth century. A small community of researchers built foundational knowledge about the structure and function of primary cilia. The discovery that proteins move along cilia, followed by insights into the many links between primary cilia function and health stimulated expansion of cilia research. Investigators in disparate fields have found that cilia biology is central to many aspects of cell signaling and development. These, in turn, have provided new insights into fundamental aspects of cilia biology. Technological advances, including super‐resolution microscopy, have provided new insights into the movement and molecular architecture of cilia. However, many exciting questions remain. Emerging studies will likely provide additional insights into how defects in cilia form and function contribute to disease and will hopefully lead to novel interventions to alleviate the symptoms of ciliopathies.

ACKNOWLEDGMENT The author thanks Christine Kettenhofen, Corey Valinsky, and Shu‐Hsien Sheu for helpful suggestions on the text.

59

REFERENCES 1. Reiter, J.F. and Leroux, M.R. Genes and molecular pathways underpinning ciliopathies. Nat Rev Mol Cell Biol, 2017;18(9):533–47. 2. Battle, C., Ott, C.M., Burnette, D.T. et al. Intracellular and extracellular forces drive primary cilia movement. Proc Natl Acad Sci U S A, 2015;112(5): 1410–15. 3. Yeh, C., Li, A., Chuang, J.‐Z. et al. IGF‐1 activates a cilium‐localized noncanonical Gβγ signaling pathway that regulates cell‐cycle progression. Dev Cell, 2013;26(4):358–68. 4. Li, Q., Montalbetti, N., Wu, Y. et al. Polycystin‐2 cation channel function is under the control of microtubular structures in primary cilia of renal epithelial cells. J Biol Chem, 2006;281(49):37566–75. 5. Delling, M., Indzhykulian, A.A., Liu, X. et al. Primary cilia are not calcium‐ responsive mechanosensors. Nature, 2016;531(7596):656–60. 6. Nachury, M.V. How do cilia organize signalling cascades? Philos Trans R Soc Lond B Biol Sci, 2014;369(1650):20130465. 7. Wloga, D., Joachimiak, E., Louka, P., and Gaertig, J. Posttranslational modifications of tubulin and cilia. Cold Spring Harb Perspect Biol, 2017;9(6):a028159. 8. Gluenz, E., Höög, J.L., Smith, A.E. et al. Beyond 9+0: noncanonical axoneme structures characterize sensory cilia from protists to humans. FASEB J, 2010;24(9):3117–21. 9. Lechtreck, K.F. IFT–cargo interactions and protein transport in cilia. Trends Biochem Sci, 2015;40(12):765–78. 10. Kozminski, K.G., Johnson, K.A., Forscher, P., and Rosenbaum, J.L. A motility in the eukaryotic flagellum unrelated to flagellar beating. Proc Natl Acad Sci U S A, 1993;90(12):5519–23. 11. Cole, D.G., Diener, D.R., Himelblau, A.L. et al. Chlamydomonas kinesin‐II‐ dependent intraflagellar transport (IFT): IFT particles contain proteins required for ciliary assembly in Caenorhabditis elegans sensory neurons. J Cell Biol, 1998;141(4):993–1008. 12. Stepanek, L. and Pigino, G. Microtubule doublets are double‐track railways for intraflagellar transport trains. Science, 2016;352(6286):721–4. 13. Benmerah, A. The ciliary pocket. Curr Opin Cell Biol, 2012;25(1):78–84. 14. Gonçalves, J. and Pelletier, L. The ciliary transition zone: finding the pieces and assembling the gate. Mol Cells, 2017;40(4):243–53. 15. Jana, S.C., Mendonça, S., Machado, P. et al. Differential regulation of transition zone and centriole proteins contributes to ciliary base diversity. Nat Cell Biol, 2018;20(8):928–41. 16. Sang, L., Miller, J.J., Corbit, K.C. et  al. Mapping the NPHP‐JBTS‐MKS protein network reveals ciliopathy disease genes and pathways. Cell, 2011;145(4):513–28. 17. Dean, S., Moreira‐Leite, F., Varga, V., and Gull, K. Cilium transition zone proteome reveals compartmentalization and differential dynamics of ciliopathy complexes. Proc Natl Acad Sci U S A, 2016;113(35):E5135–43. 18. Diener, D.R., Lupetti, P., and Rosenbaum, J.L. Proteomic analysis of isolated ciliary transition zones reveals the presence of ESCRT oroteins. Curr Biol, 2015;25(3):379–84. 19. Yang, T.T., Su, J., Wang, W.‐J. et  al. Superresolution pattern recognition reveals the architectural map of the ciliary transition zone. Sci Rep, 2015;5(1):14096. 20. Williams, C.L., Li, C., Kida, K. et al. MKS and NPHP modules cooperate to establish basal body/transition zone membrane associations and ciliary gate function during ciliogenesis. J Cell Biol, 2011;192(6):1023–41. 21. Craige, B., Tsao, C.‐C., Diener, D.R. et al. CEP290 tethers flagellar transition zone microtubules to the membrane and regulates flagellar protein content. J Cell Biol, 2010;190(5):927–40. 22. Awata, J., Takada, S., Standley, C. et al. NPHP4 controls ciliary trafficking of membrane proteins and large soluble proteins at the transition zone. J Cell Sci, 2014;127(Pt 21):4714–27. 23. Lambacher, N.J., Bruel, A.‐L., van Dam, T.J.P. et al. TMEM107 recruits ciliopathy proteins to subdomains of the ciliary transition zone and causes Joubert syndrome. Nat Cell Biol, 2016;18(1):122–31. 24. Shi, X., Garcia, G., III, Van De Weghe, J.C. et al. Super‐resolution microscopy reveals that disruption of ciliary transition‐zone architecture causes Joubert syndrome. Nat Cell Biol, 2017;19(10):1178–88. 25. Yang, T.T., Hampilos, P.J., Nathwani, B. et  al. Superresolution STED microscopy reveals differential localization in primary cilia. Cytoskeleton (Hoboken), 2012;70(1):54–65. 26. Takao, D. and Verhey, K.J. Gated entry into the ciliary compartment. Cell Mol Life Sci, 2016;73(1):119–27.

60

THE LIVER: REFERENCES

27. Kee, H.L., Dishinger, J.F., Blasius, T.L. et al. A size‐exclusion permeability barrier and nucleoporins characterize a ciliary pore complex that regulates transport into cilia. Nat Cell Biol, 2012;14(4):431–7. 28. Endicott, S.J. and Brueckner, M. NUP98 Sets the size‐exclusion diffusion limit through the ciliary base. Curr Biol, 2018;28(10):1643–50.e3. 29. Dishinger, J.F., Kee, H.L., Jenkins, P.M. et al. Ciliary entry of the kinesin‐2 motor KIF17 is regulated by importin‐β2 and RanGTP. Nat Cell Biol, 2010;12(7):703–10. 30. Souto‐Padron, T. and de Souza, W. Freeze‐fracture localization of filipin‐ cholesterol complexes in the plasma membrane of Trypanosoma cruzi. J Parasitol, 1983;69(1):129. 31. Cuevas, P. and Gutierrez Diaz, J.A. Absence of filipin‐sterol complexes from the ciliary necklace of ependymal cells. Anat Embryol, 1985;172(1). 32. Vieira, O.V., Gaus, K., Verkade, P. et al. FAPP2, cilium formation, and compartmentalization of the apical membrane in polarized Madin‐Darby canine kidney (MDCK) cells. Proc Natl Acad Sci U S A, 2006;103(49):18556–61. 33. Garcia‐Gonzalo, F.R., Phua, S.C., Roberson, E.C. et al. Phosphoinositides regulate ciliary protein trafficking to modulate hedgehog signaling. Dev Cell, 2015;34(4):400–9. 34. Chávez, M., Ena, S., Van Sande, J. et  al. Modulation of ciliary phosphoinositide content regulates trafficking and sonic hedgehog signaling output. Dev Cell, 2015;34(3):338–50. 35. Bielas, S.L., Silhavy, J.L., Brancati, F. et al. Mutations in INPP5E, encoding inositol polyphosphate‐5‐phosphatase E, link phosphatidyl inositol signaling to the ciliopathies. Nat Genet, 2009;41(9):1032–6. 36. Jacoby, M., Cox, J.J., Gayral, S. et al. INPP5E mutations cause primary cilium signaling defects, ciliary instability and ciliopathies in human and mouse. Nat Genet, 2009;41(9):1027–31. 37. Fansa, E.K., Kösling, S.K., Zent, E. et  al. PDE6δ‐mediated sorting of INPP5E into the cilium is determined by cargo‐carrier affinity. Nat Commun, 2016;7:11366. 38. Jensen, V.L., Li, C., Bowie, R.V. et al. Formation of the transition zone by Mks5/Rpgrip1L establishes a ciliary zone of exclusion (CIZE) that compartmentalises ciliary signalling proteins and controls PIP2 ciliary abundance. EMBO J, 2015;34(20):2537–56. 39. Delling, M., DeCaen, P.G., Doerner, J.F. et al. Primary cilia are specialized calcium signalling organelles. Nature, 2013;504(7479):311–14. 40. Shen, P.S., Yang, X., DeCaen, P.G. et al. The structure of the polycystic kidney disease channel PKD2 in lipid nanodiscs. Cell, 2016;167(3):763–73. e11. 41. Liu, X., Vien, T., Duan, J. et al. Polycystin‐2 is an essential ion channel subunit in the primary cilium of the renal collecting duct epithelium. Elife, 2018;7:E2363. 42. Moore, B.S., Stepanchick, A.N., Tewson, P.H. et al. Cilia have high cAMP levels that are inhibited by Sonic Hedgehog‐regulated calcium dynamics. Proc Natl Acad Sci U S A, 2016;113(46):13069–74. 43. Sigg, M.A., Menchen, T., Lee, C. et al. Evolutionary proteomics uncovers ancient associations of cilia with signaling pathways. Dev Cell, 2017;43(6):744–62.e11. 44. Singla, V. and Reiter, J.F. The primary cilium as the cell’s antenna: signaling at a sensory organelle. Science, 2006;313(5787):629–33. 45. Sipe, C.W. and Lu, X. Kif3a regulates planar polarization of auditory hair cells through both ciliary and non‐ciliary mechanisms. Development, 2011;138(16):3441–9. 46. Huangfu, D., Liu, A., Rakeman, A.S. et al. Hedgehog signalling in the mouse requires intraflagellar transport proteins. Nature, 2003;426(6962):83–7. 47. Nonaka, S., Tanaka, Y., Okada, Y. et al. Randomization of left‐right asymmetry due to loss of nodal cilia generating leftward flow of extraembryonic fluid in mice lacking KIF3B motor protein. Cell, 1998;95(6):829–37. 48. Davenport, J.R., Watts, A.J., Roper, V.C. et al. Disruption of intraflagellar transport in adult mice leads to obesity and slow‐onset cystic kidney disease. Curr Biol, 2007;17 (18):1586–94. 49. Haycraft, C.J., Zhang, Q., Song, B. et al. Intraflagellar transport is essential for endochondral bone formation. Development, 2007;134(2):307–16. 50. Dinsmore, C. and Reiter, J.F. Endothelial primary cilia inhibit atherosclerosis. EMBO Rep, 2016;17(2):156–66. 51. Foerster, P., Daclin, M., Asm, S. et al. mTORC1 signaling and primary cilia are required for brain ventricle morphogenesis. Development, 2017;144(2):201–10. 52. Liu, H., Kiseleva, A.A., and Golemis, E.A. Ciliary signalling in cancer. Nat Rev Cancer, 2018;18(8):511–24.

53. Berbari, N.F., Johnson, A.D., Lewis, J.S. et  al. Identification of ciliary localization sequences within the third intracellular loop of G protein‐coupled receptors. Mol Biol Cell, 2008;19(4):1540–7. 54. Loktev, A.V. and Jackson, P.K. Neuropeptide Y family receptors traffic via the Bardet‐Biedl syndrome pathway to signal in neuronal primary cilia. Cell Rep, 2013;5(5):1316–29. 55. Corbit, K.C., Aanstad, P., Singla, V. et al. Vertebrate Smoothened functions at the primary cilium. Nature, 2005;437(7061):1018–21. 56. Badgandi, H.B., Hwang, S.‐H., Shimada, I.S. et al. Tubby family proteins are adapters for ciliary trafficking of integral membrane proteins. J Cell Biol, 2017;216(3):743–60. 57. Sun, X., Haley, J., Bulgakov, O.V. et al. Tubby is required for trafficking G protein‐coupled receptors to neuronal cilia. Cilia, 2012;1(1):21. 58. Lu, L. and Madugula, V. Mechanisms of ciliary targeting: entering importins and Rabs. Cell Mol Life Sci, 2017;75(4):597–606. 59. Nachury, M.V., Loktev, A.V., Zhang, Q. et al. A core complex of BBS proteins cooperates with the GTPase Rab8 to promote ciliary membrane biogenesis. Cell, 2007;129(6):1201–13. 60. Khan, S.A., Muhammad, N., Khan, M.A. et al. Genetics of human Bardet– Biedl syndrome, an updates. Clin Genet, 2016;90(1):3–15. 61. Berbari, N.F., Lewis, J.S., Bishop, G.A. et al. Bardet‐Biedl syndrome proteins are required for the localization of G protein‐coupled receptors to primary cilia. Proc Natl Acad Sci U S A, 2008;105(11):4242–6. 62. Mykytyn, K. and Askwith, C. G‐Protein‐coupled receptor signaling in cilia. Cold Spring Harb Perspect Biol, 2017;9(9):a028183. 63. Masyuk, A.I., Huang, B.Q., Radtke, B.N. et al. Ciliary subcellular localization of TGR5 determines the cholangiocyte functional response to bile acid signaling. Am. J Physiol Gastrointest Liver Physiol, 2013;304(11):G1013–24. 64. Green, J.A., Schmid, C.L., Bley, E. et al. Recruitment of β‐arrestin into neuronal cilia modulates somatostatin receptor subtype 3 ciliary localization. Mol Cell Biol, 2016;36(1):223–35. 65. Ye, F., Nager, A.R., and Nachury, M.V. BBSome trains remove activated GPCRs from cilia by enabling passage through the transition zone. J Cell Biol. 2018;217(5):1847–68. 66. Pal, K., Hwang, S.‐H., Somatilaka, B. et al. Smoothened determines β‐arrestin–mediated removal of the G protein–coupled receptor Gpr161 from the primary cilium. J Cell Biol, 2016;212(7):861–75. 67. Mashukova, A., Spehr, M., Hatt, H., and Neuhaus, E.M. β‐Arrestin2‐mediated internalization of mammalian odorant receptors. J Neurosci, 2006;26(39):9902–12. 68. Luttrell, L.M., Ferguson, S.S., Daaka, Y. et al. Beta‐arrestin‐dependent formation of beta2 adrenergic receptor‐Src protein kinase complexes. Science, 1999;283(5402):655–61. 69. Ye, F., Nager, A.R., and Nachury, M.V. BBSome trains remove activated GPCRs from cilia by enabling passage through the transition zone. J Cell Biol, 2018;217(5):1847–68. 70. Wood, C.R. and Rosenbaum, J.L. Ciliary ectosomes: transmissions from the cell’s antenna. Trends Cell Biol, 2015;25 (5):276–85. 71. LaVail, M.M. Rod outer segment disk shedding in rat retina: relationship to cyclic lighting. Science, 1976;194(4269):1071–4. 72. Hogan, M.C., Manganelli, L., Woollard, J.R. et al. Characterization of PKD protein‐positive exosome‐like vesicles. J Am Soc Nephrol, 2009;20(2):278–88. 73. Nager, A.R., Goldstein, J.S., Herranz‐Pérez, V. et al. An actin network dispatches ciliary GPCRs into extracellular vesicles to modulate signaling. Cell, 2017;168(1–2):252–63.e14. 74. Phua, S.C., Chiba, S., Suzuki, M. et al. Dynamic remodeling of membrane composition drives cell cycle through primary cilia excision. Cell, 2017;168(1–2):264–79.e15. 75. Hilgendorf, K.I., Johnson, C.T., and Jackson, P.K. The primary cilium as a cellular receiver: organizing ciliary GPCR signaling. Curr Opin Cell Biol, 2016;39:84–92. 76. Grzelak, C.A., Martelotto, L.G., Sigglekow, N.D. et al. The intrahepatic signalling niche of hedgehog is defined by primary cilia positive cells during chronic liver injury. J Hepatol, 2014;60(1):143–51. 77. Mukhopadhyay, S. and Rohatgi, R. G‐protein‐coupled receptors, Hedgehog signaling and primary cilia. Semin Cell Dev Biol, 2014;33:63–72. 78. Haycraft, C.J., Banizs, B., Aydin‐Son, Y. et al. Gli2 and Gli3 localize to cilia and require the intraflagellar transport protein polaris for processing and function. PLoS Genet, 2005;1(4):e53.



5:  Primary Cilia

79. Liu, J., Zeng, H., and Liu, A. The loss of Hh responsiveness by a non‐ciliary Gli2 variant. Development, 2015;142(9):1651–60. 80. Nakamura, T. and Gold, G.H. A cyclic nucleotide‐gated conductance in olfactory receptor cilia. Nature, 1987;325(6103):442–4. 81. Fesenko, E.E., Kolesnikov, S.S., and Lyubarsky, A.L. Induction by cyclic GMP of cationic conductance in plasma membrane of retinal rod outer ­segment. Nature, 1985;313(6000):310–13. 82. Choi, E.J., Xia, Z., and Storm, D.R. Stimulation of the type III olfactory adenylyl cyclase by calcium and calmodulin. Biochemistry, 1992;31(28): 6492–8. 83. Pablo, J.L., DeCaen, P.G., and Clapham, D.E. Progress in ciliary ion channel physiology. J Gen Physiol, 2017;149(1):37–47. 84. Torres, V.E. and Harris, P.C. Strategies targeting cAMP signaling in the treatment of polycystic kidney disease. J Am Soc Nephrol, 2014; 25(1):18–32. 85. Wann, A.K.T., Zuo, N., Haycraft, C.J. et al. Primary cilia mediate mechanotransduction through control of ATP‐induced Ca2+ signaling in compressed chondrocytes. FASEB J, 2012;26(4):1663–71. 86. Wood, C.R., Huang, K., Diener, D.R., and Rosenbaum, J.L. The cilium secretes bioactive ectosomes. Curr Biol, 2013;23 (10):906–11. 87. Ott, C., Elia, N., Jeong, S.Y. et al. Primary cilia utilize glycoprotein‐dependent adhesion mechanisms to stabilize long‐lasting cilia‐cilia contacts. Cilia, 2012;1(1):3. 88. McDermott, K.M., Liu, B.Y., Tlsty, T.D., and Pazour, G.J. Primary cilia regulate branching morphogenesis during mammary gland development. Curr Biol, 2010;20(8):731–7. 89. Jensen, C.G., Poole, C.A., McGlashan, S.R. et  al. Ultrastructural, tomographic and confocal imaging of the chondrocyte primary cilium in situ. Cell Biol Int, 2004;28(2):101–10. 90. Sorokin, S.P. Reconstructions of centriole formation and ciliogenesis in mammalian lungs. J Cell Sci, 1968;3(2):207–30. 91. Lu, L. and Madugula, V. Mechanisms of ciliary targeting: entering importins and Rabs. Cell Mol Life Sci, 2017;75(4):597–606. 92. Yang, T.T., Chong, W.M., Wang, W.‐J. et al. Super‐resolution architecture of mammalian centriole distal appendages reveals distinct blade and matrix functional components. Nat Commun, 2018;9(1):2023. 93. Sánchez, I. and Dynlacht, B.D. Cilium assembly and disassembly. Nat Cell Biol, 2016;18(7):711–17. 94. Garcia‐Gonzalo, F.R. and Reiter, J.F. Open sesame: how transition fibers and the transition zone control ciliary composition. Cold Spring Harb Perspect Biol, 2017;9(2):a028134. 95. Yadav, S.P., Sharma, N.K., Liu, C. et al. Centrosomal protein CP110 controls maturation of mother centriole during cilia biogenesis. Development, 2016;143(9):1491–501. 96. Schmidt, K.N., Kuhns, S., Neuner, A. et al. Cep164 mediates vesicular docking to the mother centriole during early steps of ciliogenesis. J Cell Biol., 2012;199(7):1083–101. 97. Wei, Q., Xu, Q., Zhang, Y. et al. Transition fibre protein FBF1 is required for the ciliary entry of assembled intraflagellar transport complexes. Nat Commun, 2013;4(1):2750. 98. Kubo, T., Brown, Jason M., Bellve, Karl et  al. Together, the IFT81 and IFT74 N‐termini form the main module for intraflagellar transport of tubulin. J Cell Sci, 2016;129(10):2106–19.

61

99. Park, S.M., Lim, J.S., Ramakrishina, S. et al. Brain somatic mutations in MTOR disrupt neuronal ciliogenesis, leading to focal cortical dyslamination. Neuron, 2018;99(1):83–97.e7. 100. Anderson, C.T. and Stearns, T. Centriole age underlies asynchronous primary cilium growth in mammalian cells. Curr Biol, 2009;19(17):1498–502. 101. Paridaen, J.T.M.L., Wilsch‐Bräuninger, M., and Huttner, W.B. Asymmetric inheritance of centrosome‐associated primary cilium membrane directs ciliogenesis after cell division. Cell, 2013;155(2):333–44. 102. Bernabé‐Rubio, M. and Alonso, M.A. Routes and machinery of primary cilium biogenesis. Cell Mol Life Sci, 2017;74 (22):4077–95. 103. He, Q., Wang, G., Dasgupta, S. et al. Characterization of an apical ceramide‐enriched compartment regulating ciliogenesis. Mol Biol Cell, 2012;23(16): 3156–66. 104. Wu, B. and Guo, W. The exocyst at a glance. J Cell Sci, 2015; 128(16):2957–64. 105. Bernabé‐Rubio, M., Andrés, G., Casares‐Arias, J. et al. Novel role for the midbody in primary ciliogenesis by polarized epithelial cells. J Cell Biol, 2016;214(3):259–73. 106. Sharma, N., Kosan, Z.A., Stallworth, J.E., Berbari, N.F., and Yoder, B.K. Soluble levels of cytosolic tubulin regulate ciliary length control. Mol Biol Cell, 2011;22(6):806–16. 107. Gadadhar, S., Dadi, H., Bodakuntla, S. et al. Tubulin glycylation controls primary cilia length. J Cell Biol, 2017;216(9):2701–13. 108. Kim, S. and Dynlacht, B.D. Assembling a primary cilium. Curr Opin Cell Biol, 2013;25(4):506–11. 109. Chien, A., Shih, S.M., Bower, R. et al. Dynamics of the IFT machinery at the ciliary tip. Elife, 2017;6:979. 110. Hendel, N.L., Thomson, M., and Marshall, W.F. Diffusion as a ruler: modeling kinesin diffusion as a length sensor for intraflagellar transport. Biophys J, 2018;114(3):663–74. 111. Mirvis, M., Stearns, T., and Nelson, W.J. Cilium structure, assembly, and disassembly regulated by the cytoskeleton. Biochem J, 2018;475(14):2329–53. 112. Yan, X. and Zhu, X. Branched F‐actin as a negative regulator of cilia formation. Exp Cell Res, 2013;319(2):147–51. 113. Kohli, P., Höhne, M., Jüngst, C. et  al. The ciliary membrane‐associated proteome reveals actin‐binding proteins as key components of cilia. EMBO Rep, 2017;18(9):1521–35. 114. Drummond, M.L., Li, M., Tarapore, E. et al. Actin polymerization controls cilia‐mediated signaling. J Cell Biol, 2018;217(9):3255–66. 115. Assis, L.H.P., Silva‐Junior, R.M.P., Dolce, L.G. et al. The molecular motor Myosin Va interacts with the cilia‐centrosomal protein RPGRIP1L. Sci Rep, 2017;7(1):43692. 116. Ou, Y., Ruan, Y., Cheng, M. et al. Adenylate cyclase regulates elongation of mammalian primary cilia. Exp Cell Res, 2009;315(16):2802–17. 117. Gerhardt, C., Leu, T., Lier, J.M., and Rüther, U. The cilia‐regulated proteasome and its role in the development of ciliopathies and cancer. Cilia, 2016;5(1):14. 118. Rieder, C.L., Jensen, C.G., and Jensen, L.C. The resorption of primary cilia during mitosis in a vertebrate (PtK1) cell line. J Ultrastruct Res, 1979;68(2):173–85. 119. Seeger‐Nukpezah, T., Liebau, M.C., Höpker, K. et  al. The centrosomal kinase plk1 localizes to the transition zone of primary cilia and induces phosphorylation of nephrocystin‐1. PLoS One, 2012;7(6):e38838.

6

Endocytosis in Liver Function and Pathology Micah B. Schott, Barbara Schroeder, and Mark A. McNiven Department of Biochemistry and Molecular Biology, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA

INTRODUCTION A prominent function of the hepatocyte is the regulated uptake of extracellular material for subsequent processing and/or transport into bile. This process, known as endocytosis, depends on elaborate vesicle trafficking machinery that is linked to specific lipid–membrane subdomains and the cytoskeletal matrix. This provides a mechanism to sequester and internalize transmembrane receptor/ligand complexes, such as epidermal growth factor, hepatocyte growth factor, and iron‐bound transferrin, and to help maintain normal lipid serum levels through the endocytosis of low‐density lipoproteins (LDLs). Of equal importance is the fact that this highly evolved machinery can be “hijacked” by many pathogens including bacteria, viruses, and parasites to infect the liver, leading to inflammation and hepatitis. This review will outline the molecular and cell biological mechanisms that underlie endocytosis in the liver, including the diverse roles of endocytic vesicles in the cytoplasm and how these pathways are altered in liver disease.

ENDOCYTIC VESICLE FORMATION AT THE PLASMA MEMBRANE Endocytic vesicles are formed at the plasma membrane as inward budding events that invaginate toward the cytoplasm. These membrane structures form vesicles that are packed with various types of “cargo” such as integral membrane proteins, receptor– ligand complexes, lipids, fluid, and nutrients. The cargo content of endocytic vesicles differs considerably between the various modes of uptake. For example, nonselective endocytic pathways, such as macropinocytosis, mediate the import of nutrient‐rich

extracellular fluid, whereas selective pathways, such as receptor‐mediated endocytosis, import specific soluble ligands and transmembrane receptors. Regardless of the mechanism of internalization, endocytic vesicles deliver cargo to the early ­ endosome (EE) for sorting to different destinations, including recycling back to the cell surface, secretion to the extracellular milieu, or degradation by late endosomes and lysosomes. During receptor‐mediated endocytosis, extracellular ligands, such as LDL, transferrin, epidermal growth factor (EGF), hormones, and many others, bind with high affinity to specific receptors at the plasma membrane to propagate intracellular signaling cascades that lead to gene transcription and other processes. In addition to signal transduction, receptor–ligand binding also initiates endocytic uptake, an event that “desensitizes” signaling by reducing receptor availability at the cell surface. More recently, insights into “housekeeping” trophic receptors, such as the transferrin receptor (TfR) and the LDL receptor (LDLR), which have been assumed not to play a prominent role in cell signaling, have been observed to activate specific signaling cascades. The formation of endocytic vesicles requires the coordination of a wide variety of adaptor proteins that link cargo to the ­endocytic machinery and cytoskeleton. As a brief summary, the various modes of endocytosis and their molecular machinery are described in the following section.

CLATHRIN‐DEPENDENT ENDOCYTOSIS Clathrin‐dependent endocytosis is a central and intensely studied process by which most surface receptors are internalized. The initial steps of this event require receptor–ligand interactions at the plasma

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



6:  Endocytosis in Liver Function and Pathology

membrane, which trigger the recruitment of adaptor proteins that orchestrate the assembly of a clathrin coat (Figure 6.1). Clathrin‐coated vesicles are relatively uniform in size (100– 150 nm diameter) and are present in all cell types. They were first observed by Roth and Porter, who noted that the endocytosis of yolk protein into the oocyte of the mosquito was associated with a marked increase in invaginations of the oocyte cell membrane, which was coated on the cytoplasmic face with what they termed a bristle coat [3]. Soon thereafter, clathrin was identified as the major protein component of the coat [4]. Clathrin‐dependent endocytosis is a highly coordinated process that remains a topic of great research interest. First, plasma membrane subdomains undergo phospholipid remodeling at sites of vesicle formation, a step that is critical for the recruitment of adaptors that link surface receptors with clathrin and other endocytic proteins. Next, inward membrane curvature is achieved by the actin cytoskeleton and curvature‐sensing BAR domain proteins. Clathrin is recruited from the cytosol to stabilize the forming vesicle and aid in its displacement from the plasma membrane. Forming vesicles are separated from the cell surface by dynamin oligomers that induce vesicle scission by generating constrictive force around the thin vesicle neck. Once internalized, the clathrin coat is disassembled to allow for vesicle fusion with downstream target endosomes. All of these steps occur rapidly on the order of 1–5 minutes and require spatial (a)

(b)

Adaptor, cargo recruitment

Membrane bending

63

and temporal coordination of over 50 different adaptors [1]. A model of these events is depicted in Figure 6.1.

Coated pit/vesicle formation and associated factors The plasma membrane is markedly enriched in the phospholipid phosphatidylinositol 4,5‐bisphosphate (PI(4,5)P2) compared to other cellular compartments, which aids in recruitment of adaptors during the initial stages of vesicle formation. The mechanisms that initiate receptor sequestration and membrane bending at these sites are not clearly understood. At minimum, these early steps seem to require the PI(4,5)P2‐binding protein Fes/ CIP4 homology (FCH) domain only 1 and 2 (FCHO1 and FCHO2), as well as the adaptors epidermal growth factor receptor (EGFR) pathway substrate 15 (Eps15) and intersectins 1 and 2 [1]. FCHO proteins also contain curvature‐sensing BAR domains that likely aid in their localization at early budding vesicles. Eps15, along with other cargo‐specific adaptors, helps to recruit a critical endocytosis protein known as adaptin 2 (AP2), the main link between surface receptors and clathrin. The role of receptor–ligand interactions in signaling clathrin‐ coated pit formation is unclear, but some studies suggest that receptors may directly recruit a specific subset of endocytic adaptors. For example, transferrin receptor contains a tyrosine

Scission

Uncoating

(c)

BAR-domain proteins (endophilin, amphiphisin)

Actin

Synd a

pin

Dynamin

Cortactin Arp2/3

AP2

Clathrin GPCR AP2 Arrestin

RTK

Clathrin

Figure 6.1  Model of clathrin‐mediated endocytosis. (a) Endocytosis occurs at plasma membrane sites of receptor–ligand cargo where clathrin and clathrin adaptors are recruited. Invaginating membranes are separated from the plasma membrane by scission machinery in cooperation with the cytoskeleton. Vesicle uncoating allows for participation in downstream endocytic trafficking routes. Cartoon adapted from [1] with permission of Springer Nature. (b) Scanning electron micrograph of clathrin‐coated vesicles show an intracellular view of clathrin‐coated pits. Reprinted from [2] with permission of Rockefeller University Press. (c) Components of clathrin‐coated vesicles, including fission machinery and associated actin cytoskeleton. These components come together to complete the process of cargo sequestration, vesicle formation, and membrane scission.

64

THE LIVER: HEPATOCELLUAR TRAFFICKING OF ENDOCYTIC VESICLES

recognition motif that activates AP2 to recruit clathrin and stimulate endocytosis [5]. Another example is β‐adrenergic ­ receptors that signal the recruitment of the clathrin adaptor β‐ arrestin [6]. Clathrin is recruited from the cytosol to bind AP2 or other cargo‐specific adaptors that are directly linked to surface receptors destined for internalization. Consisting of both heavy and light chains, clathrin forms heterodimers that further assemble into three‐legged triskelions. Clathrin assembly can occur around nascent vesicle pits, but has also been shown to assemble at the plasma membrane prior to vesicle formation as flat, planar lattices [7, 8]. Whereas clathrin heavy chains are capable of binding to adaptins and other accessory proteins, the light chains prevent premature lattice assembly in the cytosol and are regulated by calcium binding and phosphorylation [8–13]. Clathrin does not provide sufficient force for inward ­membrane curvature. Thus, inward curvature on the plasma membrane is thought to require actin polymerization that synergizes with the clathrin coat and vesicle scission machinery. In yeast, actin binds clathrin adaptors such as Sla2 and Ent1 to “pull” coated pits inward, while actin polymerization and myosin at the plasma membrane aids in constriction of the endocytic vesicle neck [1]. BAR domain proteins, such as endophilin and amphiphysin, also sense membrane curvature and facilitate endocytic neck constriction prior to vesicle scission [14, 15]. Other adaptors are also known to link the actin cytoskeleton to the endocytic vesicles, including the actin‐binding proteins profilin, synapsin, syndapin, and cortactin, some of which interact with the molecular machinery that drives vesicle scission. Scission of clathrin‐coated vesicles marks their separation from the plasma membrane. This is accomplished by a family of large GTPase mechanoenzymes known as dynamins. Dynamin binds to GTP near its N‐terminus, while membrane attachment is mediated by a plekstrin homology (PH) domain. Dynamin proteins also bind a variety of effectors via a C‐terminal proline‐rich domain (PRD) [16–18], including BAR domain‐containing proteins and cytoskeletal adaptors. Dynamin has been referred to as a “pinchase” because of its ability to generate discrete vesicles from invaginated coated pits. Dynamin proteins can self‐assemble into oligomeric rings along lipid vesicles [19] and membrane tubules [20]. In vitro studies of various dynamin mutations have revealed that GTP hydrolysis generates the constrictive force for vesicle scission [21]. Following vesicle scission, the clathrin coat is disassembled, and nascent vesicles fuse with other peripheral endosomes. Clathrin disassembly requires the ATPase Hsc70 and the remodeling of membrane PI(4,5)P2 phospholipids to phosphatidylinositol 4‐phosphate (PI(4)P) by synaptojanin. Hsc70 works together with its co‐chaperone GAK/auxilin that is recruited in part by PI(4)P phospholipids [22, 23]. After GAK/auxilin recruitment, the clathrin coat is released from the vesicle. It is worth noting that auxilin/GAK activity is not restricted to clathrin uncoating but is also required for the dynamin‐dependent constriction of coated pits [24].

Clathrin‐independent endocytosis Clathrin‐independent endocytosis is a broad category of ­internalization pathways that can be dynamin dependent and independent [25]. This includes fluid‐phase endocytosis

that  nonselectively imports extracellular fluid and nutrients, ­phagocytosis that imports extracellular pathogens, and caveolin‐­ mediated endocytosis. Although relatively little is known regarding these mechanisms in comparison to clathrin‐mediated endocytosis, perhaps the most broadly appreciated is the formation of caveolae, which are small flask‐shaped vesicles at the plasma membrane that are enriched in sphingolipids and cholesterol. In contrast to clathrin‐mediated endocytosis, caveolae are relatively stable structures at the plasma membrane and are internalized at a very slow rate [26, 27]. This has led to some controversy regarding the role of caveolae in hepatocellular endocytosis, especially given the moderate expression of caveolin, the main structural component of caveolae, in the liver. Nonetheless, hepatocyte caveolae are internalized by dynamin scission machinery [28, 29], and caveolin has been shown to play several important hepatic functions, including lipid metabolism and liver regeneration [30].

HEPATOCELLUAR TRAFFICKING OF ENDOCYTIC VESICLES Once internalized by either clathrin‐dependent or clathrin‐independent endocytosis, endocytic vesicles fuse with EEs that serve as a central hub for sorting and trafficking of endocytic cargo. From here, cargo can be directed through multiple sorting pathways, such as recycling back to the plasma membrane, retrograde transport to the Golgi, degradation by lysosomes and late endosomes, or secretion from endolysosomal vesicles that fuse with the plasma membrane (Figure 6.2). Recycling requires the sorting and concentration of cargo along membrane tubules that extend from endosomal vesicles. Cargo that is not recycled can be directly internalized within small intraluminal vesicles (ILVs), forming distinct endosomal structures known as multivesicular bodies (MVBs). These cargo‐rich ILVs are then degraded by the late endosomal pathway, marked by the acidification of MVBs by fusion with lysosomes. Hepatocytes may also secrete the contents of MVBs, late endosomes, and lysosomes into the bile as these vesicles fuse with the apical plasma membrane. These diverse events are dictated by the spatial and temporal coordination of various protein complexes, post‐translational modifications, phospholipid dynamics, and cytoskeletal regulators. In addition, it has become increasingly appreciated that endosomal vesicles play diverse roles in cellular homeostasis beyond that of cargo trafficking [31].

Cargo sorting at early endosomes Newly formed endocytic vesicles converge at the early endosome (EE), also called the “sorting endosome,” that decides the  fate of endocytosed cargo. The importance of this sorting process cannot be overstated as it is essential to hepatocyte function and insures the proper degradation, recycling, or ­transcytosis of trophic and growth factor receptors, bile acid transporters and other integral membrane proteins [32]. EEs contain multiple subdomains that are thought to mediate different ­ sorting pathways [33]. For example, tubular membrane extensions arising from specific endosomal subdomains are



6:  Endocytosis in Liver Function and Pathology Pl(4,5)P2

65

Basolateral membrane

Rab4

TJ

Rab5

Rab11

WASH retromer

BC

ESCRT EHD1 EHD3 Rab11

EHD1 EHD3

Early/sorting endosome

Endocytic recycling compartment

MVB

Pl(3)P Pl(3,4)P2

Rab5

Rab27b

Rab27a

WASH retromer

Nucleus

Lys

Golgi Rab7 Lys

Late endosome

Pl(3,5)P2

Figure 6.2  Hepatocyte endocytic trafficking routes and the distribution of Rab GTPases and phosphatidylinositides at the compartments within the cell. Rab GTPases mediate trafficking from coated vesicles to early/sorting endosomes (Rab5) or late endosomes (Rab7) as well as recycling back to the plasma membrane directly (Rab4) or indirectly via the endocytic recycling compartment (Rab11). Note that the plasma membrane and the endosomal pathways are enriched in a subset of phosphatidylinositides that contribute to Rab targeting and additional specificity of the different compartments. TJ, tight junction; BC, bile canaliculus; MVB, multivesicular body.

destined for recycling, whereas large vesicular subdomains are thought to “mature” down the late endosomal pathway. Even within a single EE, there is thought to be variation in pH and phospholipid composition at different regions that aid in sorting between recycling, secretory, and degradative routes. The trafficking routes of endosomal vesicles is largely facilitated by a family of small GTPases known as Ras‐associated binding (Rab) proteins. Approximately 60 different Rab proteins play important roles in membrane and vesicle trafficking pathways in humans, most of which are expressed in the hepatocyte [34, 35]. Rab proteins guide endocytic vesicle trafficking by binding to effector proteins that modulate endocytic vesicle functions. For EE function, Rab5 is predominant along with binding effectors such as phosphoinositide 3‐kinase (PI3K)/ VPS34, EEA1, rabenosyn‐5, and others [36]. PI3K/VPS34 changes the phospholipid composition of EEs to phosphatidylinositol 3‐phosphate (PI(3)P), giving these vesicles a distinctive membrane signature that aids in the recruitment of other Rab5 effectors. EEA1 binds both Rab5 and PI(3)P and mediates the fusion of endocytic vesicles by coordinating SNARE proteins such as syntaxin6 and syntaxin13. Rabenosyn‐5 is thought to regulate “fast recycling” of cargo such as the transferrin receptor, and delivery of cargo to the endocytic recycling center for “slow recycling” [37]. Cargo sorting is driven by dynamic membrane tubules that extend from EE subdomains [38]. Although the mechanisms that form these tubules are still unclear, several lines of evidence suggest that cargo is first concentrated at specific EE

subdomains, followed by actin polymerization that generates force for tubule initiation and extension. Membrane tubules further extend along microtubule tracks, and scission machinery separates the tubule from the EE, releasing tubular vesicle carriers that transport the sorted cargo to the plasma ­membrane, endocytic recycling compartment (ERC), or Golgi apparatus. Recent work has considerably advanced our understanding of the molecular mechanisms underlying each of these steps. In order for endocytic sorting to occur, endosomes must be able to recognize and sequester specific cargo proteins. This “cargo recognition” process appears to rely on a diverse class of proteins known as the sorting nexins (Snx) [39]. Snx proteins contain a conserved PX domain that binds PI(3)P phospholipids enriched on EEs, as well as cargo‐specific binding domains that vary between Snx family members. Snx proteins also contain additional motifs such as BAR domains that sense membrane curvature, FERM domains that aid in cargo recognition, and others [40]. Sorting nexins work in concert with the retromer, a trimeric protein complex composed of Vps26, Vps35, and Vps29 that is critical to endocytic sorting. More recently, a new retromer‐like complex known as retriever (consisting of DSCR3, C160rf62, and VPS29) was also identified to work in concert with Snx17 [41]. Retriever/Snx17 promotes the recycling of a subset of cargo molecules that is distinct from retromer‐dependent sorting. Thus, the ability of EEs to accurately decipher different types of cargo, and to organize cargo at endosomal subdomains, is absolutely critical to hepatocellular vesicle

66

THE LIVER: HEPATOCELLUAR TRAFFICKING OF ENDOCYTIC VESICLES

transport, and involves an impressive synchronization between cargo adaptors, protein complexes, and the cytoskeleton. In addition to sorting cargo at endosomal subdomains, Snx–retromer/retriever complexes coordinate actin dynamics at these sites by recruiting the WASH complex [40, 42], which consists of five individual proteins (WASH1, Fam21, Strumpellin, SWIP, and CCDC53). WASH stimulates Arp2/3 for actin nucleation at EE sites where membrane tubulation occurs. It is thought that actin generates the force required for tubule formation, extension, and/or possibly scission. Fam21 of the WASH complex seems to act as the main tethering link to other pathways. For example, Fam21 is known to bind Vps35 of the retromer complex [43]. Fam21 also links the WASH complex to retriever by binding yet another recently described protein complex known as CCC (COMMD1, CCDC22, CCDC93) [41]. Early endosomal tubules elongate and move along microtubule tracks. This requires tethering to molecular motor proteins that travel toward (minus‐end dyneins) or away from (plus‐end kinesins) the microtubule organizing center (MTOC). The small GTPase Rab7 facilitates the tethering of EE tubules to molecular motor proteins in cooperation with Snx‐BAR proteins. The tethering of EE tubules to dyneins versus kinesins plays an important role in dictating cargo destination. For example, dynein‐tethered tubules are effectively transported back to the perinuclear endocytic recycling compartment and/ or Golgi apparatus, whereas kinesin‐tethered tubules are ­transported to the plasma membrane. The scission machinery that separates nascent tubules from EEs is currently unclear, but likely involves actin, the ATPase EHD1, and/or dynamin GTPases.

Regulation of cargo sorting by post‐ translational modifications As described above, cargo sorting between recycling, degradation, and secretion pathways relies on various endosomal adaptors that bind specific cargo. These cargo–adaptor interactions, as well as cargo fate, are guided by post‐translational modifications, such as ubiquitination and phosphorylation, that direct downstream sorting pathways. Ubiqutin (Ub) is an 8 kDa, highly conserved protein that can be covalently linked to lysine residues on target proteins. This can occur as a polyubiquitin chain or as a single monoubiquitin attached to one or multiple lysine residues. Whereas ubiquitination of soluble proteins leads to their degradation by the proteasome [44, 45], ubiquitination of endocytic cargo is a signal for MVB internalization and degradation by the late endosomal pathway (described in a later section). A prototypical example of this process is the EGFR that is ubiquitinated following ligand stimulation, which prevents EGFR recycling and promotes its sorting to degradative pathways [46, 47]. Phosphorylation by nonreceptor tyrosine kinases (NRTK) and by serine–threonine kinases also play important roles in endocytic cargo trafficking. For example, the NRTK Src was shown to phosphorylate the EGFR adaptor CIN85 to regulate EGFR ubiquitination and degradation by the late endosomal pathway [48]. In addition, the transferrin receptor (TfR) is

regulated by the NRTK c‐Abl kinase. Whereas TfR is normally recycled back to the plasma membrane, inhibition of c‐Abl kinase redirects TfR to late endosomes for degradation [49]. Serine–threonine kinases, such as cyclic AMP (cAMP)‐dependent protein kinase A (PKA), are well‐described to regulate GPCR sorting. For example, activation of the β‐adrenergic receptor leads to multiple PKA phosphorylation events at its C‐ terminal cytoplasmic tail, which recruits β‐arrestins that induce GPCR internalization and desensitization [6]. In addition, PKA phosphorylation of the β‐adrenergic adaptor gravin is essential for receptor recycling and resensitization back to the plasma membrane [50–53]. Thus, it is clear that cargo‐specific phosphorylation by NRTKs and serine–threonine kinases drive many important aspects of endocytic regulation, including cargo internalization and sorting between degradative and recycling pathways.

Recycling endosomes and the endocytic recycling compartment Whereas early endosomal cargo can be recycled directly back to the plasma membrane (fast recycling), a “slow recycling” pathway also exists whereby cargo is shunted through tubular recycling endosomes (RE) that are located at the cell periphery or clustered within the perinuclear endocytic recycling compartment (ERC). REs are enriched in Rab11, distinguishing these vesicles from Rab5‐positive EEs. The delivery of cargo from Rab5‐positive EEs to Rab11‐ positive REs seems to rely on a family of Eps15 homology domain (EHD) proteins that play diverse roles at various stages of endosomal sorting and recycling [54, 55]. EHD1 is thought to cause scission of tubular REs arising from the ERC, and perhaps EE tubules as well [56]. Indeed, EHD1‐depleted cells show aberrant recycling of the transferrin receptor [57], as do EHD3 and EHD4, which also aid in the transition of the TfR from EE to RE [58, 59]. In addition to their well‐described roles in the RE, EHD proteins also affect sorting at the EEs by regulating Rab5 and its effectors rabenosyn‐5 and rabakyrin‐5 [58–60]. Tubular REs arising from the perinuclear ERC deliver cargo back to the plasma membrane. The generation of these structures is still under investigation, but likely involves a variety of mechanisms. Interestingly, the WASH complex, which is important in the actin‐based budding of EE tubules, is also present on REs and suggests its role in the generating tubular carriers from the ERC [61]. However, EHD proteins also promote both the scission of ERC‐derived tubular carriers (EHD1) and the stabilization of these structures (EHD3). EHD1 seems to operate within a protein complex containing the endosomal trafficking adaptors MICAL‐L1 and syndapin2, both of which can generate tubules from phosphatidylserine (PS)‐rich liposomes in vitro [62]. The EHD1–MICAL‐L1–syndapin2 complex may also work on EE‐derived tubular carriers, perhaps even through binding to the Vps26 subunit of the retromer complex [60, 63, 64]. Thus, the generation of tubules is critical to cargo recycling directly from EE vesicles and through the ERC, but the synergy between the different molecular machineries on early endosomes and REs is still unclear.



6:  Endocytosis in Liver Function and Pathology

Multivesicular bodies and the late endosomal pathway Hepatocellular cargo that is not recycled will progress down the  late endosomal pathway for degradation or secretion into the  apical bile canaliculus. Due to the acidic pH of late endosomes and their role in degradation of proteins and lipids, late endosomes are more akin to lysosomes than EEs. Late endosomes can be distinguished experimentally from EEs by their acidic luminal pH, enrichment in PI(3,5)P2 phospholipids, and association with Rab7 and other protein markers. Although the transition from Rab5‐positive EEs to Rab7‐positive late endosomes is not fully understood, the prevailing model is that EEs themselves “mature” to become gradually more acidic by fusion with smaller lysosomes that donate acid hydrolases responsible for the breakdown of proteins, lipids, and nucleic acids at low pH [65]. At the same time, endosomal cargo that is destined for degradation will be invaginated within small, 50% of acute liver failure [79]. The importance of autophagy in this process has been observed particularly in the context of acetaminophen (APAP) overdose, which induces severe liver injury [92]. Autophagy is activated by excessive APAP to clear damaged mitochondria that produce reactive oxygen species (ROS), leading to hepatocellular death. The degradation of mitochondria by autophagy is a protective mechanism, as pharmacological stimulation of autophagy protects against APAP‐induced liver injury in mice [93]. The selective autophagy of damaged mitochondria, termed “mitophagy,” seems to utilize both autophagic and endosomal machineries. The process is driven by mitochondrial fission that produces smaller sized mitochondria that are more susceptible to autophagic engulfment [94]. In canonical mitophagy, damaged mitochondria accumulate PTEN‐induced putative ­ kinase 1 (PINK1) that recruits E3 ubiquitin ligase parkin to the ­mitochondrial outer membrane [95]. This generates ubiquitin signatures on the mitochondrial surface that are recognized by



6:  Endocytosis in Liver Function and Pathology

autophagy receptors that recruit LC3‐positive phagophores for engulfment within the autophagosome [95, 96]. Recent reports indicate that this engulfment may also be mediated directly by  endosomal vesicles even in the absence of traditional autophagosome‐based mitophagy [97, 98]. In this model, ­ termed “endosomal mitophagy,” parkin recruits damaged mitochondria to Rab5‐positive EEs that utilize the ESCRT machinery to bind ubiquitylated mitochondria. ESCRT recognition of ubiquitylated mitochondria leads to an MVB‐like mitochondrial engulfment and degradation by the late endosomal pathway. In addition to mitophagy, endosomal vesicles have recently been reported to play a novel role in mitochondrial fission and fusion. These findings suggest that lysosomes make an intimate connection with mitochondria, especially at sites of mitochondrial fission [99]. The tethering of mitochondria to lysosomes requires active, GTP‐bound Rab7 on the lysosome, whereas untethering requires a protein known as TBC1D15 that inactivates Rab7 by facilitating GTP hydrolysis and dissociation. Interestingly, TBC1D15 is recruited to the mitochondria to regulate Rab7 activity, suggesting a bidirectional mechanism whereby mitochondria regulate lysosomal positioning via Rab7 inactivation, and lysosomes act on mitochondria to mediate fission.

Autophagy and lipid droplet homeostasis The liver is a central hub for the storage and regulation of neutral lipids such as triglycerides and cholesterol esters. These lipids are assembled from free fatty acids (FFA) that are synthesized de novo or imported from extracellular sources (i.e. dietary fat, lipids released from adipose tissue) and stored within specialized organelles known as lipid droplets (LDs) [32]. Dysregulation of these important hepatocellular functions leads to profound lipodystrophies that impact both liver function and whole‐body lipid homeostasis. Most common are fatty liver diseases that affect 20–30% of the world population and coincide with other metabolic diseases (obesity, diabetes, metabolic syndrome) and/or chronic alcohol consumption [100, 101]. Central to hepatic fat storage and energy utilization, LDs store intracellular neutral lipids surrounded by a phospholipid monolayer that associates with proteins involved in lipid storage, breakdown, and fusion [102–104]. In addition, several proteomic studies have shown that LDs associate with numerous Rab GTPases, as well as endosomal vesicle machinery found on endosomes, multivesicular bodies, and lysosomes [105–109]. LDs are catabolized by two different types of lipases – ­neutral lipases that reside in the cytoplasm and acid lipases that reside in lysosomes/late endosomes. Neutral lipases, such as adipose triglyceride lipase (ATGL) and hormone‐sensitive lipase (HSL), are recruited from the cytosol to the LD following β‐adrenergic receptor activation of the cAMP/PKA pathway. It was once thought that these neutral lipases played a more prominent role in adipose tissue and skeletal muscle, but numerous studies have now demonstrated their importance in liver function and metabolic disease pathology [110–114]. Acid lipases such as lysosomal acid lipase (LAL) are active within the low pH environment of lysosomes/late endosomes, and thus cannot be recruited to the LD surface in the same manner as the ­cytosolic lipases. LAL gains access to LDs via lipophagy (Figure  6.3), a selective

69

autophagic mechanism for LD breakdown [115–117]. Although the mechanisms that traffic hepatocellular LDs from the cytoplasm to the lumen of lysosomes and late endosomes are not completely understood, it is clear that numerous Rab proteins and other endocytic machinery play important roles in this process. For example, both Rab7 and Rab10 are activated on the LD surface during periods of low nutrient availability. Rab10 was shown to form a trimeric complex with its effector Eps15 homology binding protein 1 (EHBP1) and EHD2 on the LD surface, and this complex is important for phagophore recruitment and extension around the LD [118]. Rab7 drives LD interaction with multivesicular bodies and late endosomes marked by the membrane tetraspanin CD63 [119]. The direct interaction of LDs with MVBs/endosomes raises the possibility that alternative forms of “microlipophagy” may exist in parallel with traditional autophagosome‐based “macrolipophagy” (Figure  6.3) Interestingly, Rab7 was also shown to be inhibited by chronic ethanol exposure, which alters lysosomal morphology and motility to perturb the lipophagic catabolism of LDs during alcoholic fatty liver progression [120]. Other endocytic proteins also appear to be involved in hepatocellular lipophagy, including clathrin, dynamin2, and Vps4 [84, 119]. In addition to lipophagy, several studies now show a bidirectional synergy between autophagy and cytosolic lipases. First, various autophagy pathways are known to promote ATGL ­activity in hepatocytes to protect against fatty liver progression. For  example, chaperone‐mediated autophagy was recently shown to degrade LD proteins that inhibit ATGL‐mediated lipolysis [121]. Macroautophagy also seems to facilitate ­hepatocyte lipolysis, as both HSL and ATGL contain several LC3‐interacting regions that link these cytosolic lipases to autophagosomes that function to deliver HSL and ATGL to the surface of LDs [122]. Conversely, ATGL‐mediated lipolysis has been reciprocally shown to induce autophagy at the transcriptional level. In this model, ATGL activity releases FFAs that  serve as signaling ligands for PGC‐1α and PPAR‐α. The ­mechanism for this relies on the activation of NAD‐dependent deacetylase sirtuin 1, which facilitates the interaction of PPAR‐α with deacetylated PGC‐1α to promote autophagic gene transcription and other metabolic programs [123, 124].

VIRAL INFECTION AND HEPATOCELLULAR ENDOCYTOSIS Human pathogens such as hepatitis B virus (HBV) and hepatitis C virus (HCV) are a leading cause for liver diseases such as steatohepatitis, cirrhosis, and hepatocellular carcinoma. These viruses hijack normal hepatocellular processes, particularly host endocytic pathways that are utilized for infection, propagation, and  secretion to neighboring cells and tissues [reviewed in 125].  Five hepatitis viruses have been described that hijack different  endocytic pathways for replication and secretion: ­ HBV  (Hepadnaviridae family), HDV (Deltaviridae), HCV (Flaviviridae family), HAV  (Hepatovirus Picornaviridae), and  HEV (Hepervirus Heperviridae). Although the genetic and  molecular components of these hepatic viruses vary

70

THE LIVER: VIRAL INFECTION AND HEPATOCELLULAR ENDOCYTOSIS

(a)

Plasma membrane Clathrin-mediated endocytosis

Microlipophagy LD

? MVB/Late endosome

EE

Macrolipophagy ALR LD Phagophore

Lysosome Autophagosome

(b)

Dyn2 Autolysosome

(c)

Figure 6.3  Lipophagy in hepatocytes. (a) A working model of two distinct lipophagy pathways. Microlipophagy is proposed to occur by the direct uptake of lipid droplets by endosomal vesicles such as multivesicular bodies (MVBs) and late endosomes for degradation by lysosomal enzymes. Macrolipophagy utilizes traditional autophagy machinery whereby lipid droplets (LDs) are targeted by phagophores for complete engulfment within autophagosomes which fuse with lysosomes to become degradative autolysosomes. The terminal stages of macrolipophagy incorporate autolysosome reformation (ALR), whereby membrane tubules extending from the autolysosome undergo scission by dynamin 2 (Dyn2). This generates nascent lysosomes that contribute to autophagic and late endosomal degradation. (b) Transmission electron micrograph shows an LD (white arrow) encased within an autophagic membrane in Huh7 human hepatoma cells. Reproduced from [118] with permission of American Association for the Advancement of Science. (c) Electron micrograph of a siRNA Dyn2‐depleted Hep3B human hepatoma cell shows an autolysosome tubule (white arrows) that is elongated in the absence of tubule scission machinery. Reproduced from [84] with permission of Rockefeller University Press.

considerably, each virus consists of a nucleotide genome of ssRNA (D,C,A,E) or dsDNA (B) that is encapsulated by a proteinaceous core and surrounded by an outer envelope consisting of proteins and host‐derived lipids. A model for how each of these viruses are assembled as they advance through hepatocyte endocytic pathways is depicted in Figure 6.4.

Viral attachment and endocytosis Hepatitis viruses attach to the cell surface and can move ­laterally along the plasma membrane prior to endocytosis. For HBV, this interaction is facilitated by hepatocyte heparin sulfate proteoglycans (HSPGs) and by sodium taurocholate cotransporting polypeptide (NTCP), both of which bind to HBV envelope ­proteins. HCV attachment is also facilitated by HSPGs, but has also been reported to interact with apolipoproteins, junctional proteins, surface receptors, and others that help

facilitate membrane attachment. HAV or HEV attachment to the  ­hepatocyte is not well understood and seems to require HSPGs, asialoglycoprotein receptor (ASPGR), and others [reviewed in 125]. Hepatic viruses utilize clathrin‐mediated endocytosis for hepatocellular uptake, but the contributions of other endocytic routes such as caveolin‐mediated endocytosis, pinocytosis, and phagocytosis cannot be ruled out [125]. Upon internalization, these viruses traverse through hepatocyte endocytic pathways prior to genome release into the nucleus via the cytoplasm. For HBV, genome release into the nucleus is pH dependent and requires early‐to‐late endosomal transitions mediated by Rab5 and Rab7 [126]. The mechanisms supporting these important steps, including how these viruses evade late endosomal degradation, remain unclear. In contrast to HBV and other hepatic viruses, the HCV genome is translated in the cytoplasm following release from EEs in a pH‐dependent manner [127].



6:  Endocytosis in Liver Function and Pathology

BCDE

Virus

71

ABE C?

C

Clathrin-mediated endocytosis

HSPGs

Autophagosome

Endosome

B

ACDE RNA

Nucleus

C

MVB

Golgi

B

Viral genome

DNA

RNA

C?

B ER

ESCRT:

ABE

Lysosome

Figure 6.4  Model depicting how the different hepatitis viruses (ABCDE) utilize common and distinct endocytic pathways in the hepatocyte ­during internalization, infection, maturation, and release. Modified from [125] with permission of John Wiley & Sons.

Viral assembly, replication, and hepatocellular secretion HBV‐infected cells secrete both infectious virions and non‐ infectious subviral particles (SVPs) made up of envelope proteins that likely serve as “decoys” against the host immune system. SVPs are assembled in the ER and packaged for secretion within the ER–Golgi intermediate compartment (ERGIC) [128]. For HCV, assembly occurs within ER‐derived double‐ membrane vesicles (~150 nm in diameter) that are constructed in response to viral infection. Interestingly, HCV also stimulates the formation of LDs that are contained within these vesicles [129]. LDs are thought to facilitate HCV replication, as core and nonstructural proteins are found on the LD surface. LDs may also contain viral double‐stranded RNA, suggesting LDs are also a site of genomic replication. Other work has shown that LD‐associated proteins, such as Rab18 [130] and Plin3/Tip47 [131], also help facilitate HCV replication. Hepatitis viruses utilize different endo‐membrane vesicle pathways for secretion. For example, HCV secretion requires trans‐Golgi machinery and Rab11, suggesting interplay between the Golgi and RE compartments [132]. Other viruses, such as HBV, utilize the MVB for uptake and secretion through a mechanism that requires the ESCRT machinery. It was recently found that HBV induces the dramatic tubulation of MVBs and autophagosomes by modulating Rab7 activity, which stimulates lysosomal fusion and pH‐dependent secretion [133]. Recent studies reveal that hepatic viruses utilize late ­endocytic compartments for maturation and release. Some of these viruses (HAV, HEV) utilize MVB‐derived membranes to envelop nascent virions as their genomes do not encode envelope proteins. In addition to serving as a membrane source, MVBs also generate ILVs that support an exosome‐like shedding of hepatic viruses. Trafficking of viruses to MVBs, as well as MVB fusion with the plasma membrane, seems to require the participation of several Rab proteins including Rab2b, Rab5a, Rab9b, Rab27a,

and Rab27b. These later Rabs are particularly important in MVB‐based secretion, as Rab27a is thought to mediate MVB docking/fusion with the plasma membrane, whereas Rab27b participates in the transfer of MVBs from microtubules to the actin‐rich regions at the cell periphery [134]. In conclusion, hepatitis viruses utilize hepatocyte membrane trafficking machinery for infection and propagation. Despite the substantial progress that has been made toward understanding how the host hepatocellular endocytic pathways are modified by these viruses, future work will need to further define the synergy between the endosomal compartments, nucleus, ER, Golgi, and the autophagic machinery in the viral life cycle.

FUTURE DIRECTIONS New imaging and biochemical techniques have provided detailed insights into the mechanisms of vesicle formation and post‐endocytic trafficking. Nevertheless, the list of protein and lipid components of the vesicle‐forming machinery continues to expand while the functions of many endocytic components in specific transport processes are yet to be established. It is important to understand how the hepatocyte manages to ­coordinate and control a massive protein and lipid network that supports vesicular trafficking events. In this context, ­challenges for the future will be to understand how different endocytic cargos are sequestered and sorted, how membrane subdomains are maintained even within individual vesicles, and which phospho‐substrates are targeted by kinase signaling cascades to control endocytic trafficking events. Equally exciting will be  the elucidation of alternative roles for endocytic vesicles in  processes such as the autophagy of mitochondria and LDs.  Insights into these processes will provide a foundation for  understanding the function of hepatocytes in health and disease.

72

THE LIVER:  REFERENCES

REFERENCES 1. Kaksonen, M. and Roux, A. Mechanisms of clathrin‐mediated endocytosis. Nature reviews Mol Cell Biol, 2018;19(5):313–26. 2. Heuser, J. Effects of cytoplasmic acidification on clathrin lattice morphology. J Cell Biol, 1989;108(2):401–11. 3. Roth, T.F. and Porter, K.R. Yolk protein uptake in the oocyte of the mosquito Aedes aegypti. L. J Cell Biol, 1964;20:313–32. 4. Pearse, B.M. Clathrin: a unique protein associated with intracellular transfer of membrane by coated vesicles. Proc Natl Acad Sci U S A, 1976;73(4): 1255–9. 5. Kadlecova, Z., Spielman, S.J., Loerke, D. et  al. Regulation of clathrin‐­ mediated endocytosis by hierarchical allosteric activation of AP2. J Cell Biol, 2017;216(1):167–79. 6. Goodman, O.B., Jr., Krupnick, J.G., Santini, F. et al. Beta‐arrestin acts as a clathrin adaptor in endocytosis of the beta2‐adrenergic receptor. Nature, 1996;383(6599):447–50. 7. Larkin, J.M., Donzell, W.C., and Anderson, R.G. Potassium‐dependent assembly of coated pits: new coated pits form as planar clathrin lattices. J Cell Biol, 1986;103(6 Pt 2):2619–27. 8. Heuser, J. Three‐dimensional visualization of coated vesicle formation in fibroblasts. J Cell Biol, 1980;84(3):560–83. 9. Ungewickell, E. and Ungewickell, H. Bovine brain clathrin light chains impede heavy chain assembly in vitro. J Biol Chem, 1991;266(19): 12710–14. 10. Vigers, G.P., Crowther, R.A., and Pearse, B.M. Three‐dimensional structure of clathrin cages in ice. EMBO J, 1986;5(3):529–34. 11. Nathke, I., Hill, B.L., Parham, P., and Brodsky, F.M. The calcium‐binding site of clathrin light chains. J Biol Chem, 1990;265(30):18621–7. 12. Hill, B.L., Drickamer, K., Brodsky, F.M., and Parham, P. Identification of the  phosphorylation sites of clathrin light chain LCb. J Biol Chem, 1988;263(12):5499–501. 13. Schmid, S.L., Braell, W.A., Schlossman, D.M., and Rothman, J.E. A role for clathrin light chains in the recognition of clathrin cages by ‘uncoating ATPase.’ Nature, 1984;311(5983):228–31. 14. Meinecke, M., Boucrot, E., Camdere, G. et  al. Cooperative recruitment of dynamin and BIN/amphiphysin/Rvs (BAR) domain‐containing proteins leads to GTP‐dependent membrane scission. J Biol Chem, 2013;288(9):6651–61. 15. Dawson, J.C., Legg, J.A., and Machesky, L.M. Bar domain proteins: a role in tubulation, scission and actin assembly in clathrin‐mediated endocytosis. Trends Cell Biol, 2006;16(10):493–8. 16. Kim, Y. and Chang, S. Ever‐expanding network of dynamin‐interacting proteins. Mol Neurobiol, 2006;34(2):129–36. 17. McNiven, M.A., Cao, H., Pitts, K.R., and Yoon, Y. The dynamin family of  mechanoenzymes: pinching in new places. Trends Biochem Sci, 2000;25(3):115–20. 18. Praefcke, G.J. and McMahon, H.T. The dynamin superfamily: universal membrane tubulation and fission molecules? Nat Rev Mol Cell Biol, 2004;5(2):133–47. 19. Tuma, P.L. and Collins, C.A. Dynamin forms polymeric complexes in the presence of lipid vesicles. Characterization of chemically cross‐linked dynamin molecules. J Biol Chem, 1995;270(44):26707–14. 20. Takei, K., McPherson, P.S., Schmid, S.L., and De Camilli, P. Tubular membrane invaginations coated by dynamin rings are induced by GTP‐gamma S in nerve terminals. Nature, 1995;374(6518):186–90. 21. Sweitzer, S.M. and Hinshaw, J.E. Dynamin undergoes a GTP‐dependent conformational change causing vesiculation. Cell, 1998;93(6):1021–9. 22. Massol, R.H., Boll, W., Griffin, A.M., and Kirchhausen, T. A burst of auxilin recruitment determines the onset of clathrin‐coated vesicle uncoating. Proc Natl Acad Sci U S A, 2006;103(27):10265–70. 23. Lee, D.W., Wu, X., Eisenberg, E., and Greene, L.E. Recruitment dynamics of GAK and auxilin to clathrin‐coated pits during endocytosis. J Cell Sci, 2006;119(Pt 17):3502–12. 24. Sever, S., Skoch, J., Newmyer, S. et al. Physical and functional connection between auxilin and dynamin during endocytosis. EMBO J, 2006;25(18): 4163–74. 25. Mayor, S., Parton, R.G., and Donaldson, J.G. Clathrin‐independent pathways of endocytosis. Cold Spring Harb Perspect Biol, 2014;6(6):a016758. 26. Tagawa, A., Mezzacasa, A., Hayer, A. et  al. Assembly and trafficking of caveolar domains in the cell: caveolae as stable, cargo‐triggered, vesicular transporters. J Cell Biol, 2005;170(5):769–79.

27. Pelkmans, L., Burli, T., Zerial, M., and Helenius, A. Caveolin‐stabilized membrane domains as multifunctional transport and sorting devices in endocytic membrane traffic. Cell, 2004;118(6):767–80. 28. Yao, Q., Chen, J., Cao, H. et al. Caveolin‐1 interacts directly with dynamin‐2. J Mol Biol, 2005;348(2):491–501. 29. Henley, J.R., Krueger, E.W., Oswald, B.J., and McNiven, M.A. Dynamin‐ mediated internalization of caveolae. J Cell Biol, 1998;141(1):85–99. 30. Fernandez‐Rojo, M.A. and Ramm, G.A. Caveolin‐1 function in liver physiology and disease. Trends Mol Med, 2016;22(10):889–904. 31. Schroeder, B. and McNiven, M.A. Importance of endocytic pathways in liver function and disease. Compr Physiol, 2014;4(4):1403–17. 32. Schulze, R.J. et al. The cell biology of the hepatocyte. J Cell Biol, 2019;218(7):2096–112. 33. Short, B. Sorting out endosome form and function. J Cell Biol, 2015;210(6):870. 34. Zhen, Y. and Stenmark, H. Cellular functions of Rab GTPases at a glance. J Cell Sci, 2015;128(17):3171–6. 35. Stenmark, H. Rab GTPases as coordinators of vesicle traffic. Nat Rev Mol Cell Biol, 2009;10(8):513–25. 36. Jovic, M., Sharma, M., Rahajeng, J., and Caplan, S. The early endosome: a busy sorting station for proteins at the crossroads. Histol Histopathol. 2010;25(1):99–112. 37. Navaroli, D.M., Bellve, K.D., Standley, C. et  al. Rabenosyn‐5 defines the fate of the transferrin receptor following clathrin‐mediated endocytosis. Proc Natl Acad Sci U S A, 2012;109(8):E471–80. 38. van Weering, J.R. and Cullen, P.J. Membrane‐associated cargo recycling by tubule‐based endosomal sorting. Semin Cell Dev Biol, 2014;31:40–7. 39. Cullen, P.J. Endosomal sorting and signalling: an emerging role for sorting nexins. Nat Rev Mol Cell Biol, 2008;9(7):574–82. 40. Wang, J., Fedoseienko, A., Chen, B. et al. Endosomal receptor trafficking: retromer and beyond. Traffic, 2018;19(8):578–90. 41. McNally, K.E. et al. Retriever is a multiprotein complex for retromer-independent endosomal cargo recycling. Nat Cell Biol, 2017;19(10):1214–25. 42. Duleh, S.N. and Welch, M.D. WASH and the Arp2/3 complex regulate endosome shape and trafficking. Cytoskeleton (Hoboken), 2010;67(3): 193–206. 43. Harbour ME., Breusegem, S.Y., and Seaman, M.N. Recruitment of the endosomal WASH complex is mediated by the extended “tail” of Fam21 binding to the retromer protein Vps35. Biochem J, 2012;442(1):209–20. 44. Elsasser, S. and Finley, D. Delivery of ubiquitinated substrates to protein‐ unfolding machines. Nat Cell Biol, 2005;7(8):742–9. 45. Miller, J. and Gordon, C. The regulation of proteasome degradation by multi‐ ubiquitin chain binding proteins. FEBS Lett, 2005;579(15):3224–30. 46. Huang, F., Goh, L.K., and Sorkin, A. EGF receptor ubiquitination is not ­necessary for its internalization. Proc Natl Acad Sci U S A, 2007;104(43): 16904–9. 47. Ravid, T., Heidinger, J.M., Gee, P., Khan, E.M., and Goldkorn, T. c‐Cbl‐ mediated ubiquitinylation is required for epidermal growth factor receptor exit from the early endosomes. J Biol Chem, 2004;279(35):37153–62. 48. Schroeder, B., Srivatsan, S., Shaw, A., Billadeau, D., and McNiven, M.A. CIN85 phosphorylation is essential for EGFR ubiquitination and sorting into multivesicular bodies. Mol Biol Cell, 2012;23(18):3602–11. 49. Cao, H., Schroeder, B., Chen, J., Schott, M.B., and McNiven, M.A. The endocytic fate of the transferrin receptor is regulated by c‐Abl kinase. J Biol Chem, 2016;291(32):16424–37. 50. Shih, M., Lin, F., Scott, J.D., Wang, H.Y., and Malbon, C.C. Dynamic complexes of β2‐adrenergic receptors with protein kinases and phosphatases and the role of gravin. J Biol Chem, 1999;274(3):1588–95. 51. Lin, F., Wang, H., and Malbon, C.C. Gravin‐mediated formation of signaling complexes in beta 2‐adrenergic receptor desensitization and resensitization. J Biol Chem, 2000;275(25):19025–34. 52. Fan, G.F., Shumay, E., Wang, H.Y., and Malbon, C.C. The scaffold protein gravin (cAMP‐dependent protein kinase‐anchoring protein 250) binds the β2‐adrenergic receptor via the receptor cytoplasmic Arg‐329 to Leu‐413 domain and provides a mobile scaffold during desensitization. J Biol Chem, 2001;276(26):24005–14. 53. Tao, J., Wang, H.Y., and Malbon, C.C. Protein kinase A regulates AKAP250 (gravin) scaffold binding to the beta2‐adrenergic receptor. EMBO J, 2003;22(24):6419–29. 54. Mintz, L., Galperin, E., Pasmanik‐Chor, M. et al. EHD1 – an EH‐domain‐ containing protein with a specific expression pattern. Genomics, 1999;59(1): 66–76.



6:  Endocytosis in Liver Function and Pathology

55. Pohl, U., Smith, J.S., Tachibana, I. et al. EHD2, EHD3, and EHD4 encode novel members of a highly conserved family of EH domain‐containing ­proteins. Genomics, 2000;63(2):255–62. 56. Xie, S., Bahl, K., Reinecke, J.B. et al. The endocytic recycling compartment maintains cargo segregation acquired upon exit from the sorting endosome. Mol Biol Cell, 2016;27(1):108–26. 57. Naslavsky, N., Boehm, M., Backlund, P.S., Jr., and Caplan, S. Rabenosyn‐5 and EHD1 interact and sequentially regulate protein recycling to the plasma membrane. Mol Biol Cell, 2004;15(5):2410–22. 58. Naslavsky, N., Rahajeng, J., Sharma, M., Jovic, M., and Caplan, S. Interactions between EHD proteins and Rab11‐FIP2: a role for EHD3 in early endosomal transport. Mol Biol Cell, 2006;17(1):163–77. 59. Sharma, M., Naslavsky, N., and Caplan, S. A role for EHD4 in the regulation of early endosomal transport. Traffic, 2008;9(6):995–1018. 60. Zhang, J., Reiling, C., Reinecke, J.B. et  al. Rabankyrin‐5 interacts with EHD1 and Vps26 to regulate endocytic trafficking and retromer function. Traffic, 2012;13(5):745–57. 61. Duleh, S.N. and Welch, M.D. WASH and the Arp2/3 complex regulate ­endosome shape and trafficking. Cytoskeleton, 2010;67(3):193–206. 62. Giridharan, S.S., Cai, B., Vitale, N., Naslavsky, N., and Caplan, S. Cooperation of MICAL‐L1, syndapin2, and phosphatidic acid in tubular recycling endosome biogenesis. Mol Biol Cell, 2013;24(11):1776–90, S1–15. 63. Zhang, J., Naslavsky, N., and Caplan, S. EHDs meet the retromer: complex regulation of retrograde transport. Cell Logist, 2012;2(3):161–5. 64. Gokool, S., Tattersall, D., and Seaman, M.N. EHD1 interacts with retromer to stabilize SNX1 tubules and facilitate endosome‐to‐Golgi retrieval. Traffic, 2007;8(12):1873–86. 65. Scott, C.C., Vacca, F., and Gruenberg, J. Endosome maturation, transport and functions. Semin Cell Dev Biol, 2014;31:2–10. 66. Hessvik, N.P. and Llorente, A. Current knowledge on exosome biogenesis and release. Cell Mol Life Sci, 2018;75(2):193–208. 67. Schmidt, O. and Teis, D. The ESCRT machinery. Curr Biol, 2012;22(4): R116–20. 68. Carlson, L.A., Shen, Q.T., Pavlin, M.R., and Hurley, J.H. ESCRT filaments as spiral springs. Dev Cell, 2015;35(4):397–8. 69. Frankel, E.B. and Audhya, A. ESCRT‐dependent cargo sorting at multivesicular endosomes. Semin Cell Dev Biol, 2018;74:4–10. 70. Shen, J., Huang, C.K., Yu, H. et al. The role of exosomes in hepatitis, liver cirrhosis and hepatocellular carcinoma. J Cell Mol Med, 2017;21(5): 986–92. 71. Maji, S., Matsuda, A., Yan, I.K., Parasramka, M., and Patel, T. Extracellular vesicles in liver diseases. Am J Physiol Gastrointestinal Liver Physiol, 2017;312(3):G194–G200. 72. Cornu M., de Caudron de Coquereaumont, G., and Hall, M.N. mTOR signaling in liver disease, in Signaling Pathways in Liver Diseases 3rd edn (eds. J.‐F. Dufour and P.‐A. Clavien), John Wiley & Sons, Chichester, 2015, pp. 314–25. 73. Settembre, C., De Cegli, R., Mansueto, G. et al. TFEB controls cellular lipid metabolism through a starvation‐induced autoregulatory loop. Nat Cell Biol, 2013;15(6):647–58. 74. Napolitano, G. and Ballabio, A. TFEB at a glance. J Cell Sci, 2016;129(13):2475–81. 75. Gissen, P. and Arias, I.M. Structural and functional hepatocyte polarity and liver disease. J Hepatol, 2015;63(4):1023–37. 76. Sewell, R.B. et al. Pharmacologic perturbation of rat liver lysosomes: effects on release of lysosomal enzymes and of lipids into bile. Gastroenterology, 1988;95(4):1088–98. 77. Tooze, S.A., Abada, A., and Elazar, Z. Endocytosis and autophagy: exploitation or cooperation? Cold Spring Harbor Perspect Biol, 2014;6(5):a018358. 78. Mizushima, N. Autophagy: process and function. Genes Dev, 2007;21(22): 2861–73. 79. Czaja, M.J., Ding, W.X., Donohue, T.M., Jr. et al. Functions of autophagy in normal and diseased liver. Autophagy, 2013;9(8):1131–58. 80. Alers, S., Loffler, A.S., Wesselborg, S., and Stork, B. Role of AMPK‐mTOR‐ Ulk1/2 in the regulation of autophagy: cross talk, shortcuts, and feedbacks. Mol Cell Biol, 2012;32(1):2–11. 81. Mercer, T.J., Gubas, A., and Tooze, S.A. A molecular perspective of mammalian autophagosome biogenesis. J Biol Chem, 2018;293(15):5386–95. 82. Lamb, C.A., Yoshimori, T., and Tooze, S.A. The autophagosome: origins unknown, biogenesis complex. Nat Rev Mol Cell Biol, 2013;14(12):759–74. 83. Reggiori, F. and Klionsky, D.J. Autophagosomes: biogenesis from scratch? Curr Opin Cell Biol, 2005;17(4):415–22.

73

84. Schulze, R.J., Weller, S.G., Schroeder, B. et al. Lipid droplet breakdown requires dynamin 2 for vesiculation of autolysosomal tubules in hepatocytes. J Cell Biol, 2013;203(2):315–26. 85. Hyttinen, J.M., Niittykoski, M., Salminen, A., and Kaarniranta, K. Maturation of autophagosomes and endosomes: a key role for Rab7. Biochim Biophys Acta, 2013;1833(3):503–10. 86. Ravikumar, B., Moreau, K., Jahreiss, L., Puri, C., and Rubinsztein, D.C. Plasma membrane contributes to the formation of pre‐autophagosomal structures. Nat Cell Biol, 2010;12(8):747–57. 87. Szatmari, Z., Kis, V., Lippai, M., et al. Rab11 facilitates cross‐talk between autophagy and endosomal pathway through regulation of Hook localization. Mol Biol Cell, 2014;25(4):522–31. 88. Fader, C.M. and Colombo, M.I. Autophagy and multivesicular bodies: two closely related partners. Cell Death Differ, 2009;16(1):70–8. 89. Madrigal‐Matute, J. and Cuervo, A.M. Regulation of liver metabolism by autophagy. Gastroenterology, 2016;150(2):328–39. 90. Tasset, I. and Cuervo, A.M. Role of chaperone‐mediated autophagy in metabolism. FEBS J, 2016;283(13):2403–13. 91. Schneider, J.L., Suh, Y., and Cuervo, A.M. Deficient chaperone‐mediated autophagy in liver leads to metabolic dysregulation. Cell Metab, 2014;20(3):417–32. 92. Chao, X., Wang, H., Jaeschke, H., and Ding, W.X. Role and mechanisms of  autophagy in acetaminophen‐induced liver injury. Liver Int, 2018;38(8):1363–74. 93. Ni, H.M., Bockus, A., Boggess, N., Jaeschke, H., and Ding, W.X. Activation of autophagy protects against acetaminophen‐induced hepatotoxicity. Hepatology, 2012;55(1):222–32. 94. Gomes, L.C., Di Benedetto, G., and Scorrano, L. During autophagy mitochondria elongate, are spared from degradation and sustain cell viability. Nat Cell Biol, 2011;13(5):589–98. 95. Lazarou, M., Sliter, D.A., Kane, L.A. et al. The ubiquitin kinase PINK1 recruits autophagy receptors to induce mitophagy. Nature, 2015; 524(7565):309–14. 96. Harper, J.W., Ordureau, A., and Heo, J.M. Building and decoding ubiquitin chains for mitophagy. Nat Rev Mol Cell Biol, 2018;19(2):93–108. 97. Hammerling, B.C., Najor, R.H., Cortez, M.Q. et  al. A Rab5 endosomal pathway mediates Parkin‐dependent mitochondrial clearance. Nat Commun, 2017;8:14050. 98. Yamano, K., Wang, C., Sarraf, S.A. et al. Endosomal Rab cycles regulate Parkin‐mediated mitophagy. eLife, 2018;7. 99. Wong, Y.C., Ysselstein, D., and Krainc, D. Mitochondria‐lysosome contacts regulate mitochondrial fission via RAB7 GTP hydrolysis. Nature, 2018;554(7692):382–6. 100. Brunt, E.M., Wong, V.W., Nobili, V. et al. Nonalcoholic fatty liver disease. Nat Rev Dis Primers, 2015;1:15080. 101. Spengler, E.K. and Loomba, R. Recommendations for diagnosis, referral for liver biopsy, and treatment of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Mayo Clin Proc, 2015;90(9):1233–46. 102. Thiam, A.R., Farese, R.V., Jr., and Walther, T.C. The biophysics and cell biology of lipid droplets. Nat Rev Mol Cell Biol, 2013;14(12):775–86. 103. Walther, T.C. and Farese, R.V., Jr. Lipid droplets and cellular lipid metabolism. Annu Rev Biochem, 2012;81:687–714. 104. Guo, Y., Cordes, K.R., Farese, R.V., Jr., and Walther, T.C. Lipid droplets at a glance. J Cell Sci, 2009;122(Pt 6):749–52. 105. Schmidt, C., Ploier, B., Koch, B., and Daum, G. Analysis of yeast lipid droplet proteome and lipidome. Meth Cell Biol, 2013;116:15–37. 106. Zhang, H., Wang, Y., Li, J. et al. Proteome of skeletal muscle lipid droplet reveals association with mitochondria and apolipoprotein a‐I. J Proteome Res, 2011;10(10):4757–68. 107. Sato, S., Fukasawa, M., Yamakawa, Y. et  al. Proteomic profiling of lipid droplet proteins in hepatoma cell lines expressing hepatitis C virus core protein. J Biochem, 2006;139(5):921–30. 108. Cermelli, S., Guo, Y., Gross, S.P., and Welte, M.A. The lipid‐droplet proteome reveals that droplets are a protein‐storage depot. Curr Biol, 2006;16(18):1783–95. 109. Brasaemle, D.L., Dolios, G., Shapiro, L., and Wang, R. Proteomic analysis of proteins associated with lipid droplets of basal and lipolytically stimulated 3T3‐L1 adipocytes. J Biol Chem, 2004;279(45):46835–42. 110. Schott, M.B., Rasineni, K., Weller, S.G. et  al. β‐Adrenergic induction of lipolysis in hepatocytes is inhibited by ethanol exposure. J Biol Chem, 2017;292(28):11815–28.

74

THE LIVER:  REFERENCES

111. Ong, K.T., Mashek, M.T., Bu, S.Y., and Mashek, D.G. Hepatic ATGL knockdown uncouples glucose intolerance from liver TAG accumulation. FASEB J, 2013;27(1):313–21. 112. Ong, K.T., Mashek, M.T., Bu, S.Y., Greenberg, A.S., and Mashek, D.G. Adipose triglyceride lipase is a major hepatic lipase that regulates triacylglycerol turnover and fatty acid signaling and partitioning. Hepatology, 2011;53(1):116–26. 113. Wu, J.W., Wang, S.P., Alvarez, F. et al. Deficiency of liver adipose triglyceride lipase in mice causes progressive hepatic steatosis. Hepatology, 2011;54(1):122–32. 114. Reid, B.N., Ables, G.P., Otlivanchik, O.A. et al. Hepatic overexpression of hormone‐sensitive lipase and adipose triglyceride lipase promotes fatty acid oxidation, stimulates direct release of free fatty acids, and ameliorates steatosis. J Biol Chem, 2008;283(19):13087–99. 115. Schulze, R.J., Drizyte, K., Casey, C.A., and McNiven, M.A. Hepatic lipophagy: new insights into autophagic catabolism of lipid droplets in the liver. Hepatol Commun, 2017;1(5):359–69. 116. Singh, R. and Cuervo, A.M. Lipophagy: connecting autophagy and lipid metabolism. Int J Cell Biol, 2012;2012:282041. 117. Singh, R., Kaushik, S., Wang, Y. et al. Autophagy regulates lipid metabolism. Nature, 2009;458(7242):1131–5. 118. Li, Z., Schulze, R.J., Weller, S.G. et  al. A novel Rab10‐EHBP1‐EHD2 ­complex essential for the autophagic engulfment of lipid droplets. Sci Adv, 2016;2(12):e1601470. 119. Schroeder, B., Schulze, R.J., Weller, S.G. et al. The small GTPase Rab7 as a central regulator of hepatocellular lipophagy. Hepatology, 2015;61(6):1896–907. 120. Schulze, R.J., Rasineni, K., Weller, S.G. et  al. Ethanol exposure inhibits hepatocyte lipophagy by inactivating the small guanosine triphosphatase Rab7. Hepatol Commun, 2017;1:140–52. 121. Kaushik, S. and Cuervo, A.M. Degradation of lipid droplet‐associated proteins by chaperone‐mediated autophagy facilitates lipolysis. Nat Cell Biol, 2015;17(6):759–70.

122. Martinez‐Lopez, N., Garcia‐Macia, M., Sahu, S. et  al. Autophagy in the CNS and periphery coordinate lipophagy and lipolysis in the brown a­ dipose tissue and liver. Cell Metab, 2016;23(1):113–27. 123. Sathyanarayan, A., Mashek, M.T., and Mashek, D.G. ATGL promotes autophagy/lipophagy via SIRT1 to control hepatic lipid droplet catabolism. Cell Rep, 2017;19(1):1–9. 124. Khan, S.A., Sathyanarayan, A., Mashek, M.T. et  al. ATGL‐catalyzed ­lipolysis regulates SIRT1 to control PGC‐1alpha/PPAR‐alpha signaling. Diabetes, 2015;64(2):418–26. 125. Inoue, J., Ninomiya, M., Shimosegawa, T., and McNiven, M.A. Cellular membrane trafficking machineries used by the hepatitis viruses. Hepatology, 2018;68(2):751–76. 126. Macovei, A. et al. Regulation of hepatitis B virus infection by Rab5, Rab7, and the endolysosomal compartment. J Virol, 2013;87(11):6415–27. 127. Coller, K.E. et al. RNA interference and single particle tracking analysis of hepatitis C virus endocytosis. PLoS Pathog, 2009;5(12):e1000702. 128. Selzer, L. and Zlotnick, A. Assembly and release of hepatitis B virus. Cold Spring Harb Perspect Med, 2015 Nov 9;5(12):pii: a021394. 129. Miyanari, Y. et al. The lipid droplet is an important organelle for hepatitis C virus production. Nat Cell Biol, 2007;9:1089–97 130. Salloum, S. et al. Rab18 binds to hepatitis C virus NS5A and promotes interaction between sites of viral replication and lipid droplets. PLoS Pathog, 2013;9(8):e1003513. 131. Ploen, D. et al. TIP47 plays a crucial role in the life cycle of hepatitis C virus. J Hepatol, 2013;58(6):1081–8. 132. Coller, K.E. et al. Molecular determinants and dynamics of hepatitis C virus secretion. PLoS Pathog, 2012;8(1):e1002466. 133. Inoue, J., Krueger, E.W., Chen, J. et al. HBV secretion is regulated through the activation of endocytic and autophagic compartments mediated by Rab7 stimulation. J Cell Sci, 2015;128(9):1696–706. 134. Ostrowski, M. et al. Rab27a and Rab27b control different steps of the exosome secretion pathway. Nat Cell Biol, 2010;12:19–30.

7

The Hepatocellular Secretory Pathway Catherine L. Jackson1 and Mark A. McNiven2 Institut Jacques Monod, UMR7592 CNRS Université Paris‐Diderot, Sorbonne Paris Cité, Paris, France Department of Biochemistry and Molecular Biology, and The Center for Digestive Diseases, Mayo Clinic & Foundation, Rochester, MN, USA

1 2

INTRODUCTION Hepatocytes are known to secrete scores of proteins and lipid particles into the sinusoidal space, and also play a key role in lipid homeostasis in organisms. An emerging theme over the past few years has been the intimate connection between membrane trafficking and lipid metabolism, with important implications for hepatocyte function. The secretory pathway assures processing and delivery of proteins to their correct destination, relying on highly organized vesicular trafficking machinery that includes numerous enzymes, cytoskeletal proteins, molecular motors, and coat proteins. This machinery is highly conserved in evolution, and is found in virtually all eukaryotic cells. Specialized cell types such as hepatocytes can have more specific secretion ­pathways. Very low density lipoprotein (VLDL) particles are produced uniquely by liver cells, and use the ubiquitous vesicu­ lar trafficking machinery in addition to components specific to this large cargo. Through conventional biochemical and molecu­ lar methods, the use of genetic model organisms, and recent advancements in live cell imaging, much has been learned about these trafficking pathways in hepatocytes and other epithelial cells. This chapter will focus on the molecular mechanisms by which nascent proteins and larger cargos are sequestered, pack­ aged into vesicle carriers, and targeted to specific hepatocellular destinations during the secretory process.

THE SECRETORY PATHWAY The secretory pathway begins with synthesis of proteins at the endoplasmic reticulum (ER) [1]. Next, proteins are transported to the Golgi apparatus (Figure 7.1), the central processing and

sorting station of the pathway [2]. Generally distributed ­throughout the cell and extending to the periphery, the ER is characterized by a network of interconnected membrane struc­ tures that include cisternae and tubular networks [3, 4]. The Golgi apparatus, in contrast, is localized in the perinuclear region of the cell and forms a continuous ribbon in mammalian cells [5]. This ribbon is composed of organized stacks of flat­ tened saccules, interconnected by more complex tubulo‐vesicu­ lar regions [5]. The first cis‐Golgi element and the trans‐Golgi network (TGN) are tubular membrane meshworks at the entry and exit sides of the Golgi apparatus, respectively. Early studies of the secretory pathway made use of rat liver (Figure 7.1), as well as exocrine pancreas cells, due to their high levels of secretion [6]. These studies determined that during the synthesis and transport of secreted proteins, the ER and Golgi function sequentially, and prior to packaging of secretory cargo into secretory granules or vesicles. The mechanism by which proteins are transported from a donor to an acceptor compartment in eukaryotic cells is via the budding and fusion of vesicular carriers [7] (Figure  7.2). The first step in formation of a vesicle is activation of a small G pro­ tein of the ADP ribosylation factor (Arf) family by a nucleotide exchange factor, which leads to recruitment of proteins called effectors to the membrane [10–12]. Cargo adaptors, which are coupled to a vesicle coat, are a major class of effectors of the Sar1 and Arf1 small G proteins that function at the ER and Golgi, respectively. Both Arf and Sar proteins have an N‐terminal amphipathic helix that inserts into the membrane, contributing to deformation into a highly curved bud [13]. A variety of different adaptors bind to the cargo, either transmembrane or luminal pro­ teins, to concentrate them into the forming vesicle [14]. Other effectors of Sar1 and Arf1 include lipid‐modifying enzymes, which change the composition of the membrane at the budding

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

76

THE LIVER:  EVOLUTIONARY ORIGINS OF VESICLE COATS

Figure 7.1  Elaborate vesicular processes in the hepatocyte. Thin‐­ section electron micrograph of a rat liver hepatocyte from the original collections of, Dr. Keith Porter, about 1962. Deep in the cytoplasm, parallel cisternae of the endoplasmic reticulum (ER) appear to give rise to the flattened, stacked saccules of the Golgi apparatus (G). Scores of small vesicular profiles can be seen in the Golgi region, as well as larger lipoprotein‐filled secretory vesicles budding from the trans side of the Golgi. Specific coats, accessory proteins, and address tags characterize many of these vesicles. G, Golgi; ER, endoplasmic reticulum; Lys, lysosome.

site, also contributing to membrane deformation [15–17]. Polymerization of the outer shell of the coat also aids in shaping the membrane into a highly curved bud. Once the bud has formed and undergone scission from the donor compartment membrane, uncoating of the vesicle ensues. The first step is hydrolysis of GTP on the Arf family small G protein, which can commence even before the vesicle has under­ gone fission from the donor compartment [18]. The interaction between vesicle cargo and the coat maintains the coat on the vesicle membrane until the appropriate time for uncoating, which must occur prior to fusion of the vesicle with its target membrane. Transport in the anterograde direction from the ER to the Golgi apparatus is mediated by coat protomer II (COPII)‐coated vesicles (Figure  7.2). On the ER membrane, subunits of the COPII coat are recruited upon activation of Sar1 by its exchange factor Sec12, to form buds and ultimately generate vesicles. COPII vesicle formation occurs at specialized regions of the ER that are called ER exit sites (ERES). In mammalian cells, COPII vesicles undergo homotypic fusion, and the resulting compart­ ment recruits exchange factor Golgi brefeldin A resistant factor 1 (GBF1), its substrate Arf1, and the coat protomer I (COPI) coat, which form COPI vesicles (Figure  7.2). The sorting compart­ ment formed upon GBF1–Arf1–COPI recruitment, called the ER/Golgi intermediate compartment (ERGIC), is made up of a complex network of tubules [19]. COPI‐coated vesicles are also thought to mediate the retrograde transport of resident ER pro­ teins and Golgi enzymes from the Golgi apparatus back to the ERGIC and to the ER [20, 21]. In addition to secretory proteins,

membrane proteins and lysosomal enzymes pass through the secretory apparatus to reach their final destinations, which reflects the important sorting role of the Golgi apparatus. A particularly active sorting region of the Golgi is the TGN, where several distinct types of vesicles containing specific cargo proteins and harboring particular coat proteins arise [22–24]. Several of these coats contain cargo adaptor proteins (APs) recruited to membranes by the Arf1 or Arf3 small G proteins. The APs are four‐subunit complexes containing two large and two small subunits. AP‐1, and probably AP‐3 as well, recruit clathrin upon membrane binding to form a clathrin–adaptor coat. However, this is not the case for AP‐4, which forms a non‐ clathrin coat [22, 23]. Another important set of adaptors at the TGN are the Golgi‐localized γ‐ear‐containing, ADP ribosyla­ tion factor‐binding (GGA) proteins (GGA1, GGA2, and GGA3), which bind clathrin and transport distinct cargo from the TGN to endosomes. Secretory vesicles (or granules) carry­ ing constitutively secreted proteins from the TGN to the cell surface are believed to be uncoated vesicles, as no coats have yet been identified in mammalian cells [23].

EVOLUTIONARY ORIGINS OF VESICLE COATS The COPII, COPI, and AP–clathrin coats are all thought to share a common origin, arising from a primordial coat com­ plex. All three vesicle coats have subunits containing a β‐pro­ peller domain followed by an α‐solenoid, also known as a helix‐turn‐helix domain [8] (Figure  7.2). Strikingly, only eukaryotes, and a few rare prokaryotes with rudimentary vesic­ ular trafficking pathways, have proteins containing a β‐propel­ ler motif fused to an α‐solenoid fold [8]. Hence this class of protein is strictly correlated in evolution with the necessity to bend membranes, such as in vesicle formation. This β‐ propeller–α‐solenoid structure is also found in nuclear pore components that bind to the highly curved membrane of the pore [25]. These observations led to the protocoatomer hypoth­ esis, which postulates that COPII, COPI, and clathrin–AP coats, as well as some nucleoporins, all shared a primordial ancestor [25]. Additional support for this hypothesis comes from clear sequence homology between the β‐ and γ‐COP sub­ units of COPI and the large subunits of the AP‐1 and AP‐2 adaptors [8, 26, 27]. However, despite this homology, the structures of the COPI and AP–clathrin coats are significantly different (Figure 7.2). The β‐propeller–α‐solenoid subunits of both coats polymerize due to interactions between α‐solenoids, but in the case of COPI, these subunits make direct contact with the membrane, and are positioned beside the other sub­ complex of COPI. In contrast, in the AP–clathrin coat, clathrin forms a cage around the vesicle that does not directly touch the membrane (Figure 7.2a). COPII resembles AP–clathrin in that the β‐propeller–α‐solenoid subunit (Sec31), along with Sec13, forms an outer cage that does not directly contact the mem­ brane (Figure 7.2a). However, the geometry of the COPII and AP–clathrin coats is quite different based on in vitro structural data (Figure  7.2a), and visualization of vesicles within cells (Figure 7.2b).



7:  The Hepatocellular Secretory Pathway

(a)

β-propeller α-solenoid small GTPase Sec23 Sec24 AP/COP large subunit (β) AP/COP large subunit (EGADZ) Clathrin AP/COP medium subunit AP/COP small subunit Transmembrane cargo α-COP

Sec13/Sec31

(b)

77

β’-COP

COPII

COPI

AP-1 + clathrin

COPII

COPI

Clathrin

A

B

C

D

E

F

Figure 7.2  Comparison of the COPII, COPI, and AP–clathrin coats. (a) All three coats contain β‐propeller–α‐solenoid components (red spheres and blue rods), but they are arranged in different ways in each coat. In COPII and AP–clathrin coats, these subunits form an outer polyhedral layer. In COPI, on the other hand, the β‐propeller–α‐solenoid‐containing subunits directly contact the membrane as well as the other COPI subunits. The other subunits of COPI and the adaptor subunits share both sequence and structural homology. All three coats are recruited to membranes by a member of the Arf small G protein (GTPase) family: Sar1 in the case of COPII, and Arf1 for both COPI and AP–clathrin coats. Reproduced from [8] with permission of Elsevier. (b) Visualization in their native cellular environment of COPII (A,D), COPI (B,E) and AP‐1–clathrin (C,F)‐coated vesicles by cryoelectron tomography. Imaging was performed on Chlamydomonas reinhardtii cells. Panels A–C show cross‐sections through each vesicle; panels D–F are grazing slices through the coats at the tops of the same vesicles. Note that the geometry of the COPII and clathrin coats is visible in the lower panels, revealing the triangular Sec13/31 COPII lattice (D) and clathrin triskelions (F). Scale bars, 50 nm. Reproduced from Bykov 2017 [9], https://cdn.elifesciences.org/articles/32493/elife‐32493‐v2.pdf. Licensed under CCBY 4.0.

78

THE LIVER:  COPI‐COATED VESICLES

In addition to β‐propeller–α‐solenoid subunits, coats also share similar features such as recruitment to membranes by a member of the Arf family of small G proteins, as described above. Other ancestral motifs found in all coat complexes are longin domains (present in the medium and small COPI/AP subunits, Figure  7.2a) and coiled coils [8]. This ensemble of ancient conserved features supports the idea of an ancient pri­ mordial coat, which split into the three major coats of the secre­ tory pathway: COPII, COPI, and AP–clathrin.

COPI‐COATED VESICLES COPI‐coated vesicles are required for intra‐Golgi trafficking and for recycling from the Golgi apparatus back to the ER. The COPI coat consists of seven subunits (α, β, β′, γ, δ, ε, ζ COP) that form a 680 kDa complex and associate with the small GTP‐binding protein Arf1. COPI sorts cargo proteins into form­ ing vesicles as it induces membrane curvature [28] (Figure 7.2). COPI‐coated vesicles are best known for their function in the retrieval of ER‐resident proteins from the Golgi. Selected COPI subunits are known to associate with dilysine amino acid motifs in the cytoplasmic domains of ER‐resident enzymes [20]. Furthermore, several COPI temperature‐sensitive mutants in yeast have revealed defects in transport between the ER and the Golgi at the restrictive temperature [21]. Further, COPI‐coated vesicles have been shown to mediate retrograde transport of resident ER proteins by using the KDEL (single‐letter code for amino acids) receptor [29], confirming a retrieval role for COPI coat proteins. Regulators of Arf–GTP binding and GTP hydrolysis control the spatio‐temporal localization of Arf protein activation and down­ stream signaling. GDP release from Arf proteins is catalyzed by Arf guanine nucleotide exchange factors (GEFs), allowing the more abundant GTP to bind. This nucleotide exchange activity is contained within the Sec7 domain, an evolutionarily conserved sequence first identified in yeast Sec7p [30–32]. A Sec7 domain is present in all Arf GEFs identified to date. The function of Sec7 domains was first identified in the yeast Gea1p protein [33]. The human ortholog of yeast Gea1p, GBF1, was identified by P. Melançon and colleagues [34]. GBF1 is a Golgi‐localized protein whose overexpression was found to confer resistance to the fungal toxin brefeldin A, which completely blocks secretion, and causes disassembly of the Golgi apparatus and eventually its fusion with the ER [10]. The Arf GTPase‐activating proteins (GAPs) catalyze the hydrolysis of GTP on Arf family proteins, a function carried out by a conserved GAP domain, which contains a zinc finger [10]. The primary sequence homology of the catalytic domains of the Arf GEFs and GAPs accelerated their identification and allowed studies of their evolution. Phylogenetic analyses of the Arf GAPs has revealed that they have likely co‐evolved with their Arf sub­ strates [35], and a similar conclusion very likely holds for the Arf GEFs [36]. From these results we can conclude that Arf proteins function in a tightly coordinated manner with their regulators. There are seven subfamilies of Arf GEFs in eukaryotic cells [30]. Members of the GBF/Gea and BIG/Sec7 subfamilies of Sec7 domain GEFs use class I Arf proteins as substrates [10], and in addition, GBF1 possibly functions on class II Arf pro­ teins [34, 37]. In humans, there is one GBF/Gea member, GBF1,

and two members of the BIG/Sec7 subfamily, brefeldin A inhib­ ited GEF 1 (BIG1) and BIG2. These two subfamilies have a similar domain structure, probably because they have descended from a common ancestor, but have different steady‐state locali­ zations and functions [10, 30, 31]. In animal cells, including hepatocytes, and yeast, the major steady‐state localization of  GBF/Gea GEFs is at the early Golgi, whereas BIG/Sec7 GEFs are present predominantly at the late Golgi, ­including the TGN [10, 31]. COPI vesicles mediate retrograde transport of components that need to be recycled from the Golgi and ERGIC back to earlier compartments and to the ER. Arf proteins (primarily Arf1 and Arf3) at the TGN recruit the heterotetrameric clathrin adaptor proteins, AP‐1, AP‐3, and AP‐4, and the three mono­ meric Golgi‐localized γ‐ear‐containing, ADP ribosylation fac­ tor‐binding proteins (GGAs 1–3) [24, 38]. To explain how one or two Arf proteins can recruit so many different coats to differ­ ent membrane sites, several groups have found evidence that the GEFs activating Arfs at different sites determine the coat that will be recruited. GBF1 at ERGIC and the cis‐Golgi is responsi­ ble for COPI recruitment, and the BIG1 and BIG2 proteins at the trans‐Golgi are responsible for recruiting the AP–clathrin coats [10]. Demonstration of a direct interaction between GBF1 and COPI coat subunits provides a molecular explanation for this specificity [39]. Interaction of GBF1 with COPI prior to Arf1 activation ensures that COPI is stabilized on the membrane where GBF1 activates Arf1 [39]. Removal of the coat from a vesicle (uncoating) is required for fusion of the vesicle to the acceptor membrane. The first step in vesicle uncoating occurs upon hydrolysis of GTP on the Arf family G protein that maintains the coat on the membrane [40, 41]. Even after hydrolysis of GTP on Sar1 or Arf1, the coat (or at least a portion of the coat) remains on the vesicle due to inter­ actions with other vesicle proteins such as cargo molecules [40, 42]. In the case of COPI vesicles, GTP hydrolysis on Arf1 is mediated by Arf GAP proteins [28, 43, 44]. Arf GAP1 is recruited only to highly curved membranes through its ALPS motif, which ensures that the coat remains on the budding vesi­ cle until it has achieved a mature, highly curved morphology [45, 46]. It was originally thought that the coat was lost quite quickly after budding. However, interaction of the coat with a tethering complex on the acceptor membrane has recently been demonstrated for multiple types of vesicles [47, 48]. For COPI vesicles transported to the ER, the Ds11 complex present on the ER membrane acts as a vesicle docking site through direct inter­ action with the COPI coat [49–51]. Moreover, this coat–tether interaction stimulates uncoating of the vesicle [51]. In addition to fusing with the ER, COPI vesicles also mediate intra‐Golgi retrograde trafficking. Interactions between COPI subunits and the Golgi COG tethering complex [52, 53], and with the cis‐ Golgi tether p115 [54], are also likely to be involved in targeting vesicles to these Golgi acceptor membranes. GBF1 has been shown to have a crucial function in the ­replication of numerous viruses, including hepatitis C [55] and hepatitis E [56]. Hepatitis A, B, C, D, and E viruses cause the majority of liver disease across the globe, and represent a world­ wide health problem. For the well‐studied hepatitis C virus, lipid metabolism is an important component in infection, and lipid metabolism pathways are subverted for propagation of this virus



7:  The Hepatocellular Secretory Pathway

[57]. The catalytic domain of GBF1 is required for HCV to form functional viral replication complexes [58]. Arf4 and Arf5 are also required for HCV replication, through their roles in lipid homeostasis in hepatocytes [58]. Therefore, understanding the mechanisms by which hepatitis viruses infect hepatocytes could increase our knowledge of how the secretory pathway and lipid homeostasis are coordinated in cells, in addition to providing avenues for treatment of these important human pathogens.

COPII‐COATED VESICLES COPII‐coated vesicles function in ER‐to‐Golgi anterograde transport Protein transport from the ER to the Golgi complex is mediated by vesicular carriers that originate from morphologically defined, specialized ribosome‐free regions of the ER called ER exit sites (or ER export sites) (ERES) [59]. ER buds and an abundance of small vesicular profiles at these sites, which are distinct from smooth ER could be clearly visualized in the clas­ sic electron micrographs of pancreatic acinar cells by Jamieson and Palade [60]. Studies using live cell imaging demonstrated that these transitional areas are formed transiently and randomly from the ER [61]. The cargo carriers that mediate ER‐to‐Golgi protein transport have been subjects of intense study. As for many of the protein trafficking steps along the secretory pathway, molecular compo­ nents were first identified in studies using the yeast Saccharomyces cerevisiae as a model system. Yeast has provided a powerful combination of genetics, molecular biology, and biochemistry to elucidate mechanisms both in vitro and in vivo. Isolation of con­ ditional sec (secretion) mutants allowed the identification and characterization of numerous components of trafficking path­ ways, including all essential subunits of the COPII coat [32]. By using yeast membranes for ER vesicle budding in vitro, cell‐free assays for vesicle budding and fusion could be reconstituted. These systems enabled the isolation of ER‐derived transport ves­ icles that were 60–65 nm in diameter and exhibited a distinct electron‐dense coat on their surface, termed COPII [62]. COPII coat components were identified from the purified vesicles and revealed to be distinct from those of COPI coats [62]. Early studies in various cell types including human hepatoma HepG2 cells [63] reported that secretory cargo protein was present at bud sites upon ER exit and within small 40–80 nm carrier vesicles and tubules referred to as vesicular–tubular clusters (VTCs). In pan­ creatic cells, similar structures were shown to be COPII‐­positive. The ER cisternae near VTCs often exhibited budding profiles that were decorated with an electron‐dense, honeycomb‐patterned coat (Figure 7.2b), which was identified as COPII by immunogold labe­ ling antibodies against COPII coat components [64].

COPII‐coated vesicle protein components and vesicle formation The COPII coat components, which are distinct from the COPI coatomer, were postulated to induce vesicle budding of ER‐derived transport vesicles because only the GTPase Sar1,

79

Sec13/31, and Sec23/24 were found to be necessary for the generation of COPII‐coated vesicles from washed micro­ somes [62]. Sec12, a GTP‐GEF for Sar1 that is localized to the ER initiates COPII vesicle biogenesis. After the activa­ tion of Sar1, the adaptor coat components Sec23/24 are recruited to bind cargo. Subsequently, the structural β‐ propeller–α‐solenoid Sec13/31 coats are recruited to form the coat and to contribute to membrane deformation. The selection of cargo including transmembrane proteins and v‐ SNAREs (soluble NSF attachment receptors) by Sec23/24 is mediated by three different sites for recognizing ER‐to‐Golgi v‐SNAREs on Sec24. The interaction is regulated by the assembly state of the SNAREs, which suggests that COPII proteins can be involved in vesicle fusion specificity in the ER‐to‐Golgi step [12]. In the final stages of COPII vesicle biogenesis, the GTP used during vesicle formation is hydrolyzed by Sec23, a Sar1‐ specific GAP. Upon GTP binding by Sar1, membrane inser­ tion of the N‐terminal α helix of Sar1 deforms synthetic liposomes into narrow tubules. Mutation of the helix led to a defect in membrane curvature and the formation of vesicles from native ER, although the recruitment of coat proteins appeared to be normal [13]. Moreover, inhibition of GTP hydrolysis by Sar1 resulted in COPII‐coated vesicles that failed to detach from the ER, suggesting that regulation of the N‐terminus of Sar1 by GTP binding and hydrolysis controls COPII vesicle fission [65]. Although the minimal requirements for in vitro vesicle bio­ genesis are the five molecules listed above, other components essential for COPII vesicle formation in different cell types have been identified. Sec16 was first identified in yeast, and found to be essential and required for COPII vesicle formation in vivo [12]. Sec16 is present throughout eukaryotes, and is thought to play a role in scaffolding other COPII components and organ­ izing ERES through interaction with other ERES components [14]. Sec16 is also required for formation of stress granules con­ taining COPII components that are formed when cells are starved and consequently shut down the secretory pathway [66]. Hence Sec16 has broad functions in regulation of COPII com­ ponents in various cellular physiological states, including cell growth and starvation. The TRK‐fused gene protein (TFK) and the cargo receptor transport and Golgi organization 1 (TANGO1)/cutaneous T cell lymphoma‐associated antigen (cTAGE5) proteins are more recently identified ERES components that function in ERES organization [67, 68]. TFK binds to the Sec23/24 inner coat and contributes to uncoating of COPII vesicles [67]. As described for COPI vesicles, the uncoating of COPII vesicles also occurs late in a transport step, after targeting of the vesi­ cle to the a­ cceptor compartment [47, 48]. Interestingly, TFK has an N‐­ terminally unstructured region in addition to a COPII‐binding site, which confers properties of liquid droplet phase separation in vitro. In the absence of TFK, COPII vesi­ cles disperse, suggesting that TFK might function to maintain vesicles close to their target membranes though forming a liq­ uid droplet ­structure between ERES and ERGIC target mem­ branes [67]. In addition to TRK, the COPII vesicle coat component Sec23 also interacts with the tethering complex transport protein particle (TRAPP) at the acceptor

80

THE LIVER:  TRANS‐GOLGI NETWORK‐DERIVED VESICLES

compartment membrane prior to complete release of the COPII coat and fusion [69]. TANGO1 was identified in a screen for genes affecting secre­ tion in flies, and was later demonstrated to play an essential role in collagen secretion in keratinocytes and fibroblasts [68]. Like TFK, TANGO1 interacts with the inner components of the COPII coat, Sec23 and Sec24, but not the outer cage compo­ nents, and localizes to ERES [68]. TANGO1 is also expressed in liver cells [70], where it functions in export of specific cargos, as described below. In mammalian cells, GST–hSec23‐binding columns allowed the identification of a 125 kDa protein, p125, that is expressed in liver, and shares sequence homology with p­ hosphatidic acid (PA)‐preferring phospholipase A1, which produces lysophos­ pholipids [71]. Recently, the function of lysophospholipids in lowering the rigidity of the ER membrane during COPII ­vesicle budding was demonstrated both in vitro and in cells [72]. These results indicate the importance of membrane ­com­position and membrane biophysical properties in vesicle formation.

LARGE COPII CARRIERS TRANSPORT VLDL PARTICLES FROM THE ER TO THE GOLGI Recently, it has become clear that large cargos, too large to be incorporated into classic 60 nm COPII vesicles, nevertheless use COPII for their transport out of the ER to the Golgi in the secretory pathway [73, 74]. In hepatocytes, VLDL particles are formed from the ER membrane on the luminal side, and from there are transported to the Golgi apparatus [70, 75]. However, each VLDL particle is up to 90 nm in diameter, and especially if multiple particles enter the same carrier, this cargo is too large to be incorporated into classic COPII vesicles. However, there is evidence for COPII involvement in ER–Golgi transport of VLDL [76, 77]. In addition, the TANGO proteins are impor­ tant components in this process, acting as scaffolds for the assembly of these larger COPII vesicular structures. TANGO1 was first identified as a transmembrane cargo receptor for luminal procollagen, another cargo too large for classic COPII vesicles. As a cargo receptor, TANGO1 was found to link lumi­ nal collagen to the COPII coat on the other side of the ER membrane [68]. TANGO1 and an interacting partner called TALI are both required for optimal secretion of VLDL particles in the hepato­ cyte cell line HepG2 [70]. TANGO1 and TALI not only interact, but the complex interacts with apolipoprotein B (ApoB100), required for production of VLDL particles via its interaction with triglycerides [70]. All three proteins colocalize in HepG2 cells, and interaction of TANGO/TALI with apoB present on VLDL particles results in their recruitment to ER exit sites, where they are packaged into COPII carriers [70]. Interestingly, ERGIC53, a marker of ERGIC membranes, colocalizes with apoB‐containing structures at ER exit sites [70]. This and other data support the conclusion that ERGIC is involved in forma­ tion of VLDL particles, as it is for collagen‐containing COPII carriers [78].

TRANS‐GOLGI NETWORK‐DERIVED VESICLES Protein sorting and trafficking events at the trans‐Golgi network The TGN is a tubulovesicular network that acts primarily in sorting proteins to their final destinations. Within this membra­ nous reticulum, cargo proteins are segregated efficiently into distinct vesicles that will be targeted to various compartments, including the endosome/lysosome system, the apical or basolat­ eral domains in polarized cells, and the secretory granule pool. In regulated secretory cells, this granule pool awaits the appro­ priate extracellular stimulus to exocytose its contents. The TGN is known to interact intimately and dynamically with the endosomal apparatus, especially during events related to receptor recycling, endosomal maturation, and the delivery of lysosomal enzymes [23]. A large number of different types of vesicles are formed at the trans‐Golgi and TGN, reflecting the intense sorting activity that occurs at this exit face of the Golgi. The best characterized are clathrin‐coated vesicles, which are linked to the membrane through various APs, as described briefly above. Other non‐ clathrin‐coated vesicles are characterized by adaptor proteins homologous to AP‐1, but which carry different cargo.

Formation of clathrin‐coated and clathrin‐ independent vesicles at the TGN Clathrin was the first coat to be visualized in the Golgi region of mammalian cells. An electron‐dense, spiked coat on TGN mem­ branes in liver hepatocytes could be seen clearly in the original micrographs of Novikoff and Yam [80] (Figure 7.3a). The first‐ identified and best‐characterized cargo adaptors are heterotrim­ eric adaptor protein complexes AP‐1 to AP‐5, three of which mediate sorting at the TGN (AP‐1, AP‐3, and AP‐4). All three of these coats are recruited to membranes by Arf1 and/or Arf3. AP‐1 in addition requires the phosphoinositide phosphatidylin­ ositol 4‐phosphate (PI(4)P) (Figure  7.4). TGN‐derived CCVs are now known to have a coat composed of clathrin triskelia and the AP complexes AP‐1 (made up of γ‐, β1‐, μ1‐, and σ1‐adaptin subunits) [22] and AP‐3 (δ‐, β3‐, μ3‐, and σ3‐adaptin subunits) [81]. AP‐4 does not bind to clathrin. The subunit μ1 of AP‐1 is responsible for binding to YXXΦ motifs within cargo proteins, whereas β1 interacts with dileucine‐based signals in cargo. The γ‐adaptin subunit recruits accessory proteins to the site of vesi­ cle formation [23]. There are two forms of the μ1 subunit of AP‐1, μ1A and μ1B. μ1A is ubiquitous, whereas μ1B is found only in polarized epi­ thelial cells [22, 23]. The two adaptor complexes AP‐1A and AP‐1B, containing μ1A and μ1B, respectively, are both present in polarized cells, and function in basolateral protein sorting. AP‐1B is responsible for sorting of a specific subset of cargo proteins at the TGN in epithelial cells to the basolateral plasma membrane, including the LDL receptor and interleukin 6 recep­ tor β chain [82]. A second type of coat complex recruited to TGN membranes by Arf1 and PI(4)P is the family of monomeric adaptors, the



7:  The Hepatocellular Secretory Pathway

(a)

(b)

81

VLDL transport vesicles

ER

Golgi apparatus

VLDL VLDL

Sar1b Sec23/24 Sec13/31 TANGO

VLDL VLDL

VLDL

ApoB100

VLDL

Sar1a Sec23/24 Sec13/31

Classic COPII vesicles

Figure 7.3  (a) VLDL particles are present in dilated elements (arrows) within the Golgi apparatus, and in secretory vesicles forming within the trans‐Golgi and TGN. Rough ER is evident by the presence of ribosomes. G, Golgi; arrows, Golgi elements containing VLDL particles; tGE, trans‐Golgi elements containing VLDL particles; C, clathrin‐coated vesicle; R, ribosomes; P, peroxisome. Magnification: 44,000×. Reproduced with permission from [80]. ©1978 P.M. Novikoff and A. Yam. Originally published in Journal of Cell Biology. https://doi.org/10.1083/jcb.76.1.1 (b) Schematic diagram depicting VLDL trafficking from the ER to the Golgi in hepatocytes. VLDL particles form from the ER membrane on the luminal side, and require apolipoprotein B 100 (apoB100). The particles bud into the ER lumen, then are packaged into VLDL transport vesicles using the COPII machinery (Sec23/24, Sec13/31) as well as TANGO. The diameter of VLDL transport vesicles is approximately 110 nm [79], suf­ ficient to enclose VLDL‐sized particles, and larger than classic COPII vesicles (60–70 nm diameter on average). The latter contain nascent secretory and transmembrane cargo proteins synthesized on ER‐localized ribosomes, and translocated into the ER lumen. Both VLDL transport vesicles and classic COPII vesicles require COPII proteins for their budding from the ER membrane, but use different Sar1 small G proteins: Sar1b for VLDL transport vesicles and Sar1a for classic COPII vesicles. Upon fission of the vesicle from the ER, both types of vesicles are targeted to the acceptor compartment, uncoat, then fuse with the acceptor compartment membrane.

R

ab

Golgi

4)P

PI(

LE/MVB

P2

)

,5

(4 PI

Rab9

6

Rab8

R ab

11

Basolateral membrane

b2

Ra

BC

1

Rab

Apical membrane

ER

Figure 7.4  Specific small Rab GTPases and phosphoinositides are localized to distinct membrane compartments along the hepatocyte secretory pathway. Rab1 and Rab2 are believed to mediate the traffic of nascent proteins from the ER to the cis‐Golgi. Rab6 is a kinesin‐associated Rab that mediates transport within the Golgi stacks, Rab11 regulates exit from the TGN, and Rab8 functions in the targeting of secretory vesicles from the Golgi to the plasma membrane. Rab9 has a role in late endosome‐to‐TGN transport. PI(4)P is significantly enriched in the Golgi and is believed to mediate the recruitment of multiple proteins that support vesicle formation at the TGN. PI(4,5)P2 at the Golgi functions with Arf1 and phospholipase D (PLD) and is also involved in vesicle formation. LE/MVB, late endosome/multivesicular body; ER, endoplasmic reticulum; BC, bile canalicular membrane.

82

THE LIVER:  COORDINATION BETWEEN THE SECRETORY PATHWAY AND LIPID METABOLISM

Golgi‐localized, γ‐ear‐containing, Arf‐binding proteins (GGAs). There are three known GGA proteins in mammalian cells: GGA1, GGA2, and GGA3 [23]. Each GGA protein contains three domains: VHS (Vps27, Hrs, Stam), GAT (GGA and TOM), and γ‐adaptin ear (GAE). Through the VHS domain, the GGAs recognize DXXLL motifs in specific cargo such as the mannose 6‐phosphate receptor (MPR) tail and thus provide a sorting function between the TGN and endosomes [83]. The GAT domain interacts with Arf; mutations in this domain abro­ gated recruitment of GGAs to the TGN as well as binding to Arf [84]. The GAE domain is homologous to the GAE domain of the AP‐1 complex‐recruiting accessory proteins [85]. CCVs generated from the TGN are involved in the transport of newly synthesized lysosomal enzymes. Newly synthesized, soluble lysosomal enzymes acquire mannose 6‐phosphate resi­ dues at the cis‐Golgi, and these residues are recognized by the MPRs at the TGN [23]. The GGAs are essential for packaging MPRs into CCVs and for transporting them from the TGN to endosomes [86]. Another adaptor complex at the TGN, AP‐3, is expressed ubiquitously, is localized to buds/vesicles associated with the TGN, and can interact with clathrin and with sorting signals in the cytoplasmic tails of lysosomal membrane proteins [23]. Correct targeting of lysosomal‐associated membrane pro­ tein‐1 (LAMP‐1) and lysosomal integral membrane protein‐2 (LIMP‐2) are mediated by the AP‐3 adaptor complex [81]. In addition to the heterotetrameric and GGA adaptor complexes, additional adaptors have been identified, including epsin‐related proteins [23]. In addition to Arf family small G proteins, another family of small G proteins, called Rabs, are involved in each trafficking step of the secretory pathway [87]. Rab1, along with Rab2, are known to function in trafficking from the ER and ERGIC com­ partments [59, 87]. Other Rab GTPases, including Rab6, Rab19, Rab30, Rab33, Rab36, Rab40, Rab41, and Rab43 func­ tion at the Golgi, and Rab8 (the ortholog of Sec4 in yeast) is crucial in the exocytic pathway for late Golgi to plasma mem­ brane transport. Rab3, Rab26, Ra27, Rab37, and Rab2 have also been implicated in TGN to plasma membrane trafficking [87]. Rab8 and Rab11, along with Arf4 and Arf regulators, function in trafficking from the TGN to cilia in photoreceptor cells [88] (Figure 7.4).

Scission of TGN‐derived vesicles Since the localization of the large GTPase dynamin to the Golgi apparatus, in addition to its known cell surface distribution, was reported, it has been suggested that vesicle fission at the Golgi apparatus is mediated by this molecular pinchase. Dynamin’s association with Golgi membranes was demonstrated bio­ chemically with specific antibodies that detected dynamin in immunoblots of purified rat liver Golgi fractions and by the immuno‐isolation of Golgi membranes on magnetic beads coated with those dynamin antibodies [89]. Cortactin, an actin‐ binding protein, in a complex with dynamin, functions in post‐ Golgi transport in liver cells. By using in vitro or intact cell experiments, it was shown that activation of Arf1 recruits actin, cortactin, and dynamin to Golgi membranes. Disruption of cort­ actin–dynamin interactions reduced dynamin recruitment to the Golgi and blocked the transit of nascent proteins from the TGN,

which indicates an essential role of the cortactin–dynamin com­ plex in TGN function [90]. The dynamins are believed to form large helical polymers from which many interactive proline‐rich tail domains extend. These domains bind to a variety of SH3‐domain‐containing proteins, many of which appear to be actin‐binding proteins such as Abp1 and syndapin. These findings suggest that the dynamin family acts as a “polymeric contractile scaffold” at the interface between biological membranes and filamentous actin (Figure 7.4). A number of non‐dynamin‐mediated mechanisms for scis­ sion of vesicles, including those formed at the TGN, have recently been described [91]. In the case of Rab6 vesicles formed at the TGN, cytoskeletal motors play a crucial role. Active Rab6‐GTP recruits myosin II to membranes, where its motor activity on filamentous actin creates the force necessary to mediate vesicle fission [92].

COORDINATION BETWEEN THE SECRETORY PATHWAY AND LIPID METABOLISM The secretory pathway is intimately linked to lipid metabolism pathways [93]. The ER is one of the major sites within the cell for lipid synthesis, with multiple pathways operating to shuttle lipids to different destinations. Both the phospholipids that make up cel­ lular membranes, and neutral storage lipids (triglycerides and cho­ lesterol esters) are synthesized by enzymes in the ER membrane [94, 95]. In addition to their synthesis, neutral lipids are also thought to be packaged into lipid droplets in the ER membrane [95]. A regulatory switch between channeling of fatty acids into either phospholipids for secretory membrane production or into neutral lipids for storage in lipid droplets (LDs) was first identified in yeast [96]. Subsequent work in mammalian cells has uncovered similar regulatory mechanisms. A critical control point in this switch is the metabolism of phosphatidic acid (PA), which serves as a precursor of both phospholipids and storage triglycerides [97]. Indeed, PA can either be dephosphorylated by PA phosphatases to produce diglyceride (DAG), which is then fatty acid acylated to form triglycerides, or converted by CDP‐DAG synthase to CDP‐ DAG, a precursor of all the major cellular phospholipids. Hence the action of PA phosphatases is responsible for channeling PA towards storage lipid synthesis. There is one PA phosphatase in yeast, Pah1, and three in mammalian cells, lipin1, lipin2, and lipin3. A lipin1‐knockout mouse has a lipodsystrophy phenotype, and Pah1 is required for LD formation in yeast, showing the importance of these enzymes for lipid storage [97]. Pah1 and lipins are localized both to the nucleus and cytoplasm, and are involved in transcriptional regulation of lipid metabolism genes [97]. In addition to ubiquitous regulatory pathways linking the  secretory pathway and lipid metabolism, hepatocytes also use  the secretory pathway for transport of VLDL particles (Figure 7.3a), a central function in organismal lipid homeostasis [75, 79, 98]. The liver plays a key role in the body both in stor­ age of lipids and in their distribution to other tissues. In the liver, fatty acids from the diet, from triglyceride breakdown, and from biosynthesis are incorporated into LDs for storage, and into VLDL particles for distribution. Both types of particles have a



7:  The Hepatocellular Secretory Pathway

hydrophobic lipid core made up of triglycerides surrounded by a phospholipid monolayer, likely arising from deposition of ­triglycerides between the leaflets of the bilayer of the ER mem­ brane [75, 79]. LDs surrounded by cytosolic leaflet phospholip­ ids subsequently bud towards the cytoplasm, whereas VLDL particles surrounded by phospholipids of the luminal leaflet bud into the ER lumen. VLDL particles then enter the secretory pathway, from where they are released from hepatocytes into the bloodstream (Figure 7.3b). In hepatocytes, overexpression of lipin1 has been shown to repress VLDL secretion [98, 99]. The mechanism has not been fully elucidated, but an attractive hypothesis is that high levels of lipin1 channel PA to triglyceride synthesis rather than ­phospholipid synthesis, the latter being required for secretory pathway activity. This could have important implications for the pathogenesis of alcoholic fatty liver disease (FLD), as it has been shown that ethanol increases lipin1 levels [99]. Moreover, ethanol exposure inhibits VLDL secretion in mouse liver, and the increase in lipin1 expression induced by ethanol is mediated by AMPK‐SREBP1 signaling [75, 99]. In response to changing nutritional signals, mTORC1 regulates lipin activity by control­ ling its nuclear localization in hepatocytes [100].

FUTURE DIRECTIONS Major hepatocellular functions include secretion of nascent proteins and VLDL particles into the circulation, and the forma­ tion of bile, processes for which protein and lipid trafficking are essential. This chapter has described some of the molecular components required to direct nascent proteins and more ­complex lipoprotein particles (VLDL) from the ER to the Golgi apparatus and then to their final destination. The core machin­ ery is evolutionarily conserved and operates in all cells, but is modified for the specific trafficking processes performed by the liver, such as secretion of VLDL. Attaining a better understand­ ing of how these processes are utilized by the liver under nor­ mal and pathophysiological conditions is a current and future challenge for liver cell biologists. An important development over the past few years has been elucidation of the intimate con­ nection between membrane trafficking pathways and lipid transport and metabolism. Both production of cytosolic lipid droplets and luminal VLDL particles is closely coupled to vesicular trafficking from the ER to the Golgi. An important area for future work is in understanding this crosstalk as it occurs in liver cells. Elucidating the mechanisms involved will provide avenues for therapeutic intervention not only in liver diseases such as alcoholic and non‐alcoholic FLD, but also in countering infection by viruses (such as hepatitis C and E) which subvert these processes for their propagation.

ACKNOWLEDGMENTS This chapter is based on the chapter by Susan Chi and Mark McNiven in the previous edition. CLJ was supported by grants from the ANR (ANR-13-BSV2-0013) and the “Fondation pour la Recherche Médicale” (DEQ20150934717), France.

83

REFERENCES 1. Rapoport, T.A., Li, L., and Park, E. Structural and mechanistic insights into protein translocation. Annu Rev Cell Dev Biol, 2017;33:369–90. 2. Glick, B.S. and Nakano, A. Membrane traffic within the Golgi apparatus. Annu Rev Cell Dev Biol, 2009;25:113–32. 3. Nixon‐Abell, J., Obara, C.J., Weigel, A.V. et  al. Increased spatiotemporal resolution reveals highly dynamic dense tubular matrices in the peripheral ER. Science, 2016;354(6311). 4. Shibata, Y., Hu, J., Kozlov, M.M., and Rapoport, T.A. Mechanisms shaping the membranes of cellular organelles. Annu Rev Cell Dev Biol, 2009;25:329–54. 5. Rambourg, A. and Clermont Y. Three‐dimensional electron microscopy: structure of the Golgi apparatus. Eur J Cell Biol, 1990;51(2):189–200. 6. Farquhar, M.G. and Palade, G.E. The Golgi apparatus (complex)‐(1954–1981)‐ from artifact to center stage. J Cell Biol, 1981;91(3 Pt 2):77s–103s. 7. Bonifacino, J.S. and Glick, B.S. The mechanisms of vesicle budding and fusion. Cell, 2004;116(2):153–66. 8. Dacks, J.B. and Robinson, M.S. Outerwear through the ages: evolutionary cell biology of vesicle coats. Curr Opin Cell Biol, 2017;47:108–16. 9. Bykov, Y.S., Schaffer, M., Dodonova, S.O. et al. The structure of the COPI coat determined within the cell. eLife, 2017;6. 10. Donaldson, J.G. and Jackson, C.L. ARF family G proteins and their regula­ tors: roles in membrane transport, development and disease. Nat Rev Mol Cell Biol, 2011;12(6):362–75. 11. Gillingham, A.K. and Munro S. The small G proteins of the Arf family and their regulators. Annu Rev Cell Dev Biol, 2007;23:579–611. 12. Lee, M.C., Miller, E.A., Goldberg, J., Orci, L., and Schekman R. Bi‐­ directional protein transport between the ER and Golgi. Annu Rev Cell Dev Biol, 2004;20:87–123. 13. Lee, M.C., Orci, L., Hamamoto, S. et  al. Sar1p N‐terminal helix initiates membrane curvature and completes the fission of a COPII vesicle. Cell, 2005;122(4):605–17. 14. Barlowe, C. and Helenius, A. Cargo capture and bulk flow in the early secre­ tory pathway. Annu Rev Cell Dev Biol, 2016;32:197–222. 15. Brown, H.A., Gutowski, S., Moomaw, C.R., Slaughter, C., and Sternweis, P.C. ADP‐ribosylation factor, a small GTP‐dependent regulatory protein, stimulates phospholipase D activity. Cell, 1993;75(6):1137–44. 16. De Matteis, M.A., Wilson, C., and D’Angelo, G. Phosphatidylinositol‐ 4‐phosphate: the Golgi and beyond. BioEssays, 2013;35(7):612–22. 17. Ktistakis, N.T., Brown, H.A., Waters, M.G., Sternweis, P.C., and Roth, M.G. Evidence that phospholipase D mediates ADP ribosylation factor‐dependent formation of Golgi coated vesicles. J Cell Biol, 1996;134(2):295–306. 18. Ambroggio, E., Sorre, B., Bassereau, P. et al. ArfGAP1 generates an Arf1 gradient on continuous lipid membranes displaying flat and curved regions. EMBO J, 2010;29(2):292–303. 19. Bannykh, S.I., Rowe, T., and Balch, W.E. The organization of endoplasmic reticulum export complexes. J Cell Biol, 1996;135(1):19–35. 20. Cosson, P. and Letourneur F. Coatomer interaction with di‐lysine endoplas­ mic reticulum retention motifs. Science, 1994;263(5153):1629–31. 21. Letourneur, F., Gaynor, E.C., Hennecke, S. et al. Coatomer is essential for retrieval of dilysine‐tagged proteins to the endoplasmic reticulum. Cell, 1994;79(7):1199–207. 22. Bonifacino, J.S. Adaptor proteins involved in polarized sorting. J Cell Biol, 2014;204(1):7–17. 23. Guo, Y., Sirkis, D.W., and Schekman R. Protein sorting at the trans‐Golgi network. Annu Rev Cell Dev Biol, 2014;30:169–206. 24. Manolea, F., Chun, J., Chen, D.W. et  al. Arf3 is activated uniquely at the trans‐Golgi network by brefeldin A‐inhibited guanine nucleotide exchange factors. Mol Biol Cell, 2010;21(11):1836–49. 25. Devos, D., Dokudovskaya, S., Alber, F. et al. Components of coated vesicles and nuclear pore complexes share a common molecular architecture. PLoS Biol, 2004;2(12):e380. 26. Duden, R., Griffiths, G., Frank, R., Argos, P., and Kreis, T.E. Beta‐COP, a 110 kd protein associated with non‐clathrin‐coated vesicles and the Golgi complex, shows homology to beta‐adaptin. Cell, 1991;64(3): 649–65. 27. Serafini, T., Stenbeck, G., Brecht, A. et al. A coat subunit of Golgi‐derived non‐clathrin‐coated vesicles with homology to the clathrin‐coated vesicle coat protein beta‐adaptin. Nature, 1991;349(6306):215–20. 28. Beck, R., Rawet, M., Wieland, F.T., and Cassel D. The COPI system: ­molecular mechanisms and function. FEBS Lett, 2009;583(17):2701–9.

84

THE LIVER:  REFERENCES

29. Orci, L., Stamnes, M., Ravazzola, M. et al. Bidirectional transport by distinct populations of COPI‐coated vesicles. Cell, 1997;90(2):335–49. 30. Bui, Q.T., Golinelli‐Cohen, M.P., and Jackson, C.L. Large Arf1 guanine nucle­ otide exchange factors: evolution, domain structure, and roles in membrane trafficking and human disease. Mol Genet Genomics, 2009;282(4):329–50. 31. Casanova, J.E. Regulation of Arf activation: the Sec7 family of guanine nucleotide exchange factors. Traffic, 2007;8(11):1476–85. 32. Novick, P., Field, C., and Schekman R. Identification of 23 complementation groups required for post‐translational events in the yeast secretory pathway. Cell, 1980;21(1):205–15. 33. Peyroche, A., Paris, S., and Jackson, C.L. Nucleotide exchange on ARF mediated by yeast Gea1 protein. Nature, 1996;384(6608):479–81. 34. Claude, A., Zhao, B.P., Kuziemsky, C.E. et al. GBF1: a novel Golgi‐associ­ ated BFA‐resistant guanine nucleotide exchange factor that displays speci­ ficity for ADP‐ribosylation factor 5. J Cell Biol, 1999;146(1):71–84. 35. Schlacht, A., Mowbrey, K., Elias, M., Kahn, R.A., and Dacks, J.B. Ancient complexity, opisthokont plasticity, and discovery of the 11th subfamily of Arf GAP proteins. Traffic, 2013;14(6):636–49. 36. Schlacht A. Evolution of the vesicle formation machinery. PhD thesis, Department of Cell Biology, University of Alberta, Edmonton, Canada, director Joel Dacks, 2015. 37. Lowery, J., Szul, T., Styers, M. et al. The Sec7 guanine nucleotide exchange factor GBF1 regulates membrane recruitment of BIG1 and BIG2 guanine nucleotide exchange factors to the trans‐Golgi network (TGN). J Biol Chem, 2013;288(16):11532–45. 38. Bonifacino, J.S. and Lippincott‐Schwartz, J. Coat proteins: shaping mem­ brane transport. Nat Rev Mol Cell Biol, 2003;4(5):409–14. 39. Deng, Y., Golinelli‐Cohen, M.P., Smirnova, E., and Jackson, C.L. A COPI coat subunit interacts directly with an early‐Golgi localized Arf exchange factor. EMBO Rep, 2009;10(1):58–64. 40. Sato, K. and Nakano A. Dissection of COPII subunit‐cargo assembly and disassembly kinetics during Sar1p‐GTP hydrolysis. Nat Struct Mol Biol, 2005;12(2):167–74. 41. Tanigawa, G., Orci, L., Amherdt, M. et al. Hydrolysis of bound GTP by ARF protein triggers uncoating of Golgi‐derived COP‐coated vesicles. J Cell Biol, 1993;123(6 Pt 1):1365–71. 42. Forster, R., Weiss, M., Zimmermann, T. et  al. Secretory cargo regulates the turnover of COPII subunits at single ER exit sites. Curr Biol, 2006;16(2):173–9. 43. Kliouchnikov, L., Bigay, J., Mesmin, B. et  al. Discrete determinants in ArfGAP2/3 conferring Golgi localization and regulation by the COPI coat. Mol Biol Cell, 2009;20(3):859–69. 44. Weimer, C., Beck, R., Eckert, P. et  al. Differential roles of ArfGAP1, ArfGAP2, and ArfGAP3 in COPI trafficking. J Cell Biol, 2008;183(4): 725–35. 45. Bigay, J., Casella, J.F., Drin, G., Mesmin, B., and Antonny, B. ArfGAP1 responds to membrane curvature through the folding of a lipid packing sen­ sor motif. EMBO J, 2005;24(13):2244–53. 46. Bigay, J., Gounon, P., Robineau, S., and Antonny, B. Lipid packing sensed by ArfGAP1 couples COPI coat disassembly to membrane bilayer curvature. Nature, 2003;426(6966):563–6. 47. Barrowman, J., Bhandari, D., Reinisch, K., and Ferro‐Novick S. TRAPP complexes in membrane traffic: convergence through a common Rab. Nat Rev Mol Cell Biol, 2010;11(11):759–63. 48. Schroeter, S., Beckmann, S., and Schmitt, H.D. Coat/tether interactions‐ exception or rule? Front Cell Dev Biol, 2016;4:44. 49. Andag, U. and Schmitt, H.D. Ds11p, an essential component of the Golgi‐ endoplasmic reticulum retrieval system in yeast, uses the same sequence motif to interact with different subunits of the COPI vesicle coat. J Biol Chem, 2003;278(51):51722–34. 50. Reilly, B.A., Kraynack, B.A., VanRheenen, S.M., and Waters, M.G. Golgi‐ to‐endoplasmic reticulum (ER) retrograde traffic in yeast requires Ds11p, a component of the ER target site that interacts with a COPI coat subunit. Mol Biol Cell, 2001;12(12):3783–96. 51. Ren, Y., Yip, C.K., Tripathi, A. et  al. A structure‐based mechanism for vesicle capture by the multisubunit tethering complex Ds11. Cell, ­ 2009;139(6):1119–29. 52. Suvorova, E.S., Duden, R., and Lupashin, V.V. The Sec34/Sec35p complex, a Ypt1p effector required for retrograde intra‐Golgi trafficking, interacts with Golgi SNAREs and COPI vesicle coat proteins. J Cell Biol, 2002;157(4):631–43.

53. Zolov, S.N. and Lupashin, V.V. Cog3p depletion blocks vesicle‐mediated Golgi retrograde trafficking in HeLa cells. J Cell Biol, 2005;168(5): 747–59. 54. Guo, Y., Punj, V., Sengupta, D., and Linstedt, A.D. Coat‐tether interaction in Golgi organization. Mol Biol Cell, 2008;19(7):2830–43. 55. Goueslain, L., Alsaleh, K., Horellou, P. et  al. Identification of GBF1 as a cellular factor required for hepatitis C virus RNA replication. J Virol, 2010;84(2):773–87. 56. Farhat, R., Ankavay, M., Lebsir, N. et al. Identification of GBF1 as a cellular factor required for hepatitis E virus RNA replication. Cell Microbiol, 2018;20(1). 57. Lavie, M. and Dubuisson J. Interplay between hepatitis C virus and lipid metabolism during virus entry and assembly. Biochimie, 2017;141:62–9. 58. Farhat, R., Seron, K., Ferlin, J. et al. Identification of class II ADP‐ribosylation factors as cellular factors required for hepatitis C virus replication. Cell Microbiol, 2016;18(8):1121–33. 59. Altan‐Bonnet, N., Sougrat, R., and Lippincott‐Schwartz, J. Molecular basis for Golgi maintenance and biogenesis. Curr Opin Cell Biol, 2004;16(4):364–72. 60. Jamieson, J.D. and Palade, G.E. Synthesis, intracellular transport, and ­discharge of secretory proteins in stimulated pancreatic exocrine cells. J Cell Biol, 1971;50(1):135–58. 61. Presley, J.F., Cole, N.B., Schroer, T.A. et al. ER‐to‐Golgi transport ­visualized in living cells. Nature, 1997;389(6646):81–5. 62. Barlowe, C., Orci, L., Yeung, T. et al. COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell, 1994;77(6):895–907. 63. Mizuno, M. and Singer, S.J. A soluble secretory protein is first concentrated in the endoplasmic reticulum before transfer to the Golgi apparatus. Proc Natl Acad Sci U S A, 1993;90(12):5732–6. 64. Orci, L., Ravazzola, M., Meda, P. et al. Mammalian Sec23p homologue is restricted to the endoplasmic reticulum transitional cytoplasm. Proc Natl Acad Sci U S A, 1991;88(19):8611–15. 65. Bielli, A., Haney, C.J., Gabreski, G. et al. Regulation of Sar1 NH2 terminus by GTP binding and hydrolysis promotes membrane deformation to control COPII vesicle fission. J Cell Biol, 2005;171(6):919–24. 66. Aguilera‐Gomez, A., Zacharogianni, M., van Oorschot, M.M. et al. Phospho‐ rasputin stabilization by Sec16 is required for stress granule formation upon amino acid starvation. Cell Rep, 2017;20(4):935–48. 67. Hanna, M.G., Block, S., Frankel, E.B. et  al. TFG facilitates outer coat ­disassembly on COPII transport carriers to promote tethering and fusion with ER‐Golgi intermediate compartments. Proc Natl Acad Sci U S A, 2017;114(37):E7707–E16. 68. Saito, K., Chen, M., Bard, F. et  al. TANGO1 facilitates cargo loading at endoplasmic reticulum exit sites. Cell, 2009;136(5):891–902. 69. Cai, H., Yu, S., Menon, S. et al. TRAPPI tethers COPII vesicles by binding the coat subunit Sec23. Nature, 2007;445(7130):941–4. 70. Santos, A.J., Nogueira, C., Ortega‐Bellido, M., and Malhotra V. TANGO1 and Mia2/cTAGE5 (TALI) cooperate to export bulky pre‐chylomicrons/ VLDLs from the endoplasmic reticulum. J Cell Biol, 2016;213(3):343–54. 71. Shimoi, W., Ezawa, I., Nakamoto, K. et al. p125 is localized in endoplasmic reticulum exit sites and involved in their organization. J Biol Chem, 2005;280(11):10141–8. 72. Melero, A., Chiaruttini, N., Karashima, T. et al. Lysophospholipids facilitate COPII vesicle formation. Curr Biol, 2018;28(12):1950–8 e6. 73. Brodsky, J.L., Gusarova, V., and Fisher, E.A. Vesicular trafficking of hepatic apolipoprotein B100 and its maturation to very low‐density lipoprotein par­ ticles; studies from cells and cell‐free systems. Trends Cardiovasc Med, 2004;14(4):127–32. 74. Zanetti, G., Pahuja, K.B., Studer, S., Shim, S., and Schekman R. COPII and the regulation of protein sorting in mammals. Nat Cell Biol, 2011;14(1):20–8. 75. Gluchowski, N.L., Becuwe, M., Walther, T.C., and Farese, R.V., Jr. Lipid droplets and liver disease: from basic biology to clinical implications. Nat Rev Gastroenterol Hepatol, 2017;14(6):343–55. 76. Gusarova, V., Brodsky, J.L., and Fisher, E.A. Apolipoprotein B100 exit from the endoplasmic reticulum (ER) is COPII‐dependent, and its lipidation to  very low density lipoprotein occurs post‐ER. J Biol Chem, 2003;278(48):48051–8. 77. Shoulders, C.C., Stephens, D.J., and Jones, B. The intracellular transport of chylomicrons requires the small GTPase, Sar1b. Curr Opin Lipidol, 2004;15(2):191–7.



7:  The Hepatocellular Secretory Pathway

78. Santos, A.J., Raote, I., Scarpa, M., Brouwers, N., and Malhotra V. TANGO1 recruits ERGIC membranes to the endoplasmic reticulum for procollagen export. eLife, 2015;4. 79. Tiwari, S. and Siddiqi, S.A. Intracellular trafficking and secretion of VLDL. Arterioscler Thromb Vasc Biol, 2012;32(5):1079–86. 80. Novikoff, P.M. and Yam A. Sites of lipoprotein particles in normal rat hepat­ ocytes. J Cell Biol, 1978;76(1):1–11. 81. Peden, A.A., Rudge, R.E., Lui, W.W., and Robinson, M.S. Assembly and function of AP‐3 complexes in cells expressing mutant subunits. J Cell Biol, 2002;156(2):327–36. 82. Guo, X., Mattera, R., Ren, X. et  al. The adaptor protein‐1 mu1B subunit expands the repertoire of basolateral sorting signal recognition in epithelial cells. Dev Cell. 2013;27(3):353–66. 83. Doray, B., Bruns, K., Ghosh, P., and Kornfeld, S. Interaction of the cation‐ dependent mannose 6‐phosphate receptor with GGA proteins. J Biol Chem, 2002;277(21):18477–82. 84. Puertollano, R., Randazzo, P.A., Presley, J.F., Hartnell, L.M., and Bonifacino, J.S. The GGAs promote ARF‐dependent recruitment of clathrin to the TGN. Cell, 2001;105(1):93–102. 85. Collins, B.M., Praefcke, G.J., Robinson, M.S., and Owen, D.J. Structural basis for binding of accessory proteins by the appendage domain of GGAs. Nat Struct Biol, 2003;10(8):607–13. 86. Puertollano, R., Aguilar, R.C., Gorshkova, I., Crouch, R.J., and Bonifacino, J.S. Sorting of mannose 6‐phosphate receptors mediated by the GGAs. Science, 2001;292(5522):1712–16. 87. Zhen, Y. and Stenmark H. Cellular functions of Rab GTPases at a glance. J Cell Sci, 2015;128(17):3171–6. 88. Wang, J., Morita, Y., Mazelova, J., and Deretic, D. The Arf GAP ASAP1 provides a platform to regulate Arf4‐ and Rab11‐Rab8‐mediated ciliary receptor targeting. EMBO J, 2012;31(20):4057–71.

85

89. Henley, J.R. and McNiven, M.A. Association of a dynamin‐like protein with the Golgi apparatus in mammalian cells. J Cell Biol, 1996;133(4):761–75. 90. Cao, H., Weller, S., Orth, J.D. et al. Actin and Arf1‐dependent recruitment of a cortactin‐dynamin complex to the Golgi regulates post‐Golgi transport. Nat Cell Biol, 2005;7(5):483–92. 91. Renard, H.F., Johannes, L., and Morsomme, P. Increasing diversity of biologi­ cal membrane fission mechanisms. Trends Cell Biol, 2018;28(4):274–86. 92. Miserey‐Lenkei, S., Chalancon, G., Bardin, S. et al. Rab and actomyosin‐ dependent fission of transport vesicles at the Golgi complex. Nat Cell Biol, 2010;12(7):645–54. 93. Jackson, C.L., Walch, L., and Verbavatz, J.M. Lipids and their trafficking: an integral part of cellular organization. Dev Cell, 2016;39(2):139–53. 94. Vance, J.E. Phospholipid synthesis and transport in mammalian cells. Traffic, 2015;16(1):1–18. 95. Walther, T.C., Chung, J., and Farese, R.V., Jr. Lipid droplet biogenesis. Annu Rev Cell Dev Biol, 2017;33:491–510. 96. Gaspar, M.L., Hofbauer, H.F., Kohlwein, S.D., and Henry, S.A. Coordination of storage lipid synthesis and membrane biogenesis: evidence for cross‐talk between triacylglycerol metabolism and phosphatidylinositol synthesis. J Biol Chem, 2011;286(3):1696–708. 97. Siniossoglou, S. Phospholipid metabolism and nuclear function: roles of the lipin family of phosphatidic acid phosphatases. Biochim Biophys Acta, 2013;1831(3):575–81. 98. Wang, Y., Viscarra, J., Kim, S.J., and Sul, H.S. Transcriptional regulation of hepatic lipogenesis. Nat Rev Mol Cell Biol, 2015;16(11):678–89. 99. You, M., Jogasuria, A., Lee, K. et al. Signal transduction mechanisms of alcoholic fatty liver disease: emerging role of lipin‐1. Curr Mol Pharmacol, 2017;10(3):226–36. 100. Peterson, T.R., Sengupta, S.S., Harris, T.E. et al. mTOR complex 1 ­regulates lipin 1 localization to control the SREBP pathway. Cell, 2011;146(3):408–20.

8

Mitochondrial Function, Dynamics, and Quality Control Marc Liesa, Ilan Benador, Nathanael Miller, and Orian S. Shirihai Department of Medicine, Division of Endocrinology and Department of Molecular and Medical Pharmacology, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA

INTRODUCTION The word “mitochondrion” originates from the fusion of two Greek words: mitos, meaning thread and chondros, meaning grain [1]. This term was coined by Benda after visual inspection of these organelles. It is worth noting that thread and grain are the most common shapes of bacteria: coccus was coined for grain‐shaped bacteria and bacillus for thread‐shaped bacteria. Later, it was demonstrated that mitochondria originated from bacteria engulfed by the ancestors of eukaryotic cells. The intracellular parasite Rickettsia is widely accepted as the α‐proteobacteria order from which mitochondria originated and is morphologically classified as a cocco‐bacillus bacterial species. Therefore, the endosymbiotic theory has both anatomical and DNA conservation evidence in its favor. In this regard, the main features of mitochondria that are present among all eukaryotic species are conserved in bacteria, namely: (i) Mitochondria consume oxygen to synthesize ATP, using a process named oxidative phosphorylation (OXPHOS), similar to bacterial aerobic ATP synthesis. (ii) Mitochondria have their own genome, which encodes for 13 subunits of the respiratory complexes executing OXPHOS, tRNA, and rRNAs required for their translation. (iii) Mitochondria are motile organelles that can fuse or divide into smaller organelles, similar to bacteria. In this chapter, we will provide a summary of these three conserved features of mitochondria, with a focus on hepatocyte mitochondrial OXPHOS function and the role of ­mitochondrial dynamics in hepatocytes determining OXPHOS function.

MITOCHONDRIAL OXIDATIVE PHOSPHORYLATION (OXPHOS) Mitochondria are organelles formed by two membrane layers with different compositions of lipids and proteins, separated by an intermembrane space (IMS). Over 1500 proteins are estimated to reside in human mitochondria, but only 13 are encoded by the mitochondrial DNA (mtDNA). This low number of ­proteins means that protein import through the mitochondrial membrane layers is a central process for mitochondrial biogenesis, function, and turnover. The outer mitochondrial membrane is bidirectionally permeable to small solutes, including small peptides (6 g in adults). Like many drugs, acetami­ nophen is metabolized by the liver where it reacts with cytochromes and is conjugated to glutathione, sulfate, and glucu­ ronic acid, allowing for further metabolism and excretion. However, at high doses glutathione becomes depleted, and a toxic reactive metabolite, N‐acetyl‐p‐benzoquinoneimine (NAPQI), accumulates and can instigate a chain reaction of hepatocyte necrosis that initiates around central veins of liver lobules (Figure 14.4). Notably, cytochromes CYP2E1 and CYP3A4 that catalyze the conversion of acetaminophen into NAPQI also strongly localize around central veins. The propagating cycle of necrosis involves reactive oxygen species (ROS), mitochondrial oxidative stress and dysfunction, activation of Jun N‐terminal kinase, and an inflammatory response [21, 86]. Gap junctions and gap junction proteins are expected to con­ tribute to this in multiple ways. For instance, although APAP can enter cells by diffusion, gap junction channels could distrib­ ute cellular signals, ions, and critical molecules, and hemichan­ nels could extrude inflammatory signals such as ATP and other and damage‐associated molecular pattern molecules. On the other hand, gap junctions could also distribute detoxifying com­ pounds such as glutathione or UDP‐glucuronic acid to hepato­ cytes where these may be depleted. A critical step in the progression of toxicity appears to be the spreading of centri­ lobular necrosis to the rest of the liver. In 2004 Asamoto and colleagues [87] described the generation of transgenic rats that express hepatocyte‐specific dominant‐negative Cx32 under control of the albumin promoter. This resulted in decreased gap junction activity, reduced Cx32 and Cx26 membrane localiza­ tion, and resistance to the hepatic toxins carbon tetrachloride and d‐galactosamine. A second study then found that these rats were resistant to acetaminophen toxicity, showing lowered serum aminotransferase levels and improved liver histology compared to wild type [88]. In addition, it was found that aceta­ minophen‐induced cell death in isolated hepatocytes from Cx32‐KO mice was no longer synchronized as it is in wild type. Connexin expression was also required to allow cell‐to‐cell pro­ tection of coupled hepatocytes from female to male animals. Another study observed that Cx32‐KO mice themselves were resistant to acetaminophen and thioacetamide toxicity as observed by aminotransferase levels, histology, and death of the animals. Toxicity could also be reduced by the gap junction‐ blocking drug 2‐aminoethoxy‐diphenyl‐borate (2‐APB) [89]. Although these studies appeared compelling, a number of doubts have been raised. As described earlier in the chapter, Cx32‐KO mice display increased levels of chemically induced liver tumors, suggesting that they have increased susceptibility to drugs. Another study indicated that Cx32‐KO mice have reduced levels of centrilobular glutathione, and that they are in fact more susceptible to acetaminophen. Maes and colleagues, who have studied acetaminophen toxicity extensively, weighed



14:  Gap and Tight Junctions in Liver: Structure, Function, and Pathology

in on these issues by performing their own studies. They found that Cx32‐KO mice have similar levels of cell death, inflamma­ tion, and oxidative stress in response to acetaminophen, with somewhat lower levels of protein adduct formation [90]. The gap junction blocker 2‐APB was also shown to exert protective effects through inhibition of CYP450s and c‐Jun N‐terminal kinase. Follow‐up reports by some of these investigators have now found that connexin and pannexin hemichannels may be important, as specific inhibition of hemichannels provided pro­ tection from drug‐induced liver injury [91]. Collectively, these studies indicate that all of the contributors to acetaminophen toxicity are not well resolved. Many technical details can influ­ ence the results, such as the solubilizing agent DMSO, and the species of mouse or rodent; for instance, rats are less susceptible than mice [92]. Connexin proteins and the activity of gap junc­ tions (and also Panx1) appear to feature in these processes, but at present a clear protective or damaging role in this toxicity cannot be assumed and it remains to be seen whether gap block­ age can be useful in the clinical setting.

PROSPECTS AND PERSPECTIVES The two most prominent junctional types in liver are the gap junctions, which provide direct intercellular communication, and the tight junctions, which serve to partition membrane domains of individual cells and to occlude extracellular space, restricting pericellular diffusion. The integral membrane pro­ teins of gap junctions, the connexins, have been structurally well characterized, but their newly recognized roles in binding to and organizing cytoplasmic proteins suggest that they may form nucleation sites for intracellular as well as extracellular signaling. Connexins interact with actin and microtubules as well as mitochondria, and they participate in recovery from injury in many tissues, although their precise role as both gap junctions and hemichannels (where Panx1 may also contribute) in drug‐induced liver injury remains enigmatic. By contrast, peripheral tight junction proteins were identified prior to the core proteins, and one area of intense interest is the diversity of “tightness” profiles provided by the large claudin family of integral membrane proteins. Although gap and tight junctions perform different functions, there are numerous points at which these junctional types appear to overlap. Indeed, the findings that the traditionally tight junction‐associated protein ZO‐1 also binds to connexins, and that occludin colocalizes with Cx32 indicate the possibility for either coordinate or reciprocal regulation of macromolecular complexes containing gap and tight junction proteins. Studies of protein–protein interactions and of coordinate and subordinate regulation of gene families may elucidate the intricacies of inter‐ and intracellular signal­ ing concerning growth control by gap junctions and mainte­ nance of the “blood–biliary barrier” formed by tight junctions [30, 93]. As with the HCV receptor CD81, the major gap junc­ tion components (connexins) and the major tight junction com­ ponents (occludin and claudins) are tetraspan proteins with critical intracellular and extracellular domains. The appropria­ tion of tight junction proteins as portals for virus entry suggests special properties for these proteins. As tight junctions actively

171

form the apical–basolateral junction, it is possible that this loca­ tion is a protective niche for viruses or that viruses can exploit temporary disruptions of the blood–biliary barrier for their entry. Future studies are likely to reveal new functions of connexins, occludin, and claudins that affect the processes ­ of  viral entry as well as downstream signaling and eventual induction of hepatitis.

REFERENCES 1. Kojima T., Sawada, N., Yamaguchi H., Fort A.G., and Spray D.C. Gap and tight junctions in liver: composition, regulation, and function, in The Liver: Biology and Pathobiology, 5th edn (eds. I.M. Arias et al.), Wiley‐Blackwell, Chichester, 2009, pp. 201–20. 2. Spray, D.C., Saez, J.C., Herzberg, E.L. et al. Gap junctions in liver: composi­ tion, function and regulation, in The Liver: Biology and Pathobiology, 3rd edn (eds. I.M. Arias et al.), Raven Press, New York, 1994, pp. 951–67. 3. Spray, D.C., Ginzberg, R.D., Morales, E.A., Gatmaitan, Z., and Arias, I.M. Electrophysiological properties of gap junctions between dissociated pairs of rat hepatocytes. J Cell Biol, 1986;103(1):135–44. 4. Franke, W.W. Discovering the molecular components of intercellular ­junctions  –  a historical view. Cold Spring Harb Perspect Biol, 2009;1(3): a003061. 5. Revel, J.P., Yancey, S.B., and Meyer, D.J. Cellular architecture and intercel­ lular communication in rat liver. JAMA, 1981;245(9):958–9. 6. Unwin, P.N. and Ennis, P.D. Two configurations of a channel‐forming mem­ brane protein. Nature, 1984;307(5952):609–13. 7. Perkins, G.A., Goodenough, D.A., and Sosinsky, G.E. Formation of the gap junction intercellular channel requires a 30 degree rotation for interdigitating two apposing connexons. J Mol Biol, 1998;277(2):171–7. 8. Maeda, S., Nakagawa, S., Suga, M. et al. Structure of the connexin 26 gap junction channel at 3.5 A resolution. Nature, 2009;458(7238):597–602. 9. Furuse, M., Hirase, T., Itoh, M. et  al. Occludin: a novel integral mem­ brane  protein localizing at tight junctions. J Cell Biol, 1993;123(6 Pt 2): 1777–88. 10. Saitou, M., Fujimoto, K., Doi, Y. et al. Occludin‐deficient embryonic stem cells can differentiate into polarized epithelial cells bearing tight junctions. J Cell Biol, 1998;141(2):397–408. 11. Chiba, H., Osanai, M., Murata, M. et al. Transmembrane proteins of tight junctions. Biochim Biophys Acta, 2008;1778(3):588–600. 12. Evans, M.J., von Hahn, T., Tscherne, D.M. et al. Claudin‐1 is a hepatitis C virus co‐receptor required for a late step in entry. Nature, 2007;446(7137):801–5. 13. Guttman, J.A. and Finlay, B.B. Tight junctions as targets of infectious agents. Biochim Biophys Acta, 2009;1788(4):832–41. 14. Rahner, C., Mitic, L.L., and Anderson, J.M. Heterogeneity in expression and subcellular localization of claudins 2, 3, 4, and 5 in the rat liver, pancreas, and gut. Gastroenterology, 2001;120(2):411–22. 15. Yamamoto, T., Kojima, T., Murata, M. et al. p38 MAP‐kinase regulates func­ tion of gap and tight junctions during regeneration of rat hepatocytes. J Hepatol, 2005;42(5):707–18. 16. Amasheh, S., Meiri, N., Gitter, A.H. et  al. Claudin‐2 expression induces cation‐selective channels in tight junctions of epithelial cells. J Cell Sci, 2002;115(Pt 24):4969–76. 17. Paul, D.L. Molecular cloning of cDNA for rat liver gap junction protein. J Cell Biol, 1986;103(1):123–34. 18. Kumar, N.M. and Gilula, N.B. The gap junction communication channel. Cell, 1996;84(3):381–8. 19. Zhang, J.T. and Nicholson, B.J. Sequence and tissue distribution of a second protein of hepatic gap junctions, Cx26, as deduced from its cDNA. J Cell Biol, 1989;109(6 Pt 2):3391–401. 20. Dahl, G. and Muller, K.J. Innexin and pannexin channels and their signaling. FEBS Lett, 2014;588(8):1396–402. 21. Maes, M., Crespo Yanguas, S., Willebrords, J., Cogliati, B., and Vinken M. Connexin and pannexin signaling in gastrointestinal and liver disease. Transl Res, 2015;166(4):332–43. 22. Sosinsky, G.E., Boassa, D., Dermietzel, R. et al. Pannexin channels are not gap junction hemichannels. Channels (Austin), 2011;5(3):193–7.

172

THE LIVER: REFERENCES

23. Willebrords, J., Maes, M., Crespo Yanguas, S., and Vinken M. Inhibitors of connexin and pannexin channels as potential therapeutics. Pharmacol Ther, 2017;180:144–60. 24. Stauffer, K.A. The gap junction proteins beta 1‐connexin (connexin‐32) and beta 2‐connexin (connexin‐26) can form heteromeric hemichannels. J Biol Chem, 1995;270(12):6768–72. 25. Valiunas, V., Niessen, H., Willecke, K., and Weingart, R. Electrophysiological properties of gap junction channels in hepatocytes isolated from connexin32‐ deficient and wild‐type mice. Pflugers Arch, 1999;437(6):846–56. 26. Locke, D., Liu, J., and Harris, A.L. Lipid rafts prepared by different methods contain different connexin channels, but gap junctions are not lipid rafts. Biochemistry, 2005;44(39):13027–42. 27. Duffy, H.S., Delmar, M., and Spray, D.C. Formation of the gap junction nexus: binding partners for connexins. J Physiol Paris, 2002;96(3–4): 243–9. 28. Duffy, H.S., Fort, A.G., and Spray, D.C. Cardiac connexins: genes to nexus. Adv Cardiol, 2006;42:1–17. 29. Duffy, H.S., Iacobas, I., Hotchkiss, K. et al. The gap junction protein con­ nexin32 interacts with the Src homology 3/hook domain of discs large homolog 1. J Biol Chem, 2007;282(13):9789–96. 30. Kojima, T., Yamamoto, T., Murata, M. et al. Regulation of the blood‐biliary barrier: interaction between gap and tight junctions in hepatocytes. Med Electron Microsc, 2003;36(3):157–64. 31. Fowler, S.L., Akins, M., Zhou, H., Figeys, D., and Bennett, S.A. The liver connexin32 interactome is a novel plasma membrane‐mitochondrial ­signaling nexus. J Proteome Res, 2013;12(6):2597–610. 32. Graf, J. and Boyer, J.L. The use of isolated rat hepatocyte couplets in ­hepatobiliary physiology. J Hepatol, 1990;10(3):387–94. 33. Gunzel, D., Zakrzewski, S.S., Schmid, T. et al. From TER to trans‐ and para­ cellular resistance: lessons from impedance spectroscopy. Ann N Y Acad Sci, 2012;1257:142–51. 34. Suzuki, H., Nishizawa, T., Tani, K. et al. Crystal structure of a claudin pro­ vides insight into the architecture of tight junctions. Science, 2014;344(6181): 304–7. 35. Matsumoto, K., Imasato, M., Yamazaki, Y. et  al. Claudin 2 deficiency reduces bile flow and increases susceptibility to cholesterol gallstone disease in mice. Gastroenterology, 2014;147(5):1134–45 e10. 36. Buckley, A. and Turner, J.R. Cell biology of tight junction barrier regulation and mucosal disease. Cold Spring Harb Perspect Biol, 2018;10(1). 37. Fasano, A., Fiorentini, C., Donelli, G. et  al. Zonula occludens toxin ­modulates tight junctions through protein kinase C‐dependent actin reor­ ganization, in vitro. J Clin Invest, 1995;96(2):710–20. 38. Tsukamoto, T. and Nigam, S.K. Tight junction proteins form large com­ plexes and associate with the cytoskeleton in an ATP depletion model for reversible junction assembly. J Biol Chem, 1997;272(26):16133–9. 39. Turner, J.R., Angle, J.M., Black, E.D. et  al. PKC‐dependent regulation of transepithelial resistance: roles of MLC and MLC kinase. Am J Physiol, 1999;277(3 Pt 1):C554–62. 40. Pradhan‐Sundd, T., Vats, R., Russell, J.M. et al. Dysregulated bile transport­ ers and impaired tight junctions during chronic liver injury in mice. Gastroenterology, 2018;155(4):1218–32.e24. 41. Gissen, P. and Arias, I.M. Structural and functional hepatocyte polarity and liver disease. J Hepatol, 2015;63(4):1023–37. 42. Mailly, L., Xiao, F., Lupberger, J. et al. Clearance of persistent hepatitis C virus infection in humanized mice using a claudin‐1‐targeting monoclonal antibody. Nat Biotechnol, 2015;33(5):549–54. 43. Crespo Yanguas, S., Willebrords, J., Maes, M. et al. Connexins and pannex­ ins in liver damage. EXCLI J, 2016;15:177–86. 44. Gumucio, J.J. and Miller, D.L. Functional implications of liver cell heteroge­ neity. Gastroenterology, 1981;80(2):393–403. 45. Jungermann, K. and Katz N. Functional specialization of different hepato­ cyte populations. Physiol Rev, 1989;69(3):708–64. 46. Berthoud, V.M., Iwanij, V., Garcia, A.M., and Saez, J.C. Connexins and glucagon receptors during development of rat hepatic acinus. Am J Physiol, 1992;263(5 Pt 1):G650–8. 47. Kinugasa, A. and Thurman, R.G. Differential effect of glucagon on gluco­ neogenesis in periportal and pericentral regions of the liver lobule. Biochem J, 1986;236(2):425–30. 48. Lee, S.M. and Clemens, M.G. Subacinar distribution of hepatocyte mem­ brane potential response to stimulation of gluconeogenesis. Am J Physiol., 1992;263(3 Pt 1):G319–26.

49. Seseke, F.G., Gardemann, A., and Jungermann K. Signal propagation via gap junctions, a key step in the regulation of liver metabolism by the sympathetic hepatic nerves. FEBS Lett, 1992;301(3):265–70. 50. Saez, J.C., Connor, J.A., Spray, D.C., and Bennett, M.V. Hepatocyte gap junctions are permeable to the second messenger, inositol 1,4,5‐trisphos­ phate, and to calcium ions. Proc Natl Acad Sci U S A, 1989;86(8):2708–12. 51. Nathanson, M.H. and Burgstahler, A.D. Coordination of hormone‐induced calcium signals in isolated rat hepatocyte couplets: demonstration with con­ focal microscopy. Mol Biol Cell, 1992;3(1):113–21. 52. Clapham, D.E. Calcium signaling. Cell, 2007;131(6):1047–58. 53. Bartlett, P.J., Gaspers, L.D., Pierobon, N., and Thomas, A.P. Calcium‐ dependent regulation of glucose homeostasis in the liver. Cell Calcium, 2014;55(6):306–16. 54. Guerra, M.T. and Nathanson, M.H. Calcium signaling and secretion in chol­ angiocytes. Pancreatology, 2015;15(4 Suppl):S44–8. 55. Scemes, E. and Spray, D.C. Extracellular K+ and astrocyte signaling via con­ nexin and pannexin channels. Neurochem Res, 2012;37(11):2310–16. 56. Schlosser, S.F., Burgstahler, A.D., and Nathanson, M.H. Isolated rat hepato­ cytes can signal to other hepatocytes and bile duct cells by release of nucleo­ tides. Proc Natl Acad Sci U S A, 1996;93(18):9948–53. 57. Aasen, T., Mesnil, M., Naus, C.C., Lampe, P.D., and Laird, D.W. Gap junc­ tions and cancer: communicating for 50 years. Nat Rev Cancer, 2016;16(12):775–88. 58. Yamasaki, H., Krutovskikh, V., Mesnil, M. et al. Role of connexin (gap junc­ tion) genes in cell growth control and carcinogenesis. C R Acad Sci III, 1999;322(2–3):151–9. 59. Saez, J.C., Gregory, W.A., Watanabe, T. et al. cAMP delays disappearance of gap junctions between pairs of rat hepatocytes in primary culture. Am J Physiol, 1989;257(1 Pt 1):C1–11. 60. Spray, D.C., Rozental, R., and Srinivas, M. Prospects for rational develop­ ment of pharmacological gap junction channel blockers. Curr Drug Targets, 2002;3(6):455–64. 61. Saez, J.C., Nairn, A.C., Czernik, A.J. et al. Phosphorylation of connexin 32, a hepatocyte gap‐junction protein, by cAMP‐dependent protein kinase, pro­ tein kinase C and Ca2+/calmodulin‐dependent protein kinase II. Eur J Biochem, 1990;192(2):263–73. 62. Koval, M., Isakson, B.E., and Gourdie, R.G. Connexins, pannexins and innexins: protein cousins with overlapping functions. FEBS Lett, 2014;588(8):1185. 63. Epifantseva, I. and Shaw, R.M. Intracellular trafficking pathways of Cx43 gap junction channels. Biochim Biophys Acta, 2018;1860(1):40–7. 64. Fort, A.G., Murray, J.W., Dandachi, N. et al. In vitro motility of liver con­ nexin vesicles along microtubules utilizes kinesin motors. J Biol Chem, 2011;286(26):22875–85. 65. Francis, R., Xu, X., Park, H. et al. Connexin43 modulates cell polarity and directional cell migration by regulating microtubule dynamics. PLoS One, 2011;6(10):e26379. 66. Polusani, S.R., Kalmykov, E.A., Chandrasekhar, A., Zucker, S.N., and Nicholson, B.J. Cell coupling mediated by connexin 26 selectively con­ tributes to reduced adhesivity and increased migration. J Cell Sci, 2016;129(23):4399–410. 67. Matsuuchi, L. and Naus, C.C. Gap junction proteins on the move: connexins, the cytoskeleton and migration. Biochim Biophys Acta, 2013;1828(1): 94–108. 68. Falk, M.M., Bell, C.L., Kells Andrews, R.M., and Murray, S.A. Molecular mechanisms regulating formation, trafficking and processing of annular gap junctions. BMC Cell Biol, 2016;17(Suppl 1):22. 69. Gaietta, G., Deerinck, T.J., Adams, S.R. et al. Multicolor and electron micro­ scopic imaging of connexin trafficking. Science, 2002;296(5567):503–7. 70. Bejarano, E., Yuste, A., Patel, B. et al. Connexins modulate autophagosome biogenesis. Nat Cell Biol, 2014;16(5):401–14. 71. Laird, D.W. Syndromic and non‐syndromic disease‐linked Cx43 mutations. FEBS Lett, 2014;588(8):1339–48. 72. Scherer, S.S. and Kleopa, K.A. X‐linked Charcot‐Marie‐Tooth disease. J Peripher Nerv Syst, 2012;17(Suppl 3):9–13. 73. Nelles, E., Butzler, C., Jung, D. et al. Defective propagation of signals gener­ ated by sympathetic nerve stimulation in the liver of connexin32‐deficient mice. Proc Natl Acad Sci U S A, 1996;93(18):9565–70. 74. Stumpel, F., Ott, T., Willecke, K., and Jungermann K. Connexin 32 gap junc­ tions enhance stimulation of glucose output by glucagon and noradrenaline in mouse liver. Hepatology, 1998;28(6):1616–20.



14:  Gap and Tight Junctions in Liver: Structure, Function, and Pathology

75. Niessen, H. and Willecke K. Strongly decreased gap junctional permeability to inositol 1,4, 5‐trisphosphate in connexin32 deficient hepatocytes. FEBS Lett, 2000;466(1):112–14. 76. Schwarz, M., Wanke, I., Wulbrand, U., Moennikes, O., and Buchmann A. Role of connexin32 and beta‐catenin in tumor promotion in mouse liver. Toxicol Pathol, 2003;31(1):99–102. 77. King, T.J. and Lampe, P.D. Altered tumor biology and tumorigenesis in irradiated and chemical carcinogen‐treated single and combined con­ nexin32/p27Kip1‐deficient mice. Cell Commun Adhes, 2005;12(5–6): 293–305. 78. Kojima, T., Fort, A., Tao, M., Yamamoto, M., and Spray, D.C. Gap junction expression and cell proliferation in differentiating cultures of Cx43 KO mouse hepatocytes. Am J Physiol Gastrointest Liver Physiol, 2001;281(4): G1004–13. 79. Temme, A., Ott, T., Dombrowski, F., and Willecke, K. The extent of synchro­ nous initiation and termination of DNA synthesis in regenerating mouse liver is dependent on connexin32 expressing gap junctions. J Hepatol, 2000;32(4): 627–35. 80. Dagli, M.L., Yamasaki, H., Krutovskikh, V., and Omori, Y. Delayed liver regeneration and increased susceptibility to chemical hepatocarcinogenesis in transgenic mice expressing a dominant‐negative mutant of connexin32 only in the liver. Carcinogenesis, 2004;25(4):483–92. 81. Igarashi, I., Makino, T., Suzuki, Y. et  al. Background lesions during a ­24‐month observation period in connexin 32‐deficient mice. J Vet Med Sci, 2013;75(2):207–10. 82. Li, Q., Omori, Y., Nishikawa, Y. et  al. Cytoplasmic accumulation of con­ nexin32 protein enhances motility and metastatic ability of human hepatoma cells in vitro and in vivo. Int J Cancer, 2007;121(3):536–46.

173

83. Spray, D.C., Hanstein, R., Lopez‐Quintero, S.V. et  al. Gap junctions and bystander effects: Good Samaritans and executioners. Wiley Interdiscip Rev Membr Transp Signal, 2013;2(1):1–15. 84. Zibara, K., Awada, Z., Dib, L. et al. Anti‐angiogenesis therapy and gap junc­ tion inhibition reduce MDA‐MB‐231 breast cancer cell invasion and metas­ tasis in vitro and in vivo. Sci Rep, 2015;5:12598. 85. Crespo Yanguas, S., da Silva, T.C., Pereira, I.V.A. et al. TAT‐Gap19 and car­ benoxolone alleviate liver fibrosis in mice. Int J Mol Sci, 2018;19(3). 86. Iorga, A., Dara, L., and Kaplowitz, N. Drug‐induced liver injury: cascade of events leading to cell death, apoptosis or necrosis. Int J Mol Sci, 2017;18(5). 87. Asamoto, M., Hokaiwado, N., Murasaki, T., and Shirai, T. Connexin 32 dominant‐negative mutant transgenic rats are resistant to hepatic damage by chemicals. Hepatology, 2004;40(1):205–10. 88. Naiki‐Ito, A., Asamoto, M., Naiki, T. et al. Gap junction dysfunction reduces acetaminophen hepatotoxicity with impact on apoptotic signaling and con­ nexin 43 protein induction in rat. Toxicol Pathol, 2010;38(2):280–6. 89. Patel, S.J., Milwid, J.M., King, K.R. et al. Gap junction inhibition prevents drug‐induced liver toxicity and fulminant hepatic failure. Nat Biotechnol, 2012;30(2):179–83. 90. Maes, M., McGill, M.R., da Silva, T.C. et al. Connexin32: a mediator of aceta­ minophen‐induced liver injury? Toxicol Mech Methods, 2016;26(2):88–96. 91. Maes, M., Crespo Yanguas, S., Willebrords, J. et al. Connexin hemichannel inhibition reduces acetaminophen‐induced liver injury in mice. Toxicol Lett, 2017;278:30–7. 92. Maes, M. and Vinken M. Connexin‐based signaling and drug‐induced hepa­ totoxicity. J Clin Transl Res, 2017;3(Suppl 1):189–98. 93. Lee, N.P. The blood‐biliary barrier, tight junctions and human liver diseases. Adv Exp Med Biol, 2012;763:171–85.

15

Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver Katherine I. Farley‐Barnes1 and Susan J. Baserga1,2,3 Department of Molecular Biophysics & Biochemistry, Yale University School of Medicine, New Haven, CT, USA Department of Genetics, Yale University School of Medicine, New Haven, CT, USA 3 Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, CT, USA 1 2

INTRODUCTION Ribosomes, the cellular machinery responsible for all protein synthesis, are essential for cells to grow and to divide. The biogenesis of ribosomes is thus a highly regulated process that coordinates a number of cellular cues and depends especially upon nutrient availability. Nowhere is the relationship between nutrient availability, cellular signaling, and ribosome biogenesis more critical than in the liver. Liver size decreases dramatically after fasting in rats and mice [1, 2]. When refed, however, the liver size recovers within 24 hours [1, 2]. Total liver proteins and RNA follow the same fluctuation, indicating that the liver responds to limited nutrients by downregulating ribosome production and upregulating it again once the organism is presented with an adequate diet [1–3]. Understanding ribosome biogenesis is not only essential to understanding basic cellular biology, but also to probing the pathogenesis of a number of human diseases. Such diseases include cancer, where an increase in cell growth and proliferation goes hand‐in‐hand with an increase in ribosome biogenesis (reviewed in [4]). Ribosome dysfunction has also been implicated in neurodegenerative diseases (reviewed in [5]). ­ Additionally, aberrant ribosome biogenesis causes a number of tissue‐specific disorders, termed ribosomopathies (reviewed in [6, 7]). It is intriguing that dysregulation of such an essential process like ribosome biogenesis would result in a viable organism. Indeed, how changes in ribosome biogenesis, a process that is required for every cell type, would create tissue‐specific pathologies is an outstanding question in the field. Much of the current knowledge of the steps in ribosome ­biogenesis comes from studies in the budding yeast,

Saccharomyces cerevisiae. Only recently have scientists made advances into understanding how this process works in humans. Additionally, how ribosome biogenesis changes among different tissues is largely unknown. One of the best models for understanding ribosome biogenesis at the tissue level, however, has been the liver. Foundational studies on the liver have provided insights into how ribosomes are made and how the cell responds to various stimuli to regulate the production of ribosomes. Indeed, nucleoli, the non‐membrane‐bound organelles responsible for ribosome production, were first biochemically isolated from liver cells in 1956 [8]. It is likely that in the future, the liver will continue to play a large role in our understanding of ribosome biogenesis and its relation to human disease.

OVERVIEW OF RIBOSOME BIOGENESIS Ribosome biogenesis begins in a non‐membrane‐bound organelle inside the cell nucleus called the nucleolus. It starts with transcription of the tandemly repeated ribosomal DNA (rDNA) by RNA polymerase I (RNAPI). As the rDNA is being transcribed, the nucleolus forms around it. Thus, the repeats of rDNA on the chromosomes are called nucleolar organizing regions, or NORs. The NORs are located on five different chromosomes in humans (13, 14, 15, 21, and 22). Therefore, human cells have the potential to form 10 nucleoli per cell, although far fewer nucleoli are often observed [9]. The number of nucleoli in a diploid mouse hepatocyte ranges from 3 to 6 (mice have the potential to form up to 12 nucleoli per cell), with 3 nucleoli per nucleus being seen most often [10].

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



15:  Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver

The human nucleolus consists of three different compartments: the fibrillar center (FC), the dense fibrillar component (DFC), and the granular component (GC) (Figure 15.1). A distinct function in making ribosomes can be ascribed to each of these compartments. The nucleolus forms around the FC, and at the interface of the FC and the DFC, where rDNA transcription takes place. After the transcription of the rDNA, modification of the pre‐ribosomal RNA (pre‐rRNA), processing of the pre‐ rRNA, and pre‐ribosome assembly proceed outwards from the DFC into the third nucleolar compartment, the GC. In the GC, ribosomal proteins (r‐proteins) assemble with the pre‐rRNA to form the pre‐small subunit (pre‐SSU) and pre‐large subunit (pre‐LSU) of the ribosome [11]. One explanation for the tripartite organization of the nucleolus uses the physics of liquid–liquid phase separations. Just like an oil droplet in water, the proteins of the nucleolus separate into multiple compartments due to differences in surface tension and hydrophobicity. For example, a structure similar to the DFC/GC compartments can be replicated in vitro by mixing fibrillarin (FBL), a box C/D small nucleolar ribonucleoprotein (snoRNP) methyltransferase enzyme that is enriched in the DFC, and nucleophosmin (NPM1), a protein with a number of functions that is enriched in the GC [12]. At the FC/DFC interface, the rDNA is transcribed as a 47S polycistronic precursor that contains 3 of the 4 the ribosomal RNAs (rRNAs) which eventually are incorporated into the small (18S) and large (5.8S and 28S) subunits of the ribosome. To form the mature rRNAs, the pre‐rRNA must be modified, cleaved, and processed. Two types of RNA modifications are primarily used: 2′‐O‐methylation and pseudouridylation (reviewed in [13]). Generally, the modifications are placed at functionally important regions of the ribosome, such as the tRNA‐binding sites or at the interface of the small and large subunits (reviewed in [13]). These modifications aid in folding the pre‐rRNA into the correct secondary and tertiary structures. Most of these modifications are made by snoRNPs which direct their guide RNA to base pair with the corresponding site to be modified and position the enzyme to perform the modification. The majority of pre‐rRNA modifications are carried out by two types of snoRNPs: box C/D, which perform the 2′‐O‐methylations, and box H/ACA, which perform the pseudouridylations, both named for conserved snoRNA sequences. FC DFC

175

In addition to the modifications made to the pre‐rRNA, the pre‐ rRNA is also extensively cleaved and processed. In humans, pre‐ rRNA processing occurs through multiple pathways (Figure 15.2). The difference between the two major processing pathways begins with the choice of cleavage at either site 1 or site 2. If site 1 is cleaved first, the 41S pre‐rRNA is made first until a later cleavage at site 2 separates the 21S and 32S pre‐rRNAs. If site 2 is cleaved first, the small and large subunit pre‐rRNAs are separated into the 30S and 32S pre‐rRNAs, respectively (Figure 15.2, and reviewed in [14]). Regardless of the pathway used, the end result is the mature 18S, 5.8S, and 28S rRNAs. In addition to the snoRNAs and r‐proteins, a number of trans‐ acting factors are also important for optimal pre‐rRNA processing. Best characterized in yeast, these factors include endo‐ and exonucleases, helicases, AAA‐ATPases, GTPases, and more [15]. Many of the functions of these enzymes are still being defined. In yeast, over 200 proteins are involved in ribosome assembly [15], a number that is likely to be greatly increased in humans. Indeed, a series of studies are currently striving to define all of the factors necessary for this process in humans [16–19]. Ribosome assembly concludes in the cytoplasm, after the addition of the 5S rRNA which is transcribed from chromosome 1 by RNA polymerase III (RNAPIII) in humans. The pre‐SSU and pre‐LSU are both exported through the nuclear pore complex, and export of each subunit requires common factors as well as subunit‐specific factors. Exportin 1 (XPO1, or Crm1 in yeast) is one of the best characterized proteins involved in both large and small subunit export [20]. Additionally, the final steps of pre‐rRNA processing occur in the cytoplasm, including the processing of the 18SE to the mature 18S rRNA at site 3. Overall, human ribosome biogenesis begins in the nucleolus but encompasses the nucleus and cytoplasm as well, leaving many steps in the process open to regulation.

REGULATION OF RIBOSOME BIOGENESIS Our understanding of the intricacies of the regulation of human ribosome biogenesis is ever increasing. Regulation begins at the level of the rDNA with the organization and silencing of select rDNA repeats. RNAPI transcription of the rDNA is also ­influenced by many cellular signals. Correct processing of the pre‐rRNA is controlled by a number of molecules that signal to modulate the transcription of r‐proteins, snoRNPs, and other assembly and processing factors. Finally, nuclear export of the pre‐ribosomal subunits can be monitored and regulated by the cell. In all, ribosome biogenesis requires all three RNA polymerases and hundreds of other factors, so there are many ways for the cell to control steps in this process. Just a few of these mechanisms are described in detail below.

GC

Figure 15.1  The morphology of the nucleolus in mammalian cells. On the left is the nucleolus of a cell actively engaged in ribosome biogenesis. On the right is the nucleolus of a cell undergoing inhibition of rRNA transcription. Indicated are the three nucleolar compartments: the fibrillar center (FC), the dense fibrillar component (DFC), and the granular component (GC). See text for details.

mTOR and the regulation of ribosomal protein translation One key regulator of ribosome biogenesis is the mechanistic target of rapamycin (mTOR) pathway. Essential for cell growth, the mTOR pathway responds to various cellular signals to upregulate the transcription of r‐proteins and to ultimately

176

THE LIVER:  REGULATION OF RIBOSOME BIOGENESIS A′

47S pre-rRNA 5′

A0

1

3 2a

18S

5′ETS

A0

1

3 2a

4

5.8S

ITS1

18S

45S

2

ITS2

2

28S

4

5.8S

28S

site 2 cleavage

A0

1

30S

site 1 cleavage

3 2a

3 2a

4

32S

18S

5.8S

28S

2

4 5.8S

18S

3 2a

21S

3′

3′ETS

3 2a

18S

21S

18S

18SE

18S

28S

41S

28S

32S

4 5.8S

3

18SE

18S

18S

12S

5.8S

5.8S

28S

18S

12S

5.8S

5.8S

28S

Figure 15.2  Schematic depicting the two major pre‐rRNA processing pathways in human cells. The 47S pre‐rRNA is processed and cleaved through multiple pathways to form the mature 18S, 5.8S, and 28S rRNAs that are incorporated into the synthesized ribosome. ETS and ITS ­represent external transcribed spacers and internal transcribed spacers, respectively.

increase global protein synthesis. There are two different mTOR‐containing cytoplasmic complexes, termed mTORC1 and mTORC2. The mTORC1 complex contains mTOR, RPTOR (regulatory associated protein of MTOR complex 1), AKT1S1 (AKT substrate 1), DEPTOR (DEP domain containing mTOR‐ interacting protein), and MLST8 (mTOR‐associated protein, LST8 homolog). mTORC2 contains mTOR, RICTOR (RPTOR‐ independent companion of MTOR complex 2), PRR5 (proline‐ rich 5), MAPKAP1 (mitogen‐activated protein kinase‐associated protein 1), DEPTOR, and MLST8. Although mTORC1 and mTORC2 have different complex members and substrates, mTOR functions as a serine/threonine kinase in both complexes. Importantly for ribosome biogenesis, mTORC1 signals to RPS6 kinases 1 and 2 (S6K1 and S6K2). S6K1 phosphorylates several substrates, ultimately promoting increased translation initiation and elongation, as well as increased ribosome biogenesis through the upregulation of rDNA transcription (reviewed in [21]). mTORC1 also phosphorylates eukaryotic translation initiation factor 4B pseudogene 1 (4EBP1). 4EBP1 affects cap‐ dependent translation, and its phosphorylation prevents its binding to eukaryotic translation initiation factor 4E (EIF4E). This leaves EIF4E free to bind eukaryotic initiation factor 4 gamma proteins (EIF4Gs), which then bind 5′ caps of mRNAs and recruit other factors required for translation (reviewed in [22]). Of note, some of mTORC1’s function is inhibited by the immunosuppressant rapamycin [23].

mTOR plays an important role in ribosome biogenesis through control of the translation of mRNAs encoding r‐proteins. In higher eukaryotes, mRNAs that code for r‐proteins are denoted by a 5′ terminal oligopyrimidine (5′TOP) motif [24]. This motif begins with a C and contains, on average, 12.2 nucleotides in humans [25]. When cells are stimulated with nutrients, 5′TOP mRNAs are recruited to ribosomes for increased translation [26]. Rapamycin treatment prevents the translation of 5′TOP mRNAs, and mutation of the 5′TOP ­renders the mRNA’s translation insensitive to rapamycin treatment [27–29]. More recent studies have shown that mTORC1 affects 5′TOP translation through its phosphorylation of EIF4E‐binding proteins (4E‐BPs) [30]. Phosphorylation by active mTORC1 allows EIF4E to bind EIF4G1 [30]. 5′TOP mRNAs require this EIF4E–EIF4G1 interaction for their cap binding and translation, while other mRNAs do not, providing an explanation for mTOR’s specific role in 5′TOP translation after nutrient stimulation [30]. Regarding feeding and nutrient stimulation, the liver relies heavily on mTORC1 signaling. Nutrient uptake, as well as uptake of amino acids such as leucine, stimulates ribosome ­biogenesis in the liver [31]. In the livers of rats administered leucine, mTOR signaling is upregulated and translation of 5′TOP‐containing mRNAs is increased [32]. This response can be inhibited using the drug rapamycin [32]. Translation of ribosomal protein mRNAs is also upregulated during liver ­



15:  Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver

regeneration in rats, although the precise mechanism of action is unknown [33, 34]. Additionally, in the liver, mTOR signaling acts as an important regulatory mechanism in the body’s response to ethanol. Chen et  al. found decreased levels of the protein DEPTOR, a  known inhibitor of mTORC1 phosphorylation, in both a chronic‐plus‐binge ethanol‐fed mouse model of alcoholic steatosis and in patients with alcoholic liver disease [35]. Decreased DEPTOR leads to increased levels of 4EBP1 and S6K phosphorylation [35]. Phosphorylated S6K in turn phosphorylates sterol regulator element‐binding transcription factor 1 (SREBF1), which translocates to the nucleus and upregulates the expression of proteins involved in fatty acid biosynthesis [36, 37]. Phosphorylated mTORC1 can also phosphorylate lipin 1, promoting the transcription of SREBF1 which contributes to the build up of lipids in alcoholic steatosis [38]. Overall, mTOR signaling highlights the importance of regulating ribosome ­biogenesis for proper liver health.

Regulation of RNAPI transcription RNAPI transcriptional regulation is critical in the liver, since the liver responds so dramatically to nutrient availability. Upon feeding, it has been shown that RNAPI activity quickly increases [2]. Similarly, RNAPI activity increases during liver regeneration, in addition to the increased r‐protein translation previously mentioned (cited many times, including in [39, 40]). The tandemly repeated rDNA is transcribed by RNAPI as a large (47S) polycistronic precursor that contains the 18S, 5.8S, and 28S mature rRNAs. Only about half of the rDNA repeats are active at any given time, and RNAPI transcription is tightly regulated to ensure generation of an adequate number of ribosomes for cell growth or maintenance. RNAPI transcription initiation begins with the binding of upstream binding transcription factor (UBTF, commonly referred to as UBF) to the upstream control element (UCE) (Figure 15.3). UBF binding then helps to recruit selectivity factor 1 (SL1), a species‐specific complex containing the TATA‐binding protein (TBP) and three TBP‐ associated ­factors (TAFs) in humans [41]. The pre‐initiation complex is formed when SL1 binds to the core promoter (CP) of rRNA genes along with RNAPI. The interaction between SL1 and RNAPI is mediated by TIF‐1A (Rrn3 in S. cerevisiae) [42]. Multiple signaling pathways regulate rDNA transcription. Among these is again mTOR signaling, which alters RNAPI transcription through multiple mechanisms. These mechanisms include S6K phosphorylation which in turn phosphorylates the C‐terminal tail of UBF, aiding its recruitment of SL1 and the rest of the RNAPI transcription machinery [43]. mTOR also works to activate TIF‐1A by promoting phosphorylation of

UBF UCE

SL1/TIF-IB

TIF-IA A43

PoI I

CP

Figure 15.3  Schematic representation of the transcription initiation complex at the promoter of the rRNA gene. See text for details.

177

residue S44 and decreasing the amount of S199 phosphorylation, ultimately increasing TIF‐1A’s interaction with SL1 [44]. Additionally, mTOR may bind directly to the RNAPI promoter [45]. The MAPK signaling pathway also has been shown to play a role in rDNA transcriptional regulation through multiple mechanisms. In the MAPK pathways, extracellular stimulation by factors such as epidermal growth factor (EGF) triggers a cascade of protein kinases which in turn promote factors necessary for growth and development. Like mTOR signaling, MAPK phosphorylation of two residues on TIF‐1A increases RNAPI initiation [46]. Additionally, UBF is phosphorylated via the MAPK signaling cascade. UBF phosphorylation may then enhance UBF’s interaction with SL1 and/or disrupt formation of the enhanceosomes that are promoted by UBF and slow down RNAPI [43, 47]. Both options increase RNAPI activity. Signaling through the retinoblastoma protein (pRB), a pocket family protein and tumor suppressor, can also affect ribosome biogenesis via modulation of RNAPI transcription. This occurs both directly through the RNAPI transcription machinery and indirectly through the interaction of pRB with other factors. pRB, as well as the pocket family protein p130, directly affects rDNA transcription by disrupting the interaction between SL1 and UBF [48, 49]. Indirectly, pRB modulates rDNA transcription through its interactions with E2F1. When pRB is hyperphosphorylated, it cannot bind to E2F transcription factors. These E2Fs are then free to activate transcription of several ­target genes that move the cell cycle successfully from G1 to S phase. Conversely, when the cell is stressed, pRB becomes hypophosphorylated, is able to bind and inhibit E2Fs, and the cell cycle is stopped, ultimately feeding back on ribosome biogenesis (reviewed in [50]). Additionally, at low expression of E2F1, E2F1 is able to bind directly to the rDNA promoter, activating RNAPI transcription [51]. At high E2F1 levels, p14ARF is expressed, and p14ARF binds to E2F1 and inhibits its transcriptional activity so that rDNA transcription is decreased [52]. In  turn, decreased rDNA transcription feeds back on E2F1 expression to reduce it [53].

Regulation of ribosome biogenesis by MYC MYC acts as a transcription factor to regulate expression of numerous proteins required for ribosome biogenesis. It has been studied regarding hepatocyte proliferation and growth, but MYC also plays a role in many other tissues and is especially significant in cancer [54]. While the MYC transcription factor family contains multiple MYC proteins (including c‐MYC, n‐MYC, and l‐MYC), c‐MYC is the best studied with regards to ribosome biogenesis. c‐MYC modulates the transcription mediated by all three RNA polymerases. To enhance RNAPI activity, c‐MYC can regulate UBF and TIF‐1A expression [55, 56]. c‐MYC is also located in the nucleolus and can directly a­ ssociate with SL1 to stimulate RNAPI transcription [57, 58]. Additionally, c‐MYC and n‐MYC regulate RNAPII transcription of r‐proteins [59–61]. c‐MYC regulates transcription of the 5S rRNA as well by interacting with the RNAPIII transcription machinery [62]. MYC’s role in ribosome biogenesis is critical for proper liver function. Overexpression of c‐MYC in hepatocytes results in enlarged nucleoli as well as increased expression of several

178

THE LIVER:  REGULATION OF RIBOSOME BIOGENESIS

proteins required for ribosome biogenesis [63]. Additionally, c‐MYC plays a role in several liver diseases [54]. For example, c‐MYC expression is upregulated in the livers of patients with alcoholic liver disease (ALD) and after ethanol feeding in a mouse model that mimics the early stages of ALD [64]. Interestingly, combined c‐MYC overexpression and ethanol feeding results in decreased p53 levels [64], indicating a possible mechanism for the liver dysplasia observed in ALD patients.

Nucleolar stress regulated through p53 Regulation of ribosome biogenesis is also controlled by the tumor suppressor p53 in a process known as nucleolar stress (reviewed in [65, 66]). In a normally functioning cell, r‐proteins are engaged in the process of translation as structural components of ribosomes. In this scenario, MDM2 ubiquitinates the p53 protein, targeting p53 for degradation (Figure  15.4). However, when there are perturbations in ribosome biogenesis, r‐proteins become disengaged. Two of these r‐proteins, uL18 (RPL5) and uL5 (RPL11), along with the 5S rRNA, form the 5S ribonucleoprotein complex (5S RNP). The 5S RNP then binds to MDM2, blocking MDM2’s ability to ubiquitinate p53, ­ultimately resulting in stabilization of p53 levels (Figure 15.4) [67, 68]. p53 stabilization causes cellular senescence and ­apoptosis through regulating the transcription of several known proteins [69]. p53 also plays a key role in a number of liver diseases (reviewed in [70]). Notably, p53 is important in liver regeneration, where it must first be stabilized after injury to remove the affected tissue, and then downregulated to allow for proliferation of new tissue [71]. p53 stabilization also occurs in the livers  of mice during nutrient deprivation [72]. While scientists have proposed a 5′ adenosine monophosphate kinase (AMPK)‐ dependent mechanism of p53 stabilization, AMPK‐depleted

cells are still able to stabilize p53, leaving open the possibility of an additional aspect of the nucleolar stress response in the liver during starvation [72].

Regulation of ribosome biogenesis is coordinated with circadian rhythms In mice, humans, and birds, it has been shown that liver size fluctuates according to daily rhythms [73–76]. These daily rhythms, governed by the organism’s internal circadian clock, control a number of cellular cues that result in liver size changes. For example, mice generally eat at night, during their most active phase. Throughout the night, therefore, mouse liver size increases and, conversely, liver size decreases during the day when mice are not eating (resting phase) [75]. Interestingly, findings by Sinturel et al. have shown that when an organism eats is crucial for regulating this liver mass cycle, since mice fed during the day (not during their normal feeding time at night) do not undergo these changes [75]. At the biochemical level, liver size fluctuations correlate with the changes in protein translation controlled by the number of ribosomes present in the hepatocytes (i.e. more ribosomes leads to larger cell size and thus larger liver mass) [75]. Interestingly, the increased number of ribosomes is not due to the production of more ribosomes by increased RNAPI transcription, but rather to the decreased ability to eliminate excess ribosomes. During the active phase of the mice, more small subunit r‐proteins are  translated, likely through increased mTOR signaling as ­discussed above [75, 77–79]. During the resting phase, these r‐proteins are not made, resulting in excess pre‐18S rRNA. To reduce the number of ribosomes in the cell, pre‐18S rRNAs are polyadenylated [80]. In the resting liver cells, this polyadenylation then targets these RNAs for degradation by the

Normal growing cell MDM2

Ub

Cell growth and proliferation (steady state p53 levels)

p53 Functional ribosomes

Nucleolar stress uL18 5S rRNA

p14ARF

uL18 5S rRNA

uL5

uL5

p14ARF p53

uL18

Free ribosomal proteins

MDM2

5S rRNA

Cell cycle arrest and apoptosis (p53 levels increased)

MDM2

uL5

Figure 15.4  The nucleolar stress response in human cells. During normal growth conditions, MDM2 ubiquitinates p53, marking it for degradation and allowing cell growth and proliferation (top). Under conditions of nucleolar stress, r‐proteins are no longer engaged in functional ribosomes. The 5S RNP containing uL18, uL5, and the 5S rRNA binds MDM2, either with or without p14ARF, to prevent MDM2 from ubiquitinating p53. p53 levels are therefore stabilized, triggering cell cycle arrest and apoptosis (bottom). Reproduced from [65] with permission of Portland Press Limited.



15:  Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver

catalytic component of the exosome, the 3′‐to‐5′ exonuclease EXOSC10, and fewer ribosomes are produced overall [75, 80]. This strategy for controlling ribosome biogenesis may be especially useful in understanding the tissue‐specific intricacies of the liver diseases discussed below.

ROLE OF RIBOSOME BIOGENESIS IN LIVER DISEASE North American Indian childhood cirrhosis North American Indian childhood cirrhosis (NAIC, OMIM 604901) is a liver‐specific ribosomopathy that results from the mutation of the ribosome biogenesis factor UTP4 (previously Cirhin) [81, 82]. Identified in the Ojibway‐Cree First Nations children from Quebec, Canada, NAIC is characterized by ­transient neonatal jaundice that progresses to biliary cirrhosis [81, 83]. Thus far, the only effective treatment for this disease is liver transplantation [83]. In this autosomal recessive disorder, all patients have a homozygous missense mutation on chromosome 16 (16q22) in the UTP4 gene (R565W) [81, 84]. The UTP4 protein is a member of the t‐Utp subcomplex, which is required for pre‐rRNA transcription and processing [85, 86]. Specifically, human UTP4 is required for small subunit pre‐rRNA processing at sites A′, A0, 1, and 2a (Figure 15.2) [87]. It is intriguing, therefore, that a defect in the function of this ubiquitous and essential protein leads to liver‐specific pathology. Few models have been developed for studying NAIC. In a zebrafish model of NAIC, no pre‐rRNA processing defects were observed in zebrafish depleted of UTP4 using a morpholino targeting either the start site or the splice acceptor site between exons 14 and 15 [88], despite the established role of UTP4 in yeast and human pre‐rRNA processing [85–87]. However, modest increases in p53 levels were observed in the UTP4 zebrafish model, and the biliary defects seen were abrogated in a p53‐ mutated background [88]. While NAIC is the only disorder of ribosome biogenesis, or ribosomopathy, known to affect liver function specifically, this mechanism of p53 stabilization also occurs in other ribosomopathies [65]. It has also been reported that the Utp4 homozygous knockout mouse is embryonic lethal, while the heterozygous mouse develops normally [89], although these results have not been fully described. Further definition of the role of UTP4 in liver pathogenesis is therefore needed.

Shwachman–Diamond syndrome Shwachman–Diamond syndrome (SDS, OMIM 260400) is a ribosomopathy that usually presents in early childhood with bone marrow failure (neutropenia) and/or pancreatic insufficiency manifested as failure to thrive [90, 91]. Ninety percent of patients have mutations in the Shwachman–Bodian–Diamond syndrome gene (SBDS; SDO1 in yeast), though recently mutations have also been found in DNAC21 and EFL1 [90, 92]. All of the proteins encoded by these genes participate in the late steps in cytoplasmic maturation of the LSU (60S) subunit of the ribosome. SDS can progress to myelodysplastic syndrome and

179

acute myelogenous leukemia with increasing patient age [90, 91, 93]. The median survival of a cohort of SDS patients was recently found to be only 41 years [93]. SDS patients often have liver abnormalities such as hepatomegaly, elevated transaminases and mildly elevated bile acid levels, particularly early in childhood [94]. Most often the transaminase levels normalize after infancy though the bile acid levels often remain elevated. In some patients over the age of 30, liver microcysts, but not fatty liver or cirrhosis, can be visualized by magnetic resonance imaging. Taken together, these findings suggest a connection between the liver and SDS. This is a ripe topic for further follow‐up using new liver imaging techniques. In addition, the development of the gut microbiome in children may also play an important role [95].

Hepatitis B/C and hepatocellular carcinoma For hundreds of years, pathologists have used the nucleolus to prognosticate cancer without truly understanding the connection between them [96]. What is known is that many diverse cancer cells have increased numbers and size of nucleoli, and that these findings correlate with a worse cancer prognosis (reviewed in [97]). In hepatocellular carcinoma (HCC), increased nucleolar size has been shown to predict the development of HCC in patients with chronic liver disease [98–100]. This correlation is especially strong in patients infected with hepatitis B virus (HBV). Interestingly, the HBx oncoprotein of HBV has been known to directly affect nucleolar function by binding to the protein NPM1, which aids in bringing HBx to the nucleolus [101]. Once in the nucleolus, HBx–NPM1 works to increase rDNA transcription, leading to increased cellular proliferation and transformation [101]. The function of the hepatitis C virus (HCV) is also linked to ribosome biogenesis. rDNA transcription is upregulated in HCV‐infected cells [102, 103] via UBF, where UBF is phosphorylated after HCV activation of cell cycle proteins cyclin D1 (CCND1) and cyclin‐dependent kinase 4 (CDK4). Additionally, NS5B, the RNA polymerase required for HCV replication, also interacts with the nucleolar phosphoprotein nucleolin (NCL) [104]. There is thus a strong link between the nucleolus and viruses responsible for chronic liver disease and HCC.

Fragile X syndrome Fragile X syndrome is caused by mutation in the FMR1 gene on the X chromosome that results in a greater frequency of a CGG repeated sequence, leading to transcriptional silencing of FMR1 gene expression [105]. Loss of the FMR1 protein (FMRP) causes a range of symptoms including intellectual disability and obesity [106]. The cell responds to FMRP loss by increasing mTOR and ERK signaling, both of which are crucial to ribosome biogenesis and protein synthesis [107, 108]. Decreases in eIF4E phosphorylation cause a concomitant increase in the translation of a protein important for neuronal function, matrix metalloproteinase 9 (MMP‐9). Thus, FMRP loss in fragile X causes an increase in MMP‐9, which in turn results in intellectual disability.

180

THE LIVER:  REFERENCES

Although fragile X syndrome is not a liver‐specific disorder, recent studies have suggested that it may be treated with metformin, a drug that affects liver function [109, 110]. Metformin is prescribed to manage high blood sugar in patients with type 2 diabetes [111]. Metformin was tested as a therapeutic agent for fragile X syndrome because it inhibits the mTORC1 and MAPK/ ERK pathways that are activated in fragile X patients and that are important for liver development [112]. Interestingly, in Frm1−/y mice, metformin treatment did restore many of the cognitive defects, including increasing social preference and decreasing repetitive behaviors. Additionally, defects in insulin signaling and in circadian rhythms were also rescued by metformin treatment in a Drosophila model of fragile X syndrome [113]. Interestingly, metformin treatment did not rescue levels of ERK phosphorylation in the livers of Frm1−/y mice, despite rescuing these levels in the brain [109]. Additionally, it is possible that metformin treatment is influencing other pathways that are affected by FMRP loss, such as gluconeogenesis and the gut microbiome, and it is the rescue of these other pathways that contributes to metformin’s effect on the signs and symptoms of fragile X syndrome [109]. In liver cancer cells, the drug metformin has recently been suggested to have an anticancer effect through the DEPTOR–mTOR pathway [114], suggesting an unexplored mechanism of metformin action as related to FMRP loss.

CONCLUSION Several liver‐specific disorders are directly influenced by ribosome biogenesis in humans. Understanding the cellular biology of human ribosome biogenesis therefore has far‐reaching consequences for treating liver disease. However, we have only a modest understanding of the complex process of ribosome biogenesis itself in humans, and we are only beginning to understand the layers of regulation needed to modulate ribosome biogenesis at the tissue level. Indeed, there are many questions remaining, including how a such a ubiquitous process as ribosome biogenesis may cause liver‐specific defects. Regardless, it is clear that the process of making ribosomes and the regulation of this process are important for liver metabolism and disease pathogenesis. Therefore, targeting the nucleolus for therapeutic intervention may be a viable option in the future of treatment of liver‐associated pathologies.

REFERENCES 1. Anand, P. and Gruppuso, P.A. Rapamycin inhibits liver growth during ­refeeding in the rat via control of ribosomal protein translation but not cap‐ dependent translation initiation. J Nutr, 2006;136(1):27–33. 2. Conde, R.D. and Franze‐Fernandez, M.T. Increased transcription and decreased degradation control and recovery of liver ribosomes after a period of protein starvation. Biochem J, 1980;192(3):935–40. 3. Addis, T., Poo, L.J., and Lew, W. The rate of protein formation in the organs  and tissues of the body: I. after casein refeeding. J Biol Chem, 1936;116(1):343–52. 4. Derenzini, M., Montanaro, L., and Trere, D. Ribosome biogenesis and cancer. Acta Histochem, 2017;119(3):190–7.

5. Parlato, R. and Kreiner, G. Nucleolar activity in neurodegenerative diseases:  a missing piece of the puzzle? J Mol Med (Berl), ­ 2013;91(5):541–7. 6. McCann, K.L. and Baserga, S.J. Genetics. Mysterious ribosomopathies. Science, 2013;341(6148):849–50. 7. Danilova, N. and Gazda, H.T. Ribosomopathies: how a common root can cause a tree of pathologies. Dis Models Mech, 2015;8(9):1013–26. 8. Monty, K.J., Litt, M., Kay, E.R., and Dounce, A.L. Isolation and properties of liver cell nucleoli. J Biophys Biochem Cytol, 1956;2(2):127–45. 9. Farley, K., Surovtseva, Y., Merkel, J., and Baserga, S. Determinants of mammalian nucleolar architecture. Chromosoma, 2015;124(3):323–31. 10. Shea, J.R. and Leblond, C.P. Number of nucleoli in various cell types of the mouse. J Morphol, 1966;119(4):425–33. 11. Krüger, T., Zentgraf, H., and Scheer, U. Intranucleolar sites of ribosome ­biogenesis defined by the localization of early binding ribosomal proteins. J Cell Biol, 2007;177(4):573. 12. Feric, M., Vaidya, N., Harmon, T.S. et al. Coexisting liquid phases underlie nucleolar subcompartments. Cell, 2016;165(7):1686–97. 13. Sloan, K.E., Warda, A.S., Sharma, S. et al. Tuning the ribosome: The influence of rRNA modification on eukaryotic ribosome biogenesis and function. RNA Biol, 2017;14(9):1138–52. 14. Henras, A.K., Plisson‐Chastang, C., O’Donohue, M.‐F., Chakraborty, A., and Gleizes, P.‐E. An overview of pre‐ribosomal RNA processing in eukaryotes. Wiley Interdiscip Rev RNA, 2015;6(2):225–42. 15. Woolford Jr., J.L. and Baserga, S.J. Ribosome biogenesis in the yeast Saccaromyces cerevisiae. Genetics [Internet], 2013;195:1–39. 16. Farley‐Barnes, K.I., McCann, K.L., Ogawa, L.M. et al. Diverse regulators of human ribosome biogenesis discovered by changes in nucleolar number. Cell Rep, 2018;22(7):1923–34. 17. Badertscher, L., Wild, T., Montellese, C. et al. Genome‐wide RNAi screening identifies protein modules required for 40S subunit synthesis in human cells. Cell Rep, 2015;13(12):2879–91. 18. Wild, T., Horvath, P., Wyler, E. et al. A protein inventory of human ribosome biogenesis reveals an essential function of exportin 5 in 60S subunit export. PLoS Biol, 2010;8(10):e1000522. 19. Tafforeau, L., Zorbas, C., Langhendries, J.L. et al. The complexity of human ribosome biogenesis revealed by systematic nucleolar screening of pre‐ rRNA processing factors. Mol Cell, 2013;51(4):539–51. 20. Zemp, I. and Kutay, U. Nuclear export and cytoplasmic maturation of ­ribosomal subunits. FEBS Lett, 2007;581(15):2783–93. 21. Gentilella, A., Kozma, S.C., and Thomas, G. A liaison between mTOR signaling, ribosome biogenesis and cancer. Biochim Biophys Acta, ­ 2015;1849(7):812–20. 22. Hay, N. and Sonenberg, N. Upstream and downstream of mTOR. Genes Dev, 2004;18(16):1926–45. 23. Thoreen, C.C., Kang, S.A., Chang, J.W. et  al. An ATP‐competitive mammalian target of rapamycin inhibitor reveals rapamycin‐resistant functions of mTORC1. J Biol Chem, 2009;284(12):8023–32. 24. Yoshihama, M., Uechi, T., Asakawa, S. et al. The human ribosomal protein genes: sequencing and comparative analysis of 73 genes. Genome Res, 2002;12(3):379–90. 25. Perry, R.P. The architecture of mammalian ribosomal protein promoters. BMC Evolutionary Biol, 2005;5(1):15. 26. Meyuhas, O. and Kahan, T. The race to decipher the top secrets of TOP mRNAs. Biochim Biophys Acta, 2015;1849(7):801–11. 27. Jefferies, H.B., Reinhard, C., Kozma, S.C., and Thomas, G. Rapamycin selectively represses translation of the “polypyrimidine tract” mRNA family. Proc Natl Acad Sci U S A, 1994;91(10):4441–5. 28. Terada, N., Patel, H.R., Takase, K. et  al. Rapamycin selectively inhibits translation of mRNAs encoding elongation factors and ribosomal proteins. Proc Natl Acad Sci U S A, 1994;91(24):11477–81. 29. Jefferies, H.B., Fumagalli, S., Dennis, P.B., Reinhard, C., Pearson, R.B., and Thomas, G. Rapamycin suppresses 5’TOP mRNA translation through ­inhibition of p70s6k. EMBO J, 1997;16(12):3693–704. 30. Thoreen, C.C., Chantranupong, L., Keys, H.R. et al. A unifying model for mTORC1‐mediated regulation of mRNA translation. Nature, 2012;485:109. 31. Anthony, T.G., Anthony, J.C., Yoshizawa, F., Kimball, S.R., and Jefferson, L.S. Oral administration of leucine stimulates ribosomal protein mRNA translation but not global rates of protein synthesis in the liver of rats. J Nutr, 2001;131(4):1171–6.



15:  Ribosome Biogenesis and its Role in Cell Growth and Proliferation in the Liver

32. Reiter, A.K., Anthony, T.G., Anthony, J.C., Jefferson, L.S., and Kimball, S.R. The mTOR signaling pathway mediates control of ribosomal protein mRNA translation in rat liver. Int J Biochem Cell Biol, 2004;36(11):2169–79. 33. Aloni, R., Peleg, D., and Meyuhas, O. Selective translational control and nonspecific posttranscriptional regulation of ribosomal protein gene expression during development and regeneration of rat liver. Mol Cell Biol, 1992;12(5):2203–12. 34. Yo‐ichi, N. and Kikuo, O. Stimulation of the synthesis of ribosomal proteins  in regenerating rat liver with special reference to the increase in the amounts of effective mRNAs for ribosomal proteins. Eur J Biochem, 1980;107(2):323–9. 35. Chen, H., Shen, F., Sherban, A. et al. DEPTOR suppresses lipogenesis and ameliorates hepatic steatosis and acute‐on‐chronic liver injury in alcoholic liver disease. Hepatology, 2018. 36. Porstmann, T., Santos, C.R., Griffiths, B. et al. SREBP activity is regulated by mTORC1 and contributes to Akt‐dependent cell growth. Cell Metab, 2008;8(3–3):224–36. 37. Düvel, K., Yecies, J.L., Menon, S. et al. Activation of a metabolic gene regulatory network downstream of mTOR complex 1. Mol Cell, 2010;39(2):171–83. 38. Peterson, T.R., Sengupta, S.S., Harris, T.E. et al. mTOR complex 1 regulates lipin 1 localization to control the SREBP pathway. Cell, 2011;146(3):408–20. 39. Tsukada, K. and Lieberman, I. Synthesis of ribonucleic acid by liver nuclear and nucleolar preparations after partial hepatectomy. J Biol Chem, 1964;239(9):2952–6. 40. Dabeva, M.D. and Dudov, K.P. Transcriptional control of ribosome production in regenerating rat liver. Biochem J, 1982;208(1):101–8. 41. Comai, L., Tanese, N., and Tjian, R. The TATA‐binding protein and associated factors are integral components of the RNA polymerase I transcription factor, SL1. Cell, 1992;68(5):965–76. 42. Miller, G., Panov, K.I., Friedrich, J. et al. hRRN3 is essential in the SL1‐ mediated recruitment of RNA Polymerase I to rRNA gene promoters. EMBO J, 2001;20(6):1373–82. 43. Hannan, K.M., Brandenburger, Y., Jenkins, A. et al. mTOR‐dependent regulation of ribosomal gene transcription requires S6K1 and is mediated by phosphorylation of the carboxy‐terminal activation domain of the nucleolar transcription factor UBF. Mol Cell Biol, 2003;23(23):8862–77. 44. Mayer, C., Zhao, J., Yuan, X., and Grummt, I. mTOR‐dependent activation of the transcription factor TIF‐IA links rRNA synthesis to nutrient availability. Genes Dev, 2004;18(4):423–34. 45. Tsang, C.K., Liu, H., and Zheng, X.F.S. mTOR binds to the promoters of RNA polymerase I‐ and III‐transcribed genes. Cell Cycle, 2010;9(5):953–7. 46. Zhao, J., Yuan, X., Frödin, M., and Grummt, I. ERK‐dependent phosphorylation of the transcription initiation factor TIF‐IA is required for RNA polymerase I transcription and cell growth. Mol Cell, 2003;11(2):405–13. 47. Stefanovsky, V., Langlois, F., Gagnon‐Kugler, T., Rothblum, L.I., and Moss, T. Growth factor signaling regulates elongation of RNA polymerase I transcription in mammals via UBF phosphorylation and r‐chromatin remodeling. Mol Cell, 2006;21(5):629–39. 48. Hannan, K.M., Hannan, R.D., Smith, S.D. et al. Rb and p130 regulate RNA polymerase I transcription: Rb disrupts the interaction between UBF and SL‐1. Oncogene, 2000;19(43):4988–99. 49. Cavanaugh, A.H., Hempel, W.M., Taylor, L.J. et al. Activity of RNA polymerase I transcription factor UBF blocked by Rb gene product. Nature, 1995;374(6518):177–80. 50. Donati, G., Montanaro, L., and Derenzini, M. Ribosome biogenesis and control of cell proliferation: p53 is not alone. Cancer Res, 2012;72(7):1602. 51. Ayrault, O., Andrique, L., and Séité, P. Involvement of the transcriptional factor E2F1 in the regulation of the rRNA promoter. Exp Cell Res, 2006;312(7):1185–93. 52. Eymin, B., Karayan, L., Seite, P. et al. Human ARF binds E2F1 and inhibits its transcriptional activity. Oncogene, 2001;20(9):1033–41. 53. Donati, G., Brighenti, E., Vici, M. et al. Selective inhibition of rRNA transcription downregulates E2F‐1: a new p53‐independent mechanism linking cell growth to cell proliferation. J Cell Sci, 2011;124(17):3017. 54. Zheng, K., Cubero, F.J., and Nevzorova, Y.A. c‐MYC – making liver sick: role of c‐MYC in hepatic cell function, homeostasis and disease. Genes, 2017;8(4):123. 55. Poortinga, G., Wall, M., Sanij, E. et  al. c‐MYC coordinately regulates ­ribosomal gene chromatin remodeling and Pol I availability during granulocyte differentiation. Nucl Acids Res, 2011;39(8):3267–81.

181

56. Poortinga, G., Hannan, K.M., Snelling, H. et al. MAD1 and c‐MYC regulate UBF and rDNA transcription during granulocyte differentiation. EMBO J, 2004;23(16):3325–35. 57. Arabi, A., Wu, S., Ridderstråle, K. et al. c‐Myc associates with ribosomal DNA and activates RNA polymerase I transcription. Nat Cell Biol, 2005;7:303. 58. Grandori, C., Gomez‐Roman, N., Felton‐Edkins, Z.A. et al. c‐Myc binds to human ribosomal DNA and stimulates transcription of rRNA genes by RNA polymerase I. Nat Cell Biol, 2005;7(3):311–18. 59. Boon, K., Caron, H.N., van Asperen, R. et al. N‐myc enhances the expression of a large set of genes functioning in ribosome biogenesis and protein synthesis. EMBO J, 2001;20(6):1383–93. 60. Coller, H.A., Grandori, C., Tamayo, P. et al. Expression analysis with oligonucleotide microarrays reveals that MYC regulates genes involved in growth, cell cycle, signaling, and adhesion. Proc Natl Acad Sci U S A, 2000;97(7):3260–5. 61. Menssen, A. and Hermeking, H. Characterization of the c‐MYC‐regulated transcriptome by SAGE: Identification and analysis of c‐MYC target genes. Proc Natl Acad Sci U S A, 2002;99(9):6274–9. 62. Gomez‐Roman, N., Grandori, C., Eisenman, R.N., and White, R.J. Direct activation of RNA polymerase III transcription by c‐Myc. Nature, 2003;421:290. 63. Kim, S., Li, Q., Dang, C.V., and Lee, L.A. Induction of ribosomal genes and hepatocyte hypertrophy by adenovirus‐mediated expression of c‐Myc in vivo. Proc Natl Acad Sci U S A, 2000;97(21):11198–202. 64. Nevzorova, Y.A., Cubero, F.J., Hu, W. et al. Enhanced expression of c‐myc in hepatocytes promotes initiation and progression of alcoholic liver disease. J Hepatol, 2016;64(3):628–40. 65. Farley, K.I. and Baserga, S.J. Probing the mechanisms underlying human diseases in making ribosomes. Biochem Soc Trans, 2016;44(4):1035–44. 66. Woods, S.J., Hannan, K.M., Pearson, R.B., and Hannan, R.D. The nucleolus as a fundamental regulator of the p53 response and a new target for cancer therapy. Biochim Biophys Acta, 2015;1849(7):821–9. 67. Sloan Katherine, E., Bohnsack Markus, T., and Watkins Nicholas, J. The 5S RNP couples p53 homeostasis to ribosome biogenesis and nucleolar stress. Cell Rep, 2013;5(1):237–47. 68. Donati, G., Peddigari, S., Mercer, C.A., and Thomas, G. 5S ribosomal RNA is an essential component of a nascent ribosomal precursor complex that regulates the Hdm2‐p53 checkpoint. Cell Rep, 2013;4(1):87–98. 69. Fischer, M. Census and evaluation of p53 target genes. Oncogene, 2017;36:3943. 70. Krstic, J., Galhuber, M., Schulz, T., Schupp, M., and Prokesch, A. p53 as a dichotomous regulator of liver disease: the dose makes the medicine. Int J Mol Sci, 2018;19(3):921. 71. Fan, X., Chen, P., Tan, H. et  al. Dynamic and coordinated regulation of keap1‐NRF2‐ARE and p53/p21 signaling pathways is associated with acetaminophen injury responsive liver regeneration. Drug Metab Dispos, 2014;42(9):1532. 72. Prokesch, A., Graef, F.A., Madl, T. et al. Liver p53 is stabilized upon starvation and required for amino acid catabolism and gluconeogenesis. FASEB J, 2017;31(2):732–42. 73. Leung, N.W.Y., Farrant, P., and Peters, T.J. Liver volume measurement by ultrasonography in normal subjects and alcoholic patients. J Hepatol, 1986;2(2):157–64. 74. Wilson, W.O. and McFarland, L.Z. Diurnal changes in livers and digestive systems of Coturnix as related to three photoperiodic regimens. Poultry Sci, 1969;48(2):477–82. 75. Sinturel, F., Gerber, A., Mauvoisin, D. et al. Diurnal oscillations in liver mass and cell size accompany ribosome assembly cycles. Cell, 2017;169(4):651– 63.e14. 76. Fisher, H.I. and Bartlett, L.M. Diurnal cycles in liver weights in birds. The Condor, 1957;59(6):364–72. 77. Jouffe, C., Cretenet, G., Symul, L. et  al. The circadian clock coordinates ribosome biogenesis. PLoS Biol, 2013;11(1):e1001455. 78. Janich, P., Arpat, A.B., Castelo‐Szekely, V., Lopes, M., and Gatfield, D. Ribosome profiling reveals the rhythmic liver translatome and circadian clock regulation by upstream open reading frames. Genome Res, 2015;25(12):1848–59. 79. Atger, F., Gobet, C., Marquis, J. et al. Circadian and feeding rhythms differentially affect rhythmic mRNA transcription and translation in mouse liver. Proc Natl Acad Sci U S A, 2015;112(47):E6579–88.

182

THE LIVER:  REFERENCES

80. Shcherbik, N., Wang, M., Lapik, Y.R., Srivastava, L., and Pestov, D.G. Polyadenylation and degradation of incomplete RNA polymerase I transcripts in mammalian cells. EMBO Rep, 2010;11(2):106–11. 81. Betard, C., Rasquin‐Weber, A., Brewer, C. et al. Localization of a recessive gene for North American Indian childhood cirrhosis to chromosome region  16q22‐and identification of a shared haplotype. Am J Hum Genet, 2000;67(1):222–8. 82. Weber, A.M., Tuchweber, B., Yousef, I. et al. Severe familial cholestasis in North American Indian children: a clinical model of microfilament dysfunction? Gastroenterology, 1981;81(4):653–62. 83. Drouin, E., Russo, P., Tuchweber, B., Mitchell, G., and Rasquin‐Weber, A. North American Indian cirrhosis in children: a review of 30 cases. J Pediatr Gastroenterol Nutr, 2000;31(4):395–404. 84. Chagnon, P., Michaud, J., Mitchell, G. et al. A missense mutation (R565W) in cirhin (FLJ14728) in North American Indian childhood cirrhosis. Am J Hum Genet, 2002;71(6):1443–9. 85. Gallagher, J.E., Dunbar, D.A., Granneman, S. et al. RNA polymerase I transcription and pre‐rRNA processing are linked by specific SSU processome components. Genes Dev, 2004;18(20):2506–17. 86. Prieto, J.L. and McStay, B. Recruitment of factors linking transcription and processing of pre‐rRNA to NOR chromatin is UBF‐dependent and occurs independent of transcription in human cells. Genes Dev, 2007; 21(16):2041–54. 87. Freed, E.F., Prieto, J.L., McCann, K.L., McStay, B., and Baserga, S.J. NOL11, implicated in the pathogenesis of North American Indian childhood cirrhosis, is required for pre‐rRNA transcription and processing. PLoS Genet, 2012;8(8):e1002892. 88. Wilkins, B.J., Lorent, K., Matthews, R.P., and Pack, M. p53‐mediated biliary defects caused by knockdown of cirh1a, the zebrafish homolog of the gene responsible for North American Indian childhood cirrhosis. PLoS One, 2013;8(10):e77670. 89. Yu, B., Mitchell, G.A., and Richter, A. Cirhin up‐regulates a canonical NF‐ kappaB element through strong interaction with Cirip/HIVEP1. Exp Cell Res, 2009;315(18):3086–98. 90. Warren, A.J. Molecular basis of the human ribosomopathy Shwachman‐ Diamond syndrome. Adv Biol Regul, 2018;67:109–27. 91. Myers, K.C., Bolyard, A.A., Otto, B. et al. Variable clinical presentation of Shwachman‐Diamond syndrome: update from the North American Shwachman‐Diamond Syndrome Registry. J Pediatr, 2014;164(4):866–70. 92. Stepensky, P., Chacon‐Flores, M., Kim, K.H. et al. Mutations in EFL1, an SBDS partner, are associated with infantile pancytopenia, exocrine pancreatic insufficiency and skeletal anomalies in aShwachman‐Diamond like syndrome. J Med Genet, 2017;54(8):558–66. 93. Alter, B.P., Giri, N., Savage, S.A., and Rosenberg, P.S. Cancer in the National Cancer Institute inherited bone marrow failure syndrome cohort after fifteen years of follow‐up. Haematologica, 2018;103(1):30–9. 94. Toiviainen‐Salo, S., Durie, P.R., Numminen, K. et al. The natural history of Shwachman‐Diamond syndrome‐associated liver disease from childhood to adulthood. J Pediatr, 2009;155(6):807–11 e2. 95. Leung, D.H. and Yimlamai, D. The intestinal microbiome and paediatric liver disease. Lancet Gastroenterol Hepatol, 2017;2(6):446–55. 96. Pianese, G. Beitrag zur Histologie und Aetiologie der Carcinoma. Histologische und experimentelle Untersuchungen. Beitr Pathol Anat Allg Pathol, 1896;142:1–193.

97. Derenzini, M., Montanaro, L., and Trere, D. What the nucleolus says to a tumour pathologist. Histopathology, 2009;54(6):753–62. 98. Trere, D., Borzio, M., Morabito, A. et al. Nucleolar hypertrophy correlates with hepatocellular carcinoma development in cirrhosis due to HBV infection. Hepatology, 2003;37(1):72–8. 99. Derenzini, M., Trerè, D., Oliveri, F. et al. Is high AgNOR quantity in hepatocytes associated with increased risk of hepatocellular carcinoma in chronic liver disease? J Clin Pathol, 1993;46(8):727–9. 100. Borzio, M., Trere, D., Borzio, F. et  al. Hepatocyte proliferation rate is a powerful parameter for predicting hepatocellular carcinoma development in liver cirrhosis. Mol Pathol, 1998;51(2):96–101. 101. Ahuja, R., Kapoor, N.R., and Kumar, V. The HBx oncoprotein of hepatitis B virus engages nucleophosmin to promote rDNA transcription and cellular proliferation. Biochim Biophys Acta, 2015;1853(8):1783–95. 102. Raychaudhuri, S., Fontanes, V., Barat, B., and Dasgupta, A. Activation of ribosomal RNA transcription by hepatitis C virus involves upstream ­binding factor phosphorylation via induction of cyclin D1. Cancer Res, 2009;69(5):2057–64. 103. Kao, C.F., Chen, S.Y., and Lee, Y.H. Activation of RNA polymerase I transcription by hepatitis C virus core protein. J Biomed Sci, ­ 2004;11(1):72–94. 104. Hirano, M., Kaneko, S., Yamashita, T. et  al. Direct interaction between nucleolin and hepatitis C virus NS5B. J Biol Chem, 2003;278(7):5109–15. 105. Kremer, E., Pritchard, M., Lynch, M. et al. Mapping of DNA instability at the fragile X to a trinucleotide repeat sequence p(CCG)n. Science, 1991;252(5013):1711–14. 106. Rajaratnam, A., Shergill, J., Salcedo‐Arellano, M. et al. Fragile X s­ yndrome and fragile X‐associated disorders. F1000Res, 2017;6:2112. 107. Gkogkas, C.G., Khoutorsky, A., Cao, R. et al. Pharmacogenetic inhibition of eIF4E‐dependent Mmp9 mRNA translation reverses fragile X s­ yndrome‐ like phenotypes. Cell Rep. 2014;9(5):1742–55. 108. Hou, L., Antion, M.D., Hu, D. et al. Dynamic translational and proteasomal regulation of fragile X mental retardation protein controls mGluR‐dependent long‐term depression. Neuron, 2006;51(4):441–54. 109. Gantois, I., Khoutorsky, A., Popic, J. et  al. Metformin ameliorates core deficits in a mouse model of fragile X syndrome. Nature Med, 2017;23(6):674–7. 110. Dy ABC, Tassone, F., Eldeeb, M., Salcedo‐Arellano, M.J., Tartaglia, N., and Hagerman, R. Metformin as targeted treatment in fragile X syndrome. Clin Genet, 2018;93(2):216–22. 111. Samocha‐Bonet, D., Debs, S., and Greenfield, J.R. Prevention and t­ reatment of type 2 diabetes: a pathophysiological‐based approach. Trends Endocrinol Metab, 2018;29(6):370–9. 112. Soares, H.P., Ni, Y., Kisfalvi, K., Sinnett‐Smith, J., and Rozengurt, E. Different patterns of Akt and ERK feedback activation in response to rapamycin, active‐site mTOR inhibitors and metformin in pancreatic cancer cells. PLoS One, 2013;8(2):e57289. 113. Monyak, R.E., Emerson, D., Schoenfeld, B.P. et  al. Insulin signaling ­misregulation underlies circadian and cognitive deficits in a Drosophila fragile X model. Mol Psychiatry, 2017;22(8):1140–8. 114. Obara, A., Fujita, Y., Abudukadier, A. et  al. DEPTOR‐related mTOR ­suppression is involved in metformin’s anti‐cancer action in human liver cancer cells. Biochem Biophys Res Commun, 2015;460(4): 1047–52.

16

miRNAs and Hepatocellular Carcinoma Yusuke Yamamoto1, Isaku Kohama1, and Takahiro Ochiya1,2 Division of Molecular and Cellular Medicine, National Cancer Center Research Institute, Tsukiji, Tokyo, Japan 2 Institute of Medical Science, Tokyo Medical University, Shinjuku, Tokyo, Japan 1

INTRODUCTION Several types of small noncoding RNAs have been identified in eukaryotic cells, including microRNAs (miRNAs), small interfering RNAs (siRNAs), and Piwi‐interacting RNAs (piRNAs) [1, 2]. Revolutionary advances in next‐generation sequencing have made substantial contributions to the discovery of new classes of small RNAs and the identification of a large number of these nucleotides. miRNAs have a length of 18–24 n­ ucleotides, are endogenously expressed in almost all types of eukaryotic cells, and primarily silence gene expression at the post‐ transcriptional level. In general, miRNAs are incorporated into the RNA‐induced silencing complex (RISC) that functions as an RNA silencer and is located in the cytoplasm of cells [3]. Based on a computational prediction model, the expression of approximately 30% of all genes is influenced by miRNA‐induced gene silencing [4]. Notably, miRNAs play important roles in regulating nearly every physiological cellular process, and their dysregulation is strongly correlated with various types of disease, including cancer [3]. The expression patterns and levels of miRNAs are tightly regulated in different types of cells during development; for example, miR‐122 expression is limited to hepatocytes [5]. A number of miRNAs have been shown to regulate cellular differentiation and proliferation under physiological conditions [6, 7]. Likewise, many miRNAs have been reported to play key roles in cancer initiation and progression, including invasion and metastasis processes in almost all types of cancers [8–10]. The purpose of this chapter is to provide an overview of the functions of miRNAs in hepatocellular carcinoma (HCC) and their clinical applications, such as diagnostics and therapeutics.

THE BASICS OF miRNA BIOLOGY AND BIOGENESIS The sequences of miRNAs are highly conserved across species [11]. Primary miRNA genes (pri‐miRNAs) are transcribed by the RNA polymerase II (Pol II) transcription units, and the transcripts are called pri‐miRNAs (Figure 16.1). The pri‐miRNAs possess a cap and a poly‐A structure, as well as protein‐coding genes. The size of the pri‐miRNA usually exceeds 1 kilobase, and it contains an RNA hairpin loop structure with a mature miRNA sequence [12]. In most cases, one pri‐miRNA contains one mature miRNA; however, some pri‐miRNAs include several mature miRNAs, such as the miR‐17–92 cluster and miR‐ 106a–363 cluster [13]. Some miRNA families, particularly the let‐7 family, are widely conserved in many species and include more than 10 mature miRNAs in humans [14]. Some miRNAs are also located in the intron regions of protein‐coding genes, and intronic RNAs become pri‐miRNAs during the process of splicing; thus, the expression levels of these types of miRNAs correlate with the parent gene expression [15]. During miRNA biogenesis, pri‐miRNAs, which are transcribed by a Pol II RNA polymerase, contain hairpin loop structures (approximately 70 nucleotides) that are recognized by a DiGeorge syndrome critical region 8 (DGCR8) nuclear protein in the nucleus. DGCR8 interacts with Drosha [16], an RNase III endonuclease, to form the microprocessor complex. In the complex, Drosha cleaves the pri‐miRNAs at the base of stem‐loop structures and creates 60–70 nucleotide stem‐loop intermediates, called a precursor miRNA (pre‐miRNA) [17]. The pre‐miRNAs in the nucleus are transported to the cytoplasm by the nucleocytoplasmic shuttler exportin‐5 (Exp‐5) [18]. In the cytoplasm, the hairpin structure of pre‐miRNA is cleaved by the RNase III

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

184

THE LIVER:  miRNAs IN HEPATOCELLULAR CARCINOMA

DNA Transcribed by Pol RNA polymerase Pri-miRNA (> 1 kb)

Nucleus

Processed by Drosha, DGCR8

Pre-miRNA (60–70 nt)

Exported by Exp5

Cytoplasm

Mature-miRNA (22 nt)

5’UTR

RISC

3’UTR Protein coding sequence

AAAAAAA Translational inhibition or mRNA degradation

Figure 16.1  The biogenesis and function pathways of miRNAs. Primary miRNA genes (pri‐miRNAs) are transcribed by the Pol II transcription units, and the transcripts are called pri‐miRNAs with a cap and a poly‐A structure. In the process of miRNA biogenesis, pri‐miRNAs are recognized by a DGCR8 nuclear protein in the nucleus. DGCR8 interacts with Drosha, an RNase III endonuclease, which cleaves the pri‐miRNAs at the base of stem‐loop structures and creates a pre‐miRNA. The pre‐miRNAs in the nucleus are transported to the cytoplasm by Exp‐5. In the cytoplasm, Dicer, an RNase III enzyme cleaves pre‐miRNA, yielding an imperfect miRNA duplex of approximately 22 nucleotides. The “guide” strand of the duplex incorporates into RISC. Based on the “seed” sequence in miRNAs, the expression of target genes are inhibited by either translational inhibition (partial complementarity) or mRNA degradation (perfect complementarity).

enzyme Dicer, yielding an imperfect miRNA duplex of approximately 22 nucleotides [17]. The imperfect miRNA duplex contains a “guide” strand that is the antisense strand of the target gene sequence and a “passenger” strand. Basically, the “guide” strand of the duplex incorporates into the RNA‐induced silencing complex (RISC), although “passenger” strand of the duplex might also function as an active miRNA in some cases. The RISC contains primary miRNA‐binding proteins, namely, a member of the argonaute protein family. In mouse and human, four Ago (Ago I–IV) proteins are conserved [19]. Ago proteins possess a PAZ domain that is responsible for binding both single‐stranded and duplexed RNA; Ago II is a major protein in the RISC with helicase and endonuclease activities [20]. In the process of RISC assembly, the “passenger” strand of the imperfect miRNA duplex that was cleaved by Dicer is removed and eventually degraded. The “guide” strand is selectively incorporated into the RISC and acts as a gene silencer. When the “guide” strand of the miRNA tightly binds to the Ago protein in the RISC, the miRNA recognizes target genes through homologous sequences at the 5′ end of the miRNA, which is designated the “seed” sequence (generally, 6–8 nucleotides) [18]. Primarily, miRNAs negatively regulate the expression target genes through an interaction between the “seed” sequence at their 5′ ends and the 3′ untranslated region (3′ UTR) of their target genes. Based on the complementarity of the “seed” sequence with the target gene sequence, the mature miRNA causes either translational inhibition (partial complementarity) or mRNA degradation

(perfect complementarity) (Figure  16.1). A single miRNA is considered to possess the capacity to control hundreds of target genes and influence a wide range of gene pathways. These unique characteristics indicate that miRNAs are potentially critical modulators of nearly all physiological events and the development of various diseases. Thus, one of the key issues in miRNA biology is to identify the target genes. The initial process used to predict target genes is to observe the pairing of the 3′ UTR of target genes and miRNA “seed” sequence located from nucleotides 2 to 8 at the 5′ end of the miRNA. Several online bioinformatics tools have been developed to predict the miRNA target genes: Target Scan 7.1 (http://www.targetscan. org/vert_72/), miRD (http://mirdb.org/), miRBase (http://www. mirbase.org/help/targets.shtml), and miRWalk2.0 (http://zmf. umm.uni‐heidelberg.de/apps/zmf/mirwalk2/). These websites predict both (i) the genes that are potentially targeted by a specific miRNA and (ii) miRNAs that bind to a specific gene.

miRNAs IN HEPATOCELLULAR CARCINOMA Hepatocellular carcinoma The primary malignancies of the liver are mainly HCC, cholangiocarcinoma, and hepatic angiocarcinoma. More than 90% of

16:  miRNAs AND HEPATOCELLULAR CARCINOMA

liver cancer cases are HCC, and it frequently develops in patients with hepatitis B virus (HBV) and hepatitis C virus (HCV) infections and cirrhosis (80–90% of the HCC cases) [21]. Although the treatment outcomes are continuously improving, HCC is the third leading cause of cancer‐related death worldwide. Due to the higher prevalence rates of HBV and HCV, the incidence rates of HCC are higher in Asia and Africa than elsewhere; the total number of the HCC cases worldwide is expected to increase in the next few years. HCV is the most common cause of HCC in Japan and Western countries, but in other Asian countries and Africa, HBV is the most common cause of HCC development. The main cause of hepatitis is a virus infection, but it is also caused by alcoholism (alcoholic liver disease, ALD) and steatosis (non‐alcoholic steatohepatitis, NASH). HBV and HCV infections cause chronic hepatitis, leading to abnormal cell proliferation and death. Hepatitis also gradually results in genetic alterations in the liver. Many people who become infected with these viruses do not even know it. Some become asymptomatic long‐term HBV or HCV carriers and develop liver fibrosis, which subsequently leads to liver cirrhosis within 20 years. In the last stage of the disease, the cirrhotic liver eventually develops HCC. Approximately 80% of the HCC cases develop from liver cirrhosis. In the process of HCC development from hepatitis via liver cirrhosis, hepatocytes repeatedly grow and undergo cell death, leading to the accumulation of genetic aberrations, such as point mutations, genomic deletions, and amplifications, which are the main causes of HCC development [22]. When the patients have a sufficient hepatic function reserve, the optimal treatment for HCC is partial surgical resection. Although the 5‐year survival rates after treatment have substantially improved over the past few decades, the recurrence rates is still very high: approximately 70%. Another HCC treatment is liver transplantation. Generally, patient who are eligible for liver transplantation have multiple liver dysfunctions. Additionally, many HCC cases are detected at an advance stage; thus, less than 40% of patients with HCC are eligible for surgery and transplantation. In general, radiological approaches, such as ultrasonography and computed tomography, are used for the diagnosis and surveillance of HCC. However, these methods are not optimal to detect a small lesion in the liver. Other approaches for HCC screening are serological tests of tumor biomarkers, such as α‐ fetoprotein (AFP) and protein induced by vitamin K absence or antagonist‐II (PIVKA‐II). These methods are less invasive and can be readily applied for HCC screening; however, the low sensitivity and specificity of serological biomarkers are critical issues, particularly for the detection of early‐stage HCC [23, 24]. Recently, expression profiling of miRNAs has been considered as a novel biomarker to detect cancer. The first studies were performed using an miRNA microarray, which provided substantial contributions to miRNA biology and cancer biology. In these studies, miRNAs whose expression levels in cancer tissues were decreased compared with patient‐matched normal tissues, target oncogenes, and thus the miRNAs originally function as tumor suppressors. In contrast, miRNAs whose expression levels are increased in cancer target tumor suppressor genes and function as oncogenes in cancer development. Importantly, these differential expression profiles of miRNAs are closely associated with the clinical outcomes in patients with many types of cancer, including HCC.

185

Although the earliest studies were mainly based on the miRNA microarray platform, recent technological advances in next‐generation sequencing have also contributed to miRNA research in the field of cancer biology. Using both comprehensive approaches, several miRNAs have been identified as oncogenic and tumor‐suppressive miRNAs in HCC, and some of these miRNAs are also considered diagnostic and prognostic biomarkers. As mentioned earlier, these miRNAs likely affect various pathways by inhibiting hundreds of genes. Although many studies have been performed to investigate miRNA functions and their targets in other types of cancer, not specifically HCC, theoretically, the target genes identified for certain miRNAs in other cancers are potentially the same targets in HCC and thus the pathways affected by certain miRNAs would also be similar. Therefore, miRNA research in the field of HCC will likely contribute not only to improving our understanding of miRNA functions in HCC but also to development of innovative and feasible therapeutic methods and novel diagnostic strategies for patients with HCC, particularly for the early and accurate detection of HCC.

Oncogenic roles of miRNAs in HCC A comprehensive analysis of miRNA profiling in HCC cases provides clear evidence that miR‐21 functions as an oncogene during HCC development. Notably, miR‐21 is one of the miRNAs with the highest expression in primary HCCs, and its abundant expression is a hallmark of various types of cancers. As miR‐21 expression is substantially increased in various types of cancers, it is designated as an “oncomiR.” One of the major functions of miR‐21 in HCC is to inhibit the expression of a tumor suppressor gene called phosphatase and tensin homolog deleted from chromosome 10 (PTEN), and its expression is closely associated with the poor differentiation of tumor cells. The PTEN gene has empirically been confirmed to be a direct target of miR‐21. PTEN is a phosphatase that dephosphorylates focal adhesion kinase (FAK). When FAK is phosphorylated, it is activated. The status of FAK activation correlates with aggressive tumor behavior in HCC [25]. Therefore, miR‐21‐mediated PTEN downregulation results in an increase in FAK phosphorylation and an aggressive phenotype of HCC characterized by invasion. The inhibition of miR‐21 decreases the growth rate of HCC cells in soft agar and promotes apoptotic cell death [26]. As expected, miR‐21 overexpression promotes the growth, migration, and invasion of HCC cells [27]. The level of the miR‐21 transcript is increased by signal transducer and activator of transcription 3 (STAT3), a major mediator of interleukin 6 (IL‐6) signaling. STAT3 is involved in tumor transformation by suppressing apoptotic signaling and directly binds to the promoter region of the miR‐21 primary transcript to induce its ­transcription [28]. Another well‐known oncogenic miRNA in HCC is the miR‐17–92 polycistronic cluster. This miRNA cluster contains six miRNAs: miR‐17, miR‐18a, miR‐19a, miR‐20a, miR‐19b, and miR‐92a, which are also considered “oncomiRs.” Overexpression of the miR‐17–92 cluster was confirmed in 100% of human HCC cases, and some of these miRNAs, such as miR‐20a, were expressed in high levels in cirrhotic liver

186

THE LIVER:  miRNAs IN HEPATOCELLULAR CARCINOMA

tissues, suggesting that they might be important for the early stage of HCC development [29]. The expression of the miR‐17– 92 polycistronic cluster is positively regulated by the c‐Myc oncogene. The direct binding of c‐Myc to the promoter region in the miR‐17–92 polycistronic cluster has been reported [30]. The function of the miR‐17–92 cluster in vivo was investigated using liver‐specific miR‐17–92 transgenic mice in combination with a hepatic carcinogen (diethylnitrosamine) [31]. In this mouse model, liver‐specific miR‐17–92 transgenic mice developed a significantly greater number of HCC lesions than the control mouse group treated with diethylnitrosamine. Similar data were also obtained from in vitro experiments; this research group revealed that increased expression of the miR‐17–92 cluster clearly promoted the growth, colony formation, and invasion of human HCC cell lines [31]. Profiling of miRNA expression in HCC cell lines derived from chronic carriers of HBV and HCV and patients with nonviral‐associated HCC identified the differential expression of several miRNAs: upregulation of miR‐222, miR‐221, and miR‐31, and downregulation of miR‐223, miR‐126, and miR‐122a. Upregulated miRNAs might function as oncogenes, and in contrast, downregulated miRNAs might function as tumor suppressors. Further analyses showed that the deregulated expression patterns of miR‐223 and miR‐222 clearly differentiated HCC cases from noncancerous liver tissues, regardless of the viral infection status [32]. In another study of HCV‐infected HCC cases, the miRNA profiling in a set of 52 human primary liver tumors consisting of premalignant dysplastic liver nodules and HCCs using quantitative real‐time polymerase chain reaction (qRT‐PCR), identified 10 upregulated and 19 downregulated miRNAs compared to normal liver tissues [33]. A further analysis of these miRNAs confirmed that expression levels of miR‐122, miR‐100, and miR‐10a were significantly increased in HCV‐infected HCC cases, indicating that they functioned as putative oncogenic miRNAs. Likewise, miRNA profiling of 89 HCC samples using a ligation‐mediated amplification method identified subclasses of HCC (three main clusters: the wingless‐type MMTV integration site, interferon‐related, and proliferation) through an unsupervised clustering analysis. In a subset of the proliferation subclass, poorly characterized miRNAs from chr19q13.42 were expressed at high levels. Among these miRNAs, enhanced expression of miR‐517a and miR‐520c promoted the growth, migration, and invasion of the HCC cells in vitro. In particular, miR‐517 enhanced tumorigenesis and metastasis in an in vivo model. Thus, they are also considered oncogenic miRNAs [34]. As a new paradigm of cancer biology, many researchers have focused on tumor‐initiating cells (TICs) or cancer stem cells (CSCs), which are considered to possess a higher self‐ renewal capability and drug‐ and radio‐resistance. Ma et  al. [35] identified a CD133‐positive population of HCC cells as CSCs in HCC (approximately 1.3–13.6% of the cells in the tumor bulk) and also observed higher levels of miR‐130b expression in the CD133‐positive CSC population than in the CD133‐negative main population. Importantly, when miR‐ 130b was overexpressed in the CD133‐negative HCC cells from a lentivirus vector, the cells exhibited a higher resistance to chemotherapeutic agents and a significant increase in

tumorigenicity in vivo. In this study, TP53INP1 was identified as one of the miR‐130b target genes. Thus, miR‐130b is one of the key miRNAs regulating cancer stemness, including the self‐renewal capability and drug resistance, indicating that it is also an oncogenic miRNA [35]. The major issue in curing cancer is finding methods to prevent cancer metastasis and recurrence after curative treatment. Investigations of the molecular mechanisms of these processes will provide new insights into clinical applications for patients with HCC. In the field of miRNA biology, several miRNAs have been reported to function as upstream regulators of key molecules responsible for migration, invasion, and metastasis, which have important roles in determining liver cancer outcomes. Some oncogenic miRNAs have been described as pro‐metastatic miRNAs. For example, overexpressed oncogenic miR‐17–5p promotes the migration of HCC cells in vitro and in vivo by activating the p38 mitogen‐ activated protein kinase (MAPK) pathway and increasing the phosphorylation of heat shock protein 27 (HSP27) [36]. Ding et  al. [37] identified 22 miRNAs located at amplified or deleted genomic DNA regions in HCC. Among them, miR‐151, which is frequently amplified on chromosome 8q24.3 and is coexpressed with its host gene FAK, is closely correlated with the intrahepatic metastasis of HCC. Overexpression of miR‐151 induced the migration and invasion of HCC cells in vitro and in vivo by inhibiting the expression of RhoGDIA, a putative metastasis suppressor in HCC [37]. Another pro‐metastatic miRNA, miR‐143, has been reported, and fibronectin type III domain‐containing 3B (FNDC3B) is one of its direct target genes [38]. In this study, the intratumor administration of miR‐143 significantly enhanced HCC metastasis in an HCC cell‐xenografted mouse model. Additionally, a study using p21‐HBx transgenic mice showed that miR‐143 inhibition significantly blocked local liver metastasis and distant lung metastasis [38]. Oncogenic miR‐517 is also defined as a pro‐metastatic miRNA whose overexpression causes metastatic dissemination in vivo [34]. A study profiling the expression of 156 miRNAs in HCC cases revealed that miR‐222 upregulation was commonly observed in an HCC cohort. The suppression of miR‐222 expression led to reduced cell motility by negatively regulating AKT signaling. In this study, protein phosphatase 2A subunit B (PPP2R2A) was identified as one of the miR‐222 target genes using an in silico assay and luciferase reporter assay of the PPP2R2A 3′UTR. Based on these data, miR‐222 is also a pro‐metastatic miRNA [39]. The oncogenic miRNAs have also been reported to be key factors in intrahepatic cholangiocarcinoma (ICC). For example, miR‐191 plays an important role in tumorigenesis. Next‐ generation sequencing technology with five pairs of ICC and matched‐to‐normal bile duct tissues revealed higher levels of miR‐191 expression in ICC than in adjacent normal bile duct tissues. Overexpression of miR‐191 promotes the growth, invasion, and migration of ICC cells in vitro and in vivo. Moreover, ten–eleven translocation 1 (TET1) was identified as a direct target gene of miR‐191, and reduced TET1 expression causes p53 promoter methylation, suppressing p53 transcription. The expression of miR‐191 correlates with poor prognosis for patients with ICC [40].

16:  miRNAs AND HEPATOCELLULAR CARCINOMA

Tumor‐suppressive effect of miRNAs on HCC In the processes of cancer initiation and progression, both oncogenes and tumor suppressor genes are essential components of the mechanism regulating the cell cycle. In miRNA biology, a number of reports identified specific miRNAs that negatively regulate the cell cycle and function as tumor suppressors. Most of tumor‐suppressive miRNAs target cyclin–cyclin‐dependent‐ kinase (cyclin–CDK) complexes, a class of positive modulators of the cell cycle. A class of miRNAs were reported to function as major negative regulators of cell cycle progression; this group includes miR‐1, miR‐22, miR‐34a, miR‐122, miR‐375, and let‐7. The most interesting of these miRNAs is miR‐122, because it is abundantly expressed in the liver and comprises over 70% of the total amount of miRNAs. The expression of miR‐122 is specifically and significantly downregulated in human HCC cases compared with adjacent normal liver tissues [41]. As shown in the study by Tsai et al., miR‐122 functions as a tumor suppressor in HCC, and the authors experimentally identified 32 target genes related to cell movement, cell morphology, cell–cell signaling, and transcription [41]. The miR‐122 expression level is downregulated in most HCC cases [42]. In this study, cyclin G1 was one of the target genes of miR‐122; the expression levels of cyclin G1 inversely correlated with miR‐122 expression. Obviously, cyclin G1 is a positive regulator of cell cycle progression, leading to p53 downregulation and the induction of hepatocarcinogenesis, and as miR‐122 inhibits cyclin G1, it functions as a tumor suppressor in HCC. In contrast, the aberrant expression of miR‐122 has also been reported in patients with both HCC and HCV infection [33], presumably because miR‐122 expression in hepatocytes is essential for HCV RNA accumulation [43]. During liver development, miR‐122 expression is associated with four liver‐selective transcription factors, hepatocyte nuclear factor 1α (HNF1α), HNF3β, HNF4α, and CCAAT/enhancer‐ binding protein α (C/EBPα), and regulates the balance between the proliferation and differentiation of hepatocytes by directly inhibiting CUTL1, a transcriptional repressor of genes specifying terminal differentiation in multiple cell lineages [44]. Therefore, miR‐122 is thought to generally function as a tumor suppressor in HCC, but its roles in the specific circumstances remain unclear. Primary HCC exhibits a variety of genomic abnormalities, such as chromosomal instability, CpG hypermethylation, DNA rearrangements associated with HBV integration, DNA hypomethylation, and, to a lesser extent, microsatellite instability [45]. The hypermethylation of the promoter regions of tumor‐ suppressive miRNAs causes decreased expression and/or silencing in cancer cells. The expression of miR‐1 is downregulated in primary HCC cases compared with patient‐matched liver tissues, and miR‐1 is also a well‐known tumor suppressor miRNA. Initially, miR‐1 was identified as a methylated miRNA in HCC; thus, its expression is inhibited. Following treatment with 5‐ azacytidine (DNA hypomethylating agent) and/or trichostatin A (histone deacetylase inhibitor), miR‐1 expression is restored in HCC cells. Overexpression of miR‐1 suppresses cell growth and induces apoptotic cell death by directly targeting the FoxP1, MET, and HDAC4 genes [46]. Likewise, miR‐148 was originally identified as a hepatospecific miRNA that was expressed

187

at high levels in the mature liver and controlled the differentiation of fetal hepatoblasts by directly targeting DNA methyltransferase 1 (DNMT1), a major enzyme responsible for epigenetic silencing [47]. The authors also reported a beneficial effect of miR‐148a on HCC suppression as a tumor suppressor miRNA. Regardless of the DNMT1 expression level, miR‐148a functions as a tumor suppressor by controlling the expression of the c‐Met oncogene [47]. The miR‐375 expression level was also downregulated in HCC tissues compared with adjacent nontumor tissues from patients with HCC. Yes‐associated protein (YAP) is a potent oncogenic driver, and miR‐375 overexpression decreases the transcriptional activity of YAP, thus exerting a tumor‐suppressive effect on HCC cells [48]. The research group also identified stathmin 1 as one of downstream targets of miR‐223 using luciferase reporter assay with the 3′ UTR; stathmin 1 expression inversely correlates with miR‐223 expression in HCC cells [32]. miR‐34a is one of the most well‐known tumor suppressor miRNAs whose transcription is directly regulated by p53, the guardian of the genome. A qRT‐PCR analysis of 83 HCC formalin‐fixed paraffin‐embedded (FFPE) tissue samples revealed decreased miR‐34a expression in HCC samples compared to the patient‐matched liver tissues. As expected, ectopic expression of miR‐34a inhibits cell proliferation and induces apoptosis in HCC cells by influencing phospho‐ERK1/2 and phospho‐ STAT5 signaling. Thus, miR‐34a also functions as a tumor suppressor in HCC [49]. Tumor suppressor miRNAs also function as anti‐metastatic miRNAs to some extent. These miRNAs mainly modulate the epithelial‐to‐mesenchymal transition (EMT). As a typical EMT‐ related miRNA, the transcription of the miR‐200 family is positively regulated by p53, which suppresses tumor progression and metastasis. The upregulation of the miR‐200 family by p53 was confirmed by a profiling analysis of 92 primary HCCs and 9 HCC cell lines. The miR‐200 family targets ZEB1 and ZEB2, which are master regulators of the EMT. Thus, p53 inhibits EMT through miR‐200 family‐mediated ZEB1 and ZEB2 repression. Additionally, p53‐mediated activation of the miR‐192 family suppresses ZEB2 expression. These miRNAs could function as anti‐metastatic miRNAs by repressing the EMT phenotype [50]. c‐Met is a transmembrane tyrosine kinase receptor and an inducer of the EMT. Other tumor suppressor miRNAs, such as miR‐34a and miR‐23b, have also been reported to possess anti‐metastatic functions; miR‐34a suppresses tumor invasion and migration by directly targeting c‐Met in HepG2 cells [51], and miR‐23b expression leads to urokinase‐type plasminogen activator (uPA) and c‐Met downregulation and subsequently reduces cancer invasion and metastasis [52]. Notably, miR‐122 has been reported to function as an anti‐metastatic miRNA by directly inhibiting disintegrin and metalloproteinase domain‐containing protein 10 (ADAM10) [53] and ADAM17 [41]. ADAM family genes are closely associated with cancer metastasis; thus, miR‐122 functions as an anti‐metastatic miRNA. Recently, Su et al. identified miR‐217, whose expression level was decreased in highly invasive MHCC‐97H HCC cells and metastatic HCC tissues. Ectopic expression of miR‐217 suppressed the invasion of MHCC‐97H cells by directly repressing E2F3 [54]. More recently, miR‐ 501–3p has been described to be involved in the metastasis of

188

THE LIVER:  ROLES OF miRNAs IN HEPATITIS VIRUS INFECTION

HCC. Significantly lower expression of miR‐501–3p was observed both in metastatic HCC cell lines and recurrent and metastatic HCC tissue samples. Similar to other research reports of anti‐metastatic miRNAs, ectopically enhanced expression of miR‐501‐3p suppresses the growth, migration, invasion and EMT of metastatic HCC cells. Lin‐7 homolog A (LIN7A) was identified as a direct target of miR‐501‐3p, and its inhibition suppressed metastasis in HCC cells [55]. These tumor suppressor and/or anti‐metastatic miRNAs represent potentially useful prognostic biomarkers and treatment strategies for HCC.

ROLES OF miRNAs IN HEPATITIS VIRUS INFECTION Although miRNAs are involved in the process of HCC initiation and progression, they are also closely associated with infections with the hepatitis viruses HBV and HCV. HBV is a nuclear DNA virus, and no miRNAs have been identified in the HBV genome, although one miRNA sequence in the HBV genome was computationally predicted [56]. In contrast, HBV influences the expression of miRNAs in the host cells to create a favorable environment for HBV replication and immune evasion. HBV infection is basically characterized by two phases: the acute and chronic phases. In the acute phase, HBV actively replicates while evading the host immune attack. In the chronic phase, the HBV infection becomes dormant and the virus replicates stably and escapes from the immune system. In both infection phases, miRNAs in the host cells play important roles in the interaction between the host and virus. The most well‐studied miRNA involved in HBV infection is miR‐122, which is a liver‐ specific miRNA that is abundantly expressed in hepatocytes. Although miR‐122 is essential for the process of HCV infection, the loss of miR‐122 unexpectedly promotes the replication of HBV [57, 58]. miR‐122 directly binds to the HBV genome sequence, which is highly conserved, and suppresses the expression of the viral genes. Moreover, HBV X protein (HBx), which is essential for the transcription of the HBV genome, binds peroxisome proliferator‐activated receptor‐gamma (PPARγ) and suppresses the transcription of miR‐122, presumably inducing HBV replication [59]. The transfection of miRNA mimics showed that overexpression of miR‐1 increases HBV replication and upregulates HBV core promoter transcription, antigen expression, and progeny secretion. These effects of miR‐1 on HBV replication are not directly mediated by targeting the HBV genome, as revealed by bioinformatics and luciferase reporter analyses. Thus, miR‐1 regulates the expression of several host genes to enhance HBV replication and reverse the cancer cell phenotype [60]. Moreover, a screen with an miRNA library containing 2048 miRNAs identified 39 miRNAs that repress HBV replication by examining the intracellular and extracellular DNA and HBsAg levels. In particular, miR‐204 substantially decreases both HBV DNA and HBsAg levels in HCC cells. Rab22a was identified as one of the targets of miR‐204, and the loss of Rab22a also suppresses intracellular and extracellular HBV DNA expression [61]. On

the other hand, HBx regulates miRNA expression in the host cells. Notably, let‐7a has been identified to be negatively regulated by HBx; their expression levels are inversely correlated. As let‐7a directly targets STAT3, HBx‐mediated downregulation of let‐7a and upregulation of STAT3 support cell proliferation, leading to hepatocarcinogenesis [62]. Additionally, an miRNA microarray analysis revealed 10 miRNAs that were differentially expressed between a stable HBV‐producing cell line (HepG2.2.15) and its control cell line (HepG2). High expression of miR‐501 has been observed in HBV‐producing cells. The loss of miR‐501 expression suppresses HBV replication, but does not affect the growth of HBV‐producing cells. According to the results of a luciferase reporter assay, HBXIP, an inhibitor of HBV replication, is a potential target of miR‐501. Thus, miR‐501 would be an important miRNA regulating HBV replication in the host cells [63]. miRNAs are involved in the physiology and function of the immune system. Host innate antiviral immunity is the first line of defense against hepatitis virus infection. The defense mechanism is precisely controlled by a number of genes at multiple stages. The expression of miRNAs in the host cells was consistently shown to play a key regulatory role in the process of HBV infection and its defense. Notably, miR‐155 is upregulated during HBV infection and affects the host immune response; miR‐155 regulates the acute inflammatory response after the recognition of pathogens by toll‐like receptors. Overexpression of miR‐155 induces the expression of several interferon (IFN)‐ inducible antiviral genes and inhibits suppressor of cytokine signaling 1 (SOCS1) expression, causing STAT1 and STAT3 phosphorylation in human hepatoma cells. Additionally, miR‐155 overexpression partially inhibits HBx gene expression in vitro [64]. Likewise, a study of 98 patients with HBV‐related HCC revealed that higher expression of miR‐200c blocked HBV‐mediated PD‐L1 expression by directly targeting the 3′UTR of PD‐L1 and resulted in a better prognosis. PD‐L1 expression correlates negatively with miR‐200c expression in HBV‐related HCC cases [65]. In particular, the natural history of HBV infection in young children frequently exhibits a shift from an acute to chronic infection as the virus becomes dormant in the host cells: infected hepatocytes. When HBV forms a covalently closed circular DNA (cccDNA) in the nucleus of the host cells, it stably survives until the eventual reactivation of its life cycle, leading to the chronic infection phase [66]. Some miRNAs have been reported to be required for the chronic HBV infection. Methylation of the CpG islands in the cccDNA by DNA methyltransferase 1 (DNMT1) prevents viral gene expression. DNMT1 was also identified as a direct target of miR‐152, whose expression is frequently downregulated in HBV‐related HCCs [67]. Thus, since the inhibition of miR‐152 causes global DNA hypermethylation and increases the methylation levels of two tumor suppressor genes, glutathione S‐transferase pi 1 (GSTP1) and E‐cadherin 1 (CDH1), miR‐152 functions as a tumor suppressor of the epigenetic aberrations associated with HBV‐ related HCC [67]. In addition, a computational analysis of the HBV genome has identified seven sites that are potential targets of human liver miRNAs. These miRNA target sites are located in the cluster of a 995 bp segment within the viral polymerase open reading frame (ORF) and the overlapping surface antigen

16:  miRNAs AND HEPATOCELLULAR CARCINOMA

ORF and are conserved among the most common HBV subtypes [68]. The validation experiment using a luciferase reporter assay identified a direct interaction between miR‐125a‐5p and the HBV sequence that clearly inhibited the reporter activity of the surface antigen [68]. Liver‐specific miRNAs also participate in the HCV infection process. HCV is a single‐stranded RNA virus that infects hepatocytes and develops persistent infections. The major miRNA in the liver, miR‐122, is essential for HCV replication. The HCV RNA genome comprises a 5′ UTR, a long polyprotein ORF, and a 3′ UTR. A genetic interaction between miR‐122 and the 5′ UTR of the viral RNA genome was identified in a bioinformatics search, based on an analysis of the predicted miRNA‐binding sites [57]. Notably, HCV RNA replicates in Huh‐7 cells expressing miR‐122, but not in the HepG2 cells that do not express miR‐122. When miR‐122 is inactivated with a 2′‐O‐ methylated RNA oligonucleotide, the expression of the HCV replicon and core protein are significantly suppressed [57]; thus, miR‐122 represents an attractive target for the development of antiviral treatments.

miRNAs AS THERAPEUTIC STRATEGIES FOR HCC The biological significance of miRNAs in HCC initiation and progression, as well as HBV and HCV viral replication, could provide new opportunities for developing the therapeutic targets for HCC. When considering a therapeutic strategy for HCC, liver cirrhosis offers a better window for therapeutics because complete recovery of cirrhosis may prevent hepatocarcinogenesis. All three members of the miR‐29‐family are significantly downregulated in the cirrhotic liver and are associated with TGF‐β‐mediated fibrosis [69]. Overexpression of miR‐29b in mouse hepatic stellate cells (HSCs) decreases the expression of α‐SMA, collagen I, and TIMP‐1. Overexpression of miR‐29b in activated HSCs suppresses cell viability and colony formation, and causes cell cycle arrest in G1 phase by downregulating cyclin D1 and p21cip1. Therefore, the administration of miR‐29 might prevent hepatic fibrogenesis by targeting activated HSCs in the liver [70]. On the other hand, the HCV replication process is also targeted by miRNA‐based therapeutics. Chronically infected chimpanzees were treated with a locked nucleic acid (LNA)‐modified oligonucleotide (SPC3649) complementary to miR‐122 and exhibited long‐ lasting suppression of HCV viremia, without side‐effects or viral resistance in the chimpanzees. The prolonged effects of SPC3649 are promising results for its use as an antiviral therapy for HCV infection [43]. Regarding miRNA therapeutics for HCC, a number of target miRNAs have been proposed in both experimental and preclinical settings. Basically, two options for the target have been identified: tumor suppressor miRNAs and oncogenic miRNAs. When a target miRNA is a tumor suppressor, its expression level is generally decreased in HCC tissues and increased in normal liver tissues. Thus, the administration of the miRNA into HCC tissues may exert therapeutic effects. In contrast, when a target miRNA is an oncogene, its expression level is increased in HCC

189

and thus the inhibition of the miRNA in HCC represents a potential therapeutic strategy. Kota and colleagues used miR‐26a as the target of HCC therapy and reported the efficacy of an miRNA replacement in HCC. The expression of miR‐26a is normally decreased in HCC, and overexpression of miR‐26a induces cell cycle arrest by directly targeting cyclins D2 and E2. The systemic administration of miR‐26a with an adeno‐associated virus (AAV) inhibits cancer growth and induces tumor‐specific apoptosis in a mouse model of HCC. The results of the delivery of miRNAs provided a new therapeutic strategy for HCC [71]. For the therapeutics focusing on oncogenic miRNAs, one of the targets is miR‐191. It has been reported as a therapeutic candidate miRNA for HCC therapy. Indeed, miR‐191 expression is increased by TCDD (2,3,7,8‐tetrachlorodibenzo‐p‐dioxin), a known liver carcinogen, and it regulates a number of cancer‐related pathways. The administration of anti‐miR‐191 into the orthotopic HCC xenograft inhibits the growth of tumor cells and reduces the tumor mass [72]. Another target is miR‐21, which is overexpressed in HCC and designated as an onco‐miR. Wagenaar and colleagues developed potent and specific single‐stranded oligonucleotide inhibitors of miR‐21 (anti‐miR‐21). Treatment with anti‐miR‐21 significantly decreases cell viability in the majority of HCC cell lines by inducing apoptosis and necrosis. Similar to the in vitro experiments, the effect of anti‐miR‐21 was confirmed in HCC tumor xenograft models [73]. Moreover, oncogenic miRNAs of the miR‐17 family were also targeted for pharmacological inhibition in an HCC model. Notably, miR‐17 itself was blocked by a tough decoy inhibitor in HCC cell lines, leading to a global derepression of direct targets of miR‐17. A lipid nanoparticle represented one of the most advanced platforms for the systemic delivery of an oligonucleotide to tumor tissues in vivo and was used to encapsulate a potent anti‐miR‐17 family oligonucleotide, which was designated as RL01‐17(5). RL01‐17(5) was systemically administered to orthotopic HCC xenografts, and significant tumor suppression was observed. Thus, these findings potentially represent proof‐of‐concept evidence for HCC therapies targeting the miR‐17 family [26]. Therapeutic strategies based on miRNAs or anti‐miRNAs hold great promise due to their abilities to regulate a large number of genes and pathways. Prior to the therapeutic use of miRNAs or anti‐miRNAs in the clinic, target identification and validation should be performed. Technological advances in the in vivo delivery systems, such as nanoparticles, virus vectors, and use of exosomes, will enable efficient and safe miRNA‐ or anti‐miRNA‐based gene therapy in HCC.

EXOSOMAL miRNAs IN HCC In 2007, an exemplary study by Valadi and colleagues reported that miRNAs and mRNAs were packaged in extracellular vesicles (EVs) [74]. These EVs are designated as exosomes, microvesicles, apoptotic bodies, and other extracellular particles, according to their size, density, and secretion mechanism. Secreted EVs contain cytoplasmic components; they can form at the plasma membrane by direct budding into the extracellular environment and produce large (100–1000 nm), irregularly

190

THE LIVER:  EXOSOMAL miRNAs IN HCC

shaped microvesicles. In contrast, another type of particle, exosomes, are thought to be released from multivesicular bodies (MVBs), whose size is 30–100 nm. EVs are secreted from cells and transferred to another cell [74]. They are secreted from almost all types of cell and play a key role in intercellular communication by transporting RNA molecules and proteins (Figure 16.2). Thus, EVs are a novel tool for the exchange of genetic information between cells, such as miRNAs. Notably, EVs have been detected in body fluids, including blood, saliva, urine, and breast milk. Although RNases are present in large amounts in body fluids such as the blood, circulating extracellular RNAs are protected by the bilayer lipid membrane of EVs. Prior to this research, miRNAs were considered to function intracellularly, but they are even active and function in the extracellular space via EVs. Indeed, cancer cell‐derived EVs containing miRNAs are transferred into endothelial cells and regulate angiogenesis [75]. Currently, many studies have proposed that RNAs, including miRNAs, function as novel humoral factors in intercellular communication, and some have focused on the effect of EV‐derived miRNAs in cancer biology [76]. Cancer cells generally express oncogenic miRNAs at higher levels and tumor‐suppressive miRNAs at lower levels, and EVs from the cancer cells basically reflect the miRNA expression profiles of the cells themselves. For example, p53‐mutated colon cancer cells secrete oncogenic miR‐1246 via EVs. EVs carrying miR‐1246 are delivered to the neighboring macrophages for reprogramming into a cancer‐promoting state, creating a favorable microenvironment for the cancer cells [77]. In

contrast, miR‐143, which functions as a tumor suppressor, is expressed at high levels in normal prostate cells, and the EVs from normal prostate cells contain miR‐143. EVs from normal prostate cells are transferred to cancer cells to block cancer cell growth [78]. To date, a large number of studies have reported that extracellular particles are an important source of miRNAs in the circulation and suggest that EV‐derived miRNAs also play key roles in carcinogenesis, including HCC initiation and progression. According to Kogure and colleagues, HCC cells secrete EVs with distinct profiles of both RNAs and proteins compared with their cells of origin. Similar to EVs derived from other cancer cells, the HCC cell‐derived EVs were confirmed to be internalized by other cells and modulate gene expression in recipient cells [79]. Liver‐specific miR‐122, a tumor suppressor miRNA, is also transferred via EVs between Huh7 and HepG2 human hepatoma cells. In cell culture models, EV‐derived miR‐122 from Huh7 cells was internalized by miR‐122‐deficient HepG2 cells and subsequently downregulated miR‐122 target mRNAs, indicating that intercellular communication via EVs occurred between neighboring cells [80]. EVs secreted from HCC cells are also transferred to the surrounding fibroblasts. Cancer‐associated fibroblasts (CAFs) are well‐known to play a critical role in positively regulating the tumor microenvironment. The miRNAs sequencing of CAF‐ derived exosomes from patients with HCC revealed a significant loss of miR‐320a expression from CAF‐derived exosomes [81]. An exogenous miRNA transfection experiment revealed that CAFs transferred miRNAs to HCC cells. Notably,

Fibroblast Conversion to CAFs

Macrophage, etc.

Block cell growth

HCC cells

EVs containing miRNAs

Endothelial cells 1. Increased vascular permeability 2. Angiogenesis

Figure 16.2  Intercellular communication in hepatocellular carcinoma via extracellular vesicles. HCC‐derived EVs modulate the properties of microenvironmental cells such as immune cells, endothelial cells, fibroblasts, epithelial cells, and mesenchymal stem cells. Cancer cells use EVs to build a favorable environment to form tumors. For example, HCC‐derived EVs containing miRNAs control CAF conversion in an endocrine manner [82], increase vascular permeability in a paracrine manner [83], and induce angiogenesis [84]. In contrast, HCC cells also receive EVs from surrounding cells. Macrophage‐derived EVs containing miRNAs block the proliferation of HCC cells. EVs could work as an intercellular communication tool between neighboring cells in an autocrine manner [80].

16:  miRNAs AND HEPATOCELLULAR CARCINOMA

miR‐320a functions as a tumor suppressor that inhibits HCC cell proliferation, migration, and metastasis by directly targeting the PBX3 homeobox gene. Therefore, the transfer of normal stromal cell‐derived miR‐320a via EVs is probably important for the prevention of HCC development; however, CAF‐mediated HCC tumor progression is at least partially caused by the loss of the tumor‐suppressive miR‐320a from exosomes [81]. Since exosomes function locally and systemically, EV‐derived miR‐1247–3p from highly metastatic HCC cells has been described to regulate metastatic niche formation in the lung by converting normal fibroblasts to CAFs [82]. EV‐derived miR‐ 1247–3p suppresses B4GALT3 expression, resulting in the activation of β1‐integrin–NF‐κB signaling in fibroblasts. Interestingly, serum miR‐1247‐3p levels in EVs correlate with lung metastasis in patients with HCC [82]. Furthermore, EV‐derived miR‐103 has been reported to correlate with increased vascular permeability required for metastasis. Deep sequencing and qRT‐PCR validation identified a correlation between a high serum miR‐103 level and a high metastatic capacity in patients with HCC. When endothelial cells were retreated with EVs derived from hepatoma cells expressing miR‐103 at high levels, their permeability significantly increased, facilitating the transendothelial invasion of tumor cells [83]. Cancer cell‐derived miR‐103 transfer decreases the integrity of endothelial junctions by directly targeting VE‐ cadherin (VE‐Cad), p120‐catenin (p120) and zonula occludens 1 [83]. In addition, an evaluation of serum miRNA levels revealed a high level of miR‐210‐3p in exosomes isolated from the sera of patients with HCC, and the miR‐210 level positively correlated with the microvessel density in HCC tissues [84]. By directly targeting the SMAD4 and STAT6 genes, exosomal miR‐210 enhanced angiogenesis, as confirmed by an in vitro tubulogenesis assay using endothelial cells [84]. On the other hand, macrophage‐derived EVs are taken up by HCC cells. Two miRNAs, miR‐142 and miR‐223, are contained in EVs, and transfer of these miRNAs contributes to the inhibition of the proliferation of HCC cells by inhibiting the expression of stathmin‐1 and insulin‐like growth factor‐1 receptor (IGF1R). Thus, the transfer of EVs carrying specific miRNAs from immune cells represents a potentially new defense system protecting against HCC initiation and progression [85].

miRNAs AS DIAGNOSTIC BIOMARKERS IN HCC At the dawn of the use of miRNA profiling for cancer diagnostics, whole transcriptomic analyses were applied to investigate the molecular classification of the HCC tissues by determining the clinical and genetic properties of the tumor [34]. Within the last decade, a number of studies clearly indicated that miRNA profiling provides a specific miRNA expression fingerprint of tissues stratified according to malignancy, risk factors, and oncogene/tumor suppressor gene alterations, displaying the clinical and pathological features of HCC initiation and progression; thus, miRNA profiles could be exploited as potential cancer biomarkers [86]. Because the miRNA signature likely represents the HCC status, an understanding of miRNA profiles

191

would be useful for establishing new diagnostic and/or prognostic tools for HCC. Although the molecular landscapes of tumors were initially established using surgical or biopsy specimens, more recently, an analysis of circulating nucleic acids, known as a liquid biopsy strategy, has received increasing attention as a diagnostic tool for solid tumors. The liquid biopsy has several advantages; for example, it is less invasive and can be applied to any methods, such as the detection of cell‐free DNA (cfDNA) and miRNAs. A single biopsy provides spatially and temporally limited information about the tumors; however, the liquid biopsy provides the average genetic information and may be able to reflect intratumor heterogeneity [87, 88]. Because the detection of tumor‐associated genomic DNA mutations is useful for diagnostic decisions regarding treatment strategies, such as the selection of molecular target drugs, the methods for detecting cfDNA were established for preclinical use (e.g., CancerSEEK, which distinguishes eight cancer types, even at early stages) [89]. In addition to cfDNA, the liquid biopsy approach is highly effective at detecting miRNAs, proteins, and lipids packaged in EVs in the body fluids. An early analysis of miRNAs detected in blood samples from patients with HCC revealed that oncofetal miR‐500 is present at high levels [90]. Therefore, the detection and/or quantification of serum miRNA levels using liquid biopsy represents a potentially feasible application for the diagnosis of HCC, assessments of prognosis, surveillance of recurrent tumors, and prediction of drug responses. Deep sequencing followed by qRT‐PCR validation was applied to identify serum biomarkers for HCC and identified three miRNAs – miR‐25, miR‐375, and let‐7f – as biomarkers for HCC. The receiver operating characteristic (ROC) curve analysis of these miRNAs yielded an area under the ROC curve (AUC) of 0.997, with a 97.9% sensitivity and 99.1% specificity in distinguishing HCC cases from controls [91]. A logistic regression model was established with seven miRNAs (miR‐122, miR‐192, miR‐21, miR‐223, miR‐26a, miR‐27a, and miR‐801) by randomly separating samples into a training cohort and a validation cohort to discover a plasma miRNA combination that discriminated HBV‐related HCC cases. The miRNA combination discriminated the HCC cases from healthy individuals (AUC: 0.941), patients with chronic hepatitis B (AUC: 0.842), and patients with cirrhosis (AUC: 0.884) [92]. Significantly higher plasma levels of a single miRNA, miR‐21, were observed in patients with HCC than in patients with chronic hepatitis and healthy individuals. Interestingly, the ROC analysis of plasma miR‐21 levels displayed an AUC of 0.773, with a 61.1% sensitivity and 83.3% specificity in distinguishing HCC from chronic hepatitis, and an AUC of 0.953 with an 87.3% sensitivity and 92.0% specificity in distinguishing HCC from healthy individuals [93]. These results are better than a conventional plasma HCC biomarker, AFP, suggesting that miR‐21 is a promising miRNA biomarker [93]. In 2015, Lin and colleagues reported a comprehensive study of the serum miRNA profile for HCC diagnosis in five groups: patients with HCC, patients with HBV‐related cirrhosis, patients with chronic hepatitis (HBV), inactive HBsAg carriers, and healthy individuals. Using six serum samples from patients with HCC and eight serum samples from patients with chronic hepatitis (control), 19 highly expressed miRNAs were identified

192

THE LIVER:  REFERENCES

with the miRNA microarray. Using a combination of four models, such as a linear support vector machine, nonlinear support vector machine, linear discriminant analysis, and logistical regression, the authors developed an miRNA classifier (seven miRNAs: miR‐29a, miR‐29c, miR‐133a, miR‐143, miR‐145, miR‐192, and miR‐505) for detecting HCC [94]. This miRNA classifier showed higher accuracy than the conventional AFP method and detected small‐sized, early‐stage, and α‐fetoprotein‐negative HCC cases in at‐risk patients. Moreover, the circulating miRNA signature was applied to estimate the risk of HCC in patients with cirrhosis [95]. By comparing serum miRNA profiles between 330 cirrhotic liver samples and 42 early‐stage HCC samples, the calculated score based on the expression levels of five miRNAs in serum predicted patients with cirrhosis who were at high and low risk of developing HCC. The follow‐ up study (median period: 752 days) showed a good prediction accuracy (AUC: 0.725, P < 0.001) [95]. As described above, miRNA levels in blood samples are valuable and useful for the preclinical detection of HCC, providing patients with an opportunity to undergo curative resection and achieve a better prognosis [94]. Recently, diagnostic models based on serum miRNA levels have reported the successful discrimination of specific cancer types from other types of cancers [96] and the prediction of cancer metastasis [97], as well as the differentiation of malignancy from benign tissues. Thus, the liquid biopsy approach is a very promising method and a technology that will likely be applied from the bench to the bedside in the near future.

SUMMARY AND PERSPECTIVES This chapter focused on the functions and potential clinical applications, such as therapeutics and diagnostics, of miRNAs in liver cancer. A brief overview of researchers’ progress in improving our understanding of the function of miRNAs in hepatocarcinogenesis has also been provided. The research field of miRNA biology in HCC has developed quickly and broadly from basic studies to clinical studies. The importance of miRNAs in basic biology for the promotion and inhibition of HCC development is based on the unique and fascinating features of miRNAs, which influence thousands of genes and regulate a variety of pathways. For instance, a liver‐specific miRNA, miR‐122, constitutes approximately 70% of all the miRNA molecules in the liver. As expected, miR‐122 is important for liver development, and surprisingly is essential for the replication of HCV. In contrast, it functions as a tumor suppressor during HCC development. Thus, one miRNA spatially and temporally exerts multiple functions in different circumstances. Due to their abilities to regulate complex gene networks, miRNAs represent potential therapeutic targets in HCC, as evidenced by findings from studies overexpressing miR‐26 and inhibiting miR‐17 [71, 73]. The recent and most critical breakthrough in the miRNA research field is the discovery of extracellular RNAs, such as EV‐packaged miRNAs. The detection of circulating miRNAs in the blood is performed using a microarray or deep sequencing, resulting in the diagnosis of HCC. As shown, studies with a large cohort have identified HCC

biomarkers. Nevertheless, the published data are still inconsistent. Several explanations for this discrepancy are likely, such as the use of different detection platforms, normalization methods, sample preparation methods, and sample types (serum or plasma), as well as variability in the cohort size. Therefore, standardization is necessary to obtain a reliable liquid biopsy test for HCC. Furthermore, EVs have the potential to serve as a natural vector; thus, they could be applied to deliver specific miRNA molecules, proteins, or drugs to the disease site. In the next decade, we will look ahead to the future of miRNA research towards clinical applications of circulating miRNAs as diagnostic, prognostic, and therapeutic tools.

ACKNOWLEDGMENTS This work was supported in part by the Japan Agency for Medical Research and Development (AMED): P‐CREATE; 17cm0106217h0002 and Development and New Energy and  Industrial Technology Development Organization; 16ae0101011h0003, a research grant from the Uehara Memorial Foundation and a research grant from the Naito Foundation.

REFERENCES 1. Groszhans, H. and Filipowicz, W. Molecular biology: the expanding world of small RNAs. Nature, 2008;451(7177):414. 2. Aravin, A.A., Hannon, G.J., and Brennecke, J. The piwi‐piRNA pathway provides an adaptive defense in the transposon arms race. Science, 2007;318(5851):761–64. 3. Bartel, D.P. MicroRNAs: target recognition and regulatory functions. Cell, 2009;136:215–33. 4. Lim, L.P., Lau, N.C., Garrett‐Engele, P. et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature, 2005;433(7027):769. 5. Chang, J., Nicolas, E., Marks, D. et al. miR‐122, a mammalian liver‐specific microRNA, is processed from hcr mRNA and may downregulate the high affinity cationic amino acid transporter CAT‐1. RNA Biol, 2004;1(2):106–13. 6. Schratt, G.M., Tuebing, F., Nigh, E.A. et  al. A brain‐specific microRNA regulates dendritic spine development. Nature, 2006;439(7074):283. 7. Chen, C.Z., Li, L., Lodish, H.F., and Bartel, D.P. MicroRNAs modulate hematopoietic lineage differentiation. Science, 2004;303(5654):83–86. 8. Lu, J., Getz, G., Miska, E.A. et al. MicroRNA expression profiles classify human cancers. Nature, 2005;435:834–38. 9. Voorhoeve, P.M. A genetic screen implicates miRNA‐372 and miRNA‐373 as oncogenes in testicular germ cell tumors. Cell, 2006;124:1169. 10. Mayr, C., Hemann, M.T., and Bartel, D.P. Disrupting the pairing between let‐7 and Hmga2 enhances oncogenic transformation. Science, 2007; 315:1576. 11. Farh, K.K. Grimson, A., Jan, C. et al. The widespread impact of mammalian microRNAs on mRNA repression and evolution. Science, 2005;310(5755): 1817–21. 12. Lee, Y., Jeon, K., Lee, J.T., Kim, S., and Kim, V.N. MicroRNA maturation: stepwise processing and subcellular localization. EMBO J, 2002;21: 4663–70. 13. Truscott, M., Islam, A.B., and Frolov, M.V. Novel regulation and functional interaction of polycistronic miRNAs. RNA, 2016;22:129–38. 14. Pasquinelli, A.E., Reinhart, B.J., Slack, F. et al. Conservation of the sequence and temporal expression of let‐7 heterochronic regulatory RNA. Nature, 2000;408(6808):86. 15. Ruby, J.G., Jan, C.H., and Bartel, D.P. Intronic microRNA precursors that bypass Drosha processing. Nature, 2007;448(7149):83.

16:  miRNAs AND HEPATOCELLULAR CARCINOMA 16. Gregory, R.I., Yan, K.P., Amuthan, G. et  al. The microprocessor complex mediates the genesis of microRNAs. Nature, 2004;432(7014):235–40. 17. Lee, Y., Ahn, C., Han, J. et al. The nuclear RNaseIII Drosha initiates microRNA processing. Nature, 2003;425:415–19. 18. Meister, G. and Tuschl, T. Mechanisms of gene silencing by double‐stranded RNA. Nature, 2004;431:343. 19. Chen, P.Y. and Meister, G. microRNA‐guided posttranscriptional gene regulation. Biol Chem, 2005;386(12):1205–18. 20. Song, J.J., Smith, S.K., Hannon, G.J., and Joshua‐Tor, L. Crystal structure of argonaute and its implications for RISC slicer activity. Science, 2004;305: 1434. 21. Siegel, R.L., Miller, K.D., and Jemal, A. Cancer statistics 2017. CA Cancer J Clin, 2017;67:7–30. 22. Forner, A., Llovet, J.M., and Bruix, J. Hepatocellular carcinoma. The Lancet, 2012;379:1245–55. 23. Oka, H., Tamori, A., Kuroki, T., Kobayashi, K., and Yamamoto, S. Prospective study of alpha‐fetoprotein in cirrhotic patients monitored for development of hepatocellular carcinoma. Hepatology, 1994;19(1):61–6. 24. Seo, S.I., Kim, H.S., Kim, W.J. et  al. Diagnostic value of PIVKA‐II and alpha‐fetoprotein in hepatitis B virus‐associated hepatocellular carcinoma. World J Gastroenterol, 2015;21:3928–35. 25. Meng, F., Henson, R., Wehbe‐Janek, H. et  al. MicroRNA‐21 regulates expression of the PTEN tumor suppressor gene in human hepatocellular cancer. Gastroenterology, 2007;133:647–58. 26. Wagenaar, T.R., Zabludoff, S., Ahn, S.M. et  al. Anti‐miR‐21 suppresses hepatocellular carcinoma growth via broad transcriptional network deregulation. Mol Cancer Res, 2015;13:1009–21. 27. Xu, G, Zhang, Y., Wei, J. et al. MicroRNA‐21 promotes hepatocellular carcinoma HepG2 cell proliferation through repression of mitogen‐activated protein kinase‐kinase 3. BMC Cancer, 2013;13:469. 28. Iliopoulos, D., Jaeger, S.A., Hirsch, H.A. et al. STAT3 activation of miR‐21 and miR‐181b‐1, via PTEN and CYLD, are part of the epigenetic switch linking inflammation to cancer. Mol Cell, 2010;39:493–506. 29. Connolly, E., Melegari, M., Landgraf, P. et  al. Elevated expression of the miR‐17–92 polycistron and miR‐21 in hepadnavirus‐associated hepatocellular carcinoma contributes to the malignant phenotype. Am J Pathol, 2008;173:856–64. 30. Hayashita, Y., Osada, H., Tatematsu, Y. et  al. A polycistronic microRNA cluster, miR‐17–92, is overexpressed in human lung cancers and enhances cell proliferation. Cancer Res, 2005;65:9628–32. 31. Zhu, H., Han, C., and Wu, T. MiR‐17‐92 cluster promotes hepatocarcinogenesis. Carcinogenesis, 2015;36:1213–22. 32. Wong, Q.W., Lung, R.W., Law, P.T. et  al. MicroRNA‐223 is commonly repressed in hepatocellular carcinoma and potentiates expression of Stathmin1. Gastroenterology, 2008;135:257–69. 33. Varnholt, H., Drebber, U., Schulze, F. et al. MicroRNA gene expression profile of hepatitis C virus‐associated hepatocellular carcinoma. Hepatology, 2008;47:1223–32. 34. Toffanin, S., Hoshida, Y., Lachenmayer, A. et al. MicroRNA‐based classification of hepatocellular carcinoma and oncogenic role of miR‐517a. Gastroenterology, 2011;140:1618–28. 35. Ma, S., Tang, K.H., Chan, Y.P. et  al. miR‐130b promotes CD133+ liver tumor‐initiating cell growth and self‐renewal via tumor protein 53‐induced nuclear protein 1. Cell Stem Cell, 2010;7: 694–707. 36. Yang, F., Yin, Y., Wang, F. et al. miR‐17‐5p promotes migration of human hepatocellular carcinoma cells through the p38 mitogen‐activated protein kinase‐heat shock protein 27 pathway. Hepatology, 2010;51:1614–23. 37. Ding, J., Huang, S., Wu, S. et al. Gain of miR‐151 on chromosome 8q24.3 facilitates tumour cell migration and spreading through downregulating RhoGDIA. Nat Cell Biol, 2010;12:390–9. 38. Zhang, X., Liu, S., Hu, T. et al. Up‐regulated microRNA‐143 transcribed by nuclear factor kappa B enhances hepatocarcinoma metastasis by repressing fibronectin expression. Hepatology, 2009;50:490–9. 39. Wong, Q.W., Ching, A.K., Chan, A.W. et al. MiR‐222 overexpression confers cell migratory advantages in hepatocellular carcinoma through enhancing AKT signaling. Clin Cancer Res, 2010;16:867–75. 40. Li, H., Zhou, Z.Q., Yang, Z.R. et  al. MicroRNA‐191 acts as a tumor promoter by modulating the TET1‐p53 pathway in intrahepatic cholangiocarcinoma. Hepatology, 2017;66:136–51. 41. Tsai, W.C., Hsu, P.W., Lai, T.C. et al. MicroRNA‐122, a tumor suppressor microRNA that regulates intrahepatic metastasis of hepatocellular carcinoma. Hepatology, 2009;49:1571–82.

193

42. Gramantieri, L., Ferracin, M., Fornari, F. et al. Cyclin G1 is a target of miR‐ 122a, a microRNA frequently down‐regulated in human hepatocellular carcinoma. Cancer Res, 2007;67:6092–9. 43. Lanford, R.E., Hildebrandt‐Eriksen, E.S., Petri, A. et al. Therapeutic silencing of microRNA‐122 in primates with chronic hepatitis C virus infection. Science, 2010;327:198–201. 44. Xu, H., He, J.H., Xiao, Z.D. et al. Liver‐enriched transcription factors regulate microRNA‐122 that targets CUTL1 during liver development. Hepatology, 2010;52:1431–42. 45. Herath, N.I., Leggett, B.A., and MacDonald, G.A. Review of genetic and epigenetic alterations in hepatocarcinogenesis. J Gastroenterol Hepatol, 2006;21:15–21. 46. Datta, J., Kutay, H., Nasser, M.W. et al. Methylation mediated silencing of MicroRNA‐1 gene and its role in hepatocellular carcinogenesis. Cancer Res, 2008;68:5049–58. 47. Gailhouste, L., Gomez‐Santos, L., Hagiwara, K. et al. miR‐148a plays a pivotal role in the liver by promoting the hepatospecific phenotype and suppressing the invasiveness of transformed cells. Hepatology, 2013;58:1153–65. 48. Liu, A.M., Poon, R.T., and Luk, J.M. MicroRNA‐375 targets Hippo‐signaling effector YAP in liver cancer and inhibits tumor properties. Biochem Biophys Res Commun, 2010;394:623–7. 49. Dang, Y., Luo, D., Rong, M. et al. Underexpression of miR‐34a in hepatocellular carcinoma and its contribution towards enhancement of proliferating inhibitory effects of agents targeting c‐MET. PLoS One, 2013;8:e61054. 50. Kim, T., Veronese, A., Pichiorri, F. et al. p53 regulates epithelial–mesenchymal transition through microRNAs targeting ZEB1 and ZEB2. J Exp Med, 2011;208:875–83. 51. Li, N., Fu, H., Tie, Y. et  al. miR‐34a inhibits migration and invasion by down‐regulation of c‐Met expression in human hepatocellular carcinoma cells. Cancer Lett, 2009;275:44–53. 52. Salvi, A., Sabelli, C., Moncini, S. et al. MicroRNA‐23b mediates urokinase and c‐met downmodulation and a decreased migration of human hepatocellular carcinoma cells. FEBS J, 2009;276:2966–82. 53. Bai, S., Nasser, M.W., Wang, B. et al. MicroRNA‐122 inhibits tumorigenic properties of hepatocellular carcinoma cells and sensitizes these cells to sorafenib. J Biol Chem, 2009;284:32015–27. 54. Su, J., Wang, Q., Liu, Y. et al. miR‐217 inhibits invasion of hepatocellular carcinoma cells through direct suppression of E2F3. Mol Cell Biochem, 2014;392:289–96. 55. Luo, C., Yin, D., Zhan, H. et al. MicroRNA‐501‐3p suppresses metastasis and progression of hepatocellular carcinoma through targeting LIN7A. Cell Death Dis, 2018;9:535. 56. Jin, W.B., Wu, F.L., Kong, D. et al. HBV‐encoded microRNA candidate and its target. Comput Biol Chem, 2007;31:124–6. 57. Jopling, C.L., Yi, M., Lancaster, A.M. et al. Modulation of hepatitis C virus RNA abundance by a liver‐specific microRNA. Science, 2005;309:1577–81. 58. Wang, S., Qiu, L., Yan, X. et  al. Loss of microRNA 122 expression in patients with hepatitis B enhances hepatitis B virus replication through cyclin G(1)‐modulated P53 activity. Hepatology, 2012;55:730–41. 59. Song, K., Han, C., Zhang, J. et al. Epigenetic regulation of MicroRNA‐122 by peroxisome proliferator activated receptor‐gamma and hepatitis b virus X protein in hepatocellular carcinoma cells. Hepatology, 2013;58:1681–92. 60. Zhang, X., Zhang, E., Ma, Z. et al. Modulation of hepatitis B virus replication and hepatocyte differentiation by microRNA‐1. Hepatology, 2011;53:1476–85. 61. Naito, Y., Hamada‐Tsutsumi, S., Yamamoto, Y. et al. Screening of microRNAs for a repressor of hepatitis B virus replication. Oncotarget, 2018;9:29857–68. 62. Wang, Y., Lu, Y., Toh, S.T. et al. Lethal‐7 is down‐regulated by the hepatitis B virus x protein and targets signal transducer and activator of transcription 3. J Hepatol, 2010;53:57–66. 63. Jin, J., Tang, S., Xia, L. et al. MicroRNA‐501 promotes HBV replication by targeting HBXIP. Biochem Biophys Res Commun. 2013;430:1228–33. 64. Su, C., Hou, Z., Zhang, C. et  al. Ectopic expression of microRNA‐155 enhances innate antiviral immunity against HBV infection in human hepatoma cells. Virol J, 2011;8:354. 65. Sun, C., Lan, P., Han, Q. et al. Oncofetal gene SALL4 reactivation by hepatitis B virus counteracts miR‐200c in PD‐L1‐induced T cell exhaustion. Nat Commun, 2018;9:1241. 66. Ganem, D. and Prince, A.M. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med, 2004;350:1118–29.

194

THE LIVER:  REFERENCES

67. Huang, J., Wang, Y., Guo, Y. et al. Down‐regulated microRNA‐152 induces aberrant DNA methylation in hepatitis B virus‐related hepatocellular carcinoma by targeting DNA methyltransferase 1. Hepatology, 2010;52:60–70. 68. Potenza, N., Papa, U., Mosca, N. et al. Human microRNA hsa‐miR‐125a‐5p interferes with expression of hepatitis B virus surface antigen. Nucleic Acids Res, 2011;39:5157–63. 69. Roderburg, C., Urban, G.W., Bettermann, K. et  al. Micro‐RNA profiling reveals a role for miR‐29 in human and murine liver fibrosis. Hepatology, 2011;53:209–18. 70. Wang, J., Chu, E.S., Chen, H.Y. et al. MicroRNA‐29b prevents liver fibrosis by attenuating hepatic stellate cell activation and inducing apoptosis through targeting PI3K/AKT pathway. Oncotarget, 2015;6:7325–38. 71. Kota, J., Chivukula, R.R., O’Donnell, K.A. et  al. Therapeutic microRNA delivery suppresses tumorigenesis in a murine liver cancer model. Cell, 2009;137:1005–17. 72. Elyakim, E., Sitbon, E., Faerman, A. et al. hsa‐miR‐191 is a candidate oncogene target for hepatocellular carcinoma therapy. Cancer Res, 2010;70: 8077–87. 73. Huang, X., Magnus, J., Kaimal, V. et al. Lipid nanoparticle‐mediated delivery of anti‐miR‐17 family oligonucleotide suppresses hepatocellular carcinoma growth. Mol Cancer Ther, 2017;16:905–13. 74. Valadi, H., Ekström, K., Bossios, A. et  al. Exosome‐mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol, 2007;9:654–9. 75. Kosaka, N., Iguchi, H., Hagiwara, K. et  al. Neutral sphingomyelinase 2 (nSMase2)‐dependent exosomal transfer of angiogenic microRNAs regulate cancer cell metastasis. J Biol Chem, 2013;288:10849–59. 76. Kosaka, N., Yoshioka, Y., Fujita, Y. et al. Versatile roles of extracellular vesicles in cancer. J Clin Invest, 2016;126:1163–72. 77. Cooks, T., Pateras, I.S., Jenkins, L.M. et al. Mutant p53 cancers reprogram macrophages to tumor supporting macrophages via exosomal miR‐1246. Nat Commun, 2018;9:771. 78. Kosaka, N., Iguchi, H., Yoshioka, Y. et al. Competitive interactions of cancer cells and normal cells via secretory microRNAs. J Biol Chem, 2012;287:1397–405. 79. Kogure, T., Lin, W.L., Yan, I.K. et  al. Intercellular nanovesicle‐mediated microRNA transfer: a mechanism of environmental modulation of hepatocellular cancer cell growth. Hepatology, 2011;54:1237–48. 80. Basu, S. and Bhattacharyya, S.N. Insulin‐like growth factor‐1 prevents miR‐122 production in neighbouring cells to curtail its intercellular transfer to ensure proliferation of human hepatoma cells. Nucleic Acids Res, 2014;42:7170–85. 81. Zhang, Z., Li, X., Sun, W. et al. (2017) Loss of exosomal miR‐320a from cancer‐associated fibroblasts contributes to HCC proliferation and metastasis. Cancer Lett, 2017;397:33–42.

82. Fang, T., Lv, H., Lv, G. et al. Tumor‐derived exosomal miR‐1247–3p induces cancer‐associated fibroblast activation to foster lung metastasis of liver cancer. Nat Commun, 2018;9:191. 83. Fang, J.H., Zhang, Z.J., Shang, L.R. et al. Hepatoma cell‐secreted exosomal microRNA‐103 increases vascular permeability and promotes metastasis by targeting junction proteins. Hepatology, 2018;68:1459–75. 84. Lin, X.J., Fang, J.H., Yang, X.J. et al. Hepatocellular carcinoma cell‐secreted exosomal microRNA‐210 promotes angiogenesis in vitro and in vivo. Mol Ther Nucleic Acids, 2018;11:243–52. 85. Aucher, A., Rudnicka, D., and Davis, D.M. MicroRNAs transfer from human macrophages to hepato‐carcinoma cells and inhibit proliferation. J Immunol, 2013;191:6250–60. 86. Ladeiro, Y., Couchy, G., Balabaud, C. et al. MicroRNA profiling in hepatocellular tumors is associated with clinical features and oncogene/tumor suppressor gene mutations. Hepatology, 2008;47:1955–63. 87. Crowley, E., Di Nicolantonio, F., Loupakis, F., and Bardelli, A. Liquid biopsy: monitoring cancer‐genetics in the blood. Nat Rev Clin Oncol, 2013;10:472–84. 88. Siravegna, G., Marsoni, S., Siena, S. et al. Integrating liquid biopsies into the management of cancer. Nat Rev Clin Oncol, 2017;14:531–48. 89. Cohen, J.D., Li, L., Wang, Y. et al. Detection and localization of surgically resectable cancers with a multi‐analyte blood test. Science, 2018;359: 926–30. 90. Yamamoto, Y., Kosaka, N., Tanaka, M. et  al. (2009) MicroRNA‐500 as a potential diagnostic marker for hepatocellular carcinoma. Biomarkers, 2009;14:529–38. 91. Li, L.M., Hu, Z.B., Zhou, Z.X. et  al. Serum microRNA profiles serve as novel biomarkers for HBV infection and diagnosis of HBV‐positive hepatocarcinoma. Cancer Res, 2010;70:9798–807. 92. Zhou, J., Yu, L., Gao, X. et al. Plasma microRNA panel to diagnose hepatitis B virus‐related hepatocellular carcinoma. J Clin Oncol, 2011;29:4781–8. 93. Tomimaru, Y., Eguchi, H., Nagano, H. et al. Circulating microRNA‐21 as a novel biomarker for hepatocellular carcinoma. J Hepatol, 2012;56:167–75. 94. Lin, X.J., Chong, Y., Guo, Z.W. et al. A serum microRNA classifier for early detection of hepatocellular carcinoma: a multicentre, retrospective, longitudinal biomarker identification study with a nested case‐control study. Lancet Oncol, 2015;16:804–15. 95. Huang, Y.H., Liang, K.H., Chien, R.N. et al. A circulating microRNA signature capable of assessing the risk of hepatocellular carcinoma in cirrhotic patients. Sci Rep, 2017;7:523. 96. Yokoi, A., Matsuzaki, J., Yamamoto, Y. et al. Integrated extracellular microRNA profiling for ovarian cancer screening. Nat Commun, 2018;9:4319. 97. Shiino, S., Matsuzaki, J., Shimomura, A. et al. Serum miRNA‐based prediction of axillary lymph node metastasis in breast cancer. Clin Cancer Res, 2019;25(6):1817–27.

17

Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases Harmeet Malhi and Gregory J. Gores College of Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA

INTRODUCTION Apoptosis is a ubiquitous form of cell death occurring in human liver diseases (Figure  17.1). It has historically been defined morphologically by the presence of cytoplasmic shrinkage (pyknosis), chromatin condensation, nuclear fragmentation (karyorhexis), the presence of plasma membrane blebbing, and the maintenance of an intact plasma membrane that retains its integrity as the cell fragments into apoptotic bodies. Indeed, apoptotic bodies were first described in the liver in patients with yellow fever where they were referred to as Councilman bodies. A more current biochemical description of apoptosis is caspase‐ dependent cell death [1]. Caspases are cysteine proteases that cleave proteins at sites next to aspartic acid residues. The above described apoptosis morphology is due to the cellular effects of caspase activation. Caspase activation, and hence apoptosis, is a highly regulated form of cell death, with multiple checkpoints and molecular mediators, activated via two distinct pathways, the extrinsic pathway and the intrinsic pathway. The extrinsic pathway is initiated via death receptor activation and the intrinsic pathway by intracellular perturbations that result in caspase activation (Figure  17.2). In hepatocytes, both pathways converge on mitochondria. Multiple intracellular molecules transmit apoptotic signals and regulate the apoptotic signaling cascades, upstream and downstream of mitochondria [2]. Mitochondrial permeabilization is not only requisite but also sufficient for apoptosis; therefore, intracellular regulators downstream of mitochondrial permeabilization such as caspase inhibitors cannot prevent cell death [3]. Unlike developmental apoptosis, which is carefully regulated in a spatiotemporal pattern and does not invoke secondary events, pathologic apoptosis activates secondary

signaling events and can be massive, which can result in organ failure. Secondary signaling events include pathologic apoptosis‐ induced tissue inflammation, injury, and fibrosis. For example, in acute liver injury apoptosis is massive and correlates with outcome (i.e. liver transplantation or death) [4]. In chronic liver injury apoptosis is continuous, modulates the inflammatory response and promotes fibrogenesis, resulting in cirrhosis [5, 6]. Hepatocyte apoptosis is evident in liver injury related to viral hepatitis, metabolic diseases, alcoholic steatohepatitis, autoimmune hepatitis, and drug‐induced liver injury [7–9], emphasizing the shared pathogenic role of hepatocyte apoptosis in liver injury from multiple, varied, acute, and chronic insults. Apoptosis of other cellular compartments, such as sinusoidal endothelial cells and stellate cells, also plays a role in liver injury. Here we discuss the signaling mediators and regulators of hepatocyte apoptosis and the inclusion of injury stimulus‐ specific information within each mechanism.

THE EXTRINSIC PATHWAY Death receptors are cell surface transmembrane proteins that belong to the tumor necrosis factor/nerve growth factor (TNF/ NGF) receptor superfamily, and are defined on the basis of ligand specificity (i.e. their affinity for tumor necrosis factor alpha (TNF‐α), Fas ligand (FasL), or tumor necrosis factor‐ related apoptosis‐inducing ligand (TRAIL) [10]. The extracellular N‐terminal domain binds their respective ligands; there is a membrane‐spanning region and then the intracellular C‐terminal domain, which contains a conserved sequence known as the death domain (DD). The ligand‐bound trimerized receptor

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

196

THE LIVER:  THE EXTRINSIC PATHWAY

(a)

(b)

mitochondrial amplification categorizes hepatocytes as type II cells, in contrast to type I cells in which caspase 8 or 10 can directly activate caspase 3 and 7 without mitochondrial involvement [12]. Caspase 8 proteolytically cleaves the proapoptotic BH3‐only protein of the Bcl‐2 family Bid to tBid (truncated Bid), which leads to activation of Bax and Bak (proapoptotic multidomain members of the Bcl‐2 family), and pore formation in the outer mitochondrial membrane [13]. Multiple levels of signal transduction and amplification present opportunities for regulation of death receptor‐mediated apoptosis at many levels. Availability of cell surface receptor and ligand is one level, for example the hepatocyte growth factor (HGF) receptor Met associates with and regulates the availability of Fas for binding its ligand [14]. Moreover, the basal expression of death receptors in hepatocytes is low, and known to increase in both acute and chronic liver diseases. Cellular caspase 8 (FLICE)‐like inhibitory protein (cFLIP) can inhibit cytotoxic signaling by death receptors [15]. cFLIP is an enzymatically inactive homolog of caspase 8 with conserved structural homology in the DED that allows binding to FADD. This binding precludes maximal cellular activation of caspase 8. Pro‐ and antiapoptotic members of the Bcl‐2 family regulate the extrinsic pathway by modulating the ability of tBid to activate Bax and Bak (discussed later) [15].

Tumor necrosis factor‐α

Figure 17.1  Hepatocyte apoptosis in non‐alcoholic steatohepatitis. (a) Photomicrograph of a hematoxylin and eosin‐stained liver section from a dietary mouse model of obesity‐associated non‐alcoholic steatohepatitis, demonstrating the presence of hepatocyte apoptosis. The black arrow points to an apoptotic hepatocyte. (b) The TUNEL stain detects DNA strand breaks, a feature of cell death, in a corresponding liver section from this mouse model of non‐alcoholic steatohepatitis.

complex brings together the DD, allowing recruitment of other adaptor proteins to form a death‐inducing signaling complex (DISC). For death signaling, Fas‐associated protein with death domain (FADD) must be recruited to the DISC complex [10]. FADD contains a death effector domain (DED) through which it binds inactive initiator caspases 8 and 10, in their procaspase form. The procaspases form homodimers and undergo autoproteolytic cleavage with formation of active caspase 8 or 10 [11]. This process of activation of initiator caspases is referred to as the induced‐proximity model of caspase activation. Activation of initiator caspase 8 is essential for the transmission of apoptotic signals from death receptors to mitochondria. This occurs via the Bcl‐2 family protein Bid and is discussed below. In hepatocytes, mitochondrial permeabilization with amplification of the apoptotic cascade occurs in death receptor‐initiated apoptosis. This involves release of mitochondrial mediators of apoptosis which leads to the eventual activation of the effector caspases 3 and 7, with positive feedback amplification of caspase 8 activation. The requirement of

TNF‐α is a circulating cytokine, primarily produced by cells of the immune system, including Kupffer cells in the liver, although it can also be produced by other cell types, such as hepatocytes. Hepatocytes express both tumor necrosis factor receptor 1 (TNFR1), a 55 kDa protein, and tumor necrosis factor receptor 2 (TNFR2), a 75 kDa protein, but their functional significances differ [16]. TNFR1 is thought to mediate most of the biologic effects of TNF‐α; it expresses a cytoplasmic death domain (DD) and executes the apoptotic program by interacting with adaptor proteins [17] (Figure 17.3). Ligand‐activated TNFR1 generates sequential signaling events referred to as complex I and complex II. On binding TNF‐α, TNFR1 recruits the adaptor protein tumor necrosis factor receptor‐associated death domain (TRADD). Signaling then proceeds in two steps. The first step, or complex I, involves recruitment of tumor necrosis factor receptor‐associated protein (TRAF‐2) and receptor‐interacting protein 1 (RIP1), leading to rapid activation of nuclear factor κB (NFκB) [18]. NFκB transcriptionally activates expression of prosurvival (e.g. Bcl‐xL, A1, XIAP, and cFLIP) and proinflammatory genes (e.g. interleukin 6). Complex I also activates the c‐Jun N‐terminal kinase (JNK) pathways. Complex I is internalized with dissociation of TRAF2, RIP1, and TRADD from the ligated receptor. TRADD then recruits FADD and procaspase 8 to initiate apoptotic signaling; this signaling pathway is referred to as complex II. TRADD does not interact with TNFR2, nor does FADD directly interact with TNFR1. Therefore, TNF‐α/ TNFR1 signaling first leads to NFκB‐mediated transcriptional activation of prosurvival and proinflammatory genes prior to the activation of apoptosis signaling. In cells resistant to NFκB, or in the presence of a transcriptional inhibitor such as actinomycin D, which prevents the synthesis of prosurvival proteins, the apoptotic effect of TNF‐α is unmasked.



17:  Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases

197

Figure 17.2  The extrinsic and intrinsic pathways of hepatocyte apoptosis. Mitochondrial permeabilization is required for hepatocyte apoptosis. The extrinsic pathway is mediated by death receptors. Fas or TRAIL, upon ligation with their cognate receptors, activate events leading to mitochondrial permeabilization. The death‐inducing signaling complex is formed on the intracellular domain of ligated homotrimerized receptors in conjunction with adaptor proteins, leading to caspase 8 activation, Bid cleavage, and activation of Bax and Bak. TNF‐α signaling pathway can promote apoptosis by Bid‐induced lysosomal permeabilization. Intracellular perturbations such as ER stress, lysosomal permeabilization, or JNK activate the intrinsic pathway of cell death. ER stress‐induced apoptosis is partly mediated by the transcription factor CHOP, which can upregulate TRAIL‐R2 or Bim expression. JNK activation can be induced by TNF‐α, ER stress, or reactive oxygen species. These pathways are regulated by the proapoptotic and antiapoptotic proteins of the Bcl‐2 family.

Necroptosis is a form of kinase‐regulated, caspase‐independent cell death best described following activation of TNFR1 by TNF‐α, but characterized by morphology similar to necrosis – hence the name [19]. Necroptosis is known to occur only under conditions of caspase 8 absence or inhibition and is mediated by the receptor‐interacting protein (RIP) family kinases, specifically RIP1 and RIP3. RIP1 has a DD and a caspase recruitment domain which allows interactions at the DISC with the adaptor protein TRADD. That is why RIP1 is present in complex I. When caspase 8 is absent or inhibited, RIP1 is deubiquitinated and can interact with RIP3 to form the necrosome, which also recruits mixed‐lineage kinase domain‐like pseudokinase (MLKL), which is phosphorylated by RIP3 and translocates to the plasma membrane where its trimerized form mediates calcium influx and necroptosis [20]. When RIP1 is polyubiquinated it promotes cell survival and inflammation. The pathophysiologic contribution of necroptosis to the liver is controversial due to the lack of RIP3 in normal hepatocytes. Liver nonparenchymal cells and infiltrating immune cells

express RIP3, and this may account for some of the incongruous observations in mouse models. However, RIP1 has many roles, including kinase activity‐dependent and scaffolding function‐ dependent roles that can promote apoptosis or inflammation. TNF‐α has pleiotropic effects in vivo, including hepatocyte proliferation, liver inflammation, and modulation of hepatocyte apoptosis. In a murine model of TNF‐α‐induced liver injury (TNF‐α + d‐galactosamine), liver injury is Bax‐dependent [21]. TNF‐α‐associated caspase 8 activation can also cause lysosomal permeabilization with release of intralysosomal cathepsin B into the cytosol, which causes mitochondrial dysfunction [22]. Mice deficient in cathepsin B are protected from the injurious effects of TNF‐α [23]. c‐Jun N‐terminal kinase (JNK), a stress‐activated kinase, is activated by TNF‐α. Sustained activation of JNK can lead to apoptosis by modulation of the Bcl‐2 family of proteins. JNK can also transcriptionally activate death receptor expression (i.e. TRAIL receptor 2/death receptor 5). Furthermore, JNK can promote TNF‐α‐induced apoptotic signaling at complex II by facilitating degradation of cFLIP, thus

198

THE LIVER:  THE EXTRINSIC PATHWAY

Figure 17.3  Complex I and complex II of tumor necrosis factor alpha signaling. Tumor necrosis factor receptor 1 (TNFR1), upon binding TNF‐α on its extracellular domain, activates complex I and complex II. Complex I is formed by the adaptor proteins TNFR1‐associated death domain protein (TRADD) and receptor‐interacting protein (RIP), which recognize and bind via their death domains (DD) and TNF receptor‐associated factor (TRAF2) via its kinase domain or an intermediate domain. Complex I mediates the activation of nuclear factor κB (NFκB) and transient c‐Jun N‐terminal kinase (JNK) activation. NFκB translocates to the nucleus transcriptionally, activating antiapoptotic and inflammatory genes, such as cellular FLICE‐like inhibitory protein (cFLIP), Bcl‐xL, Mcl‐1, A1, and XIAP, which regulate apoptosis at multiple levels. Sustained JNK activation requires the adaptor protein RIP and is mediated in part by oxidative stress. Complex II is formed by receptor dissociation of TRADD, RIP, and TRAF2 and ligand‐independent recruitment of Fas‐associated death domain (FADD) via its DD. FADD contains a death effector domain (DED), leading to recruitment and activation of procaspase 8. In select conditions in the absence or inhibition of caspase 8 signaling, receptor‐interacting protein 1 (RIP1) interacts with RIP3, leading to the recruitment and phosphorylation of the mixed‐lineage kinase domain‐like pseudokinase (MLKL), which leads to necroptosis by translocating to the cell membrane.

antagonizing an antiapoptotic TNF‐α‐induced NFκB target gene. Similarly, loss of cellular inhibitors of apoptosis proteins 1 and 2, also antiapoptotic NFκB target genes, sensitizes carcinoma cells to TNF‐α‐mediated cytotoxicity [24]. TNF‐α can lead to superoxide formation and caspase‐independent cell death by TRADD and RIP1‐mediated activation of Nox1 NADPH oxidase leading to reactive oxygen species formation [25]. This process is independent of FADD, and caspase 8 activation. Thus, a multitude of complex processes contribute to TNF‐α cytotoxicity. In experimental models of liver injury, a role for TNF‐α cell death has been elucidated. Following partial hepatectomy, massive hepatocyte cell death occurs after completion of cell cycle progression due to sustained TNF‐α signaling in mice lacking tissue inhibitor of metalloproteinase 3 (Timp3), a model characterized by abnormal chronically elevated TNF‐α activity [26]. In TNFR1‐deficient ethanol‐fed mice, hepatocyte apoptosis, serum alanine aminotransferase levels (ALT),

and inflammatory foci were decreased as compared to wild‐type ethanol‐fed mice; TNFR2‐deficient mice developed liver injury and apoptosis comparable to those in wild‐type controls [27]. In ischemia reperfusion injury mice lacking TNFR1 and treated with pentoxyfylline, a pharmacologic TNF‐α inhibitor, liver injury and apoptosis are significantly reduced [28]. Liver samples from patients with alcoholic steatohepatitis or non‐ alcoholic steatohepatitis demonstrate enhanced TNFR1 expression [29]. Serum levels of TNFR1 in patients with alcoholic hepatitis are predictive of three‐month survival [30]. Thus, the TNF‐α cascade is activated in patients with many liver diseases, including fulminant hepatic failure, alcoholic steatohepatitis, non‐alcoholic steatohepatitis, chronic hepatitis C, and chronic hepatitis B [29, 31, 32]; it is indeed a hallmark of inflammatory changes in these conditions and likely contributes to hepatocyte apoptosis in vivo. Our understanding of why the TNFR1‐initiated NFκB cell survival pathways fail in these diseases remains rudimentary.



17:  Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases

Fas Fas (also known as Apo‐1, CD95) is ubiquitously expressed in liver cell types. Hepatocytes are exquisitely sensitive to Fas‐ induced apoptosis, and exogenously administered Fas agonistic antibody results in fulminant hepatic failure in mice [33]. Fas signaling usually results in hepatocyte apoptosis, although there are reports of Fas‐induced proliferation of T cells and fibroblasts, chemokine secretion from macrophages, and Fas‐mediated acceleration of liver regeneration after partial hepatectomy in mice [34]. Fas–Fas ligand (FasL) binding leads to receptor oligomerization, bringing together the intracellular DD, recruitment of FADD, and procaspase 8 or 10 at the DISC (Figure 17.4). This leads to activation and autoproteolytic activation of procaspase 8 or 10, generation of tBid, activation of Bax and Bak, mitochondrial permeabilization with eventual activation of caspase 3 and 7. Fas can be activated by soluble or circulating as well as membrane‐bound FasL. FasL is expressed by cells of the immune system, such as cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells [35]. The liver is enriched in both these cell populations, therefore under constant “Fas attack.” However Fas‐induced signaling is regulated at many levels. Cell surface expression of Fas, levels of FasL, and cFLIP inhibition of caspase 8 activation at the DISC are potential regulatory sites. Of interest in hepatocytes is the sequestration of Fas by the hepatocyte growth factor receptor (HGF) Met [14]. Met–Fas complexes prevent binding of FasL to Fas; however, Fas does not

199

affect HGF binding to its receptor Met. Pretreatment of cells with HGF releases Fas from this complex, and enhances FasL binding and toxicity at lower concentrations of FasL. High concentrations of FasL are maximally toxic even in the absence of HGF. Thus, the Met–Fas complex fine tunes and regulates the biologic availability of Fas in hepatocytes. In embryonic hepatocytes, Met prevents Fas‐induced cFLIP degradation, thus preventing apoptosis. In adult mice, genetic deficiency of Fas leads to hepatic hyperplasia, in addition to enlargement of lymph nodes and spleen [36]. The induction of fulminant hepatic failure in mice by exogenous administration of Fas agonistic antibody is further regulated by the Bcl‐2 family of proteins. It can be abrogated by overexpression of Bcl‐2 and enhanced by genetic inhibition of Bcl‐xL [37]. Genetic inhibition of Fas itself or Bid mitigates liver injury by Fas agonists [38]. Neutralization of Fas reduces warm ischemia/reperfusion‐related liver injury [39]. Circulating levels of serum Fas are elevated in patients with fulminant hepatic failure [40]. Levels of serum Fas vary by etiology, and the highest levels occur in patients with drug‐induced liver injury. Fas expression and apoptosis are enhanced in liver samples from patients with chronic hepatitis C [41]. Circulating levels of soluble Fas correlate with histologic activity, and along with levels of caspase 3 activity, are predictive of response to therapy [42]. Similarly, in patients with chronic hepatitis B hepatocyte Fas levels and circulating levels of soluble Fas are elevated [41, 43]. Fas expression is enhanced in liver samples

Figure 17.4  Fas and TRAIL receptor signaling: Fas and TRAIL receptors are activated by ligand binding, which leads to receptor oligomerization, bringing together their conserved death domains (DD). The adaptor protein Fas‐associated death domain (FADD) binds to the trimerized intracellular DD and via its death effector domain (DED) leads to activation of procaspase 8. Active caspase 8 leads to proteolytic cleavage of Bid to tBid and downstream mitochondrial permeabilization via activation of Bax and Bak. Mitochondrial permeabilization leads to release of the contents of the intermembrane space including cytochrome c, smac/DIABLO, Apaf‐1, and endonuclease G, culminating in the activation of caspase 3/7 and cleavage of cellular proteins.

200

THE LIVER:  THE INTRINSIC PATHWAY

from patients with non‐alcoholic fatty liver disease [7]. In experimental models of dietary and genetic fatty liver, steatotic livers are sensitized to exogenous Fas administration. Indeed, in patients with non‐alcoholic fatty liver disease, the inhibition of Fas by Met is diminished, providing another mechanism to explain the enhanced sensitivity to Fas‐induced hepatocyte apoptosis. Furthermore, free fatty acid treatment can increase Fas expression in vitro in cell culture models of hepatocyte steatosis, sensitizing cells to Fas‐induced apoptosis. In the bile duct‐ ligated mouse model of cholestatic liver injury, hepatocyte apoptosis is mediated by Fas, and Fas‐induced apoptosis promotes hepatic fibrosis [44]. Toxic bile acids promote cell surface expression of Fas, and can lead to ligand‐independent Fas oligomerization and induction of hepatocyte apoptosis [45, 46]. In bile salt‐mediated ligand‐independent hepatocyte apoptosis, Fas phosphorylation is required for its translocation to the cell surface; this can occur in a Yes kinase/epidermal growth factor receptor‐dependent and JNK‐dependent manner [47].

disease due to a reduction in hepatocyte apoptosis [54, 55]. Free fatty acids, which are elevated in the metabolic syndrome, transcriptionally enhance TRAIL‐R2 expression in cell culture and render steatotic cells sensitive to TRAIL toxicity [51]. In acute hepatitis B‐induced liver failure in humans and experimental adenoviral acute hepatitis in mice, TRAIL‐R2 expression is enhanced, as is sensitivity to TRAIL. This occurs independently of Kupffer cells and NK cells, suggesting a hepatocyte‐generated paracrine loop for elimination of virally infected cells [56]. Circulating soluble TRAIL levels are elevated in patients with chronic viral hepatitis B. Hepatitis B X antigen increases TRAIL‐R1 expression in cell culture experiments, conferring sensitivity to TRAIL. In liver samples from patients with chronic hepatitis C, TRAIL‐R1 and TRAIL‐R2 expression and TRAIL‐ induced apoptosis were enhanced [50]. Hepatitis C virus core protein also selectively modulates cellular responsiveness to TRAIL by promoting TRAIL‐induced Bid cleavage [57].

Tumor necrosis factor‐related apoptosis‐ inducing ligand

THE INTRINSIC PATHWAY

The role of tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL, also known as Apo‐2 ligand) and its receptors in liver disease is an area with remarkable recent advances. TRAIL binds with several receptors [48]. TRAIL receptor 1 (TRAIL‐ R1/death receptor (DR) 4) and TRAIL receptor 2 (TRAIL‐R2/ DR5/killer/TRICK2) are complete receptors and can induce apoptosis via caspase activation, similar to Fas [49]. This occurs via the adaptor protein FADD, recruitment of procaspase 8 and 10 to the TRAIL receptor DISC, in a cFLIP‐regulated manner (Figure  17.4). TRAIL receptor 3 (TRAIL‐R3/Apo‐3/TRAMP/ WSL‐1/LARD, decoy receptor 1 (DcR1)) and TRAIL receptor 4 (TRAIL‐R4, DR6, decoy receptor 2 (DcR2)) are incomplete cell surface receptors and cannot stimulate apoptotic signaling. Normal human hepatocytes, in situ and in vivo, are considered resistant to TRAIL‐induced apoptosis, though there are occasional reports of in vitro TRAIL‐induced hepatocyte apoptosis [50]. This resistance to cell death may be secondary to cFLIP‐ induced inhibition of caspase 8 activation at the DISC or cell surface expression/availability of TRAIL‐R1 or TRAIL‐R2. However, diseased hepatocytes are sensitized to TRAIL‐induced apoptosis [51]. TRAIL also sensitizes to Fas‐induced hepatocyte apoptosis by activating JNK and the proapoptotic BH3‐ only protein Bim. TRAIL‐induced hepatocyte apoptosis has been demonstrated in cholestatic, viral, and metabolic liver diseases. Toxic bile acids transcriptionally regulate hepatocyte cell surface TRAIL‐ R2 expression in Fas‐deficient cells and inactivate cFLIP by phosphorylation, thus dually sensitizing cells to TRAIL‐induced apoptosis [52]. In the bile duct‐ligated mouse model of cholestasis, hepatocyte TRAIL‐R2 expression is enhanced and hepatocytes are sensitized to exogenously administered TRAIL [53]. By corollary, liver injury and hepatocyte apoptosis are significantly reduced in TRAIL‐deficient mice following bile duct ligation. Steatosis is also associated with increased hepatocyte expression of TRAIL‐R2 and TRAIL‐R1, which imparts sensitivity to TRAIL toxicity, and TRAIL or TRAIL receptor deletion in mice improves dietary obesity‐associated fatty liver

Intracellular stress leads to the activation of the intrinsic pathway of apoptosis. Stress can be perceived and transduced by any membrane‐defined organelle in the cell. For example, lysosomes can mediate steatotic liver cell death, as can the endoplasmic reticulum (ER). DNA damage can lead to genotoxic stress and steatosis can activate JNK, also a mediator of the intrinsic pathway of apoptosis. These processes converge on mitochondria and are transduced by the Bcl‐2 family of proteins, and so are usually referred to as the Bcl‐2‐regulated or mitochondrial pathway of apoptosis. The Bcl‐2 family consists of proapoptotic and antiapoptotic proteins, which are classified into groups on the basis of the number of shared Bcl‐2 homology (BH) domains. The proapoptotic proteins are structurally divided based on the number of shared BH domains into multidomain (Bak and Bax, display BH1, 2, and 3 domains) and BH3‐only proteins (Bid, Noxa, Puma, Bim, Bmf, Bik, Hrk, and Bad). The antiapoptotic proteins include Bcl‐2, Bcl‐xL, Bcl‐w, A1, Mcl‐1, and Boo, and share four BH domains with the exception of Mcl‐1, which shares three BH domains with the rest of the antiapoptotic Bcl‐2 family members. The liver expresses Bcl‐xL and Mcl‐1; Bcl‐2 is not expressed by hepatocytes. Bax and Bak are both abundantly expressed by hepatocytes. The antiapoptotic members of this family are located on the cytoplasmic aspect of membrane‐bound organelles, primarily the mitochondria, though also on other organelles such as the ER. They protect cells from death by preventing spontaneous oligomerization of Bax and Bak, and may be necessary for survival of certain cell types. Given this constitutive role in preventing apoptosis, predictably, hepatocyte‐specific Bcl‐xL‐ or Mcl‐1‐ knockout mouse models demonstrate an increase in spontaneous hepatocyte apoptosis, liver injury, and fibrosis [58, 59]. Bax and Bak are required for mitochondrial permeabilization, while Bax is located in the cytosol and translocates to mitochondria upon activation; Bak is a resident mitochondrial outer membrane protein. The activation of Bax and Bak is regulated by interactions between the antiapoptotic Bcl‐2 proteins and the BH3‐only proapoptotic proteins. Several models have been



17:  Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases

proposed to explain the biochemical activation of Bax or Bak by proapoptotic BH3‐only proteins. Using Bim as an example, upon activation, Bim is released from the dynein motor complex, and can directly engage and activate Bax and Bak. Alternatively, Bim can bind and negate the inhibitory effect of Bcl‐2 or Bcl‐xL, releasing Bax and Bak from inhibition by these proteins (the derepression model). Although the large number of BH3‐only proteins imparts redundancy, its primary effect is to impart stimulus specificity. For example, free fatty acids activate Bim and Puma [60]; Puma and Noxa are target genes of the tumor suppressor p53 [61]. Cytosolic Bid must be cleaved by active caspase 8 or 10 to generate truncated Bid (tBid), which translocates to the mitochondria to activate Bak or Bax. Bim is sequestered by the microtubule‐associated dynein motor complex in the cytosol, from which it is dissociated by proapoptotic stimuli. For example, JNK‐mediated Bim phosphorylation promotes its mitochondrial translocation. Bim is also regulated transcriptionally, and known to be increased in ER stress‐induced apoptosis. Cytosolic Bad is bound to the death inhibitory protein 14‐3‐3. Activating and inhibitory phosphorylation events have been described for both Bim and Bad.

Mitochondria In addition to the metabolic functions of mitochondria, hepatocytes require mitochondria to die. The mitochondrial intermembrane space sequesters a number of proapoptotic proteins including cytochrome c, SMAC/DIABLO (second mitochondrial activator of caspase/direct IAP‐binding protein with low pI), HtrA2/Omi, AIF (apoptosis‐inducing factor), and endonuclease G [2, 15]. Current models suggest that active Bax or Bak form pores in the outer mitochondrial membrane leading to mitochondrial outer membrane permeabilization (MOMP) and release of these mediators into the cytosol [62]. In the absence of Bak and Bax, cells are resistant to apoptotic stimuli; thus, Bak and Bax are critical effectors of MOMP. Though previously thought to be a rapid and complete phenomenon wherein all mitochondria within a cell would experience MOMP within minutes, recent work has challenged this paradigm by describing partial MOMP in cells undergoing apoptotic stress. Since partial MOMP does not result in apoptosis, this type of cellular stress has been named sublethal stress. Inactive Bak and Bax support mitochondrial fusion, whereas active Bak and Bax shift mitochondrial dynamics toward fission. MOMP can also occur secondary to permeability transition (PT) of the inner mitochondrial membrane (IMM) via the PT pore (PTP). The identity of the PTP has received much attention and several candidate proteins have been evaluated, including adenine nucleotide transporter and the voltage‐dependent anion channel. Current evidence suggests that the PTP is formed from FOF1 ATP synthase dimers in the IMM [63], though these is some evidence to suggest that spastic paraplegia 7 could also form the PTP. Opening of the permeability transition pore leads to rapid fluxes of ions and water, dissipation of the mitochondrial inner transmembrane potential, swelling of the mitochondria, and outer mitochondrial membrane rupture leading to the release of the contents into the intermembrane space. MOMP releases intermembrane contents into the cytosol and commits the cell to

201

apoptosis. SMAC inactivates post‐mitochondrial inhibitors of apoptosis proteins (IAP). Cytosolic cytochrome c, apoptotic protease‐activating factor 1 (Apaf) and ATP form a complex called the apoptosome, leading to activation of procaspase 9 and effector caspases 3 and 7 [64]. These effector caspases cleave over 500 substrates resulting in cellular demolition. Cytokeratin 18 is a structural protein expressed in most epithelial cells that is cleaved by caspase 3 at aspartate positions 238 and 396. The fragment generated by this cleavage, cytokeratin 18–aspartate 396 (CK18‐asp396) forms a neo‐epitope that is recognized by the M30 antibody. This neoepitope can be detected in apoptotic tissues as well as serum by a commercially available ELISA. Indeed circulating levels of CK18‐asp396 are elevated in patients with liver injury and can correlate with outcome [4]. Thus this biomarker presents a noninvasive, simple, and mechanistic tool to monitor progress and response to therapy in liver injury.

Lysosomes Lysosomes are intracellular organelles with acid intravesicular pH that contain lysosomal proteases, known as cathepsins [65]. Cathepsin B and D, two of 11 known human cathepsins, are stable and active at neutral pH. Methodical dissection of pathways that mediate intracellular death signals demonstrates that lysosomes can be involved in the intrinsic pathway of cell death. Typically, lysosomal permeabilization, when it mediates apoptosis, is selective and partial and is observed upstream of mitochondrial permeabilization. Cathepsin B‐induced mitochondrial permeabilization can occur via caspase 2 (in mice) and via proteolytic cleavage of Bid similar to death receptor‐induced activation of Bid [66, 67]. Indeed Bid also links death receptors to lysosomal permeabilization; providing crosstalk between death receptors and their engagement of the lysosomal and mitochondrial pathways [66]. Bax activation by intracellular stress can also result in lysosomal permeabilization [68]. Cathepsin D levels were elevated in serum from patients with fulminant hepatic failure as well as chronic hepatitis [69, 70]. Cathepsin B‐deficient mice are resistant to TNF‐α‐induced hepatocyte apoptosis [23]. In models of cellular steatosis, cathepsin B inhibition prevents mitochondrial permeabilization and apoptosis. In cathepsin B‐deficient mice liver apoptosis, injury, and fibrosis are diminished following bile duct ligation [71]; liver apoptosis and injury are abrogated in ischemia reperfusion injury as well [72].

Endoplasmic reticulum The ER has an inbuilt mechanism to cope with excess or altered unfolded proteins that serves to correct the inciting imbalance. This process is termed the unfolded protein response (UPR). The UPR can also be activated by stimuli that affect the function of the ER, such as calcium depletion, glycosylation inhibition (tunicamycin), ultraviolet radiation, and insulin resistance. The ER stress response consists of a series of compensatory processes to correct both the excess and the stress of the unfolded proteins. Global translation is attenuated to reduce the functional protein load of the ER. There is also selective translation of UPR target genes aimed at protecting the ER [73, 74]. The transducers of ER stress are membrane proteins that have an ER

202

THE LIVER:  THE INTRINSIC PATHWAY

luminal domain and a cytosolic domain. Inositol‐requiring protein 1 alpha (IRE1α) and protein kinase RNA‐like ER kinase (PERK) auto‐transphosphorylate when released from the ER chaperone BiP/Grp78. IRE1α possesses endoribonucleolytic activity leading to excision of an intron within X‐box binding protein 1 (XBP1) mRNA to generate spliced XBP1 (sXBP1), a transcription factor that activates a subset of UPR target genes. IRE1α also recruits TRAF2, leading to JNK activation. PERK phosphorylates and inactivates the eukaryotic translation initiation factor 2α (eIF2α), resulting in global translation attenuation with selective translation of activating transcription factor 4 (ATF4), which leads to transcription of C/EBP‐homologous protein (CHOP), and the ER chaperone BiP/Grp78. Activating transcription factor 6 (ATF6) is cleaved within the ER membrane, generating an ATF6 fragment that translocates to the nucleus and activates a subset of UPR target genes. It is not known if ATF6 also regulates apoptotic signaling. ER stress also activates a negative feedback regulatory loop that terminates the UPR; however in the setting of sustained ER stress, proapoptotic signaling occurs [75]. Bax and Bak both bind to the cytoplasmic domain of IRE1α, and in cells lacking Bax and Bak, IRE1 stress‐generated JNK activation and XBP1 splicing are reduced [76], thus linking the core apoptotic machinery to ER stress response. Bax and Bak localize on the ER membrane, in addition to mitochondrial membranes. In cells lacking both Bax and Bak, the ER is depleted of calcium and unable to respond to certain death stimuli [77]. The proapoptotic transcription factor CHOP can increase Bim expression transcriptionally and by inhibiting its proteasomal degradation, leading to Bim‐dependent ER stress‐induced apoptosis [78]. CHOP can also upregulate TRAIL‐R2 expression, sensitizing cancer cells to TRAIL‐induced apoptosis [79]. The involvement of the ER stress‐induced apoptotic pathway in liver diseases is an area of emerging research. In the bile duct‐ ligated mouse model of cholestasis, an early and transient induction of CHOP expression is observed [80]. Mice deficient in CHOP are protected from hepatocyte apoptosis, liver injury, and liver fibrosis. In cell culture, the toxic bile acid glycochenodeoxycholic acid also induces ER stress and CHOP expression in isolated rat hepatocytes [81]. In transgenic mice expressing hepatitis C viral core and E2 proteins, hepatocyte apoptosis is associated with CHOP expression [82]. Cycloheximide, an inhibitor of protein synthesis, induces ER stress, induction of CHOP expression and apoptotic hepatocyte cell death in rat livers [83]. In non‐alcoholic fatty liver disease, markers of ER stress were variably activated [84]. Toxic saturated fatty acids also induce ER stress and apoptosis in liver cell lines [85]. In a mouse model of alcohol‐induced liver injury, CHOP‐deficient mice are protected from hepatocyte apoptosis, though able to mount an ER stress response [86].

c‐Jun N‐terminal kinase Given the role of JNK in multiple models of cell death, it warrants a separate discussion as a final common cell death mediator. JNK1 and 2 are ubiquitously expressed, including liver, whereas JNK3 is not expressed in the liver [87]. JNK activation occurs downstream of kinase cascades that can be activated by multiple stimuli including TNF‐α, IRE1α, reactive oxygen

species, free fatty acids, and bile acids [88, 89]. JNK involvement in apoptosis is temporally regulated and stimulus specific [90]. The same inciting stimulus (e.g. TNF‐α) can induce biphasic JNK activation mediated by distinct intracellular pathways. Transient and early JNK activation promotes survival; and sustained and late activation of JNK promotes apoptosis [91]. In the case of TNF‐α, production of reactive oxygen species mediates the delayed and sustained activation of JNK. Other stimuli, such as toxic free fatty acids, result in early and sustained JNK activation, culminating in apoptotic signaling [92]. JNK‐stimulated proapoptotic signaling converges on mitochondria via the activation of Bax and Bak. In the absence of Bax and Bak, JNK‐ induced cell death is mitigated. Furthermore, mitochondrial permeabilization and release of cytochrome c are abolished in cells derived from mice lacking Jnk1 and 2 genes, in response to stimuli that cause intracellular stress [90]. JNK‐mediated phosphorylation of pro‐ and antiapoptotic proteins upstream of mitochondria also regulates apoptotic sensitivity. JNK can phosphorylate and activate the BH3‐only proteins; for example, Bim phosphorylation releases it from binding to the dynein motor complex and promotes apoptosis [93]. Sustained JNK activation promotes caspase 8 formation at the DISC by activation of the E3 ubiquitin ligase Itch, which ubiquinates and degrades cFLIP, promoting liver cell death [94]. JNK can phosphorylate the antiapoptotic proteins Bcl‐2, Bxl‐xL, and Mcl‐1, and the proapoptotic proteins Bmf and Bad. JNK1 and JNK2 can both mediate liver injury in a stimulus‐ specific manner. In a murine model of steatohepatitis induced by methionine‐ and choline‐deficient diet, JNK1 plays a predominant role. In high‐fat diet‐induced obesity and genetic obesity in mice, JNK was activated; this was found to be predominantly JNK1, though JNK2 plays a role that is unmasked in the absence of JNK1 [95]. In free fatty acid‐based cellular models of hepatocyte steatosis, JNK2 is the predominant isoform that mediates apoptosis [92]. Oleic acid, a minimally toxic free fatty acid, also sensitizes steatotic hepatocytes to TRAIL‐induced apoptosis by JNK‐dependent transcriptional upregulation of the death receptor TRAIL‐R2 [51]. This mechanism is shared by toxic bile acids, which also sensitize hepatocytes to TRAIL‐induced apoptosis by transcriptionally activating TRAIL‐R2 expression in a JNK‐dependent manner [53, 96]. Liver injury induced by ischemia reperfusion is also mediated by JNK, and pharmacologic inhibition of JNK in donor livers improved graft survival and decreased apoptosis after orthotopic liver transplantation [97]. In acetaminophen‐ induced acute liver injury JNK activation was robust and sustained, and led to Bax translocation to mitochondria and poor animal survival. Pharmacologic inhibition of JNK decreased liver injury and hepatocyte cell death and improved survival; utilizing genetically deficient models of Jnk1 or Jnk2 it was demonstrated that both mediate liver injury, though JNK2 was predominant. JNK activation was observed in hepatocytes in human liver samples from patients with acetaminophen‐ induced acute liver failure. JNK inhibition was more effective in decreasing hepatocyte cell death than N‐acetylcysteine in a murine model of acetaminophen‐induced liver injury. In a murine model of TNF‐α‐induced liver injury utilizing galactosamine and lipopolysaccharide, JNK2 mediated caspase 8 activation and mitochondrial permeabilization.



17:  Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases

THE CONSEQUENCES OF HEPATOCYTE APOPTOSIS Apoptosis, inflammation, and injury are in some ways inseparable, and it is difficult sometimes to dissect the primary event. However, based on the inciting stimulus, apoptosis or inflammatory signaling may be the primary event, each stimulating the other. The liver has a large population of Kupffer cells, NK cells, and NK T cells. These cells are a ready source of TNF‐α and other cytokines that mediate inflammation, such as Fas, TRAIL, and TNF‐α that mediate hepatocyte apoptosis and transforming growth factor beta (TGF‐β) that activates stellate cells. Apoptotic hepatocytes can be engulfed by Kupffer cells, leading to generation of cytokines [6]. In keeping with this the pharmacologic inhibition of apoptosis prevents Kupffer cell activation. Also, in the bile duct‐ligated mouse, Kupffer cell depletion decreases hepatocyte apoptosis, liver injury, and liver inflammation. In addition, stressed hepatocytes increase expression of NKG2D ligands, thus inviting NK‐ and NKT cell‐mediated destruction. Fibrosis is the hallmark of ongoing liver injury. Hepatic stellate cells mediate hepatic fibrosis. In the normal liver, stellate cells maintain a quiescent phenotype. On activation, they undergo a metamorphosis to become myofibroblasts, secreting collagen which leads to liver fibrosis. Stellate cells in vitro can engulf apoptotic hepatocytes, leading to their activation, and increased expression of TGF‐β, alpha smooth muscle actin, and collagen alpha 1 [98]. Similarly, in vivo hepatocyte apoptosis is a fibrogenic stimulus. Several experimental studies have demonstrated that the inhibition of hepatocyte apoptosis abrogates liver fibrosis [5, 71, 99]. By corollary, apoptosis of activated stellate cells should decrease liver fibrosis and dissociate ongoing hepatocyte apoptosis from the ensuing fibrogenic response. Indeed, activated stellate cells are sensitized to apoptotic signaling. This can be achieved by inhibition of NFκB, TRAIL‐mediated stellate cell apoptosis, and NK cell‐mediated stellate cell apoptosis. Indeed, the resolution phase of fibrosis requires apoptosis of activated hepatic stellate cells. Lastly, the clinical applications of apoptosis are discussed in the conclusion of this chapter. The cytokeratin 18‐derived M30 neoantigen reflects epithelial cell apoptosis, is abundant in hepatocytes, can easily be measured in serum by a commercially available ELISA, and correlates with hepatocyte apoptosis in diverse liver diseases. In a study with a small number of patients with chronic hepatitis C, pretreatment M30 levels were predictive of response to therapy, inferring from this that patients with an apoptotic response to virally infected hepatocytes are more likely to have a treatment response. In another study with chronic hepatitis C patients with normal transaminases, serum M30 levels correlated with fibrosis. In patients with non‐alcoholic fatty liver disease, serum M30 levels offer reliable discrimination of patients with steatohepatitis from simple steatosis, and increasing levels are predictive of a higher likelihood of inflammation. Caspase inhibitors have demonstrated efficacy in preventing hepatocyte apoptosis and injury in experimental models of liver injury [99, 100]. In patients with chronic hepatitis C, orally administered caspase inhibitor was found to be safe, and lowered transaminases. The caspase inhibitor Emricasan (IDUN‐6556) is currently in clinical trials for non‐alcoholic steatohepatitis.

203

In conclusion, hepatocyte apoptosis is a key mediator of liver injury and inflammation in most forms of liver disease. Multiple apoptotic pathways are activated by a given injurious stimulus in a vulnerable hepatocyte. The predominant signaling pathway that results in mitochondrial dysfunction in a given cell is difficult to discern; however, multiple pathways could potentially cooperate or oppose each other, to eventually result in mitochondrial permeabilization. Once mitochondrial permeabilization occurs, the hepatocyte is committed to cell death. Evidence of hepatocyte apoptosis can be demonstrated by serum markers and early studies demonstrate prognostic significance of apoptosis markers. Lastly, therapeutic manipulation of apoptosis is of benefit in preventing liver injury and fibrosis.

ACKNOWLEDGMENTS Support for this work is provided by NIH grant DK 41876 (GJG), DK 111378 (HM), and the Mayo Foundation.

REFERENCES 1. Malhi, H., Gores, G.J., and Lemasters, J.J. Apoptosis and necrosis in the liver: a tale of two deaths? Hepatology, 2006;43(2 Suppl 1):S31–44. 2. Green, D.R. and Kroemer, G. The pathophysiology of mitochondrial cell death. Science, 2004;305(5684):626–9. 3. Xiang, J., Chao, D.T., and Korsmeyer, S.J. BAX‐induced cell death may not require interleukin 1 beta‐converting enzyme‐like proteases. Proc Natl Acad Sci U S A, 1996;93(25):14559–63. 4. Rutherford, A.E., Hynan, L.S., Borges, C.B. et al. Serum apoptosis markers in acute liver failure: a pilot study. Clin Gastroenterol Hepatol, 2007; 5(12):1477–83. 5. Canbay, A., Higuchi, H., Bronk, S.F. et al. Fas enhances fibrogenesis in the bile duct ligated mouse: a link between apoptosis and fibrosis. Gastroenterology, 2002;123(4):1323–30. 6. Canbay, A., Friedman, S., and Gores, G.J. Apoptosis: the nexus of liver injury and fibrosis. Hepatology, 2004;39(2):273–8. 7. Feldstein, A.E., Canbay, A., Angulo, P. et al. Hepatocyte apoptosis and fas expression are prominent features of human nonalcoholic steatohepatitis. Gastroenterology, 2003;125(2):437–43. 8. Natori, S., Rust, C., Stadheim, L.M. et al. Hepatocyte apoptosis is a pathologic feature of human alcoholic hepatitis. J Hepatol, 2001;34(2):248–53. 9. Kohli, V., Selzner, M., Madden, J.F., Bentley, R.C., and Clavien, P.A. Endothelial cell and hepatocyte deaths occur by apoptosis after ischemia‐ reperfusion injury in the rat liver. Transplantation, 1999;67(8):1099–105. 10. Guicciardi, M.E. and Gores, G.J. Life and death by death receptors. FASEB J. 2009;23(6):1625–37. 11. Muzio, M., Stockwell, B.R., Stennicke, H.R., Salvesen, G.S., and Dixit, V.M. An induced proximity model for caspase‐8 activation. J Biol Chem. 1998;273(5):2926–30. 12. Scaffidi, C., Fulda, S., Srinivasan, A. et al. Two CD95 (APO‐1/Fas) signaling pathways. EMBO J, 1998;17(6):1675–87. 13. Yin, X.M. Bid, a critical mediator for apoptosis induced by the activation of Fas/TNF‐R1 death receptors in hepatocytes. J Mol Med (Berl), 2000 ;78(4):203–11. 14. Wang, X., DeFrances, M.C., Dai, Y. et  al. A mechanism of cell survival: sequestration of Fas by the HGF receptor Met. Mol Cell, 2002;9(2):411–21. 15. Danial, N.N. and Korsmeyer, S.J. Cell death: critical control points. Cell, 2004;116(2):205–19. 16. Yamada, Y., Webber, E.M., Kirillova, I., Peschon, J.J., and Fausto, N. Analysis of liver regeneration in mice lacking type 1 or type 2 tumor necrosis factor receptor: requirement for type 1 but not type 2 receptor. Hepatology, 1998;28(4):959–70.

204

THE LIVER:  REFERENCES

17. Tartaglia, L.A., Ayres, T.M., Wong, G.H., and Goeddel, D.V. A novel domain within the 55 kd TNF receptor signals cell death. Cell, 1993;74(5):845–53. 18. Micheau, O. and Tschopp, J. Induction of TNF receptor I‐mediated apoptosis via two sequential signaling complexes. Cell, 2003;114(2):181–90. 19. Grootjans, S., Vanden Berghe, T., and Vandenabeele, P. Initiation and execution mechanisms of necroptosis: an overview. Cell Death Differ, 2017;24(7):1184–95. 20. Cai, Z., Jitkaew, S., Zhao, J. et al. Plasma membrane translocation of trimerized MLKL protein is required for TNF‐induced necroptosis. Nat Cell Biol, 2014;16(1):55–65. 21. Sass, G., Shembade, N.D., Haimerl, F. et  al. TNF pretreatment interferes with mitochondrial apoptosis in the mouse liver by A20‐mediated down‐ regulation of Bax. J Immunol, 2007;179(10):7042–9. 22. Guicciardi, M.E., Deussing, J., Miyoshi, H. et al. Cathepsin B contributes to TNF‐alpha‐mediated hepatocyte apoptosis by promoting mitochondrial release of cytochrome c. J Clin Invest, 2000;106(9):1127–37. 23. Guicciardi, M.E., Miyoshi, H., Bronk, S.F., and Gores, G.J. Cathepsin B knockout mice are resistant to tumor necrosis factor‐alpha‐mediated hepatocyte apoptosis and liver injury: implications for therapeutic applications. Am J Pathol, 2001;159(6):2045–54. 24. Varfolomeev, E., Blankenship, J.W., Wayson, S.M. et  al. IAP antagonists induce autoubiquitination of c‐IAPs, NF‐kappaB activation, and TNFalpha‐ dependent apoptosis. Cell, 2007;131(4):669–81. 25. Kim, Y.S., Morgan, M.J., Choksi, S., and Liu, Z.G. TNF‐induced activation of the Nox1 NADPH oxidase and its role in the induction of necrotic cell death. Mol Cell, 2007;26(5):675–87. 26. Mohammed, F.F., Smookler, D.S., Taylor, S.E. et al. Abnormal TNF activity in Timp3‐/‐ mice leads to chronic hepatic inflammation and failure of liver regeneration. Nat Genet, 2004;36(9):969–77. 27. Yin, M., Wheeler, M.D., Kono, H. et al. Essential role of tumor necrosis factor alpha in alcohol‐induced liver injury in mice. Gastroenterology, 1999;117(4):942–52. 28. Rudiger, H.A. and Clavien, P.A. Tumor necrosis factor alpha, but not Fas, mediates hepatocellular apoptosis in the murine ischemic liver. Gastroenterology, 2002;122(1):202–10. 29. Ribeiro, P.S., Cortez‐Pinto, H., Sola, S. et al. Hepatocyte apoptosis, expression of death receptors, and activation of NF‐kappaB in the liver of nonalcoholic and alcoholic steatohepatitis patients. Am J Gastroenterol, 2004;99(9):1708–17. 30. Spahr, L., Giostra, E., Frossard, J.L. et al. Soluble TNF‐R1, but not tumor necrosis factor alpha, predicts the 3‐month mortality in patients with alcoholic hepatitis. J Hepatol, 2004;41(2):229–34. 31. Torre, F., Rossol, S., Pelli, N. et al. Kinetics of soluble tumour necrosis factor (TNF)‐alpha receptors and cytokines in the early phase of treatment for chronic hepatitis C: comparison between interferon (IFN)‐alpha alone, IFN‐alpha plus amantadine or plus ribavirin. Clin Exp Immunol, 2004; 136(3):507–12. 32. Fang, J.W., Shen, W.W., Meager, A., and Lau, J.Y. Activation of the tumor necrosis factor‐alpha system in the liver in chronic hepatitis B virus infection. Am J Gastroenterol, 1996;91(4):748–53. 33. Ogasawara, J., Watanabe‐Fukunaga, R., Adachi, M. et al. Lethal effect of the anti‐Fas antibody in mice. Nature, 1993;364(6440):806–9. 34. Desbarats, J. and Newell, M.K. Fas engagement accelerates liver regeneration after partial hepatectomy. Nat Med, 2000;6(8):920–3. 35. Berke, G. The CTl’s kiss of death. Cell, 1995;81(1):9–12. 36. Adachi, M., Suematsu, S., Kondo, T. et al. Targeted mutation in the Fas gene causes hyperplasia in peripheral lymphoid organs and liver. Nat Genet, 1995;11(3):294–300. 37. Lacronique, V., Mignon, A., Fabre, M. et  al. Bcl‐2 protects from lethal hepatic apoptosis induced by an anti‐Fas antibody in mice. Nat Med, 1996;2(1):80–6. 38. Zhang, H., Cook, J., Nickel, J. et al. Reduction of liver Fas expression by an antisense oligonucleotide protects mice from fulminant hepatitis. Nat Biotechnol, 2000;18(8):862–7. 39. Al‐Saeedi, M., Steinebrunner, N., Kudsi, H. et al. Neutralization of CD95 ligand protects the liver against ischemia‐reperfusion injury and prevents acute liver failure. Cell Death Dis, 2018;9(2):132. 40. Ryo, K., Kamogawa, Y., Ikeda, I. et al. Significance of Fas antigen‐mediated apoptosis in human fulminant hepatic failure. Am J Gastroenterol, 2000;95(8):2047–55. 41. Kiyici, M., Gurel, S., Budak, F. et  al. Fas antigen (CD95) expression and apoptosis in hepatocytes of patients with chronic viral hepatitis. Eur J Gastroenterol Hepatol, 2003;15(10):1079–84.

42. Toyoda, M., Kakizaki, S., Horiguchi, N. et al. Role of serum soluble Fas/ soluble Fas ligand and TNF‐alpha on response to interferon‐alpha therapy in chronic hepatitis C. Liver, 2000;20(4):305–11. 43. Song, le H., Binh, V.Q., Duy, D.N. et al. Variations in the serum concentrations of soluble Fas and soluble Fas ligand in Vietnamese patients infected with hepatitis B virus. J Med Virol, 2004;73(2):244–9. 44. Miyoshi, H., Rust, C., Roberts, P.J., Burgart, L.J., and Gores, G.J. Hepatocyte apoptosis after bile duct ligation in the mouse involves Fas. Gastroenterology, 1999;117(3):669–77. 45. Sodeman, T., Bronk, S.F., Roberts, P.J., Miyoshi, H., and Gores, G.J. Bile salts mediate hepatocyte apoptosis by increasing cell surface trafficking of Fas. Am J Physiol Gastrointest Liver Physiol, 2000;278(6):G992–9. 46. Faubion, W.A., Guicciardi, M.E., Miyoshi, H. et al. Toxic bile salts induce rodent hepatocyte apoptosis via direct activation of Fas. J Clin Invest, 1999;103(1):137–45. 47. Reinehr, R., Becker, S., Wettstein, M., and Haussinger, D. Involvement of the Src family kinase yes in bile salt‐induced apoptosis. Gastroenterology, 2004;127(5):1540–57. 48. Kimberley, F.C. and Screaton, G.R. Following a TRAIL: update on a ligand and its five receptors. Cell Res, 2004;14(5):359–72. 49. Schneider, P., Thome, M., Burns, K. et al. TRAIL receptors 1 (DR4) and 2 (DR5) signal FADD‐dependent apoptosis and activate NF‐kappaB. Immunity, 1997;7(6):831–6. 50. Volkmann, X., Fischer, U., Bahr, M.J. et  al. Increased hepatotoxicity of tumor necrosis factor‐related apoptosis‐inducing ligand in diseased human liver. Hepatology, 2007;46(5):1498–508. 51. Malhi, H., Barreyro, F.J., Isomoto, H., Bronk, S.F., and Gores, G.J. Free fatty acids sensitise hepatocytes to TRAIL mediated cytotoxicity. Gut, 2007;56(8):1124–31. 52. Higuchi, H., Yoon, J.H., Grambihler, A. et  al. Bile acids stimulate cFLIP phosphorylation enhancing TRAIL‐mediated apoptosis. J Biol Chem, 2003;278(1):454–61. 53. Higuchi, H., Bronk, S.F., Taniai, M., Canbay, A., and Gores, G.J. Cholestasis increases tumor necrosis factor‐related apoptotis‐inducing ligand (TRAIL)‐ R2/DR5 expression and sensitizes the liver to TRAIL‐mediated cytotoxicity. J Pharmacol Exp Ther, 2002;303(2):461–7. 54. Idrissova, L., Malhi, H., Werneburg, N.W. et al. TRAIL receptor deletion in mice suppresses the inflammation of nutrient excess. J Hepatol, 2015;62(5):1156–63. 55. Hirsova, P., Weng, P., Salim, W. et al. TRAIL deletion prevents liver inflammation but not adipose tissue inflammation during murine diet‐induced obesity. Hepatol Commun, 2017;1(7):648–62. 56. Mundt, B., Kuhnel, F., Zender, L. et al. Involvement of TRAIL and its receptors in viral hepatitis. FASEB J, 2003;17(1):94–6. 57. Chou, A.H., Tsai, H.F., Wu, Y.Y. et  al. Hepatitis C virus core protein modulates TRAIL‐mediated apoptosis by enhancing Bid cleavage and activation of mitochondria apoptosis signaling pathway. J Immunol, 2005; 174(4):2160–6. 58. Takehara, T., Tatsumi, T., Suzuki, T. et al. Hepatocyte‐specific disruption of Bcl‐xL leads to continuous hepatocyte apoptosis and liver fibrotic responses. Gastroenterology, 2004;127(4):1189–97. 59. Weber, A., Boger, R., Vick, B. et  al. Hepatocyte‐specific deletion of the antiapoptotic protein myeloid cell leukemia‐1 triggers proliferation and hepatocarcinogenesis in mice. Hepatology, 2010;51(4):1226–36. 60. Barreyro, F.J., Kobayashi, S., Bronk, S.F. et al. Transcriptional regulation of Bim by Fox03A mediates hepatocyte lipoapoptosis. J Biol Chem, 2007;282(37):27141–54. 61. Yu, J. and Zhang, L. The transcriptional targets of p53 in apoptosis control. Biochem Biophys Res Commun, 2005;331(3):851–8. 62. Kalkavan, H. and Green, D.R. MOMP, cell suicide as a BCL‐2 family business. Cell Death Differ, 2018;25(1):46–55. 63. Giorgio, V., von Stockum, S., Antoniel, M. et al. Dimers of mitochondrial ATP synthase form the permeability transition pore. Proc Natl Acad Sci U S A, 2013;110(15):5887–92. 64. Riedl, S.J. and Salvesen, G.S. The apoptosome: signalling platform of cell death. Nat Rev Mol Cell Biol, 2007;8(5):405–13. 65. Guicciardi, M.E., Leist, M., and Gores, G.J. Lysosomes in cell death. Oncogene, 2004;23(16):2881–90. 66. Guicciardi, M.E., Bronk, S.F., Werneburg, N.W., Yin, X.M., and Gores, G.J. Bid is upstream of lysosome‐mediated caspase 2 activation in tumor necrosis factor alpha‐induced hepatocyte apoptosis. Gastroenterology, 2005; 129(1):269–84.



17:  Hepatocyte Apoptosis: Mechanisms and Relevance in Liver Diseases

67. Stoka, V., Turk, B., Schendel, S.L. et  al. Lysosomal protease pathways to apoptosis. Cleavage of bid, not pro‐caspases, is the most likely route. J Biol Chem, 2001;276(5):3149–57. 68. Feldstein, A.E., Werneburg, N.W., Li, Z., Bronk, S.F., and Gores, G.J. Bax inhibition protects against free fatty acid‐induced lysosomal permeabilization. Am J Physiol Gastrointest Liver Physiol, 2006;290(6): G1339–46. 69. Gove, C.D., Wardle, E.N., and Williams, R. Circulating lysosomal enzymes and acute hepatic necrosis. J Clin Pathol, 1981;34(1):13–16. 70. Kyaw, A., Aung, T., Htut, T., Myint, H., and Tin, K.M. Lysosomal enzyme activities in normals and in patients with chronic liver diseases. Clin Chim Acta, 1983;131(3):317–23. 71. Canbay, A., Guicciardi, M.E., Higuchi, H. et  al. Cathepsin B inactivation attenuates hepatic injury and fibrosis during cholestasis. J Clin Invest, 2003;112(2):152–9. 72. Baskin‐Bey, E.S., Canbay, A., Bronk, S.F. et  al. Cathepsin B inactivation attenuates hepatocyte apoptosis and liver damage in steatotic livers after cold ischemia‐warm reperfusion injury. Am J Physiol Gastrointest Liver Physiol, 2005;288(2):G396–402. 73. Ron, D. and Walter, P. Signal integration in the endoplasmic reticulum unfolded protein response. Nat Rev Mol Cell Biol, 2007;8(7):519–29. 74. Malhi, H. and Kaufman, R.J. Endoplasmic reticulum stress in liver disease. J Hepatol, 2011;54(4):795–809. 75. Lin, J.H., Li, H., Yasumura, D. et al. IRE1 signaling affects cell fate during the unfolded protein response. Science, 2007;318(5852):944–9. 76. Hetz, C., Bernasconi, P., Fisher, J. et al. Proapoptotic BAX and BAK modulate the unfolded protein response by a direct interaction with IRE1alpha. Science, 2006;312(5773):572–6. 77. Scorrano, L., Oakes, S.A., Opferman, J.T. et al. BAX and BAK regulation of endoplasmic reticulum Ca2+: a control point for apoptosis. Science, 2003;300(5616):135–9. 78. Puthalakath, H., O’Reilly, L.A., Gunn, P. et al. ER stress triggers apoptosis by activating BH3‐only protein Bim. Cell, 2007;129(7):1337–49. 79. He, Q., Luo, X., Jin, W. et  al. Celecoxib and a novel COX‐2 inhibitor ON09310 upregulate death receptor 5 expression via GADD153/CHOP. Oncogene, 2008;27(18):2656–60. 80. Tamaki, N., Hatano, E., Taura, K. et al. CHOP deficiency attenuates cholestasis‐induced liver fibrosis by reduction of hepatocyte injury. Am J Physiol Gastrointest Liver Physiol, 2008;294(2):G498–505. 81. Tsuchiya, S., Tsuji, M., Morio, Y., and Oguchi, K. Involvement of endoplasmic reticulum in glycochenodeoxycholic acid‐induced apoptosis in rat hepatocytes. Toxicol Lett, 2006;166(2):140–9. 82. Tumurbaatar, B., Sun, Y., Chan, T., and Sun, J. Cre‐estrogen receptor‐mediated hepatitis C virus structural protein expression in mice. J Virol Methods, 2007;146(1–2):5–13. 83. Ito, K., Kiyosawa, N., Kumagai, K. et al. Molecular mechanism investigation of cycloheximide‐induced hepatocyte apoptosis in rat livers by morphological and microarray analysis. Toxicology, 2006;219(1–3):175–86.

205

84. Puri, P., Mirshahi, F., Cheung, O. et al. Activation and dysregulation of the unfolded protein response in nonalcoholic fatty liver disease. Gastroenterology, 2008;134(2):568–76. 85. Wei, Y., Wang, D., Topczewski, F., and Pagliassotti, M.J. Saturated fatty acids induce endoplasmic reticulum stress and apoptosis independently of ceramide in liver cells. Am J Physiol Endocrinol Metab, 2006; 291(2):E275–81. 86. Ji, C., Mehrian‐Shai, R., Chan, C., Hsu, Y.H., and Kaplowitz, N. Role of CHOP in hepatic apoptosis in the murine model of intragastric ethanol feeding. Alcohol Clin Exp Res, 2005;29(8):1496–503. 87. Czaja, M.J. The future of GI and liver research: editorial perspectives. III. JNK/AP‐1 regulation of hepatocyte death. Am J Physiol Gastrointest Liver Physiol, 2003;284(6):G875–9. 88. Urano, F., Wang, X., Bertolotti, A. et al. Coupling of stress in the ER to activation of JNK protein kinases by transmembrane protein kinase IRE1. Science, 2000;287(5453):664–6. 89. Ueda, S., Masutani, H., Nakamura, H. et al. Redox control of cell death. Antioxid Redox Signal, 2002;4(3):405–14. 90. Tournier, C., Hess, P., Yang, D.D. et  al. Requirement of JNK for stress‐ induced activation of the cytochrome c‐mediated death pathway. Science, 2000;288(5467):870–4. 91. Ventura, J.J., Hubner, A., Zhang, C. et al. Chemical genetic analysis of the time course of signal transduction by JNK. Mol Cell, 2006;21(5):701–10. 92. Malhi, H., Bronk, S.F., Werneburg, N.W., and Gores, G.J. Free fatty acids induce JNK‐dependent hepatocyte lipoapoptosis. J Biol Chem, 2006;281(17):12093–101. 93. Lei, K. and Davis, R.J. JNK phosphorylation of Bim‐related members of the Bc12 family induces Bax‐dependent apoptosis. Proc Natl Acad Sci U S A, 2003;100(5):2432–7. 94. Chang, L., Kamata, H., Solinas, G. et al. The E3 ubiquitin ligase itch couples JNK activation to TNFalpha‐induced cell death by inducing c‐FLIP(L) turnover. Cell, 2006;124(3):601–13. 95. Hirosumi, J., Tuncman, G., Chang, L. et al. A central role for JNK in obesity and insulin resistance. Nature, 2002;420(6913):333–6. 96. Higuchi, H., Bronk, S.F., Takikawa, Y. et al. The bile acid glycochenodeoxycholate induces trail‐receptor 2/DR5 expression and apoptosis. J Biol Chem, 2001;276(42):38610–18. 97. Uehara, T., Xi Peng, X., Bennett, B. et al. c‐Jun N‐terminal kinase mediates hepatic injury after rat liver transplantation. Transplantation, 2004; 78(3):324–32. 98. Canbay, A., Taimr, P., Torok, N. et  al. Apoptotic body engulfment by a human stellate cell line is profibrogenic. Lab Invest, 2003;83(5):655–63. 99. Canbay, A., Feldstein, A., Baskin‐Bey, E., Bronk, S.F., and Gores, G.J. The caspase inhibitor IDN‐6556 attenuates hepatic injury and fibrosis in the bile duct ligated mouse. J Pharmacol Exp Ther, 2004;308(3):1191–6. 100. Natori, S., Higuchi, H., Contreras, P., and Gores, G.J. The caspase inhibitor IDN‐6556 prevents caspase activation and apoptosis in sinusoidal endothelial cells during liver preservation injury. Liver Transpl, 2003;9(3):278–84.

SECTION B: THE HEPATOCYTE

18

Copper Metabolism and the Liver Cynthia Abou Zeid, Ling Yi, and Stephen G. Kaler Section on Translational Neuroscience, Molecular Medicine Branch, Intramural Research Program, National Institutes of Health, Bethesda, MD, USA

OVERVIEW Among its several roles in the human body, the liver is a pivotal organ involved in metabolizing and excreting various toxic substances into the bile. One such function is elimination of trace metals, such as copper, that are necessary for several physiological processes but are also toxic if present in excess. Copper is a micronutrient acquired from the diet and plays a critical role in human growth and development. It is a divalent metal ion that can be found in both the cuprous (Cu+) and cupric (Cu2+) states. Its capacity to readily gain and donate electrons renders it important in various metabolic pathways. Copper is used as a cofactor for enzymes involved in the mitochondrial electron transport chain, detoxification of reactive oxygen species, iron transport, connective tissue metabolism, melanin pigment production, synthesis of amidated neuropeptides, and catecholamine metabolism [1–5]. However, the chemical properties of copper also expose to the risk of oxidative damage when homeostatic mechanisms are impaired. Consequently, tissue copper levels are tightly regulated by several transporters and chaperone proteins involved in systemic copper metabolism [3, 6]. The main route of systemic copper regulation is considered to be biliary excretion. Emerging data support regulation at the intestinal absorption level, but these mechanisms are not completely elucidated [7–9]. In all cases, the liver is still counted as the major organ of copper metabolism, since it functions not only in the elimination of excessive copper, but also in its distribution to several destinations following intestinal uptake.

THE LIVER: A CENTRAL ORGAN OF COPPER METABOLISM Dietary copper absorption primarily occurs in the duodenum. The process begins with uptake of reduced copper via the apical membrane of the enterocyte, through the copper transporter Ctr1 [10, 11]. DMT1 is a less‐specific divalent metal transporter that also imports iron, manganese, and nickel, and can play an additional role in copper uptake [12]. Once inside the enterocyte, copper is shuttled by its chaperone ATOX1 to the copper‐ transporting ATPase ATP7A. The latter is responsible for copper export via the enterocyte’s basolateral pole into the portal circulation (Figure  18.1). ATP7A has ubiquitous distribution and important roles in several cell types [13]. In hepatocytes, ATP7A has higher expression in the neonatal period with a reduction over time [14]. In the liver, the predominant copper‐transporting ATPase is ATP7B, a protein closely related to ATP7A with similarities in its structure and function [15]. The general role of these pumps depends on copper levels in the cell. In physiological copper states, ATP7A and ATP7B localize to the trans‐Golgi network and are responsible for conveying copper into the secretory pathway, for metalation of copper‐dependent enzymes. In case of elevated intracellular levels of copper, these transporters traffic to the plasma membrane to export copper to the extracellular milieu. The latter mechanism is how ATP7A works in intestinal absorption of copper, by sending the metal from the enterocyte into the portal venous circulation [16, 17]. The copper pool in blood is then bound to serum proteins such as albumin and α2‐macroglobulin [18] and directed to the liver.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

210

THE LIVER:  COPPER, THE LIVER, AND DISEASE

Figure 18.1  Copper metabolism in enterocytes and hepatocytes. (a) In enterocytes, copper (Cu+) uptake is mediated by CTR1, possibly in concert with the metalloreductases STEAP1, STEAP2, and dCYTB. The roles of CTR2 (not shown) and DMT1 in this process are less certain. Within enterocytes, GSH and MT function in copper sequestration and storage. The chaperones CCS, ATOX1, Cox17, Cox11, and Sco1 ferry copper to specific proteins or organelles. With an increase in copper levels, ATP7A traffics to the basolateral surface and pumps copper into the blood. (b) In hepatocytes, ATOX1 provides copper to ATP7B for metalation of ceruloplasmin (CP), and traffics to the apical membrane to pump copper into the bile, the body’s major mechanism for copper removal. CCS has been proposed to deliver copper to XIAP, which may interact with COMMD1, a protein mutated in hepatic copper toxicosis of Bedlington terriers and which may modulate ATP7B activity. This putative pathway is denoted by dashed lines. Modified from [3] with permission of Springer Nature.

Copper is imported through the basolateral membrane of hepatocytes by Ctr1 and DMT1 (Figure 18.1) [3]. For release into the bloodstream, copper needs to be bound to proteins, the main species in blood being ceruloplasmin, a secreted glycoprotein that acts as a ferroxidase and functions in systemic iron metabolism [19]. It is synthetized in the liver, and is initially devoid of copper (apoceruloplasmin) [20]. In the trans‐Golgi network of hepatocytes, copper is loaded onto apoceruloplasmin by ATP7B and subsequently secreted into the blood circulation. When liver copper levels increase, ATP7B trafficks to the apical membrane of hepatocytes, from where it excretes excess copper into the biliary canaliculi. Through biliary excretion, the liver is the major organ responsible for copper level regulation by elimination of the metal. Renal excretion of copper also occurs, and becomes clinically significant only in the presence of pathological mechanisms of copper overload from metabolic disorders [3], or in conditions resulting in biliary obstruction [21–24]. The liver is therefore the central organ in copper metabolism, which explains how dysfunction in one can impact the other. Disorders of copper metabolism manifest in the liver with clinical, laboratory, and pathological signs of hepatic dysfunction, as discussed below.

COPPER, THE LIVER, AND DISEASE Copper metabolism disorders involving the liver Wilson disease Wilson disease was first described in 1912 by Samuel Alexander Kinnier Wilson as a familial disorder associating liver cirrhosis with neurological manifestations, also referred to as

hepato‐lenticular degeneration [25]. It was not until three decades later that elevated copper levels and accumulation in the brain and liver were identified, which led to the classification of this disease as a disorder of copper metabolism [26]. This biochemical phenotyping facilitated discovery of the putative role of ATP7B in Wilson disease, supported by the similarities with ATP7A [27]. Wilson disease is an autosomal recessive disorder caused by a dysfunction of the copper transporter ATP7B, and more than 600 disease‐causing mutations have been identified to date [27, 28]. The distribution and frequency of specific mutations vary depending on the geographic region and population studied [29–32]. The types of amino acid changes most frequently found are missense mutations, small insertions or deletions, and splice site junction mutations [33]. Pathological variations in the sequence of this copper transporter cause copper overload and injury in the hepatocytes. Excessive copper eventually spills into the blood circulation and accumulates in several extrahepatic tissues, namely the brain, kidneys, and cornea, which helps explain the signs and symptoms of this disease [34]. Wilson disease classically presents in individuals ranging in age from 3 to 50 years, with a variable combination of neurological, psychiatric, and hepatic manifestations [35, 36]. Clinical signs and symptoms vary greatly among affected individuals, even within a single family [30]. Liver disease classically occurs in Wilson disease individuals who are symptomatic and diagnosed at an early age from childhood through early adulthood. Neurological and psychiatric manifestations are more frequent in older individuals, where hepatic involvement may be less pronounced [37]. In addition, very late onset (after age 70) and very early onset (nine months of age) of Wilson disease have been reported [38, 39]. Hepatic presentations can be both acute or chronic. Acute Wilsonian liver disease manifests as a rapidly occurring jaundice from a hepatitis‐like injury or a



18:  Copper Metabolism and the Liver

Coombs‐negative hemolytic disease and progresses to fulminant liver failure. The illness may also present as chronic liver disease, with or without cirrhosis. Especially since there are treatments available, Wilson disease should be considered in the differential diagnosis of any unexplained chronic liver disease or new‐onset neuropsychiatric symptoms. Psychiatric manifestations of Wilson disease include mood and personality disorders and sometimes cognitive deterioration. Neurological presentations are dominated by extrapyramidal symptoms including tremor, chorea, and choreoathetosis, problems in coordination and fine motor control, as well as rigidity and gait disturbance. From the ophthalmological standpoint, copper accumulation in the cornea may form a brown pathognomonic circle called the Kayser–Fleischer ring visible on slit lamp examination [30] (Figure  18.2). If clearly documented, this finding can confirm the diagnosis in a patient suspected of having Wilson disease. Confirmatory laboratory tests in Wilson disease include low levels of serum copper and ceruloplasmin that result from the inability of mutated ATP7B to load copper onto apoceruloplasmin, and increased 24‐hour urinary copper excretion. If liver biopsy is obtained during diagnostic workup of unexplained liver disease, hepatic copper levels are usually markedly elevated. A scoring system for the diagnosis of Wilson disease based on both clinical and laboratory criteria is available to guide clinicians faced with a possible case of this disorder [40]. Definitive confirmation is obtained by molecular analysis of the ATP7B gene. Treatment of Wilson disease should begin as soon as the diagnosis is made to prevent evolution to fulminant liver disease, cirrhosis, or irreversible neuropsychiatric damage. Therapy in symptomatic patients is based on the administration of copper chelators such as d‐penicillamine or trientine to prevent its accumulation in organs. Zinc salts can also be used as means to inhibit copper absorption in the intestines. Oral zinc induces

Figure 18.2  Kayser–Fleischer ring in the cornea of an adult patient with Wilson disease, resulting from copper deposition in Descemet’s membrane. Image courtesy of T.U. Hoogenraad, M.D., Department of Neurology, University Hospital, Utrecht, The Netherlands. Reproduced from Kaler, Wilson disease, in Cecil’s Textbook of Medicine, 23rd edn. (eds. L. Goldman and D. Ausiello), Saunders, Philadelphia, 2008, ch. 230, pp. 1593–5 with permission of Elsevier.

211

expression of metallothionein, an intracellular metal chelator that binds copper with an even higher affinity than zinc, and impairs its absorption. Pharmacotherapy using copper chelators is ideally continued as a lifelong treatment. However, side effects such as dermatological changes, hypersensitivity ­reactions, autoimmune disease, bone marrow suppression, and nephrotoxicity can be an obstacle for treatment adherence. Resistant cases that do not respond to pharmacological therapy and life‐threatening cases of Wilson disease may be treated with liver transplantation. A liver transplant that replaces the major dysfunctional organ is usually curative [41, 42]. A number of new therapies for Wilson disease are currently in preclinical stage studies. Targets are inhibition of ATP7B mutant retention and degradation in the endoplasmic reticulum, promotion of copper excretion [43, 44], and introduction of engineered hepatic cells with the hope of repopulating the liver [45]. Viral gene therapy for Wilson disease aims to introduce working copies of ATP7B by delivering the normal ATP7B complementary DNA (cDNA) to hepatocytes. This is accomplished by incorporating the normal gene’s cDNA in a nonpathogenic virus that targets the liver [46–48]. Gene therapy showed promise in a mouse model with an advanced stage of Wilson disease; liver enzymes and copper excretion were normalized [49]. This approach has potential for future application in clinical trials.

Huppke–Brendel syndrome Huppke–Brendel syndrome is a recently described autosomal recessive copper metabolism disorder with a biochemical phenotype similar to that of Wilson disease, but with different etiology and clinical manifestations. Patients with Huppke–Brendel syndrome are infants and children with low serum levels of copper and ceruloplasmin associated with syndromic features including congenital cataracts, hearing loss, and severe developmental delay [50]. On brain imaging, cerebellar atrophy, widened subarachnoid spaces, and hypomyelination are evident. This disorder is lethal in early childhood, with death reported between 22 months and 6 years of age in patients from the original cohort. Initially, several known copper transporters and proteins were investigated for their potential role in the pathogenesis of this syndrome, with no success [51]. Homozygosity mapping and deep sequencing later identified the genetic basis of the disorder, mutations in SLC33A1, a gene that encodes the acetyl‐ CoA transporter AT‐1 [52]. It is thought that AT‐1 is needed to acetylate glycoproteins and gangliosides in the endoplasmic reticulum and/or Golgi apparatus [53]. Copper metabolism imbalance in Huppke–Brendel patients could be secondary to the effects of a defective AT‐1 on trafficking of copper pumps ATP7A and ATP7B, since both proteins normally undergo acetylation of specific lysine residues (L. Yi and S.G. Kaler, unpublished data).

MEDNIK syndrome MEDNIK is an acronym used to describe an autosomal recessive neurocutaneous syndrome that was first defined in French‐ Canadian families sharing common ancestors [54]. MEDNIK stands for mental retardation, enteropathy, deafness, neuropathy, ichthyosis, and keratodermia. In addition to these clinical signs

212

THE LIVER:  CONCLUSIONS

and symptoms, dysfunction in copper metabolism was later identified in one of the index cases reported with this disease, and subsequently confirmed in the original cohort [55]. Patients with MEDNIK present a mixed biochemical phenotype with elements of both copper deficiency and copper excess. Developmental delays and brain atrophy found in MEDNIK subjects are reminiscent of Menkes disease, a neurodegenerative disorder secondary to mutations of ATP7A [3]. At the same time, MEDNIK patients show features of copper overload and liver disease with hypoceruloplasminemia, accumulation of copper in the liver, and intrahepatic cholestasis [55]. These changes are similar to those seen in Wilson disease caused by ATP7B mutations. The molecular basis of MEDNIK syndrome is mutation in the AP1S1 gene that encodes the sigma 1A (σ1A) subunit of adaptor protein complex 1 (AP‐1), a complex that normally mediates intracellular trafficking of certain transmembrane proteins between the trans‐Golgi network, endosomes, and plasma membrane of cells. The ATP7A and ATP7B copper ATPases are both transmembrane proteins with similar structures, and their function depends on their ability to translocate to different cellular compartments in response to changing cellular copper levels. In studies of ATP7A trafficking, the AP‐1 complex was implicated as a key player in this intracellular itinerary by interaction with a di‐leucine motif near the C‐terminus [56, 57]. This conclusion was extrapolated to ATP7B, given the similarities between these two ATPases, including the di‐leucine motif to which AP‐1 binds. With a defective σ1A subunit of the AP‐1 complex, ATP7A/B appear unable to remain docked in the trans‐Golgi network under normal or low copper states [57]. This defect in copper ATPase localization and trafficking may represent the link between AP1S1 mutations and the copper metabolism disturbances found in patients with MEDNIK. It seems conceivable that ATP7B function in the liver could be impacted more than ATP7A in other tissues, since two other isoforms of the σ1 subunit, σ1B and σ1C, exist and may substitute for σ1A, depending on their expression levels (L. Yi and S.G. Kaler, unpublished data). The cutaneous and other clinical features of MEDNIK presumably relate to impaired function of other transmembrane protein cargos. Zinc acetate therapy has been suggested and used to reduce copper overload in the liver of MEDNIK patients [55]. Further understanding of copper ATPase trafficking in MEDNIK subjects could shed light on the precise pathophysiology of the disorder, and allow the development of more precisely targeted therapies.

Primary disorders of the liver and copper imbalance Since the liver is the main regulatory organ of copper metabolism, the consequences of an imbalance in the metal’s levels are expected in hepatocytes, as previously described in diseases such as Wilson disease and MEDNIK. Conversely, evidence of copper level dysregulation is also present in primary hepatic diseases.

significant fat deposits accompanied by hepatocellular inflammation and necrosis (non‐alcoholic steatohepatitis or NASH). The latter can potentially progress to fibrosis and cirrhosis, end‐ stage liver disease, and hepatocellular carcinoma [58]. NAFLD is closely related to the metabolic syndrome, which encompasses abdominal obesity, hypertension, dyslipidemia, and impaired glucose tolerance or diabetes [59]. The pathophysiology of this disorder is complex and incompletely understood, with a possible role of copper in the process. Interestingly, a study of 124 patients with NAFLD showed reduced hepatic copper concentrations compared to controls and other liver diseases. This correlation was more pronounced in patients with advanced hepatic steatosis, non‐alcoholic steatohepatitis, and additional features of the metabolic syndrome [60]. Copper deficiency has previously been linked to atherogenic dyslipidemia, and a recent study of ATP7B showed additional data connecting copper and lipid metabolism at the enterocyte level [7, 61, 62]. Copper‐deficient diets in rats induced hepatic steatosis and insulin resistance, which further supports the role of low copper in the pathogenesis of NAFLD [60–62]. Additionally, beta‐oxidation occurs in peroxisomes and mitochondria, and the process is downregulated in states of copper deficiency [63]. Copper is needed for normal mitochondrial function, since it serves as a critical cofactor for cytochrome c oxidase, the last enzyme complex in the mitochondrial electron transport chain. Copper‐deficient mice show mitochondrial dysfunction [64], and similar alterations can be seen in humans with NAFLD. Finally, copper is also a cofactor of superoxide dismutase 1 (SOD1), an important enzyme for controlling oxidative stress, which has been linked to the hepatic damage in NAFLD and NASH [4, 60].

Hepatocellular carcinoma The role of copper in cancer development and progression has been widely studied, with several reports showing high serum and tumor copper levels in multiple cancer types [65–67]. High copper levels promote angiogenesis and oxidative stress generation, processes which create a suitable environment for tumor growth, and which could be counteracted by lowering copper levels [67–70]. These realizations led to the investigation of copper chelation as a potential anticancer therapy [71, 72], with applications in hepatocellular carcinoma (HCC). HCC is one of the cancers in which studies have documented high tumor copper levels [73–75]. Serum copper and ceruloplasmin levels were also found to be elevated in patients with HCC, and higher levels correlated with worse prognosis [76]. Interestingly, other types of hepatic neoplasms, including cholangiocarcinoma and metastatic liver disease, did not show significant copper level elevation, which helps distinguish them from HCC [77, 78]. The therapeutic strategy of lowering copper levels to treat cancer tested in vivo with HCC yielded promising results showing suppression of angiogenesis and tumor growth [72, 79]. Additional study of this therapeutic approach for other cancers is warranted.

Non‐alcoholic fatty liver disease Non‐alcoholic fatty liver disease (NAFLD) is the accumulation of fat in the liver with structural disorganization, in the absence of excessive alcohol consumption. The severity of steatosis and histological changes varies, ranging from benign steatosis to

CONCLUSIONS The liver is the main regulator of copper levels in the blood, through its role in the distribution and excretion of this trace



18:  Copper Metabolism and the Liver

metal. Hepatocytes process copper by directing it to various pathways, depending on the body’s needs and on intracellular levels. Copper is shuttled by the transporter ATP7B into the secretory compartment of hepatic cells, where it is incorporated into the serum glycoprotein ceruloplasmin. In conditions of excessive intracellular copper, the metal is excreted into the bile from the canalicular (apical) pole of hepatocytes. Defects in ATP7B function or trafficking lead to copper accumulation in the liver with consequent toxicity and hepatocellular injury. These mechanisms explain the liver manifestations seen in Wilson disease, Huppke– Brendel syndrome, and MEDNIK syndrome. Copper dyshomeostasis is also found in HCC and NAFLD, with implications for lipid metabolism, but the mechanisms underlying these conditions are incompletely understood. Further insights into copper metabolism at the cellular level, as well as carefully designed clinical trials involving novel copper chelators or viral gene therapy are needed to further advance understanding of these disease mechanisms and assess potential therapeutic remedies.

REFERENCES 1. Baker, Z.N., Cobine, P.A., and Leary, S.C. The mitochondrion: a central architect of copper homeostasis. Metallomics, 2017;15;9(11):1501–12. 2. Bousquet‐Moore, D., Mains, R.E., and Eipper, B.A. PAM and copper  –  a gene/nutrient interaction critical to nervous system function. J Neurosci Res, 2010;88(12):2535–45. 3. Kaler, S.G. ATP7A‐related copper transport diseases‐emerging concepts and future trends. Nat Rev Neurol, 2011;7(1):15–29. 4. McCord, J.M. and Fridovich, I. Superoxide dismutase an enzymic function for erythrocuprein (hemocuprein). J Biol Chem, 1969;244(22):6049–55. 5. Rucker, R.B., Kosonen, T., Clegg, M.S. et  al. Copper, lysyl oxidase, and extracellular matrix protein cross‐linking. Am J Clin Nutr, 1998;67(5 Suppl):996S–1002S. 6. Lutsenko, S. Human copper homeostasis: a network of interconnected pathways. Curr Opin Chem Biol, 2010;14(2):211–17. 7. Pierson, H., Muchenditsi, A., Kim B.‐E. et al. The function of ATPase copper transporter ATP7B in intestine. Gastroenterology, 2018;154(1):168–180.e5. 8. Zerounian, N.R., Redekosky, C., Malpe, R., and Linder, M.C. Regulation of copper absorption by copper availability in the Caco‐2 cell intestinal model. Am J Physiol Gastrointest Liver Physiol, 2003;284(5):G739–747. 9. Berghe, V.D., Ve, P., and Klomp, L.W. New developments in the regulation of intestinal copper absorption. Nutr Rev, 2009;67(11):658–72. 10. Nose, Y., Kim, B.‐E., and Thiele, D.J. Ctr1 drives intestinal copper absorption and is essential for growth, iron metabolism, and neonatal cardiac function. Cell Metab, 2006;4(3):235–44. 11. Lee, J., Peña, M.M.O., Nose, Y., and Thiele, D.J. Biochemical characterization of the human copper transporter Ctr1. J Biol Chem, 2002;277(6): 4380–7. 12. Garrick, M.D., Singleton, S.T., Vargas, F. et al. DMT1: which metals does it transport? Biol Res, 2006;39(1):79–85. 13. Lutsenko, S., Barnes, N.L., Bartee, M.Y., and Dmitriev, O.Y. Function and regulation of human copper‐transporting ATPases. Physiol Rev, 2007;87(3): 1011–46. 14. Kim, B.‐E., Turski, M.L., Nose, Y. et al. Cardiac copper deficiency activates a systemic signaling mechanism that communicates with the copper acquisition and storage organs. Cell Metab, 2010;11(5):353–63. 15. Linz, R. and Lutsenko, S. Copper‐transporting ATPases ATP7A and ATP7B: cousins, not twins. J Bioenerg Biomembr, 2007;39(5–6):403–7. 16. Monty, J.‐F., Llanos, R.M., Mercer, J.F.B., and Kramer, D.R. Copper exposure induces trafficking of the menkes protein in intestinal epithelium of ATP7A transgenic mice. J Nutr, 2005;135(12):2762–6. 17. Ravia, J.J., Stephen, R.M., Ghishan, F.K., and Collins, J.F. Menkes copper ATPase (Atp7a) is a novel metal‐responsive gene in rat duodenum, and immunoreactive protein is present on brush‐border and basolateral membrane domains. J Biol Chem, 2005;280(43):36221–7.

213

18. Liu, N., Lo, L.S., Askary, S.H. et al. Transcuprein is a macroglobulin regulated by copper and iron availability. J Nutr Biochem, 2007;18(9):597–608. 19. Hellman, N.E. and Gitlin, J.D. Ceruloplasmin metabolism and function. Annu Rev Nutr, 2002;22:439–58. 20. Hellman, N.E., Kono, S., Mancini, G.M. et  al. Mechanisms of Copper Incorporation into Human Ceruloplasmin. J Biol Chem, 2002;277(48): 46632–8. 21. Kuo, Y.M., Gitschier, J., and Packman, S. Developmental expression of the mouse mottled and toxic milk genes suggests distinct functions for the Menkes and Wilson disease copper transporters. Hum Mol Genet, 1997;6(7):1043–9. 22. Pyatskowit, J.W. and Prohaska, J.R. Copper deficient rats and mice both develop anemia but only rats have lower plasma and brain iron levels. Comp Biochem Physiol C Toxicol Pharmacol, 2008;147(3):316–23. 23. Schilsky, M.L. and Thiele, D.J. Copper metabolism and the liver, in The Liver (eds. I.M. Arias et  al.), Wiley‐Blackwell, Chichester, 2009, pp. 221–33. 24. Smallwood, R.A., Williams, H.A., Rosenoer, V.M., and Sherlock, S. Liver‐ copper levels in liver disease: studies using neutron activation analysis. Lancet, 1968;292(7582):1310–13. 25. Wilson, S.A.K. Progressive lenticular degeneration: a familial nervous disease associated with cirrhosis of the liver. Brain, 1912;34(4):295–507. 26. Cumings, J.N. The copper and iron content of brain and liver in the normal and in hepato‐lenticular degeneration. Brain, 1948;71(4):410–15. 27. Bull, P.C., Thomas, G.R., Rommens, J.M., Forbes, J.R., and Cox, D.W. The Wilson disease gene is a putative copper transporting P‐type ATPase similar to the Menkes gene. Nat Genet, 1993;5(4):327–37. 28. Bennett, J. and Hahn, S.H. Clinical molecular diagnosis of Wilson disease. Semin Liver Dis, 2011;31(3):233–8. 29. Ferenci, P. Regional distribution of mutations of the ATP7B gene in patients with Wilson disease: impact on genetic testing. Hum Genet, 2006;120(2):151–9. 30. Weiss, K.H. Wilson disease. GeneReviews [internet], 2016. https://www. ncbi.nlm.nih.gov/books/NBK1512/ (Accessed Oct 26, 2018). 31. Gomes, A. and Dedoussis, G.V. Geographic distribution of ATP7B mutations in Wilson disease. Ann Hum Biol, 2016;43(1):1–8. 32. Chang, I.J. and Hahn, S.H. The genetics of Wilson disease. Handb Clin Neurol, 2017;142:19–34. 33. Bandmann, O., Weiss, K.H., and Kaler, S.G. Wilson disease and other neurological copper disorders. Lancet Neurol, 2015;14(1):103–13. 34. Roberts, E.A. and Schilsky, M.L. A practice guideline on Wilson disease. Hepatology, 2003;37(6):1475–92. 35. Ferenci, P., Członkowska, A., Merle, U. et  al. Late‐onset Wilson disease. Gastroenterology, 2007;132(4):1294–8. 36. Kaler, S.G. Inborn errors of copper metabolism, in Handbook of Clinical Neurology (eds. O. Dulac, M. Lassonde, and H.B. Sarnat), Elsevier, New York, 2013, pp. 1745–54. 37. Machado, A., Chien, H.F., Deguti, M.M. et al. Neurological manifestations in Wilson disease: report of 119 cases. Mov Disord, 2006;21(12):2192–6. 38. Ala, A., Borjigin, J., Rochwarger, A., Schilsky, M. Wilson disease in septuagenarian siblings: raising the bar for diagnosis. Hepatology, 2005;41(3): 668–70. 39. Kim, J.W., Kim, J.H., Seo, J.K. et al. Genetically confirmed Wilson disease in a 9‐month old boy with elevations of aminotransferases. World J Hepatol, 2013;5(3):156–9. 40. Ferenci, P., Caca, K., Loudianos, G. et al. Diagnosis and phenotypic classification of Wilson disease. Liver Int, 2003;23(3):139–42. 41. Roberts, E.A. and Schilsky, M.L. Diagnosis and treatment of Wilson disease: an update. Hepatology, 2008;47(6):2089–111. 42. Schilsky, M.L. Liver transplantation for Wilson disease. Ann N Y Acad Sci, 2014;1315:45–9. 43. Kaler, S.G. Microbial peptide de‐coppers mitochondria: implications for Wilson disease. J Clin Invest, 2016;126(7):2412–4. 44. Lichtmannegger, J., Leitzinger, C., Wimmer, R. et  al. Methanobactin reverses acute liver failure in a rat model of Wilson disease. J Clin Invest, 2016;126(7):2721–35. 45. Ranucci, G., Polishchuck, R., and Iorio, R. Wilson disease: prospective developments towards new therapies. World J Gastroenterol, 2017;23(30): 5451–6. 46. Meng, Y., Miyoshi, I., Hirabayashi, M. et al. Restoration of copper metabolism and rescue of hepatic abnormalities in LEC rats, an animal model of Wilson

214

THE LIVER:  REFERENCES

disease, by expression of human ATP7B gene. Biochim Biophys Acta, 2004;1690(3):208–19. 47. Roybal, J.L., Endo, M., Radu, A. et al. Early gestational gene transfer with targeted ATP7B expression in the liver improves phenotype in a murine model of Wilson disease. Gene Ther, 2012;19(11):1085–94. 48. Murillo, O., Luqui, D.M., Gazquez, C. et al. Long‐term metabolic correction of Wilson disease in a murine model by gene therapy. J Hepatol, 2016;64(2):419–26. 49. Murillo, O., Moreno, D., Gazquez, C. et al. Improvement of gene therapy for Wilson disease. Mol Ther, 2016;24:S64. 50. Huppke, P., Brendel, C., Kalscheuer, V. et al. Mutations in SLC33A1 cause a lethal autosomal‐recessive disorder with congenital cataracts, hearing loss, and low serum copper and ceruloplasmin. Am J Hum Genet, 2012;90(1):61–8. 51. Horváth, R., Freisinger, P., Rubio, R. et  al. Congenital cataract, muscular hypotonia, developmental delay and sensorineural hearing loss associated with a defect in copper metabolism. J Inherit Metab Dis, 2005; 28(4):479–92. 52. Kanamori, A., Nakayama, J., Fukuda, M.N. et  al. Expression cloning and characterization of a cDNA encoding a novel membrane protein required for the formation of O‐acetylated ganglioside: a putative acetyl‐CoA transporter. Proc Natl Acad Sci U S A, 1997;94(7):2897–902. 53. Hirabayashi, Y., Nomura, K.H., and Nomura, K. The acetyl‐CoA transporter family SLC33. Mol Aspects Med, 2013;34(2–3):586–9. 54. Montpetit, A., Côté, S., Brustein, E. et al. Disruption of AP1S1, causing a novel neurocutaneous syndrome, perturbs development of the skin and spinal cord. PLoS Genet, 2008;4(12):e1000296. 55. Martinelli, D., Travaglini, L., Drouin, C.A. et  al. MEDNIK syndrome: a novel defect of copper metabolism treatable by zinc acetate therapy. Brain, 2013;136(Pt 3):872–81. 56. Holloway, Z.G., Velayos‐Baeza, A., Howell, G.J. et  al. Trafficking of the Menkes copper transporter ATP7A is regulated by clathrin‐, AP‐2‐, AP‐1‐, and Rab22‐dependent steps. Mol Biol Cell, 2013 Jun 1;24(11):1735–48. 57. Yi, L. and Kaler, S.G. Direct interactions of adaptor protein complexes 1 and 2 with the copper transporter ATP7A mediate its anterograde and retrograde trafficking. Hum Mol Genet, 2015;24(9):2411–25. 58. Adams, L.A., Angulo, P., and Lindor, K.D. Nonalcoholic fatty liver disease. CMAJ, 2005;172(7):899–905. 59. Greenfield, V., Cheung, O., and Sanyal, A.J. Recent advances in nonalcholic fatty liver disease. Curr Opin Gastroenterol, 2008;24(3):320–7. 60. Aigner, E., Strasser, M., Haufe, H. et al. A role for low hepatic copper concentrations in nonalcoholic fatty liver disease. Am J Gastroenterol, 2010;105(9):1978–85. 61. al‐Othman, A.A., Rosenstein, F., Lei, K.Y. Copper deficiency alters plasma pool size, percent composition and concentration of lipoprotein components in rats. J Nutr, 1992;122(6):1199–204.

62. al‐Othman, A.A., Rosenstein, F., and Lei, K.Y. Copper deficiency increases in vivo hepatic synthesis of fatty acids, triacylglycerols, and phospholipids in rats. Proc Soc Exp Biol Med, 1993;204(1):97–103. 63. Tosco, A., Fontanella, B., Danise, R. et al. Molecular bases of copper and iron deficiency‐associated dyslipidemia: a microarray analysis of the rat intestinal transcriptome. Genes Nutr, 2010;5(1):1–8. 64. Ibdah, J.A., Perlegas, P., Zhao, Y. et al. Mice heterozygous for a defect in mitochondrial trifunctional protein develop hepatic steatosis and insulin resistance. Gastroenterology, 2005;128(5):1381–90. 65. Apelgot, S., Coppey, J., Fromentin, A. et al. Altered distribution of copper (64Cu) in tumor‐bearing mice and rats. Anticancer Res, 1986;6(2):159–64. 66. Daniel, K.G., Harbach, R.H., Guida, W.C., and Dou, Q.P. Copper storage diseases: Menkes, Wilsons, and cancer. Front Biosci, 2004;9:2652–62. 67. Gupte, A. and Mumper, R.J. Elevated copper and oxidative stress in cancer cells as a target for cancer treatment. Cancer Treat Rev, 2009;35(1):32–46. 68. Brewer, G.J. Copper lowering therapy with tetrathiomolybdate as an antiangiogenic strategy in cancer. Curr Cancer Drug Targets, 2005;5(3):195–202. 69. Lowndes, S.A. and Harris, A.L. The role of copper in tumour angiogenesis. J Mammary Gland Biol Neoplasia, 2005;10(4):299–310. 70. Sproull, M., Brechbiel, M., and Camphausen, K. Antiangiogenic therapy through copper chelation. Expert Opin Ther Targets, 2003;7(3):405–9. 71. Goodman, V.L., Brewer, G.J., and Merajver, S.D. Copper deficiency as an anti‐cancer strategy. Endocr Relat Cancer, 2004;11(2):255–63. 72. Yoshii, J., Yoshiji, H., Kuriyama, S. et al. The copper‐chelating agent, trientine, suppresses tumor development and angiogenesis in the murine hepatocellular carcinoma cells. Int J Cancer, 2001;94(6):768–73. 73. Vecchio, F.M., Federico, F., and Dina, M.A. Copper and hepatocellular carcinoma. Digestion, 1986;35(2):109–14. 74. Ebara, M., Fukuda, H., Hatano, R. et al. Relationship between copper, zinc and metallothionein in hepatocellular carcinoma and its surrounding liver parenchyma. J Hepatol, 2000;33(3):415–22. 75. Ebara, M., Fukuda, H., Hatano, R. et al. Metal contents in the liver of patients with chronic liver disease caused by hepatitis C virus. Reference to hepatocellular carcinoma. Oncology, 2003;65(4):323–30. 76. Casaril, M., Capra, F., Marchiori, L. et al. Serum copper and ceruloplasmin in early and in advanced hepatocellular carcinoma: diagnostic and prognostic relevance. Tumori, 1989;75(5):498–502. 77. Haratake, J., Horie, A., Takeda, S. et al. Tissue copper content in primary and metastatic liver cancers. Acta Pathol Jpn, 1987;37(2):231–8. 78. Gurusamy, K. and Davidson, B.R. Trace element concentration in metastatic liver disease: a systematic review. J Trace Elem Med Biol, 2007;21(3):169–77. 79. Yoshiji, H., Kuriyama, S., Yoshii, J. et al. The copper‐chelating agent, trientine, attenuates liver enzyme‐altered preneoplastic lesions in rats by angiogenesis suppression. Oncol Rep, 2003;10(5):1369–73.

19

The Central Role of the Liver in Iron Storage and Regulation of Systemic Iron Homeostasis Tracey A. Rouault1, Victor R. Gordeuk2, and Gregory J. Anderson3 Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA 2 University of Illinois at Chicago, Chicago, IL, USA 3 QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia 1

THE LIVER AS A MAJOR IRON REPOSITORY In healthy adult humans, the liver is estimated to store from 0.07 to 0.4 g of iron. A combined analysis of previous studies using quantitative phlebotomy indicated that total body median storage iron was approximately 0.8 g among 39 normal men predominantly in the third and fourth decades of life, and 0.4 g among 20 normal women of similar age [1–5]. Based on the assumption that one‐third of iron stores are normally in the liver, this would translate to a normal median hepatic iron content of 0.27 g for men and 0.13 g for women. Iron is sequestered within the iron storage protein ferritin, a large 24‐subunit protein composed of ferritin L (light) and H (heavy) chains (Figure 19.1). The ferritin subunits combine to form a spherical protein shell that contains numerous channels. On the inside of the spherical protein, iron is oxidized from the ferrous (Fe2+) form to the ferric (Fe3+) form by the ferroxidase activity of the ferritin H chain, and insoluble ferric iron deposits grow from their initial deposition sites on side‐ chains of ferritin L subunits [6, 7]. As ferritin iron uptake and oxidation continue, thousands of ferric iron atoms accumulate within the ferritin sphere, which has the capacity to store up to 4500 iron atoms. As the iron contained within ferritin is not very bioavailable, ferritin sequesters iron and reduces the availability of cytosolic iron to interact with oxygen and generate harmful reactive oxygen species. Ferritin also serves as a source of iron for the metabolic needs of the cell. Iron can be released from intact ferritin that has been mono‐ubiquitinated and will be ultimately degraded by the proteasome [8].

Alternatively, ferritin multimers can undergo degradation in lysosomes to release iron. This process is guided by interaction of ferritin with NCOA4, which binds to ferritin and ­delivers it to the lysosome (Figure 19.1) [9].

SERUM TRANSFERRIN IS THE MAJOR SOURCE OF IRON FOR TISSUES Transferrin (TF) is a high‐affinity binding protein for ferric iron that circulates in the plasma with a concentration of 9–28 nmol L−1. It is synthesized mainly by hepatocytes and is secreted into the circulation. The plasma usually contains an excess of TF, such that less than one‐third of the iron‐binding sites on TF are occupied by iron. TF contains two lobes, each of which contains a ferric (Fe3+) iron‐binding site [11]. Cells in tissues throughout the body can obtain iron from circulating TF by increasing expression of transferrin receptors (TFRs). TFR1 is the major receptor for TF and binds diferric TF with high affinity (107–109 M−1) at physiological pH. Monoferric and apoTF are bound with much lower affinity (106 and q22) [11], whereas the BVRB‐encoding gene is on chromosome 19. The catalytic site for biliverdin reduction is located within the N‐terminal domain of both isozymes [12]. However, the C‐terminal domain of BVRA contains a basic‐leucine‐zipper (bZiP) domain to function as a transcription factor, which does not exist in BVRB. The C‐terminal domain of BVRA also contains both a nuclear localization sequence (NLS) and a nuclear export sequence (NES), enabling BVRA to bind to DNA sequences such as the antioxidant response element (ARE) and hypoxia response elements (HRE), thereby recruiting Nrf2. Nrf2 induces HO‐1, which protects against oxidative injury. Additionally, BVRA is a member of the insulin receptor substrate family with serine/threonine/tyrosine kinase activity and through its C‐terminal region interacts with both major arms of the insulin signaling pathway, namely the phosphatidylinositol 3‐kinase (PI3‐kinase)/Akt pathway and the IRK/IRS/PI3‐ kinase/MAPK pathway [10]. The evolutionary conservation of the energy‐consuming process of generation and excretion of the nonpolar bilirubin suggests that the stronger antioxidant activity of bilirubin may be particularly important during the neonatal period, when concentrations of other intracellularly available antioxidants are low in body fluids. Another potential advantage of bilirubin formation may be that, being more nonpolar, bilirubin is more efficiently extracted by the placenta in intrauterine life, although the mechanism of bilirubin extraction by the placenta from the fetal circulation has not been elucidated fully. Cord blood bilirubin concentrations of fetuses born to mothers who have unconjugated hyperbilirubinemia due to bilirubin glucuronidation deficiency are similar to the maternal serum bilirubin levels, suggesting that the placenta does not pose a barrier to equilibration of maternal and fetal serum unconjugated bilirubin [13]. As heme oxygenase acts specifically at the α‐bridge of heme, physiologically generated biliverdin and bilirubin are designated as biliverdin IXα and bilirubin IXα, respectively. The two

dipyrrolic halves are joined by a central methane bridge. Each half has a propionic acid side‐chain. X‐ray diffraction crystallography and nuclear magnetic resonance studies have revealed that the propionic acid side‐chains of bilirubin are internally hydrogen‐bonded to the pyrrolic and lactam sites on the opposite half of the molecule [14]. The internal hydrogen bonding engages all polar groups and “buries” the central methane bridge that joins the two dipyrrolic halves of the molecule. Physiologically, the hydrogen bonds can be disrupted by enzyme‐catalyzed conjugation of one or both propionic acid side‐chains, forming bilirubin mono‐ and diglucuronides, respectively. In the van den Bergh reaction [15] that is commonly employed in clinical analysis of serum bilirubin levels, the diazo reagents, which attack the central methane bridge, act rapidly on the non‐ hydrogen‐bonded conjugated bilirubin (“direct”‐reacting bilirubin), whereas the unconjugated fraction reacts rapidly only when the hydrogen bonds are disrupted by a chemical accelerator (“total” bilirubin). The hydrogen bonds can also be disrupted transiently by configurational isomerization of bilirubin, which occurs upon exposure to light. These bilirubin photoproducts can be excreted into bile without conjugation, which explains the efficacy of phototherapy in reducing serum unconjugated bilirubin levels.

POTENTIAL BENEFICIAL EFFECTS OF BILIRUBIN Although clinicians are generally concerned with serum bilirubin levels as an indicator of liver function and disease, and the toxicity of bilirubin, within a near‐physiological range of plasma concentrations, the antioxidative action of bilirubin may provide beneficial effects. Serum bilirubin levels were found to be inversely related to the risk of obesity and metabolic syndrome [16] and ischemic coronary artery disease in middle‐aged men [17]. An inverse relationship between serum bilirubin levels and cancer mortality was reported [18]. The study of a large number of subjects in the United States revealed that the odds ratio for history of colorectal cancer was reduced by 0.295 in men and 0.186 in women per 1 mg dL−1 increment in serum bilirubin concentrations [19]. Subjects with mildly elevated plasma bilirubin levels were found to have lower levels of abdominal obesity and reduced risk of metabolic syndrome. Consistent with this, obese individuals with elevated insulin and visceral adiposity had lower plasma bilirubin [20]. It should be noted that these convincing statistical associations do not, by themselves, establish a causative role of bilirubin in reducing the incidence of a number of common diseases in the population. Mechanistically, HO‐1 induction by cobalt‐protoporphyrin administration in leptin‐deficient ob/ob mice led to the recruitment of FGF21, PPARα, and Glut1, thereby reducing hepatic heme, body weight gain, plasma glucose, fatty acid synthase, and hepatic steatosis [21]. Many of these effects could be reproduced by bilirubin administration in mice with diet‐induced obesity or leptin receptor deficiency (db/db) [22].



20:  Disorders of Bilirubin Metabolism

HEPATIC METABOLISM AND ELIMINATION OF BILIRUBIN Bilirubin circulates in plasma bound to albumin. Albumin binding keeps unconjugated bilirubin in solution and prevents its diffusion into tissues and all its toxic effects. Unconjugated bilirubin binds more tightly to albumin than conjugated bilirubin. Therefore, in the absence of proteinuria, unconjugated bilirubin does not undergo glomerular filtration significantly. As albumin is normally present in approximately threefold molar excess to bilirubin, there is a significant reserve bilirubin binding capacity, which acts as a buffer for fluctuations of serum bilirubin levels. However, a number of metabolites and drugs affect albumin binding of bilirubin and thereby risk of neurotoxicity. Therefore, measurement of unbound plasma bilirubin and the reserve bilirubin binding capacity could provide a more accurate estimate of the risk of bilirubin‐induced neurological damage (BIND). This is particularly important in premature infants, in whom the threshold total plasma bilirubin concentrations used for instituting phototherapy and/or exchange transfusion in full‐term infants may be misleading. Directly measured free (unbound) serum bilirubin levels (Bf) are more sensitive and specific predictors of BIND, as evidenced by incidence of hearing defect than total serum bilirubin or bilirubin/albumin ratio [23]. Peroxidase treatment, gel chromatography, electrophoretic analysis, and direct fluorometry have been used for Bf determination [24], However, these approaches are not in clinical use, except for Bf determination by peroxidase treatment, which is used routinely in Japan [25]. Because of the tight binding of unconjugated bilirubin to albumin, bilirubin is not excreted in the urine in the absence of proteinuria. Conjugated bilirubin binds less tightly to albumin. Therefore, when conjugated bilirubin accumulates in plasma

1

Bilirubin glucuronides 6

3 GST’s

Contiguous membrane

5

7

2 Bilirubin UDPGA

4 UGT1A1

UDP

Bilirubin glucuronides ABCC2 (MRP2)

because of acquired or inherited liver diseases, a significant amount of conjugated bilirubin is excreted in urine. In the presence of prolonged accumulation of conjugated bilirubin in plasma, a fraction of the conjugated bilirubin becomes covalently bound to albumin. The covalently bound fraction, termed delta‐bilirubin, is not excreted in bile or urine, and may persist in plasma even after biliary obstruction or intrahepatic cholestasis is relieved. In the liver, bilirubin dissociates from albumin and is internalized by hepatocytes via facilitated diffusion. Although the transport protein SLC21A6 (OATP‐2) was implicated in bilirubin internalization, this could not be confirmed in subsequent studies [26]. Within the hepatocyte, binding of bilirubin to glutathione‐ S‐transferases (GSTs) inhibits its efflux, thereby increasing the net uptake. Microsomal uridinediphosphoglucuronate glucuronosyltransferase type 1 (UGT1A1) catalyzes the transfer of glucuronic acid from UDP‐glucuronate to bilirubin, forming mono‐ and diglucuronides. Glucuronidation makes bilirubin water soluble, reduces its toxicity, and promotes its secretion into bile. Glucuronide conjugation is critical in biliary excretion of bilirubin. Significant reduction of hepatic bilirubin glucuronidating activity results in the accumulation of unconjugated bilirubin in plasma (see later). Finally, the bilirubin glucuronides are transported into the bile canaliculi by an energy‐consuming process mediated by ABCC2 (also termed MRP2) (Figure 20.1). Among the many isoforms of uridinediphosphoglucuronate glucuronosyltransferase (UGT), UGT1A1 is the only one that contributes significantly to bilirubin glucuronidation. UGT1A1 [27] also accepts estradiol and several drugs as substrates. UGT1A1 is expressed from an unusually organized gene that expresses eight UGT1 isoforms from eight different promoters, each next to a unique exon encoding the N‐terminal half of a specific isoform. The transcript

Albumin + Bilirubin

Albumin-Bilirubin Sinusoidal surface

231

ABCC3

OATP01B1

OATP01B3

Bilirubin glucuronides ABCC2 (MRP2) Canalicular surface

Figure 20.1  Summary of hepatic metabolism of bilirubin. Bilirubin is strongly bound to albumin in the circulation (1). At the sinusoidal surface of the hepatocyte, this complex dissociates, and bilirubin enters hepatocytes by fascilitated diffusion (2). This process is non‐adenosine triphosphate (ATP)‐dependent and bidirectional. Within the hepatocyte, bilirubin binds to a group of cytosolic proteins, mainly to glutathione‐S‐transferases (GSTs) (3). GST binding inhibits the efflux of bilirubin from the cell, thereby increasing the net uptake. A specific form of uridine diphosphoglucuronate glucuronosyltransferase, UGT1A1, located in the endoplasmic reticulum, catalyzes the transfer of the glucuronic acid moiety from UDP‐ glucuronic acid (UDPGA) to bilirubin, forming bilirubin glucuronides (diglucuronide and monoglucuronide) (4). Glucuronidation is necessary for efficient excretion of bilirubin in bile. Canalicular excretion of bilirubin and other organic anions (except most bile acids) is primarily an energy‐ dependent process, mediated by the ATP‐utilizing transporter ABCC2, also known as multidrug‐resistance‐related proteins (MRP2) (5). Excess bilirubin glucuronides are pumped back into the plasma by ABCC3 located at the sinusoidal membrane (6) and undergo reuptake by hepatocytes located downstream to the portal blood flow via sinusoidal surface organic anion transport proteins, OATP01B1 and OATP01B3 (7).

232

THE LIVER:  DISORDERS OF BILIRUBIN METABOLISM

initiated from each promoter is spliced to four common region exons (exons 2–5) that encode the identical C‐terminal half of all UGT1A isoforms [28]. The presence of a different promoter for each isoform permits its independent regulation. Mutations in any of the five exons (1A1–5) encoding UGT1A1 can cause complete or partial deficiency of bilirubin glucuronidation, resulting in Crigler–Najjar syndrome type 1 (CN1) or Crigler–Najjar syndrome type 2 (CN2), which are characterized by hyperbilirubinemia with potential cerebral injury [29, 30]. Gilbert syndrome, a milder and much more common type of inherited hyperbilirubinemia, results from a variant TATA element within the UGT1A1 promoter, which leads to reduced synthesis of structurally normal UGT1A1 protein [31]. Finally, the bilirubin glucuronides are transported into the bile canaliculi against a steep concentration gradient by the ATP‐hydrolyzing canalicular pump ABCC2 (also termed MRP2), which is rate limiting in bilirubin throughput. Thus, efficient uptake and conjugation of bilirubin by the periportal (zone 1) hepatocytes that are initially exposed to bilirubin in portal blood may saturate the storage capacity of these cells. A portion of the bilirubin glucuronides produced in these cells is secreted into the sinusoidal plasma by the multispecific sinusoidal export pump MRP3 (ABCC3). Conjugated bilirubin flowing toward the central vein undergoes reuptake via the organic anion‐transporting polypeptides OATP1B1 (SLC01B1) and OATP1B3 (SLC01B3) located in hepatocyte sinusoidal membranes. This process recruits additional hepatocytes, thereby increasing the bilirubin‐excreting capacity of the liver [32]. Bilirubin is degraded by intestinal bacteria into a series of urobilinogen and related products. Most of the urobilinogen reabsorbed from the intestine is excreted in bile, but a small fraction is excreted in urine. Absence of urobilinogen in stool and urine indicates complete obstruction of the bile duct. In liver disease and states of increased bilirubin production, urinary urobilinogen excretion is increased. Urobilinogen is colorless; its oxidation product, urobilin, contributes to the color of normal urine and stool. Normally, the capacity of the liver for bilirubin uptake, conjugation, and excretion is closely balanced, so that reduction of any of these processes can limit the rate of bilirubin throughput by the liver. On the other hand, increase of the bilirubin handling capacity of the liver, for example, in response to increased bilirubin load, requires that all these pathways are coordinately upregulated. Several nuclear receptor proteins, such as CAR and PXR, may regulate such coordinated regulation [33–35].

DISORDERS OF BILIRUBIN METABOLISM At higher concentrations, unconjugated bilirubin is toxic to many cells and organelles. Since albumin binding abrogates the toxic effect of bilirubin, the harmful effects occur when unconjugated bilirubin is present in a molar excess over albumin. Such conditions are usually limited to neonates with a high level of

unconjugated hyperbilirubinemia and subjects with inherited disorders of bilirubin conjugation.

Neonatal hyperbilirubinemia Normally, newborns have higher serum bilirubin levels than adults. Eighty percent of all term infants exhibit clinical jaundice during the first 5 days of life. For clinical management, serum bilirubin levels in newborns need to be interpreted according to the infant’s age in hours. Bilirubin levels increase during the first few days of life, typically peaking at the age of 96 hours, when the 50th percentile in healthy newborns in Western countries ranges from 8 to 9 mg dL−1 and the 95th percentile is 15–17.5 mg dL−1 [36]. These levels are considered to be harmless. After this, serum bilirubin levels decline to less than 1 mg dL−1 in 7–10 days. Exaggeration of this physiological jaundice increases the risk of BIND. Neonatal hyperbilirubinemia results from a combination of increased bilirubin production and lower hepatic bilirubin excretory capacity. Exacerbation of these factors, with or without additional complicating disorders, increases BIND risk. In addition to prematurity, common risk factors for severe hyperbilirubinemia in babies born at 35 weeks or more gestation are exclusive breastfeeding (particularly with excessive weight loss), clinical jaundice noted within the first 24 hours, hemolytic diseases (e.g. glucose 6‐phosphate dehydrogenase (G6PD) deficiency), and cephalohematoma or significant bruising during birth [37]. Contribution of genetic factors is suggested by increased incidence of hyperbilirubinemia in infants of East Asian ethnicity and history of neonatal jaundice in a previous sibling [38]. Mechanisms of neonatal jaundice are briefly considered here.

Increased bilirubin production Bilirubin production, as measured by CO excretion in breath, is increased during the newborn period [39]. The excess bilirubin is derived from shortened erythrocyte half‐life and also from nonerythroid sources [40]. Mother–fetus Rh incompatibility has become infrequent since the availability of anti‐Rh immunoglobulins [41], but ABO blood group incompatibility continues to be a common cause of exaggerated neonatal hyperbilirubinemia. Other common hemolytic disorders include sickle cell disease, glucose 6‐phosphate dehydrogenase deficiency, hereditary spherocytosis, and toxic or allergic drug reactions. Ineffective erythropoiesis, as in thalassemia, vitamin B12 deficiency, and congenital dyserythropoietic anemias can also result in excessive bilirubin production.

Low hepatic bilirubin uptake Low uptake rate during the first few days of life may result from delayed closure of the ductus venosus and low levels of cytosolic GSTs [42], which increase net bilirubin uptake by reducing its efflux.

Reduced bilirubin glucuronidation Both full‐term and premature infants are born with approximately 1% of adult hepatic UGT1A1 activity, which increases to adult levels by 14 weeks, regardless of the gestational age at birth [43]. Inhibition of UGT1A1 exacerbates and prolongs ­neonatal jaundice.



20:  Disorders of Bilirubin Metabolism

Maternal milk jaundice In general, breastfed infants have higher serum bilirubin levels than formula‐fed babies [44]. Mild maternal milk jaundice may abate despite continuation of breastfeeding and even in more severe cases, resolves promptly upon discontinuation of breastfeeding, If breastfeeding is continued, the hyperbilirubinemia may persist for weeks and, in some cases, may increase to 15–24 mg dL−1 by the age of 10–19 days. Although maternal milk jaundice is usually benign [45], kernicterus can occur in rare cases  [46]. The mechanism of maternal milk jaundice is not clear. Polyunsaturated free fatty acids produced by lipolytic enzymes in some maternal milk samples may be responsible for the inhibition of bilirubin glucuronidation, as suggested by increased inhibitory effect of maternal milk upon storage and marked reduction of the inhibitory effect by heating the milk to 56°C [47].

Maternal serum jaundice Lucey and associates [48] described a syndrome manifested by moderate to severe unconjugated hyperbilirubinemia (8.9–65 mg dL−1) within the first 4 days of life, which is thought to be caused by an unidentified inhibitor of UGT1A1 present in maternal serum. Jaundice may persist several weeks and is occasionally associated with kernicterus.

Delayed maturation of canalicular bilirubin excretion Maturation of canalicular excretion mechanism may lag behind the maturation of uptake and conjugation, so that in the late newborn period, canalicular excretion becomes rate limiting in hepatic bilirubin throughput. In these cases, conjugated bilirubin may accumulate in serum [49].

Increased intestinal reabsorption Deconjugation of bilirubin by intestinal β‐glucuronidase releases unconjugated bilirubin, which is not further degraded because of the absence of an established intestinal microbiota in the newborn. This results in increased bilirubin absorption [50], which may be enhanced by feeding maternal milk.

233

common benign disorder called Gilbert syndrome, which is characterized by a mild, fluctuating, and often intermittent unconjugated hyperbilirubinemia.

Crigler–Najjar syndrome type 1 This rare syndrome with autosomal recessive inheritence was described in 1952 by Crigler and Najjar in six infants from three unrelated families [51] and was later found to result from lack of UGT1A1 activity [52]. All patients had lifelong severe nonhemolytic unconjugated hyperbilirubinemia resulting in bilirubin‐ induced encephalopathy and death within 15 months in five of the six originally reported cases. The remaining patient developed kernicterus for the first time at the age of 15 years, and died six months after that [52]. A related patient remained without brain damage until 18 years of age, but then developed kernicterus and died at the age of 24 [53]. As the original families described by Crigler and Najjar had a high degree of consanguinity, several other recessively inherited disorders, ­ including Morquio syndrome, homocystinuria, metachromatic leukodystrophy, and bird‐headed dwarfism existed in these families. However, in other families CN1 was not associated with additional inherited disorders. Several hundred CN1 patients were described subsequently in all races. After Arias reported a milder variant of this condition (see below under Crigler–Najjar syndrome type 2), the original potentially lethal Crigler–Najjar syndrome was designated Crigler–Najjar syndrome type 1 (CN1), whereas the milder variant of the disease was termed Crigler–Najjar syndrome type 2 (CN2). Jaundice is often the only clinical finding, although some patients may have residual neurologic abnormalities, from previous episodes of bilirubin encephalopathy. With routine use of phototherapy and plasmapheresis during acute bilirubin encephalopathy, many patients with CN1 now survive beyond childhood, but many survivors develop kernicterus around puberty or in early adult life [54]. Orthotopic or auxilliary liver transplantation results in normalization of serum bilirubin. Because of the relatively high concentration of unconjugated bilirubin excreted in bile as a result of phototherapy, pigment gallstones are common.

Laboratory tests

DISORDERS ASSOCIATED WITH UNCONJUGATED HYPERBILIRUBINEMIA Predominantly unconjugated hyperbilirubia can result from increased bilirubin production due to hemolytic disorders or ineffective erythropoiesis. In the presence of normal liver function, increased bilirubin production does not lead to plasma levels of bilirubin exceeding 5 mg dL−1. Three inherited disorders associated with UGT1A1 deficiency and consequent reduction of bilirubin glucuronidation have been described. A near‐complete deficiency of UGT1A1 activity results in Crigler–Najjar syndrome type 1 (CN1); severe but incomplete deficiency of UGT1A1 activity is termed Crigler–Najjar syndrome type 2 (CN2), also known as Arias syndrome; and a mild reduction of UGT1A1 activity results in a

Serum bilirubin levels usually range from 20 to 25 mg dL−1, but may reach 50 mg dL−1 [51, 52]. Serum bilirubin is all unconjugated and tightly bound to albumin, therefore bilirubinuria is absent. The bile contains only small amounts of unconjugated bilirubin [55]. Although fecal urobilinogen excretion is reduced, the stool color remains normal [51]. Normal bile canalicular transport is evidenced by normal plasma clearance of bromosulfophthalein and indocyanin green, as well as visualization of the biliary tree by cholecystographic agents [51].

Liver histology Historically, liver histology has been reported as normal other than bilirubin plugs in bile canaliculi and bile ducts [51, 55]. However, a recent systematic analysis of 22 CN1 cases undergoing liver transplantation at a single center showed liver fibrosis of various degrees in 41% of the explanted livers [56]. The

234

THE LIVER:  DISORDERS ASSOCIATED WITH UNCONJUGATED HYPERBILIRUBINEMIA

liver fibrosis was not associated with portal hypertension and there was no significant correlation with gallstones. The transplant recipients with liver fibrosis were older in age, suggesting an accrual of the risk of fibrosis with age.

Abnormalities of hepatic UGTs Hepatic UGT activity toward bilirubin is virtually absent in all patients with CN1. In addition, many of these patients have reduced glucuronidation of phenolic substrates [57], which can be explained by the location of mutation in UGT1A1 exons (see later). By immunological analysis, expression of UGT isoforms may vary in the livers of different CN1 patients [58].

Molecular bases of CN1 and CN2 Description of the layout of the UGT1A locus in 1992 by Ritter et al. [59] and Bosma et al. [60] explained the structural characteristics and different substrate specificities of various isoforms of the UGT1A family. Each isoform has a unique N‐terminal domain but they all have identical C‐terminal domains. Processed UGT1 mRNAs consist of five exons. The UGT1A locus comprises four exons encoding the common C‐terminal domain of all UGT1A isoforms that contains the binding site of the glucuronic acid donor substrate uridine diphosphoglucuronic acid and the single transmembrane region of the protein. Twelve unique exon 1 sequences are located 5′ to these common region exons, only one of which is used in a given UGT1A isoform, encoding its unique N‐terminal domain that imparts aglycone substrate specificity to specific isoforms. Each unique exon 1 has a separate proximal upstream promoter, so that depending on which promoter is used for transcription, RNA transcripts of different lengths are produced. In each transcript only the exon 1 located at the 5′ end of the transcript is spliced to the common region exon 2, and all other exon 1’s are treated as intervening sequences and deleted. This results in the translation of nine UGT1 isoforms, of which only UGT1A1 contributes significantly to bilirubin glucuronidation [61]. Consequently, genetic lesions within any of the five exons comprising the UGT1A1 gene may lead to complete or near‐complete loss of hepatic bilirubin glucuronidation. Such genetic lesions may consist of point mutations, deletions, or insertions within the coding region or introns at splice donor or acceptor sites [29, 54]. The genetic lesions may result in mutation of a single critical amino acid or deletion of segments of the enzyme. The various genetic lesions described in the literature have been reviewed [29, 62]. In cases where the genetic lesions are present in exons 2–5, all isoforms expressed from the UGT1A locus are affected, whereas mutation located in the unique first exon of UGT1A1 affects the gluronidation of only the substrates of this isoform. As CN1 or CN2 can be caused by any of a large number of mutations, deletions, or insertions, no particular mutation is very common in any ethnic group or community. An exception to this is seen where there is a strong founder effect and a high level of consanguinity, such as in the Amish–Mennonite community, in which there is a high incidence of CN1 and all people with CN1 carry a specific nonsense mutation in exon 1 of UGT1A1 [63]. In nearly all cases molecular diagnosis of CN1 and CN2 can be made by sequence determination of products of

polymerase chain reaction (PCR) amplification of the five UGT1A1 exons and their flanking splice sites. The same strategy can be used for prenatal diagnosis by analyzing DNA extracted from amniotic cells or chorionic villus samples [64]. As in other recessively inherited inherited disorders, in nearly all CN1 and CN2 cases, one mutant allele is derived from each heterozygous parent, giving rise to homozygous or compound homozygous states. However, a person with CN1 with uniparental isodisomy has been reported, in whom both mutant alleles were inherited from the father [57]. The mother’s UGT1A1 genotype was normal. This case highlights the desirability of analyzing the genotype of both parents to determine the mode of inheritance of CN syndromes.

Animal models of CN1 A mutant strain of Wistar rats exhibiting lifelong nonhemolytic unconjugated hyperbilirubinemia inherited as an autosomal recessive characteristic was reported by Gunn in 1938 [65]. Subsequently, the cause of jaundice was found to be deficiency of UGT‐mediated glucuronidation of bilirubin. Jaundiced Gunn rats are homozygous for the deletion of a single guanosine residue in the common region exon 4, causing a frameshift and a premature termination codon that results in the expression of a truncated protein lacking 150 amino acid residues at the C‐ terminus. This results in the loss of activity of all UGT isoforms expressed from the UGT1 locus, but UGT isoforms expressed from other loci (e.g. UGT2) are not affected [66]. Experiments using Gunn rats have provided important information on bilirubin toxicity and have helped in developing novel cell and gene‐ based therapies of CN1 [67–69]. A mouse knockout model of CN1 was created by disrupting exon 4 of the mouse UGT1 locus [70]. The knockout mice have higher bilirubin levels than Gunn rats, and have a high spontaneous mortality rate unless treated with intensive phototherapy. The UGT1‐KO mice have enabled the pathophysiological study of BIND and development of novel therapeutic strategies.

Treatment of CN1 Management of CN1 centers on maintaining serum bilirubin concentrations below neurotoxic levels. Partial or whole liver transplantation cures the disease, but commits the pateint to prolonged immunosuppressive therapy. Therapies based on hepatocyte transplantation and gene therapy are still considered experimental. A brief discussion of these treatment modalities follows.

Phototherapy Phototherapy is routinely used to reduce the level of plasma unconjugated bilirubin [55]. Banks of fluorescent lamps, or more recently light‐emitting diode (LED) lamps, are used with devices for shielding the eyes. LED “light blankets,” or “light jackets,” have been also devised. Light converts bilirubin IXα‐ ZZ to its photoisomers (see the section on Bilirubin chemistry), which are excreted in bile and partly degraded. During the neonatal period, use of phototherapy is based on age‐related serum bilirubin concentrations: 15 mg dL−1 (260 μM) at age 24–48 hours, 18 mg dL−1 (310 μM) at 49–72 hours, and 20 mg dL−1



20:  Disorders of Bilirubin Metabolism

(340 μM) above 72 hours [24]. If serum bilirubin remains above these levels and is not reduced by at least 1–2 mg dL−1 within 4–6 hours despite intensive phototherapy, plasmapheresis is considered. The efficiency of phototherapy is reduced beyond the age of 3 or 4 years, because of the thickening of skin, pigmentation, and reduction of surface area relative to body mass, requiring readjustment of photherapy intensity and duration.

Plasmapheresis During neurologic emergencies, serum bilirubin concentration can be acutely reduced by plasmapheresis [55]. If serum bilirubin levels exceed the target levels described in the previous paragraph by 5 mg dL−1, despite intensive phototherapy, plasmapheresis is added to continued intensive photherapy, because after removal of albumin‐bound bilirubin from blood, bilirubin is mobilized from tissue stores to the plasma, leading to secondary increase of bilirubin levels.

Orthotopic liver transplantation At present, the transplantation of whole liver or a segment of the liver is the only available curative therapy for CN1 [71]. Although this procedure commits the patient to prolonged immunosuppressive therapy, liver transplantation has dramatically improved the outlook for CN1 patients.

235

disorders [75]. Because of this and the increasing shortage of good‐quality donor livers for hepatocyte isolation, novel strategies are being explored to induce preferential proliferation of transplanted normal hepatocytes over the mutant host cells. As adult hepatocytes retain a remarkable capacity to proliferate and the liver to body weight ratio is regulated tightly by physiological mechanisms, transplanted hepatocytes must compete with host liver hepatocytes for preferential proliferation. Controlled regional irradiation of the liver, in combination with a variety of mitotic stimuli can enhance both initial hepatocyte engraftment in Gunn rats and subsequent proliferation of the donor cells, leading to normalization of serum bilirubin levels [76]. Following initial evaluation in non‐human primates [77], a clinical trial has been initiated to evaluate preparative hepatic irradiation for hepatocyte transplantation in patients with inherited metabolic disorders of the liver [78]. Hepatocyte‐like cells (iHep) generated by differentiating human‐induced pluripotent stem cells derived by reprogramming somatic cells (e.g. skin fibroblasts, bone marrow cells, peripheral blood mononuclear cells, or epithelial cells shed in the urine) can be transplanted into the livers of experimental animals. Transplantation of human iHep cells into immunosuppressed Gunn rats subjected to preparative X‐irradiation resulted in significant reduction of serum bilirubin levels [69].

Gene therapy Experimental methods for reducing serum bilirubin levels Inhibiting heme oxygenase activity Non‐iron metalloporphyrins are dead‐end inhibitors of microsomal heme oxygenase [72]. Tin‐protoporphyrin injection suppresses neonatal hyperbilirubinemia in rhesus monkeys [73]. Injection of tin‐mesoporphyrin (0.5 μmol kg−1, three times a week for 13–23 weeks) in two 17‐year‐old male CN1 patients reduced serum bilirubin concentrations modestly. However, the  duration and safety of this treatment of CN1 are not yet established.

Hepatocyte transplantation As UGT1A1 activity is present in excess in normal liver, partial replacement of the enzyme activity in livers of CN1 patients should reduce serum bilirubin to nontoxic levels. After extensive validation in Gunn rats [68], isolated allogeneic human hepatocytes were transplanted into the liver of an adolescent CN1 patient through a catheter placed percutaneously into the portal vein [74]. Transplantation of 7.5 × 109 hepatocytes reduced bilirubin levels by about 50% and enabled reduction of the duration of phototherapy [74]. However, although bilirubin glucuronides were detectable in bile for up to two and a half years, serum bilirubin level gradually increased to pretransplantation levels. The patient received an auxiliary liver transplantation, which rapidly reduced the serum bilirubin levels to normal (J. Roy Chowdhury, personal communication). Experience in this case and a number of other cases of hepatocyte transplantation for CN1, as well as for a number of other inherited liver diseases indicated that the number of adult hepatocytes that can be transplanted at a single procedure is insufficient for fully curing inherited liver‐based metabolic

Gene therapy approaches aimed at reconstitution of the missing UHT1A1 activity include: (i) ex vivo gene therapy, which consists of transduction of primary hepatocytes using viral vectors, followed by transplantation into the liver; (ii) systemic administration of viral or nonviral vectors to transduce hepatocytes in vivo with transcription units expressing UGT1A1; and (iii) targeted gene editing through homologous recombination for correction of a specific mutation in the UGT1A1 gene, or for insertion of a UGT1A1 transcription unit at a genomic “safe haven” site of choice, or for targeted insertion of a UGT1A1 open reading frame downstream to a highly expressed gene (e.g. albumin) to take advantage of the strong endogenous promoter. After validation in preclinical experiments, a clinical trial has been initiated in CN1 patients using recombinant adeno‐associated viral vectors.

CN2 (Arias syndrome) In this milder variant of Crigler–Najjar syndrome described by Arias in 1962 [79], serum bilirubin, which is mostly unconjugated, usually ranges from 8 to 18 mg dL−1. During intercurrent illness, general anesthesia, or prolonged fasting, serum bilirubin can increase to 40 mg dL−1 [80]. Kernicterus is unusual, but has been reported during episodes of exacerbated hyperbilirubinemia [79–81]. As in CN1, there is no evidence of hemolysis or other liver dysfunction. CN2 can be clinically differentiated from CN1 by at least 25% decrease of serum bilirubin after induction of the residual UGT1A1 activity by administration of drugs, such as phenobarbital. Also in contrast to CN1, bile contains a significant amount of bilirubin glucuronides. In CN2, as in Gilbert syndrome (see later), bilirubin monoglucuronide exceeds 30% of total conjugated bilirubin (normal, ~10%),

236

THE LIVER:  DISORDERS ASSOCIATED WITH UNCONJUGATED HYPERBILIRUBINEMIA

reflecting a reduced hepatic UGT1A1 activity. Liver histology is normal, and UGT1A1 activity is usually reduced to 10% normal [82].

Genetic lesions Molecular genetic studies are consistent with autosomal recessive inheritance [83]. In CN2, UGT1A1 coding region mutations always result in single amino acid transitions that significantly reduce the UGT1A1 activity, without completely abolishing it. In some cases, the mutation has been shown to increase the Km for bilirubin [54]. The various lesions causing CN2 have been reviewed [29].

Gilbert syndrome Gilbert and Lereboullet described this common disorder, associated with a mild, fluctuating unconjugated hyperbilirubinemia in 1901 [84]. More than a century later, Bosma and associates discovered that this syndrome is caused by a polymorphism in the proximal promoter region that reduces the expression of UGT1A1 [85]. Gilbert syndrome is usually diagnosed in young adults during blood tests for routine check up, screening, or investigation of unrelated illnesses. In many cases, hyperbilirubinemia is intermittent and is usually less than 3 mg dL−1. Bilirubin levels increase during intercurrent illness, stress, fasting, or menstruation [86]. Mild icterus is the only positive clinical finding, and predominantly unconjugated hyperbilirubinemia is the only abnormality found on routine blood tests. Some patients report fatigue and abdominal discomfort, which may be manifestations of anxiety or superimposed illnesses. Oral cholecystrography visualizes the gallbladder normally, but there may be a higher incidence of gallstones. Liver biopsy is not necessary for diagnosis, but when performed, shows normal histology, except for some nonspecific lipofuscin accumulation in the centrilobular zone.

Incidence Gilbert syndrome is one of the most common inherited disorder in humans, the reported incidence ranging from 3% to 7% of most populations [87]. As males have higher average serum bilirubin levels than females, Gilbert syndrome is diagnosed more frequency in males [87]. Gilbert syndrome is often recognized around puberty, which may be related to increased red cell mass and consequent increased bilirubin production, as well as inhibition of bilirubin glucuronidation by endogenous steroid hormones.

Diagnosis Gilbert syndrome is diagnosed in individuals with mild unconjugated hyperbilirubinemia in the absence of normal liver serology and without evidence of hemolysis. Hepatic UGT1A1 activity is consistently low (around 30% of normal) in Gilbert syndrome [88]. Although hemolysis is not a feature of Gilbert syndrome, coexistent hemolytic disorders, such as glucose 6‐ phosphate dehydrogenase deficiency can make jaundice clinically obvious, thereby bringing the patient to the attention of a physician. Sequence determination of the TATAA element

within the proximal promoter region of UGT1A1 provides a simple means of diagnosis (see section on Genetic basis of Gilbert syndrome). If confirmation of diagnosis is required, chromatographic analysis of bile for determination of the bilirubin monoglucuronide to diglucuronide ratio can be used. As in CN2, bile in Gilbert syndrome contains an increased proportion (over 10%) of bilirubin monoglucuronide [82], reflecting reduced hepatic UGT1A1 activity in these syndromes. Reduction of caloric intake to 400 kcal day−1 for 2 days or nicotinic acid administration [89] increase serum bilirubin levels in Gilbert syndrome as well as in normal subjects, therefore these tests do not provide definitive diagnosis. Needle biopsy of the liver is not recommended for the diagnosis.

Genetic basis of Gilbert syndrome Gilbert syndrome is associated with a varient TATAA box in the promoter upstream to exon 1 of UGT1A1. The normal TATAA element has the sequence A[TA]6TAA, whereas subjects with Gilbert syndrome are homozygous for insertion of two nucleotides, making its sequence T[TA]7TAA, which reduces the expression of UGT1A1 to approximately 30% of normal [85]. The Gilbert‐type TATAA element has been designated as UGT1A1*28 [90]. Approximately 9% of most populations are homozygous for UGT1A1*28 and about 42% are heterozygous carriers. However, all subjects who are homozygous for the UGT1A1*28 allele do not exhibit the clinical phenotype, indicating that other variables, particularly the rate of bilirubin production may be necessary for manifestation of hyperbilirubinemia. Thus, although the inheritance is autosomal (chromosome 2q37), jaundice is uncomon in women, probably because of lower daily bilirubin production. The gene frequency for UGT1A1*28 may be lower in Japan. Some UGT1A1 coding region mutations may cause mild hyperbilirubinemia, compatible with the clinical diagnosis of Gilbert syndrome [91, 92]. These mutations have been reported only in populations of East Asian origin. Some of these mutations were rported to be dominant negative, suggesting that they reduce the activity of a normal allele [93]. Some heterozygous carriers of structural mutations (CN1 or CN2) may also carry the Gilbert‐type promoter on the structurally normal allele, which reduces the expression of the only structurally normal allele. Such combinations may give rise to intermediate levels of hyperbilirubinemia [94, 95], explaining the long‐standing observation that intermediate levels of hyperbilirubinemia are common among relatives of patients with CN1 or CN2. Because of the rarity of CN1 and CN2 mutations and the very high frequency of the UGT1A1*28 allele, this type of combination is a more common cause of intermediate levels of hyperbilirubinemia than the presence of a coding region mutation on both UGT1A1 alleles.

Health implications of Gilbert syndrome Gilbert syndrome is innocuous, but its recognition is considered important mainly for reassuring the patient and the physician that no underlying liver disease is responsible for the mild hyperbilirubinemia. In fact, mild elevation of serum bilirubin may have health benefits (see section on Potential beneficial effects of bilirubin). However, reduced UGT1A1 activity can



20:  Disorders of Bilirubin Metabolism

affect detoxification of certain drugs [96]. Gilbert syndrome is associated with a high incidence of diarrhea in patients treated with the anticancer drug irinotecan [97]. Oxidative paracetamol metabolism is associated with drug toxicity. There is conflicting evidence in literature for increased oxidative metabolism of paracetamol (acetaminophen) and its decreased glucuronidation in subjects with the UGT1A1*28 allele [98, 99]. Nilotinib and probably other tyrosine kinase inhibitors used in the treatment of leukemia are not metabolized by UGT1A1, but may inhibit the enzyme activity, thereby enhancing hyperbilirubinemia in patients with Gilbert syndrome, as well as CN2 [100].

Animal model The Bolivian population of squirrel monkeys (Saimiri sciureus) has higher serum unconjugated bilirubin concentrations and a greater hyperbilirubinemic response to fasting than does a closely related Brazilian population [101, 102]. Plasma clearance of intravenously administered bilirubin is lower in the Bolivian squirrel monkey population. As in subjects with Gilbert syndrome, hepatic UGT activity toward bilirubin, and bilirubin diglucuronide:monoglucuronide ratio in bile are lower in the Bolivian population. Fasting hyperbilirubinemia is rapidly reversed by oral or intravenous administration of carbohydrates, but not lipids [101].

DISORDERS PREDOMINANTLY ASSOCIATED WITH CONJUGATED HYPERBILIRUBINEMIA Normally, by chromatographic measurement, approximately 4% of bilirubin in plasma is conjugated. The proportion of conjugated bilirubin increases in plasma when bilirubin glucuronides formed inside hepatocytes are transferred back to plasma as a result of biliary obstruction, inflammatory or ischemic injury of the liver, inherited disorders of bile canalicular excretion, or failure of reuptake of bilirubin glucuronides transported out of the hepatocytes into sinusoidal blood. In addition, there are several disorders, collectively termed progressive familial intrahepatic cholestasis, that are caused by inherited abnormalities of bile canalicular transport proteins or a tight junction protein. In another genetic condition, paucity of bile ducts leads to conjugated hyperbilirubinemia. A brief description of these disorders follows.

Dubin–Johnson syndrome Dubin and Johnson [103] and Sprinz and Nelson [104] described a syndrome characterized by chronic conjugated hyperbilirubinemia and grossly pigmented, but otherwise histologically normal liver. Mild icterus is the only consistent physical finding, but rarely, hepatosplenomegaly has been observed [105, 106]. Patients are usually asymptomatic, but an occasional patient complains of weakness and vague abdominal pain. Serum bile acid levels are nearly normal [105], and pruritus is absent. Serum bilirubin levels are increased during intercurrent illness, intake of oral contraceptives, and pregnancy [105]. Diagnosis is

237

usually made after puberty and, in some cases, during pregnancy or contraceptive use [105, 107].

Laboratory tests Liver function tests, including serum bile acid levels, are normal [107]. Serum bilirubin levels usually range from 2 to 5 mg dL−1, but can rarely reach 20–25 mg dL−1. Over 50% of total serum bilirubin is direct‐reacting, and bilirubin is excreted in urine. Because the hepatocellular canalicular excretion is abnormal for many organic anions, except bile acids, oral cholecystography, even using a “double dose” of the contrast material, fails to visualize the gallbladder. Gallbladder may be visualized 4 to 6 hours after intravenous administration of meglumine iodipamide (Biligrafin) [108]. Macroscopically, the liver is black, and light microscopy reveals a dense pigment. After intravenous infusion of 3H‐epinephrine into mutant Corriedale sheep (an animal model for Dubin–Johnson syndrome), radioactivity is incorporated into the dark brown pigment [109], which is not authentic melanin, but may consist of polymers of epinephrine metabolites [110]. Following liver regeneration after hepatocellular apoptosis, such as after acute viral hepatitis, the pigment is cleared from the liver and reaccumulates slowly after recovery.

Organic anion transport Organic anions other than bile acids are transported into the bile canaliculus from the hepatocyte against a concentration gradient by an ATP‐dependent energy‐consuming process, mediated by a protein that had been originally termed the canalicular multispecific organic anion transporter (cMOAT), and is now termed MRP2 or ABCC2 [111–114]. These anions include bilirubin glucuronides, the leukotriene LTC4, oxidized and reduced glutathione, and numerous glucuronide and glutathione conjugates. In contrast, with some exceptions, bile acids are secreted normally. MRP2 is one of the ABC transporters [112]. Direct evidence for its involvement in canalicular transport came from the discovery of a frameshift mutation in the gene encoding Mrp2 in the TR− rat [115]. Studies in TR− mice indicated that some sulfated or glucuronidated bile acids require Mrp2 for biliary excretion [116]. Consistent with this, in a multicenter study in Japan, neonates with Dubin–Johnson syndrome had significantly increased serum total bile acid levels [117]. Despite the lack of MRP2 function, the serum bilirubin is only mildly elevated in Dubin–Johnson syndrome, suggesting the presence of additional mechanisms of biliary excretion of bilirubin conjugates. In addition, accumulation of organic anions within the hepatocytes due to MRP2 deficiency leads to upregulation of the expression of MRP1 and MRP3 in the basolateral surface of the hepatocytes. These and possibly additional ATP‐consuming pumps may actively export both unconjugated and conjugated bilirubin from the hepatocyte to plasma via the space of Disse [118]. After intravenous injection of the organic anion bromosulfophthalein (BSP), plasma BSP concentration decreases at near‐ normal rate for 45 minutes, indicating normal uptake across the sinusoidal surface of hepatocytes and normal storage capacity of hepatocytes. However, in 90% of patients, plasma BSP concentration exhibits a secondary increase, so that plasma BSP concentration at 90 minutes exceeds that at 45 minutes. This

238

THE LIVER:  DISORDERS PREDOMINANTLY ASSOCIATED WITH CONJUGATED HYPERBILIRUBINEMIA

secondary rise is due to reflux of glutathione‐conjugated BSP from hepatocytes into the circulation [119]. A similar secondary rise occurs after intravenous administration of bilirubin. However, such secondary rise of plasma BSP can also occur in some other hepatobiliary disorders [120], therefore, this phenomenon is not pathognomonic of Dubin–Johnson syndrome.

Urinary coproporphyrin excretion Coproporphyrins excreted in urine consists of two isomers, coproporphyrin I and coporophyrin III. Normally, in adults approximately 75% of the urinary coproporphyrin is isomer III, which is the precursor of heme. Total urinary coproporphyrin excretion is normal in Dubin–Johnson syndrome, but over 80% is isomer I [121]. Neonates have a higher proportion of urinary coproporphyrin I than adults, but the proportion is not as high as in Dubin–Johnson syndrome. The relationship of the abnormal pattern of urinary porphyrin excretion to the organic anion transport defect is not fully understood. When correlated with history and physical examination, the urinary coproporphyrin excretion pattern is diagnostic of Dubin–Johnson syndrome.

Genetic basis and inheritance Dubin–Johnson syndrome is inherited as an autosomal recessive trait, and has been reported in all races and both sexes. It is rare in all races, except in Iranian, Iraqi, and Moroccan Jews living in Israel, who have an incidence of 1 in 1300 [106], in whom it is associated with clotting factor VII deficiency [122]. Dubin– Johnson syndrome may be relatively common in some areas of Japan, where consanguinity is frequent. Dubin–Johnson syndrome is caused by insertions, deletions, and nonsense mutations of the MRP2 (ABCC2) gene, or abnormal splicing of the RNA transcript that causes loss of MRP2 expression, or missense mutations of the gene that interfere with normal localization of MRP2 in the bile canalicular plasma membrane of hepatocytes [115, 117, 123, 124]. Human MRP2 gene has been localized to chromosome 10q23–q24 [125]. Single amino acid transitions most commonly involve the crucial ATP‐binding region. Some mutations may lead to impaired glycosylation of MRP2, leading to premature proteasome‐ dependent degradation [126].

Animal models In mutant Corriedale sheep, biliary excretion of conjugated bilirubin, glutathione‐conjugated BSP, iopanoic acid, and indocyanine green is decreased, whereas the transport of taurocholate [127] and unconjugaed BSP [128] is normal. The secretion of organic cations, such as procaine amide ethobromide, is unaffected. Serum bilirubin is mildly increased, with 60% of the bilirubin being conjugated. Other than dark brown pigmentation, liver histology is normal [109]. As in human Dubin–Johnson syndrome, total urinary coproporphyrin excretion is normal, but the proportion of coproporphyrin I is increased. Thus, the mutant Corriedale sheep is a model of human Dubin–Johnson syndrome. As in humans with Dubin–Johnson syndrome, Eisai hyperbilirubinemic (EHBR) and TR− rats lack Mrp2 and exhibit reduced biliary excretion of conjugated bilirubin, leukotriene LTC4 and glutathione, as glucuronic acid and glutathione

conjugates [129] and many other organic anions [130], Coproporphyrin I is the major porphyrin isomer excreted in urine [131]. The liver is not normally pigmented, but intracellular pigments are deposited upon feeding a diet enriched in tryptophan, tyrosine, and phenylalanine. Impaired excretion of anionic metabolites of these amino acids may result in their retention, oxidation, polymerization, and subsequent lysosomal accumulation in hepatocytes [132]. Experiments in these rat models showed that the pathway for secretion of bilirubin conjugates and many other organic anions into the bile canaliculus differed from that for secretion of most bile acids. Bile acids with free 3‐OH groups are transported by the bile salt export pump (BSEP), but sulfated or glucuronide‐conjugated bile acids are transported by Mrp2 [133]. The golden lion tamarin monkey (Leontopithecus rosalia), which manifests elevated serum conjugated bilirubin levels, is a non‐human primate model of Dubin–Johnson syndrome [134].

Rotor syndrome Rotor, Manahan, and Florentin reported two families with several patients with lifelong predominantly conjugated hyperbilirubinemia without evidence of hemolysis [135]. Other routine blood biochemistries and hematologic tests were normal and, in contrast to Dubin–Johnson syndrome, there was no pigmentation in the liver. Liver histology was normal. Rotor syndrome is harmless [135]. Although rare, it has been described in several races.

Organic anion excretion The organic anion excretion defect in Rotor syndrome is different from that in Dubin–Johnson syndrome. As in many acquired liver diseases, over 25% of injected BSP is retained in serum at 45 minutes, and there is no secondary rise of plasma BSP level [136]. Plasma clearance of intravenously administrated unconjugated bilirubin and indocyanine green is also delayed. In contrast to the findings in Dubin–Johnson syndrome, bile canalicular export pumps are normal in Rotor syndrome, therefore, the gallbladder is visualized by oral cholecystography [137, 137A].

Urinary coproporphyrin excretion Total urinary coproporphyrin is increased two‐ to fivefold over normal in Rotor syndrome, of which approximately 65% is isomer I [138]. These findings are similar to those seen in many other hepatobiliary disorders, and distinguish Rotor syndrome from Dubin–Johnson syndrome. However, two brothers with clinical features of Rotor syndrome were reported to have over 80% of urinary coproporphyrins as isomer I [139], raising doubts about the usefulness of using urinary coproporphyrin analysis in distinguishing the two syndromes. More recently, the discovery of the genetic basis of Rotor syndrome has clarified the defect in organic anion handling by hepatocytes in this disorder (see later).

Molecular basis of Rotor syndrome Van de Steeg et  al. found a tight linkage between simultaneous mutations in the SLC01B1 and SLC01B3 genes and Rotor syndrome in six families [140]. These mutations result in deficiency



20:  Disorders of Bilirubin Metabolism

of the organic anion transporting polypeptides OATP1B1 and OATP1B3, respectively. Using mice deficient in the multispecific sinusoidal export pump Abcc3, as well as Oatp1a/1b, van de Steeg et  al. demonstrated that Abcc3 secretes bilirubin glucuronides from hepatocytes to blood, whereas Oatp1a/1b mediate their reuptake. Because of the efficient uptake and conjugation of bilirubin by zone 1 hepatocytes from blood entering the hepatic sinusoids, these bilirubin glucuronides formed in these cells may exceed their canalicular excretory capacity. A portion of bilirubin glucuronides formed in the zone 1 hepatocytes is transported back into the sinusoidal blood and is subsequently taken up by hepatocytes located downstream to the blood flow (toward Zzone 3) via Oatp1b1/3. By recruiting additional hepatocytes, this mechanism increases the capacity of liver to handle a bilirubin load. In humans, OATP1B1 and OATP1B3 belong to the 11‐member OATP family that have 12 membrane‐spanning domains and mediate the uptake of a wide variety of compounds. They are almost exclusively localized in the sinusoidal domain of hepatocyte plasma membranes. OATP1B1 and OATP1B3 transport different, but partly overlapping substrates. In addition to bilirubin glucururonides, substrates for both transporters include drugs such as hydroxymethylglutaryl (HMG)‐CoA reductase inhibitors (statins), angiotensin II receptor blockers (sartans), angiotensin‐ converting enzyme (ACE) inhibitors, and antidiabetes (glinides) as substrates [141]. In addition to organic anions, OATP1B1 and OATP1B3 accept neutral compunds, such as digoxin, ouabain, lopinavir, as well as zwitter ionic drugs such as fexofenadine. OATP1B1 preferentially recognizes estrone‐3‐sulfate , whereas cholecystokinin octapeptide CCK‐8, telmisartan, paclitaxel, and docetaxel are specifically transported via OATP1B3. OATPs are structurally unrelated to ABCC2 (MRP2), which is located in the bile canalicular domain of hepatocyte plasma membranes, but they share many of their substrates, including bilirubin glucuronides, which coordinates hepatocellular uptake with and canalicular excretion. For certain substrates, OATP1B1 and OATP1B3 cannot compensate for the lack of each other. For example, simvastatin‐ associated myopathy is strongly linked to a single nucleotide polymorphism of SLC01B1 [142]. On the other hand, for the reuptake of bilirubin glucuronides, OATP1B1 and OATP1B3 can compensate for the lack of one another, whereby simultaneous mutation of both is required for the manifestation of Rotor syndrome.

Animal model A subgroup of Southdown sheep exhibits mild conjugated hyperbilirubinemia and photosensitivity [143]. Southdown sheep with these clinical features were found to have a missense (glycine to arginine) mutation in Slc01b3 [144]. This glycine residue is conserved in seven other mammalian species and is thought to be functionally critical.

INHERITED CHOLESTASIS SYNDROMES Progressive familial intrahepatic cholestasis Three life‐threatening disorders, collectively termed progressive familial intrahepatic cholestasis (PFIC1, PFIC2, PFIC3, and PFIC4), can cause cholestasis of varying severities by affecting

239

the secretion of bile acids or other components of bile across the bile canaliculi of hepatocytes [145]. Benign recurrent intrahepatic cholestasis (BRIC) is a disorder that is genetically related to PFIC1 or PFIC2. The PFIC syndromes usually present during infancy or childhood, and often lead to growth failure and progressive liver disease. Several additional genetic cholestatic disorders have been described during the last few years. PFIC syndromes and the other inherited cholestatic diseases have been reviewed recently [146] and are discussed in brief here.

Progressive familial intrahepatic cholestasis type 1 PFIC1 was originally described in an Amish–Mennonite family and was named Byler disease, after the family name of the first patient [147]. PFIC1 is associated with severe life‐threatening cholestasis. It is caused by mutations in the P‐type ATPase gene FIC1 (also termed ATP8B1), which is located on chromosome 18q21 [148]. FIC1‐mediated ATP hydrolysis is coupled with the translocation of acidic phospholipids. How FIC1 mutation causes cholestasis is not fully understood [149]. FIC1 deficiency may interfere with the nuclear translocation of the nuclear receptor FXR [150], thereby downregulating the expression of the bile salt export protein (BSEP) (see section on Progressive familial intrahepatic cholestasis type 2).

Benign recurrent intrahepatic cholestasis BRIC was first described in 1959 [151]. It presents in adolescence or early adulthood with recurrent episodes of cholestasis, characterized by conjugated hyperbilirubinemia, together with malaise, anorexia, pruritus, weight loss, and malabsorption. During the attacks, which last weeks to months, laboratory tests reveal biochemical evidence of cholestasis without severe hepatocellular injury [152–154]. These attacks are followed by a complete clinical, biochemical, and histological remission. The clinical features and duration of the episodes in a given patient resemble those in previous attacks. Liver biopsy shows noninflammatory intrahepatic cholestasis without fibrosis, notwithstanding the number and severity of the attacks. During remission, light or electron microscopy of the liver tissue returns to normal [155]. Like PFIC1, the inheritance shows an autosomal recessive pattern. Interestingly, this relatively benign disorder is also caused by certain missense mutations in the FIC1 gene, other mutations of which cause the much more severe ­disorder PFIC1. There is no specific treatment for BRIC. Some cases are associated with mutations of ABCB11, which is associated with PFIC2. Because of this, some authors classify BRIC into BRIC‐I and BRIC‐II.

Progressive familial intrahepatic cholestasis type 2 This disorder, which resembles Byler disease clinically, occurs in mainly in Middle Eastern and European races. It is caused by defects of the ABCB11 gene that encodes the bile salt export pump (BSEP), previously termed sister P‐glycoprotein. BSEP is a canalicular ATP‐dependent transporter of bile acids from the hepatocytes into the bile [156]. The ABCB11 gene is located on chromosome 2q24 [157]. A large number of different point mutations in ABCB11 have been described in patients with PFIC2.

240

THE LIVER:  ACKNOWLEDGMENT

Although both PFIC1 and PFIC2 are associated with life‐ threatening cholestasis, serum γ‐glutamyl transpeptidase (GGT) levels are normal or nearly so in both disorders, which differentiates them from PFIC3 [145, 157]. BSEP is expressed specifically in the liver and, as expected, liver transplantation ameliorates all manifestations of PFIC2. However, in several cases, PFIC‐like symptoms recur after liver transplantation consequent to development of immunoglobulin G‐type antibodies against BSEP. Although antibodies develop against both the N‐terminal and the C‐terminal regions of the protein, only sera recognizing the first extracellular loop (ECL1) inhibit transepithelial taurocholate transport [158].

Progressive familial intrahepatic cholestasis type 3 PFIC3 involves mutations in ABCB4, an ATP‐utilizing bile canalicular pump (also known as the multidrug resistance protein‐3 P‐glycoprotein (PGY3, MDR3). ABCB4 transports phosphatidylcholine from the inner lipid leaflet of the bile canaliculus to the outer leaflet, thereby replenishing the phospholipid in the outer leaflet, which is continuously removed by contact with bile acids. Failure of phosphatidylcholine translocation, the canalicular membrane, and small bile ducts are chronically injured by bile salts [158, 159]. In contrast to PFIC1 and PFIC2, in which the canalicular membrane is protected by reduced secretion of bile salts in the canalicular bile, serum GGT activity is increased in PFIC3 [145]. Human ABCB4 deficiency can present with a variety of hepatobiliary disorders, including small‐ duct primary sclerosing cholangitis [160] and cholesterol gallstones [161]. Interestingly, asymptomatic female heterozygous carriers of nonsense or missense mutations of the ABCB4 gene have been reported to manifest familial intrahepatic cholestasis of pregnancy [162, 163], the cause of which was unknown heretofore. Liver transplantation is the only treatment option for PFIC3 available at this time. Transplantation of normal hepatocytes in a knockout mouse model of PFIC3 has resulted in spontaneous massive liver repopulation with the transplanted cells, resulting in long‐term amelioration of the phospholipid transport defect [164, 165].

OTHER INHERITED CHOLESTATIC DISORDERS Mutation of CIRH1A, the gene encoding a mitochondrial scaffold protein, has been reported to manifest as cholangiopathy of North American Indian childhood cirrhosis. Clinically, this disorder resembles extrahepatic biliary atresia, but there is no evidence of biliary tract obstruction [168].

GRACILE syndrome Genetic lesions of another mitochondrial scaffold gene, BCS1L, presents potentially lethal intrahepatic cholestasis associated with fetal growth retardation, amino aciduria, iron overload, and lactic acidosis (GRACILE) [169]. Interestingly, neither North American Indian childhood cirrhosis nor GRACILE exhibit the usual features of mitochondriopathy, such as central nervous system (CNS) lesions or hepatocellular steatosis.

Alagille syndrome Alagille syndrome affects the development of multiple organ systems, including liver, heart, kidneys, eyes, vertebrae (sagittal, cleft), and CNS [170]. Paucity of bile ducts is found in 89% of cases [171]. Most patients exhibit a characteristic facies. Although the CNS lesions can be the result of malnutrition, even after careful nutritional balance, intracranial bleeding is the most common CNS complication [171]. Alagille syndrome is inherited as an autosomal dominant characteristic with markedly variable penetrance. Lesions of the Jagged1 (JAG1) gene, on chromosome 20p12 [172] are responsible for this disease. Jagged‐1 is a ligand of Notch, and is important in the Notch signaling pathway, which is critical in organ development. Coding region mutations of JAG1 have been identified in about 70% of the patients with Alagille syndrome. Surprisingly, 50–70% of these mutations are de novo, and not found in the parents. About 21–50% of patients with Alagille syndrome who manifest hepatic symptoms in infancy eventually require liver transplantation.

Progressive familial intrahepatic cholestasis type 4 As discussed earlier, PFIC syndromes can be classified clinically by serum GGT levels that reflect bile salt‐induced injury of bile canaliculi or smallest bile ductules. In PFIC1 and PFIC2 serum GGT is normal, whereas in PFIC3 it is elevated. In 2014, Sambrota et al. [166] reported a multi‐institution collaborative study of 33 children from 29 families who had severe chronic cholestatic liver disease, with low GGT for the degree of cholestasis. However, these children did not have mutations of either ABCB11 and ATP8B1, which cause PFIC1 and PFIC2, respectively. In these 29 families, 18 parents were consanguineous. Genetic analysis by targeted resequencing (TRS), whole‐exome sequencing (WES), or both revealed protein‐truncating mutations in the tight junction protein 2 gene (TJP2), which resulted in failure of incorporation of TJP2 protein in desmosomes delimiting the bile canaliculi. PFIC resulting from TJP2 mutations has been designated PFIC4. Two cases of hepatocellular carcinoma in infants with this disorder have been reported [167].

Villin disease Villin is a tissue‐specific actin‐modifying protein, critical in the bundling, nucleation, capping, and severing of actin filaments [173]. In vertebrates, villin is expressed at the apical surface of epithelial cells and is a major structural component of the brush border cytoskeleton. Several children with features of cholestasis, who later developed cirrhosis of the liver requiring liver transplantation, were found to have lesions of the VILLIN1 gene [174].

ACKNOWLEDGMENT This work was supported in part by NIH grants RO1‐DK‐092469, RO1‐DK 100490 and P30‐DK‐41296.



20:  Disorders of Bilirubin Metabolism

REFERENCES 1. Kappas, A. and Drummond, G.S. Direct comparison of Sn‐mesoporphyrin, an inhibitor of bilirubin production, and phototherapy in controlling hyperbilirubinemia in term and near‐term newborns. Pediatrics, 1995;95:468. 2. Valaes, T., Petmezaki, S., Henschke, C., Drummond, G.S., and Kappas, A. Control of jaundice in preterm newborns by an inhibitor of bilirubin production: studies with tin‐mesoporphyrin. Pediatrics, 1994;93:1. 3. Westlake, D.W.S., Roxburgh, J.M., and Talbot, G. Microbial production of carbon monoxide from flavonoids. Nature, 1961;189:510. 4. Itoh, K., Chiba, T., Takahashi, S. et al. An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements. Biochem Biophys Res Commun, 1997;236(2):313–22. 5. Naito, Y., Takagi, T., Uchiyama, K., and Yoshikawa, T. Heme oxygenase‐1: a novel therapeutic target for gastrointestinal diseases. J Clin Biochem Nutr, 2011;48:126. 6. Yachie, A., Niida, Y., Wada, T. et  al. Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase‐1 deficiency. J Clin Invest, 1999;103:129–35. 7. Mougiakakos, D., Jitschin, R., Johansson, C.C. et al. The impact of inflammatory licensing on heme oxygenase‐1‐mediated induction of regulatory T cells by human mesenchymal stem cells. Blood, 2011;117:4826–35. 8. Basuroy, S., Bhattacharya, S., Tcheranova, D. et al. HO‐2 provides endogenous protection against oxidative stress and apoptosis caused by TNF‐α in cerebral vascular endothelial cells. Am J Physiol Cell Physiol, 2006;291:C897–C908. 9. Chowdhury, R., Roy Chowdhury, N., and Arias, I.M. Bilirubin conjugation in the spiny dogfish Squalus acanthias, the small skate Raja erinacea and the winter flounder Pseudopleuronectes americanus. Comp Biochem Physiol, 1980;668:523. 10. O’Brien, L., Hosick, P.A., John, K., Stec, D.E., and Hinds, T.D., Jr. Biliverdin reductase isozymes in metabolism. Trends Endocrinol Metab, 2015;26:212–20. 11. Parkar, M., Jeremiah, S.J., Povey, S. et al. Confirmation of the assignment of human biliverdin reductase to chromosome 7. Ann Human Genet, 1984;48: 57–60. 12. Fu, G., Liu, H., and Doerksen, R.J. Molecular modeling to provide insight into the substrate binding and catalytic mechanism of human biliverdin‐ IXalpha reductase. J Phys Chem B, 2012;116:9580–94. 13. Hannam, S., Moriaty, P., O’Reilly, H. et al. Normal neurological outcome in two infants treated with exchange transfusions born to mothers with Crigler‐ Najjar Type 1 disorder. Eur J Pediatr, 2009;168:427–9. 14. Bonnet, R.J., Davis, E., and Hursthouse, M.B. Structure of bilirubin. Nature, 1976;262:326. 15. Van den Bergh, A.A.H. and Muller, P. Uber eine direkte und eine indirekte Diazoreaktion auf Bilirubin. Biochem Z, 1916;77:90. 16. Choi, S.H., Yun, K.E., and Choi, H.J. Relationships between serum total bilirubin levels and metabolic syndrome in Korean adults. Nutr Metab Cardiovasc Dis, 2013;23:31. 17. Breimer, L.H., Wannamethee, G., Ebrahim, S., and Shaper, A.G. Serum bilirubin and risk of ischemic heart disease in middle‐aged British men. Clin Chem, 1995;41:1504. 18. Temme, E.H.N., Zhang, J., Schouten, E.G. et al. Serum bilirubin and 10‐year mortality risk in a Belgian population. Cancer Causes Control, 2001;12:887. 19. Zucker, S.D., Horn, P.S., and Sherman, K.E. Serum bilirubin levels in the US population: gender effect and inverse correlation with colorectal cancer. Hepatology, 2004;40:827. 20. Torgerson, J.S., Lindroos, A.K., Sjostrom, C.D. et  al. Are elevated aminotransferases and decreased bilirubin additional characteristics of the metabolic syndrome? Obesity Res, 1997;5:105–14. 21. Hinds, T.D. Jr. Sodhi, K., Meadows, C. et al. Increased HO‐1 levels ameliorate fatty liver development through a reduction of heme and recruitment of FGF21. Obesity, 2014;22:705–12. 22. Dong, H., Huang, H., Yun, X. et al. Bilirubin increases insulin sensitivity in leptin‐receptor deficient and diet‐induced obese mice through suppression of ER stress and chronic inflammation. Endocrinology, 2014;155:818–28. 23. Amin, S.B., Saluja, S., Saili, A. et al. Chronic auditory toxicity in late preterm and term infants with significant hyperbilirubinemia. Pediatrics, 2017;(4):pii:e20164009. 24. Ahlfors, C.E., Vreman, H.J., Wong, R.J. et  al. Effects of sample dilution, peroxidase concentration, and chloride ion on the measurement of unbound bilirubin in premature newborns. Clin Biochem, 2007;40:261–7.

241

25. Nakamura, H., Yonetani, M., Uetani, Y., Funato, M., and Lee, Y. Determination of serum unbound bilirubin for prediction of kernicterus in low birthweight infants. Acta Paediatr Jpn, 1992;34: 642–7. 26. Wang, P., Kim, R.B., Roy‐Chowdhury, J., and Wolkoff, A.W. The human organic anion transport protein SLc21A6 is not sufficient for bilirubin transport. J Biol Chem, 2003;278:20695. 27. Bosma, P.J., Seppen, J., Goldhoorn, B. et al. Bilirubin UDP‐glucuronosyltransferase 1 is the only relevant bilirubin glucuronidating isoform in man. J Biol Chem, 1994;269:17960. 28. Ritter, J.K., Chen, F., Sheen, Y.Y. et  al. A novel complex locus UGT1 encodes human bilirubin, phenol, and other UDP‐glucuronosyltransferase isozymes with identical carboxyl termini. J Biol Chem, 1992;267:3257. 29. Kadakol, A., Ghosh, S.S., Sappal, B.S. et  al. Genetic lesions of bilirubin uridinediphosphoglucuronate glucuronosyltransferase causing Crigler‐ Najjar and Gilbert’s syndromes: correlation of genotype to phenotype. Hum Mutat, 2000;16:297. 30. Gantla, S., Bakker, C.T.M., Deocharan, B. et  al. Splice site mutations: a novel genetic mechanism of Crigler‐Najjar syndrome type 1. Am J Hum Genet, 1998;62:585. 31. Bosma, P.J., Roy Chowdhury, J., Bakker, C. et al. A sequence abnormality in the promoter region results in reduced expression of bilirubin‐UDP‐glucuronosyltransferase‐1 in Gilbert syndrome. N Engl J Med, 1995;333:1171. 32. van de Steeg, E., Stranecky, V., Hartmannova, H. et al. Complete OATP1B1 and OATP1B3 deficiency causes human Rotor syndrome by interrupting conjugated bilirubin reuptake into the liver. J Clin Invest, 2012;122:519. 33. Huang, W., Zhang, J., Chua, S.S. et al. Induction of bilirubin clearance by the constitutive androstane receptor. Proc Natl Acad Sci U S A, 2003;100:4156. 34. Xie, W., Yeuh, M.F., Radominska‐Pandya, A. et al. Control of steroid, heme, and carcinogen metabolism by nuclear pregnane X receptor and constitutive androstane receptor. Proc Natl Acad Sci U S A, 2003;100:4150. 35. Roy‐Chowdhury, J., Locker, J., and Roy‐Chowdhury, N. Nuclear receptors orchestrate detoxification pathways. Dev Cell, 2003;4:607. 36. Maisels, M.J. Managing the jaundiced newborn: a persistent challenge. Can Med Assoc J, 2015;187:335–43. 37. American Academy of Pediatrics Subcommittee on Hyperbilirubinemia. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics, 2004;114:297–316. Erratum October 01, 2004. 38. Bhutani, V.K., Stark, A.R., Lazzeroni, L.C. et al. Initial clinical testing evaluation and risk assessment for universal screening for Hyperbilirubinemia Screening Group. Predischarge screening for severe neonatal hyperbilirubinemia identifies infants who need phototherapy. J Pediatr, 2013;162:477–82. 39. Maisels, M.J., Pathak, A., Nelson, N.M. et  al. Endogenous production of carbon monoxide in normal and erythroblastic newborn infants. J Clin Invest, 1971;50:1. 40. Vest, M., Strebel, L., and Hauensiein, D. The extent of “shunt” bilirubin and erythrocyte survival in the newborn infant measured by the administration of (15N) glycine. Biochem J, 1965;95:11c. 41. Clarke, C.A., Donohoe, W.T.A., Finn, R. et al. Prevention of Rh hemolytic disease: Final results of the “high risk” clinical trial. A combined study from centers in England and Baltimore. Br Med J, 1971;2:607. 42. Levi, A.J., Gatmaitan, Z., and Arias, I.M. Deficiency of hepatic organic anion‐binding protein, impaired organic anion uptake by liver and “physiologic” jaundice in newborn monkeys. N Engl J Med, 1970;283:1136. 43. Kawade, N. and Onishi, S. The prenatal and postnatal development of UDP‐glucuronosyltransferase activity towards bilirubin and the effect of premature birth on this activity in human liver. Biochem J, 1981;196:257. 44. Arthur, L.J., Bevan, B.R., and Holton, J.B. Neonatal hyperbilirubinemia and breast feeding. Dev Med Child Neurol, 1966;8:279. 45. Arias, I.M., Gartner, L.M., Seifter, S., and Furman, M. Prolonged neonatal unconjugated hyperbilirubinemia associated with breast feeding and a steroid, pregnane‐3α, 20β‐diol, in maternal milk that inhibits glucuronide formation in vitro. J Clin Invest, 1964;42:2037. 46. Maisels, M.J. and Newman, T.B. Kernicterus in otherwise healthy breast‐fed term newborns. Pediatrics, 1995;96:730. 47. Foliot, A., Ploussard, J.P., Housett, E. et al. Breast milk jaundice: In vitro inhibition of rat liver bilirubin‐uridine diphosphate glucuronosyltransferase activity and Z protein‐bromosulfophthalein binding by human breast milk. Pediatr Res, 1976;10:594. 48. Lucey, J.F. and Driscol, J.J. Physiological jaundice re‐examined, in Kernicterus (ed. A. Sass‐Kortsak), University of Toronto Press, Toronto, 1961.

242

THE LIVER:  REFERENCES

49. Nies, A.T., Gatmaitan, Z., and Arias, I.M. ATP‐dependent phosphatidylcholine translocation in rat liver canalicular plasma membrane vesicles. J Lipid Res, 1996;37:1125. 50. Poland, R.L. and Odell, G.B. Physiologic jaundice: the enterohepatic circulation of bilirubin. N Engl J Med, 1971;284:1 51. Crigler, J.F. and Najjar, V.A. Congenital familial non‐hemolytic jaundice with kernicterus. Pediatrics, 1952;10:169. 52. Childs, B., Sidbury, J.B., and Migeon, C.J. Glucuronic acid conjugation by patients with familial non‐hemolytic jaundice and their relatives. Pediatrics, 1959;23:903. 53. Berk, P.D., Martin, F., Blaschke, T.F. et al. Unconjugated hyperbilirubinemia: physiological evaluation and experimental approaches to therapy. Ann Intern Med, 1975;82:552. 54. Seppen, J., Bosma, P.J., Goldhoorn, B.G. et  al. Discrimination between Crigler‐Najjar type I and II by expression of mutant bilirubin uridine diphosphate‐glucuronosyltransferase. J Clin Invest, 1994;94:2385. 55. Wolkoff, A.W., Roy Chowdhury, J., Gartner, L.A. et al. Crigler‐Najjar syndrome (type I) in an adult male. Gastroenterology, 1979;76:3380. 56. Mitchell, E., Ranganathan, S., McKiernan, P. et  al. Hepatic parenchymal injury in Crigler‐Najjar type I. J Pediatr Gastroenterol Nutr, 2018;66: 588–94. 57. Petit, F.M., Gajdos, V., Parisot, F. et al. Parental isodisomy for chromosome 2 as the cause of Crigler‐Najjar syndrome type I syndrome. Eur J Hum Genet, 2005;13:278–82. 58. Van Es, H.H.G., Goldhoorn, B., Paul‐Abrahamse, M. et al. Immunochemical characterization of UDP‐glucuronosyltransferase in four patients with the Crigler‐Najjar type I syndrome. J Clin Invest, 1990;85:1199. 59. Ritter, J.K., Crawford, J.M., and Owens, I.S. Cloning of two human liver bilirubin‐UDP‐glucuronosyltransferase cDNAs with expression in COS‐1 cells. J Biol Chem, 1991;266:1043. 60. Bosma, P.J., Roy Chowdhury, N., Goldhoorn, B.G. et al. Sequence of exons and the flanking regions of human bilirubin‐UDP‐glucuronosyltransferase gene complex and identification of a genetic mutation in a patient with Crigler‐Najjar syndrome, type I. Hepatology, 1992;15:941–7. 61. Bosma, P.J., Seppen, J., Goldhoorn, B. et al. Bilirubin UDP‐glucuronosyltransferase 1 is the only relevant bilirubin glucuronidating isoform in man. J Biol Chem, 1994;269:17960. 62. Servedio, V., d’Apolito, M., Maiorano, N. et al. Spectrum of UGT1A1 mutations in Crigler‐Najjar (CN) syndrome patients: identification of twelve novel alleles and genotype‐phenotype correlation. Hum Mutat, 2005;25:325. 63. Strauss, K.A., Robinson, D.L., Vreman, H.J. et al. Management of hyperbilirubinemia and prevention of kernicterus in 20 patients with Crigler‐Najjar disease. Eur J Pediatr, 2006;165:306–19. 64. Kadakol, A., Deocharan, B., Mukhopadhyay, L. et al. Rapid prenatal diagnosis of Crigler‐Najjar syndrome type 1 by genetic analysis of chorionic villus samples. Hepatology, 1998;28:316A. 65. Gunn, C.H. Hereditary acholuric jaundice in a new mutant strain of rats. J Hered, 1938;29:137. 66. Roy Chowdhury, N., Kondapalli, R., and Roy Chowdhury, J. The Gunn rat: An animal model for inherited deficiency of bilirubin glucuronidation, in Animal Models in Liver Research (eds. C.E. Cornelius, R.R. Marshak, and E.C. Melby), Academic Press, New York, 1993, p. 150. 67. Roy Chowdhury, N., Kondapalli, R., and Roy Chowdhury, J. The Gunn rat: an animal model for inherited deficiency of bilirubin glucuronidation, in Animal Models in Liver Research (eds. C.E. Cornelius, R.R. Marshak, and E.C. Melby), Academic Press, New York, 1993, pp. 150–75. 68. Polgar, Z., Li, Y., Li Wang, X. et al. Gunn rats as a surrogate model for evaluation of hepatocyte transplantation‐based therapies of Crigler‐Najjar syndrome type 1. Methods Mol Biol, 2017;1506:131–47. 69. Chen, Y., Li, Y., Wang, X. et al. Amelioration of hyperbilirubinemia in Gunn rats after transplantation of human induced pluripotent stem cell‐derived hepatocytes. Stem Cell Reports, 2015;5:22–30. 70. Nguyen, N., Bonzo, J.A., Chen, S. et al. Disruption of the Ugt1 locus in mice resembles human Crigler‐Najjar type I disease. J Biol Chem, 2008; 283:7901–11. 71. Mazariegos, G., Shneider, B., Burton, B. et al. Liver transplantation for pediatric metabolic disease. Mol Genet Metab, 2014;111:418–27. 72. Kappas, A., Drummond, G.S., Manola, T., Petmezaki, S., and Valaes T Sn‐ protoporphyrin use in the management of hyperbilirubinemia in term newborns with direct Coombs‐positive ABO incompatibility. Pediatrics, 1988;81(4):485–97.

73. Galbraith, R.A., Drummond, G.S., and Kappas A. Suppression of bilirubin production in the Crigler‐Najjar, type I syndrome: studies with heme oxygenase inhibitor tin‐mesoporphyrin. Pediatrics, 1992;89:175. 74. Fox, I.J., Roy Chowdhury, J., Kaufman, S.S. et  al. Treatment of Crigler‐ Najjar syndrome type I with hepatocyte transplantation. N Engl J Med, 1998;333:1422. 75. Iansante, V., Mitry, R.R., Filippi, C., Fitzpatrick, E., and Dhawan, A. Human hepatocyte transplantation for liver disease: current status and future perspectives. Pediatr Res, 2018;83:232–40. 76. Zhou, H., Dong, X., Kabarriti, R. et al. Single liver lobe repopulation with wildtype hepatocytes using regional hepatic irradiation cures jaundice in Gunn rats. PLoS One, 2012;7:e46775. 77. Yannam, G.R., Han, B., Setoyama, K. et al. A nonhuman primate model of human radiation‐induced venocclusive liver disease and hepatocyte injury. Int J Radiat Oncol Biol Phys, 2014;88:404–11. 78. Soltys, K.A., Setoyama, K., Tafaleng, E.N. et  al. Host conditioning and rejection monitoring in hepatocyte transplantation in humans. J Hepatol, 2016;66:987–1000. 79. Arias, I.M. Chronic unconjugated hyperbilirubinemia without overt signs of hemolysis in adolescents and adults. J Clin Invest, 1962;41:2233. 80. Gollan, J.L., Huang, S.M., Billing, B., and Sherlock, S. Prolonged survival in three brothers with severe type II Crigler‐Najjar syndrome: ultrastructural and metabolic studies. Gastroenterology, 1975;68:1543. 81. Arias, I.M., Gartner, L.M., Cohen, M. et al. Chronic nonhemolytic unconjugated hyperbilirubinemia with glucuronosyltransferase deficiency: clinical, biochemical, pharmacologic, and genetic evidence for heterogeneity. Am J Med 1969;47:395. 82. Fevery, J., Blanckaert, N., Heirwegh, K.P.M. et al. Unconjugated bilirubin and an increased proportion of bilirubin monoconjugates in the bile of patients with Gilbert’s syndrome and Crigler‐Najjar syndrome. J Clin Invest, 1977;60:970. 83. Bosma, P.J., Golhoorn, B., Oude Elferink, R.P. et al. A mutation in bilirubin uridine 5′‐diphosphate glucuronosyltransferase isoforms 1 causing Crigler‐ Najjar syndrome type II. Gastroenterology, 1993;105:216. 84. Gilbert, A. and Lereboullet, P. La cholamae simple familiale. Semin Med, 1901;21:241. 85. Bosma, P.J., Roy Chowdhury, J., Bakker, C. et al. A sequence abnormality in the promoter region results in reduced expression of bilirubin‐UDP‐glucuronosyltransferase‐1 in Gilbert syndrome. N Engl J Med, 1995;333:1171. 86. Thompson, R.P.H. (1981) Genetic transmission of Gilbert’s syndrome, in Familial Hyperbilirubinemia (ed. L. Okolicsanyl), Wiley, New York, 1981, p. 91. 87. Powell, L.W., Hemingway, E., Billing, B.H., and Sherlock, S. Idiopathic unconjugated hyperbilirubinemia (Gilbert’s syndrome): a study of 42 families. N Engl J Med, 1967;277:1108. 88. Arias, I.M. and London, I.M. Bilirubin glucuronide formation in vitro: Demonstration of a defect in Gilbert’s disease. Science, 1957;126:563. 89. Davidson, A.R., Rojas‐Beuno, A., Thompson, R.P.H., and Williams, R. Reduced caloric intake and nicotinic acid provocation tests in diagnosis of Gilbert’s syndrome. Br Med J, 1975;2:480. 90. Bosma, P.J., Roy Chowdhury, J., Bakker, C. et al. A sequence abnormality in the promoter region results in reduced expression of bilirubin‐UDP‐glucuronosyltransferase‐1 in Gilbert syndrome. N Engl J Med, 1995;333:1171. 91. Maruo, Y., Sato, H., Yamano, T. et al. Gilbert’s syndrome caused by homozygous missense mutation (Tyr486Asp) of bilirubin‐UDP glucuronosyl transferase. J Pediatr, 1998;132:1045. 92. Soeda, Y., Yamamoto, K., Adachi, Y. et al. Predicted homozygous mis‐sense mutation in Gilbert’s syndrome. Lancet, 1995;346:1494. 93. Koiwai, O., Nishizawa, M., Hasada, K. et al. Gilbert’s syndrome is caused by a heterozygous missense mutation in the gene for bilirubin UDP‐glucuronosyltransferase. Hum Mol Genet, 1995;4:1183. 94. Kadakol, A., Sappal, B.S., Ghosh, S.S. et  al. Interaction of coding region mutations and the Gilbert‐type promoter abnormality of the UGT1A1 gene causes moderate degrees of unconjugated hyperbilirubinemia and may lead to neonatal kernicterus. J Med Genet, 2001;38:244. 95. Labrune, P., Myara, A., Chalas, J. et al. Association of a homozygous (TA)8 promoter polymorphism and a N400D mutation of UGT1A1 in a child with Crigler‐Najjar type II syndrome. Hum Mutat, 2002;20:399. 96. Wells, P.G., Mackenzie, P.I., Roy‐Chowdhury, J. et al. Glucuronidation and the UDP‐glucuronosyltransferases in health and disease. Drug Metab Dispos, 2004;32:281.



20:  Disorders of Bilirubin Metabolism

97. Iyer, L., King, C.D., Whitington, P.F. et al. Genetic predisposition to the metabolism of irinotecan (CPT‐11): Role of uridine diphosphate glucuronosyltransferase isoform 1A1 in the glucuronidation of its active metabolite (SN‐38) in human liver microsomes. J Clin Invest, 1998;101:847. 98. de Morais, S.M., Uetrecht, J.P., and Wells, P.G. Decreased glucuronidation and increased bioactivation of acetaminophen in Gilbert’s syndrome. Gastroenterology, 1992;102:577. 99. Ullrich, D., Sieg, A., Blume, R. et al. Normal pathways for glucuronidation, sulphation and oxidation of paracetamol in Gilbert’s syndrome. Eur J Clin Invest, 1987;17:237. 100. Abumiya, M., Takahashi, N., Niioka, T. et al. Influence of UGT1A1 6, 27, and 28 polymorphisms on nilotinib‐induced hyperbilirubinemia in Japanese patients with chronic myeloid leukemia. Drug Metab Pharmacokinet, 2014;29:449–54. 101. Portman, O.W., Alexander, M., Roy Chowdhury, J. et al. Effects of nutrition on hyperbilirubinemia in Bolivian squirrel monkeys. Hepatology, 1984;4:454. 102. Portman, O.W., Roy Chowdhury, J., Roy Chowdhury, N. et  al. A non‐ human primate model for Gilbert’s syndrome. Hepatology, 1984;4:175. 103. Dubin, I.N. and Johnson, F.B. (1954) Chronic idiopathic jaundice with unidentified pigment in liver cells: a new clinicopathologic entity with a report of 12 cases. Medicine (Baltimore) 1954;33:155. 104. Sprinz, H. and Nelson, R.S. Persistent nonhemolytic hyperbilirubinemia associated with lipochrome‐like pigment in liver cells: report of four cases. Ann Intern Med, 1954;41:952. 105. Dubin, I.N. Chronic idiopathic jaundice: a review of fifty cases. Am J Med, 1958;23:268. 106. Shani, M., Seligshon, U., Gilon, E. et  al. Dubin‐Johnson syndrome in Israel: clinical, laboratory, and genetic aspects of 101 cases. West J Med, 1970;39:549. 107. Come, S.E., Shohet, S.B., and Robinson, S.H. Surface remodeling vs. whole‐cell hemolysis of reticulocytes produced with erythroid stimulation or iron deficiency anemia. Blood, 1974;44:817. 108. Morita, M. and Kihava, T. Intravenous cholecystography and metabolism of meglumine iodipamide (biligrafin) in Dubin‐Johnson syndrome. Radiology, 1971;95:57. 109. Arias, I.M., Bernstein, L., Roffler, R., and Ben Ezzer, J. Black liver diseases in Corriedale sheep: metabolism of tritiated epinephrine and incorporation of isotope into the hepatic pigment in vivo. J Clin Invest, 1965;44:1026. 110. Swartz, H.M., Sarna, T., and Varma, R.R. On the nature and excretion of the hepatic pigment in the Dubin‐Johnson syndrome. Gastroenterology, 1979;76:958. 111. Ishikawa, T., Muller, M., Klunemann, C. et  al. ATP‐dependent primary active transport of cysteinyl leukotrienes across liver canalicular membranes. J Biol Chem, 1990;265:19279. 112. Jedlitschky, G., Leier, I., Buchholz, U. et al. ATP‐dependent transport of glutathione S‐conjugates by the multidrug resistance‐associated protein. Cancer Res, 1994;54:4833. 113. Kobayashi, K., Sogame, Y., Hara, H., and Hayashi, K. Mechanism of glutathione S‐conjugate transport in canalicular and basolateral rat liver plasma membranes. J Biol Chem, 1990;265:7737. 114. Nishida, T., Hardenbrook, C., Gatmaitan, Z., and Arias, I.M. ATP‐dependent organic anion transport system in normal and TR− rat liver canalicular membranes. Am J Physiol, 1992;262:G629. 115. Paulusma, C.C., Bosma, P.J., Zaman, G.J.R. et al. Congenital jaundice in rats with a mutation in a multidrug resistance‐associated protein gene. Science, 1996;271:1126. 116. Takikawa, H., Sano, N., Narita, T. et al. Biliary excretion of bile acid conjugates in a hyperbilirubinemic mutant sprague‐dawley rat. Hepatology, 1991;14:352–60. 117. Togawa, T., Mizuochi, T., Sugiura, T. et al. Clinical, pathologic, and genetic features of neonatal Dubin‐Johnson syndrome: a multicenter study in Japan. J Pediatr, 2018;196:161–7. 118. Konig, J., Rost, D., Cui, Y., and Keppler, D. Characterization of the human multidrug resistance protein isoform MRP3 localized to the basolateral hepatocyte membrane. Hepatology, 1999;29:1156. 119. Charbonnier, A. and Brisbois, P. Etude chromatographique de la BSP au cours de l’epreuve clinique d’epuration plasmatique de ce colorant. Rev Intern Hepatol, 1960;10:1163. 120. Rodes, J., Zubizarreta, A., and Bruguera, M. Metabolism of bromsulphophthalein in Dubin‐Johnson syndrome: Diagnostic value of the paradoxical increase in plasma levels of BSP. Am J Dig Dis, 1972;17:545.

121. 122. 123.

124.

125.

126.

127.

128.

129. 130.

131.

132.

133.

134.

135. 136.

137.

137A. 138. 139. 140.

141.

142.

143.

144.

243

Kaplowitz, N., Javitt, N., and Kappas, A. Coproporphyrin I and III excretion in bile and urine. J Clin Invest, 1972;51:2895. Seligsohn, U., Shani, M., Ramot, B. et al. Dubin‐Johnson syndrome in Israel. II. Association with factor‐VII deficiency. Q J Med, 1970;39:569. Kamisako, T., Leier, I., Cui, Y. et al. Transport of monoglucuronosyl and bisglucuronosyl bilirubin by recombinant human and rat multidrug resistance protein 2. Hepatology, 1999;30:485. Kartenbeck, J., Leuschner, U., Mayer, R., and Keppler, D. Absence of the canalicular isoform of the MRP gene‐encoded conjugate export pump from the hepatocytes in Dubin‐Johnson syndrome. Hepatology, 1996; 23:1061. Allikmets, R., Gerrard, B., Hutchinson, A., and Dean, M. Characterization of the human ABC superfamily: isolation and mapping of 21 new genes using the Expressed Sequence Tags database. Hum Mol Gen, 1996;5:1649. Keitel, V., Karenbeck, J., Nies, A.T. et al. Impaired protein maturation of the conjugate export pump multidrug resistance protein 2 as a consequence of a deletion mutation in Dubin‐Johnson syndrome. Hepatology, 2000;32:1317. Cornelius, C.E., Arias, I.M., and Osburn, B.I. Hepatic pigmentation with photosensitivity: a syndrome in Corriedale sheep resembling Dubin‐ Johnson syndrome in man. J Am Vet Med Assoc, 1965;146:709. Barnhart, J.L., Gronwall, R.R., and Combes, B. Biliary excretion of ­sulfobromophthalein compounds in normal and mutant Corriedale sheep: evidence for a disproportionate transport defect for conjugated sulfobromophthalein. Hepatology, 1981;1:441. Bohan, A. and Boyer, J.L. Mechanisms of hepatic transport of drugs: Implications for cholestatic drug reactions. Semin Liver Dis, 2002;22:123. Jansen, P.L.M., van Klinken, J.W., van Gelder, M. et al. Preserved organic anion transport in mutant TR‐ rats with a hepatobiliary secretion defect. Am J Physiol, 1993;265:G445. Jansen, P.L.M., Peters, W.H.M., and Lamers, W.H. Hereditary chronic conjugated hyperbilirubinemia in mutant rats caused by defective hepatic anion transport. Hepatology, 1985;5:573. Kitamura, T., Alroy, J., Gatmaitan, Z. et al. Defective biliary excretion of epinephrine metabolites in mutant (TR‐) rats: Relation to the pathogenesis of black liver in the Dubin‐Johnson syndrome and Correidale sheep with an analogous excretory defect. Hepatology, 1992;15:1154. Takikawa, H., Nishikawa, K., Sano, N. et  al. Mechanisms of biliary excretion of lithocholate‐3‐sulfate in Eisai hyperbilirubinemic rats (EHBR). Dig Dis Sci, 1995;40:1792. Schulman, F.Y., Montali, R.J., Bush, M. et al. Dubin‐Johnson‐like syndrome in Golden Lion tamarins (Leontopithecus rosalia rosalia). Vet Pathol, 1993;30:491. Rotor, A.B., Manahan, L., and Florentin, A. Familial nonhemolytic jaundice with direct van den Bergh reaction. Acta Med Philos, 1948;5:37. Wolpert, E., Pascasio, F.M., Wolkoff, A.W., and Arias, I.M. Abnormal sulfobromophthalein metabolism in Rotor’s syndrome and obligate heterozygotes. N Engl J Med, 1977;296:1099. Schiff, L., Billing, B.H., and Oikawa Y (1959) Familial nonhemolytic jaundice with conjugated bilirubin in the serum; a case study. N Engl J Med, 1959;260(26):1315–18. Porush, J.G., Delman, A.J., and Feuer, M.M. Chronic idiopathic jaundice with normal liver histology. Arch Intern Med, 1962;109:102. Wolkoff, A.W., Wolpert, E., Pascasio, F.N., and Arias, I.M. Rotor’s syndrome: A distinct inheritable pathophysiologic entity. Am J Med, 1976;60:173 Rapacini, G.L., Topi, G.C., Anti, M. et al. Porphyrins in Rotor syndrome: A study on an Italian family. Hepatogastroenterology, 1986;33:11. van de Steeg, E., Stranecky’, V., Hartmannova, H. et al. Complete OATP1B1 and OATP1B3 deficiency causes human Rotor syndrome by interrupting conjugated bilirubin reuptake into the liver. J Clin Invest, 2012;122:519–28. Maeda, K. Organic anion transporting polypeptide (OATP)1B1 and OATP1B3 as important regulators of the pharmacokinetics of substrate drugs. Biol Pharm Bull, 2015;38:155–68. Link, E., Parish, S., Armitage, J. et al. SEARCH Collaborative Group. SLC01B1 variants and statin‐induced myopathy: a genomewide study. N Engl J Med, 2008;359:789–99. Cornelius, C.E. and Gronwall, R.R. Congenital photosensitivity and hyperbilirubinemia in Southdown sheep in the United States. Am J Vet Res, 1968;29:291–5. Posbergh, C.J.,.Kalla, S.E., Sutter, N.B., Tennant, B.C., and Huson, H.J. Mutation responsible for congenital photosensitivity and hyperbilirubinemia in Southdown sheep. Am J Vet Res, 2018;79:538–45.

244

THE LIVER:  REFERENCES

145. Jansen, P.L. and Muller, M.M. Progressive familial intrahepatic cholestasis types 1, 2, and 3. Gut, 1998;42:766. 146. Sticova, E., Jirsa, M., and Pawlowska, J. New insights in genetic cholestasis: from molecular mechanisms to clinical implications. Can J Gastroenterol Hepatol, 2018, Article ID 2313675. https://doi.org/10.1155/2018/2313675 147. Clayton, R.J., Iber, F.L., Ruebner, B.H. et al. Byler disease: Fatal familial intrahepatic cholestasis in an Amish kindred. Am J Dis Child, 1969;117:112. 148. Bull, L.N., Van Eijk, M.J., Pawlikowaska, L. et al. A gene encoding a P‐ type ATPase mutated in two forms of hereditary cholestasis. Nat Genet, 1998;18:219. 149. Klomp, L.W., Vargas, J.C., van Mil, S.W. et al. Characterization of mutations in ATP8B1 associated with hereditary cholestasis. Hepatology, 2004;40:27. 150. Chen, F., Ananthanarayanan, M., Emre, S. et al. Progressive familial intrahepatic cholestasis, type 1, is associated with decreased farnesoid X receptor activity. Gastroenterology, 2004;126:756. 151. Summerskill, W.H.J. and Walshe, J.M. Benign recurrent intrahepatic obstructive jaundice. Lancet 1959;2:686. 152. Tygstrup, N. and Jensen, B. Intermittent intrahepatic cholestasis of unknown etiology in five young males from the Faroe Islands. Acta Med Scand, 1969;185:523. 153. Summerfield, J.A., Scott, J., Berman, M. et al. Benign recurrent intrahepatic cholestasis: Studies of bilirubin kinetics, bile acids, and cholangiography. Gut, 1980;21:154. 154. Summerskill, W.H.J. The syndrome of benign recurrent cholestasis. Am J Med, 1965;38:298. 155. Biempica, L., Gutstein, S., and Arias, I.M. Morphological and biochemical studies of benign recurrent cholestasis. Gastroenterology, 1967; 52:521. 156. Thompson, R. and Strautnieks, S. BSEP: Function and role in progressive familial intrahepatic cholestasis. Semin Liver Dis, 2001;21:545. 157. Strautnieks, S.S., Kagalwalla, A.F., Tanner, M.S. et al. Identification of a locus for progressive familial intrahepatic cholestasis (PFIC2) on chromosome 2q24. Am J Hum Genet, 1997;61:630. 158. Stindt, J., Kluge, S., Dröge, C. et al. Bile salt export pump‐reactive antibodies form a polyclonal, multi‐inhibitory response in antibody‐induced bile salt export pump deficiency. Hepatology, 2016;63:524–37. 159. Nies, A.T., Gatmaitan, Z., and Arias, I.M. ATP‐dependent phosphatidylcholine translocation in rat liver canalicular plasma membrane vesicles. J Lipid Res, 1996;37:1125. 160. Strautnieks, S.S., Knisley, A.S., Gerred, S. et al. Mutations in MDR3 in adult onset cholangiopathy, in Bile Acids: From Genomics to Disease and Therapy.

Proceedings of Falk Symposium 129 (eds. D. Keppler, U. Leuschner, G. Paumgartner, and A. Stiehl), Kluwer, Dordrecht, 2003, p. 184. 161. Rosmorduc, O., Hermelin, B., and Poupon, R. MDR3 gene defect in adults with symptomatic intrahepatic and gallbladder cholesterol cholelithiasis. Gastroenterology, 2001;120:1459. 162. Dixon, P.H., Weerasekera, N., Linton, K.J. et al. Heterozygous MDR3 missense mutation associated with intrahepatic cholestasis of pregnancy: evidence for a defect in protein trafficking. Hum Mol Genet, 2000;9:1209. 163. Jacquemin, E., Cresteil, D., Manouvrier, S. et al. Heterozygous non‐sense mutation associated with intrahepatic cholestasis of pregnancy. Lancet, 1999;353:210. 164. de Vree, J., Jacquemin, E., Strum, E. et al. Mutations in the MDR3 gene cause progressive familial intrahepatic cholestasis. Proc Acad Natl Sci U S A, 1998;25:282. 165. de Vree, J.M.L., Ottenhoff Smith, A.J., Aten, J. et al. Rapid correction of Mdr2 deficiency by transplantation of Mdr3 transgenic hepatocytes. Hepatology, 1998;28:387A. 166. Sambrotta, M., Strautnieks, S., Papouli, E. et al. Mutations in TJP2 cause progressive cholestatic liver disease. Nat Genet, 2014;46:326–8. 167. Zhou, S., Hertel, P.M., Finegold, M.J. et al. Hepatocellular carcinoma associated with tight‐junction protein 2 deficiency. Hepatology, 2015;62:1914–16. 168. Drouin, E., Russo, P., Tuchweber, B. et al. North American Indian cirrhosis in children: a review of 30 cases. J Pediatr Gastroenterol Nutr, 2000;31:395. 169. Rapola, J., Heikkila, P., and Fellman, V. (2002) Pathology of lethal fetal growth retardation syndrome with aminoaciduria, iron overload and lactric acidosis (GRACILE). Pediatr Pathol Mol Med, 2002;21:183. 170. Emerick, E.M., Rand, E.B., Goldmuntz, E. et al. Features of Alagille syndrome in 92 patients: frequency and relation to prognosis. Hepatology, 1999;29:822. 171. Piccoli, D.A. and Spinner, N.A. Alagille syndrome and the Jagged1 gene. Semin Liver Dis, 2001;21:525. 172. Oda, T., Elkaholoun, A.G., Meltzer, P.S., and Chandrashekharappa, S.C. Identification and cloning of the human homolog (Jag1) of the rat jagged1 gene from the Alagille syndrome critical region at 20p12. Genomics, 1997;43:376. 173. Friederich, E., Vancompernolle, K., Louvard, D., and Vandekerckhove, J. Villin function in the organization of the actin cytoskeleton. Correlation of in vivo effects to its biochemical activities in vitro. J Biol Chem, 1999;274:26751–60. 174. Phillips, M.J., Azuma, T., Meredith, S.L. et al. Abnormalities in villin gene expression and canalicular microvillus structure in progressive cholestatic liver disease if childhood. Lancet, 2003;362:1090.

21

Hepatic Lipid Droplets in Liver Function and Disease Douglas G. Mashek1,2, Wenqi Cui1, Linshan Shang1, and Charles P. Najt1 Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, MN, USA 2 Department of Medicine, Division of Diabetes, Endocrinology and Metabolism, University of Minnesota, Minneapolis, MN, USA 1

INTRODUCTION Lipid droplet (LD) accumulation is the defining characteristic of non‐alcoholic and alcoholic fatty liver disease. Historically con­ sidered to be inert and simply a sign of disease, LDs are increas­ ingly recognized as etiological factors in numerous liver diseases as well as having important non‐pathological roles in cell signal­ ing and function. These dynamic properties of LDs are highly regulated by the hundreds of proteins that coat the LD surface and control lipid trafficking and flux. In this chapter, we will highlight the major pathways of lipid metabolism that influence LD accumulation, explore key proteins that regulate LD turnover and that link LDs to cellular dysfunction and liver disease.

LIPID UPTAKE AND SYNTHESIS Fatty acids (FAs) serve as the building blocks for the biosynthesis of many complex lipids including triacylglycerols (TAGs), phospholipids, and cholesterol esters. The liver derives these FAs from three main sources  –  direct uptake from the blood, uptake and degradation of lipoprotein remnants, and de novo lipogenesis (DNL). Under fasting conditions, nearly all of the FAs entering the liver are derived from direct uptake as a result of enhanced adipose tissue lipolysis [1]. In the fed state, there is a decreased reliance from adipose‐derived FAs that coincides

with an increase in the contribution of DNL and remnant uptake, both of which are highly influenced by diet composition. Interestingly, the contribution of FAs from DNL is markedly increased in subjects with non‐alcoholic fatty liver disease (NAFLD) and can account for up to nearly 30% of hepatic FAs [2]. The increased flux through the DNL pathway likely contrib­ utes to NAFLD development and its comorbidities. Of special note is the role of dietary sucrose or its monosaccharide con­ stituent fructose in DNL flux. Numerous studies have linked sucrose and/or fructose to NAFLD etiology, which results from high rate of flux of fructose carbons to the liver and its unique metabolism relative to glucose (see [3] for review). Intracellular FAs must first be activated to their respective acyl‐CoAs prior to entrance in downstream metabolic pathways. This occurs in an ATP‐dependent reaction catalyzed by a family of long‐chain acyl‐CoA synthetases (ACSL). In addition, the family of fatty acid transport proteins (FATP) also possess acyl‐ CoA synthetase activity and contribute to FA uptake. Importantly, both the ACSL and FATP families consist of multiple isoforms with different intracellular localizations and substrate specificity [4]. As examples, several isoforms including ACSL3 and 5 and FATP2 and 5 promote TAG synthesis in the liver [5–8], highlighting important roles in trafficking FAs to anabolic pathways. ACSL1 activates and channels FAs to the mitochondria for oxidation in heart and adipose tissue [9, 10], but has minimal effects on FA oxidation in the liver [11]. The isoform responsible for this initial step in hepatic FA β‐oxidation has yet to be elucidated.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

246

THE LIVER:  LIPOLYSIS

TRIACYLGLYCEROL BIOSYNTHESIS AND LIPID DROPLET BIOGENESIS The primary route for disposal of FAs is their esterification to TAG and its subsequent storage in LDs. The coordination of numerous acyltransferase enzymes and the phosphatase enzyme lipin are required for the synthesis of TAG [12]. Each of the enzymes in this pathway has numerous isoforms with different substrate specificity, which influence TAG composition. Examples are the initial (glycerol‐3‐phosphate acyltransferase, GPAT) and terminal (diacylglycerol acyltransferase, DGAT) enzymes in the classical Kennedy pathway of TAG synthesis. GPAT1 shows substrate specificity for saturated FAs, which accounts for the high percentage of saturated FAs in the sn‐1 position of TAG [13]. Additionally, DGAT1 channels exoge­ nous FAs into TAG, whereas DGAT2 is more selective to FAs derived from DNL [14–16]. Thus, the relative activity levels of these various isoforms dictate the acyl composition of TAG. It should also be noted that more recent work has identified mona­ cylglycerol acyltransferases (MGAT), historically thought to exist solely in the intestine, to be present in the liver and have increased expression in models of hepatic steatosis [17]. Although the limited studies on the role of these enzymes are not conclusive [18, 19], the contribution of the monacylglycerol synthetic pathway to hepatic TAG synthesis will likely be a focus of future studies. LDs are thought to develop in the endoplasmic reticulum (ER), where neutral lipids accumulate within the leaflets of the membrane bilayer. As TAG begins to accumulate, it forms a neutral lipid core surrounded by a phospholipid monolayer embedded with proteins. This core and associated phospholipid monolayer eventually bud from the ER, thus forming a cytosolic LD (see [20, 21] for excellent reviews on LD biogenesis). The nascent LDs formed at the ER are small in diameter and are coated with DGAT1 [22]. For LDs to increase in size, they can either undergo fusion with other LDs or synthesize TAG at the surface of LDs. Cell death‐inducing DFF45‐like effector C (CIDEC) is a key protein involved in the fusion of LDs [23, 24]. In response to FAs, CIDEC translocates from the ER to LDs where it initiates fusion and the formation of large LDs [25]. The LD monolayer membrane contains numerous enzymes involved in TAG synthesis. For example, several ACSLs as well as GPAT4 and DGAT2 are present on the LD surface to promote TAG synthesis and expansion of preexisting LDs [22]. As the LD grows with TAG, the phospholipid monolayer must also expand. As such, the synthesis of phosphatidylcholine (PC), the most abundant LD phospholipid, is a critical compo­ nent regulating LD growth and dynamics [26]. Deposition of TAG in the LD core increases the size of the droplet, thereby diluting the relative amount of PC in the phospholipid mon­ olayer. As diacylglycerol (DAG) and phosphatidic acid become enriched at the LD surface, the rate‐limiting enzyme in PC syn­ thesis, CTP:phosphor‐choline cytidylyltransferase α (CCTα), translocates from the nucleus and binds to the LD monolayer. Once there, CCTα becomes activated to facilitate CDP‐choline synthesis. Interestingly, other enzymes critical for PC synthesis are not localized on LDs, suggesting a unique role of CCTα in controlling LD PC metabolism [26].

LIPOLYSIS Catabolism of TAG stored within LDs is a major determinant of LD size and number. Numerous proteins directly or indirectly influence this process and, as a result, impact the development or resolution of hepatic steatosis. Although named after the tis­ sue where it has its highest expression, adipose triglyceride lipase (ATGL) is also involved in hepatic TAG turnover. First discovered in 2004, ATGL is now widely recognized as the pri­ mary cytosolic TAG hydrolase in numerous tissues [27]. Ablation of ATGL in the liver results in TAG accumulation and reduced FA oxidation, while ATGL overexpression alleviates hepatic steatosis; ATGL does not influence very low density lipoprotein (VLDL) secretion [28–30]. The increased FA oxida­ tion is driven via enhanced PGC‐1α/PPAR‐α signaling. We have shown that ATGL drives this transcriptional network through the protein deacetylase sirtuin 1 (SIRT1), which is required for ATGL‐mediated induction of oxidative metabolism [31]. Thus, ATGL appears to be an important signaling node that links LD catabolism to cell signaling as a means to coordinate down­ stream transcriptional networks that control metabolism. The next enzyme in the classical lipolysis pathway is hormone‐­ sensitive lipase (HSL), however, the contribution of hepatocyte HSL to liver TAG catabolism is not known; HSL has been shown to catalyze cholesterol ester hydrolysis in hepatocytes [32]. Monoacylglycerol lipase (MAGL), which catalyzes the last step in hepatic TAG breakdown, has not been studied exclu­ sively in the liver. However, studies in mice lacking MAGL or utilizing MAGL inhibitors reveal that MAGL plays a key role in hepatic injury via decreasing endocannabinoids and, thereby, promoting inflammation [33]. Given the importance of ATGL in catalyzing the initial step in cytosolic TAG hydrolysis, perhaps it is not surprising that ATGL activity is highly regulated. A host of proteins that directly interact with ATGL to influence its activity have been identified. CGI‐58 is widely accepted as the primary coactiva­ tor of ATGL [34]. Although ATGL activity is an important mechanism through which CGI‐58 elicits its effects, it clearly has functions beyond ATGL. For example, CGI‐58 ablation results in an age‐dependent and robust (8‐ to 52‐fold) increase in liver TAG, whereas ATGL ablation only increases hepatic TAG by ~3‐fold [29, 35]. Moreover, CGI‐58 ablation regu­ lates hepatic TAG hydrolysis independent of ATGL, suggest­ ing, at least in the liver, a more complicated role of these proteins in LD turnover [36]. The perilipin (PLIN) family were the first LD proteins to be characterized. They consist of five family members, with vary­ ing structural homology, that act to antagonize ATGL‐catalyzed lipolysis among other functions [37]. PLIN2 is the isoform with the highest expression in the liver and is induced in response to FA loading in cells and NAFLD in both humans and rodents [38]. Liver‐specific ablation of PLIN2 prevents hepatic steatosis and inflammation [39, 40]. Similar to PLIN2, ablation of PLIN3 in the liver also reduces steatosis, but does not influence inflam­ matory markers [41]. PLIN5 is highly expressed in oxidative tissues including the liver and its expression is robustly upregu­ lated in response to fasting [42]. Liver‐specific PLIN5 ablation reduces steatosis, whereas its overexpression promotes steatosis, which is likely due to its anti‐lipolytic effects [43, 44].



21:  Hepatic Lipid Droplets in Liver Function and Disease

In addition to the perilipin proteins discussed above, a host of other proteins have been shown to directly interact with and antagonize ATGL. G0/G1 switch gene 2 (G0S2) is a potent inhibitor of ATGL that influences hepatic energy metabolism. G0S2 overexpression in the liver promotes steatosis and reduces FA oxidation, while ablation of hepatic G0S2 does the opposite [45–47]. Other interacting proteins including pigment epithelial derived factor [48], hypoxia‐inducible protein 2 [49], and CIDEC [50] inhibit ATGL activity directly while UBXD8 tar­ gets it for degradation [51].

Very low density lipoprotein secretion Approximately 70% of hepatic FAs that are destined for VLDL– TAG first transit through the cytosolic LD pool before lipopro­ tein assembly [52, 53]. Members of the carboxylesterase (Ces) family of enzymes contribute to turnover of cytosolic LDs and their trafficking to LDs destined for VLDL secretion. In particu­ lar, ablation of hepatic Ces1d (also known as triglyceride hydro­ lase, TGH), which colocalizes to ER luminal LDs, reduces VLDL secretion and increases the size, but decreases the num­ ber, of cytosolic LDs [54, 55]. It appears that Ces1d reduces the transfer of lipids to existing cytosolic LDs, which may be due to transfer of ER lipids to VLDL for packaging. While other CES proteins have been studied, their role specifically in liver and VLDL secretion has not been characterized. In addition to the carboxylesterases, the LD protein CIDEB plays an essential role in VLDL lipidation. Specifically, ablation of CIDEB reduces VLDL–TAG secretion via an apolipoprotein B (ApoB)‐ dependent mechanism [56]. In contrast, ancient ubiquitous protein 1, which partially colocalizes to LDs, antagonizes ApoB‐medi­ ated VLDL lipidation as a means to suppress VLDL production [57]. These studies also highlight the importance of ApoB as a nexus between LDs and VLDL lipidation. Indeed, ApoB colocal­ izes to LDs on extensions of the ER, which is likely critical in cytosolic transfer of lipids to the developing VLDL particle [58].

LIPOPHAGY Autophagy is a highly conserved and well‐characterized mecha­ nism through which cellular organelles, protein aggregates, and macronutrients are degraded during times of nutrient insuffi­ ciency. Although an extensive body of literature characterizing autophagy of numerous organelles exists, autophagic degrada­ tion of LDs, termed lipophagy, was relatively more recently identified [59] (see [60] for an extensive review). In this pro­ cess, multiple autophagic arms contribute to LD degradation. Macrolipophagy is the classical process through which autophagosomes bud off part of the LD prior to fusing with lysosomes to form autolysosomes, which can then degrade the lipid cargo. Microlipophagy describes the direct interaction and transfer of lipids between LDs and lysosomes. Finally, chaperone‐mediated autophagy is characterized by degradation of select proteins, which also may indirectly influence lipophagy. For example, degradation of specific LD proteins such as PLIN2 via chaperone‐mediated autophagy allows for the subse­ quent degradation of LDs [61]. Although macroautophagy and

247

microautophagy are precisely regulated and considered as pro‐survival mechanisms in response to oxidative stress and nutrient limitation, chaperone‐mediated autophagy serves a housekeeping role in the clearance and recycling of misfolded protein aggregates to maintain cellular integrity. Through the process of macro‐ and micro‐lipophagy autophagosomes and lysosomes, respectively, target LDs for degradation. Once inter­ nalized, the lipids are hydrolyzed by lysosomal acid lipase, the only known lysosomal lipase with activity towards neutral lipids such as TAG and cholesterol ester. The FAs produced from the hydrolysis of TAG and cholesterol ester are then available for β‐oxidation or other downstream pathways. Since its initial discovery, a rapidly growing body of litera­ ture has characterized autophagy/lipophagy in the liver. To date, most studies have evaluated lipophagy in the broader context of global autophagy induction or inhibition. Lipophagy is broadly regulated similar to autophagy; nutrients, insulin, and mTOR antagonize autophagy/lipophagy, whereas depletion of nutrients or increase in activity of low energy sensors (SIRT1, AMPK, etc.) promotes autophagy/lipophagy. At the transcriptional level, autophagy is regulated by a host of transcription factors coinciding with their known roles in nutrient/energy sensing [60]. Consistent with nutrient regulation, many small molecules (caffeine, ginsenoside, and quercetin to name a few) reduce hepatic steatosis through alterations in lipophagy [62–64]. Figure 21.1 provides an overview of the above major pathways involving LDs and key protein mediators of each pathway. Lipolysis of TAG via cytosolic lipases and lipophagy are the two pathways thought to contribute to hepatic LD degradation. We recently explored the interrelationship between the two pathways. These studies identified ATGL to act via SIRT1 sign­ aling as an upstream driver of autophagy/lipophagy [65]. Moreover, autophagy/lipophagy is required for ATGL to drive LD degradation. Consistent with these studies, ablation of PLIN2 also promotes LD degradation in a lipophagy‐dependent manner [66]. Thus, these studies point towards a dominant role of lipophagy in hepatic LD degradation. As our understanding of the lipophagic process and its significance to hepatic LD turnover evolves, this pathway will be undoubtedly targeted through small molecules, as mentioned above, or transcript/pro­ tein targeting as a means to alleviate hepatic steatosis.

LIPID DROPLETS AND LIVER DISEASE Non‐alcoholic fatty liver disease Lipid storage in LDs is a hallmark of NAFLD. Alterations in the balance of LD anabolism and catabolism are integral to the development and resolution of NAFLD. Thus, dozens of studies show that reducing substrates for TAG synthesis, inhibiting enzymes involved in TAG synthesis, or increasing TAG hydrol­ ysis can prevent or alleviate hepatic steatosis. Indeed, ongoing clinical trials are testing these pathways, especially those involved in TAG synthesis. PNPLA3 is an LD protein that belongs to the same family of enzymes as ATGL (i.e. PNPLA2). A nucleotide substitution in PNPLA3 (rs738409, I148M) is widely recognized as the single

248

THE LIVER:  LIPID DROPLETS AND LIVER DISEASE

Figure 21.1  Key lipid droplet (LD) protein mediators of specific lipid metabolic pathways influencing LDs. Proteins that antagonize specific pathways are highlighted in red.

largest genetic predictor of NAFLD [67]. The prevalence of this mutation is ~15–25% in most populations but increases to ~50% in Hispanics [67]. PNPLA3 is highly upregulated in response to high carbohydrate diets and the I148M single‐nucleotide poly­ morphism (SNP) is predictive of steatosis, especially when ­carriers consume a high carbohydrate/sugar diet [68]. Similarly, the PNPLA3 variant is robustly increased on LDs in mice fed high sucrose diets [69]. The mechanism(s) through which PNPLA3 conveys its phenotype is currently under intense debate. Although ample studies show that the I148M variant attenuates LD catabolism [70], other studies suggest a lipogenic role of I148M through its lysophosphatidic acid acyltransferase activity [71]. Regardless of the mechanism of action, the pres­ ence of the I148M variant also progresses NAFLD to liver dam­ age and injury in numerous models of liver disease [72] as discussed in more detail later. In addition to PNPLA3, several other LD proteins are altered in NAFLD and appear to be involved in disease etiology. The first discovered and perhaps most studied perilipin isoform is PLIN1. In adipose tissue, PLIN1 binds CGI‐58 under basal con­ ditions, but phosphorylation in response to lipolytic stimuli dis­ rupts this interaction allowing CGI‐58 to partner with ATGL to promote lipolysis [73]. Although PLIN1 is normally expressed at very low or undetectable levels in the liver, it is upregulated in livers of humans with NAFLD [38, 74]. It remains unknown if this increase is merely a consequence of increased LDs or if PLIN1 plays an etiological role in NAFLD. PLIN2 and PLIN3 are also increased in human NAFLD [38]. Consistent with their anti‐lipolytic roles, liver‐specific PLIN2 or PLIN3 ablation pre­ vents hepatic steatosis [39–41]. The surface of the LD is coated with hundreds of proteins whose presence and/or abundance are dynamic. Although LD proteomics has been conducted in numerous cell types, several studies have characterized the hepatic LD proteome and how it changes in response to fasting/refeeding [75] or NAFLD

[76–78]. From these studies, we can glean that LDs interact with numerous cellular organelles to coordinate metabolism as evidenced by the abundant presence of numerous organelle‐spe­ cific proteins in the LD proteome. Moreover, these studies have expedited the discovery of novel proteins with pathological roles in liver disease. One such example is 17β‐hydroxysteroid dehydrogenase‐13 (17β‐HSD13), which was identified and characterized to be increased in a LD proteomics screen of liver biopsies from NAFLD patients relative to non‐steatotic controls [77]. Overexpression of 17β‐HSD13 in the liver of mice increases steatosis confirming a direct effect on etiology of the disease [77]. Although the underlying mechanism of action is still under investigation, 17β‐HSD13 overexpression in hepato­ cytes increases lipogenesis, suggesting the increased FA supply and subsequent TAG synthesis may be a driving factor underly­ ing its ability to promote steatosis [77]. Since 17β‐HSD13 is thought to be involved in estrogen and androgen metabolism, it has been proposed that it may act to alter local pools of hor­ mones, leading to downstream changes in lipogenesis [79]. Genome‐wide association studies also identified a polymor­ phism (rs641738 C>T) that contains the membrane‐bound O‐acyltransferase domain‐containing 7 gene (MBOAT7, or LPIAT1) and transmembrane channel‐like 4 gene (TMC4) to be associated with increased hepatic fat content mediated by changes in the phosphatidylinositol acyl‐chain remodeling [80, 81]. MBOAT7 is localized on intracellular membranes such as ER, mitochondria, and LDs and functions as a lysophosphoino­ sitol acyltransferase [80]. Another prominent NAFLD‐associated polymorphism (rs58542926, E167K) was identified in trans­ membrane 6 superfamily member 2 (TM6SF2) [82]. Despite being localized in the ER and the ER–Golgi intermediate com­ partment, but not on LDs, TM6SF2 regulates cholesterol metab­ olism and incorporation of polyunsaturated FAs into hepatic TAG [83, 84]. TM6SF2 deficiency causes LD accumulation due, at least in part, to reduced TAG secretion [84].



21:  Hepatic Lipid Droplets in Liver Function and Disease

249

Alcoholic fatty liver disease Relative to NAFLD, our understanding of LDs in alcoholic fatty liver disease (AFLD) is less developed. The I148M polymor­ phism of PNPLA3 is positively associated with AFLD in genome‐wide association studies [85, 86] consistent with its broad role in promoting liver disease. In addition, TM6SF2 (rs58542926) and MBOAT7 (rs641738) genes are also identi­ fied as risk loci for alcohol‐related cirrhosis [87], suggesting the importance of lipid metabolism in the pathogenesis of AFLD. The expression level of PLIN2 increases to coincide with expan­ sion of the LD pool following ethanol exposure [88], suggesting reduced lipolysis may be one factor contributing to increased LDs. Lipophagy and expression of RAG GTPase Rab7, which is required for lipophagy, are greatly inhibited by ethanol, further supporting reduced LD catabolism as a contributing factor to AFLD [83]. Moreover, ethanol exposure inhibits β‐adrenergic induced phosphorylation of HSL and the recruitment of ATGL to the LD surface, leading to decreased lipolysis [89]. More recently, a novel role of ceramide synthase 6 (CerS6) has been uncovered. Previous studies have shown that increased cera­ mide synthesis and accumulation are linked to AFLD [90]. The expression of CerS6, which is involved in ceramide synthesis, is increased in response to alcohol [91], and inhibition of ceramide synthesis attenuates alcohol‐mediated steatosis and liver dys­ function [92, 93]. Moreover, CerS6 drives PLIN2 expression and knockdown of PLIN2 attenuates ceramide production fol­ lowing ethanol exposure [91]. Finally, CerS6 localizes to LDs and interacts with ACSL5, suggesting that ceramides are pro­ duced at the LD surface [91, 94]. Although excess efflux of ethanol carbons to DNL has long been thought to be the driving force in AFLD, these more recent yet limited studies suggest a more complex etiology.

Non‐alcoholic steatohepatitis The transition from simple steatosis to non‐alcoholic steato­ hepatitis (NASH) is a pivotal point in disease pathology. Inflammation and reactive oxygen species are key components in advancing progression to NASH. Since LDs directly interact with mitochondria and can generate inflammatory eicosanoids at the LD surface [95], perhaps it is not surprising that numerous LD proteins are implicated in NAFLD progression. The I148M mutation of PNPLA3, in addition to promoting steatosis, is also a major driver of progression from steatosis to NASH in both adult and pediatric populations [96, 97]. Ablation of CGI‐58 causes NASH in mice [35], while ATGL ablation has no effect on liver fibrosis [29], suggesting lipolysis‐independent effects of CGI‐58 on NAFLD progression. Liver‐specific knockout of PLIN2 alleviates NASH pathologies in a methionine choline‐ deficient (MCD) model of NASH [40]. Although its role in NASH etiology is not known, PLIN1 is highly expressed in adults and children with NASH [98]. In contrast to proteins involved in TAG breakdown, the DGAT enzymes appear to have opposing effects on NASH. Treatment of mice with antisense oligonucleotides (ASOs) targeting DGAT1 attenuates fibrosis, induced by the MCD diet, and reduces stellate cell activation without influencing hepatic steatosis [99]. In contrast, DGAT2 ablation in the same model reduces steatosis, but increases

Figure 21.2  Lipid droplet proteins that link progression of steatosis to non‐alcoholic steatohepatitis (NASH), hepatitis C virus (HCV), hepato­ cellular carcinoma (HCC), and insulin/glucose homeostasis.

hepatic inflammation, lipid peroxidation, and fibrosis [100]. Thus, this terminal step in the TAG synthetic pathway appears to play an important role in coupling/uncoupling steatosis from the liver dysfunction that precedes NASH. Serine/threonine protein kinase 25 (STK25) is one of the few kinases that have been characterized to reside on LDs in hepato­ cytes [101]. Overexpression of STK25 promotes steatosis, liver inflammation, and fibrosis in mice [84]. Moreover, STK25 pro­ tein abundance correlates with steatosis in humans [101]. Although the proteins downstream of STK25 remain to be elu­ cidated, these initial studies suggest that STK25 could play a major role in hepatic LD biology and disease progression through its phosphorylation of LD proteins. Another important aspect of LDs during NASH development is their lipid composition. Specifically, Ioannou et al. observed the presence of cholesterol crystals in LDs subjects with NASH but not simple steatosis [102]. Increasing dietary cholesterol intake in rodents also increases LD cholesterol crystals and pro­ motes inflammation and fibrosis [103, 104]. These studies are consistent with human studies that show increased dietary cho­ lesterol to be an independent risk factor for NASH and cirrhosis [105]. Activated Kupffer cells surround dead hepatocytes laden with cholesterol crystals, suggesting that cell death and signal­ ing to immune cells may be directly involved in explaining how cholesterol crystals promote inflammation [102]. Additionally, PC levels are reduced in subjects with NAFLD and NASH, which suggests that alterations in the LD monolayer could also influence disease progression [106]. Figure 21.2 highlights the key LD proteins involved in the development of NASH and other complications downstream of steatosis.

Insulin resistance and glucose dysregulation Ectopic LD accumulation is correlated with insulin resistance in numerous tissues including the liver. The consequences of hepatic insulin resistance include excess hepatic glucose pro­ duction and VLDL secretion, which contribute to the hyper­ glycemia and hypertriglyceridemia, respectively, commonly present in prediabetes and type 2 diabetes. Although numerous mechanisms underlying the development of insulin resistance exist, alterations in lipid metabolism are a common theme.

250

THE LIVER:  LIPID DROPLETS AND LIVER DISEASE

Accumulation of intermediates in lipid metabolic pathways are thought to be the key drivers of hepatic insulin resistance rather than TAG itself. Intermediates in the lipid biosynthesis pathway including phosphatidic acid, DAG, and ceramides may directly antagonize insulin signaling. Reactive oxygen species may result from the oversupply of FAs to mitochondria and/or mito­ chondrial dysfunction, leading to insulin resistance. Finally, production of inflammatory signaling molecules within hepato­ cytes or hepatic immune cells may also interfere with normal hepatocyte function and insulin signaling. LD dynamics appear to play a major role in the above pro­ cesses that promote insulin resistance. PLIN proteins have sig­ nificant but differing effects on hepatic insulin sensitivity. For example, ablation of PLIN2 or PLIN3 in the liver improves insulin sensitivity, coinciding with reduced steatosis [39–41]. An SNP within the PLIN2 gene impairs VLDL assembly and correlates with type 2 diabetes [107]. In contrast, overexpres­ sion of PLIN5 in the liver increases steatosis, but improves insu­ lin sensitivity [43]. These effects are consistent with studies on PLIN5 in other tissues, including muscle and heart, where LD accumulation is dissociated from insulin resistance [108, 109]. Although the exact role and function of the perilipin proteins remains an active area of research, much of their effects are thought to occur via direct interactions with other LD proteins to influence TAG turnover. One mechanism through which perilipin proteins may work is their interaction with lipases, such as ATGL, to inhibit lipol­ ysis under basal conditions [37]. ATGL overexpression in the liver improves hepatic insulin signaling and reduces DAGs and ceramides in response to high fat feeding without altering glucose tolerance [30]. Knockdown of liver ATGL in high fat fed mice does not affect insulin signaling, but reduces gluco­ neogenesis and improves glucose tolerance without changing DAG levels [110]. ATGL is now recognized to form a regula­ tory loop with FoxO1 in which ATGL drives FoxO1 activity and FoxO1 drives ATGL expression reciprocally [111]. This regulatory circuitry promotes lipolysis and FA oxidation, which support hepatic gluconeogenesis [111]. Knockdown of the ATGL coactivator CGI‐58 increases liver TAG, DAG, and ceramides but improves glucose tolerance and insulin resist­ ance [112]. The inhibitor of ATGL, G0S2, also influences insulin resistance. G0S2 overexpression improves glucose tol­ erance in mice with increased hepatic glucose uptake [45], whereas G0S2 ablation in liver reduces steatosis, but improves whole‐body insulin sensitivity and glucose tolerance despite increased hepatic gluconeogenesis [46]. Ablation of hypoxia‐ inducible factor 2, an ATGL inhibitor, also promotes TAG catabolism and FA oxidation while improving glucose toler­ ance [113]. Taken as a whole, inhibiting lipolysis causes hepatocytes to preferentially take up and burn glucose, which may explain the generally improved glucose tolerance in mod­ els of reduced hepatic lipolysis. In contrast, promoting lipoly­ sis causes less steatosis as a result of enhanced FA oxidation, but also increases gluconeogenesis with variable response to systemic glucose homeostasis depending on the protein manipulated to regulate lipolysis. Similarly, DGATs also influ­ ence insulin sensitivity. DGAT2 overexpression increases TAG, DAG, and ceramides but maintains insulin sensitivity [114]. Thus, while complicated, these studies suggest that

accumulation of TAG and other signaling lipids in the liver can be dissociated from insulin resistance or glucose intolerance, suggesting that there are likely many mechanisms that can link NAFLD to insulin/glucose homeostasis. In contrast to its effects on liver diseases, the PNPLA3 I148M polymorphism appears to uncouple NAFLD from its commonly associated metabolic complications. I148M carriers do not have elevated serum TAG, which is common in subjects with NAFLD [67]. Most studies have reported I148M carriers do not have increased risk for insulin resistance or type 2 diabetes despite the presence of NAFLD [115–117]. The latter study concluded that individuals with the I148M SNP, compared to noncarrier steatotic controls, have increased saturated and monosaturated hepatic lipids and reduced polyunsaturated lipid species, which may explain why the lipids are less toxic [89, 90]. Similar to PNPLA3 I148M, the TM6SF2 E167K variant also does not increase type 2 diabetes risk [118]. It will be of interest for future studies to further clarify the role of the PNPLA3 and TM6SF2 polymorphisms in NAFLD‐related complications and identify other dietary, genetic, or environmental modifiers to this relationship.

Hepatitis C virus The majority of hepatitis C virus (HCV)‐infected patients have steatosis, which is even more pronounced in those with genotype 3A [119]. The increase in LD accumulation is likely due to the fact that the HCV life cycle is tightly intertwined with LD metabolism. Several HCV proteins bind directly to LDs and the blockade of de novo LD formation via DGAT1 inhibition prevents the translocation of the proteins from the ER to LDs and blocks viral replication [120, 121]. Similarly, the LD proteins PLIN3 and Rab18 are also required for direct interaction with HCV proteins and viral replication [122, 123]. HCV itself can promote hepatic steatosis and it appears to do so through the direct inhibition of LD catabolism. HCV infection reduces TAG hydrolysis, in part through indirect inhibition of ATGL‐mediated TAG hydrolysis [124]. Similarly, HCV suppressed the expression of the putative triglyceride lipase arylacetamide deacetylase (AADAC) to reduce lipolysis in the early stages of infection [125]. Although LDs are required for viral replication, VLDL secretion is essential for packaging and export of newly formed virus. Along these lines, ablation of AADAC, which also suppresses VLDL secretion, prevents virus production, suggesting that lipolysis may be required for late‐stage packaging and export of the virus [125]. Consistent with this logic, the ATGL coacti­ vator CGI‐58 is also essential for viral replication [126]. These effects may be due to increased CGI‐58‐mediated VLDL production. Given its role in VLDL secretion, perhaps it is not surprising that CIDEB is also required for HCV release [127]. Nonetheless, CIDEB also directly interacts with the core protein of HCV and is essential for entry and replication of the virus [127], suggesting a prominent role for CIDEB in the entire HCV life cycle. Although these data clearly highlight an essential role of LD proteins and LD metabolism in regulating the HCV life cycle, much remains to be elucidated regarding the molecular underpinnings of this relationship.



21:  Hepatic Lipid Droplets in Liver Function and Disease

Hepatocellular carcinoma Although LD accumulation is a common observation in hepatic neoplastic samples, the specific role of LDs and LD proteins in development of hepatocellular carcinoma (HCC) has not been extensively studied. Given its high penetrance, the I148M poly­ morphism in PNPLA3 also is likely a major driver of HCC since it is positively associated with the development of HCC [128, 129]. These effects are expected based on the robust effects of the I148M SNP on the development of NASH, a known risk factor for HCC [130]. In addition to PNPLA3 (rs738409), TM6SF2 (rs58542926) is also a risk factor for the development of HCC in alcohol‐related cirrhosis [131]. Similar to its pres­ ence in human NAFLD, PLIN1 expression is also increased in neoplastic liver cells although it is not clear if this merely reflects more LDs or if it has a pathological role [132]. Finally, MAGL has been shown to promote HCC via increased inflam­ mation and has been used as a prognostic indicator of HCC [133, 134]. Clearly, much remains to be understood regarding the role of LDs in HCC and related hepatic malignancies.

Not all fatty liver (or LDs) is created equal As our understanding of NAFLD progresses, there is a growing appreciation that NAFLD is a very heterogeneous disease. As discussed earlier, numerous factors including diet, genetics, pre­ disposing diseases (obesity, diabetes, etc.) and HCV can lead to NAFLD through different etiological paths. As a result, mani­ festation of the disease itself, as well as its complications, can vary widely among individuals. These differences also likely contribute to the difficulties in identifying efficacious treatment regiments for NAFLD. Clearly, an important focus of research and diagnostic medicine moving forward should be towards characterizing these distinct forms of NAFLD and targeting therapies specific for each case. Analogous to NAFLD, there is also immense variability in LDs both across the liver and within cells. LD proteins are com­ monly observed to have uneven distribution across LDs within an individual cell. Similarly, the lipid composition of LDs varies greatly; Raman spectroscopy of NAFLD reveals substantial het­ erogeneity of LD composition [135]. Moreover, the liver is highly zonated with periportal and perivenous hepatocytes hav­ ing unique and very different gene signatures and metabolic functions [136]. Beyond differential expression of PLIN pro­ teins [132], almost nothing is known regarding the differences in LDs among these different populations of cells. It would be logical to assume that LD metabolism and signaling differ greatly among and within hepatocytes, which is an area that begs for future investigation.

Perspective LDs are the distinguishing trait of NAFLD and play important roles in both the development and progression of NAFLD, a disease without efficacious and approved treatments. Under­ standing how alterations in the LD proteome and lipidome change to influence disease development and severity and define unique subtypes of NAFLD will be a critical area of research moving forward. New therapeutic strategies including ASOs

251

and CRISPR/Cas9 technologies hold great potential for the liver‐specific targeting of proteins to reverse or attenuate ­d isease pathology. Undoubtedly, the LD will likely be a key target to help alleviate the burden of NAFLD and its comorbidities.

REFERENCES 1. Barrows, B.R. and Parks, E.J. Contributions of different fatty acid sources to very low‐density lipoprotein‐triacylglycerol in the fasted and fed states. J Clin Endocrinol Metab, 2006;91(4):1446–52. 2. Donnelly, K.L., Margosian, M.R., Sheth, S.S., Lusis, A.J., and Parks, E.J. Increased lipogenesis and fatty acid reesterification contribute to hepatic triacylglycerol stores in hyperlipidemic Txnip‐/‐ mice. J Nutr, 2004; 134(6):1475–80. 3. Hannou, S.A., Haslam, D.E., McKeown, N.M., and Herman, M.A. Fructose metabolism and metabolic disease. J Clin Invest, 2018;128(2):545–55. 4. Grevengoed, T.J., Klett, E.L., and Coleman, R.A. Acyl‐CoA metabolism and partitioning. Annu Rev Nutr, 2014;34:1–30. 5. Bu, S.Y. and Mashek, D.G. Hepatic long‐chain acyl‐CoA synthetase 5 medi­ ates fatty acid channeling between anabolic and catabolic pathways. J Lipid Res, 2010;51(11):3270–80. 6. Bu, S.Y., Mashek, M.T., and Mashek, D.G. Suppression of long chain acyl‐ CoA synthetase 3 decreases hepatic de novo fatty acid synthesis through decreased transcriptional activity. J Biol Chem, 2009;284(44):30474–83. 7. Falcon, A., Doege, H., Fluitt, A. et al. FATP2 is a hepatic fatty acid trans­ porter and peroxisomal very long‐chain acyl‐CoA synthetase. AJP Endocrinol Metab, 2010;299:E384–93. 8. Doege, H., Baillie, R.A., Ortegon, A.M. et al. Targeted deletion of FATP5 reveals multiple functions in liver metabolism: alterations in hepatic lipid homeostasis. Gastroenterology, 2006;130:1245–58. 9. Ellis, J.M., Mentock, S.M., Depetrillo, M.A. et al. Mouse cardiac acyl coen­ zyme a synthetase 1 deficiency impairs fatty acid oxidation and induces car­ diac hypertrophy. Mol Cell Biol, 2011;31(6):1252–62. 10. Ellis, J.M., Li, L.O., Wu, P.‐C. et al. Adipose acyl‐CoA synthetase‐1 directs fatty acids toward beta‐oxidation and is required for cold thermogenesis. Cell Metab, 2010;12(1):53–64. 11. Li, L.O., Ellis, J.M., Paich, H.A. et al. Liver‐specific loss of long chain acyl‐ CoA synthetase‐1 decreases triacylglycerol synthesis and beta‐oxidation and alters phospholipid fatty acid composition. J Biol Chem, 2009;284(41): 27816–26. 12. Coleman, R.A. and Mashek, D.G. Mammalian triacylglycerol metabolism: synthesis, lipolysis, and signaling. Chem Rev, 2011;111(10):6359–86. 13. Vancura, A. and Haldar, D. Purification and characterization of glycerophos­ phate acyltransferase from rat liver mitochondria. J Biol Chem, 1994; 269(44):27209–15. 14. Wurie, H.R., Buckett, L., and Zammit, V.A. Diacylglycerol acyltransferase 2 acts upstream of diacylglycerol acyltransferase 1 and utilizes nascent diglyc­ erides and de novo synthesized fatty acids in HepG2 cells. FEBS J, 2012;279(17):3033–47. 15. Qi, J., Lang, W., Geisler, J.G. et al. The use of stable isotope‐labeled glycerol and oleic acid to differentiate the hepatic functions of DGAT1 and ‐2. J Lipid Res, 2012;53(6):1106–16. 16. Villanueva, C.J., Monetti, M., Shih, M. et  al. Specific role for acyl CoA:diacylglycerol acyltransferase 1 (Dgat1) in hepatic steatosis due to exogenous fatty acids. Hepatology, 2009;50(2):434–42. 17. Hall, A.M., Kou, K., Chen, Z. et al. Evidence for regulated monoacylglyc­ erol acyltransferase expression and activity in human liver. J Lipid Res, 2012;53(5):990–9. 18. Soufi, N., Hall, A.M., Chen, Z. et al. Inhibiting monoacylglycerol acyltrans­ ferase 1 ameliorates hepatic metabolic abnormalities but not inflammation and injury in mice. J Biol Chem, 2014;289(43):30177–88. 19. Hall, A.M., Soufi, N., Chambers, K.T. et al. Abrogating monoacylglycerol acyltransferase activity in liver improves glucose tolerance and hepatic insu­ lin signaling in obese mice. Diabetes, 2014;63(7):2284–96. 20. Wilfling, F., Haas, J.T., and Walther, T.C., and Farese, R.V. Jr. Lipid droplet biogenesis. Curr Opin Cell Biol, 2014;29:39–45.

252

THE LIVER:  REFERENCES

21. Pol, A., Gross, S.P., and Parton, R.G. Review: biogenesis of the multifunc­ tional lipid droplet: lipids, proteins, and sites. J Cell Biol, 2014;204(5): 635–46. 22. Wilfling, F., Wang, H., Haas, J.T. et al. Triacylglycerol synthesis enzymes mediate lipid droplet growth by relocalizing from the ER to lipid droplets. Dev Cell, 2013;24(4):384–99. 23. Xu, W., Wu, L., Yu, M. et al. Differential Roles of Cell Death‐inducing DNA fragmentation factor‐α‐like effector (CIDE) proteins in promoting lipid droplet fusion and growth in subpopulations of hepatocytes. J Biol Chem, 2016;291(9):4282–93. 24. Gao, G., Chen, F.‐J., Zhou, L. et  al. Control of lipid droplet fusion and growth by CIDE family proteins. Biochim Biophys Acta Mol Cell Biol Lipids, 2017;1862(10, Part B):1197–204. 25. Li, H., Chen, A., Shu, L. et  al. Translocation of CIDEC in hepatocytes depends on fatty acids. Genes Cells, 2014;19(11):793–802. 26. Krahmer, N., Guo, Y., Wilfling, F. et al. Phosphatidylcholine synthesis for lipid droplet expansion is mediated by localized activation of CTP:phosphocholine cytidylyltransferase. Cell Metab, 2011;14(4):504–15. 27. Zimmermann, R., Strauss, J.G., Haemmerle, G. et  al. Fat mobilization in adipose tissue is promoted by adipose triglyceride lipase. Science, 2004;306(5700):1383–6. 28. Ong, K.T., Mashek, M.T., Bu, S.Y., Greenberg, A.S., and Mashek, D.G. Adipose triglyceride lipase is a major hepatic lipase that regulates triacylg­ lycerol turnover and fatty acid signaling and partitioning. Hepatology, 2011;53(1):116–26. 29. Wu, J.W., Wang, S.P., Alvarez, F. et al. Deficiency of liver adipose triglycer­ ide lipase in mice causes progressive hepatic steatosis. Hepatology, 2011;54(1):122–32. 30. Turpin, S.M., Hoy, A.J., Brown, R.D. et al. Adipose triacylglycerol lipase is a major regulator of hepatic lipid metabolism but not insulin sensitivity in mice. Diabetologia, 2011;54(1):146–56. 31. Khan, S.A., Sathyanarayan, A., Mashek, M.T. et al. ATGL‐catalyzed lipoly­ sis regulates SIRT1 to control PGC‐1α/PPAR‐α signaling. Diabetes, 2015;64(2):418–26. 32. Sekiya, M., Osuga, J.‐I., Yahagi, N. et  al. Hormone‐sensitive lipase is involved in hepatic cholesteryl ester hydrolysis. J Lipid Res, 2008;49(8): 1829–38. 33. Cao, Z., Mulvihill, M.M., Mukhopadhyay, P. et al. Monoacylglycerol lipase controls endocannabinoid and eicosanoid signaling and hepatic injury in mice. Gastroenterology, 2013;144(4):808–17.e15. 34. Lass, A., Zimmermann, R., Haemmerle, G. et al. Adipose triglyceride lipase‐ mediated lipolysis of cellular fat stores is activated by CGI‐58 and defective in Chanarin‐Dorfman syndrome. Cell Metab, 2006;3(5):309–19. 35. Guo, F., Ma, Y., Kadegowda, A.K.G. et al. Deficiency of liver Comparative Gene Identification‐58 causes steatohepatitis and fibrosis in mice. J Lipid Res, 2013;54(8):2109–20. 36. Lord, C.C., Ferguson, D., Thomas, G. et al. Regulation of hepatic triacylg­ lycerol metabolism by CGI‐58 does not require ATGL co‐activation. Cell Rep, 2016;16(4):939–49. 37. Kimmel, A.R. and Sztalryd, C. The perilipins: major cytosolic lipid droplet‐ associated proteins and their roles in cellular lipid storage, mobilization, and systemic homeostasis. Annu Rev Nutr, 2016;36:471–509. 38. Straub, B.K., Stoeffel, P., Heid, H., Zimbelmann, R., and Schirmacher, P. Differential pattern of lipid droplet‐associated proteins and de novo per­ ilipin expression in hepatocyte steatogenesis. Hepatology, 2008;47(6): 1936–46. 39. Imai, Y., Boyle, S., Varela, G.M. et al. Effects of perilipin 2 antisense oligo­ nucleotide treatment on hepatic lipid metabolism and gene expression. Physiol Genomics, 2012;44(22):1125–31. 40. Najt, C.P., Senthivinayagam, S., Aljazi, M.B. et  al. Liver‐specific loss of perilipin 2 alleviates diet‐induced hepatic steatosis, inflammation, and fibro­ sis. Am J Physiol Gastrointest Liver Physiol, 2016;310(9):G726–38. 41. Carr, R.M., Patel, R.T., Rao, V. et al. Reduction of TIP47 improves hepatic steatosis and glucose homeostasis in mice. Am J Physiol Regul Integr Comp Physiol, 2012;302(8):R996–1003. 42. Wolins, N.E., Quaynor, B.K., Skinner, J.R. et al. OXPAT/PAT‐1 is a PPAR‐ induced lipid droplet protein that promotes fatty acid utilization. Diabetes, 2006;55(12):3418–28. 43. Trevino, M.B., Mazur‐Hart, D., Machida, Y. et al. Liver perilipin 5 expres­ sion worsens hepatosteatosis but not insulin resistance in high fat‐fed mice. Mol Endocrinol, 2015;29(10):1414–25.

44. Wang, C., Zhao, Y., Gao, X. et al. Perilipin 5 improves hepatic lipotoxicity by inhibiting lipolysis. Hepatology, 2015;61(3):870–82. 45. Wang, Y., Zhang, Y., Qian, H. et al. The g0/g1 switch gene 2 is an important regulator of hepatic triglyceride metabolism. PloS One, 2013;8(8):e72315. 46. Zhang, X., Xie, X., Heckmann, B.L. et al. Targeted disruption of G0/G1 switch gene 2 enhances adipose lipolysis, alters hepatic energy balance, and alleviates high‐fat diet‐induced liver steatosis. Diabetes, 2014;63(3):934–46. 47. Sugaya, Y. and Satoh, H. Liver‐specific G0 /G1 switch gene 2 (G0s2) expres­ sion promotes hepatic insulin resistance by exacerbating hepatic steatosis in male Wistar rats. J Diabetes, 2017;9(8):754–63. 48. Chung, C., Doll, J.A., Gattu, A.K. et  al. Anti‐angiogenic pigment epithe­ lium‐derived factor regulates hepatocyte triglyceride content through adi­ pose triglyceride lipase (ATGL). J Hepatol, 2008;48(3):471–8. 49. Das, K.M.P., Wechselberger, L., Liziczai, M. et al. Hypoxia‐inducible lipid droplet‐associated protein inhibits adipose triglyceride lipase. J Lipid Res, 2018;59(3):531–41. 50. Grahn, T.H.M., Kaur, R., Yin, J. et al. Fat‐specific protein 27 (FSP27) inter­ acts with adipose triglyceride lipase (ATGL) to regulate lipolysis and insulin sensitivity in human adipocytes. J Biol Chem, 2014;289(17):12029–39. 51. Olzmann, J.A., Richter, C.M., and Kopito, R.R. Spatial regulation of UBXD8 and p97/VCP controls ATGL‐mediated lipid droplet turnover. Proc Natl Acad Sci U S A, 2013;110(4):1345–50. 52. Yang, L.Y., Kuksis, A., Myher, J.J., and Steiner, G. Origin of triacylglycerol moiety of plasma very low density lipoproteins in the rat: structural studies. J Lipid Res, 1995;36(1):125–36. 53. Yang, L.Y., Kuksis, A., Myher, J.J., and Steiner, G. Contribution of de novo fatty acid synthesis to very low density lipoprotein triacylglycerols: evidence from mass isotopomer distribution analysis of fatty acids synthesized from [2H6]ethanol. J Lipid Res, 1996;37(2):262–74. 54. Wang, H., Wei, E., Quiroga, A.D. et  al. Altered lipid droplet dynamics in hepatocytes lacking triacylglycerol hydrolase expression. Mol Biol Cell, 2010;21(12):1991–2000. 55. Lian, J., Wei, E., Wang, S.P. et  al. Liver specific inactivation of carboxy­ lesterase 3/triacylglycerol hydrolase decreases blood lipids without causing severe steatosis in mice. Hepatology, 2012;56(6):2154–62. 56. Ye, J., Li, J.Z., Liu, Y. et al. Cideb, an ER‐ and lipid droplet‐associated pro­ tein, mediates VLDL lipidation and maturation by interacting with apolipo­ protein B. Cell Metab, 2009;9(2):177–90. 57. Zhang, J., Zamani, M., Thiele, C. et al. AUP1 (ancient ubiquitous protein 1) is a key determinant of hepatic very‐low‐density lipoprotein assembly and secretion. Arterioscler Thromb Vasc Biol, 2017;37(4):633–42. 58. Ohsaki, Y., Cheng, J., Suzuki, M., Fujita, A., and Fujimoto, T. Lipid droplets are arrested in the ER membrane by tight binding of lipidated apolipoprotein B‐100. J Cell Sci, 2008;121(14):2415–22. 59. Singh, R., Kaushik, S., Wang, Y. et al. Autophagy regulates lipid metabolism. Nature, 2009;458(7242):1131–5. 60. Schulze, R.J., Sathyanarayan, A., and Mashek, D.G. Breaking fat: the regula­ tion and mechanisms of lipophagy. Biochim Biophys Acta, 2017;1862(10 Pt B):1178–87. 61. Kaushik, S. and Cuervo, A.M. AMPK‐dependent phosphorylation of lipid droplet protein PLIN2 triggers its degradation by CMA. Autophagy, 2016;12(2):432–8. 62. Sinha, R.A., Farah, B.L., Singh, B.K. et al. Caffeine stimulates hepatic lipid metabolism by the autophagy‐lysosomal pathway in mice. Hepatology, 2014;59(4):1366–80. 63. Huang, Q., Wang, T., Yang, L., and Wang, H.‐Y. Ginsenoside Rb2 alleviates hepatic lipid accumulation by restoring autophagy via induction of Sirt1 and activation of AMPK. Int J Mol Sci, 2017;18(5). 64. Zhu, X., Xiong, T., Liu, P. et al. Quercetin ameliorates HFD‐induced NAFLD by promoting hepatic VLDL assembly and lipophagy via the IRE1a/XBP1s pathway. Food Chem Toxicol, 2018;114:52–60. 65. Sathyanarayan, A., Mashek, M.T., and Mashek, D.G. ATGL promotes autophagy/lipophagy via SIRT1 to control hepatic lipid droplet catabolism. Cell Rep, 2017;19(1):1–9. 66. Tsai, T.‐H., Chen, E., Li, L. et  al. The constitutive lipid droplet protein PLIN2 regulates autophagy in liver. Autophagy, 2017;13(7):1130–44. 67. Romeo, S., Kozlitina, J., Xing, C. et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat Genet, 2008;40(12): 1461–5. 68. Davis, J.N., Lê, K.‐A., Walker, R.W. et  al. Increased hepatic fat in over­ weight Hispanic youth influenced by interaction between genetic variation in



21:  Hepatic Lipid Droplets in Liver Function and Disease

PNPLA3 and high dietary carbohydrate and sugar consumption. Am J Clin Nutr, 2010;92(6):1522–7. 69. Smagris, E., BasuRay, S., Li, J. et al. Pnpla3I148M knockin mice accumu­ late PNPLA3 on lipid droplets and develop hepatic steatosis. Hepatology, 2015;61(1):108–11. 70. He, S., McPhaul, C., Li, J.Z. et al. A sequence variation (I148M) in PNPLA3 associated with nonalcoholic fatty liver disease disrupts triglyceride hydroly­ sis. J Biol Chem, 2010;285(9):6706–15. 71. Kumari, M., Schoiswohl, G., Chitraju, C. et al. Adiponutrin functions as a nutritionally regulated lysophosphatidic acid acyltransferase. Cell Metab, 2012;15(5):691–702. 72. Bruschi, F.V., Tardelli, M., Claudel, T., and Trauner, M. PNPLA3 expression and its impact on the liver: current perspectives. Hepatic Med Evid Res, 2017;9:55–66. 73. Subramanian, V., Rothenberg, A., Gomez, C. et al. Perilipin A mediates the reversible binding of CGI‐58 to lipid droplets in 3T3‐L1 adipocytes. J Biol Chem, 2004;279(40):42062–71. 74. Pawella, L.M., Hashani, M., Eiteneuer, E. et al. Perilipin discerns chronic from acute hepatocellular steatosis. J Hepatol, 2014;60(3):633–42. 75. Kramer, D.A., Quiroga, A.D., Lian, J., Fahlman, R.P., and Lehner, R. Fasting and refeeding induces changes in the mouse hepatic lipid droplet proteome. J Proteomics, 2018;181:213–24. 76. Khan, S.A., Wollaston‐Hayden, E.E., Markowski, T.W., Higgins, L., and Mashek, D.G. Quantitative analysis of the murine lipid droplet‐associated proteome during diet‐induced hepatic steatosis. J Lipid Res, 2015;56(12): 2260–72. 77. Su, W., Wang, Y., Jia, X. et al. Comparative proteomic study reveals 17β‐ HSD13 as a pathogenic protein in nonalcoholic fatty liver disease. Proc Natl Acad Sci U S A, 2014;111(31):11437–42. 78. Liu, M., Ge, R., Liu, W. et  al. Differential proteomics profiling identifies LDPs and biological functions in high‐fat diet‐induced fatty livers. J Lipid Res, 2017;58(4):681–94. 79. Zhang, X., Wang, Y., and Liu, P. Omic studies reveal the pathogenic lipid droplet proteins in non‐alcoholic fatty liver disease. Protein Cell, 2017;8(1):4–13. 80. Mancina, R.M., Dongiovanni, P., Petta, S. et al. The MBOAT7‐TMC4 vari­ ant rs641738 increases risk of nonalcoholic fatty liver disease in individuals of European descent. Gastroenterology, 2016;150(5):1219–30.e6. 81. Luukkonen, P.K., Zhou, Y., Hyötyläinen, T. et  al. The MBOAT7 variant rs641738 alters hepatic phosphatidylinositols and increases severity of non‐ alcoholic fatty liver disease in humans. J Hepatol, 2016;65(6):1263–5. 82. Kozlitina, J., Smagris, E., Stender, S. et al. Exome‐wide association study identifies a TM6SF2 variant that confers susceptibility to nonalcoholic fatty liver disease. Nat Genet, 2014;46(4):352–6. 83. Mahdessian, H., Taxiarchis, A., Popov, S. et al. TM6SF2 is a regulator of liver fat metabolism influencing triglyceride secretion and hepatic lipid droplet content. Proc Natl Acad Sci U S A, 2014;111(24):8913–18. 84. Luukkonen, P.K., Zhou, Y., Nidhina Haridas, P.A. et  al. Impaired hepatic lipid synthesis from polyunsaturated fatty acids in TM6SF2 E167K variant carriers with NAFLD. J Hepatol, 2017;67(1):128–36. 85. Tian, C., Stokowski, R.P., Kershenobich, D., Ballinger, D.G., and Hinds, D.A. Variant in PNPLA3 is associated with alcoholic liver disease. Nat Genet, 2010;42(1):21–3. 86. Chamorro, A.‐J., Torres, J.‐L., Mirón‐Canelo, J.‐A. et al. Systematic review with meta‐analysis: the I148M variant of patatin‐like phospholipase domain‐ containing 3 gene (PNPLA3) is significantly associated with alcoholic liver cirrhosis. Aliment Pharmacol Ther, 2014;40(6):571–81. 87. Buch, S., Stickel, F., Trépo, E. et al. A genome‐wide association study con­ firms PNPLA3 and identifies TM6SF2 and MBOAT7 as risk loci for alcohol‐ related cirrhosis. Nat Genet, 2015;47(12):1443–8. 88. Mak, K.M., Ren, C., Ponomarenko, A., Cao, Q., and Lieber, C.S. Adipose differentiation‐related protein is a reliable lipid droplet marker in alcoholic fatty liver of rats. Alcohol Clin Exp Res, 2008;32(4):683–9. 89. Schott, M.B., Rasineni, K., Weller, S.G. et  al. β‐Adrenergic induction of lipolysis in hepatocytes is inhibited by ethanol exposure. J Biol Chem, 2017;292(28):11815–28. 90. Carr, R.M., Dhir, R., Yin, X., Agarwal, B., and Ahima, R.S. Temporal effects of ethanol consumption on energy homeostasis, hepatic steatosis, and insulin sensitivity in mice. Alcohol Clin Exp Res, 2013;37(7):1091–9. 91. Williams, B., Correnti, J., Oranu, A. et al. A novel role for ceramide synthase 6 in mouse and human alcoholic steatosis. FASEB J, 2018;32(1):130–42.

253

92. Lizarazo, D., Zabala, V., Tong, M., Longato, L., and de la Monte, S.M. Ceramide inhibitor myriocin restores insulin/insulin growth factor signal­ ing for liver remodeling in experimental alcohol‐related steatohepatitis. J Gastroenterol Hepatol, 2013;28(10):1660–8. 93. Tong, M., Longato, L., Ramirez, T. et al. Therapeutic reversal of chronic alcohol‐related steatohepatitis with the ceramide inhibitor myriocin. Int J Exp Pathol, 2014;95(1):49–63. 94. Senkal, C.E., Salama, M.F., Snider, A.J. et al. Ceramide is metabolized to acylceramide and stored in lipid droplets. Cell Metab, 2017;25(3):686–97. 95. Bozza, P.T., Bakker‐Abreu, I., Navarro‐Xavier, R.A., and Bandeira‐Melo, C. Lipid body function in eicosanoid synthesis: an update. Prostaglandins Leukot Essent Fatty Acids, 2011;85(5):205–13. 96. Singal, A.G., Manjunath, H., Yopp, A.C. et al. The effect of PNPLA3 on fibrosis progression and development of hepatocellular carcinoma: a meta‐ analysis. Am J Gastroenterol, 2014;109(3):325–34. 97. Valenti, L., Alisi, A., Galmozzi, E. et al. I148M patatin‐like phospholipase domain‐containing 3 gene variant and severity of pediatric nonalcoholic fatty liver disease. Hepatology, 2010;52(4):1274–80. 98. Carr, R.M., Dhir, R., Mahadev, K. et  al. Perilipin staining distinguishes between steatosis and non‐alcoholic steatohepatitis in adults and children. Clin Gastroenterol, 2017;15(1):145–7. 99. Yamaguchi, K., Yang, L., McCall, S. et al. Diacylglycerol acyltranferase 1 anti‐sense oligonucleotides reduce hepatic fibrosis in mice with nonalco­ holic steatohepatitis. Hepatology, 2008;47(2):625–35. 100. Yamaguchi, K., Yang, L., McCall, S. et al. Inhibiting triglyceride synthesis improves hepatic steatosis but exacerbates liver damage and fibrosis in obese mice with nonalcoholic steatohepatitis. Hepatology, 2007;45(6): 1366–74. 101. Amrutkar, M., Cansby, E., Nuñez‐Durán, E. et al. Protein kinase STK25 regulates hepatic lipid partitioning and progression of liver steatosis and NASH. FASEB J, 2015;29(4):1564–76. 102. Ioannou, G.N., Haigh, W.G., Thorning, D., and Savard, C. Hepatic choles­ terol crystals and crown‐like structures distinguish NASH from simple steatosis. J Lipid Res, 2013;54(5):1326–34. 103. Ioannou, G.N., Subramanian, S., Chait, A. et al. Cholesterol crystallization within hepatocyte lipid droplets and its role in murine NASH. J Lipid Res, 2017;58(6):1067–79. 104. Savard, C., Tartaglione, E.V., Kuver, R. et  al. Synergistic interaction of dietary cholesterol and dietary fat in inducing experimental steatohepatitis. Hepatology, 2013;57(1):81–92. 105. Yasutake, K., Nakamuta, M., Shima, Y. et al. Nutritional investigation of non‐obese patients with non‐alcoholic fatty liver disease: the significance of dietary cholesterol. Scand J Gastroenterol, 2009;44(4):471–7. 106. Puri, P., Baillie, R.A., Wiest, M.M. et al. A lipidomic analysis of nonalco­ holic fatty liver disease. Hepatology, 2007;46(4):1081–90. 107. Sentinelli, F., Capoccia, D., Incani, M. et al. The perilipin 2 (PLIN2) gene Ser251Pro missense mutation is associated with reduced insulin secretion and increased insulin sensitivity in Italian obese subjects. Diabetes Metab Res Rev, 2016;32(6):550–6. 108. Bosma, M., Sparks, L.M., Hooiveld, G.J. et al. Overexpression of PLIN5 in skeletal muscle promotes oxidative gene expression and intramyocellular lipid content without compromising insulin sensitivity. Biochim Biophys Acta, 2013;1831(4):844–52. 109. Kuramoto, K., Okamura, T., Yamaguchi, T. et al. Perilipin 5, a lipid droplet‐ binding protein, protects heart from oxidative burden by sequestering fatty acid from excessive oxidation. J Biol Chem, 2012;287(28):23852–63. 110. Ong, K.T., Mashek, M.T., Bu, S.Y., and Mashek, D.G. Hepatic ATGL knockdown uncouples glucose intolerance from liver TAG accumulation. FASEB J, 2013;27(1):313–21. 111. Zhang, W., Bu, S.Y., Mashek, M.T. et al. Integrated regulation of hepatic lipid and glucose metabolism by adipose triacylglycerol lipase and FoxO proteins. Cell Rep, 2016;15(2):349–59. 112. Brown, J.M., Betters, J.L., Lord, C. et  al. CGI‐58 knockdown in mice causes hepatic steatosis but prevents diet‐induced obesity and glucose intol­ erance. J Lipid Res, 2010;51(11):3306–15. 113. DiStefano, M.T., Danai, L.V., Roth Flach, R.J. et al. The lipid droplet pro­ tein hypoxia‐inducible gene 2 promotes hepatic triglyceride deposition by inhibiting lipolysis. J Biol Chem, 2015;290(24):15175–84. 114. Monetti, M., Levin, M.C., Watt, M.J. et al. Dissociation of hepatic steatosis and insulin resistance in mice overexpressing DGAT in the liver. Cell Metab, 2007;6(1):69–78.

254

THE LIVER:  REFERENCES

115. Luukkonen, P.K., Zhou, Y., Sädevirta, S. et al. Ceramides dissociate steato­ sis and insulin resistance in the human liver in non‐alcoholic fatty liver disease. J Hepatol, 2016;64(5):1167–75. 116. Speliotes, E.K., Butler, J.L., Palmer, C.D. et al. PNPLA3 variants specifi­ cally confer increased risk for histologic nonalcoholic fatty liver disease but not metabolic disease. Hepatology, 2010;52(3):904–12. 117. Sookoian, S. and Pirola, C.J. Meta‐analysis of the influence of I148M vari­ ant of patatin‐like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease. Hepatology, 2011;53(6):1883–94. 118. Pirola, C.J. and Sookoian, S. The dual and opposite role of the TM6SF2‐ rs58542926 variant in protecting against cardiovascular disease and confer­ ring risk for nonalcoholic fatty liver: a meta‐analysis. Hepatology, 2015;62(6):1742–56. 119. Adinolfi, L.E., Gambardella, M., Andreana, A. et al. Steatosis accelerates the progression of liver damage of chronic hepatitis C patients and corre­ lates with specific HCV genotype and visceral obesity. Hepatology, 2001;33(6):1358–64. 120. Herker, E., Harris, C., Hernandez, C. et al. Efficient hepatitis C virus parti­ cle formation requires diacylglycerol acyltransferase‐1. Nat Med, 2010;16(11):1295–8. 121. Camus, G., Herker, E., Modi, A.A. et  al. Diacylglycerol acyltransferase‐1 localizes hepatitis C virus NS5A protein to lipid droplets and enhances NS5A interaction with the viral capsid core. J Biol Chem, 2013; 288(14):9915–23. 122. Vogt, D.A., Camus, G., Herker, E. et  al. Lipid droplet‐binding protein TIP47 regulates hepatitis C Virus RNA replication through interaction with the viral NS5A protein. PLoS Pathog, 2013;9(4):e1003302. 123. Salloum, S., Wang, H., Ferguson, C., Parton, R.G., and Tai, A.W. Rab18 binds to hepatitis C virus NS5A and promotes interaction between sites of viral replication and lipid droplets. PLoS Pathog, 2013;9(8):e1003513. 124. Camus, G., Schweiger, M., Herker, E. et al. The hepatitis C virus core protein inhibits adipose triglyceride lipase (ATGL)‐mediated lipid mobilization and enhances the ATGL interaction with comparative gene identification 58 (CGI‐58) and lipid droplets. J Biol Chem, 2014; 289(52):35770–80. 125. Nourbakhsh, M., Douglas, D.N., Pu, C.H. et al. Arylacetamide deacetylase: a novel host factor with important roles in the lipolysis of cellular triacylg­

lycerol stores, VLDL assembly and HCV production. J Hepatol, 2013; 59(2):336–43. 126. Vieyres, G., Welsch, K., Gerold, G. et al. ABHD5/CGI‐58, the Chanarin‐ Dorfman syndrome protein, mobilises lipid stores for hepatitis C virus production. PLoS Pathog, 2016;12(4):e1005568. 127. Cai, H., Yao, W., Li, L. et  al. Cell‐death‐inducing DFFA‐like effector B contributes to the assembly of hepatitis C virus (HCV) particles and inter­ acts with HCV NS5A. Sci Rep, 2016;6:27778. 128. Liu, Y.‐L., Patman, G.L., Leathart, J.B.S. et al. Carriage of the PNPLA3 rs738409 C >G polymorphism confers an increased risk of non‐alcoholic fatty liver disease associated hepatocellular carcinoma. J Hepatol, 2014;61(1):75–81. 129. Krawczyk, M., Stokes, C.S., Romeo, S., and Lammert, F. HCC and liver disease risks in homozygous PNPLA3 pI148M carriers approach mono­ genic inheritance. J Hepatol, 2015;62(4):980–1. 130. Cholankeril, G., Patel, R., Khurana, S., and Satapathy, S.K. Hepatocellular carcinoma in non‐alcoholic steatohepatitis: current knowledge and implica­ tions for management. World J Hepatol, 2017;9(11):533–43. 131. Stickel, F., Buch, S., Nischalke, H.D. et al. Genetic variants in PNPLA3 and TM6SF2 predispose to the development of hepatocellular carcinoma in individuals with alcohol‐related cirrhosis. Am J Gastroenterol, 2018; 113(10):1475–83. 132. Straub, B.K., Herpel, E., Singer, S. et al. Lipid droplet‐associated PAT‐ proteins show frequent and differential expression in neoplastic steatogen­ esis. Mod Pathol, 2010;23(3):480–92. 133. Zhu, W., Zhao, Y., Zhou, J. et al. Monoacylglycerol lipase promotes progression of hepatocellular carcinoma via NF‐κB‐mediated epithelial‐mesenchymal transition. J Hematol Oncol, 2016;9(1):127. 134. Zhang, J., Liu, Z., Lian, Z. et al. Monoacylglycerol lipase: a novel potential therapeutic target and prognostic indicator for hepatocellular carcinoma. Sci Rep, 2016;6:35784. 135. Kochan, K., Maslak, E., Krafft, C. et al. Raman spectroscopy analysis of lipid droplets content, distribution and saturation level in non‐alcoholic fatty liver disease in mice. J Biophotonics, 2015;8(7):597–609. 136. Guzmán, M. and Castro, J. Zonation of fatty acid metabolism in rat liver. Biochem J, 1989;264(1):107–13.

22

Lipoprotein Metabolism and Cholesterol Balance Mariana Acuña‐Aravena and David E. Cohen Division of Gastroenterology and Hepatology, Joan & Sanford I. Weill Department of Medicine, Weill Cornell Medical College, New York, NY, USA

INTRODUCTION Lipids are insoluble or sparingly soluble molecules that are essential for membrane biogenesis and maintenance of membrane integrity. They also serve as energy sources, hormone precursors, and signaling molecules. In order to facilitate transport through the relatively aqueous blood, nonpolar lipids, such as cholesteryl esters or triglycerides, are packaged within lipoproteins. Increased concentrations of certain lipoproteins in the circulation are associated strongly with atherosclerosis. Much of the prevalence of cardiovascular disease, the leading cause of death in the United States and most Western countries, can be attributed to elevated plasma concentrations of cholesterol‐rich low‐density lipoprotein (LDL) particles, as well as lipoproteins that are rich in triglycerides. Epidemiologically, decreased concentrations of high‐density lipoprotein (HDL) cholesterol also predispose to atherosclerotic disease. This chapter highlights the biochemistry and ­physiology of cholesterol and lipoproteins. Because abundant clinical outcomes data have proven that morbidity and mortality from cardiovascular disease can be reduced by the use of lipid‐­lowering drugs, mechanisms of pharmacologic interventions that can ameliorate hyperlipidemia will be discussed.

BIOCHEMISTRY AND PHYSIOLOGY OF CHOLESTEROL AND LIPOPROTEIN METABOLISM Lipoproteins are macromolecular aggregates that transport triglycerides and cholesterol through the blood. Circulating ­ lipoproteins can be differentiated on the basis of density, size,

and protein content (Table 22.1). As a general rule, larger, less dense lipoproteins have a greater percentage composition of lipids; chylomicrons are the largest and least dense lipoprotein subclass, whereas HDL are the smallest lipoproteins, containing the lowest lipid content and the highest proportion of protein. Structurally, lipoproteins are microscopic spherical particles ranging from 7 to 100 nm in diameter. Lipoprotein particles consist of a monolayer of polar, amphipathic lipids surrounding a hydrophobic core. Each lipoprotein particle also contains one or more types of apolipoprotein (Table 22.1). The polar lipids that comprise the surface coat are unesterified cholesterol and phospholipid molecules, arranged in a monolayer. The hydrophobic core of a lipoprotein contains the cholesteryl esters (cholesterol molecules linked by an ester bond to a fatty acid) and triglycerides (three fatty acids esterified to a glycerol molecule). The apolipoproteins (also referred to as apoproteins) are amphipathic proteins that intercalate into the lipid membrane of lipoproteins. In addition to stabilizing the structure of lipoproteins, apolipoproteins engage in biological functions. They may act as receptor ligands for lipoprotein particles or may activate enzymatic activities in the plasma. As will be discussed, the apolipoprotein composition determines the metabolic fate of the lipoprotein. From a metabolic perspective, lipoprotein particles can be divided into lipoproteins that participate in the delivery of triglyceride molecules to muscle and fat tissue (the apolipoprotein B (apoB)‐containing lipoproteins: chylomicrons and very‐low‐ density lipoprotein, VLDL) and lipoproteins that are involved primarily in cholesterol transport (HDL and the remnants of apoB‐containing lipoproteins). HDL also serves as a reservoir for exchangeable apolipoproteins in the plasma, including apoA‐I, apoC‐II, and apoE. The following discussion presents each lipoprotein class in the context of its function.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

256

THE LIVER:  BIOCHEMISTRY AND PHYSIOLOGY OF CHOLESTEROL AND LIPOPROTEIN METABOLISM

Table 22.1  Characteristics of plasma lipoproteins Density (g mL ) Diameter (nm) Total lipid (% wt) Composition, % dry weight Protein Triglycerides Unesterified cholesterol + cholesterol ester Phospholipids (% wt lipid) Electrophoretic mobilitya Major apolipoproteins −1

CM

VLDL

IDL

LDL

HDL

G (S78G) mutation in BSC1L [150, 151]. BCS1L encodes a mitochondrial inner‐membrane protein necessary for the assembly of mitochondrial respiratory chain complex III. Typically, patients with GRACILE have fulminant lactic acidosis soon after birth, with nonspecific aminoaciduria and cholestasis. Prominent siderosis is noted in the hepatocytes and reticuloendothelial system but typical histopathologic features of mitochondrial dysfunction such as microvesicular steatosis are not found. GRACILE syndrome patients have no neurological abnormalities or dysmorphic features and normal respiratory chain function, including normal complex III activity. Other mutations described in BSC1L cause a variety of phenotypes including neurological disease, generalized mitochondriopathies and, most recently, Bjørnstad syndrome, a form of sensorineural hearing loss associated with pili torti [152–156].



29:  Molecular Cholestasis

Severity of effects of BSC1L mutation appears correlated with increased generation of reactive oxygen species [154]. A recent review showed that whilst the collection of symptoms in GRACILE is specific to the 232A>G (S78G) mutation, other mutations in BCS1L can cause liver disease. The early diagnosis is important as data shows that a ketogenic diet may improve hepatic symptoms [157].

SLC25A13 disease (citrullinemia type II, neonatal intrahepatic cholestasis caused by citrin deficiency, NICCD) SLC25A13 (CITRIN) deficiency was initially described in adult‐onset citrullinemia type II (CTLN2), an autosomal recessive disorder with variable age of presentation characterized by intermittent encephalopathy with hyperammonemia [158–160]. CTLN2 is due to decreased activity of hepatic argininosuccinate synthase, which leads to hyperammonemia, coma, and may cause death due to acute metabolic decompensation within a few years of onset [160]. More recently mutations in SLC25A13 were discovered in patients with an inherited form of NH, now known as neonatal intrahepatic cholestasis caused by CITRIN deficiency (NICCD). Although initially described predominantly in Japanese patients it was then recognized in patients from neighboring Korea, Taiwan, and China. We now appreciate that this disorder occurs in all ethnic groups and is important to diagnose as specialized management may not only cause clinical improvement but could prevent the onset of the adult form of the disease [161, 162]. A summary of the clinical characteristics of NICCD in 75 patients reported intrahepatic cholestasis with raised gGT, acholic stools, and poor weight gain in most. A NH syndrome was associated with hypocoagulability, hypoglycemia, disturbances of liver synthetic function, galactosuria, and, rarely, hyperammonemia. A typical but transient feature in NICCD was found to be abnormal plasma amino acid concentrations [161]. Most patients had raised citrulline, methionine, arginine, and threonine to serine ratio values. Raised tyrosine and lysine values are also seen in some patients [163]. Timely diagnosis of NICCD is important in view of specific dietary treatment (increased protein intake, initial treatment with a galactose‐free diet, and avoidance of high carbohydrate intake). The standard dietary management of NH could be partially toxic in this condition, precipitating deterioration of liver disease and even requiring transplantation [161]. NICCD is typically not severe although some cases required liver transplantation [163]. The proportion of patients who progress from NICCD to CTLN2 is unknown. An additional ­benefit of accurate diagnosis may be in the introduction of cancer surveillance and early intervention as hepatocellular carcinoma and cholangiocarcinoma are known to complicate CITRIN deficiency [164]. Citrin is a liver‐specific mitochondrial aspartate‐glutamate carrier. Investigation of the Slc25a13 knockout (Ctrn−/−) mouse model revealed markedly decreased activities in aspartate transport and in the malate‐aspartate mitochondrial shuttle. Deficits in ureagenesis from ammonia and in gluconeogenesis from lactate also could be demonstrated [165]. Ctrn−/− mice failed to show CTLN2‐like symptoms; this suggests that citrin

357

deficiency alone may not be sufficient to produce a CTLN2‐ like phenotype in mice. These observations are in parallel with the variable age of onset and incomplete penetrance of CTLN2, where involvement of additional environmental or genetic ­factors is suspected.

CIRH1A deficiency (North American Indian childhood cirrhosis, NAIC) NAIC is a form of neonatal cholestasis that progresses to cirrhosis. It has been recognized only in the Ojibway‐Cree ­ ­population from northwestern Quebec [166, 167]. Children may initially present with transient neonatal jaundice which progresses to cirrhosis. Management of the liver disease requires liver transplantation in childhood or early adulthood. Patients have elevated gGT levels and the characteristic histological appearance of this condition, with portal‐tract fibrosis, may resemble that of extrahepatic biliary atresia. A missense mutation in CIRH1A, which encodes CIRHIN, was identified in these children [167]. It was found that CIRHIN is a ribonucleoprotein that is required for maturation of the 18S rRNA of the ribosome. The NAIC mutation is likely to affect the function of CIRHIN in ribosome biogenesis and pre‐ribosome assembly by disrupting interaction with NOL11 [168, 169].

Lymphoedema‐cholestasis syndrome Lymphoedema‐cholestasis syndrome (LCS, also known as Aagenaes syndrome) was originally described in Norwegian patients [170], and refers to an autosomal recessive disorder characterized by severe neonatal cholestasis with an elevated ­ serum gGT. Although in most patients investigated liver disease progressed to cirrhosis, some requiring liver transplantation in childhood, more than 50% of patients survive into adulthood [171, 172]. In addition, patients develop severe lymphoedema, which may be manifest at birth or appear later in childhood [173]. Norwegian families with LCS share genetic linkage to LCS1 locus on chromosome 15 [173], however, the causal gene mutations have not yet been identified. Mutations in CCBE1 (collagen and calcium‐binding EGF domain‐1) were also reported in pediatric and adult patients with LCS from different ethnic backgrounds [174, 175]. CCBE1 is known to be important for embryonic lymphangiogenesis and mutations in CCBE1 were identified in cases of lymphatic dysplasia and hydrops fetalis [176, 177].

Other inherited conditions associated with cholestasis A number of other conditions not mentioned here but thoroughly reviewed elsewhere may present with neonatal and infantile cholestasis. The list of the disorders includes alpha‐1 antitrypsin deficiency, Niemann–Pick disease type C, galactosemia, hereditary fructose deficiency, fatty acid oxidation defects, cystic fibrosis, and indeed many other infective or immune‐mediated conditions in which hepatocellular injury leads nonspecifically to impairment of pathways involved in handling of the constituents of bile [7, 178]. Each of the conditions mentioned above has specific distinguishing features and baseline metabolic investigations should help in diagnosis.

358

THE LIVER:  INFANTILE CHOLESTATIC SYNDROMES WITHOUT OBVIOUS GENETIC CAUSE

Table 29.3  Syndromic forms of cholestasis Clinical syndromes

Hepatic manifestations

Extrahepatic manifestations

Genetic defect

Alagille syndrome

Neonatal giant‐cell hepatitis with high GGT in some patients. Chronic cholestasis with bile duct loss in most patients with fully penetrant phenotype. Hepatocellular carcinoma may occur as early as four years of age

JAG1 or NOTCH2 Monoallelic mutation, mostly with complete loss of function

NISCH

Neonatal cholestasis with high gGT and hepatomegaly. Liver biopsy shows extensive fibrosis and cholangiography shows sclerosing cholangitis. Portal hypertension detected after age five years NH with high GGT. Liver biopsy features of hepatocyte microvesicular steatosis and cholestasis. Patients may develop hepatocellular carcinoma

Cardiovascular: typically peripheral pulmonary stenosis, various other congenital cardiac malformations. Renal: possibly more common in NICCD. Various forms of dysplasia, cystic kidneys, renal tubular insufficiency (hematuria, proteinuria), renal tubular acidosis. Ocular: posterior embryotoxon. Facial: pointed chin, straight nose, deep set eyes, broad forehead. Skeletal: butterfly vertebrae Dry and scaly skin. Ichthyosis most prominent on abdomen and limbs. Hypotrichosis and alopecia. Intracytoplasmic vacuoles in peripheral blood eosinophils. Significant improvement after liver transplantation Intermittent hyperammonemia, confusion, coma, cerebral edema occurs in the adult form of CITRIN deficiency

NICCD Citrullinemia type II (adult onset) GRACILE syndrome

Hepatocellular cholestasis and giant‐cell changes, progressing to intralobular fibrosis; prominent siderosis. Bile duct paucity may be seen ARC syndrome Neonatal cholestasis with low GGT, abnormal bile excretion on hepatobiliary scan with trimethylbromoimino‐diacetic acid (TBIDA) scan, lipofuscin granule accumulation in hepatocytes and interlobular bile duct hypoplasia. Patients typically have very severe failure to thrive

Lymphoedema‐ Transient neonatal cholestasis with high GGT. cholestasis Liver disease progresses to cirrhosis in most. syndrome Transient episodes of cholestasis in adulthood NAIC

Transient neonatal cholestasis progressing to cirrhosis. High GGT. Liver transplantation is required in childhood/early adulthood. Histopathological features of severe cholangiopathy

Growth retardation, lactic acidosis, aminoaciduria, and iron deposition in hepatocytes and reticuloendothelial system. Pancreatic fibrosis. Death in infancy Neuromuscular: neurogenic arthrogryposis. The severity may vary from unilateral talipes to multiple joint involvement. Dysplasia of corpus callosum in at least 20%. Renal: renal fanconi syndrome and nephrogenic diabetes insipidus. Both can vary in severity. Hematology: deficiency of alpha granules resulting in hypogranular appearance of platelets. Abnormal platelet aggregation. Skin: ichthyosis, cutis laxa. Dysmorphism: low set ears, hirsutism, proximal insertion of thumbs Peripheral lymphoedema

Not known

INFANTILE CHOLESTATIC SYNDROMES WITHOUT OBVIOUS GENETIC CAUSE Cholestatic jaundice is the commonest presentation of liver ­disease in infancy, affecting approximately 1 in every 2500 infants. Cholestasis has multiple etiologies with only subtle clinical differences among various diseases. Obstruction to bile flow – whether mechanical, as in extrahepatic biliary atresia, or functional, as in PFIC  –  increases intrahepatocyte bile acid ­concentration and causes damage which, if not corrected, leads to cell death. Jaundice, acholic stools, hepatomegaly, and the consequences of fat malabsorption are among the manifestations, joint or several, of cholestasis [179, 180]. The most populated categories of cholestatic jaundice in the first months of life, historically accounting for up to 70% of cases [179, 180], are the syndromes of extrahepatic biliary atresia (EHBA) and NH. In most patients with either syndrome the cause was then unknown, although some cases were already linked to genetic, infectious, or other environmental factors [181, 182]. EHBA is the most frequent form of neonatal cholestasis, occurring in approximately 1 in 10 000 births, and is more common in

CLDN1 Mutation of both alleles SLC25A13 (CITRIN) Mutation of both alleles BCSL1 Mutation of both alleles VPS33B or VIPAS39 Mutation of both alleles

LCS1 locus on chromosome 15q26.1 CCBE1 mutations, on both alleles, in patients without linkage to LCS1 CIRH1A Mutation of both alleles

girls. Concentrations of GGT activity in serum rise in EHBA. EHBA is characterized by complete fibrotic obliteration of the lumen of all or part of the extrahepatic biliary tree within the first three postnatal months [183]. Luminal obliteration and fibrosis of intrahepatic bile ducts also may be seen [184]. Approximately 20% of patients with biliary atresia also have at least one other major congenital anomaly. This finding suggests that genetically determined defects underlie some of these cases and that the same genes are involved in regulation of development of both the biliary tract and other organs [185, 186]. Polysplenia syndrome (polysplenia, midline liver, interrupted inferior vena cava, situs inversus, preduodenal portal vein, and malrotation of the intestine) in particular is present in 10% of all children with biliary atresia [187, 188]. Abnormal situs coupled with bile‐duct disease suggests that genes involved in shaping the laterality of thoracic and abdominal organs are involved in bile duct development. Numerous authors have suggested virus‐induced mechanisms in the etiopathogenesis of EHBA. The rationale for this is reviewed elsewhere [189, 190]. Interestingly, EHBA has been reported as a rare association of primary ciliary dyskinesia [191]. Indeed, the inv mouse, with



29:  Molecular Cholestasis

a defect in the murine orthologue of INVS, has EHBA associated with situs inversus [192, 193]. INVS mutations were identified as one of the causes of nephronophthisis, an autosomal recessive syndrome of cystic kidney disease often associated with congenital hepatic fibrosis and cholestasis [194, 195]. Many other inherited renal cystic disorders are associated with hepatic fibrosis, which suggests analogous pathways for the development of tubular structures in the liver and kidneys [196, 197]. Primary cilia act not only to move fluid: apart from mucociliary clearance, cilia also take part in pattern formation during embryonic development, in left–right axis orientation and in retinal photoreception [198, 197]. A particularly important role in development of bile ducts and renal tubules is reserved for primary cilia; most of the known genes inactivated in hepatorenal cystic diseases are associated with primary cilia function [199]. Defects in other molecular processes, such as tight junction protein complex formation and intercellular signaling during bile‐duct morphogenesis, may also be among contributors to EHBA. Deficiency of the proteins involved in these processes causes specific forms of neonatal sclerosing cholangitis (absence of claudin‐1 in NISCH syndrome, [138]) and paucity of interlobular bile ducts (haploinsufficiency for JAG1/NOTCH2 in Alagille syndrome [113, 122]. Assessment of synthesis of dysfunctional forms of these proteins in patients with EHBA remains to be undertaken. A description of villin deficiency associated with EHBA‐like disease [200] has not been followed by a report of a genetic correlate. The proportion of patients with a final diagnosis of NH is getting smaller, as a result of the identification of the diseases described above. NH is evidently a syndrome, or clinicopathologic diagnosis of last resort. NH is diagnosed if cholestasis cannot be assigned to infection, genetic/metabolic defect, or ­ mechanical impediment to bile flow and if microscopy of liver finds widespread giant‐cell transformation of hepatocytes. Giant‐ cell change of hepatocytes is common, however, to many cholestatic disorders in neonates, including EHBA. Most instances of NH resolve without sequelae; they may represent constitutive dysfunction of bile‐handling mechanisms unmasked by perinatal stress, a hypothesis that awaits experimental assessment. Previously, 20% of cases of idiopathic NH that are progressive, appeared familial, and had a poor prognosis [184]. This group in particular has been eroded by the discoveries described above, combined with the widespread use of genetic testing [201–203].

REFERENCES 1. Heubi, J.E., Setchell, K.D., and Bove, K.E. Inborn errors of bile acid metabolism. Semin Liver Dis, 2007;27(3):282–94. 2. Rossi, M., Vajro, P., Iorio, R. et al. Characterization of liver involvement in defects of cholesterol biosynthesis: long‐term follow‐up and review. Am J Med Genet A, 2005;132A(2):144–51. 3. Steinberg, S.J., Dodt, G., Raymond, G.V., Braverman, N.E., Moser, A.B., and Moser H.W. Peroxisome biogenesis disorders. Biochim Biophys Acta, 2006;1763(12):1733–48. 4. Russell, D.W. The enzymes, regulation, and genetics of bile acid synthesis. Annu Rev Biochem, 2003;72:137–74. 5. Hofmann, A.F. and Hagey, L.R. Bile acids: chemistry, pathochemistry, biology, pathobiology, and therapeutics. Cell Mol Life Sci, 2008;65(16):2461–83.

359

  6. Sedel, F., Tourbah, A., Fontaine, B. et al. Leukoencephalopathies associated with inborn errors of metabolism in adults. J Inherit Metab Dis, 2008;31(3): 295–307.   7. Clayton, P.T., Verrips, A., Sistermans, E., Mann, A., Mieli‐Vergani, G., and Wevers, R. Mutations in the sterol 27‐hydroxylase gene (CYP27A) cause hepatitis of infancy as well as cerebrotendinous xanthomatosis. J Inherit Metab Dis, 2002;25(6):501–13.   8. Clayton, P.T., Leonard, J.V., Lawson, A.M. et al. Familial giant cell hepatitis associated with synthesis of 3 beta, 7 alpha‐dihydroxy‐ and 3 beta,7 alpha, 12 alpha‐trihydroxy‐5‐cholenoic acids. J Clin Invest, 1987;79(4):1031–8.  9. Ichimiya, H., Egestad, B., Nazer, H., Baginski, E.S., Clayton, P.T., and Sjövall, J. Bile acids and bile alcohols in a child with hepatic 3 beta‐hydroxy‐ delta 5‐C27‐steroid dehydrogenase deficiency: effects of chenodeoxycholic acid treatment. J Lipid Res, 1991;32(5):829–41. 10. Buchmann, M.S., Kvittingen, E.A., Nazer, H. et al. Lack of 3 beta‐hydroxy‐ delta 5‐C27‐steroid dehydrogenase/isomerase in fibroblasts from a child with urinary excretion of 3 beta‐hydroxy‐delta 5‐bile acids. A new inborn error of metabolism. J Clin Invest, 1990;86(6):2034–7. 11. Jacquemin, E., Setchell, K.D., O’Connell N.C. et al. A new cause of progressive intrahepatic cholestasis: 3 beta‐hydroxy‐C27‐steroid dehydrogenase/ isomerase deficiency. J Pediatr, 1994;125(3):379–84. 12. Cheng, J.B., Jacquemin, E., Gerhardt, M. et al. Molecular genetics of 3beta‐ hydroxy‐delta5‐C27‐steroid oxidoreductase deficiency in 16 patients with loss of bile acid synthesis and liver disease. J Clin Endocrinol Metab, 2003;88(4):1833–41. 13. Setchell, K.D., Suchy, F.J., Welsh, M.B., Zimmer‐Nechemias, L., Heubi, J., and Balistreri W.F. Delta 4–3‐oxosteroid 5 beta‐reductase deficiency described in identical twins with neonatal hepatitis. A new inborn error in bile acid synthesis. J Clin Invest, 1988;82(6):2148–57. 14. Setchell, K.D., Schwarz, M., O’Connell N.C. et al. Identification of a new inborn error in bile acid synthesis: mutation of the oxysterol 7alpha‐hydroxylase gene causes severe neonatal liver disease. J Clin Invest, 1998;102(9): 1690–703. 15. Tsaousidou, M.K., Ouahchi, K., Warner, T.T. et  al. Sequence alterations within CYP7B1 implicate defective cholesterol homeostasis in motor‐neuron degeneration. Am J Hum Genet, 2008;82(2):510–5. 17. Ferdinandusse, S., Rusch, H., van Lint, A.E., Dacremont, G., Wanders, R.J., and Vreken, P. Stereochemistry of the peroxisomal branched‐chain fatty acid alpha‐ and beta‐oxidation systems in patients suffering from different peroxisomal disorders. J Lipid Res, 2002;43(3):438–44. 18. Van Veldhoven, P.P., Meyhi, E., Squires, R.H. et al. Fibroblast studies documenting a case of peroxisomal 2‐methylacyl‐CoA racemase deficiency: possible link between racemase deficiency and malabsorption and vitamin K deficiency. J Clin Invest Eur J Clin Invest, 2001;31(8):714–22. 19. Setchell, K.D., Heubi, J.E., O’Connell N.C., Hofmann, A.F., and Levine J.A. Absence of bile acid conjugation: a new inborn error of bile acid metabolism. Hepatol, 1996;24(4):371A. 20. Carlton, V.E., Harris, B.Z., Puffenberger, E.G., Batta, A.K., Knisely, A.S., Robinson, D.L. et al. Complex inheritance of familial hypercholanemia with associated mutations in TJP2 and BAAT. Nat Genet, 2003;34(1):91–6. 21. Tazawa, Y., Yamada, M., Nakagawa, M., Konno, T., and Tada, K. Bile acid profiles in siblings with progressive intrahepatic cholestasis: absence of biliary chenodeoxycholate. J Pediatr Gastroenterol Nutr, 1985;4(1):32–7. 22. Bull, L.N., Carlton, V.E., Stricker, N.L. et  al. Genetic and morphological findings in progressive familial intrahepatic cholestasis (Byler disease [PFIC‐1] and Byler syndrome): evidence for heterogeneity. Hepatol, 1997; 26(1):155–64. 23. Houwen, R.H., Baharloo, S., Blankenship, K. et al. Genome screening by searching for shared segments: mapping a gene for benign recurrent intrahepatic cholestasis. Nat Genet, 1994;8(4):380–6. 24. van Mil, S.W., van der Woerd, W.L., van der Brugge, G. et al. Benign recurrent intrahepatic cholestasis type 2 is caused by mutations in ABCB11. Gastroenterol, 2004;127(2):379–84. 25. van Ooteghem, N.A., Klomp, L.W., van Berge‐Henegouwen, G.P., and Houwen R.H. Benign recurrent intrahepatic cholestasis progressing to progressive familial intrahepatic cholestasis: low GGT cholestasis is a clinical continuum. J Hepatol, 2002;36(3):439–43. 26. Takahashi, A., Hasegawa, M., Sumazaki, R. et al. Gradual improvement of liver function after administration of ursodeoxycholic acid in an infant with a novel ABCB11 gene mutation with phenotypic continuum between BRIC2 and PFIC2. Eur J Gastroenterol Hepatol, 2007;19(11):942–6.

360

THE LIVER:  REFERENCES

27. Meier, Y., Zodan, T., Lang, C. et al. Increased susceptibility for intrahepatic cholestasis of pregnancy and contraceptive‐induced cholestasis in carriers of the 1331T>C polymorphism in the bile salt export pump. World J Gastroenterol, 2008;14(1):38–45. 28. Müllenbach, R., Bennett, A., Tetlow, N. et al. ATP8B1 mutations in British cases with intrahepatic cholestasis of pregnancy. 1. Gut, 2005;54(6):829–34. 29. Lammert, F., Marschall, H.U., Glantz, A., and Matern, S. Intrahepatic cholestasis of pregnancy: molecular pathogenesis, diagnosis and management. J Hepatol, 2000;33(6):1012–21. 30. Milkiewicz, P., Gallagher, R., Chambers, J., Eggington, E., Weaver, J., and Elias, E. Obstetric cholestasis with elevated gamma glutamyl transpeptidase: incidence, presentation and treatment. J Gastroenterol Hepatol, 2003;18(11):1283–6. 31. Clayton, R.J., Iber, F.L., Ruebner, B.H., and McKusick V.A. Byler disease. Fatal familial intrahepatic cholestasis in an Amish kindred. Am J Dis Child, 1969;117(1):112–24. 32. Bull, L.N., van Eijk, M.J., Pawlikowska, L. et al. A gene encoding a P‐type ATPase mutated in two forms of hereditary cholestasis. Nat Genet, 1998;18(3):219–24. 33. Eppens, E.F., van Mil, S.W., de Vree, J.M. et al. FIC1, the protein affected in two forms of hereditary cholestasis, is localized in the cholangiocyte and the canalicular membrane of the hepatocyte. J Hepatol, 2001;35(4):436–43. 34. Ujhazy, P., Ortiz, D., Misra, S. et al. Familial intrahepatic cholestasis 1: studies of localization and function. Hepatol, 2001;34(4 Pt 1):768–75. 35. van Mil, S.W., Houwen, R.H., and Klomp L.W. Genetics of familial intrahepatic cholestasis syndromes. J Med Genet, 2005;42(6):449–63. 36. Paulusma, C.C., Folmer, D.E., Ho‐Mok, K.S. et  al. ATP8B1 requires an accessory protein for endoplasmic reticulum exit and plasma membrane lipid flippase activity. Hepatol, 2008;47(1):268–78. 37. Frankenberg, T., Miloh, T., Chen, F.Y. et al. The membrane protein ATPase class I type 8B member 1 signals through protein kinase C zeta to activate the farnesoid X receptor. Hepatol, 2008;44(1):195–204. 39. Demeilliers, C., Jacquemin, E., Barbu, V. et al. Altered hepatobiliary gene expressions in PFIC1: ATP8B1 gene defect is associated with CFTR downregulation. Hepatol, 2006;43(5):1125–34. 40. Knisely, A.S., Agostini, R.M., Zitelli, B.J., Kocoshis, S.A., and Boyle J.T. Byler’s syndrome. Arch Dis Child, 1997;77(3):276–7. 41. Oshima, T., Ikeda, K., and Takasaka, T. Sensorineural hearing loss associated with Byler disease. Tohoku J Exp Med, 1999;187(1):83–8. 42. Winklhofer‐Roob, B.M., Shmerling, D.H., Solèr, R., and Briner, J. Progressive idiopathic cholestasis presenting with profuse watery diarrhoea and recurrent infections (Byler’s disease). Acta Paediatr, 1992;81(8):637–40. 43. Knisely A.S. Progressive familial intrahepatic cholestasis: a personal perspective. Pediatr Dev Pathol, 2000;3(2),113–25. 44. Walkowiak, J., Jankowska, I., Pawlowska, J., Bull, L., Herzig, K.H., and Socha, J. Normal pancreatic secretion in children with progressive familial intrahepatic cholestasis type 1. Scand J Gastroenterol, 2006;41(12):1480–3. 45. Klomp, L.W., Vargas, J.C., van Mil, S.W. et al. Characterization of mutations in ATP8B1 associated with hereditary cholestasis. Hepatol, 2004;40(1):27–38. 46. Strautnieks, S.S., Bull, L.N., Knisely et al. A gene encoding a liver‐specific ABC transporter is mutated in progressive familial intrahepatic cholestasis. Nat Genet, 1998;20(3):233–8. 47. Stieger, B., Meier, Y., and Meier, P.J. The bile salt export pump. Pflugers Arch, 2007;453(5):611–20. 48. Noe, J., Kullak‐Ublick, G.A., Jochum, W. et  al. Impaired expression and function of the bile salt export pump due to three novel ABCB11 mutations in intrahepatic cholestasis. J Hepatol, 2005;43(3):536–43. 49. Lam, C.W., Cheung, K.M., Tsui, M.S., Yan, M.S., Lee, C.Y., and Tong, S.F. A patient with novel ABCB11 gene mutations with phenotypic transition between BRIC2 and PFIC2. J Hepatol, 2006;44(1):240–2. 50. Strautnieks, S.S., Byrne, J.A., Pawlikowska, L. et  al. Severe bile salt export pump deficiency: 82 different ABCB11 mutations in 109 families. Gastroenterol, 2008;134(4):1203–14. 51. Chen, H.L., Chang, P.S., Hsu, H.C. et  al. FIC1 and BSEP defects in Taiwanese patients with chronic intrahepatic cholestasis with low gamma‐ glutamyltranspeptidase levels. J Pediatr, 2002;140(1):119–24. 52. Kagawa, T., Watanabe, N., Mochizuki, K. et  al. Phenotypic differences in PFIC2 and BRIC2 correlate with protein stability of mutant Bsep and impaired taurocholate secretion in MDCK II cells. Am J Physiol Gastrointest Liver Physiol, 2008;294(1):G58–67. 53. Lam, P., Pearson, C.L., Soroka, C.J., Xu, S., Mennone, A., and Boyer, J.L. Levels of plasma membrane expression in progressive and benign mutations

of the bile salt export pump (Bsep/Abcb11) correlate with severity of cholestatic diseases. Am J Physiol Cell Physiol, 2007;293(5):C1709–16. 54. Dixon, P.H., Wadsworth, C.A., Chambers, J. et al. A comprehensive analysis of common genetic variation around six candidate loci for intrahepatic cholestasis of pregnancy. Am J Gastroenterol, 2014;109(1):76–84. 55. Dixon, P.H., Sambrotta, M., Chambers, J. et al. An expanded role for heterozygous mutations of ABCB4, ABCB11, ATP8B1, ABCC2 and TJP2 in intrahepatic cholestasis of pregnancy. Sci Rep, 2017;7(1):11823. 56. Vallejo, M., Briz, O., Serrano, M.A., Monte, M.J., and Marin J.J. Potential role of trans‐inhibition of the bile salt export pump by progesterone metabolites in the etiopathogenesis of intrahepatic cholestasis of pregnancy. J Hepatol, 2006;44(6):1150–7. 57. Iwanaga, T., Nakakariya, M., Yabuuchi, H., Maeda, T., and Tamai, I. Involvement of bile salt export pump in flutamide‐induced cholestatic hepatitis. Biol Pharm Bull, 2007;30(4):739–44. 58. Schafmayer, C., Tepel, J., Franke, A. et al. Investigation of the Lith1 candidate genes ABCB11 and LXRA in human gallstone disease. Hepatol, 2006;44(3):650–7. 59. Wang, R., Lam, P., Liu, L. et al. Severe cholestasis induced by cholic acid feeding in knockout mice of sister of P‐glycoprotein. Hepatol, 2003; 38(6):1489–99. 60. Lam, P., Wang, R., and Ling, V. Bile acid transport in sister of P‐glycoprotein (ABCB11) knockout mice. Biochemistry, 2005;44(37):12598–605. 61. Paulusma, C.C., Bosma, P.J., Zaman, G.J. et al. Congenital jaundice in rats with a mutation in a multidrug resistance‐associated protein gene. Science, 1996;271(5252):1126–8. 62. Ito, K., Suzuki, H., Hirohashi, T., Kume, K., Shimizu, T., and Sugiyama, Y. Molecular cloning of canalicular multispecific organic anion transporter defective in EHBR. Am J Physiol, 1997;272(1 Pt 1):G16–22. 63. Paulusma, C.C., Kool, M., Bosma, P.J. et al. A mutation in the human canalicular multispecific organic anion transporter gene causes the Dubin– Johnson syndrome. Hepatol, 1997;25(6):1539–42. 64. Ito, K., Suzuki, H., Hirohashi, T., Kume, K., Shimizu, T., and Sugiyama, Y. Functional analysis of a canalicular multispecific organic anion transporter cloned from rat liver. J Biol Chem, 1998;273(3):1684–8. 65. Nies, A.T. and Keppler, D. The apical conjugate efflux pump ABCC2 (MRP2). Pflugers Arch, 2007;453(5):643–59 66. Dubin, I.N. and Johnson F.B. Chronic idiopathic jaundice with unidentified pigment in liver cells; a new clinicopathologic entity with a report of 12 cases. Medicine (Baltimore), 1954;33(3):155–97. 67. Lee, J.H., Chen, H.L., Chen, H.L., Ni, Y.H., Hsu, H.Y., and Chang M.H. Neonatal Dubin–Johnson syndrome: long‐term follow‐up and MRP2 mutations study. Pediatr Res, 2006;59(4 Pt 1):584–9. 68. Donner, M.G. and Keppler, D. Up‐regulation of basolateral multidrug resistance protein 3 (Mrp3) in cholestatic rat liver. Hepatol, 2001;34(2):351–9. 69. Chu, X.Y., Strauss, J.R., Mariano, M.A. et al. Characterization of mice lacking the multidrug resistance protein MRP2 (ABCC2). J Pharmakos Exp Ther, 2006;317(2):579–89. 70. Kikuchi, S., Hata, M., Fukumoto, K. et al. Radixin deficiency causes conjugated hyperbilirubinemia with loss of Mrp2 from bile canalicular membranes. Nat Genet, 2002;31(3):320–5. 71. Wolpert, E., Pascasio, F.M., Wolkoff, A.W., and Arias I.M. Abnormal sulfobromophthalein metabolism in Rotor’s syndrome and obligate heterozygotes. N Engl J Med, 1977;296(19):1099–101. 72. Hrebícek, M., Jirásek, T., Hartmannová, H. et al. Rotor‐type hyperbilirubinaemia has no defect in the canalicular bilirubin export pump. Liver Int, 2007;27(4):485–91. 73. van de Steeg, E., Stránecký, V., Hartmannová, H. et al. Complete OATP1B1 and OATP1B3 deficiency causes human Rotor syndrome by interrupting conjugated bilirubin reuptake into the liver. J Clin Invest, 2012;122(2):519–28. 74. Deleuze, J.F., Jacquemin, E., Dubuisson, C. et al. Defect of multidrug‐resistance 3 gene expression in a subtype of progressive familial intrahepatic cholestasis. Hepatol, 1996;23(4):904–8. 75. Oude Elferink, R.P. and Paulusma, C.C. Function and pathophysiological importance of ABCB4 (MDR3 P‐glycoprotein). Pflugers Arch, 2007;453(5):601–10. 76. Lamireau, T., Bouchard, G., Yousef, I.M. et  al. Dietary lecithin protects against cholestatic liver disease in cholic acid‐fed Abcb4‐deficient mice. Pediatr Res, 2007;61(2):185–90. 77. Baghdasaryan, A., Fickert, P., Fuchsbichler, A. et al. Role of hepatic phospholipids in development of liver injury in Mdr2 (Abcb4) knockout mice. Liver Int, 2008;28(7):948–58.



29:  Molecular Cholestasis

  78. Smit, J.J., Schinkel, A.H., Oude Elferink, R.P. et al. Homozygous disruption of the murine mdr2 P‐glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell, 1993;75(3):451–62.   79. Ziol, M., Barbu, V., Rosmorduc, O. et al. ABCB4 heterozygous gene mutations associated with fibrosing cholestatic liver disease in adults. Gastroenterol, 2008;135(1):131–41.   80. Gotthardt, D., Runz, H., Keitel, V. et al. A mutation in the canalicular phospholipid transporter gene, ABCB4, is associated with cholestasis, ductopenia, and cirrhosis in adults. Hepatol, 2008;48(4):1157–66.   81. Jacquemin, E. Role of multidrug resistance 3 deficiency in pediatric and adult liver disease: one gene for three diseases. Semin Liver Dis, 2001;21(4):551–62   82. de Vree, J.M., Jacquemin, E., Sturm, E. et al. Mutations in the MDR3 gene cause progressive familial intrahepatic cholestasis. Proc Natl Acad Sci USA, 1998;95(1):282–7.   83. Dixon, P.H., Weerasekera, N., Linton, K.J. et al. Heterozygous MDR3 missense mutation associated with intrahepatic cholestasis of pregnancy: evidence for a defect in protein trafficking. Hum Mol Genet, 2000;9(8):1209–17.   84. Pauli‐Magnus, C., Lang, T., Meier, Y. et al. Sequence analysis of bile salt export pump (ABCB11) and multidrug resistance p‐glycoprotein 3 (ABCB4, MDR3) in patients with intrahepatic cholestasis of pregnancy. Pharmacogenet, 2004;14(2):91–102.   85. Wasmuth, H.E., Glantz, A., Keppeler, H. et al. Intrahepatic cholestasis of pregnancy: the severe form is associated with common variants of the hepatobiliary phospholipid transporter ABCB4 gene. Gut, 2007;56(2):265–70.   87. Rosmorduc, O. and Poupon, R. Low phospholipid associated cholelithiasis: association with mutation in the MDR3/ABCB4 gene. Orphanet J Rare Dis, 2007;2:29.  88. Rosmorduc, O., Hermelin, B., Boelle, P.Y., Parc, R., Taboury, J., and Poupon, R. ABCB4 gene mutation‐associated cholelithiasis in adults. Gastroenterol, 2003;125(2):452–9.   89. Kano, M., Shoda, J., Sumazaki, R., Oda, K., Nimura, Y., and Tanaka, N. Mutations identified in the human multidrug resistance P‐glycoprotein 3 (ABCB4) gene in patients with primary hepatolithiasis. Hepatol Res, 2004;29(3):160–6.   90. Shoda, J., Oda, K., Suzuki, H. et  al. Etiologic significance of defects in cholesterol, phospholipid, and bile acid metabolism in the liver of patients with intrahepatic calculi. Hepatol, 2001;33(5):1194–205.   91. Lang, C., Meier, Y., Stieger, B. et al. Mutations and polymorphisms in the bile salt export pump and the multidrug resistance protein 3 associated with drug‐induced liver injury. Pharmacogenet Genomics, 2007;17(1):47–60.   92. Bramow, S., Ott, P., Thomsen Nielsen, F., Bangert, K., Tygstrup, N., and Dalhoff, K. Cholestasis and regulation of genes related to drug metabolism and biliary transport in rat liver following treatment with cyclosporine A and sirolimus (rapamycin). Pharmakon Toxicol, 2001;89(3):133–9.   93. Morton, D.H., Salen, G., Batta, A.K. et al. Abnormal hepatic sinusoidal bile acid transport in an Amish kindred is not linked to FIC1 and is improved by ursodiol. Gastroenterol, 2000;119(1):188–95.   94. Sambrotta, M., Strautnieks, S., Papouli, E. et al. Mutations in TJP2 cause progressive cholestatic liver disease. Nat Genet, 2014;46(4):326–8   95. Zhou, S., Hertel, P.M., Finegold, M.J. et al. Hepatocellular carcinoma associated with tight‐junction protein 2 deficiency. Hepatol, 2015;62(6):1914–6.   96. Knowles, B.C., Roland, J.T., Krishnan, M. et  al. Myosin Vb uncoupling from RAB8A and RAB11A elicits microvillus inclusion disease. J Clin Invest, 2014;124(7):2947–62.   97. Lapierre, L.A., Kumar, R., Hales, C.M. et al. Myosin vb is associated with plasma membrane recycling systems. Mol Biol Cell, 2001;12(6):1843–57.   98. Wakabayashi, Y., Dutt, P., Lippincott‐Schwartz, J. et al. Rab11a and myosin Vb are required for bile canalicular formation in WIF‐B9 cells. Proc Natl Acad Sci USA, 2005;102(42):15087–92.   99. Wakabayashi, Y., Lippincott‐Schwartz, J., and Arias I.M. Intracellular trafficking of bile salt export pump (ABCB11) in polarized hepatic cells: constitutive cycling between the canalicular membrane and rab11‐positive endosomes. Mol Biol Cell, 2004;15(7):3485–96. 100. Muller, T., Hess, M.W., Schiefermeier, N. et al. MYO5B mutations cause microvillus inclusion disease and disrupt epithelial cell polarity. Nat Genet, 2008;40(10):1163–5. 101. Ruemmele, F.M., Muller, T., Schiefermeier, N. et al. Loss‐of‐function of MYO5B is the main cause of microvillus inclusion disease: 15 novel mutations and a CaCo‐2 RNAi cell model. Hum Mutat, 2010;31(5):544–51.

361

102. Thoeni, C.E., Vogel, G.F., Tancevski, I. et al. Microvillus inclusion disease: loss of Myosin vb disrupts intracellular traffic and cell polarity. Traffic, 2014;15(1):22–42. 103. Girard, M., Lacaille, F., Verkarre, V. et  al. MYO5B and bile salt export pump contribute to cholestatic liver disorder in microvillous inclusion disease. Hepatol, 2014;60(1):301–10. 104. Grammatikopoulos, T., Sambrotta, M., Strautnieks, S. et al. Mutations in DCDC2 (doublecortin domain containing protein 2) in neonatal sclerosing cholangitis. J Hepatol, 2016;65(6):1179–1187. 105. Girard, M., Bizet, A.A., Lachaux, A. et al. DCDC2 mutations cause neonatal sclerosing cholangitis. Hum Mutat, 2016;37(10):1025–9. 106. Carrion‐Castillo, A., Franke, B., and Fisher S.E. Molecular genetics of dyslexia: an overview. Dyslexia, 2013;19(4):214–40. 107. Schueler, M., Braun, D.A., Chandrasekar, G. et  al. DCDC2 mutations cause a renal‐hepatic ciliopathy by disrupting Wnt signaling. Am J Hum Genet, 2015;96(1):81–92. 107. Shueler, M., Braun, D.A., Chandrasekar, G. et al. DCDC2 mutations cause a renal‐hepatic ciliopathy by disrupting Wnt signaling. Am J Hum Genet, 2015;96(1):82–92. 108. Alagille, D., Odievre, M., Gautier, M., and Demerges J.P. Hepatic ductular hypoplasia associated with characteristic facies, vertebral malformations, retarded physical, mental and sexual development, and cardiac murmur. J Pediatr 1975;86:63–71. 109. Emerick, K.M., Rand, E.B., Goldmuntz, E., Krantz, I.D., Spinner, N.B., and Piccoli D.A. Features of Alagille syndrome in 92 patients: frequency and relation to prognosis. Hepatol, 1999;29(3):822–9. 110. Chiaretti, A., Zampino, G., Botto, L., and Polidori, G. Alagille syndrome and hepatocarcinoma: a case report. Acta Paediatr, 1992;81(11):937. 111. Békássy, A.N., Garwicz, S., Wiebe, T., Homestand, I., and Jensen O.A. Hepatocellular carcinoma associated with arteriohepatic dysplasia in a 4‐ year‐old girl. Med Pediatr Oncol, 1992;20(1):78–83. 112. Kamath, B.M., Bason, L., Piccoli, D.A., Krantz, I.D., and Spinner N.B. Consequences of JAG1 mutations. J Med Genet, 2003;40(12):891–5. 113. Oda, T., Elkahloun, A.G., Pike, B.L. et al. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat Genet, 1997;16(3):235–42. 114. Warthen, D.M., Moore, E.C., Kamath, B.M. et al. Jagged1 (JAG1) mutations in Alagille syndrome: increasing the mutation detection rate. Hum Mutat, 2006;27(5):436–43. 115. Lanford, P.J., Lan, Y., Jiang, R. et al. Notch signalling pathway mediates hair cell development in mammalian cochlea. Nat Genet, 1999;21(3):289–92. 116. Davies, J.A. and Fisher C.E. Genes and proteins in renal development. Exp Nephrol, 2002;10(2):102–13. 117. Hu, Q.D., Cui, X.Y., Ng, Y.K., and Xiao, Z.C. Axoglial interaction via the notch receptor in oligodendrocyte differentiation. Ann Acad Med Singapore, 2004;33(5):581–8. 118. Bray, S.J. Notch signalling: a simple pathway becomes complex. Nat Rev Mol Cell Biol, 2006;7(9):678–89. 119. Ehebauer, M., Hayward, P., and Arias A.M. Notch, a universal arbiter of cell fate decisions. Science, 2006;314(5804):1414–5. 120. McCright B., Gao, X., Shen, L. et al. Defects in development of the kidney, heart and eye vasculature in mice homozygous for a hypermorphic Notch2 mutation. Development, 2001;128(4):491–502. 121. McCright B., Lozier, J., and Gridley, T. A mouse model of Alagille syndrome: Notch2 as a genetic modifier of Jag1 haploinsufficiency. Development, 2002;129(4):1075–82. 122. McDaniell R., Warthen, D.M., Sanchez‐Lara, P.A. et al. NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. Am J Hum Genet, 2006;79(1):169–73. 123. Lozier, J., McCright B., and Gridley, T. Notch signaling regulates bile duct morphogenesis in mice. PLoS ONE, 2008;3(3):e1851. 124. Gissen, P., Johnson, C.A., Morgan, N.V. et al. Mutations in VPS33B., encoding a regulator of SNARE‐dependent membrane fusion, cause arthrogryposis‐renal dysfunction‐cholestasis (ARC) syndrome. Nat Genet, 2004;36(4):400–4. 125. Cullinane, A.R., Straatman‐Iwanowska, A., Zaucker, A. et al. Mutations in VIPAR cause an arthrogryposis, renal dysfunction and cholestasis syndrome phenotype with defects in epithelial polarization. Nat Genet, 2010;42(4):303–12. 126. Banushi, B., Forneris, F., Straatman‐Iwanowska, A. et  al. Regulation of post‐Golgi LH3 trafficking is essential for collagen homeostasis. Nat Commun, 2016;7:12111.

362

THE LIVER:  REFERENCES

127. Lo, B., Li, L., Gissen, P. et al. Requirement of VPS33B., a member of the Sec1/Munc18 protein family, in megakaryocyte and platelet alpha‐granule biogenesis. Blood, 2005;106(13):4159–66. 128. Bem, D., Smith, H., Banushi, B. et al. VPS33B regulates protein sorting into and maturation of α‐granule progenitor organelles in mouse megakaryocytes. Blood, 2015;126(2):133–43. 129. Hershkovitz D., Mandel, H., Ishida‐Yamamoto, A. et al. Defective lamellar granule secretion in arthrogryposis, renal dysfunction, and cholestasis syndrome caused by a mutation in VPS33B. Arch Dermatol, 2008;144(3): 334–40. 130. Rogerson, C. and Gissen, P. VPS33B and VIPAR are essential for epidermal lamellar body biogenesis and function. Biochim Biophys Acta Mol Basis Dis, 2018;1864(5):1609–21. 131. Horslen, S.P., Quarrell, O.W., and Tanner M.S. Liver histology in the arthrogryposis multiplex congenita, renal dysfunction, and cholestasis (ARC) syndrome: report of three new cases and review. J Med Genet, 1994;31(1):62–4. 132. Bull, L.N., Mahmoodi, V., Baker, A.J. et  al. VPS33B mutation with ichthyosis, cholestasis, and renal dysfunction but without arthrogryposis: incomplete ARC syndrome phenotype. J Pediatr, 2006;148(2):269–71. 133. Hanley, J., Dhar, D.K., Mazzacuva, F. et al. Vps33b is crucial for structural and functional hepatocyte polarity. J Hepatol, 2017;66(5):1001–11. 134. Wang C., Cheng, Y., Zhang, X. et  al. Vacuolar protein sorting 33B is a tumor suppressor in hepatocarcinogenesis. Hepatol, 2018;68(6):2239–53. 135. Gissen, P., Tee, L., Johnson, C.A. et al. Clinical and molecular genetic features of ARC syndrome. Hum Genet, 2006;120(3):396–409. 136. Akbar, M.A., Mandraju, R., Tracy, C., Hu, W., Pasare, C., and Krämer, H. ARC syndrome‐linked Vps33B protein is required for inflammatory endosomal maturation and signal termination. Immunity, 2016;45(2):267–79. 137. Deal, J.E., Barratt, T.M., and Dillon, M.J. Fanconi syndrome, ichthyosis, dysmorphism, jaundice and diarrhoea‐‐a new syndrome. Pediatr Nephrol, 1990;4(4):308–13. 138. Hadj‐Rabia, S., Baala, L., Vabres, P. et  al. Claudin‐1 gene mutations in neonatal sclerosing cholangitis associated with ichthyosis: a tight junction disease. Gastroenterol, 2004;127(5):1386–90. 139. Feldmeyer, L., Huber, M., Fellmann, F., Beckmann, J.S., Frenk, E., and Hohl, D. Confirmation of the origin of NISCH syndrome. Hum Mutat, 2006;27(5):408–10. 140. Szepetowski, S., Lacoste, C., Mallet, S., Roquelaure, B., Badens, C., and Fabre, A. NISCH syndrome, a rare cause of neonatal cholestasis: a case report. Arch Pediatr, 2017;24(12):1228–34. 141. Baala, L., Hadj‐Rabia, S., Hamel‐Teillac, D. et al. Homozygosity mapping of a locus for a novel syndromic ichthyosis to chromosome 3q27‐q28. J Invest Dermatol, 2002;119(1):70–6. 142. Mazzon, E., Puzzolo, D., Caputi, A.P., and Cuzzocrea, S. Role of IL‐10 in hepatocyte tight junction alteration in mouse model of experimental colitis. Mol Med, 2002;8(7):353–66. 143. Maly, I.P. and Landmann, L. Bile duct ligation in the rat causes upregulation of ZO‐2 and decreased colocalization of claudins with ZO‐1 and occludin. Histochem Cell Biol, 2008;129(3):289–99. 144. Fickert, P., Fuchsbichler, A., Marschall, H.U. et al. Lithocholic acid feeding induces segmental bile duct obstruction and destructive cholangitis in mice. Am J Pathol, 2006;168(2):410–22. 145. Anderson, J.M. Leaky junctions and cholestasis: a tight correlation. Gastroenterol, 1996;110(5):1662–5. 146. Mottino, A.D., Hoffman, T., Crocenzi, F.A., Sánchez Pozzi, E.J., Roma, M.G., and Vore, M. Disruption of function and localization of tight junctional structures and Mrp2 in sustained estradiol‐17beta‐D‐glucuronide‐ induced cholestasis. Am J Physiol Gastrointest Liver Physiol, 2007;293(1): G391–402. 147. Kojima, T., Yamamoto, T., Murata, M., Chiba, H., Kokai, Y., and Sawada, N. Regulation of the blood‐biliary barrier: interaction between gap and tight junctions in hepatocytes. Med Electron Microsc, 2003;36(3):157–64. 148. Furuse, M., Hata, M., Furuse, K. et  al. Claudin‐based tight junctions are crucial for the mammalian epidermal barrier: a lesson from claudin‐1‐ deficient mice. J Cell Biol, 2002;156(6):1099–111. 149. Smeitink, J., van den Heuvel, L., and DiMauro S. The genetics and pathology of oxidative phosphorylation. Nat Rev Genet, 2001;2(5):342–52. 150. Visapää, I., Fellman, V., Vesa, J. et al. GRACILE syndrome, a lethal metabolic disorder with iron overload, is caused by a point mutation in BCS1L. Am J Hum Genet, 2002;71(4):863–76.

151. Fellman, V., Lemmelä, S., Sajantila, A., Pihko, H., and Järvelä, I. Screening of BCS1L mutations in severe neonatal disorders suspicious for mitochondrial cause. J Hum Genet, 2008;53(6):554–8. 152. Baker, R.A., Priestley, J.R.C., Wilstermann, A.M., Reese, K.J., and Mark P.R. Clinical spectrum of BCS1L Monopathies and their underlying structural relationships. Am J Med Genet A, 2018. 153. Fernandez‐Vizarra, E., Bugiani, M., Goffrini, P. et al. Impaired complex III assembly associated with BCS1L gene mutations in isolated mitochondrial encephalopathy. Hum Mol Genet, 2007;16(10):1241–52. 154. Hinson, J.T., Fantin, V.R., Schönberger, J. et al. Missense mutations in the BCS1L gene as a cause of the Björnstad syndrome. N Engl J Med, 2007; 356(8):809–19. 155. De Meirleir, L., Seneca, S., Damis, E. et al. Clinical and diagnostic characteristics of complex III deficiency due to mutations in the BCS1L gene. Am J Med Genet A, 2003;121A(2):126–31. 156. Selvaag, E. Pili torti and sensorineural hearing loss. A follow‐up of Bjørnstad’s original patients and a review of the literature. Eur J Dermatol, 2000;10(2):91–7. 157. Purhonen J,, Rajendran, J., Mörgelin, M. et al. Ketogenic diet attenuates hepatopathy in mouse model of respiratory chain complex III deficiency caused by a Bcs1l mutation. Sci Rep,. 2017;7(1):957. 158. Sase, M., Kobayashi, K., Imamura, Y. et al. Level of translatable messenger RNA coding for argininosuccinate synthetase in the liver of the patients with quantitative‐type citrullinemia. Hum Genet, 1985;69(2):130–4. 159. Saheki, T., Nakano, K., Kobayashi, K. et al. Analysis of the enzyme abnormality in eight cases of neonatal and infantile citrullinaemia in Japan. J Inherit Metab Dis, 1985;8(3):155–6. 160. Kobayashi, K., Sinasac, D.S., Iijima, M. et al. The gene mutated in adult‐ onset type II citrullinaemia encodes a putative mitochondrial carrier protein. Nat Genet, 1999;22(2):159–63. 161. Ohura, T., Kobayashi, K., Tazawa, Y. et al. Clinical pictures of 75 patients with neonatal intrahepatic cholestasis caused by citrin deficiency (NICCD). J Inherit Metab Dis, 2007;30(2):139–44. 162. Tabata, A., Sheng, J.S., Ushikai, M. et al. Identification of 13 novel mutations including a retrotransposal insertion in SLC25A13 gene and frequency of 30 mutations found in patients with citrin deficiency. J Hum Genet, 2008;53(6):534–45. 163. Kobayashi, K., Iijima, M., Ushikai, M., Ikeda, S., and Saheki, T. Citrin deficiency. J Jpn Pediatr Soc, 2006;110:1047–59. 164. Soeda, J., Yazaki, M., Nakata, T. et al. Primary liver carcinoma exhibiting dual hepatocellular‐biliary epithelial differentiations associated with citrin deficiency: a case report. J Clin Gastroenterol, 2008;42(7):855–60. 165. Sinasac, D.S., Moriyama, M., Jalil, M.A. et  al. Slc25a13‐knockout mice harbor metabolic deficits but fail to display hallmarks of adult‐onset type II citrullinemia. Mol Cell Biol, 2004;24(2):527–36. 166. Weber, A.M., Tuchweber, B., Yousef, I. et al. Severe familial cholestasis in North American Indian children: a clinical model of microfilament dysfunction? Gastroenterol, 1981;81(4):653–62. 167. Chagnon, P., Michaud, J., Mitchell, G. et al. A missense mutation (R565W) in cirhin (FLJ14728) in North American Indian childhood cirrhosis. Am J Hum Genet, 2002;71(6):1443–9. 168. Freed, E.F. and Baserga, S.J. The C‐terminus of Utp4, mutated in childhood cirrhosis, is essential for ribosome biogenesis. Nucleic Acids Res, 2010;38(14):4798–806. 169. Freed, E.F., Prieto, J.L., McCann, K.L., McStay, B., and Baserga, S.J. NOL11, implicated in the pathogenesis of North American Indian childhood cirrhosis, is required for pre‐rRNA transcription and processing. PLoS Genet, 2012;8(8):e1002892. 170. Aagenaes, O., Sigstad, H., and Bjorn‐Hansen, R. Lymphoedema in hereditary recurrent cholestasis from birth. Arch Dis Child,1970; ­ 45(243):690–5. 171. Drivdal, M., Trydal, T., Hagve, T.A., Bergstad, I., and Aagenaes, O. Prognosis, with evaluation of general biochemistry, of liver disease in lymphoedema cholestasis syndrome 1 (LCS1/Aagenaes syndrome). Scand J Gastroenterol, 2006;41(4):465–71. 172. Iversen, K., Drivdal, L.M., Billaud Feragen, K.J., and Geirdal, A.Ø. Quality of life in adults with lymphedema cholestasis syndrome 1. Health Qual Life Outcomes, 2018;16(1):146. 173. Bull, L.N., Roche, E., Song, E.J. et al. Mapping of the locus for cholestasis‐ lymphedema syndrome (Aagenaes syndrome) to a 6.6‐cM interval on ­chromosome 15q. Am J Hum Genet, 2000;67(4):994–9.



29:  Molecular Cholestasis

174. Viveiros, A., Reiterer, M., Schaefer, B. et al. CCBE1 mutation causing sclerosing cholangitis: Expanding the spectrum of lymphedema‐cholestasis syndrome. Hepatol, 2017;66(1):286–288. 175. Shah, S., Conlin, L.K., Gomez, L. et al. CCBE1 mutation in two siblings, one manifesting lymphedema‐cholestasis syndrome, and the other, fetal hydrops. PLoS One,. 2013;8(9):e75770. 176. Alders, M., Hogan, B.M., Gjini, E. et al. Mutations in CCBE1 cause generalized lymph vessel dysplasia in humans. Nat Genet, 2009;41(12):1272–4. 177. Hogan, B.M., Bos, F.L., Bussmann, J. et al. Ccbe1 is required for embryonic lymphangiogenesis and venous sprouting. Nat Genet, 2009;41(4):396–8. 178. Clayton, P.T. Diagnosis of inherited disorders of liver metabolism. J Inherit Metab Dis, 2003;26(2–3):135–46. 179. Balistreri, W.F. Neonatal cholestasis. J Pediatr, 1985;106(2):171–84. 180. Dick, M.C. and Mowat, A.P. Hepatitis syndrome in infancy – an epidemiological survey with 10 year follow up. Arch Dis Child,1985;60(6):512–6. 181. Ozkan, T.B., Mistik, R., Dikici, B., and Nazlioglu H.O. Antiviral therapy in neonatal cholestatic cytomegalovirus hepatitis. BMC Gastroenterol, 2007;7:9. 182. Sokol, R.J., Shepherd, R.W., Superina, R., Bezerra, J.A., Robuck, P., and Hoofnagle J.H. Screening and outcomes in biliary atresia: summary of a National Institutes of Health workshop. Hepatol, 2007;46(2):566–81. 183. Balistreri, W.F., Grand, R., Hoofnagle, J.H., Suchy, F.J., Ryckman, F.C., Perlmutter, D.H., et al. Biliary atresia: current concepts and research directions. Summary of a symposium. Hepatol, 1996;23(6):1682–92. 184. Sokol, R.J., Mack, C., Narkewicz, M.R., and Karrer F.M. Pathogenesis and outcome of biliary atresia: current concepts. J Pediatr Gastroenterol Nutr, 2003;37(1):4–21. 185. Sokol, R.J. and Mack, C. Etiopathogenesis of biliary atresia. Semin Liver Dis, 2001;21(4):517–24. 186. Ohi, R. Surgery for biliary atresia. Liver, 2001;21(3):175–82. 187. Davenport, M., Savage, M., Mowat, A.P., and Howard E.R. Biliary atresia splenic malformation syndrome: an etiologic and prognostic subgroup. Surgery, 1993;113(6):662–8. 188. Karrer, F.M. and Lilly J.R. Correction of biliary atresia and jejunal atresia in an infant. J Pediatr Surg, 1993;168(4):469–76. 190. Roach, J.P. and Bruny J.L. Advances in the understanding and treatment of biliary atresia. Curr Opin Pediatr, 2008;20(3):315–9.

363

191. Gershoni‐Baruch, R., Gottfried, E., Pery, M., Sahin, A., and Etzioni, A. Immotile cilia syndrome including polysplenia, situs inversus, and extrahepatic biliary atresia. Am J Med Genet, 1989;33(3):390–3. 192. Mazziotti, M.V., Willis, L.K., Heuckeroth, R.O. et al. Anomalous development of the hepatobiliary system in the Inv mouse. Hepatol, 1999;30(2):372–8. 193. Shimadera, S., Iwai, N., Deguchi, E., Kimura, O., Fumino, S., and Yokoyama, T. The inv mouse as an experimental model of biliary atresia. J Pediatr Surg, 2007;42(9):1555–60. 194. Otto, E.A., Schermer, B., Obara, T., O’Toole J.F., Hiller, K.S., Mueller, A.M. et al. Mutations in INVS encoding inversin cause nephronophthisis type 2, linking renal cystic disease to the function of primary cilia and left‐ right axis determination. Nat Genet, 2003;34(4):413–20. 195. Olbrich, H., Fliegauf, M., Hoefele, J. et  al. Mutations in a novel gene, NPHP3, cause adolescent nephronophthisis, tapeto‐retinal degeneration and hepatic fibrosis. Nat Genet, 2003;34(4):455–9. 196. Johnson, C.A., Gissen, P., and Sergi, C. Molecular pathology and genetics of congenital hepatorenal fibrocystic syndromes. J Med Genet, 2003;40(5): 311–9. 197. Adams, M., Smith, U.M., Logan, C.V., and Johnson C.A. Recent advances in the molecular pathology, cell biology and genetics of ciliopathies. J Med Genet, 2008;45(5):257–67. 198. Ibañez‐Tallon, I., Heintz, N., and Omran, H. To beat or not to beat: roles of cilia in development and disease. Hum Mol Genet, 2003;12(1):R27–35. 199. Lina, F. and Satlinb, L.M. Polycystic kidney disease: the cilium as a common pathway in cystogenesis. Curr Opin Pediatr, 2004;16(2):171–6. 200. Phillips, M.J., Azuma, T., Meredith, S.L. et al. Abnormalities in villin gene expression and canalicular microvillus structure in progressive cholestatic liver disease of childhood. Lancet, 2003;362(9390):1112–9. 201. Nicastro, E. and D’Antiga, L. Next generation sequencing in pediatric hepatology and liver transplantation. Liver Transpl, 2018;24(2):282–93. 202. Chen, H.L., Li, H.Y., Wu, J.F. et al. Panel‐based next‐generation sequencing for the diagnosis of cholestatic genetic liver diseases: clinical utility and challenges. J Pediatr, 2018;171:171–7. 203. Togawa, T., Sugiura, T., Ito, K. et al. Molecular genetic dissection and neonatal/infantile intrahepatic cholestasis using targeted next‐generation sequencing. J Pediatr, 2016;171:171–7.

30

Pathophysiologic Basis for Alternative Therapies for Cholestasis Claudia D. Fuchs, Emina Halilbasic, and Michael Trauner Hans Popper Laboratory of Molecular Hepatology, Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Austria

INTRODUCTION Cholestasis is characterized by impaired bile formation with insufficient amounts of bile reaching the duodenum resulting in intrahepatic and systemic accumulation of bile acids (BAs) and other potentially toxic cholephiles [1]. Important causes include disturbances of hepatocellular and/or cholangiocellular bile secretion and destruction/obstruction of smaller and/or larger bile ducts by mechanical processes (e.g. stones, tumors) or immune‐mediated fibrosing cholangiopathies such as primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC). Ideally, causal therapy of cholestasis resolves the ­underlying cholestatic insult and promotes recovery which is currently possible only in some cases (e.g. removal of stone, surgical removal or stenting of tumor/stricture, delivery in intrahepatic cholestasis of pregnancy, discontinuation of causative drug in drug‐induced liver injury [DILI]). Even when the underlying cause has been resolved, recovery can be slow (e.g. persistent hepatocellular secretory failure after DILI) requiring supportive therapy. Since the etiologies of many cholestatic disorders such as PBC and PSC are not fully understood and causal therapies are still lacking, stimulation of bile secretion and adaptive mechanisms may attenuate liver injury irrespective of the cause of cholestasis. Of note, targeting the immune mechanisms underlying or at least contributing to the pathogenesis of PBC and PSC has so far been rather disappointing. The first‐line therapy for many cholestatic liver diseases is ursodeoxycholic acid (UDCA), a hydrophilic BA that reduces the toxicity of the endogenous BA pool and has several ­additional beneficial therapeutic mechanisms [2]. As a potent intracellular signaling molecule UDCA stimulates hepatobiliary secretion (mainly by promoting vesicular targeting of

transporters in a post‐transcriptional manner) thereby counteracting cholestasis [2]. In addition, UDCA stimulates biliary bicarbonate (HCO3−) secretion supporting the “biliary HCO3− umbrella” [3], an important protective mechanism for hepatocytes and cholangiocytes against highly toxic BA monomers in  bile by maintaining alkaline pH along the apical surface. In  addition, UDCA may also promote biliary phospholipid secretion thereby facilitating the formation of mixed micelles and limiting the amounts of free, non‐micellar bound BAs. As chaperone UDCA also reduces endoplasmatic reticulum stress, exerts anti‐apoptotic effects, and its activation of the glucocorticoid receptor (GR/NR3C1) may be responsible for some direct anti‐inflammatory effects, it further contributes to the cytoprotective action of UDCA in cholestatic liver disease [2]. As such, UDCA has found application in the treatment of a broad range of cholestatic diseases but it has been approved only as first‐line treatment for PBC, where UDCA improves survival [4]. In other cholestatic disorders, such as PSC, UDCA often improves serum liver tests, without proven survival benefit. Increasing its dose (28–30 mg kg−1/day) was even harmful in PSC [2] although the mechanisms are still incompletely understood. However, amelioration in liver enzymes upon UDCA treatment (in moderate dose) and their exacerbation after UDCA withdrawal may point toward potential beneficial effects. The field has made impressive progress since introduction of UDCA [2]. The identification of specific BA transport systems and dedicated BA receptors [5] now allows a more specific targeting of impaired BA homeostasis (including adaptive/alternative transport, metabolic pathways, and regulatory networks) irrespective of the etiology of cholestasis. This chapter will focus on the pathophysiologic basis for alternative therapies for cholestasis beyond UDCA, the first available drug for cholestasis.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

NUCLEAR RECEPTORS Nuclear receptors (NRs) are the largest family of transcription factors comprising 48 members in humans and 49 members in mice, which share a common structural organization and are activated upon ligand binding [6]. They coordinate several key hepatic functions including the regulation of BA synthesis and hepatobiliary excretory function, glucose and lipid metabolism, inflammatory processes, fibrosis as well as liver regeneration, and tumorigenesis [7]. After ligand (e.g. BA) binding, NRs change their conformation, facilitating the recruitment of ­cofactors and dissociation of corepressors, implementing DNA binding and stimulating transcription of genes typically involved in metabolism and/or transport of the ligand (e.g. BAs) thus constituting the molecular basis for feedback regulation of metabolism [6]. Due to their central role in hepatic (patho)physiology, NRs have become attractive targets for anticholestatic drugs serving as ligands for these receptors. The most relevant NRs involved in BA signaling and hepatobiliary homeostasis include the farnesoid X receptor (FXR/NR1H4), pregnane X  receptor (PXR/NR1I2), constitutive androstane receptor (CAR/NR1I3), and vitamin D receptor (VDR/NR1I1). Furthermore, the glucocorticoid receptor (GR/NR3C1) and the fatty acid activated peroxisome proliferator‐activated receptors (PPARα, PPARγ, and PPARδ) as regulators of inflammation, fibrosis, and energy metabolism may also impact on biliary homeostasis and thereby counteract cholestatic liver injury. Notably the current standard therapy of cholestasis, UDCA, does not activate FXR (but may even counteract its activation by reducing the relative concentration of BAs with more potent FXR ligand function [8]), and has low affinity to GR and VDR mediating its anti‐inflammatory effects and may indirectly activate PXR after being metabolized into lithocholic acid (LCA) [2].

Nuclear bile acid receptor (FXR) FXR serves as the main NR for BAs [8–10] regulating their ­synthesis and hepatic uptake as well as elimination from the liver. It forms heterodimers with the retinoid X receptor (RXR, NR2B), binding to the inverted repeat (IR)‐1 within the promoter sequence of target genes. FXR is predominantly present in liver, gastrointestinal tract, kidney, and adrenal glands acting as a central regulator of BA homeostasis, lipid, glucose [11], and amino acid metabolism [12] as well as inflammation [7] and liver regeneration [13]. FXR has a key role in controlling BA synthesis by regulating the expression of rate limiting enzymes of both classical (CYP7A1) and alternative pathways (CYP8B1 and CYP27A1). This is mediated by induced expression of two key regulators, an orphan nuclear receptor, small heterodimer partner (SHP, NR0B2) [11] and an intestinal hormone, fibroblast growth factor 15/19 (human FGF19, mouse Fgf15) [14, 15]. BA‐activated FXR induces the hepatic expression of SHP, an atypical nuclear receptor lacking a DNA binding domain that functions as a potent transcriptional repressor, which in turn, interacts with transcription factors such as hepatocyte nuclear factor (HNF)‐4α/ NR2A1 and liver receptor homolog (LRH)‐1/NR5A2 to inhibit their function and suppress CYP7A1 expression [16]. The ileal hormone FGF15/19 is secreted in the portal circulation upon

365

BA activation of FXR in enterocytes and reaches the liver, where it binds to fibroblast growth factor receptor 4 (FGFR4)/β‐ Klotho complex and activates c‐Jun N‐terminal kinase‐­ dependent pathway which ultimately leads to CYP7A1 repression [15] (Figure  30.1). Of note, SHP is required for FGF15/19 to efficiently repress BA synthesis and mice lacking SHP are refractory to the inhibitory effects of either FXR ­agonists or FGF15/19 on Cyp7a1 expression [7]. FXR further fine‐tunes BA levels through induction of other target genes, including transcription factor V‐Maf avian musculoaponeurotic  fibrosarcoma oncogene homolog G (MAFG), a global transcriptional repressor of BA synthesis, lysine‐specific histone demethylase (LSD1), a key repressive epigenetic component in a SHP complex, and β‐Klotho, the obligate co‐receptor for FGF15/19 [17]. Apart from regulation of BA synthesis, the intrahepatic BA  concentration is controlled by FXR‐mediated inhibition of  the basolateral BA uptake transporter sodium/taurocholate co‐transporting polypeptide (NTCP/SLC10A1) and upregulation of the canalicular BA transporter bile salt export pump (BSEP/ABCB11) in hepatocytes (Figure 30.1), promoting BA elimination via induction of bile flow. Beside this induction of BA dependent bile flow, FXR also increases BA‐independent bile flow promoting biliary HCO3− transport by facilitating complex formation of carboanhydrase 14 and anion exchanger 2 (AE2) [18]. FXR also induces alternative basolateral BA transport in hepatocytes via organic solute transporter alpha and beta (OSTα/β; SLC51A/SLC51B) and detoxification via enzymes such as Cyp3a11/CYP3A4, Sult2a1/SULT2a1 (sulfotransferase family 2A member 1), and Ugt2b4/UGT2b4 [7] further preventing the accumulation of BAs within target cells (Figure 30.1). In the liver, FXR is also present in cells other than hepatocytes. In cholangiocytes FXR regulates ductular bile formation by alkalization and fluidization via induction of vasoactive intestinal polypeptide receptor 1 (VPAC‐1) [7]. Moreover, FXR activation in cholangiocytes induces expression of the antimicrobial peptide cathelicidin, indicating an important role of BAs/FXR in maintaining sterility of biliary tree and protection against bile duct inflammation [19]. Activation of FXR has antifibrotic effects [18], the underlying mechanisms and FXR expression in hepatic stellate cells (HSC) are discussed controversially. FXR has been reported to have a direct antifibrotic effect via activation of SHP in HSC [7], but other studies, however, have revealed only very low or even no FXR and SHP expression in human HSCs and murine periductal myofibroblasts [20], arguing for a more indirect antifibrotic effect through control of cholestasis and inflammation. Present also in liver sinusoidal endothelial cells [21], FXR ameliorates portal hypertension not by targeting fibrosis, but by directly interfering with endothelial function. FXR activation reduces intrahepatic vascular resistance by upregulation of eNOS, cystathionase, and dimethylaminohydrolase 1, promoting the generation of vasodilators such as nitric oxide and hydrogen sulfide, as well as a reduction of intrahepatic vasoconstrictors endothelin‐1 and p moesin [18]. Mice lacking FXR display increased hepatic inflammation at  baseline and are more susceptible to lipopolysaccharide (LPS)‐induced inflammation [18]. FXR may have direct anti‐ inflammatory effects by interference with nuclear factor

FXR PXR PPARα

PXR CAR

PXR

Inflammation e.g. NFκB βKlotho

BSEP

FGF 15/19

FGFR4

MDR2/3

FXR FXR SHP

BA synthesis

MRP2 PPAR α/δ

OSTβ

GR

PL HCO3 -

PPARα

JNK

NTCP

GR FXR

secretion of hydroxylated & sulfated BAs

BAs

FXR BAs

Hepatic stellate cell

Alternative BA export

FXR FGF 15/19

OSTα

MRP4

THE LIVER:  NUCLEAR RECEPTORS

MRP3

366

PXR

Fibrosis

FXR PXR CAR

Bilirubin

VDR

PPARγ: MCP1, migration, proliferation PXR: TGF1β, transdifferentiation, proliferation VDR: SMAD pathway

AE2

conjugated

PPARγ PXR

PPAR α/δ

Immune

cells

BA detoxification CYPs, SULTs, UGTs

CYP7A1

Macrophage

CD4+ T cell

Hepatocyte PPARα PPARγ PPARδ

PPARγ Reactive cholangiocyte Phenotype VCAM1

FXR VDR GR Cholangioprotective VPAC1, AE2, Cathelecithin

Cholangiocyte

VDR

Inflammation PPARα: phosphorylation of p38 in CD4+ T cells PPARγ: cytocine secretion from macrophages PPARδ: M2 activation of macropages VDR: production of Tregs

Figure 30.1  Nuclear receptors as therapeutic targets of alternative therapies in cholestasis. Nuclear receptor activation in hepatocytes maintains the balance between BA synthesis, detoxification, uptake, and excretion by regulation of expression of hepatobiliary transporters and metabolic pathways. FXR (in a SHP dependent manner) represses hepatic BA uptake (NTCP) and BA synthesis (CYP7A1). Moreover intestine‐derived FGF15/19 to the FGFR4/bKlotho dimer counteracts CYP7A1 expression. Of note, human hepatocytes (in addition to enterocytes) also express FGF19. FXR promotes BA secretion via induction of canalicular transporters (BSEP, MRP2, and MDR2/3) and induces BA elimination via alternative basolateral BA transporter OSTα/β. CAR and PXR regulate the adaption to increased intracellular BA concentrations via upregulation of MRP3 and 4 (alternative BA export) as well as induction of detoxification/hydroxylation enzymes. PPARα stimulates phospholipid secretion (via MDR2/3) and is also involved in regulation of detoxification pathways. Activation of AE2 expression by GR stimulates biliary bicarbonate secretion, thus reducing bile toxicity. Apart from regulating BA homeostasis, NRs have additional anti‐inflammatory and antifibrotic effects (right‐hand panel). Their activation may result in induction of defensive mechanisms in bile duct epithelial cells. Activation of PPARγ in cholangiocytes reduces vascular cell adhesion molecule (VCAM‐1) expression, thereby counteracting reactive cholangiocyte phenotype. Activation of FXR, VDR, and GR in cholangiocytes exert cholangioprotective effects via upregulation of VPAC1, AE2, and Cathelicidin, respectively. Antifibrotic effects of NRs are related to their activation in hepatic stellate cells. PPARγ, PXR, and VDR reduce expression of profibrogenic genes such as MCP1, TGFβ1, and SMAD, respectively. Furthermore, they reduce migration, proliferation, as well as transdifferentiation of HSC into myofibroblasts. Anti‐inflammatory effects of NRs are related to their activation in immune cells such as macrophages and CD4+ T cells. Green arrows indicate stimulatory effects and red lines indicate suppressive effects. AE, anion exchanger; BAs, bile acids; Bili‐glu, bilirubin glucuronide; BSEP, bile salt export pump; CAR, constitutive androstane receptor; CYP7A1, cholesterol‐7α‐hydroxylase; CYPs, cytochrome P450 enzymes; FGF, fibroblast growth factor; FXR, farnesoid X receptor; GR, glucocorticoid receptor; HSC, hepatic stellate cell; MDR3, multidrug resistance protein 3; MRP2, multidrug resistance‐ associated protein 2; MRP3, multidrug resistance‐associated protein 3; MRP4, multidrug resistance‐associated protein 4; NTCP, sodium taurocholate cotransporting polypeptide; OSTα/β, organic solute transporter α and β; PC, phosphatidylcholine; PXR, pregnane X receptor; PPARα, peroxisome proliferator‐activated receptor α; PPARγ, peroxisome proliferator‐activated receptor γ; SHP, small heterodimer partner; SULTs, sulfatation enzymes; UGTs, glucuronidation enzymes; VDR, vitamin D receptor.

kappa‐b (NFκB) [18]. Recently, FXR has also been implicated in activation of hepatic natural killer T cells and hepatic accumulation of myeloid‐derived suppressor cells, counteracting immune‐mediated liver injury in rodents [18]. Anti‐­ inflammatory effects of FXR were not only observed in liver, but also in intestine since pharmacological FXR activation improved intestinal inflammation and permeability in colitis models [7]. Reciprocal interactions between BAs and immune cells may exist, since the BA receptors FXR and TGR5/GBAR‐1 (Takeda G protein‐coupled receptor/G protein‐coupled bile acid receptor 1) are expressed in macrophages [22]. BAs have been identified on one hand as damage‐associated molecular patterns (DAMPs), triggering inflammatory response by binding to

inflammasomes [22] and on the other hand specific BA species (LCA) have been shown to inhibit inflammasome activation [23]. Beside its direct anti‐inflammatory properties and modulation of immune cell function, FXR represents a central molecular switch controlling intestinal microbiota by protecting the gut from bacterial overgrowth and disruption of the epithelial barrier [24]. Conversely, FXR dysfunction drives bacterial translocation, facilitated by increased intestinal permeability and intestinal inflammation [18]. Pharmacological activation of FXR improved gene expression of tight junction proteins such as claudin‐1 and occludin [25] as well as expression of inducible isoform nitric oxide synthase (iNOS) and angiopoietin 1 (Ang1), genes involved in antibacterial defense by producing



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

antimicrobial peptides [26]. Activation (or intestinal overexpression [24]) of FXR decreased bacterial overgrowth and improved liver injury in the bile duct ligation (BDL) [24] as well as in the α‐naphthyl‐isothiocyanate (ANIT) model [27]. In line, obstructive jaundice with virtually complete absence of BAs from the intestinal lumen promotes bacterial translocation in patients [28]. Loss of FXR results in severe disruption of BA homeostasis as reflected by increased BA synthesis as well as increased BA uptake and reduced hepatobiliary BA secretion in FXR knockout (KO) mice [29]. Although metabolic adaption to cholestatic conditions does not entirely rely only on FXR and – at least in mice – can be partially compensated by CAR/PXR‐dependent overexpression of detoxifying enzymes and alternative efflux systems [18, 30], a loss‐of‐function mutation in the FXR gene results in severe neonatal progressive familial intrahepatic cholestasis 5 (PFIC5) [31]. While impaired function of FXR leads to profound dysregulation of BA homeostasis, pharmacological activation of FXR has shown multiple beneficial effects in preclinical models of cholestasis. Collectively, FXR: (i) suppresses hepatic BA synthesis, reduces intestinal and hepatic BA uptake, while stimulating canalicular and basolateral BA efflux, thus resulting in a net reduction of the toxic hepatic BA overload; (ii) alters bile composition by increasing phospholipid and HCO3− secretion finally resulting in less toxic bile; (iii) mediates direct anti‐inflammatory effects in hepatocytes and non‐parenchymal liver cells, together with immunomodulatory effects in the adaptive immune system; (iv) impacts on the gut–liver axis by induction of FGF15/19, a suppressor of BA synthesis; (v) reduces bacterial overgrowth and intestinal permeability; (vi) ameliorates the complications of advanced liver disease such as portal hypertension and possibly carcinogenesis. These features make FXR ligands attractive compounds for alternative therapies in cholestasis and immune‐mediated disorders such as PBC and PSC including associated/underlying inflammatory bowel ­disease (IBD). Various steroidal (BA derived) and non‐steroidal (non‐BA derived) FXR activators have been developed and have shown a range of beneficial effects in animal models of cholestasis. More specifically, the non‐steroidal FXR agonist GW4064 and the steroidal FXR agonist obeticholic acid (OCA/INT747  –  a 6‐ethyl derivative of the primary BA chenodeoxycholic acid [18]) showed positive effects in chemically induced cholestatic liver injury (α‐naphthyl isothiocyanate (ANIT) and estradiol induced cholestasis), in bile duct ligated animals as well as in the Mdr2 KO mouse model of sclerosing cholangitis [18]. Based on these encouraging preclinical data, FXR ligands have already entered clinical development. Add‐on treatment with OCA in PBC patients, who biochemically did not respond sufficiently to UDCA or OCA monotherapy, improved liver enzymes and markers of inflammation [7, 32, 33]. These studies led to conditional approval of OCA as a second‐line therapy for the treatment of PBC patients with incomplete biochemical response or intolerance to standard therapy with UDCA, which is already reflected by current guidelines [4]. Long‐term clinical effects of OCA treatment in PBC patients are currently under investigation. Since OCA functions as systemic FXR ligand undergoing enterohepatic circulation the risk of overdosing in cirrhotic PBC patients needs to be taken into account. The role

367

of OCA in other cholestatic liver disorders such as PSC is undergoing clinical evaluation [34]. Non‐steroidal FXR agonists (such as LJN452/Tropifexor and GS‐9674) have different pharmacokinetic properties favoring intestinal (instead of hepatic) FXR activation, and also do not undergo enterohepatic circulation. Differential targeting of FXR in hepatocytes, cholangiocytes, and/or enterocytes could critically determine the pharmacological effects of steroidal versus non‐steroidal FXR ligands. Whether non‐steroidal FXR ligands lack possible BA‐structure‐related side‐effects, such as pruritus, needs to be determined [35]. An open question remains whether activation of intestinal FXR resulting mainly in inhibition of BA synthesis via FGF15/19 signaling is sufficient to counteract cholestatic liver disease or whether more systemic anti‐­ inflammatory effects are necessary. Another important G‐protein coupled BA receptor, TGR5, is expressed in cholangiocytes, gallbladder epithelium, as well as endothelial cells and Kupffer cells (but not in hepatocytes), where it regulates biliary HCO3− secretion and has anti‐inflammatory effects [18] (also see Chapter  24 for more details). In contrast to FXR [18], TGR5 stimulates production of intestinal glucagon like peptide 1 (GLP‐1) [7]. GLP‐1 signaling may protect cholangiocytes from apoptosis and thus modulate their adaptive response to cholestasis in mice [18]. However, TGR5 expression is reduced in Mdr2 KO mouse model of sclerosing cholangitis [36] and a pure TGR5 agonist did not improve liver disease in this model [18]. Additional concerns include a ­potential role of TGR5 in development of pruritus [18], potential pro‐proliferative and anti‐apoptotic effects in isolated human cholangiocytes and cholangiocellular carcinoma cells [18] as well as TGR5 overexpression in human cholangiocarcinoma [37] (for more details see Chapter 24).

Fibroblast growth factor 19 (FGF19) Activation of FXR stimulates production of endogenous FGF15/19 in the ileum, which is transported to the liver where it binds to FGFR4 and its co‐receptor β‐Klotho in hepatocytes, contributing to suppression of hepatic BA synthesis via c‐Jun N‐terminal kinase (JNK) and SHP, in addition to other effects on lipid and glucose metabolism [15]. Interestingly, FGF19 may also exert anti‐inflammatory effects by counteracting NFκB signaling [38]. Selective overexpression of intestinal FXR improved liver injury in the Mdr2 KO mouse model of sclerosing cholangitis via induction of intestinal FGF15 expression and subsequent reduction of BA pool size. In line, FGF19 application to bile duct ligated mice reduced total BA pool size [39], improved jaundice and serum levels of liver transaminases as well as necrosis. However, stimulation of endogenous FGF15/19 could also have side‐effects such as pro‐proliferative or pro‐carcinogenic effects. Endogenous FGF19 and its receptor FGFR4 have been implicated in development of hepatocellular carcinoma (HCC) and inhibitors of FGFR4‐related signaling are even developed as anticancer drugs [18]. As such, FGF19 is amplified in human HCC and in mice ectopic overexpression of FGF19 accelerates tumor development [18]. Importantly, FGF19 mimetics such as M70/NGM282 have amino acid modifications at the N‐terminus allowing the dissection of metabolic from pro‐proliferative actions thus maintaining their BA

368

THE LIVER:  NUCLEAR RECEPTORS

metabolic regulatory properties while lacking their pro‐­ proliferative effects. In line, adeno‐associated virus mediated expression of M70 in Mdr2 KO mouse model of sclerosing cholangitis decreased hepatic inflammation and biliary fibrosis [40]. Notably, prolonged M70 treatment did not accelerate HCC formation in the Mdr2 KO mouse and even had tumor preventive effects, while FGF19 application promoted hepatic carcinogenesis [40]. In PBC patients with insufficient response to UDCA NGM282 improved cholestatic liver enzymes [40]. Diarrhea, headache, and nausea were observed as side‐effects, suggesting that FGF19 may also impact on intestinal motility. Of note, NGM282 has been shown to improve markers of hepatic fibrosis without reducing cholestatic liver enzymes such as alkaline phosphatase in patients with PSC [40] suggesting that FGF19 could also mediate potential direct anti‐inflammatory and antifibrotic actions. To understand the full (patho)physiological and therapeutic implications of FGF19 it has to be kept in mind that  –  in contrast to rodents  –  human FGF19 is also produced by hepatocytes, bile ducts, and gallbladder epithelium [41]. Interestingly, serum FGF19 correlates positively with severity of the cholestasis [18]. The extent to which patients with elevated endogenous levels of FGF19 may benefit from  exogenous FGF19 mimetic application remains to be ­elucidated. For example, the FGF19 mimetic could also compete with endogenous FGF19 at the FGFR4 binding site [18] potentially counteracting development of HCC, while maintaining the beneficial metabolic effects on BA homeostasis.

Peroxisome proliferator‐activated receptors (PPARs) PPARα (NR1C1), PPARγ (NR1C3), and PPARδ (NR1C2), are structural homologues which are activated by endogenous fatty acids and their derivatives [7]. PPARδ is ubiquitously expressed, while PPARα is predominantly found in organs responsible for fatty acid catabolism such as liver, heart, kidney, brown adipose tissue, small and large intestine. PPARγ is highly expressed in adipose tissue and immune cells [42]. The main function of PPARs is the regulation of lipid and energy homeostasis. Beyond the realm of metabolism, PPARs also exert direct anti‐ inflammatory effects. Apart from being a key regulator of hepatic lipid metabolism, PPARα is also involved in BA metabolism and BAs can indirectly stimulate human PPARα via FXR [43]. PPAR activation via bezafibrate represses BA synthesis (through reduction of HNF4α binding to the CYP7A1 promoter) [7] as well as hepatocellular BA uptake (reduced Ntcp expression by direct mechanisms) [44]. Bezafibrates (PPARα ligand with additional broader pan PPAR activity) also induce expression of enzymes involved in BA detoxification (Sult2a1, Ugt2b4, and Ugt1a3) [7]. In vitro, the PPARα agonist ciprofibrate also induced the BA uptake transporter ASBT (apical sodium dependent bile salt transporter) in human cholangiocytes and Caco2 cells [2]. PPARα stimulates expression of multidrug resistant 3 (MDR3) [44] thereby potentially contributing to elevated biliary ­phospholipid (PL) secretion and counteracting BA toxicity by promoting formation of mixed micelles. In addition, PPARα exerts anti‐inflammatory effects by ­transrepressing NFκB signaling via induction of IκBα (which

sequesters NFκB) [45]. Fenofibrate (a more selective PPARα ligand) downregulates the IL6 receptor subunits gp80 and gp130 in the liver [46], thereby reducing phosphorylation of STAT3 and c‐Jun [47]. In addition to PPARα, PPARγ and PPARδ activation reduces production of inflammatory mediators and cytokines in various immune cells including monocytes/macrophages, dendritic cells, lymphocytes, as well as cholangiocytes [7]. Activation of all PPAR isoforms also has antifibrotic effects. Although HSC do not show pronounced PPARα expression, its activation led to reduced HSC activation and repression of fibrotic markers [48]. PPARγ is expressed only in quiescent HSC and has antifibrotic effects by suppressing their activation, proliferation, and migration [49]. Accordingly, loss of PPARγ is a typical feature of HSC activation [49].Treatment with thiazolidinedione (a PPARγ agonist) improved bile duct proliferation and hepatic fibrosis in bile duct ligated rats [7]. The plant extract curcumin, the yellow pigment of the spice turmeric, also activated PPARγ in cholangiocytes, decreased infiltration of inflammatory cells, and improved portal inflammation and fibrosis in Mdr2 KO mouse model [7]. In contrast to PPARα and PPARγ, PPARδ activation had pro‐ as well as antifibrotic effects. Application of KD3010, but not GW501516, (two different PPARδ agonists) was hepatoprotective and antifibrotic in mice subjected to BDL and CCl4. In vitro that KD3010, in contrast to GW501516, increased the expression of several CYP enzymes resulting in reduced reactive oxygen species production, possibly counteracting hepatic fibrosis [50]. Patients with obstructive jaundice undergoing percutaneous transhepatic biliary drainage that received bezafibrate treatment showed increased biliary PL secretion. However, the same study showed that MDR3 expression is already increased in PBC patients at baseline and was not further upregulated by bezafibrate treatment [51], indicating that bezafibrate may induce beneficial clinical effects via PL‐independent mechanisms. Fibrates such as bezafibrate and fenofibrate have shown beneficial effects in several smaller, mostly uncontrolled studies in patients with PBC (not responding to UDCA) and  –  to lesser extent – in PSC [52]. Recently, a larger randomized controlled trial demonstrated pronounced improvement in liver enzymes, liver stiffness, and pruritus by bezafibrate in PBC patients with insufficient response to UDCA [53] and fibrates may be considered a second‐line treatment in PBC [54]. Future long‐term studies will have to explore the safety and efficacy of fibrates in PBC and other cholestatic disorders. A selective and potent PPARδ agonist MBX‐8025, Seladelpar, reduced alkaline phosphatase in a phase II proof‐of‐concept study in PBC [55] and further clinical trials are ongoing. Another promising PPARα/δ dual agonist, elafibranor, with beneficial effects in non‐alcoholic steatohepatitis (NASH) and diabetes [56], is currently also being tested in PBC.

Glucocorticoid receptor (GR) GR/NR3C1 is ubiquitously expressed. Apart from its well‐ known anti‐inflammatory and immunosuppressive function, GR regulates several metabolic pathways including not only carbohydrate and protein, but also BA homeostasis [2]. Glucocorticoids (via GR) induce BA uptake systems such as NTCP, ASBT, and



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

OSTα/β [2]. Ligand‐activated GR also upregulates AE2 expression, which should result in increasing cholangiocellular bicarbonate secretion and restoring the bicarbonate umbrella [57]. This observation is of particular interest since AE2 expression is reduced in liver and inflammatory cells of PBC patients [2], possibly responsible for making cholangiocytes more susceptible to an autoimmune first hit. Besides glucocorticoids, UDCA also activates GR. The combination of UDCA and another GR ligand dexamethasone (but not UDCA or dexamethasone alone) increased expression and function of AE2 via interaction with HNF1 and GR on the AE2 alternate promoter [57]. UDCA‐­ activated GR is also anti‐inflammatory through suppressing NFκB‐dependent transcription by preventing GR – p65 interaction [2] representing another example for the dichotomous effects of NRs on both metabolism and inflammation. Glucocorticoids signal not only via GR, but also other NRs such as CAR, and glucocorticoids increase expression of PXR and RXRα [43]. Budesonide also activates PXR in mice and humans, while dexamethasone directly activates PXR in mice but not in humans [58]. Clinically prednisolone and budesonide were tested in combination with UDCA in PBC patients. While the combination of prednisolone with UDCA was not superior to UDCA monotherapy [2], budesonide UDCA combination significantly improved serum liver tests compared to UDCA alone [2].

Xenobiotic sensors PXR and CAR Pregnane X receptor (PXR/NR1I2) and constitutive androstane receptor (CAR/NR1I3) are key regulators of enzymes involved in BA hydroxylation/detoxification [2] and alternative BA export. Both receptors are low affinity, broad specificity xenosensors, activated by a range of compounds which are structurally unrelated such as antibiotics including rifampicin, clotrimazole, the antidepressant St. John’s wort, and synthetic steroids such as 5b‐pregnane‐3,20‐dione, pregnenolone 16a‐carbonitrile (PCN) and dexamethasone [2, 59]. In addition to xenobiotics, also potentially toxic endogenous cholephiles such as secondary BAs (e.g. LCA) and bilirubin activate these receptors [60]. PXR and CAR mitigate potentially harmful effects of toxic BAs by coordinated activation of detoxification pathways such as hydroxylation (Cyp3a11/CYP3A4, and CYP2B10), sulfation (SULT2a1/Sult2a1), conjugation (BACS/Bacs, BAT/Bat), and subsequent elimination via basolateral export pumps multidrug resistance‐associated protein 3 (MRP3) and MRP4 (Figure 30.1) [43]. In addition, they regulate bilirubin metabolism in the liver inducing its glucuronidation (UGT1A1/Ugt1a1) and canalicular excretion (MRP2) (Figure 30.1) [43]. PXR was shown to inhibit the interaction between PGC1α and HNF4α, thereby repressing CYP7A1 gene transcription and BA synthesis [7]. Interestingly, PXR was identified as FXR target gene [61], pointing toward a potential NR crosstalk in the protection from BA toxicity. Besides beneficial effects on BA detoxification and elimination, PXR may also exert direct anti‐inflammatory and antifibrotic effects. In mice, activation of PXR inhibits inflammation and fibrosis, while PXR KO mice are more susceptible to inflammatory stimuli [7]. In human HSCs, activation of PXR inhibits expression of the major profibrogenic cytokine TGF‐1β, preventing their transdifferentiation to fibrogenic myofibroblasts and slows down proliferation [62].

369

CAR ligands such as phenobarbital and PXR ligands such as rifampicin have been used to treat pruritus for decades [43]. These antipruritogenic effects could be explained by their impact on BA detoxification and alternative BA elimination outlined above. Of note, rifampicin is recommended as second‐line treatment for pruritus by the current guidelines [4]. Moreover, PXR and CAR activation may not only alleviate pruritus, but also have anticholestatic effects. As such, rifampicin (PXR ligand) improved liver function in PBC patients [63] and p­ henobarbital (CAR ligand) reduced serum BA concentration [43]. Interestingly, traditional Chinese medicines which were used to treat neonatal jaundice by enhancing bilirubin clearance have been identified as CAR ligands (e.g. Yin Chin/Artemisia capillaris) [64]. Stimulation of an adaptive response to cholestasis may ­particularly be useful when the underlying cause cannot be improved. Moreover, persistent secretory failure after DILI and prolonged mechanical cholestasis may be overcome by PXR ligands such as rifampicin [2]. Stimulation of detoxification (e.g. via CAR and PXR agonists) may complement other therapies (e.g. FXR agonists) targeting mainly hepatobiliary flow and excretion of BAs and other cholephiles. Therefore, more specific and less toxic CAR/PXR agonists could be a valuable addition to the therapeutic armamentarium in cholestasis.

Vitamin D and A receptors Vitamin D receptor (VDR/NR1I1) is not only involved in the regulation of calcium homeostasis, but also in cell proliferation and differentiation. Expression of VDR is high in intestine and low in liver where it is mainly found in components of the immune system including Kupffer cells, monocytes, natural killer cells, T and B lymphocytes, and dendritic cells, endothelial cells, cholangiocytes, and HSCs [43]. Vitamin D represses Th1 and Th17 response and increases differentiation of ­regulatory T cells and NK cells, implying a critical role in the pathogenesis of autoimmune diseases [65]. In line, VDR polymorphisms are associated with several immune‐mediated liver diseases such as PBC and autoimmune hepatitis [7]. VDR, ­acting as a potential BA sensor along the gut–liver axis, also functions as a receptor for the secondary BA LCA, which is rather hydrophobic/cytotoxic. Activation of VDR by LCA ­stimulates Cyp3a11 (and its human homologue of CYP3A4) expression, an enzyme that detoxifies LCA in liver and intestine [7]. Additionally, SULT2A1/Sult2a1 (an enzyme catalyzing ­sulfation of BAs) as well as basolateral export pump MRP3 was upregulated by vitamin D application in enterocytes [7]. The role of the VDR pathway in BA detoxification and elimination may have also emerged to shield first‐pass tissues from the ­cytotoxic effects of BA overload. Notably, VDR activation in cholangiocytes also increased expression of the antimicrobial peptide cathelicidin, which could contribute to the anti‐inflammatory properties of VDR in immune‐mediated bile duct injury [7]. Cathelicidin is able to neutralize the deleterious effect of LPS which accumulates in the biliary tree in fibrosing cholangiopathies [19]. Furthermore, activation of VDR ameliorated fibrosis, possibly by epigenetic modifications of the SMAD pathway in HSCs [2]. However, administration of vitamin D to Mdr2 KO mice did not improve hepatic fibrosis despite ameliorating serum liver enzymes [2].

370

THE LIVER:  TRANSPORTER MODULATION

Low vitamin D levels have been linked to progression of fibrosis in various liver diseases including cholestatic liver injury [7]. Low serum vitamin D level is also an indicator of advanced disease and a predictor of UDCA non‐response in PBC patients [66]. Whether (high dose) vitamin D supplementation may also improve hepatic immune‐tolerance, thereby reducing biliary injury and subsequent fibrosis in PBC remains to be evaluated. Vitamin A is the umbrella term for diverse retinoids and their precursor forms. Retinoids are important regulators of cell proliferation and differentiation that bind to two families of NRs, retinoic acid receptors (RARα, β, and γ/NR1B1, B2, and B3) and retinoid X receptors (RXRα, β, and γ/NR2B1, B2, and B3) [65]. A biologic active form of vitamin A  –  all‐trans retinoic acid (ATRA) and its stereoisomer, 9‐cis‐retinoic acid – activate RARs, whereas RXRs are mainly activated by 9‐cis‐retinoic acid [67, 68]. RARs and RXRs are known to be key regulators in activation of HSC, transdifferentiating them from vitamin A‐storing cells to myofibroblasts, which are proliferative, contractile, inflammatory, and chemotactic and characterized by enhanced matrix production [69]. In line with loss of vitamin A, activation of HSC is accompanied by downregulation of RAR and RXR expression [70]. On the other hand, activation of RAR and RXR may attenuate HSC activation and proliferation [71], and has been shown to suppress TGFβ and IL6 mRNA expression, thereby reducing fibrosis [72]. ATRA also interferes with the innate immune system and inhibits proinflammatory cytokine expression of macrophages [73]. ATRA also repressed CYP7a1 promoter activity [74], which was complemented by UDCA [75], indicating potential synergistic effects. In line, combination of ATRA with UDCA attenuated biliary fibrosis, bile duct proliferation, as well as hepatic inflammation in BDL rats and Mdr2 KO mice [72, 76]. However, in a clinical pilot study in PSC combination of ATRA and UDCA did not achieve the primary endpoint (reduction of alkaline phosphatase) despite reduction of the bile acid intermediate C4 and inflammation [77].

Since the enterohepatic circulation of BAs depends on active BA transport systems in liver and intestine, inhibition or stimulation of their transport function not only alters the BA load to the liver but also BA signaling. Key transporter targets include the apical sodium dependent bile salt transporter (ASBT/SLC10A2) for intestinal reabsorption, the sodium taurocholate cotransporting polypeptide (NTCP/SLC10A1) for hepatocellular reuptake of conjugated BAs and organic anion transporter (OATPs/SLCO/ SCL21) mainly for unconjugated BAs, and perhaps under certain circumstances also bile salt export pump (BSEP/ABCB11) as rate limiting canalicular BA transport system.

BA concentrations [18, 78]. Surgical interruption of the enterohepatic BA circulation has been a longstanding therapeutic strategy in PFIC [79]. In addition, increasing fecal BA concentrations and exposure of gut epithelial cells to BAs can stimulate potential beneficial colonic BA signaling such as TGR5‐­regulated GLP‐1 secretion from enteroendocrine L‐cells [80]. GLP‐1 has cholangioprotective effects protecting these cells from apoptosis and attenuating the reactive cholangiocyte phenotype [18]. Application of highly potent and selective ASBT inhibitors (Lopixipat and A4250) to Mdr2 KO mice increased fecal BA loss which cannot be compensated by increased endogenous BA synthesis. Thus, despite increased Cyp7a1 and repressed Fgf15 expression by ASBT inhibition, hepatic as well as biliary BA concentration and bile flow were reduced, leading to a beneficial BA/PL ratio, resulting in an improvement of liver and bile duct injury in this model of biliary toxicity [18, 78]. Furthermore, ASBT inhibitor treatment also reduced recruitment of Kupffer cells and neutrophils to the liver and promoted expansion of anti‐ inflammatory monocytes [18]. Similarly, application of the BA sequestrant colesevelam completely reversed liver and bile duct injury [78], further indicating that interruption of enterohepatic circulation of BAs may have therapeutic potential in attenuating cholestatic liver disease beyond their role in treatment of pruritus. Although both therapeutic options result in reduced hepatic BA uptake and thereby lowering hepatobiliary BA toxicity, their mechanism of action may differ profoundly since enhanced conversion of primary into secondary BAs (seen under colesevelam, but not under ASBT inhibitor treatment), may change intestinal BA signaling. GLP‐1 expression in enteroendocrine L‐cells was increased in colesevelam treated Mdr2 KO mice, pointing toward elevated TGR5 activity [78]. Enteroendocrine L‐cell FXR activity and TGR5 mediated GLP‐1 secretion appear to be inversely related [18]. Resin‐bound BAs may activate colonic TGR5 but not intracellular FXR [80], while non‐bound BAs signal intracellularly via FXR [18]. This may have critical implications on GLP‐1 secretion (stimulated by TGR5, but suppressed by FXR). In addition to the differences in GLP‐1 signaling between ASBT inhibitors and BA sequestrants hepatic and biliary BA composition also varies. Interestingly, colesevelam (but not ASBT inhibitor) resulted in a “hydrophilization” (thereby detoxification) of the biliary BA composition [18, 78]. Potential mechanistic differences between these two treatment strategies may deserve ­further clinical evaluation. Several ASBT inhibitors have recently been investigated in early phase clinical trials for treatment of pruritus in PBC and a range of pediatric cholestatic syndromes. While pharmacological target engagement such as reduced serum BA levels, reduced FGF‐19 and increased levels of C4 (as marker of BA synthesis) was mostly observed, the clinical effects on pruritus were rather variable (often not superior to placebo) and confounded by GI side‐effects such as diarrhea [81, 82].

ASBT inhibitors and BA sequestrants

NTCP blockers

Interruption of enterohepatic circulation of BAs  –  either ­pharmacologically with ASBT inhibitors or with BA sequestrants – may be advantageous not only for treating pruritus but also the underlying cholestatic liver disease since depleting BAs from the enterohepatic circulation reduces hepatic and biliary

Reduction of NTCP function may also reduce intrahepatic BA  levels, subsequently changing hepatobiliary BA excretion and, thereby, also hepatobiliary toxicity. BAs can induce ­inflammation‐based liver injury by stimulating the expression and secretion of inflammatory cytokines in cultured mouse

TRANSPORTER MODULATION



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

hepatocytes [83], thereby initiating hepatic neutrophil recruitment which contributes to progression of cholestatic liver disease [84]. Pharmacological inhibition of NTCP may reduce intracellular accumulation of BAs, resulting in amelioration of cholestasis and subsequent hepatocellular driven inflammatory response. Genetic absence of NTCP in mice and humans results in increased levels of unconjugated BAs in blood, but is well tolerated and no pruritus, fat malabsorption, or liver dysfunction were observed [18]. Myrcludex B, is a NTCP blocker, a small peptide inhibitor designed to prevent hepatitis B virus uptake which, like BAs, need NTCP to enter hepatocytes [18]. In DDC (3,5‐diethoxycarbonyl‐1,4‐dihydrocollidine) fed mice as model for sclerosing cholangitis NTCP inhibition improved serum biochemistry as well as hepatic inflammation and fibrosis. This protective effect may be based on increased biliary PL/BA ratio, making the bile less toxic. Consistent with this hypothesis, no beneficial effects of NTCP inhibition were seen in the Mdr2 KO mouse model lacking biliary PL secretion to begin with [85]. Based on these preclinical findings, combination therapy with Myrcludex B (to reduce hepatocellular BA uptake) and PPARα agonists (upregulation of Mdr2/MDR3) might be of particular interest to counteract biliary toxicity (at least in conditions with residual MDR2/3 functionality). Besides Myrcludex B, several drugs have been identified to prevent hepatitis B virus (HBV) infection via inhibition of NTCP, however, only some such as rosiglitazone, zafirlukast, and sulfasalazine inhibited simultaneously NTCP‐mediated BA uptake, making them attractive ­therapeutic approaches for cholestasis [86, 87].

Potential indirect benefits of BSEP blockage in mice Although in humans BSEP deficiency results in the very severe liver disease such as progressive familial intrahepatic cholestasis (PFIC) type 2, mice lacking BSEP are surprisingly protected from development of severe cholestatic liver injury [18]. Moreover, these mice are also resistant to acquired cholestasis induced either by bile duct ligation or DDC feeding [18]. Already at baseline, BSEP KO mice exhibit increased ­expression of enzymes Cyp2b10 and Cyp3a11, involved in BA hydroxylation, resulting in a very hydrophilic, less toxic BA pool composition, mainly consisting of tetrahydroxylated BAs [18]. This metabolic preconditioning with a  –  potential non‐toxic  – ­hydrophilic BA pool, may protect these animals from development of cholestatic liver injury. Thus, tetrahydroxylated BAs or activation of enzymes involved in formation of these BAs may have therapeutic potential to treat cholestatic liver disease. This hypothesis is in line with the fact that liver injury in bile duct‐ ligated mice can be attenuated with agonists for CAR and PXR (activating enzymes involved in BA hydroxylation) [43]. Under rare circumstances, direct transient blockage of BSEP export function could even directly protect bile ducts from BA‐induced injury. This might be relevant in the context of liver transplantation, where BSEP expression recovers faster than MDR3, resulting in less favorable biliary BA/PL ratio especially when MDR3 variants are present [18]. Altered bile ­composition after liver transplantation has been associated with the development of non‐anastomotic biliary strictures [88]. In  such exceptional conditions transient inhibition of BSEP

371

together with activation of MDR3 and bicarbonate secretion could theoretically help to protect bile ducts from harmful BA concentration during the immediate post‐transplant period [18], although this highly controversial hypothesis still needs to be tested. Interestingly, drugs such as cyclosporine A used for immune suppression inhibit BSEP function [89].

Chaperones Impaired bile secretory function in cholestasis does not only result from changes in gene expression/transcription but may also be due to post‐transcriptional changes with disturbed cell polarity and impaired targeting and function of transporters to the canalicular (and – to a lesser degree – the basolateral) membrane [5]. Therefore, stimulation of transporter targeting and/or function may also have important therapeutic implications. Chaperones are mainly used as therapies for hereditary transport defects, but could also be of interest in counteracting acquired cholestasis. They may directly interact with the variant of the ABC transporter proteins such as ABCB11/BSEP or ABCC2/MRP2, thereby decreasing its degradation and increasing its folding rate (correctors), or impact indirectly on protein synthesis via interaction with endogenous ER protein chaperones (potentiators) [90]. Interestingly, tauroUDCA has been shown to enhance the secretory capacity of cholestatic hepatocytes by stimulation of exocytosis and insertion of canalicular transport proteins into the apical membrane via PKC dependent mechanisms (Figure 30.2) [91]. In MDCKII cells, 4‐phenylbutyrate (4‐PBA) was shown to increase cell surface expression and transport capacity not only from BSEP variants but also from the wild‐type form [91]. Treatment of rats with 4‐PBA enhanced expression of wild‐type ABCB11 at the canalicular membrane [93]. In patients with ABCB11 mutations (PFIC‐2) 4‐PBA treatment increased canalicular expression of ABCB11 [94]. Further promising observations are made with ivacaftor, originally known as a cystic fibrosis transmembrane conductance regulator (CFTR) potentiator [95] which has been shown to increase phosphatidyl concentration in patients carrying certain mutations in the MDR3 gene [95]. Chaperones targeting ABC transporters such as ABCB11/BSEP and ABCB4/MDR2/3 ­ could therefore also be of particular interest to improve acquired cholestasis.

Tight junction sealers Hepatocyte tight junctions are crucial for maintaining the structure and function of bile canaliculi and osmotic gradients generated by active transport ultimately driving bile flow, while bile duct epithelium tight junctions are required to maintain epithelial cell polarity and prevent leakage of potentially toxic bile into the periportal space [96]. Loss of tight junction integrity may result in regurgitation of bile and bile duct injury [96]. Apart from several experimental models, disruption of tight junctions have also been observed in PBC, PSC, and biliary atresia [97]. Notably, restoration of abnormal ZO1 staining was observed in PBC patients treated by UDCA [97]. Sulfasalazine, an anti‐inflammatory agent which is extensively used in long‐ term therapy for chronic IBD increases expression of (intestinal) tight junctions in vivo [98] and protects against BA‐induced

372

THE LIVER:  TARGETING GUT–LIVER AXIS FOR CHOLESTATIC LIVER DISEASE TUDCA NTCP

Hepatocyte

TUDCA & cAMP

εPKC Wild type

Corrector

WT Transcription mRNA

Nucleus

Translation mRNA

TLC

Corrector ER

Mutant

Folding canaliculus

Misfolding

Mutant

Potentiator

Subapical endosomal recycling compartement

Golgi

Folded mutant

Figure 30.2  Therapeutic targeting of transporter trafficking and function at the post‐transcriptional level. After post‐translational modification in the Golgi apparatus, wild‐type (WT, blue boxes) and mutated forms (green boxes) of transporter are shuttled to a subapical recycling compartment. TauroUDCA, entering the hepatocytes via NTCP, and cAMP enhances the secretory capacity of cholestatic hepatocytes by stimulation of ­exocytosis and insertion of transport proteins into the apical membrane (green arrow) via εPKC‐dependent mechanisms. Conversely, transporter internalization (back to the subapical recycling compartment) is accelerated by taurolithocholic acid (TLC, red arrows). Correctors can partially rescue misfolding by improving folding at the ER and delaying the turnover to the plasma membrane with a presently poorly understood m ­ echanism. Although rescued mutants retain in part their function, they are conformationally unstable and might be eliminated by ubiquitination. Potentiators (e.g. Ivacaftor) may correct this phenotype.

apoptosis [99]. Therefore, this drug might be an attractive therapeutic for cholestatic liver diseases, especially for PSC patients also suffering from IBD and this is also currently being studied in ongoing clinical trials.

norUDCA – A CHOLEHEPATIC DRUG 24‐norursodeoxycholic acid (norUDCA) is a side‐chain shortened derivative of UDCA lacking a methyl group and conferring resistance to taurine and glycine conjugation, facilitating its penetration into cholangiocytes [100]. This results in cholehepatic shunting between hepatocytes and cholangiocytes (bypassing the classical enterohepatic circulation), and stimulates biliary bicarbonate secretion, resulting in alkalization of bile (Figure 30.3). NorUDCA is actively secreted in its anionic form (norUDCA−) into the canaliculus. NorUDC is converted to the protonated form (norUDCA) by accepting a hydrogen cation (generated by ionization of carbonic acid into hydrogen and bicarbonate) (Figure  30.3). Protonated norUDCA is then absorbed passively into cholangiocytes (where it gets deprotonated again) subsequently entering periductular capillary plexus and returns back to the sinusoid to be resecreted into the canaliculus in its anionic form (norUDCA−). Acceptance of the proton of carbonic acid by norUDCA− results in increased alkalization of the bile by generation of bicarbonate [101]. Alkalization of bile by bicarbonate enrichment may be a key protective mechanism (“bicarbonate umbrella”) against BA toxicity to cholangiocytes [3], preventing uncontrolled membrane permeation of

protonated (in particular glycine) conjugated BAs [2]. Taurine conjugated and bi‐norUDCA, (lacking two methyl groups) do not undergo cholehepatic shunt and exert less (Tauro-norUDCA) or no (bi‐norUDCA) effect on biliary bicarbonate secretion compared to unconjugated norUDCA, emphasizing the unique therapeutic properties of norUDCA in the Mdr2 KO mouse model of sclerosing cholangitis [2]. Anti‐inflammatory effects of norUDCA were seen on isolated cholangiocytes and macrophages, where it directly interferes with NFκB signaling [102]. In several mouse models (Mdr2 KO, BDL, NEMO KO, Schistosomiasis) norUDCA treatment reduced hepatic infiltration of inflammatory cells [102], indicating that norUDCA may even have direct anti‐inflammatory effects. Moreover, norUDCA also exerts antifibrotic and antiproliferative effects [103], including downregulation of cyclins and inhibition of mTOR signaling with induction of autophagy [18]. Notably a recent phase 2 study demonstrated improvement of cholestatic liver enzymes independent from the previous exposure and response to UDCA [104], encouraging an ongoing phase 3 study.

TARGETING GUT–LIVER AXIS FOR CHOLESTATIC LIVER DISEASE In healthy conditions an enormous amount of gut‐derived molecules reach the liver via the portal blood without triggering any inflammatory response. Under pathophysiological conditions, when the intestinal barrier function is defective or in case of imbalanced gut bacterial homeostasis, an altered composition of



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

BA detoxification

norUDC–

H2O + CO2 Anti-proliferative Anti-inflammatory Anti-fibrotic

norUDC– Cholehepatic shunting

BAs

Blood

Hepatocyte

norUDC–

Basolateral export

MRP3 CYPs SULTs, UGTs MRP4

norUDC–

373

Kidney

H2CO3 norUDC–

H+ + HCO3–

norUDCA + HCO3– AE2 HCO3– Bile

Cholangiocyte

HCO3–

Figure 30.3  Therapeutic mechanisms of norUDCA. The anionic form of norUDCA (norUDC−) is secreted into the canaliculus by unknown mechanisms. norUDC− accepts a hydrogen molecule deriving from ionization of carbonic acid (H2CO3). The acceptance of the proton of H2CO3 by norUDC− leaves behind a HCO3− in the ductular lumen. NorUDCA is passively absorbed into cholangiocytes and transported to the periductular capillary plexus, returning to the sinusoids and to being resecreted into the canaliculus, thereby completing a cholehepatic shunt, allowing “ductal targeting” of anti‐inflammatory, antifibrotic, and antiproliferative effects to injured bile ducts. Through the process of cholehepatic shunting norUDCA also increases biliary bicarbonate (HCO3−) concentration, thereby stabilizing the bicarbonate umbrella and facilitates BA detoxification and elimination via basolateral efflux pumps facilitating their subsequent renal excretion. BAs, bile acid.

gut‐derived products or even aberrant gut‐primed lymphocytes can reach the liver, where it may elicit or exacerbate hepatic inflammation. PBC and PSC are cholestatic autoimmune liver diseases in which the end results are caused by immune‐­ mediated bile duct injury. Both conditions are frequently associated with gut inflammation, PSC with IBD and PBC with coeliac disease [105], and dysbiosis [106]. This clinical observation has stimulated several intriguing pathogenic concepts in which gut commensals, pathogens, and intestinal antigens have been implicated in causing bile duct injury, in particular in PSC. One of the first theories connecting intestinal inflammation with liver injury is passive leakage of proinflammatory microbial components to the portal circulation and the possibility of an antigenic trigger of microbial origin. A study in rats with small bowel bacterial overgrowth described significant hepatic inflammation leading to fibrosis due to toxin translocation [107]. On the other side, obstructive cholestasis with absence of intraluminal BAs leads to bacterial overgrowth promoting bacterial translocation [61]. This observation could be explained by the well‐known direct antibacterial effects of BAs. Furthermore, it has been recently shown that BAs exert also indirect effects of microbiota via activation of intestinal FXR mediating the expression of anti‐inflammatory genes such as Ang1, iNOS, and IL18 [24]. The fact that BAs are able to shape microbiota communities by growth inhibition of certain bile sensitive bacteria [108] as well as by anti‐inflammatory effects via FXR activation open opportunities for BA‐based therapies in liver diseases associated with intestinal inflammation. In addition to intestinal inflammation‐related liver injury changes in microbiota (dysbiosis) have been described in a wide

range of liver diseases. Several studies using high‐throughput sequencing technology reported reduced diversity and significant shifts in the overall composition of the gut microbiota in PSC patients and appear to be distinct from IBD in the absence of PSC [61]. The role of dysbiosis in pathogenesis of cholangiopathies has also been emphasized by recent reports of a changed microbiome in PBC [106]. Several animal models demonstrated that enteric dysbiosis and/or administration of bacterial antigens can result in hepatobiliary inflammation with features resembling PSC [61]. Interestingly, a recent GWAS uncovered several new PSC risk loci related to immune regulation including fucosyltransferase‐2, known to influence microbiota and affecting susceptibility to microbial infection [61]. The importance of the presence of gut microbiota was reflected in aggravation of cholestatic liver phenotype and cholangiocyte senescence in germ free Mdr2 KO mice (as an established model of PSC) [61]. However, the key question remains whether the alterations in gut microbiota cause liver and bile duct disease or only appear to be secondary to cholestasis or when the disease has already progressed. Since reduced diversity dysbiosis is observed in multiple inflammatory and metabolic diseases it is tempting to assume that these aspects of observation are secondary to disease and disease‐related inflammation. Future evidence for the potential cause–effect relationship between the microbiome and liver disease may come from interventional studies targeting gut microbiota. Several absorbable and non‐absorbable antibiotics modulating microbiota have been tested in PSC [109] and revealed biochemical improvement, with vancomycin being one of the most promising agents [110]. In addition probiotics have been

374

THE LIVER:  CONCLUSIONS AND OUTLOOK

explored but so far not recalled convincing clinical results [111]. Finally, there has been an increasing interest in fecal microbiota transplantation which is currently being tested as a treatment option for PSC [112]. Given the key role of BA in gut microbiome homeostasis an effect of UDCA on the microbiota profile could be suspected. In an interventional trial in PBC [106] UDCA treatment partially relieved microbiome dysbiosis after six months [106]. Interestingly, it was also found that patients with PBC and inadequate response to UDCA had increased abundance of disease‐ associated genera compared with those with adequate response. In line it has also been shown that UDCA attenuated disease and normalized microbiota changes in mouse models of colitis [113]. Homing of gut primed T cells has been implicated in the pathogenesis of PSC [61]. Effector T cells activated during colitis express receptors such as α4β7 integrin for MADCAM and CCR9 for CCL25, facilitating homing to the gut and from the gut to the liver [61]. Vedolizumab, a gut‐specific α4β7 integrin‐neutralizing monoclonal antibody and CCX282‐B, a small molecule inhibiting CCR9, have been developed for treatment of IBD. Activation of VAP‐1 stimulates hepatic expression of MADCAM‐1 and CCL25, subsequently initiating the recruitment of mucosal T cells to the liver [61]. Recently, serum levels of VAP‐1 have been shown to correlate with severity of disease in PSC patients [61]. Therefore, the drugs that may regulate migration of activated gut mucosal lymphocytes to the liver appear to be an attractive option for treating PSC. Several small studies that tested anti‐TNFα agents on reducing biliary inflammation have been disappointing [114]. A retrospective study in IBD patients with PSC, who have been treated with vedolizumab, did not have any beneficial effect on liver enzymes [114]. Given its role in mediating α4β7/MADCAM‐1 interactions VAP‐1 inhibition may also represent a therapeutic target for counteracting cholestatic liver disease [61]. VAP‐1 antibody BTT1023 is currently being clinically tested in PSC patients [115].

rationale (indicated by association of PBC and genetic variants in IL12A and IL12RB2 loci [117]) and good efficacy in other diseases such as psoriasis in Crohn’s disease [118]. In small studies including PBC patients with an incomplete response to UDCA, selective B cell depletion with rituximab had limited biochemical efficacy although a significant decrease of autoantibody production was observed [119]. In addition, since fatigue in PBC has been thought to be driven by muscle bioenergetics, abnormality related to AMA was reduced by rituximab treatment; rituximab was also tested for fatigue in PBC, but no effect was observed [120]. Moreover, other biologicals such as abatacept (chimeric CTLA4 protein), ­ FFP104 (anti‐CD40 antibody), NI0801 (anti‐Cxcl10 antibody), E6011 (anti‐Cxcl1 antibody), and Etrasimod (S1P receptor modulator) are tested in PBC. Cenicriviroc (CVC) is a novel antagonist for both C‐C chemokine receptors 2 and 5 (CCR2 and 5) that are expressed in many inflammatory cells including neutrophils, T cells, and monocytes [121]. Peribiliary‐recruited inflammatory cells may contribute to the pathogenesis of cholangiopathies. CVC treatment improved cholestatic liver injury in the Mdr2 KO mouse model of sclerosing cholangitis by reducing macrophage recruitment to the liver [122]. Combining treatments that reduce BA pool size with the one that blocks inflammation may have superior effects in the treatment of cholestasis. Combining CVC and ATRA in cholestatic models (BDL rats and Mdr2 KO mouse) potentiated the effect of monotherapies, indicating a multitargeted therapy as an important paradigm for treating cholestatic liver injury [123]. CVC is currently being tested in patients with PSC [124]. Among the emerging antifibrotic inhibitors αVβ6 and lysyl oxidase homolog 2 (LOXL2) are attractive targets. αVβ6 is expressed during epithelial repair in cholestatic hepatocytes and cholangiocytes and has a critical role in activation of TGFb1. Protective effects of αVβ6 inhibition (STX100) were observed in the Mdr2 KO mouse model of sclerosing cholangitis as well as in bile duct ligated animals. The humanized monoclonal anti‐αVβ6 antibody STX100 is under clinical investigation for TARGETING LIVER INFLAMMATION idiopathic pulmonary fibrosis and chronic allograft nephropathy AND FIBROSIS [125]. LOXL2, promoting collagen and elastin cross‐linking [126], as well as cholangiocellular tight junction permeability Apart from their detergent and proapoptotic properties, BAs [127], has been shown to be critical for development of fibrosis. cause liver injury in cholestasis by stimulating cytokine‐­ However, despite promising preclinical findings the clinical mediated inflammatory response and fibrosis [83]. In this results in simtuzumab (LOXL2 inhibitor) treated PSC patients ­process increased hepatocellular BA levels initiated the event by were negative [128]. releasing proinflammatory cytokines (e.g. by activating Egr‐1) subsequently attracting inflammatory cells which then lead to tissue injury. Conclusively, reduction of BA load to the liver and/or inflammatory response by different treatment strategies CONCLUSIONS AND OUTLOOK would diminish cholestatic liver injury. Since PSC and PBC are considered to be immune‐mediated Our progress in understanding the molecular pathophysiology bile duct diseases, several classic immunosuppressive and newer of bile formation and cholestasis has led to the development immunomodulatory approaches have been tested over time with of  novel therapeutic options as alternative and second‐line rather disappointing results in regard to efficacy and/or safety ­ therapies to UDCA. While UDCA acts mainly at a post‐­ profile [116]. Newer approaches have targeted IL12 [117, 118] transcriptional level, some of these new alternative therapies and IL23 [118] which have both been implicated in the develop- target key regulatory transcription factors such as FXR, GR, and ment of PBC [117]. However, ustekinumab, a monoclonal PPARs. Other candidates include key regulators of adaptive BA ­antibody targeting the shared subunit IL12p40, showed rather metabolic and transport pathways such as the xenobiotic sensors disappointing effects in PBC [118], despite a good scientific CAR and PXR, the FXR downstream target FGF19, inhibitors



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

of BA uptake systems within the enterohepatic circulation (e.g. ASBT), or novel BA derivatives (e.g. norUDCA) undergoing cholehepatic shunting even bypassing the enterohepatic circulation. Direct immunological approaches with immunosuppressive or biological agents have so far been rather disappointing in treating cholestatic disorders such as PBC and PSC, but novel approaches target the gut–liver axis, proinflammatory monocytes, and integrin‐signaling in fibrosis. Moreover, several of the novel alternative therapeutic approaches targeting BA transport and metabolism exert not only anticholestatic but also profound anti‐inflammatory as well as immune‐modulatory actions which could be key in treating immune‐mediated cholangiopathies such as PBC/PSC and inflammatory/immune responses secondary to cholestasis irrespective of their underlying etiology. Finally, the gut–liver axis offers additional opportunities for therapeutic modulation of the microbiome (e.g. in PSC and IBD). A major challenge will be to clinically test and apply these expanding therapeutic opportunities and their combinations in an individualized personalized medicine approach.

ACKNOWLEDGMENTS This work was supported by grants F3517‐B20 and I 2755 from the Austrian Science Foundation (to MT).

REFERENCES 1. Erlinger, S. What is cholestasis in 1985? J Hepatol, 1985;1:687–93. 2. Beuers, U., Trauner, M., Jansen, P., and Poupon, R. New paradigms in the treatment of hepatic cholestasis: from UDCA to FXR PXR and beyond. J Hepatol, 2015;62:S25–37. 3. Beuers, U., Hohenester, S., de Buy Wenniger, L.J., Kremer, A.E., Jansen, P.L., and Elferink, R.P. The biliary HCO3(‐) umbrella: a unifying hypothesis on pathogenetic and therapeutic aspects of fibrosing cholangiopathies. Hepatol, 2010;52:1489–96. 4. European Association for the Study of the Liver. EASL clinical practice guidelines: the diagnosis and management of patients with primary biliary cholangitis. J Hepatol, 2017;67:145–72. 5. Trauner, M. and Boyer, J.L. Bile salt transporters: molecular characterization, function, and regulation. Physiol Rev, 2003;83:633–71. 6. Evans, R.M. and Mangelsdorf, D.J. Nuclear receptors, RXR and the big bang. Cell, 2014;157:255–66. 7. Halilbasic, E., Baghdasaryan, A., and Trauner, M. Nuclear receptors as drug targets in cholestatic liver diseases. Clin Liver Dis 2013;17:161–189. 8. Parks, D.J., Blanchard, S.G., Bledsoe, R.K. et al. Bile acids: natural ligands for an orphan nuclear receptor. Science, 1999;284:1365–8. 9. Makishima, M., Okamoto, A.Y., Repa, J.J. et al. Identification of a nuclear receptor for bile acids. Science, 1999;284:1362–5. 10. Wang, H., Chen, J., Hollister, K., Sowers, L.C., and Forman, B.M. Endogenous bile acids are ligands for the nuclear receptor FXR/BAR. Mol Cell, 1999;3:543–53. 11. Thomas, C., Pellicciari, R., Pruzanski, M., Auwerx, J., and Schoonjans, K. Targeting bile‐acid signalling for metabolic diseases. Nat Rev Drug Discov, 2008;7:678–93. 12. Massafra, V., Milona, A., Vos, H.R. et  al. Farnesoid X receptor activation promotes hepatic amino acid catabolism and ammonium clearance in mice. Gastroenterol, 2017;152:1462–76. 13. Huang, W., Ma, K., Zhang, J. et al. Nuclear receptor‐dependent bile acid signaling is required for normal liver regeneration. Science, 2006;312:233–6. 14. Holt, J.A., Luo, G., Billin, A.N. et al. Definition of a novel growth factor‐ dependent signal cascade for the suppression of bile acid biosynthesis. Genes Dev, 2003;17:1581–91.

375

15. Inagaki, T., Choi, M., Moschetta, A. et  al. Fibroblast growth factor 15 functions as an enterohepatic signal to regulate bile acid homeostasis. ­ Cell Metab, 2005;2:217–25. 16. Lu, T.T., Makishima, M., Repa, J.J. et al. Molecular basis for feedback regulation of bile acid synthesis by nuclear receptors. Mol Cell, 2000;6:507–15. 17. Byun, S., Kim, D.H., Ryerson, D. et al. Postprandial FGF19‐induced phosphorylation by Src is critical for FXR function in bile acid homeostasis. Nat Commun, 2018;9:2590. 18. Trauner, M., Fuchs, C.D., Halilbasic, E., and Paumgartner, G. New therapeutic concepts in bile acid transport and signaling for management of cholestasis. Hepatol, 2017;65:1393–404. 19. D’Aldebert, E., Biyeyeme Bi Mve, M.J., Mergey, M. et al. Bile salts control the antimicrobial peptide cathelicidin through nuclear receptors in the human biliary epithelium. Gastroenterol, 2009;136:1435–43. 20. Fickert, P., Fuchsbichler, A., Moustafa, T. et al. Farnesoid X receptor critically determines the fibrotic response in mice but is expressed to a low extent in human hepatic stellate cells and periductal myofibroblasts. Am J Pathol, 2009;175:2392–405. 21. Schwabl, P., Hambruch, E., Seeland, B.A. et al. The FXR agonist PX20606 ameliorates portal hypertension by targeting vascular remodelling and ­sinusoidal dysfunction. J Hepatol, 2017;66:724–33. 22. Hao, H., Cao, L., Jiang, C. et  al. Farnesoid X receptor regulation of the NLRP3 inflammasome underlies cholestasis‐associated sepsis. Cell Metab, 2017;25:856–67. 23. Guo, C., Xie, S., Chi, Z. et al. bile acids control inflammation and metabolic disorder through inhibition of NLRP3 inflammasome. Immunity, 2016;45:802–16. 24. Inagaki, T., Moschetta, A., Lee, Y.K. et  al. Regulation of antibacterial defense in the small intestine by the nuclear bile acid receptor. Proc Natl Acad Sci USA, 2006;103:3920–5. 25. Verbeke, L., Farre, R., Verbinnen, B. et al. The FXR agonist obeticholic acid prevents gut barrier dysfunction and bacterial translocation in cholestatic rats. Am J Pathol, 2015;185:409–19. 26. Ding, L., Yang, L., Wang, Z., and Huang, W. Bile acid nuclear receptor FXR and digestive system diseases. Acta Pharm Sin B, 2015;5:135–44. 27. Liu, Y., Binz, J., Numerick, M.J. et al. Hepatoprotection by the farnesoid X receptor agonist GW4064 in rat models of intra‐ and extrahepatic cholestasis. J Clin Invest, 2003;112:1678–87. 28. Kuzu, M.A., Kale, I.T., Col, C., Tekeli, A., Tanik, A., and Koksoy, C. Obstructive jaundice promotes bacterial translocation in humans. Hepatogastroenterology, 1999;46:2159–64. 29. Sinal, C.J., Tohkin, M., Miyata, M., Ward, J.M., Lambert, G., and Gonzalez, F.J., Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell, 2000;102:731–44. 30. Wagner, M., Fickert, P., Zollner, G. et  al. Role of farnesoid X receptor in determining hepatic ABC transporter expression and liver injury in bile duct‐ ligated mice. Gastroenterol, 2003;125:825–38. 31. Gomez‐Ospina, N., Potter, C.J., Xiao, R. et al. Mutations in the nuclear bile acid receptor FXR cause progressive familial intrahepatic cholestasis. Nat Commun, 2016;7:10713. 32. Nevens, F., Andreone, P., Mazzella, G. et  al. A placebo‐controlled trial of  obeticholic acid in primary biliary cholangitis. N Engl J Med, 2016;375:631–43. 33. Hirschfield, G.M., Mason, A., Luketic, V. et al. Efficacy of obeticholic acid in patients with primary biliary cholangitis and inadequate response to ­ursodeoxycholic acid. Gastroenterol, 2015;148:751–61. 34. Kris, V., Kowdley, C.L.B., Levy, C. et al. The AESOP trial: a randomized, double‐blind, placebo‐controlled, phase 2 study of obeticholic acid in patients with primary sclerosing cholangitis. Hepatol, 2017;66:1254A. 35. Papazyan, R., Liu, X., Liu, J. et al. FXR activation by obeticholic acid or nonsteroidal agonists induces a human‐like lipoprotein cholesterol change in mice with humanized chimeric liver. J Lipid Res, 2018;59:982–93. 36. Reich, M., Spomer, L., Hoehne, J. et al. Expression of the bile acid receptor TGR5 in livers of PSC patients and Mdr2−/− (Abcb4−/−) mice. J Hepatol, 2017;66:S555. 37. Reich, M., Klindt, C., Deutschmann, K., Spomer, L., Haussinger, D., and Keitel, V. Role of the G protein‐coupled bile acid receptor TGR5 in liver damage. Dig Dis, 2017;35:235–40. 38. Drafahl, K.A., McAndrew, C.W., Meyer, A.N., Haas, M., and Donoghue, D.J. The receptor tyrosine kinase FGFR4 negatively regulates NF‐kappaB signaling. PLoS One, 2010;5:e14412.

376

THE LIVER:  REFERENCES

39. Modica, S., Petruzzelli, M., Bellafante, E. et  al. Selective activation of nuclear bile acid receptor FXR in the intestine protects mice against cholestasis. Gastroenterol, 2012;142:355–65. 40. Hirschfield, G.M., Chazouilleres, O., Drenth, J.P. et al. Effect of NGM282, a FGF19 analogue, in primary sclerosing cholangitis: a multicentre, randomized, double‐blind, placebo‐controlled phase 2 trial. J Hepatol, 2019;70(3):483–93. 41. Zweers, S.J., Booij, K.A., Komuta, M. et  al. The human gallbladder secretes fibroblast growth factor 19 into bile: towards defining the role of fibroblast growth factor 19 in the enterobiliary tract. Hepatology, 2012;55:575–83. 42. Varga, T., Czimmerer, Z., and Nagy, L. PPARs are a unique set of fatty acid regulated transcription factors controlling both lipid metabolism and inflammation. Biochim Biophys Acta, 2011;1812:1007–22. 43. Zollner, G. and Trauner, M. Nuclear receptors as therapeutic targets in cholestatic liver diseases. Br J Pharmacol, 2009;156:7–27. 44. Honda, A., Ikegami, T., Nakamuta, M. et  al. Anticholestatic effects of ­bezafibrate in patients with primary biliary cholangitis treated with ursodeoxycholic acid. Hepatology, 2013;57:1931–41. 45. Delerive, P., Gervois, P., Fruchart, J.C., and Staels, B. Induction of IkappaBalpha expression as a mechanism contributing to the anti‐inflammatory activities of peroxisome proliferator‐activated receptor‐alpha activators. J Biol Chem, 2000;275:36703–7. 46. Gervois, P., Kleemann, R., Pilon, A. et  al. Global suppression of IL‐6‐ induced acute phase response gene expression after chronic in vivo treatment with the peroxisome proliferator‐activated receptor‐alpha activator fenofibrate. J Biol Chem, 2004;279:16154–60. 47. Ricote, M. and Glass, C.K. PPARs and molecular mechanisms of transrepression. Biochim Biophys Acta, 2007;1771:926–35. 48. Ip, E., Farrell, G.C., Robertson, G., Hall, P., Kirsch, R., and Leclercq, I. Central role of PPARalpha‐dependent hepatic lipid turnover in dietary steatohepatitis in mice. Hepatology, 2003;38:123–32. 49. Marra, F., Efsen, E., Romanelli, R.G. et al. Ligands of peroxisome proliferator‐activated receptor gamma modulate profibrogenic and proinflammatory actions in hepatic stellate cells. Gastroenterol, 2000;119:466–78. 50. Iwaisako, K., Haimerl, M., Paik, Y.H. et al. Protection from liver fibrosis by a peroxisome proliferator‐activated receptor delta agonist. Proc Natl Acad Sci USA, 2012;109:E1369–76. 51. Nakamuta, M., Fujino, T., Yada, R. et al. Therapeutic effect of bezafibrate against biliary damage: a study of phospholipid secretion via the PPARalpha‐ MDR3 pathway. Int J Clin Pharmacol Ther, 2010;48:22–8. 52. Ghonem, N.S., Assis, D.N., and Boyer, J.L. Fibrates and cholestasis. Hepatology, 2015;62:635–43. 53. Corpechot, C., Chazouilleres, O., Rousseau, A. et al. A placebo‐controlled trial of bezafibrate in primary biliary cholangitis. N Engl J Med, 2018;378:2171–81. 54. Lindor, K.D., Bowlus, C.L., Boyer, J., Levy, C., and Mayo, M. Primary biliary cholangitis: 2018 practice guidance from the American Association for the Study of Liver Diseases. Hepatology, 2019;69(1):394–419. 55. Jones, D., Boudes, P.F., Swain, M.G. et al. Seladelpar (MBX‐8025), a selective PPAR‐delta agonist, in patients with primary biliary cholangitis with an inadequate response to ursodeoxycholic acid: a double‐blind, randomised, placebo‐controlled, phase 2, proof‐of‐concept study. Lancet Gastroenterol Hepatol, 2017;2:716–26. 56. Sookoian, S. and Pirola, C.J. Elafibranor for the treatment of NAFLD: one pill, two molecular targets and multiple effects in a complex phenotype. Ann Hepatol, 2016;15:604–9. 57. Arenas, F., Hervias, I., Uriz, M., Joplin, R., Prieto, J., and Medina, J.F. Combination of ursodeoxycholic acid and glucocorticoids upregulates the  AE2 alternate promoter in human liver cells. J Clin Invest, 2008;118:695–709. 58. Zimmermann, C., van Waterschoot, R.A., Harmsen, S., Maier, A., Gutmann, H., and Schinkel, A.H. PXR‐mediated induction of human CYP3A4 and  mouse Cyp3a11 by the glucocorticoid budesonide. Eur J Pharm Sci, 2009;36:565–71. 59. Moore, L.B., Parks, D.J., Jones, S.A. et al. Orphan nuclear receptors constitutive androstane receptor and pregnane X receptor share xenobiotic and steroid ligands. J Biol Chem, 2000;275:15122–7. 60. Huang, W., Zhang, J., Chua, S.S., Qatanani, M., Han, Y., Granata, R., and Moore, D.D. Induction of bilirubin clearance by the constitutive androstane receptor (CAR). Proc Natl Acad Sci USA, 2003;100:4156–61.

61. Hov, J.R. and Karlsen, T.H. The microbiome in primary sclerosing cholangitis: current evidence and potential concepts. Semin Liver Dis, 2017;37:314–31. 62. Haughton, E.L., Tucker, S.J., Marek, C.J. et al. Pregnane X receptor activators inhibit human hepatic stellate cell transdifferentiation in vitro. Gastroenterology, 2006;131:194–209. 63. Bachs, L., Pares, A., Elena, M., Piera, C., and Rodes, J. Comparison of rifampicin with phenobarbitone for treatment of pruritus in biliary cirrhosis. Lancet, 1989;1:574–6. 64. Huang, W., Zhang, J., and Moore, D.D. A traditional herbal medicine enhances bilirubin clearance by activating the nuclear receptor CAR. J Clin Invest, 2004;113:137–43. 65. Yang, C.Y., Leung, P.S., Adamopoulos, I.E., and Gershwin, M.E. The implication of vitamin D and autoimmunity: a comprehensive review. Clin Rev Allergy Immunol, 2013;45:217–26. 66. Guo, G.Y., Shi, Y.Q., Wang, L. et  al. Serum vitamin D level is associated with disease severity and response to ursodeoxycholic acid in primary biliary cholangitis. Aliment Pharmacol Ther, 2015;42:221–30. 67. Levin, A.A., Sturzenbecker, L.J., Kazmer, S. et al. 9‐cis retinoic acid stereoisomer binds and activates the nuclear receptor RXR alpha. Nature, 1992;355:359–61. 68. Idres, N., Marill, J., Flexor, M.A., and Chabot, G.G. Activation of retinoic acid receptor‐dependent transcription by all‐trans‐retinoic acid metabolites and isomers. J Biol Chem, 2002;277:31491–8. 69. Tsuchida, T. and Friedman, S.L. Mechanisms of hepatic stellate cell activation. Nat Rev Gastroenterol Hepatol, 2017;14:397–411. 70. Ohata, M., Lin, M., Satre, M., and Tsukamoto, H. Diminished retinoic acid signaling in hepatic stellate cells in cholestatic liver fibrosis. Am J Physiol, 1997;272:G589–96. 71. Sharvit, E., Abramovitch, S., Reif, S., and Bruck, R. Amplified inhibition of stellate cell activation pathways by PPAR‐gamma, RAR and RXR agonists. PLoS One, 2013;8:e76541. 72. He, H., Mennone, A., Boyer, J.L., and Cai, S.Y. Combination of retinoic acid and ursodeoxycholic acid attenuates liver injury in bile duct‐ligated rats and human hepatic cells. Hepatol, 2011;53:548–57. 73. Wang, X., Allen, C., and Ballow, M. Retinoic acid enhances the production of IL‐10 while reducing the synthesis of IL‐12 and TNF‐alpha from LPS‐ stimulated monocytes/macrophages. J Clin Immunol, 2007;27:193–200. 74. Crestani, M., Sadeghpour, A., Stroup, D., Galli, G., and Chiang, J.Y. The opposing effects of retinoic acid and phorbol esters converge to a common response element in the promoter of the rat cholesterol 7 alpha‐hydroxylase gene (CYP7A). Biochem Biophys Res Commun, 1996;225:585–92. 75. Le Vee, M., Jouan, E., Stieger, B., and Fardel, O. Differential regulation of drug transporter expression by all‐trans retinoic acid in hepatoma HepaRG cells and human hepatocytes. Eur J Pharm Sci, 2013;48:767–74. 76. Cai, S.Y., Mennone, A., Soroka, C.J., and Boyer, J.L. All‐trans‐retinoic acid improves cholestasis in alpha‐naphthylisothiocyanate‐treated rats and Mdr2−/− mice. J Pharmacol Exp Ther, 2014;349:94–8. 77. Assis, D.N., Abdelghany, O., Cai, S.Y. et al. Combination therapy of all‐trans retinoic acid with ursodeoxycholic acid in patients with primary sclerosing cholangitis: a human pilot study. J Clin Gastroenterol, 2017;51:e11–16. 78. Fuchs, C.D., Paumgartner, G., Mlitz, V. et al. Colesevelam attenuates cholestatic liver and bile duct injury in Mdr2(−/−) mice by modulating composition, signalling and excretion of faecal bile acids. Gut, 2018;67:1683–91. 79. Whitington, P.F., Freese, D.K., Alonso, E.M., Schwarzenberg, S.J., and Sharp, H.L. Clinical and biochemical findings in progressive familial intrahepatic cholestasis. J Pediatr Gastroenterol Nutr, 1994;18:134–41. 80. Harach, T., Pols, T.W., Nomura, M., Maida, A., Watanabe, M., Auwerx, J., and Schoonjans, K. TGR5 potentiates GLP‐1 secretion in response to anionic exchange resins. Sci Rep, 2012;2:430. 81. Al‐Dury, S. and Marschall, H.U. Ileal bile acid transporter inhibition for the treatment of chronic constipation, cholestatic pruritus, and NASH. Front Pharmacol, 2018;9:931. 82. Shneider, B.L., Spino, C., Kamath, B.M. et  al. Placebo‐controlled randomized trial of an intestinal bile salt transport inhibitor for pruritus in Alagille syndrome. Hepatol Commun, 2018;2:1184–98. 83. Cai, S.Y., Ouyang, X., Chen, Y. et  al. Bile acids initiate cholestatic liver injury by triggering a hepatocyte‐specific inflammatory response. JCI Insight, 2017;2:e90780. 84. Saito, J.M. and Maher, J.J. Bile duct ligation in rats induces biliary expression of cytokine‐induced neutrophil chemoattractant. Gastroenterology, 2000;118:1157–68.



30:  Pathophysiologic Basis for Alternative Therapies for Cholestasis

85. Slijepcevic, D., Roscam Abbing, R.L.P. et al. Na(+) ‐taurocholate cotransporting polypeptide inhibition has hepatoprotective effects in cholestasis in mice. Hepatol, 2018. doi: 10.1002/hep.29888. 86. Donkers, J.M., Zehnder, B., van Westen, G.J.P. et al. Reduced hepatitis B and D viral entry using clinically applied drugs as novel inhibitors of the bile acid transporter NTCP. Sci Rep, 2017;7:15307. 87. Kaneko, M., Futamura, Y., Tsukuda, S. et al. Chemical array system, a platform to identify novel hepatitis B virus entry inhibitors targeting sodium taurocholate cotransporting polypeptide. Sci Rep, 2018;8:2769. 88. Buis, C.I., Geuken, E., Visser, D.S. et  al. Altered bile composition after liver transplantation is associated with the development of nonanastomotic biliary strictures. J Hepatol, 2009;50:69–79. 89. Byrne, J.A., Strautnieks, S.S., Mieli‐Vergani, G., Higgins, C.F., Linton, K.J., and Thompson, R.J. The human bile salt export pump: characterization of substrate specificity and identification of inhibitors. Gastroenterology, 2002;123:1649–58. 90. Martin, G.M., Chen, P.C., Devaraneni, P., and Shyng, S.L., Pharmacological rescue of trafficking‐impaired ATP‐sensitive potassium channels. Front Physiol, 2013;4:386. 91. Beuers, U., Bilzer, M., Chittattu, A. et al. Tauroursodeoxycholic acid inserts the apical conjugate export pump, Mrp2, into canalicular membranes and stimulates organic anion secretion by protein kinase C‐dependent mechanisms in cholestatic rat liver. Hepatology, 2001;33:1206–116. 92. Hayashi, H. and Sugiyama, Y. 4‐phenylbutyrate enhances the cell surface expression and the transport capacity of wild‐type and mutated bile salt export pumps. Hepatology, 2007;45:1506–16. 93. Hayashi, H. and Sugiyama, Y. Short‐chain ubiquitination is associated with the degradation rate of a cell‐surface‐resident bile salt export pump (BSEP/ ABCB11). Mol Pharmacol, 2009;75:143–50. 94. Gonzales, E., Grosse, B., Schuller, B. et al. Targeted pharmacotherapy in progressive familial intrahepatic cholestasis type 2: Evidence for improvement of cholestasis with 4‐phenylbutyrate. Hepatology, 2015;62:558–66. 95. Vauthier, V., Housset, C., and Falguieres, T. Targeted pharmacotherapies for defective ABC transporters. Biochem Pharmacol, 2017;136:1–11. 96. Rao, R.K. and Samak, G. Bile duct epithelial tight junctions and barrier function. Tissue Barriers, 2013;1:e25718. 97. Sakisaka, S., Kawaguchi, T., Taniguchi, E. et al. Alterations in tight junctions differ between primary biliary cholangitis and primary sclerosing cholangitis. Hepatology, 2001;33:1460–8. 98. Xu, B., Li, Y.L., Xu, M. et  al. Geniposide ameliorates TNBS‐induced experimental colitis in rats via reducing inflammatory cytokine release and restoring impaired intestinal barrier function. Acta Pharmacol Sin, 2017;38:688–98. 99. Rust, C., Bauchmuller, K., Bernt, C. et al. Sulfasalazine reduces bile acid induced apoptosis in human hepatoma cells and perfused rat livers. Gut, 2006;55:719–27. 100. Hofmann, A.F. and Hagey, L.R., Key discoveries in bile acid chemistry and biology and their clinical applications: history of the last eight decades. J Lipid Res, 2014;55:1553–95. 101. Yoon, Y.B., Hagey, L.R., Hofmann, A.F., Gurantz, D., Michelotti, E.L., and Steinbach, J.H. Effect of side‐chain shortening on the physiologic properties of bile acids: hepatic transport and effect on biliary secretion of 23‐nor‐ ursodeoxycholate in rodents. Gastroenterology, 1986;90:837–52. 102. Halilbasic, E., Steinacher, D., and Trauner, M. Nor‐ursodeoxycholic acid as a novel therapeutic approach for cholestatic and metabolic liver diseases. Dig Dis, 2017;35:288–92. 103. Fickert, P., Wagner, M., Marschall, H.U. et al. 24‐norUrsodeoxycholic acid is superior to ursodeoxycholic acid in the treatment of sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology, 2006;130:465–81. 104. Fickert, P., Hirschfield, G.M., Denk, G. et  al. norUrsodeoxycholic acid improves cholestasis in primary sclerosing cholangitis. J Hepatol, 2017;67:549–58. 105. Trivedi, P.J. and Adams, D.H. Mucosal immunity in liver autoimmunity: a comprehensive review. J Autoimmun, 2013;46:97–111. 106. Tang, R., Wei, Y., Li Y. et  al. Gut microbial profile is altered in primary biliary cholangitis and partially restored after UDCA therapy. Gut, 2018;67:534–41.

377

107. Lichtman, S.N., Sartor, R.B., Keku, J., and Schwab, J.H. Hepatic inflammation in rats with experimental small intestinal bacterial overgrowth. Gastroenterology, 1990;98:414–23. 108. Watanabe, M., Fukiya, S., and Yokota, A. Comprehensive evaluation of the bactericidal activities of free bile acids in the large intestine of humans and rodents. J Lipid Res, 2017;58:1143–52. 109. Elfaki, D.A. and Lindor, K.D. Antibiotics for the treatment of primary sclerosing cholangitis. Am J Ther, 2011;18:261–5. 110. Tabibian, J.H., Weeding, E., Jorgensen, R.A. et  al. Randomised clinical trial: vancomycin or metronidazole in patients with primary sclerosing cholangitis – a pilot study. Aliment Pharmacol Ther, 2013;37:604–12. 111. Shen, J., Zuo, Z.X., and Mao, A.P. Effect of probiotics on inducing remission and maintaining therapy in ulcerative colitis, Crohn’s disease, and pouchitis: meta‐analysis of randomized controlled trials. Inflamm Bowel Dis, 2014;20:21–35. 112. Ali, A.H., Carey, E.J., and Lindor, K.D. The microbiome and primary sclerosing cholangitis. Semin Liver Dis, 2016;36:340–8. 113. Van den Bossche, L., Hindryckx, P., Devisscher, L. et al. Ursodeoxycholic acid and its taurine‐ or glycine‐conjugated species reduce colitogenic dysbiosis and equally suppress experimental colitis in mice. Appl Environ Microbiol, 2017;83. 114. Tse, C.S., Loftus, E.V., Jr., Raffals, L.E., Gossard, A.A., and Lightner, A.L. Effects of vedolizumab, adalimumab and infliximab on biliary inflammation in individuals with primary sclerosing cholangitis and inflammatory bowel disease. Aliment Pharmacol Ther, 2018;48:190–5. 115. Arndtz, K., Corrigan, M., Rowe, A. et al. Investigating the safety and activity of the use of BTT1023 (Timolumab), in the treatment of patients with primary sclerosing cholangitis (BUTEO): a single‐arm, two‐stage, open‐label, multi‐centre, phase II clinical trial protocol. BMJ Open, 2017;7:e015081. 116. Mousa, H.S., Carbone, M., Malinverno, F., Ronca, V., Gershwin, M.E., and Invernizzi, P. Novel therapeutics for primary biliary cholangitis: toward a disease‐stage‐based approach. Autoimmun Rev, 2016;15:870–6. 117. Hirschfield, G.M., Liu, X., Xu, C. et al. Primary biliary cirrhosis associated with HLA, IL12A, and IL12RB2 variants. N Engl J Med, 2009;360:2544–55. 118. Mousa, H.S., Lleo, A., Invernizzi, P., Bowlus, C.L., and Gershwin, M.E. Advances in pharmacotherapy for primary biliary cholangitis. Expert Opin Pharmacother, 2015;16:633–43. 119. Myers, R.P., Swain, M.G., Lee, S.S., Shaheen, A.A., and Burak, K.W. B‐ cell depletion with rituximab in patients with primary biliary cholangitis refractory to ursodeoxycholic acid. Am J Gastroenterol, 2013;108:933–41. 120. Khanna, A., Jopson, L., Howel, D. et al. Rituximab is ineffective for treatment of fatigue in primary biliary cholangitis: a phase 2 randomized controlled trial. Hepatology, 2018. 121. Mack, M., Cihak, J., Simonis, C. et al. Expression and characterization of the chemokine receptors CCR2 and CCR5 in mice. J Immunol, 2001;166:4697–4704. 122. Guicciardi, M.E., Trussoni, C.E., Krishnan, A. et al. Macrophages contribute to the pathogenesis of sclerosing cholangitis in mice. J Hepatol, 2018;69:676–86. 123. Yu, D., Cai, S.Y., Mennone, A., Vig, P., and Boyer, J.L. Cenicriviroc, a cytokine receptor antagonist, potentiates all‐trans retinoic acid in reducing liver injury in cholestatic rodents. Liver Int, 2018;38:1128–38. 124. Bennett, L.D., Fox, J.M., and Signoret, N. Mechanisms regulating chemokine receptor activity. Immunology, 2011;134:246–56. 125. Dyson, J.K., Hirschfield, G.M., Adams, D.H. et al. Novel therapeutic targets in primary biliary cholangitis. Nat Rev Gastroenterol Hepatol, 2015;12:147–58. 126. Barry‐Hamilton, V., Spangler, R., Marshall, D. et al. Allosteric inhibition of lysyl oxidase‐like‐2 impedes the development of a pathologic microenvironment. Nat Med, 2010;16:1009–17. 127. Pollheimer, M.J., Racedo, S., Mikels‐Vigdal, A. et al. Lysyl oxidase‐like protein 2 (LOXL2) modulates barrier function in cholangiocytes in cholestasis. J Hepatol, 2018;69:368–77. 128. Muir, A.J., Levy, C., Janssen, H.L.A. et al. Simtuzumab for primary sclerosing cholangitis: phase 2 study results with insights on the natural history of the disease. Hepatology, 2019;69(2):684–98.

31

Adaptive Regulation of Hepatocyte Transporters in Cholestasis James L. Boyer Department of Internal Medicine and Liver Center, Yale University School of Medicine, New Haven, CT, USA

INTRODUCTION Bile secretion is a unique and vital function of the liver. It delivers bile acids to the intestine for the solubilization and absorption of dietary lipids and serves as an excretory pathway for endogenous metabolic products including cholesterol, bilirubin, and porphyrins as well as foreign compounds and xenobiotics. Many liver disorders can impair this secretion, resulting in retention of bile and the syndrome of cholestasis. In turn, cholestatic liver injury results in adaptive responses in the expression of membrane transport proteins and hepatic enzymes that synthesize and excrete bile acids, bilirubin, and other solutes in an attempt to reduce the hepatic accumulation of these products and thus attenuate liver tissue injury. Ultimately these adaptive responses are not sufficient and chronic cholestatic injury leads to the development of biliary cirrhosis and progressive liver failure. Developing therapeutic strategies that might enhance these intrinsic adaptations is an unmet medical need. Here we review current knowledge of the mechanism of these adaptive responses. The major membrane transport systems that result in the formation of bile are illustrated in Figure 31.1. Their gene designations, membrane locations, and primary functions are briefly summarized in Table 31.1. Hepatic enzymes that are involved in the regulation of bile salt synthesis and metabolism are summarized in Table  31.2. A more detailed set of references may be found in Chapter  23 of the earlier 5th edition. My apologies to authors whose articles were not able to be included in this 6th edition.

Steps in bile formation The overall process of hepatic bile secretion can be divided functionally into four different phases: Phase 0: Hepatic uptake mechanisms, Phase I: Hydroxylating enzyme reactions, Phase

II: Conjugating enzyme reactions, for example, sulfation, glucuronidation, and amidation, and Phase III: Hepatic efflux mechanisms. Each of these steps is regulated by a series of transporters (Phase 0 and III) or enzyme reactions (Phase I and II). This chapter focuses on the molecular adaptations in these hepatocyte membrane transporters and enzymatic reactions that serve to attenuate cholestatic liver injury and its systemic effects. Cellular events that impair signal transduction pathways, cytoskeleton structures, tight junction and gap junctional proteins, and the targeting of intracellular vesicles to the apical canalicular domain that maintains the secretory polarity of the hepatocyte all contribute to the cholestatic phenotype [1] but are not considered here. Much information has been obtained from animal models of cholestasis and from discoveries of the genetic basis of several hereditary and acquired human cholestatic disorders. Studies in human hepatocyte cell lines and cholestatic liver disorders, as well as mouse knockout models have contributed significantly to advances in this field.

GENERAL OVERVIEW – ACQUIRED DEFECTS IN BILE TRANSPORT PROTEINS The molecular characterization of hepatic transport systems led to analyses of their response to cholestatic liver injury resulting in the following paradigm: Determinants of bile formation adapt to cholestatic liver injury in order to minimize hepatic injury by: (i) diminishing the hepatic uptake of bile salts and other solutes, (ii) reducing bile acid synthesis, (iii) augmenting bile acid detoxification mechanisms, and (iv) upregulating

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

BSEP NTCP

Na+ BA– BA–,

BA

FIC1

PS

379

MRP4 BA-S–

MRP2

OA–

GSH HCO3–

MDR1

OA– GSH

OATPs

OC+

BCRP

OC+ PL

OA–S

OCT1

OC+

H+,Na+

OATs

Bil-G–

MDR3 OSTα-β

Chol OA–

MRP3

ABCG5/8

OC+ MATE-1

?

Figure 31.1  Membrane transporters that determine the uptake and excretion of bile acids and other organic solutes in hepatocytes (see tables for full terminology). Table 31.1  Nomenclature, location, and function of the major hepatocyte plasma membrane bile acid and organic solute transporters involved in bile secretion (Phase 0 and III)

a

Name

Abbreviation (gene)

Phase

Location

Function

Sodium‐taurocholate cotransporter polypeptide Organic anion‐transporting polypeptides

NTCP (SLC10A1)

0

OATPs (SLCO1B1,1B3 and 3A1)

0 and III

Basolateral membrane of hepatocytes Basolateral membrane of hepatocytes

Organic solute transporter alpha/beta

OSTα/β (SLC51A and SLC51B)

III

Organic cation transporter‐1 Organic anion transporter 2 Multidrug‐resistance protein-1 (P‐glycoprotein)a Multidrug‐resistance protein-3 (phospholipid transporter)a Bile salt export pumpa

OCT‐1 (SLC22A1)

0

OAT‐2 (SLC22A7)

0

MDR1 (ABCB1)

III

MDR3 (ABCB4)

III

Basolateral membrane of hepatocytes, cholangiocytes, ileum, and proximal tubule of kidney Basolateral membrane of hepatocytes Basolateral membrane of hepatocytes Canalicular and cholangiocyte apical membrane Canalicular membrane

BSEP (ABCB11)

III

Canalicular membrane

Multidrug‐resistance‐ associated protein 2 (canalicular multispecific organic anion transporter)a Multidrug‐resistance‐ associated protein 3a

MRP2 (ABCC2)

III

Canalicular membrane

MRP3 (ABCC3)

III

Multidrug‐resistance‐ associated protein 4a

MRP4 (ABCC4)

III

Multidrug‐resistance‐ associated protein‐6a

MRP6 (ABCC6)

III

Basolateral membrane of hepatocytes and cholangiocytes Basolateral membrane of hepatocyte; apical membrane of proximal tubule of kidney Basolateral membrane of hepatocyte

Breast cancer resistance proteina

BRCP (ABCG2)

III

Canalicular membrane, proximal tubule of kidney

Sterolin‐1 and 2a

ABCG5/G8

III

Multidrug and toxin extrusion protein 1

MATE‐1 (SLC47A1)

III

Canalicular membrane and apical membrane of intestine Canalicular membrane and brush border of kidney

Primary carrier for conjugated bile salt uptake from portal blood Broad substrate carriers for sodium‐ independent uptake of bile salts, organic anions, and other amphipathic organic solutes from portal blood. OATP3A1 is induced in cholestasis Heteromeric solute carrier for facilitated transport of bile acids across basolateral membrane of ileum. Expression induced in liver in cholestasis Facilitates sodium‐independent hepatic uptake of small organic cations Facilitates sodium‐independent hepatic uptake of drugs and prostaglandins ATP‐dependent excretion of various organic cations, xenobiotics, and cytotoxins into bile; barrier function in cholangiocytes ATP‐dependent translocation of phosphatidylcholine from inner to outer leaflet of membrane bilayer (a floppase) ATP‐dependent bile salt transport into bile; stimulates bile salt‐dependent bile flow Mediates ATP‐dependent multispecific organic anion transport (e.g. bilirubin diglucuronide) into bile; contributes to bile salt‐independent bile flow by GSH transport Expression induced in cholestasis. Transports bilirubin and bile acid glucuronide conjugates Expression induced in cholestasis – transports sulfated bile acid conjugates and cyclic nucleotides ATP‐dependent transport of organic anions and small peptides. Mutations of MRP6 gene result in pseudoxanthoma elasticum ATP‐dependent multispecific drug transporter, particularly sulfate conjugates; protoporphyrins are endogenous substrates. Substrates overlap with MRP2 Heteromeric ATP dependent transporter for cholesterol and plant sterols (stilbestrol) Organic cation/H+ exchanger extrudes cationic xenobiotics

 These transporters are members of the ATP‐binding cassette superfamily.

380

THE LIVER: GENERAL OVERVIEW – ACQUIRED DEFECTS IN BILE TRANSPORT PROTEINS

Table 31.2  Hepatic enzymes involved in the regulation of bile salt synthesis and bile salt and bilirubin metabolism (Phase I and II) and their adaptive responses to cholestasis Name

Abbreviation (gene)

Cytochrome P450 7A1

(CYP7A1)

Cytochrome P450 8B1

(CYP8B1)

Cytochrome P450 27A1

(CYP27A1)

Cytochrome P450 3A4

Function

Adaptive response in cholestasisa Downregulation limits the synthesis of bile acids More favored pathway resulting in increased % of bile acids as cholic acid Unchanged

(CYP3A4)

Cholesterol 7α‐hydroxylase, the rate‐limiting step in bile acid synthesis from cholesterol Sterol 12α‐hydroxylase, the pathway to cholic acid synthesis. Controls ratio of cholic acid to chenodeoxycholic acid Sterol 27‐hydroxylase, the mitochondrial “acidic” alternative pathway of bile acid side‐chain oxidation. Favors chenodeoxycholic acid formation Mediates Phase I bile acid hydroxylation

Sulfotransferase

SULT2A1

Mediates Phase II conjugation with sulfate

Uridine glucuronyl transferase

(UGT1A1/ B4/B7)

Mediates Phase II conjugation with glucuronic acid

Bile acid CoA synthetase Bile acid CoA:amino acid N‐acyltransferase

BACS BAAT/Bat

Forms bile acid CoA‐thioesters, substrates for BAAT Mediates taurine and glycine conjugation of bile acids (amidation)

Unchanged or stimulated; may facilitate ability of bile acids to undergo Phase II conjugations Unchanged or stimulated; facilitates bile acid conjugation to substrates for MRP2 and MRP4 Unchanged or stimulated: mediates bile acid and bilirubin conjugation to substrates for MRP2 and MRP3 —a —a; reduced in sepsis

 The adaptive response in cholestasis is not known.

a

mechanisms that facilitate the export of bile salts and other toxic substances from hepatocytes [2]. Thus, transport proteins on the basolateral sinusoidal membrane which normally function to remove bile salts and other cholephiles selectively from portal blood are usually downregulated during cholestatic liver injury by both transcriptional and post‐transcriptional mechanisms [3, 4]. In contrast, some canalicular transport proteins, particularly the multidrug resistance protein (MDR) homologs, are either not severely impaired or may actually be upregulated; these include MDR1/Mdr1, BSEP/Bsep (bile salt export pump), and MDR3/Mdr2. These later findings suggest that the cholestatic hepatocyte attempts to maintain canalicular export function. Perhaps more importantly, transporters such as MRP3/ Mrp3, MRP4/Mrp4 (5–10) and OSTα‐OSTβ (the organic solute transporter) [11] that are expressed at the basolateral sinusoidal membrane at low levels in normal liver are substantially upregulated in cholestatic hepatocytes, facilitating the efflux of hydrophobic bile salts and other products back to the circulation where they may be cleared in part by the kidney [12]. Thus, cholestasis results in a partial reversal of bile secretory polarity. Less is known concerning molecular responses in cholangiocytes during cholestasis. However, bile duct proliferation is characteristic of most cholestatic disorders. The apical sodium‐ dependent bile salt transporter (ASBT), originally identified in  the ileum, is also expressed on the luminal membranes of cholangiocytes and the proximal tubules of the kidney and functions to remove bile salts from bile and glomerular filtrate, respectively [13]. Because the cholestatic hepatocyte continues to excrete bile salts, albeit at a reduced rate, ASBT may remove bile salts from an obstructed biliary tree. MRP3, MRP4, OSTα‐β, and OATP3A1 are also located on the blood side of the cholangiocyte and are upregulated in the cholestatic liver [5, 11, 14]. MRP3/Mrp3 has a high affinity for glucuronidated conjugates and may be the means by which bilirubin glucuronide is effluxed back into the blood in cholestatic liver. MRP4/Mrp4 may provide the same role for sulfated conjugates [15, 16].

Sulfated bile salt conjugates in particular are formed in the cholestatic human liver albeit to a lesser extent in mouse liver. OSTα‐OSTβ, which mirrors the tissue expression of ASBT, also plays a major role in bile acid efflux from the cholestatic hepatocyte in human liver [11]. Downregulation of Asbt in the renal proximal tubule results in diminished absorption of bile acids from the glomerular filtrate, thereby facilitating bile salts excretion in the urine [17]. Mrp2 and Mrp4 (expressed on the apical luminal membrane of the proximal tubule) may also facilitate the tubular excretion of bile acid and other divalent conjugates. Less is known about the response of intestinal bile acid transport proteins in cholestasis, although ASBT is generally downregulated on the luminal brush border of the ileum, reducing the return of bile acids to the liver in the enterohepatic circulation [18], and intestinal MRP2/Mrp2 is diminished in both humans and rats with obstructive cholestasis [19]. Adaptations also occur in hepatic enzymes that determine the synthesis and metabolism of bile acids. In general, these changes result in a smaller and less hydrophobic bile salt pool, although significant species differences exist in composition and response. For example, in cholestatic mice, bile acid pools are enriched in the highly hydrophilic bile acid, muricholic acid. Enzymes that regulate bile acid and bilirubin conjugation are also significantly upregulated in cholestasis. The major enzymes and their function and adaptive responses are summarized in Table 31.1. Thus, a complex pattern of adaptive responses in transporter expression and metabolism occurs in hepatocytes, cholangiocytes, kidney, and intestine, that attempt to mitigate tissue damage from the retention of bile salts and other toxic substrates. Much of this adaptive regulatory response is mediated by transcriptional events that are regulated by liver‐enriched transcription factors (hepatocyte nuclear factors [HNFs]) and nuclear receptors (NRs) [4, 2]. Members of the HNF family tend to regulate constitutive expression of genes whereas NRs induce adaptive responses in gene expression and are activated by specific ligands. In cholestasis, these ligands are bile acids,



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

bilirubin, oxysterols, and drugs or xenobiotics [21]. Chief among these are bile acids which regulate gene expression primarily via the farnesoid X receptor (FXR). When bile acids accumulate in the cholestatic liver, the expression of genes that are regulated by FXR are usually enhanced. Other NRs that play an important role in regulating gene expression are the pregnane X receptor (PXR), the constitutive androstane receptor (CAR), the liver X receptor (LXR) [22], the vitamin D receptor (VDR), and peroxisome proliferator‐activated receptor alpha (PPARα) [23]. Knockout animals for FXR, PXR, CAR, and LXR are each more susceptible to cholestatic injury than their respective wild‐ type [4, 22–24]. Other ligand stimulated NRs involved in the adaptive response to cholestasis include the retinoic X receptor alpha (RXRα) and the glucocorticoid receptor (GR). RXRα plays a particularly important role in these responses since it is the obligate heterodimeric partner for the class II low‐affinity NRs that include FXR, PXR, LXR, CAR, and PPARα. Cholestasis also can affect the expression of NRs. Studies of inflammatory and fibrotic models of cholestasis indicate that RXRα expression is reduced and may be translocated out of the

381

nucleus into the cytoplasm, where it may be degraded, thus influencing the expression of genes dependent on its NR partners [25]. Reductions in mRNA levels of FXR, and its target gene SHP (short heterodimeric partner), have also been reported in patients with cholestasis from PFIC1 and 2 (progressive familial intrahepatic cholestasis type 1 and type 2) and biliary atresia [26, 27]. Reductions in PXR and CAR have also been reported in late stages of biliary atresia [27]. PPAR agonists (fibrates) stimulate MDR3, inhibit bile acid synthesis, have anti‐ inflammatory effects [28], and improve liver function in patients with primary biliary cholangitis (PBC) [29, 30]. Several other regulators of transcription for which no specific ligands are known, but that also influence the expression of bile transporters and enzymes, include the small heterodimer partner‐1 (SHP‐1), liver receptor homologue‐1 (LRH‐1), and HNF‐4α. HNF‐4α is a primary determinant of HNF‐1α expression, which determines the transcription of genes that encode for cytochrome P4507A (CYP7A), NTCP, and OATPs (organic anion transporting polypeptides) as examples  [31]. Table  31.3 lists these NRs and their activating ligands and key transcription

Table 31.3  Nuclear receptors, important ligands, transcription factors, and target genes regulating the adaptive response in cholestasis (? = from rodents only)a Nuclear receptors

Ligands

Major target genes

Anticipated function in cholestasis

FXR (farnesoid X receptor) – NR1H4

Bile acids, GW4064, 6α‐ethyl‐CDCA, fexaramines

BSEP, OATP1B3, MRP2, MDR3, OSTα/β, CYP3A4, UGT2B4 and B7, SULT2A1, BACS and BAAT, SHP, I‐BABP, FGF‐15/19, PXR

PXR (pregnane X receptor) – NRII

Xenobiotics, LCA, ursodeoxycholic acid, rifampicin Xenobiotics, bilirubin, phenobarbital, Yin Chin, TCPOBOP Oxysterols (metabolites of cholesterol) All‐trans‐retinoic acid

CYP3A4, OATP1B1?, SULT2A1, UGT1A1 MRP2, MRP3, MDR1, GST CYP2B, CYP3A4, OATP1B1, MRP2, MRP4, UGT1A1, SULT2A1, GSTs CYP7A1, CYP8B, UGT1A3 ABC5/8?, SHP, OSTα/β?, LRH‐1 NTCP?, MRP2?, ASBT, MRP3

VDR (vitamin D receptor) – NRIII GR (glucocorticoid receptor)

Vitamin D, lithocholic acid Corticosteroids

CYP3A4, SULTs?, MRP3?

RXRα (retinoic X receptor) – NR2B1 Others SHP‐1 small heterodimer partner) – NR0B2

9‐cis‐Retinoic acid

Obligate heterodimeric partner for all class II NRs

Inhibition of bile acid synthesis and ileal bile acid uptake via SHP and FGF‐15/19. Upregulation of Phase I and II bile acid hydroxylation and conjugation, induction of bile acid canalicular and basolateral transporters Induction of Phase I and II bile acid and bilirubin conjugation reactions and bile acid and bilirubin alternative export pumps Induction of Phase I and II bile acid and bilirubin conjugation reactions and bile acid and bilirubin alternative export pumps Inhibits bile acid synthesis while stimulating Phase II and III steps ? Induction of RXR partners via metabolism to 9‐cis‐retinoic acid; loss in cholestasis upregulates MRP3 Induction of Phase I and II bile acid hydroxylation ? Induction of AE2 (together with ursodeoxycholate) Reduction of RXR in cholestasis has variable effects depending on the gene

None; Upregulated by FXR

Interacts with LRH‐1 to suppress CYP7A1, CYP8B1, CYP27A1. Also inhibits NTCP?, and ASBT?, OATP1B1 CYP7A1, CYP8B1. MRP3, ASBT?

CAR (constitutive androstane receptor) – NR3 LXRα (liver X receptor) – NR1H3 RARα (retinoic acid receptor) – NR1B1

LRH‐1(FTF) – liver receptor homologue‐1; fetal transcription factor – NR5A2 HNF‐4α (hepatocyte nuclear factor‐4) – NR2A1 PPARα (peroxisome proliferator‐activated receptor‐alpha) – NR1C1

? Blocked by SHP None Fatty acids, eicosanoids, fibrates, statins

NTCP, ASBT, MRP2, BSEP, AE2

HNF‐1, CYP7A1, CYP8B1, CYP27A1 OATP1B1, NTCP? BACS and BAAT SULT2A1, UGT2B4, MDR2/3 CYP7A

Suppresses bile acid synthesis, and hepatic and ileal bile acid uptake Inhibition of bile acid synthesis (CYP8B1) and ileal uptake (ASBT?) via SHP; upregulates MRP3 Inhibition by SHP via FXR of bile acid synthesis and bile acid uptake Induces Phase II bile acid and bilirubin conjugation reactions. Inhibits bile acid synthesis

 Transporter and enzyme gene abbreviations include: ABCG5 and G8 (sterolin 1 and 2); ASBT (apical sodium‐dependent bile acid transporter, SLC10A2); BAAT (bile acid CoA:amino acid N‐acyltransferase); BACS (bile acid CoA‐synthase); BSEP (bile salt export pump, ABCB11); CYP7A (cytochrome P450 7A); CYP8B (cytochrome P450 8B); CYP27A1 (cytochrome P450 27A1); FGF‐15/19 (fibroblast growth factor 15 or 19); GSTs (glutathione S‐transferases); I‐BABP (ileal bile acid binding protein); MDR1 or 2/3 (multidrug resistance protein 1 or 2/3, ABCB1or 4); MRP2, 3, and 4 (multidrug resistance associated protein 2, 3, or 4, ABCC2,3, or 4); NTCP (sodium taurocholate cotransporting polypeptide, SLC10A1); OATP‐1B1, 1B3 (organic anion transporting polypeptide C or 8, SLC01B1 or 1B3); OSTα‐OSTβ (organic solute transport protein alpha and beta); SHP‐1 (small heterodimer partner 1); SULTs (sulfotransferases); UGTs (uridine 5′‐glucuronosyl‐transferases).

a

382

THE LIVER: GENERAL OVERVIEW – ACQUIRED DEFECTS IN BILE TRANSPORT PROTEINS

factors and illustrates their major target genes and also summarizes the anticipated regulatory responses in cholestasis. Further details of these interactions are provided below. However, because of differences in gene expression between species, it is difficult to generalize. For additional information on adaptive responses in animal models of cholestasis and the role of NRs, the reader is also referred to several recent comprehensive reviews [4, 32].

Regulation of transporter genes in the cholestatic hepatocyte Ion transporters: (Na+K+‐ATPase and Na+/H+ exchange) Several major ion transport proteins are located on the basolateral membrane and regulate important homeostatic housekeeping functions in the hepatocyte, including maintenance of the electrical potential, cell volume, and intracellular pH. These transporters include Na+/K+‐ATPase and the Na+/H+ exchanger‐ isoform 1 (NHE‐1). Although many cholestatic agents have been shown to inhibit the sodium pump in vitro, the molecular expression of the sodium pump is either not significantly affected or somewhat upregulated, as discerned from studies in several different animal models of cholestasis [33–35]. This adaptive response could result from an attempt to counteract increased sodium entry and cell swelling that may result from the detergent properties of retained bile salts. Na+/H+ exchange is also upregulated at both transcriptional and post‐transcriptional levels following common bile duct ligation in the rat, resulting in an increase in intracellular pH. This response may also contribute to sodium entry and cell swelling in the cholestatic liver [36].

Transporters on the basolateral membrane of the hepatocyte (phase 0) Sodium taurocholate cotransporting polypeptide, NTCP/Ntcp (SLC10A1/Slc10a1) Sodium taurocholate cotransporting polypeptide (Ntcp) and its human homolog NTCP is the major determinant of the selective hepatic uptake of conjugated bile salts from the portal circulation and is substantially downregulated in both PBC [27, 37], biliary atresia [7], PFIC, and animal models of cholestasis [33, 35, 38]. NTCP mRNA is also reduced in patients with alcoholic hepatitis [37] and other inflammatory conditions. suggesting mediation by cytokines. The transcriptional regulation of NTCP/Ntcp differs between species [39]. In human tissue, the mechanism is complex as bile acids induce SHP via FXR, which reduces HNF4α binding to bile acid response elements (BAREs) in the NTCP promoter which, in turn, inhibits its transactivating effects on HNF1α [31]. HNF1α expression is highly dependent on activation by HNF4α, which is a main regulator of NTCP expression [31]. Bile acids also have SHP‐independent effects on HNF4α binding. However, SHP has no direct effect on NTCP/ Ntcp ­promoter activity, and bile acid‐induced signaling pathways via c‐Jun N‐terminal kinase may be involved. Other studies suggest that the human NTCP gene is also activated by the glucocorticoid receptor and PPARγ coactivator‐1α, and

can be suppressed by bile acids via a small heterodimer partner‐dependent ­mechanism [40]. In contrast, the rat Ntcp promoter contains binding regions for the transactivators, HNF1, and retinoid receptors (RARα/ RXRα) [41, 42]. Bile acids downregulate rat Ntcp via FXR‐ dependent Shp expression and subsequent inhibition of retinoid activation of RARα/RXRα [43]. Several additional upstream regions are involved in cytokine responses. Cholestasis produced by administration of endotoxin (LPS) results in loss of HNF1 and the RXRα : RARα heterodimer [42], which decreases Ntcp expression. LPS also results in release of cytokines (tumor necrosis factor alpha [TNF‐α] and IL‐1β) that may contribute to the diminished Ntcp expression [35]. In mice, Ntcp repression by common bile duct ligation (CBDL) and cholic acid or taurocholate feeding is mediated by FXR and does not depend on cytokines, whereas Ntcp repression by LPS is independent of FXR. Although impairment of Ntcp in rodents and NTCP in humans can explain the reduced Na+‐dependent uptake of conjugated bile salts, sodium‐independent mechanisms for hepatic bile salt uptake persist due to continued expression of several other sinusoidal membrane organic anion transporters such as Oatp2/Oatp1a4 and Oatp4/Oatp1b2 in mice and possibly OATP1B3 in humans [44, 45].

Organic anion transporting polypeptides, OATPs/Oatps (SLCO/Slco) The OATP/Oatps are members of a large super family of transporters (solute carrier organic anion transporter family [SLCO]) with broad substrate specificity that are capable of translocating a wide range of organic anions, including unconjugated and conjugated bile salts, bulky organic cations, and even certain uncharged organic substrates [46]. These proteins appear to function as anion exchangers, exchanging the extracellular anion with either intracellular bicarbonate or glutathione [41,42], although the exact driving forces are still not known. OATPs in human liver consist of OATP1B1 (SLCO1B1, ­formally OATP‐C), OATP1B3 (SLCO1B3, formally OATP8), OATP1A2 (SLCO1A2, formally OATP‐A), OATP2B1 (SLCO2B1, formally OATP‐B), and OATP3A1. Human OATP1B1 is the most widely studied and the major sodium‐independent uptake mechanism for bile acids. Its expression is reduced in alcoholic hepatitis [37] and PBC, particularly in later stages of the disease [47]. Although transcriptional mechanisms are not known for most OATPs in cholestatic patients, in vitro studies indicate that  OATP1B1, like NTCP, is regulated by both FXR/SHP‐dependent and ‐ independent mechanisms where HNF1α is also a primary transcription factor [33]. In contrast to OATP1B1, studies in primary sclerosing cholangitis suggested that OATP1A2, (formally OATP‐A), mRNA expression is actually increased [44]. However, OATP1A2 is expressed in cholangiocytes in human liver and not hepatocytes [48]. Bile acids also stimulate expression of OATP1B3 via FXR, suggesting that OATP1B3 might function in reverse and extrude organic anions from the cholestatic human hepatocyte. Recently, OATP3A1 was found to be upregulated in the liver of patients with obstructive cholestasis and in primary human hepatocytes via FGF19 activation of transcription factors SP1 and



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

nuclear factor‐kappa B where it functions as a bile acid efflux transporter [14]. Several cholestatic animal models have been used to evaluate the expression of Oatp1al (Slco1a1, Oatp1), Oatp1a4 (Slc1a4, Oatp2), Oatp1b2 (Slco1b2), and Oatp 4. CBDL, ethinylestradiol (EE), and LPS all result in downregulation of Oatp1al, although mRNA levels remain unchanged after EE treatment, suggesting that the mechanism of EE cholestasis is mainly post‐transcriptional (see Chapter 23 in the 5th edition for specific references). Estrogen‐induced cholestasis results in downregulation of all basolateral Oatps. Cytokines, particularly IL‐1β and TNF‐α, play a major role in the regulation of OATP/Oatps and also other bile acid transporters in cholestasis [49]. Basolateral and canalicular transporter systems are downregulated by both cytokines. Decreased binding activities of nuclear receptor heterodimers may also be explained in part by a reduction of the ubiquitous heterodimerization partner, RXR‐α.

Organic cation transporter (OCT‐1, SLC22a1) OCT‐1 (organic cation transporter) is the major hepatic uptake transporter for small organic cations. OCT‐1 expression in human cholestatic liver has not been assessed but animal models of cholestasis (CBDL and LPS) indicate downregulation of rOct‐1 mRNA and protein and impaired uptake of Oct‐1 substrates [50]. The human OCT‐1 gene is transactivated by HNF‐4α and, like other basolateral hepatic uptake transporters, is suppressed by bile acids via SHP [51]. Thus, hOCT‐1 is also likely to be downregulated in cholestasis affecting uptake of substrates like metformin.

383

During cholestasis, elevated hepatic levels of bile acids a­ ctivate FXR, which induces SHP and inhibits CYP7A1 gene transcription and bile acid synthesis [53]. In the intestine reductions in bile salt concentrations in the intestinal lumen decrease production of FGF15/19, an FXR regulated hormone in the ileum, that interacts with the FGFR4 receptor in hepatocytes in conjunction with β‐klotho [54]. Reductions in FGF15/19 lead to an upregulation of CYP7A1/Cyp7a1. A decline in FGF15/19 also reduces gallbladder filling and contractions [55]. FGF19 is upregulated in extrahepatic obstruction in human liver, inhibiting bile acid synthesis [56]. In addition, proinflammatory cytokines (TNFα and IL‐1β) also play a role via c‐Jun N‐terminal kinase (JNK)/cJun post‐transcriptional signal transduction pathways [57]. Thus, multiple transcriptional and post‐transcriptional mechanisms exist to reduce the formation of bile acids during cholestasis. However, the master switch appears to be at the level of  the intestine, where bile acid homeostasis is exquisitely ­regulated via FGF15/19. In rat models of CBDL and α‐naphthyl isothiocyanate (ANIT) administration, Cyp8b1 expression, but not Cyp7a1, was significantly inhibited [58] whereas 17α‐ethinylestradiol‐ induced cholestasis decreased Cyp7a1 but not the acidic pathway via Cyp27a. In patients with primary biliary cholangitis, CYP7A1 mRNA decreased to 10–20% of control levels with no changes in CYP27A or CYP8B1 [9]. Thus, adaptive changes that serve to reduce bile acid synthesis occur in both rodents and patients with cholestasis.

Bile acid hydroxylation (phase I) Organic anion transporter 2 and 3 (OAT‐2 and ‐3, Slc22a6 and Slc22a8) These two members of the OAT family are expressed in liver and kidney, and transport a variety of organic anions including bile acids. However, the effects of cholestasis on the expression of these transporters in liver is not well studied.

Bile acid synthesis Bile acids are formed from cholesterol in the liver by a series of enzymatic pathways consisting of the classic (neutral) pathway or the alternative (acidic) pathway [52]. The former is mediated by CYP7A1 and is rate limiting in the overall process. This pathway also involves CYP8B1, which leads to the production of CA and which determines the relative hydrophobicity of bile by determining the ratio of chenodeoxycholic acid (CDCA) to CA that normally is approximately equal. CYP27A1 mediates the alternative pathway which leads ­primarily to the formation of CDCA. All bile acid species are conjugated (amidated) with either glycine or taurine by bile acid CoA‐synthase (BACS) and bile acid CoA:amino acid N‐ acyltransferase (BAAT). These bile acid conjugates are then excreted into bile via the BSEP and can undergo both deamidation and 7α‐dehydroxylation by intestinal bacteria, producing deoxycholic acid and lithocholic acid. Unconjugated bile acids are reamidated when they recycle back to the liver in the enterohepatic circulation [52].

Studies in bile duct ligated rats indicate that most Cyp450‐­mediated reactions are diminished. In vitro studies show that unconjugated hydrophobic bile acids are more potent inhibitors than conjugated bile acids [59]. However, during cholestasis hydroxylation reactions mediated predominantly by CYP3A4 attempt to decrease the hydrophobicity of the bile acid pool accounting for the significant levels of (poly)hydroxylated bile acids in the urine of cholestatic animals and patients [60]. CYP3A4 is regulated by the NRs PXR, FXR [60], VDR, and CAR [61]. Bile acids also induce Cyp3a11 after CBDL in mice [60]. However, CYP3A4 mRNA is only mildly altered in patients with PBC [9].

Bile acid conjugation (phase II) Glucuronidation, sulfation, and amidation of bile acids (major Phase II reactions) also diminish their toxicity during cholestasis and facilitates their excretion back into blood via MRP3 and MRP4 with subsequent elimination by the kidney. Human ­uridine glucuronyl transferase 1A3 (UGT1A3) forms acyl chenodeoxycholic acid 24‐glucuronide. Cytosolic sulfotransferase 2A (SULT2A) also plays a significant role in sulfation reactions, particularly in females. This enzyme may be upregulated by Car‐dependent mechanisms [61] in rodents. In contrast, SULT2A is negatively regulated through CDCA‐mediated FXR activation in mice and humans [62]. Amidation reactions for bile acids are determined by both BACS and BAAT and both are regulated by FXR [63]. Few

384

THE LIVER: GENERAL OVERVIEW – ACQUIRED DEFECTS IN BILE TRANSPORT PROTEINS

studies have examined these enzymes during cholestatic liver injury. Mutations in BAAT have been described in hypercholanemia in childhood [64].

Hepatic efflux mechanisms (phase III) Basolateral membrane of the hepatocyte (the multidrug resistance associated proteins MRPS/Mrps (ABCC) and the organic solute transporter, OSTα‐OSTβ Cholestatic liver injury results in reversal of the secretory polarity of the hepatocyte. This phenomenon is associated with the  adaptive upregulation of transporters on the basolateral membrane that facilitates the extrusion of bile salts and other organic solutes back into the hepatic sinusoids where they can be eliminated in the urine. Mrp1, ‐3, and ‐4 and Ostα‐β are normally expressed at low levels in normal liver [65, 66] but are generally upregulated during cholestasis [5, 6, 11]. Mrp1, ‐5, and ‐6 are of little functional importance in the cholestatic liver while Mrp3 and ‐4 are capable of preferentially transporting glucuronidated and sulfated bile acids, respectively [15], and their induction during cholestasis may account for the appearance of these bile acid conjugates in the urine [12]. Induction of Mrp3 after BDL in mice is due to TNF‐α dependent upregulation of Lrh‐1 with increased binding of Lrh‐1 to the Mrp3 promoter [67]. Hepatic injury is more severe in bile duct ligated TNF‐ receptor knockout mice [67]. Studies in HepG2 cells indicate that human MRP3 expression is repressed by RXR α : RARα which occupies Sp1 activator sites in the MRP3 promoter. Because RXRα : RARα expression is diminished by cholestatic liver injury, loss of RXRα : RARα may lead to upregulation of MRP3/Mrp3 expression by derepressing Sp1 activation in cholestatic disorders [68]. More recent studies in Mrp4 and Mrp3 knockout mice suggest that Mrp4, rather than Mrp3, may be more important in protecting the hepatocytes from bile acid toxicity [8, 69]. MRP4 protein, but not mRNA, levels are significantly increased in patients with stage III and IV primary biliary cholangitis [9], PFIC1 [7] and late‐stage biliary atresia [27], suggesting post‐transcriptional regulation [9]. Mrp3 in mice is regulated by CAR, PXR, and VDR whereas Mrp4 is induced primarily by CAR and PPARα [70]. Both CAR and aryl hydrocarbon receptor (AHR) response elements are present in the MRP4 promoter so that bilirubin, a CAR activator, as suggested by studies in Car knockout mice [70], may stimulate MRP4 expression. There is no evidence that other basolateral Mrps expressed in the liver, including Mrp5 and ‐6, play any protective role in the adaptive response in cholestatic liver injury. OSTα‐β/Ostα‐β is a facilitated heteromeric transporter of bile acids and other organic solutes where the direction of transport is dependent on the electrochemical gradient between the cell and blood. It is weakly expressed in the liver in rodents but is significantly upregulated by FXR regulated mechanisms at both mRNA and protein levels primarily in patients with stage III and IV PBC, biliary atresia, and in bile duct ligated mice and rats [11, 27]. Ostα‐β is also regulated by Lxr in the mouse which shares a DR‐1 binding site with Fxr in the mouse promoter [71]. Paradoxically

hepatic injury is reduced after bile duct obstruction in Ostα knockout mice [72].

Canalicular membrane of the hepatocyte (MDR1/ Mdr1a,b; MDR3/Mdr2; MRP2/Mrp2; BSEP/Bsep; BCRP/Bcrp; ABCG5/G8;Abcg5/8) Despite downregulation of hepatic uptake mechanisms and upregulation of basolateral efflux transporters that function to retard the accumulation of bile acids and other toxic substrates, hepatic levels of bile acids and other constituents of bile continue to accumulate in the cholestatic liver and are primary substrates for the apical canalicular membrane efflux pumps. These transporters are all members of the ABC superfamily and the  inability of these rate‐limiting transporters to effectively excrete these substrates into bile becomes the major determinant of the cholestatic phenotype.

Canalicular organic solute transporters Multidrug resistance protein, MDR1/Mdr1a,b (ABCB1/Abcb1) MDR1 encodes for the drug efflux pump also known as P‐glycoprotein 170 (Pgp‐170). MDR1 transports a variety of drugs, xenobiotics and lipids although its importance in excretion of endogenous substrates remains unclear. Mdr1 is capable of transporting bile acids, albeit with a much lower affinity than Bsep [73]. Disease causing MDR1 mutations have not been found. However, polymorphisms have significant effects on drug absorption, excretion, and toxicity [74]. Most forms of cholestasis result in significant upregulation of Pgp‐170 Mdr1a/b mRNA in both animal models and humans with the level of expression correlating with the severity of the cholestasis [75]. MDR1 protein is increased in late‐stage PBC and in biliary atresia [9, 27]. Molecular regulation of MDR1/ Mdr1 is thought to be mediated by NF‐κB transcriptional mechanisms [76]. Both CAR and PXR activators can also stimulate MDR1 expression [77].

MDR3/Mdr2 (ABCB4/Abcb4) The importance of this phospholipid export pump in the pathogenesis of cholestasis is dramatically demonstrated by mutations in the MDR3/Mdr2 gene that result in PFIC3 in children [78] and biliary cirrhosis in the mouse knockout model, Mdr2−/− [79]. MDR3/ Mdr2/is a phospholipid floppase. In its absence phosphatidylcholine cannot be excreted into bile, and bile salts cannot form mixed micelles resulting in progressive injury to the bile duct epithelium and, in the mouse, histological findings reminiscent of primary sclerosing cholangitis [80]. Over time a biliary‐type cirrhosis develops and, in some cases, hepatocellular carcinoma. Children with PFIC3 also have defective phospholipid excretion in bile, bile duct proliferation by histological examination, and elevated γ‐glutamyl transferase, thus distinguishing them from PFIC1 and 2, where bile duct proliferation is absent and γ‐glutamyl transferase is normal [81]. Phospholipid excretion is partially deficient in heterozygotes who normally do not have a cholestatic phenotype [78, 82].



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

However, mothers of PFIC3 patients are obligate heterozygotes (MDR3+/−) and are at risk for developing cholestasis during the third trimester of pregnancy when high levels of estrogens are present [83, 84]. Polymorphisms and mutations in MDR3 can predispose patients to cholestatic liver injury when exposed to other potential cholestatic agents, including drugs and environmental toxins [85]. A missense mutation in ABCB4 has been described in ductopenic cholestatic liver disease in adults. MDR3‐deficient patients present a wide spectrum of disease including intrahepatic cholestasis of pregnancy, low phospholipid associated cholelithiasis (LPAC), cholestatic cirrhosis, and death in childhood and adults [86]. MDR3/Mdr2 is upregulated in most forms of cholestasis, including bile duct obstruction and ANIT [58] and in patients with biliary atresia [27]. MDR3/Mdr2 is regulated in large part by Fxr/FXR‐mediated mechanisms [87], as well as PPARs in mice and in humans [88]. Trials with fibrates in patients with PBC are based in part on this finding.

MRP2/Mrp2 (ABCC2/Abcc2) MRP2 encodes for the conjugate drug export pump, also known as multidrug resistance protein 2 or the canalicular multidrug organic anion transporter (cMOAT) [65, 66]. This ABC transporter is the major export pump for bilirubin diglucuronides, glutathione conjugates, and divalent bile acids conjugated with sulfates and glucuronides. It is a major determinant of bile acid independent bile flow and drug conjugate excretion. Expression varies considerably in human liver [89] and single nucleotide polymorphisms (SNPs) affect drug clearance. Mutations in the Mrp2 gene result in a stop codon and premature termination of protein translocation in the mutant TR–/GY [90] and the Eisai hyperbilirubinemic rat (EHBR) [91]. Mutations in the MRP2 gene in humans result in the Dubin–Johnson syndrome [92] a genetically determined cause of conjugated hyperbilirubinemia. Although not cholestatic by definition, since bile salt excretion is normal, this mutation results in impaired excretion of a variety of amphipathic organic anions, including leukotrienes, conjugated bilirubin, divalent bile acid conjugates, and ­ coproporphyrin isomer series 1, in addition to a variety of other compounds including bromosulfophthalein (BSP), indocyanine green, and oral cholecystographic agents [65]. Antibiotics, such as ampicillin and ceftriaxone, and heavy metals are also excreted by MRP2. Although it is not known if bile secretion is impaired in the Dubin–Johnson syndrome, bile salt independent bile flow is reduced in the rat model as a result of impaired glutathione excretion [93]. Glutathione is a low‐affinity substrate for Mrp2 , the major pathway for excretion of glutathione, oxidized GSSG and glutathione conjugates [94]. Polymorphisms in MRP2 can result in diminished glutathione excretion and thus may contribute to other forms of toxic and cholestatic liver injury [89, 95]. Indeed, the mRNA and protein expression levels of Mrp2, are markedly downregulated in animal models of cholestasis, including CBDL, EE, and particularly following administration of LPS [38]. The latter explains the increase in serum conjugated bilirubin characteristic of sepsis‐induced jaundice [38,  96]. Post‐transcriptional events can result in retrieval of Mrp2 from the canalicular membrane to a submembranous localization early in cholestasis prior to changes in mRNA

385

expression. MRP2 protein is also markedly reduced in liver biopsies from patients with inflammatory cholestasis [37]. The Mrp2 promoter is activated by RXRα : RARα, and depressed by IL‐1β following bile duct ligation in rats [42]. Thus, Mrp2 expression in obstructive cholestasis is associated with cytokine‐dependent alterations in the RXRα : RARα NRs. Response elements for CAR, PXR, FXR, and AHR also are present in the MRP2/Mrp2 promoter so the molecular regulation of this transporter in cholestasis remains complex [4]. Since bilirubin and glutathione are antioxidants, their hepatic retention in cholestasis may have cytoprotective effects.

The bile salt export pump (BSEP/Bsep, ABCB11/Abcb11) This gene encodes for the “sister of P‐glycoprotein”. the canalicular membrane bile salt export pump and the major, if not the sole determinant of bile salt dependent bile flow. Mutations in BSEP result in PFIC2 [97], benign recurrent intrahepatic cholestasis (BRIC), and intrahepatic cholestasis of pregnancy in adults. More than 100 different mutations have been described in families with these disorders [97]. PFIC2 resembles the Byler’s disease (PFIC1) phenotype, with the absence of bile duct proliferation and normal γ‐glutamyl transferase levels in serum. Bsep knockout mice also demonstrate impaired bile salt transport into bile. Altogether the evidence is compelling that BSEP/Bsep is the canalicular transporter that determines bile salt dependent bile formation. Experimental observations suggest that BSEP/Bsep is variably preserved during cholestatic liver injury. Bsep is only moderately impaired in several rat models of cholestasis [96]. Although a variety of cholestatic agents, including LPS, EE, cyclosporin A, and rifamycin, profoundly inhibit ATP dependent bile salt transport in vitro in isolated rat liver canalicular membrane vesicles [98], downregulation in vivo is less pronounced. After 3 days of CBDL in the rat Bsep mRNA and ­protein expression are inhibited by ~30 and 50%, respectively, but after 7–14 days recover to ~60 and 80% of control ­values, respectively. Immunofluorescence studies indicate that the  transporter remains at the canalicular membrane [96]. Furthermore, bile salt excretion continues in the face of complete bile duct obstruction, albeit at a reduced rate [96] and is mediated by TNF‐α and Il‐1β [99]. LPS and EE administration in vivo to rats also results in only partial inhibition of Bsep expression [100]. Although reduced staining of BSEP protein is seen in patients with inflammatory cholestasis [37], BSEP expression is maintained in patients with primary biliary cholangitis [47] and is reduced in early but not late stages of biliary atresia [7, 27], where it is maintained in its normal amount and location [101]. BSEP/Bsep is strongly regulated by FXR/Fxr in human, rat, and mouse [102] and Bsep induction by bile acids is absent in the FXR knockout mouse [103]. Thus, differences in levels of BSEP/Bsep expression in cholestasis may be related in part to differences in the levels of expression of this nuclear receptor. For example, while initially reduced, FXR and BSEP levels return to normal in late‐stage cases of biliary atresia [27]. LRH‐1 also can regulate BSEP expression and thus may play a supporting role to FXR in maintaining hepatic bile acid levels and in coordinating the expression of both CYP7A1 and BSEP to determine bile acid synthesis and excretion [104]. The nuclear

386

THE LIVER:  REFERENCES

factor erythroid 2‐related factor 2 (NRF2) also transactivates the human BSEP promoter by binding to a Maf recognition element (MARE). NRF2 is activated by oxidative stress which can result from the accumulation of “toxic” bile acids [105]. Taurocholate transport is competitively cis‐inhibited by ­various cholestatic drugs, including cyclosporin A, rifamycin, rifamycin SV, and glibenclamide in Sf9 cells expressing rat Bsep [98]. Altogether these findings suggest that the cholestatic effects of these compounds may be determined in part by the extent to which the canalicular export pumps continue to function as export pumps. Interestingly, the cholestatic metabolite estradiol‐17β‐ glucuronide inhibited ATP‐dependent taurocholate transport only when Mrp2 was co‐expressed in Sf9 cells, suggesting that this compound results in trans‐inhibition of Bsep‐mediated bile salt transport only after excretion into the bile canaliculi by Mrp2 [98]. Thus, some drugs may produce cholestasis by inhibiting BSEP only after excretion into bile. Polymorphisms in BSEP, particularly V444A, also affect the level of expression of BSEP in human liver [89] and may predispose to some forms of drug‐ induced cholestasis and cholestasis of pregnancy [85, 106]. Mutations of BSEP are associated with cholestatic liver diseases of varying severity including PFIC2 and BRIC2. Genetic polymorphisms are also linked to intrahepatic cholestasis of pregnancy (ICP) and drug‐induced liver injury [106].

Other canalicular solute and lipid transporters/flippases (BCRP, ABCG2) and ABCG5/8, MATE‐1, and FIC1) The breast cancer resistance protein, BRCP (ABCG2), is also expressed on the canalicular membrane. BRCP shares broad substrate specificity with MRP2 and excretes sulfated drug conjugates [107]. Bcrp does not appear to have a significant role in the adaptive response to cholestasis in the liver but may be more important for solute export in the kidney and intestine. BRCP is downregulated in the duodenum of patients with obstructive cholestasis but not in bile duct ligated rats [108]. Polymorphisms in BRCP are speculated to play a role in troglitazone sulfate excretion and could possibly result in cholestasis induced by this metabolite. Sterolin 1 and 2 (ABCG5/8) are two ABC transporters that form a heterodimer at the canalicular membrane and account in large part for the excretion of cholesterol and plant sterols [109]. Little is known about its role in cholestasis, although estrogen‐ induced cholestasis results in diminished mRNA expression [110]. Reduced expression of ABCG5/8 would be expected to contribute to the hypercholesterolemia that develops during cholestasis. The multidrug and toxic compound extrusion protein‐1 (MATE‐1) is also expressed in the liver on the apical canalicular membrane. Members of the MATE family are organic cation exporters that excrete metabolic or xenobiotic organic cations from the body by way of a H+ or Na+ exchange mechanism [111]. However, little is known about its specific role in the liver and whether it is regulated during cholestasis. The gene product of familial intrahepatic cholestasis‐1 (FIC1, ATP8B1) is a P‐type ATPase which functions as a phosphatidylserine flippase at the canalicular membrane [112]. Mutations in FIC1 result in PFIC‐1, also called Byler’s disease [113]. PFIC‐1 is phenotypically similar to PFIC‐2. Whether FIC1’s function is impaired in other forms of cholestasis and contributes further to cell injury is not clear.

Canalicular ion transporters (AE2, also SLC4A2) Anion Exchanger‐2 (AE2) This transporter encodes for the canalicular Cl−/HCO3− exchanger that regulates the excretion of bicarbonate, a partial determinant of canalicular bile salt independent bile formation [114]. Anion exchanger‐2 (AE2) is also expressed on the luminal membrane of the cholangiocyte and is a determinant of bicarbonate excretion from this epithelium, providing a protective “biliary bicarbonate umbrella” [115]. Thus, impairment of AE2 would be expected to reduce bile flow and thus might ­predispose to other cholestatic insults. Initial studies showed that AE2 gene expression and protein immunoreactivity at the canaliculus and in bile ducts were reduced in patients with PBC but not in other cholestatic and non‐cholestatic liver diseases [116]. Reduced expression of AE2 mRNA has also been observed in salivary glands in patients with PBC and sicca syndrome, suggesting that there may be a generalized deficiency of this transporter in this disease. Measurements of the Cl−/HCO3− anion exchanger (AE) in isolated cholangiocytes showed that the cAMP‐stimulated AE activity is diminished in PBC compared to both healthy and diseased controls. MicroRNA‐506 (miR‐506) is upregulated in cholangiocytes of PBC patients and AE2 may be a target of miR‐506 [117]. Stimulation of the activity of AE2 may be one of several mechanisms by which ursodeoxycholic acid (UDCA) may improve outcomes in PBC patients.

CONCLUSION This chapter has focused on the adaptive response that membrane transporters in hepatocytes play in inherited and acquired forms of cholestasis and the role of NRs in this process. Information about the mechanisms of these transcriptional events is rapidly expanding, revealing the complexity and the interrelatedness of these responses in the form of extended networks. Adaptative mechanisms also occur in cholangiocytes, kidney, and intestine that contribute to the ability to modulate this disorder but are beyond the scope of this review. In the future, progress will come from new therapeutic strategies, most likely combinations of novel nuclear receptor ligands that stimulate these protective pathways.

ACKNOWLEDGMENTS Publications cited from this laboratory were supported by USPHS DK 25636 and DK P30–34989.

REFERENCES 1. Trauner, M. and Boyer, J.L. Bile salt transporters: molecular characterization, function and regulation. Physiol Rev, 2003;83:633–71. 2. Boyer, J.L. New perspectives for the treatment of cholestasis: lessons from basic science applied clinically. J Hepatol, 2007;46:365–71.



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

 3. Stahl, S., Davies, M.R., Cook, D.I. and Graham, M.J. Nuclear hormone receptor‐dependent regulation of hepatic transporters and their role in the adaptive response in cholestasis. Xenobiotica, 2008;38:725–77.  4. Halibasic, E.C.T. and Trauner, M. Bile acid transporters and regulatory nuclear receptors in the liver and beyond. J Hepatol, 2013;58:155–68.   5. Soroka, C.J., Lee, J.M., Azzaroli, F. and Boyer, J.L. Cellular localization and up‐regulation of multidrug resistance‐associated protein 3 in hepatocytes and cholangiocytes during obstructive cholestasis in rat liver. Hepatology, 2001;33:783–91.   6. Denk, G.U., Soroka, C.J., Takeyama, Y. et al. Multidrug resistance‐associated protein 4 is upregulated in liver but downregulated in kidney in obstructive cholestasis in the rat. J Hepatol, 2004;40:585–91.   7. Keitel, V., Burdelski, M., Waqrskulat, U. et al. Expression and localization of hepatobiliary transport proteins in progressive familial intrahepatic cholestasis. Hepatology, 2005;41:1160–72.   8. Mennone, A., Soroka, C.J., Cai, S.Y. et al. Mrp4‐/‐ mice have an impaired cytoprotective response in obstructive cholestasis. Hepatology, 2006;43:1013–21.   9. Zollner, G., Wagner, M., Fickert, P. et al. Expression of bile acid synthesis and detoxification enzymes and the alternative bile acid efflux pump MRP4 in patients with primary biliary cirrhosis. Liver Int, 2007;27:920–9. 10. Takeyama, Y., Uehara, Y., Inomata, S. et al. Alternative transporter pathways in patients with untreated early‐stage and late‐stage primary biliary cirrhosis. Liver Int, 2009;29:406–14. 11. Boyer, J.L., Trauner, M., Mennone, A. et al. Upregulation of a basolateral FXR‐dependent bile acid efflux transporter OSTalpha‐OSTbeta in cholestasis in humans and rodents. Am J Physiol Gastrointest Liver Physiol, 2006;290:G1124–30. 12. Makino, I., Hashimoto, H., Shinozaki, K., Yoshino, K., and Nakagawa, S. Sulfated and nonsulfated bile acids in urine, serum, and bile of patients with hepatobiliary diseases. Gastroenterol, 1975;68:545–53. 13. Lee, J., Azzaroli, F., Wang, L. et al. Adaptive regulation of bile salt transporters in kidney and liver in obstructive cholestasis in the rat. Gastroenterology, 2001;121:1473–84. 14. Pan, Q., Zhang, X., Zhang, L. et al. Solute carrier organic anion transporter family member 3A1 is a bile acid efflux transporter in cholestasis. Gastroenterology, 2018; 155(5):1578–92. 15. Kruh, G.D., Belinsky, M.G., Gallo, J.M., and Lee, K. Physiological and pharmacological functions of Mrp2, Mrp3 and Mrp4 as determined from recent studies on gene‐disrupted mice. Cancer Metastasis Rev, 2007;26:5–14. 16. Zamek‐Gliszczynski, M.J., Hoffmaster, K.A., Nezasa, K., Tallman, M.N. and Brouwer, K.L. Integration of hepatic drug transporters and phase II metabolizing enzymes: mechanisms of hepatic excretion of sulfate, glucuronide, and glutathione metabolites. Eur J Pharm Sci, 2006;27:447–86. 17. Lee, J.M., Azzaroli, F., Gigliozzi, A., Mennone, A., and Boyer, J.L. Adaptive regulation of the ileal sodium‐dependent bile salt transporter (ISBT), and the multispecific organic anion transporter (MRP2), in kidney and cholangiocytes in chronic cholestasis  –  alternative pathways for bile salt excretion. Hepatology, 1999;30:417A. 18. Hruz, P., Zimmermann, C., Gutmann, H. et  al. Adaptive regulation of the ileal apical sodium dependent bile acid transporter (ASBT) in patients with obstructive cholestasis. Gut, 2006;55:395–402. 19. Dietrich, C.G., Geier, A., Salein, N. et al. Consequences of bile duct obstruction on intestinal expression and function of multidrug resistance‐associated protein 2. Gastroenterology, 2004;126:1044–53. 20. Chiang, J.Y. Bile acid metabolism and signaling. Compr Physiol, 2013;3: 1191–1212. 21. Boyer, J.L. Nuclear receptor ligands: rational and effective therapy for chronic cholestatic liver disease? Gastroenterology, 2005;129:735–40. 22. Guo, G.L., Lambert, G., Negishi, M. et al. Complementary roles of farnesoid X receptor, pregnane X receptor, and constitutive androstane receptor in protection against bile acid toxicity. J Biol Chem, 2003;278:45062–71. 23. Trottier, J., Milkiewicz, P., Kaeding, J., Verreault, M., and Barbier, O. Coordinate regulation of hepatic bile acid oxidation and conjugation by nuclear receptors. Mol Pharm, 2006;3:212–22. 24. Stedman, C.A., Liddle, C., Coulter, S.A. et al. Nuclear receptors constitutive androstane receptor and pregnane X receptor ameliorate cholestatic liver injury. Proc Natl Acad Sci USA, 2005;102:2063–8. 25. Zimmerman, T.L., Thevananther, S., Ghose, R., Burns, A.R., and Karpen, S.J. Nuclear export of retinoid X receptor alpha in response to interleukin‐ 1beta‐mediated cell signaling: roles for JNK and SER260. J Biol Chem, 2006;281:15434–40.

387

26. Demeilliers, C., Jacquemin, E., Barbu, V. et al. Altered hepatobiliary gene expressions in PFIC1: ATP8B1 gene defect is associated with CFTR downregulation. Hepatology, 2006;43:1125–34. 27. Chen, H.L., Liu, Y.J., Chen, H.L. et al. Expression of hepatocyte transporters and nuclear receptors in children with early and late‐stage biliary atresia. Pediatr Res, 2008. 28. Ghonem, N.S., Assis, D.N., and Boyer, J.L. Fibrates and cholestasis. Hepatology, 2015;62:635–43. 29. Levy, C. and Lindor, K.D. Editorial: itching to know: role of fibrates in PBC. Am J Gastroenterol, 2018;113:56–7. 30. Levy, C., Peter, J.A., Nelson, D.R. et al. Pilot study: fenofibrate for patients with primary biliary cirrhosis and an incomplete response to ursodeoxycholic acid. Aliment Pharmacol Ther, 2011;33:235–42. 31. Jung, D. and Kullak‐Ublick, G.‐A. Hepatocyte nuclear factor 1a: a key ­mediator of the effect of bile acids on gene expression. Hepatology, 2003;37 :622–31. 32. Staudinger, J.L., Woody, S., Sun, M., and Cui, W. Nuclear‐receptor‐mediated regulation of drug‐ and bile‐acid‐transporter proteins in gut and liver. Drug Metab Rev,2013;45:48–59. 33. Gartung, C., Ananthanarayanan, M., Rahman, M.A. et al. Down‐regulation of expression and function of the rat liver Na+/bile acid cotransporter in extrahepatic cholestasis. Gastroenterology, 1996;110:199–209. 34. Landmann, L. Cholestasis‐induced alterations of the trans‐ and paracellular pathways in rat hepatocytes. Histochemistry,1995;103:3–9. 35. Green, R.M., Beier, D., and Gollan, J.L. Regulation of hepatocyte bile salt transporters by endotoxin and inflammatory cytokines in rodents. Gastroenterology, 1996;111:193–8. 36. Elsing, C., Reichen, J., Marti, U., and Renner, E.L. Hepatocellular Na+/H+ exchange is activated at transcriptional and posttranscriptional levels in rat biliary cirrhosis. Gastroenterology, 1994;107:468–78. 37. Zollner, G., Fickert, P., Zenz, R. et al. Hepatobiliary transporter expression in percutaneous liver biopsies of patients with cholestatic liver diseases. Hepatology, 2001;33:633–46. 38. Trauner, M., Arrese, M., Lee, H., Boyer, J.L., and Karpen, S. Endotoxin downregulates rat hepatic Ntcp gene expression via decreased activity of critical transcription factors. J Clin Invest, 1998;101:2092–100. 39. Jung, D., Hagenbuch, B., Fried, M., Meier, P.J., and Kullak‐Ublick, G.A. Role of liver‐enriched transcription factors and nuclear receptors in regulating the human, mouse, and rat NTCP gene. Am J Physiol Gastrointest Liver Physiol, 2004;286:G752–61. 40. Eloranta, J.J., Jung, D., and Kullak‐Ublick, G.A. The human Na+‐taurocholate cotransporting polypeptide gene is activated by glucocorticoid receptor and peroxisome proliferator‐activated receptor‐gamma coactivator‐1alpha, and suppressed by bile acids via a small heterodimer partner‐dependent mechanism. Mol Endocrinol,2006;20:65–79. 41. Karpen, S.J., Sun, A., Kudish, B. et al. Multiple factors regulate the rat liver basolateral sodium‐dependent bile acid cotransporter gene promoter. J Biol Chem, 1996;271:15211–21. 42. Denson, L.A., Auld, K.L., Schiek, D.S., McClure, M.H., Mangelsdorf, D.J., and Karpen, S.J. Interleukin‐1β suppresses retinoid transactivation of two hepatic transporter genes involved in bile formation. J Biol Chem, 2000;275: 8835–43. 43. Denson, L.A., Sturm, E., Echevarria, W. et al. The orphan nuclear receptor, shp, mediates bile acid‐induced inhibition of the rat bile acid transporter, Ntcp. Gastroenterology, 2001;121:140–7. 44. Kullak‐Ublick, G.‐A., Beuers, U., Fahney, C., Hagenbuch, B., Meier, P.J., and Paumgartner, G. Identification and functional characterization of the promoter region of the human organic anion transporting polypeptide gene. Hepatology, 1997;26:991–7. 45. Jung, D., Podvinec, M., Meyer, U.A. et al. Human organic anion transporting polypeptide 8 promoter is transactivated by the farnesoid X receptor/bile acid receptor. Gastroenterology, 2002;122:1954–66. 46. Hagenbuch, B. and Stieger, B. The SLCO (former SLC21) superfamily of transporters. Mol Aspects Med, 2013;34:396–412. 47. Zollner, G., Fickert, P., Silbert, D. et al. Adaptive changes in hepatobiliary transporter expression in primary biliary cirrhosis. J Hepatol, 2003;38: 717–27. 48. Lee, W., Glaeser, H., Smith, L.H. et  al. Polymorphisms in human organic anion‐transporting polypeptide 1A2 (OATP1A2): implications for altered drug disposition and central nervous system drug entry. J Biol Chem, 2005;280:9610–7.

388

THE LIVER:  REFERENCES

49. Geier, A., Dietrich, C.G., Voigt, S. et  al. Effects of proinflammatory cytokines on rat organic anion transporters during toxic liver injury and cholestasis. Hepatology, 2003;38:345–54. 50. Denk, G.U., Soroka, C.J., Mennone, A., Koepsell, H., Beuers, U., and Boyer, J.L. Down‐regulation of the organic cation transporter 1 of rat liver in obstructive cholestasis. Hepatology, 2004;39:1382–9. 51. Saborowski, M., Kullak‐Ublick, G.A., and Eloranta, J.J. The human organic cation transporter 1 gene is transactivated by hepatocyte nuclear factor‐ 4{alpha}. J Pharmacol Exp Ther, 2006;317(2):778–85. 52. Hofmann, A.F. and Hagey, L.R. Bile acids: chemistry, pathochemistry, biology, pathobiology, and therapeutics. Cell Mol Life Sci, 2008;65:2461–83. 53. Goodwin, B., Jones, S.A., Price, R.R. et  al. A regulatory cascade of the nuclear receptors FXR, SHP‐1, and LRH‐1 represses bile acid biosynthesis. Mol Cell, 2000;6:517–26. 54. Lin, B.C., Wang, M., Blackmore, C., and Desnoyers, L.R. Liver‐specific activities of FGF19 require Klotho beta. J Biol Chem, 2007;282:27277–84. 55. Inagaki, T., Choi, M., Moschetta, A. et al. Fibroblast growth factor 15 functions as an enterohepatic signal to regulate bile acid homeostasis. Cell Metab, 2005;2:217–25. 56. Schaap, F.G., van der Gaag, N.A., Gouma, D.J., and Jansen, P.L. High expression of the bile salt‐homeostatic hormone fibroblast growth factor 19 in the liver of patients with extrahepatic cholestasis. Hepatology, 2009;49:1228–35. 57. Li, T., Jahan, A., and Chiang, J.Y. Bile acids and cytokines inhibit the human cholesterol 7 alpha‐hydroxylase gene via the JNK/c‐jun pathway in human liver cells. Hepatology, 2006;43:1202–10. 58. Liu, Y., Binz, J., Numerick, M.J. et al. Hepatoprotection by the farnesoid X receptor agonist GW4064 in rat models of intra‐ and extrahepatic cholestasis. J Clin Invest, 2003;112:1678–87. 59. Chen, J. and Farrell, G.C. Bile acids produce a generalized reduction of the catalytic activity of cytochromes P450 and other hepatic microsomal enzymes in vitro: Relevance to drug metabolism in experimental cholestasis. J Gastroenterol Hepatol, 1996;11:870–7. 60. Marschall, H.U., Wagner, M., Bodin, K. et al. Fxr(‐/‐) mice adapt to biliary obstruction by enhanced phase I detoxification and renal elimination of bile acids. J Lipid Res, 2006;47:582–92. 61. Saini, S.P., Sonoda, J., Xu, L. et al. A novel constitutive androstane receptor‐mediated and CYP3A‐independent pathway of bile acid detoxification. Mol Pharmacol, 2004;65:292–300. 62. Miyata, M., Matsuda, Y., Tsuchiya, H. et al. Chenodeoxycholic acid‐mediated activation of the farnesoid X receptor negatively regulates hydroxysteroid sulfotransferase. Drug Metab Pharmacokinet, 2006;21:315–23. 63. Pircher, P.C., Kitto, J.L., Petrowski, M.L. et al. Farnesoid X receptor regulates bile acid‐amino acid conjugation. J Biol Chem, 2003;278:27703–11. 64. Carlton, V.E., Harris, B.Z., Puffenberger, E.G. et al. Complex inheritance of familial hypercholanemia with associated mutations in TJP2 and BAAT. Nat Genet, 2003;34:91–6. 65. Konig, J., Nies, A.T., Cui, Y., Leier, I., and Keppler, D. Conjugate export pumps of the multidrug resistance protein (MRP) family: localization, substrate specificity, and MRP2‐mediated drug resistance. Biochim Biophys Acta, 1999;1461:377–94. 66. Borst, P., Evers, R., Kool, M., and Wijnholds, J. The multidrug resistance protein family. Biochim Biophys Acta, 1999;1461:347–57. 67. Bohan, A., Chen, W.S., Denson, L.A., Held, M.A., and Boyer, J.L. Tumor necrosis factor alpha‐dependent up‐regulation of Lrh‐1 and Mrp3(Abcc3) reduces liver injury in obstructive cholestasis. J Biol Chem, 2003;278:36688–98. 68. Chen, W., Cai, S.Y., Xu, S., Denson, L.A., Soroka, C.J., and Boyer, J.L. Nuclear receptors RXRalpha:RARalpha are repressors for human MRP3 expression. Am J Physiol Gastrointest Liver Physiol, 2007;292:G1221–27. 69. Belinsky, M.G., Dawson, P.A., Shchaveleva, I. et al. Analysis of the in vivo functions of Mrp3. Mol Pharmacol, 2005;68:160–8. 70. Huang, W., Zhang, Z., Chua, S.S. et al. Induction of bilirubin clearance by the constitutive androstane receptor (CAR). PNAS, 2003;100:4156–61. 71. Okuwaki, M., Takada, T., Iwayanagi, Y. et  al. LXR alpha transactivates mouse organic solute transporter alpha and beta via IR‐1 elements shared with FXR. Pharm Res, 2007;24:390–8. 72. Soroka, C.J., Mennone, A., Hagey, L.R., Ballatori, N., and Boyer, J.L. Mouse organic solute transporter alpha deficiency enhances renal excretion of bile acids and attenuates cholestasis. Hepatology, 2010;51:181–90. 73. Lam, P., Wang, R., and Ling, V. Bile acid transport in sister of P‐glycoprotein (ABCB11) knockout mice. Biochemistry, 2005;44:12598–605.

74. Maeda, K. and Sugiyama, Y. Impact of genetic polymorphisms of transporters on the pharmacokinetic, pharmacodynamic and toxicological properties of anionic drugs. Drug Metab Pharmacokinet, 2008;23:223–35. 75. Schrenk, D., Gant, T.W., Preisegger K‐H., Silverman, J.A., Marino, P.A., and Thorgeirsson, S.S. Induction of multidrug resistance gene expression during cholestasis in rats and nonhuman primates. Hepatology, 1993;17:854–60. 76. Bentires‐Alj, M., Barbu, V., Fillet, M. et al. NF‐kappaB transcription factor induces drug resistance through MDR1 expression in cancer cells. Oncogene, 2003;22:90–7. 77. Geick, A., Eichelbaum, M., and Burk, O. Nuclear receptor response elements mediate induction of intestinal MDR1 by rifampin. J Biol Chem, 2001;276: 14581–7. 78. Deleuze, J.F., Jacquemin, E., Dubuisson, C. et al. Defect on multidrug‐resistance 3 gene expression in a subtype of progressive familial intrahepatic cholestasis. Hepatology, 1996;23:904–8. 79. Smit, J.J., Schinkel, A.H., Oude Elferink, R.P. et al. Homozygous disruption of the murine mdr2 P‐glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell, 1993;75:451–62. 80. Fickert, P., Fuchsbichler, A., Wagner, M. et al. Regurgitation of bile acids from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice. Gastroenterology, 2004;127:261–74. 81. Bull, L.N., Carlton VE.H., Stricker, N.L. et al. Genetic and morphological findings in progressive familial intrahepatic cholestasis (Byler disease [PFIC‐1] and Byler syndrome): evidence for heterogeneity. Hepatology, 1997;26:155–64. 82. de Vree, J.M., Jacquemin, E., Sturm, E. et al. Mutations in the MDR3 gene cause progressive familial intrahepatic cholestasis. Proc Natl Acad Sci USA, 1998;95:282–7. 83. Jacquemin, E., Cresteil, D., Manouvrier, S., Boute, O., and Hadchouel, M. Heterozygous non‐sense mutation of the MDR3 gene in familial intrahepatic cholestasis of pregnancy. Lancet, 1999;353:210–11. 84. Pauli‐Magnus, C., Lang, T., Meier, Y. et  al. Sequence analysis of bile salt export pump (ABCB11) and multidrug resistance p‐glycoprotein 3 (ABCB4, MDR3) in patients with intrahepatic cholestasis of pregnancy. Pharmacogenet, 2004;14:91–102. 85. Lang, C., Meier, Y., Stieger, B. et al. Mutations and polymorphisms in the bile salt export pump and the multidrug resistance protein 3 associated with drug‐induced liver injury. Pharmacogenet Genomics, 2007;17:47–60. 86. Gotthardt, D., Runz, H., Keitel, V. et al. A mutation in the canalicular phospholipid transporter gene, ABCB4, is associated with cholestasis, ductopenia, and cirrhosis in adults. Hepatology, 2008;48(4):1157–66. 87. Huang, L., Zhao, A., Lew, J.L. et al. Farnesoid X receptor activates transcription of the phospholipid pump MDR3. J Biol Chem, 2003;278:51085–90. 88. Ghonem, N.S., Ananthanarayanan, M., Soroka, C.J., and Boyer, J.L. Peroxisome proliferator‐activated receptor alpha activates human multidrug resistance transporter 3/ATP‐binding cassette protein subfamily B4 transcription and increases rat biliary phosphatidylcholine secretion. Hepatology, 2014;59:1030–42. 89. Meier, Y., Pauli‐Magnus, C., Zanger, U.M. et al. Interindividual variability of canalicular ATP‐binding‐cassette (ABC)‐transporter expression in human liver. Hepatology, 2006;44:62–74. 90. Buechler, M., Koenig, J., Brom, M. et al. cDNA cloning of the hepatocyte canalicular isoform of the multidrug resistance protein, cMrp, reveals a novel conjugate export pump deficient in hyperbilirubinemic mutant rats. J Biol Chem, 1996;271:15091–8. 91. Ito, K., Suzuki, H., Hirohashi, T., Kazuhiko, K., Shimizu, T., and Sugiyama, Y. Molecular cloning of canalicular multispecific organic anion transporter defective in EHBR. Am J Physiol, 1997;272:G16–22. 92. Kartenbeck, J., Leuschner, U., Mayer, R., and Keppler, D. Absence of the canalicular isoform of the MRP gene‐encoded conjugate export pump from the hepatocytes in Dubin–Johnson syndrome. Hepatology, 1996;23:1061–6. 93. Oude Elferink, R.P.J., Ottenhoff, R., Liefting, W., Haan, J.D., and Jansen, P.L.M. Hepatobiliary transport of glutathione and glutathione conjugate in rats with hereditary hyperbilirubinemia. J Clin Invest, 1989;84:478–83. 94. Paulusma, C., van Geer, M., Evers, R. et al. Canalicular multispecific organic anion transporter/multidrug resistance protein 2 mediates low‐affinity transport of reduced glutathione. Biochem J, 1999;338:393–401. 95. Choi, J.H., Ahn, B.M., Yi, J. et al. MRP2 haplotypes confer differential susceptibility to toxic liver injury. Pharmacogenet Genomics,2007;17:403–15. 96. Lee, J.M., Trauner, M., Soroka, C.J., Stieger, B., Meier, P.J., and Boyer, J.L. Expression of the bile salt export pump is maintained after chronic cholestasis in the rat. Gastroenterology, 2000;118:163–72.



31:  Adaptive Regulation of Hepatocyte Transporters in Cholestasis

  97. Strautnieks, S.S., Byrne, J.A., Pawlikowska, L. et al. Severe bile salt export pump deficiency: 82 different ABCB11 mutations in 109 families. Gastroenterology, 2008;134:1203–14.   98. Stieger, B., Fattinger, K., Madon, J., Kullak‐Ublick G‐A., and Meier, P.J. Drug‐ and estrogen‐induced cholestasis through inhibition of the hepatocellular bile salt export pump (Bsep) of rat liver. Gastroenterology, 2000;118:422–30.   99. Donner, M.G., Schumacher, S., Warskulat, U., Heinemann, J., and Haussinger, D. Obstructive cholestasis induces TNF‐{alpha}‐ and IL‐1 ‐mediated periportal downregulation of Bsep and zonal regulation of Ntcp, Oatp1a4, and Oatp1b2. Am J Physiol Gastrointest Liver Physiol, 2007;293:G1134–46. 100. Lee, J.M., Trauner, M., Soroka, C., Steiger, B., Meier, P.J., and Boyer, J.L. The molecular expression of the bile salt excretory pump, sister of P‐glycoprotein (SPGP), is selectively preserved in cholestatic liver injury. Hepatology, 1998;28(4):429A. 101. Kogan, D., Ananthanarayanan, M., Emre, S., Suchy, F.J., and Shneider, B.L. The bile salt excretory pump (BSEP/SPGP) is not down‐regulated in human cholestasis associated with extrahepatic biliary atresia (EHBA). Hepatology, 1999;30:468A. 102. Ananthanarayanan, M., Balasubramanian, N., Makishima, M., Mangelsdorf, D.J., and Suchy, F.J. Human bile salt export pump promoter is transactivated by the farnesoid X receptor/bile acid receptor. J Biol Chem, 2001;276:28857–65. 103. Sinal, C.J., Tohkin, M., Miyata, M., Ward, J.M., Lambert, G., and Gonzalez, F.J. Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell, 2000;102:731–44. 104. Song, X., Kaimal, R., Yan, B., and Deng, R. Liver receptor homolog 1 transcriptionally regulates human bile salt export pump expression. J Lipid Res, 2008;49:973–84. 105. Weerachayaphorn, J., Cai, S.Y., Soroka, C.J., and Boyer, J.L. Nuclear factor erythroid 2‐related factor 2 is a positive regulator of human bile salt export pump expression. Hepatology, 2009;50:1588–96. 106. Kubitz, R., Droge, C., Stindt, J., Weissenberger, K., and Haussinger, D. The bile salt export pump (BSEP) in health and disease. Clin Res Hepatol Gastroenterol,2012;36:536–53.

389

107. Suzuki, M., Suzuki, H., Sugimoto, Y., and Sugiyama, Y. ABCG2 transports sulfated conjugates of steroids and xenobiotics. J Biol Chem, 2003;278:22644–9. 108. Mennone, A., Soroka, C.J., Harry, K.M., and Boyer, J.L. Role of breast cancer resistance protein in the adaptive response to cholestasis. Drug Metab Dispos, 2010;38:1673–8. 109. Graf, G.A., Yu, L., Li, W.P. et al. ABCG5 and ABCG8 are obligate heterodimers for protein trafficking and biliary cholesterol excretion. J Biol Chem, 2003;278:48275–82. 110. Kamisako, T. and Ogawa, H. Alteration of the expression of adenosine triphosphate‐binding cassette transporters associated with bile acid and cholesterol transport in the rat liver and intestine during cholestasis. J Gastroenterol Hepatol, 2005;20:1429–34. 111. Omote, H., Hiasa, M., Matsumoto, T., Otsuka, M., and Moriyama, Y. The MATE proteins as fundamental transporters of metabolic and xenobiotic organic cations. Trends Pharmacol Sci, 2006;27:587–93. 112. Paulusma, C.C., Groen, A., Kunne, C. et  al. Atp8b1 deficiency in mice reduces resistance of the canalicular membrane to hydrophobic bile salts and impairs bile salt transport. Hepatology, 2006;44:195–204. 113. Bull, L.N., van Eijk, M.J., Pawlikowska, L. et al. A gene encoding a P‐type ATPase mutated in two forms of hereditary cholestasis. Nat Genet, 1998;18:219–24. 114. Boyer, J.L. Bile duct epithelium: frontiers in transport physiology. Am J Physiol, 1996;270:G1–5. 115. Hohenester, S., Wenniger, L.M., Paulusma, C.C. et  al. A biliary HCO3‐ umbrella constitutes a protective mechanism against bile acid‐induced injury in human cholangiocytes. Hepatology, 2012;55:173–83. 116. Medina, J.F., Martinez‐Anso, E., Vazquez, J.J., and Prieto, J. Decreased anion exchanger 2 immunoreactivity in the liver of patients with primary biliary cirrhosis. Hepatology, 1997;25:12–7. 117. Concepcion, A.R., Lopez, M., Ardura‐Fabregat, A., and Medina, J.F. Role of AE2 for pHi regulation in biliary epithelial cells. Front Physiol, 2013;4:413.

SECTION D: NON‐HEPATOCYTE CELLS

32

Cholangiocyte Biology and Pathobiology Massimiliano Cadamuro1,2, Romina Fiorotto2,3, and Mario Strazzabosco2,3 Department of Molecular Medicine, University of Padua, Padova, Italy International Center for Digestive Health (ICDH), University of Milan‐Bicocca, Monza, Italy 3 Liver Center and Section of Digestive Diseases, Department of Internal Medicine, Section of Digestive Diseases, Yale University School of Medicine, New Haven, CT, USA 1 2

INTRODUCTION Cholangiocytes are the epithelial cells that line the intrahepatic and extrahepatic biliary tree. The last two decades have witnessed a significant expansion of our understanding of the function and dysfunction of this important epithelium. We have also understood that cholangiocytes are important actors in liver repair and progression of liver fibrosis and play a fundamental role in liver immunobiology. Several chapters and reviews have been recently written on a number of aspects of cholangiocyte biology. In this review we elected to focus on those mechanisms that are more pathophysiologically relevant to human diseases and that may represent possible avenues for basic and translation research in the next few years.

FUNCTIONAL ANATOMY OF THE BILIARY TREE The biliary system is a complex network of tubules coated by epithelial cells, or cholangiocytes, which starts from the canals of Hering in the liver lobules (intrahepatic biliary tree), continues outside the liver (extrahepatic biliary tree), and terminates into the ampulla of Vater. The best known function of the biliary tree is the transport and modification of the primary bile secreted by the hepatocytes, but it is now widely recognized that the biliary tree is actually a functionally complex structure endowed with many different biological functions reflected in a morpho‐ functional specialization of cholangiocytes [1, 2]. The first structure belonging to the biliary tree is represented by the canals of Hering (CoH), that is composed 50% by cholangiocytes and 50% by hepatocytes juxtaposed with each other [3, 4];

CoH connect the hepatocellular canalicular network, carrying the primary bile, with the intraportal ramifications of the biliary tree. CoH is also considered the site of the putative liver stem cell niche. The intrahepatic biliary tree gradually merges from cholangioles, interlobular, septal, areal, and segmental ducts that end in the two principal hepatic ducts and then the common hepatic duct. The latter receives the cystic duct coming from the gallbladder and connects the liver and gallbladder to the intestine. The extrahepatic biliary network is surrounded by a capillary plexus and by peribiliary glands, a stem cell niche of hepatic progenitor cells that differs from that of the CoH. Notably, the intra‐ and extrahepatic biliary tree has a different embryonic origin [5, 6]. Cholangiocytes belonging to different compartments of the biliary tree are characterized by a specialized morphology, reflecting distinct physiological function. The smallest biliary epithelial cells lining the CoH and the distal branches of the biliary tree have a cuboidal shape with a round basal nucleus, and are quickly reactive to liver and/or biliary damage. The large cholangiocytes, lining the bile ducts of higher diameter, are usually columnar and, thanks to their transport abilities, mediate alkalinization, hydration, and modification of the bile [1, 7]. Cholangiocytes in fact, have both secretory and absorptive functions, being involved in the recirculation of bile acids, glucose, and water. Notably, about 40% of human bile is generated by biliary epithelial cells, following modification of the primary product of the hepatocytes. Data in animals suggest that while secretory functions of large cholangiocytes are regulated mostly by cyclic adenosine monophosphate (cAMP) signaling, the most important second messenger in small cholangiocytes is [Ca2+]i [7, 8]. These distinctions are, however, less actual now that we have a better understanding of the microdomain restriction of second messenger signaling.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

394

THE LIVER:  DEVELOPMENT OF THE BILIARY TREE

DEVELOPMENT OF THE BILIARY TREE The extra‐ and intrahepatic branches of the biliary tree originate from different embryological areas, thus explaining their morphological and biological properties. The extrahepatic biliary tree originates from Hex−/Pdx1+/Sox17+ cells in the caudal part of the ventral foregut [9, 10]. Studies conducted on Alagille syndrome demonstrate that perturbation of the Notch pathway, one of the morphogenetic signaling pathways involved in the biliary specification of the hepatoblasts, causes intrahepatic ductopenia without affecting the extrahepatic biliary structures, and suggests that different signals govern the formation of the intra‐ and extrahepatic structures [11]. Conversely, intrahepatic bile ducts derive from Hex+/Pdx‐1−/ Sox17−/AFP+ hepatoblasts, bipotent cells present in the fetal liver bud. Starting from the eighth gestational week (GW), hepatoblasts in contact with mesenchymal cells located in the nascent portal tract, form a continue single cells rim of keratin K8+/18+/19+ cells called the ductal plate (DP). Between twelfth and sixteenth GW, the DP duplicates in discrete areas, represses the expression of hepatocellular markers and begins to express K7 and K19, that is, markers of commitment toward a biliary phenotype. The duplicating DP progressively forms a lumen, while the residual biliary monolayers are gradually reabsorbed by apoptosis or by retro‐differentiation to periportal hepatocytes, as recently shown by lineage tracking experiments [12]. Duplicating DPs are gradually incorporated into the portal mesenchyma where they are joined by a peribiliary capillary plexus, which nourishes the nascent biliary tree, and by vascular structures that will become the portal arteries [13]. The concomitant elongation and maturation of biliary epithelium and vascular structures requires the cooperation of different cell types equipped with several specific ligands, receptors, and other morphogens, among them angiogenic growth factors, such as vascular endothelial growth factor (VEGF) A, platelet‐ derived growth factor (PDGF), and the angiopoietins, Ang‐1 and Ang‐2, and their specific receptors, VEGFR1, VEGFR2, PDGFRβ, and Tie‐2 [13, 14]. In addition, a range of morphogenetic signaling and transcription factors, including Notch and WNT/β‐catenin pathways, transforming growth factor‐β (TGF‐β), and hepatocyte nuclear factors (HNFs) are activated [9, 15]. Furthermore, recent studies outlined the importance of miRNAs as regulatory mediators coordinating the interactions among all those peptides. It is important for the experimental as well as the clinical hepatologist to acquire a working knowledge of these mechanisms, since an altered development of the biliary tract is responsible for a number of liver malformative diseases, among which the so‐called DP malformations (DPM) are of particular interest [16–18]. A fundamental family of growth factors involved in biliary fate determination is TGFβ, and its receptors; in particular, TGFβ type II‐R (TβRII). All TGFβ ligands are highly expressed around the nascent portal area, and induce the transition of hepatoblasts into cholangiocytes, by binding to TβRII, which is transiently expressed by the monostratified DP cells. This effect is, at least in part, mediated by the upregulation of Jag1 and by the modulation of the transcription factors Hes1 and Hey1, two downstream effectors of Notch signaling. As demonstrated by studies on different animal models (zebrafish and mice)

[19–21], periportal hepatoblasts expressing Notch2 are induced by Jag1+ portal mesenchymal cells to acquire cholangiocyte‐like morphology and phenotype (DP cells). The interaction between Jag1 and its receptor Notch2 induces the cleavage and nuclear import of the Notch intracellular domain (NICD) that, in cooperation with TGFβ, binds to the nuclear transcription factor recombinant signal binding protein for immunoglobulin kappa J (RBP‐Jk) and activates Hes1 that, in turn promotes the expression of the biliary epithelial cell‐specific transcription factors Hnf1β, Sox4, and Sox9. Notably, mutations on Jag1 or Notch2, cause Alagille syndrome, a multiorgan disease characterized by vanishing of intrahepatic biliary structures, and by the accumulation of cells retaining both hepatic and cholangiocyte phenotype (intermediate hepatobiliary cells) [11, 22]; jaundice, itching, and hypercholesterolemia characterize the hepatic phenotype. Recent work by Diehl’s group has clarified the important role of the Hedgehog (Hh) pathway in biliary morphogenesis, repair, and cancer. This signaling is normally repressed by the interaction between Patched (Ptc) and Smoothened (Smo) [23–25]. Once the Hh ligand (i.e. Sonic, SHh) binds its receptor Patched (Ptc), the signal pathway is de‐repressed and Gli3A induces the transcription of the downstream effectors Ptc, Glioblastoma (Gli) 1 and Gli2. Recent data support an important role of Hh in bile duct morphogenesis; in fact, in the fetal mouse liver, SHh expression together with Gli1, increase at the earliest gestational days (E11.5) and rapidly decrease overtime. Its reported association with the Meckel syndrome, a lethal DP malformation, highlights the morphogenetic properties of Hh signaling. Hh also has modulatory effects on two other fundamental transcription factors, Yes‐associated protein (YAP) and Sox9. In mice, YAP is expressed at E15.5 by several cells accumulated in the periportal areas around the nascent portal tract; at E18.5 the clear majority of YAP+ cells co‐express Sox9, a marker of ­biliary differentiation, which is also downstream of Notch signaling [26]. The role of YAP as a master gene involved in differentiation of cholangiocyte precursors is also confirmed by the essential role played by the Hippo pathway in regulating liver size during liver regeneration after partial hepatectomy (PHx) [27]. Furthermore, recent data [28] show that Sox9 and Sox4, are involved in the maintenance of the apical–basal ­polarity of the epithelial sheet, in the correct development of the primary cilium and, as well as in the normal formation, elongation and branching of the biliary tree. This mechanism, which is known as “planar cell polarity” (PCP), is primarily governed by the non‐canonical Wnt/β‐catenin pathway, and is altered in some kidney and liver ciliopathies (diseases deriving from a faulty assembly of primary cilia or from the malfunction of ­proteins expressed in cilia) [29]. In summary, the available information highlight the intensive cross‐talk mechanisms among the different morphogens and likely a redundancy of the signals regulating biliary tree architecture, in health and in the reparative/regenerative response to liver damage. More recently, miRNAs have been suggested to play a role in biliary tree specification and development; Rogler and colleagues showed that the treatment of mice fetuses at E16.5 with miR‐23b, miR‐27b, and miR‐24 antagonists perturbed the correct development of the fetal liver and stimulated the expression of keratin 19 in the liver parenchyma due to the



32:  Cholangiocyte Biology and Pathobiology

upregulation of the TGFβ signaling [30]. Those miRNAs in fact target different components of the TGFβ pathway, and in particular at the level of the small mother against decapentaplegic (SMAD) [31].

PHYSIOLOGICAL FUNCTIONS OF THE BILIARY TREE Within the biliary tree there is a functional and morphological specialization between small and large ducts; in particular, the main secretory functions are sustained by cholangiocytes lining interlobular, septal, and major ducts, while the ability to react to liver damage and possibly function as bipotential progenitor cells resides in the small ducts and canals of Hering, respectively [32–36].

SECRETION AND BILE PRODUCTION The best understood function of cholangiocytes is to modify the volume, fluidity, and alkalinity of the primary bile secreted by the hepatocytes. In addition, the biliary epithelium may reabsorb water, glucose, glutathione, bile acids, and electrolytes [37]. Actually, biliary epithelial cells are able to modify the composition of the primary bile secreted by the hepatocytes into the bile canaliculi and in humans, up to 40% of the bile is produced by the biliary epithelium, depending on the digestive phases. Primary bile flows through the CoH to the biliary network where cholangiocytes lining the larger biliary structures modify it. Changes in bile volume and composition are modulated by a complex interplay among different pro‐secretory hormonal and paracrine stimuli, among which secretin and ATP are important examples [38, 39], and anti‐secretory factors such as endothelin‐1 (ET‐1) [40], gastrin, and somatostatin [41]. These secretory and anti‐secretory stimuli mostly work by increasing or decreasing the levels of cellular cAMP as will be described in detail below. The above described secretory mediators are integrated by a number of membrane‐bound or soluble adenylyl cyclase (ACs), such as AC4, 5, 6, 7, 8, 9, and SAC (soluble adenyl cyclase), resulting in given levels of intracellular levels of cAMP, and activation of PKA or epithelial Na+ channel (ENAC). Recent works have highlighted the importance of microdomains in cAMP/PKA activation. Some of these microdomains may be associated with a specific ACs that binds PKA to selected proteins. Activation of PKA stimulates the cystic fibrosis transmembrane conductance regulator (CFTR) channel [42] that mediates the luminal secretion of Cl−. Cl− secreted into the lumen creates the driving force for the activation of an apically located Na+‐independent Cl−/HCO3− exchanger (AE2) that extrudes HCO3− in exchange for Cl− from the apical surface of the biliary epithelial cells. The concentration of anions into the lumen of the bile ducts creates an osmotic gradient that attracts water into the bile ducts, through aquaporin‐1 and ‐4. To preserve cellular homeostasis, the secretion of bicarbonate is counterbalanced by the intracellular import of anions through the

395

Na+‐dependent Cl−/HCO3− exchanger (NCHE), and by the Na+/ HCO3− cotransporter (NCB1) present at the basolateral side of the cholangiocytes. The difference of potential between the intra‐ and extracellular sides of the epithelium is maintained by two Na+ transport systems: the Na+/K+/2Cl− cotransporter (NKCC1) and the Na+/K+ adenosine triphosphate (ATP)‐ase [8]. These coordinated mechanisms are necessary to increase the alkalinity and the volume of the bile and their dysfunction can be responsible for pathologic conditions. In cystic fibrosis (CF), for example, a genetic disease caused by a genetically transmitted defect in the function of CFTR, biliary complications are increasingly being recognized. In cystic fibrosis liver disease (CFLD) defective CFTR function alters the coordinated activation of AE2 and the secretion of bicarbonate and fluid into the bile. Alterations of the bile composition precipitates a cascade of events such as the formation of bile plugs, accumulation of toxicants (i.e. toxic bile acids, endotoxins) that damage the epithelium and causes focal biliary inflammation progressing to sclerosing cholangitis and eventually cirrhosis [43]. Recent findings (revised in sections on Barrier Function of Cholangiocytes and in Cholangiocytes and Immunity) have highlighted a novel function of CFTR as a regulator of endotoxin tolerance in cholangiocytes, thanks to its interaction with other proteins (i.e. Src tyrosine kinases), and this will change our view of CFLD from a classic channelopathy to an inflammatory disease (Figure 32.1). Bicarbonate secretion is not only necessary to drive bile secretory processes but it is also an important defense mechanism developed by cholangiocytes to protect themselves from several toxicants present in the bile and by the action of apolar protonated hydrophobic bile acids. The vast majority of human bile salts is glycine‐conjugated and are usually partially protonated, and apolar, becoming membrane‐permeable at pH 7.4 [44]. Thus, hydrophobic bile salts accumulate into the cell cytoplasm and may cause cell damage, including apoptosis at micromolar concentrations. Starting from these observations, it was hypothesized that the secretion of HCO3− above the apical membrane of cholangiocytes provides a sort of local bicarbonate shield (or umbrella) able to protect the epithelium. In physiologic conditions, several mechanisms are at play to generate and maintain this bicarbonate rich microenvironment [45]. In normal cholangiocytes, the primary bile acid chenodeoxycholic acid (CDCA) interacts with its receptor TGR5, a membrane‐ bound bile acid receptor coupled to a stimulatory G‐protein, localized both on the apical membrane and on the primary cilium. Activation of TGR5 is thought to increase cAMP concentration and activate the excretion of Cl−, together with ATP through CFTR. ATP released into the bile binds specific surface purinergic receptor P2Y2 that increases intracellular Ca2+ levels and stimulates by an autocrine loop the secretion of Cl− from calcium‐dependent Cl‐channels (i.e. TMEM16A). Cl− is transported back into the cytoplasm by the AE2 exchanger as described in the section on Secretion and Bile Production. Finally, the action of alkaline phosphatases, located in the apical glycocalyx in close proximity to the P2Y receptors ensures the correct amount of bicarbonate secretion by degradation of ATP in ADP and AMP [46]. It is hypothesized that a “destabilization” of the biliary bicarbonate shield could generate changes in the luminal pH that

396

THE LIVER:  BILIARY PRIMARY CILIA AND “CILIOPATHIES”

Figure 32.1  Bile secretion and modification in cholangiocytes. Cholangiocytes, through a complex network of pumps, ion exchangers, and channels are able to modify bile composition, altering its volume, pH, and hydration. Prosecretory hormones (i.e. secretin) bind their respective receptors stimulating the intracellular increase of cAMP due to the activity of ACs, leading to the activation of CFTR that extrude Cl− anions in the luminal space together with ATP. Cl− is, in turn, reabsorbed by AE2 extruding HCO3−, responsible for the alkalinization of the bile. cAMP levels are further sustained by the activation of TGR5 on the apical side of cholangiocytes. The increased osmotic gradient at the apical side of the cholangiocytes stimulates H2O flow passively through AQPs (AQP1, AQP4), as well as through a paracellular pathway; ATP, also secreted into the bile by CFTR, activates the P2Y purinergic receptor and Cl− secretion. Cellular ion homeostasis is maintained by the presence of several ion transporters at the basal side of cholangiocytes, such as NCHE, NCB1, and Na+/H− pump. This mechanism leads to HCO3− secretion into the bile, which together with the presence of a glycocalyx, constitutes a well‐known defensive mechanism that is also at the basis of the cholehepatic shunting of weak acids and bile acids (see [127]), dubbed the “bicarbonate umbrella”. Abbreviations: cAMP, cyclic adenosine monophosphate; ACs, adenylyl cyclase; CFTR, cystic fibrosis transmembrane conductance regulator; AE2, Na+‐independent Cl−/HCO3− exchanger; AQPs, aquaporins; ENAC, epithelial Na+ channel; NCB1, Na+/HCO3− cotransporter; NCHE, Na+‐dependent Cl−/HCO3− exchanger; NHE‐1, sodium‐hydrogen antiporter 1.

alter biliary homeostasis and bile salt physicochemical status and causes toxic effects to the biliary epithelium, thereby triggering a chronic cholangitis and scarring and sclerosing cholangitis akin to the pathophysiological sequence described in the Mdr3‐knockout (KO) mouse [44, 45, 47]. In chronic cholangitis, AE2 expression seems to be reduced due to the ex novo expression of miR‐506 that suppresses the translation of AE2 by binding its 3′UTR region. The reduced AE2 activity has a dual effect, that on one side destabilizes the biliary umbrella and the buffering effect on bile salts, and secondarily leads to an intracellular accumulation of HCO3− that activates the soluble AC (sAC) and sensitizes cholangiocytes to apoptosis. Treatment with ursodeoxycholic acid (UDCA) or the homolog norUDCA could in part activate the luminal secretion

of HCO3−, through a Ca2+/cPKCα/PKA mechanism and the P2Y‐dependent signaling and thus reduce the necroinflammatory damage [48].

BILIARY PRIMARY CILIA AND “CILIOPATHIES” The apical surface of cholangiocytes presents a non‐motile primary cilium. Recent studies have highlighted its importance in cholangiocyte biology and pathobiology. Primary, or sensory, cilia are non‐motile structures, characterized by a 9 + 0 configuration, in which, nine doublet microtubules, or axoneme, are



32:  Cholangiocyte Biology and Pathobiology

anchored to the basal body [49]. Primary cilia express several proteins, including polycystin 1 and 2 (PC1, and PC2), and fibrocystin (FPC) that could participate in a number of intracellular signaling cascades. Primary cilia can act as osmo‐ or mechano‐receptors in biliary structures. Cilia sense the direction of the bile flow and, by bending, activate calcium channels allowing the intracellular influx of Ca2+ ions. These structures are likely involved in cell proliferation and senescence, activation of progenitor cell compartment, regeneration, and development through the modulation of the Hh and canonical WNT pathways. Furthermore, signaling mechanisms located in cilia may supervise the correct cell planarity and polarity through non‐canonical WNT signaling [50, 51]. During biliary development, the presence and the correct conformation of the primary cilia is necessary for the correct assembly of the biliary structures; in fact, its absence is observed in three rare but lethal syndromes of infants, the Meckel [52], Joubert [53], and Jeune [54] syndromes. These are multiorgan diseases and the liver is characterized by cyst‐like, dysmorphic biliary structures surrounded by a dense fibrotic stroma, resembling congenital hepatic fibrosis (CHF). Several liver diseases, known as ciliopathies, may be related to primary cilia defects, including CHF, Caroli’s disease (CD), autosomal dominant polycystic kidney disease (ADPKD). ADPKD is a genetic liver and kidney disease due to mutation on the gene encoding for PC1 (80%) and PC2 (20%) characterized by progressive and massive enlargement of cyst covered by biliary epithelia; patients with ADPKD may develop severe complications such as mass effect, cyst hemorrhage, rupture, or infection requiring urgent liver transplantation. CHF, CD, and autosomal recessive polycystic kidney disease (ARPKD), are rare inherited diseases of the renal tubular and biliary epithelium, characterized by enlargement of the biliary tree with cyst‐ like features associated with portal fibrosis and inflammation. These diseases, all due to mutations in the Pkhd1 gene, encoding for FPC cause severe portal hypertension and may be complicated by acute cholangitis, intrahepatic lithiasis, and also cholangiocarcinoma. The influences of ciliary protein dysfunction in the pathogenesis of cystic liver diseases will be explained in the following chapters. Interestingly, mice carrying mutations for the Itf88 gene, also known as Tg737, encoding for the ciliary protein polaris, display several malformations of the kidney and biliary cystic dilatation, accompanied by pericystic fibrosis resembling the ARPKD phenotype [55].

BARRIER FUNCTION OF CHOLANGIOCYTES A fundamental function of epithelial cells, including cholangiocytes, is to selectively control the diffusion of ions and molecules through the epithelial barrier. Biliary epithelial cells, thanks to their ability to secrete ions in a polarized fashion and to their selective permeability to solute and water, actively maintain liver homeostasis. Moreover, the biliary epithelium functions as a barrier against the back‐diffusion of xenobiotics, toxic metabolites, and bile salts from the bile to the interstitial tissue. Primary or secondary changes in the correct formation and polarization of the epithelial sheet have been reported to

397

play a role the pathogenesis of several liver diseases, including primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), and CFLD [56–58]. The tight junctions (TJs) are among the main structures responsible for epithelial barrier function and are of primary importance for the correct conformation of the biliary structures; unfortunately, to date, very little is known about mechanisms controlling their function at the level of canaliculi and the bile ducts. In general, assembling and disassembling of TJ are modulated by different signaling mechanisms, including protein kinases, and G‐proteins; probiotics, glutamine, and growth factor protect the correct assembly of TJs, while disruption of TJ are induced by pathogens, toxins, and inflammatory cytokines. Experiments on polarized monolayers of cholangiocytes demonstrated that treatment with TNFα, as well as with lipopolysaccharides (LPS) and nitric oxide (NO) increase the paracellular permeability. In particular, in human and mouse in vitro cultures of cholangiocytes carrying mutations for CFTR treated with LPS increases permeability to dextrans by inducing a derangement of the normal F‐actin cytoskeletal shape, and the delocalization of E‐cadherin, a cytoskeletal structural protein typical of differentiated epithelial cells [58]. In normal cholangiocytes, CFTR is localized on the apical membrane where it functions as a channel protein but also interacts in a multiprotein complex and regulates the function of other proteins. Examples are proteins that negatively regulate the activity of Src family tyrosine kinases (SFK). In cholangiocytes that lack CFTR at the membrane, Src is indeed more active and phosphorylates the LPS receptor TLR4. An aberrant activation of TLR4 in response to LPS and the increased downstream activation of NF‐κB with production of inflammatory mediators are directly responsible for the incorrect reshaping of the F‐actin cytoskeleton and consequent increased permeability [58, 59]. Notably, the treatment with PP2, an SFK inhibitor, prevents the improper assembly of the cytoskeleton and rescues the paracellular permeability of CF‐KO cells. In contrast to Src, activation of tyrosine kinases (TKs), such as EGFR, have a protective effect on TJ conformation following hepatic noxae. The activation of the pathway mediated by EGFR inhibits Src phosphorylation by increasing concentration and activation of the phospholipase Cγ (PLCγ)/ PKC axis. Altered TJ permeability and back‐diffusion of bile acids may play a pathogenic role in experimental conditions, as well as in human biliary diseases.

CHOLANGIOCYTE REACTION TO BILIARY DAMAGE Biliary epithelial cells are usually quiescent, however, following a liver insult, cholangiocytes activate and/or proliferate as a part of the so‐called “hepatic reparative complex”. A typical element of the hepatic repair response to liver damage is the “ductular reaction” (DR), a stereotyped histopathological lesion of the biliary epithelium, which plays a fundamental role in the progression of hepatic fibrosis. DR is characterized by a marked proliferation of cholangiocytes with poor cytoplasm, arranged in cell cords without a lumen or in richly anastomosed small

398

THE LIVER:  CHOLANGIOCYTE PROLIFERATION

diameter ducts (250 mg dl−1) exhibit higher plasma FA concentrations throughout the day

467

than nondiabetic controls [84]. Similarly, subjects with T2D display higher rates of glycerol turnover and gluconeogenesis from glycerol than nondiabetic controls [85]. Importantly, increased plasma FA concentrations are a risk factor for incident T2D [86]. It is tempting to speculate that a key mechanism whereby obesity‐induced adipose tissue inflammation and insulin resistance promote T2D is by driving EGP. But studies directly addressing the role of adipose lipolysis in the transition to overt T2D are currently lacking. With regard to the second question, what role hepatocellular insulin resistance and the DAG–PKCε– INSR axis might play in fasting hyperglycemia, it is prudent to note the limitations of the rodent studies that have emphasized the importance of indirect insulin action for EGP suppression. Specifically, the use of overnight‐fasted rodents in these studies minimized the contribution of glycogenolysis to EGP and maximized the contribution of gluconeogenesis. However, as discussed above, glycogenolysis is a physiologically significant contributor to EGP in humans during the first 24 hours of a fast [60]. Hepatocellular insulin resistance potently affects hepatic glycogen metabolism [57], and insulin‐stimulated hepatic glycogen deposition is impaired in T2D [48]. Thus, for the two major components of EGP, a reasonable simplification may be to state that direct hepatocellular effects dominate insulin regulation of hepatic glycogen metabolism, while indirect effects dominate insulin regulation of gluconeogenesis (Figure 37.3).

INSULIN SENSITIZING AGENTS AND HEPATIC INSULIN RESISTANCE Metformin Metformin, a biguanide, has been used to treat patients with T2D for over 30 years and improves liver function test abnormalities, at least temporarily, in some patients with NAFLD. Since hepatic glucose production is the sum of gluconeogenesis and glycogenolysis, it is important to know which pathway is affected by metformin. Using [6,6−2H] glucose measures of EGP combined with 13 C MRS measures of net hepatic glycogenolysis and 2H2O measures of gluconeogenesis, Hundal et al. studied nine diabetic subjects before and after three months of metformin therapy and compared the results with seven age‐, sex‐, and weight‐matched controls [87]. At baseline, the diabetic subjects had increased rates of EGP compared to non‐diabetic control subjects. Using 13 C MRS measures of net hepatic glycogenolysis and 2H2O measures of gluconeogenesis, it was clear that increased hepatic gluconeogenesis accounted for the increase in EGP, consistent with previous studies [62]. After three months of metformin ­therapy, EGP was lowered in the diabetic subjects, due to a 36% reduction in hepatic gluconeogenesis [87]. Though metformin clearly reduces fasting glycemia and it is widely accepted that this occurs through decreased gluconeogenesis, the cellular mechanisms for the effect of metformin remain debated. A recent study by Madiraju et al. is unique in identifying both a specific molecular interaction for metformin and a physiological basis for the rare metformin adverse effect of lactic acidosis [88]. Madiraju et al. observed non‐competitive

468

THE LIVER:  INSULIN SENSITIZING AGENTS AND HEPATIC INSULIN RESISTANCE Adipose insulin resistance inflammation

Hepatic insulin resistance

Insulin secretion

Hepatocellular insulin signaling

Adipose lipolysis Acetyl CoA Glycerol turnover Gluconeogenesis

Net glycogenolysis

Hepatic glucose production

Figure 37.3  Effect of adipose and hepatic insulin resistance on hepatic glucose production. Hepatic glucose production is controlled through both direct and indirect mechanisms. Direct effects are mediated through the hepatocellular insulin receptor and, in the acute setting, are mediated largely through promotion of glycogen storage. Indirect effects include, but are not limited to, inhibition of adipocyte lipolysis which decreases whole‐body glycerol and fatty acid turnover. This in turn decreases allosteric activation of gluconeogenesis by hepatic mitochondrial acetyl‐CoA and substrate‐ driven gluconeogenesis from glycerol. A prediction that follows from this model is that hepatocellular insulin resistance mainly affects the glycogenolytic component of hepatic glucose production, while adipocyte insulin resistance (often associated with inflammation) mainly  affects the gluconeogenic component of hepatic glucose production.

inhibition of the redox shuttle enzyme mitochondrial glycerophosphate dehydrogenase (mGPD, or GPD2) by metformin at physiologically relevant concentrations, an effect that is predicted to cause an increase in the cytoplasmic redox state and a  reciprocal decrease in the mitochondrial redox state of the hepatocyte. These redox effects were indeed observed with ­metformin treatment in rats, and resulted in inhibition of gluconeogenesis from redox‐dependent substrates (i.e. lactate and glycerol) but not from redox‐independent substrates (i.e. pyruvate and alanine) in cultured hepatocytes. Rats treated with ­antisense oligonucleotides targeting mGPD, as well as mGPD knockout mice, phenocopied metformin‐treated animals and did not further suppress EGP upon metformin treatment. The role of mGPD inhibition in metformin‐treated humans remains to be investigated. Other hypotheses for metformin action continue to be investigated as well. Metformin has been found to inhibit mitochondrial complex I [89], though this may occur only at supraphysiologic concentrations [88]. Inhibition of the respiratory chain would be predicted to increase the cellular [AMP] : [ATP] ratio, and indeed, metformin stimulates the AMP‐activated kinase (AMPK) [88]. However, AMPK knockout mice remain responsive to metformin treatment [90] and pharmacological activation of AMPK to the same extent as metformin was insufficient to phenocopy metformin’s suppression of EGP [88]. Metformin‐induced increases in the [AMP] : [ATP] ratio have also been hypothesized to suppress EGP by antagonizing the cAMP‐mediated effects of glucagon [91]; however, the physiological relevance of this mechanism was challenged by a human study in which metformin did not inhibit glucagon‐­ stimulated increases in EGP [92].

Thiazolidinediones The thiazolidinediones (TZDs) are PPARγ agonists that improve insulin sensitivity and have been used in the treatment of T2D. Both TZDs currently approved for use in the United States, rosiglitazone and pioglitazone, have been shown to improve hepatic insulin sensitivity in human studies [93, 94]. The ability of the TZDs to improve hepatic insulin action is primarily due to their extrahepatic effects. PPARγ is mainly expressed in adipocytes with lower expression in the liver and skeletal muscle, though there are species differences in expression. Given this discordance between the site of PPARγ expression and the site of drug effects, it was hypothesized that TZDs redistribute fat from the liver and muscle into the adipocyte [95]. Mayerson et al. tested this hypothesis, giving rosiglitazone to nine patients with T2D in order to assess changes in insulin sensitivity (as assessed by the hyperinsulinemic–euglycemic clamp) and tissue fat content (as measured by 1H MRS) [96]. Insulin‐mediated suppression of lipolysis in subcutaneous fat was also assessed using microdialysis to measure glycerol release. After three months of therapy, rosiglitazone therapy was associated with around 40% decrease in hepatic triglyceride content, around 40% increase in extramyocellular triglyceride concentration, and improved suppression of adipocyte lipolysis. Though there were no detectable decreases in intramyocellular triglyceride in this study, rosiglitazone did improve insulin‐ mediated whole‐body glucose disposal. This apparent disconnect between intramyocellular triglyceride and peripheral insulin action underscores the fact that intramyocellular triglyceride is only a crude marker for the active metabolite (putatively DAG) responsible for lipid‐induced insulin resistance [97].



37:  Non‐alcoholic Fatty Liver Disease and Insulin Resistance

In  summary, TZDs exert their beneficial effects by shifting intracellular lipid from the liver and muscle into the adipose ­tissue, via PPARγ‐mediated improvements in adipose insulin sensitivity.

Weight loss Weight loss reduces intrahepatic fat content and improves hepatic insulin sensitivity in humans. Petersen et  al. studied eight obese subjects before and after weight loss induced by a hypocaloric, low‐fat diet [98]. Patients were kept on a diet until their fasting plasma glucose levels had stabilized. This required 3–12 weeks, with an average weight loss of around 8 kg. This only led to a modest change in BMI, with subjects dropping from 30 to 28 kg m−2. Thus, even after weight loss, subjects remained overweight. A comparison was made to lean, sedentary control subjects without diabetes, who had an average BMI of 24 kg m−2. Weight loss did not alter IMCL content, which was still elevated compared to the lean controls. However, intrahepatic triglyceride content dropped substantially from around 12% to 2%, approaching the less than 1% values seen in the lean controls. Weight loss and the reduction in intrahepatic triglyceride decreased the rate of fasting EGP and improved hepatic insulin sensitivity (as assessed by hyperinsulinemic–euglycemic clamp) to near normal. Thus, in this and other studies [99– 101], modest weight loss can potently correct hepatic steatosis and restore hepatic insulin sensitivity. Finally, the DiRECT trial, which compared an intensive weight loss intervention using hypocaloric meal replacement shakes to control usual care, extended these findings in overweight adults with type 2 diabetes not on insulin [102]. A subset of this trial cohort underwent measurements of liver fat content, and a large decrease (from around 16% to 3%) in liver fat content was observed in the weight loss cohort [103]. These findings underscore the reversibility of NAFLD with weight loss. As of this writing, no reports have determined whether exercise, independent of weight loss, can improve hepatic steatosis and hepatic insulin sensitivity. Several trials have tested exercise regimens of varying intensity and duration and demonstrated beneficial effects on intrahepatic triglyceride and metabolic parameters, but also on body weight [104–106]. However, in one cross‐sectional study, physical activity was inversely correlated with intrahepatic triglyceride content, even after controlling for other key variables such as age, sex, BMI, HOMA, and adiponectin [107]. It is tantalizing to hypothesize that exercise, by reversing muscle insulin resistance [108], would allow more glucose to be taken up into the muscle, with resultant decreases in de novo lipogenesis and hepatic fat content. However, more needs to be done to determine whether this hypothesis is valid, and more importantly, whether such lifestyle interventions can be implemented and adopted by an ever‐increasing at‐risk population.

REFERENCES 1. International Diabetes Federation. IDF Diabetes Atlas, 2017, International Diabetes Federation.

469

2. Gregg, E.W., Li, Y., Wang, J. et al. Changes in diabetes‐related complications in the United States, 1990–2010. N Engl J Med, 2014;370(16):1514–23. 3. Le, M.H., Devaki, P., Ha, N.B. et al. Prevalence of non‐alcoholic fatty liver disease and risk factors for advanced fibrosis and mortality in the United States. PloS One, 2017;12(3):e0173499. 4. Tilg, H., Moschen, A.R., and Roden, M. NAFLD and diabetes mellitus. Nat Rev Gastroenterol Hepatol, 2016;14(1):147. 5. Pais, R., Barritt, A.S., Calmus, Y. et  al. NAFLD and liver transplantation: Current burden and expected challenges. J Hepatol, 2016;65(6):1245–57. 6. Shulman, G.I. Ectopic fat in insulin resistance, dyslipidemia, and cardiometabolic disease. N Engl J Med, 2014;371(12):1131–41. 7. Randle, P.J., Garland, P.B., Hales, C.N. et al. The glucose fatty‐acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet, 1963;1(7285); 785–9. 8. Jucker, B.M., Rennings, A.J., Cline, G.W. et al. 13C and 31P NMR studies on the effects of increased plasma free fatty acids on intramuscular glucose metabolism in the awake rat. J Biol Chem, 1997;272(16):10464–73. 9. Randle, P.J. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle after 35 years. Diabetes Metab Rev, 1998;14(4):263–83. 10. Krssak, M., Falk Petersen, K., Dresner, A. et al. Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia, 1999;42(1):113–6. 11. Perseghin, G., Scifo, P., De Cobelli, F. et  al. Intramyocellular triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H‐13C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents. Diabetes, 1999;48(8):1600–6. 12. Rothman, D.L., Shulman, R.G., and Shulman, G.I. 31P nuclear magnetic resonance measurements of muscle glucose‐6‐phosphate. Evidence for reduced insulin‐dependent muscle glucose transport or phosphorylation activity in non‐insulin‐dependent diabetes mellitus. J Clin Invest, 1992;89(4):1069–75. 13. Rothman, D.L., Magnusson, I., Cline, G. et  al. Decreased muscle glucose transport/phosphorylation is an early defect in the pathogenesis of non‐­ insulin‐dependent diabetes mellitus. Proc Natl Acad Sci USA, 1995;92(4):983–7. 14. Cline, G.W., Petersen, K.F., Krssak, M. et al. Impaired glucose transport as a cause of decreased insulin‐stimulated muscle glycogen synthesis in type 2 diabetes. N Engl J Med, 1999;341(4):240–6. 15. Roden, M., Price, T.B., Perseghin, G. et al. Mechanism of free fatty acid‐ induced insulin resistance in humans. J Clin Invest, 1996;97(12):2859–65. 16. Dresner, A., Laurent, D., Marcucci, M. et  al. Effects of free fatty acids on  glucose transport and IRS‐1‐associated phosphatidylinositol 3‐kinase ­activity. J Clin Invest, 1999;103(2):253–9. 17. Rahimi, Y., Camporez, J.‐P.G., Petersen, M.C. et  al. Genetic activation of pyruvate dehydrogenase alters oxidative substrate selection to induce skeletal muscle insulin resistance. Proc Natl Acad Sci USA, 2014;111(46):16508–13. 18. Pollare, T., Vessby, B., and Lithell, H. Lipoprotein lipase activity in skeletal muscle is related to insulin sensitivity. Arterioscler Thromb J Vasc Biol, 1991;11(5):1192–203. 19. Jang, C., Oh, S.F., Wada, S. et al. A branched‐chain amino acid metabolite drives vascular fatty acid transport and causes insulin resistance. Nat Med, 2016;22(4):421–6. 20. Lynch, C.J., and Adams, S.H. Branched‐chain amino acids in metabolic ­signalling and insulin resistance. Nat Rev Endocrinol, 2014;10(12):723–36. 21. Petersen, K.F., Befroy, D., Dufour, S. et al. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science, 2003;300 5622;, 1140–2. 22. Petersen, K.F., Dufour, S., Befroy, D. et al. Impaired mitochondrial activity in the insulin‐resistant offspring of patients with type 2 diabetes. N Engl J Med, 2004;350(7):664–71. 23. Lee, H.‐Y., Choi, C.S., Birkenfeld, A.L. et al. Targeted expression of catalase to mitochondria prevents age‐associated reductions in mitochondrial ­function and insulin resistance. Cell Metab, 2010;12(6):668–74. 24. Camporez, J.P., Wang, Y., Faarkrog, K. et al. Mechanism by which arylamine N‐acetyltransferase 1 ablation causes insulin resistance in mice. Proc Natl Acad Sci USA, 2017;114(52):E11285–92. 25. Knowles, J.W., Xie, W., Zhang, Z. et  al. Identification and validation of N‐acetyltransferase 2 as an insulin sensitivity gene. J Clin Invest, 2015;125(4):1739–51. 26. Rusu, V., Hoch, E., Mercader, J.M. et al. Type 2 Diabetes Variants Disrupt Function of SLC16A11 through Two Distinct Mechanisms. Cell, 2017;170(1):199–212.

470

THE LIVER:  REFERENCES

27. Fink, R.I., Kolterman, O.G., Griffin, J. et  al. Mechanisms of insulin ­resistance in aging. J Clin Invest, 1983;71(6):1523–35. 28. Rowe, J.W., Minaker, K.L., Pallotta, J.A. et  al. Characterization of the ­insulin resistance of aging. J Clin Invest, 1983;71(6):1581–7. 29. Yu, C., Chen, Y., Cline, G.W. et al. Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate‐1(IRS‐1)‐associated phosphatidylinositol 3‐kinase activity in muscle. J Biol Chem, ­ 2002;277(52):50230–6. 30. Kim, J.K., Fillmore, J.J., Sunshine, M.J. et  al. PKC‐theta knockout mice are  protected from fat‐induced insulin resistance. J Clin Invest, 2004;114(6):823–7. 31. Li, Y., Soos, T.J., Li, X. et al. Protein kinase C θ inhibits insulin signaling by phosphorylating IRS1 at Ser1101. J Biol Chem, 2004;279(44):45304–7. 32. Petersen, M.C., Camporez, J.P.G., and Shulman, G.I. IRS1 Ser1101 phosphorylation impairs insulin‐stimulated muscle glucose metabolism. ­ Diabetes, 2017;66(1):A37. 33. Itani, S.I., Pories, W.J., Macdonald, K.G. et  al. Increased protein kinase C  theta in skeletal muscle of diabetic patients. Metabolism, 2001;50(5):553–7. 34. Szendroedi, J., Yoshimura, T., Phielix, E. et  al. Role of diacylglycerol ­activation of PKCθ in lipid‐induced muscle insulin resistance in humans. Proc Natl Acad Sci USA, 2014;111(26):9597–602. 35. Nowotny, B., Zahiragic, L., Krog, D. et al. Mechanisms underlying the onset of oral lipid‐induced skeletal muscle insulin resistance in humans. Diabetes, 2013;62(7):2240–8. 36. Shulman, G.I., Rothman, D.L., Jue, T. et al. Quantitation of muscle glycogen synthesis in normal subjects and subjects with non‐insulin‐dependent ­diabetes by 13C nuclear magnetic resonance spectroscopy. N Engl J Med, 1990;322(4):223–8. 37. Donnelly, K.L., Smith, C.I., Schwarzenberg, S.J. et al. Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J Clin Invest, 2005;115(5):1343–51. 38. Diraison, F., Moulin, P., and Beylot, M. Contribution of hepatic de novo lipogenesis and reesterification of plasma non esterified fatty acids to plasma triglyceride synthesis during non‐alcoholic fatty liver disease. Diabetes Metab, 2003;29(5):478–85. 39. Lambert, J.E., Ramos‐Roman, M.A., Browning, J.D. et al. Increased de novo lipogenesis is a distinct characteristic of individuals with nonalcoholic fatty liver disease. Gastroenterology, 2014;146(3):726–35. 40. Petersen, K.F., Dufour, S., Savage, D.B. et al. The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome. Proc Natl Acad Sci USA, 2007;104(31):12587–94. 41. Brown, M.S. and Goldstein, J.L. Selective versus total insulin resistance: a pathogenic paradox. Cell Metab, 2008;7(2):95–6. 42. Vatner, D.F., Majumdar, S.K., Kumashiro, N. et  al. Insulin‐independent regulation of hepatic triglyceride synthesis by fatty acids. Proc Natl Acad Sci USA, 2015;112(4):1143–8. 43. Merkel, M., Weinstock, P.H., Chajek‐Shaul, T. et  al. Lipoprotein lipase expression exclusively in liver. A mouse model for metabolism in the neonatal period and during cachexia. J Clin Invest, 1998;102(5):893–901. 44. Kim, J.K., Fillmore, J.J., Chen, Y. et  al. Tissue‐specific overexpression of lipoprotein lipase causes tissue‐specific insulin resistance. Proc Natl Acad Sci USA, 2001;98(13):7522–7. 45. Moitra, J., Mason, M.M., Olive, M. et al. Life without white fat: a transgenic mouse. Genes Dev, 1998;12(20):3168–81. 46. Kim, J.K., Gavrilova, O., Chen, Y. et al. Mechanism of insulin resistance in A‐ZIP/F‐1 fatless mice. J Biol Chem, 2000;275(12):8456–60. 47. Samuel, V.T., Liu, Z.‐X., Qu, X. et  al. Mechanism of hepatic insulin resistance in non‐alcoholic fatty liver disease. J Biol Chem, 2004; ­ 279(31):32345–53. 48. Krssak, M., Brehm, A., Bernroider, E. et  al. Alterations in postprandial hepatic glycogen metabolism in type 2 diabetes. Diabetes, 2004; 53(12):3048–56. 49. Petersen, M.C., and Shulman, G.I. Roles of diacylglycerols and ceramides in hepatic insulin resistance. Trends Pharmacol Sci, 2017;38(7):649–65. 50. Kumashiro, N., Erion, D.M., Zhang, D. et al. Cellular mechanism of insulin resistance in nonalcoholic fatty liver disease. Proc Natl Acad Sci USA, 2011;108(39):16381–5. 51. Magkos, F., Su, X., Bradley, D. et al. Intrahepatic diacylglycerol content is associated with hepatic insulin resistance in obese subjects. Gastroenterology, 2012;142(7):1444–6.

52. Luukkonen, P.K., Zhou, Y., Sädevirta, S. et al. Hepatic ceramides dissociate steatosis and insulin resistance in patients with non‐alcoholic fatty liver ­disease. J Hepatol, 2016;64(5):1167–75. 53. Ter Horst, K.W., Gilijamse, P.W., Versteeg, R.I. et al. Hepatic Diacylglycerol‐ Associated Protein Kinase Cε Translocation Links Hepatic Steatosis to Hepatic Insulin Resistance in Humans. Cell Rep, 2017;19(10):1997–2004. 54. Rando, R.R. and Young, N. The stereospecific activation of protein kinase C. Biochem Biophys Res Commun, 1984;122(2):818–23. 55. Eichmann, T.O., Kumari, M., Haas, J.T. et al. Studies on the substrate and stereo/regioselectivity of adipose triglyceride lipase, hormone‐sensitive lipase, and diacylglycerol‐O‐acyltransferases. J Biol Chem, 2012; 287(49):41446–57. 56. Samuel, V.T., Liu, Z.‐X., Wang, A. et al. Inhibition of protein kinase Cepsilon prevents hepatic insulin resistance in nonalcoholic fatty liver disease. J Clin Invest, 2007;117(3):739–45. 57. Petersen, M.C., Madiraju, A.K., Gassaway, B.M. et  al. Insulin receptor Thr1160 phosphorylation mediates lipid‐induced hepatic insulin resistance. J Clin Invest, 2016;126(11):4361–71. 58. Perry, R.J., Wang, Y., Cline, G.W. et  al. Leptin mediates a glucose‐fatty acid  cycle to maintain glucose homeostasis in starvation. Cell, 2018;172(1):234–48. 59. Taylor, R., Magnusson, I., Rothman, D.L. et al. Direct assessment of liver glycogen storage by 13C nuclear magnetic resonance spectroscopy and ­regulation of glucose homeostasis after a mixed meal in normal subjects. J  Clin Invest, 1996;97(1):126–32. 60. Rothman, D.L., Magnusson, I., Katz, L.D. et  al. Quantitation of hepatic ­glycogenolysis and gluconeogenesis in fasting humans with 13C NMR. Science, 1991;254 (5031):573–6. 61. Petersen, K.F., Price, T., Cline, G.W. et  al. Contribution of net hepatic ­glycogenolysis to glucose production during the early postprandial period. Am J Physiol, 1996;270(1 Pt 1):E186–91. 62. Magnusson, I., Rothman, D.L., Katz, L.D. et al. Increased rate of gluconeogenesis in type II diabetes mellitus. A 13C nuclear magnetic resonance study. J Clin Invest, 1992;90(4):1323–7. 63. Wajngot, A., Chandramouli, V., Schumann, W.C. et al. Quantitative contributions of gluconeogenesis to glucose production during fasting in type 2 ­diabetes mellitus. Metabolism, 2001;50(1):47–52. 64. Marchesini, G., Brizi, M., Bianchi, G. et al. Nonalcoholic fatty liver disease: a feature of the metabolic syndrome. Diabetes, 2001;50(8):1844–50. 65. Kahn, S.E. The relative contributions of insulin resistance and beta‐cell dysfunction to the pathophysiology of Type 2 diabetes. Diabetologia, ­ 2003;46(1):3–19. 66. Lee, Y.H., Wang, M.‐Y., Yu, X.‐X. et al. Glucagon is the key factor in the development of diabetes. Diabetologia, 2016;59(7):1372–5. 67. Shulman, G.I., Liljenquist, J.E., Williams, P.E. et al. Glucose disposal during insulinopenia in somatostatin‐treated dogs. The roles of glucose and ­glucagon. J Clin Invest, 1978;62(2):487–91. 68. Roden, M., Perseghin, G., Petersen, K.F. et al. The roles of insulin and glucagon in the regulation of hepatic glycogen synthesis and turnover in humans. J Clin Invest, 1996;97(3):642–8. 69. Reaven, G.M., Chen, Y.D., Golay, A. et al. Documentation of hyperglucagonemia throughout the day in nonobese and obese patients with noninsulin‐ dependent diabetes mellitus. J Clin Endocrinol Metab, 1987;64(1):106–10. 70. Raskin, P. and Unger, R.H. Hyperglucagonemia and its suppression. Importance in the metabolic control of diabetes. N Engl J Med, 1978;299(9):433–6. 71. Basu, R., Schwenk, W.F., and Rizza, R.A. Both fasting glucose production and disappearance are abnormal in people with “mild” and “severe” type 2 diabetes. Am J Physiol Endocrinol Metab, 2004;287(1):E55–62. 72. Petersen, M.C., Vatner, D.F., and Shulman, G.I. Regulation of hepatic glucose metabolism in health and disease. Nat Rev Endocrinol, ­ 2017;13(10):572–87. 73. Accili, D. and Arden, K.C. FoxOs at the crossroads of cellular metabolism, differentiation, and transformation. Cell, 2004;117(4):421–6. 74. Liu, Y., Dentin, R., Chen, D. et  al. A fasting inducible switch modulates  gluconeogenesis via activator/coactivator exchange. Nature, 2008;456(7219):269–73. 75. Koo, S.‐H., Flechner, L., Qi, L. et al. The CREB coactivator TORC2 is a key regulator of fasting glucose metabolism. Nature, 2005;437(7062):1109–11. 76. Dentin, R., Liu, Y., Koo, S.‐H. et al. Insulin modulates gluconeogenesis by inhibition of the coactivator TORC2. Nature, 2007;449 7160;, 366–9.



37:  Non‐alcoholic Fatty Liver Disease and Insulin Resistance

77. Samuel, V.T., Beddow, S.A., Iwasaki, T. et al. Fasting hyperglycemia is not associated with increased expression of PEPCK or G6Pc in patients with Type 2 Diabetes. Proc Natl Acad Sci, 2009;106(29):12121–6. 78. Perry, R.J., Camporez, J.‐P.G., Kursawe, R. et al. Hepatic acetyl CoA links adipose tissue inflammation to hepatic insulin resistance and type 2 diabetes. Cell, 2015;160(4):745–58. 79. Perry, R.J., Zhang, D., Zhang, X.‐M. et al. Controlled‐release mitochondrial protonophore reverses diabetes and steatohepatitis in rats. Science, 2015;347(6227):1253–6. 80. Zingone, A., Hiraiwa, H., Pan, C.‐J. et  al. Correction of glycogen storage disease type 1a in a mouse model by gene therapy. J Biol Chem, 2000;275(2):828–32. 81. Previs, S.F., Cline, G.W., and Shulman, G.I. A critical evaluation of mass isotopomer distribution analysis of gluconeogenesis in vivo. Am J Physiol, 1999;277(1 Pt 1):E154–60. 82. Titchenell, P.M., Chu, Q., Monks, B.R. et  al. Hepatic insulin signalling is  dispensable for suppression of glucose output by insulin in vivo. Nat Commun, 2015;12(6):7078. 83. Buettner, C., Patel, R., Muse, E.D. et al. Severe impairment in liver insulin signaling fails to alter hepatic insulin action in conscious mice. J Clin Invest, 2005;115(5):1306–13. 84. Reaven, G.M., Hollenbeck, C., Jeng, C.Y. et al. Measurement of plasma glucose, free fatty acid, lactate, and insulin for 24 h in patients with NIDDM. Diabetes, 1988;37(8):1020–4. 85. Puhakainen, I., Koivisto, V.A., and Yki‐Järvinen, H. Lipolysis and gluconeogenesis from glycerol are increased in patients with noninsulin‐dependent diabetes mellitus. J Clin Endocrinol Metab, 1992;75(3):789–94. 86. Paolisso, G., Tataranni, P.A., Foley, J.E. et al. A high concentration of fasting plasma non‐esterified fatty acids is a risk factor for the development of NIDDM. Diabetologia, 1995;38(10):1213–17. 87. Hundal, R.S., Krssak, M., Dufour, S. et al. Mechanism by which metformin reduces glucose production in type 2 diabetes. Diabetes, 2000;49(12):2063–9. 88. Madiraju, A.K., Erion, D.M., Rahimi, Y. et al. Metformin suppresses gluconeogenesis by inhibiting mitochondrial glycerophosphate dehydrogenase. Nature, 2014;510(7506): 542–6. 89. El‐Mir, M.Y., Nogueira, V., Fontaine, E. et al. Dimethylbiguanide inhibits cell respiration via an indirect effect targeted on the respiratory chain ­complex I. J Biol Chem, 2000;275(1):223–8. 90. Foretz, M., Hébrard, S., Leclerc, J. et  al. Metformin inhibits hepatic ­gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state. J Clin Invest, 2010;120(7):2355–69. 91. Miller, R.A., Chu, Q., Xie, J. et al. Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP. Nature, 2013;494(7436):256–60. 92. Konopka, A.R., Esponda, R.R., Robinson, M.M. et al. Hyperglucagonemia mitigates the effect of metformin on glucose production in prediabetes. Cell Rep, 2016;15(7):1394–1400. 93. Tiikkainen, M., Häkkinen, A.‐M., Korsheninnikova, E. et  al. Effects of rosiglitazone and metformin on liver fat content, hepatic insulin resistance,

471

insulin clearance, and gene expression in adipose tissue in patients with type 2 diabetes. Diabetes 2004; 53(8):2169–76. 94. Basu, R., Basu, A., Chandramouli, V. et  al. Effects of pioglitazone and metformin on NEFA‐induced insulin resistance in type 2 diabetes. ­ Diabetologia, 2008;51(11):2031–40. 95. Shulman, G.I. Cellular mechanisms of insulin resistance. J Clin Invest, 2000;106(2):171–6. 96. Mayerson, A.B., Hundal, R.S., Dufour, S. et al. The effects of rosiglitazone on insulin sensitivity, lipolysis, and hepatic and skeletal muscle triglyceride content in patients with type 2 diabetes. Diabetes, 2002;51(3):797–802. 97. Samuel, V.T. and Shulman, G.I. Nonalcoholic fatty liver disease as a nexus of metabolic and hepatic diseases. Cell Metab, 2018;27(1):22–41. 98. Petersen, K.F., Dufour, S., Befroy, D. et  al. Reversal of nonalcoholic hepatic steatosis, hepatic insulin resistance, and hyperglycemia by ­moderate weight reduction in patients with type 2 diabetes. Diabetes, 2005;54(3):603–8. 99. Klein, S., Mittendorfer, B., Eagon, J.C. et  al. Gastric bypass surgery improves metabolic and hepatic abnormalities associated with nonalcoholic fatty liver disease. Gastroenterology, 2006;130(6):1564–72. 100. Sato, F., Tamura, Y., Watada, H. et  al. Effects of diet‐induced moderate weight reduction on intrahepatic and intramyocellular triglycerides and glucose metabolism in obese subjects. J Clin Endocrinol Metab, 2007;92(8):3326–9. 101. Viljanen, A.P.M., Iozzo, P., Borra, R. et al. Effect of weight loss on liver free fatty acid uptake and hepatic insulin resistance. J Clin Endocrinol Metab, 2009;94(1):50–5. 102. Lean, M.E.J., Leslie, W.S., Barnes, A.C. et  al. Primary care‐led weight management for remission of type 2 diabetes (DiRECT): an open‐label, cluster‐randomized trial. Lancet, 2018;391(10120):541–51. 103. Taylor, R., Al‐Mrabeh, A., Zhyzhneuskaya, S. et al. Remission of human type 2 diabetes requires decrease in liver and pancreas fat content but is  dependent upon capacity for beta cell recovery. Cell Metab, 2018;28:547–56. 104. Oh, S., Shida, T., Yamagishi, K. et  al. Moderate to vigorous physical ­activity volume is an important factor for managing nonalcoholic fatty liver disease: A retrospective study. Hepatology, 2015;61(4):1205–15. 105. Zhang, H.‐J., He, J., Pan, L.‐L. et  al. Effects of moderate and vigorous exercise on nonalcoholic fatty liver disease: a randomized clinical trial. JAMA Intern Med, 2016;176(8):1074–82. 106. Bacchi, E., Negri, C., Targher, G. et al. Both resistance training and aerobic training reduce hepatic fat content in type 2 diabetic subjects with nonalcoholic fatty liver disease (the RAED2 randomized trial). Hepatol Baltim Md, 2013;58(4):1287–95. 107. Perseghin, G., Lattuada, G., De Cobelli, F. et al. Habitual physical activity is associated with intrahepatic fat content in humans. Diabetes Care, 2007;30(3):683–8. 108. Perseghin, G., Price, T.B., Petersen, K.F. et al. Increased glucose transport‐ phosphorylation and muscle glycogen synthesis after exercise training in insulin‐resistant subjects. N Engl J Med, 1996;335(18):1357–62.

38

AMPK: Central Regulator of Glucose and Lipid Metabolism and Target of Type 2 Diabetes Therapeutics Daniel Garcia, Maria M. Mihaylova, and Reuben J. Shaw Molecular and Cell Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, CA, USA

AMPK STRUCTURE AND MECHANISM OF ACTIVATION A fundamental requirement of all cells is that they couple the availability of nutrients to signals emanating from growth factors to drive proliferation only when nutrients are in sufficient abundance to guarantee successful cell division. Even in non‐ dividing cells, nutrients in the environment supply the necessary building blocks for cellular metabolism and survival, and fuel the bioenergetic needs of the cell by providing substrates to produce intracellular ATP to be used for all cellular processes. When nutrient levels fall, ATP levels fall and unless ATP‐consuming biosynthetic processes are curtailed, a critical shortage of ATP will cause catastrophic cellular demise. Eukaryotic cells all share a highly conserved metabolic checkpoint that acts as a sensor of ATP levels in the cell, the AMP‐activated protein kinase (AMPK). As intracellular ATP levels fall due to pathological stresses such as glucose or oxygen shortages, osmotic stress, or disruptions of glycolysis or mitochondrial oxidative phosphorylation will all result in decreased ATP. Upon activation under low ATP conditions, AMPK acts a metabolic checkpoint in the cell, suppressing ATP‐consuming biosynthetic processes while stimulating ATP‐generating processes to repair the initiating loss of ATP [1]. Upon activation, AMPK initiates

acute effects on metabolic enzymes as well as prolonged adaptions in glucose and lipid metabolism through modulation of transcriptional programs of metabolic enzymes. In addition to ubiquitous roles as an energy checkpoint, AMPK also plays additional key roles in glucose and lipid metabolism in specialized metabolic tissues in mammals and higher eukaryotes such as liver, muscle, and adipose [2]. Thus AMPK not only governs cellular energetics, but indeed overall organismal bioenergetics by coordinating the response between tissues to nutritional input. AMPK was first discovered as a mammalian protein kinase that is activated by changes in intracellular adenosine nucleotide levels [3]. However, it was not until later, that the yeast ortholog SNF‐1 (sucrose non‐fermenting complex) was annotated from a Saccharomyces cerevisiae mutant screen for cells that failed to grow on non‐fermentable carbon sources or sucrose [4, 5]. The AMP‐activated protein kinase (AMPK) is an obligate heterotrimeric kinase complex composed of a catalytic (α) subunit and two regulatory (β and γ) subunits. AMPK is activated under conditions of energy stress, when intracellular ATP levels decline and intracellular AMP increases, as occurs during nutrient deprivation or hypoxia [2]. Upon energy stress, AMP directly binds to tandem repeats of cystathionine‐β‐synthase (CBS) domains in the AMPK γ subunit. Binding of AMP is

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS

thought to prevent dephosphorylation of the critical activation loop threonine in the α subunit [6]. The phosphorylation of the activation loop threonine is absolutely required for AMPK activation. In mammals, there are seven mammalian genes encoding each of the α, β, γ subunits, allowing for 12 distinct heterotrimeric variations (see Figure 38.1). There are two genes encoding catalytic subunits, α1 and α2, two regulatory β and three γ subunits that participate in the heterotrimer. Of these, γ3 appears mostly skeletal muscle specific and α2 appears to be most highly expressed in key metabolic tissues including muscle and liver. The catalytic α1 and α2 subunits contain a kinase domain within their N‐terminus, as well as a critical region for binding β and γ subunits within their C‐terminus. All kinases possess an activation loop which is often a target for upstream kinases, creating a conformation change that allows substrate assess to the catalytic pocket, and the activation loop of the AMPK α subunits contains a single threonine (Thr172 in mammalian AMPKα) that is the key regulatory site whose phosphorylation is absolutely required across all species for AMPK activation. It has been shown that α1 catalytic isoform is found mainly in the cytoplasm, whereas the α2 isoform appears to be nuclear in some cell types. In liver, phosphorylation of both the α1 and α2 subunits accounts for half of the total AMPK activity and there appears to be no preferential binding of the α1 and α2 subunits with the different β or γ subunits [7]. The crystal structures of S. cerevisiae Snf1 and the human α2 kinase domains

were first revealed (8, 9), but in the past decade our molecular understanding of the regulation and function of AMPK has significantly been advanced by the elucidation of the crystal structures of a variety of AMPK holoenzymes [10–16]. Although some details differ, together these studies provide a detailed view of the architecture of the AMPK complex. The structure of the AMPK trimeric complex consists of three major segments or “modules”: the catalytic module, the carbohydrate‐binding module (CBM), and the nucleotide‐binding module (also called “regulatory fragment”). The activation loop of the α‐subunit resides at the interface between the catalytic and nucleotide‐ binding modules, in close proximity to the C‐terminus of the β‐subunit and the CBS repeats of the γ‐subunit. This structural arrangement ensures that phosphorylation and dephosphorylation of Thr172 is sensitive to conformational rearrangements induced by nucleotide binding. The catalytic domain exhibits a typical eukaryotic serine/threonine kinase domain structure with a small N‐lobe and a large C‐lobe. The CBM directly contacts the N‐lobe of the kinase domain and the interface between these two modules forms a discrete pocket that was identified as the binding site for many direct AMPK‐activating compounds. It is speculated that natural metabolites might bind this site to regulate AMPK, however, no such metabolite has been yet identified. The nucleotide‐binding module is made up mostly by the γ‐subunit, which forms a flattened disk with the CBS repeats symmetrically arranged around the disk, one in each quadrant. Length

α1

473

Major Site Chromosome of Expression Location

T172

βγ Binding

Kinase domain

α-CTD

550

ubiquitous

α-CTD

552

liver, muscle widespread

1p31

5p11-14

T172

α2

Kinase domain glycogen binding β1

GBD

270

ubiquitous

12q24.1-.3

β2

GBD

272

ubiquitous?

1q21.1

Bateman domain 1 AMP γ1 γ2 long

Bateman domain 2 AMP AMP

CBS1

CBS2

CBS3

CBS4

331

ubiquitous

12q12-14

CBS1

CBS2

CBS3

CBS4

569

ubiquitous

7q36

γ2 short

CBS1

CBS2

CBS3

CBS4

328

ubiquitous

7q36

γ3-NTD

CBS1

CBS2

CBS3

CBS4

489

skeletal muscle

2q35

CBS1

CBS2

CBS3

CBS4

464

skeletal muscle

2q35

γ2-NTD

γ3 long

αγ Binding

γ3 short γ3-NTD

Figure 38.1  Human AMPK subunit isoforms. Domain structure, expression pattern, and alternative splice isoforms of the two catalytic kinase (α) isoforms, the two beta regulatory subunits which contain a glycogen‐binding domain (GBD), and the three genes encoding the gamma subunits, which each contain 4 CBS domains which directly bind to AMP as drawn.

474

THE LIVER:  UPSTREAM REGULATORS OF AMPK: LKB1 AND CAMKK

Mechanistically, these crystallographic studies reveal the molecular details of how adenine nucleotides and small molecule activators activate AMPK. In the case of nucleotides, the crystal structures show that when AMP is bound to site 3, the γ‐subunit forms stable interactions with a few amino acids within the α‐ linker’s α‐RIM1 and α‐RIM2, which interact with the unoccupied site 2 and the AMP molecule bound at site 3, respectively [12, 15, 16]. The binding of the α‐RIM motifs to the γ‐subunit restricts the flexibility of the α‐linker, resulting in tighter association of the catalytic and nucleotide‐binding modules, which physically protects Thr172 from dephosphorylation. Interestingly, the same effect is proposed to occur when ADP binds site 3, raising the possibility that in some contexts ADP might be the relevant AMPK activating signal [14]. Moreover, the binding of the α‐ RIM motifs to the γ‐subunit shifts the autoinhibitory domain (AID) in the α‐subunit away from the kinase domain when AMP is bound, thus releasing the AID’s negative effects on the kinase domain [10, 12, 13]. This rearrangement of the AID domain may represent the molecular basis for the allosteric activation effect of AMP. According to this model, the AID can shift between kinase domain‐bound (inactive AMPK) and nucleotide‐module‐bound states (active AMPK) depending on the nucleotide binding status. In summary, the published crystal structures concur that binding of AMP, especially at site 3, induces a conformational change that is transmitted to the kinase domain by changes in the interaction of the α‐RIM motifs and the AID with the nucleotide‐binding module. These structural changes, which are opposed by ATP, result in the allosteric activation of AMPK and a compaction of the interface between the catalytic module and the nucleotide‐ binding module, which protects Thr172 from dephosphorylation. However, it is not clear whether these structural rearrangements also promote Thr172 phosphorylation. On the other hand, activating compounds, such as A769662, activate AMPK by a different mechanism. Binding of these compounds, together with phosphorylation of serine 108 in the β‐subunit, stabilizes the CBM and strengthens the interaction of the CBM with the kinase domain [10, 13, 15]. Specifically, binding of activating compounds induces the formation of an α‐helix in the β‐subunit, termed the C‐interacting helix, which interacts with the so‐called C‐helix of the kinase domain (a conserved helix across multiple kinases which is important for ATP binding). This conformational change results in a shift toward a closed, active conformation of the kinase domain, protection from Thr172 dephosphorylation, and increased substrate affinity. Interestingly, glycogen inhibits the CBM‐KD interaction and this may be the mechanism by which glycogen inhibits AMPK [17].

or oxygen (hypoxia). It is now evident, that as complex organisms developed, various circulating hormones gained function to act as whole‐organism sensors and are capable of turning on AMPK in response to metabolic stresses such as starvation. One well‐known adipokine, adiponectin has been shown to activate AMPK in liver, leading to fatty‐acid oxidation and decrease in blood glucose levels, consistent with previous findings that adiponectin is able to suppress hepatic glucose production [18]. In addition to hormonal input from adiponectin, AMPK activity has been shown to be modulated by leptin, resistin, ghrelin, epinephrine, and cannabinoids [19]. Exercise is another metabolic stress that has been shown to activate AMPK in response to muscle contraction. Such activation may be due to increased AMP to ATP ratios caused by movement and muscle contraction, correlating to studies in mice where electrical muscle stimulation increased AMP levels and turned on AMPK [20]. Two different classes of drugs for treatment of diabetes mellitus type 2, biguanides such as metformin [21] and thiazolidinediones (TZDs) such as rosiglitazone [22] and pioglitazone [23] have been shown to activate hepatic AMPK, most likely through perturbation of mitochondrial ATP output via mild inhibition of complex I of the mitochondrial respiratory chain. Metformin is a biguanide that has been structurally modified from galegine, a naturally occurring compound found in the French lilac (Galega officinalis). Although it was not until 1918 that biguanides were discovered to have a blood glucose lowering effect, people have been using the French lilac to ameliorate a variety of maladies since the Middle Ages [24]. Bitter melon or Momordica charantia, like the French lilac, has been used for hundreds of years in traditional Chinese medicine to treat many different ailments and it was not until recently that scientists isolated triterpenoid compounds from M. charantia that are able to activate AMPK and facilitate fatty acid oxidation and glucose utilization when administrated in mice [25]. In a recent study, AMPK has been also been shown to become active in response to resveratrol, a polyphenol that is found in the skin of red grapes, certain nuts and berries, and has been linked to longevity in model organisms such as yeast, nematodes, drosophila, fish, and mice. Further, it was shown that resveratrol could mimic the benefits of dietary restriction in mice by perturbing fatty liver phenotype and increasing insulin sensitivity in animals fed high calorie diet [26]. Although direct mutations in AMPK have not been found in diabetic patents, studies show that AMPK is implicated in various pathways deregulated in such metabolic disorders and makes it an attractive therapeutic target in treatment of such diseases.

STRESS, HORMONES, AND THERAPEUTICS ACTIVATE AMPK

UPSTREAM REGULATORS OF AMPK: LKB1 AND CAMKK

Animals, in their multi‐cell complexity and multi‐organ utilization, have evolved intricate mechanisms to sense and initiate immediate responses to energy requirement or nutrient deprivation. Multiple studies have now shown that cellular stress caused by starvation or exercise can activate AMPK. In single‐cell eukaryotes as well as mammalian cell culture conditions, AMPK is activated in response to nutrient depletion of glucose

There was evidence for an upstream kinase activating AMPK as early as 1978 [27] and in the years to follow scientists intently labored to identify the kinase responsible for phosphorylation and activation of catalytic subunits α1 or α2. However, it was not until 1996 that an “AMP‐ activated protein kinase” (AMPKK) was partially purified from rat liver and shown to phosphorylate AMPK on Thr172 [28]. Quickly after the budding yeast genome



38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS

was completed, three upstream kinases Sak1 (Pak1), Elm1, and Tos3 were identified by whole genome screening methods and were shown to act upstream of the yeast AMPK orthologue, the SNF1 complex. When these kinases were genetically knocked out in yeast, the results yielded the same phenotype as a snf1 mutant, again placing them upstream of the SNF1 complex [29, 30]. In the human genome the closest related kinases to the yeast ones were found to be Calmodulin‐dependent protein kinases, CaMKKα and CaMKKβ and the protein kinase LKB1. Several groups simultaneously showed that LKB1 [1, 31, 32] and CAMKKβ [33–35] were indeed the upstream kinases acting on AMPK in mammals and were capable of phosphorylating both AMPKα1 and AMPKα2 subunits on Thr172. Interestingly, LKB1 was first identified in humans as a serine/ threonine tumor suppressor kinase that is defective in the cancer predisposing Peutz–Jeghers Syndrome [36]. In mammalian cells, LKB1 exists in a complex with two other proteins, STRAD (sterile‐20‐related adaptor) and MO25 (mouse protein‐25) [31] and when bound to these accessory proteins, it is stabilized and constitutively activated. Recent data suggest that when AMP nucleotides bind allosterically to the Bateman domains in the γ subunit of AMPK complex, conformational changes occur that protect the LKB1‐mediated phosphorylation Thr172 in the α subunit [6], although effects on localization and complex formation between LKB1 and AMPK following energy stress remain to be fully explored. In 2005, scientists showed that LKB1 is indeed the major upstream AMPK kinase in liver [37] and that genetic deletion of hepatic LKB1 almost completely reduces hepatic AMPK activity. Lack of LKB1 in mouse liver rapidly leads to hyperglycemia and increased levels of gluconeogenic and lipogenic gene expression. It was also shown in these animals that activation of AMPK by the antidiabetic therapeutic metformin is dependent on LKB1, and in the absence of hepatic LKB1, metformin is unable to lower blood glucose levels [37]. However, it is important to note that AMPK is not the only substrate of LKB1. LKB1 similarly phosphorylates the activation loop of a family of 11 kinases all related to AMPK, also resulting in their activation [38]. Importantly, of these 14 LKB1‐dependent kinases, only AMPKα1 and AMPKα2 are activated under low ATP conditions, probably due to the fact that only they interact with AMPKγ which contains the AMP‐binding domains [39]. Little is currently known about what stimuli direct LKB1 toward any of these AMPK‐related kinases and current evidence suggests that LKB1 is constitutively active and these other kinases may be regulated through phosphorylation at other sites outside of their activation loops. Collectively, these findings map AMPK on the axis of a major tumor suppressor pathway and provide an interesting link between cancer and metabolism. In addition to LKB1, the calmodulin‐dependent protein kinases, CaMKKα and CaMKKβ also regulate AMPK activity, though only in response to calcium flux and not in response to changes in AMP, which appears to work completely through LKB1 based on genetic knockout and RNAi studies. Genetic deletion of LKB1 dramatically reduces AMPK activation in liver suggesting little role for CAMKK in this tissue, though in hypothalamic neurons controlling food intake, CAMKKβ (CAMKK2) appears to be the dominant upstream kinase for AMPK [40]. This is consistent with the fact that CaMKKα and CaMKKβ are most highly expressed in neurons, whereas LKB1 is more ubiquitously expressed.

475

DOWNSTREAM TARGETS I: REGULATION OF ACUTE METABOLIC RESPONSE – ENZYME IN LIPOGENESIS In the liver, AMPK phosphorylates and regulates multiple downstream targets involved in lipogenesis and lipid homeostasis (Figure 38.2). One of the first identified downstream targets of AMPK was acetyl‐coenzyme A carboxylase (ACC) [41]. ACC is an enzyme involved in the generation of fatty acid precursor malonyl‐CoA, a key metabolite in the regulation of energy homeostasis. Two genes encoding two different ACC isoforms are found in mammals – ACC1 and ACC2 – and they appear to have distinct tissue specificity. It has been shown that ACC1 and ACC2 control the synthesis of two different pools of malonyl‐CoA production. ACC1 is thought to suppress the production of malonyl‐CoA used in fatty acid synthesis whereas ACC2 stimulates fatty acid oxidation (reviewed in [42]). AMPK inhibits both ACC1 and ACC2 through direct phosphorylation of their homologous residues Ser79 in ACC1 and Ser218 in ACC2. Downregulation of ACC1 activity leads to reduced malonyl‐CoA levels and a decrease in lipogenesis. AMPK phosphorylation of ACC2 inhibits its enzymatic activity and decreases cellular levels of malonyl‐CoA levels, which leads to direct inhibition of mitochondrial fatty acid uptake and increased fatty acid oxidation and ATP production through carnitine palmitoyltransferase 1 (CPT‐1) (reviewed in [19]). Interestingly, it has been also shown that calorie restriction can increase AMPK activity causing a decrease of fatty acid synthesis or upregulation of fatty acid oxidation through inhibition of ACC. In a recent study, various AMPK activating polyphenol compounds led to lowering of lipid accumulation in HepG2 cells grown in high glucose and inhibited atherosclerosis in diabetic LDLR−/− mice treated with the various polyphenols [43]. In addition to reporting ACC as a downstream target of AMPK, HMG‐CoA reductase was also found as one of the first key downstream substrates [41, 44]. It had been already known for over ten years at the point that HMG‐CoA reductase kinase (3‐hydroxy‐3‐ methyl‐glutaryl‐CoA reductase or HMGR) activity was regulated by an upstream kinase [27], however, that kinase had not yet been discovered. Today, it is known that HMG‐CoA reductase is a rate‐ limiting enzyme involved in the production of cholesterol and other isoprenoids, and more specifically functions in converting HMG‐CoA to mevalonic acid. By phosphorylating HMGR, AMPK blocks anabolic or ATP consuming processes such as cholesterol synthesis in order to preserve intracellular ATP levels. Strikingly, the AMPK phosphorylation site in HMGR is conserved throughout eukaryotes including plants.

Downstream targets II: regulation of metabolic adaptation: control of transcription In response to changes in AMP : ATP ratios, AMPK can rapidly regulate downstream targets via phosphorylation. However, in addition to these fast post‐translational modifications, AMPK can also promote long‐term transcriptional changes and reprogram transcription of certain genes in response to the cellular state. These effects are thought to be mediated by the direct

476

THE LIVER:  DOWNSTREAM TARGETS I: REGULATION OF ACUTE METABOLIC RESPONSE – ENZYME IN LIPOGENESIS

Low Nutrients (glucose, O2) Mitochondrial inhibitors: TZDs, biguanides

Exercise adiponectin

LKB1 STRAD MO25

Ghrelin, cannabinoids Resistin

AICAR (amp mimetic) Resveratrol, polyphenols

AMP

CAMKKβ Ca2 +

Glycolysis inhibitors: 2-DG Abbott A769662

γ β

P AMPKα

Glycolysis/ Glucose Uptake iPFK2 Ser461 TBC1D1 Ser237

Fatty Acid Oxidation ACC2 Ser221

Transcriptional Control of Glucose Metabolism

Lipid Synthesis HMG CoR Ser872 ACC1 Ser79

Insulin Sensitivity Protein Synthesis TSC2 Ser1387 Raptor Ser792 IRS1 Ser794

CRTC2 FOXO3 p300 AREBP HNF4a Chrebp

Ser171 Ser413 Ser89 Ser470 Ser313 Ser568

Figure 38.2  The AMPK signaling pathway. AMPK is phosphorylated on its kinase activation loop threonine and activated by two distinct upstream kinases in response to different stimuli. LKB1 activates AMPK in response to all stimuli that lower intracellular ATP and increase AMP. CAMKKb activates AMPK in response to calcium flux in an AMP‐independent manner. AMPK is activated physiologically by exercise, low nutrients such as lowered glucose or lowered oxygen, and hormones including ghrelin, leptin, adiponectin, and cannabinoids. Leptin is reported to activate AMPK in peripheral tissues but inhibit AMPK in the central nervous system through poorly‐understood mechanisms. In addition, AMPK is activated by agents that disrupt ATP production by inhibiting or poisoning the mitochondria, including uncouplers, or agents that inhibit glycolysis such as the glucose analog 2‐deoxyglucose that as a competitive inhibitor for hexokinase. Additional pharmacological agents that activate AMPK include resveratrol and related polyphenols as well as the cell‐permeable AMP‐mimetic AICAR and the first small molecule direct activator Abbott A‐769662. Upon activation, AMPK serves to inhibit anabolic, ATP‐consuming biosynthetic processes such as protein, lipid, and glucose synthesis, while upregulating catabolic processes to generate ATP production, including increased glycolysis, glucose uptake, and fatty acid oxidation. AMPK modulates cell growth and insulin sensitivity through the mTOR signaling pathway. All current in best established in vivo direct AMPK substrates and their AMPK phosphorylation sites are listed. All the sites listed conform to the identified optimal AMPK substrate motif.

phosphorylation of sequence‐specific transcription factors and transcriptional coactivators by AMPK. Of the two best‐studied transcriptional programs controlled by AMPK in liver, gluconeogenesis and lipogenesis, a number of direct substrates for AMPK in gluconeogenesis have been identified. Interestingly, two different coactivators which modulate cyclic AMP‐responsive element binding protein (CREB)‐ dependent transcription in response to glucagon induced by fasting are the histone acetyltransferase p300 and the CREB coactivator CRTC2 (also known as TORC2 or transducer of regulated CREB activity 2) [45–47]. CRTC2 is a critical rate‐ limiting transcriptional regulator of gluconeogenesis in mice via effects on hepatic CREB targets including the PGC‐1α promoter. When AMPK is active, it can phosphorylate CRTC2 on residue Ser171, which in turn allows CRCT2 to associate to 14–3–3 proteins and be sequestered from the nucleus to the cytoplasm, an event that shuts off transcription of gluconeogenic enzymes [48, 49]. If AMPK activity is decreased, hypophosphorylated CRTC2 can conversely move into the nucleus where it binds to transcriptional coactivator CREB and promotes transcription of PGC1α and downstream gluconeogenic targets PEPCK and G6Pase (see Figure 38.3). Strikingly, both p300 and CRTC2 have been shown to be phosphorylated at the same regulatory sites by either AMPK or its related family member SIK1 (salt‐inducible kinase 1), both of which are

LKB1‐dependent. In mice lacking hepatic LKB1, CRTC2 is hypophosphorylated and predominantly nuclear compared to wild‐type mice [37]. Furthermore, LKB1−/− mice in liver had dramatic increases in fasting blood glucose levels, which were greatly attenuated upon introduction of an shRNA reducing CRTC2 levels in the liver, reinforcing the idea that gluconeogenesis in liver is controlled by LKB1‐dependant kinases and that CRTC2 is a key downstream target of these kinases. Consistent with these findings in the LKB1 liver‐specific knockout mice, animals bearing a liver‐specific knockout of the AMPKα2 isoform also exhibited elevated hepatic glucose output, glucose intolerance, impaired leptin‐ and adiponectin‐regulated hepatic glucose production [50] (see Table 38.1). Future studies will be needed to dissect the temporal and spatial regulation of the contexts in which AMPKa2 or SIK1 control these key modulators of CREB and gluconeogenesis. Interestingly, SIK1 itself is a CREB target, providing a time‐delayed mechanism to attenuate chronic CREB‐dependent transcription. Phosphorylation of p300 and CRTC2 by AMPK and SIK kinases may be key effectors of metformin in the control of type 2 diabetes. How AMPK phosphorylation of p300 may regulate its many other downstream interacting transcription factors key to hepatic metabolism remains to be examined, although mice lacking LKB1 or AMPK in liver provide excellent tools for such future studies.



38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS

477

14-3-3 P

LKB1

CRTC2 cytoplasm

P P

SIK

P

AMPK

p300

P P ChREBP

nucleus

AMPK

14-3-3

P

FASN

PGC1α

CRTC2

CREB

PGC-1α

P FoxO3

HNF4a

P AREBP

PEPCK G6Pase

Inhibition of Gluconeogenesis

Inhibition of Lipogenesis

Figure 38.3  LKB1 and AMPK‐mediated control of metabolic transcriptional programs. AMPK and its related family members SIK1 and SIK2 (not shown) phosphorylate a common set of substrates including the CREB coactivator CRTC2 (CREB‐regulated transcriptional coactivator 2), previously known as TORC2 (transducer of regulated CREB 2) and the histone acetyltransferase p300. Phosphorylation of CRTC2 creates a 14‐3‐3 docking site which then results in 14‐3‐3 mediated nuclear export of CRTC2. This cytoplasmic sequestration of CRTC2 by AMPK or SIK kinases causes an inhibition of CREB‐dependent transcriptional targets including key mediators of gluconeogenesis like the PGC‐1α coactivator. In addition to suppressing PGC‐1a mRNA expression, AMPK also directly phosphorylates a handful of transcription factors (FOXO3, HNF4a, and AREBP) that directly bind to the promoters of the two key gluconeogenic enzymes PEPCK and G6Pase. In addition to these gluconeogenesis regulators, AMPK also is reported to phosphorylate Chrebp, a key lipogenic transcription factor that controls levels of fatty acid synthase (FASN), acetyl‐CoA carboxylase 1 (ACC1), and L‐pyruvate kinase mRNA. Table 38.1  Mouse models of AMPK/LKB1 function in liver Mouse model AMPKα1 knockout AMPKα2 knockout

Metabolic phenotype

None Glucose uptake in muscle Hyperglycemia, low insulin AMPKα2 liver knockout Hyperglycemia Glucose intolerance Hyperlipidemia LKB1 liver knockout Hyperglycemia Glucose intolerance Hyperlipidemia

Reference [48] [49] [48] [46] [30]

In addition to direct phosphorylation of coactivators, AMPK also phosphorylates the hepatocyte nuclear factor 4 alpha (HNF4α), which is a key transcription factor of the nuclear receptor superfamily which binds to the promoters of the two key gluconeogenic enzymes, phosphoenolpyruvate carboxykinase (PEPCK) and G6Pase. It is hypothesized that AMPK phosphorylation of HNF4α on residue Ser304 decreases the protein’s stability by interfering with its ability to dimerize and bind to DNA and may further promote its degradation [51, 52]. In addition to PEPCK and G6Pase, HNF4α controls gene expression of glucose transporter GLUT2 and glycolytic enzymes such as aldolase B and liver‐type pyruvate kinase (L‐PK), which are diminished when AMPK is activated by 5‐aminoimidazole‐4‐ carboxamide‐1‐β‐D‐ribofuranoside (AICAR) in hepatocytes [51].

It has been also shown that a subset of patients with MODY (maturity‐onset diabetes of the young) form of diabetes harbor mutations in HNF4α [53]. Interestingly, HNF4α liver specific knockout mice do not develop hyperglycemia unlike MODY patients, but do develop lipid accumulation, consistent with altered triglyceride levels in MODY patients [54]. These findings further reinforce the possible role of AMPK in lipid and glucose homeostasis through modulation of downstream transcriptional regulators such as HNF4α. In addition to HNF4α, in a 2006 report AMPK was shown to directly phosphorylate another transcription factor named the AICAR response‐element binding protein (AREBP) that directly binds the PEPCK promoter [55]. AMPK phosphorylation of AREBP on Ser470 reduces its ability to bind DNA and in turn lowers expression of PEPCK. Thus, like HNF4α, AMPK phosphorylation of AREBP controls its ability to bind to DNA and promote transcription of downstream target genes. In addition to effects on gluconeogenesis, AMPK activation is also known to inhibit hepatic lipogenesis. While some of that is due to acute effects on lipogenic enzymes as previously discussed, it is also known that AMPK activation leads to decreased transcription of key hepatic lipogenic enzymes, including fatty acid synthase (FASN) as well as ACC1. Two sequence–sequence transcription factors are known to coregulate many of these lipogenic enzymes: the sterol‐responsive binding protein

478

THE LIVER:  DOWNSTREAM TARGETS I: REGULATION OF ACUTE METABOLIC RESPONSE – ENZYME IN LIPOGENESIS

(SREBP‐1c) and the carbohydrate response binding protein (ChREBP). ChREBP is highly expressed in liver and has essential roles in glucose‐induced transcription of liver pyruvate kinase (L‐PK), in addition to its effects on lipogenic enzyme promoters. Phosphorylation by AMPK on Ser568 of ChREBP [56], promotes decreased DNA binding which causes a decrease in the transcription of glycolytic and lipogenic genes. It is known now that glucose metabolism can be repressed by fatty acids, which act as another source of energy when glucose needs to be preserved. This effect has been named the fatty acid “sparing effect” on glucose. Like ChREBP, SREBP‐1 is also directly regulated by AMPK [21, 57, 61]. Due to its rate‐limiting effects on lipogenic enzyme expression, SREBP‐1 has been linked to insulin resistance, dyslipidemia and type 2 diabetes [58, 59]. In one study, treatment of rat hepatocytes with AMPK activators such as metformin or AICAR led to suppressed SREBP‐1 mRNA expression, and also lowered hepatic mRNA levels for SREPB‐1 controlled genes FAS and S14 [21]. This suggests that AMPK activation through metformin can inhibit the expression of lipogenic genes. Indeed, metformin treatment or overexpression of an activated allele of AMPK was found to be sufficient to reduce triglyceride content in insulin resistant HepG2 cells [60]. Mice lacking hepatic AMPK function due to liver‐specific LKB1 deletion show elevated SREBP1 and SREBP1 target genes resulting in lipid accumulation and hepatic steatosis [37]. In 2011, Srebp1 was found to be directly phosphorylated by AMPK and its highly related kinase SIK1, both of which appear to suppress activation of full‐length Srebp1a in the ER membrane, preventing appearance of Srebp1c in the nucleus [61]. While the activity of the SREBP proteins is primarily regulated through intracellular concentrations of unsaturated fatty acids and cholesterol, AMPK phosphorylation of Ser372 on SREBP1 may also inhibit its activity by preventing proteolytic processing and thus transcriptional activity. Additional studies are needed however to truly sort out the requirement and roles for AMPK‐dependent phosphorylation of Srebp1 in vivo. As seen from the above‐mentioned studies, AMPK plays a major role in the gluconeogenic transcriptional program through p300, CRTC2, HNF4α, and AREBP, as well as the lipogenic transcriptional program through ChREBP and SREBP‐1. Finally, studies have hinted that AMPK may play a more global role in transcription through direct phosphorylation and regulation of transcriptional repressors histone deacetylases class IIa (HDACs IIa) enzymes [62] and transcriptional activator histone acetyltransferase p300 [45]. It has been shown that AMPK can phosphorylate class IIa HDAC5 on residues Ser259 and Ser498 in human primary myotubes, promoting 14‐3‐3 binding and dissociation from DNA binding transcription partner myocyte enhancer factor‐2 (MEF2), which in turn allows expression of downstream target genes [62]. Although direct regulation of class IIa HDACs by AMPK has not yet been implicated in liver metabolism, this provides an attractive postulate where AMPK can simultaneously control multiple downstream transcriptional events in response to upstream metabolic stresses. Such analysis will not happen without challenges, since it has been shown that multiple upstream kinases are capable of phosphorylating the same critical residues of HDAC 5

[63]. Indeed, we previously demonstrated that AMPK‐related kinases downstream of LKB1 can phosphorylate HDAC4, HDAC5, and HDAC7 in mouse liver, which suppresses their ability to enhance FOXO‐dependent gluconeogenesis [64].

Downstream targets III: regulation of cell growth and insulin signaling via mTOR Environmental factors cue cells to cease growing and dividing when conditions are unfavorable. These mechanisms are conserved from the smallest and simplest of eukaryotes to the most complex ones. When nutrients are scarce, cellular energy sensor AMPK gets activated and inhibits energy demanding processes such as protein synthesis and cell growth. One of the ways AMPK accomplishes that task is by negatively regulating the mechanistic target of rapamycin (mTOR) pathway. mTOR is a serine/threonine kinase highly conserved in all eukaryotes and a central regulator of cell growth. Whereas AMPK is active under nutrient‐poor conditions and inactive under nutrient‐rich conditions, mTOR is activated in the inverse pattern. In higher eukaryotes, mTOR activation requires positive signals from both nutrients (glucose, amino acids) and growth factors. mTOR, like its budding yeast orthologs, is found in two biochemically and functionally discrete signaling complexes [65]. In mammals, the mTORC1 complex is composed of four known subunits: raptor (regulatory associated protein of mTOR), PRAS40, mLST8, and mTOR. The mTORC2 complex contains rictor (rapamycin insensitive companion of mTOR), mSIN1, PRR5/ Protor, mLST8, and mTOR [66]. Signaling from mTOR complex 1 (mTORC1) is nutrient‐sensitive, acutely inhibited by the bacterial macrolide rapamycin, and controls cell growth, angiogenesis, and metabolism. In contrast, mTORC2 is not sensitive to nutrients, nor acutely inhibited by rapamycin, and its known substrates include the hydrophobic motif phosphorylation sites in AGC kinases including Akt and PKC family members. Downstream of the AMPK‐ and rapamycin‐sensitive raptor‐ mTOR (mTORC1) complex are its two well‐characterized substrates: 4EBP1 and the p70 ribosomal S6 kinase. Phosphorylation of 4EBP‐1 by mTORC1 suppresses its ability to bind and inhibit the translation initiation factor eIF4E. mTORC1 mediates phosphorylation of S6K at a Thr residue in a hydrophobic motif at the C‐terminus of the kinase domain. A specific motif (TOS motif) found in both 4EBP1 and S6K was shown to mediate direct binding of these proteins to raptor allowing them to be phosphorylated in the mTORC1 complex. Mechanistic details of how mTORC1 regulates the assembly of translational initiation complexes via a number of ordered phosphorylation events were recently discovered [67]. mTORC1‐dependent translation is known to control a number of specific cell growth regulators, including cyclin D1, the HIF‐1α transcriptional factor and c‐ myc, which in turn promote processes including cell cycle progression, cell growth, glycolysis, and angiogenesis, all contributing to enhanced tumorigenesis [66]. Upstream components of the mTORC1 complex were initially discovered through classical cancer genetics. The TSC2 tumor suppressor tuberin and its obligate binding partner hamartin (TSC1), are mutated in a familial tumor syndrome called tuberous sclerosis complex (TSC). TSC patients are predisposed to widespread benign tumors termed hamartomas in kidney, lung,

38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS

brain, and skin. Genetic studies in Drosophila and mammalian cells identified the TSC tumor suppressors as critical upstream inhibitors of the mTORC1 complex. TSC2 (also known as tuberin) contains a GTPase‐activating protein (GAP) domain at its C‐terminus that inactivates the small Ras‐like GTPase Rheb, which has been shown to associate with and directly activate the mTORC1 complex in vitro [68]. Loss of TSC1 or TSC2 therefore leads to hyperactivation of mTORC1. Phosphorylation of TSC1 and TSC2 serves as an integration point for a wide variety of environmental signals that regulate mTORC1 [69]. One of the key activators of the mTORC1 pathway is PI3‐kinase, which plays a key role in promoting cell growth and insulin‐mediated effects on metabolism. PI3‐kinase activates the serine/threonine kinase Akt which directly phosphorylates and inactivates both TSC2 and an inhibitor of the mTORC1 complex named PRAS40 [68, 70]. In addition to these growth factor cues that activate mTORC1, the complex is rapidly inactivated by a wide variety of cell stresses, thereby ensuring that cells do not continue to grow under unfavorable conditions. One of the unique aspects of the mTORC1 complex is that unlike many of the aforementioned growth factor activated kinases, it is dependent on nutrient availability for its kinase activity. Withdrawal of glucose, amino acids, or oxygen leads to rapid suppression of mTORC1 activity even in the presence of full growth factors [69]. Upon LKB1‐ and AMP‐dependent activation of AMPK by nutrient loss, AMPK directly phosphorylates the TSC2 tumor suppressor on conserved serine sites distinct from those targeted by other kinases, which constitutes one mechanism by which glucose and oxygen control mTORC1 activation [71–74]. Interestingly, TSC2 orthologs are absent from lower eukaryotes such as S. cerevisiae and Caenorhabditis elegans and mammalian cells lacking TSC2 still remain partially sensitive to AMPK activation, indicating that there may be an alternative and more ancient

479

backup mechanism which allows AMPK to inhibit cell growth and proliferation through the mTORC1 pathway. Collectively, these observations prompted the discovery of yet another novel mechanism of inhibition. The mTOR kinase exists in complex consisting of mLST8/Gbl, PRAS40, and the scaffold protein Raptor. In a recent study, it was shown that AMPK is able to directly phosphorylate Raptor on two conserved residues Ser722 and Ser792, which in turn induces binding to 14‐3‐3 proteins and inactivation the mTORC1 complex [75]. Taken together with previous studies, these findings indicate that AMPK directly phosphorylates both TSC2 and raptor to inhibit mTORC1 activity by a dual‐pronged mechanism (see Figure 38.4). Importantly, metformin treatment of mice leads to robust phosphorylation of raptor Ser792 in murine liver, an effect which is ablated in the LKB1‐liver specific knockout mice [75]. LKB1‐liver specific knockout mice which lack AMPK activity in liver exhibit hyperactivation of mTORC1 signaling in the liver including increased phosphorylation of S6K1 and 4EBP1 under ad libitum fed conditions [37]. In addition, hormones that activate AMPK in liver including glucagon [76] and adiponectin [77] have been reported to suppress mTORC1 signaling. Very recently, our laboratory collaborating with the lab of Brendan Manning demonstrated that AMPK is genetically required in primary hepatocytes and in the livers of mice for metformin to suppress mTORC1 signaling [78]. Using liver‐specific knockouts of AMPKα1 and AMPKα2 we found that metformin was completely unable to suppress mTORC1 after metformin treatment in mice. However, in cell culture, at higher doses and at later timepoints, metformin induced suppression of mTORC1 signaling suggestive of a transcriptional response which may be due to activation of the mito ER stress pathway and ATF4‐dependent activation of REDD1 mRNAs and other mTORC1 inhibitor mRNAs [79].

Insulin R IRS1/2 PI3K

mTOR rictor

LKB1

PIP3

AMP

FOXO GSK3

TSC2 TSC1

AMPK

raptor

-3

-3

Rheb

-3

rapamycin

14

-3

metformin

PTEN

Akt

14



PRAS40

mTOR

Srebp1

4EBP1

S6K

EIF4E

S6

Figure 38.4  AMPK regulates the mechanistic target of rapamycin (mTOR) pathway to control cell growth and insulin sensitivity. AMPK directly phosphorylates the TSC2 (tuberous sclerosis 2) tumor suppressor and the key mTOR scaffold subunit raptor to inhibit the activity of the mTOR‐raptor kinase complex (mTORC1) toward its downstream substrates 4EBP1 and S6K1. PI3K‐activity activates mTORC1 by Akt‐dependent phosphorylation of TSC2 and the mTORC1 inhibitor PRAS40. Under conditions of nutrient excess, overstimulated mTORC1 and its substrate S6K both directly phosphorylate the insulin receptor substrate 1 and 2 proteins, resulting in their degradation. Thus too much mTORC1 activity attenuates insulin signaling, leading to cellular insulin resistance. AMPK reverses this resistance by inactivating mTORC1 and can also directly phosphorylate IRS1 itself. mTORC1 has also been recently shown to play a role in the control of lipogenesis through regulation of SREBP1.

480

THE LIVER:  DOWNSTREAM TARGETS I: REGULATION OF ACUTE METABOLIC RESPONSE – ENZYME IN LIPOGENESIS

Beyond effects on cell growth, mTOR also has effects on lipid metabolism that may be particularly important in liver. One key regulator of lipogenesis is the aforementioned SREBP‐1 transcription factor. Recently, mTORC1 signaling was shown to be required for nuclear accumulation of SREBP1 and the induction of SREBP1 target genes [80]. Consistent with previous results with metformin, treatment with the AMPK activators AICAR and 2DG, or the mTORC1 inhibitor rapamycin resulted in suppression of nuclear SREBP1 accumulation [80]. In future studies, it will be important to define how much of the lipid‐ reducing effects of AMPK are due to direct phosphorylation of lipogenic enzymes such as acetyl‐CoA carboxylase (ACC), and how much are due to effects on SREBP‐1 or Chrebp‐dependent transcription through effects of AMPK on mTORC1, versus direct phosphorylation of Srebp‐1 and Chrebp by AMPK. One final aspect of liver physiology that may be under control of the AMPK–mTOR axis is insulin signaling. A major route by which excess nutrients downregulate insulin signaling leading to cellular insulin resistance is through hyperactivation of the mTORC1 complex. Excess glucose leads to hyperactivation of mTOR via suppression of AMPK, and excess fats and excess amino acids also act to hyperactivate mTORC1 [81]. The mTOR/ Raptor complex, along with its key downstream substrate S6K, have been shown to directly phosphorylate the insulin receptor substrate‐1 and 2 (IRS1 and IRS2), leading to their proteasome degradation. The same is observed under conditions of hyperinsulinemia, as insulin signaling itself also leads to increases in mTORC1 activity as described earlier. The net effect is a negative feedback loop whereby too much mTOR/raptor activity leads to hyperphosphorylation of IRS1/IRS2 and suppression of PI3‐kinase and Akt signaling downstream of the insulin receptor [81]. This nutrient induced hyperactivation of mTORC1 and consequent downregulation of Akt signaling is observed in a cultured cell systems and also in peripheral tissues of mice on a high fat diet. Illustrating its importance downstream of mTORC1 in the IRS1/IRS2 inhibition, this effect is lost in peripheral tissues from an S6K1−/− mouse [82]. As one of the key direct substrates of AMPK is raptor and TSC2, AMPK activation leads to a inhibition of mTORC1 and its phosphorylation of IRS1 Ser636/639 and negative feedback loop on PI3‐kinase/Akt signaling. Exogenous LKB1 expression in HEK293 cells and metformin treatment of human hepatocellular carcinoma HepG2 cells can suppress phosphorylation of IRS‐1 on Ser636/639 and induce Akt phosphorylation [83]. In addition to suppressing phosphorylation of IRS‐1 by mTORC1, AMPK has also been shown to directly phosphorylate IRS1 itself on S789, though the functional consequence of that phosphorylation event on IRS function remains uncertain [83, 84]. Taken altogether, these studies demonstrate a mechanism by which AMPK activation can promote PI3‐kinase/Akt activity while simultaneously reducing mTORC1 activity. This provides cells with a negative feedback switch that integrates upstream signals from both nutrients and growth factors and allows the cells to maintain energy homeostasis and insulin sensitivity.

Downstream targets IV: autophagy and mitophagy Autophagy is a cellular process in which proteins, organelles, and other macromolecules are delivered to the lysosomes for

degradation. It is a process used by cells both for normal turnover and for the generation of nutrients in response to energy shortages. AMPK potently promotes autophagy through several mechanisms. AMPK phosphorylates and activates ULK1 (unc‐51‐like autophagy‐activating kinase 1), which triggers the initiation of the autophagic cascade [85–87]. Importantly, mTOR strongly suppresses autophagy, in part by directly phosphorylating and inhibiting ULK1 [86]. Accordingly, AMPK promotes autophagy not only by direct activation of ULK1 but also by negatively regulating mTORC1 and blocking its inhibitory effect on ULK1. Thus, ULK1 is yet another node at which AMPK and mTOR regulate a specific metabolic process in opposing fashion. AMPK also stimulates autophagy initiation by differential regulation of VPS34 (vacuolar protein sorting 34) containing complexes [88], which are important for the initiation and formation of autophagosomes. AMPK was reported to directly phosphorylate and inhibit VPS34 in non‐autophagic complexes that do not contain autophagy adaptor proteins, while enhancing VPS34 activity in pro‐autophagic complexes that contain Beclin‐1 by directly phosphorylating Beclin‐1 [88]. In this way, AMPK presumably suppresses nonessential vesicle trafficking in favor of membrane trafficking into the autophagy pathway during nutrient starvation. Given that both AMPK and ULK1 have been reported to directly phosphorylate distinct sites in both Beclin‐1 and Vps34, much remains to be clarified about the temporal and spatial control of autophagy initiation in response to different stresses. In addition, AMPK and ULK1 have also both been reported to phosphorylate and control the localization of Atg9, a transmembrane protein involved in early autophagosome formation [87, 89, 90]. AMPK has also recently been shown to promote autophagy through transcriptional mechanisms, via regulation of Tfeb (transcription factor EB), a master transcriptional regulator of lysosomal genes and autophagy. Although no direct link between AMPK and Tfeb has been reported, AMPK can activate Tfeb through inhibition of mTORC1, thus blocking the ability of mTOR to phosphorylate and translocate Tfeb out of the nucleus [91]. Furthermore, through phosphorylation and activation of the transcription factor FOXO3a (Forkhead box O3) [92], AMPK has been reported to increase the levels of CARM1 (coactivator‐associated arginine methyltransferase 1), an important cofactor for Tfeb transcription [93]. In addition to general autophagy, several lines of evidence indicate that AMPK promotes mitophagy, the process of degradation of defective mitochondria. Indeed, activation of ULK1 by AMPK was shown to be required for proper removal of damaged mitochondria via mitophagy, though the details of how ULK1 regulates mitophagy are not fully resolved [85]. A necessary step preceding removal of damaged mitochondria is the fragmentation of mitochondria in response to mitochondrial insults, in order to separate and target damaged mitochondrial fragments to turnover via the mitophagy pathway. This highly conserved process is known as mitochondrial fission. Recently, a novel mechanism was elucidated by which AMPK promotes mitochondrial fission [94]. In this study, AMPK was demonstrated to induce mitochondrial fission during energy stress through direct phosphorylation of MFF (mitochondrial fission factor), which then serves as a receptor for DRP1 (dynamin‐ related protein 1), the enzyme that catalyzes mitochondrial



38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS

fission [94]. Once at the mitochondria, DRP1 splits damaged mitochondria into smaller fragments that are presumably more efficiently cleared by autophagosomes. Furthermore, AMPK activates PGC1α (peroxisome proliferator‐activated receptor gamma, coactivator 1α), a master regulator of mitochondrial biogenesis, reportedly via direct phosphorylation of PGC1α [95] but also by promoting NAD+‐dependent activation of PGC1α by Sirt1 (sirtuin 1) [96]. Interestingly, Tfeb, similar to its family member Tfe3, was recently reported to drive mitochondrial biogenesis as well [97, 98], which offers the possibility that activation of Tfeb, or Tfe3, might be yet another mechanism by which AMPK can promote the regeneration of mitochondria. In all, AMPK coordinates mitochondrial fission and mitophagy in the acute response to mitochondrial insults, and after sustained energy stress, AMPK promotes transcriptional induction of mitochondrial biogenesis. In this fashion, AMPK serves as a central mediator of mitochondrial quality, ensuring metabolic efficiency in cells and tissues.

Downstream targets V: polarization of hepatocytes The predominant upstream activating kinase for AMPK, LKB1, is a well‐conserved key regulator of cell polarity and trafficking, and metabolism, due at least in part to its ability to phosphorylate and activate the MARK/Par1 subfamily of AMPK related kinases [37]. In an intestinal cell line in culture, genetic activation of LKB1 induced apical–basolateral polarity in the absence of cell–cell or cell–matrix cues [99], though evidence of which of the 14 AMPK‐related kinase family were involved has not been determined. Nonetheless, AMPK itself has been connected to cell polarization, particularly in hepatocytes and its activation enhanced bile canalicular formation, whereas inhibition resulted in loss of polarity and mislocalization of apical transporters [100–102]. In MDCK cells, AMPK regulates tight junction assembly and disassembly in response to calcium depletion [103–104]. LKB1‐AMPK activation phosphorylates the core tight junction protein cingulin on Ser137 [105] while at the same time stabilizing existing cell junctions to maintain cell polarity through phosphorylation of Gα‐interacting vesicle‐ associated protein (GIV/Girdin) on Ser245 [106]. Interestingly, GIV/Girdin was previously reported as an Akt substrate [107], reinforcing the idea that AMPK and mTORC1/Akt converge on a common small set of effector proteins (ULK1, MFF, Srebp1, Foxo, Girdin) central to metabolism and growth [108]. Additional studies are needed to delineate the roles of AMPK and mTORC1/Akt in hepatocyte polarity and liver zonation in vivo.

Therapeutics and future perspectives AMPK has received a lot of attention as a potential target for treating diseases associated with metabolic perturbation. This includes diabetes, obesity, and fatty liver diseases and also cancer, which is often associated with changes in metabolism. The type 2 diabetes drug metformin had been used for decades when a study showed that its mechanism of action involved the activation of AMPK in hepatocytes [21]. Mechanistically, metformin induces energy stress through inhibition of complex I of the

481

respiratory chain in mitochondria [109]. This leads to a change in the ATP‐to‐AMP ratio and canonical AMPK activation. There has been extensive research to evaluate the contribution of AMPK to the effect of metformin on circulating glucose and lipids. AMPK phosphorylation of ACCs has been proposed as a master contributor to the changes in lipid synthesis that are induced by metformin, which in turn modulates insulin sensitivity and glucose uptake in muscle. One key piece of evidence in favor of this hypothesis came from the generation of a mouse knock‐in mutant lacking the AMPK site on both ACC1 and ACC2 [110]. This mouse model revealed that these phosphorylation events mediate the insulin‐sensitizing effect of metformin, thus establishing AMPK as a relevant target in the action of metformin. Despite controversy [111], AMPK is widely viewed as an essential component of the action of metformin at physiological concentrations [78, 112]. Given the role of the AMPK–ACC pathway in regulating fatty acid synthesis, AMPK activation is also an attractive treatment option for conditions associated with increased fatty acid production, such as non‐alcoholic fatty liver disease (NAFLD) [113]. In addition to diabetes, retrospective studies have revealed that patients taking metformin had a decreased occurrence of cancer [114]. However, whether direct activation of AMPK would be sufficient to replicate the beneficial effect of metformin was not known until recent advances in the generation of potent and specific small‐molecule activators of AMPK. Most notably, two recent studies have revealed that direct AMPK activation indeed improves symptoms of type 2 diabetes in multiple animal models. One study revealed that a pan‐specific AMPK activator improved diabetes symptoms in several animal models, including rodents and monkeys [115]. In another study, a direct AMPK activator increased glucose uptake in muscle and reduced blood glucose in type 2 diabetes models [116]. Interestingly, inactivation of AMPK in the liver had no effect on the efficacy of the drug, whereas muscle‐specific deletion of AMPK abolished the effect of the drug, establishing muscle AMPK as a key therapeutic target in type 2 diabetes [116]. Although hepatic AMPK may not be as central as originally hypothesized for metformin effects on glucose homeostasis in type 2 diabetes, hepatic AMPK is still a very attractive target in the emerging field of therapeutics for NASH and NAFLD [113]. Indeed in every system from C. elegans to mammals, all studies concur that AMPK activation suppresses de novo lipogenesis and lipid accumulation. While much has been learned, there is still even more yet to be learned about this ancient energy sensor and central metabolic regulator.

REFERENCES 1. Shaw, R.J., Kosmatka, M., Bardeesy, N. et al. The tumor suppressor LKB1 kinase directly activates AMP‐activated kinase and regulates apoptosis in response to energy stress. Proc Natl Acad Sci USA, 2004;101:3329–35. 2. Kahn, B.B., Alquier, T., Carling, D., and Hardie, D.G. AMP‐activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism. Cell Metab, 2005;1:15–25. 3. Carlson, C.A. and Kim, K.H. Regulation of hepatic acetyl coenzyme A carbozylase by phosphorylation and dephosphorylation. J Biol Chem, 1973;248:378–80. 4. Mitchelhill, K.I. et al. Mammalian AMP‐activated protein kinase shares structural and functional homology with the catalytic domain of yeast Snf1 protein kinase. J Biol Chem, 1994;269:2361–4.

482

THE LIVER:  REFERENCES

5. Woods, A., Munday, M.R., Scott, J., Yang, X., Carlson, M., and Carling, D. Yeast SNF1 is functionally related to mammalian AMP‐activated protein kinase and regulates acetyl‐CoA carboxylase in vivo. J Biol Chem, 1994;269:19509–15. 6. Sanders, M.J., Grondin, P.O., Hegarty, B.D., Snowden, M.A., and Carling, D. Investigating the mechanism for AMP activation of the AMP‐activated protein kinase cascade. Biochem J, 2007;403:139–48. 7. Cheung, P.C., Salt, I.P., Davies, S.P., Hardie, D.G., and Carling, D. Characterization of AMP‐activated protein kinase gamma‐subunit isoforms and their role in AMP binding. Biochem J, 2000;346(3):659–69. 8. Amodeo, G.A., Rudolph, M.J., and Tong, L. Crystal structure of the heterotrimer core of Saccharomyces cerevisiae AMPK homologue SNF1. Nature, 2007;449:492–5. 9. Xiao, B., Heath, R., Saiu, P. et al. Structural basis for AMP binding to mammalian AMP‐activated protein kinase. Nature, 2007;449:496–500. 10. Calabrese, M.F., Rajamohan, F., Harris, M.S. et al. Structural basis for AMPK activation: natural and synthetic ligands regulate kinase activity from opposite poles by different molecular mechanisms. Fold Des, 2014;22:1161–72. 11. Chen, L., Wang, J., Zhang, Y.‐Y. et al. AMP‐activated protein kinase undergoes nucleotide‐dependent conformational changes. Nat Struct Mol Biol, 2012;19:716–8. 12. Chen, L., Xin, F.‐J., Wang, J. et al. Conserved regulatory elements in AMPK. Nature, 2013;498:E8–10. 13. Li, X., Wang, L., Zhou, X.E. et al. Structural basis of AMPK regulation by adenine nucleotides and glycogen. Cell Res, 2015.25:50–66. 14. Xiao, B., Sanders, M.J., Underwood, E. et  al. Structure of mammalian AMPK and its regulation by ADP. Nature, 2011;472:230–3. 15. Xiao, B., Sanders, M.J., Carmena, D. et al. Structural basis of AMPK regulation by small molecule activators. Nat Commun, 2013;4:3017. 16. Xin, F.‐J., Wang, J., Zhao, R.‐Q., Wang, Z.‐X., and Wu, J.‐W. Coordinated regulation of AMPK activity by multiple elements in the α‐subunit. Cell Res, 2013;23:1237–40. 17. McBride, A., Ghilagaber, S., Nikolaev, A., and Hardie, D.G. The glycogen‐ binding domain on the AMPKβ subunit allows the kinase to act as a glycogen sensor. Cell Metab, 2009;9:23–34. 18. Yamauchi, T., Kamon, J., Minokoshi, Y. et al. Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat Med, 2002;8:1288–95. 19. Hardie, D.G. AMP‐activated protein kinase as a drug target. Annu Rev Pharmacol Toxicol, 2007;47:185–210. 20. Sakamoto, K., McCarthy, A., Smith, D. et al. Deficiency of LKB1 in skeletal muscle prevents AMPK activation and glucose uptake during contraction. EMBO J, 2005;24:1810–20. 21. Zhou, G., Myers, R., Li, Y. et al. Role of AMP‐activated protein kinase in mechanism of metformin action. J Clin Invest, 2001;108:1167–74. 22. Fryer, L.G., Parbu‐Patel, A., and Carling, D. The anti‐diabetic drugs rosiglitazone and metformin stimulate AMP‐activated protein kinase through distinct signaling pathways. J Biol Chem, 2002;277:25226–32. 23. Saha, A.K., Avilucea, P.R., Ye, J.M., Assifi, M.M., Kraegen, E.W., and Ruderman, N.B. Pioglitazone treatment activates AMP‐activated protein kinase in rat liver and adipose tissue in vivo. Biochem Biophys Res Commun, 2004;314:580–5. 24. Witters, L.A. The blooming of the French lilac. J Clin Invest, 2001;108:1105–7. 25. Tan, M.J., Ye, J.M., Turner, N. et al. Antidiabetic activities of triterpenoids isolated from bitter melon associated with activation of the AMPK pathway. Chem Biol, 2008;15:263–73. 26. Baur, J.A., Pearson, K.J., Price, N.L. et al. Resveratrol improves health and survival of mice on a high‐calorie diet. Nature, 2006;444:337–42. 27. Ingebritsen, T.S., Lee, H.S., Parker, R.A., and Gibson, D.M. Reversible modulation of the activities of both liver microsomal hydroxymethylglutaryl coenzyme A reductase and its inactivating enzyme. Evidence for regulation by phosphorylation‐dephosphorylation. Biochem Biophys Res Commun, 1978;81:1268–77. 28. Hawley, S.A., Davison, M., Woods, A. et al. Characterization of the AMP‐ activated protein kinase from rat liver and identification of threonine 172 as the major site at which it phosphorylates AMP‐activated protein kinase. J Biol Chem, 1996;271:27879–87. 29. Hong, S.P., Leiper, F.C., Woods, A., Carling, D., and Carlson, M. Activation of yeast Snf1 and mammalian AMP‐activated protein kinase by upstream kinases. Proc Natl Acad Sci USA, 2003;100:8839–43. 30. Sutherland, C.M., Hawley, S.A., McCartney, R.R. et  al. Elm1p is one of three upstream kinases for the Saccharomyces cerevisiae SNF1 complex. Curr Biol, 2003;13:1299–305.

31. Hawley, S.A., Boudeau, J., Reid, J.L. et al. Complexes between the LKB1 tumor suppressor, STRADalpha/beta and MO25alpha/beta are upstream kinases in the AMP‐activated protein kinase cascade. J Biol, 2003;2:28. 32. Woods, A., Johnstone, S.R., Dickerson, K. et al. LKB1 is the upstream kinase in the AMP‐activated protein kinase cascade. Curr Biol, 2003;13:2004–8. 33. Hawley, S.A., Pan, D.A., Mustard, K.J. et al. Calmodulin‐dependent protein kinase kinase‐beta is an alternative upstream kinase for AMP‐activated protein kinase. Cell Metab, 2005;2:9–19. 34. Hurley, R.L., Anderson, K.A., Franzone, J.M., Kemp, B.E., Means, A.R., and Witters, L.A. The Ca2+/calmodulin‐dependent protein kinase kinases are AMP‐activated protein kinase kinases. J Biol Chem, 2005;280:29060–6. 35. Woods, A., Dickerson, K., Heath, R. et al. C(Ca2+)/calmodulin‐dependent protein kinase kinase‐beta acts upstream of AMP‐activated protein kinase in mammalian cells. Cell Metab, 2005;2:21–33. 36. Hemminki, A., Markie, D., Tomlinson, I. et  al. A serine/threonine kinase gene defective in Peutz‐Jeghers syndrome. Nature, 1998;391:184–7. 37. Shaw, R.J., Lamia, K.A., Vasquez, D. et al. The kinase LKB1 mediates glucose homeostasis in liver and therapeutic effects of metformin. Science, 2005;310:1642–6. 38. Lizcano, J.M., Goransson, O., Toth, R. et al. LKB1 is a master kinase that activates 13 kinases of the AMPK subfamily, including MARK/PAR‐1. EMBO J, 2004;23:833–43. 39. Al‐Hakim, A.K., Goransson, O., Deak, M. et  al. 14‐3‐3 cooperates with LKB1 to regulate the activity and localization of QSK and SIK. J Cell Sci, 2005;118:5661–73. 40. Anderson, K.A., Ribar, T.J., Lin, F. et al. Hypothalamic CaMKK2 contributes to the regulation of energy balance. Cell Metab, 2008;7:377–88. 41. Carling, D., Zammit, V.A., and Hardie, D.G. A common bicyclic protein kinase cascade inactivates the regulatory enzymes of fatty acid and cholesterol biosynthesis. FEBS Lett, 1987;223:217–22. 42. Saggerson, D. Malonyl‐CoA a key signaling molecule in mammalian cells. Annu Rev Nutr, 2008;28:253–72. 43. Zang, M., Xu, S., Maitland‐Toolan, K.A. et al. Polyphenols stimulate AMP‐ activated protein kinase, lower lipids, and inhibit accelerated atherosclerosis in diabetic LDL receptor‐deficient mice. Diabetes, 2006;55:2180–91. 44. Sato, R., Goldstein, J.L., and Brown, M.S. Replacement of serine‐871 of hamster 3‐hydroxy‐3‐methylglutaryl‐CoA reductase prevents phosphorylation by AMP‐activated kinase and blocks inhibition of sterol synthesis induced by ATP depletion. Proc Natl Acad Sci USA, 1993;90:9261–5. 45. Yang, W., Hong, Y.H., Shen, X.Q., Frankowski, C., Camp, H.S., and Leff, T. Regulation of transcription by AMP‐activated protein kinase: phosphorylation of p300 blocks its interaction with nuclear receptors. J Biol Chem, 2001;276:38341–4. 46. Koo, S.H. et al. The CREB co‐activator TORC2 is a key regulator of fasting glucose metabolism. Nature, 2005;437:1109–11. 47. Liu, Y., Dentin, R., Chen, D. et al. A fasting inducible switch modulates gluconeogenesis via activator/coactivator exchange. Nature, 2008;456:269–73. 48. Bittinger, M.A., McWhinnie, E., Meltzer, J. et  al. Activation of cAMP response element‐mediated gene expression by regulated nuclear transport of TORC proteins. Curr Biol, 2004;14:2156–61. 49. Screaton, R.A., Conkright, M.D., Katoh, Y. et  al. The CREB coactivator TORC2 functions as a calcium‐ and cAMP‐sensitive coincidence detector. Cell, 2004;119:61–74. 50. Andreelli, F., Foretz, M., Knauf, C. et al. Liver adenosine monophosphate‐ activated kinase‐alpha2 catalytic subunit is a key target for the control of hepatic glucose production by adiponectin and leptin but not insulin. Endocrinology, 2006 147, 243241. 51. Leclerc, I., Lenzner, C., Gourdon, L., Vaulont, S., Kahn, A., and Viollet, B. Hepatocyte nuclear factor‐4a involved in type 1 maturity‐onset diabetes of the young is a novel target of AMP‐activated protein kinase. Diabetes, 2001;50. 52. Hong, Y.H., Varanasi, U.S., Yang, W., and Leff, T. AMP‐activated protein kinase regulates HNF4alpha transcriptional activity by inhibiting dimer formation and decreasing protein stability. J Biol Chem, 2003;278:27495–501. 53. Doria, A., Patti, M.E., and Kahn, C.R The emerging genetic architecture of type 2 diabetes. Cell Metab, 2008;8:186–200. 54. Hayhurst, G. P., Lee, Y.H., Lambert, G., Ward, J.M., and Gonzalez, F.J. Hepatocyte nuclear factor 4a (nuclear receptor 2a1) is essential for maintenance of hepatic gene expression and lipid homeostasis. Mol Cell Biol, 2001:1393–1403. 55. Inoue, E. and Yamauchi, J. AMP‐activated protein kinase regulates PEPCK gene expression by direct phosphorylation of a novel zinc finger transcription factor. Biochem Biophys Res Commun, 2006;351:793–9.



38:  CENTRAL REGULATOR OF GLUCOSE AND LIPID METABOLISM AND TARGET OF TYPE 2 DIABETES THERAPEUTICS 56. Kawaguchi, T., Osatomi, K., Yamashita, H., Kabashima T., and Uyeda, K. Mechanism of fatty acid “sparing” effect on glucose‐induced transcription: regulation of carbohydrate‐responsive element‐binding protein by AMP‐ activated protein kinase. J Biol Chem, 2002;277:3829–35. 57. Foretz, M., Ancellin, N., Andreelli, F. et al. Short‐term overexpression of a constitutively active form of AMP‐activated protein kinase in the liver leads to mild hypoglycemia and fatty liver. Diabetes, 2005;54:1331–9. 58. Shimomura, I. et  al. Decreased IRS‐2 and increased SREBP‐1c lead to mixed insulin resistance and sensitivity in livers of lipodystrophic and ob/ob mice. Mol Cell, 2000;6:77–86. 59. Kakuma, T. et al. Leptin, troglitazone, and the expression of sterol regulatory element binding proteins in liver and pancreatic islets. Proc Natl Acad Sci USA, 2000;97:8536–41. 60. Zang, M., Zuccollo, A., Hou, X. et  al. AMP‐activated protein kinase is required for the lipid‐lowering effect of metformin in insulin‐resistant human HepG2 cells. J Biol Chem, 2004;279:47898–905. 61. Li, Y., Xu, S., Mihaylova, M. et  al. AMPK phosphorylates and inhibits SREBP activity to attenuate hepatic steatosis and atherosclerosis in diet‐ induced insulin resistant mice. Cell Metab, 2011;13:376–88. 62. McGee, S.L., van Denderen, B.J., Howlett, K.F. et al. AMP‐activated protein kinase regulates GLUT4 transcription by phosphorylating histone deacetylase 5. Diabetes, 2008;57:860–7. 63. Chang, S., Bezprozvannaya, S., Li, S., and Olson, E.N. An expression screen reveals modulators of class II histone deacetylase phosphorylation. Proc Natl Acad Sci USA, 2005;102:8120–5. 64. Mihaylova, M.M., Vasquez, D.S., Ravnskjaer, K. et al. Class IIa histone deacetylases are hormone‐activated regulators of FOXO and mammalian glucose homeostasis. Cell, 2011;145:607–21. 65. Wullschleger, S., Loewith, R., and Hall, M.N. TOR signaling in growth and metabolism. Cell, 2006;124:471–84. 66. Guertin, D.A. and Sabatini, D.M. Defining the role of mTOR in cancer. Cancer Cell, 2007;12:9–22. 67. Holz, M.K., Ballif, B.A., Gygi, S.P., and Blenis, J. mTOR and S6K1 mediate assembly of the translation preinitiation complex through dynamic protein interchange and ordered phosphorylation events. Cell 2005;123:569–80. 68. Sancak, Y., Thoreen, C.C., Peterson, T.R. et al. PRAS40 is an insulin‐regulated inhibitor of the mTORC1 protein kinase. Mol Cell, 2007;25:903–15. 69. Shaw, R.J. and Cantley, L.C. Ras, PI(3)K and mTOR signalling controls tumour cell growth. Nature, 2006;441:424–30. 70. Vander Haar, E., Lee, S.I., Bandhakavi, S., Griffin, T.J., and Kim, D.H. Insulin signalling to mTOR mediated by the Akt/PKB substrate PRAS40. Nat Cell Biol, 2007;9:316–23. 71. Inoki, K., Zhu, T., and Guan, K.L. TSC2 mediates cellular energy response to control cell growth and survival. Cell, 2003;115:577–90. 72. Corradetti, M.N., Inoki, K., Bardeesy, N., DePinho, R.A., and Guan, K.L Regulation of the TSC pathway by LKB1: evidence of a molecular link between tuberous sclerosis complex and Peutz–Jeghers syndrome. Genes Dev, 2004;18:1533–8. 73. Shaw, R.J., Bardeesy, N., Manning, B.D. et al. The LKB1 tumor suppressor negatively regulates mTOR signaling. Cancer Cell, 2004;6:91–9. 74. Liu, L., Cash, T.P., Jones, R.G., Keith, B., Thompson, C.B., and Simon, M.C. Hypoxia‐induced energy stress regulates mRNA translation and cell growth. Mol Cell, 2006;21:521–31. 75. Gwinn, D.M., Shackelford, D.B., Egan, D.F. et al. AMPK phosphorylation of raptor mediates a metabolic checkpoint. Mol Cell, 2008;30:214–26. 76. Kimball, S.R., Siegfried, B.A., and Jefferson, L.S. Glucagon represses signaling through the mammalian target of rapamycin in rat liver by activating AMP‐activated protein kinase. J Biol Chem, 2004;279:54103–9. 77. Wang, C., Mao, X., Wang, L. et al. Adiponectin sensitizes insulin signaling by reducing p70 S6 kinase‐mediated serine phosphorylation of IRS‐1. J Biol Chem, 2007;282:7991–6. 78. Howell, J.J., Hellberg, K., Turner, M. et  al. Metformin inhibits hepatic mTORC1 signaling via dose‐dependent mechanisms involving AMPK and the TSC complex. Cell Metab, 2017;25:463–71. 79. Kimball, S.R. and Jefferson, L.S. Induction of REDD1 gene expression in the liver in response to endoplasmic reticulum stress is mediated through a PERK eIF2α phosphorylation, ATF4‐dependent cascade. Biochem Biophys Res Commun, 2012;427:485–9. 80. Porstmann, T., Santos, C.R., Griffiths, B. et al. SREBP activity is regulated by mTORC1 and contributes to Akt‐dependent cell growth. Cell Metab, 2008;8:224–36.

483

81. Manning, B.D. Balancing Akt with S6K: implications for both metabolic diseases and tumorigenesis. J Cell Biol, 2004;167:399–403. 82. Um, S.H., Frigerio, F., Watanabe, M. et  al. Absence of S6K1 protects against age‐ and diet‐induced obesity while enhancing insulin sensitivity. Nature, 2004;431:200–5. 83. Tzatsos, A. and Kandror, K.V. Nutrients suppress phosphatidylinositol 3‐ kinase/Akt signaling via raptor‐dependent mTOR‐mediated insulin receptor substrate 1 phosphorylation. Mol Cell Biol, 2006;26:63–76. 84. Jakobsen, S.N., Hardie, D.G., Morrice, N., and Tornqvist, H.E 5’‐AMP‐ activated protein kinase phosphorylates IRS‐1 on Ser‐789 in mouse C2C12 myotubes in response to 5‐aminoimidazole‐4‐carboxamide riboside. J Biol Chem, 2001;276:46912–6. 85. Egan, D.F., Shackelford, D.B., Mihaylova, M.M. et al. Phosphorylation of ULK1 (hATG1) by AMP‐activated protein kinase connects energy sensing to mitophagy. Science, 2011;331:456–61. 86. Kim, J., Kundu, M., Viollet, B., and Guan, K.‐L. AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nat Cell Biol, 2011;13:132–41. 87. Mack, H.I.D., Zheng, B., Asara, J.M., and Thomas, S.M. AMPK‐dependent phosphorylation of ULK1 regulates ATG9 localization. Autophagy, 2012;8:1197–1214. 88. Kim, J., Kim, Y.C., Fang, C. et al. Differential regulation of distinct Vps34 complexes by AMPK in nutrient stress and autophagy. Cell, 2013;152:290–303. 89. Weerasekara, V.K., Panek, D.J., Broadbent, D.G. et  al. Metabolic‐stress‐ induced rearrangement of the 14–3‐3ζ interactome promotes autophagy via a ULK1‐ and AMPK‐regulated 14–3‐3ζ interaction with phosphorylated Atg9. Mol Cell Biol, 2014;34:4379–88. 90. Zhou, C., Ma, K., Gao, R. et al. Regulation of mATG9 trafficking by Src‐ and ULK1‐mediated phosphorylation in basal and starvation‐induced autophagy. Cell Res, 2017;27:184–201. 91. Young, N.P., Kamireddy, A., Van Nostrand, J.L. et al. AMPK governs lineage specification through Tfeb‐dependent regulation of lysosomes. Genes Dev, 2016;30:535–52. 92. Greer, E.L., Oskoui, P.R., Banko, M.R. et al. The energy sensor AMP‐activated protein kinase directly regulates the mammalian FOXO3 transcription factor. J Biol Chem, 2007;282:30107–19. 93. Shin, H.‐J.R., Kim, H., Oh, S. et al. AMPK–SKP2–CARM1 signalling cascade in transcriptional regulation of autophagy. Nature, 2016;534:553–7. 94. Toyama, E.Q., Herzig, S., Courchet, J. et al. Metabolism. AMP‐activated protein kinase mediates mitochondrial fission in response to energy stress. Science, 2016;351:275–81. 95. Jäger, S., Handschin, C., St‐Pierre, J., and Spiegelman, B.M. AMP‐activated protein kinase (AMPK) action in skeletal muscle via direct phosphorylation of PGC‐1alpha. PNAS, 2007;104:12017–22. 96. Cantó, C., Gerhart‐Hines, Z., Feige, J.N. et  al. AMPK regulates energy expenditure by modulating NAD+ metabolism and SIRT1 activity. Nature, 2009;458:1056–60. 97. Mansueto, G., Armani, A., Viscomi, C. et al. Transcription factor EB controls metabolic flexibility during exercise. Cell Metab, 2017;25:182–96. 98. Wada, S., Neinast, M., Jang, C., et al. The tumor suppressor FLCN mediates an alternate mTOR pathway to regulate browning of adipose tissue. Genes Dev, 2016;30:2551–64. 99. Baas, A.F. et al. Complete polarization of single intestinal epithelial cells upon activation of LKB1 by STRAD. Cell, 2004;116:457–66. 100. Fu, D., Wakabayashi, Y., Ido, Y., Lippincott‐Schwartz, J., and Arias, I.M. Regulation of bile canalicular network formation and maintenance by AMP‐activated protein kinase and LKB1. J Cell Sci, 2010;123:3294–302. 101. Fu, D., Wakabayashi, Y., Lippincott‐Schwartz, J., and Arias, I.M. Bile acid stimulates hepatocyte polarization through a cAMP‐Epac‐ MEK‐LKB1‐ AMPK pathway. Proc Natl Acad Sci USA, 2011;108:1403–8. 102. Homolya, L., Fu, D., Sengupta, P. et al. LKB1/AMPK and PKA control ABCB11 Trafficking and polarization in hepatocytes. PLoS One, 2014:9. 103. Zhang, L., Li, J., Young, L.H., and Caplan, M.J. AMP‐activated protein kinase regulates the assembly of epithelial tight junctions. Proc Natl Acad Sci USA, 2006;103:17272–7. 104. Zheng B. and Cantley, L.C. Regulation of epithelial tight junction assembly and disassembly by AMP‐activated protein kinase. Proc Natl Acad Sci USA, 2007;104:819–22. 105. Yano, T., Matsui, T., Tamura, A., Uji, M., and Tsukita, S. The association of microtubules with tight junctions is promoted by cingulin phosphorylation by AMPK. J Cell Biol, 2013;203:605–14.

484

THE LIVER:  REFERENCES

106. Aznar, N., Patel, A., Rohena, C.C. et al. AMP‐activated protein kinase fortifies epithelial tight junctions during energetic stress via its effector GIV/ Girdin. Elife, 2016;5:e20795. 107. Enomoto, A., Murakami, H., Asai, N. et al. Akt/PKB regulates actin organization and cell motility via Girdin/APE. Dev Cell, 2005;9:389–402.111. 108. Mihaylova, M.M. and Shaw, R.J. The AMPK signalling pathway coordinates cell growth, autophagy, and metabolism. Nat Cell Biol, 2011;13:1016–23. 109. Owen, M.R., Doran, E., and Halestrap, A.P. Evidence that metformin exerts its anti‐diabetic effects through inhibition of complex 1 of the mitochondrial respiratory chain. Biochem J, 2000;348(3):607–14. 110. Fullerton, M.D., Galic, S., Marcinko, K. et al. Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin‐sensitizing effects of metformin. Nat Med, 2013;19:1649–54. 111. Foretz, M. et al. Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state. J Clin Invest, 2010;120:2355–69.

112. An, H. and He, L. Current understanding of metformin effect on the control of hyperglycemia in diabetes. J Endocrinol, 2016;228:R97–106. 113. Smith, B.K. Marcinko, K., Desjardins, E.M., Lally, J.S., Ford, R.J., and Steinberg, G.R. Treatment of nonalcoholic fatty liver disease: role of AMPK. Am J Physiol Endocrinol Metab, 2016;311:E730–40. 114. Quinn, B.J., Kitagawa, H., Memmott, R.M., Gills, J.J., and Dennis, P.A. Repositioning metformin for cancer prevention and treatment. Trends Endocrinol Metab, 2013;24:469–80. 115. Myers, R.W. et al. Systemic pan‐AMPK activator MK‐8722 improves glucose homeostasis but induces cardiac hypertrophy. Science, 2017;357:507–11. 116. Cokorinos, E.C. et al. Activation of skeletal muscle AMPK promotes glucose disposal and glucose lowering in non‐human primates and mice. Cell Metab. 2017;25:1147–59.

39

Insulin‐Mediated PI3K and AKT Signaling Hyokjoon Kwon1 and Jeffrey E. Pessin2,3 Department of Medicine, Division of Endocrinology, Metabolism and Nutrition, Rutgers‐Robert Wood Johnson Medical School, New Brunswick, NJ, USA 2 Albert Einstein‐Mount Sinai Diabetes Research Center and the Fleischer Institute for Diabetes and Metabolism, Albert Einstein College of Medicine, Bronx, NY, USA 3 Department of Medicine and Department of Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, NY, USA 1

INTRODUCTION

CHARACTERISTICS OF PI3K AND AKT

Obesity is a pandemic in modern society and closely linked with diverse metabolic diseases such as cardiovascular dis­ ease, type 2 diabetes mellitus (T2D) and non‐alcoholic fatty liver disease (NAFLD). Thus, the cost of managing obesity and related metabolic diseases is a burden on public health care systems in modern society. T2D is a quickly growing global metabolic disorder characterized by impaired insulin secretion from pancreatic β cells and insulin resistance in peripheral tissues such as liver, muscle, and adipose tissue. Insulin resistance diminishes insulin‐stimulated glucose uptake in muscle and glycogen synthesis in liver and also impairs insulin‐mediated suppression of gluconeogenesis in liver, resulting in hyperglycemia and hyperinsulinemia. As insulin is a crucial endocrine hormone for modulating glucose homeostasis, the molecular mechanism of insulin signaling in glucose homeostasis has been a central theme in biomedical research. Insulin signaling is initiated by the activation of the insulin receptor (IR) through autophosphorylation of the tyrosine residue in the IR, and then many signaling molecules including insulin receptor substrate 1 and 2 (IRS1/2), phosph­ oinositide‐3‐kinase (PI3K) and AKT/protein kinase B (PKB) are involved to modulate downstream signaling pathways. In this chapter, we will focus on the role of PI3K and AKT/PKB in insulin signaling related to pathophysiology in T2D, NAFLD, and liver cancer.

PI3K and AKT/PKB are key signaling molecules, mediating insulin signaling in diverse tissues including liver, muscle, and adipose tissue. In general, the intracellular responses to extracel­ lular signals are mediated by small second‐messenger molecules such as cAMP, Ca2+, and lipid molecules that are responsible for transducing signals between the extracellular environment and intracellular compartments. In this pathway, PI3K activation induces the formation of insotitol‐3,4,5‐trisphosphate (PI(3,4,5) P3) that serves as a second‐messenger to activate AKT in the insulin signaling pathway, resulting in glucose disposal, glyco­ gen synthesis, and suppression of lipolysis. To understand the physiologic role of these signaling molecules in insulin action and glucose homeostasis, in the following sections we review the biochemical characteristics of the PI3K and AKT kinases.

Phosphoinositide 3‐kinase (PI3K) Phosphatidylinositol has free hydroxyl groups that are ­phosphorylated by diverse series of kinases including phosphoi­ nositide 3‐kinase (PI3K) for the generation of crucial second‐ messenger in cell signaling. Early studies demonstrated that the cleavage of phosphatidylinositol‐4,5‐bisphosphate (PI(4,5)P2) by a membrane bound phospholipase C (PLC) generated diacylg­ lycerol (DAG) and inositol‐1,4,5‐trisphosphate. Diacylglycerol activates protein kinase C (PKC), and inositol‐1,4,5‐trisphosphate

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

486

THE LIVER:  CHARACTERISTICS OF PI3K AND AKT

(a)

Class 1A Catalytic

p110α/β/δ

ABD

RBD

Regulatory

p85α/β

SH3

BH

Helical

Kinase

N-SH2

iSH2

C-SH2

p55α/p50α

N-SH2

iSH2

C-SH2

p55γ

N-SH2

iSH2

C-SH2

C2

Class 1B Catalytic

p110γ

Regulatory

p101

RBD

C2

Helical

Kinase

GβγBD

p84

(b)

T308

AKT1

1

PH

Kinase Domain

AKT2

1

PH

Kinase Domain

AKT3

1

PH

Kinase Domain

T309

T305

T450

S473

HD T451

S474

HD T447

480

481

S472

HD

479

Figure 39.1  Schematic structure of PI3K and AKT/PKB. (a) Domain structure of class I PI3K catalytic and regulatory subunits. ABD, adaptor‐ binding domain; RBD, Ras‐binding domain; SH2, Src homology‐2; BH, breakpoint cluster region homology; iSH, coiled‐coil inter‐SH2. (b) Domain structure AKT isoforms. PH, pleckstrin homology domain; HD, hydrophobic domain.

induces Ca2+ influx from intracellular calcium stores. In contrast, PI3K mediates phosphorylation of phosphoinositides at the D3  position to generate various 3‐phosphorylated phosphoi­ nositides such as phosphatidyl‐3,5‐bisphosphate (PI(3,5)P2) and phosphatidyl‐3,4,5‐triphosphate (PI(3,4,5)P3) in the absence of phospholipase‐mediated cleavage. In insulin signaling, PI3K generates plasma membrane‐bound PI(3,4,5)P3 to activate AKT through both the binding of the AKT pleckstrin homology (PH) domain to PI(3,4,5)P3 and AKT site‐specific phosphorylations by the 3‐phosphoinositide‐dependent kinase 1 (PDK1) and by the mechanistic target of rapamycin (mTOR) in complex 2 (mTORC2). PI3K is classified into three classes (class I, class II, and class III) according to structure and lipid substrate preferences [1]. Class II PI3K including PI3K‐C2α, PI3K‐C2β and PI3K‐ C2γ was discovered in mammals on the basis of sequence homology with class I and III PI3K instead of functional con­ text. Thus, their functional role is unclear although class II PI3K is constitutively associated with intracellular membranes. Class II PI3K consists of two subclasses α and β and contains a C‐terminal C2 domain as observed in PKC molecules for phospholipid binding. Vps34 was originally identified in Saccharomyces cerevisiae for the endosomal sorting and is the only member of class III PI3K. Vps34 generates a constitutive

heterodimer with Vps15/p150 to locate on the intracellular membrane, and the biological function of Vps34 in mammals relates to the regulation of vesicle trafficking such as autophagy, endocytosis, and phagocytosis. In mammals, class I PI3K is present in all cell types and mediates insulin signaling. Class I PI3K consists of a heterodimer with a catalytic subunit (110– 120 kDa) and regulatory subunit, and phosphatidylinositol‐4, 5‐bisphosphate (PI(4,5)P2) is a preferred substrate in vivo. Catalytic subunits contain C‐terminal catalytic and phosphati­ dylinositol kinase domains, N‐terminal adaptor‐binding domain (ABD), and Ras‐binding domain (RBD) (Figure 39.1a). Low concentrations (nanomolar) of Wortmannin irreversibly inhibits the catalytic subunit of class I PI3K by Schiff base for­ mation with a lysine in the C‐terminal kinase domain [2]. Regulatory subunits have C‐terminal Src homology‐2 (SH2) domains that are separated by an inter‐SH2 (iSH2) region to provide a binding site for the ABD of the catalytic subunit. The SH2 domain is a module of about 100 amino acids to bind phosphotyrosine‐containing motifs. Regulatory subunits also have N‐terminal proline‐binding SH3 and a breakpoint cluster region homology domain (BH). Class I PI3K catalytic p110 subunits are divided into a class IA group such as p110α, p110β, and p110δ, which bind the regulatory p85 type subunit



39:  Insulin‐Mediated PI3K and AKT Signaling

and into a class IB p110γ to interact with regulatory p101 and p84. The p110α and p110β are expressed ubiquitously, whereas p110γ and p110δ are expressed in immune cells. Each catalytic subunit generates a dimer with a regulatory subunit to modulate the catalytic activity and subcellular localization. The p110α catalytic subunit generates a heterodimer with one of five dif­ ferent regulatory subunits including p85α, p85δ, p55α, p55γ, and p50α. In the p85α/p110α heterodimer, the ABD of p110α catalytic subunit interacts with coiled‐coil inter‐SH2 (iSH) domain of the p85α regulatory subunit to maintain a stable het­ erodimer with a low kinase activity state [3]. Thus, binding of the SH2 domain in a p85α regulatory subunit with phosphoty­ rosine residues such as phosphotyrosine in IRS1/2 releases inhibitory interactions between p85α and p110α, resulting in the activation of PI3K in insulin signaling [4, 5].

AKT/protein kinase B The v‐AKT oncogene known as AKT8 was identified from transforming retrovirus in a spontaneous thymoma of AKR mice, and then cellular homologue serine and threonine kinase AKT (roughly 57 kDa), also termed as protein kinase B (PKB), was cloned and characterized [6–9]. Human AKT has three isoforms (AKT1/PKBα, AKT2/PKBβ, AKT3/PKBγ), and each isoform is encoded from separate human genes: AKT1/ PKBα in Chr14 q32.33, AKT2/PKBβ in Chr19 q13.2, and AKT3/PKBγ in Chr1 q44 (Figure 39.1b). For functional stud­ ies, murine models demonstrate that each AKT isoform has distinct and overlapping signaling functions. AKT1 deficient mice show reduced body size, AKT2 deficiency impairs glu­ cose homeostasis, and AKT3 deficient mice show diminished brain size. In contrast, all AKT isoforms contribute to phos­ phorylation of GSK3α and GSK3β in 3T3‐L1 adipocytes and regulation of forkhead box O1 (FOXO1), TSC2, and GSK3β in H157 cells. AKT kinases belong to a class of AMP/GMP kinase and protein kinase C (AGC) kinases. AKT has a PH domain, kinase domain, and hydrophobic domain (HD) with PH and kinase domains con­ nected with an α‐helical linker. The PH domain found as a major phosphorylation site by PKC in pleckstrin proteins at its N‐termi­ nus interacts with PI(3,4,5)P3 to localize AKT into the plasma membrane and then to interact with 3‐phosphoinositide‐dependent kinase 1 (PDK1) that also has a PH domain to be enriched in plasma membrane. The PH domain is a 100–120 amino acids motif with seven anti‐parallel β‐sheets forming a hydrophobic pocket that interacts with the C‐terminal amphipathic helix. The PH domain is a primary lipid binding site although it is also involved in protein– protein interactions. As the PH domain has different phosphoi­ nositide binding specificity, PH domains of dynamin bind to PI(4) P1 and PI(4,5)P2, whereas PH domains of AKT and PDK1 bind to PI(3,4)P2 and PI(3,4,5)P3 [10, 11]. The kinase domain of AKT shares high homology with other AGC kinase members, and the kinase domain of three AKT kinase isoforms displays around 87% sequence homology. HD is a characteristic feature of AGC kinases including protein kinase A (PKA), protein kinase C (PKC), and PDK1 and contains a F‐X‐X‐F/Y‐S/T‐Y/F motif where X is any amino acid. Upon biosynthesis of AKT, the nascent AKT poly­ peptide is phosphorylated at the ribosome by mTORC2 at the C‐terminal turn motif [12], enhancing the protein stability in

487

cytosol. Indeed, mTORC2‐induced phosphorylation of Thr‐450 in the AKT1 turn motif localizes AKT1 to the cytosol as an inactive conformation through the interaction between PH and kinase domains [13]. AKT activity is regulated through the phospho­ rylation in kinase domain at Thr‐308 by PDK1 and in the hydro­ phobic domain at Ser‐473 by mTORC2, necessary for the full induction of Akt substrate kinase activity [14].

INSULIN‐MEDIATED PI3K AND AKT SIGNALING AND INSULIN RESISTANCE Glucose homeostasis is maintained by delicate regulation of the pancreatic exocrine and endocrine systems. The pancreatic exo­ crine cells composed of acinar and ductal cells secrete digestive enzymes into the duodenum for nutrient digestion. In contrast, the pancreatic endocrine cells in the islets of Langerhans secrete endocrine hormones into the blood circulation to regulate nutri­ ent metabolism. Pancreatic islets include several endocrine cells such as α cells (glucagon production), β cells (insulin produc­ tion), δ cells (somatostatin production), PP cells (pancreatic polypeptide production), and ε cells (ghrelin production) for specific endocrine functions. The secretion of glucagon and insulin is tightly regulated in response to circulating glucose levels to maintain normoglycemia during fasting and feeding, respectively. In peripheral tissues, insulin stimulates glucose uptake (skeletal muscle and adipose tissue) and glycogen syn­ thesis (skeletal muscle and liver) and inhibits gluconeogenesis and glycogenolysis (liver). Insulin also increases lipogenesis in hepatocytes and adipocytes and diminishes lipolysis in adipo­ cytes to reduce circulating free fatty acid and glycerol [15]. Thus, impaired regulation of insulin secretion and IR‐mediated signaling pathway results in T2D.

Insulin‐mediated PI3K and AKT signaling Identification and characterization of the IR initiated a huge effort to understand the molecular mechanisms of insulin action in glucose homeostasis. At the molecular level, insulin binds to the cell surface IR to initiate intracellular signaling cascades that ultimately results in specific cellular biological responses [16]. The IR is a transmembrane tyrosine kinase receptor encoded by Chr19 p13.2 in humans and has two isoforms, IR‐A and IR‐B, generated by alternative splicing of exon11. The IR‐A isoform excludes exon 11 and is predominantly expressed in fetal tissues and the brain with high affinity for insulin and insu­ lin‐like growth factor 2 (IGF‐2), whereas IR‐B includes exon 11 and is highly expressed in liver. To interact with insulin, the insulin receptor consists of two α subunits and two β subunits that are bound by a disulfide link into an α2β2 heterotetrameric complex. The α and β subunits are derived from a single large precursor by one or more proteolytic cleavages. The extracellu­ lar α subunits directly bind insulin that allows for a transmem­ brane conformational change that activates the intracellular tyrosine kinase domain of the IR β subunits, resulting in intra­ molecular transphosphorylation of the β subunit to phosphoryl­ ate its adjacent partner on specific tyrosine residues [17, 18]. Autophosphorylated tyrosine residues in the tyrosine kinase

488

THE LIVER:  INSULIN‐MEDIATED PI3K AND AKT SIGNALING AND INSULIN RESISTANCE

domain of β subunits recruit receptor substrates such as insulin receptor substrate (IRS) that is a scaffold for organizing the proximal signaling complex. Phosphotyrosine residues in acti­ vated IRs interact with IRS through its phosphotyrosine binding domain (PTB) [19]. IRS also has several tyrosine residues, which are phosphorylated by activated IR tyrosine kinase activ­ ity, for the interaction with the SH2 domain of the regulatory subunit of PI3K [5]. IRS isoforms (IRS1‐IRS4) show differen­ tial function in physiology. IRS‐1 knockout mice have growth retardation and insulin resistance particularly in muscle, albeit with normal whole body glucose tolerance [20], whereas IRS‐2 knockout mice show impaired insulin action in liver and pancre­ atic β cell loss, resulting in T2D [21]. IRS1/2 phosphorylated on specific tyrosine residues activates two major signaling path­ ways; (i) the PI3K‐AKT/PKB pathway and (ii) Ras/mitogen‐ activated protein kinase (MAPK) pathway. In addition, there are inhibitory molecules for insulin signaling such as protein tyros­ ine phosphatase 1B (PTP1B), suppressor of cytokine signaling (SOCS), and growth factor receptor bound protein 10 (Grb10) that suppress insulin signaling by inducing IR dephosphoryla­ tion, physical blocking substrate phosphorylation, and degrada­ tion of the IR and/or IRS substrates. Metabolites such as DAG also mediate the inhibition of insulin signaling (Figure 39.2).

Insulin‐mediated PI3K activation PI3K‐AKT/PKB signaling pathway modulates most metabolic functions of insulin. Activation of PI3K by extracellular stimu­ lation including insulin induces the activation of AKT. In the basal state, p85α/p110α heterodimeric PI3K has interactions between ABD of the p110α catalytic subunit and coiled‐coil iSH domain of p85α regulatory subunit to maintain stable PI3K with a low kinase activity state. However, insulin‐induced phos­ phorylation on tyrosine residue in IRS1/2 mediates interaction between phosphotyrosine of IRS1/2 and the SH2 domain in p85α regulatory subunit to release inhibitory interactions between p85α and p110α [5, 22]. Thus, PI3K accesses the ­membrane to produce phosphatidylinositol‐3,4,5‐triphosphate (PI(3,4,5)P3) from PI(4,5)P2 at the plasma membrane. Although p110α or p110β knockout mice are embryonic lethal, liver spe­ cific p110β deficient mice showed little effect in insulin signal­ ing. However, liver specific p110α deficient mice have glucose intolerance and insulin resistance, suggesting that p110α is criti­ cal to mediate insulin signaling in hepatocytes [23]. Deletion of the p85 regulatory subunits in heterozygous deletion of p85α, p85β, or p50α/p55α showed enhanced insulin sensitivity as ­regulatory subunits are in excess concentrations to catalytic subunits, thereby competing with IRS for the formation of

Figure 39.2  Insulin signaling and insulin resistance. IRS1/2 phosphorylated on specific tyrosine residues activates the phosphoinositide 3‐kinase (PI3K)‐AKT/protein kinase B (PKB) pathway and Ras/mitogen‐activated protein kinase (MAPK) pathway. PI3K‐AKT signaling pathway regulates metabolic processes such as glycogen synthesis (muscle and liver), glucose uptake (muscle and adipocytes), protein synthesis (muscle and liver), and gluconeogenesis (liver). Inflammatory signals such as TNF‐α, saturated free fatty acid, IL‐6, LPS, and diacylglycerol (DAG) activate inhibitory molecules such as suppressor of cytokine signaling (SOCS) and cJun N‐terminal kinase (JNK) to suppress insulin signaling, resulting in insulin resistance.



39:  Insulin‐Mediated PI3K and AKT Signaling

p85α/p110α heterodimer [24]. PI(3,4,5)P3 generated by insulin‐ induced activation of PI3K is metabolized rapidly into PI(4,5)P2 by lipid phosphatase including tumor suppressor phosphatase and tensin homologue (PTEN) and SH2‐containing inositol 5′‐phosphatase‐2 (SHIP2) to terminate proximal signaling [25, 26]. PTEN encoded in chromosome 10q23 was identified as a tumor suppressor that is inactivated in diverse tumors including endometrial, prostate, and mammary carcinomas.

Insulin‐mediated AKT activation PI(3,4,5)P3 generated by PI3K triggers the activation of 3‐phos­ phoinositide‐dependent protein kinase 1 (PDK1) that is respon­ sible for the phosphorylation and activation of AKT. PDK1 contains two domains, an N‐terminal kinase domain and a C‐ terminal PH domain. Autophosphorylation of Ser‐241 in kinase domain by PDK1 is required for kinase activity, and the small PH domain binds to membrane bound PI(3,4,5)P3 and PI(3,4)P2. As AKT also has a PH domain to interact with membrane‐bound PI(3,4,5)P3, AKT is subsequently recruited from the cytosol to the plasma membrane through binding of its PH domain to PI(3,4,5)P3, resulting in a conformational change that separates the PH and kinase domains from inactive conformation and exposes two key regulatory residues whose phosphorylations are required for maximal activation of AKT kinase. The acti­ vated PKD1 closely localized with AKT through PI(3,4,5)P3 binding phosphorylates Thr‐308 in the activation loop of AKT [13]. AKT is also phosphorylated on Ser‐473 by the mTORC2, and this dual phosphorylation results in full activation of AKT kinase activity [14]. Deletion of mTORC2‐specific subunits such as Rictor or Sin1 abrogates AKT phosphorylation at both the C‐terminal turn motif (Thr‐450 of AKT1) and HD (Ser‐473 of AKT1), resulting in the impaired proximal signaling of AKT [14, 27]. Interestingly, the AKT isoforms show differential func­ tions according to the distinctive expression profiles in tissues. AKT1 and AKT2 are broadly expressed with AKT2 being more closely linked to metabolic processes. AKT1 deficient mice have growth retardation without defects in metabolism. In contrast, AKT2 deficient mice show insulin resistance due to interrupted insulin signaling [28].

Proximal signaling of insulin‐induced AKT activation Activated AKT/PKB regulates diverse insulin‐mediated meta­ bolic pathways such as glucose transport, glycogen synthesis, gluconeogenesis, protein synthesis, and cell growth. Several AKT substrates have been identified by the AKT consensus motif (R‐X‐R‐X‐X‐S/T‐B) where X is any amino acid, and B represents bulky hydrophobic amino acids [29]. AKT promotes cell growth and proliferation through the suppression of p27 to activate cyclin D1 along with the activation of MAPK signaling pathway (Figure 39.2). AKT phosphorylates AKT substrate of 160 kDa (AS160) to activate the Rab family of small GTPases that initiate the translocation of the glucose transporter 4 (GLUT4), resulting in glucose uptake in muscle and adipocytes. AKT also suppresses glycogen synthase kinase‐3 (GSK3) via phosphorylation on Ser‐21 or Ser‐9 to activate glycogen syn­ thase in muscle and liver [30]. AKT phosphorylates FOXO1

489

that induces FOXO1 association with nuclear 14‐3‐3 proteins to exclude FOXO1 from the nucleus to the cytosol in an inactive state. In the liver, this suppresses gluconeogenic gene expres­ sion and thereby inhibits hepatic glucose production in feeding. AKT phosphorylates tuberous sclerosis complex 1 and 2 (TSC1/2), which releases the inhibition of Ras homologue enriched in brain (Rheb) for the activation of mTORC1 complex [14], that in turn enhances protein synthesis through the acti­ vation of eukaryotic translation initiation factor 4E binding protein‐1 (4E‐BP) and p70 ribosomal protein S6 kinase 1 (p70S6K1). Furthermore, mTORC1 activation induces SREBP1c activation for the expression of lipogenic genes including FAS and ACC to induce lipogenesis in liver [31]. Fasting‐induced β‐adrenergic receptor signaling activates PKA to phosphorylate perilipin1 (PLIN1) and hormone sensitive lipase (HSL), facilitating lipolysis in adipocytes. In feeding, however, insulin‐induced AKT activation mediates phospho­ rylation of phosphodieterase‐3B (PDE‐3B) to decrease cAMP, resulting in the suppression of PKA activity and HSL activity in adipocytes [32]. Thus, insulin‐induced suppression of ­adipose tissue lipolysis mediates acute suppression of gluconeo­ genesis in liver as suppressed lipolysis reduces acetyl‐CoA ­levels in liver [33, 34].

Insulin resistance in liver The molecular mechanisms accounting for insulin resistance are still unclear although there is substantial progress in our under­ standing of insulin signaling. Insulin resistance is the integral result of alterations in insulin secretion in pancreatic β cells, IR expression, ligand binding, and downstream IR signaling, resulting in diverse metabolic disease such as T2D and cardio­ vascular and NAFLD diseases. Liver is one of major tissues for regulating glucose homeostasis in response to pancreatic hor­ mones such as insulin and glucagon produced by pancreatic β and α cells, respectively. Insulin‐mediated PI3K and AKT acti­ vation regulate hepatic glucose and lipid metabolism. In insulin sensitive individuals, regular meals increase the blood glucose that in turn releases pancreatic insulin to activate insulin signal­ ing and glucose uptake through IR and GLUT4 in muscle and adipose tissue, respectively. In the liver, PI3K and AKT activa­ tion enhance glycogen synthesis and de novo lipogenesis (DNL) and also suppresses gluconeogenesis through FOXO inactiva­ tion (Figure  39.3a). Thus, hepatic insulin signaling may sup­ press gluconeogenesis through transcriptional suppression of gluconeogenic genes such as PCK1 and G6PC for long‐term regulation. However, the ability of insulin to acutely suppress gluconeogenesis is mostly mediated by an inhibition of adipose tissue lipolysis [34]. First, insulin reduces hepatic glucose pro­ duction within 30 minutes but does not reduce gluconeogenic protein levels [35]. Second, insulin suppresses adipose tissue lipolysis to regulate hepatic gluconeogenesis as acetyl‐CoA (Ac‐CoA) generated by adipose tissue lipolysis allosterically activates pyruvate carboxylase (PyC) activity for hepatic gluco­ neogenesis. Thus, suppressed adipose tissue lipolysis lowers hepatic acetyl‐CoA and glycerol contents to reduce PyC activity and glucose production [33, 34, 36]. In contrast, in insulin resistance status, glycogenolysis and gluconeogenesis are enhanced to produce hepatic glucose due to dysfunctional

490

THE LIVER:  INSULIN‐MEDIATED PI3K AND AKT SIGNALING AND INSULIN RESISTANCE

(a)

(b)

Figure 39.3  Insulin resistance in liver. (a) In insulin sensitive individuals, insulin derived from pancreatic β cells activates AKT to enhance gly­ cogen synthesis, de novo lipogenesis, and suppress gluconeogenesis through FOXO1 degradation. (b) In insulin resistance, impaired insulin signal­ ing increases glycogenolysis and gluconeogenesis, and flux of free fatty acids and glycerol from adipose tissue and intestine results in glucose production and fat accumulation in liver. DNL, de novo lipogenesis; GS, glycogen synthase; GP, glycogen phosphorylase; PyC, pyruvate carboxy­ lase; Ac‐CoA, acetyl‐CoA; FA‐CoA, fatty acyl‐CoA; TAG, triacylglycerol; CM, chylomicron; VLDL, very low density lipoprotein.

insulin signaling (Figure  39.3b). Impaired activation of PI3K and AKT induces activation of glycogen phosphorylase (GP) to induce glycogenolysis along with suppressed glycogen syn­ thase (GS) activity. Impaired AKT‐mediated phosphorylation of FOXO1 also generates nuclear active FOXO1 to mediate hepatic gluconeogenesis. Flux of free fatty acids and glycerol from adi­ pose tissue and intestine supplies substrate for glucose produc­ tion and fat accumulation in liver, resulting in hyperglycemia and hyperlipidemia. Hepatic insulin resistance is reproducibly correlated with increased hepatic triglyceride content as shown in NAFLD. High levels of DAG generated by incomplete synthesis to tri­ glyceride or the breakdown of triglyceride to DAG have been proposed to inhibit insulin signaling through protein kinase Cε (PKCε) activation, which phosphorylates IR at Thr‐1160 (Thr‐1150 in mouse) to suppress tyrosine kinase activity of IR [37–41]. Thus, intervention of accumulation of triglyceride in liver results in the reversal of hepatic insulin resistance in humans and rodent models with NAFLD. In this regard, adipose

triglyceride lipase (ATGL) deficient mice that have reduced ability to convert triglyceride to DAG show enhanced glucose tolerance and insulin sensitivity [42]. More recently, an alterna­ tive model of increased ceramide levels has also been shown to be associated with insulin resistance [43]. Decreased IRS pro­ tein levels also contribute to insulin resistance in rodents and humans although complete molecular understanding of the mechanisms of reduced IRS levels are still under investigation [44]. SOCS1/3 increased by obesity‐induced inflammatory cytokines such as TNF‐α and IL‐6 enhances the degradation of IRS1/2 through E3 ubiquitin ligase activation, resulting in insu­ lin resistance [45, 46]. IRS phosphorylation on serine residues is another mechanism that induces insulin resistance as the IRS proteins contain several serine residues that are phosphorylated by kinases such as extracellular signal regulated kinase (ERK), cJun N‐terminal kinase (JNK), protein kinase Cζ (PKCζ), and p70S6K [47]. The phosphorylation of IRS on Ser‐307 is a ­typical inhibitory signal to suppress insulin signaling as Ser‐307 locates in the phosphotyrosine‐binding (PTB) domain of IRS [48].



39:  Insulin‐Mediated PI3K and AKT Signaling

Thus, increased TNF‐α, saturated free fatty acids, and endoplas­ mic reticulum (ER) stress in obese individuals activate JNK and inhibitor of nuclear factor κB kinase β (IKKβ) to phosphorylate Ser‐307 of IRS. In addition, ERK activated by insulin also phos­ phorylates IRS1 on Ser‐612 to attenuate AKT activation [49].

PI3K AND AKT IN GLUCOGENOGENESIS AND LIPOGENESIS The secretion of insulin and glucagon are tightly regulated to modulate gluconeogenesis and lipogenesis in hepatocytes, and dysregulation of the proximal signaling pathways of these ligands results in the hyperglycemia and hyperlipidemia seen in metabolic diseases. In fasting, glucagon and catecholamines regulate gluconeogenesis through the activation of cAMP‐ dependent PKA. Activated PKA phosphorylates CREB‐Ser133 to induce the expression of gluconeogenic genes such as Pck1 and G6pc and also phosphorylates liver pyruvate kinase (L‐PK) to suppress glycolysis. Glucagon decreases the level of fruc­ tose‐2,6‐bisphosphate, an allosteric activator of phosphofruc­ tokinase and an inhibitor of fructose‐1,6‐bisphosphatase to suppress glycolysis via a phosphorylation of phosphofructoki­ nase 2 (PFK2) [50]. Glucagon‐induced activation of PKA regu­ lates gene expression mediated by the CREB‐CBP‐CRTC2 complex, and CRTC2 is a key regulator in CREB‐CBP‐CRTC2 complex regulation. Glucagon‐induced PKA activation sup­ presses constitutively active serine/threonine kinase SIK2 to decrease CRTC2‐Thr171 phosphorylation and also phospho­ rylates inositol‐1,4,5‐triphsophate receptors (IP3R) to increase Ca2+ influx for CRTC2‐specific phosphatase calcineurin acti­ vation, resulting in the formation of the CREB‐CBP‐CRTC2 complex in the nucleus to induce gluconeogenic target gene expression [51, 52]. In contrast, insulin suppresses the secretion of glucagon from pancreatic α‐cells, and phosphorylated CRTC2 through insulin signaling interacts with 14‐3‐3 protein to be sequestered in cytoplasm, thereby suppressing the forma­ tion of the CREB‐CBP‐CRTC2 complex [53]. Thus, insulin regulates glucose homeostasis after feeding through the activa­ tion of PI3K and AKT to suppress gluconeogenesis and enhance lipogenesis in hepatocytes.

PI3K and AKT‐mediated modulation in gluconeogenesis Gluconeogenesis in liver contributes approximately half of the hepatic glucose production in humans during overnight fasting and is the primary mechanism responsible for the increased fasting glucose levels in T2D patients. Gluconeogenesis is ­ ­regulated by complex mechanisms: (i) availability of gluconeo­ genic substrate such as lactate, alanine and glycerol, (ii) meta­ bolic intermediate‐induced allosteric regulation in metabolic enzymes, and (iii) the balance of hormones such as insulin, glucagon, and catecholamines. Glycerol and non‐esterified fatty acids (NEFA) generated by lipolysis are involved in the regula­ tion of gluconeogenesis. Glycerol, one of major substrates in  gluconeogenesis, is converted to glycerol‐3‐phosphate by glycerol kinase and then dihydroxyacetone‐3‐phosphate to

491

participate glucose production in liver. Although acetyl‐CoA generated by β‐oxidation of NEFA in mitochondria does not contribute to the substrates for glucose production, acetyl‐CoA allosterically activates pyruvate carboxylase to enhance gluco­ neogenesis [33]. Thus, insulin‐mediated suppression of gluca­ gon secretion and adipose tissue lipolysis is important in the regulation of hepatic gluconeogenesis. Experimentally, infusion of insulin in fasted rats decreased the concentration of glycerol and acetyl‐CoA along with the suppression of hepatic glucose production. However, when infusion of glycerol and acetate was added to the infusion of insulin, hepatic glucose production was restored as increased glycerol and acetyl‐CoA enhanced gluco­ neogenesis [33–35]. These results suggest that lipolysis is also involved in the regulation of gluconeogenesis. FOXO1 and transcriptional coactivator peroxisome prolif­ erator‐activated receptor‐γ coactivator‐1α (PGC‐1α) enhance the expression of gluconeogenic genes to mediate gluconeo­ genesis in liver. To suppress gluconeogenesis in the post­ prandial state, insulin‐dependent PI3K and AKT activation phosphorylates FOXO1 (Thr‐24, Ser‐253, and Ser‐316 for murine FoxO1) to exclude FOXO1 from the nucleus that is then subsequently ubiquitinated and degraded by proteasome, resulting in the suppression of gluconeogenesis [54, 55]. However, the role of insulin in gluconeogenesis through FOXO1‐PGC1α is still unclear as liver specific IR and FOXO1 gene double knockout mice show normal glucose homeostasis although liver specific IR knockout mice have glucose intoler­ ance and insulin resistance [56, 57]. In addition, T2D patients and insulin resistant high‐fat diet fed rodent models do not show differential expression of gluconeogenic genes including PCK1 and G6PC [58]. In contrast to the high‐fat diet, fructose fed rodents display increased G6pc expression that is the tran­ scriptional target of both FOXO1 (primarily a gluconeogenic gene transcriptional activator) and ChREBPβ (primarily a lipogenic gene transcriptional regulator) [59]. Since insulin inhibits G6pc expression through suppression of FOXO1, and fructose is a poor inducer of insulin secretion, it is hypothe­ sized that one mechanism for enhanced gluconeogenesis by fructose is due to an imbalance between FOXO1 and ChREBPβ regulation.

PI3K and AKT‐mediated modulation in lipogenesis Glucose from excess dietary carbohydrate undergoes glycoly­ sis in liver and is eventually converted into fatty acids through DNL and then esterified to triglyceride for very low density lipoprotein (VLDL) secretion into blood circulation. DNL is abnormally increased in NAFLD and closely linked to the pathogenesis of T2D. Regulation of DNL is mediated by tran­ scriptional regulation of enzymes such as fatty acid synthase (FAS) and acetyl‐CoA carboxylase (ACC) for fatty acid syn­ thesis and allosteric regulation of ACC. The transcriptional regulation of critical enzymes in DNL is modulated by two major transcriptional regulators including sterol regulatory ele­ ment binding protein 1c (SREBP1c) and carbohydrate response element binding protein (ChREBP) that are activated by enhanced insulin signaling and glucose concentration, respec­ tively [60, 61]. Thus, insulin‐induced PI3K and AKT activation

492

THE LIVER:  PI3K AND AKT IN LIVER DISEASES: NAFLD AND CANCER

closely regulate the DNL in liver. mTORC1 activated by AKT suppresses Lipin1, a phosphatidic acid phosphatase, to increase phosphorylation of nascent SREBP1c in ER, resulting in the activation of SREBP1c to induce the expression of fatty acid synthesis enzymes such as FAS and ACC in liver [62, 63]. In contrast, ChREBP regulation is mediated by glucose imported by insulin independent GLUT2. Thus, glucose metabolites such as glucose‐6‐phosphate and fructose‐2,6‐bisphosphate and dephosphorylation of Ser‐196 in ChREBP are proposed to regulate ChREBP activity [64, 65]. ACC is also one of the key regulators modulating DNL in liver as ACC converts acetyl‐ CoA to malonyl‐CoA to provide monomer for fatty acid syn­ thesis in DNL. ACC has two isoforms: ACC1 in adipose tissue, mammary gland, and liver and ACC2 in skeletal and cardiac muscle. ACC1 activity is regulated by several different levels. First, ACC expression is enhanced by insulin signaling through SREBP1c activation along with ChREBP and LXRs [61]. Second, ACC1 exists as a low activity dimer, and allosteric activators such as citrate and glutamate stimulate polymeriza­ tion of ACC1 to generate high activity polymers to enhance DNL [66], whereas malonyl‐CoA and fatty acyl‐CoA inhibits ACC1 polymerization as a feedback inhibition mechanism. Third, dephosphorylation and phosphorylation of ACC1 by insulin and glucagon are also important for the regulation of ACC1 activity although the molecular mechanism is still unclear [67–69]. Furthermore, malonyl CoA, a product of ACC, inhibits carnitine : palmitoyl‐CoA transferase‐1 (CPT1) to modulate the entry of long‐chain fatty acyl‐CoA into mito­ chondria for β‐oxidation. Thus, ACC suppression is important to reduce lipid stores in muscle and liver and hence to enhance insulin sensitivity.

PI3K AND AKT IN LIVER DISEASES: NAFLD AND CANCER NAFLD is a common hepatic disorder characterized by fat accumulation in the liver in the absence of excessive alcohol intake. Steatosis, which has benign fat accumulation within the cytoplasm of at least 5% of the hepatocytes in the liver, pro­ gresses to non‐alcoholic steatohepatitis (NASH), fibrosis, cir­ rhosis, and liver cancer. For the diagnosis of NASH, hepatocyte ballooning and inflammation appear with steatosis. However, NASH is not necessary for liver fibrosis, and in general NASH and fibrosis frequently occur together. Liver fibrosis character­ ized by the deposition of excess extracellular matrix (ECM) such as collagen by activated hepatic stellate cells progresses to cirrhosis, which demonstrates the loss of hepatic structure, por­ tal hypertension, and the formation of regenerative nodules. Cirrhosis is irreversible and induces liver organ failure and hepatocellular carcinoma (HCC). NAFLD is closely linked with hepatic and adipose tissue insulin resistance, showing that about 50% reduction in glucose disposal and impaired insulin‐medi­ ated suppression in endogenous gluconeogenesis, and approxi­ mately 80% of the T2D patients demonstrates NAFLD [70, 71]. Thus, understanding the molecular mechanisms linking these pathophysiologic symptoms is one of major targets of current biomedical research.

PI3K and AKT signaling in non‐alcoholic fatty liver disease The development of NAFLD is an early step in the pathogenesis of T2D, and most obese T2D patients have NAFLD. As lipid accumulation induces insulin resistance in liver, weight loss leads to the resolution of NAFLD and normalization of fasting glucose level [72]. Thus, hyperinsulinemia associated with hepatic insulin resistance is a hallmark feature of NAFLD, espe­ cially in humans with T2D. Diverse tissue specific IR knockout mice demonstrate that muscle specific IR knockout mice or muscle/adipose tissue specific IR knockout mice have normal glucose levels. However, liver specific IR knockout mice result in hyperglycemia with peripheral insulin resistance and NASH, suggesting that liver insulin resistance is critical in pathogenesis of NAFLD [73]. Peripheral insulin resistance induces lipolysis in adipose tissue to increase free fatty acid flux into the liver, and liver insulin resistance induces DNL through SREBP1c activation, resulting in fat accumulation in liver. The liver‐specific overexpression in mice of a mutated cata­ lytic subunit p110α (Pik3ca) found in human HCC results in severe liver steatosis in six months and liver tumors in a year [74]. In contrast liver specific Pik3ca deficient mice show sup­ pressed liver steatosis with a high‐fat diet [75], suggesting that PI3K is involved in NAFLD. Unlike Pik3ca, Pik3cb (encoding for p110β) deficiency mice show normal hepatic lipid levels, indicating that Pik3ca is specifically involved in NAFLD and HCC development. PTEN dephosphorylates PI(3,4,5)P3 to sup­ press PI3K‐AKT signaling pathways, and PTEN deficient mice demonstrate early onset of hepatic microvesicular steatosis due to increased DNL and free fatty acid uptake that progresses into NASH, liver fibrosis, and cancer [76]. As Pik3ca deficient mice show steatosis without NASH and liver fibrosis, PTEN deficient mice are useful rodent models to recapitulate human NAFLD pathogenesis. AKT, a downstream PI3K signaling molecule, is also involved in NAFLD development. Adenoviral overexpres­ sion of AKT significantly induces micro and macrovesicular steatosis and NASH in 12 weeks and then results in HCC [77]. High‐fat diet‐induced hepatic fibrosis enhances deposition of ECM, resulting in interactions between the ECM and AKT through the integrin‐linked protein kinase (ILK) [78]. Thus, liver‐specific ILK deletion shows improved HFD‐induced liver steatosis and insulin resistance.

PI3K and AKT signaling in cancer Platelet‐derived growth factor (PDGF) stimulates PI3K to gener­ ate PI(3,4)P2 and PI(3,4,5)P3 in smooth muscle, suggesting that PI3K is important for cellular response to growth factors and tumorigenesis [79], and the PI3K pathway is one of the most fre­ quently activated signaling pathways in human cancer. Thus, PI3K mutations are commonly found in a variety of cancers, and development of PI3K inhibitors is currently being conducted to treat various cancers. Systemic identification of somatic muta­ tions in cancer genomes uncovers diverse mutations in human cancer to help understand the molecular mechanisms of tumori­ genesis and develop targeted therapeutics. Class I PI3K, espe­ cially p110α/p85α heterodimer, activation is closely linked with tumorigenesis [80, 81]. Most of the reported mutations in p110α



39:  Insulin‐Mediated PI3K and AKT Signaling

encoded by the PIK3CA cluster are conserved in the region for the helical (exon 9, E542K and E545K) and kinase domains (exon20, H1047R) of p110α [33]. These mutations in PIK3CA constitutively activate p110α kinase activity without upstream stimuli by growth factors, promoting uncontrolled proliferative signaling through the constitutive activation of AKT, S6K, and 4E‐BP in tumorigenesis. PIK3CA mutations in the helical and kinase domains also show a distinct pattern in gender and tissue specificity. In colorectal cancer, PIK3CA mutations are more fre­ quently found in women, and helical domain mutations (exon 9) demonstrate more effects than kinase domain mutations (exon 20). The mutations of PIK3CB encoding for the p110β subunit on E633K in the helical domain are linked with increased kinase activity and enhanced cell proliferation. Like the p110α H1047R mutation, the p110β E633K mutation enhances membrane tar­ geting to activate downstream signaling. Increased expression of p110γ has been reported in chronic myeloid leukemia, invasive breast carcinoma, and pancreatic ductal adenocarcinoma. Recently, somatic mutations in PIK3R1, which encodes p85α, are also found in cancers, and these mutations cluster in the iSH2 region (Figure 39.1a) releasing p110α to enhance kinase activity, resulting in the phosphorylation of AKT on S473 [82]. Tumor suppressor PTEN suppresses PI3K and AKT by dephosphoryla­ tion of PI(3,4,5)P3 and PI(3,4)P2, thus suppressing tumorigenesis. In addition to the alteration in PI3Ks, mutations of other signaling molecules in PI3K/AKT/mTORC axis such as PTEN, AKT, TSC1/,2 and mTOR are also present. PTEN is frequently mutated in human cancers including HCC (about 5% of HCC), and decreased PTEN levels are observed in human hepatic steatosis [83, 84]. Thus, liver‐specific deletion of the PTEN gene results in hyperplastic, fatty degeneration of the liver and increased cell proliferation [76]. Thus, the class I PI3K, p100α, appears to be an ideal target for drug development, whereas little is known about the genetic modification in class II and III PI3K in cancer. AKT regulates cell proliferation and survival and is one of the most activated downstream effectors of PI3K activation in tumo­ rigenesis. AKT phosphorylates GSK3β to prevent degradation of cyclin D1 and activates the mTORC pathway to enhance the translation of cyclin D1 and D3, resulting in cell proliferation. AKT also prevents programmed cell death through the phospho­ rylation of Bcl‐2‐associated death promoter (BAD) to suppress cytochrome c release from mitochondria and caspase activation by phosphorylation of pro‐caspase 9. AKT promotes additional processes such as angiogenesis by increased nitric oxide produc­ tion and metastasis through the increased secretion of matrix metalloproteinases and epithelial‐mesenchymal transition (EMT) [85]. Thus, AKT is closely linked with tumorigenesis. Indeed, the AKT activation is observed in human cancers through the amplification, overexpression, or point mutation of the AKT kinase genes. AKT1 amplification causes the resistance to cis­ platinum treatment in gastric carcinomas. Somatic mutations of AKT1 on E17K causes localization of AKT1 at the plasma mem­ brane to increase phosphorylation on Ser‐473 and Thr‐308, resulting in leukemia in the mouse model [86]. In HCC, enhanced expression of AKT2 protein was detected in about 40% tumors, whereas AKT1 expression was similar in all cases. As the PI3K/AKT/mTORC signaling axis is a major determi­ nant of tumorigenesis, several molecules for suppressing this pathway have been developed and evaluated in clinical trials

493

[87]. Pan‐PI3K inhibitors such as buparlisib (BKM120) and XL147 suppress all p110 isoforms, whereas isoform‐specific PI3K inhibitors are developed to minimize the side‐effects of pan‐PI3K inhibitors. p110α‐specific alpelisib and MLN1117 and p110β‐specific taselisib (GDC‐0032) are more effective to tumors with PIK3CA mutation. In contrast, p110α‐specific GSK2636771 is more effective in PTEN‐deficient tumors. AKT inhibitors such as AZD‐5363, GDC‐0068, and MK‐2206 are in clinical trials. MK‐2206, an allosteric pan‐AKT inhibitor, is in clinical trial to treat cancers such as breast and colorectal can­ cers. MK‐2206 suppresses phosphorylation of AKT Thr‐308 and Ser‐473, resulting in the suppression of cell proliferation in cell lines, which has PIK3CA activating mutations, PTEN inac­ tivation, and amplification of AKT [88]. As MK‐2206 treatment demonstrates limited antitumor activity in phase II clinical trial, combined treatment with chemotherapeutics and small mole­ cule inhibitors are in trials. Although rapalogs, the rapamycin analogs to suppress mTOR, are developed as immunosuppres­ sants, they also suppress cell proliferation and angiogenesis to treat cancer. Temsirolimus is the first rapalog that is approved by the Food and Drug Administration (FDA) for the treatment of renal cell carcinoma as it arrests the cell cycle in the G1 phase and suppresses angiogenesis by reducing vascular endothelial growth factor (VEGF) synthesis.

SUMMARY Numerous studies over the past decade have begun to unravel the complex molecular details and specific signaling pathways that regulate normal liver function with respect to hepatic glu­ cose and fatty acid metabolism. These efforts have also begun to elucidate the primary basis for dysregulated liver metabolism that occurs during states of insulin resistance and T2D. We now know that elevated fatty acid levels can serve as a significant contributor to hepatic glucose production. When coupled with liver insulin resistance this combination becomes a strong driver of unrestrained gluconeogenesis and hyperglycemia. On the other hand, hyperinsulinemia associated with insulin resistance activates the transcriptional network that drives lipogenic gene expression and DNL. When coupled with other dietary factors, fructose further enhances DNL through ChREBP that at the same time exacerbates hepatic glucose production through tran­ scriptional activation of G6PC. Several challenges remain but further understanding of the network and interplay between liver gluconeogenic and lipogenic signaling should provide new avenues to develop specific therapeutic approaches for the treat­ ment of metabolic liver disease.

REFERENCES 1. Vanhaesebroeck, B., Guillermet‐Guibert, J., Graupera, M., and Bilanges, B. The emerging mechanisms of isoform‐specific PI3K signalling. Nat Rev Mol Cell Biol, 2010;11(5):329–41. 2. Wymann, M.P., Bulgarelli‐Leva, G., Zvelebil, M.J. et  al. Wortmannin ­inactivates phosphoinositide 3‐kinase by covalent modification of Lys‐802, a  residue involved in the phosphate transfer reaction. Mol Cell Biol, 1996;16(4):1722–33.

494

THE LIVER:  REFERENCES

3. Backer, J.M. The regulation of class IA PI 3‐kinases by inter‐subunit interac­ tions. Curr Top Microbiol Immunol, 2010;346:87–114. 4. Carpenter, C.L., Auger, K.R., Chanudhuri, M. et  al. Phosphoinositide 3‐ kinase is activated by phosphopeptides that bind to the SH2 domains of the 85‐kDa subunit. J Biol Chem, 1993;268(13):9478–83. 5. Myers, M.G., Jr., Backer, J.M., Sun, X.J. et al. IRS‐1 activates phosphati­ dylinositol 3’‐kinase by associating with src homology 2 domains of p85. Proc Natl Acad Sci USA, 1992;89(21):10350–4. 6. Bellacosa, A., Testa, J.R., Staal, S.P., and Tsichlis, P.N. A retroviral onco­ gene, akt, encoding a serine‐threonine kinase containing an SH2‐like region. Science, 1991;254(5029):274–7. 7. Coffer, P.J. and Woodgett, J.R. Molecular cloning and characterisation of a novel putative protein‐serine kinase related to the cAMP‐dependent and pro­ tein kinase C families. Eur J Biochem, 1991;201(2):475–81. 8. Jones, P.F., Jakubowicz, T., Pitossi, F.J., Maurer, F., and Hemmings, B.A. Molecular cloning and identification of a serine/threonine protein kinase of the second‐messenger subfamily. Proc Natl Acad Sci USA, 1991;88(10):4171–5. 9. Staal, S.P. Molecular cloning of the akt oncogene and its human homologues AKT1 and AKT2: amplification of AKT1 in a primary human gastric adeno­ carcinoma. Proc Natl Acad Sci USA, 1987;84(14):5034–7. 10. Klarlund, J.K., Guilherme A., Holik, J.J., Virbasius, J.V., Chawla, A., and Czech, M.P. Signaling by phosphoinositide‐3,4,5‐trisphosphate through pro­ teins containing pleckstrin and Sec7 homology domains. Science, 1997;275(5308):1927–30. 11. Stephens, L., Anderson, K., Stokoe, D. et al. Protein kinase B kinases that mediate phosphatidylinositol 3,4,5‐trisphosphate‐dependent activation of protein kinase B. Science, 1998;279(5351):710–4. 12. Keranen, L.M., Dutil, E.M., and Newton, A.C. Protein kinase C is regulated in vivo by three functionally distinct phosphorylations. Curr Biol, 1995;5(12):1394–403. 13. Calleja, V., Alcor, D., Laguerre, M. et al. Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol, 2007;5(4):e95. 14. Sarbassov, D.D., Guertin, D.A., Ali, S.M., and Sabatini, D.M. Phosphorylation and regulation of Akt/PKB by the rictor‐mTOR complex. Science, 2005;307(5712):1098–101. 15. Pessin, J.E. and Saltiel, A.R. Signaling pathways in insulin action: molecular targets of insulin resistance. J Clin Invest, 2000;106(2):165–9. 16. Taniguchi, C.M., Emanuelli, B., and Kahn, C.R. Critical nodes in signalling pathways: insights into insulin action. Nat Rev Mol Cell Biol, 2006;7(2): 85–96. 17. Ullrich, A., Bell, J.R., Chen, E.Y. et al. Human insulin receptor and its rela­ tionship to the tyrosine kinase family of oncogenes. Nature, 1985;313(6005):756–61. 18. Ebina, Y., Ellis, L., Jarnagin, K. et al. The human insulin receptor cDNA: the structural basis for hormone‐activated transmembrane signalling. Cell, 1985;40(4):747–58. 19. Voliovitch, H., Schindler, D.G., Hadari, Y.R., Taylor, S.I., Accili, D., and Zick, Y. Tyrosine phosphorylation of insulin receptor substrate‐1 in vivo depends upon the presence of its pleckstrin homology region. J Biol Chem, 1995;270(30):18083–7. 20. Araki, E., Lipes, M.A., Patti, M.E. et al. Alternative pathway of insulin sig­ nalling in mice with targeted disruption of the IRS‐1 gene. Nature, 1994;372(6502):186–90. 21. Withers, D.J., Gutierrez, J.S., Towery, H. et al. Disruption of IRS‐2 causes type 2 diabetes in mice. Nature, 1998;391(6670):900–4. 22. Zhang, X., Vadas, O., Perisic, O. et al. Structure of lipid kinase p110beta/ p85beta elucidates an unusual SH2‐domain‐mediated inhibitory mechanism. Mol Cell, 2011;41(5):567–78. 23. Sopasakis, V.R., Liu, P., Suzuki, R. et  al. Specific roles of the p110alpha isoform of phosphatidylinsositol 3‐kinase in hepatic insulin signaling and metabolic regulation. Cell Metab, 2010;11(3):220–30. 24. Terauchi, Y., Tsuji, Y., Satoh, S. et al. Increased insulin sensitivity and hypo­ glycaemia in mice lacking the p85 alpha subunit of phosphoinositide 3‐ kinase. Nat Genet, 1999;21(2):230–5. 25. Wijesekara, N., Konrad, D., Eweida, M. et al. Muscle‐specific Pten deletion protects against insulin resistance and diabetes. Mol Cell Biol, 2005;25(3):1135–45. 26. Sleeman, M.W., Wortley, K.E., Lai, K.M. et  al. Absence of the lipid phosphatase SHIP2 confers resistance to dietary obesity. Nat Med, ­ 2005;11(2):199–205.

27. Jacinto, E., Facchinetti, V., Liu, D. et al. SIN1/MIP1 maintains rictor‐mTOR complex integrity and regulates Akt phosphorylation and substrate specific­ ity. Cell, 2006;127(1):125–37. 28. Cho, H., Mu, J., Kim, J.K., Thorvaldsen, J.L. et al. Insulin resistance and a diabetes mellitus‐like syndrome in mice lacking the protein kinase Akt2 (PKB beta). Science, 2001;292(5522):1728–31. 29. Manning, B.D. and Cantley, L.C. AKT/PKB signaling: navigating down­ stream. Cell, 2007;129(7):1261–74. 30. Cross, D.A., Alessi, D.R., Cohen, P., Andjelkovich, M., and Hemmings, B.A. Inhibition of glycogen synthase kinase‐3 by insulin mediated by protein kinase B. Nature, 1995;378(6559):785–9. 31. Saxton, R.A. and Sabatini, D.M. mTOR signaling in growth, metabolism, and disease. Cell, 2017;169(2):361–71. 32. Czech, M.P., Tencerova, M., Pedersen, D.J., and Aouadi, M. Insulin ­signalling mechanisms for triacylglycerol storage. Diabetologia, 2013;56(5): 949–64. 33. Perry, R.J., Camporez, J.P., Kursawe, R. et al. Hepatic acetyl CoA links adi­ pose tissue inflammation to hepatic insulin resistance and type 2 diabetes. Cell, 2015;160(4):745–58. 34. Rebrin, K., Steil, G.M., Mittelman, S.D., and Bergman, R.N. Causal linkage between insulin suppression of lipolysis and suppression of liver glucose output in dogs. J Clin Invest, 1996;98(3):741–9. 35. Ramnanan, C.J., Edgerton, D.S., Rivera, N. et al. Molecular characterization of insulin‐mediated suppression of hepatic glucose production in vivo. Diabetes, 2010;59(6):1302–11. 36. Perry, R.J., Zhang, X.M., Zhang, D. et  al. Leptin reverses diabetes by ­suppression of the hypothalamic‐pituitary‐adrenal axis. Nat Med, 2014;20(7): 759–63. 37. Petersen, M.C., Madiraju, A.K., Gassaway, B.M. et  al. Insulin receptor Thr1160 phosphorylation mediates lipid‐induced hepatic insulin resistance. J Clin Invest, 2016;126(11):4361–71. 38. Chin, J.E., Liu, F., and Roth, R.A. Activation of protein kinase C alpha inhibits insulin‐stimulated tyrosine phosphorylation of insulin receptor ­ ­substrate‐1. Mol Endocrinol, 1994;8(1):51–8. 39. Badin, P.M., Vila, I.K., Louche, K. et al. High‐fat diet‐mediated lipotoxicity and insulin resistance is related to impaired lipase expression in mouse skel­ etal muscle. Endocrinology, 2013;154(4):1444–53. 40. Griffin, M.E., Marcucci, M.J. et al. Free fatty acid‐induced insulin resistance is associated with activation of protein kinase C theta and alterations in the insulin signaling cascade. Diabetes, 1999;48(6):1270–4. 41. Jornayvaz, F.R. and Shulman, G.I. Diacylglycerol activation of protein kinase Cepsilon and hepatic insulin resistance. Cell Metab, 2012;15(5): 574–84. 42. Haemmerle, G., Lass, A., Zimmermann, R. et  al. Defective lipolysis and altered energy metabolism in mice lacking adipose triglyceride lipase. Science, 2006;312(5774):734–7. 43. Chavez, J.A. and Summers, S.A. A ceramide‐centric view of insulin resist­ ance. Cell Metab, 2012;15(5):585–94. 44. Shimomura, I., Matsuda, M., Hammer, R.E., Bashmakov, Y., Brown, M.S., and Goldstein, J.L. Decreased IRS‐2 and increased SREBP‐1c lead to mixed insulin resistance and sensitivity in livers of lipodystrophic and ob/ob mice. Mol Cell, 2000;6(1):77–86. 45. Rui, L., Yuan, M., Frantz, D., Shoelson, S., and White, M.F. SOCS‐1 and SOCS‐3 block insulin signaling by ubiquitin‐mediated degradation of IRS1 and IRS2. J Biol Chem, 2002;277(44):42394–8. 46. Kwon, H. and Pessin, J.E. Adipokines mediate inflammation and insulin resistance. Front Endocrinol, 2013;4:71. 47. Boura‐Halfon, S. and Zick, Y. Phosphorylation of IRS proteins, insulin action, and insulin resistance. Am J Physiol Endocrinol Metab, 2009;296(4): E581–91. 48. Hirosumi, J., Tuncman, G., Chang, L. et al. A central role for JNK in obesity and insulin resistance. Nature, 2002;420(6913):333–6. 49. Bard‐Chapeau, E.A., Hevener, A.L., Long, S., Zhang, E.E., Olefsky, J.M., and Feng, G.S. Deletion of Gab1 in the liver leads to enhanced glucose toler­ ance and improved hepatic insulin action. Nat Med, 2005;11(5):567–71. 50. Rider, M.H., Bertrand, L., Vertommen, D., Michels, P.A., Rousseau, G.G., and Hue, L. 6‐phosphofructo‐2‐kinase/fructose‐2,6‐bisphosphatase: head‐ to‐head with a bifunctional enzyme that controls glycolysis. Biochem J, 2004;381(Pt 3):561–79. 51. Koo, S.H., Flechner, L., Qi, L. et al. The CREB coactivator TORC2 is a key regulator of fasting glucose metabolism. Nature, 2005;437(7062):1109–11.



39:  Insulin‐Mediated PI3K and AKT Signaling

52. Wang, Y., Li G., Goode, J. et  al. Inositol‐1,4,5‐trisphosphate receptor ­regulates hepatic gluconeogenesis in fasting and diabetes. Nature, 2012; 485(7396):128–32. 53. Dentin, R., Liu, Y., Koo, S.H. et al. Insulin modulates gluconeogenesis by inhibition of the coactivator TORC2. Nature, 2007;449(7160):366–9. 54. Brunet, A., Bonni, A., Zigmond, M.J. et al. Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell, 1999;96(6):857–68. 55. Tzivion, G., Dobson, M., and Ramakrishnan, G. FoxO transcription ­factors;  regulation by AKT and 14‐3‐3 proteins. Biochim Biophys Acta, 2011;1813(11):1938–45. 56. O’Sullivan, I., Zhang, W., Wasserman, D.H. et al. FoxO1 integrates direct and indirect effects of insulin on hepatic glucose production and glucose utilization. Nat Commun, 2015;6:7079. 57. Titchenell, P.M., Chu, Q., Monks, B.R., and Birnbaum, M.J. Hepatic insulin signalling is dispensable for suppression of glucose output by insulin in vivo. Nat Commun, 2015;6:7078. 58. Samuel, V.T., Beddow, S.A., Iwasaki, T. et al. Fasting hyperglycemia is not associated with increased expression of PEPCK or G6Pc in patients with Type 2 diabetes. Proc Natl Acad Sci USA, 2009;106(29):12121–6. 59. Kim, M.S., Krawczyk, S.A., Doridot, L. et al. ChREBP regulates fructose‐ induced glucose production independently of insulin signaling. J Clin Invest, 2016;126(11):4372–86. 60. Solinas, G., Boren, J., and Dulloo, A.G. De novo lipogenesis in metabolic homeostasis: more friend than foe? Mol Metab, 2015;4(5):367–77. 61. Wang, Y., Viscarra, J., Kim, S.J., and Sul, H.S. Transcriptional regulation of hepatic lipogenesis. Nat Rev Mol Cell Biol, 2015;16(11):678–89. 62. Hegarty, B.D., Bobard, A., Hainault, I., Ferre, P., Bossard, P., and Foufelle, F. Distinct roles of insulin and liver X receptor in the induction and cleavage of  sterol regulatory element‐binding protein‐1c. Proc Natl Acad Sci USA, 2005;102(3):791–6. 63. Peterson, T.R., Sengupta, S.S., Harris, T.E. et al. mTOR complex 1 regulates lipin 1 localization to control the SREBP pathway. Cell, 2011;146(3): 408–20. 64. Dentin, R., Tomas‐Cobos, L., Foufelle, F. et al. Glucose 6‐phosphate, rather than xylulose 5‐phosphate, is required for the activation of ChREBP in response to glucose in the liver. J Hepatol, 2012;56(1):199–209. 65. Uyeda, K. and Repa, J.J. Carbohydrate response element binding protein, ChREBP, a transcription factor coupling hepatic glucose utilization and lipid synthesis. Cell Metab, 2006;4(2):107–10. 66. Brownsey, R.W., Boone, A.N., Elliott, J.E., Kulpa, J.E., and Lee, W.M. Regulation of acetyl‐CoA carboxylase. Biochem Soc Trans, 2006;34(2):223–7. 67. Peng, I.C., Chen, Z., Sun, W. et al. Glucagon regulates ACC activity in adi­ pocytes through the CAMKKbeta/AMPK pathway. Am J Physiol Endocrinol Metab, 2012;302(12):E1560–8. 68. Witters, L.A., Watts, T.D., Daniels, D.L., and Evans, J.L. Insulin stimulates the dephosphorylation and activation of acetyl‐CoA carboxylase. Proc Natl Acad Sci USA, 1988;85(15):5473–7. 69. Zammit, V.A. Role of insulin in hepatic fatty acid partitioning: emerging concepts. Biochem, J., 1996;314(1):1–14.

495

70. Lazo, M. and Clark, J.M. The epidemiology of nonalcoholic fatty liver dis­ ease: a global perspective. Semin Liver Dis, 2008;28(4):339–50. 71. Marchesini, G., Brizi, M., Bianchi, G. et al. Nonalcoholic fatty liver disease: a feature of the metabolic syndrome. Diabetes, 2001;50(8):1844–50. 72. Henry, R.R., Scheaffer, L., and Olefsky, J.M. Glycemic effects of intensive caloric restriction and isocaloric refeeding in noninsulin‐dependent diabetes mellitus. J Clin Endocrinol Metab, 1985;61(5):917–25. 73. Biddinger, S.B. and Kahn, C.R. From mice to men: insights into the insulin resistance syndromes. Annu Rev Physiol, 2006;68:123–58. 74. Kudo, Y., Tanaka, Y., Tateishi, K. et al. Altered composition of fatty acids exacerbates hepatotumorigenesis during activation of the phosphatidylinosi­ tol 3‐kinase pathway. J Hepatol, 2011;55(6):1400–8. 75. Chattopadhyay, M., Selinger, E.S., Ballou, L.M., and Lin, R.Z. Ablation of PI3K p110‐alpha prevents high‐fat diet‐induced liver steatosis. Diabetes, 2011;60(5):1483–92. 76. Horie, Y., Suzuki, A., Kataoka, E. et al. Hepatocyte‐specific Pten deficiency results in steatohepatitis and hepatocellular carcinomas. J Clin Invest, 2004;113(12):1774–83. 77. Calvisi, D.F., Wang, C., Ho, C. et al. Increased lipogenesis, induced by AKT‐ mTORC1‐RPS6 signaling, promotes development of human hepatocellular carcinoma. Gastroenterology, 2011;140(3):1071–83. 78. Williams, A.S., Trefts, E., Lantier, L. et al. Integrin‐linked kinase is neces­ sary for the development of diet‐induced hepatic insulin resistance. Diabetes, 2017;66(2):325–34. 79. Auger, K.R., Serunian, L.A., Soltoff, S.P., Libby, P., and Cantley, L.C. PDGF‐dependent tyrosine phosphorylation stimulates production of novel polyphosphoinositides in intact cells. Cell, 1989;57(1):167–75. 80. Yuan, T.L. and Cantley, L.C. PI3K pathway alterations in cancer: variations on a theme. Oncogene, 2008;27(41):5497–510. 81. Samuels, Y., Wang, Z., Bardelli, A. et al. High frequency of mutations of the PIK3CA gene in human cancers. Science, 2004;304(5670):554. 82. Jaiswal, B.S., Janakiraman, V., Kljavin, N.M. et  al. Somatic mutations in p85alpha promote tumorigenesis through class IA PI3K activation. Cancer Cell, 2009;16(6):463–74. 83. Xu, Z., Hu, J., Cao, H. et al. Loss of Pten synergizes with c‐Met to promote hepatocellular carcinoma development via mTORC2 pathway. Exp Mol Med, 2018;50(1):e417. 84. Whittaker, S., Marais, R., and Zhu, A.X. The role of signaling pathways in the development and treatment of hepatocellular carcinoma. Oncogene, 2010;29(36):4989–5005. 85. Grille, S.J., Bellacosa, A., Upson, J. et  al. The protein kinase Akt induces epithelial mesenchymal transition and promotes enhanced motility and inva­ siveness of squamous cell carcinoma lines. Cancer Res, 2003;63(9):2172–8. 86. Carpten, J.D., Faber, A.L., Horn, C. et  al. A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature, 2007;448(7152): 439–44. 87. Mayer, I.A. and Arteaga, C.L. The PI3K/AKT pathway as a target for cancer treatment. Annu Rev Med, 2016;67:11–28. 88. Sangai, T., Akcakanat, A., Chen, H. et  al. Biomarkers of response to Akt  inhibitor MK‐2206 in breast cancer. Clin Cancer Res, 2012;18(20): 5816–28.

40

Ca2+ Signaling in the Liver Mateus T. Guerra1, M. Fatima Leite2, and Michael H. Nathanson1 Departments of Medicine and Cell Biology, Yale University School of Medicine, New Haven, CT, USA Department of Physiology and Biophysics, UFMG, Belo Horizonte, Brazil

1 2

MECHANISMS OF Ca2+ SIGNALING Hormone receptors and initiation of Ca2+ signals Hormones, neurotransmitters, and growth factors initiate cytosolic Ca2+ (Cai2+) signaling through a variety of mechanisms [1]. Most commonly, these factors trigger intracellular signaling cascades by binding to either G protein‐coupled receptors (GPCRs) or receptor tyrosine kinases (RTKs) on the surface of liver cells. Several GPCRs in the liver have been particularly well studied. These include the V1a vasopressin receptor, the α1B adrenergic receptor, several subtypes of the P2Y class of purinergic receptors, and the angiotensin receptors. RTKs for several types of growth factors are expressed in liver as well, such as insulin, epidermal growth factor (EGF), and hepatocyte growth factor (HGF). Upon activation, each of these receptors initiates signaling events that lead to an increase in cytosolic and/or nuclear Ca2+. The binding of Ca2+ mobilizing hormones and growth factors to their specific plasma membrane receptors activates phospholipase C (PLC), which is membrane‐associated. α, β, and γ subtypes of PLC are recognized [2]. Many isoforms of these subtypes have been identified; PLCβ1 and PLCβ2 are the isoforms activated by G proteins, while PLCγ is activated by RTKs [2]. Activation of PLC by GPCRs hydrolyzes the pool of phospholipid phosphatidylinositol‐4,5‐bisphosphate (PIP2) within the plasma membrane. The hydrolysis of PIP2 by  PLC results in the formation of diacylglycerol (DAG) and inositol 1,4,5 trisphosphate (InsP3). DAG remains at the plasma membrane to activate protein kinase C (PKC), while InsP3 diffuses into the cytosol to release Ca2+ from intracellular stores via its interaction with the InsP3 receptor (InsP3R). Activation of PLC by RTKs was thought to similarly act on the plasma

membrane pool of PIP2 [1], but several lines of evidence now suggest it instead hydrolyzes nuclear PIP2, which generates InsP3 in the nucleus to increase free Ca2+ within the nucleoplasm [3]. The sequence of events linking hormone receptors and RTKs to Cai2+ signaling is summarized in Figure 40.1.

Inositol 1,4,5‐trisphosphate receptor The InsP3R is an InsP3‐gated Ca2+ channel located on the endoplasmic reticulum (ER), although there may also be active channels at the plasma membrane. In hepatocytes, increases in Cai2+ are initiated by binding of InsP3 to the InsP3R. Full length sequences for three distinct InsP3R genes have been determined and knockout mice have been generated for each of these three isoforms [4]. These isoforms share considerable sequence homology, but each subtype is expressed and regulated in a distinct fashion. There also are isoform‐specific differences in tissue expression and subcellular distribution, suggesting that the isoforms serve distinct roles in Cai2+ signaling. InsP3Rs are homotetramers consisting of 313, 307, or 304 KDa subunits, corresponding to the type I, II, or III isoforms, respectively, and heterotetramers can form as well [5]. The InsP3R has six membrane spanning domains, oriented so that the N‐terminus of the protein is in the cytoplasm. The receptor has an uneven bell shape as deduced by cryo‐electron microscopy with the bulkier N‐terminus directed toward the cytoplasm and the narrower end facing the ER lumen. Deletion analysis studies of the mouse InsP3R1 have revealed three functional regions within the InsP3R: an N‐terminal InsP3‐binding domain, a Ca2+ channel‐forming C‐terminal domain, and a regulatory domain flanked by the InsP3‐binding domain and the channel region. Several phosphorylation sites are found along the InsP3R’s amino acid sequence. Moreover,

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



40: Ca2+ SIGNALING IN THE LIVER

497

Growth factor

Hormone

PLC γ

PLC β G-protein coupled receptor

InsP3

Gprotein

Receptor tyrosine kinase

Ca2+

InsP3R

Endoplasmic reticulum

To nucleus

Ca2+ Figure 40.1  Mechanisms of hormone‐ and growth factor‐induced Ca2+ signaling in hepatocytes. Upon binding to its specific G protein‐coupled plasma membrane receptor, a hormone induces phospholipase C (PLC)‐beta to hydrolyze phosphatidylinositol 4,5 bisphosphate (PIP2) to form diacylglycerol (DAG) and inositol 1,4,5‐trisphosphate (InsP3). Alternatively, upon binding to its specific receptor tyrosine kinase (RTK), a growth hormone induces PLC‐gamma to hydrolyze PIP2 to form DAG and InsP3. InsP3 then binds to its tetrameric receptor (InsP3R) in the ER, which acts as a Ca2+ channel to allow Ca2+ to enter the cytosol. RTKs such as c‐Met and the insulin receptor may also translocate to the nucleus in hepatocytes to activate nuclear PLC and increase Ca2+ within the nucleoplasm.

the regulatory domain interacts with several protein partners responsible for modulating channel activity. The C‐terminus also interacts with protein partners, which influences the subcellular localization of the receptor [6]. The InsP3 binding domain includes multiple sequences scattered throughout the N‐terminal region and key residues responsible for the interaction between InsP3 and its receptor have been identified by site‐directed mutagenesis and X‐ray crystallography [7]. Upon InsP3 binding, the receptor undergoes a conformational change that opens the Ca2+ channel, so that Ca2+ in the ER is released into the cytosol. Although each of the three InsP3R isoforms acts as an InsP3‐gated Ca2+ channel, the isoforms are not uniformly sensitive to InsP3. The relative order of affinity is type II > type I > type III. The Kd for InsP3R2 is 27 nM, which is twice as great as the affinity of InsP3R1, and ten times the affinity of InsP3R3. InsP3 is absolutely required for Ca2+ release via the InsP3R, but the concentration of Ca2+ in the cytosol modulates the open probability of the Ca2+ channel [8]. This dependence of the InsP3R on the cytosolic Ca2+ concentration is important for organizing the spatial and temporal pattern of Cai2+ signals. The activity of InsP3R Ca2+ channels is highly dependent on post‐translational modifications such as phosphorylation, binding of protein cofactors [6], and O‐linked β‐N‐acetylglucosamine glycosylation (O‐GlcNAcylation) [9], which operates by

reversible addition of N‐acetylglucosamine monosaccharide to serine/threonine residues. It is speculated that O‐GlcNAcylation of InsP3R in liver may be particularly important in states of high nutrient availability such as NAFLD or diabetes. InsP3R activity can also be modulated by protein degradation through the proteasome pathway or by selective proteolysis, providing yet another level of regulation of this Ca2+ release channel. Many cell types express more than one InsP3R isoform. Hepatocytes normally express only InsP3R1 and InsP3R2. InsP3R3 is not detected in hepatocytes, but is the predominant isoform in bile duct epithelia [10]. Moreover, InsP3R2 is most concentrated in the apical region of hepatocytes while InsP3R1 is dispersed throughout the cell (Figure 40.2a) [11]. In contrast, in bile duct epithelia InsP3R3 is most concentrated in the apical region, while InsP3R1 and InsP3R2 are dispersed throughout the cell (Figure 40.2b) [12]. The apical localization of InsP3R2 in hepatocytes depends upon lipid rafts in the canalicular membrane [13], but the mechanism by which InsP3Rs associate with lipid rafts is not yet known.

Ryanodine receptor The other major class of intracellular Ca2+ release channels is the ryanodine receptor (RyR). Like InsP3R, RyR has three

498

THE LIVER:  MECHANISMS OF Ca2+ SIGNALING

(a)

(b)

Figure 40.2  InsP3 receptors (InsP3Rs) are concentrated in the apical region of hepatocytes and cholangiocytes. (a) Confocal immunofluorescence image of rat liver shows the distribution of the InsP3R2 (green) in hepatocytes. Cells also are labeled with the actin stain phalloidin (red), which outlines individual hepatocytes and labels their apical region most intensely. InsP3R2 colocalizes with periapical actin (arrowheads), and thus is most concentrated in the apical region. (b) Confocal immunofluorescence image of a bile duct from a biopsy of a normal human liver shows the distribution of InsP3R3 (green). Cholangiocytes also are labeled to reveal CFTR (red), which resides on their apical membrane. The InsP3R3 is most concentrated in the apical region, just beneath CFTR. Reproduced from [105] with permission of Elsevier.

family members (RyR1, RyR2, and RyR3). These channels contribute to Cai2+ signaling by releasing Ca2+ from the lumen of the sarcoplasmic/endoplasmic reticulum into the cytosol [1]. RyRs autocatalytically release Ca2+ via Ca2+‐induced Ca2+ release (CICR). Rat hepatocytes express only a truncated RyR1 isoform that by itself is not able to elicit Ca2+ release but can increase the frequency of InsP3‐induced Ca2+ oscillations [14]. These findings suggest there is a novel RyR‐like protein which contributes to Ca2+ signaling in hepatocytes. Nicotinic acid adenine dinucleotide phosphate (NAADP) is a second messenger that mediates release of Ca2+ from intracellular acidic compartments such as lysosomes and secretory granules. The role of NAADP in hepatocyte Cai2+ signaling is not clear, although there is in vitro evidence demonstrating NAADP‐dependent Ca2+ release from reconstituted hepatocyte lysosomes and microsomes [15].

Mitochondria Mitochondria are best known for the metabolic and respiratory role they play in cells. However, mitochondria strongly influence Cai2+ signaling as well, by taking up Ca2+ from and releasing it back into the cytosol. Mitochondria have their own Ca2+ transport machinery, involving Ca2+ influx through a uniporter, and Ca2+ efflux via both a Na+ exchanger and a H+ exchanger [16]. The uniporter is driven by the potential gradient across the mitochondrial membrane, while the Ca2+ efflux mechanisms are  active transport systems. The CCDC109A gene encodes the  mitochondrial calcium uniporter (MCU) protein, and Ca2+ influx in mitochondria occurs via a macromolecular complex on the inner mitochondrial matrix composed of pore‐forming subunits (MCU, MCUb, and essential MCU regulator [EMRE]), plus regulatory proteins MICU1/2 that modulate the activity of the uniporter according to the concentration of cytosolic Ca2+ in  the vicinity of the uniporter [16]. Pathological Ca2+ efflux from mitochondria also can occur through the permeability transition pore (PTP). Formation of this pore results in a sudden, marked increase in the permeability of the mitochondrial inner

membrane to ions and small molecules. Irreversible formation of the PTP can dissipate the potential gradient across the mitochondrial membrane, leading to mitochondrial swelling and irreversible cell injury, including apoptosis and necrosis. Indeed, liver regeneration after partial hepatectomy is halted due to massive necrotic cell death in mice lacking MICU1 in hepatocytes, which undergo sustained elevations in mitochondrial Ca2+ [17]. Mitochondria, like the ER, are densely distributed in hepatocytes. There are regions of close proximity between these two organelles (Figure 40.3), and InsP3R1 clustered in these regions are largely responsible for mitochondrial Ca2+ signaling in hepatocytes [18]. The discovery of interactions between apoptotic proteins, InsP3R, and mitochondrial proteins helped uncover an essential interplay between ER and mitochondrial Ca2+ signaling in the process of programmed cell death. The current view is that anti‐ apoptotic proteins belonging to the Bcl‐2 family such as Bcl‐XL and Bcl‐2 induce ER Ca2+ leak through direct interaction with InsP3R1. This Ca2+ leak then reduces the ER Ca2+ available to be released into the cytosol and consequently, decreases the amount of Ca2+ taken up by surrounding mitochondria. Because mitochondria are less likely to become overloaded with Ca2+, PTP formation is hampered and there is diminished cellular sensitivity to apoptotic stimuli [19]. Accordingly, buffering of mitochondrial matrix Ca2+ prevents apoptotic death of hepatocytes and accelerates liver regeneration after partial hepatectomy [20]. Mcl‐1 is another anti‐apoptotic protein that acts in part by diminishing mitochondrial Ca2+ signals. Unlike other Bcl‐2 family members, Mcl‐1 is expressed in mitochondria and ­inhibits mitochondrial Ca2+ signals directly [21]. Mcl‐1 is the principal anti‐apoptotic protein in cholangiocytes, and its overexpression may promote the development of cholangiocarcinoma. In contrast, overexpression of the pro‐apoptotic proteins Bax and Bak lead to ER Ca2+ overload and greater susceptibility to Ca2+‐mediated apoptosis. Mitochondria can sequester significant amounts of Ca2+ from the cytosol and mitochondrial Ca2+ can closely parallel the ­cytosolic Ca2+ increase induced by receptor activation. This



40: Ca2+ SIGNALING IN THE LIVER

(a)

499

(b)

Figure 40.3  Mitochondria and endoplasmic reticulum form regions that are in close proximity in hepatocytes. (a) Histologically normal human liver biopsy specimen stained with markers for an ER protein (PDI, green) and a mitochondrial protein (Tom‐22, red) show the close association between these two organelles in hepatocytes. Regions in white represent areas of colocalization, scale bar = 20 μM. (b) EM micrograph of a mouse hepatocyte showing regions of close (less than 40 nm) contact between the ER (green) and mitochondria (magenta). Mitochondrial Ca2+ signals are derived from Ca2+ that is released from InsP3Rs that cluster in these regions. Reproduced from [18] with permission of John Wiley and Sons.

functional effect is consistent with electron microscopy data that show a close proximity between mitochondria and ER and a dynamic physical interaction between these two organelles [22]. There is also functional evidence for microdomains where mitochondria and InsP3Rs are in close apposition, so that mitochondria take up a significant fraction of the Ca2+ released by the InsP3R [23]. In hepatocytes, InsP3R1 is the isoform in proximity to mitochondria and is responsible for mitochondrial Ca2+ signals [18]. Ca2+ uptake by mitochondria affects multiple factors in cell metabolism, including the mitochondrial proton motive force, electron transport, and the activity of dehydrogenases associated with the TCA cycle, adenine nucleotide translocase, the F1‐ATPase, and PDH phosphorylase. Slow or small Cai2+ elevations are not transmitted effectively into mitochondria, and are less able to activate mitochondrial metabolism. In contrast, Cai2+ oscillations trigger mitochondrial Ca2+ oscillations and sustained NAD(P)H formation [24]. Thus, the frequency rather than the amplitude of Cai2+ oscillations regulates mitochondrial metabolism [24].

ORGANIZATION OF Ca SIGNALS 2+

Detection of Ca signals in hepatocytes 2+

Our understanding of the complexity of Cai2+ signals has evolved dramatically, largely due to two technical advances: (i) development of Ca2+‐sensitive fluorescent dyes and proteins has permitted Cai2+ to be monitored continuously in live cells, and (ii) improvements in fluorescence imaging techniques allows Cai2+ to be detected not only in cell populations but in single cells and in distinct subcellular regions of individual cells [25]. As observations move from cell populations to single cells to subcellular regions, the complexity of Cai2+ signaling patterns increases and the time scale of signaling events decreases (Figure 40.4). It is now appreciated that Cai2+ signaling is regulated at the subcellular level, and that this level of regulation is necessary for Cai2+

in turn to act as a second messenger that regulates multiple cell functions simultaneously. Ca2+ signaling was initially studied with fluorescent dyes and the bioluminescent protein aequorin, but genetically encoded fluorescent Ca2+ indicators (GECI) are now more widely used. These GECI are based on modified versions of the circularly permuted (cp) green fluorescent protein (GFP) and other fluorescent proteins such as blue fluorescent protein (BFP) and mApple red fluorescent protein (RFP) [26]. These cp proteins are fused with the Ca2+ binding protein calmodulin (CaM) and a calmodulin (CaM)–binding region of chicken myosin light chain kinase (M13). Ca2+ binding induces a conformational change that results in increased fluorescence. Additional refinements have yielded ratiometric GECI and sensors with different affinities for Ca2+. Other GECI have been developed that are based on the principle of fluorescence resonance energy transfer (FRET). These indicators provide the advantage of better signal to noise ratio and photostability as compared to the aequorin reporters and yet retain the ability to be expressed in diverse tissues and subcellular locations.

Ca2+ signaling patterns in hepatocytes Peptide hormones such as vasopressin or angiotensin generally induce biphasic increases in Cai2+ in populations of isolated hepatocytes (Figure 40.4). These increases typically consist of two components. The first component is the rapid peak, then fall in Cai2+ which takes place over a period of seconds. This is due to release of Ca2+ from InsP3‐sensitive stores and occurs even in Ca2+‐free medium. The second component is the sustained plateau in Cai2+, which follows the rapid peak. This occurs only in the presence of extracellular Ca2+ and is due mostly to influx of Ca2+ to replenish depleted intracellular stores. Although this population response is highly reproducible, Ca2+ signaling patterns vary markedly among single hepatocytes. Different stimuli evoke distinct responses, and additional variation occurs among hepatocytes stimulated under identical conditions [27]. The range of signaling patterns seen among single hepatocytes

500

THE LIVER:  ORGANIZATION OF Ca2+ SIGNALS

Cytosolic Ca2+ (nM)

(a)

700 600 500 400 300 200 100 0 −50

Fluorescence intensity (pixel value)

(b)

0

50

100 150 Time (sec)

200

250

250 200 150

200

100 50

150

0

100 200 300 Time (sec)

100 50

0

(c)

10

20 30 Relative time (sec)

40

0.5 1.0 Relative time (sec)

1.5

Fluorescence intensity (pixel value)

250 200 150 100 50 0.0

Figure 40.4  Different views of cytosolic Ca2+ signaling in hepatocytes. In each case, isolated rat hepatocytes were stimulated with the α1B‐adrenergic agonist phenylephrine. (a) In a population of hepatocytes, a single transient peak is observed, followed by a sustained elevation. (b) In a single hepatocyte, a series of repetitive peaks (oscillations) are observed. (c) In different regions of the same hepatocyte, the increase in Ca2+ occurs at different times. This represents a Ca2+ wave. Notice that these Ca2+ signaling events occur over progressively shorter  time intervals as the level of focus moves from populations to  single cells to subcellular regions. Reproduced from [25] with ­permission of Elsevier.

includes single transient or sustained Cai2+ increases and repetitive Cai2+ spikes (i.e. oscillations). For example, lower concentrations of vasopressin induce Cai2+ oscillations, while higher concentrations induce sustained increases in Cai2+ [28], while stimulation of hepatocytes with phenylephrine typically evokes Cai2+ oscillations, but the oscillation frequency is dose‐dependent [28]. The duration of individual Cai2+ spikes also depends upon the agonist. For example, Cai2+ spikes induced by phenylephrine are short (around seven seconds) compared to the duration of spikes induced by vasopressin (around ten seconds) or angiotensin (around fifteen seconds). The frequency of vasopressin‐ induced Cai2+ oscillations tends to be greater than that of

phenylephrine‐induced oscillations as well. Differences in the frequency of Cai2+ oscillations regulate gene transcription in  some cell systems [29], although this has not yet been ­demonstrated in hepatocytes. It is currently thought that Cai2+ oscillations do not depend upon InsP3 oscillations. Instead, oscillations are thought to result from the bell‐shaped dependence of the open probability of the InsP3R on Cai2+. Extracellular Ca2+ contributes to Cai2+ oscillations in hepatocytes, because Cai2+ oscillations gradually dissipate in Ca2+‐free medium. Extracellular Ca2+ thus serves to maintain internal Ca2+ stores, which are the primary source of Ca2+ for oscillations in hepatocytes. The mechanism by which Ca2+ release from the ER triggers Ca2+ influx through the plasma membrane has been described in many cell types including hepatocytes [30]. The process of store‐operated Ca2+ entry (SOCE) involves the interaction of the integral ER membrane protein stromal interacting molecule 1 (STIM1) and its binding partner on the plasma membrane, the ORAI calcium release‐ activated calcium modulator (ORAI) channels [31]. STIM1 works as a Ca2+ sensor in the ER; once ER Ca2+ is reduced due to InsP3‐dependent activation or pharmacological inhibition of the sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) pump, STIM1 forms homodimers that interact with and gate ORAI channels on the plasma membrane, leading to influx of Ca2+ into the cytosol. This Ca2+ influx ultimately assists in replenishing ER Ca2+ stores. Increases in Cai2+ in hepatocytes typically begin near the apical membrane, where InsP3R2 is most concentrated [32]. Both vasopressin‐ and phenylephrine‐induced Cai2+ waves began there, then spread in a non‐diminishing fashion across the cell. Immunofluorescence studies in rat liver, isolated rat hepatocyte couplets, and hepatocytes in collagen sandwich culture established that InsP3R2 is concentrated subapically [33, 34] and that this localization depends on intact lipid rafts (specialized membrane patches rich in cholesterol) [13] (Figure 40.2). Examination of Cai2+ signaling in polarized preparations of isolated hepatocytes has shown that this is the region where Cai2+ waves originate, and that this depends on local clustering of InsP3R2 [13, 32]. Ca2+ signals in cholangiocytes begin in the apical region, just as they do in hepatocytes, even though InsP3R3 rather than InsP3R2 is concentrated near their apical surface, where the Ca2+ waves originate [12].

Spread of Ca2+ signals among hepatocytes Cai2+ signals occur asynchronously among isolated hepatocytes. For example, the lag time between stimulation with vasopressin and initiation of Cai2+ signaling varies among isolated hepatocytes by up to several seconds, while the frequency of Ca2+ oscillations can vary by up to 50% among isolated hepatocytes stimulated with phenylephrine [27]. However, hepatocytes that communicate via gap junctions coordinate their Cai2+ signals. For example, stimulation of isolated hepatocyte couplets with vasopressin induces a single Cai2+ wave that crosses both of the cells, while stimulation with phenylephrine induces Cai2+ oscillations that are synchronized in the two cells [27]. In the isolated perfused rat liver, Cai2+ signaling displays an even higher level of organization, because vasopressin induces Cai2+ waves that



40: Ca2+ SIGNALING IN THE LIVER

501

(a) 0 sec.

4 sec.

30 sec.

(b)

[Ca2+]i (F/F0)

1.5

1 0

100 200 Time (s)

Figure 40.5  Organization of Ca2+ signals in the intact liver. (a) Serial confocal images of a vasopressin‐induced Ca2+ wave in the isolated perfused rat liver demonstrates that the increase in Ca2+ begins in the pericentral region, then spreads as a wave toward the periportal regions. Images were obtained at baseline and after 4 and 30 seconds of stimulation with vasopressin (20 nM). Reproduced from [35] with permission of the American Physiological Society. (b) Ca2+ signals in individual hepatocytes within an isolated perfused rat liver stimulated with vasopressin. Tracing shows Ca2+ oscillations detected in three hepatocytes sequentially arranged along the hepatic lobule. The distance between the first (red tracing) and third (green tracing) cell is 40 μm. The phase difference between the cells is due to the time needed for a Ca2+ wave to spread across the hepatic lobule. Reproduced from [106] with permission of Elsevier.

cross the entire lobule [35] (Figure 40.5). Cai2+ waves cross individual hepatocytes at the same speed, regardless of whether the hepatocytes are isolated or within the liver. Vasopressin‐induced Cai2+ waves cross the hepatic lobule in a pericentral‐to‐periportal direction [36], presumably directed by the V1a vasopressin receptor gradient present from the pericentral to periportal region [36]. In contrast, ATP induces Cai2+ signals in a random fashion across the hepatic lobule, consistent with the lack of a P2Y receptor gradient across the lobule [37]. Thus, sophisticated patterns of Cai2+ signaling are induced in the intact liver, and these patterns are agonist‐specific. This may permit different Ca2+ agonists to have distinct effects in liver even though Cai2+ signals induced by these agonists appear similar in isolated hepatocytes. The basis for organization of Cai2+ signals in liver has been studied in multicellular systems of hepatocytes. Studies in isolated rat hepatocyte couplets demonstrate that hepatocytes communicate via gap junctions, and that both Ca2+ and InsP3 can cross these gap junctions [38]. Hormone‐induced Cai2+ signaling is highly coordinated in such couplets, and this coordination depends upon gap junction conductance as well [27]. Hepatocytes express two gap junction isoforms, connexin 32 (Cx32) and connexin 26 (Cx26) [39]. Expression of both isoforms is dramatically reduced after bile duct ligation, and coordination of Cai2+ signals is impaired under this condition as well [39]. Furthermore, cell‐to‐cell spread of InsP3 and Cai2+ waves is markedly impaired in hepatocytes isolated from Cx32 knockout mice [40]. Moreover, the expression of Cx32 or Cx43 in a liver cell line, where intercellular Ca2+ signals are not normally observed, causes propagation of Ca2+ from cell to cell [41]. Thus, gap junctions play an essential role in organizing Cai2+

signals among adjacent hepatocytes. This also is important for associated physiological responses. For example, bile secretion in isolated perfused rat liver is reduced by pharmacological inhibition of connexin function [42]. Another physiological process that requires gap junction communication among hepatocytes is glucose output [43] and this is impaired in Cx32 knockout mice stimulated with either glucagon or norepinephrine [44]. Cx32 function also is involved in drug‐induced liver injury, because mice defective in Cx32 are partially protected from acetaminophen‐induced hepatocyte cell death and liver failure [45]. Organization of cell‐to‐cell Cai2+ waves depends on other factors in addition to gap junctions. For example, increases in InsP3 are required in each cell across which a Ca2+ wave spreads [46]. Moreover, neither InsP3 nor Ca2+ alone is sufficient to support the spread of a Cai2+ wave across a hepatocyte [47]. The presence of agonist binding to its specific receptor at the surface of the cell also is required to support the spread of Cai2+ waves. This was demonstrated in experiments in which one or both cells of a hepatocyte couplet were microperfused with norepinephrine. Stimulation of individual cells evoked Cai2+ oscillations only in the perfused cell, and perfusion of an entire couplet was necessary to evoke Cai2+ oscillations in both cells [47]. Thus, the presence of hormone at each cell ensures that sufficient levels of intracellular messengers are generated in order to reach a level of excitability necessary for supporting the propagation of a Cai2+ wave. Other studies have focused on the mechanism by which intercellular Cai2+ waves become oriented within the hepatic acinus. Vasopressin‐induced waves begin in the region of the central venule [36], while ATP‐induced Cai2+ waves begin in a

502

THE LIVER:  ORGANIZATION OF Ca2+ SIGNALS

seemingly random pattern across the hepatic acinus [35]. Studies in isolated hepatocytes similarly show that pericentral hepatocytes are more sensitive to vasopressin but not to ATP [37]. Studies in isolated hepatocyte couplets and triplets stimulated with vasopressin or norepinephrine show that one of the cells generally has increased sensitivity to a particular hormone and acts as a pacemaker to drive Cai2+ oscillations in neighboring cells. This increased sensitivity is likely due to increased expression of hormone receptor rather than differences in downstream signaling components such as G proteins or InsP3R [48]. Cells with increased expression of hormone receptor produce higher concentrations of InsP3, so they respond sooner than other cells stimulated with the same concentration of hormone. As a result, cells with the greatest level of hormone receptor expression act as pacemakers, and different cells can act as the pacemaker for different hormones. Thus, the pattern of Cai2+ waves and oscillations in the intact liver depends upon multiple factors, which include: (i) the establishment of pacemaker cells by virtue of increased expression of hormone receptors, (ii) simultaneous stimulation of both pacemaker and non‐pacemaker cells, and (iii) communication of second messengers among these cells via gap junctions [49]. The liver also possesses paracrine mechanisms for generating and regulating Cai2+ signals. Hepatocytes and bile duct cells each secrete ATP [50, 51], and both cell types express P2Y ATP receptors [52, 53]. Because P2Y receptors are GPCRs that link to InsP3‐mediated Cai2+ signaling, secretion of ATP by hepatocytes stimulates Cai2+ signaling in neighboring hepatocytes and bile duct cells [50]. This paracrine signaling mechanism thus permits increases in Cai2+ to spread among neighboring cells independent of communication via gap junctions. Hepatocytes and bile duct cells express P2X receptors as well [54], which are plasma membrane ATP‐gated Ca2+ channels, but the physiological role of these receptors in liver is less clear. Cai2+ signaling in liver also can be modified rather than initiated by paracrine pathways. For example, bradykinin does not mobilize Cai2+ in isolated hepatocytes, yet in the intact liver it modifies the propagation of Cai2+ waves induced by vasopressin [55], likely via nitric oxide (NO) release from endothelial cells, which diffuses to hepatocytes where it stimulates generation of cGMP.

Nuclear Ca2+ signaling The nucleus is separated from the cytosol by the nuclear envelope, which is a specialized region of the ER. Like the ER, the nuclear envelope is able to store and release Ca2+. The nuclear envelope accumulates Ca2+ via a Ca2+‐ATPase pump [56], and releases Ca2+ via channels that are sensitive to InsP3, cADPR, and NAADP [57, 58]. These Ca2+ storage pumps and release channels have distinct distributions within the nuclear envelope. The Ca2+‐ ATPase pump is located only in the outer leaflets of the envelope, while the InsP3R is located only in the inner membrane, and cADPR‐sensitive channels are present on both sides of the nuclear envelope. Both InsP3Rs and RyRs are localized along invaginations of the nuclear envelope, denoted the nucleoplasmic reticulum, which work as a regulatory Ca2+ domain within the nucleus [59]. The importance of nuclear Ca2+ is also highlighted by findings demonstrating that different transcription factors are directly or indirectly dependent on Ca2+ in the nucleus [60].

Evidence suggests that InsP3 releases Ca2+ directly from the nuclear envelope into the nucleus. InsP3 releases 45Ca2+ into isolated hepatocyte nuclei, and InsP3 increases free nuclear Ca2+ even if the nucleus is surrounded by a Ca2+ chelator. In addition, extracellular ATP preferentially activates nuclear Ca2+ release in HepG2 cells, via an InsP3R‐dependent mechanism [61]. The nuclear envelope possesses the machinery necessary to produce InsP3, including PIP2 and PLC [62], and this machinery may be activated selectively through RTK pathways. In one study, IGF‐1 and integrins caused PIP2 breakdown in the nucleus but not at the plasma membrane, while activation of G protein‐linked receptors caused breakdown of PIP2 in the cytosol but not the nucleus. Similarly, activation of the HGF receptor c‐Met in a liver cell line caused PIP2 breakdown in the nucleus resulting in nuclear Ca2+ signals. Moreover, activation of this highly localized cascade was dependent on translocation of the activated receptor to the nucleus [3]. Nuclear Ca2+ signaling also occurs by spread of Cai2+ signals into the nucleus. The nuclear envelope contains pores that are permeable to molecules up to 60 KDa in size. In the absence of a gating mechanism, a pore this size would allow rapid equilibration of Ca2+ between the nucleus and cytosol. Under certain circumstances, free diffusion of Ca2+ through the nuclear pore indeed occurs. However, a nuclear‐cytosolic Ca2+ gradient has been demonstrated in a number of cell types [63], suggesting that the permeability of nuclear pores can be regulated. Moreover, electrophysiological studies suggest that Ca2+ permeability through nuclear pores is restricted. Atomic force microscopy studies similarly suggest that nuclear pore permeability is regulated, and that depletion of Ca2+ within the nuclear envelope closes the pores. Other work using fluorescent dyes or aequorin also demonstrates that depletion of Ca2+ attenuates the permeability of the pores to intermediate‐sized molecules that lack a nuclear localization sequence. Studies monitoring diffusion of photoactivatable GFP across the nuclear envelope suggest that permeability of the nuclear pore may be regulated by cytosolic Ca2+ rather than by Ca2+ stores within the nuclear envelope [64]. EF‐hand Ca2+ binding motifs are present in proteins of the nuclear pore, so it is possible that these function as Ca2+‐gating sensors for the pore. There is additional evidence that nuclear Ca2+ sometimes passively follows Cai2+ [65]. For example, stimulation of hepatocytes with vasopressin results in a Ca2+ wave that appears to spread from the cytosol into the nucleus. Moreover, a mathematical analysis of Ca2+ waves in hepatocytes stimulated with vasopressin suggests that nuclear Ca2+ signals can be described simply by diffusion of Ca2+ inward from the nuclear envelope. Nuclear Ca2+ also can contribute to Ca2+ signals in the cytosol. For example, localized increases of Ca2+ in the cytosol (Ca2+ puffs) can spread across the cell by diffusing across the nucleus. Ca2+ puffs are highly transient and localized Cai2+ signals that result from the coordinated opening of small clusters of InsP3Rs. Puffs can be triggered by a subthreshold concentration of agonist and the resulting Cai2+ signal rapidly dissipates by diffusion in the cytosol and sequestration of Ca2+ into intracellular stores. However, the range of diffusion of Ca2+ in the nucleus can be much greater than in cytosol, so Ca2+ puffs generated near the nuclear envelope can spread into and across the nucleus in order to spread to other, more distant regions of the cytosol [65]. Thus, the nucleus may function as a tunnel that helps distribute Ca2+ to the cytosol.



40: Ca2+ SIGNALING IN THE LIVER

Ca2+ signaling in bile duct cells Cai2+ signaling has been examined to a lesser extent in bile duct epithelia than in hepatocytes. ATP and UTP both increase Cai2+ in the Mz‐ChA‐1 cholangiocarcinoma cell line, a model for bile duct epithelium. ATP and UTP also increase Cai2+ in primary cultures of rat bile duct epithelia, and acetylcholine (ACh) increases Cai2+ in these cells via M3 muscarinic receptors [52]. As in hepatocytes and other epithelia, the range of patterns of agonist‐ induced Cai2+ signals include sustained and transient Cai2+ increases and Cai2+ oscillations. Cai2+ spikes induced by ACh are longer in duration and lower in frequency than those induced by ATP [52]. Cai2+ signaling is mediated by InsP3 in bile duct cells, because increases in Cai2+ are blocked by InsP3R antagonists. As discussed above, all three InsP3R isoforms are expressed in these cells [12]; Ca2+ waves begin in their apical region, where InsP3R3 is concentrated, and then spread basolaterally via the types I and II InsP3Rs. Bile duct cells are coupled via connexin 43 (Cx43) gap junctions, and expression of Cx43 synchronizes their Ca2+ oscillations. Moreover, Cx43 permeability is under hormonal control, and activation of either protein kinase A or C decreases permeability and impairs intercellular communication [66].

EFFECTS OF Ca2+ SIGNALS IN HEALTH AND DISEASE Ca2+ regulates a wide range of functions in liver. The following sections are meant to provide illustrative examples of the different ways in which Ca2+ regulates normal and abnormal liver function, rather than to provide an exhaustive list of Ca2+‐mediated functions.

Energy metabolism Storage and release of glucose was among the first functions of the liver shown to be regulated by Cai2+. Synthesis of glycogen is regulated by glycogen synthase, while phosphorylase is the rate‐limiting enzyme for glycogenolysis. Both enzymes are regulated by phosphorylation and dephosphorylation, and an increase in Cai2+ is one of the most important signals for regulating these events [67]. For example, hormones such as vasopressin and angiotensin increase InsP3 in hepatocytes, which mobilize Ca2+, leading to phosphorylation and activation of glycogen phosphorylase, and then glycogenolysis. Similarly, extracellular nucleotides activate glycogen phosphorylase and thus stimulate glycogenolysis by binding to P2Y nucleotide receptors [68]. Glucagon and beta‐adrenergic agonists also stimulate glycogen phosphorylase activity in liver, but through an alternative, cAMP‐dependent pathway. Ca2+‐mobilizing bile acids such as UDCA, TLCA, and LCA activate phosphorylase to the same extent as hormones such as vasopressin. These bile acids activate phosphorylase through a Ca2+‐dependent but InsP3‐independent mechanism, consistent with the observation that they increase Cai2+ in an InsP3‐independent fashion [69]. Gluconeogenic enzymes are preferentially located in the periportal region [70], although other factors also may be involved in regional differences in glycogenolytic capacity. ATP

503

mobilizes glucose mainly from the periportal zone, while norepinephrine and vasopressin preferentially release glucose from pericentral hepatocytes. This may in part reflect the fact that pericentral hepatocytes are more sensitive than periportal hepatocytes to vasopressin and norepinephrine [37]. When hepatocytes are not uniformly sensitive to a particular hormone, then intercellular communication via gap junctions enhances glucose release. For example, glucose release is impaired in perfused livers from Cx32‐deficient mice upon stimulation with either norepinephrine or glucagon [44]. Similarly, vasopressin‐ or glucagon‐induced glucose release is impaired in perfused rat livers treated with the gap junction blocker 18α‐glycyrrhetinic acid (αGA) [42]. However, glucose release is not altered in αGA‐treated livers stimulated with dibutyryl cAMP or 2,5‐ di(tert‐butyl)‐1,4‐benzohydroquinone (tBuBHQ), both of which stimulate glucose release in a receptor‐independent fashion [42]. Hormone‐induced glucose release also is impaired if gap junctions are blocked in isolated rat hepatocytes, or if hepatocytes are dispersed. Therefore, hepatocytes may contribute differently to glucose metabolism across the hepatic lobule, although there is some integration of metabolic activity via gap junctions. This integration of metabolic function is particularly important in times of stress, because fasting induces hypoglycemia in Cx32‐deficient but not wild‐type mice, and because endotoxin‐induced hypoglycemia is exacerbated in the knockout mice as well [40]. Similarly, liver glucose production upon stimulation with glucagon and norepinephrine is reduced in Cx32 knockout mice [44]. Glucose output by the liver is also regulated by Ca2+‐calmodulin kinase II gamma (CamKIIγ), a downstream effector of intracellular Ca2+ signaling [71]. In a physiological context, this kinase is activated by InsP3R‐ dependent Ca2+ release during fasting and promotes translocation of the transcription factor FoxO1 to the nucleus of hepatocytes. This translocation in turn controls a transcriptional program that potentiates glycogenolysis and gluconeogenesis and leads to increased glucose output by the liver. This mechanism is potentiated in experimental obesity and contributes to the high serum levels of glucose found in obese mice. Accordingly, genetic deletion of CamKIIγ or deletion of FoxO1 in mice are associated with lower blood concentrations of glucose. A second Ca2+‐dependent transcription program that modulates glucose production by the liver is operated by the CREB regulated transcription coactivator 2 (CRTC2) [72]. Here, glucagon activates cAMP formation, PKA activation, and phosphorylation of InsP3R in hepatocytes. This phosphorylation renders the receptor more susceptible to activation by InsP3 and thus Ca2+ is more readily released in the cytosol. Free Ca2+ can then bind to calcineurin which dephosphorylates CRTC2 allowing its translocation to the nucleus. The final result is the activation of genes that promote glucose production. Termination of this mechanism is mainly by insulin‐dependent activation of Akt. In diabetes, insulin resistance thus ensures prolonged activation of the CRTC2 and sustained glucose output by the liver. CREB also increases transcription and thus expression of InsP3R2 [73]. Hepatic InsP3R2 expression also increases during fasting, and it has been postulated that this occurs via glucagon‐ and cAMP‐mediated activation of CREB [73]. Ca2+ signaling also plays a role in lipid metabolism and the pathogenesis of non‐alcoholic fatty liver disease (NAFLD).

504

THE LIVER:  EFFECTS OF Ca2+ SIGNALS IN HEALTH AND DISEASE

Relevant studies mainly describe changes in expression of Ca2+ handling proteins, such as intracellular channels and pumps, and their effect on lipid metabolism. Studies in obese mice showed that expression of Serca2b is specifically downregulated in the liver [74]. The physiological function of this sarco(endo)plasmic reticulum Ca2+‐ATPase is to maintain high levels of Ca2+ within the ER, which are required for proper protein folding and secretion. This is achieved by active transport of free Ca2+ from the cytosol to the ER cisternae. In experimental obesity, the downregulation of Serca2b leads to reduced ER Ca2+ content and ER stress, which collectively activates glucose production and triggers a lipogenic gene expression program. Conversely, re‐expression of Serca2b in obese mice restores ER protein folding and improves both glucose and fatty acid metabolism. InsP3Rs have also been implicated in lipid processing and droplet formation in hepatocytes. Liver specific InsP3R1 knockout (LSKO1) mice are resistant to the development of diet‐ induced liver steatosis and this effect is due to impaired Ca2+ transfer from ER to mitochondria [18]. Similarly, mice fed a high‐ fat diet (HFD) or genetically obese mice display a reorganization of ER‐mitochondria contact sites [75]. These contact regions are known as mitochondrial‐associated membranes (MAMs) and facilitate the exchange of phospholipids and transmission of Ca2+ between the two organelles. In obesity, the overall percentage of MAMs is increased, leading to InsP3R1‐dependent Ca2+ overload in mitochondria and formation of reactive oxygen species (ROS) and mitochondrial dysfunction. Defects in SOCE also occur in experimental fatty liver disease. In obese mice, STIM1 translocation to the vicinity of the plasma membrane is impaired, which results in markedly diminished SOCE. These alterations in SOCE in turn are associated with lipid droplet formation and lipid toxicity in hepatocytes [76]. Overall, pathological lipid accumulation in hepatocytes and alterations in the expression of Ca2+ signaling proteins are interrelated, and there is evidence that this occurs in human disease as well. In particular, there is a progressive increase in the colocalization between ER and mitochondria in hepatocytes from normal liver to simple steatosis to non‐alcoholic steatohepatitis (NASH). There also is increased expression of InsP3R1, which is responsible for mitochondrial Ca2+ signals, in liver biopsies from patients with NASH [18].

Bile flow Cai2+ regulates fluid and electrolyte secretion in many types of epithelia [25]. This is mediated in part by polarized Cai2+ waves, which also occur in hepatocytes [77]. Cai2+ has multiple effects on bile flow. Early studies in isolated perfused rat liver showed that the net effect of Cai2+ on bile acid‐independent bile flow is inhibitory [78], but more recent studies demonstrate that both bile acid‐independent and bile acid‐dependent flow are potentiated by Ca2+ signaling in hepatocytes [33, 34]. First, the membrane localization and activity of the main apical bile‐acid independent solute transporter, the multidrug resistance‐associated protein 2 (Mrp2) is potentiated by Ca2+ agonists in in vitro systems. Moreover, bile flow is decreased in InsP3R2 knockout animals suggesting that Ca2+ release near the apical membrane is important for bile secretion [33]. Total internal reflection fluorescence (TIRF) microscopy studies furthermore suggest that periapical Ca2+ signals induce Mrp2‐containing vesicles to fuse

with the plasma membrane [33]. Bile‐salt dependent flow, which relies on the activity of the bile salt export pump (Bsep), is similarly decreased in rat hepatocytes with reduced expression of InsP3R2 or in cells treated with an intracellular Ca2+ buffering agent [34]. Moreover, disruption of lipid rafts in the apical membrane induces InsP3R2 to migrate away from the periapical region, impairs Ca2+ signals, and reduces secretion [13, 34]. Finally, InsP3R2 expression is decreased and the remaining receptor moves from the apical region in two separate models of cholestasis induced by estrogen and endotoxin [34]. Collectively, these studies suggest that subapical Ca2+ signals in hepatocytes potentiate secretion of bile solutes. Bile acids such as lithocholic acid (LCA), taurolithocholic acid (TLCA), taurodeoxycholic acid and taurochenodeoxycholic, as well as the therapeutic bile acids ursodeoxycholic acid (UDCA) and tauroursodeoxycholic acid (TUDCA) increase intracellular Ca2+ in hepatocytes [69, 79]. LCA and TLCA are cholestatic, but UDCA and TUDCA are choleretic. In fact, TUDCA can reverse the cholestatic effect of TLCA, and this depends on InsP3R2 [33]. UDCA and TUDCA also induce hepatocytes to secrete ATP, which may relate to their therapeutic effect [80].

Cell proliferation Ca2+ signals are associated with progression through the cell cycle [81]. In sea urchin embryos, there are two prominent cell cycle‐related Ca2+ signals. The first Ca2+ transient occurs just prior to entry into mitosis, and the second one occurs during the metaphase–anaphase transition. Intracellular injection of Ca2+ chelators such as BAPTA or an InsP3R antagonist such as heparin abolish these Ca2+ signals and prevent entry into mitosis. Introduction of InsP3 or Ca2+ in the cytosol has the opposite effect, to accelerate entry into mitosis. Ca2+ transients are also observed during cell cycle progression in somatic cells, although the relationship between these Ca2+ signals and progression through the cell cycle is less established. In Swiss 3T3 cells, serum withdrawal suppresses these Ca2+ transients but does not affect progression though mitosis. However, mitosis is blocked by specific inhibition of Ca2+ transients through injection of BAPTA plus incubation in calcium‐ free media. Moreover, photorelease of caged Ca2+ induces premature entry into mitosis. Downstream targets of Ca2+ have also been implicated in cell cycle progression. Calcineurin is essential for Xenopus laevis embryonic development [82]. In addition, pharmacological inhibition of Ca2+/calmodulin kinase II (CaMKII) arrests cells at the G2/M transition. Moreover, calmodulin overexpression accelerates the cell cycle in mouse C127 cells and its downregulation extends the cell cycle. Heterologous expression of the Ca2+ binding protein parvalbumin has also been used to study Ca2+ signaling in the regulation of the cell cycle. This protein is normally expressed in skeletal muscle and neurons; in myocytes it modulates relaxation of fast twitch muscle fibers due to its Ca2+ buffering capacity. The first report using this protein as a molecular tool showed that buffering Ca2+ slowed progression through the cell cycle in mouse C127 cells. More recently, parvalbumin variants targeted to the nucleus or the cytoplasm [60] were used to investigate the relative importance of Ca2+ signals in each of these cellular



40: Ca2+ SIGNALING IN THE LIVER

compartments for regulation of the cell cycle in a liver cell line. It was found that nucleoplasmic rather than cytosolic Ca2+ is essential for cellular proliferation, and is necessary in particular for progression through early prophase [83]. Recent findings suggest that HGF and insulin, two potent growth factors in liver, selectively form InsP3 in the nucleus to initiate nuclear Ca2+ signals [3, 84]. These findings suggest that certain growth factors may stimulate proliferation of hepatocytes by selectively inducing Ca2+ signals in the nucleus. A common model of hepatocyte proliferation in vivo is liver regeneration after partial hepatectomy. Following 70% hepatectomy, the liver undergoes a coordinated regenerative response involving all hepatic cell types, but this is punctuated by marked proliferation of hepatocytes. In rats, 24 hours after partial hepatectomy there is a decrease in InsP3R2 expression accompanied by a decrease in the frequency of Ca2+ oscillations. This decrease in receptor expression normalizes within 4 days and is then associated with an increase in the frequency of Ca2+ oscillations. This remodeling of the Ca2+ signaling machinery is thought to be essential for the initial regenerative response [85]. However, complete loss of InsP3R2 results in delayed liver regeneration and also impaired Ca2+ signaling in the hepatocyte nucleus [86]. Fatty liver causes a c‐Jun mediated decrease in InsP3R2 expression in hepatocytes [86], which may contribute in part to the impaired liver regeneration seen in patients with NAFLD. At least three growth factors, HGF, EGF, and insulin, mobilize Cai2+ in hepatocytes and contribute to liver regeneration [87–89]. Liver regeneration after partial hepatectomy is significantly delayed in rats expressing parvalbumin in the cytosol of hepatocytes [90]. Buffering of InsP3 in the nucleus of hepatocytes also delays liver regeneration [89], suggesting that RTK ‐mediated, InsP3‐dependent Ca2+ signals in the nucleus are essential for proper hepatocyte proliferation. On the other hand, liver regeneration is enhanced by expression of parvalbumin in mitochondria, because this mitigates apoptosis that is occurring concomitant with proliferation [20]. Therefore, the rate of liver regeneration is determined by a careful balance between Ca2+ signals in the cytosol, nucleus, and mitochondria. Ca2+‐dependent processes have been associated with progression of hepatocellular carcinoma (HCC). For instance, expression of Ca2+/calmodulin‐dependent kinase kinase II (CamKKII) is upregulated in liver cancer and its experimental downregulation or pharmacological inhibition reduce tumor growth in a mouse model of HCC [91]. Also, overexpression of hepatitis B virus X protein (HBx), a known oncogenic viral protein, promotes liver cancer cell invasion and metastasis through a Ca2+‐mediated increase in secretion of HMGB1 [92]. The role of intracellular Ca2+ channels and downstream targets of Ca2+ signaling in the pathogenesis of liver cancer is a topic of active investigation.

Ductular secretion Cai2+ directly regulates fluid and electrolyte secretion in bile duct epithelia. Cholangiocytes express the apical, Ca2+‐activated Cl− channel TMEM16a [93], which is the primary mechanism for Ca2+‐stimulated secretion in these cells. The other principal mechanism for regulating cholangiocyte secretion is via the apical, cAMP‐activated, cystic fibrosis transmembrane conductance regulator (CFTR) Cl− channel [51]. Together, these two mechanisms create the Cl− gradient responsible for driving anion

505

exchanger 2 (AE2)‐mediated HCO3−/Cl− exchange that is responsible for secreting bicarbonate into the bile. Ca2+ potentiates adenylyl cyclase activity and cAMP formation in cholangiocytes via a calcineurin‐dependent pathway [94] and indirectly through cAMP formation induced by SOCE [95]. This crosstalk between the Ca2+ and cAMP signaling pathways has been considered a secondary, indirect mechanism for Ca2+‐stimulated secretion. However, cAMP‐ and CFTR‐mediated secretion may rely on Ca2+ in a more direct way as well [96]. Activation of CFTR in cholangiocytes is linked to exocytosis of vesicles rich in ATP [97], which releases ATP into the bile, resulting in stimulation of apical P2Y receptors [51, 96]. Activation of these P2Y receptors links to GPCRs and InsP3 formation and then Ca2+ release from apical, InsP3R3, which in turn drives Ca2+‐dependent Cl− channels and then HCO3− secretion [96]. Conversely, neurotransmitters such as acetylcholine stimulate basolateral M3 muscarinic receptors on cholangiocytes, leading to Ca2+ release from InsP3R1 and InsP3R2. This Ca2+ similarly activates apical Ca2+‐dependent Cl− channels, which ultimately stimulate biliary HCO3− secretion. Peptide hormones such as secretin stimulate basolateral secretin receptors, which induce the formation of cAMP, so this pathway serves to activate CFTR leading to paracrine, ATP‐mediated HCO3− secretion. This putative universal role of InsP3R‐mediated Ca2+ signaling in ductular secretion is in agreement with the observation that loss of InsP3R3 is a common molecular event in cholestatic disorders including primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), extrahepatic biliary obstruction, and biliary atresia [98]. Factors that contribute to this specific loss of InsP3R3 in biliary diseases include the transcription factors Nrf2 (nuclear factor erythroid 2‐related factor 2) and NF‐κB, plus the microRNA miR‐506. Nrf2 is activated during conditions of oxidative stress and specifically downregulates InsP3R3 in a human cholangiocyte cell line. Accordingly, it is translocated to the nucleus of bile duct cells in a range of ductular diseases including PBC and PSC [99]. The NF‐κB pathway is activated by endotoxin‐induced stimulation of TLR4, and this mechanism has been implicated in loss of InsP3R3 from cholangiocytes in patients with sepsis or alcoholic hepatitis [100]. The specific loss of InsP3R3 in PBC might also be mediated in part by miR‐506, because this microRNA decreases InsP3R3 expression in a human cholangiocyte cell line and is increased in bile ducts in liver biopsies from PBC patients [101]. Cholangiocytes express primary cilia on their apical membranes, and this sensory organelle also links to Ca2+ signaling and secretion. For example, primary cilia can sense and transduce mechanical stimuli from within the bile duct lumen, so that increases in bile flow activate both cAMP production and Ca2+ release [102]. Cholangiocyte cilia also sense increases in osmolarity within the bile duct lumen, and this links to activation of TRPV4 channels, which allow entry of extracellular Ca2+ into the cytoplasm [103]. The role of cilia and ciliary Ca2+ signaling in cholestatic disorders remains an area of investigation. The GPCR for bile acids (TGR5) also is expressed at the apical membrane of cholangiocytes and it is enriched in cilia of rodent and human bile duct cells [104]. TGR5 activation couples to the formation of cAMP in cholangiocytes, so pharmacological agonists of this receptor are currently under study as choleretic agents in a variety of acquired and genetic cholestatic disorders.

506

THE LIVER:  REFERENCES

ACKNOWLEDGMENTS This work was supported by NIH grants DK57751, DK34989, DK112797, and DK114041, and by grants from CNPq and FAPEMIG.

REFERENCES 1. Berridge, M.J., Bootman, M.D., and Roderick, H.L. Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol, 2003;4(7):517–29. 2. Rhee, S.G. and Choi, K.D. Regulation of inositol phospholipid‐specific phospholipase C isozymes. J Biol Chem, 1992;267:12393–6. 3. Gomes, D.A., Rodrigues, M.A., Leite, M.F. et al. c‐Met must translocate to the nucleus to initiate calcium signals. J Biol Chem, 2008;283(7):4344–51. 4. Futatsugi, A., Nakamura, T., Yamada, M.K. et  al. IP3 receptor types 2 and  3  mediate exocrine secretion underlying energy metabolism. Science, 2005;309(5744):2232–4. 5. Chandrasekhar, R., Alzayady, K.J., Wagner, L.E., 2nd, and Yule, D.I. Unique regulatory properties of heterotetrameric inositol 1,4,5‐trisphosphate receptors revealed by studying concatenated receptor constructs. J Biol Chem, 2016;291(10):4846–60. 6. Choe, C.U. and Ehrlich, B.E. The inositol 1,4,5‐trisphosphate receptor (IP3R) and its regulators: sometimes good and sometimes bad teamwork. Sci STKE, 2006;2006(363):re15. 7. Bosanac, I., Alattia, J.R., Mal, T.K. et al. Structure of the inositol 1,4,5‐trisphosphate receptor binding core in complex with its ligand. Nature, 2002;420(6916):696–700. 8. Hagar, R.E., Burgstahler, A.D., Nathanson, M.H., and Ehrlich, B.E. Type III InsP3 receptor channel stays open in the presence of increased calcium. Nature, 1998;396(6706):81–4. 9. Bimboese, P., Gibson, C.J., Schmidt, S., Xiang, W., and Ehrlich, B.E. Isoform‐specific regulation of the inositol 1,4,5‐trisphosphate receptor by O‐linked glycosylation. J Biol Chem, 2011;286(18):15688–97. 10. Dufour J‐F., Luthi, M., Forestier, M., and Magnino, F. Expression of inositol 1,4,5‐trisphosphate receptor isoforms in rat cirrhosis. Hepatology, 1999;30: 1018–26. 11. Hirata, K., Pusl, T., O’Neill A.F., Dranoff, J.A., and Nathanson, M.H. The type II inositol 1,4,5‐trisphosphate receptor can trigger Ca2+ waves in rat hepatocytes. Gastroenterology, 2002;122(4):1088–100. 12. Hirata, K., Dufour, J.F., Shibao, K. et al. Regulation of Ca(2+) signaling in rat bile duct epithelia by inositol 1,4,5‐trisphosphate receptor isoforms. Hepatology, 2002;36(2):284–96. 13. Nagata, J., Guerra, M.T., Shugrue, C.A., Gomes, D.A., Nagata, N., and Nathanson, M.H. Lipid rafts establish calcium waves in hepatocytes. Gastroenterology, 2007;133(1):256–67. 14. Pierobon, N., Renard‐Rooney, D.C., Gaspers, L.D., and Thomas, A.P. Ryanodine receptors in liver. J Biol Chem, 2006;281(45):34086–95. 15. Zhang, F. and Li, P.L. Reconstitution and characterization of a nicotinic acid adenine dinucleotide phosphate (NAADP)‐sensitive Ca2+ release channel from liver lysosomes of rats. J Biol Chem, 2007;282(35):25259–69. 16. De Stefani, D., Rizzuto, R., and Pozzan, T. Enjoy the trip: calcium in mitochondria back and forth. Annu Rev Biochem, 2016;85:161–92. 17. Antony, A.N., Paillard, M., Moffat, C. et al. MICU1 regulation of mitochondrial Ca(2+) uptake dictates survival and tissue regeneration. Nat Comm, 2016;7:10955. 18. Feriod, C.N., Oliveira, A.G., Guerra, M.T. et al. Hepatic inositol 1,4,5 trisphosphate receptor type 1 mediates fatty liver. Hepatol Commun, 2017;1(1):23–35. 19. White, C., Li, C., Yang, J. et al. The endoplasmic reticulum gateway to apoptosis by Bcl‐X(L) modulation of the InsP3R. Nat Cell Biol, 2005;7(10):1021–8. 20. Guerra, M.T., Fonseca, E.A., Melo, F.M. et al. Mitochondrial calcium regulates rat liver regeneration through the modulation of apoptosis. Hepatology, 2011;54(1):296–306. 21. Minagawa, N., Kruglov, E.A., Dranoff, J.A., Robert, M.E., Gores, G.J., and Nathanson, M.H. The anti‐apoptotic protein Mcl‐1 inhibits mitochondrial Ca2+ signals. J Biol Chem, 2005;280(39):33637–44.

22. Csordas, G., Renken, C., Varnai, P. et al. Structural and functional features and significance of the physical linkage between ER and mitochondria. J Cell Biol, 2006;174(7):915–21. 23. Hajn¢czky, G., Hager, R., and Thomas, A.P. Mitochondria suppress local feedback activation of inositol 1,4,5‐trisphosphate receptors by Ca2+. J Biol Chem, 1999;274:14157–62. 24. Robb‐Gaspers, L.D., Burnett, P., Rutter, G.A., Denton, R.M., Rizzuto, R., and Thomas, A.P. Integrating cytosolic calcium signals into mitochondrial metabolic responses. EMBO J, 1998;17:4987–5000. 25. Nathanson, M.H. Cellular and subcellular calcium signaling in gastrointestinal epithelium. Gastroenterology, 1994;106:1349–64. 26. Zhao, Y., Araki, S., Wu, J. et al. An expanded palette of genetically encoded Ca(2)(+) indicators. Science, 2011;333(6051):1888–91. 27. Nathanson, M.H. and Burgstahler, A.D. Coordination of hormone‐induced calcium signals in isolated rat hepatocyte couplets: demonstration with confocal microscopy. Mol Biol Cell, 1992;3:113–21. 28. Rooney, T.A., Sass, E.J., and Thomas, A.P. Characterization of cytosolic calcium oscillations induced by phenylephrine and vasopressin in single Fura‐2 loaded hepatocytes. J Biol Chem, 1989;264:17131–41. 29. Dolmetsch, R.E., Xu, K., and Lewis, R.S. Calcium oscillations increase the efficiency and specificity of gene expression. Nature, 1998;392(6679):933–6. 30. Jones, B.F., Boyles, R.R., Hwang, S.Y., Bird, G.S., and Putney, J.W. Calcium influx mechanisms underlying calcium oscillations in rat hepatocytes. Hepatology, 2008;48(4):1273–81. 31. Feske, S., Skolnik, E.Y., and Prakriya, M. Ion channels and transporters in lymphocyte function and immunity. Nat Rev Immunol, 2012;12(7):532–47. 32. Hernandez, E., Leite, M.F., Guerra, M.T. et  al. The spatial distribution of inositol 1,4,5‐trisphosphate receptor isoforms shapes Ca2+ waves. J Biol Chem, 2007;282(13):10057–67. 33. Cruz, L.N., Guerra, M.T., Kruglov, E. et al. Regulation of multidrug resistance‐associated protein 2 by calcium signaling in mouse liver. Hepatology, 2010;52(1):327–37. 34. Kruglov, E.A., Gautam, S., Guerra, M.T., and Nathanson, M.H. Type 2 inositol 1,4,5‐trisphosphate receptor modulates bile salt export pump activity in rat hepatocytes. Hepatology, 2011;54(5):1790–9. 35. Motoyama, K., Karl, I.E., Flye, M.W., Osborne, D.F., and Hotchkiss, R.S. Effect of Ca2+ agonists in the perfused liver: determination via laser scanning confocal microscopy. Am J Physiol Regul Integr Comp Physiol, 1999;276:R575–85. 36. Nathanson, M.H., Burgstahler, A.D., Mennone, A., Fallon, M.B., Gonzalez, C.B., and Saez, J.C. Ca2+ waves are organized among hepatocytes in the intact organ. Am J Physiol, 1995;269(1 Pt 1):G167–71. 37. Tordjmann, T., Berthon, B., Combettes, L., and Claret, M. The location of hepatocytes in the rat liver acinus determines their sensitivity to calcium‐ mobilizing hormones. Gastroenterology, 1996;111:1343–52. 38. Saez, J.C., Connor, J.A., Spray, D.C., and Bennett, M.V.L. Hepatocyte gap junctions are permeable to the second messenger, inositol 1,4,5‐triphosphate, and to calcium ions. Proc Natl Acad Sci USA, 1989;86:2708–12. 39. Fallon, M.B., Nathanson, M.H., Mennone, A., Sáez J.C., Burgstahler, A.D., and Anderson, J.M. Altered expression and function of hepatocyte gap junctions after common bile duct ligation in the rat. Am J Physiol Cell Physiol, 1995;268:C1186–94. 40. Correa, P.R., Guerra, M.T., Leite, M.F., Spray, D.C., and Nathanson, M.H. Endotoxin unmasks the role of gap junctions in the liver. Biochem Biophys Res Commun, 2004;322(3):718–26. 41. Leite, M.F., Hirata, K., Pusl, T. et al. Molecular basis for pacemaker cells in epithelia. J Biol Chem, 2002;277:16313–23. 42. Nathanson, M.H., Rios‐Velez, L., Burgstahler, A.D., and Mennone, A. Communication via gap junctions modulates bile secretion in the isolated perfused rat liver. Gastroenterology, 1999;116:1176–83. 43. Bartlett, P.J., Gaspers, L.D., Pierobon, N., and Thomas, A.P. Calcium‐ dependent regulation of glucose homeostasis in the liver. Cell Calcium, 2014;55(6):306–16. 44. Stumpel, F., Ott, T., Willecke, K., and Jungermann, K. Connexin 32 gap junctions enhance stimulation of glucose output by glucagon and noradrenaline in mouse liver. Hepatology, 1998;28:1616–20. 45. Patel, S.J., Milwid, J.M., King, K.R. et al. Gap junction inhibition prevents drug‐induced liver toxicity and fulminant hepatic failure. Nat Biotechnol, 2012;30(2):179–83. 46. Boitano, S., Dirksen, E.R., and Sanderson, M.J. Intercellular propagation of calcium waves mediated by inositol trisphosphate. Science, 1992;258:292–5.



40: Ca2+ SIGNALING IN THE LIVER

47. Tordjmann, T., Berthon, B., Claret, M., and Combettes, L. Coordinated ­intercellular calcium waves induced by noradrenaline in rat hepatocytes: dual control by gap junction permeability and agonist. EMBO J, 1997;16: 5398–407. 48. Tordjmann, T., Berthon, B., Jacquemin, E. et al. Receptor‐oriented intercellular calcium waves evoked by vasopressin in rat hepatocytes. EMBO J, 1998;17:4695–703. 49. Burgstahler, A.D. and Nathanson, M.H. Coordination of calcium waves among hepatocytes: teamwork gets the job done. Hepatology, 1998;27:634–5. 50. Schlosser, S.F., Burgstahler, A.D., and Nathanson, M.H. Isolated rat hepatocytes can signal to other hepatocytes and bile duct cells by release of nucleotides. Proc Natl Acad Sci USA, 1996;93:9948–53. 51. Fiorotto, R., Spirli, C., Fabris, L., Cadamuro, M., Okolicsanyi, L., and Strazzabosco, M. Ursodeoxycholic acid stimulates cholangiocyte fluid secretion in mice via CFTR‐dependent ATP secretion. Gastroenterology, 2007;133(5):1603–13. 52. Nathanson, M.H., Burgstahler, A.D., Mennone, A., and Boyer, J.L. Characterization of cytosolic Ca2+ signaling in rat bile duct epithelia. Am J Physiol Gastrointest Liver Physiol, 1996;271:G86–96. 53. Kitamura, T., Brauneis, U., Gatmaitan, Z., and Arias, I.M. Extracellular ATP, intracellular calcium and canalicular contraction in rat hepatocyte doublets. Hepatology, 1991;14:640–7. 54. Gonzales, E., Prigent, S., Abou‐Lovergne, A. et al. Rat hepatocytes express functional P2X receptors. FEBS Lett, 2007;581(17):3260–6. 55. Patel, S., Robb‐Gaspers, L.D., Stellato, K.A., Shon, M., and Thomas, A.P. Coordination of calcium signalling by endothelial‐derived nitric oxide in the intact liver. Nat Cell Biol, 1999;1:467–71. 56. Lanini, L., Bachs, O., and Carafoli, E. The calcium pump of the liver nuclear membrane is identical to that of endoplasmic reticulum. J Biol Chem, 1992;267(16):11548–52. 57. Gerasimenko, J.V., Maruyama, Y., Yano, K. et al. NAADP mobilizes Ca2+ from a thapsigargin‐sensitive store in the nuclear envelope by activating ryanodine receptors. J Cell Biol, 2003;163(2):271–82. 58. Gerasimenko, O.V., Gerasimenko, J.V., Tepikin, A.V., and Petersen, O.H. ATP‐dependent accumulation and inositol trisphosphate‐ or cyclic ADP‐ ribose‐mediated release of Ca2+ from the nuclear envelope. Cell, 1995;80:439–44. 59. Malhas, A., Goulbourne, C., and Vaux, D.J. The nucleoplasmic reticulum: form and function. Trends Cell Biol, 2011;21(6):362–73. 60. Pusl, T., Wu, J.J., Zimmerman, T.L. et al. Epidermal growth factor‐mediated activation of the ETS domain transcription factor Elk‐1 requires nuclear calcium. J Biol Chem, 2002;277(30):27517–27. 61. Leite, M.F., Thrower, E.C., Echevarria, W. et al. Nuclear and cytosolic calcium are regulated independently. Proc Natl Acad Sci USA, 2003;100(5):2975–80. 62. Divecha, N., Rhee S‐G., Letcher, A.J., and Irvine, R.F. Phosphoinositide signalling enzymes in rat liver nuclei: phosphoinositidase C isoform ·1 is specifically, but not predominantly, located in the nucleus. Biochem J, 1993;289:617–20. 63. Waybill, M.M., Yelamarty, R.V., Zhang, Y. et al. Nuclear calcium gradients in cultured rat hepatocytes. Am J Physiol Endocrinol Metab, 1990;261:E49–57. 64. O’Brien E.M., Gomes, D.A., Sehgal, S., and Nathanson, M.H. Hormonal regulation of nuclear permeability. J Biol Chem, 2007;282(6):4210–7. 65. Lipp, P., Thomas, D., Berridge, M.J., and Bootman, M.D. Nuclear calcium signalling by individual cytoplasmic calcium puffs. EMBO J, 1997;16:7166–73. 66. Bode, H.P., Wang, L., Cassio, D. et  al. Expression and regulation of gap junctions in rat cholangiocytes. Hepatology, 2002;36(3):631–40. 67. Blackmore, P.F., Strickland, W.G., Bocckino, S.B., and Exton, J.H. Mechanism of hepatic glycogen synthase in activation induced by Ca2+‐ mobilizing hormones. Biochem J, 1986;237:235–42. 68. Keppens, S. and DeWulf, H. Characterization of the liver P2‐purinoceptor involved in the activation of glycogen phosphorylase. Biochem J, 1986;240:367–71. 69. Combettes, L., Dumont, M., Berthon, B., Erlinger, S., and Claret, M. Release of calcium from the endoplasmic reticulum by bile acids in rat liver cells. J Biol Chem, 1988;263:2299–303. 70. Halpern, K.B., Shenhav, R., Matcovitch‐Natan, O. et al. Single‐cell spatial reconstruction reveals global division of labour in the mammalian liver. Nature, 2017;542(7641):352–6.

507

71. Ozcan, L., Wong, C.C., Li, G. et  al. Calcium signaling through CaMKII regulates hepatic glucose production in fasting and obesity. Cell Metabol, 2012;15(5):739–51. 72. Wang, Y., Li, G., Goode, J. et al. Inositol‐1,4,5‐trisphosphate receptor regulates hepatic gluconeogenesis in fasting and diabetes. Nature, 2012;485(7396): 128–32. 73. Kruglov, E., Ananthanarayanan, M., Sousa, P., Weerachayaphorn, J., Guerra, M.T., and Nathanson, M.H. Type 2 inositol trisphosphate receptor gene expression in hepatocytes is regulated by cyclic AMP. Biochem Biophys Res Commun, 2017;486(3):659–64. 74. Park, S.W., Zhou, Y., Lee, J., Lee, J., and Ozcan, U. Sarco(endo)plasmic reticulum Ca2+‐ATPase 2b is a major regulator of endoplasmic reticulum stress and glucose homeostasis in obesity. Proc Natl Acad Sci USA, 2010;107(45):19320–5. 75. Arruda, A.P., Pers, B.M., Parlakgul, G., Guney, E., Inouye, K., and Hotamisligil, G.S. Chronic enrichment of hepatic endoplasmic reticulum‐ mitochondria contact leads to mitochondrial dysfunction in obesity. Nat Med, 2014;20(12):1427–35. 76. Arruda, A.P., Pers, B.M., Parlakgul, G. et al. Defective STIM‐mediated store operated Ca(2+) entry in hepatocytes leads to metabolic dysfunction in ­obesity. Elife, 2017;6. 77. Nathanson, M.H., Burgstahler, A.D., and Fallon, M.B. Multi‐step mechanism of polarized Ca2+ wave patterns in hepatocytes. Am J Physiol Gastrointest Liver Physiol, 1994;267:G338–49. 78. Nathanson, M.H., Gautam, A., Bruck, R., Isales, C.M., and Boyer, J.L. Effects of Ca2+ agonists on cytosolic Ca2+ in isolated hepatocytes and on  bile secretion in the isolated perfused rat liver. Hepatology, 1992;15:107–16. 79. Beuers, U., Nathanson, M.H., and Boyer, J.L. Effects of tauroursodeoxycholic acid on cytosolic Ca2+ signals in isolated rat hepatocytes. Gastroenterology, 1993;104:604–12. 80. Nathanson, M.H., Burgstahler, A.D., Masyuk, A., and Larusso, N.F. Stimulation of ATP secretion in the liver by therapeutic bile acids. Biochem, J., 2001;358(1):1–5. 81. Poenie, M., Alderton, J., Steinhardt, R., and Tsien, R. Calcium rises abruptly and briefly throughout the cell at the onset of anaphase. Science, 1986;233(4766): 886–9. 82. Nishiyama, T., Yoshizaki, N., Kishimoto, T., and Ohsumi, K. Transient activation of calcineurin is essential to initiate embryonic development in Xenopus laevis. Nature, 2007;449(7160):341–5. 83. Rodrigues, M.A., Gomes, D.A., Leite, M.F. et al. Nucleoplasmic calcium is required for cell proliferation. J Biol Chem, 2007;282(23):17061–8. 84. Rodrigues, M.A., Gomes, D.A., Andrade, V.A., Leite, M.F., and Nathanson, M.H. Insulin induces calcium signals in the nucleus of rat hepatocytes. Hepatology, 2008;48(5)1621–31. 85. Nicou, A., Serriere, V., Hilly, M. et al. Remodelling of calcium signalling during liver regeneration in the rat. J Hepatol, 2007;46(2):247–56. 86. Khamphaya, T., Chukijrungroat, N., Saengsirisuwan, V. et al. Nonalcoholic fatty liver disease impairs expression of the type II inositol 1,4,5‐trisphosphate receptor. Hepatology, 2018;67(2):560–74. 87. Paranjpe, S., Bowen, W.C., Bell, A.W., Nejak‐Bowen, K., Luo, J.H., and Michalopoulos, G.K. Cell cycle effects resulting from inhibition of hepatocyte growth factor and its receptor c‐Met in regenerating rat livers by RNA interference. Hepatology, 2007;45(6):1471–7. 88. Rodrigues, M.A., Gomes, D.A., Andrade, V.A., Leite, M.F., and Nathanson, M.H. Insulin induces calcium signals in the nucleus of rat hepatocytes. Hepatology, 2008. 89. Amaya, M.J., Oliveira, A.G., Guimaraes, E.S. et  al. The insulin receptor translocates to the nucleus to regulate cell proliferation in liver. Hepatology, 2014;59(1):274–83. 90. Lagoudakis, L., Garcin, I., Julien, B. et al. Cytosolic calcium regulates liver regeneration in the rat. Hepatology, 2010;52(2):602–11. 91. Lin, F., Marcelo, K.L., Rajapakshe, K. et al. The camKK2/camKIV relay is  an essential regulator of hepatic cancer. Hepatology, 2015;62(2): 505–20. 92. Chen, S., Dong, Z., Yang, P. et al. Hepatitis B virus X protein stimulates high mobility group box 1 secretion and enhances hepatocellular carcinoma metastasis. Cancer Lett, 2017;394:22–32. 93. Li, Q., Dutta, A., Kresge, C., Bugde, A., and Feranchak, A.P. Bile acids stimulate cholangiocyte fluid secretion by activation of transmembrane member 16A Cl(‐) channels. Hepatology, 2018;68(1):187–99.

508

THE LIVER:  REFERENCES

94. Alvaro, D., Alpini, G., Jezequel, A.M. et al. Role and mechanisms of action of acetylcholine in the regulation of rat cholangiocyte secretory function. J Clin Invest, 1997;100:1349–62. 95. Spirli, C., Locatelli, L., Fiorotto, R. et  al. Altered store operated calcium entry increases cyclic 3’,5’‐adenosine monophosphate production and extracellular signal‐regulated kinases 1 and 2 phosphorylation in polycystin‐2‐ defective cholangiocytes. Hepatology, 2012;55(3):856–68. 96. Minagawa, N., Nagata, J., Shibao, K. et al. Cyclic AMP regulates b­ icarbonate secretion in cholangiocytes through release of ATP into bile. Gastroenterology, 2007;133(5):1592–602. 97. Sathe, M.N., Woo, K., Kresge, C. et al. Regulation of purinergic signaling in biliary epithelial cells by exocytosis of SLC17A9‐dependent ATP‐enriched vesicles. J Biol Chem, 2011;286(28):25363–76. 98. Shibao, K., Hirata, K., Robert, M.E., and Nathanson, M.H. Loss of inositol 1,4,5‐trisphosphate receptors from bile duct epithelia is a common event in cholestasis. Gastroenterology, 2003;125(4):1175–87. 99. Weerachayaphorn, J., Amaya, M.J., Spirli, C. et  al. Nuclear factor, erythroid  2‐like 2 regulates expression of type 3 inositol 1,4,5‐trisphosphate receptor and calcium signaling in cholangiocytes. Gastroenterology, 2015;149(1):211–22.

100. Franca, A., Filho, A., Guerra, M.T. et  al. Effects of endotoxin on type 3 inositol 1,4,5‐trisphosphate receptor in human cholangiocytes. Hepatology, 2019;69(2):817–830. 101. Ananthanarayanan, M., Banales, J.M., Guerra, M.T. et  al. Post‐translational regulation of the type III inositol 1,4,5‐trisphosphate receptor by miRNA‐506. J Biol Chem, 2015;290(1):184–96. 102. Masyuk, A.I., Masyuk, T.V., Splinter, P.L., Huang, B.Q., Stroope, A.J., and  LaRusso, N.F. Cholangiocyte cilia detect changes in luminal fluid flow  and transmit them into intracellular Ca2+ and cAMP signaling. Gastroenterology, 2006;131(3):911–20. 103. Gradilone, S.A., Masyuk, A.I., Splinter, P.L. et  al. Cholangiocyte cilia express TRPV4 and detect changes in luminal tonicity inducing bicarbonate secretion. Proc Natl Acad Sci USA, 2007;104(48):19138–43. 104. Keitel, V., Ullmer, C., and Haussinger, D. The membrane‐bound bile acid receptor TGR5 (Gpbar‐1) is localized in the primary cilium of cholangiocytes. Biol Chem, 2010;391(7):785–9. 105. Pusl, T. and Nathanson, M.H. The role of inositol 1,4,5‐trisphosphate receptors in the regulation of bile secretion in health and disease. Biochem Biophys Res Commun, 2004;322(4):1318–25. 106. Gaspers, L.D. and Thomas, A.P. Calcium signaling in liver. Cell Calcium, 2005;38(3–4):329–42.

41

Clinical Genomics of NAFLD Frank Lammert Department of Medicine II, Saarland University Medical Center, Homburg, Germany

INTRODUCTION To date fatty liver is a common liver disease worldwide and is to be included in the differential diagnosis of elevated liver enzymes, in particular in the setting of indicators of the metabolic syndrome, such as central obesity, dyslipidemia, hypertension, or increased fasting plasma glucose. The prevalence of non‐alcoholic fatty liver disease (NAFLD) depends on the definition, since liver fat content represents a continuous parameter. In liver biopsies, steatosis (NAFL) is diagnosed when hepatic steatosis exceeds 5%, and non‐alcoholic steatohepatitis (NASH) is defined as the presence of at least 5% hepatic steatosis plus inflammation with hepatocyte injury [1]. In the Dallas heart study (a multi‐ethnic population‐based probability sample of Dallas County residents), the upper 95th percentile of liver fat measured by proton magnetic spectroscopy (1H‐MRS) in healthy subjects was 5.6%, corresponding to 15% histological liver fat, and according to this definition, 31% of the cohort ­displayed hepatic steatosis [2]. When measured by magnetic resonance imaging‐based techniques such as the proton density fat fraction (PDFF), 5% steatosis corresponds to a PDFF of 6.0 to 6.4% [3]. The presence of cirrhosis with current or previous histopathological evidence of steatosis or NASH defines NASH‐cirrhosis. These definitions of fatty liver diseases imply that there is a gradient from common benign NAFL to defined subgroups of patients with more advanced disease such as NASH or cirrhosis. Given the lack of reliable non‐invasive markers, a liver biopsy is still formally required to detect or exclude the presence of NASH [4]. Modeling the epidemic of NAFLD until 2030 in a Markov model based on historical and projected changes in adult prevalence of obesity and type 2 diabetes mellitus demonstrates an exponential increase in burden of NAFLD [5]. In the United

States, prevalent NAFLD cases are forecasted to increase 21% from 83 million in 2015 to 101 million 2030, while prevalent NASH cases will increase 63% from 17 million to 27 million cases. The overall NAFLD prevalence among the population aged greater than or equal to 15 years is projected at 34% in 2030. In 2015, approximately 20% of NAFLD cases were classified as NASH, increasing to 27% by 2030. Incidence of decompensated cirrhosis increases 170% to 105 000 cases by 2030, while incidence of hepatocellular cancer (HCC) will increase by 140% to 12 000 cases. Liver deaths will increase 180% to an estimated 78 000 deaths in 2030, and during 2015– 2030, there are projected to be nearly 800 000 excess liver deaths [5]. Family studies show that first degree relatives of patients with NAFLD display a markedly higher risk to develop NAFLD than the general population, independent of obesity. Recently, ­magnetic resonance imaging was used to quantify hepatic steatosis (PDFF) and liver fibrosis in 60 pairs of twins residing in Southern California [6]. The presence of hepatic steatosis and the stage of liver fibrosis correlated between monozygotic but not dizygotic twins. In multivariate models adjusted for age, sex, and ethnicity, the heritability of hepatic steatosis (based on PDFF) and liver fibrosis was 50%.

PNPLA3 (ADIPONUTRIN) VARIANT AND FATTY LIVER DISEASE Adiponutrin, the enzyme encoded by the PNPLA3 gene (aka calcium‐independent phospholipase A2ε), is a 481‐amino acid member of the patatin‐like phospholipase domain‐containing family (PNPLA). This domain was originally discovered in

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

510

THE LIVER:  PNPLA3 (ADIPONUTRIN) VARIANT AND FATTY LIVER DISEASE

lipid hydrolases of potato and named after the most abundant protein of the potato tuber, patatin. However, since several family members are not phospholipases, a more appropriate gene designation has been called for [7]. PNPLA3 is expressed ­predominantly in the liver, retina, skin, and adipose tissue [8]. In a series of seminal genetic studies, the common non‐synonymous variant p.I148M (c.444C>G, rs738409) of the PNPLA3 gene has emerged as the key genetic determinant of fatty liver disease in pediatric and adult patients [9] (Table  41.1). Figure  41.1 summarizes the geographical distribution of PNPLA3 p.I148M genotypes in patients with NAFLD. Of note, the risk allele frequency is 25% in Europeans [10]. The ­prevalence differs among ethnic groups, and these differences generally parallel those for NASH and its sequelae. The variant is most prevalent in Hispanics (49%), is less common in non‐ Hispanic Europeans (23%), and is least common in African Americans (17%) [2, 11]. The first genome‐wide association study (GWAS) in the ­population‐based Dallas heart study comprising 2111 individuals with different ethnic backgrounds demonstrated the p.I148M variant to be associated (P = 5.9 × 10−10) with increased liver fat content, as determined by 1H‐magnetic resonance spectrometry [2]. There was a stepwise increase of hepatic triglyceride ­content with the number of p.I148M risk alleles. The association was most prominent in Hispanic patients, who are in general at a greater risk of developing fatty liver as compared to European and African Americans. Of note, the PNPLA3 variant p.I148M is also associated with serum activities of liver enzymes. The analyses of two large population‐based cohorts demonstrated that PNPLA3 polymorphisms are associated with serum alanine aminotransferase (ALT) and γ‐glutamyl transpeptidase activities in healthy individuals [12, 13]. The genetic association between the PNPLA3 mutation p.I148M and fatty liver disease was subsequently ­replicated in many studies [14–17]. Further studies showed that the PNPLA3 variant not only increases the odds of developing fatty liver itself but it also determines the degree of hepatic injury and the full spectrum of histopathological consequences of NAFLD [18]. Valenti et al. [19, 20] reported that the PNPLA3 variant was not only associated with hepatic steatosis but with NAFLD severity as determined by liver biopsy, in particular the presence of NASH, steatosis grade greater than S1, and fibrosis stage greater than F1, independent of age, body mass index (BMI), or diabetes. Another detailed analysis of histopathological severity of NAFLD was performed by Rotman et al. [21], demonstrating

that the PNPLA3 variant is associated with steatosis, portal and lobular inflammation, NAFLD activity score (NAS), and fibrosis. The first meta‐analysis by Sookoian and Pirola [9] included data from 16 studies and concluded that the PNPLA3 p.I148M variant is associated with fatty liver (odds ratio [OR] for homozygous carriers 3.3 and heterozygous carriers 1.9), NASH (OR 3.3 and 2.7, respectively), necroinflammation (OR 3.2 and 2.6, respectively), and fibrosis (OR 3.3 and 2.1, respectively); the association with necroinflammation and fibrosis was independent of the severity of steatosis. The recent meta‐analysis by Xu et al. [22] confirmed these results, with subgroup and sensitivity analyses showing that the results were neither influenced by ethnicities nor the age of subjects or the source of controls. Using two histopathologically characterized cohorts encompassing steatosis, steatohepatitis, fibrosis, and cirrhosis (N = 1074), Liu et al. [23] confirmed that the PNPLA3 variant is associated with advanced fibrosis/cirrhosis, which is independent of age, BMI, and type 2 diabetes as confounders (Table 41.2). Availing of transient elastography to quantify liver fibrosis in 899 patients with chronic liver diseases, an association between the PNPLA3 mutation and enhanced liver stiffness was identified in a wide spectrum of viral and nonviral chronic liver ­diseases [24]. Sensitivity analysis showed that the association was present across a broad range of stiffness values (12–40 kPa) [24], indicating that the variant affects not only fibrogenesis but also cirrhosis severity. Non‐obese NAFLD is defined as NAFLD that develops in patients with a BMI less than 25 kg m−2 [25]. In patients with non‐obese NAFLD, the PNPLA3 p.I148M allele is more ­frequent than in other NAFLD patients [11, 26–28] and independently associated with both NASH and fibrosis stage greater than or equal to F2 [29]. Carriers of the PNPLA3 mutation who are obese [30] or present with NASH [31] are predisposed to HCC development. In multivariate analysis adjusted for age, sex, BMI, diabetes, and cirrhosis, carriage of each copy of the PNPLA3 risk allele conferred an additive risk for HCC (OR 2.3), with homozygotes exhibiting a fivefold increased risk over the wild‐type genotype. When compared to the UK general population, the risk‐effect was more pronounced (OR 12.2) [31]. Further analysis of the data found the positive predictive value of p.I148M to be weak, but the negative predictive value of the absence of p.I148M is strong (97%). Hence, prospective studies are warranted to ­validate the clinical utility of PNPLA3 genotyping to select the patients who are least likely to develop HCC and therefore least likely to benefit from surveillance [32]. In a series of newly

Table 41.1  Key examples of liver diseases associated with variant PNPLA3 Disease

Study

Year

Reference

Non‐alcoholic fatty liver disease Alcoholic liver cirrhosis

Romeo et al. Tian et al. Stickel et al. Krawczyk et al. Vigano et al. Cai et al. Müller et al. Valenti and Fargion Nischalke et al.

2008 2010 2011 2011 2013 2011 2011 2011 2011

[2] [98] [99] [24] [100] [101] [102] [103] [104]

Liver fibrosis HBV steatosis HCV steatosis Alcohol and HCV cirrhosis HCV cirrhosis Hepatocellular cancer

HBV, hepatitis B virus; HCV, hepatitis C virus; PNPLA3, patatin‐like phospholipase domain‐containing protein 3.



41:  Clinical Genomics of NAFLD

511

Figure 41.1  Worldwide estimated prevalence of NAFLD and distribution of PNPLA3 genotypes. PNPLA3 genotype is presented as minor risk allele frequency (light blue section of the pie chart). Reproduced with permission of Springer Nature from [105].

Table 41.2  Multivariate analysis of association between PNPLA3 and TM6SF2 genotypes and fibrosis stages F0–1 (mild) versus F2–4 (advanced) Discovery cohort (n = 349)

Validation cohort (n = 725)

Combined cohort (n = 1047)

Variables

OR (95% CI)

P‐value

OR (95% CI)

P‐value

OR (95% CI)

P‐value

TM6SF2 genotype PNPLA3 genotype Age Gender (female) BMI T2DM

2.94 (1.76–4.89) 1.57 (1.21–2.19) 1.03 (1.01–1.06) 0.94 (0.57–1.56) 1.05 (1.00–1.10) 2.39 (1.49–3.84)

3.44 × 10−5 0.0086 0.0045 0.8297 0.0368 0.0003

1.46 (1.03–2.09) 1.32 (1.05–1.66) 1.02 (1.01–1.04) 1.81 (1.30–2.50) 1.03 (1.01–1.05) 2.73 (1.93–3.88)

0.0362 0.0183 0.0041 4.50 × 10−4 9.80 × 10−4 1.68 × 10−8

1.88 (1.41–2.5) 1.40 (1.16–1.69) 1.03 (1.01–1.04) 1.43 (1.09–1.89) 1.04 (1.02–1.05) 2.57 (1.95–3.39)

1.63 × 10−5 4.84 × 10−4 1.57 × 10−5 0.0096 3.78 × 10−5 1.78 × ×10−11

BMI, body mass index; CI, confidence interval; OR, odds ratio; T2DM, type 2 diabetes mellitus. Additive model including age, gender, BMI, T2DM, and PNPLA3 rs738409 genotype as covariates. Discovery/validation/combined cohorts: stage F0–1 (mild) n = 198/439/637, stage F2–4 (advanced) n = 151/286/437.

diagnosed HCC cases  [33], homozygosity for the genotype p.I148M was even an independent risk factor for death; HCC patients with this PNPLA3 genotype displayed reduced median survival (16.8 months) in comparison to carriers of the wild‐ type allele (25.9 months). Overall the above studies document that the PNPLA3 mutation increases the risk of developing severe hepatic fat accumulation, progressive inflammation, advanced fibrosis, and HCC across different ethnicities worldwide (Figure  41.2). Quantitative analyses showed that adiposity was shown to amplify the genetic risk conferred by PNPLA3 across the full spectrum of NAFLD, ranging from increased hepatic triglyceride content to cirrhosis (Figure 41.3) [34]. Additional studies are needed to determine the cost effectiveness and utility of multifactorial risk stratification that incorporates PNPLA3 genotype and other risk factors for HCC.

Variant PNPLA3 and metabolic traits In general, patients with fatty liver disease often present with dyslipidemia and insulin resistance [35]. However, the association of PNPLA3 with metabolic traits in humans is more complex. Several studies did not detect a relationship between the p.I148M polymorphism and serum glucose or lipid concentrations [17, 19, 36, 37]. In contrast, a Danish analysis of more than 4000 individuals with normal glucose tolerance demonstrated an association of the risk allele with increased fasting glucose levels, but the same allele was associated with lower serum concentrations of triglycerides and cholesterol in patients with impaired glucose intolerance [38]. Lower fasting triglyceride levels were also observed in severely overweight patients carrying the risk variant [39]. In line with these results,

512

THE LIVER:  PNPLA3 (ADIPONUTRIN) VARIANT AND FATTY LIVER DISEASE

Figure 41.2  Model illustrating the regulation of PNPLA3 gene expression in liver and the link between the p.I148M variant and progressive liver disease. The PNPLA3 variant drives the full phenotypic spectrum from steatosis to steatohepatitis, fibrosis, cirrhosis, and HCC. Note that ChREBP controls Pnpla3 mRNA levels in mouse hepatocytes, whereas the specific ChREBP response element is missing in the human PNPLA3 promoter. Abbreviations: ChREBP, carbohydrate response element binding protein; HCC, hepatocellular cancer; SREBP1c, sterol regulatory element binding protein 1c. Reproduced with permission of Elsevier from [106].

25

HTGC (%)

20

(b) M variant II IM MM

10

OR (95%CI)

(a)

15 10

M variant II IM MM

5

2 1

5

35

BMI (kg/m2) 22,159 13,079 1,851 67 69 20

7,034 4,156 608 24 34 7

2,026 1,232 173 10 8 4

Figure 41.3  (a) Hepatic triglyceride content (HTGC) measured by magnetic resonance spectroscopy, stratified PNPLA3 genotype, and BMI in the Dallas heart study. The triglyceride increasing effect of the variant is amplified by increasing obesity (Pinteraction = 4 × 10−5). Data are presented as median ± interquartile range. (b) Risk of cirrhosis by PNPLA3 genotype and BMI in the Copenhagen cohort. The risk‐increasing effect of the variant is amplified by increasing obesity (Pinteraction = 0.026). Data are OR ± 95% confidence interval (CI). Reproduced with permission of Springer Nature from [34].



41:  Clinical Genomics of NAFLD

Krawczyk et  al. [10] and other groups [40, 41] identified a possible association between higher fasting glucose levels and the PNPLA3 p.I148M mutation. These results contradict the general paradigm that insulin resistance represents the main driver of common NAFLD. Indeed, a dissociation between the presence of fatty liver and insulin resistance appears to be present among carriers of the PNPLA3 risk variant [36]. Specific circulating triacylglycerol signatures were observed in carriers of the variant, which resembled those of obese subjects with NAFLD [42]. PNPLA3‐associated NAFLD is associated with a relative deficiency of triacylglycerols, supporting the idea that the variant impedes intrahepatocellular lipolysis rather than stimulating lipid synthesis. NAFLD in obese patients is associated with multiple changes in triacylglycerols, which can be attributed to obesity and insulin resistance but not increased liver fat content per se.

Functional analyses of the PNPLA3 variant The close similarity to adipose triglyceride lipase and the presence of typical structural motifs (α–β–α sandwich structure, consensus serine lipase GXSXG motif, catalytic dyad S‐D) indicate a lipase function of PNPLA3 [7]. Alternatively, lysophosphatidic acyltransferase activity has been proposed

513

[43]. Furthermore, Pirazzi et  al. [44] demonstrated that PNPLA3 has retinyl‐palmitate lipase activity in hepatic stellate cells. The expression of human PNPLA3 is induced by carbohydrates and fatty acids via the sterol response element binding protein 1c (SREBP1c, Figure  41.3) as transcription factor [43, 45–47]. Livers from NAFLD patients are characterized by the increased number and size of lipid droplets within hepatocytes. Of note, PNPLA3 is predominantly localized in the endoplasmic reticulum (ER) and on the surface of lipid droplets [48] (Figure  41.4). The rs738409 variant of PNPLA3 results in an isoleucine to methionine substitution at amino acid residue 148. Structural modeling indicates that this substitution occludes access of substrates to the catalytic dyad, resulting in loss of function [49, 50]. In carriers of the p.I148M mutation, the variant PNPLA3 evades ubiquitylation and proteosomal degradation but accumulates on lipid droplets, which increase in number and median size and display impaired triglyceride mobilization [49, 51–53]. Studies performed in transgenic mice overexpressing variant PNPLA3148M but not wild‐type PNPLA3 in liver demonstrated that these animals develop steatosis due to triacylglycerol accumulation as well as several alterations of hepatic lipid metabolism, including increased synthesis of fatty acids and

Figure 41.4  (a) PNPLA3 is induced with increased free fatty acid availability during hyperinsulinemia and expressed in the ER and on the surface of lipid droplets. It is rapidly ubiquitylated and degraded upon fasting. (b) PNPLA3 p.I148M on lipid droplets escapes degradation and its accumulation favors triglyceride accumulation and steatohepatitis. (c) Reduced expression of PNPLA3 p.I148M due to the copresence of the p.E434K variant reduces liver damage. Wild‐type PNPLA3 (p.148I) with intact protective enzymatic activity is illustrated as green curves, and variant PNPLA3 p.148M causing lipid droplet formation as red curves. The number is proportional to protein levels. Red arrows indicate damaging pathways, green arrows protective pathways, and dashed lines denote suppressed pathways. Abbreviations: TAG, triglyceride; Ub, ubiquitin. Reproduced with permission of John Wiley and Sons from [48].

514

THE LIVER:  TM6SF2 AS SECOND NAFLD RISK GENE

triacylglycerol, impaired hydrolysis of triacylglycerol, and depletion of triacylglycerol long‐chain polyunsaturated fatty acids [54]. To determine whether the mutant p.148M allele causes fat accumulation in the liver when expressed at physiological levels, Pnpla3148M knockin gene were generated, which displayed normal levels of hepatic fat on chow diet, but when challenged with a high‐sucrose diet the liver fat content increased two‐ to threefold compared to wild‐type controls without changes in glucose homeostasis. The increased liver fat in the knockin mice was accompanied by a fortyfold increase in PNPLA3 on hepatic lipid droplets with no increase in hepatic Pnpla3 mRNA [55].

Genotype‐guided treatment of NAFLD The in vitro and in vivo studies indicate that suppression of mutant PNPLA3 would have beneficial effects in NAFLD and represent a novel therapeutic target for NAFLD. Until then, lifestyle modification represents the cornerstone of NAFLD treatment. Although to date large prospective studies concerning long‐term effects of the common PNPLA3 p.I148M variant on liver phenotypes are lacking, the first pilot studies indicate that weight loss has positive effects in carriers of the risk allele [56, 57]. To explore whether the PNPLA3 variant modulates the effects of weight loss on liver fat and insulin sensitivity, eight homozygous carriers were placed on a hypocaloric low‐carbohydrate diet for six days [57]. Liver fat content (measured by PDFF), whole‐body insulin sensitivity of glucose metabolism (euglycemic clamp technique), and lipolysis ([2H5]glycerol infusion) were measured before and after the diet. At baseline, fasting serum insulin and C‐peptide concentrations were lower in mutation carriers. However after a mean weight loss of 3.1 kg, liver fat decreased by 45% in patients with PNPLA3148M but by 18% only in controls. The PNPLA3 variant also modulated the changes in metabolic profile and intrahepatic triglycerides (IHTG, measured by proton magnetic resonance spectroscopy) in 154 NAFLD patients [58]. The presence of the mutation and BMI were independently associated with greater reduction in IHTG (genotype II: 3.7 ± 5.2%, IM: 6.5 ± 3.6%, MM: 11.3 ± 8.8%). Although the PNPLA3 risk allele confers a higher risk of NAFLD, these patients are more sensitive to the beneficial effects of lifestyle modification. In the  same line, obese patients carrying the prosteatotic PNPLA3148M allele lose more weight and liver fat one year after bariatric surgery, as compared to carriers of PNPLA3 wild‐type alleles [59]. The PNPLA3 genotype and the initial grade of steatosis were independent predictors of improvement of steatosis after surgery. In the WELCOME trial, 103 patients with NAFLD were randomized to ω3 fatty acids or placebo for 15–18 months in a randomized controlled trial. Adjusting for baseline measurement and covariates, the PNPLA3 p.I148M variant was independently associated with percentage of docosahexaenoic acid enrichment and end of study liver fat percentage, but did not influence the change in serum triglyceride concentrations [60]. The PNPLA3 p.I148M variant has also been reported to reduce the beneficial effects of statins on steatohepatitis [61].

TM6SF2 AS SECOND NAFLD RISK GENE A GWAS with a denser panel of exonic polymorphisms in three independent populations (N > 80 000) identified the TM6SF2 variant p.E167K (c.449C>T, rs58542926) to confer susceptibility to NAFLD in addition to PNPLA3 [23]. The TM6SF2 variant is markedly less frequent than the PNPLA3 polymorphism in all ethnic groups (3–7%). TM6SF2 encodes a protein of 351 amino acids with 7–10 predicted transmembrane domains. TM6SF2 is localized in the ER and the Golgi complex of hepatocytes and enterocytes (Figure  41.5), which synthesize apolipoprotein B‐ containing lipoproteins [52]. In contrast to PNPLA3, the expression of TM6SF2 is not altered by dietary challenges [52]. The polymorphism is associated with higher liver fat content but lower serum levels of total and low‐density lipoprotein (LDL) cholesterol, and triglycerides. The TM6SF2 p.E167K variant does not exert its effect on hepatic fat content or circulating lipid profiles by causing hepatic insulin resistance [62]. In patients with histologically proven NAFLD, an association of this variant with the degree of steatosis, necroinflammation, ballooning, and advanced fibrosis was observed, after adjustment for age, sex, BMI, and diabetes [23, 63] (Table 41.2). Accordingly, the TM6SF2 variant was independently associated with advanced fibrosis, when assessed noninvasively by transient elastography in 890 individuals [64]. TM6SF2 protein expression was decreased markedly in liver  of NAFLD patients, compared to controls [65]. siRNA knockdown of Tm6sf2 in mice decreases very‐low‐density ­lipoprotein (VLDL) secretion and increases cellular triglyceride concentration and lipid droplet content, whereas overexpression reduces liver cell steatosis [66, 67]. Chronic inactivation of Tm6sf2 in mice is associated with hepatic steatosis and inflammation as well as decreased plasma levels of total and LDL cholesterol, thus recapitulating the phenotypes observed in humans [52, 68]. These observations indicate that TM6SF2 activity is required for normal VLDL assembly and that carriers of the TM6SF2 p.E167K variant have fatty liver as a result of reduced VLDL lipidation (Figure 41.5). As a result, these patients display lower circulating lipids and reduced risk of developing carotid plaques and cardiovascular events [63]. In addition, the effect of insulin on glucose production and lipolysis is better in patients carrying the TM6SF2 p.E167K variant. Hence, they display a distinct subtype of NAFLD, which is characterized by preserved insulin sensitivity with regard to lipolysis, hepatic glucose production, and lack of hypertriglyceridemia despite clearly increased hepatic fat ­content [62]. The dual and opposite role of the TM6SF2 variant in protecting against cardiovascular disease and conferring risk for NAFLD was confirmed in a meta‐analysis of 10 studies. Pooled estimates of random effects in more than 100 000 individuals showed that homozygous or heterozygous carriers of the minor T allele are protected from cardiovascular disease, showing lower levels of total and LDL cholesterol as well as triglycerides, while hepatic lipid fat content is about 2% higher [69]. Accordingly, there is a moderate overall effect on the risk of NAFLD (OR 2.1). Hence, the variant confers protection against cardiovascular disease at the expense of an increased risk of NAFLD.



41:  Clinical Genomics of NAFLD

515

Figure 41.5  Role of TM6SF2 in VLDL lipidation. VLDL synthesis is initiated in the ER with the co‐translational addition of phospholipids to apolipoprotein (Apo) B. The addition of triglycerides to the particle also begins in the ER, a process that requires MTTP. The partially lipidated VLDL particle is packaged into vesicles and exported to the Golgi, where they appear to undergo further bulk phase lipidation. TM6SF2 promotes this bulk phase lipidation, either by transporting neutral lipids from lipid droplets to the particle by transferring lipid to MTTP (➀), to neutral lipid droplets in the ER lumen (➁), or directly to the nascent VLDL particle (➂); alternatively, TM6SF2 could participate in the transfer of lipid to the particle en route to or within the Golgi complex (➃ and ➄). Abbreviations: ER, endoplasmic reticulum; LD, lipid droplet; MTTP, microsomal triglyceride transfer protein; PL, phospholipid; PLTP, phospholipid transfer protein; TG, triglycerides; VLDL, very‐low‐density lipoprotein. Reproduced with permission of Smagris from [52].

Interestingly, the PNPLA3 and TM6SF2 variants are not only associated with clinical phenotypes but also with health services utilization in the general population [70]. In 3759 participants from the study of health in Pomerania (SHIP), homozygous carriers of the PNPLA3 risk allele had an increased odds of hospitalization (OR 1.5) as compared to major allele homozygous subjects, and carriers of the TM6SF2 risk allele had higher outpatient utilization (+68%) and inpatient days than major allele homozygous subjects. These findings highlight the strong genetic effects that can even be detected in health economic analysis. Recently the PNPLA3 p.I148M polymorphism was studied in four large cohorts with extensive health information (23andMe, UK biobank, FINRISK, CHOP) for association with 1683 binary endpoints in up to 700 000 individuals [71]. This study design has been termed phenome‐wide association study (PheWAS), which is an unbiased approach to test for associations between genetic variants and a wide range of phenotypes in large population cohorts [72]. Of note, the PNPLA3 variant, which predisposes to fatty liver disease, was associated with diabetes and lower serum cholesterol levels, and an inverse association was also detected for gallstones, acne, and gout. Beyond that, the analysis also indicated that carriers of the risk allele are prone to develop drug‐ induced liver injury (OR 1.5) when treated with non‐steroidal anti‐inflammatory drugs (NSAID) such as ibuprofen or aspirin. These associations were replicated in the UK biobank cohort and remained when adjusting for elevated liver tests [71].

MBOAT7: THE THIRD MAN After a GWAS had reported that the rs641738 C>T variant linked to the 3’ untranslated region of the gene encoding membrane‐ bound O‐acyltransferase domain‐containing 7 gene (MBOAT7, aka lysophosphatidylinositol‐acyltransferase 1, LPIAT1) increases the risk for alcoholic cirrhosis [73], 2736 participants from the Dallas heart study who had undergone proton magnetic resonance spectroscopy to measure IHTG and 1149 European individuals from the liver biopsy cross‐sectional cohort were genotyped for rs641738 [74]. This variant, which encodes p.G17E in the transmembrane channel‐like 4 (TMC4), is associated with suppression of MBOAT7 at mRNA and protein levels. It was also associated with increased hepatic fat content, more severe liver damage, and increased stage of fibrosis compared with subjects without the variant. MBOAT7 transfers polyunsaturated fatty acids (PUFA) such as arachidonic acid to lysophospholipids. The MBOAT7 rs641738 risk allele was associated with altered plasma phosphatidylinositol (PI) species, consistent with decreased MBOAT7 function. Metabolic profiling indicates changes of PI side‐chain remodeling, in particular a lack of transfer of arachidonoyl‐CoA to lyso‐PI (Lands cycle) [75]. Interestingly, PNPLA3 promotes the transfer of PUFA, especially arachidonic acid, from triglycerides to phospholipids in hepatic lipid droplets [76]. These findings together indicate a potential role of arachidonic acid incorporation in phospholipids during NAFLD pathogenesis.

516

THE LIVER:  PROTECTIVE GENE VARIATION

ADDITIONAL RISK GENES GWAS and their meta‐analyses have identified additional variants that might be associated with NAFLD. For example, Speliotes and coworkers [17, 77] reported that variants associated at genome‐wide significant levels in or near the glucokinase regulatory protein (GCKR), lysophospholipase‐like 1 (LYPLAL1) and protein phosphatase 1‐regulatory subunit 3b (PPP1R3B) might be associated with liver fat content and/or histopathological NAFLD phenotypes. The common missense loss‐of‐function GCKR mutation p.P446L (c.1337T>C, rs1260326) reduces the ability of GCKR to inhibit glucokinase in response to fructose‐6‐phosphate, thereby increasing glucokinase activity and hepatic glucose uptake. The unrestricted hepatic glycolysis reduces fasting glucose and insulin levels but increases the production of malonyl‐CoA, which in turn favors hepatic fat accumulation by serving as a substrate for de novo lipogenesis and by disruption of mitochondrial fatty acid β‐oxidation [78, 79]. The variants at these loci exhibit distinct patterns of association with serum lipids as well as glycemic and anthropometric traits, which also differ with ethnicity. Again NAFLD‐associated variants are not uniformly associated with abnormalities in serum lipids or glycemic and anthropometric traits, suggesting genetic heterogeneity in the pathways influencing these traits.

PROTECTIVE GENE VARIATION Abul‐Husn et  al. [80] reported that a truncated variant of hydroxysteroid 17β dehydrogenase 13 (HSD17B13) is associated with a reduced risk of chronic liver disease and of progression from steatosis to steatohepatitis. It has to be noted that the term “protection” represents the flipside of “susceptibility”, hence it depends on the population frequencies of the alleles, with the minor allele defining the direction of the effect (predisposition vs. protection). Using exome sequence data and electronic health records from 46 544 participants in the DiscovEHR human genetics study, gene variants that are associated with serum activities of aminotransferases were identified. Replicated variants were evaluated for association with clinical diagnoses of chronic liver disease in DiscovEHR study participants as well as two independent cohorts (N = 37 173) and with histopathological severity of liver disease in 2391 human liver biopsies. In homozygous and heterozygous carriers of the variant, the risks of NAFLD and NASH‐cirrhosis were reduced by 30% and 17% and 49% and 26%, respectively. Of note, the HSD17B13 variant ameliorated liver injury associated with the PNPLA3 p.I148M risk allele. In a second detailed histopathological study of 356 patients with biopsy‐proven disease the variant protected against lobular inflammation, ballooning degeneration, and fibrosis [81]. Hydroxysteroid17β dehydrogenases (HSD17B) form an enzyme family characterized by their ability to catalyze reactions in steroid and lipid metabolism. Recently Ma et al. [82] reported that HSD17B13 is a hepatic retinol dehydrogenase that is targeted to lipid droplets. Mice deficient in Hsd17b13 develop hepatic steatosis, while serum steroid concentrations are normal

[83]. In line with these changes, the expression of key enzymes in fatty acid synthesis, such as fatty acid synthase, acetyl‐CoA carboxylase 1, and stearoyl‐CoA desaturase, is increased in livers from Hsd17b3 knockout mice, while glucose tolerance does not differ [83].

Studies in pediatric cohorts NAFLD is becoming an emerging health problem in pediatric populations. In line with the studies in adult patients, an association between fatty liver and the PNPLA3 mutation p.I148M was observed in children with NAFLD, too. In a study by Valenti et al. [20], the PNPLA3 variant was associated with the steatosis grade in 149 children with biopsy‐proven NAFLD. In this cohort, all 23 children who were homozygous carriers of the PNPLA3 mutation were diagnosed with NASH [20]. The association between the PNPLA3 mutation and pediatric NAFLD has also been detected in African American [84] and Chinese children [85]. Interestingly, children carrying the PNPLA3 risk allele seem to be predisposed to the early development of NAFLD [21]. The analysis of 6–12‐year‐old Mexican children showed that already at this age the PNPLA3 mutation is associated with increased serum ALT activities [86], and we detected the same association in a cohort of German children aged 5–9 years [87]. The TM6SF2 variant p.E167K was also shown to be associated with hepatic steatosis and increased aminotransferase activities but lower serum cholesterol and triglyceride concentrations in obese children [88, 89]. In multiethnic cohorts of obese children and adolescents [90, 91], the MBOAT7 risk allele was associated with hepatic fat content (as determined by magnetic resonance imaging), fasting insulin and plasma glucose levels, and reduced whole‐body insulin sensitivity, independent of age, sex, and BMI z‐score. Moreover, these studies detected joint effects among TM6SF2 p.E167K, PNPLA3 p.I148M, and the MBOAT7 rs626283 and GCKR rs1260326 polymorphisms in determining IHTG. The four polymorphisms combined explain about 20% of the hepatic fat fraction in Caucasian obese children. A practical consideration is that physical activity and weight reduction might substantially improve the liver status in children carrying the risk variants and therefore rescue their deleterious liver phenotypes [56]. This possibility points to the need of early detection of children carrying risk mutations who may require more careful follow‐up and tailored therapies aiming to intercept NAFLD progression.

Combined effects of gene variants Mancina et  al. [74] analyzed the combined genetic effects of PNPLA3, TM6SF2, and MBOAT7 effects on NAFLD histopathology (Table 41.3). They did not observe an interaction, instead the three gene variants appear to act in an additive fashion, with a stepwise increase in mean hepatic fat content per each additional risk allele (Figure 41.6). The combined effects of the PNPLA3, TM6SF2, and MBOAT7 variants on NAFLD severity were studied in a German multicenter biopsy‐based study that recruited 515 patients with NAFLD [92]. In the multivariate model, the PNPLA3 and TM6SF2 variants were associated with steatosis, and fibrosis stages were affected by the PNPLA3 and MBOAT7 polymorphism. Of note, a significant increase of serum AST



41:  Clinical Genomics of NAFLD

517

Table 41.3  Population attributable fractionsa of NAFLD genes

PNPLA3 TM6SF2 MBOAT7

Steatosis (confidence intervals)

NASH

Fibrosis F2–F4

23% (11–36) 4% (0–14) 15% (3–28)

19% (12–26) 4% (1–8) 7% (0–14)

18% (10–26) 3% (0–7) 11% (2–23)

a  The contribution of a risk factor to a disease or a death is quantified using PAF. PAF is the proportional reduction in population disease or mortality that would occur if exposure to a risk factor were reduced to an alternative ideal exposure scenario.

AST [U/I]

(b)

(a) 20

P < 0.0001

500 300 100 100

50

0

1

2

3

4

5

627

1066

747

248

41

6

Number of risk alleles

s al

le

le

s k 5

ris

k

al

le

le

s ris 4

3

ris

k

al

le

le

s le le al k

P = 0.08

400 300 200 150 ALT [U/I]

n=

ris

Number of risk alleles

5

0

2

10

1

ris

k

ris

k

al

al

le

le

le

s

le

0

15

0

Mean hepatic TG%

P-trend < .0001

100 50

le le al

5

ris

k

k ris 4

k ris 3

s

s le al

le al

le al k ris 2

le

s le

s le

le k ris

1

0

ris

k

al

al

le

n = 515

le

le

s

0

Number of risk alleles

Figure 41.6  Combined genetic effects in NAFLD. (a) Association between number of PNPLA3, TM6SF2, and MBOAT7 risk alleles and hepatic triglyceride (TG) content in the Dallas heart study. The graph shows mean hepatic TG content by the number of risk alleles (error bars: SEM). From [74]. (b) Association between number of PNPLA3, TM6SF2, and MBOAT7 risk alleles and liver function tests in the NAFLD CSG study. The graphs illustrate median aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activities by the number of risk alleles (error bars: range). Analyses were performed using trend tests. The following frequencies of carriers of risk alleles were detected: 0 risk alleles, N = 56; one risk allele, N = 142; two risk alleles, N = 170; three risk alleles, N = 117; four risk alleles, N = 27; five risk alleles, N = 3. Reproduced with permission of Elsevier from [92].

activities was associated with the increment of risk alleles of either of the genotypes, and similar trends existed for ALT and γ‐glutamyl transpeptidase levels. The same observation was made in Korea: the PNPLA3 and TM6SF2 variants, but not the MBOAT7 variant were associated with both NASH and significant fibrosis (≥ F2), even after adjustment for insulin resistance [93].

MONOGENIC DISEASES CAUSING NAFLD Monogenic diseases that are associated with NAFLD are rare and usually result in extrahepatic manifestations that dominate

over hepatic steatosis. However, they provide insight into the pathogenesis of hepatic steatosis and can be considered in rare cases when the etiology is unclear. Examples include abetalipoproteinemia (microsomal triglyceride transfer protein, MTTP), hypobetalipoproteinemia (apolipoprotein B, APOB), citrullinemia type 2 (solute carrier SLC25A13), Wilson disease (ATPase ATP7B), neutral lipid storage disease (PNPLA2), and cholesteryl ester storage disease (lysosomal acid lipase, LIPA). Rare inherited mitochondriopathies have also been associated with NASH and point to the role of impaired fatty acid oxidation in the pathogenesis of NASH. Genetic testing is helpful where a rare monogenic inherited disease of lipid metabolism or mitochondrial function is suspected. More recently exome

518

THE LIVER:  REFERENCES

and whole genome sequencing is applied to identify novel gene variants causing liver diseases, including severe types of NAFLD.

PNPLA3‐ AND TM6SF2‐ASSOCIATED STEATOHEPATITIS AND FUTURE DIRECTIONS PNPLA3 (and TM6SF2 plus MBOAT7) genotyping may be used as novel non‐invasive indicator for an increased risk of progressive fatty liver disease and could be included in the clinical decision‐making in patients with NAFLD. Because in patients who carry the PNPLA3 risk variant increased lipid content and inflammation in liver can be driven predominantly by PNPLA3 together with environmental risk factors, the name PNPLA3‐NAFLD or PASH (i.e. PNPLA3‐associated steatohepatitis) might be used as a novel gene‐based liver disease entity [94, 95]. PNPLA3‐ NAFLD/PASH represents an example of how to reclassify disease according to molecular pathways and pathophysiological changes in the era of personalized medicine [96]. Carriers of the genetic risk factors could benefit from a more systematic, early, and careful surveillance of complications of progressive fatty liver disease, including HCC in the absence of cirrhosis [97].

REFERENCES   1. Chalasani, N., Younossi, Z., Lavine, J.E. et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology, 2018;67(1):328–57.   2. Romeo, S., Kozlitina, J., Xing, C. et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nat Genet, 2008;40(12): 1461–5.   3. Petaja, E.M. and Yki‐Järvinen, H. Definitions of normal liver fat and the association of insulin sensitivity with acquired and genetic NAFLD‐A systematic review. Int J Mol Sci, 2016;17(5).   4. Verhaegh, P., Bavalia, R., Winkens, B. et al. Noninvasive tests do not accurately differentiate nonalcoholic steatohepatitis from simple steatosis: a systematic review and meta‐analysis. Clin Gastroenterol Hepatol, 2018;16(6):837–61.   5. Estes, C., Razavi, H., Loomba, R. et al. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology, 2018;67(1):123–33.   6. Loomba, R., Schork, N., Chen, C.H. et al. Heritability of hepatic fibrosis and steatosis based on a prospective twin study. Gastroenterology, 2015;149(7): 1784–93.   7. Zechner, R., Zimmermann, R., Eichmann, T.O. et al. Fat signals – lipases and lipolysis in lipid metabolism and signaling. Cell Metab, 2012;15(3):279–91.   8. Huang, Y., He, S., Li, J.Z. et al. A feed‐forward loop amplifies nutritional regulation of PNPLA3. Proc Natl Acad Sci USA, 2010;107(17):7892–7.   9. Sookoian, S. and Pirola, C.J. Meta‐analysis of the influence of I148M variant of patatin‐like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease. Hepatology, 2011;53(6):1883–94. 10. Krawczyk, M., Grünhage, F., Mahler, M. et  al. The common adiponutrin variant p.I148M does not confer gallstone risk but affects fasting glucose and triglyceride levels. J Physiol Pharmacol, 2011;62(3):369–75. 11. Diehl, A.M. and Day, C. Nonalcoholic steatohepatitis. N Engl J Med, 2018;378(8):781. 12. Chambers, J.C., Zhang, W., Sehmi, J. et al. Genome‐wide association study identifies loci influencing concentrations of liver enzymes in plasma. Nat Genet, 2011;43(11):1131–8.

13. Yuan, X., Waterworth, D., Perry, J.R. et al. Population‐based genome‐wide association studies reveal six loci influencing plasma levels of liver enzymes. Am J Hum Genet, 2008;83(4):520–8. 14. Kawaguchi, T., Sumida, Y., Umemura, A. et al. Genetic polymorphisms of the human PNPLA3 gene are strongly associated with severity of non‐alcoholic fatty liver disease in Japanese. PLoS One, 2012;7(6), e38322. 15. Kitamoto, T., Kitamoto, A., Yoneda, M. et al. Genome‐wide scan revealed that polymorphisms in the PNPLA3, SAMM50, and PARVB genes are associated with development and progression of nonalcoholic fatty liver disease in Japan. Hum Genet, 2013;132(7):783–92. 16. Kotronen, A., Johansson, L.E., Johansson, L.M. et al. A common variant in PNPLA3, which encodes adiponutrin, is associated with liver fat content in humans. Diabetologia, 2009;52(6):1056–60. 17. Speliotes, E.K., Yerges‐Armstrong, L.M., Wu, J. et al. Genome‐wide association analysis identifies variants associated with nonalcoholic fatty liver disease that have distinct effects on metabolic traits. PLoS Genet, 2011;7(3), e1001324. 18. Sookoian, S., Castano, G.O., Burgueno, A.L. et al. A nonsynonymous gene variant in the adiponutrin gene is associated with nonalcoholic fatty liver disease severity. J Lipid Res, 2009;50(10):2111–6. 19. Valenti, L., Al‐Serri, A., Daly, A.K. et al. Homozygosity for the patatin‐like phospholipase‐3/adiponutrin I148M polymorphism influences liver fibrosis in patients with nonalcoholic fatty liver disease. Hepatology, 2010;51(4): 1209–17. 20. Pelusi, S., Cespiati, A., Rametta, R. et al. Prevalence and risk factors of significant fibrosis in patients with nonalcoholic fatty liver without steatohepatitis. Clin Gastroenterol Hepatol, 2019;epub. 21. Rotman, Y., Koh, C., Zmuda, J.M. et al. The association of genetic variability in patatin‐like phospholipase domain‐containing protein 3 (PNPLA3) with histological severity of nonalcoholic fatty liver disease. Hepatology, 2010;52(3):894–903. 22. Xu, R., Tao, A., Zhang, S. et al. Association between patatin‐like phospholipase domain containing 3 gene (PNPLA3) polymorphisms and nonalcoholic fatty liver disease: a HuGE review and meta‐analysis. Sci Rep, 2015;5, 9284. 23. Liu, Y.L., Reeves, H.L., Burt, A.D. et  al. TM6SF2 rs58542926 influences hepatic fibrosis progression in patients with non‐alcoholic fatty liver disease. Nat Commun, 2014;5, 4309. 24. Krawczyk, M., Grünhage, F., Zimmer, V. et al. Variant adiponutrin (PNPLA3) represents a common fibrosis risk gene: non‐invasive elastography‐based study in chronic liver disease. J Hepatol, 2011;55(2):299–306. 25. Kim, D. and Kim, W.R. Nonobese fatty liver disease. Clin Gastroenterol Hepatol, 2017;15(4):474–85. 26. Wei, J.L., Leung, J.C., Loong, T.C. et al. Prevalence and severity of nonalcoholic fatty liver disease in non‐obese patients: a population study using proton‐magnetic resonance spectroscopy. Am J Gastroenterol, 2015;110(9): 1306–14. 27. Leung, J.C., Loong, T.C., Wei, J.L. et al. Histological severity and clinical outcomes of nonalcoholic fatty liver disease in nonobese patients. Hepatology, 2017;65(1):54–64. 28. Krawczyk, M., Bantel, H., Rau, M. et  al. Could inherited predisposition drive non‐obese fatty liver disease? Results from German tertiary referral centers. J Hum Genet, 2018;63(5):621–6. 29. Fracanzani, A.L., Petta, S., Lombardi, R. et  al. Liver and cardiovascular damage in patients with lean nonalcoholic fatty liver disease, and association with visceral obesity. Clin Gastroenterol Hepatol, 2017;15(10):1604–11. 30. Burza, M.A., Pirazzi, C., Maglio, C. et  al. PNPLA3 I148M (rs738409) genetic variant is associated with hepatocellular carcinoma in obese individuals. Dig Liver Dis, 2012;44(12):1037–41. 31. Liu, Y.L., Patman, G.L., Leathart, J.B. et  al. Carriage of the PNPLA3 rs738409 C >G polymorphism confers an increased risk of non‐alcoholic fatty liver disease associated hepatocellular carcinoma. J Hepatol, 2014;61(1):75–81. 32. Anstee, Q.M., Seth, D., Day, C.P. Genetic factors that affect risk of alcoholic  and nonalcoholic fatty liver disease. Gastroenterology, ­ 2016;150(8):1728–44. 33. Hassan, M.M., Kaseb, A., Etzel, C.J. et al. Genetic variation in the PNPLA3 gene and hepatocellular carcinoma in USA: risk and prognosis prediction. Mol Carcinog, 2013;52(1), E139–47. 34. Stender, S., Kozlitina, J., Nordestgaard, B.G. et al. Adiposity amplifies the genetic risk of fatty liver disease conferred by multiple loci. Nat Genet, 2017;49(6):842–7.



41:  Clinical Genomics of NAFLD

35. Palasciano, G., Moschetta, A., Palmieri, V.O. et al. Non‐alcoholic fatty liver disease in the metabolic syndrome. Curr Pharm Des, 2007;13(21):2193–8. 36. Kantartzis, K., Peter, A., Machicao, F. et al. Dissociation between fatty liver and insulin resistance in humans carrying a variant of the patatin‐like phospholipase 3 gene. Diabetes, 2009;58(11):2616–23. 37. Speliotes, E.K., Butler, J.L., Palmer, C.D. et al. PNPLA3 variants specifically confer increased risk for histologic nonalcoholic fatty liver disease but not metabolic disease. Hepatology, 2010;52(3):904–12. 38. Krarup, N.T., Grarup, N., Banasik, K. et al. The PNPLA3 rs738409 G‐allele associates with reduced fasting serum triglyceride and serum cholesterol in Danes with impaired glucose regulation. PLoS One, 2012;7(7), e40376. 39. Stojkovic, I.A., Ericson, U., Rukh, G. et al. The PNPLA3 Ile148Met interacts with overweight and dietary intakes on fasting triglyceride levels. Genes Nutr, 2014;9(2):388. 40. Palmer, C.N., Maglio, C., Pirazzi, C. et al. Paradoxical lower serum triglyceride levels and higher type 2 diabetes mellitus susceptibility in obese individuals with the PNPLA3 148M variant. PLoS One, 2012;7(6), e39362. 41. Rembeck, K., Maglio, C., Lagging, M. et al. PNPLA 3 I148M genetic variant associates with insulin resistance and baseline viral load in HCV genotype 2 but not in genotype 3 infection. BMC Med Genet, 2012;13, 82. 42. Hyysalo, J., Gopalacharyulu, P., Bian, H. et  al. Circulating triacylglycerol signatures in nonalcoholic fatty liver disease associated with the I148M variant in PNPLA3 and with obesity. Diabetes, 2014;63(1):312–22. 43. Kumari, M., Schoiswohl, G., Chitraju, C. et al. Adiponutrin functions as a nutritionally regulated lysophosphatidic acid acyltransferase. Cell Metab, 2012;15(5):691–702. 44. Pirazzi, C., Valenti, L., Motta, B.M. et  al. PNPLA3 has retinyl‐palmitate lipase activity in human hepatic stellate cells. Hum Mol Genet, 2014;23(15):4077–85. 45. Huang, Y., Cohen, J.C., Hobbs, H.H. Expression and characterization of a PNPLA3 protein isoform (I148M) associated with nonalcoholic fatty liver disease. J Biol Chem, 2011;286(43):37085–93. 46. Dubuquoy, C., Robichon, C., Lasnier, F. et al. Distinct regulation of adiponutrin/PNPLA3 gene expression by the transcription factors ChREBP and SREBP1c in mouse and human hepatocytes. J Hepatol, 2011;55(1):145–53. 47. Perttila, J., Huaman‐Samanez, C., Caron, S. et al. PNPLA3 is regulated by glucose in human hepatocytes, and its I148M mutant slows down triglyceride hydrolysis. Am J Physiol Endocrinol Metab, 2012;302(9), E1063–9. 48. Valenti, L. and Dongiovanni, P. Mutant PNPLA3 I148M protein as pharmacological target for liver disease. Hepatology, 2017;66(4):1026–8. 49. He, S., McPhaul, C., Li, J.Z. et al. A sequence variation (I148M) in PNPLA3 associated with nonalcoholic fatty liver disease disrupts triglyceride hydrolysis. J Biol Chem, 2010;285(9):6706–15. 50. Xin, Y.N., Zhao, Y., Lin, Z.H. et  al. Molecular dynamics simulation of PNPLA3 I148M polymorphism reveals reduced substrate access to the catalytic cavity. Proteins, 2013;81(3):406–14. 51. Chamoun, Z., Vacca, F., Parton, R.G. et al. PNPLA3/adiponutrin functions in lipid droplet formation. Biol Cell, 2013;105(5):219–33. 52. Smagris, E., Gilyard, S., BasuRay, S. et al. Inactivation of Tm6sf2, a gene defective in fatty liver disease, impairs lipidation but not secretion of very low density lipoproteins. J Biol Chem, 2016;291(20):10659–76. 53. BasuRay, S., Smagris, E., Cohen, J.C. et al. The PNPLA3 variant associated with fatty liver disease (I148M) accumulates on lipid droplets by evading ubiquitylation. Hepatology, 2017;66(4):1111–24. 54. Li, J.Z., Huang, Y., Karaman, R. et al. Chronic overexpression of PNPLA3I148M in mouse liver causes hepatic steatosis. J Clin Invest, 2012;122(11):4130–44. 55. Smagris, E., BasuRay, S., Li, J. et al. Pnpla3I148M knockin mice accumulate PNPLA3 on lipid droplets and develop hepatic steatosis. Hepatology, 2015;61(1):108–18. 56. Marzuillo, P., Grandone, A., Perrone, L. et al. Weight loss allows the dissection of the interaction between abdominal fat and PNPLA3 (adiponutrin) in the liver damage of obese children. J Hepatol, 2013;59(5):1143–4. 57. Sevastianova, K., Kotronen, A., Gastaldelli, A. et  al. Genetic variation in PNPLA3 (adiponutrin) confers sensitivity to weight loss‐induced decrease in liver fat in humans. Am J Clin Nutr, 2011;94(1):104–11. 58. Shen, J., Wong, G.L., Chan, H.L. et al. PNPLA3 gene polymorphism and response to lifestyle modification in patients with nonalcoholic fatty liver disease. J Gastroenterol Hepatol, 2015;30(1):139–46. 59. Krawczyk, M., Jimenez‐Aguero, R., Alustiza, J.M. et al. PNPLA3 p.I148M variant is associated with greater reduction of liver fat content after bariatric surgery. Surg Obes Relat Dis, 2016;12(10):1838–46.

519

60. Scorletti, E., West, A.L., Bhatia, L. et al. Treating liver fat and serum triglyceride levels in NAFLD, effects of PNPLA3 and TM6SF2 genotypes: Results from the WELCOME trial. J Hepatol, 2015;63(6):1476–83. 61. Dongiovanni, P., Petta, S., Mannisto, V. et al. Statin use and non‐alcoholic steatohepatitis in at risk individuals. J Hepatol, 2015;63(3):705–12. 62. Zhou, Y., Llaurado, G., Oresic, M. et al. Circulating triacylglycerol signatures and insulin sensitivity in NAFLD associated with the E167K variant in TM6SF2. J Hepatol, 2015;62(3):657–63. 63. Dongiovanni, P., Petta, S., Maglio, C. et al. Transmembrane 6 superfamily member 2 gene variant disentangles nonalcoholic steatohepatitis from cardiovascular disease. Hepatology, 2015;61(2):506–14. 64. Petta, S., Di Marco, V., Pipitone, R.M. et al. Prevalence and severity of nonalcoholic fatty liver disease by transient elastography: genetic and metabolic risk factors in a general population. Liver Int, 2018;38(11):2060–8. 65. Sookoian, S., Castano, G.O., Scian, R. et al. Genetic variation in transmembrane 6 superfamily member 2 and the risk of nonalcoholic fatty liver disease and histological disease severity. Hepatology, 2015;61(2):515–25. 66. Kozlitina, J., Smagris, E., Stender, S. et al. Exome‐wide association study identifies a TM6SF2 variant that confers susceptibility to nonalcoholic fatty liver disease. Nat Genet, 2014;46(4):352–6. 67. Mahdessian, H., Taxiarchis, A., Popov, S. et al. TM6SF2 is a regulator of liver fat metabolism influencing triglyceride secretion and hepatic lipid droplet content. Proc Natl Acad Sci USA, 2014;111(24):8913–8. 68. Fan, Y., Lu, H., Guo, Y. et al. Hepatic transmembrane 6 superfamily member 2 regulates cholesterol metabolism in mice. Gastroenterology, 2016;150(5):1208–18. 69. Pirola, C.J. and Sookoian, S. The dual and opposite role of the TM6SF2‐ rs58542926 variant in protecting against cardiovascular disease and conferring risk for nonalcoholic fatty liver: a meta‐analysis. Hepatology, 2015;62(6):1742–56. 70. Kopp, J., Flessa, S., Lieb, W. et al. Association of PNPLA3 rs738409 and TM6SF2 rs58542926 with health services utilization in a population‐based study. BMC Health Serv Res, 2016;16, 41. 71. Diogo, D., Tian, C., Franklin, C.S. et al. Phenome‐wide association studies across large population cohorts support drug target validation. Nat Commun, 2018;9(1):4285. 72. Bush, W.S., Oetjens, M.T., Crawford, D.C. Unravelling the human genome‐ phenome relationship using phenome‐wide association studies. Nat Rev Genet, 2016;17(3):129–45. 73. Buch, S., Stickel, F., Trepo, E. et al. A genome‐wide association study confirms PNPLA3 and identifies TM6SF2 and MBOAT7 as risk loci for alcohol‐ related cirrhosis. Nat Genet, 2015;47(12):1443–8. 74. Mancina, R.M., Dongiovanni, P., Petta, S. et al. The MBOAT7‐TMC4 variant rs641738 increases risk of nonalcoholic fatty liver disease in individuals of european descent. Gastroenterology, 2016;150(5):1219–30. 75. Wang, B. and Tontonoz, P. Phospholipid remodeling in physiology and disease. Annu Rev Physiol, 2018;81:165–88. 76. Mitsche, M.A., Hobbs, H.H., and Cohen, J.C. Patatin‐like phospholipase domain‐containing protein 3 promotes transfer of essential fatty acids from triglycerides to phospholipids in hepatic lipid droplets. J Biol Chem, 2018;293(18):6958–68. 77. Hernaez, R., McLean, J., Lazo, M. et al. Association between variants in or near PNPLA3, GCKR, and PPP1R3B with ultrasound‐defined steatosis based on data from the third National Health and Nutrition Examination Survey. Clin Gastroenterol Hepatol, 2013;11(9):1183–90. 78. Beer, N.L., Tribble, N.D., McCulloch L.J. et al. The P446L variant in GCKR associated with fasting plasma glucose and triglyceride levels exerts its effect  through increased glucokinase activity in liver. Hum Mol Genet, 2009;18(21):4081–8. 79. Eslam, M., Valenti, L., and Romeo, S. Genetics and epigenetics of NAFLD and NASH: clinical impact. J Hepatol, 2018;68(2):268–79. 80. Abul‐Husn, N.S., Cheng, X., Li, A.H. et al. A protein‐truncating HSD17B13 variant and protection from chronic liver disease. N Engl J Med, 2018;378(12):1096–106. 81. Pirola, C.J., Garaycoechea, M., Flichman, D. et al. Splice variant rs72613567 prevents worst histologic outcomes in patients with nonalcoholic fatty liver disease. J Lipid Res, 2019;60(1):176–85. 82. Ma, Y., Belyaeva, O.V., Brown, P.M. et  al. 17‐Beta hydroxysteroid dehydrogenase 13 is a hepatic retinol dehydrogenase associated with ­ histological features of nonalcoholic fatty liver disease. Hepatology, ­ 2019;69(4):1504–19.

520

THE LIVER:  REFERENCES

83. Adam, M., Heikela, H., Sobolewski, C. et al. Hydroxysteroid (17beta) dehydrogenase 13 deficiency triggers hepatic steatosis and inflammation in mice. FASEB J, 2018;32(6):3434–47. 84. Santoro, N., Kursawe, R., D’Adamo E. et al. A common variant in the patatin‐like phospholipase 3 gene (PNPLA3) is associated with fatty liver disease in obese children and adolescents. Hepatology, 2010;52(4):1281–90. 85. Lin, Y.C., Chang, P.F., Hu, F.C. et al. A common variant in the PNPLA3 gene is a risk factor for non‐alcoholic fatty liver disease in obese Taiwanese children. J Pediatr, 2011;158(5):740–4. 86. Larrieta‐Carrasco, E., Leon‐Mimila, P., Villarreal‐Molina, T. et  al. Association of the I148M/PNPLA3 variant with elevated alanine transaminase levels in normal‐weight and overweight/obese Mexican children. Gene, 2013;520(2):185–8. 87. Krawczyk, M., Liebe, R., Maier, I.B., et  al. The frequent adiponutrin (PNPLA3) variant p.Ile148Met is associated with early liver injury: analysis of a German pediatric cohort. Gastroenterol Res Pract, 2015;205079. 88. Grandone, A., Cozzolino, D., Marzuillo, P. et al. TM6SF2 Glu167Lys polymorphism is associated with low levels of LDL‐cholesterol and increased liver injury in obese children. Pediatr Obes; 2016;11(2):115–9. 89. Goffredo, M., Caprio, S., Feldstein, A.E. et al. Role of TM6SF2 rs58542926 in the pathogenesis of nonalcoholic pediatric fatty liver disease: a multiethnic study. Hepatology, 2016;63(1):117–25. 90. Di Sessa, A., Umano, G.R., Cirillo, G. et al. The membrane‐bound O‐acyltransferase7 rs641738 variant in pediatric nonalcoholic fatty liver fisease. J Pediatr Gastroenterol Nutr, 2018;67(1):69–74. 91. Umano, G.R., Caprio, S., Di Sessa, A. et  al. The rs626283 variant in the MBOAT7 gene is associated with insulin resistance and fatty liver in Caucasian obese youth. Am J Gastroenterol, 2018;113(3):376–83. 92. Krawczyk, M., Rau, M., Schattenberg, J.M. et al. Combined effects of the PNPLA3 rs738409, TM6SF2 rs58542926, and MBOAT7 rs641738 variants on NAFLD severity: a multicenter biopsy‐based study. J Lipid Res, 2017;58(1):247–55. 93. Koo, B.K., Joo, S.K., Kim, D. et  al. Additive effects of PNPLA3 and TM6SF2 on the histological severity of non‐alcoholic fatty liver disease. J Gastroenterol Hepatol, 2018;33(6):1277–85.

  94. Krawczyk, M., Portincasa, P., Lammert, F. PNPLA3‐associated steatohepatitis: toward a gene‐based classification of fatty liver disease. Semin Liver Dis, 2013;33(4):369–79.   95. Yki‐Järvinen, H. Non‐alcoholic fatty liver disease as a cause and a consequence of metabolic syndrome. Lancet Diabetes Endocrinol, 2014;2(11):901–10.   96. Mirnezami, R., Nicholson, J., Darzi, A. Preparing for precision medicine. N Engl J Med, 2012;366(6):489–91.   97. Valenti, L., Dongiovanni, P., Ginanni Corradini, S. et al. PNPLA3 I148M variant and hepatocellular carcinoma: a common genetic variant for a rare disease. Dig Liver Dis, 2013;45(8):619–24.   98. Tian, C., Stokowski, R.P., Kershenobich, D. et  al. Variant in PNPLA3 is associated with alcoholic liver disease. Nat Genet, 2010;42(1):21–3.   99. Stickel, F., Buch, S., Lau, K. et al. Genetic variation in the PNPLA3 gene is associated with alcoholic liver injury in caucasians. Hepatology, 2011;53(1):86–95. 100. Vigano, M., Valenti, L., Lampertico, P. et  al. Patatin‐like phospholipase domain‐containing 3 I148M affects liver steatosis in patients with chronic hepatitis B. Hepatology, 2013;58(4):1245–52. 101. Cai, T., Dufour, J.F., Müllhaupt, B. et  al. Viral genotype‐specific role of PNPLA3, PPARG, MTTP, and IL28B in hepatitis C virus‐associated steatosis. J Hepatol, 2011;55(3):529–35. 102. Müller, T., Buch, S., Berg, T. et al. Distinct, alcohol‐modulated effects of PNPLA3 genotype on progression of chronic hepatitis C. J Hepatol, 2011;55(3):732–3. 103. Valenti, L. and Fargion, S. Patatin‐like phospholipase domain containing‐3 Ile148Met and fibrosis progression after liver transplantation. Hepatology, 2011;54(4):1484. 104. Nischalke, H.D., Berger, C., Luda, C. et al. The PNPLA3 rs738409 148M/M genotype is a risk factor for liver cancer in alcoholic cirrhosis but shows no or weak association in hepatitis C cirrhosis. PLoS One, 2011;6(11):e27087. 105. Younossi, Z., Anstee, Q.M., Marietti, M. et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol, 2018;15(1):11–20. 106. Dubuquoy, C., Burnol, A.F., Moldes, M. PNPLA3, a genetic marker of progressive liver disease, still hiding its metabolic function? Clin Res Hepatol Gastroenterol, 2013;37(1):30–5.

SECTION B: LIVER GROWTH AND REGENERATION

42

Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis Wencheng Zhang1, Amanda Allen1, Eliane Wauthier1, Xianwen Yi2, Homayoun Hani1, Praveen Sethupathy3,10, David Gerber2, Vincenzo Cardinale5, Guido Carpino6,7, Juan Dominguez‐Bendala8,9, Giacomo Lanzoni9, Domenico Alvaro5,8,*, Eugenio Gaudio7,*, and Lola Reid1,4,* Department of Cell Biology and Physiology, University of North Carolina School of Medicine, Chapel Hill, NC, USA 2  Department of Surgery, University of North Carolina School of Medicine, Chapel Hill, NC, USA 3  Department of Genetics, University of North Carolina School of Medicine, Chapel Hill, NC, USA 4  Program in Molecular Biology and Biotechnology, University of North Carolina School of Medicine, Chapel Hill, NC, USA 5  Department of Medico‐Surgical Sciences and Biotechnologies, Sapienza University of Rome, Latina, Italy 6  Department of Movement, Human and Health Sciences, Division of Health Sciences, University of Rome “Foro Italico”, Rome, Italy 7  Department of Anatomical, Histological, Forensic Medicine and Orthopedics Sciences, Sapienza University of Rome, Rome, Italy 8  Department of Medicine and Medical Specialties, Sapienza University of Rome, Rome, Italy 9  Diabetes Research Institute, Miller School of Medicine, University of Miami, Miami, FL, USA 10  Department of Biomedical Sciences, Cornell University, Ithaca, NY, USA 1 

INTRODUCTION Liver, biliary tree, and pancreas are mid‐gut endodermal organs central to handling glycogen and lipid metabolism, detoxification of xenobiotics, processing of nutrients for optimal utilization, regulation of energy needs, and synthesis of diverse factors ranging from coagulation proteins to carrier proteins (e.g. alpha‐fetoprotein (AFP), albumin, transferrin) [1]. The integrity

* L.M. Reid was the only one who wrote most of the chapter with the qualifier that the last segment, that on plasticity was written by Domenico Alvaro and Lola Reid and with input also from Eugenio Gaudio. Amanda Allen, an artist and animator, did the figures. All of the other authors have done the experiments resulting in the discoveries that are summarized in the review. The senior authors in the management of the experiments comprise G. Lanzoni and J. Dominguez‐ Bendala at the Diabetes Research Institute, D. Alvaro and E. Gaudio at Sapienza Medical Center, P. Sethupathy at Cornell, and L. Reid at UNC.

of the body depends heavily on liver, biliary tree, and pancreatic functions, and failure in any of them results in rapid death. The focus of this review is on current knowledge of the newly discovered network of multiple stem/progenitor cell populations present primarily in the biliary tree and giving rise to maturational lineages of cells forming liver and pancreas throughout life. In addition, the chapter summarizes information regarding key paracrine signals from lineage‐dependent epithelial–mesenchymal cell interactions in the maturational processes and found useful for ex vivo maintenance and differentiation of hepato/biliary and pancreatic cells at particular lineage stages. In prior reviews and previous editions of this book are given more details on hepatic stem cells and hepatoblasts, on early studies on the biliary tree stem cells, and on facets of extracellular matrix chemistry and biology, key requirements for ex vivo maintenance of the cells [1–9]. For the sake of brevity, we will not discuss findings regarding the lineage restriction of embryonic stem (ES) cells or induced

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

524

THE LIVER:  THE BILIARY TREE AS A ROOT SYSTEM FOR HEPATIC/PANCREATIC ORGANOGENESIS

pluripotent stem (iPS) cells to an hepatic or pancreatic fate. Information from these studies is provided in representative recent articles and reviews [10–13]. This chapter focuses primarily on studies in human tissues and with some references to similar investigations in mice, rats, and/or pigs. The phenomena associated with stem/progenitor cell maturational lineages in hepatic and pancreatic organogenesis are evident in all mammals studied (mice, rats, pigs, and humans) but with a few distinctive variations in some species as noted below. The stem cell or progenitor cell populations are indicated by an acronym which is preceded by a small letter indicating the species: m, murine; r, rat; p, porcine; h, human.

EMBRYONIC DEVELOPMENT During early development, definitive endoderm derives from embryonic stem cells through the effects of a number of key transcription factors including Goosecoid, MIXL1, SMAD2/3, SOX7, and SOX17 (a transcription factor essential for differentiation of liver) [14]. Endoderm subsequently segregates into foregut (lung, thyroid), midgut (pancreas, biliary tree, and liver), and both foregut and hindgut (intestine) through the effects of specific mixes of transcription factors. Those dictating the midgut organs include SOX9 (transcription factor associated with endodermal tissues), SOX17, FOXA1/FOXA2 (forkhead box a1/2), Onecut2/OC‐2, and others [15–18]. The liver, biliary tree, and pancreas derive from midgut endoderm established at the gastrulation stage of early embryonic development [19]. Among the other organs of endodermal origin, endogenous adult stem cells have been identified in most, including the small and large intestines [20], the stomach [21], and the lungs [22, 23]. The pancreas is distinct in that lineage tracing experiments indicate that there are only very rare stem cells postnatally [24– 26]. Clarification of this paradox has emerged with the many studies on the biliary tree: the stem cells for the pancreas are not  in the pancreas proper but rather within the biliary tree, especially within the peribiliary glands (PBGs) of the hepato‐ pancreatic common duct. These pancreatic stem cells are anatomically linked and give rise to committed progenitors ­ located in pancreatic duct glands (PDGs) within the pancreas proper [27, 28]. During fetal development, the formation of the liver and pancreas occurs with outgrowths on either side of the duodenum that extend and ramify into a branching biliary tree structure that, at its ends, is influenced by the cardiac mesenchyme to form liver and by the retroperitoneal mesenchymal to form pancreas [28]. The gallbladder is a major branch point connected to the common duct via the cystic duct [29]. The pancreas derives from two separate anlage: the dorsal pancreas connected to the duodenum via the accessory duct. The ventral pancreas begins as a branch from the common duct and nearer to the duodenum. The formation of the intestine involves a twisting motion that swings the ventral pancreas anlage to the other side where it subsequently merges with the dorsal pancreas anlage to form the complete pancreatic organ. The liver cannot swing to the opposite side, given its size and its connections into the cardiac

mesenchyme, connections associated with rapid vascularization of the forming liver tissue. This results in the liver and the ­ventral pancreas sharing the hepato‐pancreatic common duct connecting them to the duodenum.

THE BILIARY TREE AS A ROOT SYSTEM FOR HEPATIC/PANCREATIC ORGANOGENESIS (GENERAL ANATOMY AND IN SITU STUDIES) Newly discovered within the last decade is that hepatic and pancreatic organogenesis are ongoing throughout life “fueled” by a network of stem cells within the biliary tree, the ramifying ducts connecting the liver and the pancreas to the duodenum [5, 9, 27, 30–33] (Figure 42.1). The biliary tree, long assumed to be the conduits managing removal of bile from the liver and digestive enzymes from the pancreas, has proven also to be a “root” ­system, a reservoir of stem cells giving rise to maturational ­lineages of cells contributing to hepatic and pancreatic organogenesis and to regenerative processes in these organs. The most primitive of the stem cells are located within extramural peribiliary glands (PBGs, stem cell niches for biliary tree stem cells) tethered to the outside of large ducts (i.e. in humans: greater than 300 μm) [34]. Their functions are not known but are hypothesized to be “seeds” for organogenesis given that they sprout into ducts with regenerative demands (Reid and associates, unpublished observations). The network of intramural (within the duct walls) niches begins with the Brunner’s glands, submucosal glands found in the duodenum [35, 36] (Cardinale et  al. personal communication). These are located between the major papilla (the papilla of Vater), the entranceway to the hepato‐pancreatic duct, and the minor papilla, the port connecting the duodenum to the dorsal pancreatic duct. The Brunner’s glands are not found elsewhere within the intestinal tract. Indeed, they are used to define the transition from the duodenum to the beginning of the small intestine. The intramural network of stem cells extends throughout the biliary tree in the PBGs found in high numbers behind branch points in the biliary tree, in the cystic duct connecting the gallbladder to the common duct, in the large intrahepatic bile ducts, and in the hepato‐pancreatic common duct [9, 30, 34, 37]. The gallbladder does not have PBGs, but its stem cells are found in crypts at the base of villi in the gallbladder [29]. The second largest numbers of PBGs are found within the large, intrahepatic bile ducts and connect anatomically to the canals of Hering and thence to the sinusoidal plates of parenchymal cells within liver acini [34]. The largest numbers of PBGs are found within the hepato‐pancreatic common duct and connect to the PDGs within the ventral pancreas [27, 34]. Although not yet well defined, they are found also in the accessory duct connecting to the dorsal pancreas. The network of stem cells transition to niches of bipotent, transit amplifying cells, cells with considerable proliferative potential but questionable self‐replicative ability. These comprise hepatoblasts, located next to the canals of Hering [38–41] and to pancreatic ductal progenitors in the PDGs [42–44].



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

525

Figure 42.1  The anatomy of the connections between duodenum, biliary tree, gallbladder, liver, and pancreas in humans. Most of the anatomical features are self‐evident from the figure. The connections of the biliary tree to the duodenum are via two ports, the ampulla of Vater for the hepato‐pancreatic common duct, and the minor duodenal papilla for the accessory duct (also called the duct of Santorini) connecting to the dorsal pancreas.

These, in turn give rise to unipotent committed progenitors that link to mature hepatic or pancreatic cell populations.

LINEAGE‐DEPENDENT EPITHELIAL– MESENCHYMAL CELL PARTNERSHIPS The cells throughout the biliary tree, liver, and pancreas are in epithelial–mesenchymal cell partnerships that undergo coordinate maturation resulting in lineage‐stage‐dependent paracrine signaling influencing the stage‐dependent phenotypic traits as summarized in more detail in prior reviews [1, 45, 46]. In brief, the beginning stages are comprised of epithelial stem cells ­partnered with angioblasts (CD117+, prominin 1 [CD133]+, vascular endothelial cell growth factor [VEGF]‐receptor+, Van Willebrand Factor+, CD31-) [47]. These give rise to cellular descendants that mature coordinately and then split into two branches: epithelia partnered with endothelia (hepatocytes, islets) and epithelia partnered with stellate cells and their descendants, stroma and myofibroblasts (cholangiocytes, acinar cells). The endothelial cell precursors (CD133+, VEGF‐receptor+, Van Willebrand Factor+, CD31+) and the stellate cell precursors (CD146+, intercellular adhesion molecule [ICAM]‐1+, desmin+, alpha‐smooth muscle actin [ASMA]+, glial fibrillary acidic protein [GFAP]-) give rise to descendants with distinct phenotypic traits [1] The lineage‐dependent traits comprise morphology, cell size, DNA content (ploidy), growth potential, antigenic profile, gene expression, and lineage‐dependent paracrine signaling by the synergistic effects of extracellular matrix components, matrix‐bound signals (growth factors and cytokines linked primarily to glycosaminoglycan [GAG] chains bound to core proteins in proteoglycans) and soluble signals present in

blood and interstitial fluid. The net sum of the activities of cells at the sequential maturational lineage stages yields that for the composite tissue. The phenomena associated with stem cell networks in hepatic/pancreatic organogenesis are common to all mammals. However, there are some species‐associated distinctions. Examples include the species‐specific variations with respect to ploidy profiles in the liver (note: these have not been analyzed yet in pancreas). The lineages in all mammals begin in fetal and neonatal tissues with entirely diploid cells that transition to increasing proportions of polyploid cells in adulthood and especially in geriatric hosts. Within the liver acinus, diploid cells are always found periportally; the polyploid cells are always found pericentrally. What varies is the proportion of the cells of the liver plates that are polyploid. In humans, by around 20 years of age, the shift is to a minor fraction of tetraploid cells, all in zone three of the liver acinus; in rats, by four weeks of age, it transitions to predominantly tetraploid cells (80%), all in zone two and with a small fraction (10%) of octoploid cells in zone three; and in mice the shift occurs by three to four weeks of age and results in 97% polyploid cells comprised mostly of tetraploid and octoploid in a part of zone one, all of zone two, and with a minor fraction of cells that are 16N and 32N in zone three. The exceptions to this ploidy profile are the newly discovered diploid parenchymal cells linked on their lateral borders to endothelia encompassing the central vein; these constitute a reservoir of committed (unipotent) hepatocytic progenitors that replace apoptotic (senescing) hepatocytes [48]. Another notable variation is that pigs and oxen do not have an hepato‐pancreatic common duct, the location of the stem cell niches for the ventral pancreas for all other mammals. Ongoing studies are exploring possible locations for these stem cell niches for the ventral pancreas in pigs.

526

THE LIVER:  RADIAL AXIS OF MATURATION

MICROENVIRONMENT OF THE STEM CELL AND PROGENITOR CELL NICHES Stem and progenitor cells reside in discrete locations called niches, with unique microenvironments that are poorly defined for the hepato/biliary/pancreatic network (Table  42.1 and Figure  42.2). The niches include PBGs in the extrahepatic and intrahepatic ­biliary tree [4, 27, 30, 34] and in the cystic duct connecting the gallbladder to the biliary tree [29]; the ductal plates in fetal and neonatal livers and transitioning into the canals of Hering in pediatric and adult livers [49–53]; and the PDGs found throughout the pancreas but with an especial concentration in the head of the pancreas [28, 54–56]. The gallbladder does not have PBGs, but does have crypts at the base of villi and contains stem cells similar to late stage biliary tree stem cells (BTSCs). Collectively, these niches form a network that is continuous throughout the biliary tree and anatomically connects directly through to the canals of Hering to the sinusoidal plates within the liver and through to the PDGs, the reservoirs of committed progenitors within the pancreas, and then to the acinar cells and islets. As noted above, a key paradigm to the tissue organization is the epithelial–mesenchymal cell partnerships. It can be mimicked in vitro by use of feeder cells of the relevant mesenchymal type or by use of defined mixes of matrix components and soluble signals [47]. Matrix and soluble signals in the stem cell and progenitor cell niches have been only partially defined [40, 41, 47, 57, 58]. That which is known is summarized in Table 42.1 and in Figure 42.2. The known matrix chemistry of these stem/progenitor cell niches is dominated by hyaluronans (HA) [59], non‐sulfated glycosaminoglycans (GAGs), and minimally sulfated forms of chondroitin sulfate proteoglycans (CS‐PGs) [60]. In association with the transit amplifying cells, there are also fetal collagens (e.g. type IV), and fetal adhesion molecules (e.g. fetal laminins) along with more sulfated proteoglycans (e.g. heparan sulfate proteoglycans [HS‐PGs] and CS‐PGs) [61, 62]. Receptors for one or another of the known hyaluronan receptors (e.g. CD44) are common features of the stem cells [63]. With maturation to adult cell types, there is a branching to separate lineages (hepatocytes, islets) associated with endothelia versus cholangiocytes and acinar cells associated with stellate cells. The matrix chemistry changes parallel the split: those in epithelial–endothelial relations contain matrix chemistry dominated by network collagens (type IV, type VI), by forms of laminin, and by HS‐PGS with increasing sulfation resulting in the expression of heparin proteoglycans (HP‐PGs) associated with the most mature cells (polyploid hepatocytes, mature islets). The branch of the epithelia‐stellate cells gives rise to cholangiocytes and acinar cells and is associated with a matrix chemistry comprised of fibrillar collagens (e.g. type III, type I), adhesion molecules (forms of fibronectins, entactin), and with CS‐PGs that transition in late stages of cells to highly sulfated forms, dermatan sulfate proteoglycans (DS‐PGs).

PERIBILIARY GLANDS The PBGs occur throughout the biliary tree as intramural glands, found within the bile duct walls, and as extramural glands that are tethered by extensions to the large bile ducts [64]. They occur in highest frequencies at the branching points of the

b­iliary tree, with the greatest numbers being present in the hepato‐pancreatic common duct and the large intrahepatic bile ducts [9, 34]. Other than findings from the pioneering studies of Nakanuma and associates [64–66], almost nothing used to be known of the possible roles of the PBGs until the recent studies within the last decade analyzing them as reservoirs, crypts, containing stem cell populations. Each PBG contains a ring of cells at its perimeter and is replete with mucous production and periodic acid‐Schiff (PAS)‐positive material in its center. The cells in the ring are phenotypically quite homogeneous in the PBGs in some sites (e.g. near the fibromuscular layers of the hepato‐pancreatic common duct and the large intrahepatic bile ducts), but are quite heterogeneous in other sites (e.g. neared to the lumens of all ducts, especially the cystic duct, hilum, common duct). The pattern of the variations implicate maturational lineages for which there are two axes [27, 34].

RADIAL AXIS OF MATURATION A radial axis of maturation is evident within the duct walls (the intramural PBGs), beginning with the cells in PBGs near the fibromuscular layer (center of the duct walls) and ending with cells at the ducts’ lumens (Figure 42.3). The most primitive of the stem cells, those in the PBGs at the fibromuscular layer, have no markers of adult liver or pancreas but instead have moderately high levels of expression of pluripotency genes (e.g. SOX2 [a transcription factor that is essential for maintaining self‐renewal, or pluripotency in embryonic and determined stem cells], SALL4 [Sal‐like protein 4 found to be important for self‐replication of stem cells], BMI-1, NANOG, KLF4/5, octamer‐binding transcription factor 4 [OCT4, also known as POU5F1 {POU domain, class 5, transcription factor 1}, a gene expressed by stem cells]); co‐expression of transcription factors for both liver and pancreas (e.g. SOX17, PDX1 [pancreatic and duodenal homeobox 1, a transcription factor critical for pancreatic development]); express classic markers indicative of active proliferation such as Ki‐67; have high levels of CD44 (hyaluronan receptors); and express sodium iodide symporter (NIS). NIS is hypothesized to transport anionic intermediates needed in the synthesis of hyaluronans [67]. As one progresses toward the lumen, the cells in the PBGs lose expression of the pluripotency traits, diminish the evidence for proliferation, and acquire intermediate markers associated with stemness such as LGR5 and epithelial cell adhesion molecule (EpCAM). At the lumens, all traits of stemness have faded and been replaced with hepatic traits in the vicinity of the cardiac mesenchyme or pancreatic traits for ducts in the vicinity of the retroperitoneal mesenchyme. We hypothesize that a new nomenclature will be required for the PBGs and PDGs, since they have traits that are less those of glands and instead have features of crypts. The phenomenon parallels that observed in the intestinal crypts containing stem cells that mature in a radial axis along the villi to yield adult intestinal cells (e.g. enterocytes and goblet cells). Temporarily we will abstain from converting to a new nomenclature to enable further studies to be done that should help with establishing a logical new nomenclature for the components of the network.

Table 42.1  Representative biomarkers of major lineage stages of precursors in organogenesis of liver and pancreas

Periportal region of liver acini Lineage stage of cells

Canals of Hering

Hepatoblasts, hepatic committed progenitors

HpSCs

Endodermal transcription factors

Sox9

SOX 9, SOX 17

Pluripotency genes

Low to none

Intermediate level of expression

Cell adhesion molecules

EpCAM, NCAM

Hedgehog proteins

EpCAM, ICAM‐1 Weak CXCR4 CD133, LGR5 Weak Indian and Sonic

Matrix proteins

Laminin, type IV collagen

GAGs/PGS

HA, CD44, syndecans (HS‐PGs and CS‐PGs) MDRI‐negative; ABCG2‐moderate Albumin +, AFP +++, P450A7, Glycogen

Other stem cell markers

Multidrug resistance genes Hepatic traits

Pancreatic traits

PBGs in large intrahepatic bile ducts

PBGs in extrahepatic biliary tree

Stage 1–3 BTSCs HpSCs SOX 9, SOX 17, PDX1 in Stages 1 and 2 BTSCs SOX 9 and SOX 17 in Stage 3

CD133, LGR5 Strong Indian and Sonic Laminin 5, type III collagen HA, CD44, minimally sulfated CS‐PGs MDR1 moderate, ABCG2 very strong Albumin +/‐ AFP‐, P450A7‐

PBGs in hepato‐ pancreatic common duct

PDGs in pancreas Pancreatic ductal progenitors

PSCs SOX 9, PDX1 SOX9 and PDX in Stage 3 Intermediate level of expression

Moderate to strong expression in stages 1 and 2 BTSCs; intermediate level of expression in stage 3 BTSCs OCT4, SOX2, NANOG, SALL4, BMi‐1, KLF4/KLF5 NCAM in stage 1 and 2 BTSCs, EpCAM and NCAM in stage 3 BTSCs

None

CXCR4 in stages 1 and 2 BTSCs, CD133 in all of the stages of BTSCS; LGR5 in stages 2 and 3 BTSCs Not yet studied

CXCR4, CD133, CD24

Not yet studied

Islets: network collagens; acinar cells: fibrillar collagens Islets: syndecans and glypicans; acinar cells: CS‐PGs, DS‐PGs None

HA, CD44, others not yet studied Not yet studied

EpCAM, ICAM‐1

Weak Sonic

Negative for hepatic traits

Negative for pancreatic traits

ISL1, PROX1, NeuroD, PAX4

NGN3, MAFA, MUC6, Nkxx6.1/NKx6.2, Ptf1a, Glut2

HA, hyaluronans; HS‐PGs, heparan sulfate proteoglycan; CS‐PGs, chondroitin sulfate proteoglycan; DS‐PGs, dermatan sulfate proteoglycan; syndecans, proteoglycans with a transmembrane core protein; glypicans, proteoglycans linked to plasma membrane by PI linkages; MDR1, multidrug resistance genes; HpSCs, hepatic stem cells; HBs, hepatoblasts; BTSCs, biliary tree stem cells; PSCs, pancreatic stem cells; PBGs, peribiliary glands; PDGs, pancreatic duct glands. All of the known stages of the BTSCs are in the large intrahepatic bile ducts including in the extrahepatic biliary tree and in the hepato‐pancreatic common duct. The precursors to these are in the extramural peribiliary glands and in the Brunner’s glands in the submucosa of the duodenum.

0004435634.INDD 527

11/5/2019 6:46:31 PM

528

THE LIVER:  RADIAL AXIS OF MATURATION A. Stem Cell Niches relevant to Hepatic and Pancreatic Organogenesis Stem Cells = self-replicate; multipotent; all express moderate levels of pluripotency genes; all express one or more of the isoforms of CD44 (hyaluronan receptors) and the most primitive ones found also express NIS (sodium iodide symporter). NIS is hypothesized to be a transporter of anionic intermediates in the hyaluronan biosynthesis pathway. Determined Endodermal Stem Cell subpopulations identified to date: Extramural Peribiliary Gland Stem Cells: extremely primitive. With regenerative demands, produce ductules. Intramural Stem Cell Populations: Brunner’s Glands’ Stem Cells --submucosa of the duodenum; extremely primitive Biliary Tree Stem Cells (BTSCs). Three phenotypically distinct stages within bile duct walls Stages: 1) NIS+, Negative for LGR5 and EpCAM

2) NIS+, LGR5+ EpCAM-

3) NIS-, LGR5+, EpCAM+

Gallbladder stem cells --similar to stage 3-BTSCs Hepatic stem cells --primarily within canals of Hering and in PBGs of biliary tree Pancreatic stem cells --primarily within PBGs of hepato-pancreatic common duct Note: Only committed progenitors found within pancreas proper (so, no true stem cells within pancreas) Epithelial-mesenchymal Cell Partnerships. All are comprised of epithelial stem cells coupled with mesenchymal cell partners; there is lineage-stage-specific paracrine signaling and coordinate maturation. Stem Cell stages= epithelial stem cells partnered with angioblasts Stem Cell Niche Microenvironment = known components comprise hyaluronans, non-sulfated glycosaminoglycans, plus minimally sulfated chondroitin sulfate proteoglycans (no heparan sulfate proteoglycans).

B. Intermediates In Hepatic and Pancreatic Organogenesis Transit Amplifying Cells = highly proliferative but debatable if they self-replicate. Weak to negligible levels of pluripotency genes. Hepatoblasts --bipotent (yield hepatocytes and cholangiocytes); signature feature= alpha-fetoprotein (AFP); located adjacent to canals of Hering; anatomically connected through the hepatic stem cells to network of stem/progenitor cells within peribiliary glands (PBGs) of the large intrahepatic bile ducts. Pancreatic Ductal Progenitor Cells – bipotent (yield acinar cells and islets); found within pancreatic duct glands (PDGs) within the pancreas proper; are anatomically connected to the network of stem cells within PBGs of the hepato-pancreatic common duct of the biliary tree. Committed Progenitor Cells = do not self-replicate; highly proliferative; unipotent; no expression of pluripotency genes. Anatomically linked to the mature cells Epithelial-mesenchymal cell partnerships: Hepatocytic or islet progenitors and descendants coupled to endothelial cell lineage stages Cholangiocytic or acinar progenitors and descendants coupled to stellate cell lineage stages; late lineage stages are associated with myofibroblasts Niche Microenvironment = known components include hyaluronans + minimally sulfated sulfated chondroitin sulfate proteoglycans and heparan sulfate proteoglycans + laminin + type IV collagen

Figure 42.2  The stem/progenitor cell niches in the biliary tree, liver, and pancreas. (a) A chart shows the known connections between the niches throughout the biliary tree and within the organs. (b) Summarizes the features of the stem cell niches. (c) A summary of the transit amplifying cells.

Proximal‐to‐distal axis in maturation A proximal‐to‐distal axis in the maturation occurs beginning at the duodenum with the stem cells in the Brunner’s glands; connects into the stem cells within the PBGs within the hepato‐­ pancreatic common duct; and then transitions into the pancreas to connect with the PDGs. The branch for the liver is via the common duct and once past the cystic duct leading to the gallbladder connects into the PBGs of the large intrahepatic bile ducts. These give rise to links with the canals of Hering that transition into the sinusoidal plates of the liver acini.

The maturational progression of the cells is always within the radial axes in the duct walls, but that in or near the liver is within the domain of the cardiac mesenchyme and results in mature cells of an hepatic fate. The radial axes within ducts near the pancreas are in the domain of the retroperitoneal mesenchyme and result in mature cells of pancreatic fates. Those in between liver and pancreas yield cells with mature bile duct markers. The implications are that isolated BTSCs, especially the most primitive of them, should be able ex vivo to give rise both to liver and pancreas if provided the requisite paracrine signals from either cardiac mesenchyme versus retroperitoneal mesenchyme. This



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

529

C. Hepato/Pancreatic Stem/Progenitor Cell Niches Extramural PBGs Tethered to outside of large Bile Ducts Very primitive Endodermal Stem Cells Common Duct

Brunner’s Glands (Submucosa of the Duodenum) BGSCs, CPs

Intramural PBGs (Extrahepatic Biliary Tree) BTSCs, HpSCs, PSCs, CPs

PBGs (Large Intrahepatic Bile Ducts) BTSCs, HpSCs, HBs, CPs Ductal Plates Canals of Hering (Liver) HpSCs, HBs, CPs

Hepatocytes and Cholangiocytes

Hepato-Pancreatic Common Duct

PBGs (Cystic Duct) BTSCs, CPs

Crypts-bottoms of villi (Gallbladder) Stage 3-BTSCs, CPs

Cholangiocytes

PBG= Peribiliary Gland Extramural PBGs = tethered to duct wall Intramural PBGs = inside the duct wall

PBGs (Hepato-pancreatic Common Duct) BTSCs, PSCs, CPs

Pancreatic Duct Glands (Pancreas) Pancreatic Ductal Progenitors (CPs)

Islets and Acinar Cells BGSCs = Brunner’s glands stem cells BTSCs = biliary tree stem cells (3 stages) HpSCs = hepatic stem cells; HBs = Hepatoblasts PSCs = pancreatic stem cells CPs = committed progenitors (unipotent)

Figure 42.2  (Continued)

(a)

Figure 42.3  The radial axis of maturation in (a) liver, (b) pancreas, and (c) gallbladder. The peribiliary glands (PBGs) near to the fibromuscular layer within the bile duct walls are crypts containing the most primitive of the stem cells, ones that can give rise both to liver and to pancreas. There are hints that the cells move along the walls of the PBGs and progress upwards towards the lumens of the ducts. With that progression, the phenotypic traits transition from those for stem cells to those for mature cells. Some of the traits for the cells in the radial axis maturation are noted. The gallbladder does not have PBGs, but does have late stage BTSCs located at the base of villi in the gallbladder. These move upwards toward the tops of the villi.

is indeed what has been found [27, 30]. It also implicates the potential for using the BTSCs in cell therapies for both liver and for pancreas, an hypothesis that is currently under investigation. The network provides a biological framework for ongoing organogenesis of liver, biliary tree and pancreas throughout life.

These phenomena parallel the well‐described, intestinal maturational lineage system. The radial axis of maturation in the intestine progresses from stem cells in the crypts to fully differentiated cells at the tops of the villi. The proximal‐to‐distal axis follows the length of the intestine and results in distinct

530

THE LIVER:  EX VIVO STUDIES ON THE STEM/PROGENITORS

(b)

(c)

Figure 42.3  (Continued)

mature cells depending on whether the radial axis is located in the esophagus, stomach, duodenum, small or large intestine.

EX VIVO STUDIES ON THE STEM/ PROGENITORS The in situ and anatomical findings summarized above are complemented by ex vivo (monolayer and organoid) cultures of the various lineage stages including those of Brunner’s glands (Cardinale et al. personal communication), of two of the BTSC stages [30], of the hepatic stem cells (HpSCs) and hepatoblasts (HBs) [1, 39, 40, 68], pancreatic ductal progenitors and islets [28], and adult hepatocytes and adult cholangiocytes [47, 69].

The culture conditions required for the different lineage stages are distinct. The stem cell stages require conditions with soluble forms of hyaluronans for organoids and with hyaluronan hydrogel substrata for monolayers or for embedded organoids; these hydrogel must be very soft, under 100 Pa to maintain stemness traits [68]. The medium required for the stem cells is serum‐free, devoid of growth factors and cytokines, and tailored for the cells at this stage. One of the best ones was that established by Hiroshi Kubota [38]. It is comprised of any rich basal medium with low calcium (around 0.3 mM), no copper, selenium (10−10 M), zinc (10−12 M), insulin (around 5 μg mL−1), transferrin/Fe (around 5 μg mL−1), high density lipoprotein (around 10 μg mL−1), and a defined mixture of purified free fatty acids bound to highly purified albumin. Mature cells do not ­survive in Kubota’s medium, only the stem cells from both



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

epithelial and mesenchymal cell lineages. However, more rapid expansion of the mesenchymal partners of the epithelial stem cells, such as the angioblasts, and conditions permissive for self‐ replication of both are facilitated by the addition of leukemia inhibitory factor (LIF) and by VEGF [30, 38, 41, 47, 69]. Thus, the conditions co‐select for any of the true stem cell subpopulations and their mesenchymal cell partners, angioblasts and their descendants, precursors of stellate cells, or endothelia [38, 47] Under these conditions, we observed two major types of BTSC colonies in cultures that correlate with stage two and stage three BTSCs from the in situ studies; we have yet to identify conditions for the stage one BTSCs. The stage two category consist of cells that undulate (“dancing cells”), are very motile, and initially do not express EpCAM (CD326) but acquire it at the edges (the perimeters) of the colonies, corresponding to slight cellular differentiation [30]. These are precursors to stage three BTSCs that show uniform expression of EpCAM from the outset and display a carpet‐like appearance with cells of uniform morphology [30]. The stage three BTSCs found throughout the biliary tree are also the same or similar to the stem cell subpopulation found in the gallbladder [29]. The expansion potential of the cells in culture in Kubota’s medium is considerable: two to three cells can grow to colonies of more than 500 000 cells in around eight weeks [30]. The cells retain a stable stem cell phenotype (i.e. self‐renew) throughout months of culture, and may be subcultured (“passaged”). Initially cells show a typical division time of about one to two days, but within a week, they slow to a division every two to three days. At eight weeks the colonies contain cells in the centers that are morphologically uniform, small (7–9 μm) and express high levels of stem cell markers. Cells at the edges of the large colonies are slightly larger (around 10–12 μm), have weak expression of EpCAM and expression of markers intermediate in the differentiation pathways, indicating potential loss of stemness and transition to more mature progenitors. Using three‐dimensional (3D) hyaluronan hydrogels containing appropriate signaling molecules, the BTSCs can be induced to differentiate to hepatocytes, cholangiocytes, or pancreatic neo‐islets [30]. We have not succeeded yet to lineage restrict the BTSCs to acinar cells. The differentiation is achieved by embedding the BTSCs in specific mixes of extracellular matrix ­components (hyaluronans and type I collagen for bile ducts; hyaluronans and type IV collagen and laminin for hepatocytes or islets) and providing a serum‐free, hormonally defined medium (HDM) tailored for each specific transit amplifying cells or mature cell type. The HDM are prepared by supplementing Kubota’s medium with copper (10−12 M), higher calcium (0.6 mM), bFGF (basic fibroblast growth factor, 10 ng mL−1) and then adding a unique set of hormones and growth factors for hepatoblasts (EGF, hepatocyte growth factor [HGF], T3, soluble [or substrata] forms of type IV collagen and laminin), hepatocytes (substrata of type IV collagen and laminin plus glucagon, galactose, T3, oncostatin M, HGF, EGF, glucocorticoids), cholangiocytes (substrata of type I and III collagen + HGF, EGF, VEGF, glucocorticoids), versus for pancreatic islets (soluble [or substrata] forms of type IV collagen and laminin) plus B27, ascorbic acid, cyclopamine, retinoic acid, HGF, and, after four days, replacement of bFGF with Exendin‐4] [30].

531

The gene expression profiles of cells in 3D hydrogels complement the morphological observations. For example, cells cultured under conditions for hepatocytes produce albumin, transferrin, and P450s. Cells in conditions for cholangiocytes express anion exchanger 2 (AE2), cystic fibrosis transmembrane conductancy regulator (CFTR), gamma glutamyl transpeptidase (GGT), and secretin receptor. Cells in conditions for pancreatic islets express transcription factor PDX1 and the hormones glucagon, somatostatin, and insulin. Specific staining for human C‐peptide confirms de novo synthesis of proinsulin, and its secretion can be regulated in response to the level of glucose. In vivo studies in which the cultured cells are transplanted provide further evidence for the multipotency of the BTSCs for hepatic, biliary tree, and pancreatic fates [4, 27, 30, 70]. To confirm endocrine pancreatic differentiation, pre‐ induced neo‐islet structures were implanted into mouse fat pads, and the animals were treated with a toxin (streptozotocin) at a dose sufficient to destroy their own pancreatic beta cells, but not human beta cells. Those mice transplanted with the human neo‐islets showed significant resistance to hyperglycemia compared to controls that did not receive cell ­therapy. The presence of functional beta‐like cells derived from the biliary tree stem cells produced serum‐levels of human C‐peptide, that was regulated appropriately in response to a glucose challenge [30]. Further studies have confirmed and expanded upon these initial findings, leading us to conclude that the hepato‐pancreatic common duct is the major reservoir of stem cells giving rise to committed progenitors found in pancreatic duct glands, and thence to pancreatic islets throughout life [27].

HEPATIC STEM CELLS Those familiar with the myth of Prometheus will recall that the liver possesses a remarkable capacity for regeneration [2, 3, 71]. Yet liver diseases, potentially leading to organ failure due to hepatitis viruses, alcohol consumption, diet and metabolic disorders, and other causes, constitute a major medical burden [72–74]. Cell‐based therapies and tissue engineering represent possible approaches to address these needs [1, 75–78]. Sourcing of cells for such applications is a significant challenge. In some countries it is possible to obtain fetal tissues. In others neonatal or adult tissues can be used. Given the newly discovered source of stem cell populations in the biliary tree and precursors for both liver and pancreas, the need for liver or pancreatic tissue as a source is reduced. There is likely to be a transition to biliary tree tissue as a primary source for clinical programs, and it is considerably easier to obtain. The role(s) of hepatic stem cells in the normal maintenance of the liver and in regeneration from various insults remains a subject of active research and debate [52, 53, 71, 78–85]. Although there is now general acceptance that hepatic stem cells exist postnatally, their relevance is compared and debated with that of the plasticity of the postnatal parenchymal cells [86]. Further discussion of this is given below.

532

THE LIVER:  HEPATIC STEM CELL ISOLATION AND EXPANSION

HEPATIC STEM CELL ISOLATION AND EXPANSION Information on the findings on HpSCs is important, given that human hepatic stem cells (hHpSCs) have already been used in clinical trials in India and in Italy for patients with diverse liver diseases and conditions [70, 87–89]. The isolation of hHpSCs from fetal, neonatal, and adult human livers was achieved by immunoselection with a monoclonal antibody for the surface marker EpCAM [41]. These cells constitute approximately one percent (0.5–1.5%) of the total liver population from early childhood onwards. Unlike mature hepatocytes, they survive extended periods of ischemia, allowing collection even several days after cardiac arrest [90]. The hHpSCs express additional surface markers often found on stem/progenitor cells, such as CD133 (prominin), CD56 (neural cell adhesion molecule, NCAM), and CD44 (the hyaluronan receptor); they also express characteristic endodermal transcription factors SOX9, SOX17, and HES1. They are small (diameter 7–9 μm, which is less than half that of mature parenchymal cells) and express weak or negligible levels of adult liver‐specific functions such as albumin, cytochrome P450s, and transferrin. The hHpSCs display far greater capacity to proliferate in culture than hepatocytes or cholangiocytes, and can continue to expand for months with a doubling time of 36–40 hours. The colonies that form look remarkably similar to those of embryonic stem (ES) cells or induced pluripotent stem (iPS) cells [5, 1, 47, 91]. The hHpSCs serve as immediate precursors of human hepatoblasts (hHBs). The hHBs are readily distinguished from hHpSCs by the expression of AFP and intercellular adhesion molecule‐1 (ICAM‐1), for which the hHpSCs are entirely negative [41, 47, 51]. The hHBs, in turn, are precursors of committed unipotent progenitors for hepatocytes and cholangiocytes. When injected into livers of immune‐deficient mice, the ­hHpSCs give rise to cells expressing characteristic human liver and bile duct proteins, especially after the host’s liver has been damaged by treatment with carbon tetrachloride [41]. Whereas there has been limited success to achieve ex vivo expansion of hematopoietic stem cells, the hHpSCs proliferate for a sustained period in Kubota’s medium [38] which, as stated previously, contains no additional growth factor or cytokine. Pathways important for hHpSC survival in vivo, such as Hedgehog (Hh) signaling [40], are activated through autocrine loops. The expanded hHpSCs maintain a stable marker phenotype and also express the enzyme telomerase. The telomerase mRNA and the protein encoded are localized to the nucleus of the hHpSCs and the hHBs and telomeric enzymatic activity correlate well with both the mRNA and protein levels [92]. However, later lineage stages (committed progenitors to late ­lineage stage mature cells) have no evidence of synthesis of telomerase but have large amounts of telomerase protein localized cytoplasmically. Telomeric enzymatic activity does not correlate with total telomerase protein levels; we hypothesize that it correlates with the nuclear levels of the protein. Therefore, we further hypothesize that regenerative demands will result in small amounts of the cytoplasmic reserves of telomerase relocating to the nucleus. If we are correct, the enzymatic activity levels should correlate with the amount of telomerase (protein) in the nucleus [92].

More recently, we have observed that Sal‐like protein 4 (SALL4, found to be important for self‐replication of stem cells) is strongly expressed in the hBTSCs and hHpSCs but not in committed progenitors of either liver or pancreas [93]. SALL4 is a member of a family of zinc finger transcription factors and a regulator of embryogenesis, organogenesis, and pluripotency. It can elicit reprogramming of somatic cells and is a marker of stem cells. A crucial prerequisite for successful expansion of hHpSCs is to mimic an appropriate microenvironment. When selected for growth in vitro on tissue culture plastic and in Kubota’s medium, the hHpSCs grow as colonies with feeders of angioblasts (CD117+, VEGF‐receptor+, CD133+, Von Willebrand Factor+) [41, 47, 58]; the feeders can be replaced with soft forms of hyaluronans (under 100 Pa) and type III collagen [47, 68, 91, 94]. The cells expand for months under these conditions. By contrast, hHBs survive for only about a week under these same conditions, but they can survive if they are co‐cultured with stellate cell precursors (CD146+, alpha‐smooth muscle actin+, desmin+, VCAM+, ICAM‐1+, GFAP‐negative) or feeders of mesenchymal stem cells (MSCs). The stellate [95] feeder cells (or feeders of MSCs) can be replaced with hyaluronans, type IV collagen, and/or laminin [47, 96, 97]. The medium and matrix conditions described above allow for flow cytometrically‐purified hHpSCs or hHBs to survive and proliferate in culture and without the need for feeders. Both type III collagen and hyaluronans are constituents of the normal liver stem cell niche [47, 57, 91]. A systematic study of hHpSC behavior in 3D cultures using hyaluronan hydrogels of differing stiffness indicated that rigidity of the microenvironment is an important parameter in regulating maintenance of stemness versus differentiation to more restricted progenitors [68]. This was studied previously in differentiation of progenitors for bone and other hard tissues, but this report is the first for internal organs such as the liver. The hHpSCs, like human ES cells, grow in tight colonies. Dissociating either type of stem cells has proven to be an important practical problem for their efficient expansion ex vivo and for cryopreservation [98]. When treated enzymatically to generate a single cell suspension, both of these stem cell types undergo a high level of cell death. Ding’s laboratory screened for chemicals that would enable ES cells to survive enzymatic dissociation and remain pluripotent. They identified two compounds, a 2,4‐disubstituted thiazole (Thiazovivin) and a 2,4‐disubstituted pyrimidine (Tyrintegin), that met these criteria [99]. They found that Thiazovivin inhibits the Rho‐associated kinase (ROCK), a key component of the pathway that controls cytoskeleton remodeling, and a likely regulator of cell–ECM and cell–cell interactions. Tyrintegin enhances attachment of dissociated ES cells to ECM and stabilizes E‐cadherin. The investigators concluded that ES cell interactions in the normal niche generate signals essential to survival, and that small molecules modulating those signals can maintain viability of dissociated cells. Interestingly, we have observed that hyaluronans, a normal component of stem cell niches, can protect hHpSCs during dissociation and cryopreservation [98]. Thus, hyaluronans, a natural molecule, can mediate the needed protection of the cells as well as the artificial ones described above. The addition of hyaluronans was found to protect cell adhesion mechanisms including the hyaluronan receptor, E‐cadherin, and certain integrins, markers shared by many other stem cell subpopulations [100].



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

THE NEED FOR GRAFTING STRATEGIES IN TRANSPLANTATION OF CELLS FROM SOLID ORGANS Cell therapies attempted to date are usually by delivery of the cells by a vascular route. This is logical for hemopoietic cell types. Long‐lived hemopoietic cells have evolved to be able to flip between splice forms of matrix molecules (e.g. fibronectins) from ones with no cell binding domains (resulting in cells floating freely in blood or interstitial fluid) versus ones with cell binding domains (resulting in attachment  –  a process referred to as “homing”). By contrast, transplantation of cells from solid organs involves cell types in which their attachment proteins always have cell binding domains and so they rapidly (within seconds) aggregate. When delivered to a target organ/ tissue by a vascular route, the aggregates cause an embolus essential for engraftment but if too large resulting potentially in lethal consequences for the host. Moreover, even when successful, there is inefficient engraftment (typically around 10–20%) and with the remainder of the cells either dying or dispersing to ectopic sites [94]. Our studies and those conducted by many others have found that infusion of mature hepatocytes achieves only around 20% engraftment if injected into the portal vein of the liver [73, 101, 102]. Stem cells are even more challenging, with approximately only 3% of the cells engrafting if administered via the portal vein (or via the spleen that connects directly to the portal vein). This can be improved to around 20–25% engraftment in the liver if the hHpSCs are injected into the hepatic artery [89]. The remaining majority of the hHpSCs either die or engraft in ectopic sites, most commonly the lung. Cells that lodge in the vascular beds of ectopic sites can survive for months [100], a finding of unknown significance at this time, but of potential clinical concerns. We have devised grafting strategies for transplantation of hHpSCs embedded into a mix of soluble signals and extracellular matrix biomaterials (e.g. hyaluronans) found in stem cell niches [100]. The hHpSCs maintain a stable stem cell phenotype under the graft conditions. The grafts were transplanted by injection grafting into the livers of immuno‐compromised murine hosts, with and without carbon tetrachloride treatment, to assess the effects of quiescent versus injured liver conditions. Grafted cells remained localized to the livers, resulting in a larger bolus of engrafted cells in the host livers under quiescent conditions and demonstrated more rapid expansion upon liver injury. We therefore have proposed grafting as a preferred strategy for cell therapies for solid organs such as liver [94, 100]. Ongoing studies are resulting in assessment of other forms of grafting (e.g. patch grafting) that enable transplantation of large numbers of cells.

533

microenvironments. Cytokines and other soluble factors necessary for liver development and for the maintenance of differentiated hepatocytes have been known for some time [46, 103, 104]. However, the specific and efficiently directed differentiation of stem or progenitor cells to fully mature hepatocytes and cholangiocytes ex vivo has remained a difficult challenge. This, in fact, is a general problem in much of stem cell biology, whether starting with lineage‐restricted adult stem cells or pluripotent ES and iPS cells. The most promising strategies are to make use of complex extracellular matrix scaffolds, effectively solid‐state signaling apparatuses that can guide the differentiation of the cells.

EXTRACELLULAR MATRIX SCAFFOLDS In recent years, complex matrix scaffolds are being utilized for optimal differentiation of cells [5–8]. However, all of the scaffold types reported are limited and inefficient in their effects due to their methods for isolation, ones resulting in the loss of critical matrix components such as the proteoglycans. The only known method by which to isolate a matrix scaffold with retention of these critical factors is one developed by the Reid lab in partnership with collagen chemists [105]. They designed a method tailored to the known solubility constants of given collagens using the strategy to isolate a matrix complex with a salt buffer at a concentration to keep insoluble all of the known types of collagens in a given tissue. The insoluble complex so isolated was termed “biomatrices” [106]. Frozen sections or pulverized liver biomatrices used as cell culture substrata enabled the long‐term survival of highly functional hepatocytes, far beyond what could be achieved on plastic or with simple type I collagen gels. Recently, we have revisited the method and established an improved protocol, one involving perfusion strategies and an improved delipidation method along with the high salt  strategies, to prepare decellularized organs/tissues called “­biomatrix scaffolds” [5]. They are tissue‐specific but minimally (if at all) species‐specific, and they potently induce cell differentiation [5]. The biomatrix scaffolds contain more than 98% of the collagens and known collagen‐bound matrix components, including most of the fibronectins, laminins, nidogen, entactin, elastin, and so on, and essentially all the proteoglycans (PGs). They retain physiological levels of the known cytokines and growth factors found in the tissue. Mature parenchymal cells plated on biomatrix scaffolds in a serum‐free HDM remained stable for many weeks and continued to express liver‐ specific functions equivalent to that achieved by freshly isolated cells [107, 108].

LIVER REGENERATION DIFFERENTIATION The pharmacology of stem cell differentiation also must encompass both soluble signals (i.e. conventional biologics and/or drugs) and matrix components corresponding to the cells’ 3D

The renowned regenerative capacity of the liver has inspired countless studies on mechanisms associated with the process [71]. We will not summarize that considerable literature but refer the readers to some recent reviews [3, 109, 110]. Here we will note only the known responses of the stem cells and progenitors in two distinct forms of liver regeneration: that after

534

THE LIVER:  PLASTICITY ISSUES AND REGENERATIVE PHENOMENA

partial hepatectomy, and that after selective loss of cells in acinar zone three (the pericentral zone). (We assume that a parallel process occurs in pancreatic regeneration, though it has been studied in far less detail.) A key to understanding the responses is recognition of “feedback loop signals,” factors produced by the most mature liver cells, those in zone three of the liver acinus, and secreted into the bile. The bile flows from pericentral zone to periportal zone and then into the biliary tree and finally into the gut. The signaling molecules include bile acids and salts that affect differentiation [111]; acetylcholinesterase [112], which is produced by mature hepatocytes and serves to inactivate acetylcholine produced by periportal cells [113, 114]; and heparins, which are produced by mature hepatocytes [115] (J. Esko, A. Cadwallader, and L. Reid, unpublished observations) and are relevant in ­control of stem cells and of tissue‐specific gene expression [116, 117]. In addition the flow of the bile mechanically affects ­primary cilia on periportal cells and thereby influences signal transduction processes mediated by these organelles [118–120]. In the presence of feedback signals, the stem cells remain in a quiescent state. Diminution or loss of these signals results in dis‐inhibition of the stem/progenitor cell compartments. This leads to hyperplasia of the stem cells and other early lineage stage cells. Factors that may release the stem cell compartment from the normal feedback signaling control loops include viruses, toxins, or radiation that selectively kill cells in zone three, the pericentral zone of the acinus. The hyperplasia transitions into differentiation of the cells. The resulting fully mature cells produce bile, and the restoration or enhancement of feedback loop signals then inactivates the proliferative response. Regeneration of the liver after partial hepatectomy is distinct from that described above [71, 110, 121]. The tissue remaining after surgical removal of a portion of the liver (e.g. two‐thirds of its mass) continues to have feedback loop signals, and the early lineage stage cells remain competent to respond to these signals. The depletion below threshold levels of various liver functions and secreted products triggers DNA synthesis as a wave across the liver plates [110]. However, the DNA synthesis in most of the cells of the liver (especially those in zones two and three) is not accompanied by cytokinesis [122]. So these cells increase their level of ploidy and demonstrate hypertrophic growth [123]. The polyploidy triggers an increased rate of apoptosis resulting in turnover of the liver. With the loss of the apoptotic cells, there is a low level of proliferation of the stem cells and early lineage stage cells to replace those cells eliminated during apoptotic processes. In mammalian species examined, this turnover occurs in weeks.

PLASTICITY ISSUES AND REGENERATIVE PHENOMENA Kopp et al. [86] and others [3, 124] have hypothesized that plasticity is dominant or even the sole mechanism mediating regenerative responses for liver and pancreas. It is an hypothesis emanating from discoveries that somatic cells can be reprogrammed by artificial means to dedifferentiate or to transdifferentiate to other cell types by transfection of cells with multiple transcription factors such as those identified by Takahashi and Yamanaka [125].

Although cellular reprogramming is achievable under conditions ex vivo or in extreme artificial conditions in vivo such as hosts with suicidal transgenes, it has yet to be demonstrated in transplanted adult cells in common diseases. By contrast, there is evidence that stem/progenitors give rise to maturational lineages of cells supporting tissue regeneration [2, 32]. Stem cell functions of BTSCs and HpSCs are manifested both by their genetic signatures and by their clonogenic, self‐ replicative capacity ex vivo for months when under specific serum‐free, wholly defined conditions. In addition, these cell populations are multipotent, the other trait required for proof of stemness, as revealed by their ability to lineage restrict in vitro or in vivo to hepatocytes, cholangiocytes, or islets depending on their microenvironment before or after culture expansion [30, 41]. This is consistent with the data of Dorrell et al. [126] who showed phenotypic and functional similarity of organoid‐initiating cells in mouse pancreas and liver. Another issue contributing to misunderstandings is that liver regenerative phenomena in murine versus human tissues can be distinct [2, 32]. In long‐lasting chronic human liver diseases, a severe and progressive impairment of hepatocyte proliferative capabilities is common. Indeed, specific insults exhaust hepatocyte proliferation, induce cellular senescence, and/or arrest the hepatocyte cell cycle [32]. In human pathologic tissue, HpSCs activate to produce the so‐called ductular reactions that give rise to nascent hepatocytes repopulating cirrhotic livers through formation of hepatocyte buds [2, 32]. By contrast, murine models of liver injury typically do not result in a severe blockade of hepatocyte proliferation [2]. In a novel mouse model [32] in which apoptosis, necrosis, and senescence are induced in nearly all hepatocytes, HpSC activation proved crucial for survival and functional liver reconstitution. Evidence promoting plasticity in the liver is based on studies in a single publication by Tarlow et al. [128] These studies made use of a highly artificial model of liver regeneration: a murine model of Type I tyrosinemia, caused by a shortage of the enzyme fumarylacetoacetate hydrolase (FAH). Moreover, Tarlow et al. [128] subjected the FAH mice to a second insult: the administration of DDC (2′‐3′‐dideoxycytidine). DDC causes alterations of the biliary tree mimicking the spectra of primary sclerosing cholangitis; the consequent secondary cholestasis heavily impacts the hepatic lineages, leading to the hyperactivation of ductular reactions shown in Tarlow et  al. [128]. The findings in such an extreme model should not be used to make generic statements about stem/ progenitors under either normal or ­typical disease conditions. Additional evidence used to promote the concept of plasticity is based on experimental findings that cholangiocarcinomas can originate from dedifferentiated hepatocytes. This provocative assumption is based on observations by genetic lineage tracing studies that cholangiocarcinomas can arise from cells expressing albumin or transthyretin [129], markers erroneously ascribed only to mature hepatocytes. Albumin and transthyretin are expressed also in HpSCs and HBs subpopulations [30]. Similarly, low levels of insulin are expressed in BTSC subpopulations in the hepato‐pancreatic common duct and in multipotent progenitors within the PDGs in the pancreas [32, 130]. Therefore, claims that new beta cells derive exclusively from pre‐existing beta cells and based on insulin expression [24] ignore the fact that insulin lineage tracing also labels stem/progenitors.



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

The claim that mature liver cells are capable of extensive, complete cell division is also not true except when these cells are transplanted into livers with the FAH mutation [131] or ones expressing suicide transgenes [132], all of them having highly artificial microenvironments that, theoretically, could be resulting in reprogramming of the cells. Transplantation of mature cells into the livers of normal animals results in ­negligible cell division. Transplantation into hosts with classic disease or regeneration conditions (e.g. carbon tetrachloride or partial hepatectomy) results in transient, moderate amounts of cell division. These findings in experimental systems are in line with those from clinical trials of mature hepatocyte transplantation into patients with diverse liver dysfunctions or of islets into patients with diabetes [32]. Transplantation of adult hepatocytes provides effects lasting only a few months. Transplanted islets tend to be insufficient to maintain long‐term glucose homeostasis and exhibit limited proliferation in vivo. By contrast, early findings of stem cell therapies for liver diseases indicate that transplantation with BTSCs or HpSCs into the livers of patients with diverse liver dysfunctions results in long‐term effects over years with a steady improvement in liver functions [31]. In summary, reports from clinical trials using stem cells ­versus mature cells for cell therapies offer the most substantive and incontrovertible evidence for rejection of the claim that plasticity alone mediates regenerative responses in liver and pancreas. We prefer the assumption that mechanisms for stem/ progenitors and their maturational lineages, along with minor contributions from epigenetic phenomena, contribute to tissue turnover and repair.

535

Professor Alvaro was supported by FIRB grant #RBAP10Z7FS_ 004 and by PRIN grant #2009X84L84_002. The study was also supported by Consorzio Interuniversitario Trapianti d’Organo, Rome, Italy. Cornell University (Ithaca, NY). Dr. Sethupathy is funded in part by a UNC genetics and molecular biology curriculum T32 training grant (T32‐GM‐007092–41); a grant (SRA‐60486) from Vesta Therapeutics (awarded to L.M.R); a grant (A11–0552) from Vesta Therapeutics (awarded to P.S); and a grant (A16–0311) from the Fibrolamellar Cancer Foundation (awarded to P.S.).

Intellectual property Findings from the studies summarized in this review have been included in patent applications belonging to one or more institutions including the following: University of North Carolina (UNC) at Chapel Hill, Sapienza University in Rome, Italy, and the Diabetes Research Institute (DRI) of the  University of Miami, Florida. The IP has been licensed to Vesta Therapeutics (Bethesda, MD) for clinical uses and to  PhoenixSongs Biologicals (PSB, Branford, CT) for non‐clinical, commercial uses. None of the authors have equity or a position in Vesta, and none are paid consultants to the company. By contrast, LMR is one of the founders, is the ­scientific director, and does hold an equity position in PSB; to  date she has received no salary or consulting fees for these efforts. Other than LMR’s connections with PSB, the authors declare no conflicts of interest. This review is derivative of  and updated from a book chapter in a book on stem cells and edited by Stewart Sell [45].

ACKNOWLEDGMENTS

REFERENCES

Financial support

  1. Turner, R., Lozoya, O., Wang, Y.F. et al. Hepatic stem cells and maturational liver lineage biology. Hepatology, 2011;53:1035–45.   2. Miyajima, A., Tanaka, M., and Itoh, T. Stem/progenitor cells in liver development, homeostasis, regeneration, and reprogramming. Cell Stem Cell, 2014;14(5):561–74.  3. Itoh, T. Stem/progenitor cells in liver regeneration. Review. Hepatology. 2016;64(2):663–8.  4. Carpino, G., Renzi, A., Franchitto, A. et  al. Stem/progenitor cell niches involved in hepatic and biliary regeneration. Review. Stem Cell International, 2017:3658013.   5. Wang, Y., Cui, C., Miguez, P. et al. Lineage restriction of hepatic stem cells to mature fates is made efficient by tissue‐specific biomatrix scaffolds. Hepatology, 2011;53(1):293–305.   6. Badylak, S.F., Taylor, D., and Uygun, K. Whole‐organ tissue engineering: decellularization and recellularization of three‐dimensional matrix scaffolds. Review. Ann Rev Biomed Eng, 2011;13:27–53.   7. Baptista, P.M., Siddiqui, M.M., Lozier, G. et  al. The use of whole organ decellularization for the generation of a vascularized liver organoid. Hepatology, 2011;53(2):604–17.   8. Uygun, B., Soto‐Gutierrez, A., Yagi, H. et al. Organ re‐engineering through development of a transplantatble recellularized liver graft using decellularized liver matrix. Nat Med, 2010;16(7):814–20.   9. Cardinale, V., Wang, Y., Gaudio, E. et  al. The biliary tree: a reservoir of multipotent stem cells. Nat Rev Gastroenterol Hepatol, 2012;9:231–40. 10. Zhang, D., Jiang, W., Liu, M. et al. Highly efficient differentiation of human ES cells and iPS cells into mature pancreatic insulin‐producing cells. Cell Res, 2009;19(4):429–38. 11. Polo, J.M., Anderssen, E., Walsh, R.M. et al. A molecular roadmap of reprogramming somatic cells into iPS cells. Cell, 2012;151(7):1617–32.

UNC School of Medicine (Chapel Hill, NC). Funding derived from Vesta Therapeutics (Bethesda, MD), a wholly owned subsidiary of Toucan Capital Investments, and from the ­ Fibrolamellar Carcinoma Foundation (Greenwich, CT). Additional support was provided through discounted rates for core services via federal funding of the cores: a microscopy services laboratory in pathology and laboratory medicine core facility grant (NIH P30DK34987), core director Victoria Madden, PhD; a histology core funded by the Center for Gastrointestinal and Biliary Disease Biology (CGIBD) via an NIDDK Grant (DK34987); the Lineberger Cancer Center grant (NCI grant #CA016086); the Carolina Center for Genome Sciences (Katherine Hoadley, director), and the UNC Center for Bioinformatics (Hemant Kelkar, director). Diabetes Research Institute (Miami, FL). Studies were funded by grants from NIH, the Juvenile Diabetes Research Foundation, ADA, and the Diabetes Research Institute Foundation. Sapienza University Medical Center (Rome, Italy). Professor Gaudio was supported by research project grant from the University “Sapienza” of Rome and FIRB grant #RBAP10Z7FS_001 and by  PRIN grant #2009X84L84_001.

536

THE LIVER:  REFERENCES

12. Wang, Y., Qin, J., Wang, S. et al. Conversion of human gastric epithelial cells to multipotent endodermal progenitors using defined small molecules. Cell Stem Cell, 2016;19(4):449–61. 13. Rezania, A., Bruin, J.E., Riedel, M.J. et al. Maturation of human embryonic stem cell‐derived pancreatic progenitors into functional islets capable of treating pre‐existing diabetes in mice. Diabetes, 2012;61:2016–29. 14. Zorn, A.M. and Wells, J.M. Molecular basis of vertebrate endoderm development. Int Rev Cytol, 2007;259:49–111. 15. McLin V.A., Zorn, A.M. Organogenesis: making pancreas from liver. Curr Biol, 2003;13(3):R96–8. 16. Sinner, D., Rankin, S., Lee, M., and Zorn, A.M. SOX 17 and beta‐catenin cooperate to regulate the transcription of endodermal genes. Development, 2004;131(13):3069–80. 17. Zaret, K. Developmental competence of the gut endoderm: genetic potentiation by GATA and HNF3/fork head proteins. Dev Biol, 1999;209(1):1–10. 18. Wandzioch, E. and Zaret, K.S. Dynamic signaling network for the specification of embryonic pancreas and liver progenitors. Science, 2009;324(5935): 1707–10. 19. Tremblay, K.D. and Zaret, K.S. Distinct populations of endoderm cells converge to generate the embryonic liver bud and ventral foregut tissues. Dev Biol, 2005;280(1):87–99. 20. Barker, N., van de Wetering, M., and Clevers, H. The intestinal stem cell. Genes Dev, 2008;22:1856–64. 21. Barker, N., Huch, M., Kujala, P. et al. Lgr5(+ve) stem cells drive self‐renewal in the stomach and build long‐lived gastric units in vitro. Cell Stem Cell, 2010;6:25–36. 22. Lange, A.W., Keiser, A.R., Wells, J.M., Zorn, A.M., and Whitsett, J.A. SOX17 promotes cell cycle progression and inhibits TGF‐beta/Smad3 signaling to initiate progenitor cell behavior in the respiratory epithelium. PLoS One, 2009;4(5):e5711. 23. Snyder, J.C., Teisanu, R.M., and Stripp, B.R. Endogenous lung stem cells and contribution to disease. J Pathol, 2009;217:254–64. 24. Dor, Y., Brown, J., Martinez, O.I., and Melton, D.A. Adult pancreatic beta‐ cells are formed by self‐duplication rather than stem‐cell differentiation. Nature, 2004;429(6987):41–6. 25. Houbracken, I. and Bouwens, L. The quest for tissue stem cells in the pancreas and other organs, and their application in beta‐cell replacement. Rev Diabet Stud, 2010;7:112–23. 26. Xu, X., D’Hoker, J., Stange, G. et  al. Beta cells can be generated from endogenous progenitors in injured adult mouse pancreas. Cell, 2008; 132(2):197–207. 27. Wang, Y., Lanzoni, G., Carpino, G. Biliary tree stem cells, precursors to pancreatic committed progenitors: evidence for life‐long pancreatic organogenesis. Stem Cells, 2013;31(9):1966–79. 28. Carpino, G., Renzi, A., Cardinale, V. et  al. Progenitor cell niches in the human pancreatic duct system and associated pancreatic duct glands: an anatomical and immunophenotyping study. J Anat, 2016;228(3):474–86. 29. Carpino, G., Cardinale, V., Gentile, R. et al. Evidence for multipotent endodermal stem/progenitor cell populations in human gallbladder. J Hepatology, 2014;60(6):1194–2020. 30. Cardinale, V., Wang, Y., Carpino, G. Multipotent stem cells in the extrahepatic biliary tree give rise to hepatocytes, bile ducts and pancreatic islets. Hepatology, 2011;54(6):2159–72. 31. Lanzoni, G., Cui, C., Oikawa, T. et al. Clinical programs of stem cell therapies for liver and pancreas. Stem Cells, 2013;31(10):2047–60. 32. Lanzoni, G., Cardinale, V., Carpino, G. The hepatic, biliary and pancreatic network of stem/progenitor cells niches in humans: a new reference frame for disease and regeneration. Hepatology, 2016;64(1):277–86. 33. Alvaro, D. and Gaudio, E. Liver capsule. Biliary tree stem cell subpopulations. Hepatology, 2016;64(2)644. 34. Carpino, G., Cardinale, V., Onori, P. et al. Biliary tree stem/progenitor cells in glands of extrahepatic and intraheptic bile ducts: an anatomical in situ study yielding evidence of maturational lineages. J Anat, 2012;220(2): 186–99. 35. Leeson, T.S. and Leeson, C.R. The fine structure of Brunner’s glands in man. J Anat, 1968;103(2):263–76. 36. Krause, W.J. Brunner’s glands: a structural, histochemical and pathological profile. Prog Histochem Cytochem, 2000;35(4):259–367. 37. Semeraro, R., Carpino, G., Cardinale, V. et al. Multipotent stem/progenitor cells in the human foetal biliary tree. J Hepatol, 2012;220(2):186–99.

38. Kubota, H. and Reid, L.M. Clonogenic hepatoblasts, common precursors for hepatocytic and biliary lineages, are lacking classical major histocompatibility complex class I antigens. Proc Natl Acad Sci (USA), 2000;97(22): 12132–7. 39. Schmelzer, E., Wauthier, E., and Reid, L.M. Phenotypes of pluripotent human hepatic progenitors. Stem Cell, 2006;24(8):1852–8. 40. Sicklick, J.K., Li, Y.X., Melhem, A. et  al. Hedgehog signaling maintains resident hepatic progenitors throughout life. Am J Physiol Gastrointest Liver Physiol, 2006;290(5):G859–70. 41. Schmelzer, E., Zhang, L., Bruce, A. et al. Human hepatic stem cells from fetal and postnatal donors. J Exp Med, 2007;204(8):1973–87. 42. Bonner‐Weir, S. and Sharma, A. Pancreatic stem cells. J Pathol, 2002; 197(4):519–26. 43. Lysy, P.A., Weir, G.C., and Bonner‐Weir, S. Concise review: pancreas regeneration: recent advances and perspectives. Stem Cells Translat Med, 2012; 1(2):150–9. 44. Jiang, W., Sui, X., Zhang, D. et al. CD24: a novel surface marker for PDX1‐ positive pancreatic progenitors derived from human embryonic stem cells. Stem Cells, 2011;29(4):609–17. 45. Furth, M.E., Wang, Y., Cardinale, V. et al. Stem cell populations giving rise to liver, biliary tree and pancreas, in, The Stem Cells Handbook, 2nd edn, (ed. S. Sell), Springer Science Publishers, New York, 2013, pp. 75–126. 46. Macdonald, J.M., Xu, A., Kubota, H. et  al. Liver cell culture and lineage biology, in Methods of Tissue Engineering, (eds., A. Atala and R.P. Lanza), Academic Press, London, 2002, pp. 151–202. 47. Wang, Y., Yao, H., Barbier, C. et al. Paracrine signals from mesenchymal cell populations govern the expansion and differentiation of human hepatic stem cells to adult liver fates. Hepatology, 2010;52(4):1443–54. 48. Nusse, R. Wnt signaling and stem cell control. Cell Res, 2008;18:523–7. 49. Roskams, T. and Desmet, V. Embryology of extra‐ and intrahepatic bile ducts, the ductal plate. Anat Rec, 2008;291(6):628–35. 50. Carpentier, R., Español‐Suñer, R., van Hul, N. et al. Embryonic ductal plate cells give rise to cholangiocytes, periportal hepatocytes, and adult liver progenitor cells. Gastroenterology, 2011;141(4):1432–8. 51. Zhang, L., Theise, N., Chua, M., and Reid, L.M. The stem cell niche of human livers: symmetry between development and regeneration. Hepatology, 2008;48(5):1598–607. 52. Saxena, R. and Theise, N. Canals of Hering: recent insights and current knowledge. Semin Liver Dis,2004;24(1):43–8. 53. Theise, N.D., Saxena, R., Portmann, B.C. et al. The canals of Hering and hepatic stem cells in humans. Hepatology, 1999;30(6):1425–33. 54. Strobel, O., Rosow, D.E., Rahaklin, E.Y. J. et al. Pancreatic duct glands are distinct ductal compartments that react to chronic injury and mediate Shh‐ induced metaplasia. Gastroenterology, 2010;138:1166–77. 55. Kushner, J.A., Weir, G.C., and Bonner‐Weir, S. Ductal origin hypothesis of pancreatic regeneration under attack. Cell Metab, 2010;11(1):2–3. 56. Bonner‐Weir, S., Tosch, E., Inada, A. et al. The pancreatic ductal epithelium serves as a potential pool of progenitor cells. Pediatric Diabetes, 2004; 5:16–22. 57. McClelland, R., Wauthier, E., Uronis, J., and Reid, L.M. Gradient in extracellular matrix chemistry from periportal to pericentral zones: regulation of hepatic progenitors. Tissue Eng,2008;14(1):59–70. 58. Kubota, H., Yao, H., and Reid, L.M. Identification and characterization of vitamin A‐storing cells in fetal liver. Stem Cell, 2007;25:2339–49. 59. Lesley, J., Hascall, V.C., Tammi, M., and Hyman, R. Hyaluronan binding by cell surface CD44. J Biol Chem, 2000;275(35):26967–75. 60. Hayes, A., Tudor, D., Nowell, M., Caterson, B., and Hughes, C. Chondroitin sulfate sulfation motifs as putative biomarkers for isolation of articular cartilage progenitor cells. J Histochem Cytochem, 2007;56:125–38. 61. McClelland R., Wauthier E., Uronis J, et al. Gradients in the liver’s extracellular matrix chemistry from periportal to pericentral zones: influence on human hepatic progenitors. Tissue Eng Part A, 2008;14(1):59–70. 62. Zern, M. and Reid, Z. Extracellular Matrix Chemistry and Biology, Academic Press, New York, 1993. 63. Aruffo, A., Stamenkovic, I., Melnick, M., Underhill, C.B., and Seed, B. CD44 is the principal cell surface receptor for hyaluronate. Cell, 1990; 61:1303–13. 64. Nakanuma, Y., Hoso, M., Sanzen, T., and Sasaki, I.M. Microstructure and development of the normal and pathologic biliary tract in humans, including blood supply. A review. Microsc Res Tech, 1997;15(38):552–70.



42:  Stem Cell‐Fueled Maturational Lineages in Hepatic and Pancreatic Organogenesis

65. Nakanuma, Y., Katayanagi, K., Terada, T., and Saito, K. Intrahepatic peribiliary glands of humans. I. Anatomy, development and presumed functions. A review. J Gastroenterol Hepatol, 1994;9(1):75–9. 66. Nakanuma, Y., Sasaki, M., Terada, T., and Harada, K. Intrahepatic peribiliary glands of humans. II. Pathological spectrum. Review. J Gastroenterol Hepatol, 1994;9(1):80–6. 67. Portulano, C., Paroder‐Belenitsky, M., and Carrasco, N. The Na+/I‐ symporter (NIS): mechanism and medical impact. Endocrine Rev,2014;35(1): 106–49. 68. Lozoya, O.A., Wauthier, E., Turner, R. et  al. Regulation of hepatic stem/ progenitor phenotype by microenvironment stiffness in hydrogel models of the human liver stem cell niche. Biomaterials, 2011;32(30):7389–402. 69. Wauthier, E., McClelland, R., Turner, W. et al. Hepatic stem cells and hepatoblasts: identification, isolation and ex vivo maintenance. Methods Cell Biol, 2008;86:137–225. 70. Cardinale, V., Carpino, G., Gentile, R. et al. Transplantation of human fetal biliary tree stem/progenitor cells into two patients with advanced liver cirrhosis. BMC Gastroenterology, 2014;14:204. 71. Michalopoulos, G.K. Liver regeneration: alternative epithelial pathways. Int J Biochem Cell Biol, 2011;43:173–9. 72. Cohen, D.E. and Melton, D. Turning straw into gold: directing cell fate for regenerative medicine. Nat Rev Genet, 2011;12:243–52. 73. Puppi, J., Strom, S.J., Hughes, R.D. et  al. Improving the techniques for human hepatocyte transplantation: report from a consensus meeting in London. Cell Transplant, 2012;21(1)1–10. 74. Washburn, M.L., Bility, M.T., Zhang, L. et al. A humanized mouse model to study hepatitis C virus infection, immune response, and liver disease. Gastroenterology, 2011;140(4):1334–44. 75. Fukumitsu, K., Yagi, H., and Soto‐Gutierrez, A. Bioengineering in organ transplantation: targeting the liver. Transplant Proc,2011;43:2137–8. 76. Gerlach, J.C. Bioreactors for extracorporeal liver support. Cell Transplant, 2006;15(1):S91–103. 77. Russo, F.P. and Parola, M. Stem and progenitor cells in liver regeneration and repair. Cytotherapy, 2011;13:135–44. 78. Parveen, N., Aleem, A.K., Habeeb, M.A., and Habibullah, C.M. An update on hepatic stem cells: bench to bedside. Curr Pharm Bioltechnol, 2011;12(2):226–30. 79. Duncan, A., Hickey, R.D., Paulk, N.K. et  al. Ploidy reductions in murine fusion‐derived hepatocytes. PloS Genet, 2009;5(2):e1000385. 80. Tanaka, M., Itoh, T., Tanimizu, N., and Miyajima, A. Liver stem/progenitor cells: their characteristics and regulatory mechanisms. J Biochem, 2011; 149(3):231–9. 81. Formin, M.E., Tai, L.K., Bárcena, A., and Muench, M.O. Coexpression of CD14 and CD326 discriminate hepatic precursors in the human fetal liver. Stem Cells, 2011;20(7):1247–57. 82. Thorgeirsson, S., Factor, V., and Grisham, J. Early activation and expansion of hepatic stem cells, in Handbook of Stem Cells, 2nd edn, (eds. R. Lanza, H. Blau, D.A. Melton et al.) Elsevier, New York, 2004, pp. 497–512. 83. Navarro‐Alvarez, N., Soto‐Gutierrez, A., and Kobayashi, N. Hepatic stem cells and liver development. Method Mol Biol, 2010;640:181–236. 84. Vessey, C.J. and de la Hall, P.M. Hepatic stem cells: a review. Pathology, 2001;33(2):130–41. 85. Reid, L.M., Fiorino, A.S., Sigal, S.H., Brill, S., and Holst, P.A. Extracellular matrix gradients in the space of Disse: relevance to liver biology. Hepatology, 1992;15(6):1198–203. 86. Kopp, J.L., Grompe, M., and Sanders, M. Stem cells versus plasticity in liver and pancreas regeneration. Nat Cell Biol, 2016;18(3):238–45. 87. Khan, A.A., Parveen, N., Mahaboob, V.S. et al. Management of hyperbilirubenemia in biliary atresia by hepatic progenitor cell transplantation through hepatic artery: a case report. Transplant Proc, 2008;40(4):1153–5. 88. Khan, A.A., Parveen, N., Mahaboob, V.S. et al. Treatment of Crigler‐Najjar syndrome type 1 by hepatic progenitor cell therapy: a simple procedure for hyperbilirubinemia. Transplant Proc, 2008;40(4):1148–50. 89. Khan, A.A., Shaik, M.V., Parveen, N. et al. Human fetal liver‐derived stem cell transplantation as supportive modality in the management of end‐stage decompensated liver cirrhosis. Cell Transplant, 2010;19(4):409–18. 90. Stachelscheid, H., Urbaniak, T., Ring, A., Spengler, B., Gerlach, J.C., and Zeilinger, K. Isolation and characterization of adult human liver progenitors from ischemic liver tissue derived from therapeutic hepatectomies. Tissue Eng Part C Methods, 2009;15(7):1633–43.

537

 91. McClelland, R., Wauthier, E., Zhang, L. et  al. Ex vivo conditions for self‐replication of human hepatic stem cells. Tissue Eng Part C Methods, 2008;14(4):314–51.   92. Schmelzer, E. and Reid, L.M. Telomerase activity in human hepatic stem cells, hepatoblasts and hepatocytes from neonatal, pediatric, adult and geriatric donors. Eur J Hepatol Gastroenterol, 2009;21(10):1191–8.   93. Oikawa, T., Kamiya, A., Zeniya, M. et  al. Sal‐like protein 4 (SALL4), a stem cell biomarker in liver cancers. Hepatology, 2012;57(4):1469–83.   94. Turner, R., Gerber, D., and Reid, L.M. Transplantation of cells from solid organs requires grafting protocols. Transplantation, 2010;90:807–10.   95. Hodgman, C.D. Handbook of Chemistry and Physics, 37th ed, Chemical Rubber Publishing Co., 1955–1956, pp. 3156.   96. Turner, W.S., Schmelzer, E., McClelland, R., Wauthier, E., Chen, W., and Reid, L.M. Human hepatoblast phenotype maintained by hyaluronan hydrogels. J Biomed Mater Res, 2007;82(1):156–68.   97. Turner, W.S., Seagle, C., Galanko, J. et  al. Metabolomic footprinting of human hepatic stem cells and hepatoblasts cultured in engineered hyaluronan‐matrix hydrogel scaffolds. Stem Cell, 2008;26:1547–55.   98. Turner, R., Mendel, G., Wauthier, E., Barbier, C., and Reid, L.M. Hyaluronan‐ supplemented buffers preserve adhesion mechanisms facilitating cryopreservation of human hepatic stem/progenitor cells. Cell Transplant, 2012;21(10): 2257–66.   99. Li, W. and Ding, S. Small molecules that modulate embryonic stem cell fate and somatic cell reprogramming. Trends Pharmacol Sci,2010;31(1):36–45. 100. Turner, R., Wauthier, E., Lozoya, O. et  al. Successful transplantation of human hepatic stem cells with restricted localization to liver using hyaluronan grafts. Hepatology, 2013;57:775–84. 101. Weber, A., Mahieu‐Caputo, D., Michelle Hadchoue, M., and Franco, D. Hepatocyte transplantation: studies in preclinical models. J Inherit Metab Dis, 2006;29(2–3):436–41. 102. Weber, A., Groyer‐Picard, M.T., Franco, D., and Dagher, I. Hepatocyte transplantation in animal models. Liver Transplant, 2009;15(1):7–14. 103. Kinoshita, T. and Miyajima, A. Cytokine regulation of liver development. Biochim Biophys Acta, 2002;1592(3):303–12. 104. Kamiya, A., Kinoshita, T., Ito, Y. et al. Fetal liver development requires a paracrine action of oncostatin M through the gp130 signal transducer. EMBO J, 1999;18(8):2127–36. 105. Rojkind, M., Gatmaitan, Z., Mackensen, S., Giambrone, M.A., Ponce, P., and Reid, L.M. Connective tissue biomatrix: its isolation and utilization for long‐ term cultures of normal rat hepatocytes. J Cell Biol, 1980;87(1):255–63. 106. Reid, L.M., Gaitmaitan, Z., Arias, I., Ponce, P., and Rojkind, M. Long‐term cultures of normal rat hepatocytes on liver biomatrix. Ann N T Acad Sci, 1980;349:70–6. 107. Purushothaman, A., Hurst, D.R., Pisano, C., Mizumoto, S., Sugahara, K.,and Sanderson, R.D. Heparanase‐mediated loss of nuclear syndecan‐1 enhances histone acetyltransferase (HAT) activity to promote expression of genes that drive an aggressive tumor phenotype. J Biol Chem, 2011;286(35):30377–83. 108. Capila, I. and Linhardt, R.J. Heparin ± protein interactions. Angewandte Chemie Int Ed, 2002;41:390–412. 109. Itoh, T. and Miyajima, A. Liver regeneration by stem/progenitor cells. Hepatology, 2014;58(4):1617–26. 110. Michalopoulos, G.K. and DeFrances, M.C. Liver regeneration. Review. Science, 1997;276(5309):60–6. 111. Chiang, J.Y. Bile acid regulation of gene expression: roles of nuclear hormone receptors. Endocrine Rev, 2002;23(4):443–63. 112. Perelman, A. and Brandan, E. Different membrane‐bound forms of acetylcholinesterase are present at the cell surface of hepatocytes. Eur J Biochem, 1989;182(1):203–7. 113. Alvaro, D., Alpini, G., Jezequel, A.M. et al. Role and mechanisms of action of acetylcholine in the regulation of rat cholangiocyte secretory functions. J Clin Invest, 1997;100(6):1349–62. 114. LeSage, E.G., Alvaro, D., Benedetti, A. et  al. Cholinergic system modulates growth, apoptosis, and secretion of cholangiocytes from bile duct‐ ligated rats. Gastroenterology, 1999;117(1):191–9. 115. Vongchan, P., Warda, M., Toyoda, H., Toida, T., Marks, R.M., and Linhardt, R.J. Structural characterization of human liver heparan sulfate. Biochim Biophys Acta, 2005;1721(1–3):1–8. 116. Fujita, M., Spray, D.C., Choi, H. et al. Glycosaminoglycans and proteoglycans induce gap junction expression and restore transcription of tissue‐­ specific mRNAs in primary liver cultures. Hepatology, 1987;7(1):1S–9S.

538

THE LIVER:  REFERENCES

117. Spray, D.C., Fujita, M., Saez, J.C. et al. Proteoglycans and glycosaminoglycans induce gap junction synthesis and function in primary liver cultures. J Cell Biol, 1987;105(1):541–51. 118. Masyuk, A.I., Masyuk, T.V., and LaRusso, N.F. Cholangiocyte primary cilia in liver health and disease. Dev Dynam, 2008;237:2007–12. 119. Huang, B.Q., Masyuk, T.V., Muff, M.A., Tietz, P.S., Masyuk, A.I., and Larusso, N.F. Isolation and characterization of cholangiocyte primary cilia. Am J Physiol Gastrointest Liver Physiol, 2006;291(3):G500–9. 120. Cervantes, S., Lau, J., Cano, D.A., Borromeo‐Austin, C., and Hebrok, M. Primary cilia regulate Gli/Hedgehog activation in pancreas. Proc Nat Acad Sci USA, 2010;107(22):10109–14. 121. Michalopoulos, G.K. and Appasamy, R. Metabolism of HGF‐SF and its role in liver regeneration. Review. EXS, 1993;65:275–83. 122. Liu, H., Di Cunto, F., Imarisio, S., and Reid, L.M. Citron kinase is a cell cycle‐dependent, nuclear protein required for G2/M transition of hepatocytes. J Biol Chem, 2003;278(4):2541–8. 123. Sigal, S.H., Rajvanshi, P., Gorla, G.R. et al. Partial hepatectomy‐induced polyploidy attenuates hepatocyte replication and activates cell aging events. Am J Physiol Gastrointest Liver Physiol, 1999;276(5):G1260–72. 124. Huch, M. and Dollé, L. The plastic cellular states of liver cells: are EpCAM and Lgr5 fit for purpose? Review. Hepatology, 2016:64(2);652–62.

125. Takahashi, K. and Yamanaka, S. A developmental framework for induced pluripotency. Review. Development, 2015;142(19):3274–85. 126. Dorrell, C., Tarlow, B., Wang, Y. et  al. The organoid‐initiating cells in mouse pancreas and liver are phenotypically and functionally similar. Stem Cell Res, 2014;13(2):275–83. 127. Lu, W.Y., Bird, T.G., Boulter, L. et al. Hepatic progenitor cells of biliary origin with liver repopulation capacity. Nat Cell Biol, 2015;17(8):971–83. 128. Tarlow, B.D., Pelz, C., Naugler, W.E. et al. Bipotential adult liver progenitors are derived from chronically injured mature hepatocytes. Cell Stem Cell, 2014;15(5):605–18. 129. Fan, B., Malato, Y., Calvisi, D.F. et al. Cholangiocarcinomas can originate from hepatocytes in mice. J Clin Invest, 2012;122(8):2911–5. 130. Smukler, S.R., Arntfield, M.E., Razavi, R. et  al. The adult mouse and human pancreas contain rare multipotent stem cells that express insulin. Cell Stem Cell, 2011;8:281–93. 131. Grompe, M. and Strom, S. Mice with human livers. Gastroenterology, 2013;145(6):1209–14. 132. Rhim, J.A., Sandgren, E.P., Palmiter, R.D., and Brinster, R.L. Complete reconstitution of mouse liver with xenogeneic hepatocytes. Proc Nat Acad Sci USA, 1995;92(11):4942–6.

43

Developmental Morphogens and Adult Liver Repair Mariana Verdelho Machado1 and Anna Mae Diehl2 Gastroenterology and Hepatology Department, Hospital de Santa Maria, CHLN, Lisbon, and Faculty of Medicine, Lisbon University, Lisbon, Portugal 2 School of Medicine, Duke University, Durham, NC, USA 1

HEDGEHOG SIGNALING Pathway overview The canonical Hh pathway is a conserved, highly complex signaling cascade with four fundamental components: (i) the ligand Hedgehog, (ii) the receptor Patched (Patch), (iii) the ­signal transducer Smoothened (Smo), and (iv) the effector transcription factor, Gli (Figure 43.1). Canonical Hh signaling occurs along the primary cilium (PC) with components of the Hh pathway concentrating in PC [1] and a complex PC trafficking system regulating the interaction of Hh pathway components to enhance, or block, the Hh‐initiated signal [2]. Hh is a protein produced as a 45 kDa precursor that undergoes proteolytic processing in the endoplasmic reticulum (ER) [3] and subsequent lipid modification to acquire cholesterol and palmitoyl groups [4, 5]. Hh is secreted into the extracellular space, diffusing away from the ligand‐producing cell to bind to other cells whereby it determines their fate according to the concentration and duration of exposure [6]. Extracellular matrix proteins, such as proteoglycans, modulate the diffusion of Hh through the extracellular space and thus, regulate the concentration of Hh to which target cells are exposed [7]. Mammals have three different Hh proteins: Sonic (Shh), Indian (Ihh), and Desert (Dhh) hedgehog. The three ligands similarly activate the Hh pathway in Hh‐responsive cells, however their expression is differently regulated. While Shh and Ihh are widely expressed, Dhh is thought to be expressed mainly in the nervous system and testis [8]. Patch, the Hh receptor, is a protein with 12 transmembrane domains. When Hh ligands are absent, Patch localizes to the PC and constitutively inhibits the Hh pathway by blocking Smo, the signal transducer protein, from being activated and entering the  PC. When Hh ligand binds to Patch, these inhibitory actions of Patch are relieved and Smo becomes active. Hh‐Patch

interactions regulate Smo activity by controlling cholesterol modification of Smo. Smo is directly activated by binding cholesterol. Patch suppresses this cholesterol modification of Smo in the absence of Hh ligands, and this inhibition is relieved when Hh binds to Patch [9]. The Hh–Patch complex is subsequently internalized and degraded [10]. Three Hh co‐receptors, CAM‐ related downregulated by oncogene (Cdo), brother of Cdo (Boc), and growth‐arrest‐specific (GAS)‐2, potentiate Hh ­signaling by enhancing Hh‐Ptch interaction [11]. Conversely, Hhip, a soluble Hh receptor, inhibits Hh signaling by preventing Hh‐Patch binding [12]. Smo is a transmembrane G‐protein coupled receptor that mediates activation of Gli transcription factors in Hh‐responsive cells. Gli proteins promote transcription of several genes important in the regenerative/repair process, including vascular endothelial growth factors, angiopoietin‐1 and ‐2, snail, twist‐2, α‐smooth muscle actin, vimentin, nanog, sox2 and sox9 [8]. In the absence of Hh, Smo activity is repressed by Patch, and Gli binds to fused kinase (Fu), suppressor of fused (Sufu) and Costal‐2, to form a suppressor protein complex which prevents Gli from entering the nucleus [13]. Arrested in the cytoplasm by the suppressor protein complex, Gli is sequentially phosphorylated by protein kinase A (PKA), glycogen synthase kinase‐3 (GSK3), and calmodulin kinase‐1 (CK1). Phosphorylated Gli then binds to β‐transducin repeat containing protein (βTrCp) and the Gli‐βTrCp complex is targeted to the proteasome where Gli can be either degraded entirely or processed to generate a truncated transcription repressor (Gli‐R) [14]. When Hh binds to Patch, Smo is de‐repressed and activated Smo dissociates Gli from the suppressor protein complex, preventing Gli phosphorylation and subsequent degradation. This enables full‐length Gli to move to the nucleus where it acts as a transcription factor. Mammals have three known Gli proteins: Gli1, ‐2 and ‐3. Gli1 does not undergo proteosomal degradation and hence, remains untruncated and always promotes transcription. Gli1 is

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

540

THE LIVER:  HEDGEHOG SIGNALING

PC

PC

Hh

Smo

Patch Plasma membrane

CKS GKA

Smo

Plasma membrane

PKA CKS

Phosphorylation Ubiquitination

Gli

PKA

GKA

SCF-βTrCP

Gli

Partial Processing Proteasome

Nucleus

Nucleus

STOP Gli-R

Hh-Patch complex

Gli-A

PATHWAY OFF

PATHWAY ON

Figure 43.1  The Hedgehog signaling pathway. (a) Pathway off: Patched (Patch) blocks the entry of Smoothened (Smo) into the primary cilium (PC), which represses Smo activity and allows sequential phosphorylation of Gli by several kinases: protein kinase A (PKA), glycogen synthase‐3β (GSK3β) and casein kinase‐1 (CK1). Phosphorylated Gli undergoes ubiquitination by Skp‐Cullin‐F‐box (SCF) protein/β‐transducing repeat containing protein (TrCP), which primes Gli to limited degradation in the proteasome. Truncated Gli (Gli‐R) is a repressor of gene transcription. (b) Pathway on: Hedgehog binds to Patch and removes it from the PC, which allows Smo entering into the PC and Smo activation. Active Smo blocks phosphorylation and subsequent degradation of Gli. Full length Gli translocates to the nucleus and promotes transcription of several target genes.

an important target gene for Gli2 [15]. Full‐length Gli2 accumulates when Smo is activated because activated Smo protects Gli2 from proteosomal degradation. When Smo is inactive, both Gli2 and Gli3 are targeted to the proteasome; Gli2 is usually fully degraded but Gli3 is frequently partially processed to a truncated form that represses transcription [16]. Hence, Gli3 can act as either a transcriptional repressor (when Smo is inactive) or as an activator of transcription (when activated Smo protects it from proteosomal degradation). In contrast, Gli1 and Gli2 act predominantly as transcription promoters. Emerging evidence suggests that there may be a “Gli code” whereby the different Gli factors interact to control their expression. According to this model, Gli3 and Gli1 promote each other’s expression when Gli2 is low but become mutually antagonistic when Gli2 accumulates [17]. Besides the canonical Hh pathway, there are also two known types of non‐canonical Hh signaling. Type 1 non‐canonical Hh signaling depends on Patch but is Smo‐independent. In the absence of Hh, Patch has direct pro‐apoptotic and antiproliferative effects, by activating caspase‐3 [18] and preventing nuclear localization of cyclin D [19], respectively. Both effects of Patch are lost when Hh binds to Patch. Type 2 noncanonical Hh signaling depends on Smo but it does not require PC [20]. This noncanonical signaling depends on the Gαi activity of Smo that directly regulates metabolism (e.g. it promotes a Warburg‐like effect promoting glycolysis in muscle, adipose tissue, and myofibroblasts [21, 22]), proliferation, calcium flux, and migration (in myofibroblasts and endothelial cells [23, 24]). Additionally, Gli signaling can occur in the absence of Hh via a process that also appears to be Patch and Smo‐independent, as demonstrated by evidence that Gli induction is a direct downstream consequence of transforming growth factor (TGF) beta and RAS signaling [25, 26].

Liver development The exact contribution of the Hh pathway to liver embryogenesis remains unknown although it is clear that the pathway is crucial for differentiation of the liver bud from the ventral foregut endoderm [27]. Later in development, activation and repression of Hh signaling control hepatoblast fate, with activation stimulating a proliferative undifferentiated phenotype and repression permitting hepatoblasts to differentiate into mature hepatocytes [28]. Hedgehog is also crucial for the proper development of the muscle layer and functioning of the gallbladder and disruption of the Hh pathway at that point in development abrogates the protective effect of the gallbladder against the toxic effects of bile acids on the biliary system, which may contribute to the development of biliary atresia [29].

Liver neoplasia Primary liver cancers Hedgehog pathway activation has been demonstrated in various types of primary liver cancer, including hepatocellular carcinoma [30, 31], cholangiocarcinoma [32], and fibrolamellar hepatocellular carcinoma [33]. In all of these cancers, the level of pathway activity correlates with worse clinical outcomes and reduced cancer‐free, and overall, survival [31]. Relatively uncommon genetic mechanisms, and more prevalent epigenetic mechanisms, contribute to hedgehog dysregulation in liver cancers. For example, liver cancer cell growth is increased both by certain mutations of Smo that activate it constitutively [30], and by hypermethylation of Hh signaling inhibitors which epigenetically suppresses expression of factors that typically constrain pathway activity [34]. Dysregulated Hh signaling has been demonstrated in diverse cell types of liver cancers,



43:  Developmental Morphogens and Adult Liver Repair

including the malignant epithelial cells themselves, cancer‐ associated fibroblasts (CAFs), and tumor‐associated macrophages (TAMs). Hedgehog signaling in CAFs promotes fibrogenesis and generation of factors that maintain a supportive niche for cancer stem‐like cells, responses that likely foster cancer growth [35]. Similarly, Hh signaling in TAMs promotes immune tolerance that permits cancer cells to escape immune surveillance mechanisms [36]. On the other hand, Hh signaling also promotes vasculogenesis and this might enhance chemotherapy bioavailability [37, 38]. These opposing actions of Hh make it difficult to predict if treatments that inhibit Hh signaling might be beneficial in primary liver cancers. Enthusiasm has also been dampened by inconsistent efficacy and relatively prevalent toxicities observed when Hedgehog was inhibited in other cancers [39]. Clinicaltrials.gov lists ongoing trials of hedgehog inhibitors in hepatocellular carcinoma.

Hepatic adenomas A subtype of hepatic adenomas was recently identified in which a chromosomal deletion results in constitutive activation of Gli1 in hepatocytes. Because Gli1 is both a downstream target and a proximal effector of the Hh signaling pathway, neoplasia is attributed to dysregulated Hedgehog signaling. Hedgehog signaling drives vasculogenesis during development and these adenomas are particularly prone to clinically‐significant hemorrhage [40]. Interestingly, the Hedgehog‐high adenoma subclass is also marked by high expression of arginosuccinate synthase 1, a urea cycle enzyme [41]. This finding supports emerging evidence that Hh signaling activity in hepatocytes may regulate metabolic zonation in healthy adult liver [42]. According to this model, Hh activity is higher in periportal hepatocytes (which use the urea cycle to detoxify ammonia) than in perivenous hepatocytes (which exhibit negligible urea cycle activity) [41]. Conversely, Wnt signaling is known to be higher in zone three hepatocytes (which synthesize glutamine to scavenge ammonia) than in zone one hepatocytes (which do not) [43]. Differential Hh and Wnt activity also occurs along the intestinal crypt–villus axis. In that tissue, Hh and Wnt appear to be mutually antagonistic, with Hh restricting propagation of Wnt signals and thereby restricting the stem cell compartment to the crypt [44].

Liver repair The adult liver is designed to resist injury. Hence, hepatocyte attrition is mainly due to natural senescence and liver mass is easily maintained by a small subpopulation of hepatocytes that are more proliferative than the vast majority of adult hepatocytes, which are devoted to performing various liver‐specific functions [45]. In contrast, liver injuries that kill hepatocytes dramatically increase the demand for hepatocyte replacement and this triggers multifaceted wound healing responses that aim to restore functional hepatocyte mass [46]. Upregulation of Hh signaling has been demonstrated in injured livers, regardless of injury etiology [47]. Further, the level of pathway activity typically parallels the severity of liver damage and fibrosis [48]. These observations suggest that Hh critically regulates how adult livers respond to injury. The mechanisms involved are summarized below.

541

Hepatocytes produce Hedgehog ligands One mechanism that may explain how liver injury is tightly coupled to Hh pathway activation was suggested by the discovery that adult hepatocytes produce and release large amounts of Shh and Ihh ligands when subjected to challenges that promote ER stress and apoptosis [49, 50]. In such circumstances, ­biologically‐active Hh ligands appear to localize in hepatocyte‐ derived exosomes and microparticles [51]. Once released into the hepatic microenvironment, bile, and blood, these membrane‐bound Hh ligands are capable of activating Hh signaling in local and distant Hedgehog‐responsive cells and hence, function as hepatocyte‐derived hormones [51]. In injured livers, increased exposure to Hh ligands triggers changes in proliferation, viability, and differentiation of various types of local target cells to orchestrate reconstruction of adult liver tissue via regenerative mechanisms that resemble those involved in fetal liver development [47]. The impact of liver‐derived Hh ligands on Hh signaling in extrahepatic tissues is less studied but might be significant because Hh signaling suppresses adipogenesis in  white adipose depots [52], promotes vasculogenesis [53], and modulates immune function [54, 55], and advanced liver disease is characterized by cachexia, vascular remodeling, and immune dysfunction.

Hepatic stellate cells both produce and respond to Hedgehog ligands The major fibrogenic cell type in liver, the hepatic stellate cell (HSC), produces and is highly responsive to, Hedgehog ligands [56]. Coupled with the fact that HSCs live in the space of Disse immediately adjacent to hepatocytes and sinusoidal endothelial cells (other liver‐resident cell types that produce and/or respond to Hh ligands), HSC are positioned to be a key node in Hh‐ regulated regenerative responses. HSCs are also capable of ­noncanonical Hh signaling because they express various G ­protein‐coupled receptors that regulate Smo independently of Hedgehog–Patched interaction [22]. In addition, they can activate Gli2 via morphogen‐driven mechanisms that do not require Smo [57]. Such signaling flexibility supports the concept that net Hh pathway activity must be tightly regulated in HSC in  order to assure optimal liver repair. Indeed, Hh pathway activation is necessary for quiescent HSC to become and remain myofibroblasts and in mice, liver fibrosis is inhibited by various approaches that inhibit Smo and/or suppress Gli1/2 activity [56, 58, 59]. Conversely, simply overexpressing Shh ligand in hepatocytes is sufficient to induce progressive liver fibrosis in mice [60].

Liver sinusoidal cells (LSECs) respond to Hh ligands Liver sinusoidal cells (LSECs) respond to Hh ligands with a repertoire of behaviors that results in blood vessel formation and growth during liver injury. In LSECs, increased Hh ­signaling induces loss of fenestrae (capillarization) [61] and may promote portal hypertension/portal–systemic shunting [62]. Liver endothelial cells can also produce Hh ligands which modulate liver repair via autocrine and paracrine mechanisms [61].

542

THE LIVER:  NOTCH SIGNALING

The innate immune system is also highly Hedgehog‐responsive T cells and natural killer T (NKT) cells require Hh signaling to remain viable [54]. Hedgehog signaling also promotes the synthesis of certain chemotactic factors for NKT cells (CXCL16), and macrophages (CCL2) [63, 64]. Further, Hedgehog‐sensitive mechanisms modulate immune polarization of T cells, NKT cells, and macrophages: Hh signaling typically promotes Th2/ M2 polarization that enhances immune tolerance [55]. Tolerance is mediated, in part, via increased production of cytokines that also induce fibrogenesis (e.g. IL‐4, IL‐13, TGFβ) by re‐enforcing Hh signaling in hepatic stellate cells [54].

Ductal cells produce and respond to Hh ligands Paracrine Hh signaling between ductal cells and neighboring myofibroblasts induces a migratory, less mature, more proliferative phenotype in ductal cells and re‐enforces the fibrogenic phenotype of the myofibroblasts [65]. Together with the aforementioned effects on Hh responsive immune cells, these actions promote the ductular reaction, a fibroproliferative inflammatory response to severe acute or chronic liver injury [66].

Liver metabolism The Hh pathway is known to regulate metabolism in adipocytes [67], fibroblasts [22, 68], stem cells [69], immune cells [70], and many cancer cells [35]. Pathway activity generally stimulates glycolysis [22] and inhibits lipogenesis [71] but the mechanisms involved are multi‐faceted and not fully understood. For example, Smo can directly activate AMP kinase, a master regulator of cellular energy balance [21]. Hedgehog signaling may also control mitochondrial mass [72] and has been shown to interact with other pathways that modulate metabolism, such as Notch [73] and Wnt [44]. Hence, contextual factors are likely to influence exactly how Hh impacts metabolism in any given cell. Until recently, Hh signaling was not thought to be directly involved in hepatocyte metabolism. However, emerging evidence suggests that Hh may be a critical regulator of hepatic lipid metabolism [17]. This activity appears to be exquisitely controlled by circadian forces [17] and might have systemic, as well as local, implications for tissue health. This realization is triggering research to delineate the mechanisms that control Hh signaling in hepatocytes. Healthy hepatocytes may produce Hh ligands, although this has been difficult to demonstrate by immunostaining intact liver tissue. Failure to visualize Hh ligands in such cells could reflect the fact that healthy hepatocytes generally produce low levels of Hh ligands and/or release them efficiently. The latter possibility is supported by two lines of recent evidence. First, Hh ligands can traffic along cytonemes, providing a mechanism to shuttle ligands from ligand‐producing hepatocytes to immediately adjacent ligand‐receiving cells without releasing ligands into the extracellular space [74]. Second, VLDL and other lipoprotein particles harbor biologically‐active Hh ligands in healthy adults [75]. The association of Hh ligands with cholesterol‐ containing lipoproteins is particularly intriguing because cholesterol can bind to the extracellular domain of Smo and activate it directly [76], while Hh ligands activate Smo only after

engaging their receptor, Patched [77]. Interestingly, Smo is directly inhibited by certain other lipoprotein‐associated lipids [78] and thus, the healthy liver may modulate Hh signaling by varying lipoprotein particle composition. Hedgehog ligand producing cells that traffic into and out of the liver (e.g. immune cells, bone marrow‐derived mononuclear cells) might be another source of Hh ligands for healthy hepatocytes. Hedgehog ligands might also be provided by subpopulations of liver‐resident cells that are capable of generating Hh ligands (e.g. HSCs, LSECs, ductal cells). Signaling downstream of Hedgehog–Patched could also be modulated by non‐canonical Hh signaling that operates independently of Patched or Smo, affording hepatocytes multiple methods to titrate pathway activity. Indeed, Hh signaling is likely to be tightly regulated in most cells because pathway activity critically modulates DNA methylation and chromatin remodeling by regulating one carbon metabolism and redox state to exert epigenetic control of cell fate decisions [79].

Liver diseases with hedgehog dysregulation Hedgehog pathway activity increases with the severity of liver damage and fibrosis in multiple human liver diseases, including non‐alcoholic fatty liver disease (NAFLD) [80], alcoholic liver disease [81], viral hepatitis [82], schistosomiasis [55], primary biliary cholangitis (PBC) [83], primary sclerosing cholangitis (PSC) [84], and biliary atresia [85]. Further, liver damage and fibrosis were improved by inhibiting Hh signaling in animal models of some of these conditions (e.g. NAFLD [86, 87], schistosomiasis [88], PSC [56], biliary atresia [89]). To date, clinical trials of Hh inhibitors have not been done in any of these diseases. The major impediment seems to be the risk of adverse events caused by inhibiting Hh activity in extrahepatic tissues. Sustained inhibition of Hh might also have negative impacts on the liver itself because genetic defects that globally impair Smo activation promote the metabolic syndrome [90] and hepatic steatosis [91] in humans and inhibiting Smo blocks liver regeneration after partial hepatectomy in mice [92]. The aggregate data suggest that future therapeutic approaches to target excessive Hh signaling must bring pathway activity back to physiological levels, rather than silencing signaling completely.

NOTCH SIGNALING Pathway overview The Notch signaling pathway is fundamentally different from the Hh pathway in that it requires cell‐to‐cell contact and the interaction of a transmembrane receptor in the Notch target cell with a transmembrane ligand in the ligand‐producing cell [93]. Both pathways ultimately activate transcription factors. The canonical Notch pathway has three fundamental components: (i) the ligands (delta‐like ligands and jagged), (ii) the receptor (Notch), (iii) the effector (transcription factor CBF‐1/Su(H)/ LAG1 [CSL] and coactivator master‐mind‐like [MAML]) [94] (Figure 43.2). Both ligands and receptors undergo complex post‐translational modifications before reaching the cell membrane. Ligands are



43:  Developmental Morphogens and Adult Liver Repair

543

Notch ligand expressing cell Jag Dll Notch

Notch receptor expressing cell

γ-secretase NICD

NICD

Nucleus

NICD RBP-Jk

Hes Hey

Figure 43.2  The Notch signaling pathway. Transmembrane ligands (Jagged or Dll) in one cell, bind to transmembrane receptor (Notch) in another cell, and exposes Notch to cleavage by γ‐secretases, resulting in the release of Notch intracellular domain (NICD). NICD enters the nucleus and binds to the transcription factor RBP‐Jk, promoting the expression of several target genes such as Hes and Hey.

expressed by the afferent cell as transmembrane proteins from the Delta/Serrate/LAG‐2 (DSL) family. Mammals have three Delta‐like ligands (Dll1, Dll3, and Dll4) and two Jagged ligands (Jag1 and Jag2) [94]. Before reaching maturation, ligands undergo a cycle of endocytosis and recycling to the cell surface via a process that involves ubiquitylation by the E3 ubiquitin ligases, Neuralized and Mindbomb [93]. Interactions between ligand and receptor can occur between two cells (in trans) leading to receptor activation, or in the same cell (in cis), resulting in signaling inhibition [95]. There are four receptor paralogs (Notch1–4), with different signal strength and thus, ligands may also have different effects depending on which receptor they engage [96]. The extracellular domain of Notch is composed of epidermal growth factor (EGF)‐like repeats and a negative regulatory region (NRR). The Notch receptor is modified by O‐glycans: O‐fucose adducts generated by Pofut enzyme are required for Notch function; O‐glucose modification by Rumi enzymes enhances Notch cleavage; Fringe proteins elongate O‐fucose by addition of GlcNac, and O‐xylose adducts. All of these glycations modulate ligand‐receptor interaction [93]. Other post‐translational modifications also ­ modulate cell signaling. For example, methylation by methyltransferase CARM1 enhances Notch activation [97]; Notch hydroxylation by protein factor inhibiting HIF‐1 modulates the hypoxia response [98]. Acetylation of Notch by PCAF and p300 increases Notch stability, and deacetylation by SIRT1 reduces Notch stability [99]; ubiquitylation targets Notch to rapid proteasome degradation; phosphorylation of Notch1

mediated by GSK3β enhances stability of Notch1, whereas it diminishes activity of Notch2 [100]. The NRR region in Notch prevents cleavage by proteases [101]. The interaction of ligand with Notch receptor changes the structure of the receptor, exposing cleavage sites to the sequential action of ADAM metalloproteases and the γ‐secretase complex [102]. This results in the release of Notch intracellular domain (NICD), which translocates into the nucleus due to its nuclear localization sequence. NICD can translocate directly to the nucleus or traffic there indirectly via the endosome compartment. The latter process is highly regulated; for example, Numb negatively regulates Notch through modulation of NICD endosomal trafficking [103]. In contrast, atypical protein kinase C amplifies Notch signaling by enhancing NICD relocalization from late endosomes to the nucleus [104]. Besides translocating into the nucleus, endosomal Notch can also recycle to the cell membrane or undergo lysosomal degradation [105]. In the nucleus NICD interacts with the DNA‐binding protein CSL (also known as DNA‐binding recombination signal‐binding protein Jκ [RBP‐Jκ]), and together recruit MAML [106]. Unbound CSL acts as a transcription repressor by binding several corepressors. In contrast, the complex NICD/CSL/MAML act as a transcription factor, binding to activating cofactors [93]. Classical target genes from the Notch pathway are hairy enhancer of split (Hes) and hairy related (Hey) families, but Notch regulates the expression of components of many important pathways in liver disease, such as hepatocyte nuclear ­factors (HNF), sox9, Hh, Wnt, PDGF, TGF, and VEGF [93].

544

THE LIVER:  NOTCH SIGNALING

Notch signaling can also occur via three noncanonical routes: type 1, ligand independent; type 2, CSL‐independent, and type 3, Notch independent. Regarding the type 1 noncanonical pathway, other ligands have been shown to activate Notch (for example MAGP1,2 and DLK1) but the functional consequences are not fully understood [107, 108]. On the other hand, types 2 and 3 noncanonical signaling seem likely to have functional effects: CSL‐independent Notch‐dependent signaling upregulates IL‐6 [109] and some viral proteins directly activate CSL activation independently of Notch [110].

Liver development Notch is crucial during liver development because it coordinates biliary differentiation and morphogenesis of intra‐ and extrahepatic bile ducts, as well as the gallbladder [111]. Mutations in Jag1 [112], or less frequently Notch2 [113], cause Alagille syndrome, an autosomal disease characterized by intrahepatic bile duct paucity and cholestasis, as well as heart, ocular, and vertebral defects. Mice heterozygous for mutations in Jag1 or Notch2 have an Alagille‐like phenotype [114, 115]. Severe cases of extrahepatic biliary atresia have also been associated with missense mutations in Jag1 [116]. Large extrahepatic bile ducts derive from branching of the primitive gut‐derived diverticulum, whereas small intrahepatic bile ducts derive from tubular structures within the ductal plate. The ductal plate is a layer of hepatoblasts surrounding the portal vein branches in a ring‐like array with ductal commitment, as opposed to hepatoblasts distant to the portal vein, which differentiate into hepatocytes [117]. Interestingly, some studies suggest that hepatoblasts express the Notch2 receptor and cholangiocyte differentiation is driven by interaction with Jag1‐ expressing periductal and periportal myofibroblasts [118]. Notch induces transcription factors critical for biliary specification (including Hes1, HNF1β and Sox9) and downregulates the expression of HNF1α and HNF4 [119]. The role of Notch in biliary differentiation is highly regulated and dependent on an intricate network of intercommunicating signaling pathways such as TGFβ [120], Wnt [121], and Hippo [122].

Liver neoplasia The effect of the Notch pathway in liver carcinogenesis is diverse and receptor‐dependent. Mouse models of liver cancer suggest that Notch1 and Notch2 promote hepatocellular carcinoma [123, 124]. However, while Notch2 also seems to promote the development of cholangiocarcinoma, Notch1 seems to inhibit it [123]. The Notch oncogenic effect was linked with upregulation of insulin growth factor 2 and Sox9 [124]. Interestingly, cancer associated fibroblasts may promote carcinogenesis by expressing Notch ligands that act on neighbor tumor cells [125]. However, the role of Notch in hepatocellular tumorigenesis is complex. For example, other authors demonstrated a p53‐dependent proapoptotic TRAIL‐sensitizing and chemotherapy‐sensitizing effect of Notch activation in tumor cells in vitro [126]. Regarding human hepatocellular carcinoma, one third of patients present a genetic signature of Notch activation, with upregulation of Jag1, α‐secretase ADAM17, Notch effectors

(e.g. MAML1), and target genes (e.g. Sox9, Spp1, and Hey1) [124]. Although the Notch genetic signature does not correlate with overall prognosis, it might help select patients who would respond to Notch‐targeted therapies, since cell lines that share Notch activation signature respond to Notch inhibition/inactivation with decreased proliferation [124]. A subset of hepatocellular carcinomas also showed increased components of Notch pathway expression at the protein level [126]. Those tumors are enriched with CK19+ and Sox9+ cells, markers of biliary epithelial cells and progenitor cells with bi‐phenotypic potential and/or stem characteristics. Furthermore, tumor expression of Notch components correlates with reduced patient survival [127, 128] and Notch polymorphisms in tumor cells correlate with increased susceptibility to tumor recurrence post‐ surgery [128]. Preclinical studies also suggest that Notch can be a target for cancer therapy, since inhibition of different Notch receptors or Jag1 in the mice after hepatocellular tumorigenesis had occurred, inhibited proliferation, blocked epithelial‐to‐ mesenchymal transition, and resulted in reduced tumor burden and metastization [123]. The Notch pathway also seems to play a key role in the pathogenesis of intrahepatic cholangiocarcinoma (ICC) given that the Notch pathway is a top overexpressed pathway [129], and Jag1 overexpression has been demonstrated in virtually all ICC [130]. ICC can derive from transformation of cholangiocytes [131] or from the transdifferentiation of fully mature hepatocytes [132], both processes being dependent on Notch activation [133]. Interestingly, Jag1‐expressing Kupffer cells stimulate hepatocyte to cholangiocyte transition and hence, cholangiocarcinogenesis [134]. In human ICC, the overexpression of different Notch receptors has been shown to correlate with tumor burden, tumor aggressiveness, and reduced overall survival [135]. There is also increasing evidence linking Notch deregulation with extrahepatic cholangiocarcinoma and gallbladder cancer [136, 137].

Liver repair The Notch pathway is crucial to liver regeneration/repair, regulating biliary repair, progenitor cell‐mediated liver repair, ­vascular remodeling, and fibrogenesis. All Notch receptors are expressed in the liver [138]. Notch1 and ‐2 predominate in cholangiocytes and hepatic progenitor cells (HPC) and increase after biliary damage, whereas Notch3 and ‐4 expression increase after endothelial cell damage [139]. Interestingly, quiescent HSC express Notch1 but activated myofibroblasts downregulate Notch1 and the Notch‐inhibitor Numb, while upregulating Notch2 and ‐3 [73, 140]. The main Notch ligands expressed in the liver are Jag1 on HPC, biliary cells and HSC, and Dll4 on endothelial cells [139]. In the partial hepatectomy animal model of liver regeneration, the Notch pathway is upregulated and promotes biliary and vascular regeneration. After partial hepatectomy, Notch‐RBP/Jk activation on HPC stimulates these cells to become cholangiocytes and blocks their hepatocyte differentiation [141]. This regulation of cell‐fate resembles embryonic morphogenesis, being similarly mediated by RBP/Jk downregulation of YAP (a factor that is known to promote hepatocyte regeneration [142]), suppression of HNF1α and HNF4 (factors that promote hepatocytic differentiation), and upregulation of factors that promote



43:  Developmental Morphogens and Adult Liver Repair

biliary differentiation (e.g. HNF1β and Sox9) [143]. The Notch pathway also promotes hepatocyte regeneration, although not directly through activation of Notch signaling in hepatocytes [144]. Rather, activation of Notch on endothelial cells induces their de‐differentiation, proliferation, vascular remodeling, and release of hepatocyte trophic factors such as Wnt2 and hepatic growth factor (HGF) [145], and profibrogenic factors leading to HSC activation [146]. Furthermore, Notch signaling promotes the homing of bone‐marrow derived endothelial progenitor cells, which are also major sources of trophic factors such as HGF that are essential for hepatocyte proliferation and restoration of liver weight [146, 147]. Regarding wound‐healing responses, HSC respond to biliary damage with upregulation of Jag1, which promotes Notch signaling on HPC and biliary specification to cholangiocytes [148]. On the other hand, after hepatocyte injury, macrophages engulf hepatocyte debris and upregulate Wnt3, which induces Numb to block Notch signaling on HPC and promote specification to hepatocytes [148]. Evidence also suggests that during biliary injury, hepatocytes can transdifferentiate into primitive ductular cells and regenerate the biliary epithelium via a Notch‐dependent process [149]. The Notch pathway promotes fibrogenesis in liver repair, acting synergically with the Hh pathway [73]. Notch promotes Shh signaling through the regulation of transport in and out the PC [150], and Shh promotes Notch signaling through direct upregulation of Hes1 and Jag2 [151]. Both in vitro and in vivo studies showed that the Notch pathway drives an epithelial to mesenchymal‐like transition that transdifferentiates HSC into myofibroblasts, promoting fibrogenesis [73, 152]. In patients with chronic liver disease, activation of Notch correlates with the severity of fibrosis, regardless of etiology of liver disease [153]. Lastly, Notch also enhances inflammatory responses and M1 polarization on macrophages [153].

Liver metabolism There is growing evidence that Notch signaling reconfigures cellular metabolic pathways [154], with consequences for the metabolic syndrome, as well as cancer biology. The effect of Notch on adipocyte metabolism/function seems to be vary according to the receptor. Activation of Notch pathway generally blocks expansion of the adipose tissue, rendering it less able to cope with energy surplus. Notch also impairs adipocyte function leading to insulin resistance, decreased fatty acid oxidation, spill out of fatty acids from adipocytes, and ectopic fat accumulation in hepatocytes [155, 156]. Paradoxically, Notch‐1 activation promotes adipogenesis through upregulation of peroxisome proliferator‐activated receptor (PPAR)‐δ and ‐γ. Interestingly, Notch‐1 also associates with an adipogenic PPARγ dependent quiescent phenotype in HSC [73]. Both obesity/energy surplus, two conditions associated with increased free serum fatty acids, induce Notch pathway activity in hepatocytes [157]. This results in increased hepatic glucose production due to NCID enhancement of forkhead box protein O1 (FOXO1), with consequent transcriptional upregulation of key enzymes of gluconeogenesis (e.g. glucose 6‐phosphatase and phosphoenolpyruvate carboxykinase‐1). Simultaneously, Notch activation promotes mTORC1/ raptor pathway of lipogenesis, upregulating expression of

545

SREBP‐1 regulated lipogenic genes (e.g. acetyl‐CoA carboxylase‐1 and fatty acids synthase). Additionally, Notch inhibits fatty acid oxidation. The aggregate outcomes of increased Notch activity in hepatocytes promote insulin resistance and hepatic steatosis [157]. The Notch pathway can also change metabolism to be cancer cell‐like, modulating mitochondrial function and deregulating glutamine catabolism such that cell growth becomes independent of exogenous glutamine [154].

Liver diseases with notch dysregulation The role of Notch dysregulation on human disease was first described in diseases associated with ductal plate malformations such as Alagille syndrome and biliary atresia. However, recent evidence has linked Notch to other chronic liver diseases. Adult primary cholangiopathies differ with regards to Notch signaling. For example, PBC has more striking Notch upregulation and less Wnt activation than PSC. The ductular reaction also differs in these two diseases: it tends to be exuberant earlier in the course of PBC and associates with increased expression of ductal markers (Sox9 and K‐19) but reduced expression of hepatocyte markers [158]. NAFLD animal models show that Notch signaling upregulation contributes to obesity‐induced hepatic steatosis [157]. In human steatohepatitis, Notch pathway activation not only correlates with severity of steatosis, but also with severity of liver disease (necroinflammatory activity and aminotransferases levels) [159]. Further, Notch‐3 expression associates with ductular reaction and Notch‐4 with sinusoidal neovessel proliferation and Kupffer cell activation [160].

SUMMARY Hedgehog and Notch are among the morphogenic signaling pathways that are typically inactivated once liver development finishes. Both pathways reactivate in adulthood in response to liver injury so that healthy liver tissue can be regenerated. Because these pathways have pleiotropic actions that control cell fate, they are tightly regulated. Dysfunction of these regulatory mechanisms can lead to either insufficient, or excessive, pathway activity – both result in progressive liver damage. For example, hypoactivation of Notch signaling can impair biliary regeneration and thereby promote progressive ductopenia and/or ductal sclerosis, as occurs in certain primary cholangiopathies. Conversely, hyperactivation of the Hh pathway promotes progressive liver fibrosis in many chronic liver diseases, as well as aggressive cancer biology in different types of primary liver cancer. Hence, both Hedgehog and Notch are emerging as targets for biomarker and therapeutic development in adult liver diseases.

REFERENCES   1. Roy, S. Cilia and Hedgehog: when and how was their marriage solemnized? Differentiation, 2012;83(2):S43–8.   2. Liu, A., Wang, B., and Niswander, L.A. Mouse intraflagellar transport proteins regulate both the activator and repressor functions of Gli transcription factors. Development, 2005;132(13):3103–11.

546

THE LIVER:  REFERENCES

3. Chen, Y., Sasai, N., Ma, G. et al. Sonic Hedgehog dependent phosphorylation by CK1alpha and GRK2 is required for ciliary accumulation and activation of smoothened. PLoS Biol, 2011;9(6):e1001083. 4. Porter, J.A., Young, K.E., and Beachy, P.A. Cholesterol modification of hedgehog signaling proteins in animal development. Science, 1996;274(5285):255–9. 5. Pepinsky, R.B., Zeng, C., Wen, D. et  al. Identification of a palmitic acid‐ modified form of human Sonic hedgehog. J Biol Chem, 1998;273(22): 14037–45. 6. Briscoe, J. and Therond, P.P. The mechanisms of Hedgehog signalling and  its roles in development and disease. Nat Rev Mol Cell Biol, 2013;14(7):416–29. 7. Ayers, K.L., Gallet, A., Staccini‐Lavenant, L., and Therond, P.P. The long‐ range activity of Hedgehog is regulated in the apical extracellular space by the glypican Dally and the hydrolase Notum. Dev Cell, 2010;18(4):605–20. 8. Merchant, J.L. and Saqui‐Salces, M. Inhibition of Hedgehog signaling in the gastrointestinal tract: targeting the cancer microenvironment. Cancer Treat Rev, 2014;40(1):12–21. 9. Xiao, X., Tang, J.J., Peng, C. et al. Cholesterol Modification of Smoothened is required for Hedgehog signaling. Mol Cell,Cell, 2017;66(1):154–62. 10. Denef, N., Neubuser, D., Perez, L., and Cohen, S.M. Hedgehog induces opposite changes in turnover and subcellular localization of patched and smoothened. Cell, 2000;102(4):521–31. 11. Izzi, L., Levesque, M., Morin, S. et  al. Boc and Gas1 each form distinct Shh receptor complexes with Ptch1 and are required for Shh‐mediated cell proliferation. Dev Cell, 2011;20(6):788–801. 12. Chuang, P.T. and McMahon, A.P. Vertebrate Hedgehog signalling modulated by induction of a Hedgehog‐binding protein. Nature, 1999;397(6720):617–21. 13. Teperino, R., Aberger, F., Esterbauer, H., Riobo, N., and Pospisilik, J.A. Canonical and non‐canonical Hedgehog signalling and the control of metabolism. Semin Cell Dev Biol, 2014.;33:81–92. 14. Wang, B. and Li, Y. Evidence for the direct involvement of {beta}TrCP in Gli3 protein processing. Proc Nat Acad Sci USA, 2006;103(1):33–8. 15. Ikram, M.S., Neill, G.W., Regl, G., Eichberger, T., Frischauf, A.M., Aberger, F. et al. GLI2 is expressed in normal human epidermis and BCC and induces GLI1 expression by binding to its promoter. J Invest Dermatol, 2004;122(6):1503–9. 16. Pan, Y. and Wang, B. A novel protein‐processing domain in Gli2 and Gli3 differentially blocks complete protein degradation by the proteasome. J Biol Chem, 2007;282(15):10846–52. 17. Matz‐Soja, M., Rennert, C., Schonefeld, K. et al. Hedgehog signaling is a potent regulator of liver lipid metabolism and reveals a GLI‐code associated with steatosis. Elife, 2016;5. 18. Chinchilla, P., Xiao, L., Kazanietz, M.G., and Riobo, N.A. Hedgehog proteins activate pro‐angiogenic responses in endothelial cells through non‐ canonical signaling pathways. Cell Cycle, 2010;9(3):570–79. 19. Barnes, E.A., Kong, M., Ollendorff, V., and Donoghue, D.J. Patched1 interacts with cyclin B1 to regulate cell cycle progression. EMBO J, ­ 2001;20(9):2214–23. 20. Yuan, X., Cao, J., He, X. et  al. Ciliary IFT80 balances canonical versus non‐canonical hedgehog signalling for osteoblast differentiation. Nat Commun, 2016;7:11024. 21. Teperino, R., Amann, S., Bayer, M. et  al. Hedgehog partial agonism drives  Warburg‐like metabolism in muscle and brown fat. Cell, 2012;151(2):414–26. 22. Chen, Y., Choi, S.S., Michelotti, G.A. et al. Hedgehog controls hepatic stellate cell fate by regulating metabolism. Gastroenterology, 2012;143(5):1319– 29 e1–11. 23. Polizio, A.H., Chinchilla, P., Chen, X., Manning, D.R., and Riobo, N.A. Sonic Hedgehog activates the GTPases Rac1 and RhoA in a Gli‐independent manner through coupling of smoothened to Gi proteins. Sci Signal, 2011;4(200):pt7. 24. Bijlsma, M.F., Borensztajn, K.S., Roelink, H., Peppelenbosch, M.P., and Spek, C.A. Sonic hedgehog induces transcription‐independent cytoskeletal rearrangement and migration regulated by arachidonate metabolites. Cell Signal, 2007;19(12):2596–604. 25. Dennler, S., Andre, J., Verrecchia, F., and Mauviel, A. Cloning of the human GLI2 Promoter: transcriptional activation by transforming growth factor‐beta via SMAD3/beta‐catenin cooperation. J Biol Chem, 2009;284(46):31523–31. 26. Nolan‐Stevaux, O., Lau, J., Truitt, M.L. et  al. GLI1 is regulated through Smoothened‐independent mechanisms in neoplastic pancreatic ducts and mediates PDAC cell survival and transformation. Genes Dev, 2009;23(1):24–36.

27. Zhao, R. and Duncan, S.A. Embryonic development of the liver. Hepatology, 2005;41(5):956–67. 28. Hirose, Y., Itoh, T., and Miyajima, A. Hedgehog signal activation coordinates proliferation and differentiation of fetal liver progenitor cells. Exp Cell Res, 2009;315(15):2648–57. 29. Higashiyama, H., Ozawa, A., Sumitomo, H. et al. Embryonic cholecystitis and defective gallbladder contraction in the Sox17‐haploinsufficient mouse model of biliary atresia. Development, 2017;144(10):1906–17. 30. Sicklick, J.K., Li, Y.X., Jayaraman, A. et al. Dysregulation of the Hedgehog pathway in human hepatocarcinogenesis. Carcinogenesis, 2006;27(4):748–57. 31. Della Corte, C.M., Viscardi, G., Papaccio, F. et  al. Implication of the Hedgehog pathway in hepatocellular carcinoma. World J Gastroenterol, 2017;23(24):4330–40. 32. Jinawath, A., Akiyama, Y., Sripa, B., and Yuasa, Y. Dual blockade of the Hedgehog and ERK1/2 pathways coordinately decreases proliferation and survival of cholangiocarcinoma cells. J Cancer Res Clin Oncol,2007; 133(4):271–8. 33. Li, Y.C., Deng, Y.H., Guo, Z.H. et al. Prognostic value of hedgehog signal component expressions in hepatoblastoma patients. Eur J Med Res, 2010;15(11):468–74. 34. Tada, M., Kanai, F., Tanaka, Y. et al. Down‐regulation of hedgehog‐interacting protein through genetic and epigenetic alterations in human hepatocellular carcinoma. Clin Cancer Res, 2008;14(12):3768–76. 35. Chan, I.S., Guy, C.D., Chen, Y. et  al. Paracrine Hedgehog signaling drives  metabolic changes in hepatocellular carcinoma. Cancer Res, 2012;72(24):6344–50. 36. Wan, S., Zhao, E., Kryczek, I. et al. Tumor‐associated macrophages produce interleukin 6 and signal via STAT3 to promote expansion of human hepatocellular carcinoma stem cells. Gastroenterology, 2014;147(6):1393–404. 37. Li, W., Miao, S., Miao, M. et al. Hedgehog signaling activation in hepatic stellate cells promotes angiogenesis and vascular mimicry in hepatocellular carcinoma. Cancer Invest. 2016;34(9):424–30. 38. Olive, K.P., Jacobetz, M.A., Davidson, C.J. et  al. Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science, 2009;324(5933):1457–61. 39. Gan, G.N. and Jimeno, A. Emerging from their burrow: Hedgehog pathway inhibitors for cancer. Expert Opin Investig Drugs, 2016;25(10):1153–66. 40. Nault, J.C., Couchy, G., Balabaud, C. et al. Molecular classification of hepatocellular adenoma associates with risk factors, bleeding, and malignant transformation. Gastroenterology, 2017;152(4):880–94. 41. Nault, J.C., Couchy, G., Caruso, S. et al. Argininosuccinate synthase 1 and periportal gene expression in sonic hedgehog hepatocellular adenomas. Hepatology, 2018;68(3):964–76. 42. Schmidt‐Heck, W., Matz‐Soja, M., Aleithe, S., Marbach, E., Guthke, R., and Gebhardt, R. Fuzzy modeling reveals a dynamic self‐sustaining network of the GLI transcription factors controlling important metabolic regulators in adult mouse hepatocytes. Mol Biosyst, 2015;11(8):2190–7. 43. Gebhardt, R. and Coffer, P.J. Hepatic autophagy is differentially regulated in periportal and pericentral zones ‐ a general mechanism relevant for other tissues? Cell Commun Signal, 2013;11(1):21. 44. van den Brink, G.R., Bleuming, S.A., Hardwick, J.C. et al. Indian Hedgehog is an antagonist of Wnt signaling in colonic epithelial cell differentiation. Nat Genet, 2004;36(3):277–82. 45. Lin, S., Nascimento, E.M., Gajera, C.R. et  al. Distributed hepatocytes expressing telomerase repopulate the liver in homeostasis and injury. Nature, 2018;556(7700):244–8. 46. Machado, M.V. and Diehl, A.M. Liver renewal: detecting misrepair and optimizing regeneration. Mayo Clin Proc, 2014;89(1):120–30. 47. Machado, M.V. and Diehl, A.M. Hedgehog signalling in liver pathophysiology. J Hepatol, 2018;68(3):550–62. 48. Fleig, S.V., Choi, S.S., Yang, L. et al. Hepatic accumulation of Hedgehog‐ reactive progenitors increases with severity of fatty liver damage in mice. Lab Invest, 2007;87(12):1227–39. 49. Rangwala, F., Guy, C.D., Lu, J. et al. Increased production of sonic hedgehog by ballooned hepatocytes. J Pathol, 2011;224(3):401–10. 50. Machado, M.V., Michelotti, G.A., Pereira, T.A. et al. Reduced lipoapoptosis, hedgehog pathway activation and fibrosis in caspase‐2 deficient mice with non‐alcoholic steatohepatitis. Gut, 2015;64(7):1148–57. 51. Witek, R.P., Yang, L., Liu, R. et al. Liver cell‐derived microparticles activate hedgehog signaling and alter gene expression in hepatic endothelial cells. Gastroenterology, 2009;136(1):320–30 e2.



43:  Developmental Morphogens and Adult Liver Repair

52. Kopinke, D., Roberson, E.C., and Reiter, J.F. Ciliary Hedgehog signaling restricts injury‐induced adipogenesis. Cell, 2017;170(2):340–51. 53. Byrd, N. and Grabel, L. Hedgehog signaling in murine vasculogenesis and angiogenesis. Trends Cardiovasc Med, 2004;14(8):308–13. 54. Syn, W.K., Witek, R.P., Curbishley, S.M. et al. Role for hedgehog pathway in regulating growth and function of invariant NKT cells. Eur J Immunol, 2009;39(7):1879–92. 55. Pereira, T.A., Xie, G., Choi, S.S. et  al. Macrophage‐derived Hedgehog ligands promotes fibrogenic and angiogenic responses in human schistosomiasis mansoni. Liver Int, 2013;33(1):149–61. 56. Michelotti, G.A., Xie, G., Swiderska, M. et al. Smoothened is a master regulator of adult liver repair. J Clin Invest, 2013;123(6):2380–94. 57. Fingas, C.D., Bronk, S.F., Werneburg, N.W. et  al. Myofibroblast‐derived PDGF‐BB promotes Hedgehog survival signaling in cholangiocarcinoma cells. Hepatology, 2011;54(6):2076–88. 58. Kumar, V., Mundra, V., and Mahato, R.I. Nanomedicines of Hedgehog inhibitor and PPAR‐gamma agonist for treating liver fibrosis. Pharm Res, 2014;31(5):1158–69. 59. El‐Agroudy, N.N., El‐Naga, R.N., El‐Razeq, R.A., and El‐Demerdash, E. Forskolin, a hedgehog signalling inhibitor, attenuates carbon tetrachloride‐ induced liver fibrosis in rats. Br J Pharmacol, 2016;173(22):3248–60. 60. Chung, S.I., Moon, H., Ju, H.L. et al. Hepatic expression of Sonic Hedgehog induces liver fibrosis and promotes hepatocarcinogenesis in a transgenic mouse model. J Hepatol, 2016;64(3):618–27. 61. Xie, G., Choi, S.S., Syn, W.K. et  al. Hedgehog signalling regulates liver sinusoidal endothelial cell capillarisation. Gut, 2013;62(2):299–309. 62. Uschner, F.E., Ranabhat, G., Choi, S.S. et al. Statins activate the canonical hedgehog‐signaling and aggravate non‐cirrhotic portal hypertension, but inhibit the non‐canonical hedgehog signaling and cirrhotic portal hypertension. Sci Rep, 2015;5:14573. 63. Omenetti, A., Syn, W.K., Jung, Y. et al. Repair‐related activation of hedgehog signaling promotes cholangiocyte chemokine production. Hepatology, 2009;50(2):518–27. 64. Xie, J. The hedgehog’s trick for escaping immunosurveillance: The molecular mechanisms driving myeloid‐derived suppressor cell recruitment in hedgehog signaling‐dependent tumors. Oncoimmunology, 2014;3:e29180. 65. Omenetti, A., Porrello, A., Jung, Y. et al. Hedgehog signaling regulates epithelial‐mesenchymal transition during biliary fibrosis in rodents and humans. J Clin Invest, 2008;118(10):3331–42. 66. Omenetti, A., Popov, Y., Jung, Y. et  al. The hedgehog pathway regulates remodelling responses to biliary obstruction in rats. Gut, 2008;57(9): 1275–82. 67. Fleury, A., Hoch, L., Martinez, M.C. et al. Hedgehog associated to microparticles inhibits adipocyte differentiation via a non‐canonical pathway. Sci Rep, 2016;6:23479. 68. Du, K., Hyun, J., Premont, R.T. et  al. Hedgehog‐YAP signaling pathway regulates glutaminolysis to control activation of hepatic stellate cells. Gastroenterology, 2018;154(5):1465–79. 69. Fu, X., Zhu, M.J., Dodson, M.V., and Du, M. AMP‐activated protein kinase stimulates Warburg‐like glycolysis and activation of satellite cells during muscle regeneration. J Biol Chem, 2015;290(44):26445–56. 70. Song, M., Han, L., Chen, F.F. et al. Adipocyte‐derived exosomes carrying Sonic Hedgehog mediate M1 macrophage polarization‐induced insulin resistance via Ptch and PI3K pathways. Cell Physiol Biochem, 2018;48(4): 1416–32. 71. Bhatia, B., Hsieh, M., Kenney, A.M., and Nahle, Z. Mitogenic Sonic hedgehog signaling drives E2F1‐dependent lipogenesis in progenitor cells and medulloblastoma. Oncogene, 2011;30(4):410–22. 72. Yao, P.J., Manor, U., Petralia, R.S. et al. Sonic hedgehog pathway activation increases mitochondrial abundance and activity in hippocampal neurons. Mol Biol Cell, 2017;28(3):387–95. 73. Xie, G., Karaca, G., Swiderska‐Syn, M. et  al. Cross‐talk between Notch and  Hedgehog regulates hepatic stellate cell fate in mice. Hepatology, 2013;58(5):1801–13. 74. Gonzalez‐Mendez, L., Seijo‐Barandiaran, I., and Guerrero, I. Cytoneme‐ mediated cell‐cell contacts for Hedgehog reception. Elife, 2017;6. 75. Panakova, D., Sprong, H., Marois, E., Thiele, C., and Eaton, S. Lipoprotein particles are required for Hedgehog and Wingless signalling. Nature, 2005;435(7038):58–65. 76. Huang, P., Nedelcu, D., Watanabe, M. et al. Cellular cholesterol directly activates Smoothened in Hedgehog signaling. Cell, 2016;166(5):1176–87 e14.

547

77. Murone, M., Rosenthal, A., and de Sauvage, F.J. Sonic hedgehog signaling by the patched‐smoothened receptor complex. Curr Biol, 1999;9(2):76–84. 78. Jiang, X.L., Chen, T., and Zhang, X. Activation of sonic hedgehog signaling attenuates oxidized low‐density lipoprotein‐stimulated brain microvascular endothelial cells dysfunction in vitro. Int J Clin Exp Pathol, 2015;8(10): 12820–8. 79. Jeon, S. and Seong, R.H. Anteroposterior limb skeletal patterning requires the bifunctional action of SWI/SNF chromatin remodeling complex in Hedgehog pathway. PLoS Genet, 2016;12(3):e1005915. 80. Guy, C.D., Suzuki, A., Zdanowicz, M. et al. Hedgehog pathway activation parallels histologic severity of injury and fibrosis in human nonalcoholic fatty liver disease. Hepatology, 2012;55(6):1711–21. 81. Jung, Y., Brown, K.D., Witek, R.P. et  al. Accumulation of hedgehog‐ responsive progenitors parallels alcoholic liver disease severity in mice and humans. Gastroenterology, 2008;134(5):1532–43. 82. Pereira Tde, A., Witek, R.P., Syn, W.K. et al. Viral factors induce Hedgehog pathway activation in humans with viral hepatitis, cirrhosis, and hepatocellular carcinoma. Lab Invest, 2010;90(12):1690–703. 83. Jung, Y., McCall, S.J., Li, Y.X., and Diehl, A.M. Bile ductules and stromal cells express hedgehog ligands and/or hedgehog target genes in primary biliary cirrhosis. Hepatology, 2007;45(5):1091–6. 84. Carpino, G., Cardinale, V., Renzi, A. et al. Activation of biliary tree stem cells within peribiliary glands in primary sclerosing cholangitis. J Hepatol, 2015;63(5):1220–8. 85. Jung, H.Y., Jing, J., Lee, K.B., and Jang, J.J. Sonic hedgehog (SHH) and glioblastoma‐2 (Gli‐2) expressions are associated with poor jaundice‐free survival in biliary atresia. J Pediatr Surg, 2015;50(3):371–6. 86. Syn, W.K., Jung, Y., Omenetti, A. et al. Hedgehog‐mediated epithelial‐to‐ mesenchymal transition and fibrogenic repair in nonalcoholic fatty liver disease. Gastroenterology, 2009;137(4):1478–88 e8. 87. Hirsova, P., Ibrahim, S.H., Bronk, S.F., Yagita, H., and Gores, G.J. Vismodegib suppresses TRAIL‐mediated liver injury in a mouse model of nonalcoholic steatohepatitis. PloS One, 2013;8(7):e70599. 88. de Almeida Pereira, T., Borthwick, L., Xie, G. et  al. Crosstalk between IL13 and Hedgehog pathways contributes to Schistosomiasis mansoni fibrosis. J Hepatol, 2016;62(2):S204–S. 89. Tang, V., Cofer, Z.C., Cui, S., Sapp, V., Loomes, K.M., and Matthews, R.P. Loss of a candidate biliary atresia susceptibility gene, add3a, causes biliary developmental defects in zebrafish. J Pediatr Gastroenterol Nutr, 2016;63(5): 524–30. 90. Ali, O., Cerjak, D., Kent, J.W., Jr. et al. An epigenetic map of age‐associated autosomal loci in northern European families at high risk for the metabolic syndrome. Clin Epigenetics, 2015;7:12. 91. Sacoto, M.J.G., Martinez, A.F., Abe, Y. et al. Human germline Hedgehog pathway mutations predispose to fatty liver. J Hepatol, 2017;67(4):809–17. 92. Swiderska‐Syn, M., Syn, W.K., Xie, G. et al. Myofibroblastic cells function as progenitors to regenerate murine livers after partial hepatectomy. Gut, 2014;63(8):1333–44. 93. Siebel, C. and Lendahl, U. Notch Signaling in development, tissue homeostasis, and disease. Physiol Rev, 2017;97(4):1235–94. 94. Meurette, O. and Mehlen, P. Notch signaling in the tumor microenvironment. Cancer Cell, 2018;34(4):536–48. 95. Jacobsen, T.L., Brennan, K., Arias, A.M., and Muskavitch, M.A. Cis‐interactions between Delta and Notch modulate neurogenic signalling in Drosophila. Development, 1998;125(22):4531–40. 96. Van de Walle, I., Waegemans, E., De Medts, J. et  al. Specific Notch receptor‐ligand interactions control human TCR‐alphabeta/gammadelta ­ development by inducing differential Notch signal strength. J Exp Med, 2013;210(4):683–97. 97. Hein, K., Mittler, G., Cizelsky, W. et al. Site‐specific methylation of Notch1 controls the amplitude and duration of the Notch1 response. Sci Signal, 2015;8(369):ra30. 98. Coleman, M.L., McDonough, M.A., Hewitson, K.S. et  al. Asparaginyl hydroxylation of the Notch ankyrin repeat domain by factor inhibiting hypoxia‐inducible factor. J Biol Chem, 2007;282(33):24027–38. 99. Guarani, V., Deflorian, G., Franco, C.A. et  al. Acetylation‐dependent regulation of endothelial Notch signalling by the SIRT1 deacetylase. ­ Nature, 2011;473(7346):234–8. 100. Foltz, D.R., Santiago, M.C., Berechid, B.E., and Nye, J.S. Glycogen synthase kinase‐3beta modulates notch signaling and stability. Curr Biol, 2002;12(12):1006–11.

548

THE LIVER:  REFERENCES

101. Kovall, R.A., Gebelein, B., Sprinzak, D., and Kopan, R. The canonical Notch signaling pathway: structural and biochemical insights into shape, sugar, and force. Dev Cell, 2017;41(3):228–41. 102. Gordon, W.R., Zimmerman, B., He, L. et al. Mechanical allostery: evidence for a force requirement in the proteolytic activation of Notch. Dev Cell, 2015;33(6):729–36. 103. Couturier, L., Mazouni, K., and Schweisguth, F. Numb localizes at endosomes and controls the endosomal sorting of notch after asymmetric division in Drosophila. Curr Biol, 2013;23(7):588–93. 104. Sjoqvist, M., Antfolk, D., Ferraris, S. et al. PKCzeta regulates Notch receptor routing and activity in a Notch signaling‐dependent manner. Cell Res, 2014;24(4):433–50. 105. Leitch, C.C., Lodh, S., Prieto‐Echague, V., Badano, J.L., and Zaghloul, N.A. Basal body proteins regulate Notch signaling through endosomal trafficking. J Cell Sci, 2014;127(11):2407–19. 106. Nam, Y., Weng, A.P., Aster, J.C., and Blacklow, S.C. Structural requirements for assembly of the CS.L.intracellular Notch1.Mastermind‐like 1 transcriptional activation complex. J Biol Chem, 2003;278(23):21232–9. 107. Miyamoto, A., Lau, R., Hein, P.W., Shipley, J.M., and Weinmaster, G. Microfibrillar proteins MAGP‐1 and MAGP‐2 induce Notch1 extracellular domain dissociation and receptor activation. J Biol Chem, 2006;281(15): 10089–97. 108. Traustadottir, G.A., Jensen, C.H., Thomassen, M. et al. Evidence of non‐ canonical NOTCH signaling: delta‐like 1 homolog (DLK1) directly interacts with the NOTCH1 receptor in mammals. Cell Signal, 2016;28(4): 246–54. 109. Jin, S., Mutvei, A.P., Chivukula, I.V. et al. Non‐canonical Notch signaling activates IL‐6/JAK/STAT signaling in breast tumor cells and is controlled by p53 and IKKalpha/IKKbeta. Oncogene. 2013;32(41):4892–902. 110. Zimber‐Strobl, U. and Strobl, L.J. EBNA2 and Notch signalling in Epstein‐ Barr virus mediated immortalization of B lymphocytes. Semin Cancer Biol, 2001;11(6):423–34. 111. Zong, Y., Panikkar, A., Xu, J. et al. Notch signaling controls liver development by regulating biliary differentiation. Development, 2009;136(10): 1727–39. 112. Oda, T., Elkahloun, A.G., Pike, B.L. et al. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat Genet, 1997;16(3):235–42. 113. McDaniell, R., Warthen, D.M., Sanchez‐Lara, P.A. et al. NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. Am J Hum Genet, 2006;79(1):169–73. 114. McCright, B., Lozier, J., and Gridley, T. A mouse model of Alagille syndrome: Notch2 as a genetic modifier of Jag1 haploinsufficiency. Development, 2002;129(4):1075–82. 115. Andersson, E.R., Chivukula, I.V., Hankeova, S. et  al. Mouse model of Alagille syndrome and mechanisms of Jagged1 missense mutations. Gastroenterology, 2018;154(4):1080–95. 116. Kohsaka, T., Yuan, Z.R., Guo, S.X. et al. The significance of human jagged 1 mutations detected in severe cases of extrahepatic biliary atresia. Hepatology, 2002;36(4 Pt 1):904–12. 117. Lemaigre, F. and Zaret, K.S. Liver development update: new embryo models, cell lineage control, and morphogenesis. Curr Opin Genet Dev, 2004;14(5):582–90. 118. Furubo, S., Sato, Y., Harada, K., and Nakanuma, Y. Roles of myofibroblasts and notch and hedgehog signaling pathways in the formation of intrahepatic bile duct lesions in polycystic kidney rats. Pediatr Dev Pathol, 2013;16(3):177–90. 119. Strazzabosco, M. and Fabris, L. Notch signaling in hepatocellular carcinoma: guilty in association! Gastroenterology, 2012;143(6):1430–4. 120. Wang, W., Feng, Y., Aimaiti, Y., Jin, X., Mao, X., and Li, D. TGFbeta signaling controls intrahepatic bile duct development may through ­ regulating the Jagged1‐Notch‐Sox9 signaling axis. J Cell Physiol, ­ 2018;233(8):5780–91. 121. So, J., Khaliq, M., Evason, K. et  al. Wnt/beta‐catenin signaling controls intrahepatic biliary network formation in zebrafish by regulating notch activity. Hepatology, 2018;67(6):2352–66. 122. Wu, N., Nguyen, Q., Wan, Y. et al. The Hippo signaling functions through the Notch signaling to regulate intrahepatic bile duct development in mammals. Lab Invest, 2017;97(7):843–53. 123. Huntzicker, E.G., Hotzel, K., Choy, L. et al. Differential effects of targeting Notch receptors in a mouse model of liver cancer. Hepatology, 2015;61(3):942–52.

124. Villanueva, A., Alsinet, C., Yanger, K. et al. Notch signaling is activated in human hepatocellular carcinoma and induces tumor formation in mice. Gastroenterology, 2012;143(6):1660–9. 125. Xiong, S., Wang, R., Chen, Q. et al. Cancer‐associated fibroblasts promote stem cell‐like properties of hepatocellular carcinoma cells through IL‐6/ STAT3/Notch signaling. Am J Cancer Res, 2018;8(2):302–16. 126. Giovannini, C., Gramantieri, L., Chieco, P. et  al. Selective ablation of Notch3 in HCC enhances doxorubicin’s death promoting effect by a p53 dependent mechanism. J Hepatol, 2009;50(5):969–79. 127. Wu, T., Jiao, M., Jing, L. et al. Prognostic value of Notch‐1 expression in hepatocellular carcinoma: a meta‐analysis. Onco Targets Ther, 2015;8:3105–14. 128. Yu, T., Han, C., Zhu, G. et al. Prognostic value of Notch receptors in postsurgical patients with hepatitis B virus‐related hepatocellular carcinoma. Cancer Med, 2017;6(7):1587–600. 129. Xue, T.C., Zhang, B.H., Ye, S.L., and Ren, Z.G. Differentially expressed gene profiles of intrahepatic cholangiocarcinoma, hepatocellular carcinoma, and combined hepatocellular‐cholangiocarcinoma by integrated microarray analysis. Tumour Biol, 2015;36(8):5891–9. 130. Che, L., Fan, B., Pilo, M.G. et al. Jagged 1 is a major Notch ligand along cholangiocarcinoma development in mice and humans. Oncogenesis, 2016;5(12):e274. 131. Guest, R.V., Boulter, L., Kendall, T.J. et al. Cell lineage tracing reveals a biliary origin of intrahepatic cholangiocarcinoma. Cancer Res, 2014; 74(4):1005–10. 132. Fan, B., Malato, Y., Calvisi, D.F. et al. Cholangiocarcinomas can originate from hepatocytes in mice. J Clin Invest, 2012;122(8):2911–5. 133. Wang, J., Dong, M., Xu, Z. et al. Notch2 controls hepatocyte‐derived cholangiocarcinoma formation in mice. Oncogene, 2018;37(24):3229–42. 134. Terada, M., Horisawa, K., Miura, S. et al. Kupffer cells induce Notch‐mediated hepatocyte conversion in a common mouse model of intrahepatic cholangiocarcinoma. Sci Rep, 2016;6:34691. 135. Wu, W.R., Shi, X.D., Zhang, R. et al. Clinicopathological significance of aberrant Notch receptors in intrahepatic cholangiocarcinoma. Int J Clin Exp Pathol, 2014;7(6):3272–9. 136. Aoki, S., Mizuma, M., Takahashi, Y. et  al. Aberrant activation of Notch signaling in extrahepatic cholangiocarcinoma: clinicopathological features and therapeutic potential for cancer stem cell‐like properties. BMC Cancer, 2016;16(1):854. 137. Yoon, H.A., Noh, M.H., Kim, B.G. et al. Clinicopathological significance of altered Notch signaling in extrahepatic cholangiocarcinoma and gallbladder carcinoma. World J Gastroenterol, 2011;17(35):4023–30. 138. Nijjar, S.S., Crosby, H.A., Wallace, L., Hubscher, S.G., and Strain, A.J. Notch receptor expression in adult human liver: a possible role in bile duct formation and hepatic neovascularization. Hepatology, 2001;34(6):1184–92. 139. Morell, C.M., Fiorotto, R., Fabris, L., and Strazzabosco, M. Notch signalling beyond liver development: emerging concepts in liver repair and oncogenesis. Clin Res Hepatol Gastroenterol, 2013;37(5):447–54. 140. Sawitza, I., Kordes, C., Reister, S., and Haussinger, D. The niche of stellate cells within rat liver. Hepatology, 2009;50(5):1617–24. 141. Lu, J., Zhou, Y., Hu, T. et al. Notch signaling coordinates progenitor cell‐ mediated biliary regeneration following partial hepatectomy. Sci Rep, 2016;6:22754. 142. Yimlamai, D., Christodoulou, C., Galli, G.G. et al. Hippo pathway activity influences liver cell fate. Cell, 2014;157(6):1324–38. 143. Fabris, L., Cadamuro, M., Guido, M. et al. Analysis of liver repair mechanisms in Alagille syndrome and biliary atresia reveals a role for notch signaling. Am J Pathol, 2007;171(2):641–53. 144. Dill, M.T., Rothweiler, S., Djonov, V. et al. Disruption of Notch1 induces vascular remodeling, intussusceptive angiogenesis, and angiosarcomas in livers of mice. Gastroenterology, 2012;142(4):967–77 e2. 145. Ding, B.S., Nolan, D.J., Butler, J.M. et  al. Inductive angiocrine signals from sinusoidal endothelium are required for liver regeneration. Nature, 2010;468(7321):310–5. 146. DeLeve, L.D. Liver sinusoidal endothelial cells and liver regeneration. J Clin Invest, 2013;123(5):1861–6. 147. Wang, L., Wang, Y.C., Hu, X.B. et al. Notch‐RBP‐J signaling regulates the mobilization and function of endothelial progenitor cells by dynamic modulation of CXCR4 expression in mice. PloS One, 2009;4(10):e7572. 148. Boulter, L., Govaere, O., Bird, T.G. et al. Macrophage‐derived Wnt opposes Notch signaling to specify hepatic progenitor cell fate in chronic liver disease. Nat Med, 2012;18(4):572–9.



43:  Developmental Morphogens and Adult Liver Repair

149. Jeliazkova, P., Jors, S., Lee, M. et  al. Canonical Notch2 signaling determines biliary cell fates of embryonic hepatoblasts and adult hepatocytes independent of Hes1. Hepatology, 2013;57(6):2469–79. 150. Ingram, W.J., McCue, K.I., Tran, T.H., Hallahan, A.R., and Wainwright, B.J. Sonic Hedgehog regulates Hes1 through a novel mechanism that is independent of canonical Notch pathway signalling. Oncogene, 2008;27(10):1489–500. 151. Stasiulewicz, M., Gray, S.D., Mastromina, I. et  al. A conserved role for  Notch signaling in priming the cellular response to Shh through ciliary  localisation of the key Shh transducer Smo. Development, 2015;142(13):2291–303. 152. Chen, Y., Zheng, S., Qi, D. et al. Inhibition of Notch signaling by a gamma‐ secretase inhibitor attenuates hepatic fibrosis in rats. PloS One, 2012;7(10):e46512. 153. He, F., Guo, F.C., Li, Z. et al. Myeloid‐specific disruption of recombination signal binding protein Jkappa ameliorates hepatic fibrosis by attenuating inflammation through cylindromatosis in mice. Hepatology, 2015;61(1):303–14. 154. Basak, N.P., Roy, A., and Banerjee, S. Alteration of mitochondrial proteome due to activation of Notch1 signaling pathway. J Biol Chem, 2014;289(11):7320–34.

549

155. Chartoumpekis, D.V., Palliyaguru, D.L., Wakabayashi, N. et  al. Notch intracellular domain overexpression in adipocytes confers lipodystrophy in mice. Mol Metab, 2015;4(7):543–50. 156. Song, N.J., Yun, U.J., Yang, S. et al. Notch1 deficiency decreases hepatic lipid accumulation by induction of fatty acid oxidation. Sci Rep, 2016;6:19377. 157. Pajvani, U.B., Qiang, L., Kangsamaksin, T., Kitajewski, J., Ginsberg, H.N., and Accili, D. Inhibition of Notch uncouples Akt activation from hepatic  lipid accumulation by decreasing mTorc1 stability. Nat Med, 2013;19(8):1054–60. 158. Carpino, G., Cardinale, V., Folseraas, T. et al. Hepatic stem/progenitor cell activation differs between primary sclerosing and primary biliary cholangitis. Am J Pathol, 2018;188(3):627–39. 159. Valenti, L., Mendoza, R.M., Rametta, R. et al. Hepatic notch signaling correlates with insulin resistance and nonalcoholic fatty liver disease. Diabetes, 2013;62(12):4052–62. 160. Liew, P.L., Wang, W., Lee, Y.C., Huang, M.T., and Lee, W.J. Roles of  hepatic progenitor cells activation, ductular reaction proliferation and  Notch signaling in morbid obesity. Hepatogastroenterology, 2012; 59(118):1921–7.

44

Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology David A. Shafritz1 and Markus Grompe2 Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY, USA Papé Pediatric Research Institute, Oregon Health Sciences University, Portland, OR, USA

1 2

INTRODUCTION The major impetus for trying to reconstitute the liver by cell transplantation is based on the very high regenerative capacity of this organ. The very first experiments were done in the early 1900s, when liver fragments were transplanted into the anterior chamber of the eye, but the transplanted liver tissue degenerated rapidly and disappeared within a few days [1]. The first successful report of liver cell transplantation was in the 1970s, when isolated hepatocytes were transplanted to the liver, leading to transient reduction in serum bilirubin in the Gunn rat model for Crigler–Najjar syndrome, type 1 [2]. More recently, substantial progress has been made defining the cell types that can repopulate the liver and restore function, and the host liver conditions under which effective repopulation can be achieved. Both adult hepatocytes and hepatic stem‐like progenitor cells (stem/progenitor cells), as well as stem and progenitor cells of non‐hepatic origin, have been explored for transplantation into the liver. Studies concerning hepatocyte regeneration during both homeostasis and injury, including cell plasticity and interconversion between hepatocytes and bile duct epithelial cells and the current status of hepatic cell transplantation in humans will be covered, as well as expectations for the future.

HEPATOCYTE TRANSPLANTATION: RATIONALE AND EARLY STUDIES Currently, orthotopic liver transplantation (OLT) is the only effective method available to cure acquired and inherited hepatic disorders, especially when these diseases reach their end stages [3, 4]. However, the number of patients who are fortunate enough to receive a liver transplant is limited by donor organ availability. Liver transplantation is also expensive, carries significant morbidity and mortality, and requires long‐term immunosuppression. Hepatocyte transplantation has been shown to be therapeutic in animal models of acute liver failure [5] and limited human clinical studies have been promising [6]. Inherited liver disorders are also candidates for cell therapy because they are caused by loss of expression or dysfunction of a single hepatocyte‐specific gene (i.e. monogenic). Replacement of these “diseased” hepatocytes by normal hepatocytes should be therapeutic. Specific disease examples include Crigler–Najjar syndrome, type 1, ornithine transcarbamylase deficiency and other urea cycle disorders, familial hypercholesterolemia, factors VII and IX deficiency, and phenylketonuria (diseases in which there is no underlying liver injury or damage) and other diseases such as

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

Wilson’s disease, α1‐antitrypsin deficiency, and hemochromatosis (diseases in which there is extensive liver injury). In inherited monogenic liver disorders, autologous cell transplantation may be possible, in which the patient’s cells are genetically manipulated ex vivo and then transplanted back into his/her liver without the need for immunosuppression (and indeed this has been done) [7]. However, only a small percentage of the total hepatocellular mass (around 1–2% maximum) can be replaced by simple hepatocyte transplantation without causing portal hypertension and/or hepatic infarction. Therefore, in most instances, it will be necessary to selectively expand the therapeutic cells in vivo after their engraftment. The therapeutic threshold will be different for each disease, but generally cell replacement levels of at least 5–10% will be needed for most disorders. Attempts to increase the proportion of transplanted hepatocytes in the liver simply by stimulating liver regeneration (e.g. through partial hepatectomy (PH) or carbon tetrachloride (CCl4)‐ induced hepatic necrosis) have generally failed to show significant benefit. This is not surprising, considering that hepatocytes need to undergo only one or two rounds of cell division to replace the mass of liver removed by two‐thirds PH [8, 9]. Performing repeated PH or CCl4 administration does not significantly increase repopulation by donor hepatocytes [10], since both transplanted and host hepatocytes respond similarly to this regenerative stimulus. Repeated cell transplantations have also been performed, [11, 12] but this also did not significantly increase the efficacy of liver replacement by transplanted cells.

CLINICAL TRIALS OF HEPATOCYTE TRANSPLANTATION To date, allogeneic hepatocyte transplantation in humans has had limited clinical success. In 1998, a child with Crigler–Najjar syndrome type I, suffering from severe hyperbilirubinemia, was infused with 7.5 × 109 allogeneic donor hepatocytes via a portal vein catheter [13]. This resulted in a significant reduction of serum bilirubin. However, after two and a half years serum bilirubin progressively increased to pretreatment levels, although bilirubin glucuronides were still detectable in the bile (J. Roy Chowdhury, personal communication). Studies have also been conducted in patients with chronic liver disease and cirrhosis, again with only marginal, if any, success [14]. Hepatocyte transplantation has also been used in conjunction with ex vivo retroviral gene therapy in five patients with a defect in the low‐density lipoprotein (LDL) receptor [7]. In this study, the proportion of liver cell mass replaced was estimated to be around 1%. The procedure resulted in a very modest decrease in plasma cholesterol in several of the patients. The worldwide experience with human hepatocyte transplantation has been updated recently [15, 16]. Usually, suspensions of adult hepatocytes (autographs or allographs) were utilized in a single injection, although some patients received multiple infusions. Many patients had inherited metabolic disorders, including familial hypercholesterolemia, Crigler–Najjar ­syndrome type 1, factor VII deficiency, urea cycle defects, glycogen storage diseases, and others [14, 16]. Initial improvement of metabolic function was frequently observed. Some patients

551

subsequently received a liver transplant, so that the long‐term effects of hepatocyte transplantation could not be assessed. However, in the remaining patients, the function of transplanted hepatocytes was not sustained. Whether this resulted from rejection of transplanted cells, their lack of proliferation, or eventual apoptosis has not been determined. In no patient was expansion of transplanted hepatocytes or progressive improvement of metabolic function over time documented. Hepatocyte transplantation therapy for fulminant hepatic failure as a bridge to orthotopic liver transplantation has had some encouraging results [5, 6]. The numbers of patients treated by this modality are low, the diseases are of varying etiologies and therefore it is not possible to draw definitive conclusions regarding the efficacy of this intervention. Data from several pilot studies have been reported, many with reductions in both blood ammonia and encephalopathy and successful organ transplant [5]. In several cases, the patients recovered without the need for a transplant, but others died. At this stage it can be said that allogeneic hepatocyte transplantation in humans is safe and that the available data suggest partial efficacy for both genetic and acquired hepatic disorders.

BASIC REQUIREMENTS FOR LIVER REGENERATION In the normal adult liver, hepatocytes are quiescent, turning over very slowly (only two to three times per year). However, following surgical reduction of liver mass or extensive acute toxic liver injury, hepatocytes rapidly enter the cell cycle and proliferate to restore liver mass. During liver regeneration, 70–90% of the residual mature hepatocytes engage in DNA synthesis and undergo cell division [9]. Liver regeneration is a highly organized, complex process involving growth factors and cytokines, transcription factors, cell signaling pathways, and expression of cell cycle regulatory genes [17]; (see Chapter 45). From many studies, it has been concluded that the proliferative activity of adult hepatocytes is sufficient to regenerate the liver following two‐thirds PH without participation of stem cells [18].

LIVER SIZE CONTROL AND “HIPPO SIGNALING PATHWAY” For many years, it has been known that the liver size (mass) is proportional to total body weight, ranging from 3 to 5% in different mammalian species (see Chapter  45). After two‐thirds PH, cellular proliferation begins within 12–18 hours and the liver size returns to normal within 1–2 weeks in rodents and after 4–12 months in humans. When an undersized liver is transplanted, it grows to the expected full size for the host, and when an oversized liver is transplanted, its size is reduced to the expected mass compared with total body weight (see Chapter  45). However, until very recently, little was known concerning how this process is controlled. In 2007, Pan and colleagues [19] showed that mammalian genes comparable to those in the Drosophila Hippo kinase signaling cascade, that regulates wing mass during development,

552

THE LIVER:  ANIMAL MODELS FOR THERAPEUTIC LIVER REPOPULATION

control hepatocyte proliferation. When YAP, the mammalian counterpart to Yorki, the last gene in the Drosophila Hippo kinase cascade, is overexpressed in a transgenic mouse model, hepatocyte proliferation becomes unchecked and there is massive liver hyperplasia, leading to hepatic carcinogenesis. When YAP hyperexpression is turned off or blocked within four to six weeks after birth, liver size returns to normal [19]. YAP is synthesized in the cytoplasm and translocated to the nucleus, where it functions as a transcriptional coactivator of TEADs, p73, RUNX2, and other genes. Control of YAP function is achieved through phosphorylation by upstream kinases in the Hippo signaling pathway, Mst1/2 that phosphorylates Lats1/2, and pLats1/2 that phosphorylates YAP at amino acid S127. pYAP phosphorylated at S127 is retained in the cytoplasm, where it is subsequently degraded [19, 21]. YAP is a proliferative gene with around 500 known downstream targets, including many that effect cell growth and the cell cycle. YAP also induces two anti‐apoptotic genes, Birc2 and Birc5, so that it also increases survival of cells in which YAP nuclear function is enhanced (i.e. by reduced phosphorylation of YAP at S127). Most studies reporting on the tumorigenic properties of YAP in hepatocytes have been performed in transgenic mice or in hepatocytes expressing a non‐phosphorylatable mutant of YAP (YAP S127A), in which YAP expression/function is constitutive and under control of a strong promoter (i.e. the tet operon) [20].Therefore, it unclear whether the YAP is inherently oncogenic or whether this reported phenotype is the result of aberrant or uncontrolled YAP hyperexpression/ function.

ANIMAL MODELS FOR THERAPEUTIC LIVER REPOPULATION For many years, it was thought that mature hepatocytes could undergo only two or three divisions, after which they become terminally differentiated and incapable of further proliferation. More recently, however, it has been shown in several rodent model systems that extensive modifications of the liver microenvironment permit hepatocytes to retain high proliferative capability and to effectively repopulate the liver. Under these conditions, the level of cell replacement can be 90% and higher, providing proof of principle for therapeutic liver repopulation by transplanted cells. Originally, Sandgren et al. developed a transgenic mouse model in which a protease, urokinase plasminogen activator (uPA), is expressed exclusively in hepatocytes under control of the albumin (Alb) promoter [22]. In this model, tissue protease activity caused continuous and extensive liver injury and sub‐fulminant liver failure, leading to death of the mice at four to six weeks of age. However, some mice survived and, in these there were scattered nodules of normal liver tissue distributed throughout the hepatic parenchyma (Figure  44.1a). This occurred by deletion or inactivation of the uPA transgene from individual hepatocytes, which then clonally expanded into large clusters that replaced damaged tissue. These findings prompted the  investigators to transplant normal hepatocytes (containing a β‐galactosidase marker gene) into uPA transgenic mice,

after  which they observed extensive liver repopulation (Figure  44.1b). They estimated that each donor hepatocyte engrafted into the albumin uPA host liver, underwent 12–14 cell divisions [23]. Another mouse model for liver repopulation was generated by targeted disruption of the last gene in tyrosine catabolism, fumarylacetoacetate hydrolase (Fah) [24]. Deletion of Fah leads to accumulation of upstream intermediates in tyrosine catabolism, some of which (namely, fumarylacetoacetate) are toxic and cause extensive and continuous liver injury. The Fah null mouse represents an animal model for the human metabolic disorder hereditary tyrosinemia, type 1 (HT1), which causes extensive liver injury, hepatocellular carcinoma, and death at an early age in affected individuals. Administration of 2‐(2‐nitro‐4‐trifluoromethylbenzoyl) cyclohexane‐1,3‐dione (NTBC), a pharmacological inhibitor of tyrosine catabolism upstream of homogentisic acid, prevents accumulation of fumarylacetoacetate and is successful in treating patients with HT1 [25]. NTBC also allows Fah null mice to survive and liver failure occurs in these mice only when NTBC is discontinued. After transplanting syngeneic wild‐type (wt) hepatocytes into Fah null mice maintained on NTBC, only scattered small clusters of transplanted hepatocytes are detected (Figure 44.1c left panel). However, if NTBC treatment is discontinued shortly after cell transplantation, liver injury resumes and transplanted cells proliferate extensively, forming large clusters within three weeks (middle panel) and replacing most of the liver mass within six weeks (right panel). Fah null mice with repopulated livers remain healthy, have normal liver function tests and show a relatively normal liver structure for many months after wt hepatocyte transplantation [24]. These studies provide proof of principle that liver repopulation can effectively cure a monogenic liver disease, namely, the mouse equivalent to HT1. Transplanted repopulating hepatocytes integrate into the host hepatic architecture and express all functions necessary for normal health of the animals. In the Fah null mouse, not only do transplanted wt hepatocytes replace Fah null hepatocytes, but the transplanted cells can also be serially transplanted through up to twelve consecutive Fah null mice, while retaining full ability to proliferate and replace host hepatocytes [26, 27]. In these studies, it was calculated that each serially transplanted hepatocyte underwent an average of at least 69 cell divisions. Thus, murine hepatocytes exhibit essentially infinite capacity to proliferate and restore liver function under circumstances in which there is both massive and continuous liver injury and the transplanted hepatocytes have a significant selective advantage for survival compared with host hepatocytes. Therefore, under the specialized circumstances existing in the liver of Fah null mice, hepatocytes exhibit many properties of stem cells, except for the ability to differentiate into more than one lineage, in this case into hepatocytes but not into bile duct epithelial (BDE) cells. More recently, several additional murine liver repopulation models have been developed (reviewed in [28]). In AFC8 [29] and TK‐NOG [30] transgenic mice selection can be induced by administration of a small molecule drug. Wild‐type hepatocytes can also repopulate the livers of mice expressing a mutant form of human alpha1‐antitrypsin [31]. Similarly, transgenic mice



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

553

Figure 44.1  Liver regeneration and hepatocyte transplantation in the uPA transgenic mouse. (a) Wt mouse liver (left), regenerating uPA transgenic mouse liver by deletion of uPA transgene (middle), uPA transgenic liver (right). (b) LacZ transgenic mouse liver (left), uPA transgenic mouse liver (middle), uPA transgenic mouse liver repopulated with LacZ hepatocytes (right). (c) Repopulation of the Fah null mouse liver by wt hepatocytes under NTBC administration. Immunohistochemistry with FAH staining (dark), 400× magnification: (left panel) 2 days, (middle) 3 weeks, and (right) 6 weeks after cell transplantation.

overexpressing the p53 regulator mdm2 can be repopulated by transplanted donor cells [32]. In all of these genetic‐based liver repopulation models, successful liver repopulation by adult hepatocytes required two experimental conditions: (i) the liver was under massive and continuous liver injury and (ii) the transplanted hepatocytes had a strong cell‐autonomous selective advantage for survival in the host liver. In the Fah model, it has been shown that this selective advantage is based on p21‐dependent cell cycle arrest in host hepatocytes, resulting from p21 induction caused by DNA damage [33]. Thus, an alternative strategy to obtain a high level of liver repopulation by transplanted hepatocytes is to block proliferation of endogenous hepatocytes using exogenous DNA‐damaging agents and then transplant normal hepatocytes in conjunction with a liver proliferative stimulus. The first method described to achieve this effect was to treat rats with retrorsine, a plant alkaloid that is taken up and metabolized selectively by hepatocytes to produce a DNA alkylating agent that cross‐links cellular DNA and disrupts hepatocyte division [34]. When retrorsine or a related compound, monocrotaline, is administered to rats or mice, there is a long‐lived

inhibition of hepatocyte proliferation. However, the basic metabolic functions of DNA damaged hepatocytes are maintained and the animals survive. Two to four weeks after retrorsine administration, the animals are subjected to two‐thirds PH or CCl4 administration in conjunction with transplantation of hepatocytes from normal animals. This leads to a brisk regenerative response specifically by transplanted hepatocytes and there is near total liver repopulation in three to six months [34]. Another method to achieve effective liver repopulation by transplanted hepatocytes is to induce DNA damage with ­selective irradiation of parts of the liver in conjunction with hepatocyte transplantation and either two‐thirds PH, CCl4 administration, or ischemic liver injury [35, 36]. In X‐irradiated rodents, administration of HGF has been used to replace PH as a liver regenerative stimulus [35]. As with retrorsine or monocrotaline treatment of the host liver, transplanted hepatocytes have a proliferative advantage over host hepatocytes in irradiated liver lobes. Recently, this approach is beginning to be explored in humans [37]. However, definitive data demonstrating selective liver repopulation in humans that underwent such liver conditioning have not yet been published.

554

THE LIVER:  STEM CELLS: THEIR ORIGIN, PROPERTIES, AND TRANSPLANTATION

XENOREPOPULATION MODELS Extensive repopulation of the liver by transplanted hepatocytes can also be achieved with human hepatocytes. Since human cells are rejected by other animals, this requires the use of immune‐deficient recipients or pharmacological immune suppression. Small rodents capable of harboring human hepatocytes are not only of interest as a test bed for cell therapy, but also for a multitude of preclinical pharmaceutical research applications, including models of drug metabolism, infectious diseases (including malaria, hepatitis B and C, gene therapy, and toxicology) [28]. Three murine models capable of supporting extensive repopulation of the mouse liver with human hepatocytes have been reported (reviewed in [28]). All three models have been developed commercially and are being used by the pharmaceutical industry as well as academia in preclinical experimentation. In 2001, Dandri et al. first showed that immune‐deficient uPA transgenic mice could be engrafted with human hepatocytes and used as a model for hepatitis B [38]. Subsequently, this model was further developed to permit extensive repopulation, reaching levels as high as 90% human cells [39]. Fah knockout mice can also be repopulated extensively with human cells from a variety of sources when they are crossed onto severely immune deficient backgrounds (Figure  44.2) [40]. This model has the advantage that the hepatic injury is not constitutive but can be titrated by NTBC. Most recently, a third xenorepopulation platform has emerged, the TK‐NOG mouse [30]. All of these models are of potential use to assess the therapeutic potential of transplanted cells, be they primary human hepatocytes or stem cell derivatives.

STEM CELLS: THEIR ORIGIN, PROPERTIES, AND TRANSPLANTATION Given the promising results seen in experimental cell transplantation models, the source of cells that could be used clinically has been of considerable interest. Historically, hepatocytes isolated from adult donors were used for transplantation. However, in the clinical setting the availability of such cells from cadaveric donors is even more constrained than whole organs for orthotopic transplantation, because

improved surgical techniques and perioperative management have vastly increased the percentage of harvested livers that can be used directly for transplantation. Therefore, considerable attention has been focused on the potential use of a renewable and expandable cell source, such as stem cells, to reconstitute liver mass. Several kinds of stem cells have been explored as potential hepatocyte precursors for transplantation, including fetal liver stem cells, hepatocytes derived from pluripotent stem cells, and adult liver stem/progenitor cells. While the existence of a bipotential stem cell capable of generating both cholangiocytes and hepatocytes in fetal life – the hepatoblast – is well accepted, the notion of a true adult liver stem cell is currently controversial. Much of the existing literature on liver stem/progenitor cells can be explained by cell plasticity [41]. The current state of this field will be presented in the following sections. First, the literature on hepatic stem cells will be reviewed. This will then be followed by a discussion of cell plasticity as a possible alternative mechanism for the published data, and recent studies using genetically modified hepatocytes to increase their proliferative/repopulation potential [21]. During embryogenesis, the earliest stem cells originate from the inner cell mass of the blastocyst and are pluripotent (capable of differentiating into all the cell types in mammalian species); these cells are generally referred to as embryonic stem (ES) cells [42]. These cells give rise to somatic stem cells that subsequently differentiate into multipotent tissue‐specific stem cells [42–44]. The latter give rise to lineage‐committed progenitor cells that proliferate and differentiate into mature phenotypes that ultimately become the somatic, tissue‐specific, working cells in different organs (Figure 44.3). Through studies of cell transplantation to reconstitute the hematopoietic system and studies of cell turnover in other tissues undergoing rapid and continuous regeneration, such as skin and intestinal epithelium, stem cells have been shown to exhibit four essential properties: (i) they have the capacity to maintain themselves (self‐renew), while at the same time they generate progeny that differentiate into mature cellular phenotypes; this is referred to as asymmetric cell division [45, 46]; (ii) they are multipotent, that is, capable of producing differentiated cells in at least two lineages; (iii) their progeny are stable, reconstitute organ mass, and remain functional in the tissue for a long time; and (iv) by virtue of their capability of self‐renewal, stem cells can be transplanted serially through successive hosts.

Figure 44.2  Repopulation of the Fah−/−/Rag2−/−/IL2rg−/− mouse liver by primary human hepatocytes at six weeks after cell transplantation, shown at (a) low and (b) high magnification.



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

555

Figure 44.3  Stem cells and tissue differentiation.

LIVER REPOPULATION BY FETAL LIVER STEM/PROGENITOR CELLS During days 11–15 of embryonic development a bipotent liver epithelial cell, the hepatoblast, can be found in rats and mice (see Chapter  2). These hepatoblasts co‐express markers of adult hepatocytes (albumin) and cholangiocytes (CK19) as well as AFP. The ultimate test for a putative stem cell is to demonstrate its ability to self‐renew in vivo and to repopulate functionally a tissue or organ, long‐term. Sandhu et  al. [47] reported 5–10% repopulation of DPPIV− mutant F344 rat liver by transplanting wt ED14 fetal liver epithelial cells in conjunction with two‐thirds PH. Liver repopulation by transplanted cells increased progressively over six months, and the bulk of repopulating clusters contained both hepatocytes and mature bile duct cells. The transplanted cells were integrated into the host parenchyma and formed hybrid bile canaliculi with host hepatocytes. Thus, transplanted rat ED14 fetal liver epithelial cells exhibited three major properties of liver stem cells: (i) extensive proliferation, (ii) bipotency, and (iii) long‐term repopulation in vivo [47]. Liver repopulation by transplanted rat fetal liver cells was achieved in a non‐selective host liver environment but required PH to initiate the process. This is consistent with studies in hematopoietic stem cells, in which hematopoietic reconstitution does not occur unless there is near total ablation of the host bone marrow.

In ED14 rat fetal liver, there are three distinct populations of epithelial cells, those positive for AFP and Alb but negative for CK‐19, those positive for AFP, Alb, and CK‐19, and those positive for CK‐19 but negative for AFP and Alb [48]. The number of AFP+/Alb+/CK‐19+ cells decreased dramatically at ED16, after which liver repopulation potential of rat fetal liver cells also decreased significantly [47]. The level of liver repopulation by ED14 fetal liver cells under non‐selective conditions (i.e. in a normal liver) can also be increased to 20–25% by simply increasing the number of ED14 fetal liver cells transplanted (Figure 44.4) [49]. Repopulation continues to increase for up to 1 year, reaching an average of around 30% for the total liver, and remains stable for the life of the animal [50]. This represents a several thousand‐fold amplification of transplanted fetal liver epithelial cells in the host organ. Both hepatic parenchymal cords and mature bile ducts are formed by transplanted fetal liver cells, and the progeny of the transplanted cells express normal levels of hepatocytic and cholangiocytic genes in the respective cell types [49, 51]. Since serial transplantation has not yet been demonstrated with fetal liver epithelial cells, they are referred to as fetal liver stem/progenitor cells (FLSPCs), rather than stem cells. The mechanism of liver repopulation by rat FLSPCs has been shown to be cell competition between the transplanted cells and host hepatocytes [49], a process that was originally

556

THE LIVER:  THE LIVER STEM CELL “NICHE”

Figure 44.4  Repopulation of the normal adult rat liver by fetal liver stem/progenitor cells. (a) Two examples of whole rat liver sections repopulated by ED14 fetal liver cells. (b) Selected areas of repopulated liver at higher magnification showing both hepatocytes and mature bile duct structures.

described in Drosophila during wing development [52, 53]. These cells have been cryopreserved with full ability to repopulate the normal adult liver after thawing [54] and rat FLSPCs have been enriched to 95% purity by selection with immunomagnetic beads [51].

STEM CELLS IN THE ADULT LIVER Although many studies claim to have identified, isolated, and purified hepatic epithelial stem cells from the adult liver, much of the evidence is derived from in vitro culture of clonogenic epithelial cells [55–60]. Whether these cells represent bona fide in vivo bipotential stem cells remains uncertain. In the mouse, all of the reported liver “stem cell” markers are also found in cholangiocytes and therefore the existence of true stem cells distinct from cholangiocytes in the adult mouse liver still remains to be established. Currently, the published data regarding the emergence of bipotential progenitors can all be explained by cell plasticity, that is, fate conversion of hepatocytes to cholangiocytes and vice versa [41]. In the rat, however, classic experiments are suggestive of a specialized adult liver stem cell. Unlike in the mouse, AFP becomes expressed in cells of ductal origin upon liver injury [61]. In conjunction with [3H] thymidine pulse labeling and chase, AFP expression was shown to decrease as [3H] labeled cells differentiated into basophilic hepatocytes, which were strongly positive for albumin [62]. Thus, AFP can be considered a specific marker for stem/progenitor cells in rats.

THE LIVER STEM CELL “NICHE” If bipotential liver stem cells do exist, the question remains as to where they reside [61]? The original idea of a stem cell “niche” evolved from the concept that stem cells reside in tissues within an “inductive microenvironment” that directs their differentiation [63, 64]. More recently, the stem cell niche has been described further as “a specific location in a tissue where stem cells can reside for an indefinite period and produce progeny cells, while self‐renewing” [65]. Local stromal cells and other extracellular environmental factors attract stem cells to these niches and affect their behavior (i.e. gene expression program, proliferation, and/or differentiation), and such niches have been identified in the bone marrow, brain, skin, and intestinal mucosa [66–70]. At present, the most likely candidate for a liver stem cell niche is the canal of Hering (Figure 44.5). The canals of Hering were originally identified more than 100 years ago as luminal channels linking the hepatocyte canalicular system to the biliary system [71]. These channels contain small undifferentiated epithelial cells [72, 73] that are in direct physical continuity with hepatocytes at one membrane boundary and bile duct cells at another boundary to form duct‐like structures, enclosing a lumen (i.e. the canal; Figure  44.5b). In most models of liver structure, the canals of Hering are depicted as very limited structures confined to the portal space. However, studies conducted in humans have demonstrated that the canals of Hering are much more elaborate than previously thought and extend far into the hepatic parenchyma [73, 74]. These structures contain epithelial cells with dual expression of bile ductular and fetal



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

557

Figure 44.5  (a) Canal of Hering as the proposed liver stem cell “niche”. “Oval cells” within the canal are the precursors of both hepatocytes and bile ducts. (b) Electron micrograph of the canal of Hering. The canal of Hering (*) is bounded by two mature hepatocytes (Hc) and one undifferentiated epithelial cell (Ec).

hepatocytic markers (AFP, HepPar1, CK‐19) and were thus considered to represent “facultative hepatic stem cells” [73, 75].

“OVAL CELLS” AS HEPATOCYTE PROGENITORS The term “oval cells” was coined by Farber [76] to describe non‐ parenchymal cells in the periportal region that were ­present after treatment of rats with carcinogens: ethionine, α‐­acetylaminofluorene (2‐AAF), and 3‐methyl‐4‐diethylaminobenzene. Other methods to induce proliferation of “oval cells” are to treat rats with d‐galactosamine [77,78], or allyl alcohol [79]. In each of these models, cells are induced in the adult liver that have a small, oval shaped, pale‐stained blue nucleus and very scant, lightly basophilic cytoplasm. Farber did not believe that “oval cells” were hepatocyte progenitors, but Thorgeirsson and co‐workers [61] demonstrated that “oval cells” induced to ­proliferate in the periportal region after treatment of rats with 2‐AAF followed by two‐thirds PH subsequently differentiated into distinct clusters of basophilic hepatocytes. This was demonstrated by pulse labeling of the liver with [3H]thymidine and following the progression of label from periportal “oval cells” to hepatocytic clusters in the mid‐parenchyma in conjunction with the kinetic pattern of expression of bile ductular (CK‐7 and CK‐19) and hepatocytic (α‐fetoprotein [AFP] and Alb) markers over time [61, 62]. Other indirect evidence suggesting that “oval cells” are hepatic progenitors is their expression of c‐kit [80], CD34 [81], flt3 receptor [82], and LIF [83], all known to be expressed in hematopoietic stem cells or their immediate derivatives. Studies in the rat 2‐AAF/PH model have demonstrated that proliferating “oval cells” are indeed located in the canals of Hering [84]. Thorgeirsson and colleagues [85] performed extensive immunohistochemical and ultrastructural studies in which they demonstrated that “oval cells,” induced to proliferate by 2‐AAF/PH, are derived from undifferentiated cells in the canals of Hering, after which they pass through discontinuities in the laminar basement membrane of the ductal limiting plate and join together with stellate cells as they enter the hepatic parenchyma, proliferate, and differentiate into hepatocytes.

Oval cell induction methods were also reported for mice, including a choline‐deficient (CD)/ethionine‐substituted diet [86, 87], treatment with dipin [88], or 3,5‐diethoxycarbonyl‐1, 4‐dihydrocollidine (DDC) [89]. However, their utility has been controversial, because multiple groups have shown that “oval cells” emerging in these models generally lack bipotentiality [90]. A role for “oval cells” in normal liver physiology and cell turnover has not been established. However, as indicated above, “oval cells” are induced to proliferate when liver injury is superimposed on circumstances in which hepatocyte proliferation is impaired. In the rat, these cells exhibit many features of progenitor cells, dividing rapidly and appearing to differentiate into both hepatocytes and BDE cells. Attempts to establish specific markers for “oval cells” to distinguish them from mature hepatocytes and BDE cells, and to determine their lineage origin (mesoderm or endoderm), have led to conflicting findings. All investigators agree that “oval cells” express common liver epithelial progenitor cell markers, such as AFP and Alb for hepatocyte progenitors and CK‐19 (and OV6 in the rat) for bile duct progenitor cells. They were initially thought to express hematopoietic stem cell markers, c‐kit, CD34, and Thy 1 [80, 81, 84, 92], but subsequent studies reported that both fetal liver progenitor cells and “oval cells” are negative for these markers [75, 93, 94–96]. If “oval cells” are indeed stem cells or hepatic progenitor cells, they should be able to restore liver mass after their transplantation. About 30 years ago, Faris and Hixson [97] reported that “oval cells” isolated from the liver of rats fed a CD diet, treated with 2‐ AAF, and transplanted into the liver of secondary hosts produced “colonies” or clusters of cells with an hepatocytic phenotype in recipients that had also been subjected to the CD diet but not in recipients that had received a normal diet. However, the level of liver repopulation by transplanted CD/2‐AAF “oval cells” was not determined. “Oval cells” isolated from the liver of rats treated with d‐galactose (d‐Gal) also proliferate and differentiate into hepatocytes after their transplantation into rats undergoing two‐thirds PH [98]. Duct‐like epithelial cells isolated from the atrophic pancreas of rats undergoing treatment with a copper chelating agent (trien) also proliferate modestly and differentiate into hepatocytes after transplantation into normal rat liver [99]. Isolated pancreatic cells from normal mice also repopulate the liver of Fah null mice [100]. “Oval cells” isolated from the liver of DDC‐fed mice also

558

THE LIVER: PLASTICITY, TRANSDIFFERENTIATION, AND CLONOGENIC SUBSETS OF MATURE EPITHELIAL CELLS

repopulate the liver of Fah null mice, albeit with less efficiency than with mature hepatocytes [101]. Similarly, “oval cells” from GFP transgenic mice maintained on a DDC diet repopulate the liver of wt mice treated with monocrotaline in conjunction with PH [102]. Other studies also showed effective repopulation of the liver by purified “oval cells” in both rats treated with retrorsine [75] and in Fah null mice [60], but not in animals with a normal liver. Numerous studies have reported the isolation of “oval cell” lines from mice and rats, and also from humans, that are clonal, bipotent, and exhibit other stem and progenitor cell characteristics in vitro and in vivo (for a review see [18]). This has provided valuable information concerning the basic biological properties of these cell lines, but, in general, in vivo repopulation by transplanted “oval cell” lines has been very low.

HUMAN “OVAL CELLS” AND STEM CELLS A human counterpart to “oval cell” activation has been described in liver tissue obtained from patients with extensive chronic liver injury or submassive hepatic necrosis, that is, the so‐called “ductular reaction” (for a detailed description, see [103]). “Ductular reactions” are comprised of collections of cells in ductular arrays with the morphological appearance and immunohistochemical markers comparable to those found in rodent “oval cells.” These cells are present primarily in the portal tracts with extension into the parenchyma and express both hepatocytic and bile ductular markers, and also certain neuroendocrine genes [103–106]. Using double and triple label immunohistochemistry, Zhou et  al. [107] have shown that “ductular reactions” are bipolar structures with cells at one pole exhibiting hepatocytic morphology and gene expression (HepPar1 or HepPar1/NCAM) and cells at the other pole exhibiting biliary morphology and gene expression (CK‐19 or CK‐19/NCAM), with undifferentiated epithelial cells in the center expressing only NCAM. Cells with similar morphological and immunohistochemical properties have also been identified in the human fetal liver beginning at four weeks gestation [108]. A number of investigators have isolated, cultured, and/or passaged human fetal liver epithelial cells with bipotent properties, and several of these studies have demonstrated their differentiation into hepatocytes after transplantation into SCID or nude mice [109–111]. Schmelzer et al. [55] have identified two populations of clonogenic hepatic epithelial cells from human fetal, neonatal, and pediatric liver that exhibit stem cell properties. While these cells are progenitor‐like in vitro and can extensively expand, their ability to efficiently give rise to functional hepatocytes either in vitro or upon transplantation remains unproven. In the intestine, a self‐renewing classic stem cells resides at the bottom of intestinal crypts and can be defined by expression of the wnt‐target gene Lgr5 [112]. Clevers and co‐workers developed tissue culture conditions which permit the clonal expansion of these stem cells and were able to obtain all mature intestinal cell types in “organoids” derived from single Lgr5+ cells [113]. They applied similar tissue culture conditions to other endodermal tissues, including liver, stomach, and pancreas and were able to grow immortal Lgr5+ liver organoids from both mice and

humans [56, 57]. The cell of origin has a ductal phenotype and liver organoids are extensively cholangiocytic in phenotype. Nonetheless, they can be induced to express hepatocyte markers in culture [56, 57]. Upon transplantation, they can produce bona fide hepatocytes, but the frequency is very low. The future potential of clonally derived Lgr5+ cells is discussed in Chapter 77.

PLASTICITY, TRANSDIFFERENTIATION, AND CLONOGENIC SUBSETS OF MATURE EPITHELIAL CELLS Much of the data on liver injury responses in the historic literature has been interpreted as either proliferation of existing mature epithelial cells (hepatocytes or cholangiocytes) or activation of facultative stem cells. More recently, however, the phenomenon of plasticity and transdifferentiation between ductal epithelial cells and hepatocytes has become more appreciated and experimentally proven. Several studies showed conclusively that proliferating ductal cells, previously considered to be derived from facultative stem cells, were in fact not the source of parenchymal regeneration in several classical models of oval cell injury in the mouse [90, 114–116]. To the contrary, it has been shown clearly that mature hepatocytes can efficiently convert into cholangiocyte‐like cells [115] and even form a functional biliary tree [117]. Hepatocyte‐derived ducts can extensively proliferate and retain the ability to reconvert to their cell of origin, the hepatocytes, once the injury subsides [115]. The ability of hepatocytes to differentiate into BDE cells has also been well demonstrated in several rat models under injury conditions [118, 119]. These studies have taken advantage of recently developed lineage tracing techniques to identify the origin of cells in liver tissue under different physiologic and pathophysiologic states (see Chapter 85). Taken together these findings raise the question, whether the bipotential stem/progenitor cells and oval cells previously reported may actually have also been transdifferentiated hepatocytes, not stem cells of ductal origin. Recent papers have taken advantage of this plasticity of hepatocytes. While mature hepatocytes cannot be efficiently expanded in culture, transdifferentiated hepatocytes can be grown extensively [120]. Excitingly, these cells retain the ability to redifferentiate into mature hepatocytes and function in transplantation. While non‐hepatocytes are clearly not the source of new hepatocytes in most of the classic mouse oval cell literature, newer injury models which involve activation of p21 and senescence in mature hepatocytes have shown that cholangiocytes can become hepatocytes [32, 121, 122]. Thus, transdifferentiation has now been shown to work in both directions. If cholangiocytes can produce hepatocytes and vice versa, the question arises whether typical stem/progenitor cells exist in the adult liver or whether all regeneration can be explained by plasticity of mature epithelial cells. An important question arising from the concept of plasticity is whether all mature epithelial cells are equally proliferative and plastic. Recently, several labs have demonstrated heterogeneity within hepatocytes in the adult mouse [123, 124], once again giving rise to claims of having discovered



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

“the” liver stem cell. In addition, significant differences in the proliferative ability of cholangiocyte subsets have also been found [125].

LIMITED LIVER REPOPULATION BY EXTRAHEPATIC AND EMBRYONIC STEM CELLS Various studies have reported that cells released from the bone marrow (BM) into the circulation, migrate to the liver and differentiate into hepatocytes. However, the extent to which this occurs and the mechanism(s) involved remain highly controversial (for reviews, see [126–128]). Estimates of liver repopulation by hematopoietic cells vary widely, ranging from less than 0. 01 to 40%. Originally, Petersen et al. reported that BM stem cells from DPPIV+ F344 rats transplanted into sublethally irradiated DPPIV− F344 rats repopulate the BM after which they migrate to the liver and “transdifferentiate” into hepatocytes through the liver “oval cell” progenitor pathway [129]. This mechanism was generally accepted until studies by Wang et al. [101] using lacZ marking showed that BM cells did not enter the “oval cell” pool in wt mice treated with DDC or contribute to liver repopulation by “oval cells” in secondary Fah−/− mouse recipients. Menthena et  al. [130] also showed in rats that DPPIV+ BM cells transplanted into DPPIV− rats contributed less than 1% to “oval cells” derived from three different model systems: (i) 2‐AAF/PH, (ii) retrorsine/PH, and (iii) d‐Gal induced liver injury. In Fah−/− mice as well as other model systems, it has been demonstrated that cell fusion and reprogramming, rather than transdifferentiation, is the mechanism by which hematopoietic cells acquire an hepatocytic phenotype. Initial studies in cell culture showed that BM and neuronal cells can fuse with ES cells [131, 132]. Wang et al. [133] and Vassilopoulos et al. [134] subsequently showed that hematopoietic stem cells fuse with hepatocytes in Fah null mice to produce cells expressing the deficient enzyme, which then expand massively to restore liver mass and function [133, 134]. Fusion also occurs between hematopoietic cells and neurons or muscle cells [135, 136] and it has been shown that myelomonocytic cells can fuse with hepatocytes [137, 138] or muscle cells [139] to produce somatic hybrids expressing genes from both parental cell types. Other studies reported that fusion does not appear to be required for BM‐derived cells to differentiate into hepatocytes [140–142]. Unfractionated or CD34+ enriched cells from human cord blood [143–146], multipotent adult progenitor cells (MAPCs) [147, 148], or mesenchymal stem cells [149–153] have been transplanted into the liver of immunodeficient mice. These transplanted cells express a differentiated hepatocytic phenotype [143, 154], but liver repopulation was once again very low. Several studies have reported that mesenchymal stem cells, isolated from adipose tissue and differentiated in culture along the hepatocytic lineage, can also engraft in the liver parenchyma and contribute to liver regeneration [155, 156]. One study [156] reported large repopulation clusters with hepatocyte‐differentiated mesenchymal stem cells, but this required

559

retrorsine treatment. These studies are promising, but the ability of mesenchymal stem cells to repopulate the adult liver under more normal, clinically viable circumstances has not been established.

PLURIPOTENT STEM CELLS (ES and iPSC) Because of their extensive proliferative capacity, pluripotent stem cells are an attractive potential source of transplantable hepatocytes. Not only can these cells divide indefinitely, but they also retain the ability to differentiate into multiple different mature cell types [42]. Until 2007, ES cells were the sole source of pluripotent human cells and are ethically controversial. Now, however, pluripotent cells can be derived from direct genetic reprogramming of somatic cell types, such as dermal fibroblasts [157, 158]. Pluripotent stem cells (both ES and iPSC) cells in culture can be induced along the endodermal and hepatocytic lineages by addition of specific cytokines and growth factors [159–166]. The first step typically involves the induction of definitive endoderm using activin A. Numerous slightly different step‐wise differentiation protocols have been developed by different labs with the aim of generating mature hepatocytes as the final product [159, 166]. Historically, cells produced in this fashion expressed typical markers such as Alb, but usually also express AFP. More recent protocols yield cells that have silenced AFP and express multiple mature hepatocyte markers. Despite this, genome‐wide expression analysis demonstrates that these hepatocyte‐like cells are not fully mature and are significantly different from primary human hepatocytes. Such cells can be transplanted into the liver of immune deficient recipients with differentiation into cells expressing mature hepatocyte [165, 166] and cholangiocyte markers [166]. Nonetheless, the level of liver repopulation obtained with hepatocyte‐differentiated, pluripotent stem cells is generally low and functional tests, such as blood albumin measurements, fail to demonstrate equivalence to primary hepatocytes [167]. The repopulation level is somewhat higher when the cells are transplanted into MUP‐ uPA/SCID mice [166]. To date, all pluripotent stem cell differentiation protocols generate only “hepatocyte‐like” cells, but not fully functional, mature, and transplantable equivalents of hepatocytes isolated from adult liver. More mature “hepatocyte‐ like” cells have been selected using surface markers, such as the asialoglycoprotein receptor for purification [167], but even this approach does not yield fully mature donor cells. In the future, it is hoped that conditions will be developed in which lineage‐ specified pluripotent derivatives will be therapeutically equivalent to primary cells.

DIRECT REPROGRAMMING Since it is possible to generate hepatocyte‐like cells from dermal fibroblasts via a pluripotent intermediate stage, several groups have reported direct reprogramming of somatic cells into hepatocyte‐like cells, terming them iHeps, induced hepatocytes

560

THE LIVER: FUTURE HORIZONS

[168–172]. The concept of direct reprogramming using a cocktail of cell‐type specific transcription factors was first developed for neurons [173]. For the liver, a combination Hnf4α plus Foxa1, Foxa2, or Foxa3 was successfully used to directly convert fibroblasts into transplantable hepatocyte‐like cells [171]. Soon thereafter, the generation of iHeps from human cells using a similar combination of factors (FOXA3, HNF1α, and HNF4α) was also reported [170]. While transplantation success has been reasonably good with human iHeps in rodents and other animals, the cells in the repopulated liver are not fully differentiated [174]. Nonetheless, these cells may be functional enough to be of use in an extracorporeal bioartificial liver device (BAL) (Lijian Hui, personal communication) and further developments may eventually render them useful for in vivo human transplantation in the future.

OTHER ROLES FOR BM CELLS IN LIVER REGENERATION Several studies have reported that injections of BM‐derived stem cells can restore liver function during chronic liver injury by enhancing the degradation of liver fibrosis in mice [175, 176]. This has been associated with induction of metalloproteinases, especially MMPs 2, 9, and 13 [177]. It has been reported that BM‐derived endothelial progenitor cells (EPCs), injected into the spleen during liver injury, engraft in the liver, form new blood vessels, and secrete growth factors, such as HGF, TGFα, EGF, and VEGF, stimulating liver regeneration and improving survival of animals with massive liver injury [178]. Thus, the role of BM stem cells in liver regeneration may be supportive in generating new parenchymal mass and, under some circumstances, in ameliorating hepatic fibrosis.

USE OF EX VIVO GENETICALLY MODIFIED HEPATOCYTES TO REPOPULATE THE LIVER IN A NORMAL HEPATIC MICROENVIRONMENT As noted previously, bipotent fetal hepatoblasts can repopulate the liver under normal physiologic circumstances. Because of ethical concerns, use of adult hepatocytes for this purpose would be preferred. Thus far, however, significant liver repopulation by adult hepatocytes has required extensive and ongoing genetic, physical, or chemical injury to the host liver. A substantial number of monogenic liver disorders in which there is no underlying or ongoing liver injury could be treated by transplantation of normal adult hepatocytes, if the cells to be transplanted were first genetically modified to provide a proliferative and/or survival advantage compared to host hepatocytes. Shafritz and coworkers determined that Yap1 is hyperexpressed in ED14 rat fetal liver cells, as well as the anti‐apoptotic BIRC5 (survivin) gene [21]. Based on these findings, they transduced normal adult hepatocytes with an hYap gene linked

to a modified estrogen receptor gene (ERT2), whose function could be controlled be tamoxifen. Thus, hYapERT2 expression/ function and liver repopulation would be regulated by tamoxifen administration [21]. The hYapERT2 sequence was incorporated into a lentivirus vector which produced stable Yap expression after transduction of normal adult hepatocytes. Liver repopulation in the DPPIV‐ Fischer (F) 344 rat transplantation system produced 8–10% liver repopulation after six months on continuous tamoxifen feeding (Figure  44.6). Repopulation was tightly controlled by tamoxifen administration, transplanted hepatocytes expanded into large clusters of morphologically normal hepatocytes and integrated into host hepatic parenchymal plates with no evidence for dysplasia, dedifferentiation or tumorigenesis. A major concern in using Yap to stimulate proliferation of cells is potential tumor risk. A very recent study by Peterson et  al. [180] showed increased repopulation by transplanted hYapERT2 transduced hepatocytes on continuous tamoxifen feeding for one year in the DPPIV model system with 10–20% repopulation, no evidence for tumorigenesis, and a cell competition mechanism for repopulation, as previously found with ED14 fetal liver cells [49]. Using hYapERT2 transduced Wistar RHA rat hepatocytes transplanted into the hyperbilirubinemic Gunn rat liver, there was comparable liver repopulation with a progressive 70–80% reduction in serum bilirubin on continuous tamoxifen feeding for six months. If tamoxifen feeding was initiated six months after hepatocyte transplantation, serum bilirubin levels dropped progressively over the next six months with the same slope as observed when tamoxifen administration was initiated at the time of cell transplantation. If DPPIV− rats transplanted with hYapERT2 transduced wt hepatocytes were taken off tamoxifen after six months of liver repopulation, the liver repopulation level did not decrease after an additional six months on a normal diet; thus, transplanted hepatocytes remained in the liver long‐term. These studies provide hope that transplantation of genetically modified hepatocytes will provide a therapeutic avenue to treat selected liver diseases.

FUTURE HORIZONS Many questions regarding the basic biology of stem cells in the liver remain unresolved, including the all‐important question of whether hepatic stem cells exist at all in the adult liver. Most data suggest that it is possible to generate new hepatocytes from cells other than hepatocytes themselves and that there is an intrahepatic source of these precursors. Their precise nature and location, however, remains unclear, as do the molecular mechanisms that lead to their activation. Much of this uncertainly is related to the lack of markers that would permit the dissection of the complex cellular heterogeneity of the liver. However, cell surface markers, and also genetic lineage tracing tools, are now available to address these issues. Thus, biological studies of antigenically defined and genetically marked cells are now feasible and are likely to solve some of these dilemmas. In addition, pluripotent stem cells now afford the opportunity for “developmental biology in a dish,” making



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

561

Figure 44.6  (a) Hepatocyte transplantation protocol. (b) Detection of lentivirus TTR‐hYapERT2 transduced hepatocytes in DPPIV− host liver by DPPIV enzyme histochemistry. Clusters of DPPIV+ cells are clearly visible at three months after cell transplantation in tamoxifen fed rats (b1). In the absence of tamoxifen feeding, clusters of transplanted cells are not apparent (b2). At six months after transplantation of lenti TTR‐hYapERT2 transduced hepatocytes, clusters of DPPIV+ cells are much larger in tamoxifen fed rats, and in some instances comprise whole liver lobules (b3). In the absence of tamoxifen feeding, transplanted DPPIV+ cells are still not visible (b4). (c) Quantification of liver repopulation by scanning digital images was performed in DPPIV− rats transplanted with lenti TTR‐hYapERT2 transduced hepatocytes, lenti TTR‐GFP transduced hepatocytes, or non‐transduced hepatocytes, all maintained on tamoxifen feeding (+Tam) or on rats transplanted with lenti TTR‐hYapERT2 transduced hepatocytes maintained on normal chow (−Tam).

attractive models to study molecular mechanisms of hepatic lineage development. Although substantial progress has been made during the past 10–15 years concerning the possibility of liver repopulation by transplanted cells, much still needs to be learned. Factors governing engraftment of transplanted cells into the liver and their homing to the correct niche, factors regulating proliferation and differentiation of transplanted cells into specific phenotypes required for organ function, and specific host conditions under which effective liver replacement can be achieved, all need to be determined. The best starting point for therapeutic liver repopulation will probably be a genetic disorder with ongoing liver injury, which will hopefully induce or augment proliferation of transplanted cells in the host liver. A good example of such a condition is Wilson’s disease, in which transplanted cells might also have a modest selective advantage [179], since they will not store high levels of copper. Another example is α1‐antitrypsin deficiency, in which a mutated form of α1‐antitrypsin is not secreted from the cell and causes liver injury but much less severe than in uPA transgenic mice. Most encouraging is the recent success in repopulating the liver with normal hepatocytes in a mouse model of a1‐antitrypsin deficiency [31]. To date, very few patients have been transplanted with fetal liver cells. Studies in rats have shown that these cells have the capacity to proliferate in the host, replace hepatic mass with

functional hepatocytes, and maintain differentiated hepatic function, long‐term [50]. This requires a liver proliferative stimulus or liver injury at the time the cells are transplanted, but no selective advantage other than their ability to replace host hepatocytes by cell competition. Use of stem cells from pediatric or adult cadaveric liver or from other sources, such as bone marrow, cord blood, ES cells, or iPS cells, is a possibility, and also cultured human fetal cells or adult hepatocytes modified to favor engraftment and proliferation in the host, but this will require substantial additional research. Cell lines have been established, including ES cells, iPS cells, fetal liver cells, and “oval (progenitor) cells” that also exhibit stem cell properties and differentiate into hepatocytes and/or bile ducts in vitro and in vivo. However, these cell lines have shown only limited repopulation of the normal liver at the current state of the art. In order to advance the field of liver cell therapy further, it will be necessary to find conditions under which cells and cell lines derived from ES cells, iPS cells, fetal liver, or adult liver can be expanded in culture and successfully repopulate the liver under conditions that will be clinically acceptable. Use of newly emerging three dimensional or “organoid” culture conditions to maintain and expand hepatic derived cells or cells in the hepatic lineage will help to advance this field (see Chapter 77). Although we are not yet there, restoration of liver function by therapeutic cell transplantation holds great promise for the future.

562

THE LIVER:  REFERENCES

ACKNOWLEDGMENTS The authors would like to thank Anna Caponigro for assistance in preparing the manuscript and Dr Irmin Sternlieb, a former colleague at Albert Einstein College of Medicine, who provided the electron micrograph illustrating a canal of Hering to D.A.S.

REFERENCES   1. Boeck, J. and Popper, H. Ueber Lebertransplantation in die Vorderkammer des Auges. Virchow Arch Path Anat, 1937;299:219–234.   2. Matas, A.J., Sutherland, D.E., Steffes, M.W. et al. Hepatocellular transplantation for metabolic deficiencies: decrease of plasms bilirubin in Gunn rats. Science, 1976;192:892–4.   3. Grompe, M. Liver repopulation for the treatment of metabolic diseases. J Inherit Metab Dis, 2001;24:231–44.   4. Oishi, K., Arnon, R., Wasserstein, M.P. et al. Liver transplantation for pediatric inherited metabolic disorders: Considerations for indications, complications, and perioperative management. Pediatr Transplant, 2016;20:756–69.   5. Strom, S.C., Fisher, R.A., Thompson, M.T. et al. Hepatocyte transplantation as a bridge to orthotopic liver transplantation in terminal liver failure. Transplantation, 1997;63:559–69.   6. Bilir, B.M., Guinette, D., Karrer, F. et al. Hepatocyte transplantation in acute liver failure. Liver Transpl, 2000;6:32–40.   7. Grossman, M., Rader, D.J., Muller, D.W. et al. A pilot study of ex vivo gene therapy for homozygous familial hypercholesterolaemia. Nat Med, 1995;1:1148–54.   8. Bucher, N.L. and Swaffield, M.N. The rate of incorporation of labeled thymidine into the deoxyribonucleic acid of regenerating rat liver in relation to the amount of liver excised. Cancer Res, 1964;24:1611–25.   9. Grisham, J.W. A morphologic study of deoxyribonucleic acid synthesis and cell proliferation in regenerating rat liver; autoradiography with thymidine‐ H3. Cancer Res, 1962;22:842–9. 10. Rajvanshi, P., Kerr, A., Bhargava, K.K. et al. Studies of liver repopulation using the dipeptidyl peptidase IV‐deficient rat and other rodent recipients: cell size and structure relationships regulate capacity for increased transplanted hepatocyte mass in the liver lobule. Hepatology, 1996;23:482–96. 11. Rajvanshi, P., Kerr, A., Bhargava, K.K. et al. Efficacy and safety of repeated hepatocyte transplantation for significant liver repopulation in rodents. Gastroenterology, 1996;111:1092–1102. 12. Rozga, J., Holzman, M., Moscioni, A.D. et al. Repeated intraportal hepatocyte transplantation in analbuminemic rats. Cell Transplant, 1995;4:237–43. 13. Fox, I.J., Chowdhury, J.R., Kaufman, S.S. et al. Treatment of the Crigler– Najjar syndrome type I with hepatocyte transplantation. N Engl J Med, 1998;338:1422–6. 14. Strom, S.C., Chowdhury, J.R., Fox, I.J. Hepatocyte transplantation for the treatment of human disease. Semin Liver Dis, 1999;19:39–48. 15. Iansante, V., Chandrashekran, A. and Dhawan, A. Cell‐based liver therapies: past, present and future. Philos Trans R Soc Lond B Biol Sci, 2018;373. 16. Squires, J.E., Soltys, K.A., McKiernan, P. et al. Clinical hepatocyte transplantation: what is next? Curr Transplant Rep, 2017;4:280–289. 17. Michalopoulos, G.K. Advances in liver regeneration. Expert Rev Gastroenterol Hepatol, 2014;8:897–907. 18. Miyajima, A., Tanaka, M., and Itoh, T. Stem/progenitor cells in liver development, homeostasis, regeneration, and reprogramming. Cell Stem Cell, 2014;14:561–74. 19. Dong, J., Feldmann, G., Huang, J. et al. Elucidation of a universal size‐control mechanism in Drosophila and mammals. Cell, 2007;130:1120–33. 20. Camargo, F.D., Gokhale, S., Johnnidis, J.B. et al. YAP1 increases organ size and expands undifferentiated progenitor cells. Curr Biol, 2007;17:2054–60. 21. Yovchev, M., Jaber, F.L., Lu, Z. et al. experimental model for successful liver cell therapy by lenti TTR‐YapERT2 transduced hepatocytes with tamoxifen control of Yap subcellular location. Sci Rep, 2016;6:19275. 22. Sandgren, E.P., Palmiter, R.D., Heckel, J.L. et al. Complete hepatic regeneration after somatic deletion of an albumin‐plasminogen activator transgene. Cell, 1991;66:245–56.

23. Rhim, J.A., Sandgren, E.P., Degen, J.L. et  al. Replacement of diseased mouse liver by hepatic cell transplantation. Science, 1994;263:1149–52. 24. Overturf, K., Al‐Dhalimy, M., Tanguay, R. et al. Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinaemia type I. Nat Genet, 1996;12:266–73. 25. Lindstedt, S., Holme, E., Lock, E.A. et al. Treatment of hereditary tyrosinaemia type I by inhibition of 4‐hydroxyphenylpyruvate dioxygenase. Lancet, 1992;340:813–7. 26. Overturf, K., al‐Dhalimy, M., Ou, C.N. et al. Serial transplantation reveals the stem‐cell‐like regenerative potential of adult mouse hepatocytes. Am J Pathol, 1997;151:1273–80. 27. Wang, M.J., Chen, F., Li, J.X. et al. Reversal of hepatocyte senescence after continuous in vivo cell proliferation. Hepatology, 2014;60:349–61. 28. Grompe, M. and Strom, S. Mice with human livers. Gastroenterology, 2013;145:1209–14. 29. Washburn, M.L., Bility, M.T., Zhang, L. et al. A humanized mouse model to study hepatitis C virus infection, immune response, and liver disease. Gastroenterology, 2011;140:1334–44. 30. Hasegawa, M., Kawai, K., Mitsui, T. et  al. The reconstituted “humanized liver” in TK‐NOG mice is mature and functional. Biochem Biophys Res Commun, 2011;405:405–10. 31. Ding, J., Yannam, G.R., Roy‐Chowdhury, N. et  al. Spontaneous hepatic repopulation in transgenic mice expressing mutant human alpha1‐antitrypsin by wild‐type donor hepatocytes. J Clin Invest, 2011;121:1930–4. 32. Lu, W.Y., Bird, T.G., Boulter, L. et  al. Hepatic progenitor cells of biliary origin with liver repopulation capacity. Nat Cell Biol, 2015;17:971–83. 33. Willenbring, H., Sharma, A.D., Vogel, A. et al. Loss of p21 permits carcinogenesis from chronically damaged liver and kidney epithelial cells despite unchecked apoptosis. Cancer Cell, 2008;14:59–67. 34. Laconi, E., Oren, R., Mukhopadhyay, D.K. et al. Long‐term, near‐total liver replacement by transplantion of isolated hepatocytes in rats treated with retrorsine. Am J Pathol, 1998;153:319–29. 35. Guha, C., Sharma, A., Gupta, S. et  al. Amelioration of radiation‐induced liver damage in partially hepatectomized rats by hepatocyte transplantation. Cancer Res, 1999;59:5871–4. 36. Malhi, H., Gorla, G.R., Irani, A.N. et  al. Cell transplantation after ­oxidative hepatic preconditioning with radiation and ischemia‐reperfusion leads  to extensive liver repopulation. Proc Natl Acad Sci USA, 2002;99:13114–9. 37. Soltys, K.A., Setoyama, K., Tafaleng, E.N. et  al. Host conditioning and rejection monitoring in hepatocyte transplantation in humans. J Hepatol, 2017;66:987–1000. 38. Dandri, M., Burda, M.R., Torok, E. et al. Repopulation of mouse liver with human hepatocytes and in vivo infection with hepatitis B virus. Hepatology, 2001;33:981–8. 39. Tateno, C., Yoshizane, Y., Saito, N. et al. Near completely humanized liver in mice shows human‐type metabolic responses to drugs. Am J Pathol, 2004;165:901–12. 40. Azuma, H., Paulk, N., Ranade, A. et al. Robust expansion of human hepatocytes in Fah‐/‐/Rag2‐/‐/Il2rg‐/‐ mice. Nat Biotechnol, 2007;25:903–10. 41. Kopp, J.L., Grompe, M., and Sander, M. Stem cells versus plasticity in liver and pancreas regeneration. Nat Cell Biol, 2016;18:238–45. 42. Thomson, J.A., Itskovitz‐Eldor, J., Shapiro, S.S. et al. Embryonic stem cell lines derived from human blastocysts. Science, 1998;282:1145–7. 43. Fuchs, E. and Segre, J.A. Stem cells: a new lease on life. Cell, 2000;100:143–55. 44. Weissman, I.L. Stem cells: units of development, units of regeneration, and units in evolution. Cell, 2000;100:157–68. 45. Marshman, E., Booth, C., and Potten, C.S. The intestinal epithelial stem cell. Bioessays, 2002;24:91–8. 46. Lechler, T. and Fuchs, E. Asymmetric cell divisions promote stratification and differentiation of mammalian skin. Nature, 2005;437:275–80. 47. Sandhu, J.S., Petkov, P.M., Dabeva, M.D. et  al. Stem cell properties and repopulation of the rat liver by fetal liver epithelial progenitor cells. Am J Pathol, 2001;159:1323–34. 48. Dabeva, M.D., Petkov, P.M., Sandhu, J. et al. Proliferation and differentiation of fetal liver epithelial progenitor cells after transplantation into adult rat liver. Am J Pathol, 2000;156:2017–31. 49. Oertel, M., Menthena, A., Dabeva, M.D. et al. Cell competition leads to a high level of normal liver reconstitution by transplanted fetal liver stem/progenitor cells. Gastroenterology, 2006;130:507–20.



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

50. Shafritz, D.A. and Oertel, M. Model systems and experimental conditions that lead to effective repopulation of the liver by transplanted cells. Int J Biochem Cell Biol, 2011;43:198–213. 51. Oertel, M., Menthena, A., Chen, Y.Q. et al. Purification of fetal liver stem/ progenitor cells containing all the repopulation potential for normal adult rat liver. Gastroenterology, 2008;134:823–32. 52. Moreno, E. and Basler, K. dMyc transforms cells into super‐competitors. Cell, 2004;117:117–29. 53. de la Cova, C., Abril, M., Bellosta, P. et al. Drosophila myc regulates organ size by inducing cell competition. Cell, 2004;117:107–16. 54. Oertel, M., Menthena, A., Chen, Y.Q. et al. Properties of cryopreserved fetal liver stem/progenitor cells that exhibit long‐term repopulation of the normal rat liver. Stem Cells, 2006;24:2244–51. 55. Schmelzer, E., Wauthier, E., and Reid, L.M. The phenotypes of pluripotent human hepatic progenitors. Stem Cells, 2006;24:1852–8. 56. Huch, M., Dorrell, C., Boj, S.F. et al. In vitro expansion of single Lgr5+ liver stem cells induced by Wnt‐driven regeneration. Nature, 2013:1–6. 57. Huch, M., Gehart, H., van Boxtel, R. et al. Long‐term culture of genome‐stable bipotent stem cells from adult human liver. Cell, 2015;160:299–312. 58. Dorrell, C., Erker, L., Schug, J. et al. Prospective isolation of a bipotential clonogenic liver progenitor cell in adult mice. Genes Dev, 2011;25:1193–203. 59. Shin, S., Walton, G., Aoki, R. et al. Foxl1‐Cre‐marked adult hepatic progenitors have clonogenic and bilineage differentiation potential. Genes Dev, 2011;25:1185–92. 60. Suzuki, A., Sekiya, S., Onishi, M. et al. Flow cytometric isolation and clonal identification of self‐renewing bipotent hepatic progenitor cells in adult mouse liver. Hepatology, 2008;48:1964–78. 61. Evarts, R.P., Nagy, P., Marsden, E. et  al. A precursor‐product relationship exists between oval cells and hepatocytes in rat liver. Carcinogenesis, 1987;8:1737–40. 62. Evarts, R.P., Nagy, P., Nakatsukasa, H. et  al. In vivo differentiation of rat liver oval cells into hepatocytes. Cancer Res, 1989;49:1541–7. 63. Schofield, R. The relationship between the spleen colony‐forming cell and the haemopoietic stem cell. Blood Cells, 1978;4:7–25. 64. Trentin, J.J. Determination of bone marrow stem cell differentiation by stromal hemopoietic inductive microenvironments (HIM). Am J Pathol, 1971;65:621–8. 65. Ohlstein, B., Kai, T., Decotto, E. et al. The stem cell niche: theme and variations. Curr Opin Cell Biol, 2004;16:693–9. 66. Lapidot, T., Dar, A., and Kollet, O. How do stem cells find their way home? Blood, 2005;106:1901–10. 67. Christiano, A.M. Epithelial stem cells: stepping out of their niche. Cell, 2004;118:530–2. 68. Williams, E.D., Lowes, A.P., Williams, D. et al. A stem cell niche theory of intestinal crypt maintenance based on a study of somatic mutation in colonic mucosa. Am J Pathol, 1992;141:773–6. 69. Tumbar, T., Guasch, G., Greco, V. et  al. Defining the epithelial stem cell niche in skin. Science, 2004;303:359–63. 70. Parati, E.A., Pozzi, S., Ottolina, A. et al. Neural stem cells: an overview. J Endocrinol Invest, 2004;27:64–7. 71. Hering, E. Ueber den Bau der Wirbeltierleber. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften. Mathematisch‐ Naturwissenschaftliche Klasse 1866;Bd. 54:335–341. 72. Steiner, J.W. and Carruthers, J.S. Studies on the fine structure of the terminal branches of the biliary tree: II. Observations of pathologically altered bile canaliculi. Am J Pathol, 1961;39:41–63. 73. Theise, N.D., Saxena, R., Portmann, B.C. et al. The canals of Hering and hepatic stem cells in humans. Hepatology, 1999;30:1425–33. 74. Saxena, R., Theise, N.D., and Crawford, J.M. Microanatomy of the human liver‐exploring the hidden interfaces. Hepatology, 1999;30:1339–46. 75. Yovchev, M.I., Grozdanov, P.N., Joseph, B. et al. Novel hepatic progenitor cell surface markers in the adult rat liver. Hepatology, 2007;45:139–49. 76. Farber, E. Similarities in the sequence of early histological changes induced in the liver of the rat by ethionine, 2‐acetylamino‐fluorene, and 3’‐methyl‐4‐ dimethylaminoazobenzene. Cancer Res, 1956;16:142–8. 77. Lemire, J.M., Shiojiri, N., and Fausto, N. Oval cell proliferation and the origin of small hepatocytes in liver injury induced by D‐galactosamine. Am J Pathol, 1991;139:535–52. 78. Dabeva, M.D. and Shafritz, D.A. Activation, proliferation, and differentiation of progenitor cells into hepatocytes in the D‐galactosamine model of liver regeneration. Am J Pathol, 1993;143:1606–20.

563

  79. Yin, L., Lynch, D., and Sell, S. Participation of different cell types in the restitutive response of the rat liver to periportal injury induced by allyl alcohol. J Hepatol, 1999;31:497–507.   80. Fujio, K., Evarts, R.P., Hu, Z. et al. Expression of stem cell factor and its receptor, c‐kit, during liver regeneration from putative stem cells in adult rat. Lab Invest, 1994;70:511–6.   81. Omori, N., Omori, M., Evarts, R.P. et al. Partial cloning of rat CD34 cDNA and expression during stem cell‐dependent liver regeneration in the adult rat. Hepatology, 1997;26:720–7.   82. Omori, M., Omori, N., Evarts, R.P. et al. Coexpression of flt‐3 ligand/flt‐3 and SCF/c‐kit signal transduction system in bile‐duct‐ligated SI and W mice. Am J Pathol, 1997;150:1179–87.   83. Omori, N., Evarts, R.P., Omori, M. et al. Expression of leukemia inhibitory factor and its receptor during liver regeneration in the adult rat. Lab Invest, 1996;75:15–24.   84. Crosby, H.A., Kelly, D.A., and Strain, A.J. Human hepatic stem‐like cells isolated using c‐kit or CD34 can differentiate into biliary epithelium. Gastroenterology, 2001;120:534–44.   85. Paku, S., Schnur, J., Nagy, P. et al. Origin and structural evolution of the early proliferating oval cells in rat liver. Am J Pathol, 2001;158:1313–23.   86. Sells, M.A., Katyal, S.L., Shinozuka, H. et al. Isolation of oval cells and transitional cells from the livers of rats fed the carcinogen DL‐ethionine. J Natl Cancer Inst, 1981;66:355–62.   87. Akhurst, B., Croager, E.J., Farley‐Roche, C.A. et al. A modified choline‐ deficient, ethionine‐supplemented diet protocol effectively induces oval cells in mouse liver. Hepatology, 2001;34:519–22.   88. Factor, V.M., Radaeva, S.A., Thorgeirsson, S.S. Origin and fate of oval cells in dipin‐induced hepatocarcinogenesis in the mouse. Am J Pathol, 1994;145:409–22.  89. Preisegger, K.H., Factor, V.M., Fuchsbichler, A. et  al. Atypical ductular proliferation and its inhibition by transforming growth factor beta1 in the 3,5‐diethoxycarbonyl‐1,4‐dihydrocollidine mouse model for chronic alcoholic liver disease. Lab Invest, 1999;79:103–9.   90. Yanger, K., Knigin, D., Zong, Y. et al. Adult hepatocytes are generated by self‐duplication rather than stem cell differentiation. Cell Stem Cell, 2014;15:340–9.   91. Wright, N., Samuelson, L., Walkup, M.H. et al. Enrichment of a bipotent hepatic progenitor cell from naive adult liver tissue. Biochem Biophys Res Commun, 2008;366:367–72.   92. Petersen, B.E., Goff, J.P., Greenberger, J.S. et al. Hepatic oval cells express the hematopoietic stem cell marker Thy‐1 in the rat. Hepatology, 1998;27:433–45.   93. Nierhoff, D., Ogawa, A., Oertel, M. et al. Purification and characterization of mouse fetal liver epithelial cells with high in vivo repopulation capacity. Hepatology, 2005;42:130–9.  94. Suzuki, A., Zheng, Y., Kondo, R. et  al. Flow‐cytometric separation and enrichment of hepatic progenitor cells in the developing mouse liver. Hepatology, 2000;32:1230–9.   95. Dezso, K., Jelnes, P., Laszlo, V. et al. Thy‐1 is expressed in hepatic myofibroblasts and not oval cells in stem cell‐mediated liver regeneration. Am J Pathol, 2007;171:1529–37.  96. Tanimizu, N., Nishikawa, M., Saito, H. et  al. Isolation of hepatoblasts based on the expression of Dlk/Pref‐1. J Cell Sci, 2003;116:1775–86.   97. Faris, R.A. and Hixson, D.C. Selective proliferation of chemically altered rat liver epithelial cells following hepatic transplantation. Transplantation, 1989;48:87–92.   98. Dabeva, M.D., Hwang, S.G., Vasa, S.R. et al. Differentiation of pancreatic epithelial progenitor cells into hepatocytes following transplantation into rat liver. Proc Natl Acad Sci USA, 1997;94:7356–61.  99. Dabeva, M.D., Hurston, E., and Shafritz, D.A. Transcription factor and liver‐specific mRNA expression in facultative epithelial progenitor cells of liver and pancreas. Am J Pathol, 1995;147:1633–48. 100. Wang, X., Al‐Dhalimy, M., Lagasse, E. et al. Liver repopulation and correction of metabolic liver disease by transplanted adult mouse pancreatic cells. Am J Pathol, 2001;158:571–9. 101. Wang, X., Foster, M., Al‐Dhalimy, M. et al. The origin and liver repopulating capacity of murine oval cells. Proc Natl Acad Sci USA, 2003;100(1):11881–8. 102. Song, S., Witek, R.P., Lu, Y. et al. Ex vivo transduced liver progenitor cells as a platform for gene therapy in mice. Hepatology, 2004;40:918–24. 103. Roskams, T., van den Oord, J.J., De Vos, R. et al. Neuroendocrine features of reactive bile ductules in cholestatic liver disease. Am J Pathol, 1990;137:1019–25.

564

THE LIVER:  REFERENCES

104. Demetris, A.J., Seaberg, E.C., Wennerberg, A. et al. Ductular reaction after submassive necrosis in humans. Special emphasis on analysis of ductular hepatocytes. Am J Pathol, 1996;149:439–48. 105. Roskams, T., De Vos, R., Van Eyken, P. et al. Hepatic OV‐6 expression in human liver disease and rat experiments: evidence for hepatic progenitor cells in man. J Hepatol, 1998;29:455–63. 106. Roskams, T.A., Theise, N.D., Balabaud, C. et al. Nomenclature of the finer branches of the biliary tree: canals, ductules, and ductular reactions in human livers. Hepatology, 2004;39:1739–45. 107. Zhou, H., Rogler, L.E., Teperman, L. et al. Identification of hepatocytic and bile ductular cell lineages and candidate stem cells in bipolar ductular reactions in cirrhotic human liver. Hepatology, 2007;45:716–24. 108. Haruna, Y., Saito, K., Spaulding, S. et al. Identification of bipotential progenitor cells in human liver development. Hepatology, 1996;23:476–81. 109. Malhi, H., Irani, A.N., Gagandeep, S. et al. Isolation of human progenitor liver epithelial cells with extensive replication capacity and differentiation into mature hepatocytes. J Cell Sci, 2002;115:2679–88. 110. Lazaro, C.A., Croager, E.J., Mitchell, C. et al. Establishment, characterization, and long‐term maintenance of cultures of human fetal hepatocytes. Hepatology, 2003;38:1095–106. 111. Mahieu‐Caputo, D., Allain, J.E., Branger, J. et al. Repopulation of athymic mouse liver by cryopreserved early human fetal hepatoblasts. Hum Gene Ther, 2004;15:1219–28. 112. Barker, N., van Es, J.H., Kuipers, J. et  al. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature, 2007;449:1003–7. 113. Sato, T., Vries, R.G., Snippert, H.J. et al. Single Lgr5 stem cells build crypt‐ villus structures in vitro without a mesenchymal niche. Nature, 2009;459:262–5. 114. Schaub, J.R., Malato, Y., Gormond, C. et al. Evidence against a stem cell origin of new hepatocytes in a common mouse model of chronic liver injury. Cell Rep, 2014;8:933–9. 115. Tarlow, B.D., Pelz, C., Naugler, W.E. et al. Bipotential adult liver progenitors are derived from chronically injured mature hepatocytes. Cell Stem Cell, 2014;15:605–18. 116. Grompe, M. Liver stem cells, where art thou? Cell Stem Cell, 2014;15:257–8. 117. Schaub, J.R., Huppert, K.A., Kurial, S.N.T. et al. De novo formation of the biliary system by TGFbeta‐mediated hepatocyte transdifferentiation. Nature, 2018;557:247–51. 118. Michalopoulos, G.K., Barua, L., and Bowen, W.C. Transdifferentiation of rat hepatocytes into biliary cells after bile duct ligation and toxic biliary injury. Hepatology, 2005;41:535–44. 119. Yovchev, M.I., Locker, J., and Oertel, M. Biliary fibrosis drives liver repopulation and phenotype transition of transplanted hepatocytes. J Hepatol, 2016;64:1348–57. 120. Katsuda, T., Kawamata, M., Hagiwara, K. et al. Conversion of terminally committed hepatocytes to culturable bipotent progenitor cells with regenerative capacity. Cell Stem Cell, 2017;20:41–55. 121. Raven, A., Lu, W.Y., Man, T.Y. et al. Cholangiocytes act as facultative liver stem cells during impaired hepatocyte regeneration. Nature, 2017; 547:350–4. 122. Deng, X., Zhang, X., Li, W. et al. Chronic liver injury induces conversion of biliary epithelial cells into hepatocytes. Cell Stem Cell, 2018;23: 114–22. 123. Wang, B., Zhao, L., and Fish, M. et al. Self‐renewing diploid Axin2(+) cells fuel homeostatic renewal of the liver. Nature, 2015;524:180–5. 124. Lin, S., Nascimento, E.M., Gajera, C.R. et  al. Distributed hepatocytes expressing telomerase repopulate the liver in homeostasis and injury. Nature, 2018;556:244–8. 125. Li, B., Dorrell, C., Canaday, P.S. et al. Adult mouse liver contains two distinct populations of cholangiocytes. Stem Cell Reports, 2017;9:478–89. 126. Goodell, M.A. Stem‐cell “plasticity”: befuddled by the muddle. Curr Opin Hematol, 2003;10:208–13. 127. Wagers, A.J. and Weissman, I.L. Plasticity of adult stem cells. Cell, 2004;116:639–48. 128. Thorgeirsson, S.S. and Grisham, J.W. Hematopoietic cells as hepatocyte stem cells: a critical review of the evidence. Hepatology, 2006;43:2–8. 129. Petersen, B.E., Bowen, W.C., Patrene, K.D. et al. Bone marrow as a potential source of hepatic oval cells. Science, 1999;284:1168–70. 130. Menthena, A., Deb, N., Oertel, M. et al. Bone marrow progenitors are not the source of expanding oval cells in injured liver. Stem Cells, 2004;22:1049–61.

131. Terada, N., Hamazaki, T., Oka, M. et al. Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature, 2002;416:542–5. 132. Ying, Q.L., Nichols, J., Evans, E.P. et al. Changing potency by spontaneous fusion. Nature, 2002;416:545–8. 133. Wang, X., Willenbring, H., Akkari, Y. et  al. Cell fusion is the principal source of bone‐marrow‐derived hepatocytes. Nature, 2003;422:897–901. 134. Vassilopoulos, G., Wang, P.R., and Russell, D.W. Transplanted bone marrow regenerates liver by cell fusion. Nature, 2003;422:901–4. 135. Alvarez‐Dolado, M., Pardal, R., Garcia‐Verdugo, J.M. et  al. Fusion of bone‐marrow‐derived cells with Purkinje neurons, cardiomyocytes and hepatocytes. Nature, 2003;425:968–73. 136. Weimann, J.M., Johansson, C.B., Trejo, A. et al. Stable reprogrammed heterokaryons form spontaneously in Purkinje neurons after bone marrow transplant. Nat Cell Biol, 2003;5:959–66. 137. Camargo, F.D., Finegold, M., and Goodell, M.A. Hematopoietic myelomonocytic cells are the major source of hepatocyte fusion partners. J Clin Invest, 2004;113:1266–70. 138. Willenbring, H., Bailey, A.S., Foster, M. et al. Myelomonocytic cells are sufficient for therapeutic cell fusion in liver. Nat Med, 2004;10:744–8. 139. Camargo, F.D., Green, R., Capetanaki, Y. et al. Single hematopoietic stem cells generate skeletal muscle through myeloid intermediates. Nat Med, 2003;9:1520–7. 140. Newsome, P.N., Johannessen, I., Boyle, S. et al. Human cord blood‐derived cells can differentiate into hepatocytes in the mouse liver with no evidence of cellular fusion. Gastroenterology, 2003;124:1891–900. 141. Harris, R.G., Herzog, E.L., Bruscia, E.M. et al. Lack of a fusion requirement for development of bone marrow‐derived epithelia. Science, 2004;305:90–3. 142. Jang, Y.Y., Collector, M.I., Baylin, S.B. et al. Hematopoietic stem cells convert into liver cells within days without fusion. Nat Cell Biol, 2004; 6:532–9. 143. Danet, G.H., Luongo, J.L., Butler, G. et al. C1qRp defines a new human stem cell population with hematopoietic and hepatic potential. Proc Natl Acad Sci USA, 2002;99:10441–5. 144. Wang, X., Ge, S., McNamara, G. et al. Albumin‐expressing hepatocyte‐like cells develop in the livers of immune‐deficient mice that received transplants of highly purified human hematopoietic stem cells. Blood, 2003;101:4201–8. 145. Kakinuma, S., Tanaka, Y., Chinzei, R. et al. Human umbilical cord blood as a source of transplantable hepatic progenitor cells. Stem Cells, 2003; 21:217–27. 146. Kollet, O., Shivtiel, S., Chen, Y.Q. et  al. HGF, SDF‐1, and MMP‐9 are involved in stress‐induced human CD34+ stem cell recruitment to the liver. J Clin Invest, 2003;112:160–9. 147. Jiang, Y., Jahagirdar, B.N., Reinhardt, R.L. et al. Pluripotency of mesenchymal stem cells derived from adult marrow. Nature, 2002;418:41–9. 148. Schwartz, R.E., Reyes, M., Koodie, L. et al. Multipotent adult progenitor cells from bone marrow differentiate into functional hepatocyte‐like cells. J Clin Invest, 2002;109:1291–302. 149. Lee, O.K., Kuo, T.K., Chen, W.M. et al. Isolation of multipotent mesenchymal stem cells from umbilical cord blood. Blood, 2004;103:1669–75. 150. Anjos‐Afonso, F., Siapati, E.K., and Bonnet, D. In vivo contribution of murine mesenchymal stem cells into multiple cell‐types under minimal damage conditions. J Cell Sci, 2004;117:5655–64. 151. Kogler, G., Sensken, S., Airey, J.A. et al. A new human somatic stem cell from placental cord blood with intrinsic pluripotent differentiation potential. J Exp Med, 2004;200:123–35. 152. Lee, K.D., Kuo, T.K., Whang‐Peng, J. et al. In vitro hepatic differentiation of human mesenchymal stem cells. Hepatology, 2004;40:1275–84. 153. Aurich, I., Mueller, L.P., Aurich, H. et al. Functional integration of hepatocytes derived from human mesenchymal stem cells into mouse livers. Gut, 2007;56:405–15. 154. Sato, Y., Araki, H., Kato, J. et al. Human mesenchymal stem cells xenografted directly to rat liver are differentiated into human hepatocytes without fusion. Blood, 2005;106:756–63. 155. Banas, A., Teratani, T., Yamamoto, Y. et al. Adipose tissue‐derived mesenchymal stem cells as a source of human hepatocytes. Hepatology, 2007;46:219–28. 156. Sgodda, M., Aurich, H., Kleist, S. et al. Hepatocyte differentiation of mesenchymal stem cells from rat peritoneal adipose tissue in vitro and in vivo. Exp Cell Res, 2007;313:2875–86.



44:  Liver Repopulation by Cell Transplantation and the Role of Stem Cells in Liver Biology

157. Takahashi, K., Tanabe, K., Ohnuki, M. et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 2007;131:861–72. 158. Yu, J., Vodyanik, M.A., Smuga‐Otto, K. et al. Induced pluripotent stem cell lines derived from human somatic cells. Science, 2007;318:1917–20. 159. Hamazaki, T., Iiboshi, Y., Oka, M. et al. Hepatic maturation in differentiating embryonic stem cells in vitro. FEBS Lett, 2001;497:15–9. 160. Jones, E.A., Tosh, D., Wilson, D.I. et al. Hepatic differentiation of murine embryonic stem cells. Exp Cell Res, 2002;272:15–22. 161. Yamada, T., Yoshikawa, M., Kanda, S. et  al. In vitro differentiation of embryonic stem cells into hepatocyte‐like cells identified by cellular uptake of indocyanine green. Stem Cells, 2002;20:146–54. 162. Yamamoto, H., Quinn, G., Asari, A. et  al. Differentiation of embryonic stem cells into hepatocytes: biological functions and therapeutic application. Hepatology, 2003;37:983–93. 163. Rambhatla, L., Chiu, C.P., Kundu, P. et al. Generation of hepatocyte‐like cells from human embryonic stem cells. Cell Transplant, 2003;12:1–11. 164. Kubo, A., Shinozaki, K., Shannon, J.M. et al. Development of definitive endoderm from embryonic stem cells in culture. Development, 2004;131:1651–62. 165. Gouon‐Evans, V., Boussemart, L., Gadue, P. et al. BMP‐4 is required for hepatic specification of mouse embryonic stem cell‐derived definitive endoderm. Nat Biotechnol, 2006;24:1402–11. 166. Heo, J., Factor, V.M., Uren, T. et  al. Hepatic precursors derived from murine embryonic stem cells contribute to regeneration of injured liver. Hepatology, 2006;44:1478–86. 167. Basma, H., Soto‐Gutierrez, A., Yannam, G.R. et  al. Differentiation and transplantation of human embryonic stem cell‐derived hepatocytes. Gastroenterology, 2009;136:990–9. 168. Suzuki, A. Artificial induction and disease‐related conversion of the hepatic fate. Curr Opin Genet Dev, 2013;23:579–84. 169. Du, Y., Wang, J., Jia, J. et al. Human hepatocytes with drug metabolic function induced from fibroblasts by lineage reprogramming. Cell Stem Cell, 2014;14:394–403.

565

170. Huang, P., Zhang, L., Gao, Y. et  al. Direct reprogramming of human fibroblasts to functional and expandable hepatocytes. Cell Stem Cell, ­ 2014;14:370–84. 171. Sekiya, S. and Suzuki, A. Direct conversion of mouse fibroblasts to ­hepatocyte‐like cells by defined factors. Nature, 2011;475:390–3. 172. Zhu, S., Rezvani, M., Harbell, J. et  al. Mouse liver repopulation with ­hepatocytes generated from human fibroblasts. Nature, 2014;508:93–7. 173. Vierbuchen, T., Ostermeier, A., Pang, Z.P. et  al. Direct conversion of ­fibroblasts to functional neurons by defined factors. Nature, 2010;463: 1035–41. 174. Roy‐Chowdhury, N., Wang, X., Guha, C. et  al. Hepatocyte‐like cells derived from induced pluripotent stem cells. Hepatol Int, 2017;11:54–69. 175. Sakaida, I., Terai, S., Yamamoto, N. et al. Transplantion of bone marrow cells reduces CCl4‐induced liver fibrosis in mice. Hepatology, 2004;40:1304–11. 176. Ueno, T., Nakamura, T., Torimura, T. et  al. Angiogenic cell therapy for hepatic fibrosis. Med Mol Morphol, 2006;39:16–21. 177. Higashiyama, R., Inagaki, Y., Hong, Y.Y. et al. Bone marrow‐derived cells express matrix metalloproteinases and contribute to regression of liver fibrosis in mice. Hepatology, 2007;45:213–22. 178. Taniguchi, E., Kin, M., Torimura, T. et  al. Endothelial progenitor cell transplantation improves the survival following liver injury in mice. ­ Gastroenterology, 2006;130:521–31. 179. Yoshida, Y., Tokusashi, Y., Lee, G.H. et al. Intrahepatic transplantation of normal hepatocytes prevents Wilson’s disease in Long‐Evans cinnamon rats. Gastroenterology, 1996;111:1654–60. 180. Peterson, E.A., Polgar, Z., Devakanmalai, G.S. et al. Genes and pathways promoting long‐term liver repopulation by ex vivo hYAP‐ERT2 transduced hepatocytes and treatment of jaundice in Gunn rats. Hepatol Commun, 2018;3:129–46.

45

Liver Regeneration George K. Michalopoulos Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

INTRODUCTION Liver is the largest organ of the body. It functions as a portal of entry and metabolic processing for all substances absorbed through the gastrointestinal tract. It is the major biochemical transducer for body homeostasis, triaging the absorbed substances from the gut for storage, elimination, transformation into other types of organic chemicals, or packaged transport to the bloodstream for whole body utilization. To the extent that muscles work mostly on fatty acids and brain primarily on glucose, liver controls the availability of both the nutrients. The key role of liver as the “provider” for metabolism of all body tissues is perhaps the reason for an established tight relationship between liver weight and body weight. Changes in body status (cachexia, puberty, pregnancy, chronic disease, etc.) typically alter the algorithm of the “liver to body weight ratio” (LBWR). Whatever the operative algorithm of LBWR may be for a given body at any given time, the LBWR is tightly maintained, by mechanisms not fully understood. The rates of proliferation of hepatic cells under normal conditions are typically very low (for hepatocytes, less than 0.2%), but they exceed zero. Recent studies have demonstrated that under the placid calmness of the hepatic capsule, there are slow proliferative events of cells from different zones of the lobule, aiming to keep liver cell numbers and organ weight steady. The precise mechanisms controlling these processes are not fully understood, but for operative reasons, we need to accept the existence of a set of controls comprising a “hepatostat”, a set of processes that ascertain maintenance of liver weight to where it needs to be [1]. Such demands for maintenance of steady weight exist in endocrine glands controlled by the pituitary, and the feedbacks controlling those homeostatic events are better understood. Other organs (pancreas, kidneys, intestine, lungs, etc.) will increase slightly

in mass if a big portion of the organ (one kidney, one lung, part of pancreas) is lost, but the residual organ tissue will not attain the exact total weight that existed prior to the tissue loss. This is not the case for liver! If a major portion of liver tissue is acutely lost, liver will enter into a regenerative process so that the total hepatostat‐driven LWBR is precisely reestablished in the status quo ante. This is the process of liver regeneration (LR) that will be majorly examined in this chapter. In a clinical disease setting, regeneration is manifested in conditions leading to severe loss of hepatocytes. Chronic loss of hepatocytes is seen in infectious diseases (e.g. HCV and HBV viruses), chronic toxic conditions (e.g. alcohol, metabolic diseases, non‐alcoholic steatohepatitis (NASH), storage diseases, hemochromatosis, alpha‐1 antitrypsin deficiency), ischemia‐ reperfusion injury (most often following liver transplantation), or chronic immune attacks (autoimmune hepatitis). Chronic loss of hepatocytes is accompanied by compensatory proliferation of the surviving hepatocytes. As we will address below in this chapter, this occurs in potentially genotoxic environments and the chronic compensatory proliferation of the surviving hepatocytes may lead to development of neoplasia. Acute loss of hepatocytes is rarer, often caused by acute ingestion of toxins (e.g. attempted suicide by acetaminophen), trauma, or a short course of an acute hepatitis. In all these conditions, death of hepatocytes and compensatory proliferation proceed in tandem with inflammatory processes which aim to remove dead hepatocytes and provide cytokines for tissue repair. Overall, the regenerative processes in acute loss of hepatocytes operate very efficiently. Studies of patients with Dubin–Johnson (pigment similar to melanin accumulating in normal hepatocytes) have shown that approximately 90% of hepatocytes die in an acute hepatitis, but they are restored in proper numbers by the end of the disease [2]. It is highly likely that in a clinical setting, the inflammatory processes and the

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



45:  Liver Regeneration

567

regenerative events cooperate synergistically. It is difficult, however, to design experiments that would cleanly dissect the regulatory signals regulating only the compensatory proliferation of hepatocytes, from the comingled inflammatory events. Experimental models in rodents associated with acute administration of a liver toxin (typically CCl4 or acetaminophen) are more relevant to the ­clinical setting. The study of signals exclusively associated with regulation of hepatic regeneration, not affected by inflammatory processes, has been carried mostly in rodent livers. Each rodent liver is composed of several lobes, and each lobe is supported by its own vasculature (arterial and portal vein) and separate branch of the common bile duct. An easy surgical procedure, first described by Higgins and Anderson in 1933 [3], allows clean removal of the two larger lobes, comprising around two‐thirds of the liver. The non‐resected small lobes remain intact, with no damage or associated inflammatory response. In rats, this procedure allows removal of about 68% of the liver mass, whereas in mice, due to the presence of a gallbladder, typically about 60% of liver mass can be easily removed. The procedure is known as “2/3 partial hepatectomy” (PHx) and the results obtained by using this procedure will be the basis for discussion of regulatory signals associated with LR. At the end of the regenerative process, the non‐resected lobes become larger and, in aggregate, restore the hepatic mass that was present prior to PHx. The regenerated liver has a different shape; it now has fewer lobes (two or three, based on anatomic conventions), and the resected lobes are not restored.

Table 45.1  Mitogenic signals associated with liver regeneration after partial hepatectomy

SIGNALS REGULATING LR

Hepatocyte growth factor (HGF) and its receptor (c‐Met)

Signals related to LR have been identified by triggering proliferation in hepatocyte cultures, delay of regeneration associated with signal ablation in genomically modified mice, or response of intact liver in vivo by injecting specific signaling substances [4, 5]. Many of these signals appear not only in the regenerating liver, but also in the peripheral blood. Table  45.1 shows a list of the most studied mitogenic signals associated with LR. Of these signals, epidermal growth factor receptor (EGFR) and associated ligands (epidermal growth factor [EGF], transforming growth factor [TGFα], heparin binding epidermal growth factor [HBEGF], amphiregulin) and hepatocyte growth factor (HGF) and its receptor (MET) are capable of inducing proliferation of hepatocytes in serum‐free cultures [6]. They also induce hepatocyte DNA synthesis when injected into the portal vein of normal rodent livers. They should be viewed as “direct mitogens”. Others show their relevance by delays in LR following elimination of the signaling molecule or its cognate receptor. They are not directly mitogenic for hepatocytes in culture or in vivo. They should be viewed as “indirect mitogens”. Some more recently discovered signals (e.g. Wnt and Hedgehog) have not been tested in vivo or in vitro but they do have strong effects on LR and homeostasis. In this chapter, we will discuss these signals with emphasis on their effects on LR. It should be noted that LR always finds pathways to bypass elimination of any single mitogenic signal, direct or indirect, and proceeds to completion by following alternate pathways overcoming the signaling obstacle.

Complete mitogens 1. Mitogenic in hepatocyte cultures in chemically defined (serum‐free) media 2. Cause liver enlargement and hepatocyte DNA synthesis when injected into whole animals: Hepatocyte Growth Factor (HGF) and receptor (c‐Met) Ligands of the EGF R (EGF, TGFα, HB‐EGF, amphiregulin) Combined elimination of these two signaling receptors abolishes liver regeneration Auxiliary mitogens 1. Ablation of their signaling pathways causes delay but does not abolish liver regeneration. 2. They are not mitogenic in hepatocyte cultures and when injected in vivo do not cause hepatocyte DNA synthesis and liver enlargement Norepinephrine and the α1 adrenergic receptor TNF and TNFR1 IL6 Notch and Jagged (recombinant Jagged causes DNA synthesis in hepatocyte cultures) VEGF and receptors I and II Bile acids Serotonin Complement proteins Leptin Insulin PPAR gamma Very important and yet to be defined Elimination of these signaling molecules does not abolish liver regeneration. Their effects by exogenous administration to intact animals or hepatocyte cultures have not been fully tested Hedgehog Wnt/beta‐catenin Signals controlling Hippo pathway and Yap

HGF was first isolated on the basis of its capacity to induce DNA synthesis in primary cultures of hepatocytes [7]. Subsequent cloning and sequencing revealed an unusual structure, composed of four Kringle domains and a pseudo‐protease domain [8, 9]. There is no active protease function in HGF however, despite its otherwise strong homology to plasminogen. It is synthesized as a single polypeptide, cleaved to its active heterodimeric form in an Arg‐Val‐Val site, distal to the Kringle domains (identical to plasminogen). Urokinase plasminogen activator (uPA), free or bound to its receptor (uPAR) directly activates HGF; tissue plasminogen activator (TPA) can also activate HGF, but not as effectively [10]. HGF can also be activated by a soluble protein with high homology to Factor XII, known as HGF activator (HGFA). Hepsin and matriptase have also been shown to activate HGF. In normal liver, HGF is produced by hepatic stellate cells and it is deposited in large amounts in the extracellular matrix (ECM), mostly in the periportal region of the lobule, bound to glycosaminoglycans and specific collagens [11]. HGF production by stellate cells is controlled by the neurotrophin receptor p75NTR [12]. At later stages of LR, it is also produced by sinusoidal endothelial cells and bone marrow endothelial progenitor cells homing to the liver [13, 14]. HGF is produced in most peripheral tissues by mesenchymal cells; it is also produced by specific populations of neurons and glial cells in both brain (hippocampus, frontal

568

THE LIVER:  SIGNALS REGULATING LR

and parietal cortex) and spinal cord [15] and also in the ­placental syncytiotrophoblast [16]. The HGF receptor, MET, is the product of the gene c‐Met. It is also synthesized as a long polypeptide and activated into its dimeric form at the Golgi apparatus. It resides as a transmembrane protein in the plasma membrane of most epithelial cells, selective groups of neurons and glial cells, as well macrophages and endothelial cells. Both hepatocytes and cholangiocytes express MET. Activation of MET is typically the result of ligation with active HGF. This results in formation of HGF–MET dimers, based on the dimerization domain present on HGF. The formation of the HGF– MET dimers is associated with cross‐phosphorylation of tyrosine amino acids at sites 1234 and 1235. This is followed by  phosphorylation of tyrosine in position 1349, which then becomes a docking site for attachment of other proteins (including GAB1, GRB2), responsible for a multiplicity of downstream signals, which eventually cause activation of PIPK3, AKT (protein kinase B), and mammalian target of rapamycin (mTOR), rearrangement of cell polarity and enhancement of motility and/ or stimulation of cell proliferation [17]. The “signalosome” formation around Gab1 is already assembled within 30 minutes after PHx [18]. Hepatocyte MET appears always activated in the liver, presumably due to availability of HGF in abundance in the hepatic ECM [19]. Incomplete forms of HGF containing only one or two Kringles (NK1, NK2) can also ligate MET, but since they do not contain the dimerization domain of complete HGF, dimers can only form and MET activation occurs only in the presence of glycosaminoglycans.

Epidermal growth factor receptor (EGFR) and associated ligands EGFR is expressed in both hepatocytes and cholangiocytes. It is a member of the ErbB family of receptors. ERB1 (EGFR) and ERB3 are expressed in adult hepatocytes and cholangiocytes, whereas fetal hepatocytes express the above as well as ErbB2 (Her2/Neu) [20]. ERB3 does not have a kinase domain and ErbB4 is not expressed in liver. In contrast to HGF/MET, EGFR and ERB3 can form heterodimers with themselves or with each other without being ligated. There are multiple ligands of EGFR. The ones studied in liver and in the context of LR, primarily include EGF, TGFα, amphiregulin, and HB‐EGF. EGF is produced in secretions of exocrine glands, including salivary glands. Brunner’s glands, with histology similar to salivary glands, exist in the submucosa of the duodenum. They secrete EGF in the intestinal lumen. A portion of the secreted EGF is absorbed intact from the lumen of the duodenum and transported to the liver via the portal circulation [21]. Thus, hepatocytes are constantly exposed to EGF, and EGFR is found activated in normal liver [19]. EGF secretion from Brunner’s glands is enhanced by norepinephrine [22]. TGFα is minimally expressed in resting normal liver but its expression is maximally enhanced following PHx [23]. It is produced as a transmembrane protein and the mature form is the extracellular domain, released by proteolytic action by TACE/ADAM17 protease. Germline elimination of TGFα does not affect liver generation, presumably due to complementary effects by the other EGFR ligands involved in the process. Amphiregulin, a downstream

target of Yap protein, is also minimally expressed in normal liver, but its expression in hepatocytes rapidly increases after PHx. Germline elimination of amphiregulin delays liver generation [24]. Heparin binding EGF (HB‐EGF) is produced by macrophages and endothelial cells but not in hepatocytes [25]. It is also produced as a transmembrane protein, and the extracellular domain is released by the action of metalloproteinases as the mature form of HB‐EGF. It rises rapidly after PHx, prior to the appearance of TGFα. Germline elimination of HB‐EGF delays LR and the progression of the steps through the cell cycle [26]. In the normal liver, EGF is the major interacting ligand with EGFR. TGFα, amphiregulin, and HB‐EGF are rapidly mobilized after PHx and they have complementary functions, with neither one individually being critical for regeneration, but having effects related both to hepatocyte proliferation as well as to proliferation of cholangiocytes and endothelial cells, all of which abundantly express EGFR.

Fibroblast growth factors (FGF) and their receptors FGF is a large family of signaling proteins, composed of 23 members with diverse structures and cells or tissues of origin. Their common feature is their capacity to activate some or all of four receptors (FGFR1–FGR4). Mature hepatocytes express only FGFR4 [27], with the other receptors present in non‐epithelial hepatic cells. FGF1/2 are expressed by hepatocytes during LR [5]. They are slightly mitogenic in hepatocyte cultures (less than 20% of the effect of HGF or EGF). It is not clear whether, during LR, FGF1/2 are exerting autocrine or paracrine effects, especially on endothelial cells expressing FGFR1/2. In hepatocytes, FGFR4 is co‐expressed and functioning in conjunction with βKlotho. Elimination of either βKlotho or FGFR4 s had no effect on hepatocyte proliferation but it disrupted hepatic cholesterol, bile acid, and lipid metabolism [28]. In another study, however, knockdown of FGFR4 was associated with increased necrosis of non‐parenchymal cells and 25% mortality after PHx; liver weight was restored by hepatocyte hypertrophy [29]. The role of FGFR4 is crucial in hepatocyte biology. It is the receptor ligated by FGF15 (human FGF19), produced in the intestine following stimulation by bile acids binding to the intestinal farnesoid X receptor (FXR) [30]. Functioning as an endocrine signal, FGF15 enters through the portal circulation and regulates cholesterol metabolism and suppresses hepatocyte bile acid synthesis via FGFR4, by downregulating CYP7a1, the rate limiting enzyme for bile acid synthesis. Elimination of FGF15 caused massive increase of hepatic bile acids and increased mortality after PHx with delayed LR [31]. FGF21 is another endocrine signal, often viewed as a “hepatokine”, produced by hepatocytes. Its synthesis is controlled by PPAR’α and PGC‐1a. It is also produced in testes, pancreas, and adipose tissue. It has insulin‐mimetic effects and exerts general metabolic effects in a variety of tissues, including adipose tissue [32]. Stellate cells both produce and respond to FGF; evidence suggests that FGF regulate synthesis of ECM components by stellate cells [33]. All evidence suggests that FGF and their receptors are crucial for regulation of hepatic function, proliferation of non‐ parenchymal cells, and regulation of key metabolic functions in normal and regenerating liver.



45:  Liver Regeneration

569

Transforming growth factor beta (TGFβ)

Interleukin 6 (IL6)

There are three different forms of TGFβ, all present in the liver. TGFβ receptor is the same for all three forms, composed of three components (TGFβr I, II, and III), present in all hepatic cell types. TGFβR II is the site in which TGFβ binds directly. In liver, TGFβ1 has been the most studied and it is the one to be referred to in this segment. It is produced by stellate cells and Kupffer cell macrophages. In normal liver, TGFβ is bound to its receptor and also to decorin, a GPI‐linked protein on plasma membrane. Following PHx, TGFβ is gradually removed from hepatocytes as regeneration proceeds from periportal to pericentral areas of the lobule [34]. It is also massively increasing in the plasma, probably released from decorin as a result of the early remodeling of ECM following PHx (see below, section on Urokinase plasminogen activator (uPA) and ECM remodeling) [4]. Circulating alpha‐2 macroglobulin binds TGFβ and transports it to hepatocytes where it is inactivated [4, 5, 35]. TGFβ is mitoinhibitory in hepatocyte cultures, and its removal from liver soon after PHx would be a rational expectation, in order to allow hepatocytes to proliferate. Administration of exogenous TGFβ immediately after PHx delays initiation of hepatocyte DNA synthesis. Surprisingly, however, TGFβ expression increases at about 3 hours after PHx and reaches a plateau at 72 hours, at the same time as hepatocytes are at their maximal state of proliferation [36]. The successful completion of hepatocyte proliferation as TGFβ levels rise, is accomplished by the simultaneous decrease in expression of hepatocyte TGFβ receptors at the time of hepatocyte proliferation [37]. The contemporaneous rise in plasma norepinephrine likely also contributes to the hepatocyte “resistance” to TGFβ; as evidenced by studies in hepatocyte cultures, norepinephrine attenuates the effect of TGFβ and suppresses its mitoinhibitory activity [38]. Studies in hepatic organoid cultures show that the increase in TGFβ expression after PHx is linked to activation of EGFR and MET [39]. TGFβ plays an important role at the later stages of regeneration, though not by affecting termination of liver regeneration (see below, section on Termination of liver regeneration). TGFβ is important in angiogenesis, inducing formation of tubules by endothelial cells [35]. It also likely stimulates production of ECM by stellate cells, a phenomenon that occurs at the later stages of LR, restoring ECM after its remodeling immediately following PHx [40]. All evidence suggests that TGFβ expression during LR is a constructive event and important for the completion of LR by restoration of intact tissue histology. This is also bolstered by the finding that loss of beta‐2 spectrin, a component of the TGFβ signaling pathway, is associated with suppressed and delayed liver regeneration [41]. Also of interest, is a previous study in which normal rats were injected with a dominant negative construct against TGFβ receptor II. This, unexpectedly, resulted in initiation of DNA synthesis in the normal liver. The results suggest that under normal conditions, hepatocytes are under opposite “tonic” influences between ambient mitogens HGF, EGF, and so on, and the mitoinhibitory effect of TGFβ, with the combined effect resulting in maintaining hepatocytes in G0 state and in a stable state of differentiation; acute removal of TGFβ signaling creates an imbalance driving hepatocytes toward DNA ­synthesis [42].

IL6 has been studied extensively in hepatocytes, as the major ligand triggering the “acute phase response”, a rapid increase and secretion of many proteins associated with innate immunity. Circulating IL6 binds to a soluble receptor, and the complex binds to a hepatocyte receptor known as gp130, which is shared with oncostatin M, LIF, CNTF, and so on. This trimeric complex dimerizes to form a hexameric complex and is subsequently phosphorylated in Tyr residues, becoming a docking site for activation of JAK tyrosine kinases and STAT transcription factors. In the context of LR, IL6 is the major regulator of activation of STAT3 transcription factor. Germline elimination of IL6 is associated with delayed regeneration due to delayed STAT3 activation. LR however proceeds and completes normally, as other signals (MET, EGFR) can also phosphorylate and activate STAT3 [43–45]. IL6 is produced by Kupffer cells, but also by hepatocytes [46]. It rises in plasma after PHx, following the rise in circulating tumor necrosis factor (TNF), a major stimulus for IL6 secretion [47].

Tumor necrosis factor alpha (TNF) TNF is produced primarily by macrophages and binds to two receptors (TNFR1, TNFR2). Direct administration of TNF to normal animals leads to liver damage. TNF, however, rapidly rises in the plasma after PHx, with no apparent damage to the liver [48]. Studies by Fausto et al. have shown that germline elimination of either TNFR1 or TNFR2 leads to delayed LR and decreased production of IL6 [47]. Perhaps the most important function of TNF in LR is the activation of the transcription factor NFκB, which is delayed in TNFR knockout (KO) mice [49]. TNF also regulates expression of TACE/ ADAMS17, the plasma membrane protease associated with secretion of the mature (extracellular) form of TGF [50]. It is of interest that either TNFR1 or FAS activation can induce complete liver failure, independent of each other, and yet, similar to TNFR, germ line elimination of FAS also leads to delayed LR [51].

Norepinephrine (NE) This neurotransmitter is produced at the synaptic endings and, to a lesser extent, by adrenal medulla. Stellate cells also produce and respond to NE [52]. Addition of NE in serum‐free hepatocyte cultures dramatically enhances the mitogenic effects of EGF and HGF [53]. In addition, it downregulates the mitoinhibitory effects of TGFβ [38]. In “balanced” concentrations of mitogen EGF and mito‐inhibitor TGFβ, addition of NE leads to hepatocyte DNA synthesis [38]. NE acts by binding to the G‐protein coupled alpha‐1 adrenergic receptor (A1AR). Its effects on hepatocyte proliferation are mediated, at least in part, by A1AR inducing phosphorylation of EGFR and activation of STAT3 via Src kinase [54]. NE rises rapidly in plasma after PHx and exerts the above effects directly on hepatocytes at the time when both EGFR and MET are becoming increasingly activated [55]. NE, however, has effects on LR from extrahepatic sites. NE increases availability of both EGF and HGF to the regenerating liver by directly enhancing secretion of EGF from

570

THE LIVER:  SIGNALS REGULATING LR

Brunner’s duodenal glands [22] and production of HGF by mesenchymal cells [56]. Given the high levels of NE in plasma after PHx, it may be the mediating factor for the observed increased in HGF production in extrahepatic sites (lungs, kidney) after PHx [57]. Administration of prazosin, a specific A1AR inhibitor, has long‐lasting (three days) effects of suppression of hepatocyte DNA synthesis after PHx. Similar effects were seen after surgical sympathectomy of the liver prior to PHx [55].

Bile acids Bile acids increase in circulating blood after PHx and depletion of bile acids leads to decreased regeneration [58]. The rise of bile acids in plasma, however, occurs several hours after PHx, suggesting that bile acids act after LR has already been initiated. Bile acids bind to the transcription factor FXR, resulting in a feedback suppression of bile acid synthesis by downregulation of CYP7a1, the first enzyme involved in bile acid synthesis. Mice with genetic elimination of FXR have defective regeneration [59]. Despite the evidence of suppressed regeneration following bile acid depletion or germline elimination of FXR, the mediating pathways are not clear. It was originally thought that the effects of bile acids on LR were mediated by the production of FGF15 in the intestine (mediated by bile acids binding to intestinal FXR). As mentioned above, FGF15 acts as an endocrine FGF and regulates many aspects of lipid metabolism, including synthesis of bile acids, via receptor FGR4. Germline elimination of FGF15, however, has minimal effects on LR. FXR elimination and its effects on LR are probably mediated by the uncontrolled biosynthesis of bile acids observed in the FXR KO mice, associated with cholestatic damage of hepatic tissue. In FXR KO mice, bile acid synthesis is not inhibited neither via FXR, nor by FGF15 acting on FGFR4, because synthesis of FGF15 is also suppressed in FXR KO mice. There is recent evidence that bile acids may play a direct role in cholangiocyte proliferation via the G‐protein coupled receptor TGR5 [60].

Serotonin Thrombocytopenic mice have deficient LR and this is partially corrected by administration of serotonin [61]. Administration of serotonin to normal liver or hepatocytes in culture does not have direct mitogenic effects on hepatocytes. Mice with low levels of serotonin (deficient in tyrosine hydroxylase) also have deficient LR [61]. On the other hand, mice with germline elimination of the serotonin transporter resulting in severe decrease of serotonin in platelets and peripheral blood, do not have deficient LR [62]. Serotonin has multiple beneficial effects on liver in clinical studies and stimulation of VEGF production may play a role [63].

proteins. (For more information on this important group of hepatocyte signaling molecules, please see Chapter 46.)

Hedgehog (Hh) signaling Emerging literature has demonstrated multiple roles of the Hh signaling pathway in liver pathobiology. In relation to LR, all known hepatic cell types produce Hh and respond to Hh signaling. Inhibition of Hh signaling by cyclopamine delayed liver regeneration by 48 hours. There was suppression of hepatocyte and cholangiocyte proliferation and decreased activation of stellate cells [64]. Glypican 3 (GPC3), a GPI‐linked protein on hepatocyte plasma membrane binding to tetraspanin CD81, binds and retains Hh proteins in normal liver. Following PHx, GPC3 ceases binding to CD81 and releases Hh proteins, while Gli1, a transcription factor controlled by Hh, appears prominently in hepatocyte nuclei at day 2 after PHx [65]. Hh signaling regulates expression of Yap in stellate cells, and Yap itself is involved in stellate cell activation and production of ECM [66]. Elimination of Smoothened, the cytoplasmic signaling mediator of Hh, in stellate cells also dramatically decreased expression of Yap in hepatocytes after PHx [67]. These results demonstrate that much is yet to be discovered on the regulation of LR by this very complex system of important signaling regulators.

Insulin Insulin can best be described as an enabler of all hepatocyte functions, including proliferation and differentiation. Produced by the beta cells of pancreatic islets, it is supplied directly to the liver via portal vein, before it goes to any other organ in the body. In the absence of insulin, mitogenic effects of HGF and EGF in hepatocyte cultures are abolished and the viability of hepatocytes is severely curtailed [6]. Diversion of the portal circulation directly to the vena cava (portacaval shunt) causes liver atrophy. Administration of insulin in animals with portacaval shunt restores liver weight, and this process is mediated by proliferation of hepatocytes [68]. Insulin, however, is not by itself mitogenic in hepatocyte cultures [53] and does not cause hepatocyte proliferation when administered to normal animals.

Growth hormone (GH) There are isolated studies on effects of GH in liver regeneration but there is no clear evidence that GH is a major regulator. It controls production of “somatomedins” insulin‐like growth factor 1 (IGF1) and IGF2 by hepatocytes, which play important roles in liver development. GH has been linked to proper timing of hepatocyte DNA synthesis and to activation of EGFR [69] and enhances liver regeneration in aged animals [70].

Wnt/beta‐catenin

Purinergic signals and NK cells

The family of Wnt proteins has emerged as a critical signaling pathway for tissue repair and regeneration in most tissues. They operate through the Frizzled family of receptors, with LRP5/6 acting as coreceptors. There are multiple members to the Wnt family and they are expressed in different proportions from all hepatic cell types. Beta‐catenin is the signaling vehicle for Wnt

There is evidence that ATP is rapidly released immediately after PHx due to increase in portal vein pressure, derived from hepatocyte and macrophage lysosomes. The phenomenon is of short duration. The released ATP is rapidly hydrolyzed by ecto‐ATPases present on plasma membrane of most cells. Adenosine, the final derivative of ATP hydrolysis, can interact with purinergic



45:  Liver Regeneration

receptors. Blockade of P2Y2 purinergic receptor by a specific inhibitor delayed hepatocyte proliferation [71]. Other studies showed that the stimulation of purinergic receptors primarily affects hepatic natural killer (NK) cells and interference with NK cell function also delays liver regeneration [72]. Hepatocytes express two main types and multiple subtypes of purinergic receptors, and their stimulation regulates multiple signaling pathways associated with hepatocyte proliferation, including phosphorylation of ERK1 [71].

Immediate biochemical or physical signals associated with initiation of liver regeneration There are two immediate consequences of PHx: 1. Immediate increase of portal flow to the remaining liver lobes (approximately one‐third of the original liver mass). 2. Acute changes in peripheral blood and tissue constituents that inform on the required, “hepatostat” driven, liver to body weight ratio. The identity of the ultimate early signals that trigger LR must be sought in these two fundamental and rapid changes. The early observation by Moolten and Bucher that PHx in one member of a pair of parabiotic rats stimulated liver regeneration in the unoperated member of the pair [73], gave rise to all subsequent studies that resulted in isolation of HGF and documentation of PHx‐related changes in plasma concentrations of IL6, TNF, NE, bile acids, and so on. In addition to the effects of specific molecular/biochemical signals, the very physical forces generated by the increased portal flow through the remnant liver may also play a role. Preventing rise in portal vein pressure after PHx results in reduced early events of LR and deficient activation of HGF [74]. Fluid shear stress due to acute increase in flow over endothelial monolayers is associated with production of specific signals. Stability of urokinase plasminogen activator (uPA) mRNA and increase in the amount of uPA protein occurs by fluid shear stress in endothelial monolayers [75]. uPA activity increases as the first, so far, detected signal, within 1 minute after PHx [10]. uPA, however, is produced by both endothelial cells and hepatocytes and both cells may be the source of very early rise of uPA [76]. Wnt proteins are also released by endothelial cells, and this may relate to the appearance of beta‐catenin in hepatocyte nuclei within 5 minutes after PHx [77]. There is no study as yet, however, to assess the immediate impact of PHx on Wnt release by sinusoidal endothelial cells.

Immediate and early LR‐related signals following PHx (Table 45.2) Urokinase plasminogen activator (uPA) and ECM remodeling uPA activity in liver increases within 1 minute after PHx [10]. In addition to its recognized roles in initiating ECM remodeling and activation of plasminogen to plasmin (leading to a cascade of activation of metalloproteinases), uPA also activates HGF by converting the ECM‐embedded, single peptide HGF to its heterodimeric form. There are no other agents that can be invoked for

571

Table 45.2  Activation of signaling pathways and changes occurring early after partial hepatectomy • Increase in urokinase activity (first 5 minutes) • Translocation of N(otch)ICD to the nucleus (15 minutes) • Translocation of beta‐catenin to the nucleus (5–10 minutes to 6 hours) • Decrease in HGF biomatrix stores (30 minutes to 3 hours) • Activation of the HGF receptor (within 30–60 minutes) • Activation of the EGF receptor (within 30–60 minutes) • Increase of HGF, norepinephrine, IL6, TNF, TGFb1, serotonin, and hyaluronic acid in the plasma within 1–2 hours. • Activation of AP1, NFkB, and STAT3 within 2–3 hours. • Extensive gene expression reprogramming of hepatocytes within 30 minutes after PHx • Remodeling of extracellular matrix within the first 2–3 hours

this event, since anti‐uPA antibodies inhibit activation of the single peptide, pro‐ HGF in whole liver homogenates taken within 5 minutes after PHx [78]. uPA is also involved with the remodeling of ECM observed within the first 2 hours after PHx. The sequential steps (conversion of plasminogen to plasmin, subsequent degradation of fibrinogen, and activation of metalloproteinases MMP2 and MMP9 by plasmin) become noted shortly after PHx, and are eventually downregulated by a subsequent elevation of TIMP1 by 6–18 hours [5, 35, 79]. There are changes in multiple proteins of ECM (fibronectin prominently decreasing in periportal area within 5 minutes after PHx) followed by an increase in mRNA of several of these proteins, probably aiming to resynthesize them as a compensatory mechanism [80]. Hyaluronic acid, also an ECM component, is rapidly released in the peripheral blood and TGFβ1, present in the extracellular space bound to hepatocyte decorin, is also rapidly released with the same kinetics [5, 35]. HGF, embedded in the extracellular matrix as pro‐HGF form, is released to the plasma as active HGF and rises very rapidly to more than tenfold increase within 1 hour after PHx [81]. New HGF synthesis and the appearance of new HGF mRNA starts at 3 hours after PHx and reaches plateau at 24 hours. Increased expression of HGF mRNA is also seen in lung and kidneys after PHx. The rise of HGF in the plasma occurs 3 hours before any rise in HGF mRNA and is a result of release of HGF stored in hepatic ECM [4, 5, 35, 82, 83]. It should be noted that HGF is heavily embedded in hepatic ECM; following whole‐body elimination of the HGF gene, the hepatic ECM concentration of HGF did not decrease until after two sequential episodes of LR triggered by CCl4 [84]. The overall changes of ECM remodeling in LR are quite complex with multiple factors involved, including plasminogen activator inhibitors (PAI), tissue inhibitors of metalloproteinases (TIMP), regulation of HGF activation, and so on. [85]. ECM proteins and associated glycosaminoglycans act as co‐ receptors for many LR associated ligands, and also directly transmit not fully understood altered signals through integrins [86–88]. This will be discussed in relation to signals controlling termination of liver regeneration (see below, Termination of liver regeneration).

Mobilization of Wnt/beta‐catenin and Notch signaling There is increase in beta‐catenin in hepatocyte nuclei within 5 minutes after PHx and lasting beyond 6 hours [89]. This is presumably driven by Wnt derived from sinusoidal endothelial cells, although that linkage in terms of chronology is not directly

572

THE LIVER:  SIGNALS REGULATING LR

established. Notch is expressed in hepatocytes, cholangiocytes, and endothelial cells. Its ligand, Jagged‐1, is expressed in hepatocytes and cholangiocytes. The active proteolytic fragment of Notch, NICD, appears in hepatocyte nuclei within 15 minutes after PHx, and is associated with subsequent expression of Notch gene targets [90]. This event may be connected to the immediate early production by stellate cells of high levels of the non‐canonical Notch ligand DLK1 within minutes after PHx [91]. Both Notch and its ligand Jagged remain upregulated until at least day 4 after PHx. The Notch changes are clearly associated with hepatocyte proliferation; addition of Jagged‐1 protein in hepatocyte cultures induced DNA synthesis and silencing RNA against Notch or Jagged decreased hepatocyte proliferation until day 4 after PHx [90].

Changes of LR‐related signals in the peripheral blood Earlier studies by Moolten and Bucher [73] employed parabiotic rats and demonstrated that PHx in one of the members of the parabiotic pair led to hepatocyte DNA synthesis not only in the operated rat but also in the liver of the non‐operated one. Similar findings were also noted in hepatocytes transplanted to adipose tissue following PHx of the in situ normal liver [92]. These findings clearly demonstrated the rise of regenerative signals in the peripheral blood following PHx. Following several decades of studies, we now know that many signals associated with LR rise in the peripheral blood after PHx. HGF and NE rise within less than an hour, with overlapping but subsequent rise of IL6, TNF, and TGFβ, serotonin, and so on. (see above sections describing changes related to the specific signaling molecules) [4, 5, 35, 82, 83]. An often‐overlooked consequence of PHx is the acute drop of glucose concentration in the peripheral blood. The importance of this for stimulation of LR was clearly demonstrated by administration of dextrose following PHx; this resulted in temporary suppression of DNA synthesis at day 2 in mice. Surprisingly, most of the changes in hepatocyte transcription factors, activation of HGF, and so on, were not affected; there was, however, a rise in the mitoinhibitory signaling proteins p21 and p27 and suppression of FoxM1, a major regulator of cyclin D1 and many other proteins associated with hepatocyte DNA synthesis [93]. In addition to specific blood signals, platelets attach to the sinusoidal endothelial cells and become activated at the early stages of LR [94]. Platelets contain a mixture of stimulatory (HGF, serotonin) and mitoinhibitory (TGFβ) signals.

Cell proliferation kinetics and intercellular signaling Hepatocytes LR is not dependent on stem cells. The vast majority of existing hepatocytes participate in LR and undergo DNA synthesis. The percent of hepatocyte participation in DNA synthesis and cell proliferation varies with age. LR, however, is by no means impaired in older mice. More than 95% of hepatocytes participate in LR in rats younger than twenty months of age and the percent decreases only to 75% for older animals [95]. LR is also remarkable in that it can be repeated multiple times (up to 12 times reported) in the same animal and the performance of regeneration is not affected [96]. While there are profound

changes in gene expression of hepatocytes during LR, the gene expression patterns associated with LR are different than those operating during liver embryogenesis [97]. Following PHx, hepatocytes are the first cells to exhibit rapid changes and responses to mitogenic signals. As mentioned above, there is rapid migration of beta‐catenin and Notch (NICD) to hepatocyte nuclei within minutes after PHx [89, 90]. Taub et  al. have reported rapid changes in hepatocyte gene expression, within 30 minutes [98]. In preparation for mitosis, the structure of canaliculi becomes temporarily simplified. There are changes in gap and tight junctions from 24–72 hours, regulated by p38‐MAP kinase [99]. MET and EGFR are activated within 30 minutes [100]. STAT3, regulated by IL6 and JAK kinases [101] and NFκB, regulated by TNF [102] become activated in the first 3–5 hours. IL6 also regulates enhanced expression and action of C/EBPβ and has reverse effects on C/EBPα [103]. N‐terminal truncated isoforms of these factors (LIP and LAP, each other’s antagonists) are generated during LR and regulate many of the process of the mature C/EBP isoforms [104]. Hepatocyte FoxM1b plays a key role for all events associated with progression through the S‐phase and mitosis, and its expression increases at 24 hours, prior to the G1/S interphase. In mice deficient in FoxM1b, there is delay of DNA synthesis associated with increase in p21 [105, 106]. Genes associated with generation of IPS cells, including Oct3/4, Nanog, and Myc, increase very significantly after PHx [107]. Critical signals affecting hepatocyte entry into S‐phase are regulated by cyclins D1 and D2. They both turn on expression of multiple genes associated with cell proliferation. In addition, cyclin D1 regulates transcription of enzymes in pathways of metabolism of carbohydrates, lipids, and amino acids [108]. Cyclin D1 expression is regulated at the translational level by mir‐21 [109]. Cyclins D1 and D2 regulate activity of cyclin dependent kinases (CDK), which mediate many of the cyclin D1/2 events. CDK activities are inhibited by the p53‐regulated protein p21 [110]. The expression of p21, however, rapidly increases within few hours after PHx [111], suggesting a complex time‐dependent interplay of regulatory events, guaranteeing precision of chronology in the progression of the various steps associated with hepatocytes in G1. The chronology of these events varies between species. Peak of hepatocyte DNA synthesis is seen at 24 hours in the rat but at 36–48 hours in the mouse. However, not all hepatocytes enter into DNA synthesis at the same time. LR and DNA synthesis proceed from the periportal to the pericentral areas of the lobules, with time kinetics also varying between species [112]. Klochendler et al. were able to isolate cells from different stages of the cell cycle and documented decreased expression of genes characteristic of mature hepatocytes during DNA synthesis [113]. This suppression is in part explained by the decreased expression and suppression of action of HNF4α very early in LR [114]. The waveform transition of hepatocyte replication is thus guaranteeing that not all hepatocytes in the liver will undergo decrease in essential gene expression patterns and preserves sufficient homeostatic liver function during LR. A marked evidence of altered gene expression patterns in hepatocytes is the accumulation of triglyceride droplets during days 2 and 3 of LR. This is dependent on signaling changes associated with blood‐borne signals, as it can be induced in hepatocyte cultures exposed to serum of partially hepatectomized rats [115].



45:  Liver Regeneration

Lipid accumulation in hepatocytes in G1 is associated with induction of lipogenic enzymes and regulated by leptin [116] and by EGFR but not by MET [19]. Caveolin is present at the interface between lipid droplets and cytoplasm and regulates lipid processing and metabolomics of regeneration [117]. Altered gene expression patterns are also affected by expression of multiple miRNA at different phases of the cycle, operating at the translational level of gene expression [118]. Changes in Hippo pathway kinases and Yap also occur during LR. The Hippo pathway kinases MST1/2 and LATS1/2 phosphorylate and deactivate nuclear Yap, whereas nuclear levels of Yap are associated with regulation of liver size [119]. Yap controls expression of amphiregulin, an EGFR ligand involved in LR regulation (see above section on Epidermal growth factor receptor). In normal liver, Yap is expressed in cholangiocytes, with little or no detectable expression in hepatocytes [120]. Yap increases in hepatocyte nuclei within 24 hours, associated with decreased activity of the Hippo kinases. Levels of Yap and enhanced activity of Hippo pathway return to the lower normal by day 7 [119]. Through currently unexplained pathways, nuclear Yap in hepatocytes during LR is regulated by Hedgehog‐ mediated actions on stellate cells [67]. Autophagy regulation is also an important part of LR. Elimination of Atg5, an important component of the autophagic pathway, has severe suppressive effects on DNA synthesis and liver weight is restored mostly by hepatocyte hypertrophy in ATG5 KO mice [121]. Hepatocytes in the cell cycle produce and receive mitogenic signals from the other hepatic cells types. TGFα, FGF1/2, angiopoietins 1 and 2, VEGF, and GM‐CSF expressed and secreted by hepatocytes are known to be mitogenic and probably have paracrine effects on stellate cells, endothelial cells, and Kupffer macrophages, contributing to formation of proper lobular histologic microarchitecture [4, 5, 35, 82, 83]. Hepatocytes also receive newly synthesized HGF from stellate and endothelial cells throughout LR [13]. The net result is precise alignment of newly synthesized hepatocytes along the orientation of the closest sinusoid in a very precise manner [122]. Though the above events related to hepatocyte proliferation during LR, several studies suggest that even after 2/3 PHx, and more definitively after hepatectomies less than 2/3, the pathways involving restoration of liver weight involve not only hepatocyte proliferation but also hepatocyte hypertrophy, with hepatocyte enlargement being a solid component of the overall LR strategy [123, 124]. The signaling pathways controlling regulation of the precise percent of involvement of hepatocyte proliferation versus mere hypertrophy are not understood at this point.

573

PHx in mice with germline deletion of TGR5 is associated with severe jaundice, increased mortality, and hepatocyte necrosis. Melatonin has overall inhibitory effects on cholangiocyte proliferation [126] whereas histamine released by mast cells has stimulatory effects [127]. While these effects are well documented, there is no detailed analysis of their impact on cholangiocyte proliferation during LR. Cholangiocytes also produce platelet derived growth factor (PDGF), which is mitogenic for stellate cells [128]. It should be noted that even though then term “cholangiocytes” is used to characterize the entire biliary epithelial compartment, there is evidence from multiple sources that there are many subtypes, characterized by size and/or expression of specific receptors and the overall regenerative biology of these subtypes may differ [129].

Endothelial cells Sinusoidal endothelial cells enter into proliferation later than hepatocytes and cholangiocytes, with active DNA synthesis peaking at days 4–7 after PHx. Their proliferation is associated with activation of VEGF receptors 1, 2, angiopoietin receptors Tie 1, 2, PDGFRβ, EGFR, and MET [130]. Since hepatocytes are the first to enter into proliferation, they form avascular small clusters and start synthesizing VEGF, which attracts and stimulates proliferation of endothelial cells penetrating the clusters and establishing formation of sinusoids [131]. This activity is associated with stimulation of VEGFR2. Simultaneous stimulation of endothelial VEGF receptor 1, however, induces production of HGF by the endothelial cells, which contributes toward stimulation of hepatocyte proliferation [13]. The endothelial cells in the newly formed capillaries gradually transform into fenestrated endothelial cells through a complex process [132]. Recent studies have shown that LR is also associated with migration of endothelial precursors from bone marrow to the liver, which convert from typical endothelial cells into fenestrated endothelial cells and participate in the restructuring of the hepatic sinusoids in the enlarged lobules that result from LR [14]. The migrating sinusoidal progenitor cells (“sprocs”) actively produce HGF. Their migration is controlled by hepatic production of circulating VEGF and mediated via stromal cell derived factor 1 (SDF1) [133]. HGF and TGFβ1 are known to stimulate tubule formation in endothelial cells and are very likely play a role in the restructuring of the sinusoidal network [4, 5, 35]. Despite complex pathways of cell migration, endothelial cells regulate LR in many ways by producing cytokines favoring regenerative response. These “angiocrine” effects were described recently by Ding et al. [134].

Cholangiocytes

Stellate cells

Proliferation of cholangiocytes in portal ductules follows the same time frame as hepatocytes. Cells of the biliary system respond to the same tyrosine receptor kinases (MET and EGFR) as hepatocytes. LR proceeds by enlargement of existing lobules; there is no formation of new portal triads following PHx. The receptor TGR5, a G‐protein couple receptor, also regulates proliferation of cholangiocytes in culture in response to bile acids, with some of the bile acids having stimulatory or inhibitory effects [60]. TGR5 has complex effects during LR, protecting hepatocytes from increase in biliary flux from the intestine [125].

These cells exist in multiple organs of endodermal origin (lungs, pancreas, etc.). Hepatic stellate cells store vitamin A in lipid droplets, and they express many genes in common with glial cells of the brain, even though stellate cells are derived from nestin‐positive cells of the cardiac mesenchyme and not the neural crest. There is strong evidence that stellate cells function as part of a neuroendocrine control of normal liver and respond to parasympathetic and sympathetic innervation [135]. They also function as antigen presenting cells for NK cells [136]. Most of the studies related to stellate cells focus on pathways resulting

574

THE LIVER:  SIGNALS REGULATING LR

in stellate cell “activation” during chronic liver injury, and their contribution to liver fibrosis and cirrhosis [137]. Their participation in LR after PHx has also focused mostly on their contribution of signaling molecules (including HGF, TGFβ, epimorphin, pleiotropin, etc.) and the production of most of ECM proteins in different stages of LR. Despite their vital role in normal and regenerating liver, there is no detailed study of their proliferation rates during LR. In normal liver, stellate cells make contact with both hepatocytes and sinusoidal endothelial cells by very long processes (Figure  45.1). This relationship suggests that there are regulatory pathways operating by direct communication between stellate cells and the cell they contact; these interactions need to be explored. Histologic observations demonstrate that stellate cell numbers increase in parallel with the restoration of the sinusoidal network. Stellate cell numbers peak in

regenerated lobules at day 7 of LR, in parallel with the sinusoidal endothelial cells. Increase in desmin‐negative cells resembling immature stellate cells is seen at day 1 of LR [138].

Kupffer cells (hepatic macrophages) Several studies have documented that the Kupffer cell macrophages lining the hepatic sinusoids, in addition to standard macrophage functions (phagocytosis of particles, participation in eliminating tissue debris after local damage, etc.) also have some unique properties, related to aspects of immunity in which liver is an active participator. During LR, there is evidence that many Kupffer cells proliferate locally [139]. There is evidence, however, that mononuclear cells deriving from bone marrow also enter the liver and become typical Kupffer cells during LR [140]. Kupffer cells are recognized as F4/80+ and there distinguished into two categories. The CD11b+ Kupffer cells appear to be derived from bone marrow and their proportion increases at day 3, after the peak of hepatocyte proliferation. The CD68+ Kupffer cells do not seem to change in numbers during LR [141]. Proliferating hepatocytes produce GM‐CSF, a mitogenic signal for Kupffer cells; stellate cells produce M‐CSF, and Kupffer cells produce IL6, TNF, TGFβ, and TGFα during LR [4, 5, 35]. Overall, there are no precise measurements of population changes of Kupffer cells during LR but their numbers vary along with their participation in restoration of sinusoidal vascular network.

Termination of liver regeneration

Figure 45.1  Intricate connections between stellate cells (SC), sinusoidal endothelial cells (SEC) and hepatocytes (Hep). A single stellate cell (green) is making contacts with multiple other cells. The nature and purpose of these contacts is not fully understood. They are, however, likely to be involved in bilateral communications regulating extracellular matrix deposition and exchange of growth‐regulatory signaling molecules. Image courtesy of Dr. Donna Stolz, University of Pittsburgh.

Initiation of LR can be precisely timed by the performance of PHx. Termination of LR, however, is more difficult to define. If the liver weight to body ratio is to be used as a criterion, normal ratios (depending on species) are approached by end of the second week. In terms of hepatocyte proliferation, a return of nuclear DNA synthesis to normal levels is approached by days 5–6 [4, 5, 35]. Analysis of liver transcriptomics, however shows a slower return to normal gene expression values by some time beyond day 14 after PHx [88] (Figure 45.2). Several signaling molecules have been considered as potentially associated with “termination” of LR. Hepatocyte‐specific TGFβ1 transgenic mice have normal liver regeneration, even though TGFβ is mito‐inhibitory for hepatocytes [142]. Hepatocyte targeted elimination of TGFβ receptors or activin receptors does not prolong liver regeneration; extended liver regeneration is seen only when both receptors are eliminated [143]. An important regulator associated with termination of liver regeneration and restoration of normal liver function is likely to be ECM. Addition of ECM preparations (collagen gels, Matrigel) stabilizes differentiation but inhibits proliferation of hepatocytes in culture [6]. ECM is partially degraded and remodeled at the beginning of LR, but it is restored toward the end of LR [40, 144, 145]. Signaling between ECM and hepatocytes is mediated by integrin α3β1; the β1 intracellular domain makes contact with integrin linked kinase (ILK). The latter transmits growth suppressor and differentiation enhancement signals through multiple pathways [87]. Hepatocyte‐specific elimination of ILK prolongs liver regeneration and results in livers of size 158% of the ­original. All protein partners associated with ILK increase during LR until the end of hepatocyte proliferation, by



45:  Liver Regeneration

575

Figure 45.2  Expression of the top 250 genes expressed in normal mouse liver at different times after 2/3 partial hepatectomy (PHx). There is rapid up and down change in expression of most of the 250 genes within 1 day after PHx, with changes in both direction continuing until beyond day 14 of regeneration. This makes it difficult to determine a precise point at which the processes of LR should be considered as “terminated”. A few specific genes are massively overexpressed during LR, exceeding expression values of any gene seen in normal resting liver. The reason for the selective massive overexpression of a few genes is not clear. Mup1 is not present in humans. Selenoprotein P1 is the major vehicle by which liver sends selenium to peripheral tissues, essential for regulation of redox functions in most cells of body. Transthyretin is the major transport vehicle in plasma for thyroxin (T4) and retinol bound to the retinol binding protein. Cyp2c9 protein is approximately 18% of all proteins of members of the liver Cyp family and is associated with metabolism of most xenobiotics and many endogenous metabolic compounds including arachidonic acid. Jak3 gene encodes JAK3 kinase protein, associated with gp130 receptor of IL6 and activating STAT3 transcription factor. Psg28 is a glycoprotein produced in liver during pregnancy. There is no known function for this protein in liver regeneration.

day 6 [88]. These results suggest that ECM signaling is important for regulation of termination of liver regeneration, but the pathways remain to be further understood. Glypican 3 (GPC3), a GPI‐ linked plasma membrane protein over‐expressed in liver cancer [146], is nonetheless a growth suppressor for most organs, as demonstrated by the Simpson–Golabi–Bechmel syndrome in humans with loss of function of GPC3 [147]. Hepatocyte‐specific transgenic expression of GPC3 results in substantially suppressed LR [148]. GPC3, normally interacting with tetraspanin CD81, is involved during LR with many complex interactions with different pathways, including Hedgehog, Wnt, and Hippo signaling. GPC3 dissociates from CD81 during LR but associates with it again at days 6–7 [65]. Recent evidence has shown that GPC3 bound to CD81 enhances activity of the Hippo pathway and results in decreased levels of nuclear Yap. This may be a very important regulating signal by which GPC3, via Yap, contributes to the termination of LR [149]. LSP‐1 is another protein, deleted in about 50% of HCC, and regulating the RAF‐MEK‐ERK pathway. Deletion of hepatocyte LSP‐1 prolongs LR and transgenic expression suppresses LR [150]. Beta‐catenin is an important mitogenic signal contributing to LR; beta‐catenin signaling is downregulated by Wnt5a whose expression also increases toward the end of LR and restricts further beta‐catenin signaling [151].

Overcoming deficiencies of extracellular signals and the critical role of receptor tyrosine kinases Elimination of single extracellular signals (IL6, TNF, Wnt/beta‐ catenin, Hedgehog, bile acids, etc.) involved in regulation of LR, delays but does not abolish LR. This also applies to

elimination of single receptor tyrosine kinases (MET and EGFR). The flexibility of cybernetics of hepatocyte transcriptomics, which allows overcoming such obstacles, is remarkable. Absence of many of the extracellular signals normally involved to initiate hepatocyte DNA synthesis (e.g. IL6 and STAT3; TNF and NFκB) can be corrected by MET and EGFR [45]. Systemic, whole‐body, elimination of the HGF gene does not affect LR, because the heavily embedded HGF in ECM is sufficient in quantity to support several LR episodes prior to be depleted [84]. The only extracellular signaling intervention that completely abolishes LR is the combined elimination of signaling of both MET and EGFR [19]. When signaling of both of these receptors is eliminated 5 days prior to PHx, there is a small amount of hepatocyte proliferation at post‐PHx day 2; proliferation, however, stops entirely after that. This is followed by decrease in hepatocyte size to 35% of normal and decrease in the size of hepatic lobules. There is no net gain in liver size by the end of day 14. There is significant decrease in hepatocyte‐ specific gene expression, especially for genes involved in lipid and sterol biosynthesis, glycogen synthesis, and urea cycle. There is no activation of mTOR and AKT, enhanced activation of PTEN and AMPKα, and inactivation of enzymes related to fatty acid synthesis and redox regulation. Despite these profound changes, there is dramatic upregulation of expression of genes involved in plasma protein synthesis and all members of CYP family genes (with the exception of Cyp7a1, the first enzyme involved in bile acid synthesis). Mice die by day 14 after PHx, with low glucose, marked ascites, and high ammonia levels in peripheral blood. Remarkably, there is no increase in  hepatocyte death, but there is increased apoptosis of non‐ parenchymal cells, especially stellate cells. Mice die not because

576

THE LIVER:  SIGNALS REGULATING LR

liver dies, but because liver ceases to function in the absence of activation of EGFR and MET. The same phenomena, with some variations in activation pathways, were observed and resulted in death of normal, non‐hepatectomized mice [152]. In both of those studies, MET and EGFR signaling was eliminated throughout the body. Yet no abnormalities were detected in other tissues. Proliferation of stem cells in intestinal crypts was unaffected. Why is liver so uniquely dependent on activation of EGFR and MET? (Figure 45.3). Both receptors are activated by Tyr phosphorylation in normal resting liver [19]. Activation of MET can be explained by the heavy presence of HGF in hepatic ECM [84]. The story for EGFR is more complex. As mentioned above, EGF is synthesized by the Brunner’s glands residing under the duodenal mucosa [21]. EGF protein is directly secreted in the intestinal lumen, and a portion is absorbed intact and provided to hepatocytes continually through the portal circulation [21]. The observations for constant supply of HGF and EGF to hepatocytes suggest that the combined signaling of HGF and EGF generates a fundamental cybernetic platform required for the proper coordination of all other signaling pathways characteristic of normal hepatocytes. Collapse of coordination of the

other signaling pathways occurs when these two receptor tyrosine kinases (RTK) are simultaneously abolished. The findings may have implications for pathogenesis in human liver failure. Plasma HGF in fulminant hepatitis rises to high levels, but is not activated [153], in contrast to the active HGF released in plasma after PHx [154]. There is also closure of the fenestrae of the sinusoidal endothelial cells in fulminant hepatic failure, potentially interfering with EGF availability [155].

Facultative stem cell relations between hepatocytes and cholangiocytes Hepatocytes and cholangiocytes are the two epithelial cell types in the liver. They typically enter into proliferation to restore the proper cell numbers in their own compartment, as required for normal liver function. This is done without participation of stem cells. There are situations, however, in which either hepatocytes or cholangiocytes cannot enter into proliferation to repair deficiencies in their own compartment. In such situations, hepatocytes and cholangiocytes follow steps of transdifferentiation to transform themselves into each other [156]. This plasticity of

Figure 45.3  Dependence of hepatocytes on HGF/MET and EGFR signaling. Hepatocytes are continually exposed to HGF and EGF. HGF is produced by stellate cells and deposited as inactive single peptide form in the extracellular matrix. EGF is produced by Brunner’s glands of the duodenum, shown by black arrows in the histologic picture (hematoxylin eosin stain; magnification 100×). The histology of Brunner’s glands is similar to salivary glands, which produce and secrete EGF in saliva. EGF is produced and secreted in the intestinal lumen, and then a portion is absorbed intact and enters the liver through the portal circulation. HGF receptor MET and EGFR are activated in normal, resting liver. For details and references, see text.



45:  Liver Regeneration

phenotype confers regenerative advantages to the liver, especially in situations of liver failure. It is much faster to restore a hepatocyte or cholangiocyte compartment by a rapid transdifferentiation of one type of cell to another, than to rely on a slow growth and expansion of a true stem cell compartment. There are now numerous studies that have documented the capacity of either cell type to enter into proliferation, transdifferentiate, and restore the numbers of cells in the other compartment suffering with defective proliferation [156]. This has been shown in several experimental models in rats and mice. Blockade of rat liver regeneration by acetylaminofluorene (AAF), resulting in expression of DNA adducts and expression of p21 in hepatocytes [157], causes the appearance of a transitory population of “oval” cells with morphologic properties intermediate between hepatocytes and cholangiocytes [158]. These cells eventually evolve into mature hepatocytes [159]. The emergence of the oval cells from cholangiocytes in rats is supported by the expression of hepatocyte‐associated transcription factors in cholangiocytes immediately after AAF‐PHx [160] and the lack of emergence of oval cells by pretreatment with the cholangiocyte‐ specific toxin DAPM (4,4’‐diaminodiphenylmethane) prior to AAF‐PHx [161]. Mice cannot activate AAF, and the evidence for cholangiocyte conversion to hepatocytes had to exploit more difficult pathways that cause enhanced expression of p21 in hepatocytes. This was achieved by Forbes et  al. in mice with hepatocyte specific deletion of Mdm2, an E3 ubiquitin‐protein ligase which specifically degrades p53. In the absence of MDM2, there is enhanced expression of p53 which subsequently induces p21. There is increased hepatocyte death and senescence followed by the appearance of “progenitor” cells with properties similar to the oval cells seen in rats. The progenitor cells converted into hepatocytes and restored hepatocyte populations. The biliary origin of these progenitor cells was established by detecting the newly emergent hepatocytes as carrying lineage tagging performed on cholangiocytes [162]. Extended studies have also documented the transformation of hepatocytes to cholangiocytes in settings where cholangiocyte proliferation is expected but cannot occur, as, for example, in rats pretreated with DAPM and subjected to bile duct ligation (BDL). Up to 50% of the newly produced biliary ductules in the DAPM–BDL model carry markers specific for hepatocytes [163]. Studies with hepatic organoids demonstrated that this conversion is dependent on EGFR and MET [164]. Similar results have been shown in mice. Conversion of hepatocytes to cholangiocytes was shown in mice fed diethyldithiocarbamate (DDC) diet [165]. The most recent evidence for hepatocyte transdifferentiation to cholangiocytes was recently presented by Huppert and Willenbring, who demonstrated a complete de novo generation of a biliary system from hepatocytes by pathways controlled by TGFβ [166]. The pathways mediating this interconversion are complex and, in addition to TGFβ, also involve Hippo/Yap [120] and Notch [167]. It is not clear whether all hepatocytes, anywhere in the lobule, can participate in these interconversions. Demetris et  al. demonstrated cells with mixed  expression of hepatocyte and cholangiocyte transcription factors at the terminal end of the canals of Hering, suggesting a population of pre‐existing cells in normal liver with progenitor cell phenotypes [168]. Also, surprisingly, in LR suppressed by AAF, there are more progenitor cells appearing after

577

centrilobular damage induced by CCL4 than after regeneration following PHx. In AAF‐suppressed centrilobular damage, the progenitor cells are likely to appear from the cholangioles that penetrate deep into the hepatic lobule (canals of Hering) [169]. Lemaigre has traced the histology of formation of the ductal plate during embryogenesis and demonstrated that most newly formed ductal cells migrate into the space next the portal and arterial branches to form the ductules of the portal triads. Cells that fail to migrate, revert to hepatocytes and remain immediately proximal to the portal triad [170]. Studies in the rat demonstrated that these hepatocytes preferentially contribute to formation of biliary ductules [163]. In hepatic organoids in roller bottle cultures maintained in the presence of insulin, HGF, and EGF, epithelial cells retain a hepatobiliary phenotype with mixed gene expression patterns. Addition of corticosteroids induces the separation of the mixed lineage and the emergence of mature biliary cells immediately under the culture medium and mature hepatocytes in the tissue underlying the biliary cells. Conversion of hepatocytes to biliary cells in that system cannot occur in the absence of HGF and EGF (either one alone can sustain the conversion) [39, 171]. In the same system, either HGF or EGF induce TGFβ expression. Given the above recent findings by Schaub et al. [166] on the role of TGFβ involved in this transdifferentiation, it is likely that not only HGF and EGF, but also TGFβ, are involved in the conversion of hepatocytes to cholangiocytes in the organoid cultures. It should be noted that whereas cells of mixed hepatobiliary phenotype are rarely seen in normal human liver, they become the predominant type in acute fulminant hepatitis, regardless of etiology [172]. It is not clear whether this is a pathway to healing of fulminant hepatitis and restoration of normal structure, because the evolution of the hepatobiliary cells has not been traced in the few cases of fulminant hepatitis that heal spontaneously.

Regeneration following xenobiotic‐induced centrilobular liver injury Hepatocytes in the central portion of hepatic lobules express the (phase I) CYP family and (phase II) other enzymes involved with metabolism and processing of xenobiotics. In most instances, xenobiotics are appropriately processed for elimination through blood (kidneys) or bile (feces). In rare instances, however, the processing of some xenobiotics results in generation of reactive electrophiles, capable of reacting with the nucleophile residues of proteins and nucleic acids, resulting in death of centrilobular hepatocytes. Experimental models to study this injury and its repair pathways have most often utilized carbon tetrachloride (CCl4) or acetaminophen as the damaging agents, in rats and mice [173, 174]. The extent of centrilobular necrosis and the lethality depend on the dose of the xenobiotic, and the lethality of doses vary with the strain of the animals, other dietetic components, and so on, as these parameters affect gene expression patterns in the centrilobular regions [175]. Overall, this type of injury is more representative of the type of damage seen in human livers, especially given the fact that inappropriately high doses of acetaminophen are often used for suicide, resulting in massive liver necrosis, salvaged, whenever possible, with liver transplantation. The first evidence of this type of injury is a hepatocyte necrosis in the centrilobular areas. This is

578

THE LIVER:  SIGNALS REGULATING LR

followed by invasion of monocytes, which proliferate very actively in the site of damage and produce active macrophages. The latter proceed to remove the area of necrosis. There is rapid activation of EGFR within minutes after acetaminophen dose [176]. Hepatocytes enter into the cell cycle (proliferating cell nuclear antigen [PCNA]‐positive) by the end of day 1, primarily around the area of necrosis. Most of the cells of the rest of the lobule, however, become PCNA‐positive by the end of day 2. Hepatocytes positive for Ki‐67, a biomarker for active DNA synthesis, appear at day 2 and increase to cover the entire lobule. HGF levels increase within the liver and in the peripheral blood [81]. mRNA for TGFα and HGF increased in two peaks, at 12 and 48 hours [177]. In addition to contributions of growth factors and stimuli seen in LR, macrophages may also contribute mitogenic stimuli, as they are known to produce HGF, TGFα, TGFβ, PDGF, IL6, and TNF. The chronology of expression of these signals from macrophages in the context of repair of this type of injury has not been assessed. The fact that the first PCNA‐positive hepatocytes at day 1 first emerge around the area of necrosis suggests that mitogenic stimuli produced by macrophages, actively infiltrating and proliferating in that area, must play a very early, perhaps the first, role. The rate of removal of the necrotic hepatocytes and the restoration of normal ­histology depends on the dose of the offending agent. Normal histology, with moderate experimental doses, is usually seen within one week. The mechanistic details of repair of this type of injury, though clinically relevant, are less defined compared to LR. This is due to the comingling of inflammatory events at the earliest stages, making it difficult to distinguish signals strictly related to cell proliferation versus inflammation. An additional issue is the variation of the response based on the dose of the offending agent [175]. It should be stated, however, that all mitogenic signals identified in LR have also been identified in this type of injury, though their sources and time‐ kinetics are not be identical.

Maintenance of steady mass in normal resting livers In all previous sections of this chapter we discussed the mitogenic signals and complex responses triggered by massive loss of liver parenchyma. The elicited hepatocyte proliferation is a compensatory response, aimed to restore the full functional capacity of liver. The loss of hepatocytes is the primary reason for the emergence of signaling pathways driving the compensatory response. In a normal resting rodent liver, however, at any given moment, the percent of dead or proliferating hepatocytes typically does not exceed 0.2% of the total number of cells. Even though this would involve very small numbers of cells, there must be pathways that trigger compensatory proliferation at a micro‐scale, so that the total loss over time does not result in decreased liver mass. The mechanisms underlying this “hepatostat” are not well understood. It is reasonable to postulate that the signals regulating this process may be different than the signals elicited by a large hepatocyte loss. As mentioned above, based on constant availability of HGF (from ECM) and EGF (from the Brunner glands of duodenum), MET and EGFR are always activated in normal liver, though we do not know if this applies to all or a fraction of hepatocytes. Early studies using

continuous DNA labeling of hepatocytes in normal liver claimed a slow streaming of proliferating hepatocytes arising in periportal areas and descending to pericentral areas of the lobule [178]. Subsequent studies with pulse labeling of DNA, however, showed that replicating hepatocytes could be found in all areas of the hepatic lobule, not just in periportal, and that “streaming” of hepatocytes was not the mechanism for maintaining liver size [179]. There have been several major recent studies, that have provided new information about the role of different hepatocyte subpopulations in maintenance of the “hepatostat”. Studies by Nusse et  al. documented a unique population of hepatocytes immediately around the central venules. These cells express Axin2, a protein which is part of the beta‐catenin ubiquitination complex. These cells express Tbx3, a biomarker for fetal hepatoblasts, and are diploid. They are also positive for the beta‐ catenin dependent glutamine synthetase. Wang et  al. tagged these cells and documented that over time they generate a lineage of cells which expand around the central area and lead to progeny covering up to 40% of the cells of the lobule [180]. The results were impressive, though there is a concern that beta‐ catenin levels may be elevated in these cells. There is only one of the Axin2 genes active in this model, and the results of potential Axin2 haplo‐insufficiency may cause elevated levels of beta‐catenin, enhancing the proliferative rate of the affected hepatocytes. From the other pole of the lobule, the periportal area, Furuyama et al. provided evidence that Sox9+ lineage precursors in liver and pancreas generate progeny of cells that gradually expand toward the other areas of the lobule, differentiating into Sox− hepatocytes and pancreatic acinar cells [181]. Sox9 is expressed in cholangiocytes, and in that study, it was thought that cholangiocytes continually differentiate into hepatocytes maintaining liver mass homeostasis. Other studies, however, demonstrated that immediate periportal hepatocytes also express low levels of Sox9 and they also proliferate to establish expanded progeny migrating toward the midzonal and centrilobular areas, especially after chronic liver injury [182]. These cells are the same hepatocytes earlier identified as being responsible for generating biliary ductules in inhibited biliary repair [163], previously identified by Lemaigre et al. as remnant cells of the ductal plate which failed to enter into the portal triad and returned to hepatocyte differentiation [170]. Thus, the findings of Furuyama et al. could be explained in a different way; the lineage described by Furuyama did not originate from cholangiocytes, but from Sox9+ immediate periportal hepatocytes. Given, however, the earlier studies with pulse labeling which demonstrated that hepatocytes can be tagged in DNA synthesis in all areas of the lobule, it is logical to consider that both the centrilobular and periportal hepatocyte lineages operate, proceeding in opposite directions, and replenishing hepatocytes. In this scenario, DNA pulse labeling would tend to label hepatocytes in all areas of the lobule. Supporting evidence from a pan‐lobular hepatocyte participation in the process of continual renewal came from another study. Tchorz et al. demonstrated that liver zonation depends on a Wnt/beta‐catenin gradient with intensity increasing from periportal to centrilobular areas, and that the formation of this gradient is dependent on angiocrine signals mediated through R‐spondin (RSPO) ligands and their receptors (Lgr4/5). Lgr4+ hepatocytes had enhanced proliferative capacities during LR and they were located throughout the lobule [183]. Finally, in a



45:  Liver Regeneration

very recent study by Artandi et al., lineage tagging of subset of hepatocytes expressing high levels of telomerase demonstrated that they exist in all zones of the lobule and they give rise to tagged progeny that can over time cover the entire lobule [184]. It is very likely that all these pathways (periportal, pericentral, pan‐lobular) operate at the same time to give progeny that maintains a sufficient population of hepatocytes to guarantee adequate mass for liver function. What is not identified however, is perhaps the most crucial function of the hepatostat. What are the stimuli that regulate all these processes and guarantee that liver size does not exceed the required size?

Chronic regeneration in chronic liver disease Most chronic liver diseases are associated with attrition of hepatocytes, to a variable extent. This occurs regardless of etiology, be that viral, toxic (including alcohol), metabolic, steatohepatitis, and so on. Loss of hepatocytes triggers regenerative responses. These responses can be documented by enhanced immunohistochemistry for biomarkers associated with hepatocyte proliferation (PCNA, Ki67), as well as biomarkers associated with hepatocyte death (TUNEL assays, expression of activated caspases). The precise signaling pathways triggering the regenerative response are likely to be similar with those discovered following acute injury, discussed above. But, there also likely to be peculiarities related to the injurious agent, probably affecting regenerative signaling in different ways. The supreme “prerogative” of loss of hepatocytes is to cause compensatory hepatocyte proliferation, in order to maintain the required liver to body weight ratio, best thought of as a “hepatostat”. In contrast to any other organ in the body, a required and fixed amount of liver tissue is needed for body homeostasis, and on that regulatory principle, loss of liver “mass” is not tolerated. This phenomenon is unique to the liver; it is not the case with other organs (with the exception of endocrine glands under control by pituitary). Loss of half of pancreas, or one of the kidneys, is not associated with regeneration to increase the remnant tissue to the original size of the total pancreas, and the residual kidney becomes slightly larger but not twice the weight [185]. As a consequence of the expectations of the hepatostat, there is enhanced proliferation of the surviving hepatocytes at any time there is continuous hepatocyte loss. This chronic compensatory proliferation of surviving hepatocytes to make up for the loss of hepatocytes induced by the chronic disease has several adverse consequences. Several studies have shown that continuous proliferation in chronic liver disease is associated with decrease in hepatocyte ploidy. Hepatocytes in both in humans and rodents are typically polyploid. The average ploidy of human hepatocytes is 4n binucleate. In mice, ploidy levels reaching 8n or 16n are commonly seen. Experimental studies in rodents have shown that during chronic hepatocyte proliferation, the “ploidy conveyor” reverses course and most polyploid hepatocytes revert to diploid status [186]. Several studies in human liver have documented increased diploidy levels in most chronic liver diseases, including cirrhosis [187]. Polyploidy reversal to diploidy, in and by itself, should not be expected to have negative consequences. Unfortunately, however, a large percentage of human and rodent polyploid hepatocytes are also aneuploid, randomly missing single copies of chromosomes [188]. In a polyploidy status, the

579

functions of such missing alleles are provided by the presence of several copies of the single missing chromosome, present in the polyploid status. As hepatocytes revert to diploidy, however, some of the diploid cells are likely to randomly have only single copies of some chromosomes. For example, an 8n polyploid hepatocyte missing a single copy of chromosome 13 would generate four diploid hepatocytes of which one would only have a single copy of chromosome 13. This has implications for development of liver cancer. In chronic liver disease, hepatocyte proliferation takes place in an inflammatory environment with associated genotoxic products, including reactive electrophiles, lipid peroxides, O2− radicals, and so on. Any inflicted genotoxic damage on the single copy chromosome at the diploid stage would not be balanced by the genes of the missing chromosome. In this setting, otherwise non‐dominant mutations in the existing single chromosome copy may function as dominant (driver) mutations. This enhances the risk of developing neoplasia. In view of this, it should not be a surprise that all types of chronic liver disease are associated with increased risk for development of hepatocellular carcinoma [1]. The risk for developing liver cancer is also enhanced in situations in which proliferation of hepatocytes is overall impaired. The classic experiment by Solt et al. demonstrated that “resistant hepatocytes” against any particular mito‐inhibitory block will rapidly develop into nodules of monoclonal hepatocyte growth when challenged with acute liver injury (partial hepatectomy), and the resultant nodules will allow restoration of the hepatostat (liver to body weight ratio) [189]. These “resistant hepatocytes” are more exposed to the risks of developing aneuploid diploidy due to their enhanced rates of proliferation. The few “resistant cells” are the only ones that can proliferate to meet hepatostat demands for the whole liver. The phenotypes of liver cancer in chronic liver disease due to errors of metabolism associated with chronic inflammation, are a clear example of this situation. Most cancers developing in hemochromatosis do not store excessive iron; cancers emerging in glycogen storage diseases, lipid storage diseases, and so on, in their majority do not store the abnormal metabolic product; most cancers seen in patients with alpha‐1 antitrypsin deficiency do not carry the characteristic PAS‐diastase positive globules characteristic of the misfolded ATZ protein [1]. In all of these situations, hepatocytes that for random reasons do not carry the metabolic abnormality are the “resistant hepatocytes” of Solt et al. In their effort to maintain the hepatostat, their proliferation is faster than the vast majority of hepatocytes affected with the metabolic disease. In a recent study, the same phenomenon appears to apply to hepatocellular carcinomas associated with HCV infection. HCV impairs hepatocyte proliferation by interacting with tetraspanin CD81 and activating the Hippo pathway, thus causing decrease in nuclear Yap [149]. Most hepatocellular carcinomas do not express CD81 in plasma membrane and are thus uninfectable by HCV [190]. This makes them resistant to the mito‐inhibitory effects of Hippo activation and Yap decrease affecting the normal hepatocytes infected by HCV. The “HCV resistant” CD81‐negative hepatocytes would proliferate faster to maintain the hepatostat, and compensate for the loss of hepatocytes caused by HCV infection. In this situation, HCV functions as a “promoter” for enhanced development of hepatocellular carcinomas, with most of the latter being composed of “HCV resistant” neoplastic hepatocytes.

580

THE LIVER:  REFERENCES

CONCLUDING REMARKS Even though LR after partial hepatectomy is not representative of types of liver injury most commonly seen in liver disease, it has been a very useful model for understanding the fundamental signaling mechanisms controlling liver regeneration. It is not complicated by considerations of inflammatory phenomena associated with acute hepatocyte necrosis. Much recent knowledge has surfaced for homeostatic events related to liver mass maintenance under normal conditions. The signaling pathways, however, regulating the hepatostat are not yet well understood [1]. We understand that nuclear Yap plays a role in regulating liver size and that combined signaling of HGF/MET and EGF/ EGFR are essential for maintaining the hepatostat; combined elimination of these RTK signals causes liver decompensation and functional collapse. It should not be forgotten at the end of this chapter, that hepatocytes have very high, almost forever, proliferative capacity in liver recolonization models. In the Fah−/− model, after ten sequential recolonization events in which it was shown that both polyploid and diploid hepatocytes equally participate, calculations demonstrated that 1 hepatocyte could, in mathematic principle, generate 50 mouse livers [191]. In that sense, hepatocytes are unique amongst the epithelial cells in our  body, and much more needs to be learned not only about what turns on their proliferation, but also what harnesses that unlimited proliferative capacity and channels it to meaningful, hepatostat‐defined, boundaries, so the liver continues to ­function normally.

REFERENCES 1. Michalopoulos, G.K. Hepatostat: liver regeneration and normal liver tissue maintenance. Hepatology, 2017;65(4):1384–92. 2. Ware, A.J., Eigenbrodt, E.H., Shorey, J., and Combes, B. Viral hepatitis complicating the Dubin–Johnson syndrome. Gastroenterology, 1972;63(2):331–9. 3. Higgins, G.M. and Anderson, R.M. Experimental pathology of the liver, 1: Restoration of the liver of the white rat following partial surgical removal. Arch Pathol, 1931;12:186–202. 4. Michalopoulos, G.K. and DeFrances, M.C. Liver regeneration. Science, 1997;276(5309):60–6. 5. Michalopoulos, G.K. Liver regeneration. J Cell Physiol, 2007;213(2):286–300. 6. Block, G.D., Locker, J., Bowen, W.C. et  al. Population expansion, clonal growth, and specific differentiation patterns in primary cultures of hepatocytes induced by HGF/SF, EGF and TGF alpha in a chemically defined (HGM) medium. J Cell Biol, 1996;132(6):1133–49. 7. Zarnegar, R. and Michalopoulos, G. Purification and biological characterization of human hepatopoietin A., a polypeptide growth factor for hepatocytes. Cancer Res, 1989;49(12):3314–20. 8. Nakamura, T., Nishizawa, T., Hagiya, M. et  al. Molecular cloning and expression of human hepatocyte growth factor. Nature, 1989;342(6248): 440–3. 9. Liu, Y., Michalopoulos, G.K., and Zarnegar, R. Molecular cloning and characterization of cDNA encoding mouse hepatocyte growth factor. Biochim Biophys Acta, 1993;1216(2):299–303. 10. Mars, W.M., Liu, M.L., Kitson, R.P., Goldfarb, R.H., Gabauer, M.K., and Michalopoulos, G.K. Immediate early detection of urokinase receptor after partial hepatectomy and its implications for initiation of liver regeneration. Hepatology, 1995;21(6):1695–701. 11. Schuppan, D., Schmid, M., Somasundaram, R. et  al. Collagens in the liver  extraular matrix bind hepatocyte growth factor. Gastroenterology,. 1998;114(1):139–52. 12. Passino, M.A., Adams, R.A., Sikorski, S.L., and Akassoglou, K. Regulation of hepatic stellate cell differentiation by the neurotrophin receptor p75NTR. Science, 2007;315(5820):1853–6.

13. LeCouter, J., Moritz, D.R., Li, B. et al. Angiogenesis‐independent endothelial protection of liver: role of VEGFR‐1. Science, 2003;299(5608):890–3. 14. Wang, L., Wang, X., Xie, G., Wang, L., Hill, C.K., and DeLeve, L.D. Liver sinusoidal endothelial cell progenitor cells promote liver regeneration in rats. J Clin Invest, 2012;122(4):1567–73. 15. Achim, C.L., Katyal, S., Wiley, C.A. et al. Expression of HGF and cMet in the developing and adult brain. Brain Res Dev Brain Res 1997;102(2): 299–303. 16. Wolf, H.K., Zarnegar, R., Oliver, L., and Michalopoulos, G.K. Hepatocyte growth factor in human placenta and trophoblastic disease. Am J Pathol, 1991;138(4):1035–43. 17. Trusolino, L., Bertotti, A., and Comoglio, P.M. MET signalling: principles and functions in development, organ regeneration and cancer. Nat Rev Mol Cell Cell Biol, 2010;11(12):834–48. 18. Bard‐Chapeau, E.A., Yuan, J., Droin, N. et al. Concerted functions of Gab1 and Shp2 in liver regeneration and hepatoprotection. Mol Cell Cell Biol, 2006;26(12):4664–74. 19. Paranjpe, S., Bowen, W.C., Mars, W.M. et al. Combined systemic elimination of MET and epidermal growth factor receptor signaling completely abolishes liver regeneration and leads to liver decompensation. Hepatology, 2016;64(5):1711–24. 20. Carver, R.S., Stevenson, M.C., Scheving, L.A., and Russell, W.E. Diverse expression of ErbB receptor proteins during rat liver development and regeneration. Gastroenterology, 2002;123(6):2017–27. 21. Skov Olsen, P., Boesby, S., Kirkegaard, P. et  al. Influence of epidermal growth factor on liver regeneration after partial hepatectomy in rats. Hepatology, 1988;8(5):992–6. 22. Olsen, P.S., Poulsen, S.S., and Kirkegaard, P. Adrenergic effects on secretion of epidermal growth factor from Brunner’s glands. Gut, 1985;26(9):920–7. 23. Mead, J.E. and Fausto, N. Transforming growth factor alpha may be a physiological regulator of liver regeneration by means of an autocrine mechanism. Proc Natl Acad Sci USA, 1989;86(5):1558–62. 24. Berasain, C., Garcia‐Trevijano, E.R., Castillo, J. et  al. Amphiregulin: an early trigger of liver regeneration in mice.[see comment]. Gastroenterology, 2005;128(2):424–32. 25. Khai, N.C., Takahashi, T., Ushikoshi, H. et  al. In vivo hepatic HB‐EGF gene transduction inhibits Fas‐induced liver injury and induces liver regeneration in mice: a comparative study to HGF. J Hepatol, 2006;44(6): 1046–54. 26. Mitchell, C., Nivison, M., Jackson, L.F. et  al. Heparin‐binding epidermal growth factor‐like growth factor links hepatocyte priming with cell cycle progression during liver regeneration. J Biol Chem, 2005;280(4):2562–8. 27. Huang, X., Yu, C., Jin, C. et al. Ectopic activity of fibroblast growth factor receptor 1 in hepatocytes accelerates hepatocarcinogenesis by driving proliferation and vascular endothelial growth factor‐induced angiogenesis. Cancer Res, 2006;66(3):1481–90. 28. Luo, Y., Yang, C., Lu, W. et al. Metabolic regulator betaKlotho interacts with fibroblast growth factor receptor 4 (FGFR4) to induce apoptosis and inhibit tumor Cell, proliferation. J Biol Chem, 2010;285(39):30069–78. 29. Padrissa‐Altes, S., Bachofner, M., Bogorad, R.L. et al. Control of hepatocyte proliferation and survival by Fgf receptors is essential for liver regeneration in mice. Gut, 2015;64(9):1444–53. 30. Cicione, C., Degirolamo, C., and Moschetta, A. Emerging role of fibroblast growth factors 15/19 and 21 as metabolic integrators in the liver. Hepatology, 2012;56(6):2404–11. 31. Kong, B., Huang, J., Zhu, Y. et  al. Fibroblast growth factor 15 deficiency impairs liver regeneration in mice. Am J Physiol Gastrointest Liver Physiol, 2014;306(10):G893–902. 32. Yang, C., Lu, W., Lin, T. et al. Activation of Liver FGF21 in hepatocarcinogenesis and during hepatic stress. BMC Gastroenterol, 2013;13:67. 33. Schumacher, J.D. and Guo, G.L. Regulation of hepatic stellate cells and fibrogenesis by fibroblast growth factors. Biomed Res Int, 2016;2016: 8323747. 34. Jirtle, R.L., Carr, B.I., and Scott, C.D. Modulation of insulin‐like growth factor‐II/mannose 6‐phosphate receptors and transforming growth factor‐ beta 1 during liver regeneration (published erratum appears in J Biol Chem, 1991;266(36):24860). J Biol Chem, 1991;266(3):22444–50. 35. Michalopoulos, G.K. Principles of liver regeneration and growth homeostasis. Compr Physiol, 2013;3(1):485–513. 36. Jakowlew, S.B., Mead, J.E., Danielpour, D., Wu, J., Roberts, A.B., and Fausto, N. Transforming growth factor‐beta (TGF‐beta) isoforms in rat liver regeneration: messenger RNA expression and activation of latent TGF‐ beta. Cell Regul, 1991;2(7):535–48.



45:  Liver Regeneration

37. Chari, R.S., Price, D.T., Sue, S.R., Meyers, W.C., and Jirtle, R.L. Down‐ regulation of transforming growth factor beta receptor type I., I.I., and III during liver regeneration. Am J Surg, 1995;169(1):126–31. 38. Houck, K.A., Cruise, J.L., and Michalopoulos, G. Norepinephrine modulates the growth‐inhibitory effect of transforming growth factor‐beta in primary rat hepatocyte cultures. J Cell Physiol, 1988;135(3):551–5. 39. Michalopoulos, G.K., Bowen, W.C., Mule, K., and Stolz, D.B. Histological organization in hepatocyte organoid cultures. Am J Pathol, 2001;159(5): 1877–87. 40. Rudolph, K.L., Trautwein, C., Kubicka, S. et al. Differential regulation of extracellular matrix synthesis during liver regeneration after partial hepatectomy in rats. Hepatology, 1999;30(5):1159–66. 41. Thenappan, A., Shukla, V., Abdul Khalek, F.J. et  al. Loss of transforming growth factor beta adaptor protein beta‐2 spectrin leads to delayed liver regeneration in mice. Hepatology, 2011;53(5):1641–50. 42. Ichikawa, T., Zhang, Y.Q., Kogure, K. et al. Transforming growth factor beta and activin tonically inhibit DNA synthesis in the rat liver. Hepatology, 2001;34(5):918–25. 43. Cressman, D.E., Greenbaum, L.E., DeAngelis, R.A. et al. Liver failure and defective hepatocyte regeneration in interleukin‐6‐deficient mice. Science, 1996;274(5291):1379–83. 44. Sakamoto, T., Liu, Z., Murase, N. et al. Mitosis and apoptosis in the liver of  interleukin‐6‐deficient mice after partial hepatectomy. Hepatology, 1999;29(2):403–11. 45. Runge, D.M., Runge, D., Foth, H., Strom, S.C., and Michalopoulos, G.K. STAT 1alpha/1beta, STAT 3 and STAT 5: expression and association with c‐ MET and EGF‐receptor in long‐term cultures of human hepatocytes. Biochem Biophys Res Commun, 1999;265(2):376–81. 46. Norris, C.A., He, M., Kang, L.I. et al. Synthesis of IL‐6 by hepatocytes is a normal response to common hepatic stimuli. PLoS One, 2014;9(4):e96053. 47. Fausto, N. Involvement of the innate immune system in liver regeneration and injury. J Hepatol, 2006;45(3):347–9. 48. Chaisson, M.L., Brooling, J.T., Ladiges, W., Tsai, S., and Fausto, N. Hepatocyte‐specific inhibition of NF‐kappaB leads to apoptosis after TNF treatment, but not after partial hepatectomy. J Clin Invest, 2002;110(2): 193–202. 49. Yamada, Y., Webber, E.M., Kirillova, I., Peschon, J.J., and Fausto, N. Analysis of liver regeneration in mice lacking type 1 or type 2 tumor necrosis factor receptor: requirement for type 1 but not type 2 receptor [comment]. Hepatology, 1998;28(4):959–70. 50. Argast, G.M., Campbell, J.S., Brooling, J.T., and Fausto, N. Epidermal growth factor receptor transactivation mediates tumor necrosis factor‐ induced hepatocyte replication. J Biol Chem, 2004;279(3):34530–6. 51. Desbarats, J. and Newell, M.K. Fas engagement accelerates liver regeneration after partial hepatectomy. Nat Med, 2000;6(8):920–3. 52. Oben, J.A., Roskams, T., Yang, S. et  al. Hepatic fibrogenesis requires ­sympathetic neurotransmitters. Gut, 2004;53(3):438–45. 53. Cruise, J.L., Houck, K.A., and Michalopoulos, G.K. Induction of DNA ­synthesis in cultured rat hepatocytes through stimulation of alpha 1 adrenoreceptor by norepinephrine. Science, 1985;227(4688):749–51. 54. Han, C., Bowen, W.C., Michalopoulos, G.K., and Wu, T. Alpha‐1 adrenergic receptor transactivates signal transducer and activator of transcription‐3 (Stat3) through activation of Src and epidermal growth factor receptor (EGFR) in hepatocytes. J Cell Physiol, 2008;216(2):486–97. 55. Cruise, J.L., Knechtle, S.J., Bollinger, R.R., Kuhn, C., and Michalopoulos, G. Alpha 1‐adrenergic effects and liver regeneration. Hepatology, 1987;7(6): 1189–94. 56. Broten, J., Michalopoulos, G., Petersen, B., and Cruise, J. Adrenergic ­stimulation of hepatocyte growth factor expression. Biochem Biophys Res Commun, 1999;262(1):76–9. 57. Yanagita, K., Nagaike, M., Ishibashi, H., Niho, Y., Matsumoto, K., and Nakamura, T. Lung may have an endocrine function producing hepatocyte growth factor in response to injury of distal organs. Biochem Biophys Res Commun, 1992;182(2):802–9. 58. Huang, W., Ma, K. Zhang, J. et al. Nuclear receptor‐dependent bile acid signaling is required for normal liver regeneration. Science, 2006;312:233–6. 59. Borude, P., Edwards, G., Walesky, C. et al. Hepatocyte‐specific deletion of farnesoid X receptor delays but does not inhibit liver regeneration after partial hepatectomy in mice. Hepatology, 2012;56(6):2344–52. 60. Keitel, V. and Haussinger, D. TGR5 in the biliary tree. Dig Dis, 2011;29(1):45–7. 61. Lesurtel, M., Graf, R., Aleil, B. et  al. Platelet‐derived serotonin mediates liver regeneration. Science, 2006;312:104–7.

581

62. Matondo, R.B., Punt, C., Homberg, J. et  al. Deletion of the serotonin transporter in rats disturbs serotonin homeostasis without impairing liver regeneration. Am J Physiol Gastrointest Liver Physiol, 2009;296(4): G963–8. 63. Furrer, K., Rickenbacher, A., Tian, Y. et al. Serotonin reverts age‐related capillarization and failure of regeneration in the liver through a VEGF‐dependent pathway. Proc Natl Acad Sci USA, 2011;108(7):2945–50. 64. Ochoa, B., Syn, W.K., Delgado, I. et al. Hedgehog signaling is critical for normal liver regeneration after partial hepatectomy in mice. Hepatology, 2010;51(5):1712–23. 65. Bhave, V.S., Mars, W., Donthamsetty, S. et al. Regulation of liver growth by glypican 3, CD81, hedgehog, and Hhex. Am J Pathol, 2013;183(1):153–9. 66. Sicklick, J.K., Li, Y.X., Choi, S.S. et  al. Role for hedgehog signaling in hepatic stellate cell activation and viability. Lab Invest, 2005;85(1):1368–80. 67. Swiderska‐Syn, M., Xie, G., Michelotti, G.A. et al. Hedgehog regulates yes‐ associated protein 1 in regenerating mouse liver. Hepatology, 2016;64(1): 232–44. 68. Francavilla, A., Starzl, T.E., Porter, K. et al. Screening for candidate hepatic growth factors by selective portal infusion after canine Eck’s fistula. Hepatology, 1991;14(4 Pt 1):665–70. 69. Zerrad‐Saadi, A., Lambert‐Blot, M., Mitchell, C. et al. GH receptor plays a major role in liver regeneration through the control of EGFR and ERK1/2 activation. Endocrinology, 2011;152(7):2731–41. 70. Jin, J., Wang, G.L., Shi, X., Darlington, G.J.,and Timchenko, N.A. The age‐ associated decline of glycogen synthase kinase 3beta plays a critical role in the inhibition of liver regeneration. Mol Cell Biol, 2009;29(14):3867–80. 71. Tackett, B.C., Sun, H., Mei, Y. et al. P2Y2 purinergic receptor activation is essential for efficient hepatocyte proliferation in response to partial hepatectomy. Am J Physiol Gastrointest Liver Physiol, 2014;307(1):G1073–87. 72. Graubardt, N., Fahrner, R., Trochsler, M, et al. Promotion of liver regeneration by natural killer cells in a murine model is dependent on extracellular adenosine triphosphate phosphohydrolysis. Hepatology, 2013;57(5):1969–79. 73. Moolten, F.L. and Bucher, N.L. Regeneration of rat liver: transfer of humoral agent by cross circulation. Science, 1967;158(798):272–4. 74. Marubashi, S., Sakon, M., Nagano, H. et al. Effect of portal hemodynamics on liver regeneration studied in a novel portohepatic shunt rat model. Surgery, 2004;136(5):1028–37. 75. Sokabe, T., Yamamoto, K., Ohura, N. et al. Differential regulation of urokinase‐type plasminogen activator expression by fluid shear stress in human coronary artery endothelial cells. Am J Physiol, 2004;287(5):H2027–34. 76. Mars, W.M., Kim, T.H., Stolz, D.B., Liu, M.L., and Michalopoulos, G.K. Presence of urokinase in serum‐free primary rat hepatocyte cultures and its  role in activating hepatocyte growth factor. Cancer Res, 1996;56(12): 2837–43. 77. Nejak‐Bowen, K., Moghe, A., Cornuet, P., Preziosi, M., Nagarajan, S., and Monga, S.P. Role and regulation of p65/beta‐catenin association during liver injury and regeneration: a "complex" relationship. Gene Expr, 2017;17(3): 219–35. 78. Mars, W.M., Zarnegar, R., and Michalopoulos, G.K. Activation of hepatocyte growth factor by the plasminogen activators uPA and tPA. Am J Pathol, 1993;143(3):949–58. 79. Mohammed, F.F., Pennington, C.J., Kassiri, Z. et  al. Metalloproteinase inhibitor TIMP‐1 affects hepatocyte cell cycle via HGF activation in murine liver regeneration. Hepatology, 2005;41(4):857–67. 80. Kim, T.H., Bowen, W.C., Stolz, D.B., Runge, D., Mars, W.M., and Michalopoulos, G.K. Differential expression and distribution of focal adhesion and cell adhesion molecules in rat hepatocyte differentiation. Exp Cell Res, 1998;244(1):93–104. 81. Lindroos, P.M., Zarnegar, R., and Michalopoulos, G.K. Hepatocyte growth factor (hepatopoietin A) rapidly increases in plasma before DNA synthesis and liver regeneration stimulated by partial hepatectomy and carbon tetrachloride administration. Hepatology, 1991;13(4):743–50. 82. Michalopoulos, G.K. Liver regeneration after partial hepatectomy: critical analysis of mechanistic dilemmas. Am J Pathol, 2010;176(1):2–13. 83. Michalopoulos, G.K. Advances in liver regeneration. Exp Rev Gastroenterol Hepatol, 2014:1–11. 84. Nejak‐Bowen, K., Orr, A., Bowen, W.C., Jr., and Michalopoulos, G.K. Conditional genetic elimination of hepatocyte growth factor in mice compromises liver regeneration after partial hepatectomy. PLoS One, 2013;8(3): e59836. 85. Hojilla, C.V., Mohammed, F.F., and Khokha, R. Matrix metalloproteinases and their tissue inhibitors direct Cell, fate during cancer development. Br J Cancer, 2003;89(10):1817–21.

582

THE LIVER:  REFERENCES

86. Gkretsi, V., Bowen, W.C., Yang, Y., Wu, C., and Michalopoulos, G.K. Integrin‐linked kinase is involved in matrix‐induced hepatocyte differentiation. Biochem Biophys Res Commun, 2007;353(3):638–43. 87. Gkretsi, V., Apte, U., Mars, W.M. et al. Liver‐specific ablation of integrin‐ linked kinase in mice results in abnormal histology, enhanced cell proliferation, and hepatomegaly. Hepatology, 2008;48(6):1932–41. 88. Apte, U., Gkretsi, V., Bowen, W.C. et al. Enhanced liver regeneration following changes induced by hepatocyte‐specific genetic ablation of integrin‐ linked kinase. Hepatology, 2009;50(3):844–51. 89. Monga, S.P., Pediaditakis, P., Mule, K., Stolz, D.B., and Michalopoulos, G.K. Changes in WNT/beta‐catenin pathway during regulated growth in rat liver regeneration. Hepatology, 2001;33(5):1098–109. 90. Kohler, C., Bell, A.W., Bowen, W.C., Monga, S.P., Fleig, W., and Michalopoulos, G.K. Expression of Notch‐1 and its ligand Jagged‐1 in rat liver during liver regeneration. Hepatology, 2004;39(4):1056–65. 91. Zhu, N.L., Asahina, K., Wang, J. et al. Hepatic stellate cell‐derived delta‐ like  homolog 1 (DLK1) protein in liver regeneration. J Biol Chem, 2012;287(13):10355–67. 92. Jirtle, R.L. and Michalopoulos, G. Effects of partial hepatectomy on ­transplanted hepatocytes. Cancer Res, 1982;42(8):3000–4. 93. Weymann, A., Hartman, E., Gazit, V. et  al. p21 is required for dextrose‐ mediated inhibition of mouse liver regeneration. Hepatology, 2009;50(1): 207–15. 94. Shido, K., Chavez, D., Cao, Z., Ko, J., Rafii, S., and Ding, B.S. Platelets prime hematopoietic and vascular niche to drive angiocrine‐mediated liver regeneration. Signal Transduct Target Ther, 2017;2. 95. Stocker, E. and Heine, W.D. Regeneration of liver parenchyma under normal and pathological conditions. Beitr Pathol, 1971;144(4):400–8. 96. Stocker, E., Wullstein, H.K., and Brau, G. Capacity of regeneration in liver  epithelia of juvenile, repeated partially hepatectomized rats. Autoradiographic studies after continous infusion of 3H‐thymidine (author’s transl). Virchows Arch B Cell Pathol, 1973;14(2):93–103. 97. Kelley‐Loughnane, N., Sabla, G.E., Ley‐Ebert, C., Aronow, B.J., and Bezerra, J.A. Independent and overlapping transcriptional activation during liver development and regeneration in mice. Hepatology, 2002;35(3): 525–34. 98. Taub, R., Greenbaum, L.E., and Peng, Y. Transcriptional regulatory signals define cytokine‐dependent and ‐independent pathways in liver regeneration. Semin Liver Dis, 1999;19(2):117–27. 99. Yamamoto, T., Kojima, T., Murata, M. et  al. p38 MAP‐kinase regulates function of gap and tight junctions during regeneration of rat hepatocytes. J Hepatol, 2005;42(5):707–18. 100. Stolz, D.B., Mars, W.M., Petersen, B.E., Kim, T.H., and Michalopoulos, G.K. Growth factor signal transduction immediately after two‐thirds partial hepatectomy in the rat. Cancer Res, 1999;59(16):3954–60. 101. Li, W., Liang, X., Kellendonk, C., Poli, V.,and Taub, R. STAT3 contributes to the mitogenic response of hepatocytes during liver regeneration. J Biol Chem, 2002;277(32):28411–7. 102. Chaisson, M.L., Brooling, J.T., Ladiges, W., Tsai, S., and Fausto, N. Hepatocyte‐specific inhibition of NF‐kappaB leads to apoptosis after TNF treatment, but not after partial hepatectomy. J Clin Invest, 2002;110(2): 193–202. 103. Greenbaum, L.E., Cressman, D.E., Haber, B.A., and Taub, R. Coexistence of C/EBP alpha, beta, growth‐induced proteins and DNA synthesis in hepatocytes during liver regeneration. Implications for maintenance of the differentiated state during liver growth. J Clin Invest, 1995;96(3):1351–65. 104. Luedde, T., Duderstadt, M., Streetz, K.L. et al. C/EBP beta isoforms LIP and LAP modulate progression of the cell cycle in the regenerating mouse liver. Hepatology, 2004;40(2):356–65. 105. Wang, X., Quail, E., Hung, N.J., Tan, Y., Ye, H., and Costa, R.H. Increased levels of forkhead box M1B transcription factor in transgenic mouse hepatocytes prevent age‐related proliferation defects in regenerating liver. Proc Natl Acad Sci USA, 2001;98(20):11468–73. 106. Wang, X., Bhattacharyya, D., Dennewitz, M.B. et  al. Rapid hepatocyte nuclear translocation of the Forkhead Box M1B (FoxM1B) transcription factor caused a transient increase in size of regenerating transgenic hepatocytes. Gene Expr, 2003;11(3–4):149–62. 107. Bhave, V.S., Paranjpe, S., Bowen, W.C. et al. Genes inducing iPS phenotype play a role in hepatocyte survival and proliferation in vitro and liver regeneration in vivo. Hepatology, 2011;54(4):1360–70.

108. Mullany, L.K., White, P., Hanse, E.A. et al. Distinct proliferative and transcriptional effects of the D‐type cyclins in vivo. Cell Cycle, 2008;7(14): 2215–24. 109. Ng, R., Song, G., Roll, G.R., Frandsen, N.M., and Willenbring, H. A microRNA‐21 surge facilitates rapid cyclin D1 translation and Cell, cycle progression in mouse liver regeneration. J Clin Invest, 2012;122(3): 1097–108. 110. Albrecht, J.H., Poon, R.Y., Ahonen, C.L., Rieland, B.M., Deng, C., and Crary, G.S. Involvement of p21 and p27 in the regulation of CDK activity and Cell, cycle progression in the regenerating liver. Oncogene, 1998;16(16):2141–50. 111. Fausto, N. Protooncogenes and growth factors associated with normal and abnormal liver growth. Dig Dis Sci, 1991;36(5):653–8. 112. Rabes, H.M. Kinetics of hepatocellular proliferation as a function of the microvascular structure and functional state of the liver. Ciba Found Symp, 1977(5):31–53. 113. Klochendler, A., Weinberg‐Corem, N., Moran, M. et al. A transgenic mouse marking live replicating Cell,s reveals in vivo transcriptional program of proliferation. Dev Cell Cell, 2012;23(4):681–90. 114. Jiao, H., Zhu, Y., Lu, S., Zheng, Y., and Chen, H. An integrated approach for  the identification of HNF4alpha‐centered transcriptional regulatory ­networks during early liver regeneration. Cell Physiol Biochem, 2015;36(6): 2317–26. 115. Michalopoulos, G., Cianciulli, H.D., Novotny, A.R., Kligerman, A.D., Strom, S.C., and Jirtle, R.L. Liver regeneration studies with rat hepatocytes in primary culture. Cancer Res, 1982;42(1):4673–82. 116. Shteyer, E., Liao, Y., Muglia, L.J., Hruz, P.W., and Rudnick, D.A. Disruption of hepatic adipogenesis is associated with impaired liver regeneration in mice (see comment). Hepatology, 2004;40(6):1322–32. 117. Fernandez‐Rojo, M.A., Restall, C., Ferguson, C. et al. Caveolin‐1 orchestrates the balance between glucose and lipid‐dependent energy metabolism: implications for liver regeneration. Hepatology, 2012;55(5):1574–84. 118. Shu, J., Kren, B.T., Xia, Z. et al. Genomewide microRNA down‐regulation as a negative feedback mechanism in the early phases of liver regeneration. Hepatology, 2011;54(2):609–19. 119. Grijalva, J.L., Huizenga, M., Mueller, K. et  al. Dynamic alterations in Hippo signaling pathway and YAP activation during liver regeneration. Am J Physiol Gastrointest Liver Physiol, 2014;307(2):G196–204. 120. Yimlamai, D., Christodoulou, C., Galli, G.G. et al. Hippo pathway activity influences liver cell fate. Cell, 2014;157(6):1324–38. 121. Toshima, T., Shirabe, K., Fukuhara, T. et al. Suppression of autophagy during liver regeneration impairs energy charge and hepatocyte senescence in mice. Hepatology, 2014;60(1):290–300. 122. Hoehme, S., Brulport, M., Bauer, A. et al. Prediction and validation of cell alignment along microvessels as order principle to restore tissue architecture in liver regeneration. Proc Nat Acad Sci USA, 2010;107(23):10371–6. 123. Miyaoka, Y. and Miyajima, A. To divide or not to divide: revisiting liver regeneration. Cell Div, 2013;8(1):8. 124. Meier, M., Knudsen, A.R., Andersen, K.J., Bjerregaard, N.C., Jensen, U.B., and Mortensen, F.V. Gene expression in the liver remnant is significantly affected by the size of partial hepatectomy: an experimental rat study. Gene Expr, 2017;17(4):289–99. 125. Pean, N., Doignon, I., Garcin, I. et al. The receptor TGR5 protects the liver from bile acid overload during liver regeneration in mice. Hepatology, 2013;58(4):1451–60. 126. Glaser, S., Han, Y., Francis, H., and Alpini, G. Melatonin regulation of biliary functions. Hepatobiliary Surg Nutr, 2014;3(1):35–43. 127. Kennedy, L., Hargrove, L., Demieville, J. et al. Blocking H1/H2 histamine receptors inhibits damage/fibrosis in Mdr2(‐/‐) mice and human cholangiocarcinoma tumorigenesis. Hepatology, 2018. 128. Grappone, C., Pinzani, M., Parola, M. et al. Expression of platelet‐derived growth factor in newly formed cholangiocytes during experimental biliary fibrosis in rats. J Hepatol, 1999;31(1):100–9. 129. Li, B., Dorrell, C., Canaday, P.S. et al. Adult mouse liver contains two distinct populations of cholangiocytes. Stem Cell Reports, 2017;9(2):478–89. 130. Ross, M.A., Sander, C.M., Kleeb, T.B., Watkins, S.C., and Stolz, D.B. Spatiotemporal expression of angiogenesis growth factor receptors during the revascularization of regenerating rat liver. Hepatology, 2001;34(6):1135–48. 131. Modis, L. and Martinez‐Hernandez, A. Hepatocytes modulate the hepatic microvascular phenotype. Lab Invest, 1991;65(6):661–70.



45:  Liver Regeneration

132. Wack, K.E., Ross, M.A., Zegarra, V., Sysko, L.R., Watkins, S.C., and Stolz, D.B. Sinusoidal ultrastructure evaluated during the revascularization of regenerating rat liver. Hepatology, 2001;33(2):363–78. 133. DeLeve, L.D., Wang, X., and Wang, L. VEGF‐sdf1 recruitment of CXCR7+ bone marrow progenitors of liver sinusoidal endothelial cells promotes rat liver regeneration. Am J Physiol Gastrointest Liver Physiol< 2016;310(9): G739–46. 134. Ding, B.S., Nolan, D.J., Butler, J.M. et  al. Inductive angiocrine signals from  sinusoidal endothelium are required for liver regeneration. Nature, 2010;468(7321):310–5. 135. Roskams, T., Cassiman, D., De Vos, R., and Libbrecht, L. Neuroregulation of the neuroendocrine compartment of the liver. Anat Rec A Discov Mol Cell Cell Evol Biol, 2004;280(1):910–23. 136. Unanue, E.R. Ito cells, stellate cells, and myofibroblasts: new actors in antigen presentation. Immunity, 2007;26(1):9–10. 137. Gandhi, C.R. Hepatic stellate cell activation and pro‐fibrogenic signals. J Hepatol, 2017;67(5):1104–5. 138. Zimmermann, A. Liver regeneration: the emergence of new pathways. Med Sci Monit, 2002;8(3):RA53–63. 139. Widmann, J.J. and Fahimi, H.D. Proliferation of mononuclear phagocytes (Kupffer cells) and endothelial cells in regenerating rat liver. A light and electron microscopic cytochemical study. Am J Pathol, 1975;80(3): 349–66. 140. Fujii, H., Hirose, T., Oe, S. et al. Contribution of bone marrow cells to liver regeneration after partial hepatectomy in mice. J Hepatol, 2002;36(5):653–9. 141. Ikarashi, M., Nakashima, H., Kinoshita, M. et  al. Distinct development and  functions of resident and recruited liver Kupffer cells/macrophages. J Leukoc Biol, 2013;94(6):1325–36. 142. Sanderson, N., Factor, V., Nagy, P. et  al. Hepatic expression of mature ­transforming growth factor beta 1 in transgenic mice results in multiple tissue lesions. Proc Nat Acad Sci USA, 1995;92(7):2572–6. 143. Oe, S., Lemmer, E.R., Conner, E.A. et al. Intact signaling by transforming growth factor beta is not required for termination of liver regeneration in mice. Hepatology, 2004;40(5):1098–105. 144. Kim, T.H., Mars, W.M., Stolz, D.B., and Michalopoulos, G.K. Expression and activation of pro‐MMP‐2 and pro‐MMP‐9 during rat liver regeneration. Hepatology, 2000;31(1):75–82. 145. Kim, T.H., Mars, W.M., Stolz, D.B., Petersen, B.E., and Michalopoulos, G.K. Extracellular matrix remodeling at the early stages of liver regeneration in the rat. Hepatology, 1997;26(4):896–904. 146. Luo, J.H., Ren, B., Keryanov, S. et al. Transcriptomic and genomic analysis of human hepatocellular carcinomas and hepatoblastomas. Hepatology, 2006;44(4):1012–24. 147. Pilia, G., Hughes‐Benzie, R.M., MacKenzie, A. et al. Mutations in GPC3, a glypican gene, cause the Simpson‐Golabi‐Behmel overgrowth syndrome. Nat Genet, 1996;12(3):241–7. 148. Liu, B., Bell, A.W., Paranjpe, S. et al. Suppression of liver regeneration and hepatocyte proliferation in hepatocyte‐targeted glypican 3 transgenic mice. Hepatology, 2010;52(3):1060–7. 149. Xue, Y., Mars, W.M., Bowen, W., Singhi, A.D., Stoops, J., and Michalopoulos, G.K. Hepatitis C virus mimics effects of Glypican‐3 on CD81 and promotes development of hepatocellular carcinomas via activation of hippo pathway in hepatocytes. Am J Pathol, 2018;188(6):1469–77. 150. Koral, K., Paranjpe, S., Bowen, W.C., Mars, W., Luo, J., and Michalopoulos, G.K. Leukocyte‐specific protein 1: a novel regulator of hepatocellular proliferation and migration deleted in human hepatocellular carcinoma. Hepatology, 2015;61(2):537–47. 151. Yang, J., Cusimano, A., Monga, J.K. et al. WNT5A inhibits hepatocyte proliferation and concludes beta‐catenin signaling in liver regeneration. Am J Pathol, 2015;185(8):2194–205. 152. Tsagianni, A., Mars, W.M., Bhushan, B. et al. Combined systemic disruption of MET and epidermal growth factor receptor signaling causes liver failure in normal mice. Am J Pathol, 2018;188(10):2223–35. 153. Arakaki, N., Kawakami, S., Nakamura, O. et al. Evidence for the presence of an inactive precursor of human hepatocyte growth factor in plasma and sera of patients with liver diseases. Hepatology, 1995;22(6):1728–34. 154. Pediaditakis, P., Lopez‐Talavera, J.C., Petersen, B., Monga, S.P., and Michalopoulos, G.K. The processing and utilization of hepatocyte growth factor/scatter factor following partial hepatectomy in the rat. Hepatology, 2001;34(4 Pt 1):688–93.

583

155. Le Bail, B., Bioulac‐Sage, P., Senuita, R., Quinton, A., Saric, J., and Balabaud, C. Fine structure of hepatic sinusoids and sinusoidal cells in disease. J Electron Microsc Tech, 1990;14(3):257–82. 156. Michalopoulos, G.K. and Khan, Z. Liver stem cells: experimental findings and implications for human liver disease. Gastroenterology, 2015;149(4):876–82. 157. Trautwein, C., Will, M., Kubicka, S., Rakemann, T., Flemming, P., and Manns, M.P. 2‐acetaminofluorene blocks cell cycle progression after hepatectomy by p21 induction and lack of cyclin E expression. Oncogene, 1999;18(47):6443–53. 158. Tatematsu, M., Ho, R.H., Kaku, T., Ekem, J.K., and Farber, E. Studies on the proliferation and fate of oval cells in the liver of rats treated with 2‐ acetylaminofluorene and partial hepatectomy. Am J Pathol, 1984;114(3): 418–30. 159. Evarts, R.P., Nagy, P., Nakatsukasa, H., Marsden, E., and Thorgeirsson, S.S. In vivo differentiation of rat liver oval cells into hepatocytes. Cancer Res, 1989;49(6):1541–7. 160. Bisgaard, H.C., Nagy, P., Santoni‐Rugiu, E., and Thorgeirsson, S.S. Proliferation, apoptosis, and induction of hepatic transcription factors are characteristics of the early response of biliary epithelial (oval) cells to chemical carcinogens. Hepatology, 1996;23(1):62–70. 161. Petersen, B.E., Zajac, V.F., and Michalopoulos, G.K. Bile ductular damage induced by methylene dianiline inhibits oval cell activation. J Pathol, 1997;151(4):905–9. 162. Lu, W.Y., Bird, T.G., Boulter, L. et al. Hepatic progenitor cells of biliary origin with liver repopulation capacity. Nat Cell Biol, 2015. 163. Michalopoulos, G.K., Barua, L., and Bowen, W.C. Transdifferentiation of rat hepatocytes into biliary cells after bile duct ligation and toxic biliary injury. Hepatology, 2005;41(3):535–44. 164. Limaye, P.B., Bowen, W.C., Orr, A.V., Luo, J., Tseng, G.C., and Michalopoulos, G.K. Mechanisms of hepatocyte growth factor‐mediated and epidermal growth factor‐mediated signaling in transdifferentiation of rat hepatocytes to biliary epithelium. Hepatology, 2008;47(5):1702–13. 165. Sekiya, S., and Suzuki, A. Hepatocytes, rather than cholangiocytes, can be the major source of primitive ductules in the chronically injured mouse liver. Am J Pathol, 2014;184(5):1468–78. 166. Schaub, J.R., Huppert, K.A., Kurial, S.N.T. et al. De novo formation of the biliary system by TGFbeta‐mediated hepatocyte transdifferentiation. Nature, 2018;557(7704):247–51. 167. Sparks, E.E., Huppert, K.A., Brown, M.A., Washington, M.K., and Huppert, S.S. Notch signaling regulates formation of the three‐dimensional architecture of intrahepatic bile ducts in mice. Hepatology, 2010;51(4):1391–400. 168. Isse, K., Lesniak, A., Grama, K. et  al. Preexisting epithelial diversity in normal human livers: a tissue‐tethered cytometric analysis in portal/periportal epithelial cells. Hepatology, 2013;57(4):1632–43. 169. Petersen, B.E., Zajac, V.F., and Michalopoulos, G.K. Hepatic oval cell activation in response to injury following chemically induced periportal or pericentral damage in rats. Hepatology, 1998;27(4):1030–8. 170. Carpentier, R., Suner, R.E., van Hul, N. et al. Embryonic ductal plate cells give rise to cholangiocytes, periportal hepatocytes, and adult liver progenitor cells. Gastroenterology,. 2011;141(4):1432–8, 8 e1–4. 171. Michalopoulos, G.K., Bowen, W.C., Mule, K., and Luo, J. HGF‐, EGF‐, and dexamethasone‐induced gene expression patterns during formation of tissue in hepatic organoid cultures. Gene Expr, 2003;11(2):55–75. 172. Hattoum, A., Rubin, E., Orr, A., and Michalopoulos, G.K. Expression of hepatocyte epidermal growth factor receptor, FAS and glypican 3 in EpCAM‐positive regenerative clusters of hepatocytes, cholangiocytes, and progenitor cells in human liver failure. Hum Pathol, 2013;44(5):743–9. 173. Jaeschke, H. and Bajt, M.L. Intracellular signaling mechanisms of acetaminophen‐induced liver cell death. Toxicol Sci, 2006;89(1):31–41. 174. Slater, T.F., Cheeseman, K.H., and Ingold, K.U. Carbon tetrachloride toxicity as a model for studying free‐radical mediated liver injury. Phil Trans R Soc Lond B Biol Sci, 1985;311(1152):633–45. 175. Bhushan, B., Walesky, C., Manley, M. et al. Pro‐regenerative signaling after acetaminophen‐induced acute liver injury in mice identified using a novel incremental dose model. Am J Pathol, 2014;184(1):3013–25. 176. Bhushan, B., Chavan, H., Borude, P. et al. Dual role of epidermal growth factor receptor in liver injury and regeneration after acetaminophen overdose in mice. Toxicol Sci, 2017;155(2):363–78. 177. Webber, E.M., FitzGerald, M.J., Brown, P.I., Bartlett, M.H., and Fausto, N. Transforming growth factor‐alpha expression during liver regeneration

584

THE LIVER:  REFERENCES

after partial hepatectomy and toxic injury, and potential interactions between transforming growth factor‐alpha and hepatocyte growth factor. Hepatology, 1993;18(6):1422–31. 178. Lamers, W.H. The streaming liver: can the age of a hepatocyte be determined from its position on the portohepatic radius? Hepatology, ­ 1990;12(2):372–4. 179. Kennedy, S., Rettinger, S., Flye, M.W., and Ponder, K.P. Experiments in transgenic mice show that hepatocytes are the source for postnatal liver growth and do not stream. Hepatology, 1995;22(1):160–8. 180. Wang, B., Zhao, L., Fish, M., Logan, C.Y., and Nusse, R. Self‐renewing diploid Axin2(+) cells fuel homeostatic renewal of the liver. Nature, 2015;524(7564):180–5. 181. Furuyama, K., Kawaguchi, Y., Akiyama, H. et al. Continuous cell supply from a Sox9‐expressing progenitor zone in adult liver, exocrine pancreas and intestine. Nat Genet, 2011;43(1):34–41. 182. Font‐Burgada, J., Shalapour, S., Ramaswamy, S. et  al. Hybrid periportal hepatocytes regenerate the injured liver without giving rise to cancer. Cell, 2015;162(4):766–79. 183. Planas‐Paz, L., Orsini, V., Boulter, L. et  al. The RSPO‐LGR4/5‐ZNRF3/ RNF43 module controls liver zonation and size. Nat Cell Biol, 2016;18(5):467–79.

184. Lin, S., Nascimento, E.M., Gajera, C.R. et  al. Distributed hepatocytes expressing telomerase repopulate the liver in homeostasis and injury. Nature, 2018;556(7700):244–8. 185. Prassopoulos, P., Cavouras, D., and Gourtsoyiannis, N. Pre‐ and post‐ nephrectomy kidney enlargement in patients with contralateral renal ­cancer. Eur Urol, 1993;24(1):58–61. 186. Duncan, A.W., Taylor, M.H., Hickey, R.D. et al. The ploidy conveyor of mature hepatocytes as a source of genetic variation. Nature, 2010;467(7316):707–10. 187. Anti, M., Marra, G., Rapaccini, G.L. et al. DNA ploidy pattern in human chronic liver diseases and hepatic nodular lesions. Flow cytometric analysis on echo‐guided needle liver biopsy. Cancer, 1994;73(2):281–8. 188. Duncan, A.W., Hanlon Newell, A.E., Smith, L. et al. Frequent aneuploidy among normal human hepatocytes. Gastroenterology, 2012;142(1):25–8. 189. Solt, D.B., Medline, A.,and Farber, E. Rapid emergence of carcinogen‐ induced hyperplastic lesions in a new model for the sequential analysis of liver carcinogenesis. Am J Pathol, 1977;88(3):595–618. 190. Zhang, Y.Y., Zhang, B.H., Ishii, K., and Liang, T.J. Novel function of CD81 in controlling hepatitis C virus replication. J Virol, 2010;84(7):3396–407. 191. Overturf, K., al‐Dhalimy, M., Ou, C.N., Finegold, M.,and Grompe, M. Serial transplantation reveals the stem‐cell‐like regenerative potential of adult mouse hepatocytes. Am J Pathol, 1997;151(5):1273–80.

46

β‐Catenin Signaling Satdarshan P.S. Monga Pittsburgh Liver Research Center, University of Pittsburgh, School of Medicine and UPMC, Pittsburgh, PA, USA

BACKGROUND Genetic studies in species such as Xenopus, Drosophila, and Caenorhabditis have lent themselves well to improve our understanding of the molecular basis of human diseases. A classic example is the identification and characterization of the Wnt/β‐ catenin pathway that is crucial in normal development including embryogenesis and organogenesis, and at the same time its deregulation is implicated in disorders such as cancers (reviewed in [1–3]). This pathway has remained conserved through the evolutionary process. In Drosophila, the role of Wnt or Wingless (Wg) was initially identified in normal wing development, however it was later recognized for multiple functions such as inducing segment polarity and anterior–posterior patterning that were imperative for a viable embryo [4–6]. As the importance of Wnt emerged, several key components of this pathway were identified. The discovery of armadillo (or β‐catenin) added a significant player to this orchestra and although circumstantial evidence suggesting such a relationship existed earlier, it was a few years later that β‐catenin was positively identified as a central component of the canonical Wg pathway [4, 7–9]. These studies led to the emergence of a model system for cell adhesion and signal transduction [10]. This was also the beginning of the understanding of the Wnt/β‐catenin pathway and its role in complex cellular processes such as cell–cell adhesion, mitogenesis, motogenesis and morphogenesis in the vertebrates. The next several years focused on the discovery of various components of this pathway that improved our understanding of  the regulation of this complex signaling cascade in normal physiology and disease. Several important players such as the Wnt receptor frizzled (Fzd), zeste‐white 3 kinase or glycogen synthase kinase 3β (GSK3β), adenomatous polyposis gene product (APC), axin, dishevelled (Dsh), and T cell factor/

lymphoid enhancement factor (TCF/LEF) family of transcription factors, and their interactions were identified that were directly influenced by the Wnt signaling. Many new components and interactions as well as expanding list of target genes are being identified. Also, research is focused on their role in regulation of this pathway in health and disease. Furthermore, crosstalk has now been established between the Wnt pathway and other prominent pathways such as the Hippo, Hedgehog, Jagged/Notch, HGF, EGF, and TGFβ signaling that could have additional implications. Presently, the role of Wnt/β‐catenin pathway is well established in vertebrates in development, tissue homeostasis, and carcinogenesis [11, 12]. β‐catenin knockout yields an embryonic lethal phenotype in mice due to a defect in gastrulation [13]. Availability of conditional knockouts to overcome embryonic lethality has been key to understanding a more ubiquitous role of β‐catenin and other Wnt components in the development of many organs such as kidneys, lungs, brain, limbs, muscles, and skin [14–19]. The role of Wnt/β‐catenin signaling in liver pathobiology has gained significant importance in the last 15–20 years. This chapter will highlight the role of the Wnt/β‐catenin signaling pathway and of β‐catenin independent of the canonical Wnt pathway in liver physiology and pathobiology (Table 46.1).

THE WNT SIGNAL TRANSDUCTION PATHWAY The binding of an extracellular secreted glycoprotein Wnt to its  cell surface receptor Fzd or alternate receptors induces ­specific downstream cascades with unique biological functions. Although the most characterized signaling downstream of the

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

586

THE LIVER:  ALTERNATE WNT SIGNALING PATHWAYS

Table 46.1  Selected Wnt/β‐catenin pathway targets in liver Target genes

Context

Change

Reference

Axin‐2 Claudin‐2

Upregulated Downregulated

Cyclooxygenase‐2 Cyp1a2 Cyp2e1 Cyp7a1

Liver tumors Liver baseline/injury (β‐catenin knockout) Liver tumors (hepatoblastoma) Liver regeneration Liver tumors (cell line) Liver tumors Liver tumors Liver baseline/injury (β‐catenin knockout)

Cyp27

Liver baseline/injury (β‐catenin knockout)

Downregulated

Epidermal growth factor receptor Fibronectin G‐protein‐coupled receptor 49 (Gpr49 or Lgr5) Glutamate transporter‐1(GLT‐1) Glutamine synthetase (GS, Glul) Lect2 Ornithine aminotransferase Regenerating iselet‐derived‐3α Regucalcin

Transgenic liver (β‐catenin overexpression) Liver tumors (hepatoblastoma) Hepatocellular cancer

Upregulated Upregulated Upregulated

[212] [109, 112, 132] [101, 216, 217] [218] [219] [219] [112, 132, 133] [112, 132, 133] [119] [216] [220]

Transgenic liver (mutant β‐catenin overexpression) Transgenic liver (mutant β‐catenin overexpression) Transgenic liver (mutant β‐catenin overexpression) Transgenic liver (mutant β‐catenin overexpression) Liver tumors (HCC and hepatoblastoma Baseline liver and liver tumors (β‐catenin transgenic and knockout) Liver tumors (hepatoblastoma) Liver regeneration Liver tumors (cell line)

Upregulated Upregulated Upregulated Upregulated Upregulated Upregulated (transgenic) Downregulated (knockout) Upregulated Upregulated Upregulated

Cyclin‐D1

Tbx3 TGFα VEGF

Wnt is via the β‐catenin‐TCF‐dependent gene expression, the signaling can occur through other effectors. In this, a broad overview of the various signaling cascades activated downstream of Wnt are discussed.

THE WNT/β‐CATENIN SIGNALING In a normal steady state, in the absence of a Wnt signal, the free monomeric form of β‐catenin in the cytoplasm is actively targeted for degradation by ubiquitination. In this “off” state, β‐catenin is phosphorylated at serine45 (Ser45), Ser33, Ser33, and threonine‐41 (Thr41) by casein kinase Iα (CKIα) and GSK3β [20, 21]. CK and GSK3β are part of a larger multiprotein degradation complex that includes axin and APC. Once phosphorylated this larger complex enables recognition and ubiquitination of β‐catenin by β‐transducin repeat‐containing protein (βTrCP) and its ensuing proteosomal degradation (Figure 46.1a, left panel) [22]. Wnt proteins usually act on a cell in a paracrine or autocrine manner. However, Wnt proteins, in order to be biologically active, need to undergo specific post‐translational modifications. Porcupine, located in the endoplasmic reticulum of a cell, is essential for glycosylation and acylation of Wnts (Figure 46.1a, right panel) [23]. Following acylation, Wnts become hydrophobic and require a specific cargo receptor, Wntless (Wls) or Evenness Interrupted (Evi), to transport Wnts from Golgi to membrane for secretion (Figure 46.1a, right panel) [24]. Once secreted, Wnts bind to cell surface Fzd receptor and low‐density lipoprotein receptor‐related protein (LRP) 5/6 co‐ receptors [25–27]. This triggers phosphorylation of LRP5/6, recruitment of Dishevelled (Dvl), and recruitment of Axin to  the  plasma membrane (Figure  46.1, right panel) [28, 29]. Interestingly, the family of R‐spondin secreted proteins enhance

Upregulated Upregulated Upregulated Upregulated Downregulated

[99] [99] [221] [99] [222] [107] [223] [217] [224]

Wnt signaling through binding to the leucine‐rich repeat‐containing G‐protein coupled receptor‐4 (LGR4) and LGR5 receptors, which in turn increase Wnt‐dependent phosphorylation of LRP6 [30]. The recruitment of Axin to the plasma membrane disrupts the β‐catenin destruction complex, promoting its cytoplasmic stabilization and nuclear translocation where it binds and transactivates the TCF/LEF family of transcription factors to mediate target gene expression [31]. Several target genes of the pathway have now been identified, which are stage‐ and tissue‐specific, and the liver related targets will be discussed throughout the text. It is important to note that several Wnt pathway components such as AXIN, DKK, dFzd7, Fzd2, FRP2, WISP, βTrCP, and TCF are themselves targets, suggesting existence of several regulatory loops within this pathway.

ALTERNATE WNT SIGNALING PATHWAYS There are 19 Wnts and 10 Frizzled receptors in the mammalian genome. Understandably, not all Wnt signaling converges on β‐catenin and various Wnt‐Fzd or Wnt‐non‐Fzd interactions can lead to activation of alternate signaling pathways that are briefly discussed here. It should be noted that some Wnts that were classically dubbed to be non‐canonical have now been shown to activate or inactivate β‐catenin based on interaction with specific receptors and the final outcome of a particular signaling is context‐dependent. The Wnt signaling usually activates but rarely can inhibit β‐catenin‐TCF transactivation. An example is that of Wnt5a, which can activate or inhibit β‐catenin based on the receptor it binds [32]. It was shown that Wnt5a activated β‐catenin signaling in the presence of Fz4 while it inhibited Wnt3a‐dependent β‐catenin activation in the presence of Ror2 (Figure 46.1b).

46:  β-CATENIN SIGNALING

(a)

WNT

WNT

sFRP

WNT

WIF LRP5/6 Dkk

GSK3

Fzd

Wls

LRP5/6 Dsh

WNT

Fzd

APC GSK3 Axin

Axin APC

587

β-catenin

P

CK

Porcn

CK

WNT

β-catenin

TCF/LEF

P

β-catenin

β-catenin

Target genes

TCF/LEF

Serine 33 37 45

β-Catenin

HGF Y654 Y670

Threonine 41

(b)

(c) β-Catenin WNT5a

WNT5a

WNT5a

Fzd4

Fzd2

Ror2

Activate β-catenin

Inactivate β-catenin

EGF

Fer

Src (pp60)

Y654 Y142

Y86 Y654

β-Catenin

AJ β-Catenin

Actin cytoskeleton

E-Cadherin TyrosinePhosphoβ-Catenin

Occludin JAM-A Claudin-2 Claudin-1

?

TJ

TCF

Figure 46.1  β‐catenin as a component of the Wnt pathway and cell–cell junctions. (a) On the left, in the absence of Wnt binding to its receptor frizzled (Fzd) and co‐receptor LDL related protein 5/6 (LRP5/6) due to either sequestration of Wnt by soluble Frizzled related protein (sFRP), or in the presence of Wnt inhibitory factor (WIF) or inhibition of Fzd‐LRP5/6 interactions by Dikkopf (Dkk), β‐catenin in cytoplasm is bound to its degradation complex composed of Axin, APC, GSK3, and casein kinase (CK), which leads to its phosphorylation sequentially at Ser45 and Sr33, Ser37 and Thr41. After phosphorylation, β‐catenin is ubiquitinated by βTrCP and degraded. On the right, Wnt is glycosylated and palmitoylated by Porcupine (Porcn) in the endoplasmic reticulum and is bound to Wntless (Wls) in the Golgi apparatus to be transported to the membrane for secretion. Wnt binds to receptor Fzd and co‐receptor LRP5/6, which recruits the β‐catenin degradation complex through Dishevelled (Dsh) to phosphor‐ LRP5/6, inactivating the degradation complex, leading to hypophosphorylation of β‐catenin, which is released and eventually translocates to the nucleus to interact with T cell factor (TCF)/lymphoid enhancer factor (LEF) family of transcription factors to induce target genes. (b) Based on the receptor binding, Wnt5a can activate or inactivate β‐catenin. If it binds to Fzd4, it can activate β‐catenin while if it binds to Ror2 or Fzd2, it can inactivate β‐catenin through Wnt/calcium pathway or other mechanisms. (c) β‐Catenin is a component of adherens junctions (AJ) where it links the cytoplasmic tail of E‐cadherin to α‐catenin and actin cytoskeleton. AJ in the liver are next to tight junctions (TJ) which are known for their barrier function and are located close to the apical surface of hepatocytes preventing leakage of bile along the lateral surface of hepatocytes into the blood. TJ are composed of proteins like occludin, junctional adhesion molecule‐A (JAM‐A) and claudins. Claudin‐2 is also a Wnt/β‐catenin target and may be one basis of junctional crosstalk between the components of the AJ and TJ. Growth factors like HGF, EGF, Fer kinase, and Src kinase can phosphorylate β‐catenin at various tyrosine residues located at the C‐terminal of β‐catenin and disrupt β‐catenin‐E‐cadherin interactions and may also induce β‐catenin nuclear translocation and activation.

The Wnt/Ca2+ pathway Wnt ligands such as Wnt5a can bind to Fzd2 or Fzd7 or the tyrosine kinase‐like orphan receptor 2 (Ror2) to activate this pathway [33]. Binding of Wnt leads to formation of a complex between Fzd, Dishevelled, and G proteins, promoting the ­activation of phospholipase C (PLC), which cleaves phosphatidylinositol

4,5 biphosphate (PIP2) into diacylglycerol (DAG) and inositol 1,4,5‐triphosphate (IP3). DAG activates protein kinase C (PKC) and IP3 promotes the levels of intracellular calcium, which in turn activates calcium‐calmodulin dependent kinase II (CaMKII) and calcineurin (CaN) to regulate cell migration and proliferation [34].

588

THE LIVER:  β-CATENIN AND PROTEIN KINASE A (PKA)

The planar cell polarity pathway In the PCP pathway, binding of Wnt ligands to the Ror2/Fzd/ Dishevelled complex triggers the activation of Rho‐family small GTPases including RhoA and Rac to then activate the Rho‐associated protein kinase (ROCK) and c‐Jun N‐terminal kinase (JNK) with an eventual impact on cell polarity and migration [34, 35].

The Wnt/STOP pathway This pathway leads to Wnt‐dependent stabilization of proteins [36]. GSK3β is the central mediator of this pathway as it can phosphorylate many additional proteins besides β‐catenin to target them for proteasomal degradation [37]. Wnt binding to its co‐receptors can trigger the sequestration of GSK3β in multivesicular bodies, allowing the cytoplasmic accumulation of GSK3β‐target proteins [38]. The biological effects of this pathway can range from effect on cell cycle, cell division, regulation of cytoskeleton, cell size regulation, and DNA remodeling [39].

Wnt/TOR signaling In this pathway GSK3β phosphorylates and activates tuberous sclerosis complex 2 (TSC2), which in turn inhibits the function of mTOR complex 1 (mTORC1). The presence of Wnt ligands inhibits GSK3β‐mediated phosphorylation and degradation of TSC2, which activates the mTORC1 signaling pathway and stimulates protein translation [40].

WNT INDEPENDENT ROLES OF β‐CATENIN Apart from playing a central role in the canonical Wnt pathway, β‐catenin can interact with a number of proteins at membrane and in the cytoplasm or nuclei to modulate their signaling and biological function. A complete discussion on such interactions is out of the scope of the current chapter but some key interactions relevant in liver pathophysiology and pathobiology are briefly discussed here.

β‐CATENIN AT ADHERENS JUNCTIONS A crucial function of β‐catenin is to act as a bridge between the cytoplasmic domain of the cadherins and the actin‐containing‐cytoskeleton as component of the adherens junctions (AJ) [41]. Cadherins consist of an extracellular domain, a transmembrane domain, and a cytoplasmic tail that is the most conserved region among various subtypes. Structurally, the cytoplasmic tails of cadherins show dimerization and connect to the actin‐cytoskeleton via p120, β‐catenin, and α‐catenin. Specific β‐catenin‐binding sites on the cytoplasmic domain of cadherins have been characterized [42, 43]. The significance of regulation of β‐catenin‐cadherin interactions at AJ is not only important in modulating cell–cell adhesion but has been extended to transcriptional activation function of β‐catenin as

well, however, this remains controversial (Figure 46.1c). The interactions between β‐catenin and cadherins are regulated by tyrosine phosphorylation at the C‐terminal of β‐catenin (reviewed in [44]). Phosphorylation of β‐catenin destabilizes cadherin‐β‐catenin bond, α‐catenin‐β‐catenin complex, uncouples cadherin from actin cytoskeleton, and promotes loss of intracellular adhesion [45, 46]. Conversely, dephosphorylating β‐catenin at tyrosine residues enhanced E‐cadherin, β‐catenin, and α‐catenin reassembly [47]. Following tyrosine phosphorylation of β‐catenin, its cytosolic pool is greatly increased, as is its ability to bind to TATA‐box binding protein (TBP) to eventually increase transcriptional activity of the β‐catenin/TCF complex [48]. This has also been narrowed down to tyrosine residue 654. Another important ramification of tyrosine phosphorylation of β‐catenin and dissociation of β‐catenin‐E‐cadherin complex is that it leaves the cytoplasmic domain of E‐cadherin unstructured and vulnerable to degradation [42]. Several factors can regulate tyrosine phosphorylation of β‐catenin including: (i) nonreceptor kinases, Src and Fer [49, 50]; (ii) transmembrane kinases, EGF receptor (EGFR) and Met (HGF receptor) [51–55]; and (iii) various protein tyrosine phosphatases (Figure 46.1c). We reported a novel Met‐β‐catenin complex at the hepatocyte membrane [56]. HGF induced tyrosine phosphorylation‐ dependent nuclear translocation of β‐catenin. In a follow‐up study, we identified tyrosine residues 654 and 670 as targets of HGF‐induced β‐catenin phosphorylation (Figure  46.1c) [57]. Other reports had identified a similar effect of HGF on positively regulating β‐catenin/TCF transactivation, albeit via other mechanisms [58–60]. These observations are relevant as high levels of HGF have been observed in patients with liver pathologies that might be influencing β‐catenin redistribution and altering the disease course [61, 62]. Similarly, direct interactions of β‐catenin with EGFR have been reported as well (Figure  46.1c) [52]. In fact, ErbB2 has also been shown to be associated with β‐catenin [53]. Although the fate of β‐catenin and nuclear signaling in this context is unclear its effect on cell–cell adhesion is well defined [63].

β‐CATENIN AND PROTEIN KINASE A (PKA) β‐catenin can also be phosphorylated by cAMP/PKA in vitro and in vivo at Ser552 and Ser675, which leads to increased β‐catenin‐ TCF activation independent of any input from GSK3β [64]. This has also been shown to occur in liver pathophysiology. In evaluating ways to activate the Wnt/β‐catenin signaling to promote liver regeneration after hepatectomy, the effect of triiodothyronine (T3) or another thyroid hormone receptor‐β agonist GC‐1 (Sobetirome) was identified [65, 66]. In fact, we showed that β‐catenin activation after T3/GC‐1 was at least partially dependent on PKA‐mediated phosphorylation on Ser552 and Ser675. Owing to the increased levels of cAMP observed in various genetic diseases associated with the loss of fibrocystin function such as congenital hepatic fibrosis and Caroli disease, there was increased Ser675‐β‐catenin evident in the cholangiocytes [67]. This contributed to the increased motility of Pkhd1‐deficient cholangiocytes and thus the cAMP/PKA/β‐catenin/TCF axis

46:  β-CATENIN SIGNALING

may be playing an important result in the disease pathogenesis and represent a therapeutic target [67].

β‐CATENIN AND NF‐κB β‐catenin and p65 subunit of NF‐κB were shown to complex in breast cancer [68]. We also discovered this complex in the liver, in hepatocytes [69]. This complex seems to be inhibitory for p65 activation in hepatocytes. In fact, hepatocyte‐specific knockouts of β‐catenin were protected from LPS injury due to increased NF‐κB activation resulting in pro‐survival gene expression. There is also an increased baseline immune cell infiltration in the liver in the absence of β‐catenin and the significance of this is unclear. Recently, we have also shown that while β‐catenin and NF‐κB are both activated after partial hepatectomy, these ensue in distinct cellular compartments with the former occurring in the hepatocytes while the latter observed in the non‐parenchymal cells [70]. Overexpression of β‐catenin has also been shown to decrease NF‐κB reporter activity [71]. Cancer cells harboring β‐catenin mutations showed reduced NF‐κB activity while enhanced activity was seen in cells lacking basal β‐catenin activity. Coimmunoprecipitation studies also confirmed formation of a complex of β‐catenin with both p65 and p50 subunits of the NF‐κB protein. Expression of NF‐κB targets like iNOS and Fas also inversely correlated with β‐catenin levels in hepatocellular cancer (HCC) samples. These findings are highly relevant as majority of HCCs occur in livers with chronic injury and inflammation, which may in part be due to the complex interplay of β‐catenin and NF‐κB at least in a subset of cases.

β‐CATENIN AND FXR More recently, a complex between farnesoid X receptor (FXR) and β‐catenin in liver cells has also been shown. Very similar to its interaction with NF‐κB, β‐catenin also complexes with FXR and suppresses it. This observation was made after β‐catenin conditional knockout mice were shown to be less injured after bile duct ligation as compared to the control littermates [72]. It appears that β‐catenin functions as both an inhibitor of nuclear translocation and a nuclear corepressor through formation of a physical complex with FXR. Loss of β‐catenin expedited FXR nuclear localization and FXR‐retinoic X receptor (RXR)‐α association, culminating in small heterodimer protein (SHP) promoter occupancy and activation in response to bile acids (BA) or FXR agonist. Conversely, excessive β‐catenin sequesters FXR, thus inhibiting its activation. This complex could be playing a role in pathogenesis of cholestatic injury and may also have therapeutic implications in various hepatic diseases.

WNT/β‐CATENIN SIGNALING IN LIVER DEVELOPMENT In 2003, the first study identified the role of Wnt/β‐catenin signaling in liver development [73]. Utilizing in vitro organ

589

cultures and a comprehensive ontogenic analysis, the role of β‐catenin was demonstrated in early liver development [73, 74]. Several groups have now complemented and furthered the understanding of how extremely tight regulation of Wnt/β‐ catenin signaling is essential at multiple steps during hepatic development [75]. Wnt signaling promotes the posterior endodermal fate and suppresses the anterior endodermal fate during gastrulation and early somitogenesis [76]. At the same time suppression of Wnt signaling by secreted frizzled‐related protein 5 (sFRP5) has been shown to maintain foregut fate in the anterior endoderm and allows for liver development to initiate [76, 77]. However, once anteroposterior endoderm patterning is established, Wnt signaling now positively regulates liver specification [76]. These highly temporal and opposite roles of Wnt/β‐catenin signaling in liver formation during early development are also observed in zebrafish [78]. In fact, an unbiased forward‐genetic screen in zebrafish led to the identification of wnt2bb mutants that have very small or no liver buds [79]. Wnt2 knockdown in wnt2bb mutants blocked liver recovery and importantly resulted in no hhex expression in the liver‐forming region [80], indicating the essential role of Wnt signaling in liver specification. Both wnt2 [80] and wnt2bb [79] are expressed in the lateral plate mesoderm adjacent to the liver‐forming region. Wnt/β‐catenin signaling is not only necessary but also sufficient for liver specification. Gain‐of‐function studies in zebrafish showed that overexpression of Wnt2bb [80] or Wnt8a [81] in entire tissues induced ectopic hepatoblast and hepatocyte formation in the posterior endoderm that normally gives rise to the intestine. Wnt8a overexpression in non‐hepatic‐destined region of dorsal endoderm in zebrafish was also shown to induce ectopic hepatoblasts [82]. A mouse model in which Wnt/β‐catenin signaling is activated or inactivated in the foregut endoderm after anteroposterior endoderm patterning but prior to liver specification, will be necessary to define the role of Wnt/β‐catenin signaling in liver specification. We have performed foxa3‐cre driven β‐catenin deletion that was evident at E9.5 in mouse hepatoblasts [83]. However, this did not affect the hepatoblast compartment and HNF4α‐positive hepatoblasts were seen unequivocally in these conditional knockout embryos. While this may imply that Wnt/β‐catenin signaling is dispensable for hepatic induction in mice, it is a technical challenge to abolish β‐catenin expression at the right time and at the right place. β‐catenin gene and protein expression peaks at E10–14 in mouse embryonic liver and during this time β‐catenin is localized in the nucleus, cytoplasm, and membrane in different epithelial cells and coincided well with ongoing cell proliferation [74]. Mouse embryonic livers from E9.5–10 stages cultured in the presence of a β‐catenin antisense oligonucleotides, showed decreased proliferation and a simultaneous increase in apoptosis, two processes vital to hepatic morphogenesis that follow hepatic specification and induction [73]. This correlated well with a subsequent study that found overexpression of β‐catenin in developing chicken livers leads to a threefold increase in liver size, which is due at least in part to an expanded hepatoblast population [84]. Role of β‐catenin in bile duct morphogenesis during development is also intriguing. Antisense against β‐catenin in

590

THE LIVER:  ROLE OF WNT/β-CATENIN AND CELL JUNCTIONS IN BLOOD BILE BARRIER

embryonic liver cultures led to decreased bile duct differentiation and the addition of Wnt3a to the embryonic liver cultures induced a biliary phenotype [85]. These observations were supported by in vivo studies that showed suboptimal biliary differentiation in β‐catenin conditional null livers [83]. Conversely, enhanced biliary differentiation of hepatoblasts was observed in APC‐null livers that showed increased β‐catenin during prenatal development, and led to an untimely demise of the embryos [86]. More recent analysis shows β‐catenin to be dispensable for differentiation of cholangiocytes to hepatocytes, although β‐catenin needs to be kept in tight control during development [87]. A cell nonautonomous role of Wnt/β‐catenin signaling in biliary morphogenesis has also been reported whereby hepatocyte‐specific suppression of Wnt/β‐catenin signaling reduced Notch activity in biliary cells by reducing expression of jag1ab and jag2b in hepatocytes [88]. Wnt5a was shown to inhibit bile duct differentiation of hepatoblasts [89]. Wnt5a deletion from mesenchymal cells in mid‐gestational liver led to increased expression of Sox9 and the numbers of hepatocyte nuclear factor (HNF)‐1β‐positive cholangiocyte precursors. In an in vitro differentiation assay this effect was CaMKII‐dependent. The authors did not address the state of β‐catenin signaling in this study. Wnt5a treatment can activate NF‐AT to induce β‐catenin degradation and thus may regulate biliary differentiation partially through such a mechanism [90]. The role of β‐catenin in hepatocyte maturation during development is important. Antisense‐mediated β‐catenin knockdown in embryonic liver cultures led to persistent expression of stem cell markers in hepatocytes [73]. Lack of β‐catenin in hepatoblasts resulted in dramatic decreases in nuclear enriched transcription factors CEBPα and HNF4α, impairing hepatocyte maturation and fetal viability [83]. How is β‐catenin spatio‐temporally regulated during development? Wnt2bb appears as the upstream effector of β‐catenin during the earliest phases of hepatic morphogenesis [79]. Wnt9a expression was reported in endothelial and stellate cells of the embryonic sinusoidal wall in the developing liver [91]. This report also showed presence of Frizzled 4, 7, and 9 on hepatocytes. Wnt5a was shown to be expressed in mesenchymal cells during hepatic development although its effect on β‐catenin was not directly addressed [89]. Based on interaction of Met and β‐ catenin and the role of HGF/Met in liver development, HGF/ Met/β‐catenin signaling may be relevant [56, 92–94]. Expression of FGF‐10 in the mouse liver correlates with peak β‐catenin activation; moreover, release of FGF‐10 from stellate cells stimulates β‐catenin expression in hepatoblasts [95]. FGF‐2, FGF‐4, and FGF‐8 can impact β‐catenin activation in embryonic liver cultures [96].

ROLE OF WNT/β‐CATENIN IN METABOLIC ZONATION Liver occupies a strategic location in the body and receives nutrient‐ and toxin‐enriched blood through portal circulation. Histologically, this enters the hepatic lobule through the portal vein located in the portal triad and mixes with blood from the hepatic artery in sinusoids to eventually travel to the central

vein. While hepatocytes that are organized radially flanking sinusoids along this porto‐central axis look grossly similar, these are destined to perform differing functions. This characteristic within a hepatic lobule is called metabolic zonation, which has been known for several decades [97]. Wnt/β‐catenin signaling has now been shown to be a key regulator of the gene expression in the pericentral or zone three hepatocytes [98]. In fact, β‐catenin activation controls expression of several genes involved in glutamine metabolism and xenobiotic metabolism including glutamine synthetase (GS), ornithine aminotransferase, and the glutamate transporter GLT‐1 [99–101]. The overall basis of pericentral Wnt/β‐catenin signaling is a culmination of several factors. APC, the protein responsible for β‐catenin degradation is expressed at highest levels in zones one and two and absent in zone three. Wnt2 and Wnt9b have now been shown to be basally secreted from endothelial cells lining the central veins [102] to activate β‐catenin in pericentral hepatocytes through as yet unknown Fzd receptor but requiring Wnt co‐receptors LRP5/6 (Figure 46.2). This has been shown in various genetic knockout mice. Mice lacking β‐catenin in hepatocytes [100, 101] or Wnt co‐receptors in hepatocytes [103], or lacking Wntless in endothelial cells lining central veins preventing all Wnt secretion from these cells [102, 104, 105], all lack Wnt/β‐catenin targets in zone three hepatocytes. Similarly, the R‐Spondin/Lgr4/5 axis was shown to optimize Wnt signaling and also contribute to activation of β‐catenin in zone three hepatocytes [106]. Nuclear regulation of β‐catenin‐TCF4 complex during zonation is also relevant to discuss. It was shown that in the presence of β‐catenin, TCF4 is able to bind to Wnt‐responsive elements on target gene promoters, while in its absence TCF4 binds HNF4‐α responsive elements. This switch back and forth of TCF4 may eventually be key to regulation of pericentral versus periportal gene expression but needs to be further ascertained. Several known and new targets of β‐catenin/TCF4 signaling in the liver revealed in the study include Axin2, GS, constitutive androstane receptor, CYP1A2, aryl hydrocarbon receptor, Mrp2, glutathione S‐transferases, and Cyp27A1. Other intriguing targets of β‐catenin reported in the liver include regucalcin or senescent marker protein‐30 and 4‐gulonolactonase, both of which are essential for ascorbic acid biosynthesis in murine liver [107]. Lect2 is also an important target of β‐catenin with relevance in liver tumors and inflammation.

ROLE OF WNT/β‐CATENIN AND CELL JUNCTIONS IN BLOOD BILE BARRIER Liver is a highly vascular organ. Likewise, it also is the largest gland, and among other factors, one of its main functions is the secretion of bile. Blood and bile are kept segregated at a microscopic level owing to the function of tight junctions (TJ) in hepatocytes which allow bile to be secreted apically into biliary canaliculi moving from the pericentral to periportal zone eventually to be carried into bile ductules, while blood traverses the sinusoids at the basolateral surface of the hepatocytes. Disruption of this barrier has been reported in a subset of cases with progressive familial intrahepatic cholestasis (PFIC) due to

46:  β-CATENIN SIGNALING

591

Figure 46.2  Cell–molecule circuitry of the β‐catenin activation in a baseline liver and after partial hepatectomy. At baseline, endothelial cells lining the central vein constitutively release Wnt2 and Wnt9b, which act in a paracrine manner on proximal hepatocytes to activate β‐catenin‐TCF4 dependent expression of genes such as glutamine synthetase (GS), cytochrome P450 2e1 (Cyp2e1), and Cyp1a2. After partial hepatectomy, the increased shear stress caused by sinusoidal blood flow on sinusoidal endothelial cells (red double sided arrows) likely triggers release of Wnt2 and Wnt9b from these cells (1) to act on Fzd‐LRP5/6 receptors on hepatocytes in proximity (2) to induce β‐catenin stabilization, nuclear translocation, and binding to TCF4 (3). This leads to increased transcription of Ccnd1 as early at 12 hours (12h) after PH (4) and can last up to 72h. The protein increase of cyclin‐D1 is observed at 24 h which in turn starts to promote G1 to S phase transition of hepatocytes which peaks at 40h (5). Around the same time, hepatocytes begin to express increased levels of Wnt5a (6) which is then released and through autocrine signaling is able to inactivate β‐catenin signaling between 40–72h and thus contribute to termination of this mitogenic signal to eventually help in cessation of liver regeneration when normal liver size is realized after partial hepatectomy.

loss‐of‐function mutation in the gene encoding for the TJ protein‐2 or ZO‐2 [108]. The role of β‐catenin as part of the AJ or of AJ in general in the barrier function within the liver is not well understood. Liver‐specific β‐catenin conditional knockout mice also showed increase in γ‐catenin protein levels, which associated with E‐cadherin in lieu of β‐catenin to maintain cell–cell junctions [109]. A follow‐up study showed that while not optimal, γ‐catenin compensation at AJ via its interaction with E‐cadherin did prevent E‐cadherin degradation from ubiquitin proteasome, but E‐cadherin levels were still overall decreased in β‐catenin knockout livers despite γ‐catenin upregulation [110]. Further, increased serine/threonine phosphorylation of γ‐catenin after β‐catenin knockdown was identified and at least partially PKA‐ dependent [110]. Despite γ‐catenin increase, all pericentral Wnt/β‐catenin targets in the liver such as GS, cyp2e1, and cyp1a2 were absent in the β‐catenin knockouts, showing lack of compensation of γ‐catenin in the Wnt pathway [109, 110]. To conclusively address the functional relevance of γ‐catenin at the AJ in the absence of β‐catenin, β‐ and γ‐catenin were dually deleted in both hepatocytes and cholangiocytes using Albumin‐Cre. This led to a dramatic phenotype consisting of intrahepatic cholestasis, cholemia, biliary fibrosis, failure to thrive, and mortality. While 80% of double knockouts died in

less than 30 days after birth, the survivors progressed to severe fibrosis and HCC and succumbed by two and a half months of age. The loss of the two catenins led to two major types of molecular perturbations. Due to loss of β‐catenin and its role within the Wnt signaling pathway, there was reduced expression of TJ proteins such as claudin‐2 which is a known target of the Wnt pathway [111]. However, such loss was also evident in β‐catenin single knockout which did not have this phenotype of the dual knockouts. Loss of γ‐catenin on top of β‐catenin loss, we posit, caused a second hit deregulating junctional integrity and decompensation. β‐catenin is known to interact with E‐ cadherin and through this association, it masks the PEST domain in E‐cadherin, thus preventing its degradation [42]. β‐catenin loss in the liver was compensated by increased γ‐catenin which bound to E‐cadherin and hence prevented its degradation although not as effectively as β‐catenin as some decrease in E‐cadherin was evident upon β‐catenin loss despite increased γ‐catenin [109, 110]. However, this prevented a major decompensation and mice were able to survive and showed only minimal increases in hepatic bile acids, serum bilirubin, and reduction in bile flow due especially with age [72, 112]. Dual loss of catenins in the liver, led to a notable E‐cadherin loss and in addition also destabilized occludin, another TJ protein, leading to its loss as well [113]. Thus, dual loss of catenins, led to

592

THE LIVER:  ROLE IN LIVER REGENERATION AFTER PARTIAL HEPATECTOMY AND TOXICANT‐INDUCED INJURY

loss of claudin‐2 and occludin, disrupting TJ and leading to disruption of the blood bile barrier and a phenotype reminiscent of PFIC. While dual loss of catenins has not been reported thus far in PFIC cases, there remains a subset of these cases whose pathogenesis remains unknown [114]. Further, alterations in catenins may be a more global phenomenon contributing to cholestatic injury and cholemia in various other hepatic pathologies and more in depth study remains to be done.

ROLE IN POSTNATAL LIVER GROWTH The liver continues to grow during neonatal stages. In fact, in mice an early postnatal hepatic growth spurt occurs during the first month after birth. Wnt/β‐catenin signaling was identified to be active during these stages and correlated well with ongoing hepatocyte proliferation through regulation of cyclin‐D1 expression [115]. Several additional models have been employed that demonstrate a positive role of β‐catenin in liver growth. A transgenic mouse overexpressing a stable mutant of β‐catenin generated under the transcriptional control of calbindin‐D9K (CaBP9K) promoter and liver‐specific enhancer of the aldolase B gene displayed three to four times larger livers due to increased cell proliferation [116]. Interestingly they did not detect any changes in any of the conventional target genes of the pathway such as c‐myc and cyclin‐D1. Subsequent analysis of transgenic livers and subtractive hybridization led to identification genes involved in glutamine metabolism as targets of the Wnt/β‐ catenin pathway. However, no signs of neoplastic transformation were reported in these animals, although APC‐conditional null mice that exhibits β‐catenin stabilization leads to development of robust HCCs [117]. Other transgenic mice have shown similar hepatic growth advantage [118, 119] and also enabled identification of new targets such as the epidermal growth factor receptor [119]. Transgenic mice expressing Ser45‐mutant β‐ catenin under an albumin promoter also showed an initial increase in liver size but then adapted via enhanced association of β‐catenin to E‐cadherin at the membrane [120]. It was only after stimulation of liver growth via hepatectomy or chemical carcinogenesis that a notable growth advantage of active mutant of β‐catenin was visible as compared to controls. Two independent groups including ours have reported conditional hepatic loss of β‐catenin affecting liver size [100, 101]. The decreased liver weight : body weight ratio was chiefly due to reduced postnatal hepatocyte proliferation that led to decreased hepatic mass which was sustained throughout the animal’s lifespan. The major mechanism of reduced proliferation appears to be lower cyclin‐D1 expression, a known target of β‐catenin in various tissues such as liver and colon [101, 121].

ROLE IN LIVER REGENERATION AFTER PARTIAL HEPATECTOMY AND TOXICANT‐INDUCED INJURY The Wnt/β‐catenin pathway has been comprehensively studied in liver regeneration (LR) that ensues after partial hepatectomy

(PH). The earliest study done in a rat model of PH showed transient stabilization of β‐catenin protein due to post‐translational modification followed by its nuclear translocation within minutes after surgery [122]. Interestingly, the overall increase was transient due to activation of β‐catenin degradation, β‐catenin persisted in the nuclei of the hepatocytes until around 24 hours. Knockdown of β‐catenin at the time of PH in rats using phospho morpholino antisense oligonucleotides led to decreased hepatocyte proliferation and recovery of liver mass [123]. In mice, β‐catenin nuclear translocation has been observed as early as 3–6 hours after PH [70]. Liver‐specific β‐catenin knockout mice when subjected to PH showed notable delay in peak hepatocyte proliferation, which occurred at 72 hours instead of 40 hours [101, 124]. This occurred chiefly due to decreased cyclin‐D1 along with other cyclins like A and E, thereby reducing the number of hepatocytes in S‐phase in the knockout by at least 50% at 40 hours after PH. While redundancy among signaling pathways ensuring successful LR after PH has been appreciated, what drives it in the absence of β‐catenin remains unknown [125]. To address what regulated β‐catenin activation during LR, we utilized liver‐specific knockouts of Wnt co‐receptors LRP5/6. Dual loss of these redundant receptors from hepatocytes led to a delay in LR after PH very similar to β‐catenin knockouts [103]. This was associated with lack of β‐catenin activation and decreased cyclin‐D1 expression in hepatocytes and hence a decrease in the number of hepatocytes in S‐phase at 40 hours after PH. LR improved by 72–96 hours in these animals. Thus, β‐catenin is under the control of Wnt signaling during LR. To determine which cells in the liver may be the source of Wnts during LR, we conditionally knocked out Wntless from hepatocytes and cholangiocytes, macrophages, and endothelial cells. Only loss of Wnt secretion capability through deletion of Wntless from endothelial cells phenocopied β‐catenin knockouts and LRP5/6‐double knockouts [105]. Deletion from hepatocytes and cholangiocytes had no effect on peak hepatocyte proliferation whereas deletion from macrophages had a modest decrease suggesting contribution from these cells as secondary source of Wnts during LR [103, 105]. Intriguingly, when endothelial cells were isolated from regenerating livers at 12 hours after PH, the same two Wnts (Wnt2 and Wnt9b) that are essential for pericentral β‐catenin activation at baseline, were many‐fold upregulated in the endothelial cells as well [105]. The most upstream effectors that are responsible for basal Wnt2 and Wnt9b expression in central vein endothelial cells versus induced Wnt2 and Wnt9b expression in sinusoidal endothelial cells after PH remain unknown. We have recently speculated that increased shear stress on sinusoidal endothelial cells immediately after PH may be upstream of increased expression and release of Wnt2 and Wnt9b. Indeed, when primary or immortalized sinusoidal endothelial cells are subjected to orbital shear stress, there was increased expression of several Wnts including Wnt2 and Wnt9b [105]. Eventually, prolonged hepatocyte proliferation was observed after PH when Wntless was deleted from hepatocytes. This led to discovery of Wnt5a as a termination signal for the Wnt/ β‐catenin pathway which is secreted by hepatocytes when β‐catenin activation is no longer needed [126]. Thus, the Wnt2/9b‐LRP5/6‐β‐catenin axis is the key regulator of both the metabolic zonation to control expression of

46:  β-CATENIN SIGNALING

genes in zone three as well as in regulating cyclin‐D1 expression and driving LR after PH (Figure  46.2). Additionally, Wnt5a appears to terminate β‐catenin activation following achievement of appropriate hepatocyte proliferation and recovery of hepatic mass. β‐catenin’s signaling in LR has also been shown in zebrafish. Transgenic zebrafish lines that express dominant‐negative TCF in the hepatocytes showed a blunted regenerative response to hepatectomy due to decreased cell proliferation [78]. Similarly, zebrafish expressing APC mutant or overexpressing Wnt8a showed enhanced LR due to increased β‐catenin [78]. In patients, we have shown β‐catenin activation to positively correlate with hepatocyte proliferation albeit in patients with acetaminophen overdose [127]. First, β‐catenin was shown to be relevant in toxicant‐induced liver injury and repair. Sublethal dose of acetaminophen causes centrilobular necrosis that is followed by spontaneous hepatocyte proliferation in adjacent zone that begins the repair process. Following administration of a single intra‐peritoneal dose of 500 mg kg−1 of acetaminophen to CD1 male mice, ALT levels increased from 3–12 hours and returned to normal by 24 hours [127]. PCNA‐positive hepatocytes were visible from 3–12 hours as well and decreased by 24 hours at which time liver appeared almost normal. β‐catenin stabilization and activation was evident at 1–6 hours after acetaminophen injury and cyclin‐D1 increased from at 3–12 hours. Liver‐specific β‐catenin knockouts lack cyp2e1 and cyp1a2, and hence are protected from acetaminophen overdose since they can’t metabolize acetaminophen to generate the reactive metabolite [100, 101]. When expression of these two P450’s was induced in the β‐catenin knockout mice, it enabled partial metabolism of acetaminophen and in turn led to a mild injury. When hepatocyte proliferation was compared between controls and β‐catenin knockouts at equitoxic doses, there was a deficit in hepatocyte proliferation in the knockouts supporting the role of β‐catenin in toxicant‐induced LR as well [127]. In a retrospective study in patients, β‐catenin nuclear/cytoplasmic localization correlated with high PCNA and spontaneous LR, while predominantly membranous β‐catenin without any nuclear/cytoplasmic redistribution correlated with decreased proliferation index and need for liver transplantation [127]. Thus β‐catenin is also relevant in regulating hepatocyte proliferation in humans. To address if β‐catenin activation could promote LR, transgenic mice expressing a stable S45‐mutant‐β‐catenin were subjected to PH. These animals showed a more pronounced regenerative response with a shift to the left in regeneration kinetics [120]. Likewise, when naked Wnt‐1 DNA was injected weekly followed by PH, there was increased β‐catenin activation and premature hepatocyte proliferation [120]. More recently we have identified an important role of triiodothyronine (T3) or small molecule T3 agonist called GC‐1 (Sobetirome) in inducing β‐catenin activation and in turn cyclin‐D1 expression. Both these agents showed a positive effect on hepatocyte proliferation and promoting LR after PH [65, 66]. Interestingly, these agents did not promote pro‐tumorigenic activity of oncogenic β‐catenin in cancer models, but only promoted activation of hepatocyte proliferation in normal hepatocytes through β‐catenin activation [128]. Thus use of these agents is a highly relevant means to stimulate β‐catenin and could have potential

593

clinical utility in various scenarios [129]. This would be relevant in transplantation settings such as to induce regeneration in living donors, small‐for‐size syndrome, or generally after liver transplantation. It may also be useful in promoting LR after toxicant‐induced liver injury.

ROLE OF WNT/β‐CATENIN SIGNALING IN HEPATIC BILE ACID HOMEOSTASIS AND INTRAHEPATIC CHOLESTASIS Hepatocytes are responsible for the conversion of cholesterol into bile acids. Bile acids, which are detergents and toxic to hepatocytes, are secreted and eliminated into bile canaliculi for eventual transport to the gut lumen to assist in the digestion of dietary lipids and cholesterol. Two of the key enzymes responsible for the bile acid biosynthesis, Cyp7a1 and Cyp27, are mainly expressed in the pericentral hepatocytes, suggesting they may be regulated by Wnt/β‐catenin signaling [130]. Intrahepatic cholestasis encompasses defects in bile flow in the liver within the intrahepatic bile ducts and biliary canaliculi. The injury is likely due to the detergent action of the retained bile acids within the liver that result in local injury to hepatocytes, ensuing ductular reaction and associated inflammation and fibrosis. It has also been reported that HCCs with CTNNB1 mutation and β‐catenin activation often display intratumoral cholestasis [131]. Liver‐specific β‐catenin knockout mice showed increased susceptibility to steatohepatitis with the methionine and choline deficient diet [132]. These mice also showed increased basal ­levels of hepatic bile acids. This is at least in part thought to be due to a defect in biliary canaliculi as they showed dilatation, tortuosity, and loss of canalicular microvilli, likely due to absence of β‐catenin targets claudin‐2 and regucalcin [112]. Initially it was thought that these increased bile acids may be responsible for decreased expression of Cyp7a1 and Cyp27, which are involved in bile acid synthesis from cholesterol, in the β‐catenin knockouts as feedback mechanism [112, 132]. However, chromatin immunoprecipitation has shown these two genes to also be direct targets of the Wnt pathway and thus playing a role in bile acid metabolism [133]. An intriguing finding was made when liver‐specific β‐catenin knockout mice were subjected to bile duct ligation [72]. There was a notable decrease in hepatic injury observed which was a result of failure of elevation of bile acid levels in the knockouts after the surgery although these levels were notably upregulated in control mice after the same procedure. It was shown that there was overall decrease in hepatic bile acid biosynthesis due to decreased expression of Cyp7a1 and Cyp27 which was in part due to excessive FXR activation. In fact, we showed that β‐catenin functioned as both an inhibitor of nuclear translocation and a nuclear corepressor through formation of a physical complex with FXR. Loss of β‐catenin promoted FXR nuclear localization and FXR/RXRα association, resulting in SHP promoter activation in response bile acid or FXR agonists. Thus β‐catenin is an important regulator of bile acid metabolism and in specific cases, therapeutic inhibition of β‐catenin especially with an FXR agonist may be beneficial specially to decrease overall intrahepatic bile acid burden and related injury.

594

THE LIVER:  ROLE OF WNT/β-CATENIN SIGNALING IN HEPATOBLASTOMAS

ROLE IN HEPATIC FIBROSIS Chronic injury to the liver results in hepatic fibrosis, a wound healing response, which can lead to cirrhosis if the insult continues. Due to lack of effective treatment, other than alleviating the injury, hepatic fibrosis and cirrhosis remain a major unmet clinical need. Hepatic fibrosis is mostly a function of hepatic stellate cells (HSCs) that are the source of extracellular matrix deposition within injured livers [134]. Wnt signaling is beginning to be assessed in hepatic stellate cell biology. The first studies that showed increased expression of the WNT pathway components in hepatic fibrosis were using genomic analysis from primary biliary cholangitis livers [135, 136]. These studies identified increased expression of Wnt13, Wnt5a, β‐catenin, and others, although a cause and effect relationship was not addressed. A study directly investigating gene expression differences in quiescent versus activated HSCs identified, among others, an upregulated expression of Wnt5a and Fz2 [137]. Subsequent studies have further identified aberrant Wnt5a expression in fibrotic livers, and its suppression led to reduced HSC activation [138, 139]. Wnt5a can have opposing roles on β‐catenin signaling based on receptors expressed in a cell. Whether the role of Wnt5a in promoting fibrosis is due to inhibition, activation or independent of β‐catenin needs further clarification. Unfortunately, there is evidence for all three in the literature at the present time. As mentioned before, one study showed no change in nuclear translocation of β‐catenin following Wnt5a increase that was evident during stellate cell activation concomitant with upregulation of Fzd2 [137]. Another study showed that activation of Wnt/β‐catenin signaling in primary rat HSC by an inhibitor of GSK3β decreased synthesis of α‐smooth muscle actin and Wnt5a, and induced the expression of glial fibrillary acidic protein [140]. This study further showed that activation of Wnt signaling lowered DNA synthesis and prevented HSCs from entering the cell cycle to eventually demonstrate the role of Wnt signaling to β‐catenin in maintaining their quiescence. However, a growing body of literature supports the activation of β‐catenin by WNT signaling during the process of HSC activation and fibrosis [141–143]. In fact, different molecules have been employed to block Wnt signaling, directly or indirectly, to demonstrate an overall anti‐ fibrotic effect both in vitro and in vivo. Activating pregnane X receptor by rifampicin, among other things, also inhibited Wnt signaling and reduced HSC proliferation and transdifferentiation to active myofibroblasts [144]. Another study shows that necdin, a melanoma antigen family protein that promotes neuronal and myogenic differentiation while inhibiting adipogenesis, which expressed in HSCs. Necdin is induced during HSC activation and its silencing reversed them to quiescence through PPARγ and suppression of Wnt/β‐catenin signaling [145]. Another study showed that SEPT4, a subunit of the septin cytoskeleton specifically expressed in quiescent HSCs, is downregulated through transdifferentiation to activated myofibroblasts. Loss of SEPT4 in HSCs coincided with decreased expression of Wnt inhibitor DKK2, increased Wnt signaling via β‐catenin, and increased fibrosis [146]. Another study supported that concept that the canonical Wnt pathway promotes fibrogenesis, based on analysis of mesoderm‐specific transcript homologue, a strong

negative regulator of Wnt/β‐catenin signaling [147]. The authors showed that mesoderm‐specific transcript homologue expression in HSCs alleviated carbon tetrachloride‐induced collagen deposition in liver tissue by decreasing the expression of β‐catenin, α‐smooth muscle actin, and Smad3 both in vivo and in vitro. Wnt‐β‐catenin signaling might lead to myofibroblastic activation of HSCs via negative regulation of adipogenesis [148]. Expression of Wnt11 or β‐catenin with the S33Y mutation in preadipocytes prevented their differentiation to adipocytes at least in part through inhibition of expression of CCAAT enhancer binding protein‐α (CEBPα) and PPARγ. Conversely, blockade of β‐catenin signaling either through dominant‐negative TCF4 or axin, facilitated adipogenic differentiation of the preadipocytes. Quiescent HSCs contain lipid droplets and their activation to myofibroblasts involves loss of adipogenic genes including PPARG and CEBPA, so this event is analogous to differentiation of adipocytes to preadipocytes. Gain‐of‐function mutations in adipogenic transcription factors such as PPARG and sterol regulatory element binding protein‐1c can reverse culture‐induced myofibroblast into a quiescent HSC phenotype. Yet another study showed HSC activation to myofibroblasts by the Wnt pathway to be mediated by miR‐132 and miR‐212 which led to epigenetic repression of methyl‐CpG binding protein 2 (MeCP2) leading to loss of PPARγ [149]. More recently, in HSCs expression of stearoyl‐CoA desaturase expression was shown to be regulated by the Wnt‐β‐catenin signaling. This enzyme generates monounsaturated fatty acids, which were shown to provide a feed forward loop to augment Wnt signaling through the stabilization of Lrp5 and Lrp6 mRNAs, which encode for Wnt co‐receptors. This was shown to eventually contribute to liver fibrosis through HSC activation [150]. Taken together these observations indicate that activation of Wnt signaling via β‐catenin could be involved in HSC activation by inhibiting the adipogenic programs; inhibition of this signaling pathway could contribute to adipogenic gene profile of a quiescent HSC. Inhibiting Wnt signaling to β‐catenin might therefore block hepatic fibrosis.

ROLE OF WNT/β‐CATENIN SIGNALING IN HEPATOBLASTOMAS Hepatoblastoma (HB) is the commonest malignant hepatic tumor of childhood. These tumors are frequently sporadic; however, the incidence is highest in patients of familial adenomatous polyposis coli [151]. This led to the identification of APC mutations in HB in familial cases [152]. Increased frequency of diverse APC mutations (57%) were reported in sporadic form of the disease as well [153]. N‐terminal mutations (missense and deletions) affecting exon 3 of CTNNB1 have also been found in up to 80–90% of all HBs and are associated with nuclear and cytoplasmic β‐catenin localization [154]. Mutations in AXIN1 have been identified in less than 10% of these tumors [155]. A recent comprehensive study in 85 HB patients showed 65 cases with missense mutations and interstitial deletions in CTNNB1 [156]. They also identified loss‐of‐function mutations in APC and AXIN1 genes. Thus 82% of HB in this cohort exhibited

46:  β-CATENIN SIGNALING

Wnt/β‐catenin activation. Thus, there is compelling data that β‐ catenin activation is an obligatory event in the etiopathogenesis of HB, although the exact mechanism of how it leads to HB is unclear [157]. Activation of Wnt signaling during normal liver development has already been discussed in this chapter and plays role in both hepatoblast proliferation and hepatocyte differentiation and maturation (also reviewed in [75, 158, 159]). While β‐catenin activation in liver development is spatio‐temporally regulated and ligand‐dependent, in HB its activation is unrestricted, sustained, and non‐ligand‐dependent due to exon‐3 mutations. How this contributes to tumors remains to be addressed although various target genes of Wnt signaling such as c‐Myc, cyclin‐D1, GS, EGFR/Axin‐2, and others have been reported in various histological subtypes of HB [160]. Very similar to β‐catenin in different stages of normal liver development, a more nuclear β‐catenin is evident in embryonal HB coinciding with lack of GS, and more membranous, cytoplasmic and nuclear with GS expression in fetal HB [161]. Intriguingly, when an N‐terminal deletion mutant of β‐catenin is overexpressed in the liver using an adenoviral approach or by generation of a transgenic mouse overexpressing β‐catenin under liver‐specific promoter, the mice never display HB [116, 118]. Also, transgenic mice expressing point‐mutant form of β‐catenin do not show HB [120]. This suggests that oncogenic β‐catenin is insufficient in inducing HB. How β‐catenin activation leads to HB was unclear until recently. A recent study demonstrates β‐catenin mutations frequently co‐occur with activation of Yes associated protein‐1 (Yap), a component of the Hippo signaling pathway. In fact, 80% of all HB cases, irrespective of histological subtype, showed nuclear β‐catenin due to mutations and nuclear Yap due to unknown mechanism [162]. Co‐expression of deletion mutant of β‐catenin and constitutively active Yap by sleeping beauty transposon/transposase and hydrodynamic tail vein injection (SB‐ HTVI), led to development of HB in mice. The role of Yap‐TEAD and β‐catenin‐TCF appears to be key in regulating target genes that may be leading to HB development and further studies will be critical in addressing the mechanism. C‐Myc was shown to be dispensable for HB development in this model although it played role in optimum HB development by regulating tumor metabolism [163]. Likewise, lipocalin‐2, identified as one of the genes that was many‐fold upregulated in Yap‐β‐ catenin HB in mice and also contained TEAD and TCF binding sites in its promoter, was also shown to be dispensable for HB development [164]. However, serum lipocalin‐2 was identified as a sensitive marker of HB tumor burden in the mouse model and may be of relevance clinically as well. Future studies would be invaluable in understanding the role of these pathways in tumor biology.

ROLE OF WNT/β‐CATENIN SIGNALING IN HEPATOCELLULAR ADENOMA Hepatocellular adenomas (HCAs) are characterized by monoclonal proliferation of well‐differentiated hepatocytes that are usually arranged in sheets and cords. There is a classical absence

595

of a portal triads and interlobular bile ducts within HCAs. While initially considered as a well‐defined homogeneous entity, it is now known that several molecular classes of this tumor‐type exist which dictates tumor origin, behavior, and eventually determines prognosis and helps determine treatment options (reviewed in [165]). Around 10–15% display Wnt/β‐catenin activation secondary to mutations in the CTNNB1 gene [166]. β‐catenin‐mutated HCAs are exclusive from the HNF1A mutation group but may occur in combination with gp130 or GNAS mutations. Around half of the β‐catenin‐active HCAs are inflammatory. GS is upregulated in β‐catenin‐mutated HCA. Since detection of nuclear β‐catenin by immunostaining is often challenging due to technical issues as well as heterogeneity in staining patterns, it may not be conclusive. GS staining can aid in the diagnosis of β‐catenin‐mutated HCA with greater sensitivity [167]. β‐catenin‐mutated HCA can occur in males. CTNNB1‐ mutated HCA histologically exhibit cholestasis and cellular dysplasia. Most importantly, these subsets of HCA have greater propensity for malignant transformation [166]. Another subset of HCA has activation of the JAK/STAT pathway and show polymorphic inflammatory infiltrates. Mutations in IL6ST (coding for gp130), STAT3, and GNAS, can activate JAK/STAT signaling. A subset of this tumor‐type also exhibits CTNNB1 mutation, irrespective of the molecular driver, and poses an increased risk of malignant transformation as well.

ROLE OF WNT/β‐CATENIN SIGNALING IN HEPATOCELLULAR CANCER Wnt/β‐catenin activation has been implicated in a major subset of HCC cases. Abnormal localization of cadherins and catenins in liver cancer was first shown by immunohistochemistry [168]. A more comprehensive study identified anomalous β‐catenin expression as well as mutations in the CTNNB1 in around 25% of all HCC cases and up to 50% of all hepatic tumors in transgenic lines such as c‐myc or H‐ras [169]. Several subsequent human studies corroborated these observations and currently around 8–44% of HCCs show mutations in β‐catenin gene, mostly in exon‐3 although recent studies have also identified a region in exon 7/8 of CTNNB1 to be a hot spot [170, 171] (Table  46.2). In fact, CTNNB1 mutations have a trunk role in HCC‐like mutations affecting TERT and p53 [172]. Mutations have also been reported in other components of the degradation complex of β‐catenin including AXIN1 in around 3–16% [155, 173, 174] and AXIN2 in around 3% of all HCC cases  [155]. Additional mechanisms have also been described and include overexpression of FRZ7 [175, 176], Wnt3 upregulation [177], inactivation of GSK3β [178], methylation of soluble frizzled related protein1 (sFRP1) [179], epigenetic inactivation of several sFRPs [180], and TGFβ‐dependent activation of β‐catenin [181]. In HCC patients, exome sequencing has also revealed cooperating mutations of CTNNB1 with mutations in ARID2, NFE2L2, TERT, APOB, and MLL2. Likewise, certain mutations were also always mutually exclusive from CTNNB1 and demonstrated lack of cooperation in development of HCC. These included TP53 and AXIN1.

596

THE LIVER:  ROLE OF WNT/β-CATENIN SIGNALING IN HEPATOBLASTOMAS

Table 46.2  List of studies showing spectra of mutations in Ctnnb1 gene

Study [171]

Cases with mutations in exon‐3 of CTNNB1 (%) Additional information

[170]

36/194 (18.5%) (TCGA) and 90/249 (36%) (French cohort) 90/249 (36%)

[225]

41/125 (33%)

[182]

9/32 (28%)

[174]

20/45 (44%)

[226] [220] [155] [227] [228]

15/45 (33%) 16/38 (42%) 14/73 (19%) 5/62 (8%) 7/60 (12%)

[229]

57/434 (13%)

[230] [186] [231] [32]

9/22 (41%) 12/35 (34%) 21/119 (18%) 9/38 (24%)

[169]

8/31 (26%)

Additional 15/194 cases had mutations in exon 7/8 of CTNNB1 Additional 3/249 cases had mutations in exon 7/8 of CTNNB1 Overall, 52% HCC cases showed activation of Wnt/β‐catenin including mutations in AXIN1 and exon 7/8 of CTNNB1 Additional 15.2% showed mutations in AXIN1 Tyrosine‐654 phosphorylated β‐catenin was observed in fibrolamellar variety of HCC Additional seven patients had AXIN1 mutations No GSK3β mutations Multiple mutations in two patients One insertion between S33 and G34 Aflatoxin study 62% of tumors had cytoplasmic β‐catenin staining 34 had mutations at GSK3β sites. 17 showed mutations at codons 32 and 34 Multiple mutations in one patient Multiple mutations in two patients Multiple mutations in a patient Aberrant accumulation of β‐catenin in nucleus, cytoplasm, and membrane was seen in 39% cases Two patients had mutations at D32

Deletions usually involved one of the key sites: S33, S37, S45, or T41.

It is presently unclear if the extent of Wnt activation due to these diverse mechanisms is comparable. At the same time, it is unknown if downstream signaling in the form of target gene expression is also varied in these disparate mechanisms of β‐catenin activation. A study showed significant correlation between CTNNB1 mutations and overexpression of target genes GS, G‐protein‐coupled receptor (GPR)49, and glutamate transporter (GLT)‐1 (P = 0.0001), but not for other target genes like ornithine aminotransferase, LECT2, c‐myc, or cyclin D1 [174]. This study showed GS to be a good immunohistochemical marker of β‐catenin activation in HCC, which was also confirmed independently [182]. However, no increase in expression of GS, GPR49, or GLT‐1 was evident in loss of Axin‐1 function due to AXIN1 mutations. Thus, it is likely that functional equivalence of various modes of β‐catenin activation is distinct and may have differing consequences in tumor phenotype. Indeed β‐catenin‐active HCC due to mutations in CTNNB1, AXIN1, or additional modes of β‐catenin activation, have all been shown to have distinct phenotype in the transcriptomic classification of HCC [174, 181, 183]. Furthermore, recent studies have even shown disparity in extent of Wnt signaling based on differences in mutations within exon‐3 of CTNNB1 [184]. The effect of β‐catenin mutations on prognosis remains debatable. β‐catenin mutations have been associated with better prognosis and a more differentiated tumor type in some studies [183, 185]. Others have noted high nuclear and cytoplasmic β‐catenin in more proliferating and poorly differentiated HCC [181, 186, 187]. Eventually, more careful meta‐analysis taking

into account geographical factors, etiology, co‐morbidities, and kind of mutations may be essential to clarify this discrepancy. Decreased fibrosis was reported in a few studies and remains an intriguing hallmark of β‐catenin mutated tumors [182, 188]. Are β‐catenin mutations a risk factor for development of HCC independent of cirrhosis or are they merely decreasing the threshold of neoplastic transformation? It is also intriguing to note that in a recent study, a small but significant subset of HCV patients developed HCC without evidence of advanced fibrosis [189]. Along the same lines, it is relevant to note that the small subset of hepatic adenomas that progress to HCC in patients often exhibit β‐catenin gene mutations [166]. This neoplastic transformation of adenomas occurs in a healthy liver without any evidence of fibrosis and further supports the role of β‐catenin in fibrosis‐independent HCC. To address the relationship of advanced fibrosis, β‐catenin mutations and HCC, we used an experimental approach in which ser‐45‐mutant β‐catenin transgenic mice and control mice were fed thioacetamide diet for a prolonged period [190]. No difference in HCC in the two groups of mice suggests that β‐catenin mutations do not enhance evolution of cirrhosis to HCC supporting β‐catenin gene mutations and cirrhosis to be independent contributors to tumorigenesis that do not cooperate. The Wnt/β‐catenin pathway in HCC in experimental ­models  deserves mention. The first question that has been answered is if β‐catenin mutation, activation, or overexpression by itself is sufficient for initiation of HCC. None of the transgenic mice overexpressing either wild‐type or stable‐mutants of β‐catenin thus far have exhibited spontaneous HCC [116, 118–120]. However, several studies now suggest that β‐catenin ­collaborates with other signaling pathways to contribute to hepatocarcinogenesis. β‐catenin was shown to cooperate with activated Ha‐ras in HCC [190]. Mice heterozygous for Lkb1 deletion showed an accelerated progression to HCC when mated with adenovirus‐inducible β‐catenin mutant mice [192]. Similarly, when chemical carcinogen diethylnitrosamine (DEN), which normally induces HCC through Ha‐ras activation [193], was injected in serine‐45‐mutant β‐catenin transgenic mouse, these animals developed HCC earlier and more  profoundly [120]. These findings strongly suggest that β‐catenin mutation is one of the hits that may be critical to the development of HCC, however, additional aberrations are necessary for tumorigenesis. This is also in agreement with clinical studies where CTNNB1 mutations significantly coexisted with mutations in ARID2, NFE2L2, TERT, APOB, and MLL2 [170, 194]. Since concomitant Met overexpression/activation and mutations in CTNNB1 were found in around 11% of all HCC cases, these two proto‐oncogenes were co‐expressed using SB‐ HTVI. These mice developed HCC with gene expression profiles that displayed high correlation with the gene profiles of a subset of human HCC patients with both CTNNB1 mutations and Met activation signatures [171]. To address if Ras activation downstream of Met could be contributing to Met‐β‐catenin HCC, G12D‐KRAS and mutant‐β‐catenin were expressed using SB‐HTVI, which also yielded HCC with around 90% molecular similarity to Met‐β‐catenin HCC [195]. It will be useful to generate relevant models using SB‐HTVI that represent subsets of human HCC which can then be assessed for biological as well as therapeutic studies.

46:  β-CATENIN SIGNALING

Several proof of principle preclinical studies have demonstrated an important benefit of therapeutic inhibition of β‐ catenin for treatment of HCC [196]. Various cox2 inhibitors such as rofecoxib have shown efficacy in decreasing β‐catenin levels along with shrinkage of tumors [197]. R‐Etodolac, an enantiomer of a cox2 inhibitor that lacks an inhibitory effect on cox2 has also shown anti‐β‐catenin effect [198]. Another group of agents including Exisulind and analogues that are inhibitors of cyclic GMP phosphodiesterases (PDE) have been shown to activate protein kinase G (PKG) that in turn decrease β‐catenin levels via a novel GSK3β‐independent processing mechanism [199]. ICG‐001, a small molecule known to inhibit β‐catenin’s interaction with CREB‐binding protein (CBP) was shown to affect β‐catenin‐TCF‐dependent target gene expression [200]. Its new generation analogue PRI‐724 is in clinical trial for various malignancies that exhibit β‐catenin activation [201]. Using computational chemoinformatics to identify another small molecule with structural similarity to ICG‐001, we identified a compound we labeled as PMED‐1 [202]. This inhibitor showed in vitro and in vivo efficacy against β‐catenin in HCC cells and in zebrafish, respectively. Several of these inhibitors may have potential therapeutic utility in the correct subset of HCC patients. While GS as a biomarker of β‐catenin mutations is helpful in identifying that subset, biopsies may not be feasible in majority of HCC patients due to underlying liver disease and cirrhosis. Thus, there is a need for secreted biomarkers to detect β‐catenin mutations to select eligible patients for anti‐β‐catenin therapies since β‐catenin is not a global therapeutic target in HCC [203]. Due to this unmet need, we attempted and identified Lect2 in cell culture and in mice to be a faithful biomarker in identifying activating β‐catenin mutations [204]. However, serum Lect2 did not correspond to β‐catenin mutations in HCC patients although levels greater than 50 ng ml−1 in patients had a positive predictive value of 97% in detecting HCC [204]. Using SB‐HTVI to co‐express mutant β‐catenin and mutant Kras, we showed HCC development in these mice that resembled by gene expression the Met‐β‐catenin HCC, which in turn showed around 70% resemblance by gene expression studies to around 11% of all human HCC [171, 195]. We used these models to examine the effect of Met inhibition on HCC development and showed a lack of any notable response using two different modalities [128, 205]. On the other hand, use of a lipid nanoparticle to deliver siRNA to β‐catenin showed a complete response due to β‐catenin suppression which in turn affected cell proliferation and survival [195]. Previous in vivo studies had also shown β‐catenin suppression albeit using locked nucleic acid antisense in a chemical carcinogenesis induced HCC model, to also lead to complete response [206]. We believe β‐catenin to be a viable therapeutic target in HCC, once patients are carefully selected and verified especially for β‐catenin gene mutations. Despite no FDA approved anti‐β‐catenin agents currently, a  recent study demonstrates the role of mTOR inhibition in β‐catenin‐mutated liver tumors [207]. This study showed localization of p‐mTOR‐S2448, marker of mTORC1 activation, in the same cells as GS in a normal adult liver. This localization of GS and p‐mTOR‐S2448 in zone three hepatocytes was absent in knockouts of hepatocyte β‐catenin or Wnt co‐receptors LRP5/6. Forced re‐expression of β‐catenin or GS led to re‐expression of p‐mTOR‐S2448. More importantly, β‐catenin‐mutated HCC

597

that showed uniform positivity for GS, also showed positivity for p‐mTOR‐S2448 in mice and in patients. Based on these observations, Met‐β‐catenin HCC model showed a dramatic response to mTOR inhibitor rapamycin. Additionally, with the reports showing an efficacy of immune checkpoint inhibitors in small subsets of HCC, a recent study showed exclusion of immune cells from HCC that are mutated for β‐catenin gene, making them unlikely candidates for these agents [208]. Thus assessing β‐catenin mutation ­status to stratify cases for receiving specific agents or excluding them from receiving specific drugs could form the basis of ­personalized medicine in HCC. Since β‐catenin is also present at the cell surface in association with E‐cadherin, a concern remains if β‐catenin suppression would lead to destabilization of the AJ and may inadvertently promote tumor cell migration and metastasis. However, multiple studies have now shown that β‐catenin knockdown is compensated by increased γ‐catenin which maintains AJ through association with E‐cadherin and actin cytoskeleton [109, 110]. Intriguingly, we did not identify any compensation by γ‐catenin in the Wnt signaling upon β‐catenin loss. We also failed to detect nuclear γ‐catenin in β‐catenin knockout mice even after PH and via TopFlash (β‐catenin‐TCF) reporter assay in vitro. Thus, the redundancy in catenins at the AJ but not in the Wnt signaling is reassuring and makes β‐catenin a viable therapeutic target in HCC. It will be critical to further address the mechanisms whereby γ‐catenin is stabilized at AJ following β‐catenin suppression [110].

WNT/β‐CATENIN SIGNALING IN BILE DUCT AND GALL BLADDER TUMORS The most common tumor that arises in the biliary tree is the cholangiocarcinoma that can either originate from the intrahepatic portion, intrahepatic cholangiocarcinoma (ICC), or the hilum (hilar cholangiocarcinoma) (reviewed in [206]). In cholangiocarcinoma, reduced expression of β‐catenin and E‐cadherin at the membrane is observed as compared to the surrounding non‐cancerous ducts [208]. Nuclear localization of β‐catenin is seen in a subset of tumors based on histology and location of the tumor (reviewed in [210]). For most ICCs, aberrant nuclear localization is observed in around 15% of cases and a decrease in membranous localization is related to poorer histological differentiation [211]. No mutations in CTNNB1 were identified although mutations in any other components of the Wnt pathway were not assessed in this study. Our personal experience in 62 ICCs showed only 2/62 tumors with nuclear β‐catenin [162]. Another study detected exon 3 mutations in 7.5% of biliary tract cancer and in 57% of gallbladder adenomas [213]. Higher frequency of mutations was seen in ampullary and gallbladder carcinomas than the bile duct cancers. A higher correlation of CTNBN1 mutations and papillary adenocarcinoma was also observed. Intraductal papillary neoplasms also showed anomalous nuclear localization of β‐catenin in around 25% of patients without any exon‐3 CTNNB1 mutation [214]. Thus, while the Wnt/β‐catenin pathway may be active in a subset of biliary tract neoplasms, more studies are needed to comprehend the mechanism of its observed deregulation.

598

THE LIVER:  REFERENCES

REFERENCES   1. Moon, R.T., Bowerman, B., Boutros, M., and Perrimon, N. The promise and perils of Wnt signaling through beta‐catenin. Science, 2002;296(5573):1644–6.   2. Peifer, M. and Polakis, P. Wnt signaling in oncogenesis and embryogenesis – a look outside the nucleus. Science, 2000;287(5458):1606–9.   3. Arias, A.M. Epithelial mesenchymal interactions in cancer and development. Cell, 2001;105(4):425–31.   4. Perrimon, N. and Mahowald, A.P. Multiple functions of segment polarity genes in Drosophila. Dev Biol, 1987;119(2):587–600.   5. Babu, P. Early developmental subdivisions of the wing disk in Drosophila. Mol Gen Genet, 1977;151(3):289–94.   6. Sharma, R.P. and Chopra, V.L. Effect of the Wingless (wg1) mutation on wing and haltere development in Drosophila melanogaster. Dev Biol, 1976;48(2):461–5.  7. Patel, N.H., Schafer, B., Goodman, C.S., and Holmgren, R. The role of ­segment polarity genes during Drosophila neurogenesis. Genes Dev, 1989; 3(6):890–904.   8. Peifer, M., Rauskolb, C., Williams, M., Riggleman, B., and Wieschaus, E. The segment polarity gene armadillo interacts with the wingless signaling pathway in both embryonic and adult pattern formation. Development, 1991;111(4):1029–43.  9. Riggleman, B., Schedl, P., and Wieschaus, E. Spatial expression of the Drosophila segment polarity gene armadillo is posttranscriptionally regulated by wingless. Cell, 1990;63(3):549–60. 10. Peifer, M., Orsulic, S., Pai, L.M., and Loureiro, J. A model system for cell  adhesion and signal transduction in Drosophila. Dev Suppl, 1993: 163–76. 11. Clevers, H. and Nusse, R. Wnt/beta‐catenin signaling and disease. Cell, 2012;149(6):1192–205. 12. Nusse, R. and Clevers, H. Wnt/beta‐catenin signaling, disease, and emerging therapeutic modalities. Cell, 2017;169(6):985–99. 13. Haegel, H., Larue, L., Ohsugi, M., Fedorov, L., Herrenknecht, K., and Kemler, R. Lack of beta‐catenin affects mouse development at gastrulation. Development, 1995;121(11):3529–37. 14. Hari, L., Brault, V., Kleber, M. et al. Lineage‐specific requirements of beta‐ catenin in neural crest development. J Cell Biol, 2002;159(5):867–80. 15. Brault, V., Moore, R., Kutsch, S. et al. Inactivation of the beta‐catenin gene by Wnt1‐Cre‐mediated deletion results in dramatic brain malformation and failure of craniofacial development. Development, 2001;128(8):1253–64. 16. Lickert, H., Kutsch, S., Kanzler, B., Tamai, Y., Taketo, M.M., and Kemler, R. Formation of multiple hearts in mice following deletion of beta‐catenin in the embryonic endoderm. Dev Cell, 2002;3(2):171–81. 17. Mucenski, M.L., Wert, S.E., Nation, J.M. et al. beta‐Catenin is required for specification of proximal/distal cell fate during lung morphogenesis. J Biol Chem, 2003;278(41):40231–8. 18. Huelsken, J., Vogel, R., Erdmann, B., Cotsarelis, G., and Birchmeier, W. beta‐Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Cell, 2001;105(4):533–45. 19. Kispert, A., Vainio, S.,and McMahon, A.P. Wnt‐4 is a mesenchymal signal for epithelial transformation of metanephric mesenchyme in the developing kidney. Development, 1998;125(21):4225–34. 20. Behrens, J., Jerchow, B.A., Wurtele, M. et al. Functional interaction of an axin homolog, conductin, with beta‐catenin, APC, and GSK3beta. Science, 1998;280(5363):596–9. 21. Amit, S., Hatzubai, A., Birman, Y. et al. Axin‐mediated CKI phosphorylation of beta‐catenin at Ser 45: a molecular switch for the Wnt pathway. Genes Dev, 2002;16(9):1066–76. 22. Aberle, H., Bauer, A., Stappert, J., Kispert, A., and Kemler, R. Beta‐catenin is a target for the ubiquitin‐proteasome pathway. EMBO, J, 1997; 16(13):3797–804. 23. Barrott, J.J., Cash, G.M., Smith, A.P., Barrow, J.R., and Murtaugh, L.C. Deletion of mouse Porcn blocks Wnt ligand secretion and reveals an ectodermal etiology of human focal dermal hypoplasia/Goltz syndrome. Proc Natl Acad Sci USA, 2011;108(31):12752–7. 24. Bartscherer, K., Pelte, N., Ingelfinger, D., and Boutros, M. Secretion of Wnt ligands requires Evi, a conserved transmembrane protein. Cell, 2006;125(3):523–33. 25. Bhanot, P., Brink, M., Samos, C.H. et al. A new member of the frizzled family from Drosophila functions as a Wingless receptor. Nature, 1996; 382(6588):225–30.

26. Pinson, K.I., Brennan, J., Monkley, S., Avery, B.J., and Skarnes, W.C. An LDL‐receptor‐related protein mediates Wnt signalling in mice. Nature, 2000;407(6803):535–8. 27. Tamai, K., Semenov, M., Kato, Y. et  al. LDL‐receptor‐related proteins in Wnt signal transduction. Nature, 2000;407(6803):530–5. 28. Tamai, K., Zeng, X., Liu, C. et al. A mechanism for Wnt coreceptor activation. Mol Cell, 2004;13(1):149–56. 29. MacDonald, B.T., Tamai, K., and He, X. Wnt/beta‐catenin signaling: components, mechanisms, and diseases. Dev Cell, 2009;17(1):9–26. 30. Carmon, K.S., Gong, X., Lin, Q., Thomas, A., and Liu, Q. R‐spondins function as ligands of the orphan receptors LGR4 and LGR5 to regulate Wnt/ beta‐catenin signaling. Proc Natl Acad Sci USA, 2011;108(28):11452–7. 31. Cadigan, K.M. and Nusse, R. Wnt signaling: a common theme in animal development. Genes Dev, 1997;11(24):3286–305. 32. Mikels, A.J. and Nusse, R. Purified Wnt5a protein activates or inhibits beta‐ catenin‐TCF signaling depending on receptor context. PLoS Biol, 2006;4(4):e115. 33. Oishi, I., Suzuki, H., Onishi, N. et al. The receptor tyrosine kinase Ror2 is involved in non‐canonical Wnt5a/JNK signalling pathway. Genes Cells, 2003;8(7):645–54. 34. Pez, F., Lopez, A., Kim, M., Wands, J.R., Caron de Fromentel, C.,and Merle, P. Wnt signaling and hepatocarcinogenesis: molecular targets for the development of innovative anticancer drugs. J Hepatol, 2013;59(5):1107–17. 35. Nishita, M., Enomoto, M., Yamagata, K., and Minami, Y. Cell/tissue‐tropic functions of Wnt5a signaling in normal and cancer cells. Trends Cell Biol, 2010;20(6):346–54. 36. Acebron, S.P., Karaulanov, E., Berger, B.S., Huang, Y.L., and Niehrs, C. Mitotic wnt signaling promotes protein stabilization and regulates cell size. Mol Cell, 2014;54(4):663–74. 37. Koch, S., Acebron, S.P., Herbst, J., Hatiboglu, G., and Niehrs, C. Post‐transcriptional Wnt signaling governs epididymal sperm maturation. Cell, 2015;163(5):1225–36. 38. Vinyoles, M., Del Valle‐Perez, B., Curto, J. et  al. Multivesicular GSK3 sequestration upon Wnt signaling is controlled by p120‐catenin/cadherin interaction with LRP5/6. Mol Cell, 2014;53(3):444–57. 39. Acebron, S.P. and Niehrs, C. Beta‐catenin‐independent roles of Wnt/LRP6 signaling. Trends Cell Biol, 2016;26(12):956–67. 40. Inoki, K., Ouyang, H., Zhu, T. et al. TSC2 integrates Wnt and energy signals via a coordinated phosphorylation by AMPK and GSK3 to regulate cell growth. Cell, 2006;126(5):955–68. 41. Knudsen, K.A., Soler, A.P., Johnson, K.R., and Wheelock, M.J. Interaction of alpha‐actinin with the cadherin/catenin cell‐cell adhesion complex via alpha‐catenin. J Cell Biol, 1995;130(1):67–77. 42. Huber, A.H. and Weis, W.I. The structure of the beta‐catenin/E‐cadherin complex and the molecular basis of diverse ligand recognition by beta‐ catenin. Cell, 2001;105(3):391–402. 43. Jou, T.S., Stewart, D.B., Stappert, J., Nelson, W.J., and Marrs, J.A. Genetic and biochemical dissection of protein linkages in the cadherin‐catenin complex. Proc Natl Acad Sci USA, 1995;92(11):5067–71. 44. Lilien, J., Balsamo, J., Arregui, C., and Xu, G. Turn‐off, drop‐out: functional state switching of cadherins. Dev Dyn, 2002;224(1):18–29. 45. Ozawa, M. and Kemler, R. Altered cell adhesion activity by pervanadate due to the dissociation of alpha‐catenin from the E‐cadherin.catenin complex. J Biol Chem, 1998;273(11):6166–70. 46. Roura, S., Miravet, S., Piedra, J., Garcia de Herreros, A., and Dunach, M. Regulation of E‐cadherin/catenin association by tyrosine phosphorylation. J Biol Chem, 1999;274(51):36734–40. 47. Hu, P., O’Keefe, E.J., and Rubenstein, D.S. Tyrosine phosphorylation of human keratinocyte beta‐catenin and plakoglobin reversibly regulates their binding to E‐cadherin and alpha‐catenin. J Invest Dermatol, 2001;117(5):1059–67. 48. Piedra, J., Martinez, D., Castano, J., Miravet, S., Dunach, M., and de Herreros, A.G. Regulation of beta‐catenin structure and activity by tyrosine phosphorylation. J Biol Chem, 2001;276(23):20436–43. 49. Rosato, R., Veltmaat, J.M., Groffen, J., and Heisterkamp, N. Involvement of the tyrosine kinase fer in cell adhesion. Mol Cell Biol, 1998;18(10):5762–70. 50. Behrens, J., Vakaet, L., Friis, R. et al. Loss of epithelial differentiation and gain of invasiveness correlates with tyrosine phosphorylation of the E‐cadherin/beta‐catenin complex in cells transformed with a temperature‐sensitive v‐SRC gene. J Cell Biol, 1993;120(3):757–66. 51. Shibamoto, S., Hayakawa, M., Takeuchi, K. et al. Tyrosine phosphorylation of beta‐catenin and plakoglobin enhanced by hepatocyte growth factor and

46:  β-CATENIN SIGNALING epidermal growth factor in human carcinoma cells. Cell Adhes Commun, 1994;1(4):295–305. 52. Hoschuetzky, H., Aberle, H., and Kemler, R. Beta‐catenin mediates the interaction of the cadherin‐catenin complex with epidermal growth factor receptor. J Cell Biol, 1994;127(5):1375–80. 53. Kanai, Y., Ochiai, A., Shibata, T. et al. c‐erbB‐2 gene product directly associates with beta‐catenin and plakoglobin. Biochem Biophys Res Commun, 1995;208(3):1067–72. 54. Takahashi, K., Suzuki, K.,and Tsukatani, Y. Induction of tyrosine phosphorylation and association of beta‐catenin with EGF receptor upon tryptic digestion of quiescent cells at confluence. Oncogene, 1997;15(1):71–8. 55. Birchmeier, C., Birchmeier, W., Gherardi, E., and Vande Woude, G.F. Met, metastasis, motility and more. Nat Rev Mol Cell Biol, 2003;4(12):915–25. 56. Monga, S.P., Mars, W.M., Pediaditakis, P. et  al. Hepatocyte growth factor induces Wnt‐independent nuclear translocation of beta‐catenin after Met‐ beta‐catenin dissociation in hepatocytes. Cancer Res, 2002;62(7):2064–71. 57. Zeng, G., Apte, U., Micsenyi, A., Bell, A., and Monga, S.P. Tyrosine residues 654 and 670 in beta‐catenin are crucial in regulation of Met‐beta‐catenin interactions. Exp Cell Res, 2006;312(18):3620–30. 58. Hiscox, S. and Jiang, W.G. Association of the HGF/SF receptor, c‐met, with the cell‐surface adhesion molecule, E‐cadherin, and catenins in human tumor cells. Biochem Biophys Res Commun, 1999;261(2):406–11. 59. Papkoff, J. and Aikawa, M. WNT‐1 and HGF regulate GSK3 beta activity and beta‐catenin signaling in mammary epithelial cells. Biochem Biophys Res Commun, 1998;247(3):851–8. 60. Danilkovitch‐Miagkova, A., Miagkov, A., Skeel, A., Nakaigawa, N., Zbar, B., and Leonard, E.J. Oncogenic mutants of RON and MET receptor tyrosine kinases cause activation of the beta‐catenin pathway. Mol Cell Biol, 2001;21(17):5857–68. 61. Shiota, G., Umeki, K., Okano, J., and Kawasaki, H. Hepatocyte growth factor and acute phase proteins in patients with chronic liver diseases. J Med, 1995;26(5–6):295–308. 62. Yamagami, H., Moriyama, M., Tanaka, N., and Arakawa, Y. Detection of serum and intrahepatic human hepatocyte growth factor in patients with type C liver diseases. Intervirology, 2001;44(1):36–42. 63. Bonvini, P., An, W.G., Rosolen, A. et  al. Geldanamycin abrogates ErbB2 association with proteasome‐resistant beta‐catenin in melanoma cells, increases beta‐catenin‐E‐cadherin association, and decreases beta‐catenin‐ sensitive transcription. Cancer Res, 2001;61(4):1671–7. 64. Taurin, S., Sandbo, N., Qin, Y., Browning, D., and Dulin, N.O. Phosphorylation of beta‐catenin by cyclic AMP‐dependent protein kinase. J Biol Chem, 2006;281(15):9971–6. 65. Alvarado, T.F., Puliga, E., Preziosi, M. et al. Thyroid hormone receptor beta agonist induces beta‐catenin‐dependent hepatocyte proliferation in mice: implications in hepatic regeneration. Gene Expr, 2016;17(1):19–34. 66. Fanti, M., Singh, S., Ledda‐Columbano, G.M., Columbano, A., and Monga, S.P. Tri‐iodothyronine induces hepatocyte proliferation by protein kinase A‐ dependent beta‐catenin activation in rodents. Hepatology, 2014;59(6):2309–20. 67. Spirli, C., Locatelli, L., Morell, C.M. et  al. Protein kinase A‐dependent pSer(675) ‐beta‐catenin, a novel signaling defect in a mouse model of congenital hepatic fibrosis. Hepatology, 2013;58(5):1713–23. 68. Deng, J., Miller, S.A., Wang, H.Y. et  al. beta‐catenin interacts with and inhibits NF‐kappa B in human colon and breast cancer. Cancer Cell, 2002;2(4):323–34. 69. Nejak‐Bowen, K., Kikuchi, A., and Monga, S.P. Beta‐catenin‐NF‐kappaB interactions in murine hepatocytes: a complex to die for. Hepatology, 2013;57(2):763–74. 70. Nejak‐Bowen, K., Moghe, A., Cornuet, P., Preziosi, M., Nagarajan, S., and Monga, S.P. Role and regulation of p65/beta‐catenin association during liver injury and regeneration: a “complex” relationship. Gene Expr, 2017;17(3):219–35. 71. Du, Q., Zhang, X., Cardinal, J. et al. Wnt/beta‐catenin signaling regulates cytokine‐induced human inducible nitric oxide synthase expression by inhibiting nuclear factor‐kappaB activation in cancer cells. Cancer Res, 2009;69(9):3764–71. 72. Thompson, M.D., Moghe, A., Cornuet, P. et al. beta‐Catenin regulation of farnesoid X receptor signaling and bile acid metabolism during murine cholestasis. Hepatology, 2018;67(3):955–71. 73. Monga, S.P., Monga, H.K., Tan, X., Mule, K., Pediaditakis, P., and Michalopoulos, G.K. Beta‐catenin antisense studies in embryonic liver cul-

599

tures: role in proliferation, apoptosis, and lineage specification. Gastroenterology, 2003;124(1):202–16. 74. Micsenyi, A., Tan, X., Sneddon, T., Luo, J.H., Michalopoulos, G.K., and Monga, S.P. Beta‐catenin is temporally regulated during normal liver development. Gastroenterology, 2004;126(4):1134–46. 75. Nejak‐Bowen, K. and Monga, S.P. Wnt/beta‐catenin signaling in hepatic organogenesis. Organogenesis, 2008;4(2):92–9. 76. McLin, V.A., Rankin, S.A., and Zorn, A.M. Repression of Wnt/beta‐catenin signaling in the anterior endoderm is essential for liver and pancreas development. Development, 2007;134(12):2207–17. 77. Li, Y., Rankin, S.A., Sinner, D., Kenny, A.P., Krieg, P.A., and Zorn, A.M. Sfrp5 coordinates foregut specification and morphogenesis by antagonizing both canonical and noncanonical Wnt11 signaling. Genes Dev, 2008;22(21):3050–63. 78. Goessling, W., North, T.E., Lord, A.M. et al. APC mutant zebrafish uncover a changing temporal requirement for wnt signaling in liver development. Dev Biol, 2008;320(1):161–74. 79. Ober, E.A., Verkade, H., Field, H.A., and Stainier, D.Y. Mesodermal Wnt2b signalling positively regulates liver specification. Nature, 2006;442(7103):688–91. 80. Poulain, M. and Ober, E.A. Interplay between Wnt2 and Wnt2bb controls multiple steps of early foregut‐derived organ development. Development, 2011;138(16):3557–68. 81. Shin, D., Lee, Y., Poss, K.D., and Stainier, D.Y.R. Restriction of hepatic competence by Fgf signaling. Development, 2011;138(7):1339–48. 82. So, J., Martin, B.L., Kimelman, D., and Shin, D. Wnt/beta‐catenin signaling cell‐autonomously converts non‐hepatic endodermal cells to a liver fate. Biol Open, 2013;2(1):30–6. 83. Tan, X., Yuan, Y., Zeng, G. et al. Beta‐catenin deletion in hepatoblasts disrupts hepatic morphogenesis and survival during mouse development. Hepatology, 2008;47(5):1667–79. 84. Suksaweang, S., Lin, C.M., Jiang, T.X., Hughes, M.W., Widelitz, R.B., and Chuong, C.M. Morphogenesis of chicken liver: identification of localized growth zones and the role of beta‐catenin/Wnt in size regulation. Dev Biol, 2004;266(1):109–22. 85. Hussain, S.Z., Sneddon, T., Tan, X., Micsenyi, A., Michalopoulos, G.K., and Monga, S.P. Wnt impacts growth and differentiation in ex vivo liver development. Exp Cell Res, 2004;292(1):157–69. 86. Decaens, T., Godard, C., de Reynies, A. et al. Stabilization of beta‐catenin affects mouse embryonic liver growth and hepatoblast fate. Hepatology, 2008;47(1):247–58. 87. Cordi, S., Godard, C., Saandi, T. et al. Role of beta‐catenin in development of bile ducts. Differentiation, 2016;91(1–3):42–9. 88. So, J., Khaliq, M., Evason, K. et  al. Wnt/beta‐catenin signaling controls intrahepatic biliary network formation in zebrafish by regulating notch activity. Hepatology, 2018;67(6):2352–66. 89. Kiyohashi, K., Kakinuma, S., Kamiya, A. et al. Wnt5a signaling mediates biliary differentiation of fetal hepatic stem/progenitor cells in mice. Hepatology, 2013;57(6):2502–13. 90. Saneyoshi, T., Kume, S., Amasaki, Y., and Mikoshiba, K. The Wnt/calcium pathway activates NF‐AT and promotes ventral cell fate in Xenopus embryos. Nature, 2002;417(6886):295–9. 91. Matsumoto, K., Miki, R., Nakayama, M., Tatsumi, N., and Yokouchi, Y. Wnt9a secreted from the walls of hepatic sinusoids is essential for morphogenesis, proliferation, and glycogen accumulation of chick hepatic epithelium. Dev Biol, 2008;319(2):234–47. 92. Apte, U., Zeng, G., Muller, P. et al. Activation of Wnt/beta‐catenin pathway during hepatocyte growth factor‐induced hepatomegaly in mice. Hepatology, 2006;44(4):992–1002. 93. Schmidt, C., Bladt, F., Goedecke, S. et al. Scatter factor/hepatocyte growth factor is essential for liver development. Nature, 1995;373(6516):699–702. 94. Uehara, Y., Minowa, O., Mori, C. et  al. Placental defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor. Nature, 1995;373(6516):702–5. 95. Berg, T., Rountree, C.B., Lee, L. et al. Fibroblast growth factor 10 is critical for liver growth during embryogenesis and controls hepatoblast survival via beta‐catenin activation. Hepatology, 2007;46(4):1187–97. 96. Sekhon, S.S., Tan, X., Micsenyi, A., Bowen, W.C., and Monga, S.P. Fibroblast growth factor enriches the embryonic liver cultures for hepatic progenitors. Am J Pathol, 2004;164(6):2229–40. 97. Gebhardt, R. and Mecke, D. Heterogeneous distribution of glutamine synthetase among rat liver parenchymal cells in situ and in primary culture. EMBO J, 1983;2(4):567–70.

600

THE LIVER:  REFERENCES

  98. Benhamouche, S., Decaens, T., Godard, C. et al. Apc tumor suppressor gene is the “zonation‐keeper” of mouse liver. Dev Cell, 2006;10(6):759–70.   99. Cadoret, A., Ovejero, C., Terris, B. et al. New targets of beta‐catenin signaling in the liver are involved in the glutamine metabolism. Oncogene, 2002;21(54):8293–301. 100. Sekine, S., Lan, B.Y., Bedolli, M., Feng, S., and Hebrok, M. Liver‐specific loss of beta‐catenin blocks glutamine synthesis pathway activity and cytochrome p450 expression in mice. Hepatology, 2006;43(4):817–25. 101. Tan, X., Behari, J., Cieply, B., Michalopoulos, G.K., and Monga, S.P. Conditional deletion of beta‐catenin reveals its role in liver growth and regeneration. Gastroenterology, 2006;131(5):1561–72. 102. Wang, B., Zhao, L., Fish, M., Logan, C.Y., and Nusse, R. Self‐renewing diploid Axin2(+) cells fuel homeostatic renewal of the liver. Nature, 2015;524(7564):180–5. 103. Yang, J., Mowry, L.E., Nejak‐Bowen, K.N. et al. Beta‐catenin signaling in murine liver zonation and regeneration: a Wnt‐Wnt situation! Hepatology, 2014;60(3):964–76. 104. Leibing, T., Geraud, C., Augustin, I. et al. Angiocrine Wnt signaling controls liver growth and metabolic maturation in mice. Hepatology, 2018;68(2):707–22. 105. Preziosi, M., Okabe, H., Poddar, M., Singh, S., and Monga, S.P. Endothelial Wnts regulate β‐catenin signaling in murine liver zonation and regeneration: a sequel to the Wnt–Wnt situation. Hepatol Commun, 2018;2(7). 106. Planas‐Paz, L., Orsini, V., Boulter, L. et  al. The RSPO‐LGR4/5‐ZNRF3/ RNF43 module controls liver zonation and size. Nat Cell Biol, 2016;18(5):467–79. 107. Nejak‐Bowen, K.N., Zeng, G., Tan, X., Cieply, B., and Monga, S.P. Beta‐ catenin regulates vitamin C biosynthesis and cell survival in murine liver. J Biol Chem, 2009;284(41):28115–27. 108. Sambrotta, M., Strautnieks, S., Papouli, E. et al. Mutations in TJP2 cause progressive cholestatic liver disease. Nat Genet, 2014;46(4):326–8. 109. Wickline, E.D., Awuah, P.K., Behari, J., Ross, M., Stolz, D.B., and Monga, S.P. Hepatocyte gamma‐catenin compensates for conditionally deleted beta‐catenin at adherens junctions. J Hepatol, 2011;55(6):1256–62. 110. Wickline, E.D., Du, Y., Stolz, D.B., Kahn, M., and Monga, S.P. Gamma‐catenin at adherens junctions: mechanism and biologic implications in hepatocellular cancer after beta‐catenin knockdown. Neoplasia, 2013;15(4):421–34. 111. Mankertz, J., Hillenbrand, B., Tavalali, S., Huber, O., Fromm, M., and Schulzke, J.D. Functional crosstalk between Wnt signaling and Cdx‐related transcriptional activation in the regulation of the claudin‐2 promoter activity. Biochem Biophys Res Commun, 2004;314(4):1001–7. 112. Yeh, T.H., Krauland, L., Singh, V. et al. Liver‐specific beta‐catenin knockout mice have bile canalicular abnormalities, bile secretory defect, and intrahepatic cholestasis. Hepatology, 2010;52(4):1410–9. 113. Pradhan‐Sundd, T., Zhou, L., Vats, R. et  al. Dual catenin loss in murine liver causes tight junctional deregulation and progressive intrahepatic cholestasis. Hepatology, 2018;67(6):2320–37. 114. Vitale, G., Gitto, S., Raimondi, F. et al. Cryptogenic cholestasis in young and adults: ATP8B1, ABCB11, ABCB4, and TJP2 gene variants analysis by high‐throughput sequencing. J Gastroenterol, 2018;53(8):945–58. 115. Apte, U., Zeng, G., Thompson, M.D. et al. Beta‐catenin is critical for early postnatal liver growth. Am J Physiol Gastrointest Liver Physiol, 2007;292(6):G1578–85. 116. Cadoret, A., Ovejero, C., Saadi‐Kheddouci, S. et al. Hepatomegaly in transgenic mice expressing an oncogenic form of beta‐catenin. Cancer Res, 2001;61(8):3245–9. 117. Colnot, S., Decaens, T., Niwa‐Kawakita, M. et al. Liver‐targeted disruption of Apc in mice activates beta‐catenin signaling and leads to hepatocellular carcinomas. Proc Natl Acad Sci USA, 2004;101(49):17216–21. 118. Harada, N., Miyoshi, H., Murai, N. et  al. Lack of tumorigenesis in the mouse liver after adenovirus‐mediated expression of a dominant stable mutant of beta‐catenin. Cancer Res, 2002;62(7):1971–7. 119. Tan, X., Apte, U., Micsenyi, A. et al. Epidermal growth factor receptor: a novel target of the Wnt/beta‐catenin pathway in liver. Gastroenterology, 2005;129(1):285–302. 120. Nejak‐Bowen, K.N., Thompson, M.D., Singh, S. et  al. Accelerated liver regeneration and hepatocarcinogenesis in mice overexpressing serine‐45 mutant beta‐catenin. Hepatology, 2010;51(5):1603–13. 121. Tetsu, O. and McCormick, F. Beta‐catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature, 1999;398(6726):422–6.

122. Monga, S.P., Pediaditakis, P., Mule, K., Stolz, D.B., and Michalopoulos, G.K. Changes in WNT/beta‐catenin pathway during regulated growth in rat liver regeneration. Hepatology, 2001;33(5):1098–109. 123. Sodhi, D., Micsenyi, A., Bowen, W.C., Monga, D.K., Talavera, J.C., and  Monga, S.P. Morpholino oligonucleotide‐triggered beta‐catenin knockdown compromises normal liver regeneration. J Hepatol, 2005; ­ 43(1):132–41. 124. Sekine, S., Gutierrez, P.J., Lan, B.Y., Feng, S., and Hebrok, M. Liver‐­ specific loss of beta‐catenin results in delayed hepatocyte proliferation after partial hepatectomy. Hepatology, 2007;45(2):361–8. 125. Michalopoulos, G.K. Principles of liver regeneration and growth homeostasis. Compr Physiol, 2013;3(1):485–513. 126. Yang, J., Cusimano, A., Monga, J.K. et al. WNT5A inhibits hepatocyte proliferation and concludes beta‐catenin signaling in liver regeneration. Am J Pathol, 2015;185(8):2194–205. 127. Apte, U., Singh, S., Zeng, G. et al. Beta‐catenin activation promotes liver regeneration after acetaminophen‐induced injury. Am J Pathol, 2009; 175(3):1056–65. 128. Puliga, E., Min, Q., Tao, J. et al. Thyroid hormone receptor‐beta agonist GC‐1 inhibits Met‐beta‐catenin‐driven hepatocellular cancer. Am J Pathol, 2017;187(11):2473–85. 129. Gebhardt, R. Speeding up hepatocyte proliferation: how triiodothyronine and beta‐catenin join forces. Hepatology, 2014;59(6):2074–6. 130. Gebhardt, R. Metabolic zonation of the liver: regulation and implications for liver function. Pharmacol Ther, 1992;53(3):275–354. 131. Audard, V., Grimber, G., Elie, C. et al. Cholestasis is a marker for hepatocellular carcinomas displaying beta‐catenin mutations. J Pathol, 2007; 212(3):345–52. 132. Behari, J., Yeh, T.H., Krauland, L. et al. Liver‐specific beta‐catenin knockout mice exhibit defective bile acid and cholesterol homeostasis and increased susceptibility to diet‐induced steatohepatitis. Am J Pathol, 2010;176(2):744–53. 133. Gougelet, A., Torre, C., Veber, P. et al. T‐cell factor 4 and beta‐catenin chromatin occupancies pattern zonal liver metabolism in mice. Hepatology, 2014;59(6):2344–57. 134. Tsuchida, T. and Friedman, S.L. Mechanisms of hepatic stellate cell activation. Nat Rev Gastroenterol Hepatol, 2017;14(7):397–411. 135. Shackel, N.A., McGuinness, P.H., Abbott, C.A., Gorrell, M.D., and McCaughan, G.W. Identification of novel molecules and pathogenic pathways in primary biliary cirrhosis: cDNA array analysis of intrahepatic ­differential gene expression. Gut, 2001;49(4):565–76. 136. Tanaka, A., Leung, P.S., Kenny, T.P. et al. Genomic analysis of differentially expressed genes in liver and biliary epithelial cells of patients with primary biliary cirrhosis. J Autoimmun, 2001;17(1):89–98. 137. Jiang, F., Parsons, C.J., and Stefanovic, B. Gene expression profile of quiescent and activated rat hepatic stellate cells implicates Wnt signaling pathway in activation. J Hepatol, 2006;45(3):401–9. 138. Rashid, S.T., Humphries, J.D., Byron, A. et al. Proteomic analysis of extracellular matrix from the hepatic stellate cell line LX‐2 identifies CYR61 and Wnt‐5a as novel constituents of fibrotic liver. J Proteome Res, 2012;11(8):4052–64. 139. Xiong, W.J., Hu, L.J., Jian, Y.C. et al. Wnt5a participates in hepatic stellate cell activation observed by gene expression profile and functional assays. World J Gastroenterol, 2012;18(15):1745–52. 140. Kordes, C., Sawitza, I., and Haussinger, D. Canonical Wnt signaling maintains the quiescent stage of hepatic stellate cells. Biochem Biophys Res Commun, 2008;367(1):116–23. 141. Ge, W.S., Wang, Y.J., Wu, J.X., Fan, J.G., Chen, Y.W., and Zhu, L. beta‐ catenin is overexpressed in hepatic fibrosis and blockage of Wnt/­ beta‐catenin signaling inhibits hepatic stellate cell activation. Mol Med Rep, 2014;9(6):2145–51. 142. Cheng, J.H., She, H., Han, Y.P. et al. Wnt antagonism inhibits hepatic stellate cell activation and liver fibrosis. Am J Physiol Gastrointest Liver Physiol, 2008;294(1):G39–49. 143. Myung, S.J., Yoon, J.H., Gwak, G.Y. et al. Wnt signaling enhances the activation and survival of human hepatic stellate cells. FEBS Lett, 2007;581(16):2954–8. 144. Haughton, E.L., Tucker, S.J., Marek, C.J. et al. Pregnane X receptor activators inhibit human hepatic stellate cell transdifferentiation in vitro. Gastroenterology, 2006;131(1):194–209.

46:  β-CATENIN SIGNALING 145. Zhu, N.L., Wang, J., and Tsukamoto, H. The Necdin‐Wnt pathway causes epigenetic peroxisome proliferator‐activated receptor gamma repression in hepatic stellate cells. J Biol Chem, 2010;285(40):30463–71. 146. Yanagida, A., Iwaisako, K., Hatano, E. et al. Downregulation of the Wnt antagonist Dkk2 links the loss of Sept4 and myofibroblastic transformation of hepatic stellate cells. Biochim Biophys Acta, 2011;1812(11):1403–11. 147. Li, W., Zhu, C., Li, Y., Wu, Q., and Gao, R. Mest attenuates CCl4‐induced liver fibrosis in rats by inhibiting the Wnt/beta‐catenin signaling pathway. Gut Liver, 2014;8(3):282–91. 148. Ross, S.E., Hemati, N., Longo, K.A. et al. Inhibition of adipogenesis by Wnt signaling. Science, 2000;289(5481):950–3. 149. Kweon, S.M., Chi, F., Higashiyama, R., Lai, K., and Tsukamoto, H. Wnt pathway stabilizes MeCP2 protein to repress PPAR‐gamma in activation of hepatic stellate cells. PLoS One, 2016;11(5):e0156111. 150. Lai, K.K.Y., Kweon, S.M., Chi, F. et al. Stearoyl‐CoA desaturase promotes liver fibrosis and tumor development in mice via a Wnt positive‐signaling loop by stabilization of low‐density lipoprotein‐receptor‐related proteins 5 and 6. Gastroenterology, 2017;152(6):1477–91. 151. Hughes, L.J. and Michels, V.V. Risk of hepatoblastoma in familial adenomatous polyposis. Am J Med Genet, 1992;43(6):1023–5. 152. Kurahashi, H., Takami, K., Oue, T. et al. Biallelic inactivation of the APC gene in hepatoblastoma. Cancer Res, 1995;55(21):5007–11. 153. Oda, H., Imai, Y., Nakatsuru, Y., Hata, J., and Ishikawa, T. Somatic mutations of the APC gene in sporadic hepatoblastomas. Cancer Res, 1996;56(14):3320–3. 154. Koch, A., Denkhaus, D., Albrecht, S., Leuschner, I., von Schweinitz, D., and Pietsch, T. Childhood hepatoblastomas frequently carry a mutated degradation targeting box of the beta‐catenin gene. Cancer Res, 1999;59(2):269–73. 155. Taniguchi, K., Roberts, L.R., Aderca, I.N. et  al. Mutational spectrum of beta‐catenin, AXIN1, and AXIN2 in hepatocellular carcinomas and hepatoblastomas. Oncogene, 2002;21(31):4863–71. 156. Cairo, S., Armengol, C., De Reynies, A. et al. Hepatic stem‐like phenotype and interplay of Wnt/beta‐catenin and Myc signaling in aggressive childhood liver cancer. Cancer Cell, 2008;14(6):471–84. 157. Bell, D., Ranganathan, S., Tao, J., and Monga, S.P. Novel advances in understanding of molecular pathogenesis of hepatoblastoma: a Wnt/ beta‐catenin perspective. Gene Expr, 2017;17(2):141–54. 158. Lade, A.G. and Monga, S.P. Beta‐catenin signaling in hepatic development and progenitors: which way does the WNT blow? Dev Dyn, 2011;240(3): 486–500. 159. Shin, D. and Monga, S.P. Cellular and molecular basis of liver development. Compr Physiol, 2013;3(2):799–815. 160. Lopez‐Terrada, D., Gunaratne, P.H., Adesina, A.M. et al. Histologic subtypes of hepatoblastoma are characterized by differential canonical Wnt and Notch pathway activation in DLK+ precursors. Hum Pathol, 2009;40(6):783–94. 161. Armengol, C., Cairo, S., Fabre, M., and Buendia, M.A. Wnt signaling and hepatocarcinogenesis: the hepatoblastoma model. Int J Biochem Cell Biol, 2011;43(2):265–70. 162. Tao, J., Calvisi, D.F., Ranganathan, S. et al. Activation of beta‐catenin and Yap1 in human hepatoblastoma and induction of hepatocarcinogenesis in mice. Gastroenterology, 2014;147(3):690–701. 163. Wang, H., Lu, J., Edmunds, L.R. et al. Coordinated activities of multiple Myc‐dependent and Myc‐independent biosynthetic pathways in hepatoblastoma. J Biol Chem, 2016;291(51):26241–51. 164. Molina, L., Bell, D., Tao, J. et al. Hepatocyte‐derived lipocalin 2 is a potential serum biomarker reflecting tumor burden in hepatoblastoma. Am J Pathol, 2018;188(8):1895–909. 165. Nault, J.C., Bioulac‐Sage, P., and Zucman‐Rossi, J. Hepatocellular benign tumors‐from molecular classification to personalized clinical care. Gastroenterology, 2013;144(5):888–902. 166. Zucman‐Rossi, J., Jeannot, E., Nhieu, J.T. et al. Genotype‐phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology, 2006;43(3):515–24. 167. Bioulac‐Sage, P., Rebouissou, S., Thomas, C. et  al. Hepatocellular adenoma subtype classification using molecular markers and immunohistochemistry. Hepatology, 2007;46(3):740–8. 168. Ihara, A., Koizumi, H., Hashizume, R., and Uchikoshi, T. Expression of epithelial cadherin and alpha‐ and beta‐catenins in nontumoral livers and hepatocellular carcinomas. Hepatology, 1996;23(6):1441–7.

601

169. de La Coste, A., Romagnolo, B., Billuart, P. et al. Somatic mutations of the beta‐catenin gene are frequent in mouse and human hepatocellular carcinomas. Proc Natl Acad Sci USA, 1998;95(15):8847–51. 170. Schulze, K., Imbeaud, S., Letouze, E. et al. Exome sequencing of hepatocellular carcinomas identifies new mutational signatures and potential therapeutic targets. Nat Genet, 2015;47(5):505–11. 171. Tao, J., Xu, E., Zhao, Y. et al. Modeling a human hepatocellular carcinoma subset in mice through coexpression of met and point‐mutant beta‐catenin. Hepatology, 2016;64(5):1587–605. 172. Torrecilla, S., Sia, D., and Harrington, A.N. Trunk mutational events present minimal intra‐ and inter‐tumoral heterogeneity in hepatocellular carcinoma. J Hepatol, 2017;67(6):1222–31. 173. Satoh, S., Daigo, Y., Furukawa, Y. et al. AXIN1 mutations in hepatocellular carcinomas, and growth suppression in cancer cells by virus‐mediated transfer of AXIN1. Nat Genet, 2000;24(3):245–50. 174. Zucman‐Rossi, J., Benhamouche, S., Godard, C. et al. Differential effects of inactivated Axin1 and activated beta‐catenin mutations in human hepatocellular carcinomas. Oncogene, 2007;26(5):774–80. 175. Merle, P., de la Monte, S., Kim, M. et al. Functional consequences of frizzled‐7 receptor overexpression in human hepatocellular carcinoma. Gastroenterology, 2004;127(4):1110–22. 176. Nambotin, S.B., Lefrancois, L., Sainsily, X. et al. Pharmacological inhibition of Frizzled‐7 displays anti‐tumor properties in hepatocellular carcinoma. J Hepatol, 2010. 177. Kim, M., Lee, H.C., Tsedensodnom, O. et al. Functional interaction between Wnt3 and Frizzled‐7 leads to activation of the Wnt/beta‐catenin signaling pathway in hepatocellular carcinoma cells. J Hepatol, 2008;48(5):780–91. 178. Ban, K.C., Singh, H., Krishnan, R., and Seow, H.F. GSK‐3beta phosphorylation and alteration of beta‐catenin in hepatocellular carcinoma. Cancer Lett, 2003;199(2):201–8. 179. Shih, Y.L., Shyu, R.Y., Hsieh, C.B. et  al. Promoter methylation of the secreted frizzled‐related protein 1 gene SFRP1 is frequent in hepatocellular carcinoma. Cancer, 2006;107(3):579–90. 180. Takagi, H., Sasaki, S., Suzuki, H. et al. Frequent epigenetic inactivation of SFRP genes in hepatocellular carcinoma. J Gastroenterol, 2008;43(5): 378–89. 181. Hoshida, Y., Nijman, S.M., Kobayashi, M. et al. Integrative transcriptome analysis reveals common molecular subclasses of human hepatocellular carcinoma. Cancer Res, 2009;69(18):7385–92. 182. Cieply, B., Zeng, G., Proverbs‐Singh, T., Geller, D.A., and Monga, S.P. Unique phenotype of hepatocellular cancers with exon‐3 mutations in beta‐ catenin gene. Hepatology, 2009;49(3):821–31. 183. Boyault, S., Rickman, D.S., de Reynies, A. et al. Transcriptome classification of HCC is related to gene alterations and to new therapeutic targets. Hepatology, 2007;45(1):42–52. 184. Rebouissou, S., Franconi, A., Calderaro, J. et al. Genotype‐phenotype correlation of CTNNB1 mutations reveals different ss‐catenin activity associated with liver tumor progression. Hepatology, 2016;64(6):2047–61. 185. Yuan, R.H., Jeng, Y.M., Hu, R.H. et al. Role of p53 and beta‐catenin mutations in conjunction with CK19 expression on early tumor recurrence and prognosis of hepatocellular carcinoma. J Gastrointest Surg, 2011;15(2): 321–9. 186. Nhieu, J.T., Renard, C.A., Wei, Y., Cherqui, D., Zafrani, E.S., and Buendia, M.A. Nuclear accumulation of mutated beta‐catenin in hepatocellular carcinoma is associated with increased cell proliferation. Am J Pathol, 1999;155(3):703–10. 187. Suzuki, T., Yano, H., Nakashima, Y., Nakashima, O., and Kojiro, M. Beta‐ catenin expression in hepatocellular carcinoma: a possible participation of beta‐catenin in the dedifferentiation process. J Gastroenterol Hepatol, 2002;17(9):994–1000. 188. Dal Bello, B., Rosa, L., Campanini, N. et al. Glutamine synthetase immunostaining correlates with pathologic features of hepatocellular carcinoma and better survival after radiofrequency thermal ablation. Clin Cancer Res,Cancer Res, 2010;16(7):2157–66. 189. Lok, A.S., Seeff, L.B., Morgan, T.R. et al. Incidence of hepatocellular carcinoma and associated risk factors in hepatitis C‐related advanced liver disease. Gastroenterology, 2009;136(1):138–48. 190. Lee, J.M., Yang, J., Newell, P. et al. beta‐Catenin signaling in hepatocellular cancer: Implications in inflammation, fibrosis, and proliferation. Cancer Lett, 2014;343(1):90–7.

602

THE LIVER:  REFERENCES

191. Harada, N., Oshima, H., Katoh, M., Tamai, Y., Oshima, M., and Taketo, M.M. Hepatocarcinogenesis in mice with beta‐catenin and Ha‐ras gene mutations. Cancer Res, 2004;64(1):48–54. 192. Miyoshi, H., Deguchi, A., Nakau, M. et  al. Hepatocellular carcinoma development induced by conditional beta‐catenin activation in Lkb1+/‐ mice. Cancer Sci, 2009;100(11):2046–53. 193. Bauer‐Hofmann, R., Klimek, F., Buchmann, A., Muller, O., Bannasch, P., and Schwarz, M. Role of mutations at codon 61 of the c‐Ha‐ras gene during diethylnitrosamine‐induced hepatocarcinogenesis in C3H/He mice. Mol Carcinog, 1992;6(1):60–7. 194. Nault, J.C., Mallet, M., Pilati, C. et  al. High frequency of telomerase reverse‐transcriptase promoter somatic mutations in hepatocellular carcinoma and preneoplastic lesions. Nat Commun, 2013;4:2218. 195. Tao, J., Zhang, R., Singh, S. et al. Targeting beta‐catenin in hepatocellular cancers induced by coexpression of mutant beta‐catenin and K‐Ras in mice. Hepatology, 2017;65(5):1581–99. 196. Zeng, G., Apte, U., Cieply, B., Singh, S., and Monga, S.P. siRNA‐mediated beta‐catenin knockdown in human hepatoma cells results in decreased growth and survival. Neoplasia, 2007;9(11):951–9. 197. Yao, M., Kargman, S., Lam, E.C. et al. Inhibition of cyclooxygenase‐2 by rofecoxib attenuates the growth and metastatic potential of colorectal carcinoma in mice. Cancer Res, 2003;63(3):586–92. 198. Behari, J., Zeng, G., Otruba, W. et al. R‐Etodolac decreases beta‐catenin levels along with survival and proliferation of hepatoma cells. J Hepatol, 2007;46(5):849–57. 199. Li, H., Pamukcu, R., and Thompson, W.J. Beta‐catenin signaling: therapeutic strategies in oncology. Cancer Biol Ther, 2002;1(6):621–5. 200. Emami, K.H., Nguyen, C., Ma, H. et al. A small molecule inhibitor of beta‐ catenin/CREB‐binding protein transcription [corrected]. Proc Natl Acad Sci USA, 2004;101(34):12682–7. 201. Lenz, H.J. and Kahn, M. Safely targeting cancer stem cells via selective catenin coactivator antagonism. Cancer Sci, 2014. 202. Delgado, E.R., Yang, J., So, J. et al. Identification and characterization of a novel small‐molecule inhibitor of beta‐catenin signaling. Am J Pathol, 2014;184(7):2111–22. 203. Nejak‐Bowen, K.N. and Monga, S.P. Beta‐catenin signaling, liver regeneration and hepatocellular cancer: sorting the good from the bad. Semin Cancer Biol, 2011;21(1):44–58. 204. Okabe, H., Delgado, E., Lee, J.M. et  al. Role of leukocyte cell‐derived chemotaxin 2 as a biomarker in hepatocellular carcinoma. PLoS One, 2014;9(6):e98817. 205. Zhan, N., Michael, A.A., Wu, K. et al. The effect of selective c‐MET inhibitor on hepatocellular carcinoma in the MET‐active, beta‐catenin‐mutated mouse model. Gene Expr, 2018;18(2):135–47. 206. Delgado, E., Okabe, H., Preziosi, M. et al. Complete response of Ctnnb1‐ mutated tumours to beta‐catenin suppression by locked nucleic acid antisense in a mouse hepatocarcinogenesis model. J Hepatol, 2014. 207. Adebayo Michael, A.O., Ko, S., and Tao, J. Inhibiting glutamine‐dependent mTORC1 activation ameliorates liver cancers driven by β‐catenin mutations. Cell Metab, 2019;29(5):1135–50. 208. Ruiz de Galarreta, M., Bresnahan, E., and Molina‐Sanchez, P. β‐catenin activation promotes immune escape and resistance to anti‐PD‐1 therapy in hepatocellular carcinoma. Cancer Discov, 2019. doi: 10.1158/2159‐8290. CD‐19‐0074 209. Nakanuma, Y., Harada, K., Ishikawa, A., Zen, Y., and Sasaki, M. Anatomic and molecular pathology of intrahepatic cholangiocarcinoma. J Hepatobiliary Pancreat Surg, 2003;10(4):265–81. 210. Ashida, K., Terada, T., Kitamura, Y., and Kaibara, N. Expression of E‐cadherin, alpha‐catenin, beta‐catenin, and CD44 (standard and variant isoforms) in human cholangiocarcinoma: an immunohistochemical study. Hepatology, 1998;27(4):974–82. 211. Rashid, A. Cellular and molecular biology of biliary tract cancers. Surg Oncol Clin N Am, 2002;11(4):995–1009. 212. Sugimachi, K., Taguchi, K., Aishima, S. et al. Altered expression of beta‐ catenin without genetic mutation in intrahepatic cholangiocarcinoma. Mod Pathol, 2001;14(9):900–5.

213. Rashid, A., Gao, Y.T., Bhakta, S. et  al. Beta‐catenin mutations in biliary tract cancers: a population‐based study in China. Cancer Res, 2001; 61(8):3406–9. 214. Abraham, S.C., Lee, J.H., Hruban, R.H., Argani, P., Furth, E.E., and Wu, T.T. Molecular and immunohistochemical analysis of intraductal papillary neoplasms of the biliary tract. Hum Pathol, 2003;34(9):902–10. 215. Lustig, B., Jerchow, B., Sachs, M. et  al. Negative feedback loop of Wnt signaling through upregulation of conductin/axin2 in colorectal and liver tumors. Mol Cell Biol, 2002;22(4):1184–93. 216. Takayasu, H., Horie, H., Hiyama, E. et al. Frequent deletions and mutations of the beta‐catenin gene are associated with overexpression of cyclin D1 and fibronectin and poorly differentiated histology in childhood hepatoblastoma. Clin Cancer Res, 2001;7(4):901–8. 217. Torre, C., Benhamouche, S., Mitchell, C. et al. The transforming growth factor‐alpha and cyclin D1 genes are direct targets of beta‐catenin signaling in hepatocyte proliferation. J Hepatol, 2011;55(1):86–95. 218. Araki, Y., Okamura, S., Hussain, S.P. et al. Regulation of cyclooxygenase‐2 expression by the Wnt and ras pathways. Cancer Res, 2003;63(3):728–34. 219. Loeppen, S., Koehle, C., Buchmann, A., and Schwarz, M. A beta‐catenin‐ dependent pathway regulates expression of cytochrome P450 isoforms in mouse liver tumors. Carcinogenesis, 2005;26(1):239–48. 220. Yamamoto, Y., Sakamoto, M., Fujii, G. et al. Overexpression of orphan G‐ protein‐coupled receptor, Gpr49, in human hepatocellular carcinomas with beta‐catenin mutations. Hepatology, 2003;37(3):528–33. 221. Ovejero, C., Cavard, C., Perianin, A. et al. Identification of the leukocyte cell‐derived chemotaxin 2 as a direct target gene of beta‐catenin in the liver. Hepatology, 2004;40(1):167–76. 222. Cavard, C., Terris, B., Grimber, G. et al. Overexpression of regenerating islet‐derived 1 alpha and 3 alpha genes in human primary liver tumors with beta‐catenin mutations. Oncogene, 2006;25(4):599–608. 223. Renard, C.A., Labalette, C., Armengol, C. et al. Tbx3 is a downstream target of the Wnt/beta‐catenin pathway and a critical mediator of beta‐catenin survival functions in liver cancer. Cancer Res, 2007;67(3):901–10. 224. Delgado, E., Bahal, R., Yang, J., Lee, J.M., Ly, D.H., and Monga, S.P. Beta‐ catenin knockdown in liver tumor cells by a cell permeable gamma guanidine‐ based peptide nucleic acid. Curr Cancer Drug Targets, 2013;13(8): 867–78. 225. Guichard, C., Amaddeo, G., Imbeaud, S. et  al. Integrated analysis of somatic mutations and focal copy‐number changes identifies key genes and pathways in hepatocellular carcinoma. Nat Genet, 2012;44(6):694–8. 226. Cui, J., Zhou, X., Liu, Y., Tang, Z., and Romeih, M. Alterations of beta‐catenin and Tcf‐4 instead of GSK‐3beta contribute to activation of  Wnt pathway in hepatocellular carcinoma. Chin Med J (Engl), 2003;116(12):1885–92. 227. Devereux, T.R., Stern, M.C., Flake, G.P. et  al. CTNNB1 mutations and beta‐catenin protein accumulation in human hepatocellular carcinomas associated with high exposure to aflatoxin B1. Mol Carcinog, 2001;31(2): 68–73. 228. Wong, C.M., Fan, S.T., and Ng, I.O. Beta‐catenin mutation and overexpression in hepatocellular carcinoma: clinicopathologic and prognostic significance. Cancer, 2001;92(1):136–45. 229. Hsu, H.C., Jeng, Y.M., Mao, T.L., Chu, J.S., Lai, P.L., and Peng, S.Y. Beta‐ catenin mutations are associated with a subset of low‐stage hepatocellular carcinoma negative for hepatitis B virus and with favorable prognosis. Am J Pathol, 2000;157(3):763–70. 230. Huang, H., Fujii, H., Sankila, A. et al. Beta‐catenin mutations are frequent in human hepatocellular carcinomas associated with hepatitis C virus infection. Am J Pathol, 1999;155(6):1795–801. 231. Legoix, P., Bluteau, O., Bayer, J. et al. Beta‐catenin mutations in hepatocellular carcinoma correlate with a low rate of loss of heterozygosity. Oncogene, 1999;18(27):4044–6. 232. Kondo, Y., Kanai, Y., Sakamoto, M. et al. Beta‐catenin accumulation and mutation of exon 3 of the beta‐catenin gene in hepatocellular carcinoma. Jpn J Cancer Res, 1999;90(12):1301–9.

47

Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer Evan R. Delgado, Elizabeth C. Stahl, Nairita Roy, Patrick D. Wilkinson, and Andrew W. Duncan Department of Pathology, McGowan Institute for Regenerative Medicine, Pittsburgh Liver Research Center, University of Pittsburgh, Pittsburgh, PA, USA

HEPATIC CHROMOSOMAL VARIATIONS AND LIVER FUNCTION The development of polyploid cells in tissues and organs is a highly regulated process. In the liver, failure to complete the final stage of mitosis, known as cytokinesis, drives the retention of nuclear content to a single cell. The heterogeneity of polyploid cells in the liver is further nuanced by the occurrence of aneuploidy, where single chromosomes may be gained or lost, and by reductive mitosis, all driving the “ploidy conveyor.”

Polyploidy in somatic tissues and the liver The majority of mammalian cells have a diploid nuclear content and contain two sets of chromosomes. However, there is a phenomenon called polyploidy, which describes cells with additional sets of chromosomes. Polyploid cells are an innate component of many mammalian species, including rodents and humans [1, 2]. The formation and location of these polyploid cells is tissue‐dependent. For example, myofibrils and osteoclasts become multinucleate, polyploid cells via cell fusion [3, 4]. Others, including megakaryocytes and trophoblast giant cells, become polyploid through endoreduplication. Here, proliferating cells progress through the cell cycle but fail to complete nuclear division during mitosis, resulting in a polyploid cell with a single nucleus [5, 6]. On the other hand, failure of cytokinesis (the final step of cell division where the parent cell’s cytoplasm divides to produce two daughter cells) primarily drives the generation of polyploid cardiomyocytes and hepatocytes [2, 7, 8].

Polyploidy levels change profoundly during liver growth and maturation. In the early stages of liver development, the liver is primarily populated with small, uniform diploid (2n) hepatocytes [9, 10]. As the hepatocytes proliferate, the diploids fail to complete cytokinesis, and in turn generate a daughter cell with two diploid nuclei (2n × 2n), referred to as a binucleate tetraploid (Figure  47.1a). Next, tetraploid cells with two diploid nuclei replicate their DNA and generate two mononucleate tetraploid (4n) daughter cells by successfully completing cytokinesis. Mononucleate tetraploid cells can produce mono‐ and binucleate octaploid (8n and 4n × 4n) hepatocytes through subsequent cell divisions with complete and incomplete cytokinesis. The process continues, producing hexadecaploid (16n) cells and hepatocytes with even greater ploidy states [2]. While failure to complete cytokinesis is the most common mechanism by which hepatocytes increase ploidy states, heterotypic cell fusion between macrophages and hepatocytes has been shown to occur, albeit infrequently (at approximately 1/100 000 hepatocytes) [11]. In general, as the liver matures the hepatocyte population becomes increasingly heterogeneous in size, function, and nuclear content, including number of nuclei per cell and chromosome sets per nucleus [2, 8, 12]. The rate and percentage of liver polyploidization differ between species [2]. At birth, the majority of human hepatocytes in the liver are diploid, with polyploids representing less than 10% [13]. Notably, tetraploid hepatocytes appear within the first few weeks of development, followed by octaploid hepatocytes at two to three months after birth [14]. By adulthood, the percentage of polyploidy is nearly doubled to 20%, and after the age of 60, approximately 50% of the human liver is comprised of polyploid hepatocytes [13, 15]. Alternatively, the

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

604

THE LIVER:  HEPATIC CHROMOSOMAL VARIATIONS AND LIVER FUNCTION

Figure 47.1  Polyploidy and aneuploidy in the liver. (a) Hepatocytes are mononucleate or binucleate, and ploidy is determined by the number of nuclei per cell and DNA content of each nucleus. Cell division associated with failed cytokinesis drives hepatic polyploidization. The cartoon shows how a single diploid hepatocyte can give rise to mononucleate and binucleate tetraploid and octaploid hepatocytes. (b) In humans, diploid, tetraploid, and octaploid hepatocytes contain 46, 92, and 184 chromosomes (Chr), respectively. Euploid karyotypes contain balanced sets of chromosomes, whereas aneuploid karyotypes contain one or more random chromosome gains and/or losses. (c) Polyploid hepatocytes can undergo chromosome segregation errors (e.g. lagging chromosomes) and multipolar cell divisions to generate aneuploid daughter cells with reduced ploidy. The flowchart outlines the various types of daughter cells that can be generated by a dividing tetraploid hepatocyte undergoing bipolar (left) or multipolar (middle and right) nuclear segregation. (d) The ploidy conveyor model describes hepatic ploidy changes that occur during postnatal development and proliferation. Cell divisions involving cytokinesis failure occur frequently and lead to increased ploidy. In contrast, cell divisions associated with multipolar nuclear segregation are infrequent and lead to ploidy reversal and formation of diploid or near‐diploid daughters.



47:  Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer

livers of mice are already 50% polyploid at weaning, whereas the livers of rats begin to polyploidize upon weaning, triggered by changes in insulin AKT‐dependent signaling that result in cytokinesis failure [16]. By eight to twelve weeks of age both mouse and rat livers become 80% polyploid, a faster rate than human polyploidization, making rodents a useful model to study polyploidization in the span of a few weeks [17]. The signals regulating the steady accumulation of polyploid hepatocytes in developing and aging humans remain unknown but may also be related to the insulin‐AKT pathway. Interestingly, the frequency of cells within each ploidy state remains relatively constant among individuals of the same age, suggesting the existence of a ploidy sensor to maintain homeostasis in the liver.

Liver aneuploidy and the ploidy conveyor While hepatocytes are heterogeneous in regard to the number of chromosome sets, they also exhibit diversity at the individual chromosome level. G‐banding, fluorescence in situ hybridization, and single cell DNA sequencing studies have indicated 4–50% of hepatocytes in healthy livers of mice and humans are aneuploid, having gained or lost individual chromosomes (Figure 47.1b) [13, 17, 18]. Recently, it was shown that the livers of mice deficient in microRNA‐122 had reduced polyploid levels and were enriched with diploid hepatocytes. Moreover, the livers had abnormally low aneuploidy levels [19]. Additionally, studies have shown that polyploidy and aneuploidy levels follow similar trends, as both increase with age and are the highest in adult murine and human livers [2, 20]. The degree of aneuploidy plateaus in mice from four to fifteen months of age [17]. Similarly, in humans, hepatic aneuploidy levels remain stable between the ages of 10 and 60 years [13]. During mitosis, polyploid hepatocytes can form multipolar spindles and drive “reductive mitosis,” or a reduction in nuclear content, a process called “ploidy reversal” (Figure  47.1c,d). One example of ploidy reversal is illustrated with a proliferating tetraploid hepatocyte that forms a tripolar spindle during mitosis (Figure 47.1c, middle panel). The chromosomes of the dividing cell are pulled toward three poles, resulting in two daughter cells with approximately diploid nuclei and one daughter cell with a ~tetraploid nucleus. Additionally, polyploid cells can perform ploidy reversal by undergoing double mitosis (Figure  47.1c, right panel). In this case, proliferating tetraploids can generate up to four ~diploid daughter cells and octaploids can generate up to eight ~diploid daughter cells [17, 21, 22]. Chromosome segregation errors often occur during multipolar cell division [17, 21]. For instance, microtubules associated with the outer poles occasionally attach to the same kinetochore during multipolar spindle construction. If left unchecked, this error prevents chromosomes from migrating to discrete poles during anaphase. Consequently, these “lagging” chromosomes are frequently excluded from the daughter cell nuclei and form what is considered a micronucleus. Cytogenetic analyses of hepatocytes have shown that liver aneuploidy is widespread, with chromosome gains and losses occurring in an unbiased fashion [13, 23]. Polyploidization, ploidy reversal and aneuploidy, which are collectively called the “ploidy conveyor”, prominently drive hepatocyte heterogeneity (Figure 47.1c,d) [2].

605

THE IMPACT OF POLYPLOIDY ON GENE EXPRESSION AND LIVER REGENERATION The liver is a vitally important organ situated to metabolize hundreds of molecules for energy regulation and drug toxicity, as well as to produce important blood anticoagulation proteins, immune complexes, and digestive bile acids. These processes require the orchestration of complex gene networks and necessitate the maintenance of liver function for survival. As such, the liver has a robust capacity to adapt to injury and undergo complete regeneration. The impact of polyploidy on gene expression and liver regeneration is just beginning to be understood.

Polyploidy in liver gene expression and function While the mechanisms that produce polyploid hepatocytes have been well characterized, the relationship between ploidy and gene expression remains unclear. Experiments have shown that polyploidy is associated with specific gene expression patterns in yeast [24]. In regards to mammals, studies have demonstrated that nuclear content affects expression levels of genes regulating megakaryocyte differentiation [25] and that giant trophoblast cells could exhibit biallelic X‐chromosome gene expression due to incomplete X‐inactivation [26]. Additionally, a large‐scale genome comparative study illustrated that cardiac stress genes were expressed at different levels between diploid and polyploid cardiomyocytes [27]. It has been postulated that hepatic ploidy state could affect gene expression. One hypothesis is that gene expression levels increase proportionally with ploidy. In this case, tetraploids and octaploids would have 2× and 4× greater gene expression levels, respectively, compared to diploids. Thus, tetraploid and octaploid cells would also potentially contain 2× and 4× the total RNA and protein content as diploid cells [2]. Another hypothesis is that diploid and polyploid hepatocytes exhibit differential gene expression. In this scenario, gene expression levels would vary on a gene to gene basis between ploidy populations, with polyploids exhibiting greater expression of specific genes versus diploids and vice versa. A 2007 study by Lu et al. compared gene expression between quiescent diploid, tetraploid, and octaploid hepatocyte populations isolated by flow cytometry. Surprisingly, the study found gene expression was mostly equivalent between ploidy populations, and the magnitude of difference was small for those genes with different expression levels [28]. This suggested that few genes are differentially expressed, at least in quiescent hepatocytes. Further studies are needed to determine if differential gene expression occurs in disease conditions or in response to specific stimuli because variations at the molecular level could ultimately endow certain ploidy subsets with unique functional capabilities [2, 29].

Liver ploidy supports hepatic adaptation It has been hypothesized that the diverse population of polyploid and aneuploid hepatocytes endows the liver with the ability to adapt to a variety of environmental stresses [23]. Interestingly,

606

THE LIVER:  THE IMPACT OF POLYPLOIDY ON GENE EXPRESSION AND LIVER REGENERATION

Figure 47.2  Proof of concept studies demonstrated that disease‐resistant aneuploid hepatocytes promote adaptation to chronic liver injury. Healthy mouse livers contain randomly aneuploid hepatocytes, which are illustrated by multicolored cells. In response to chronic injury (induced by tyrosinemia in Hgd+/− Fah−/− mice), liver function was severely impaired. Hepatocytes lacking chromosome 16 (blue) were resistant to the injury and capable of proliferation. Clonal expansion by these injury‐resistant hepatocytes spontaneously repopulated the liver and restored liver function.

studies have shown that budding yeast strains with specific chromosome numbers are able to survive toxic insults and deleterious mutations [30, 31]. In the context of the liver, it has been demonstrated that aneuploid hepatocytes could protect against chronic liver disease. Notably, a study illustrated that mice suffering from tyrosinemia due to loss of fumarylacetoacetate hydrolase (FAH) an enzyme in the tyrosine catabolic pathway, developed resistance to the disease partially through hepatic aneuploidy [23] (Figure  47.2). Mice suffering from tyrosinemia can be treated and kept healthy by blocking the tyrosine catabolic pathway upstream of FAH using the drug 2‐(2‐nitro‐4‐trifluoro‐ methylbenzoyl)‐1,3‐cyclo‐hexanedione (NTBC) or deleting the enzyme homogentisic acid dioxygenase (HGD) [32, 33]. In the aforementioned study, NTBC was withdrawn from Fah−/− Hgd+/− mice and, contrary to the expectation that all mice would succumb to liver failure, some mice developed numerous small, healthy nodules in their livers. Karyotyping and array comparative genomic hybridization analyses revealed that many of the healthy nodules were comprised of aneuploid hepatocytes lacking a copy of chromosome 16, which contains the wild‐type copy of Hgd. Thus, the healthy liver nodules were populated with Hgd‐deficient hepatocytes. It was hypothesized that chronic injury was toxic to the majority of hepatocytes except those that had previously lost the chromosome with the Hgd wild‐type copy. The disease‐resistant hepatocytes (monosomic for chromosome 16), proliferated and repopulated the liver and restored normal liver function [23].

Polyploidy and liver regeneration The regenerative capacity of the liver has been well documented, but the role of polyploidy in this process remains unclear. It is hypothesized that diploid and polyploid populations could play unique roles in the process. Originally, it was believed that polyploid hepatocytes were mature, terminally differentiated cells with little proliferative capacity. This was suggested from studies demonstrating that hepatocytes in livers of mice and rats become increasingly polyploid with age, and that more than 99% of hepatocytes in adult livers were quiescent [34, 35]. However, polyploids have been observed to proliferate in response to partial hepatectomy (PH), the surgical removal of up to two‐thirds of the liver mass [36]. While these experiments confirmed that polyploid hepatocytes can proliferate, it is unclear if the

proliferative capacities of diploids and polyploids are equivalent. One study demonstrated that livers from rats that had undergone PH had increased polyploidy levels and changes associated with cell senescence and aging [37]. Another study showed that mice with drug‐induced necrosis and cirrhosis developed regenerating nodules enriched with diploid hepatocytes [38]. The idea that diploids are more proliferative than polyploid hepatocytes is also supported by observations that diploid hepatocytes are enriched in patients and rodents with hepatocellular carcinomas (HCC) [38–41]. In contrast, it was shown that transplanted octaploid and diploid hepatocytes proliferate equivalently in the Fah−/− liver repopulation model. However, a limitation of this study was that the ploidy states of the transplanted cells changed as the octaploid cells underwent ploidy reversal and the diploids polyploidized during the course of repopulation (Figure 47.1c) [17]. Taken together, these experiments illustrate that further studies are needed to determine the precise proliferative capacities of diploid and polyploid hepatocyte populations, and the mechanisms regulating proliferation of each ploidy population.

Polyploidy in aging and impaired liver regeneration Aging is known to lead to the accumulation of senescent cells in multiple tissues and organ systems, including the liver [42]. Senescence is the irreversible exit of the cell cycle, driven by the shortening of telomeres (i.e. the protective caps of chromosomes) through exhaustive replication or other stressors. Many hypothesize that senescence is a protective mechanism against tumorigenesis that might arise from genomic instability, but senescence also has the effect of limiting tissue regeneration and promoting the secretion of inflammatory mediators that can be damaging to surrounding tissue [43]. Aged octaploid cells were found to express a higher proportion of senescence markers compared to tetraploid and diploid cells, including p16, p21, and p53, which also serve important tumor suppressor roles [44]. Methods to deplete senescent cells in aged organs and ­tissues, such as the INK‐ATTAC mouse, were not capable of depleting senescent cells in the liver but reduced the spontaneous occurrence of liver tumors by depleting senescent cells at peripheral sites, thus demonstrating the role of the aged ­microenvironment in the etiology of liver cancers and other age‐ related diseases [42, 43].



47:  Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer

Unlike senescent diploid cells, which must be cleared by the immune system, senescent polyploid cells have the unique capacity to undergo ploidy reversal [44]. Aged hepatocytes that undergo ploidy reversal also downregulate expression of senescence markers, acting as a type of cell rejuvenation [44]. Ploidy reversal might partially explain the capacity of the aged liver to regenerate, although regeneration is substantially reduced in elderly individuals. After PH in mice, hepatocytes in aged livers have delayed cell cycle entry and the livers are significantly deficient in the number of proliferating hepatocytes [45–47]. The age‐related decline in hepatocyte proliferation has been attributed to transcription factor C/EBPα complexing with Brm, a chromatin remodeling protein detected only in aged livers [48], which inhibits E2f‐regulated gene expression [49]. Notably, when the circulatory systems of young and aged mice are connected through heterochronic parabiosis, the number of hepatocytes undergoing proliferation increased in the aged livers. While the ploidy of the rejuvenated hepatocytes was not documented, a reduction in the C/EBPα–Brm complexes was observed [48]. When aged diploid or octaploid hepatocytes are implanted into young Fah−/− mice, the octaploid hepatocytes undergo ploidy reversal, giving rise to lower hepatic ploidy states that no longer express markers of senescence [17, 44]. This study not only implicated the ability of a young systemic environment to rejuvenate hepatocyte proliferative capacity, but also suggested that hepatocytes with differential ploidy states have equivalent proliferation kinetics during long‐term repopulation. Future studies will need to decipher the role of cell intrinsic factors, such as ploidy and gene stability, mitochondrial function, and autophagy, versus extrinsic factors in the local or systemic environment, in the process of hepatocyte function and regeneration after injury and during aging.

THE MITOCHONDRIA‐PLOIDY NEXUS IN LIVER DISEASE Liver regeneration and a host of liver diseases are associated with hepatic ploidy alterations and mitochondrial metabolic dysfunction [50]. These observations suggest that the regulation of energy expenditure or the breakdown of mitochondrial function is linked to the development of differential ploidy states. This section describes the metabolic states and molecular regulators that connect mitochondrial metabolism and polyploidy in the liver.

Mitochondrial metabolism and metabolic disease Mitochondria serve an important role in managing energy expenditure of the liver, through metabolism of lipids and other molecules for ATP generation, which is essential for many cell functions. Dysregulated mitochondrial metabolism has been reported in liver diseases such as drug‐induced liver injury, alcoholic liver disease, non‐alcoholic fatty liver disease (NAFLD), viral hepatitis, liver cancer, and hemochromatosis [50, 51]. Mitochondrial dysfunction can manifest in several ways,

607

including disrupted mitochondrial membranes, dysregulation of oxidative phosphorylation (OXPHOS), production of reactive oxygen species (superoxides, nitric oxides, and lipid peroxides), glutathione depletion, release of cytochrome c from the mitochondrial matrix, defects in mitochondrial fission and fusion, abnormal mitophagy, and defective retrograde or anterograde signaling between the mitochondria and nucleus [50]. Insulin resistance and hyperlipidemia are the major underlying metabolic states of NAFLD [52]. Insulin signaling fortifies the integrity of the electron transport chain in the mitochondria by suppressing FOXO1 and maintaining the NAD+ : NADH ratio [53]. This NAD+ : NADH ratio then determines the activity of PGC1α, which is the major regulator of mitochondrial biogenesis [54]. Desdouets and colleagues recently showed that advanced fatty liver disease, known as non‐alcoholic steatohepatitis, is associated with enhanced polyploidy [55]. Livers from mice fed with diets that induce NAFLD, such as the methionine‐ choline‐deficient diet or high‐fat diet, were also dramatically enriched with polyploid hepatocytes. Furthermore, in these fatty liver models, oxidative stress was found to be the primary driver of increased polyploidy. Thus, there appears to be an intricate nexus between mitochondrial dysfunction, correlating with metabolic liver disease and polyploidization.

The nexus of hepatic polyploidy and mitochondrial metabolism Liver regeneration and diseases associated with altered mitochondrial metabolism, including NAFLD and hemochromatosis, are associated with alterations in ploidy [37, 55–57]. Moreover, many genes that regulate mitochondrial metabolism are reported to alter hepatic polyploidy such as Mir122, members of the E2F family (E2F1, E2F7, and E2F8), Birc5, Ercc1, Myc, p53, Rb, and Skp2 [2, 19, 58, 59]. Table 47.1 summarizes the dual role of key molecular regulators involved in mitochondrial metabolism and hepatic polyploidy. Together, these observations strongly suggest that mitochondrial metabolism and hepatic polyploidy are closely linked and could play a role in progression of liver diseases.

Mitochondria–ploidy nexus in liver regeneration Hepatic polyploidy has been reported to transiently increase when the liver regenerates following PH, during which concurrent changes occur in mitochondrial metabolism [37, 56]. First, there is a strong positive correlation between the regenerative capacity of the liver and the efficiency of OXPHOS, as demonstrated by a subsequent increase in ATP levels and respiratory control ratio, an indicator of OXPHOS efficiency [84]. Second, expression of proteins involved in carbohydrate metabolism, lipid metabolism, and OXPHOS increases 24 hours post PH [85]. Third, there is heterogeneity among the type of mitochondrial populations in hepatocytes. These mitochondrial subtypes differ in iron uptake capacity, iron metabolism, and complex IV activity, ranging from 2–42 days post PH [86]. In addition, oxidative stress is associated with liver regeneration, and this could be caused, at least in part, by regeneration‐associated OXPHOS [56]. Collectively, these reports highlight the heavy demand for

Table 47.1  The nexus of hepatic polyploidy and mitochondrial metabolism Gene

Model

Hepatic ploidy status

Relevance to mitochondrial metabolism

E2f7, E2f8

Deletion via knockout Deletion via knockout

Reduced polyploidy [60, 61] Increased polyploidy [61]

Mir122

Knockout

Reduced polyploidy [11]

Myc

Overexpression via transgenic mouse Deletion via knockout

Increased polyploidy [58]

Trp53

Deletion via knockout

Increased polyploidy [58]

Rb1

Deletion via knockout Deletion via knockout

Increased polyploidy [58] Increased polyploidy [58]

Birc5 (Survivin)

Deletion via knockout

Increased polyploidy [58]

Ercc1

Deletion via knockout

Increased polyploidy [58]

The E2F family is known to regulate cell proliferation genes where E2F7 and E2F8 cooperate to repress E2F1 expression and function [62]. 1. E2F1 regulates metabolism in organs such as liver, muscle, pancreas, and adipose tissue that determine metabolic homeostasis. Target genes of E2F1 are responsible for lipid synthesis (FAS), oxidative metabolism (TOP1MT, EVOVL2, NANOG), and glycolysis (PFKB, Sirt6, PDK) [63] 2. E2F7 and E2F8‐mediated regulation of E2F1 also occurs in metabolic scenarios such as oxidative stress induced DNA damage response [64] Mir122 deficiency has been reported in diseases with dysregulated mitochondrial metabolism such as HCC [65] and NAFLD. 1. Mice lacking miR‐122 have a severe defect in lipid metabolism [66] 2. miR‐122 regulates the expression of PGC1α, which is a master regulator of mitochondrial biogenesis [67] 3. Other targets of miR‐122, such as LCMT1, PPP1CC, ATF4, MEKK3, and MAPKAP2, are believed to regulate the expression of PGC1α [67] 4. Another direct target of miR‐122, pyruvate kinase (PKM2) [68], is a key regulator of glycolysis and, thus, can influence mitochondrial metabolism Myc provides the clearest example of programmed expansion of mitochondrial content linked to the cell cycle [69] 1. Myc regulates the expression of TFAM [70], a major determinant of mitochondrial DNA replication and maintenance, as well as PGC‐1α and PGC‐1β, which are prime regulators of mitochondrial mass and energy metabolism 2. Myc targets include genes involved in bi‐genomic and mitochondrial transcription, mitochondrial translation, protein import, and ETC complex assembly [71] 3. Myc also indirectly regulates mitochondrial gene expression via repression of microRNAs controlling nuclear genes encoding mitochondrial proteins (e.g. miR‐23a/b and glutaminase) [72, 73] 1. p53 transcriptionally represses glucose transporters GLUT1 and GLUT4 along with the insulin receptor (IR) to inhibit cellular glucose uptake into the cell [74] 2. Through transcriptional activation of TP53‐induced glycolysis and apoptosis regulator (TIGAR), p53 decreases the rate of glycolysis and redirects glycolytic intermediates into the pentose phosphate pathway (PPP) [74] 3. Glycolysis is also dampened by negative regulation of phosphor‐glycerate‐mutase (PGM) by p53. Transcriptional activation of hexokinase II (HK II) by mutant p53 stimulates glycolysis [74] 4. p53 promotes OXPHOS through transcriptional activation of synthesis of cytochrome c oxidase 2 (SCO2), a regulator of complex IV, and apoptosis‐inducing factor (AIF), which acts directly on complex I. By regulating transcription and stability of ribonucleotide reductase subunit (p53R2), p53 maintains mitochondrial homeostasis and mitochondrial genome integrity [74] 5. p53 is able to transcriptionally regulate and interact with the nuclear encoded mitochondrial transcription factor A (TFAM) and plays a role in mitochondrial DNA (mtDNA) transcription and in regulating mtDNA content [74] 6. Genotoxic stress signals trigger cytoplasmic p53 to undergo MDM2‐dependent mono‐ubiquitination that induces translocation of p53 to the mitochondria, which can trigger mitochondrial dependent apoptosis [74] 7. Mitochondrial p53 has been demonstrated to block the antioxidant function of manganese superoxide dismutase (MnSOD), thus creating a state of oxidative stress [74] 1. Proteomic effects of Rb1 ablation reveal a striking decrease in multiple mitochondrial proteins [75] 2. Rb1‐deficient cells have decreased mitochondrial mass and reduced OXPHOS activity [75] One of the prime targets of Skp2 ubiquitination‐mediated degradation is mitochondrial sirtuin 3 (SIRT3) [76], a histone NAD+ deactylase 1. SIRT3 controls the flow of mitochondrial oxidative pathways and, consequently, the rate of production of reactive oxygen species [77] 2. SIRT3 controls energy demand during stress conditions (fasting and exercise) and has the ability to quench reactive oxygen species [77] 3. SIRT3 targets enzymes involved in energy metabolism processes, including the respiratory chain, tricarboxylic acid cycle, fatty acid β‐oxidation and ketogenesis [77] Proteomic analysis of Birc5‐deleted livers reveal overexpression of mTOR [78], which is a key player in the mitochondrial metabolism [79]. This axis of BIRC5 and mTOR can possibly act as a node between ploidy and mitochondrial metabolism. 1. mTOR responds to a number of metabolic stimuli such as insulin, insulin growth factor, nutrients (amino acids), energy status, and oxygen levels and, thus, regulates proliferation [80] 2. mTOR also inhibits mitochondrial autophagy [81]. Independent of mTOR, BIRC5 regulates mitochondrial fission and complex I activity in neuroblastoma cell lines [82], suggesting that it might be involved in determining mitochondrial metabolism in the liver as well Ercc1 deficiency leads to increased oxidative damage as well as decreased levels of genes that are responsible for oxidative metabolism in the liver [83].

E2f1

Skp2

c47.indd 608

Reduced polyploidy [58]

03-11-2019 19:52:11



47:  Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer

energy during liver regeneration and underscore an essential role for mitochondrial metabolism in this process [87]. Additional studies are required to elucidate the mechanisms that connect liver regeneration with mitochondrial metabolism and ploidy.

PLOIDY AND LIVER CANCER Polyploidy has been considered a hallmark of cancer for nearly a century [88–90]. In the liver, however, some have suggested that polyploidy serves as a protective mechanism by providing multiple copies of tumor suppressor genes in the event that one is damaged [91]. Many forms of cancer exhibit polyploidy at both early and late stages of disease (e.g. renal cell carcinoma, myeloid leukemia, glioblastoma, pre‐esophageal cancer, and lung cancer [92]), but in the context of HCC a reduction of ploidy has been observed. Although HCC is associated with chromosomal variations [93], the relationship between polyploidy, HCC development and progression, and HCC drug resistance is poorly understood.

Polyploidy and liver cancer HCC is the most common form of liver cancer and is associated with high mortality and morbidity worldwide [94]. Studies conducted in the 1990s and early 2000s of patients with HCC revealed HCC tumors contained more diploid cells than polyploids [94] (Figure 47.3a). Consistent with these observations, work conducted on rats exposed to diethylnitrosamine (DEN) or 2‐acetyl‐aminofluorene indicated that the cells within the

609

diseased, HCC tissue became predominately diploid [41, 95, 96]. It is unclear if this shift from polyploid to diploid cells after malignant HCC transformation is a cause or effect in pathogenesis, but a potential explanation for the enrichment of diploid hepatocytes in HCCs is that polyploid hepatocytes protect from oncogenic transformation. This concept may seem counterintuitive, however, because polyploidy is traditionally associated with increased disease severity in other cancers [92]. Recent studies support the idea that polyploidy plays a protective role in the liver. Using a tunable system for altering hepatic ploidy in vivo, Zhu and colleagues showed that Anln deficient mice with highly polyploid livers were protected from HCC in the DEN tumor initiation model, whereas E2f8 deficient mice with predominately diploid livers were predisposed to HCC formation [97]. Most importantly, these studies revealed that the polyploid state protects against transformation by providing extra tumor suppressor alleles. With two homologous chromosomes per cell, diploid hepatocytes have two alleles of each tumor suppressor. If one tumor suppressor allele is mutated and inactivated, the cell is protected by only one functional allele. However, the cell is highly susceptible to tumorigenesis if the second allele is subsequently inactivated. In contrast, polyploid hepatocytes have four or more homologous chromosomes and a corresponding number of tumor suppressor alleles. When a single tumor suppressor is inactivated in a polyploid hepatocyte, the cell is protected or “buffered” by three or more functional alleles. Thus, even if a second mutation event inactivates a remaining allele, the cell still retains a minimum of two functional alleles and can maintain tumor suppressor activity. Taken together, studies in patients and rodent models strongly support the notion that diploid hepatocytes are more susceptible to oncogenic

Figure 47.3  Putative role of hepatic ploidy populations in liver cancer. Diploid hepatocytes are enriched in human HCC and are speculated to have enhanced proliferative potential. Polyploid hepatocytes are reduced in HCC and may have reduced proliferative potential. Recent studies indicate that polyploid hepatocytes protect against oncogenesis induced by tumor suppressor loss/inactivation; wild‐type copies of the tumor suppressor found on additional chromosomes serve as “back‐ups” that are capable of restricting oncogenic proliferation.

610

THE LIVER:  CONCLUSION

transformation, while polyploid hepatocytes provide a protective mechanism against HCC formation (Figure 47.3a).

Polyploidy and drug resistance in cancer Increased polyploidy in cancers is associated with poor prognosis primarily driven by an association with spontaneous drug resistance [92]. In colon cancer cell lines, for instance, polyploidy is attributed to increased resistance to DNA damage caused by camptothecin, cisplatin, oxaliplatin, gamma irradiation, and ultraviolet irradiation [98, 99]. Observations supporting a link between liver ploidy and drug sensitivity have also been made in in vivo studies. For example, the livers of Mir122 liver‐specific knockout mice are predominantly diploid, and when challenged with acetaminophen, Mir122 knockouts are more susceptible to liver damage compared to wild‐type mice with polyploid hepatocytes [100]. However, the data are difficult to interpret as Mir122 knockout mice express elevated levels of cytochrome P450 family 2 subfamily E member 1 (CYP2E1) and cytochrome P450 family 1 subfamily A member 2 (CYP1A2), which metabolize acetaminophen to its toxic byproduct N‐acetyl‐p‐benzoquinone imine (NAPQI) [100]. These findings indicate that the relationship between ploidy and drug resistance in the liver needs to be further investigated. Chemoresistance is a major concern as it may lead to inappropriate treatment regimens, poor disease prognosis, and decreased quality of life for patients with liver cancer. In the clinic, the multi‐tyrosine kinase inhibitor sorafenib and the topoisomerase II inhibitor doxorubicin are used to treat patients with unresectable HCC [101]. Unfortunately, chemotherapeutic intervention is minimally effective and frequently leads to tumors adapting and becoming refractory to the agent(s) administered [101]. The mechanisms by which HCC becomes chemoresistant are poorly understood, but in other cancers, polyploidy has been associated with drug resistance [92]. The shift in human HCC from polyploid to diploid cell predominance within the diseased tissue indicates that chemoresistance in liver cancer may not be derived from polyploidy but rather by other mechanisms. Overall, further studies are necessary to determine how polyploidy affects disease severity or drug resistance, which may lead to the development of novel, innovative, and effective therapeutics.

Aneuploidy and liver cancer Aneuploidy and chromosomal mis‐segregations have been linked to cancer and neoplastic transformation for almost one hundred years [88]. Poor disease prognosis and spontaneous drug resistance are strongly linked to aneuploidy and chromosomal instability as these events can contribute to cancer cell survival (Figure 47.3b) [92]. Although polyploidy and aneuploidy involve chromosomal alterations, it is critical to understand the difference between the two when considering disease. Polyploidy is a whole number increase in complete chromosome sets in a euploid cell. Depending on the context, aneuploidy can include whole chromosome gains and losses or structural changes in individual chromosomes. For example, these changes can include complete chromosomal rearrangements or even chromosome segments that are gained or lost. In the healthy adult liver of humans and mice, whole chromosome gains and losses are a normal feature [13, 17, 18, 22].

Previous work from Michalopoulos and colleagues found 461 copy number variations (CNVs) in 98 human HCC samples, demonstrating that chromosomal variations are common in liver cancer [93]. The CNVs identified in this study contained discrete genomic amplifications and deletions but not gains or losses of entire chromosomes. While aneuploidy in patient HCC tumors is not well studied, hepatoblastoma tumors have been extensively characterized and shown to contain near‐diploid or near‐tetraploid cells that have gained or lost entire chromosomes [102]. Additionally, underlying pathologies of hepatocellular carcinogenesis, such as chronic hepatitis B virus infection, aflatoxin exposure, and oxidative damage, are associated with chromosomal variations [103, 104]. It is largely accepted that increased aneuploidy, or chromosome instability (CIN), associates with worse disease prognosis [105]. Unexpectedly, patients with lung, breast, ovarian, and gastric cancers with the highest degree of CIN were reported to have better outcomes than patients with moderate CIN [105]. One potential explanation is that high‐degree CIN leads to mitotic catastrophe and cell death, which effectively kills tumors with the most aneuploidy. It is difficult to determine if aneuploidy promotes HCC cancer cell survival because experimentally‐induced aneuploidy and chromosomal rearrangements have produced ambiguous and conflicting results [105]. To study aneuploidy’s role in HCC development and progression in vitro and in vivo, researchers have used genotoxic agents or specifically tailored genetic models [105]. Recent work suggests that the relationship between aneuploidy and HCC can be studied by modulating miR‐122 levels in vitro and in vivo because Mir122 deficiency results in an increase in mononucleate hepatocytes with a concomitant reduction in aneuploidy [19]. To effectively investigate if aneuploidy contributes to HCC development, the ability to manipulate ploidy in cancer models is critical. Ultimately, the development of innovative strategies that can evade chemoresistance and eliminate tumors will depend upon determining the relationship between aneuploidy, polyploidy, and cancer cell survival.

CONCLUSION In adults, the majority of hepatocytes in the liver are polyploid. This suggests that polyploid cells are critical to the function and homeostasis of the liver. The accumulation of polyploid hepatocytes with age suggests they may serve as a reservoir for proliferation and adaptation in the case of damage and injury. Recent studies have shown that polyploid hepatocytes act as a critical reserve for cells of lower ploidy states, which can be generated by ploidy reversal. Also, polyploid hepatocytes appear to be more resistant to toxic liver injury and cancer, and they may play a discrete role in liver metabolism and regeneration. The molecular mechanisms affecting differential activity of diploid and polyploid hepatocytes are poorly understood, and a key question that remains is whether gene expression in hepatocytes is dictated by ploidy state. A better understanding of hepatic polyploidy will ultimately benefit patients in the future by driving the improvement and development of innovative strategies that prevent disease, stimulate liver regeneration, and advance artificial organ construction.



47:  Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer

ACKNOWLEDGMENTS This work was supported by grants to AWD from the NIH (R01 DK103645), the Commonwealth of Pennsylvania, and the University of Pittsburgh Physicians Academic Foundation and to ECS from the NIH (F31 DK112633) and the NIBIB training grant (T32 EB001026) entitled “Cellular Approaches to Tissue Engineering and Regeneration.” Thanks to Frances Alencastro for helpful discussions and critical reading of the manuscript. The authors of this chapter apologize to the authors whose works were not cited because of space restrictions.

REFERENCES 1. Orr‐Weaver, T.L. When bigger is better: the role of polyploidy in organogenesis. Trends Genet, 2015;31(6):307–15. 2. Duncan, A.W. Aneuploidy, polyploidy and ploidy reversal in the liver. Semin Cell Dev Biol, 2013;24(4):347–56. 3. Abmayr, S.M. and Pavlath, G.K. Myoblast fusion: lessons from flies and mice. Development, 2012;139(4):641–56. 4. Helming, L. and Gordon, S. Molecular mediators of macrophage fusion. Trends Cell Biol, 2009;19(10):514–22. 5. Edgar, B.A. and Orr‐Weaver, T.L. Endoreplication cell cycles: more for less. Cell, 2001;105(3):297–306. 6. Lee, H.O., Davidson, J.M., and Duronio, R.J. Endoreplication: polyploidy with purpose. Genes Dev, 2009;23(21):2461–77. 7. Engel, F.B., Schebesta, M., and Keating, M.T. Anillin localization defect in cardiomyocyte binucleation. J Mol Cell Cardiol, 2006;41(4):601–12. 8. Margall‐Ducos, G., Celton‐Morizur, S., Couton, D., Bregerie, O., and Desdouets, C. Liver tetraploidization is controlled by a new process of incomplete cytokinesis. J Cell Sci, 2007;120(20):3633–9. 9. Bohm, N. and Noltemeyer, N. Development of binuclearity and DNA‐polyploidization in the growing mouse liver. Histochemistry, 1981;72(1):55–61. 10. Celton‐Morizur, S., Merlen, G., Couton, D., and Desdouets, C. Polyploidy and liver proliferation: central role of insulin signaling. Cell Cycle, 2010;9(3):460–6. 11. Wang, X., Montini, E., Al‐Dhalimy, M., Lagasse, E., Finegold, M., and Grompe, M. Kinetics of liver repopulation after bone marrow transplantation. Am J Pathol, 2002;161(2):565–74. 12. Wilson, J.W. and Leduc, E.H. The occurrence and formation of binucleate and multinucleate cells and polyploid nuclei in the mouse liver. Am J Anat, 1948;82(3):353–91. 13. Duncan, A.W., Hanlon Newell, A.E., Smith, L. et al. Frequent aneuploidy among normal human hepatocytes. Gastroenterology, 2012;142(1):25–8. 14. Wang, M.J., Chen, F., Lau, J.T.Y., and Hu, Y.P. Hepatocyte polyploidization and its association with pathophysiological processes. Cell Death Dis, 2017;8(5):e2805. 15. Watanabe, T. and Tanaka, Y. Age‐related alterations in the size of human hepatocytes. A study of mononuclear and binucleate cells. Virchows Arch B Cell Pathol Incl Mol Pathol, 1982;39(1):9–20. 16. Celton‐Morizur, S., Merlen, G., Couton, D., Margall‐Ducos, G., and Desdouets, C. The insulin/Akt pathway controls a specific cell division program that leads to generation of binucleated tetraploid liver cells in rodents. J Clin Invest, 2009;119(7):1880–7. 17. Duncan, A.W., Taylor, M.H., Hickey, R.D. et al. The ploidy conveyor of mature hepatocytes as a source of genetic variation. Nature, 2010;467(7316):707–10. 18. Knouse, K.A., Wu, J., Whittaker, C.A., and Amon, A. Single cell sequencing reveals low levels of aneuploidy across mammalian tissues. Proc Natl Acad Sci USA, 2014;111(37):13409–14. 19. Hsu, S.H., Delgado, E.R., Otero, P.A. et al. MicroRNA‐122 regulates polyploidization in the murine liver. Hepatology, 2016;64(2):599–615. 20. Kudryavtsev, B.N., Kudryavtseva, M.V., Sakuta, G.A., and Stein, G.I. Human hepatocyte polyploidization kinetics in the course of life cycle. Virchows Arch B Cell Pathol Incl Mol Pathol, 1993;64(6):387–93. 21. Duncan, A.W., Hickey, R.D., Paulk, N.K. et al. Ploidy reductions in murine fusion‐derived hepatocytes. PLoS Genet, 2009;5(2):e1000385.

611

22. Faggioli, F., Vezzoni, P., and Montagna, C. Single‐cell analysis of ploidy and centrosomes underscores the peculiarity of normal hepatocytes. PloS One, 2011;6(10):e26080. 23. Duncan, A.W., Hanlon Newell, A.E., Bi, W. et al. Aneuploidy as a mechanism for stress‐induced liver adaptation. J Clin Invest, 2012;122(9):3307–15. 24. Galitski, T., Saldanha, A.J., Styles, C.A., Lander, E.S., and Fink, G.R. Ploidy regulation of gene expression. Science, 1999;285(5425):251–4. 25. Raslova, H., Kauffmann, A., Sekkai, D. et  al. Interrelation between polyploidization and megakaryocyte differentiation: a gene profiling approach. Blood, 2007;109(8):3225–34. 26. Corbel, C., Diabangouaya, P., Gendrel, A.V., Chow, J.C., and Heard, E. Unusual chromatin status and organization of the inactive X chromosome in murine trophoblast giant cells. Development, 2013;140(4):861–72. 27. Anatskaya, O.V. and Vinogradov, A.E. Genome multiplication as adaptation to tissue survival: evidence from gene expression in mammalian heart and liver. Genomics, 2007;89(1):70–80. 28. Lu, P., Prost, S., Caldwell, H., Tugwood, J.D., Betton, G.R., and Harrison, D.J. Microarray analysis of gene expression of mouse hepatocytes of different ploidy. Mamm Genom, 2007;18(9):617–26. 29. Rajvanshi, P., Liu, D., Ott, M., Gagandeep, S., Schilsky, M.L., and Gupta, S. Fractionation of rat hepatocyte subpopulations with varying metabolic potential, proliferative capacity, and retroviral gene transfer efficiency. Exp Cell Res, 1998;244(2):405–19. 30. Pavelka, N., Rancati, G., Zhu, J. et al. Aneuploidy confers quantitative proteome changes and phenotypic variation in budding yeast. Nature, 2010;468(7321):321–5. 31. Rancati, G., Pavelka, N., Fleharty, B. et al. Aneuploidy underlies rapid adaptive evolution of yeast cells deprived of a conserved cytokinesis motor. Cell, 2008;135(5):879–93. 32. Grompe, M., Lindstedt, S., al‐Dhalimy, M. et al. Pharmacological correction of neonatal lethal hepatic dysfunction in a murine model of hereditary tyrosinaemia type I. Nat Genet, 1995;10(4):453–60. 33. Manning, K., Al‐Dhalimy, M., Finegold, M., and Grompe, M. In vivo suppressor mutations correct a murine model of hereditary tyrosinemia type I. Proc Natl Acad Sci USA, 1999;96(21):11928–33. 34. Gupta, S. Hepatic polyploidy and liver growth control. Semin Cancer Biol, 2000;10(3):161–71. 35. Fausto, N. and Campbell, J.S. The role of hepatocytes and oval cells in liver regeneration and repopulation. Mech Dev, 2003;120(1):117–30. 36. Miyaoka, Y., Ebato, K., Kato, H., Arakawa, S., Shimizu, S., and Miyajima, A. Hypertrophy and unconventional cell division of hepatocytes underlie liver regeneration. Curr Biol, 2012;22(13):1166–75. 37. Sigal, S.H., Rajvanshi, P., Gorla, G.R. et  al. Partial hepatectomy‐induced polyploidy attenuates hepatocyte replication and activates cell aging events. Am J Physiol, 1999;276(5 Pt 1):G1260–72. 38. Gandillet, A., Alexandre, E., Royer, C., Cinqualbre, J., Jaeck, D., and Richert, L. Hepatocyte ploidy in regenerating livers after partial hepatectomy, drug‐induced necrosis, and cirrhosis. Eur Surg Res, 2003;35(3): 148–60. 39. Anti, M., Marra, G., Rapaccini, G.L. et  al. DNA ploidy pattern in human chronic liver diseases and hepatic nodular lesions. Flow cytometric analysis on echo‐guided needle liver biopsy. Cancer, 1994;73(2):281–8. 40. Rua, S., Comino, A., Fruttero, A. et  al. Flow cytometric DNA analysis of cirrhotic liver cells in patients with hepatocellular carcinoma can provide a new prognostic factor. Cancer, 1996;78(6):1195–202. 41. Schwarze, P.E., Pettersen, E.O., Shoaib, M.C., and Seglen, P.O. Emergence of a population of small, diploid hepatocytes during hepatocarcinogenesis. Carcinogenesis, 1984;5(10):1267–75. 42. Baker, D.J., Childs, B.G., Durik, M. et al. Naturally occurring p16(Ink4a)‐ positive cells shorten healthy lifespan. Nature, 2016;530(7589):184–9. 43. Sturmlechner, I., Durik, M., Sieben, C.J., Baker, D.J., and van Deursen, J.M. Cellular senescence in renal ageing and disease. Nat Rev Nephrol, 2017;13(2):77–89. 44. Wang, M.J., Chen, F., Li, J.X. et al. Reversal of hepatocyte senescence after continuous in vivo cell proliferation. Hepatology, 2014;60(1):349–61. 45. Bucher, N.L., Swaffield, M.N., and Ditroia, J.F. The influence of age upon the incorporation of thymidine‐2‐C14 into the DNA of regenerating rat liver. Cancer Res, 1964;24:509–12. 46. Fry, M., Silber, J., Loeb, L.A., and Martin, G.M. Delayed and reduced cell replication and diminishing levels of DNA polymerase‐alpha in regenerating liver of aging mice. J Cell Physiol, 1984;118(3):225–32.

612

THE LIVER:  REFERENCES

47. Timchenko, N.A., Wilde, M., Kosai, K.I. et al. Regenerating livers of old rats contain high levels of C/EBPalpha that correlate with altered expression of cell cycle associated proteins. Nucleic Acids Res, 1998;26(13):3293–9. 48. Conboy, I.M., Conboy, M.J., Wagers, A.J., Girma, E.R., Weissman, I.L., and Rando, T.A. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature, 2005;433(7027):760–4. 49. Iakova, P., Awad, S.S., and Timchenko, N.A. Aging reduces proliferative capacities of liver by switching pathways of C/EBPalpha growth arrest. Cell, 2003;113(4):495–506. 50. Grattagliano, I., Russmann, S., Diogo, C. et al. Mitochondria in chronic liver disease. Curr Drug Targets, 2011;12(6):879–93. 51. Du, K., Ramachandran, A., and Jaeschke, H. Oxidative stress during acetaminophen hepatotoxicity: sources, pathophysiological role and therapeutic potential. Redox Biol, 2016;10:148–56. 52. Lonardo, A., Ballestri, S., Marchesini, G., Angulo, P., and Loria, P. Nonalcoholic fatty liver disease: a precursor of the metabolic syndrome. Dig Liver Dis, 2015;47(3):181–90. 53. Cheng, Z., Tseng, Y., and White, M.F. Insulin signaling meets mitochondria in metabolism. Trends Endocrinol Metab, 2010;21(10):589–98. 54. Scarpulla, R.C. Metabolic control of mitochondrial biogenesis through the PGC‐1 family regulatory network. Biochim Biophys Acta, 2011;1813(7): 1269–78. 55. Gentric, G., Maillet, V., Paradis, V. et al. Oxidative stress promotes pathologic polyploidization in nonalcoholic fatty liver disease. J Clin Invest, 2015;125(3):981–92. 56. Gorla, G.R., Malhi, H., and Gupta, S. Polyploidy associated with oxidative injury attenuates proliferative potential of cells. J Cell Sci, 2001;114(16): 2943–51. 57. Troadec, M.B., Courselaud, B., Detivaud, L. et al. Iron overload promotes Cyclin D1 expression and alters cell cycle in mouse hepatocytes. J Hepatol, 2006;44(2):391–9. 58. Duncan, A.W. Changes in hepatocyte ploidy during liver regeneration, in Liver Regeneration: Basic Mechanisms, Relevant Models and Clinical Applications, (ed. U. Apte) Elsevier, 2015. 59. Pandit, S.K., Westendorp, B., Nantasanti, S. et al. E2F8 is essential for polyploidization in mammalian cells. Nat Cell Biol, 2012;14(11):1181–91. 60. Kent, L.N., Rakijas, J.B., Pandit, S.K. et al. E2f8 mediates tumor suppression in postnatal liver development. J Clin Invest, 2016;126(8):2955–69. 61. Pandit, S.K., Westendorp, B., and de Bruin, A. Physiological significance of polyploidization in mammalian cells. Trends Cell Biol, 2013;23(11):556–66. 62. Moon, N.S. and Dyson, N. E2F7 and E2F8 keep the E2F family in balance. Dev Cell, 2008;14(1):1–3. 63. Denechaud, P.D., Fajas, L., and Giralt, A. E2F1, a novel regulator of metabolism. Front Endocrinol, 2017;8:311. 64. Zalmas, L.P., Zhao, X., Graham, A.L. et al. DNA‐damage response control of E2F7 and E2F8. EMBO Rep, 2008;9(3):252–9. 65. Kutay, H., Bai, S., Datta, J. et al. Downregulation of miR‐122 in the rodent and human hepatocellular carcinomas. J Cell Biochem, 2006;99(3):671–8. 66. Hsu, S.H., Wang, B., Kota, J. et al. Essential metabolic, anti‐inflammatory, and anti‐tumorigenic functions of miR‐122 in liver. J Clin Invest, 2012;122(8):2871–83. 67. Burchard, J., Zhang, C., Liu, A.M. et  al. microRNA‐122 as a regulator of mitochondrial metabolic gene network in hepatocellular carcinoma. Mol Syst Biol, 2010;6:402. 68. Liu, A.M., Xu, Z., Shek, F.H. et al. miR‐122 targets pyruvate kinase M2 and affects metabolism of hepatocellular carcinoma. PloS One, 2014;9(1):e86872. 69. Morrish, F. and Hockenbery, D. MYC and mitochondrial biogenesis. Cold Spring Harb Perspect Med, 2014;4(5). 70. Ahuja, P., Zhao, P., Angelis, E. et al. Myc controls transcriptional regulation of cardiac metabolism and mitochondrial biogenesis in response to pathological stress in mice. J Clin Invest, 2010;120(5):1494–505. 71. Li, F., Wang, Y., Zeller, K.I. et al. Myc stimulates nuclearly encoded mitochondrial genes and mitochondrial biogenesis. Mol Cell Biol, 2005; 25(14):6225–34. 72. Dang, C.V., Le, A., and Gao, P. MYC‐induced cancer cell energy metabolism and therapeutic opportunities. Clin Cancer Res, 2009;15(21):6479–83. 73. Psathas, J.N. and Thomas‐Tikhonenko, A. MYC and the art of microRNA maintenance. Cold Spring Harb Perspect Med, 2014;4(8). 74. Itahana, Y. and Itahana, K. Emerging roles of p53 family members in glucose metabolism. Int J Mol Sci, 2018;19(3).

75. Nicolay, B.N., Danielian, P.S., Kottakis, F. et al. Proteomic analysis of pRb loss highlights a signature of decreased mitochondrial oxidative phosphorylation. Genes Dev, 2015;29(17):1875–89. 76. Iwahara, T., Bonasio, R., Narendra, V., and Reinberg, D. SIRT3 functions in the nucleus in the control of stress‐related gene expression. Mol Cell Biol, 2012;32(24):5022–34. 77. Hirschey, M.D., Shimazu, T., Huang, J.Y., Schwer, B., and Verdin, E. SIRT3 regulates mitochondrial protein acetylation and intermediary metabolism. Cold Spring Harb Symp Quant Biol, 2011;76:267–77. 78. Bracht, T., Hagemann, S., Loscha, M. et al. Proteome analysis of a hepatocyte‐specific BIRC5 (survivin)‐knockout mouse model during liver regeneration. J Proteome Res, 2014;13(6):2771–82. 79. Morita, M., Gravel, S.P., Hulea, L. et al. mTOR coordinates protein synthesis, mitochondrial activity and proliferation. Cell Cycle, 2015;14(4):473–80. 80. Laplante, M. and Sabatini, D.M. mTOR signaling in growth control and disease. Cell, 2012;149(2):274–93. 81. Kim, J., Kundu, M., Viollet, B., and Guan, K.L. AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nat Cell Biol, 2011;13(2):132–41. 82. Hagenbuchner, J., Kuznetsov, A.V., Obexer, P., and Ausserlechner, M.J. BIRC5/Survivin enhances aerobic glycolysis and drug resistance by altered regulation of the mitochondrial fusion/fission machinery. Oncogene, 2013;32(40):4748–57. 83. Gregg, S.Q., Gutierrez, V., Robinson, A.R. et al. A mouse model of accelerated liver aging caused by a defect in DNA repair. Hepatology, 2012;55(2):609–21. 84. Guerrieri, F., Muolo, L., Cocco, T. et  al. Correlation between rat liver regeneration and mitochondrial energy metabolism. Biochim Biophys Acta, 1995;1272(2):95–100. 85. Sun, Q., Miao, M., Jia, X. et al. Subproteomic analysis of the mitochondrial proteins in rats 24 h after partial hepatectomy. J Cell Biochem, 2008;105(1):176–84. 86. Gear, A.R. Observations on iron uptake, iron metabolism, cytochrome c content, cytochrome a content and cytochrome c‐oxidase activity in regenerating rat liver. Biochem J, 1965;97(2):532–9. 87. Huang, J. and Rudnick, D.A. Elucidating the metabolic regulation of liver regeneration. Am J Pathol, 2014;184(2):309–21. 88. Boveri, T. Zur Frage der Entstehung maligner Tumoren (The Origin of Malignant Tumors), Jena, Gustav Fischer, 1914. 89. Wolff, J. The Theory of Cancer, Science History Publishers, Sagamore Beach, 1907. 90. Duensing, A. and Duensing, S. Centrosomes, polyploidy and cancer, in Polyploidization and Cancer, (ed. R.Y.C. Poon), Springer, New York, 2010, pp. 93–103. 91. Nowell, P.C. The clonal evolution of tumor cell populations. Science, 1976;194(4260):23–8. 92. Merlo, L.M., Wang, L.S., Pepper, J.W., Rabinovitch, P.S., and Maley, C.C. Polyploidy, aneuploidy and the evolution of cancer. Adv Exp Med Biol, 2010;676:1–13. 93. Nalesnik, M.A., Tseng, G., Ding, Y. et al. Gene deletions and amplifications in human hepatocellular carcinomas: correlation with hepatocyte growth regulation. Am J Pathol, 2012;180(4):1495–508. 94. Nagasue, N., Kohno, H., Chang, Y.C. et al. DNA ploidy pattern in synchronous and metachronous hepatocellular carcinomas. J Hepatol, 1992; 16(1–2):208–14. 95. Saeter, G., Schwarze, P.E., Nesland, J.M., Juul, N., Pettersen, E.O., and Seglen, P.O. The polyploidizing growth pattern of normal rat liver is replaced by divisional, diploid growth in hepatocellular nodules and carcinomas. Carcinogenesis, 1988;9(6):939–45. 96. Schwarze, P.E., Saeter, G., Armstrong, D. et al. Diploid growth pattern of hepatocellular tumours induced by various carcinogenic treatments. Carcinogenesis, 1991;12(2):325–7. 97. Zhang, S., Zhou, K., Luo, X. et al. The polyploid state plays a tumor‐suppressive role in the liver. Dev Cell, 2018;44(4):447–59 e5. 98. Castedo, M., Coquelle, A., Vitale, I. et al. Selective resistance of tetraploid cancer cells against DNA damage‐induced apoptosis. Ann N Y Acad Sci, 2006;1090:35–49. 99. Castedo, M., Coquelle, A., Vivet, S. et al. Apoptosis regulation in tetraploid cancer cells. EMBO J, 2006;25(11):2584–95. 100. Chowdhary, V., Teng, K.Y., Thakral, S. et al. miRNA‐122 protects mice and human hepatocytes from acetaminophen toxicity by regulating cytochrome



47:  Polyploidy in Liver Function, Mitochondrial Metabolism, and Cancer

P450 family 1 subfamily A member 2 and family 2 subfamily E member 1 expression. Am J Pathol, 2017;187(12):2758–74. 101. Nguyen, K., Jack, K., and Sun, W. Hepatocellular carcinoma: past and future of molecular target therapy. Diseases, 2016;4(1):1. 102. Weaver, B.A. and Cleveland, D.W. Does aneuploidy cause cancer? Curr Opin Cell Biol, 2006;18(6):658–67. 103. Turner, P.C., Sylla, A., Diallo, M.S., Castegnaro, J.J., Hall, A.J., and Wild, C.P. The role of aflatoxins and hepatitis viruses in the etiopathogenesis of hepatocellular carcinoma: a basis for primary prevention in

613

Guinea‐Conakry, West Africa. J Gastroenterol Hepatol, 2002;17 Suppl:S441–8. 104. Slagle, B.L., Zhou, Y.Z., and Butel, J.S. Hepatitis B virus integration event in human chromosome 17p near the p53 gene identifies the region of the chromosome commonly deleted in virus‐positive hepatocellular carcinomas. Cancer Res, 1991;51(1):49–54. 105. Zasadil, L.M., Britigan, E.M., and Weaver, B.A. 2n or not 2n: aneuploidy, polyploidy and chromosomal instability in primary and tumor cells. Semin Cell Dev Biol, 2013;24(4):370–9.

PART FOUR: PATHOBIOLOGY OF LIVER DISEASE

48

Hepatic Encephalopathy Roger F. Butterworth Department of Medicine, University of Montreal, Montreal, Canada

INTRODUCTION Hepatic encephalopathy (HE) is a serious central nervous ­system (CNS) complication of liver diseases that is characterized by a range of neurological and neuropsychiatric symptoms that include perturbations of psychomotor, neuro‐cognitive, and fine motor function. Deficits in attention, visual perception, visuo‐spatial construction in addition to motor speed and ­accuracy are among the early symptoms of HE [1]. These symptoms are generally considered to be potentially reversible. The trans‐jugular intrahepatic portosystemic shunt (TIPS) procedure is effective for the prevention and treatment of complications of cirrhosis such as portal hypertension, gastrointestinal (variceal) bleeding, and refractory ascites. However, the procedure may result in new or worsening episodes of encephalopathy in up to 50% of patients [2]. The burden of cirrhosis and HE is multidimensional and is increasing. Direct economic costs related to hospitalizations, medical procedures, and medications are substantial and to these must be added the economic cost to the patient resulting from lost income. Cognitive dysfunction is associated with worsening financial status, employment prospects, and q­ uality of life with consequent burden to the patient’s family and caregivers [3].

NEW CLASSIFICATION OF HE SYNDROMES The current practice guidelines provide a system of classification according to the nature of the underlying disease. In this way, HE is subdivided into three major types: Type A: HE associated with acute liver failure Type B: HE associated with portosystemic bypass/shunting Type C: HE associated with cirrhosis

Further subdivisions then occur as a function of the time course of HE as depicted in Table 48.1. Episodic HE Recurrent HE characterized by bouts of HE that occur within a period of six months or less Permanent HE in which symptoms are unremitting

West Haven criteria for grading of HE severity In patients with cirrhosis, clinically symptomatic (overt) HE (OHE) heralds the decompensated phase of the disease and grading of the severity of OHE has traditionally made use of a battery of neuropsychiatric tests known as the West Haven criteria where grade I characterized by shortened attention span and sleep disorders is followed by changes in personality, disorientation, and asterixis (grade II). Grade III is typified by disorientation, confusion, and semi‐stupor followed by the comatose state (grade IV). If required, assessment of the depth and duration of coma/ impaired consciousness can be made using the Glasgow coma scale, a points scale based upon motor responsiveness, verbal performance, and eye opening in response to stimuli. Minimal HE (MHE), a subtype of type C HE in which clinically‐manifest symptoms are absent, occurs in up to 80% of patients with cirrhosis and has a significant impact on health‐ related quality of life (HRQOL) of the patient [3]. MHE is diagnosed by one or more of a series of well‐established psychometric tests. MHE is associated with increased numbers of falls and road traffic accidents [3, 4]. In recent times, largely due to the inherent subjective nature of the neuropsychiatric symptoms defining grade I HE by West Haven, efforts were made to modify the nomenclature and classification of HE. In the resulting AASLD/EASL guidelines, a new term pertinent to HE classification, namely covert HE (CHE), has appeared defined as MHE together with what was formerly grade I HE [4]. This change was

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

618

THE LIVER:  NEUROPATHOLOGY

accompanied by a streamlining of the diagnosis of overt HE into grades II, III, and IV as shown in Table 48.1.

CLINICAL PRESENTATION AND DIAGNOSIS HE is a diagnosis of exclusion of other metabolic disorders, infectious diseases, intracranial vascular events, and intracranial space‐occupying lesions all of which may potentially result in neuropsychiatric symptomatology [5]. Covert HE is accompanied by alterations of performance in psychometric tests directed toward attention, working memory, psychomotor speed, and visuospatial ability. Overt HE is considered to start with asterixis or flapping tremor [6] and staging of severity in OHE by West Haven. Identification of well‐established precipitating factors for HE such as infection, gastrointestinal bleeding, and constipation may support the diagnosis of HE in cirrhosis. For patients with an evidently altered state of consciousness, the Glasgow coma scale is widely employed. Despite strong evidence for a role of ammonia in the pathogenesis of HE, measurements of venous ammonia add little information by way of diagnostic or prognostic value in patients with cirrhosis. On the other hand, arterial ammonia concentrations Table 48.1  Updated nomenclature of hepatic encephalopathy (HE) syndromes is dependent upon the type of HE (A, B, or C), the grade of HE (minimal, I, II, III, IV) that is now subdivided into covert (a combination of MHE and grade I overt) and overt (II, III, and IV). Depending upon the time course, HE is then rated as episodic, recurrent, or permanent and may fall into one of two categories namely spontaneous or precipitated Type A B C

HE grade Minimal I II III IV

(a)

Time course Covert Overt

Episodic Recurrent Persistent

may have prognostic value for prediction of the severe neurological complications of acute liver failure (ALF) [7]. Grading of HE in patients with cirrhosis using West Haven criteria has been criticized due to its inherent subjective nature with the likelihood to lead to high inter‐examiner variability. One alternative technique that is gaining in popularity makes use of critical flicker frequency (CFF), a quantitative measure based upon the correlation between cerebral processing of oscillatory visual stimuli impairment and increasing severity of HE. The CFF procedure is used extensively for the grading of neurocognitive changes in a variety of neurological disorders that includes low‐grade HE in cirrhosis where a CFF cut‐off of 39 Hz differentiates low‐grade overt HE from non‐HE patients [8].

NEUROPATHOLOGY Glial pathology Glial cells manifest the most conspicuous and reproducible morphological and functional changes related to HE in both ALF and in cirrhosis. However, the nature and extent of these changes are determined by the type of liver injury (acute versus chronic), the extent of portal‐systemic shunting, and the number and duration of the episodes of HE. Glial pathology in HE involves one of two major types of cells, namely astrocytes and microglia.

Astrocytes Pathological evaluation of brain sections from patients with end‐stage ALF reveals cytotoxic brain edema as shown in Figure 48.1a where the astrocyte manifests massive swelling of an astrocyte end foot (A) [9]. The astrocyte swelling has the potential to result in the development of intracranial hypertension leading ultimately to brain herniation which remains one of the major causes of mortality in ALF. In contrast to ALF, the cardinal neuropathological feature of decompensated cirrhosis is a characteristic morphological (b)

Figure 48.1  Characteristic ultrastructural changes in the brain in acute and in chronic liver failure. (a) Electron micrograph from a patient with acute liver failure due to acetaminophen overdose showing cytotoxic brain edema with marked swelling of perivascular astrocyte (A), and dilatation of endoplasmic reticulum (arrows) and mitochondria (M). (b) Electron micrograph from a 51‐year‐old cirrhotic patient who died in hepatic coma presenting Alzheimer type II astrocytes with a large, pale nucleus and margination of chromatin (Alz) compared with normal astrocytes with normal chromatin pattern (N).



48:  Hepatic Encephalopathy

alteration of the astrocyte nucleus known as Alzheimer type 2 astrocytosis [10]. The Alzheimer type 2 phenotype consists of nuclear pallor and swelling of both the cell itself and its nucleus. Margination of the chromatin pattern and deposition of glycogen is also characteristic of the Alzheimer type 2 change as depicted in Figure  48.1b. Such changes are unevenly distributed in brain with particularly high densities observed in cerebral cortex, basal ganglia, and cerebellum [11]. It is important to note that there is also evidence of brain edema in cirrhosis although, in contrast to the situation in ALF, the edema in cirrhosis is low grade in nature, rarely leading to alterations of intracranial pressure. Low‐grade edema in cirrhosis has been demonstrated in basal ganglia and corticospinal tract of patients with cirrhosis using magnetic resonance imaging (MRI) where decreases of magnetization transfer ratios and increases of water apparent diffusion coefficients were reported [12].

Microglia Microglia are the immunomodulatory cells of the brain and results of studies in animal models of acute or chronic liver failure as well as in material from patients with these conditions confirm that activation of microglia is a common occurrence in HE [13]. Activation of microglia is indicative of neuroinflammation and is observed in a wide range of neurodegenerative CNS disorders including HE. Neuroinflammation associated with microglial activation is currently considered to be a feature of HE in both acute and chronic liver diseases [13,14]. This issue is discussed further in a subsequent section on Neuroinflammation in this chapter.

Neuronal pathology Although the principal neuropathological change typical of HE in cirrhosis is primarily glial in nature, it is important to bear in mind that neuronal cell death frequently also occurs. Changes of neuronal morphological and functional characteristics are not uncommon [15]. These changes occur in distinct brain structures and have been attributed to a spectrum of diverse pathophysiologic mechanisms. The neuropathological and clinical characteristics of these neurodegenerative disorders are summarized in the following sections.

Acquired non‐Wilsonian hepatocerebral degeneration (ANHD) ANHD occurs in patients with cirrhosis following a protracted clinical course often with multiple episodes of coma and evidence of portal‐systemic shunting. The neuropathology consists of spongiform degeneration in deep cerebrocortical layers as well as in basal ganglia structures, cerebellum, and subcortical white matter.

Wernicke’s encephalopathy Unsuspected Wernicke‐type hemorrhagic lesions in thalamus occur in up to 30% of patients with end‐stage cirrhosis of alcoholic etiology [11] where lesions may be acute or chronic (longstanding). The higher incidence of Wernicke‐type lesions in these patients compared to noncirrhotic alcoholics likely relates

619

to the fact that liver is a key site for thiamine synthesis and storage in humans and Wernicke’s encephalopathy results solely from thiamine deficiency.

Cerebellar degeneration Mild to severe cerebellar degeneration characterized by lesions in cerebellar vermis and severe loss of Purkinje cells occurs in both alcoholic and non‐alcoholic patients with end‐stage cirrhosis although the severity of neuropathologic changes is more severe in the alcoholics [11]. There are no clear correlations between the incidence and extent of cerebellar degeneration, Wernicke‐type thalamic lesions or Alzheimer type 2 astrocytosis suggesting that these neuropathologic phenotypes are underpinned by distinct mechanisms.

Post‐shunt myelopathy Although rare, myelopathy may occur in patients with cirrhosis following multiple episodes of coma or following the TIPS procedure. Symptoms of spastic paresis or paralysis of the lower limbs are apparent and result from demyelination of both direct and crossed corticospinal tracts in this condition.

Parkinsonism in cirrhosis Characterized by extrapyramidal symptoms (hypokinesia, tremor, rigidity) the condition is rapidly progressive and may occur independent of the extent of cognitive dysfunction. With a prevalence as high as 21% in one report of patients with cirrhosis listed for transplantation [16], the disorder has been attributed to dopaminergic neuronal deficits resulting from manganese deposition in basal ganglia [17]. Extrapyramidal symptoms in these patients may respond to liver transplantation and, in some cases to L‐DOPA therapy. One useful procedure that is available for confirmation of diagnosis of Parkinsonism in cirrhosis ­consists of T1‐weighted MRI where bilateral signal hyperintensities are observed in globus pallidus and substantia nigra. Further discussion relating to brain manganese deposition in cirrhosis appears in the Manganese section of the current chapter.

Global brain reserve A novel concept composed of a structural component (brain reserve [BR]) together with a patient’s ability to tolerate changes (cognitive reserve), BR is dependent upon disease etiology, severity, and progression. Decreased BR examined by magnetic resonance techniques is related to three factors namely ­structural changes in white and gray matter of multiple brain structures, brain edema, and altered brain metabolism. Results of a multi‐modal MRI study reveal that patients with alcoholic cirrhosis have, despite abstinence, a poor BR whereas patients with non‐alcoholic cirrhosis have a greater potential for deterioration of BR following HE or TIPS. Enhancement of the direct effects of chronic alcohol on brain (Wernicke’s encephalopathy, cerebellar degeneration) or indirectly via the liver resulting in ANHD could form part of the basis for the structural changes characteristic of BR. Worsening of BR as a c­ onsequence of alcoholic etiology of cirrhosis (despite abstinence) has the potential to modulate the impact of HE on disease p­ rogression and suitability for transplantation [18].

620

THE LIVER:  PATHOPHYSIOLOGY

PATHOPHYSIOLOGY Ammonia Liver is the organ primarily tasked with the body’s removal of excess ammonia in the form of urea via the urea cycle in periportal hepatocytes and glutamine via the enzyme glutamine synthetase (GS) localized in perivenous hepatocytes as shown in a simplified manner in Figure 48.2. Patients with cirrhosis commonly develop intra‐ and extra‐hepatic portal‐systemic shunts and this shunting combined with the loss of metabolic capacity of hepatocytes results in impaired removal of ammonia and consequent hyperammonemia. One key adaptive consequence of impairment of hepatic ammonia removal in cirrhosis is activation of the gene coding for the enzyme protein GS in skeletal muscle [19]. This metabolic adaptation provides an alternate pathway for ammonia removal in the form of muscle glutamine production as shown schematically in Figure 48.2. Arterial blood and brain ammonia levels are increased several‐fold in patients with cirrhosis and HE and dynamic 13NH3‐ positron emission tomography (PET) studies demonstrate significant increases of brain ammonia and of the cerebral metabolic rate for ammonia (CMRA, defined as the rate at which ammonia is captured by brain and incorporated into metabolites) in these patients [20]. These studies went on to demonstrate that, contrary to earlier reports, this increase of brain ammonia in patients with cirrhosis was primarily a consequence of increased arterial blood concentrations rather than to altered kinetics of blood–brain transfer of ammonia. Venous ammonia measurements are of little predictive value in terms of HE occurrence or severity in patients with cirrhosis. In contrast, arterial ammonia concentrations are useful predictors of high risk for the development of severe complications of brain edema in patients with ALF such as intracranial hypertension [21] and brain herniation [7] that remains a major cause of mortality in this condition. The brain is devoid of an effective urea cycle. Consequently, removal of excess ammonia by the brain depends almost

exclusively on the synthesis of glutamine and the enzyme responsible, GS, is preferentially expressed by the astrocyte. 1 H‐magnetic resonance spectroscopic studies of patients with cirrhosis reveal increased concentrations of brain glutamine as a function of the severity of HE in these patients [22]. Consequently, the accumulation of glutamine in the astrocyte may be causally‐related to the pathogenesis of brain edema in ALF. However, studies in experimental animal models with ALF resulting from toxic liver injury failed to demonstrate a significant correlation between brain glutamine content, its synthesis or turnover, and severity of HE or brain edema [23] suggesting the presence of alternative or additional mechanisms. Established mechanisms responsible for the deleterious effects of ammonia on brain function include the following:

Direct effects of the ammonium ion (NH4 +) on the neuronal membrane Concentrations of ammonia equivalent to those reported in brain in acute and/or chronic liver failure are known to exert direct deleterious effects on both inhibitory and excitatory neurotransmission [24] by virtue of the fact that, under normal physiological conditions, ammonia exists primarily in its protonated form (NH4+) an entity with comparable ionic radius to that of potassium (K+) which is an ion with potent depolarizing properties on the neuronal membrane.

Effects of ammonia on brain energy metabolism Ammonia, in low millimolar concentrations equivalent to those observed in HE is also a potent inhibitor of the rate‐limiting tricarboxylic acid (TCA) cycle enzyme α‐ketoglutarate dehydrogenase [25] with the potential to result in impairment of glucose oxidation, brain lactate production, and impending ­ brain energy failure.

Effects of ammonia on the brain GABA system Recent studies show that ammonia is an activator of translocator protein (TLP), a protein located on the mitochondrial membrane

BRAIN

BRAIN

LIVER

urea

Periportal hepatocytes

Periportal hepatocytes

Perivenous hepatocytes

Perivenous hepatocytes NH3 MUSCLE

GLUTAMINE Normal

LIVER

urea

KIDNEY

GUT MUSCLE

GLUTAMINE

NH3

GUT

KIDNEY

Liver Failure

Figure 48.2  Schematic representation of inter‐organ trafficking of ammonia and glutamine under normal physiological conditions and in liver failure. Gut‐derived ammonia is normally removed as urea (periportal hepatocytes) or glutamine (perivenous hepatocytes). In liver failure, ammonia removal by both types of hepatocytes is decreased. Brain ammonia uptake increases but there is limited capacity for further increases in brain glutamine synthesis. In contrast, skeletal muscle becomes the principal route for ammonia removal due to post‐translational increase in the enzyme glutamine synthetase.



48:  Hepatic Encephalopathy

of glial cells (astrocytes and microglia) where it serves in the transport of cholesterol. Activation by ammonia results in increased uptake and conversion to a novel series of compounds known as neurosteroids (NS) one of which, allopregnanolone, is a potent agonist of the modulatory site on the gamma‐amino butyric acid (GABA) receptor. In this way, ammonia serves as an activator of GABAergic transmission with the potential to increase chloride flux and neuroinhibition. This mechanism is covered in greater detail in the section on Inhibitory neurotransmission of the current chapter.

Manganese T1‐weighted MRI of patients with cirrhosis consistently reveals bilateral symmetric signal hyperintensities in substructures of the basal ganglia particularly the globus pallidus and substantia nigra. In a landmark study of 51 patients with cirrhosis who were listed for liver transplantation and were assessed prospectively over a one‐year period, 11 patients (21.6%) exhibited definite Parkinsonism together with typical MRI signal hyperintensities. The degree of signal hyperintensity was not correlated with the etiology of cirrhosis, Child–Pugh scores, or fasting blood ammonia and no patients had OHE at the time of imaging. Neurological symptoms included a symmetric akinetic‐rigid syndrome, tremor, stooped posture, and gait impairment with rapid progression over months [16]. Blood manganese concentrations were elevated up to sevenfold in all of nine patients in whom measurements were made and cerebrospinal fluid (CSF) manganese concentrations were increased in all of three cases in which CSF was available. Liver transplantation resulted in normalization of MRI signal hyperintensities and circulating manganese levels as well as improved neurological symptoms. Treatment of two of these patients with L‐DOPA resulted in substantial improvements of motor function [16]. Motor impairments including the akinetic‐rigid syndrome that are characteristic of Parkinsonism in cirrhosis could contribute to the poor performance in psychometric testing using paper and pencil tests such as NCT‐A and NCT‐B in which there is a significant motor component thus casting some doubt on the sole use of these tests for the assessment of cognitive function and diagnosis of MHE or covert HE in patients with cirrhosis. Alternative tests with less dependency on unimpaired motor performance are available. In a separate study, basal ganglia tissue obtained at autopsy from patients with cirrhosis who died in hepatic coma were found to contain several‐fold increased manganese concentrations [26] as well as alterations of marker proteins and metabolite patterns related to the dopamine (DA) neurotransmitter system that are characteristic of idiopathic (non‐cirrhosis‐ related) Parkinson’s disease [17].

Systemic inflammation Systemic inflammation resulting from infection and/or hepatocellular damage is a common occurrence in acute or chronic liver failure and the acquisition of a systemic inflammatory response (SIRS) is a major predictor of HE in patients with cirrhosis or ALF [27]. Cellular damage to liver tissue resulting from toxins, particularly ethanol, has the potential to exacerbate

621

SIRS the extent of which increases as a function of the nature and severity of liver injury. SIRS results from the release into the circulation of proinflammatory cytokines such as tumor necrosis factor alpha (TNFα), interleukin‐1beta (IL‐1β), and interleukin‐6 (IL‐6). Circulating levels of TNFα are invariably increased in patients with cirrhosis and the magnitude of the increase correlates well with the grade of OHE [28]. Systemic inflammation resulting from lipopolysaccharide (LPS, endotoxin) has been shown to precipitate HE and increase blood–brain barrier (BBB) permeability in mice with ALF resulting from toxic liver injury [29]. These findings suggest that both cytotoxic (cell swelling) and vasogenic (BBB breakdown) mechanisms could contribute to the pathogenesis of brain edema and its CNS complications in ALF.

The concept of synergism Interest in synergistic mechanisms as an integral part of the pathogenesis of HE in cirrhosis began with the pioneering work of Les Zieve in the 1970s where synergistic actions of established liver‐derived toxins including ammonia, methanethiol, and octanoic acid when administered to laboratory animals were shown to act in concert leading to HE [30]. Since that time, evidence of synergism between three of the major entities implicated in the pathogenesis of HE in cirrhosis namely ammonia, manganese, and proinflammatory cytokines has been reported. In patients with cirrhosis and documented infection, induction of hyperammonemia is associated with worsening of neuropsychiatric status [31] and there is evidence indicating that synergism between ammonia, manganese, and proinflammatory cytokines represents a cascade of mechanisms resulting in the worsening of HE [32]. The trail of evidence consists of the ­following: glutamate is the principal excitatory neurotransmitter of mammalian brain and the termination of its action relies on transport into surrounding astrocytes via specific transporters. Both ammonia and manganese have been shown to inhibit these transporters resulting in an excess of glutamate in the synaptic cleft leading to hyperexcitability and activation of post‐synaptic glutamate receptors. This cascade results in nitration and consequent inactivation of key proteins via a process known as protein tyrosine nitration (PTN) [33]. One such tyrosine‐nitrated protein is the enzyme GS that is solely responsible for ammonia removal by the brain. Under conditions of PTN, brain ammonia levels become increased and the synergistic cycle is amplified.

Neuroinflammation Recent studies provide convincing evidence in support of a novel concept namely the role of neuroinflammation (inflammation of the brain per se) in the pathogenesis of HE in patients with ALF [34]. Neuroinflammation in ALF is characterized by activation of microglial cells together with increased production of proinflammatory cytokines such as TNFα and the interleukins IL‐1β and IL‐6 in the brain. The grade of HE severity correlates well with that of proinflammatory cytokine production and the use of proinflammatory gene deletion strategies results in slowing of HE progression in experimental ALF [35]. Microglial activation has also been reported in autopsied brain tissue samples from ALF patients with grade IV HE [13].

622

THE LIVER:  PATHOPHYSIOLOGY

Microglial activation also occurs in patients with cirrhosis who died in hepatic coma [14]. Activated microglia are known to express transcripts for the mitochondrial protein translocator protein (TLP) and using PET and the selective TLP ligand 11C‐PK11195, activation of  microglia indicative of a neuroinflammatory response is observed in patients with cirrhosis and MHE. Microglial activation appears to be a relatively early occurrence in patients with decompensated cirrhosis. Particularly intense signals were observed in the anterior cingulate cortex, a brain structure known to be associated with the control of attention in MHE patients [36]. Liver‐brain proinflammatory signaling occurs in HE and a growing list of possible mechanisms have been proposed [34]. At the cellular level, human cerebrovascular endothelial cells exposed to TNFα manifest increased transport of ammonia [37] and cultured neural cells exposed to combinations of ammonia and recombinant proinflammatory cytokines display increased expression of genes coding for proteins implicated in cellular dysfunction in HE suggestive of synergism. Factors other than ammonia and manganese that are increased in brain in liver failure also have the potential to elicit a neuroinflammatory response. For example, millimolar concentrations of lactate are known to substantially increase the release of TNFα and IL‐6 from cultured microglial cells [38].

Brain energy metabolism There is no convincing evidence from either biochemical or spectroscopic investigations in support of the hypothesis that HE is primarily caused by a sufficiently severe loss in concentration of high energy phosphates indicative of cerebral energy failure. However, alterations of uptake and metabolism of the brain’s primary energy substrate, glucose, have consistently been described in both acute and chronic liver diseases. Studies using PET and the non‐metabolized glucose transport ligand 18 fluorodeoxyglucose in patients with cirrhosis and mild HE show significant decreases in uptake in anterior cingulate cortex, a brain structure implicated in the monitoring of responses to visual stimuli [39]. Not surprisingly, decreased brain glucose uptake in these patients was correlated with impaired performance on psychometric testing. This decrease of brain glucose uptake is currently considered to be the consequence of decreased brain energy requirements resulting from decreased neural activation in (rather than the cause) of HE [39]. Pathophysiologically‐relevant concentrations of ammonia inhibit the TCA cycle, an essential metabolic pathway involved in the maintenance of brain energy requirements [25]. Slowing of the TCA cycle results in the shuttling of pyruvate to lactate and cerebrospinal fluid lactate concentrations correlate well with severity of HE in patients with cirrhosis [40] as well as in brain microdialysates from patients with ALF where surges of increased intracranial pressure were frequently found to accompany increased lactate concentrations [41]. Brain lactate concentrations in excess of 12 mM have been reported at coma stages of encephalopathy in animal models of ALF [25]. Such concentrations are beyond the brain’s capacity to buffer lactate with the potential to result in acidosis.

Cerebral blood flow (CBF) Although global CBF in patients with cirrhosis is generally somewhat reduced, PET studies reveal that CBF changes in these patients with mild HE occurred in a region‐selective manner whereby flow to cerebral cortical regions was decreased while flow to basal ganglia, cerebellar, and thalamic structures was significantly increased. Since this pattern of changes in CBF parallels the regional changes of brain glucose utilization, CBF autoregulation (defined as the capacity of CBF to match brain activity independent of changes of systemic arterial pressure) appears to be preserved in patients with cirrhosis [42]. In contrast to the situation in cirrhosis, loss of CBF autoregulation is common in patients with ALF where the variability of CBF rates are influenced by a variety of factors including the complex interplay between local energy demands, ammonia neurotoxicity, systemic arterial pressure, and intracranial hypertension. A five‐phase classification system based upon a retrospective analysis of sequential intracranial pressure (ICP) and CBF measurements in ALF patients has been proposed [43] as follows: Phase 1: An initial phase of low CBF resulting from lower neuronal activity Phase 2: A progressive increase of CBF Phase 3: A subsequent increase of ICP Phase 4: Final stage where increased ICP results in decreased CBF Phase 5: Brain death The classification integrates well with previously‐published data; it has a mechanistic correlation and is expected to facilitate prognostic evaluation in patients with ALF.

Cell volume regulation and brain edema Brain edema has been described in patients with cirrhosis and in patients with ALF. However, the nature and pathological characteristics of the brain edema in the two conditions are quite distinct. In ALF brain edema has been characterized consisting primarily of cytotoxic brain edema that, if uncontrolled, may progress to intracranial hypertension (ICH) and brain herniation. Brain edema in ALF appears to have multiple causes including the accumulation of brain lactate [25] as described above as well as proinflammatory mechanisms resulting from microglial activation and release of cytokines [34]. Brain edema leading to ICH is uncommon in cirrhosis since the swelling is low‐grade. However, it has the potential to result in impairments of cell–cell signalling and to cause malfunction of astrocytic proteins. NMR spectroscopic studies in patients with cirrhosis have consistently shown disturbances of brain organic osmolytes including increases of glutamate/glutamine with compensatory reductions of myo‐inositol [44].

Inhibitory neurotransmission GABA is the principal inhibitory neurotransmitter system of mammalian brain and alterations of GABAergic transmission are implicated in a wide range of neurodegenerative and metabolic brain diseases including HE.



48:  Hepatic Encephalopathy

GABAergic neurotransmission is mediated by the amino acid GABA via activation of a post‐synaptic GABA‐receptor complex (GRC), a protein complex and ligand‐gated ion channel that is selective for chloride (Cl−) ion. Activation of the GRC by GABA results in the entry of chloride through its pore resulting in hyperpolarization of the post‐synaptic neuron. This results in inhibition of central neurotransmission by diminishing the chance that an action potential occurs. Activation of GABAergic transmission (also frequently referred to as “increased GABAergic tone”) was originally proposed in the 1980s based on visual evoked response patterns in experimental animals with toxic liver injury and HE that were identical to patterns observed in normal animals treated with well‐established activators of the GRC [45]. These findings led to intense interest in activation of the GRC as a major factor implicated in the pathogenesis of HE, an interest that continues to this day. Initially, studies were focused on the measurement of integral components of the GABA system including GABA concentrations and metabolites, activities of enzymes responsible for GABA synthesis as well as post‐synaptic GABA receptors. In all cases, these parameters were found to be present in normal amounts in animal models of acute and chronic liver failure and, more importantly, in autopsied brain tissue from patients with decompensated cirrhosis who died in hepatic coma [46]. These findings led to a change in focus of research in this area and a renewal of interest in the multiple modulatory sites and associated subunit proteins of the GRC itself. The GRC is composed of an active binding site for GABA, often referred to as the GABA recognition site that, together with allosteric sites, modulate the activity of the GRC. These sites are targets for a range of substances including benzodiazepines and neuroactive steroids also known as neurosteroids (NS). The benzodiazepine modulatory site on the GRC has been extensively studied in brain tissue obtained at autopsy from patients with cirrhosis who died in stage IV HE as well as in living patients with mild HE by use of the selective benzodiazepine modulatory site antagonist PET ligand Ro15‐1788 (flumazenil). No significant alterations of binding site affinities or densities of these sites were observed when compared to data from age‐matched control material [46]. Furthermore, subsequent studies failed to demonstrate any significant alterations of the allosteric coupling between the benzodiazepine modulatory site and the GABA recognition in the HE patient material [47]. Together, these findings clearly demonstrate that the GRC with respect to the benzodiazepine modulatory site is functioning normally in patients with HE related to cirrhosis. The reports of no significant alterations in expression or coupling of the GABA receptor machinery of the brain in HE led to a renewed interest in the search for so‐called “endogenous ligands” for these receptors. Several such substances were detected. Quantitative mass spectrometric analysis followed to unequivocally establish the identity of these substances. They were identified as well‐established pharmaceutical benzodiazepines or their metabolites [48] that were traced to their administration as muscle relaxants during prior endoscopic work‐up or as sedatives. In contrast to the disappointing findings in relation to the benzodiazepine modulatory site and its potential role in the

623

pathogenesis of HE in chronic liver diseases, investigations of the other modulatory site on the GRC, the NS modulatory site in HE patient material yielded more positive results. The molecular steps involved in NS synthesis in the brain and the relationship between ammonia exposure and NS synthesis is shown in a simplified schematic form in Figure 48.3. The process starts with the activation of a mitochondrial protein located in glia (astrocytes and microglia) known as translocator protein (TLP) resulting in the entry of cholesterol into the mitochondrion. A series of well‐established metabolic steps then transform cholesterol into NS that are then available for release into the synaptic cleft. Two of these NS namely allopregnanolone (ALLO) and tetrahydro‐deoxy corticosterone (THDOC), are potent positive allosteric modulators of the GRC and, hence, are activators of GABAergic neurotransmission leading to enhanced central neuroinhibition. The first direct evidence for a role of NS in the pathogenesis of HE in cirrhosis was provided by the report of significant increases of ALLO in brain tissue obtained at autopsy from patients with decompensated cirrhosis who died in stage IV HE (coma) [49]. ALLO concentrations in control material from patients with no liver disorders, patients with decompensated cirrhosis but no encephalopathy, as well as one patient who died in uremic coma, were all within normal limits. Furthermore, ALLO concentrations in brain extracts from HE patients were Ammonia

Cholesterol

TLP Perineuronal astrocyte or microglia

MM

GABA

Pregnenolone

Nerve Terminal

SER

Astrocyte

Allopregnanolone

GABA GABA site

NS site

Cl

Post-Synaptic neuron

Neuroinhibition HE

Figure 48.3  Pathophysiologic link between ammonia toxicity and “increased GABAergic tone” in HE. A schematic representation of the sequence of steps whereby activation of the translocator protein (TLP) situated on the mitochondrial membrane (MM) of the astrocyte or microglial cell occurs. This results in increased transport of cholesterol into the cell and its conversion to pregnenolone and the NS allopregnanolone. Stimulation of the NS modulatory site on the GABA‐A receptor complex by allopregnanolone then amplifies the signal produced by the neurotransmitter (GABA) acting on the adjacent GABA recognition site. The net result is increased entry of chloride (Cl−) and enhanced neuroinhibition that contributes to the pathogenesis of HE in cirrhosis.

624

THE LIVER:  MANAGEMENT AND THERAPEUTIC STRATEGIES FOR HE

present in sufficiently high concentrations to result in a 53% increase in binding of the GABA agonist ligand 3H‐muscimol to brain membrane preparations thus satisfying the requirement that concentrations of ALLO in HE patient material were sufficient to result in “increased GABAergic tone” [50]. Novel compounds with antagonist action at the NS modulatory site on the GRC have recently been synthesized. Further information and results of preliminary clinical trials are summarized in the section, GABA receptor modulators in this chapter.

MANAGEMENT AND THERAPEUTIC STRATEGIES FOR HE The ultimate test of the pertinence of the diverse theories of the pathogenesis of HE rests in the clinic where the impact of appropriate treatments based upon the nutritional, metabolic and toxic factors identified in this chapter may be undertaken.

Nutritional management The pathogenesis of malnutrition in liver disease is complex and multifactorial involving reduction of dietary intake due to anorexia and dietary restriction, alterations of nutrient biosynthesis, impaired intestinal absorption, disturbances of substrate utilization, increased protein loss, and abnormal metabolism resulting from, for example, the proinflammatory state. The functional integrity of the liver is essential for the provision, inter‐organ transfer, and metabolism of essential nutrients. As outlined in the present review, nitrogen metabolism plays a key role in the development of HE in cirrhosis and in conditions of liver failure, skeletal muscle adapts metabolically to serve as the major back‐up organ for ammonia removal. Sarcopenia or loss of muscle mass is commonly encountered in cirrhosis where it has adverse effects on survival, health‐related quality of life, as well as outcome following liver transplantation [51]. Sarcopenia may also lead to worsening of the complications of cirrhosis including portal hypertension, ascites, and HE. Indeed, results of a prospective study reveal that muscle depletion in patients with cirrhosis increases the risk of both MHE and OHE [52]. Obesity and its associated alterations of nutrient intake are frequent occurrences in patients with non‐alcoholic fatty liver disease (NAFLD) and play an important role in determining rates of progression to non‐alcoholic steatohepatitis (NASH) and cirrhosis. A recent consensus document from the International Society for Hepatic Encephalopathy and Nitrogen Metabolism (ISHEN) on guidelines for the nutritional management of HE in cirrhosis has appeared [53]. The stated objectives of nutritional management in patients with cirrhosis include the correction of specific nutritional deficiencies, prevention and treatment of the complications of cirrhosis as well as the complications of liver transplantation, and the support of liver regeneration. Energy requirements of 35–45 kcal g−1 daily are recommended for patients with cirrhosis with small meals evenly distributed throughout the day and a late‐night snack of complex carbohydrates to minimize protein utilization. The practice of dietary protein restriction that was prevalent until the late 1990s

[54] has now been widely discontinued following the report [55] that patients with cirrhosis benefit from normal protein (1.2–1.5 g kg−1 daily). Diets rich in vegetable and dairy protein are encouraged and patients intolerant to dietary protein may consider branched chain amino acid supplements as an alternative to protein [53]. Vitamin deficiencies in cirrhosis result from impaired hepatic function and diminished hepatic reserves as well as inadequate dietary intake and malabsorption. A neuropathologic study examining brain tissue from patients with cirrhosis who died in stage IV HE revealed lesions in thalamus and mamillary bodies that are characteristic of Wernicke’s encephalopathy together with cerebellar degeneration consistent with vitamin B1 deficiency [11]. The B1 deficiency in these patients was attributed to a loss of hepatic stores of the vitamin and the prevalence of lesions was threefold higher than that reported in material from non‐liver disease controls. It was recommended that B1 supplements be administered to all patients with decompensated cirrhosis particularly those with alcoholic etiology [51]. Deficiencies of other vitamins including B2, B6, and B12 have also been reported in cirrhosis associated with alcoholic liver damage and/or diminished hepatic storage but causal relationships with complications of cirrhosis were not established. Zinc deficiency is common in cirrhosis but evidence for beneficial effects of zinc supplements on HE is equivocal [56]. It has been suggested that malnutrition is a factor that increases both costs and post‐transplant complications [57]. Neurological complications post transplantation are numerous and may include diffuse encephalopathy, seizures, intracranial hemorrhage, stroke, progressive neurological deterioration, central pontine myelinolysis, ataxia, psychosis, confabulation, and peripheral neuropathy. While some of these complications likely result from specific nutritional and metabolic deficits alluded to above, further studies are required to determine the precise nutritional factors implicated and facilitate specific nutritional recommendations.

Ammonia‐lowering strategies Given the wealth of evidence derived from biochemical, neuroimaging, and spectroscopic studies in support of a central role of ammonia in the pathogenesis of HE, it is not surprising that the therapeutic strategy that has so far stood the test of time relates to the reduction of hyperammonemia. Several treatments aimed at reducing the production of ammonia and/or its removal have been found to be effective in this regard.

Non‐absorbable disaccharides Non‐absorbable disaccharides such as lactulose and lactitol remain first‐line agents for the lowering of the production and absorption of ammonia. They are metabolized by bacteria in the colon with production of acetic and lactic acids and the consequent acidification of the colon creates a hostile environment for urease‐rich intestinal bacteria that are responsible for the production of ammonia. In addition, these agents facilitate the formation of NH4+ leading to reduced ammonia production and their cathartic properties may result in up to fourfold increases of fecal nitrogen excretion [58].



48:  Hepatic Encephalopathy

Although non‐absorbable disaccharides are widely employed as first‐line therapy for HE in cirrhosis, systematic reviews and meta‐analyses of trials have yielded mixed results. In an analysis of ten studies, it was concluded that there was insufficient evidence to recommend them for standard therapy until they were shown to confer benefit over placebo [59]. The analysis was subsequently challenged based upon shortcomings in experimental design and methodology [60] and further studies including a second meta‐analysis of lactulose revealed significant improvements in cognitive function and health‐related quality of life in MHE [61] as well for the primary prophylaxis of overt HE [62] in patients with cirrhosis. A subsequent meta‐ analysis showed that, compared with placebo/no intervention, the non‐absorbable disaccharides were associated with significant overall benefits on HE [63].

Antibiotics The aminoglycoside antibiotic neomycin is still used for the treatment of patients with cirrhosis and HE despite its potentially serious side‐effects that include ototoxicity and nephrotoxicity since the drug is partially absorbed. Rifaximin is a poorly absorbed, broad spectrum antibiotic that is effective for lowering of blood ammonia and improvement of mental state in patients with MHE as well as in those with overt HE where its efficacy has been found to exceed that of neomycin. Results of a large, randomized, double‐blind, placebo‐controlled trial demonstrated that rifaximin reduced the risk of repeat episodes of overt HE and reduced the time to first hospitalization with no serious adverse events [64]. A subsequent systematic review and meta‐analysis of 19 trials with 1390 patients confirmed that rifaximin is useful for the management of HE with a beneficial effect on HE severity and for treatment of acute episodes as well as on patient mortality [65]. Rifaximin also improves psychometric test scores and HRQOL in patients with MHE [66].

Probiotics Urease‐producing bacteria are plentiful in the gut where they convert urea to carbamate and ammonia. Modifications of the gut microflora balance by the introduction of bacteria that compete with those expressing urease (probiotics) are effective ammonia‐lowering agents. Treatment of patients with non‐alcoholic cirrhosis and MHE with probiotic yoghurt results in slowing of progression to overt HE [67]. Furthermore, randomized controlled trials comparing probiotics with placebo reveal improvements in arterial ammonia concentrations, psychometric test scores, and HRQOL in MHE patients [68, 69]. The degree of improvement was equivalent to that of lactulose [68] or rifaximin [69]. A subsequent meta‐analysis of 14 studies totaling 1152 patients confirmed the comparable beneficial effects of probiotics and lactulose and went on to demonstrate efficacy of probiotics in reducing hospitalization rates as well as rates of progression to overt HE.

L‐ornithine L‐aspartate (LOLA) LOLA is a mixture of two endogenous amino acids with established ammonia‐lowering properties. Multiple mechanisms responsible for the ammonia‐lowering actions of LOLA include

625

an ability to increase urea production by residual hepatocytes (L‐ornithine is a urea cycle substrate) and increased glutamine synthesis in skeletal muscle [70]. There is evidence that LOLA has direct hepatoprotective actions in patients with cirrhosis and HE [71]. Beneficial effects of LOLA for the management and treatment of HE in cirrhosis have been reported in over 20 randomized clinical trials over the last 20 years. The intravenous formulation of LOLA is particularly effective in the treatment of low‐grade [72] and bouts of overt [73] HE in cirrhosis. A recent meta‐analysis demonstrated that the oral formulation of LOLA is particularly effective for the treatment of MHE [74]. LOLA is reportedly equivalent or superior in efficacy compared to rifaximin [69], probiotics [68, 69], and lactulose [68] for the treatment of MHE and a network meta‐analysis confirmed these findings [75]. However, in the only trial conducted to date, LOLA did not appear to be effective for the treatment of HE in ALF [76].

Branched chain amino acids (BCAAs) Results of a Cochrane review on the efficacy of BCAAs for improvement of mental state concluded that patients treated with BCAAs are more likely to recover from HE compared to patients receiving control regimens that included lactulose and neomycin [77]. Two large studies revealed beneficial effects of BCAAs on nutrition and survival, effects that may outweigh any effects on HE per se [77, 78]. Mechanisms responsible for the beneficial effects of BCAAs have not been fully elucidated despite their continuing use for over 35 years. Suggested mechanisms include beneficial effects on their use as substrates for hepatic protein synthesis as well as stimulation of liver regeneration and increases of cerebral perfusion.

Benzoate, phenylacetate Benzoate and phenylacetate are agents that have been used extensively for the lowering of blood ammonia in children with congenital urea cycle disorders. Phenylacetate is generally given orally as its precursor phenylbutyrate. Phenylacetate condenses with glutamine to form phenylacetyl glutamine whereas benzoate condenses with glycine to form hippurate. Both condensation products are excreted in the urine. The ammonia‐lowering action results from removal of these ammoniagenic substrates as well as the diversion of available ammonia for the replenishment of the amino acid pools. Sodium benzoate was reported to manifest comparable ­efficacy to lactulose in a prospective study in patients with cirrhosis and acute HE [79]. Mixtures of phenylacetate with other agents such as glycerol or L‐ornithine have recently been introduced as possible ammonia‐lowering strategies for possible use in the treatment of HE in cirrhosis. In the case of glycerol phenylbutyrate (GBP), a multicenter phase 2 trial, patients in the GBP arm manifested lowering of blood ammonia and experienced less HE events as well as less HE hospitalizations compared to placebo [80]. Ornithine phenylacetate (OP), a mixture of L‐ornithine and phenylacetate has been shown to reduce blood ammonia in experimental animal models of HE, a result that is not surprising given that both the L‐ornithine and phenylacetate moieties are

626

THE LIVER:  MANAGEMENT AND THERAPEUTIC STRATEGIES FOR HE

independently well‐established ammonia‐lowering compounds. A study on the safety, tolerability, and pharmacokinetics of OP in patients with acute liver injury or failure showed significant urinary ammonia excretion, good tolerability, and no safety signals [81]. Randomized, controlled studies are now required to determine its use as an ammonia‐scavenging agent in patients with HE.

L‐carnitine L‐carnitine has been shown to manifest a significant protective effect in patients with cirrhosis and ammonia‐precipitated encephalopathy [82]. The protective effect of L‐carnitine was evident for patients with overt HE grades I and II as well as for patients with MHE as assessed by improvements in scores for NCT‐A. Studies in both experimental congenital hyperammonemia [83] and portal‐systemic encephalopathy [84] showed a significant protective effect of L‐carnitine on brain energy metabolism characterized by improved brain glucose oxidative capacity and prevention of cerebrospinal fluid lactate accumulations [84]. As alluded to earlier in this chapter, decreased glucose oxidation in brain in liver failure is the consequence of ammonia‐induced inhibition of the TCA cycle by ammonia leading to lactate accumulation [25]. These novel beneficial effects of L‐carnitine are the consequence of inhibition of one of the major metabolic consequences of exposure to ammonia namely that of decreased brain glucose oxidation rather than a direct effect on ammonia production or removal.

Neuropharmacological approaches

Anti‐inflammatory agents Recognition of the key roles of systemic inflammation and neuroinflammation in the pathogenesis of HE in both ALF and in cirrhosis has resulted in renewed interest in the search for agents with anti‐inflammatory properties as potential therapies. Ibuprofen attenuates the neurologic deficits in learning ability in an experimental animal model of type B HE [89]. A central role of TNFα has been proposed to explain HE in cirrhosis based upon the observations that many precipitating factors including infection, gastrointestinal bleeding, constipation, and renal failure are associated with increased circulating levels of the cytokine. Moreover, many therapies known to lower ammonia such as lactulose, antibiotics, and probiotics also effectively reduce TNFα production [28]. Following up on the report that deletion of the gene coding for the TNFα receptor resulted in delayed onset of HE and prevention of brain edema in ALF due to toxic liver injury [35], a more recent study showed that the systemic sequestration of TNFα by administration of the TNFα‐neutralizing compound etanercept resulted in reduction of microglial activation and delayed progression of HE in mice with acute liver injury [90]. Minocycline is a semi‐synthetic tetracycline antibiotic that limits microglial activation by a mechanism that is independent of its antimicrobial properties and studies in the liver ischemic rat model of ALF showed that minocycline prevents microglial activation and attenuates the increased synthesis of proinflammatory cytokines including TNFα in the brain while leading to a slowing of progression of HE and prevention of brain edema [91]. The beneficial effect of mild hypothermia in ALF involves anti‐inflammatory mechanisms at both hepatic and cerebral levels [92].

GABA receptor modulators

Other centrally‐acting agents

A clinical trial of the efficacy of flumazenil, a potent antagonist of the benzodiazepine modulatory site on the GABA receptor complex was undertaken in patients with cirrhosis and moderate to severe HE. The trial demonstrated clinical improvement of HE in a significant subset of patients [85] and this was subsequently confirmed in a systematic review and meta‐analysis [86] although, as expected, the benefit was relatively short‐acting due to the short half‐life of flumazenil. The findings of increased brain concentrations of allopregnanolone (described in the section, Inhibitory neurotransmission in the current chapter) the potent agonist of the NS site on the GABA‐A receptor in material from HE patients resulted in a renewed interest in the discovery of compounds with the necessary configuration to block the actions of NS on the NS ­modulatory site on the GRC [47]. So‐called “GABA‐receptor‐ modulating steroid antagonists” (GAMSA) were synthesized and tested in animal models. One example, 3β‐20β‐ dihydroxy‐5α‐pregnane (UC1011) was shown previously to inhibit the ALLO‐induced increases of GABA‐induced increases of chloride uptake into cerebral cortical and hippocampal preparations making it an ideal candidate for further study [87]. More recently, a second member of the GAMSA family, GR 3027 has been shown to improve spatial learning, motor coordination, and circadian rhythms in an experimental model of HE in chronic liver disease [88]. Controlled clinical trials are now required.

Evidence of a dopaminergic deficit together with neurological symptoms characteristic of Parkinson’s disease are well established findings in patients with HE associated with cirrhosis. It is therefore not surprising that attempts have been made to treat these patients with agents with the potential to reactivate the dopaminergic system. Dopamine‐replacement therapy by the precursor amino acid L‐DOPA was found to be ineffective for improvement of cognitive symptoms [93] but, subsequently, using a more appropriate Parkinson disease rating scale for assessment of clinical efficacy, L‐DOPA was found to result in improvements in motor performance [16]. Clinical trials with the dopamine receptor agonist bromocriptine have yielded conflicting results [94, 95] but, again, clinical endpoints such as the portal‐systemic encephalopathy (PSE) index rather than tests aimed at assessment of dopaminergic function and motor coordination were used making interpretation of the findings problematic. Therapeutic options for the treatment of Parkinsonism in cirrhosis have been reviewed [96].

Liver transplantation and HE Liver transplantation (LT) is the ultimate treatment for decompensated cirrhosis. Although LT generally confers a level of reversibility of some of the symptoms of HE, there is a growing body of evidence suggesting that this reversibility is far from complete [97] and patients with a history of overt HE have an increased probability of persistent neurocognitive dysfunction



48:  Hepatic Encephalopathy

post‐LT compared to those without prior overt HE. This finding begs the question whether it is the episode(s) of overt HE or the LT that is responsible for the persistent deficits. This issue was subsequently addressed in a prospective study in which it was demonstrated that even a single episode of overt HE was accompanied by a learning deficit and that multiple episodes led to deficits in reaction times, attention, psychomotor speed, and working memory [98]. Overt HE patients may suffer a loss of “cognitive reserve” that this could have important implications for the assignment of priority for LT. Other possible causes of post‐LT cognitive dysfunction include anoxic intraoperative complications, immunosuppressant medication, comorbidities, manganese neurotoxicity [17], and thiamine deficiency [11]. In a study comparing the incidence and severity of neurological complications after cadaveric versus living donor LT, the most common complications (encephalopathy and seizures) were significantly lower following living donor LT. These findings could be accounted for, at least in part, to the shorter cold ischemia time. The incidence of neurological complications post‐LT was not dependent on the nature of the immunosuppressant used [99].

Future directions for the field of HE When the author of this chapter began his research in HE some four decades ago, there were two major alternative agents employed for the management and treatment of HE in cirrhosis; a non‐absorbable disaccharide and an aminoglycoside antibiotic. Both acted at the level of the gut, both resulted in lowering of blood ammonia; both had serious side‐effects. Many of us wonder if anything has changed. Well, the disaccharide is still with us but at least we appear to have found a better antibiotic. It appears that the search for agents, some new, some recycled, targeting the lowering of blood ammonia is here to stay. Progress in the search for more rational therapies for HE is slow. Despite evidence to the contrary, it is widely believed that HE results principally, if not entirely, from the neurotoxic effects of ammonia. Consequently, search for new therapies remains focused on the gut. HE is, after all, a brain disorder and advances in our understanding of intercellular signalling involving defined neurotransmitter and neuromodulators in brain areas implicated in the maintenance of consciousness, cognitive, and motor function in liver disorders continue to be elucidated as do novel concepts involving central anti‐inflammatory molecules. Yet translation of these discoveries, and sometimes the use of new molecules themselves, remains sluggish. One exception presented in this chapter relates to the discovery of the effectiveness of an antagonist of the neurosteroid modulatory site on the GABA‐A receptor complex of brain. Clinical trials are ongoing. The future directions of HE research and patient care will undoubtedly be based to an increasing extent on targeting of the brain per se.

ACKNOWLEDGMENTS Studies and related publications from the author’s research unit were funded by operating grants from the Canadian Institutes of

627

Health Research (CIHR) and the Canadian Association for the study of the Liver (CASL). The author is grateful to Mr Jonas Eric Pilling for the design of Figure 48.3.

REFERENCES 1. Mullen, K.D. and Prakash, R.K. (Eds.) Hepatic Encephalopathy, Clinical Gastroenterology, Humana Press/Springer New York, pp. 97–102 2. Rossle, M. and Euringer, W. Hepatic encephalopathy in patients with transjugular intrahepatic portosystemic shunt (TIPS), in Hepatic encephalopathy: Clinical Gastroenterology, (eds. K.D. Mullen and R.K. Prakash), Humana Press/Springer, New York, 2012, pp. 211–20. 3. Bajaj, J.S., Wade, J.B., Gibson, D.P. et al. The multi‐dimensional burden of cirrhosis and hepatic encephalopathy on patients and caregivers. Am J Gastroenterol, 2011;10:6646–53. 4. Bajaj, J.S. Introduction and setting the scene: new nomenclature of hepatic encephalopathy and American Association for the study of liver diseases/ European Association for the study of the liver guidelines. Clin Liv Dis, 2017;9:48–51. 5. Blei, A.T. and Cordoba, J. Hepatic encephalopathy practise guidelines. Am J Gastroenterol, 2001;96:1968–75. 6. Bajaj, J.S., Cordoba, J., Mullen, K.D. et  al. Review article: the design of clinical trials in hepatic encephalopathy‐an ISHEN consensus statement. Aliment Pharmacol Ther, 2011;3:339–47. 7. Clemmesen, J.O., Larsen, F.S., Kondrup, J. et  al. Cerebral herniation in patients with acute liver failure is correlated with arterial ammonia concentration. Hepatology, 1999;2:948–53. 8. Kircheis, G., Wettstein, M., Timmermann, L. et al. Critical flicker frequency for quantification of low‐grade hepatic encephalopathy. Hepatology, 2002;35:357–66. 9. Kato, M., Hughes, R.D., and Keays, R.T. Electron microscopic study of brain capillaries in cerebral edema from fulminant hepatic failure. Hepatology, 1992;15:1060–6. 10. Butterworth, R.F., Giguere, J.F., Michaud, J. et al. Ammonia: key factor in the pathogenesis of hepatic encephalopathy. Neurochem Pathol, 1987:6–12. 11. Kril, J.J. and Butterworth, R.F. Diencephalic and cerebellar pathology in alcoholic and non‐alcoholic patients with end‐stage liver disease. Hepatology, 1997;2:637–41. 12. Cordoba, J., Alonso, J, Rovira, A. et  al. The development of low‐grade cerebral edema in cirrhosis is supported by the evolution of 1H‐magnetic resonance abnormalities after liver transplantation. J Hepatol, 2001;3:598–604. 13. Butterworth, R.F. Hepatic encephalopathy: a central neuroinflammatory disorder? Hepatology, 2011;53:372–6 14. Zemtsova, I., Gorg, B., Keitel, V. et  al. Microglial activation in hepatic encephalopathy in rats and humans. Hepatology, 2011;5:404–15. 15. Butterworth, R.F., Neuronal cell death in hepatic encephalopathy. Metab Brain Dis, 2007;2:209–20. 16. Burkhard, P.R., Delavelle, J., Du Pasquier, R. et al. Chronic Parkinsonism associated with cirrhosis. Arch Neurol, 2003;60:521–8. 17. Butterworth, R.F., Spahr, L., Fontaine, S. et al. Manganese toxicity, dopaminergic dysfunction and hepatic encephalopathy. Metab Brain Dis, 1995;10:559–67. 18. Ahluwalia, V., Wade, J.B., Moeller, F.G. et al. The etiology of cirrhosis is a strong determinant of brain reserve: a multi‐modal MR imaging study. Liver Transpl, 2015;2:1123–32. 19. Desjardins, P., Rao, K.V., Michalak, A. et al. Effect of portacaval anastomosis on glutamine synthetase protein and gene expression in brain, liver and skeletal muscle. Metab Brain Dis, 1999;14:273–80. 20. Sorensen, M. and Ott, P. Cerebral ammonia metabolism in cirrhosis. Funct Mol Imaging Hepatol, 2012:153–9. 21. Bernal, W., Hall, C., Karvellas, C.J. et al. Arterial ammonia and clinical risk factors for encephalopathy and intracranial hypertension in acute liver failure. Hepatology, 2007;46:1844–52. 22. Laubenberger, J., Haussinger, D., Boyer, S. et al. Proton magnetic resonance spectroscopy of brain in symptomatic and asymptomatic patients with liver cirrhosis. Gastroenterology, 1997;112:1610–6.

628

THE LIVER:  REFERENCES

23. Zwingmann, C., Chatauret, N., Liebfritz, D. et al. Selective increase of brain lactate synthesis in experimental acute liver failure: results of a 13C/1H nuclear magnetic resonance study. Hepatology, 2003;3:420–8. 24. Felipo, V. and Butterworth, R.F. Neurobiology of ammonia. Prog Neurobiol, 2002;67:259–79. 25. Lai, J.C.K. and Cooper, A.J.L. Brain α‐ketoglutarate dehydrogenase: kinetic properties, regional distribution and effects of inhibitors. J Neurochem, 1986;47:376–86. 26. Pomier Layrargues, G., Spahr, L. and Butterworth, R.F. Increased manganese concentrations in pallidum of cirrhotic patients, Lancet, 1995;34:735–41. 27. Rolando, N., Wade, J., Davalos, M. et  al. The systemic inflammatory response syndrome in acute liver failure. Hepatology, 2000;3:734–9. 28. Odeh, M. Pathogenesis of hepatic encephalopathy: the tumour necrosis factor‐alpha theory. Eur J Clin Invest, 2007;3:291–304. 29. Chastre, A., Belanger, M., Nguyen, B.N. et al. Lipopolysaccharide precipitates hepatic encephalopathy and increases blood‐brain barrier permeability in mice with acute liver failure. Liver Int, 2013;3:353–61 30. Zieve, L., Doizaki, W.M., and Zieve, F.J. Synergism between mercaptans and ammonia or fatty acids in the production of coma: a possible role of mercaptans in the pathogenesis of hepatic coma. J Lab Clin Med, 1974;8:26–8. 31. Shawcross, D.L., Wright, G., Olde Daminck, S.W.M. et al. Role of ammonia and inflammation in minimal hepatic encephalopathy. Metab Brain Dis, 2007;2:125–38. 32. Butterworth, R.F. Pathogenesis of hepatic encephalopathy in cirrhosis: the concept of synergism revisited. Metab Brain Dis, 2016;31(6):1211–15. 33. Schliess, F., Gorg, B., Fischer, R. et al. Ammonia induces MK801‐sensitive nitration and phosphorylation of protein tyrosine residues in rat astrocytes. FASEB J, 2002;1:739–41. 34. Butterworth, R.F. The liver‐brain axis in liver failure: neuroinflammation and encephalopathy. Nat Rev Gastroenterol Hepatol, 2013;1:522–8. 35. Bemeur, C., Qu, H., Desjardins, P. et al. IL‐1 or TNF receptor gene deletion delays onset of encephalopathy and attenuates brain edema in experimental acute liver failure. Neurochem Int, 2010;5:213–5. 36. Cagnin, A., Taylor‐Robinson, S.D., Forton, D.M. et al. In vivo imaging of cerebral “peripheral benzodiazepine binding sites” in patients with hepatic encephalopathy. Gut, 2006;5:547–53. 37. Duchini, A., Govindarajan, S., Santucci, M. et al. Effects of tumor necrosis factor‐a and interleukin‐6 on fluid‐phase permeability in CNS‐derived endothelial cells. J Invest Med, 1996;4:474–82. 38. Andersson, A.K., Adermark, L., Persson, M. et  al. Lactate contributes to ammonia‐mediated astroglial dysfunction during hyperammonemia, Neurochem Res, 2009;3:555–65. 39. Lockwood, A.H., Weissenborn, K., and Butterworth, R.F. An image of the brain in patients with liver disease. Curr Opin Neurol, 1997;1:525–33. 40. Yao, H., Sadoshima, S., Fijii, K. et al. Cerebrospinal fluid lactate in patients with hepatic encephalopathy. Eur Neurol, 1987;2:182–7. 41. Tofteng, F., Jorgensen, L., Hansen, B.A. et  al. Cerebral microdialysis in patients with fulminant hepatic failure, Hepatology, 2002;3(6):1333–40. 42. Larsen, F.S., Olsen, K.S., Ejlersen, E. et al. Cerebral blood flow autoregulation and transcranial Doppler sonography in patients with cirrhosis. Hepatology, 1995;22(3):730–6. 43. Aggarwal, S., Obrist, W., Yonas, H. et al. Cerebral hemodynamic and metabolic profiles in fulminant hepatic failure: relationship to outcome. Liver Transpl, 2005;11(11):1353–60. 44. Haussinger, D. Low grade cerebral edema and the pathogenesis of hepatic encephalopathy in cirrhosis. Hepatology, 2006;43:1187–90. 45. Jones, E.A. Ammonia, the GABA neurotransmitter system and hepatic encephalopathy. Metab Brain Dis, 2002;17:275–81. 46. Butterworth, R.F., Lavoie, J., Giguere, J.F. et al. Affinities and densities of high‐affinity [3H] muscimol (GABA‐A) binding sites and of central benzodiazepine receptors are unchanged in autopsied brain tissue from cirrhotic patients with hepatic encephalopathy. Hepatology, 1988;8(5):1084–8. 47. Ahboucha, S. and Butterworth, R.F. Pathogenesis of hepatic encephalopathy: a new look at GABA from the molecular standpoint. Metab Brain Dis, 2004;19;331–43. 48. Butterworth, R.F. and Wells, J. Detection of benzodiazepines in hepatic encephalopathy: reply. Hepatology, 1995;21:604–5. 49. Ahboucha, S., Pomier Layrargues, G., Mamer, O. et al. Increased brain concentrations of a neuroinhibitory steroid in human hepatic encephalopathy. Ann Neurol, 2005;58:169–70.

50. Butterworth, R.F. Neurosteroids in hepatic encephalopathy: novel insights and new therapeutic opportunities. J Steroid Biochem Mol Biol, 2015;160:94–7. 51. Bemeur, C. and Butterworth, R.F. Nutrition in the management of cirrhosis and its neurological complications. J Clin Exp Hepatol, 2014;4(2):141–50. 52. Merli, M., Giusto, M., Lucidi, C. et al. Muscle depletion increases the risk of overt and minimal hepatic encephalopathy: results of a prospective study. Metab Brain Dis, 2013;28:281–84. 53. Amodio, P., Bemeur, C., Butterworth, R.F. et al. The nutritional management of hepatic encephalopathy in patients with cirrhosis: International Society for Hepatic Encephalopathy and Nitrogen Metabolism consensus. Hepatology, 2013;58:325–36. 54. Soulsby, C.T. and Morgan, M.Y. Dietary management of hepatic encephalopathy in cirrhotic patients; survey of current practice in United Kingdom. BMJ, 1999;31:8391. 55. Cordoba, J., Lopez‐Hellin, J., Planas, M. et al. Normal protein diet for episodic hepatic encephalopathy: results of a randomized study. J Hepatol, 2004;41(1):38–43. 56. Bresci, G., Parisi, G., and Bant, S. Management of hepatic encephalopathy with oral zinc supplementation: a long‐term treatment. Eur J Med, 1993;2(7):414–6. 57. Merli, M., Giusto, M., Gentili, F. et al. Nutritional status: its influence on the outcome of patients undergoing liver transplantation. Liver Int, 2009;30:208–14. 58. Sharma, P. and Sharma, B.J. Management of overt hepatic encephalopathy. J Clin Exp Hepatol, 2015;5:S82–7. 59. Als‐Nielsen, B., Gluud, L.L., and Gluud, C. Non‐absorbable disaccharides for hepatic encephalopathy: systematic review of randomized trials. BMJ, 2004. doi:10.1136/bmj.38048.506134.EE 60. Morgan, M.Y., Blei, A., Grungreiff, K. et  al. The treatment of hepatic encephalopathy. Metab Brain Dis, 2007;22:389–405. 61. Prasad, S., Dhiman, R.K., Duseja, A. et  al. Lactulose improves cognitive functions and health‐related quality of life in patients who have minimal hepatic encephalopathy. Hepatology, 2007;4:549–59. 62. Sharma, P., Sharma, B.C., Agrawal, A. et al. Primary prophylaxis of overt hepatic encephalopathy in patients with cirrhosis: an open‐labelled randomized controlled trial of lactulose versus no lactulose. J Gastroenterol Hepatol, 2012;27:1329–35. 63. Gluud, L.L., Vilstrup, H., and Morgan, M.Y. Non‐absorbable disaccharides versus placebo/no intervention and lactulose versus lactitol for the prevention and treatment of hepatic encephalopathy in people with cirrhosis. Cochrane Database Syst Rev, 2016;4, CD003044. 64. Bass, N.M., Mullen, K.D., Sanyal, A. et al. Rifaximin treatment in hepatic encephalopathy. N Engl J Med, 2010;362:1071–81. 65. Kimer, N., Krag, A., Moller, S. et al. Systematic review with meta‐analysis: the effects of rifaximin in hepatic encephalopathy. Aliment Pharmacol Ther, 2014;40:123–32. 66. Sidhu, S.S., Goyal, O., Mishra, B.P. et al. Rifaximin improves psychometric performance and health‐related quality of life in patients with minimal hepatic encephalopathy (the RIME trial). Am J Gastroenterol, 2011;106:307–16. 67. Bajaj, J.S., Saoian, K., Christensen, K.M. et al. Probiotic yogurt for the treatment of minimal hepatic encephalopathy. Am J Gastroenterol, 2008;103:1707–15. 68. Mittal, V.V., Sharma, B.C., Sharma, P. et al. A randomized controlled trial comparing lactulose, probiotics and L‐ornithine L‐aspartate in treatment of minimal hepatic encephalopathy. Eur J Gastroenterol Hepatol, 2011;23:725–32. 69. Sharma, K., Pant, S., Misra, S. et al. Effect of rifaximin, probiotics and L‐ ornithine L‐aspartate on minimal hepatic encephalopathy: a randomized controlled trial. Saudi J Gastroenterol, 2014;20:225–32. 70. Rose, C., Michalak, A., Pannunzio, P. et al. L‐ornithine L‐aspartate in experimental portal‐systemic encephalopathy: therapeutic efficacy and mechanism of action. Metab Brain Dis, 1998;13:147–57. 71. Butterworth, R.F. and Grüngreiff, K. L‐ornithine L‐aspartate (LOLA) for the treatment of hepatic encephalopathy in cirrhosis: evidence for novel hepatoprotective mechanisms. JSM Liver Clin Res, 2019;3:5. 72. Kircheis, G., Nilius, R., Held, C. et al. Therapeutic efficacy of L‐ornithine L‐aspartate infusions in patients with cirrhosis and hepatic encephalopathy. Hepatology, 1997;25:1351–60.



48:  Hepatic Encephalopathy

73. Sidhu, S.S., Sharma, B.C., Goyal, O. et al. L‐ornithine L‐aspartate in bouts of overt hepatic encephalopathy. Hepatology, 2018;67:700–10. 74. Butterworth, R.F., Kircheis, G., Hilger, N. et al. Efficacy of L‐ornithine L‐ aspartate for the treatment of hepatic encephalopathy and hyperammonemia in cirrhosis: systematic review and meta‐analysis of randomized controlled trials. J Clin Exp Hepatol. 2018;8(3):301–13. 75. Zhu, G.Q., Shi, K.Q., Huang, S. et al. Systematic review with network meta‐ analysis: the comparative effectiveness and safety of interventions in patients with overt hepatic encephalopathy. Aliment Pharmacol Ther, 2015;41:624–35. 76. Acharya, S.K., Bhatia, V., Sreenivas, V. et  al. Efficacy of L‐ornithine L‐ aspartate in acute liver failure: a double‐blind, randomized, placebo‐controlled study. Gastroenterology, 2009;136:2159–68. 77. Marchesini, G., Bianchi, G., Merli, M. et  al. Nutritional supplementation with branched chain amino acids in advanced cirrhosis: a double‐blind randomized trial. Gastroenterology, 2003;124:1792–801. 78. Muto, Y., Sato, S., Watanabe, A. et al. Effects of oral branched‐chain amino acid granules on event‐free survival in patients with liver cirrhosis. Clin Gastroenterol Hepatol, 2005;3(7):705–13. 79. Sushma, S., Dasarathy, S., Tandon, R.K. et al. Sodium benzoate in the treatment of acute hepatic encephalopathy: a double‐blind randomized trial, Hepatology, 1992;16:138–44. 80. Rockey, D.C., Vierling, J.M., Mantry, P. et  al. Randomized double‐blind, controlled study of glycerol phenylbutyrate in hepatic encephalopathy, Hepatology, 2014;59:1073–83. 81. Stravitz, R.T., Gottfried, M., Durkalski, V. et al. Safety, tolerability and pharmacokinetics of L‐ornithine phenylacetate in patients with acute liver injury/ failure and hyperammonemia. Hepatology, 2018;67:1003–13. 82. Malaguarnera, M., Pisone, G., Astutu, M. et al. L‐carnitine in the treatment of mild or moderate encephalopathy. Digest Dis, 2003;21:271–75. 83. Ratnakumari, L., Qureshi, I.A., and Butterworth, R.F. Effect of L‐carnitine on cerebral and hepatic energy metabolites in congenitally hyperammonemic sparse‐fur mice and its role during benzoate therapy. Metabolism, 1993;42:1039–46. 84. Therrien, G., Butterworth, J., Rose, C. et al. Protective effect of L‐Carnitine in ammonia‐precipitated encephalopathy in portacaval‐shunted rats: evidence for a central mechanism of action. Hepatology, 1997;25:551–6. 85. Pomier Layrargues, G., Giguere, J.F., Lavoie, J. et al. Flumazenil in cirrhotic patients in hepatic coma: a randomized double‐blind placebo‐controlled crossover trial. Hepatology, 1994;19:32–7. 86. Als‐Nielsen, B., Gluud, L.L., and Gluud, C. Benzodiazepine receptor antagonists for hepatic encephalopathy. Cochrane Database Syst Rev, 2004;2:CD002798.

629

87. Turkmen, S., Lundgren, P., Birzniece, V. et  al. 3Beta‐20beta‐dihydroxy‐5‐ alpha‐pregnane (UC1011) antagonism of the GABA potentiation and the learning impairment induced in rats by allopregnanolone. Eur J Neurosci, 2004;20:1604–12. 88. Johansson, M., Agusti, A., Llansola, M. et  al. GR3027 antagonizes GABA‐A receptor‐potentiating neurosteroids and restores spatial learning and motor coordination in rats with chronic hyperammonemia and hepatic encephalopathy. Am J Physiol Gastrointest Liver Physiol, 2015;309:G400–9. 89. Cauli, O., Rodrigo, R., Piedrafita, B. et al. Inflammation and hepatic encephalopathy: ibuprophen restores learning ability in rats with portacaval shunts. Hepatology, 2007;46:514–9. 90. Chastre, A., Belanger, M., Beauchesne, E. et  al. Inflammatory cascades driven by tumor necrosis factor alpha play a major role in the progression of acute liver failure and its neurological complications. PloS One, 2012;7:e49670. 91. Jiang, W., Desjardins, P., and Butterworth, R.F. Cerebral inflammation contributes to encephalopathy and brain edema in acute liver failure: protective effect of minocycline. J Neurochem, 2009;109:485–93. 92. Vaquero, J. and Butterworth, R.F. Mild hypothermia for the treatment of acute liver failure: what are we waiting for? Nat Clin Pract Gastroenterol Hepatol, 2007;10:528–9. 93. Michel, H., Solere, M., Granier, P. et  al. Treatment of cirrhotic hepatic encephalopathy with L‐DOPA. A controlled trial. Gastroenterology, 1980;29:555–61. 94. Uribe, M., Farca, A., Marquez, M.A. et al. Treatment of chronic portal‐systemic encephalopathy with bromocriptine: a double‐blind controlled trial. Gastroenterology, 1979;76:1347–51. 95. Morgan, M.Y., Jacobovits, A.W., James, I.M. et al. Successful use of bromocriptine in the treatment of chronic hepatic encephalopathy. Gastroenterology, 1980;78:663–70. 96. Butterworth, R.F. Parkinsonism in cirrhosis: pathogenesis and current therapeutic options. Metab Brain Dis, 2013;28(2):261–7. 97. Sotil, E.U., Gottstein, J., Ayala, E. et al. Impact of preoperative overt hepatic encephalopathy on neurocognitive function after liver transplantation. Liver Transpl, 2009;15:184–92. 98. Bajaj, J.S., Schubert, C.M., Heuman, D.M. et  al. Persistence of cognitive impairment after resolution of overt hepatic encephalopathy. Gastroenterology, 2010;138:2332–40. 99. Saner, F.H., Gu, Y., Minouchehr, S. et al. Neurological complications after cadaveric and living donor transplantation. J Neurol, 2006;253:612–7.

49

The Kidney in Liver Disease Moshe Levi1, Shogo Takahashi1, Xiaoxin X. Wang1, and Marilyn E. Levi2 Department of Biochemistry and Molecular and Cellular Biology, Georgetown University, Washington, DC, USA 2 Department of Medicine, Division of Infectious Diseases, University of Colorado, Aurora, CO, USA 1

IMPAIRED RENAL FUNCTION IN LIVER DISEASE Impaired renal function is common in liver diseases, either as part of multi‐organ involvement in acute illness or secondary to advanced liver disease [1–4].

kidney’s ability to counterbalance the effects of vasoconstrictors on the renal circulation, (v) increased gut permeability resulting in bacterial translocation, and (vi) systemic inflammation with release of cytokines, damage‐associated molecular patterns (DAMPS), and reactive oxygen species (ROS). These hemodynamic factors can accelerate kidney injury in the presence of intrinsic kidney disease caused by various disorders.

ACUTE KIDNEY INJURY IN LIVER DISEASE

HEPATORENAL SYNDROME

Acute kidney injury (AKI) previously known as acute renal failure (ARF) or acute tubular necrosis (ATN) occurs in approximately 20% of hospitalized patients with cirrhosis. In contrast, the incidence of AKI in acute liver failure varies from 40 to 80%, especially in the setting of infections such as viral hemorrhagic fever, leptospirosis, bacterial peritonitis, and ­ toxin‐induced injuries such as acetaminophen poisoning, administration of nephrotoxic antibiotics (such as aminoglycoside antibiotics), or high doses of non‐steroidal anti‐inflammatory drugs. The mechanisms leading to acute renal injury are multiple and additive and include: (i) changes in the systemic arterial circulation, including systemic arterial vasodilatation and sometimes in certain alcoholic patients the concomitant presence of cardiomyopathy, (ii) portal hypertension, (iii) activation of renal vasoconstrictive hormones, including the renin–angiotensin system, the sympathetic nervous system (SNS), arginine ­vasopressin, endothelin (ET), thromboxane A2, and leukotrienes, which all additively impair renal blood flow, (iv) suppression of renal vasodilatory factors including prostacyclins which impair the

Hepatorenal syndrome (HRS) is a unique form of functional renal failure that often complicates advanced liver disease, hepatic failure, or portal hypertension. The International Club of Ascites has recently updated the diagnostic criteria for AKI and HRS‐AKI [5, 6], which are outlined in Table 49.1. In cirrhotic patients with ascites, most of the cases of acute renal failure are mediated by: (i) pre‐renal failure and (ii) acute kidney injury (acute tubular necrosis). HRS accounts for approximately 20% of renal failure. HRS is characterized by: (i) an extreme expression of the profound circulatory dysfunction in cirrhosis, with marked splanchnic arterial and systemic vasodilatation, insufficient cardiac output, severe reduction of effective blood volume, homeostatic activation of vasoactive systems, and intense renal vasoconstriction, resulting in a critical decrease in renal blood flow and glomerular filtration rate, (ii) absence of pathological changes in renal tissue (vasculitis, glomerulonephritis, or tubulointerstitial ­fibrosis), and (iii) preserved renal tubular function [7]. Precipitating events that cause the development of HRS include: (i) bacterial peritonitis and/or sepsis of other origin,

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



49:  The Kidney in Liver Disease

Table 49.1  Current diagnostic and treatment criteria for HRS‐AKI (A) Diagnostic criteria of HRS‐AKI Cirrhosis and ascites Diagnosis of AKI according to ICA‐AKI criteria: increase in SCr greater than or equal to 0.3 mg dL−1 within 48 hours Absence of shock No response after two consecutive days of diuretic withdrawal and plasma volume expansion with albumin (1 g kg−1 of body weight) No current or recent use of nephrotoxic drugs (NSAIDs, aminoglycosides, iodinated contrast media, etc.) No macroscopic signs of structural kidney injury, defined as: • absence of proteinuria (greater than 500 mg d−1) • absence of microhematuria (greater than 50 RBCs per high power field), • normal findings on renal ultrasonography (B) Treatment criteria of HRS‐AKI Meeting all the diagnostic criteria of HRS‐AKI AKI stage greater than or equal to 1B (after plasma albumin expansion) No contraindication to vasoconstrictor therapy Criteria for treatment individualized AKI, acute kidney injury; HRS‐AKI, hepatorenal syndrome‐AKI; ICA, International Club of Ascites; NSAIDs, non‐steroidal anti‐inflammatory drugs; RBCs, red blood cells. Reproduced with permission of John Wiley & Sons [5].

(ii) increased doses of diuretics, (iii) large‐volume paracentesis without volume expansion, and (iv) gastrointestinal hemorrhage, which all result in further compromise of the systemic and renal circulation. Clinically, two different forms of HRS can be distinguished: type 1 HRS (HRS‐1) is characterized by rapid progression of renal failure, whereas type 2 HRS (HRS‐2) is characterized by a less severe and more stable form of renal impairment. In recent years, there have been several potentially life‐saving and/or survival‐prolonging therapies aimed at patients with HRS type 1 (HRS‐1) [8–10].

Liver transplantation Liver transplantation, when possible, is the ideal form of treatment for HRS and end‐stage liver disease. Although subjects with HRS do have more frequent complications both pre‐ and post‐operation, including AKI requiring dialysis, nevertheless the overall survival rate averages over 50% at 3 years. However, due to ongoing organ shortage, most patients with end‐stage liver disease and HRS die before a liver becomes available. Therefore, alternative temporizing measures become very important [10].

Use of vasoconstrictors and albumin There is increasing evidence that shows that the use of vasoconstrictors especially terlipressin in combination with plasma expanders such as intravenous albumin may reverse HRS by decreasing splanchnic vasodilatation and increasing effective arterial blood volume, resulting in decreases in endogenous renal vasoconstrictors and increased renal perfusion and ­glomerular filtration rate. Recently two randomized (phase 3) studies OT‐0401 and REVERSE compared the efficacy of terlipressin plus albumin versus placebo plus albumin in patients with HRS‐1 [8]. Pooled patient‐level data were analyzed for HRS reversal (serum creatinine [SCr] value less than or equal to 133 μmol L−1], 90‐day survival, need for renal replacement therapy, and

631

predictors of HRS reversal. Patients received intravenous terlipressin 1–2 mg every 6 hours plus albumin or placebo plus albumin up to 14 days. The pooled analysis comprised 308 patients (terlipressin: n = 153; placebo: n = 155). HRS reversal was significantly more frequent with terlipressin versus placebo ­ (27% versus 14%; P = 0.004). Terlipressin was associated with a more significant improvement in renal function from baseline until end of treatment, with a mean between‐group difference in SCr concentration of −53.0 μmol L−1 (P < 0.0001). Lower SCr, lower mean arterial pressure, and lower total bilirubin and absence of known precipitating factors for HRS were independent predictors of HRS reversal and longer survival in terlipressin‐treated patients. Terlipressin plus albumin therefore resulted in a significantly higher rate of HRS reversal versus albumin alone in patients with HRS‐1 (ClinicalTrials.gov identifier: OT‐0401, NCT00089570; REVERSE, NCT01143246). While terlipressin is widely used in Europe and Asia, it is not yet available in the United States and instead intravenous albumin infusion plus noradrenaline has been used. The combination vasoconstrictive therapy of midodrine plus octreotide has also been used. However, the combination of midodrine plus octreotide is not as effective as terlipressin or noradrenaline [9, 11].

VIRAL INFECTIONS There are several viral diseases that affect both the liver and the kidney, including hepatitis B and hepatitis C. Viral infections can cause renal injury by a number of mechanisms: (i) circulating immune complexes that consist of viral antigens and the host antiviral antibodies; (ii) in situ immune‐mediated mechanisms involving viral antigens bound to glomerular structures; (iii) expression of viral proteins or pathogenic proinflammatory factors including cytokines, chemokines, and growth factors in renal tissue; and (iv) direct cytopathogenic effects on glomerular and tubulointerstitial cells.

Hepatitis B Aside from hepatitis, hepatitis B causes several extrahepatic manifestations, including reactive arthritis, skin rashes, vasculitis, and glomerulonephritis. Three forms of glomerulonephritis are associated with ­hepatitis B: (i) membranous glomerulonephritis is most frequently reported in the Asian population and in children, (ii) membranoproliferative glomerulonephritis, and (iii) IgA nephropathy, most commonly seen in adults. The causal role of hepatitis B virus (HBV) infection in glomerulonephritis has been suggested by the finding of viral antigens in immune complex deposits in the glomeruli and virus‐specific cytotoxic T lymphocytes. This is particularly relevant in kidney transplant recipients, where graft function may be impacted. Hepatitis B glomerulonephritis has a more favorable course in children than adults. Adults with nephrotic syndrome and abnormal liver function tests have the worst prognosis and are associated with progression to end‐stage renal disease (ESRD).

632

THE LIVER: VIRAL INFECTIONS

Lamivudine, emtricitabine, tenofovir, and entecavir inhibit hepatitis B DNA polymerase and are used for treatment, with the caveat that resistance may develop, particularly with lamivudine. Therefore, treatment of hepatitis B with these agents will require monitoring for breakthrough viremia. Lamivudine and emtricitabine, nucleoside reverse transcriptase inhibitors used for the treatment of hepatitis B and HIV requires a lower dose for the treatment of hepatitis B monoinfection. Tenofovir disoproxil fumarate (TDF) which is also used for the treatment of both hepatitis B and HIV is potentially nephrotoxic. The prodrug TDF is converted rapidly to tenofovir in the gut and circulates exclusively in plasma and filtered at the glomerulus. TNF is also actively transported into proximal renal tubular cells from the interstitial fluid and leads to mitochondrial toxicity and proximal tubular injury, resulting in Fanconi’s syndrome. Manifestation of Fanconi’s syndrome includes proteinuria, hypophosphatemia and phosphaturia, normoglycemic glycosuria, uricosuria, hyperuricemia, and aminoaciduria. By contrast, a newer prodrug of tenofovir, tenofovir alafenamide (TAF) has more potent anti‐HIV activity and is not considered nephrotoxic as it does not accumulate in proximal tubular cells [12]. Entecavir has similar mechanisms of nephrotoxicity although to a lesser degree. All of these agents are cleared by the kidney and may require dose adjustment.

Hepatitis C More than 80% of subjects acutely infected with hepatitis C develop chronic infection and chronic liver disease. Hepatitis C virus (HCV) infection is also associated with high prevalence of renal disease. HCV‐positive subjects have 40% higher odds for development of chronic kidney disease as compared with HCV‐ negative individuals. Hepatitis C is commonly associated with membranoproliferative glomerulonephritis, with and without cryoglobulinemia, membranous glomerulonephritis, focal segmental glomerulosclerosis (FSGS), and polyarteritis nodosa. Hepatitis C infection has also been associated with diabetes‐related nephropathy. Common laboratory findings in subjects with hepatitis associated glomerulonephritis include cryoglobulinemia C‐­ (mixed type II), increased monoclonal rheumatoid factor (IgM kappa), decreased early complements C4, C1q, and CH50, and normal or slightly decreased C3. Previous treatment of hepatitis C included the use of pegylated interferon (PEG‐IFN) and ribavirin with associated adverse reactions including fevers, anemia, and limited efficacy. More recently, PEG‐IFN has been replaced by directly acting antivirals (DAAs) that have revolutionized the treatment of hepatitis C, achieving cure in 95–98% of cases, defined as a sustained virologic response at twelve weeks, or SVR12. Successful treatment of HCV often leads to remission of cryoglobulinemic glomerulonephritis or mixed cryoglobulinemia. In the presence of severe vasculitis, rapidly progressive glomerulonephritis or nephrotic syndrome, immunosuppression should be initiated early to prevent progression of renal disease [13], including corticosteroids, rituximab, and plasma exchange [13]. There are three classes of DAAs: (i) NS3/4 protease inhibitors (PIs), (ii) NS5B polymerase inhibitors, and (iii) NS5A inhibitor class. Commonly, combination therapy using 1 or more DAAs from

different classes is used to decrease the development of resistance and enhance potency of the regimen. Regimens are determined by the HCV genotype, specifically genotypes 1, 4, 5, or 6. Some DAAs are potentially nephrotoxic such as sofosbuvir which is not recommended for patients with chronic kidney disease stage 4 or 5 with eGFR less than 30 or ESRD. Hepatitis C infection prevalence in patients with ESRD ranges between 3–70% depending on the country. Therefore, all dialysis patients require initial HCV antibody screening and if negative, repeated every six months. If the HCV antibody is positive, hepatitis C RNA should be obtained to look for evidence of viremia as well as a genotype. Treatment should be initiated in the presence of HCV viremia with regimens based on the genotype. Patients who are infected with hepatitis C will suppress hepatitis B viremia if coinfected. When these patients undergo treatment of hepatitis C with either PEG‐IFN or directly acting antivirals, the underlying hepatitis B may reactivate and progress [14]. Liver failure from hepatitis B resulting in death and liver transplantation has been reported. Consequently, the Food and Drug Administration placed a black box warning on DAAs, with recommendations to screen all patients with hepatitis C for any positive serologies for hepatitis B. The most common positive serology to be consistent with active hepatitis B is hepatitis B surface antigen (HBsAg), and if detected, indicates a risk for progression of hepatitis B when HCV is successfully treated. Therefore, if hepatitis B treatment is indicated, this should be initiated prior to DAA therapy for hepatitis C to avoid active hepatitis B ­infection. In the presence of a positive hepatitis B core antibody and negative HBsAg, the risk of hepatitis B reactivation is less likely. In this scenario, HBV DNA monitoring is performed monthly and continued for three months after DAA is completed [15].

Human immunodeficiency virus While HIV‐1 causes a wide spectrum of renal diseases, the role of HIV‐1 in causing primary hepatitis is less certain. Liver function abnormalities in HIV‐1‐infected individuals have been associated with drug adverse reactions and interactions, alcohol abuse, HCV or HBV coinfections, opportunistic infections such as HHV‐8 (associated with Kaposi’s sarcoma), and Hodgkin’s or non‐ Hodgkin’s lymphoma. With initiation of antiretroviral therapy (ART), autoimmune hepatitis and liver failure may develop as a complication of immune reconstitution syndrome (IRIS), a paradoxical worsening of the individual patient’s HIV‐associated condition at the time of initiation of antiretroviral therapy and may be responsive to corticosteroids. IRIS may develop in response to autoantigens associated with autoimmune conditions that may manifest after antiretroviral therapy is initiated such as systemic lupus erythematosus, Graves’ disease, sarcoidosis, and rheumatoid arthritis. IRIS may also develop in reaction to underlying pathogens such as Mycobacterium tuberculosis or opportunistic infections such as Pneumocystis jiroveci (formerly P. carinii), Mycobacterium avium complex, Toxoplasmosis gondii, Cryptococcus spp., Histoplasmosis capsulatum, Cytomegalovirus (CMV), and other Herpesviridae and JC virus [16]. IRIS has also been reported to unmask underlying Hodgkin’s and non‐Hodgkin’s lymphoma with potential liver infiltration within six months of initiation of ART. Management of IRIS includes treatment of



49:  The Kidney in Liver Disease

opportunistic infections, hepatitis B or C or lymphoma, and continuation of antiretroviral therapy with or without steroids. Antiretroviral drug hepatotoxicity has been well described. A previous study conducted in subjects with unexplained transaminase (ALT and AST) elevations in HIV‐1 monoinfected patients on antiviral therapy found a high rate of liver lesions, mostly consistent with non‐alcoholic steatohepatitis (NASH). It is now recognized that nucleoside reverse transcriptase inhibitors (NRTIs) including AZT, lamivudine, tenofovir, and emtricitabine may induce hepatitis steatosis. The NRTI abacavir induces a unique hypersensitivity reaction in patients who are positive for HLA B5701 and are at risk for liver failure and death. Therefore, a screening blood test to identify the presence of HLA B5701 should be performed prior to initiation of abacavir. Non‐nucleoside reverse transcriptase inhibitors (NNRTI) such as efavirenz may be associated with severe liver injury. Despite these observations, a more recent study following a cohort of 10 083 HIV infected patients in the United States between 2004–2010 showed infrequent elevations in aminotransferases. However, this study also noted that the risk of hepatotoxicity was greater with HIV protease inhibitor use in hepatitis B or C coinfected patients [17]. In HIV positive patients who are coinfected with hepatitis B and/or C, more rapid progression of hepatic decompensation including drug‐induced liver injury is seen compared to HIV monoinfected individuals. Treatment with antiretroviral therapy has been shown to control the advancement of hepatic failure and is an integral part of HBV and HCV management. Initial antiretroviral regimens in coinfected patients are the same as in HIV monoinfected patients with the caveat that drug selection may require adjustment due to potential drug interactions with directly acting agents. In addition, when treating HCV with DAA regimens that include PIs, HIV coinfected patients who are maintained on HIV protease inhibitors should be switched to an HIV non‐PI containing regimen. Kidney disease is a common complication of HIV infection. The spectrum of HIV‐associated renal disease includes direct association with infection, diseases that are linked with the systemic response to infection, diseases that develop because of superinfections, and diseases that are associated with treatment of HIV infection [12]. Specifically, AKI may result from acute tubular necrosis, acute tubulointerstitial nephritis, rhabdomyolysis, antiretroviral therapy, HIV‐associated nephropathy, HIV immune complex disease, thrombotic microangiopathies, and urinary tract obstruction. Chronic kidney disease (CKD) may also result from these causes. In addition, there are several potential opportunistic infections of the kidney parenchyma, including viruses such as cytomegalovirus (CMV) and BK, fungal, both typical and atypical mycobacterial infections, as well as infiltrative lesions of the kidney that may result from lymphoma and Kaposi’s sarcoma [12]. Current antiretroviral therapy regimens may result in chronic inflammation, premature aging, and metabolic syndrome and its complications including chronic kidney disease [18]. HIV patients of African descent have increased predisposition to HIV‐associated nephropathy due to variants in ApoL1 gene encoding apolipoprotein L1 [19, 20]. In addition to HIV‐ associated nephropathy, APOL1 risk variants are also associated with focal segmental glomerulosclerosis and hypertension‐ associated arterionephrosclerosis [19, 21].

633

Treatment of HIV‐associated nephropathy centers around use of highly active antiretroviral therapy and renin angiotensin aldosterone antagonists.

Viral infections in kidney transplant recipients Prevention of kidney graft rejection requires immunosuppressive agents that primarily inhibit T‐lymphocyte function. Consequently, these patients are at risk for viral complications so that vaccinations, including live vaccines should be administered prior to transplantation such as Zostavax for herpes zoster. Following transplantation, live vaccines are contraindicated due  to risk for vaccine‐induced infection in the presence of immunosuppression.

Specific issues related to transplantation 1. Reactivation of previous infections that have remained in the latent phase such as cytomegalovirus, herpes simplex, herpes zoster, and Epstein‐Barr viruses, all potentially manifest as acute hepatitis. Epstein‐Barr virus is also associated with post‐ transplant lymphoproliferative disorder (PTLD) that may involve the liver and responds to lowering of immunosuppression. Specific cases with CD20+ markers may respond to rituximab +/− chemotherapy including cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP). Reactivation may also occur with the use of antilymphocyte antibodies such as antithymocyte globulin (ATG) and thymoglobulin and requires the use of valganciclovir or ganciclovir for CMV prophylaxis. Reactivation may be predicted by the net state of immunosuppression, whereby excessive lowering of T and B cell function may impair the ability to control latent infections, resulting in acute reactivation [22]. Vaccination for herpes zoster reactivation Reactivation of BK virus, a polyomavirus is seen primarily in post‐kidney transplant recipients. This virus is believed to be acquired during childhood and is maintained in a latent phase in the genitourinary tract. During periods of immunosuppression, BK virus may reactivate and cause multiple complications including ureteral stenosis, hematuria, kidney graft dysfunction, and organ loss. Monitoring of BK in plasma and urine is used to predict graft dysfunction and is treated preemptively with lowering of immunosuppression. 2. Donor derived transmission of viral infections: this is commonly related to the presence of viral antibodies (e.g. CMV IgG) in the donor and absence of protective antibody in the recipient termed mismatches resulting in primary infections. Use of prophylaxis to prevent primary infections with the appropriate antivirals may prevent donor transmission such as valganciclovir or ganciclovir for cytomegalovirus for six months and acyclovir for herpes simplex and zoster. Prophylaxis for EBV is controversial, and EBV DNA monitoring is recommended. 3. Community acquired infections: The most common viral pathogens observed include Herpesviridae such as cytomegalovirus, herpes simplex, herpes zoster, and Epstein‐Barr, all of which may present as primary infections and involve the liver and kidneys. Adenovirus, a cause of hemorrhagic ­cystitis, nephritis, and graft failure in addition to hepatitis

634

THE LIVER:  NAFLD, NASH, AND CKD

responds to lowering of immunosuppression and use of cidofovir, although nephrotoxicity is a significant issue. A prodrug of cidofovir, brincidofovir is currently available on a compassionate use basis for adenoviral infections with no significant nephrotoxicity reported, although gastrointestinal intolerance is frequently seen. Influenza infections are a significant cause of morbidity and mortality in transplant recipients, and vaccination is imperative.

METABOLIC SYNDROME AND OBESITY It is now increasingly recognized that metabolic syndrome and obesity affect both the kidney and the liver. The proportion of obese adults worldwide has increased to 37% in men and similarly to 38% in women. In the United States the prevalence of obesity has similarly increased to 39.8% in adults [23].

population. The rising disease prevalence is expected to result in increasing number of patients with cirrhosis and end stage liver disease requiring liver transplantation and in addition an increase in HCC [31].

NAFLD, NASH, AND CKD There is now increasing evidence that NAFLD, is associated with a nearly 40% increase in the long‐term risk of incident CKD [1–4]. The mechanisms by which NAFLD increases CKD risk are not fully known. However, several factors including insulin resistance, altered fatty acid and cholesterol metabolism, mitochondrial dysfunction, oxidative stress, ER stress, inflammation, microbiome, altered bile acid metabolism, and increased activity of profibrogenic factors can result in progression of NAFLD and play a role on increased incidence of CKD (Figures 49.1 and 49.2).

OBESITY AND KIDNEY DISEASE

Insulin resistance

The obesity epidemic has resulted in an increased incidence of obesity‐related glomerulopathy (ORG), a distinct entity presenting as proteinuria, enlarged glomeruli, progressive glomerulosclerosis, and decline in renal function [24–26]. Pathologically glomerular hypertrophy and adaptive focal segmental glomerulosclerosis are characteristics of ORG. Obesity‐induced increases in glomerular filtration rate, renal plasma flow, filtration fraction, and renal tubular sodium reabsorption results in glomerular hypertrophy. In addition to the hemodynamic changes, insulin resistance, altered fatty acid and cholesterol metabolism, mitochondrial dysfunction, oxidative stress, ER stress, inflammation, microbiome, altered bile acid metabolism, and increased activity of profibrogenic factors also play a role in the pathogenesis of ORG. Approximately one third of the patients develop progressive increase in proteinuria and decline in renal function, resulting in ESRD. Renin angiotensin aldosterone system blockade, and especially weight loss, and bariatric surgery result in reductions of proteinuria and progression of kidney disease. Studies in experimental models of diet‐induced obesity and kidney disease indicate that activation of the bile acid regulated nuclear receptor farnesoid X receptor (FXR) and/or G protein‐ coupled receptor TGR5 protects against kidney injury. The mechanisms of action include prevention of lipid accumulation, oxidative stress, inflammation, ER stress, mitochondrial ­dysfunction, and fibrosis [27–30].

Insulin resistance is a common feature of NAFLD that also contributes to its pathogenesis. For example, insulin resistance in adipose tissue leads to release of fatty acids through dysregulated lipolysis of triglycerides. Insulin resistant adipocytes also secrete cytokines like IL6 that has proinflammatory effects in the liver [32, 33]. This results in further augmentation of insulin resistance. In addition, the fatty acids released by adipocytes are taken up by the liver, which contributes to increased lipogenesis and lipid droplet formation in the liver [32, 34]. Insulin resistance also plays an important role in pathogenesis of CKD by modulating renal hemodynamics, glomerular mesangial cells and podocytes, and renal tubular function [33, 35]. In insulin resistance despite increased renovascular resistance due to impaired nitric oxide (NO) generation, reduced tubuloglomerular feedback, and dilation of afferent arterioles due to increased reabsorption of glucose and sodium result in glomerular hyperfiltration. Reduced insulin signaling in mesangial cells results in mesangial cell hypertrophy, hypertrophy, and extracellular matrix deposition. Insulin resistance in podocytes results through lysosomal and proteasomal degradation of the insulin receptor by a nephrin dependent mechanism [36]. This induces podocyte apoptosis, effacement of its foot processes, albuminuria, thickening of the glomerular basement membrane, and increased glomerulosclerosis. The renal tubular consequences of insulin resistance are more complex but include gluconeogenesis in the proximal tubule and increased sodium absorption in the distal tubule [33, 35]. While the effects of insulin resistance per se in regulation of proximal tubule glucose reabsorption via SGLT‐2 is debatable, nevertheless SGLT‐2 expression and activity are increased in humans and animal models of obesity and diabetes [37]. SGLT‐2 inhibitors including empagliflozin and canagliflozin have renal and cardiovascular disease beneficial effects in human subject. Furthermore SGLT‐2 inhibitors decrease renal disease and liver disease in an animal model of diet‐induced obesity and insulin resistance [38].

OBESITY AND LIVER DISEASE Obesity and metabolic syndrome, especially in the presence of diabetes mellitus, has become the leading cause of non‐alcoholic fatty liver disease (NAFLD), in its whole spectrum of disease ranging from simple steatosis to NASH, cirrhosis, and hepatocellular carcinoma (HCC). The prevalence of NAFLD is projected to increase from the current estimate of 25% of the



49:  The Kidney in Liver Disease

635

Figure 49.1  Organ crosstalk in the pathophysiology of non‐alcoholic fatty liver disease (NAFLD) and chronic kidney disease (CKD). Many ­factors, such as caloric intake, dietary factors (such as high fructose consumption and low vitamin D levels), genetic factors, and inflammation of visceral adipose tissue can increase an individual’s risk of NAFLD. Increased amounts of non‐esterified fatty acids (NEFAs) resulting from the expansion of intra‐abdominal visceral adipose tissue are associated with an increase in NF‐κB and inflammatory pathways, dysregulation of ­adipokine production leading to decreased adiponectin levels, and impaired insulin signaling. Progression of liver dysfunction triggers pathways that might influence the development of CKD. For example, insulin resistance and atherogenic dyslipidemia as well as proinflammatory factors, prothrombotic factors, and profibrogenic molecules can promote vascular and renal damage. Reduced activation of the energy sensor, 5′‐AMP activated protein kinase (AMPK), in response to reduced adiponectin levels further stimulates proinflammatory and profibrogenic mechanisms. Activation of the renin‐angiotensin system (RAS) and endothelial cells might also contribute to liver and kidney dysfunction by increasing oxidative stress, inflammation, and coagulation pathways. Increased production of uremic toxins by intestinal microbiota in the setting of CKD might induce further renal, liver, and cardiovascular damage through inflammatory, oxidative, and fibrotic pathways. Dysbiosis of the intestinal microbiota, which often occurs with obesity, potentially influences NAFLD, CKD, and type 2 diabetes mellitus (T2DM) through complex mechanisms. Finally, cardiovascular disease (CVD), the risk of which is increased in the setting of NAFLD, T2DM, or intestinal dysbiosis, can influence the development of renal dysfunction (and vice versa) through cardiorenal interactions. AGE, advanced glycation end‐product; NASH, non‐alcoholic steatohepatitis; ROS, reactive oxygen species; SCFA, short‐chain fatty acid; TMAO, trimethylamine oxide. Reproduced with permission of Springer Nature, Figure 2 in [1].

Lipid metabolism Excessive lipid accumulation can occur ectopically in nonfat tissue, contributing to their damage through toxic processes named lipotoxicity. Liver lipid accumulation results from an imbalance between lipid acquisition (uptake of circulating fatty acids and de novo lipogenesis [DNL]) and lipid removal (fatty acid oxidation [FAO], export as a component of very low‐­ density lipoproteins [VLDL] particles, lipid droplet formation and lipolysis) (Figure 49.3). The hepatic lipid uptake is largely dependent on fatty acid transporters [39], predominately mediated by fatty acid transport proteins (FATPs), cluster of differentiation 36 (CD36), and caveolins located in the hepatocyte plasma membrane. Of the six

mammalian FATP isoforms, FATP2 and FATP5 are found primarily in the liver and their knockdown in mice decreases uptake of fatty acids and ameliorates hepatic steatosis. CD36 facilitates the transport of long‐chain fatty acids and is regulated by peroxisome proliferator‐activated receptor (PPAR) γ, ­ pregnane X receptor, and liver X receptor (LXR). CD36 expression in the liver is increased in NAFLD and is thought to ­stimulate increased uptake of fatty acids. The caveolins comprise a family of three membrane proteins for lipid trafficking and lipid droplets formation. Caveolin 1 is increased in the liver of mice with NAFLD, and mainly localized in the centrilobular zone three, where the steatosis is most severe. Following uptake, fatty acids in the cytosol need specific intracellular fatty acid binding proteins (FABP)

636

THE LIVER:  NAFLD, NASH, AND CKD

Figure 49.2  Potential mechanisms by which intestinal dysbiosis might promote the development of non‐alcoholic fatty liver disease (NAFLD) and chronic kidney disease (CKD). An imbalance in intestinal microbiota (dysbiosis) resulting in an increase in Gram‐negative bacteria leads to an increase in lipopolysaccharide (LPS) production, which can damage the intestinal epithelium resulting in increased gut permeability. This increased gut permeability enables further egress of bacterial content, LPS, and small molecules into the portal and systemic circulations causing inflammation. Dysbiosis also promotes the increased production of secondary bile acids such as deoxycholic acid. Following their return to the liver as part of the enterohepatic circulation, secondary bile acids have been linked to chronic inflammation, cholestasis and carcinogenesis through their actions on farnesoid X Receptor (FXR; with downstream effects on cholesterol metabolism) and by inducing cellular senescence and causing damage to mitochondrial and cell membranes. The intestinal microbiota also generates molecules such as trimethylamine (TMA), p‐cresol, and indole from dietary choline, phenylalanine/tyrosine, and tryptophan, respectively. TMA is oxidized in the liver to trimethylamine oxide (TMAO), which promotes atherosclerotic vascular disease. Indole and p‐cresol are metabolized in the liver to indole sulfate and p‐cresol sulfate, which are cleared by the proximal tubules and are potentially nephrotoxic. Other molecules produced by microbiota metabolism, such as phenylacetic acid and hippuric acid, are also potentially toxic to the kidney. Reduced levels of short‐chain fatty acids (SCFAs) as a consequence of dysbiosis can also induce decreased lipogenesis and increased gluconeogenesis, leading to insulin resistance, further dysfunction of the liver and kidney, and to the development of type 2 diabetes mellitus (T2DM). Reproduced with permission of Springer Nature, Figure 3 in [1].

to facilitate their transportation, s­ torage, and utilization. FABP1, also known as liver‐type FABP, is the predominant isoform in the liver. Its expression is increased in NAFLD but declines as the disease progresses to NASH. DNL results from the synthesis of new fatty acids from acetyl‐ CoA subunits produced primarily through glycolysis and the metabolism of carbohydrates. In addition to glucose which most commonly supplies carbon units for DNL, fructose also produces acetyl‐CoA that can enter the lipogenic pathway, important when considering the increasing use of fructose in corn syrup as a sweetener. DNL starts with acetyl‐CoA that is converted to malonyl‐ CoA by acetyl‐CoA carboxylase (ACC) and further converted to palmitate by fatty acid synthase (FASN). New fatty acid may then undergo a range of desaturation, elongation, and esterification steps before ultimately being stored as triglycerides in lipid droplets or exported as VLDL particles. Dysregulated DNL is a central feature of liver lipid accumulation in NAFLD patients [31]. Lipogenic genes are coordinately regulated at the transcription level. Transcription factors, such as sterol regulatory element‐ binding protein 1c (SREBP1c), which is activated by insulin and

LXRα, and carbohydrate regulatory element‐binding protein (ChREBP), which is activated by carbohydrates, play critical roles in this process. SREBP1c expression is enhanced in patients with NAFLD, while SREBP1c knockout mice display decreased expression of lipogenic enzymes. ChREBP is the bona fide transcription ­factor that is primarily responsive to glucose. ChREBP ­regulates not only enzymes in glucose metabolism, but also lipogenic enzymes. Therefore, ChREBP causes complex metabolic changes in loss‐ and gain‐of function studies. ChREBP expression in liver biopsies from patients with NASH is increased when steatosis is found to be greater than 50% but decreased in the presence of severe insulin resistance [40], indicating ChREBP might dissociate hepatosteatosis from insulin resistance. Recent studies further illustrate how different post‐­translational modifications, such as phosphorylation, acetylation, O‐GlcNAcylation, or ubiquitination, operate together in regulating lipogenic gene transcription. For example, deacetylation of SREBP‐1c by SIRT1 inhibited binding to its target lipogenic promoters [41]. Insulin signaling triggers activation of series of kinases downstream of PI3K, such as Akt and mTORC1/2, to regulate SREBP‐1c ­activity [42].



49:  The Kidney in Liver Disease

637

Figure 49.3  The substrate‐overload liver injury model of NASH pathogenesis. Free fatty acids are central to the pathogenesis of NASH. Free fatty acids that originate from lipolysis of triglyceride in adipose tissue are delivered through blood to the liver. The other major contributor to the free fatty acid flux through the liver is DNL, the process by which hepatocytes convert excess carbohydrates, especially fructose, to fatty acids. The two major fates of fatty acids in hepatocytes are mitochondrial beta‐oxidation and re‐esterification to form triglyceride. Triglyceride can be exported into the blood as VLDL or stored in lipid droplets. Lipid droplet triglyceride undergoes regulated lipolysis to release fatty acids back into the hepatocyte free fatty acid pool. PNPLA3 participates in this lipolytic process, and a single‐nucleotide variant of PNPLA3 is strongly associated with NASH progression, underscoring the importance of the regulation of this lipolysis. When the disposal of fatty acids through beta‐oxidation or formation of triglyceride is overwhelmed, fatty acids can contribute to the formation of lipotoxic species that lead to ER stress, oxidant stress, and inflammasome activation. These processes are responsible for the phenotype of NASH with hepatocellular injury, inflammation, stellate cell activation, and progressive accumulation of excess extracellular matrix. Lifestyle modifications that include healthy eating habits and regular exercise reduce the substrate overload through decreased intake and diversion of metabolic substrates to metabolically active tissues and can thereby prevent or reverse NASH. SCD, stearoyl CoA‐desaturase; FAS, fatty acid synthase; NKT, natural killer T cell; Tregs, regulatory T cells; PMNs, polymorphonuclear leukocytes. Reproduced with permission of Springer Nature, Figure 1 in [31].

FAO is controlled by nuclear receptor PPARα and occurs mainly in the mitochondria, providing a source of energy to generate ATP especially when circulating glucose concentrations are low [31]. Activation of PPARα induces the transcription of a range of genes related to FAO thereby reducing hepatic lipid levels. PPARα expression modulates not only lipid ­homeostasis, but inflammation as well. Expression of genes related to FAO was higher in patients with more severe steatosis compared to patients with less severe steatosis or non‐steatotic controls. FAO, measured indirectly as plasma β‐hydroxybutyrate levels, was higher in patients with NASH compared to steatosis or normal controls. Increased FAO may be an adaptive response in patients with NAFLD attempting to reduce the lipid overload and lipotoxicity, but it also produces ROS and excessive FAO may exhaust the capacity of the antioxidant defense system and induce oxidative stress. Accordingly, hepatic oxidative stress and changes in

mitochondrial ultrastructure were increased alongside FAO in patients with NASH. Antioxidant markers, glutathione, glutathione peroxidase, and superoxide dismutase were decreased in liver biopsies from NAFLD patients and in mitochondria from animal models of NAFLD [43]. Hepatic lipid can be exported from the liver into the blood after being packed into water‐soluble very low‐density lipoprotein (VLDL) particles alongside cholesterol, phospholipids, and apolipoproteins [44]. The assembly of VLDL particles occurs in the ER where apolipoprotein B100 (apoB100) is lipidated by the enzyme microsomal triglyceride transfer protein (MTTP). apoB100 and MTTP are key components in hepatic VLDL secretion. In patients with genetic defects in the apoB or MTTP gene, hepatic steatosis is common due to compromised triglyceride export. Excessive lipid not secreted into the blood can form lipid droplets in hepatocytes, a defining feature of NAFLD. The

638

THE LIVER:  NAFLD, NASH, AND CKD

triglycerides stored in lipid droplets have to be mobilized via hydrolysis to release fatty acids for utilization [45]. Triglyceride hydrolysis consists of sequential reactions of adipose triglyceride lipase (ATGL, annotated as patatin‐like phospholipase domain containing protein 2, PNPLA2 or desnutrin) hydrolyzing triglycerides to diacylglycerol (DG) and fatty acid, ­hormone‐sensitive lipase (HSL) breaking down DG to monoacylglycerol (MG) and fatty acid and monoacylglycerol lipase (MAGL) completing the lipolytic reaction of MG to glycerol and fatty acid [45]. Various studies suggest a critical role of intracellular lipid droplet homeostasis in regulating hepatic fat content. A tight regulation between lipid synthesis, hydrolysis, secretion, and FAO is required to prevent hepatic lipid overload and subsequent intracellular lipid accumulation, leading to steatosis, lipotoxicity, and liver damage, and promoting disease progression and fibrosis. The specific lipotoxic lipids that promote NASH phenotype include DG, ceramides, and lysophosphatidylcholine (LPC) species [31]. Hepatic free cholesterol is also considered a critical lipotoxic molecule in NASH [46]. The main pathways by which hepatocytes acquire cholesterol, include endogenous synthesis of cholesterol via master regulator SREBP‐2 and its target 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), the rate‐limiting enzyme; uptake of LDL and chylomicron remnants via LDL receptor‐mediated endocytosis and subsequent processing through the endosomal/lysosomal compartment; and uptake of HDL cholesterol directly via scavenger receptor class B type I (SR-BI). The main pathways by which cholesterol can be removed from hepatocytes, include conversion to bile acids by the rate‐limiting enzymes CYP7A1 and CYP27A1 and excretion of bile acids into bile by bile salt export pump (BSEP); excretion of cholesterol into bile by ATP‐binding cassette sub‐ family G member 5 and 8 (ABCG5/G8); incorporation into VLDL and secretion into the circulation; and efflux of cholesterol onto circulating apolipoprotein AI and nascent HDL particles. Extensive dysregulation of cholesterol homeostasis has been documented in NAFLD, causing both increased synthesis and uptake of cholesterol as well as decreased removal of cholesterol, and leading to increased hepatic cholesterol levels. Among the agents that target lipotoxicity [31, 47], aramchol, is an antagonist for stearoyl-CoA desaturase-1 (SCD-1), which is an enzyme that catalyzes a rate‐limiting step in the synthesis of monounsaturated fatty acids such as oleic acid. This agent also activates cholesterol efflux by stimulating the ABCA1 transporter, a widely expressed cholesterol export pump. A multicenter phase 2b trial (NCT02279524) in NASH patients is ongoing. For another enzyme in DNL, ACC, two inhibitors PF‐05221304 and GS‐0976 are in clinical trials in patients with NAFLD and NASH. In the kidney, abnormal lipid metabolism promotes increased triglyceride and cholesterol accumulation [24]. Renal triglyceride accumulation can occur because of increased fatty acid synthesis mediated by SREBP‐1c and its target enzymes including ACC, FASN, and SCD‐1; and/or mediated by ChREBP and its target enzymes including liver‐pyruvate kinase (L‐PK). In the murine high‐fat diet‐induced obesity model SREBP‐1 expression and activity results in increased renal lipid accumulation and renal disease [48–50]. The effects of high‐fat diet on kidney disease are prevented in SREBP‐1c knockout mice. Increased SREBP‐1

expression is also found in the glomeruli of patients with obesity related glomerulopathy [51]. In a search for in vivo inhibitors of SREBP‐1, the FXR agonists have been demonstrated to inhibit SREBP‐1 expression, lipid accumulation, inflammation, and fibrosis in animal models of diet‐induced obesity and insulin resistance [28, 52, 53]. Renal triglyceride accumulation can also occur by increased uptake via CD36 or FATPs. Renal CD36 expression is upregulated in patients with CKD, particularly those with diabetic nephropathy [54]. Blockade or knockout of CD36 can prevent kidney injury in experimental animals. Decreased FAO mediated by PPARα and its target enzymes is also responsible for renal triglyceride accumulation and is reported in kidney biopsy samples from patients with ORG and diabetes [55]. In this regard, PPARα agonist fenofibrate has been shown to prevent development of kidney disease [24]. Studies in mice with high fat‐induced obesity have shown increased cholesterol synthesis and accumulation as well as the  development of renal disease [49, 56] associated with increased expression and activity of SREBP‐2, a master regulator of cholesterol synthesis and cholesterol metabolism. Statin treatment has also been shown to reduce lipid accumulation in the proximal tubules in high‐fat diet‐fed mice. Increased cholesterol accumulation can also occur because of decreased cholesterol efflux. The nuclear receptor LXR plays a major role in cholesterol efflux and its expression and target enzymes that regulate cholesterol efflux, including ABCA1 and ABCG1, are decreased in human kidney biopsy samples from diabetic patients with ORG [55] and in ­animal models of diabetes [57]. Treatment with LXR agonists or induction of cholesterol efflux with cyclodextrin can reduce cholesterol accumulation and improves renal disease in diabetic mice.

Mitochondrial dysfunction Lipotoxicity is mediated by different cellular mechanisms including direct damage to mitochondria [58]. Mitochondria, the powerhouse of the cells, play a pivotal role in the final oxidation of metabolites such as fatty acids and glucose. The mitochondrial oxidative phosphorylation (OXPHOS) is responsible for the production of most cellular total energy. In addition, it is also involved in the regulation of intrinsic signaling pathways of apoptosis. Since the first studies on NAFLD, much evidence pointed out that it was primarily characterized by the presence of mitochondrial dysfunction [59]. The alteration of mitochondrial functions is evident with electron microscopy analysis by some ultrastructural changes such as megamitochondria, loss of cristae, and paracrystalline inclusion bodies in the matrix. When the disposal of fatty acids through FAO or formation of triglyceride is overwhelmed in NAFLD, fatty acids can contribute to the formation of lipotoxic species that lead to oxidative stress and inflammation [31]. Oxidative stress can lead to the peroxidation of phospholipids, such as cardiolipin (diphosphatidyl glycerol), one of the major phospholipids in the inner mitochondrial membrane. Since cardiolipin is assumed to enhance the respiratory chain activity, particularly of the complex I, the oxidation of cardiolipin leads to an imbalance of OXPHOS. Oxidative stress‐ induced damage to mitochondrial DNA can also result in the impairment of OXPHOS and further increases the generation of  ROS. The mitochondrial damage may eventually lead to



49:  The Kidney in Liver Disease

apoptotic death of hepatocytes. An increased expression of apoptotic proteins and enzymes, such as Bcl‐2, was found in the murine models of fatty liver disease. The apoptosis along with the generated cytokines from the liver stellate and Kupffer cells further augment the fibrotic changes to advance the disease. Sirtuins are a group of nicotinamide adenine dinucleotide (NAD)‐dependent deacetylases, share multiple cellular functions related to mitochondrial energy homeostasis and antioxidant activity in liver and kidney diseases [60, 61]. SIRT1, the most studied of these enzymes, has an indirect regulatory effect on oxidative stress, activating forkhead proteins and PGC‐1α, transcription factors involved in transcription of mitochondrial biogenesis genes and antioxidant enzyme genes and in ROS‐detoxifying capacity. SIRT1 level is decreased in a rat model of NAFLD. SIRT3 localizes in the mitochondrial matrix and can increase FAO by activation of long‐chain acyl‐CoA dehydrogenases, and its activity is found decreased in animal models with fatty liver. NAD+, as the co‐substrate for SIRT1 and SIRT3, controls various metabolic improvements. In this way, NAD+ depletion may contribute to mitochondrial dysfunction, obstructing the adaptive response mediated by sirtuins to high hepatic lipid levels. Hence, alleviation of the mitochondrial impairment, particularly in the early stages of NAFLD, may prevent the progression of the disease. Among the various experimentally studied mitochondrial‐targeted agents, triphenylphosphonium cation ligated ubiquinone Q10 and vitamin E, Szeto‐Scheller peptides, and superoxide dismutase mimetic‐salen manganese complexes (EUK‐8 and EUK‐134) have been found to be most promising [47]. Mitochondrial dysfunction is also the main cause of renal pathology induced by high fat diet. Given that the kidney is an organ that demands continuous high‐energy provision, mostly from FAO, lipid overload and impaired FAO lead to a disturbance in fatty acid uptake and utilization, further aggravating lipid accumulation in kidney cells and tissue [62]. This will create a vicious cycle whereby lipids can damage mitochondria and generate ROS, further limiting mitochondrial FAO and causing more cellular lipid accumulation, resulting in more mitochondrial ROS levels. As an approach to target mitochondria dysfunction, activation of membrane bile acid receptor TGR5 has been shown to decrease mitochondrial ROS generation and increase mitochondrial biogenesis, mitochondrial antioxidant generation, and mitochondrial FAO in kidney disease in obesity and diabetes [29].

Oxidative stress An overload of fatty acids into mitochondria, after an increased intake or an insulin‐resistance condition, may lead to an increase in the permeability of the inner mitochondrial membrane. This occurrence leads to the dissipation of the membrane potential and the loss of ATP synthesis capacity, resulting in mitochondrial function impairment and an enhanced ROS generation. The increase of FAO, inducing an increased electron flux in the electron transport chain (ETC), may generate an “electron leakage” due to reduction of the activity of ETC complexes, thus ensuring a direct reaction between electrons and oxygen, leading to the formation of ROS, rather than the normal reaction mediated by cytochrome c oxidase that combines oxygen and protons in order to form water [63]. The

639

incomplete or suboptimal FAO leads to accumulation of long chain acylcarnitines, ceramides, and diacylglycerols, lipotoxic intermediates that may act as indirect sources of ROS. In addition to pro‐oxidant mechanisms, in an experimental model of NASH, a decreased activity of several detoxifying enzymes was observed. Glutathione peroxidase (GPx) activity is reduced probably in consequence of GSH depletion and impaired transport of cytosolic GSH into the mitochondrial matrix [64]. The polymorphism C47T of the SOD2 gene, encoding for manganese superoxide dismutase, is associated with a reduction of activity of this enzyme resulting in an increased ROS production and a high susceptibility to developing NASH and advanced fibrosis in NAFLD [65]. The oxidative stress‐induced impairment of mitochondrial permeability can generate an influx of calcium (Ca2+) and iron into the mitochondria. Iron in the presence of H2O2 favors Fenton’s reaction to generate more hydroxyl radicals. Further, Ca2+ can stimulate the inducible NO synthase‐mediated nitric oxide radical (NO●) production and apoptotic cell death. A significant increase in the NO● level during severe hepatocyte injury may induce the progression to necrotic cell  death [66]. The peroxynitrite produced from NO● and superoxide radical is one of the important mediators of free radical toxicity.

Endoplasmic reticulum stress The ER stress and the unfolded protein response (UPR) play important roles in NAFLD/NASH and the associated kidney disease. The activation of ER stress may initially be protective, while chronic ER stress may result in apoptosis and fibrosis. The UPR activation in the ER is mediated by three major signaling pathways, including activating transcription factor 6 (ATF6), inositol‐requiring enzyme 1α (IRE1α), and PRKR‐ like ER kinase (PERK). Upon accumulation of misfolded proteins in the ER, BiP (GRP78) disassociates from the sensors and binds to unfolded proteins, which then activates the sensors. In the liver while lipid accumulation in hepatocytes can initiate ER stress, activation of ER stress can also induce lipogenesis and thus creating a vicious cycle for the progression of steatosis. ER stress induces insulin induced gene (INSIG) degradation and activation of the SREBP pathway [32, 67]. In addition, ER stress also induces increased expression of the VLDL receptor which results in increased lipoprotein delivery to the liver [32]. Increased lipid accumulation and chronic ER stress in the hepatocyte results increased inflammation and causes hepatocyte apoptosis. In addition, ER stress also stimulates hepatic stellate cell (HSC) collagen I secretion, which mediates fibrosis [68]. CHOP, the C/EBP homologous protein, plays an important role in ER stress induced apoptosis and fibrosis [68]. In the kidney chronic ER stress and UPR activation [69], as it occurs in the setting of obesity and the metabolic syndrome [30], contributes to podocyte injury, apoptosis, proteinuria, and CKD. Furthermore, albumin has been shown to induce ER stress in renal tubular cells by increasing intracellular calcium levels, inducing UPR‐dependent upregulation of lipocalin 2, resulting in apoptosis.

640

THE LIVER:  NAFLD, NASH, AND CKD

Inflammation Inflammation is a physiological response to tissue injury or infection that leads to secretion of various inflammatory mediators, such as cytokines, chemokines, and eicosanoids, which coordinate cellular defense mechanisms and tissue repair. The persistence of inflammatory activity over time results in chronic inflammatory changes that exacerbate tissue injury and may result in an abnormal wound healing response. In the case of NAFLD, inflammation contributes to the development of NASH and liver fibrosis. NAFLD refers to a spectrum of histological abnormalities in the liver, ranging from isolated steatosis with no or minimal inflammatory activity and no evidence of cell damage (NAFL) to NASH, which is characterized by steatosis, inflammation, and hepatocellular injury, hallmarked by the presence of hepatocyte ballooning, with different degrees of fibrosis [70]. Many factors are associated with chronic systemic immune response, a phenotype that is common in many individuals with NAFLD, CKD, T2DM, or CVD. A central issue in this field is identification of those factors that trigger inflammation, thus fueling the transition from non‐alcoholic fatty liver to NASH [71]. These triggers of liver inflammation may have their origins inside as well as outside of the liver. Lipotoxicity is one of the major mechanisms underlying hepatocyte dysfunction leading to disease progression in NASH [72]. Increased hepatocyte free fatty acid (FFA) influx results in the formation of lipotoxic intermediates in the liver, such as ceramides, diacylglycerols, LPC, and oxidized fatty acid and cholesterol metabolites, which act as ROS [73]. Recent studies have revealed that hepatocyte inflammasome activation may be an important link between the initial metabolic stress and subsequent hepatocyte death and stimulation of fibrosis in NASH [74]. The accumulation of free cholesterol in Kupffer cells and hepatic stellate cells can also activate the NLRP3 inflammasome [75]. While liver and kidney are important organs for lipid metabolism, the increased proinflammatory cytokine production by lipotoxicity, which causes ectopic lipid deposition that occurs in more advanced forms of NAFLD is also likely to have a pathogenetic role in the development of extra‐hepatic complications, such as CVD and CKD. Deterioration to NASH results in the production or activation of multiple inflammatory mediators such as NF‐κB pathway, cytokines [76], lipopolysaccharides (LPS), and ROS, which can lead to insulin resistance, endothelial dysfunction, and a tissue inflammatory infiltrate that can amplify systemic chronic inflammation. A study in patients with T2DM, with or without persistent hepatic inflammation (owing to chronic HBV infection) suggested that the presence of liver inflammation is a key mediator in the increased risk of CKD. Although the presence of T2DM undoubtedly increases the risk of CVD in patients with NAFLD, several studies have shown that potential mediators of vascular and renal damage occur more frequently in patients with NAFLD regardless of whether or not they also have T2DM [77]. Administration of sera from patients with CKD or uninephrectomy to rodents or cultured adipocytes induces lipodystrophy, ectopic fat redistribution from adipose tissue to liver and muscle, insulin resistance, and glucose intolerance [78, 79].

Although obesity and NAFLD injure the kidney, growing e­vidence suggests CKD may also contribute to NAFLD and insulin resistance. Taken together, increased levels of lipids and  the increased inflammatory response are both thought to be important pathogenetic factors that are not only involved in the development of NASH but are also significant factors in the development and progression of CKD.

Microbiome The liver is the key metabolic organ exposed to high levels of intestinal products through the tight bidirectional linkages in the biliary tract, portal vein, and systemic circulation. An altered microbiome (or “dysbiosis”) may lead to an increase in intestinal permeability, which can amplify many of these effects derived from intestine [80]. Studies of the microbiome are relatively nascent; however, it is anticipated that there will be considerable progress in linking its role to NASH. The changes in specific microbial products, secondary to altered gut microbial composition, and the changes in intestinal permeability and function can affect hepatic structure and function to further increase the risk of NAFLD. Patients with NAFLD have a higher prevalence of microbial dysbiosis. Human studies have documented that the gut microbiome among patients with NASH is less complex than that of healthy subjects [81, 82] and indicate that weight loss also alters the microbiome. Adults with NAFLD showed increased serum trimethylamine N‐oxide (TMAO) [83] and decreased production of phosphatidylcholine [84]. Plausible mechanistic links between an altered microbiome and fatty liver are emerging and include the potential for bacterial proteins to function as ligands for G protein‐coupled receptors [85, 86] (Figure 49.4). Dysbiosis has been described in patients with obesity [87] or other features of metabolic syndrome [88] and in those with established NAFLD [89], T2DM [90, 91], CVD [92], or CKD [93]. Several potential pathways, factors, and processes might link dysbiosis, or mediators of the gut microbiota, and NAFLD to CKD risk factors and vascular and renal diseases (Figure 49.2) [1]. Dysbiosis can potentially influence NAFLD, CKD, and obesity through multiple and complex mechanisms. It has been reported that changes in the gut microbiome of patients with CKD relate to lower levels of Bifidobacteriaceae and Lactobacillaceae and to higher levels of Enterobacteriaceae [94]. Microbial fermentation of dietary fiber in the intestine by anaerobic bacteria such as Bifidobacteria and Lactobacilli results in the formation of short‐chain fatty acids (SFCAs) including acetate, propionate, and butyrate, which have the potential to influence hepatic lipogenesis and gluconeogenesis. Thus, decreased levels of SCFAs in the setting of NAFLD might contribute to the development of liver adiposity and hepatic insulin resistance [95]. The intestinal microbiota also produces trimethylamine, p‐cresol, and indole from dietary nutrients such as choline, phenylalanine/tyrosine, and tryptophan, respectively. Further metabolism in the liver by oxidation or sulfation produces ionically charged water‐soluble molecules, such as TMAO, p‐cresol sulfate, and indole sulfate, which can be excreted in the urine. Plasma TMAO levels are elevated in patients with CKD and portend poorer long‐term survival [96]. Indole sulfate, which is cleared by the proximal tubules, is

Figure 49.4  Interplay between the liver and gut microbiota in alcoholic liver disease and NAFLD. Intestinal dysbiosis and bacterial overgrowth are observed in both alcoholic liver disease (ALD) (part a) and non‐alcoholic fatty liver disease (NAFLD) (part b). Bacterial overgrowth causes an increase in secondary bile acids (BAs), which modulates farnesoid X receptor (FXR)‐mediated hepatic synthesis of BA, leading to an overall increase in hepatic BA synthesis. A reduction in hepatic phosphatidylcholine is also seen in both ALD and NAFLD, which causes triglyceride accumulation in the liver (fatty liver). While ALD‐associated dysbiosis is characterized by reduction in Lactobacillus and Candida overgrowth, patients with NAFLD have a higher abundance of Lactobacillus (effects on fungal population remain to be investigated). In ALD and NAFLD, increased ethanol and its metabolite acetaldehyde in the intestinal lumen mediate weakening of intestinal tight junctions. Consequently, increased translocation of microbial‐associated molecular patterns (MAMPs) (seen in ALD and NAFLD) and gut metabolites, such as acetaldehyde, acetate (seen in ALD) and trimethylamine (TMA, seen in NAFLD), elicits intestinal and hepatic inflammatory responses, leading to progressive liver damage. AMP, antimicrobial peptides; EtOH, ethanol; HFD, high‐fat diet; LCFAs, long‐chain fatty acids, TMAO, trimethylamine N‐oxide. Reproduced with permission of Springer Nature, Figure 3 in [86].

642

THE LIVER:  REFERENCES

proinflammatory and potentially toxic to the kidneys as it increases the risk of tubulointerstitial fibrosis [97]. It is now widely accepted that liver and kidney damage can result from extensive interaction with the gut microbiota through specialized molecules, such as TMA. As the role of the microbiota in liver disease progression, prognosis, and treatment is increasingly recognized, it is necessary to make focused and conscious efforts to study the effect of the microbiome to efficiently address the socioeconomic burden of this spectrum of liver diseases. Increasing evidence suggests that the liver and kidneys share pathways including microbiome and dysbiosis that are intrinsically linked to each other. These indicate that the prevalence of CKD is markedly increased among patients with NAFLD, and that the presence and severity of NAFLD is associated with an increased incidence of CKD, and that this association could be independent of multiple cardiorenal risk factors. Taken together, more active and systematic research for the link between the microbiome and liver and kidney disease are needed to understand the mechanisms and causal association between these diseases.

Bile acid signaling Bile acids are produced in the liver from cholesterol and are metabolized by enzymes derived from the gut microbiome. Evidence shows that bile acids are critically important for maintaining a healthy gut microbiota [98]. The synthesis of bile acids is tightly regulated by negative feedback inhibition through the nuclear receptor FXR [99], which is a transcription factor that binds to the promoter region and initiates the expression of a wide range of target genes [100]. FXR is expressed in several tissues including the liver, intestine, and kidney [101]. In the liver, bile acid‐activated FXR induces the expression of small heterodimer partner (SHP), which binds to liver receptor homolog‐1 (LRH‐1) and thereby inhibits expression of the Cyp7a1 gene [102]. The primary bile acids produced in humans are chenodeoxycholic acid (CDCA) and cholic acid (CA), while rodents produce CA and muricholic acids (MCAs), predominantly beta‐MCA (βMCA) [103]. The most potent endogenous ligand for FXR is CDCA, followed by CA, DCA, and LCA [104]. UDCA, TαMCA, TβMCA, and GβMCA inhibit FXR activation [98]. A positive phase 2b clinical trial of obeticholic acid (OCA) (25 mg per day) was terminated early because of demonstration of efficacy during an interim analysis planned a priori [105]. However, OCA, in 25 mg per day doses, causes both pruritus and moderate increases in low‐density lipoprotein (LDL) cholesterol in some individuals. Several small‐molecule agonists of FXR that do not have a bile acid structural backbone have been developed on the premise that they will not increase LDL cholesterol or cause pruritus, but this has not yet been proven. Additionally, FXR induces the release of the growth factor FGF19 from the intestine upon bile acid binding to the receptor [106], which has beneficial effects on NASH in animal models, although some results are conflicting [107, 108]. In a phase 2a study, FGF19 analog yielded a dramatic reduction in hepatic fat and liver enzymes in patients with biopsy‐ confirmed NASH [107]. TGR5 is another bile acid‐responsive receptor involved in host metabolism. TGR5 is a plasma membrane‐bound G

protein‐coupled receptor ubiquitously expressed with high expression in gallbladder, placenta, lung, spleen, intestine, liver, brown and white adipose tissue, skeletal muscle, and bone marrow [109]. TGR5 is activated mainly by the secondary bile acids LCA and DCA [110]. TGR5 signaling controls glucose homeostasis by increased energy expenditure in brown adipose tissue and muscle and by increased GLP‐1 release in intestinal L cells [111]. At present it is unclear whether FXR and TGR5 signal directly or indirectly via alterations in the microbiota and bile acid metabolism that produce agonistic or antagonistic signaling through each other. Glucagon‐like peptide (GLP)‐1 is an intestinal hormone generated through the proteolytic processing of proglucagon that stimulates insulin secretion and inhibits secretion of glucagon. GLP‐1 is also an insulin sensitizer with additional metabolic effects that contribute to its anti‐NASH activity [112]. Liraglutide, a GLP‐1 agonist requiring daily injection, improved NASH histology in a small pilot study [113]. Interestingly, L cells express FXR and FXR also regulates GLP‐1 synthesis [114].

REFERENCES 1. Targher, G. and Byrne, C.D. Non‐alcoholic fatty liver disease: an emerging driving force in chronic kidney disease. Nat Rev Nephrol, 2017;13(5): 297–310. 2. Yeung, M.W., Wong, G.L., Choi, K.C. et al. Advanced liver fibrosis but not steatosis is independently associated with albuminuria in Chinese patients with type 2 diabetes. J Hepatol, 2017. 3. Sinn, D.H., Kang, D., Jang, H.R. et al. Development of chronic kidney disease in patients with non‐alcoholic fatty liver disease: a cohort study. J Hepatol, 2017;67(6):1274–80. 4. Mantovani, A., Zaza, G., Byrne, C.D. et al. Nonalcoholic fatty liver disease increases risk of incident chronic kidney disease: a systematic review and meta‐analysis. Metabolism, 2018;79:64–76. 5. Sole, C., Pose, E., Sola, E., and Gines, P. Hepatorenal syndrome in the era of acute kidney injury. Liver Int, 2018;38(11):1891–1901. 6. Angeli, P., Gines, P., Wong, F. et  al. Diagnosis and management of acute kidney injury in patients with cirrhosis: revised consensus recommendations of the International Club of Ascites. J Hepatol, 2015;62(4):968–74. 7. Durand, F., Graupera, I., Gines, P., Olson, J.C., and Nadim, M.K. Pathogenesis of hepatorenal syndrome: implications for therapy. Am J Kidney Dis, 2016;67(2):318–28. 8. Sanyal, A.J., Boyer, T.D., Frederick, R.T. et al. Reversal of hepatorenal syndrome type 1 with terlipressin plus albumin vs. placebo plus albumin in a pooled analysis of the OT‐0401 and REVERSE randomised clinical studies. Aliment Pharmacol Ther, 2017;45(11):1390–402. 9. Glass, L. and Sharma, P. Evidence‐based therapeutic options for hepatorenal syndrome. Gastroenterology, 2016;150(4):1031–3. 10. Wong, F., Leung, W., Al Beshir, M., Marquez, M., and Renner, E.L. Outcomes of patients with cirrhosis and hepatorenal syndrome type 1 treated with liver transplantation. Liver Transpl, 2015;21(3):300–7. 11. Cavallin, M., Kamath, P.S., Merli, M. et al. Terlipressin plus albumin versus midodrine and octreotide plus albumin in the treatment of hepatorenal syndrome: a randomized trial. Hepatology, 2015;62(2):567–74. 12. Cohen, S.D., Kopp, J.B., and Kimmel, P.L. Kidney diseases associated with human immunodeficiency virus infection. N Engl J Med, 2017;377(24): 2363–74. 13. Rutledge, S.M., Chung, R.T., and Sise, M.E. Treatment of hepatitis C virus infection in patients with mixed cryoglobulinemic syndrome and cryoglobulinemic glomerulonephritis. Hemodial Int, 2018;22(1):S81–96. 14. Collins, J.M., Raphael, K.L., Terry, C. et al. Hepatitis B virus reactivation during successful treatment of hepatitis C virus with sofosbuvir and simeprevir. Clin Infect Dis, 2015;61(8):1304–6. 15. Sise, M.E. Hepatitis C virus infection and the kidney. Nephrol Dial Transplant, 2018;34(3):415–8.



49:  The Kidney in Liver Disease

16. Rodrigues, E.M.F., Fernandes, R., Susin, R., and Fior, B. Immune reconstitution inflammatory syndrome as a cause of autoimmune hepatitis and acute liver failure. Rev Bras Ter Intensiva, 2017;29(3):382–5. 17. Gowda, C., Newcomb, C.W., Liu, Q. et al. Risk of acute liver injury with antiretroviral therapy by viral hepatitis status. Open Forum Infect Dis, 2017;4(2):ofx012. 18. Nadkarni, G.N., Konstantinidis, I., and Wyatt, C.M. HIV and the aging kidney. Curr Opin HIV AIDS, 2014;9(4):340–5. 19. Genovese, G., Friedman, D.J., Ross, M.D. et al. Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science, 2010;329(5993):841–5. 20. Tzur, S., Rosset, S., Shemer, R. et al. Missense mutations in the APOL1 gene are highly associated with end stage kidney disease risk previously attributed to the MYH9 gene. Hum Genet, 2010;128(3):345–50. 21. Kopp, J.B., Nelson, G.W., Sampath, K. et  al. APOL1 genetic variants in focal segmental glomerulosclerosis and HIV‐associated nephropathy. J Am Soc Nephrol, 2011;22(11):2129–37. 22. Fishman, J.A. Infection in organ transplantation. Am J Transplant, 2017;17(4):856–79. 23. Hales, C.M., Fryar, C.D., Carroll, M.D., Freedman, D.S., and Ogden, C.L. Trends in obesity and severe obesity prevalence in US youth and adults by sex and age, 2007–2008 to 2015–2016. JAMA, 2018;319(16):1723–5. 24. D’Agati, V.D., Chagnac, A., de Vries, A.P. et al. Obesity‐related glomerulopathy: clinical and pathologic characteristics and pathogenesis. Nat Rev Nephrol, 2016;12(8):453–71. 25. Lakkis, J.I. and Weir, M.R. Obesity and kidney disease. Prog Cardiovasc Dis, 2018;61(2):157–67. 26. Whaley‐Connell, A. and Sowers, J.R. Obesity and kidney disease: from population to basic science and the search for new therapeutic targets. Kidney Int, 2017;92(2):313–23. 27. Wang, X.X., Wang, D., Luo, Y. et al. FXR/TGR5 dual agonist prevents progression of nephropathy in diabetes and obesity. J Am Soc Nephrol, 2018;29(1):118–37. 28. Wang, X.X., Jiang, T., Shen, Y., Adorini, L. et al. The farnesoid X receptor modulates renal lipid metabolism and diet‐induced renal inflammation, fibrosis, and proteinuria. Am J Physiol Renal Physiol, 2009;297(6):F1587–96. 29. Wang, X.X., Edelstein, M.H., Gafter, U. et al. G protein‐coupled bile acid receptor TGR5 activation inhibits kidney disease in obesity and diabetes. J Am Soc Nephrol, 2016;27(5):1362–78. 30. Gai, Z., Gui, T., Hiller, C., and Kullak‐Ublick, G.A. Farnesoid X receptor protects against kidney injury in uninephrectomized obese mice. J Biol Chem, 2016;291(5):2397–411. 31. Friedman, S.L., Neuschwander‐Tetri, B.A., Rinella, M., and Sanyal, A.J. Mechanisms of NAFLD development and therapeutic strategies. Nat Med, 2018;24(7):908–22. 32. Baiceanu, A., Mesdom, P., Lagouge, M., and Foufelle, F. Endoplasmic reticulum proteostasis in hepatic steatosis. Nat Rev Endocrinol, 2016;12(12): 710–22. 33. Artunc, F., Schleicher, E., Weigert, C., Fritsche, A., Stefan, N., and Haring, H.U. The impact of insulin resistance on the kidney and vasculature. Nat Rev Nephrol, 2016;12(12):721–37. 34. Samuel, V.T. and Shulman, G.I. Nonalcoholic fatty liver disease as a nexus of metabolic and hepatic diseases. Cell Metab, 2018;27(1):22–41. 35. Gnudi, L., Coward, R.J.M., and Long, D.A. Diabetic nephropathy: perspective on novel molecular mechanisms. Trends Endocrinol Metab, 2016;27(11):820–30. 36. Lay, A.C., Hurcombe, J.A., Betin, V.M.S. et  al. Prolonged exposure of mouse and human podocytes to insulin induces insulin resistance through lysosomal and proteasomal degradation of the insulin receptor. Diabetologia, 2017;60(11):2299–311. 37. Wang, X.X., Levi, J., Luo, Y. et al. SGLT2 protein expression is increased in human diabetic nephropathy: SGLT2 protein inhibition decreases renal lipid accumulation, inflammation, and the development of nephropathy in diabetic mice. J Biol Chem, 2017;292(13):5335–48. 38. Wang, D., Luo, Y., Wang, X. et al. The sodium‐glucose cotransporter 2 inhibitor dapagliflozin prevents renal and liver disease in western diet induced obesity mice. Int J Mol Sci, 2018;19(1). 39. Koo, S.H. Nonalcoholic fatty liver disease: molecular mechanisms for the hepatic steatosis. Clin Mol Hepatol, 2013;19(3):210–5. 40. Benhamed, F., Denechaud, P.D., Lemoine, M. et al. The lipogenic transcription factor ChREBP dissociates hepatic steatosis from insulin resistance in mice and humans. J Clin Invest, 2012;122(6):2176–94.

643

41. Ponugoti, B., Kim, D.H., Xiao, Z. et  al. SIRT1 deacetylates and inhibits SREBP‐1C activity in regulation of hepatic lipid metabolism. J Biol Chem, 2010;285(44):33959–70. 42. Porstmann, T., Santos, C.R., Griffiths, B. et al. SREBP activity is regulated by mTORC1 and contributes to Akt‐dependent cell growth. Cell Metab, 2008;8(3):224–36. 43. Begriche, K., Massart, J., Robin, M.A., Bonnet, F., and Fromenty, B. Mitochondrial adaptations and dysfunctions in nonalcoholic fatty liver disease. Hepatology, 2013;58(4):1497–507. 44. Fabbrini, E., Mohammed, B.S., Magkos, F., Korenblat, K.M., Patterson, B.W., and Klein, S. Alterations in adipose tissue and hepatic lipid kinetics in obese men and women with nonalcoholic fatty liver disease. Gastroenterology, 2008;134(2):424–31. 45. Quiroga, A.D. and Lehner, R. Pharmacological intervention of liver triacylglycerol lipolysis: the good, the bad and the ugly. Biochem Pharmacol, 2018;155:233–41. 46. Ioannou, G.N. The role of cholesterol in the pathogenesis of NASH. Trends Endocrinol Metab, 2016;27(2):84–95. 47. Fiorucci, S., Biagioli, M., and Distrutti, E. Future trends in the treatment of non‐alcoholic steatohepatitis. Pharmacol Res, 2018;134:289–98. 48. Sun, L., Halaihel, N., Zhang, W., Rogers, T., and Levi, M. Role of sterol regulatory element‐binding protein 1 in regulation of renal lipid metabolism and glomerulosclerosis in diabetes mellitus. J Biol Chem, 2002;277(21): 18919–27. 49. Jiang, T., Wang, Z., Proctor, G. et al. Diet‐induced obesity in C57BL/6J mice causes increased renal lipid accumulation and glomerulosclerosis via a sterol regulatory element‐binding protein‐1c‐dependent pathway. J Biol Chem, 2005;280(37):32317–25. 50. Kume, S., Uzu, T., Araki, S. et al. Role of altered renal lipid metabolism in the development of renal injury induced by a high‐fat diet. J Am Soc Nephrol, 2007;18(10):2715–23. 51. Wu, Y., Liu, Z., Xiang, Z. et  al. Obesity‐related glomerulopathy: insights from gene expression profiles of the glomeruli derived from renal biopsy samples. Endocrinology, 2006;147(1):44–50. 52. Jiang, T., Wang, X.X., Scherzer, P. et  al. Farnesoid X receptor modulates renal lipid metabolism, fibrosis, and diabetic nephropathy. Diabetes, 2007;56(10):2485–93. 53. Wang, X.X., Jiang, T., Shen, Y. et al. Diabetic nephropathy is accelerated by farnesoid X receptor deficiency and inhibited by farnesoid X receptor activation in a type 1 diabetes model. Diabetes, 2010;59(11):2916–27. 54. Yang, X., Okamura, D.M., Lu, X. et  al. CD36 in chronic kidney disease: novel insights and therapeutic opportunities. Nat Rev Nephrol, 2017;13(12): 769–81. 55. Herman‐Edelstein, M., Scherzer, P., Tobar, A., Levi, M., and Gafter, U. Altered renal lipid metabolism and renal lipid accumulation in human diabetic nephropathy. J Lipid Res, 2014;55(3):561–72. 56. Wang, X.X., Jiang, T., Shen, Y. et al. Vitamin D receptor agonist doxercalciferol modulates dietary fat‐induced renal disease and renal lipid metabolism. Am J Physiol Renal Physiol, 2011;300(3):F801–10. 57. Proctor, G., Jiang, T., Iwahashi, M., Wang, Z., Li, J., and Levi, M. Regulation of renal fatty acid and cholesterol metabolism, inflammation, and fibrosis in Akita and OVE26 mice with type 1 diabetes. Diabetes, 2006;55(9):2502–9. 58. Marra, F. and Svegliati‐Baroni, G. Lipotoxicity and the gut‐liver axis in NASH pathogenesis. J Hepatol, 2018;68(2):280–95. 59. Caldwell, S.H., Swerdlow, R.H., Khan, E.M. et al. Mitochondrial abnormalities in non‐alcoholic steatohepatitis. J Hepatol, 1999;31(3):430–4. 60. Ding, R.B., Bao, J., and Deng, C.X. Emerging roles of SIRT1 in fatty liver diseases. Int J Biol Sci, 2017;13(7):852–67. 61. Morigi, M., Perico, L., and Benigni, A. Sirtuins in renal health and disease. J Am Soc Nephrol, 2018;29(7):1799–809. 62. Tang, C., Cai, J., and Dong, Z. Mitochondrial dysfunction in obesity‐related kidney disease: a novel therapeutic target. Kidney Int, 2016;90(5):930–3. 63. Grattagliano, I., de Bari, O., Bernardo, T.C., Oliveira, P.J., Wang, D.Q., and Portincasa, P. Role of mitochondria in nonalcoholic fatty liver disease‐‐from origin to propagation. Clin Biochem, 2012;45(9):610–8. 64. Caballero, F., Fernandez, A., Matias, N. et  al. Specific contribution of methionine and choline in nutritional nonalcoholic steatohepatitis: impact on mitochondrial S‐adenosyl‐L‐methionine and glutathione. J Biol Chem, 2010;285(24):18528–36. 65. Al‐Serri, A., Anstee, Q.M., Valenti, L. et al. The SOD2 C47T polymorphism influences NAFLD fibrosis severity: evidence from case‐control and intra‐ familial allele association studies. J Hepatol, 2012;56(2):448–54.

644

THE LIVER:  REFERENCES

66. Ajith, T.A. Role of mitochondria and mitochondria‐targeted agents in non‐ alcoholic fatty liver disease. Clin Exp Pharmacol Physiol, 2018;45(5): 413–21. 67. Lebeaupin, C., Vallee, D., Hazari, Y., Hetz, C., Chevet, E., and Bailly‐Maitre, B. Endoplasmic reticulum stress signaling and the pathogenesis of non‐alcoholic fatty liver disease. J Hepatol, 2018;69(4):927–47. 68. Kropski, J.A. and Blackwell, T.S. Endoplasmic reticulum stress in the pathogenesis of fibrotic disease. J Clin Invest, 2018;128(1):64–73. 69. Cybulsky, A.V. Endoplasmic reticulum stress, the unfolded protein response and autophagy in kidney diseases. Nat Rev Nephrol, 2017;13(11):681–96. 70. Brunt, E.M., Wong, V.W., Nobili, V. et al. Nonalcoholic fatty liver disease. Nat Rev Dis Primers, 2015;1:15080. 71. Schuster, S., Cabrera, D., Arrese, M., and Feldstein, A.E. Triggering and resolution of inflammation in NASH. Nat Rev Gastroenterol Hepatol, 2018;15(6):349–64. 72. Wree, A., Broderick, L., Canbay, A., Hoffman, H.M., and Feldstein, A.E. From NAFLD to NASH to cirrhosis‐new insights into disease mechanisms. Nat Rev Gastroenterol Hepatol, 2013;10(11):627–36. 73. Han, M.S., Park, S.Y., Shinzawa, K. et  al. Lysophosphatidylcholine as a death effector in the lipoapoptosis of hepatocytes. J Lipid Res, 2008; 49(1):84–97. 74. Csak, T., Ganz, M., Pespisa, J., Kodys, K., Dolganiuc, A., and Szabo, G. Fatty acid and endotoxin activate inflammasomes in mouse hepatocytes that release danger signals to stimulate immune cells. Hepatology, 2011;54(1):133–44. 75. Wree, A., McGeough, M.D., Inzaugarat, M.E. et al. NLRP3 inflammasome driven liver injury and fibrosis: roles of IL‐17 and TNF in mice. Hepatology, 2017. 76. Sharma, M., Mitnala, S., Vishnubhotla, R.K., Mukherjee, R., Reddy, D.N., and Rao, P.N. The riddle of nonalcoholic fatty liver disease: progression from nonalcoholic fatty liver to nonalcoholic steatohepatitis. J Clin Exp Hepatol, 2015;5(2):147–58. 77. Targher, G., Day, C.P., and Bonora, E. Risk of cardiovascular disease in patients with nonalcoholic fatty liver disease. N Engl J Med, 2010; 363(14):1341–50. 78. Pelletier, C.C., Koppe, L., Croze, M.L. et al. White adipose tissue overproduces the lipid‐mobilizing factor zinc alpha2‐glycoprotein in chronic kidney disease. Kidney Int, 2013;83(5):878–86. 79. Axelsson, J., Åström, G., Sjölin, E. et al. Uraemic sera stimulate lipolysis in human adipocytes: role of perilipin. Nephrol Dial Transplant, 2011; 26(8):2485–91. 80. Leung, C., Rivera, L., Furness, J.B., and Angus, P.W. The role of the gut microbiota in NAFLD. Nat Rev Gastroenterol Hepatol, 2016;13(7): 412–25. 81. Betrapally, N.S., Gillevet, P.M., and Bajaj, J.S. Changes in the intestinal microbiome and alcoholic and nonalcoholic liver diseases: causes or effects? Gastroenterology, 2016;150(8):1745–55.e3. 82. Loomba, R., Seguritan, V., Li, W. et al. Gut microbiome based metagenomic signature for non‐invasive detection of advanced fibrosis in human nonalcoholic fatty liver disease. Cell Metab, 2017;25(5):1054–62.e5. 83. Chen, Y.M., Liu, Y., Zhou, R.F. et  al. Associations of gut‐flora‐dependent metabolite trimethylamine‐N‐oxide, betaine and choline with non‐alcoholic fatty liver disease in adults. Sci Rep, 2016;6:19076. 84. Arendt, B.M., Ma, D.W., Simons, B. et al. Nonalcoholic fatty liver disease is associated with lower hepatic and erythrocyte ratios of phosphatidylcholine to phosphatidylethanolamine. Appl Physiol Nutr Metab, 2013;38(3): 334–40. 85. Cohen, L.J., Esterhazy, D., Kim, S.H. et  al. Commensal bacteria make GPCR ligands that mimic human signalling molecules. Nature, 2017;549(7670):48–53. 86. Tripathi, A., Debelius, J., Brenner, D.A. et  al. The gut‐liver axis and the intersection with the microbiome. Nat Rev Gastroenterol Hepatol, 2018;15(7):397–411. 87. Liu, R., Hong, J., Xu, X. et  al. Gut microbiome and serum metabolome alterations in obesity and after weight‐loss intervention. Nat Med, 2017; 23:859. 88. Mehal, W.Z. The Gordian knot of dysbiosis, obesity and NAFLD. Nat Rev Gastroenterol Hepatol, 2013;10(11):637–44. 89. Tilg, H., Cani, P.D., and Mayer, E.A. Gut microbiome and liver diseases. Gut, 2016;65(12):2035–44. 90. Qin, J., Li, Y., Cai, Z. et  al. A metagenome‐wide association study of gut microbiota in type 2 diabetes. Nature, 2012;490(7418):55–60.

  91. Utzschneider, K.M., Kratz, M., Damman, C.J., and Hullar, M. Mechanisms linking the gut microbiome and glucose metabolism. J Clin Endocrinol Metab, 2016;101(4):1445–54.   92. Koopen, A.M., Groen, A.K., and Nieuwdorp, M. Human microbiome as therapeutic intervention target to reduce cardiovascular disease risk. Curr Opin Lipidol, 2016;27(6):615–22.   93. Briskey, D., Tucker, P.S., Johnson, D.W., and Coombes, J.S. Microbiota and the nitrogen cycle: Implications in the development and progression of CVD and CKD. Nitric Oxide, 2016;57:64–70.   94. Sampaio‐Maia, B., Simoes‐Silva, L., Pestana, M., Araujo, R., and Soares‐ Silva, I.J. The role of the gut microbiome on chronic kidney disease. Adv Appl Microbiol, 2016;96:65–94.   95. Scorletti, E. and Byrne, C.D. Extrahepatic diseases and NAFLD: the triangular relationship between NAFLD, type 2‐diabetes and dysbiosis. Dig Dis, 2016;34(1):11–8.   96. Tang, W.H., Wang, Z., Kennedy, D.J. et al. Gut microbiota‐dependent trimethylamine N‐oxide (TMAO) pathway contributes to both development of renal insufficiency and mortality risk in chronic kidney disease. Circ Res, 2015;116(3):448–55.   97. Nallu, A., Sharma, S., Ramezani, A., Muralidharan, J., and Raj, D. Gut microbiome in chronic kidney disease: challenges and opportunities. Transl Res, 2017;179:24–37.   98. Wahlström, A., Sayin Sama I., Marschall, H‐U., and Bäckhed, F. Intestinal crosstalk between bile acids and microbiota and its impact on host metabolism. Cell Metab, 2016;24(1):41–50.   99. Sinal, C.J., Tohkin, M., Miyata, M., Ward, J.M., Lambert, G., and Gonzalez, F.J. Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell, 2000;102(6):731–44. 100. Teodoro, J.S., Rolo, A.P., and Palmeira, C.M. Hepatic FXR: key regulator of whole‐body energy metabolism. Trends Endocrinol Metab, 2011; 22(11):458–66. 101. Lefebvre, P., Cariou, B., Lien, F., Kuipers, F., and Staels, B. Role of bile acids and bile acid receptors in metabolic regulation. Physiol Rev, 2009;89(1):147–91. 102. Goodwin, B., Jones, S.A., Price, R.R. et  al. A regulatory cascade of the nuclear receptors FXR, SHP‐1, and LRH‐1 represses bile acid biosynthesis. Mol Cell. 2000;6(3):517–26. 103. Takahashi, S., Fukami, T., Masuo, Y. et al. Cyp2c70 is responsible for the species difference in bile acid metabolism between mice and humans. J Lipid Res, 2016;57(12):2130–7. 104. Makishima, M., Okamoto, A.Y., Repa, J.J. et al. Identification of a nuclear receptor for bile acids. Science, 1999;284(5418):1362–5. 105. Neuschwander‐Tetri, B.A., Loomba, R., Sanyal, A.J. et  al. Farnesoid X nuclear receptor ligand obeticholic acid for non‐cirrhotic, non‐alcoholic steatohepatitis (FLINT): a multicentre, randomised, placebo‐controlled trial. Lancet, 2015;385(9972):956–65. 106. Nies, V.J., Sancar, G., Liu, W. et al. Fibroblast growth factor signaling in metabolic regulation. Front Endocrinol,2015;6:193. 107. Harrison, S.A., Rinella, M.E., Abdelmalek, M.F. et al. NGM282 for treatment of non‐alcoholic steatohepatitis: a multicentre, randomised, double‐ blind, placebo‐controlled, phase 2 trial. Lancet, 2018;391(10126):1174–85. 108. Jiang, C., Xie, C., Li, F. et al. Intestinal farnesoid X receptor signaling promotes nonalcoholic fatty liver disease. J Clin Invest, 2015;125(1):386–402. 109. Kawamata, Y., Fujii, R., Hosoya, M. et  al. A G protein‐coupled receptor responsive to bile acids. J Biol Chem, 2003;278(11):9435–40. 110. Chen, X., Lou, G., Meng, Z., and Huang, W. TGR5: a novel target for weight maintenance and glucose metabolism. Exp Diabetes Res, 2011; 2011:853501. 111. Thomas, C., Gioiello, A., Noriega, L. et al. TGR5‐mediated bile acid sensing controls glucose homeostasis. Cell Metab, 2009;10(3):167–77. 112. Jinnouchi, H., Sugiyama, S., Yoshida, A. et al. Liraglutide, a glucagon‐like peptide‐1 analog, increased insulin sensitivity assessed by hyperinsulinemic‐euglycemic clamp examination in patients with uncontrolled type 2 diabetes mellitus. J Diabetes Res, 2015;2015:8. 113. Armstrong, M.J., Gaunt, P., Aithal, G.P. et al. Liraglutide safety and efficacy in patients with non‐alcoholic steatohepatitis (LEAN): a multicentre, double‐blind, randomised, placebo‐controlled phase 2 study. Lancet, 2016;387(10019):679–90. 114. Trabelsi, M‐S., Daoudi, M., Prawitt, J. et al. Farnesoid X receptor inhibits glucagon‐like peptide‐1 production by enteroendocrine L cells. Nat Commun, 2015;6:7629.

50

α1‐Antitrypsin Deficiency David A. Rudnick and David H. Perlmutter Department of Pediatrics, Division of Gastroenterology, Hepatology, and Nutrition and Department of Developmental Biology, Washington University School of Medicine in St. Louis, St. Louis Children’s Hospital, St. Louis, MO, USA

INTRODUCTION The classical form of α1‐antitrypsin (α1‐AT) deficiency, homozygous for the mutant α1‐ATZ allele, is a relatively common disease. It affects approximately 1 in 1600 to 1 in 2000 live births in most populations of Northern European ancestry [1, 2]. Although only a subgroup of deficient individuals develop liver disease, it represents the most common metabolic cause of liver disease in children [3] and can be associated with chronic liver disease and hepatocellular carcinoma in adults [4]. This deficiency also causes premature development of pulmonary emphysema in adults. The α1‐AT molecule is a single‐chain secretory glycoprotein that inhibits destructive neutrophil proteases including elastase, cathepsin G, and proteinase 3. It is often referred to as a hepatic acute phase reactant in that plasma α1‐AT is predominantly derived from the liver and plasma levels increase three‐ to fivefold during the host response to tissue injury/inflammation. It is the archetype of a family of structurally related circulating serine protease inhibitors termed SERPINs. In the deficient state, there is an approximately 85% to 90% reduction in serum α1‐ AT concentrations. A single amino acid substitution results in a mutant protein unable to traverse the secretory pathway. This α1‐ATZ protein is retained in the endoplasmic reticulum (ER) rather than secreted into the blood and body fluids. The classical deficient state is unique as a genetic disease in that it causes injury to one target organ, lung, by loss‐of‐function and injury to another target organ, liver, by what appears to be a gain‐of‐function mechanism. Most data in the literature indicate that emphysema results from a decreased number of α1‐AT molecules within the lower respiratory tract, allowing unregulated elastolytic attack on the lung’s connective tissue matrix [5]. Oxidative inactivation of residual α1‐AT as a result

of cigarette smoking accelerates lung injury [5]. Moreover, the elastase–antielastase theory for the pathogenesis of emphysema is based on the concept that oxidative inactivation of α1‐AT as a result of cigarette smoking plays a key role in the emphysema of α1‐AT‐sufficient individuals, the vast majority of cases of emphysema [5]. It is more difficult to explain the pathogenesis of liver injury in this deficiency. Results of transgenic animal experiments provide further evidence that the liver disease does not result from a deficiency in antielastase activity [6, 7]. Most data in the literature corroborate the concept that liver injury in α1‐AT deficiency results from the hepatotoxicity of mutant­ α1‐ ATZ accumulation in hepatocytes. Although it is a single gene defect, there is extraordinary variation in the phenotypic expression of disease in the classical form of α1‐AT deficiency. For instance, nationwide prospective screening studies done by Sveger in Sweden showed that only 8% of an unbiased cohort developed clinically significant liver disease through the fourth decade of life [1, 8]. These data indicate that other genetic traits and/or environmental factors predispose a subgroup of PIZZ individuals, that is, those homozygous for the α1‐ATZ mutant allele, to liver injury. There is also variation in incidence and severity of lung injury among α1‐AT‐deficient individuals. Environmental factors, such as cigarette smoking, obviously play an important role [2, 5]. However, there are well‐documented examples of siblings and other relatives of deficient individuals with severe emphysema who have the same genotype, a history of heavy cigarette smoking and only mild, subclinical pulmonary function abnormalities even at advanced ages [9]. Historically, the diagnosis of α1‐AT deficiency was based on the altered migration of the abnormal α1‐ATZ molecule in serum specimens subjected to isoelectric focusing gel analysis. Contemporary diagnostic strategies may now employ genomic

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

646

THE LIVER:  CLINICAL MANIFESTATIONS OF LIVER DISEASE

analyses for specific gene mutations. Treatment of α1‐AT deficiency‐associated liver disease remains mostly supportive. Liver replacement therapy has been used successfully for severe liver injury. α1‐AT‐deficient patients with severe emphysema have undergone lung transplantation. Intravenous administration of purified plasma α1‐AT augmentation therapy was approved for α1‐AT deficiency lung disease on the basis of biochemical efficacy studies [10]. Recently, a randomized, placebo‐controlled trial provided some initial, albeit limited, evidence that augmentation therapy slows the progression of lung destruction in α1‐ AT deficient subjects [11].

CLINICAL MANIFESTATIONS OF LIVER DISEASE Liver involvement is often noticed in early infancy because of persistent jaundice, elevated serum transaminases, and increased conjugated bilirubin levels [12, 13]. Some infants are recognized because of cholestasis characterized by pruritus and hypercholesterolemia. The clinical picture resembles extrahepatic biliary atresia but liver histology shows paucity of intrahepatic bile ducts [14]. Alternatively, liver disease may first be discovered in later childhood or early adolescence when the affected individual presents with hepatosplenomegaly, ascites, or esophageal variceal hemorrhage. In some but not all cases, a history of unexplained prolonged obstructive jaundice during the neonatal period is identified. The incidence and natural history of liver disease in α1‐AT deficiency was determined in a nationwide screening study of newborns in Sweden by Sveger [1]. From 200 000 infants, 127 PIZZ newborns were identified and prospectively followed, now through age 37–40 years [8]. Initial analysis of this cohort identified 14 PIZZ infants with prolonged obstructive jaundice, another 8 with other reasons for clinical suspicion of liver disease, and the others with no clinical evidence for liver disease [1]. Five PIZZ subjects subsequently died in early childhood, two with known cirrhosis and two others with incidentally‐ detected liver disease at autopsy [15]. At reassessment of this cohort at age 37–40 years, no participating subjects showed evidence of active liver disease [8]. However, abnormalities in liver‐related laboratory and/or imaging studies were reported. One issue not addressed by the Sveger study is whether any PIZZ subjects have subclinical histologic abnormalities that lead to progressive disease in later adulthood. It is now evident that this deficiency causes liver disease with initial onset in adults. In our retrospective analysis of US liver transplantation databases, we found that 77% of α1‐AT‐deficient subjects undergoing liver transplantation over the last 20 years were adults, with peak age of transplantation in those subjects of 50–64 years [16]. Many are probably heterozygotes with other potential causes of liver disease, and some may be poorly substantiated diagnoses. Nevertheless, our analysis suggests that severe liver disease requiring transplantation in subjects with ZZ or SZ phenotypes is more frequent in adults than children. This observation is consistent with the discovery that autophagy plays a critical role in the pathobiology of this liver disease and the understanding that a physiological decline in

autophagy function in middle age may play an important role in age‐dependent degenerative diseases in general. Earlier studies also reported incidentally identified significant liver pathology at autopsy in some elderly patients with α1‐AT deficiency without known liver disease. A Swedish study showed a higher risk of cirrhosis and primary liver cancer in such subjects than previously suspected [4]. Based on these considerations, α1‐AT deficiency should be considered in the differential diagnosis of any adult who presents with chronic hepatitis, cirrhosis, portal hypertension, or hepatocellular carcinoma of unknown origin. Taken together, the overwhelming clinical experience with this disease indicates wide variation in liver disease phenotype with many “protected” from or having slowly progressing liver disease. It still remains unclear whether liver injury results from the heterozygous α1‐AT PIMZ (PI genotype heterozygous for the wild‐type M and mutant Z alleles) state by itself. Liver biopsy and transplant database studies have identified heterozygous patients with severe liver disease without other apparent explanation (reviewed in [3]). However, these studies are generally unable to exclude environmental causes of liver disease. Indeed, one cross‐sectional study of re‐examined α1‐AT deficiency patients in a referral‐based Austrian university hospital suggested that liver disease in heterozygotes was largely accounted for by hepatitis B or C virus, autoimmune disease, or alcohol abuse (reviewed in [3]). Other studies have subsequently reported significant correlations between the MZ state and development of, presentation with, or more rapid progression to advanced liver disease and liver transplantation in subjects with other chronic liver diseases [16, 17]. Liver disease has also been associated with other allelic variants of α1‐AT. For example, children with compound PISZ heterozygosity are affected by liver injury in a manner similar to PIZZ children [1], and liver disease was reported in subjects with α1‐AT deficiency type PIMmalton (PI genotype for the Mmalton allele, Table 50.1) [18]. These observations are interesting because the α1‐ATS molecule forms heteropolymers with α1‐ATZ [19], and, like α1‐ATZ, the PIMmalton molecule undergoes polymerization and ER retention [20]. Liver disease has also been detected in single patients with other α1‐AT allelic variants [3], but in those cases other causes of liver injury have generally not been completely excluded. Historically, the diagnosis was established by a serum α1‐AT phenotype determination by isoelectric focusing or by agarose electrophoresis at acid pH. Modern genomic methodologies permit specific analyses for known liver disease‐associated α1‐ AT gene mutations, and analyses for gene duplications, deletions, or previously unidentified mutations in the AT gene coding region. Liver histology is characterized by periodic acid–Schiff‐positive, diastase‐resistant globules in hepatocyte ER. These globules are most prominent in periportal hepatocytes, and may also be seen in Kupffer cells and cells of biliary ductular lineage (reviewed in [3]). There may be variable hepatocellular necrosis, inflammatory cell infiltration, periportal fibrosis, and/or cirrhosis. There is often evidence of bile duct epithelial cell destruction, and, occasionally, paucity of intrahepatic bile ducts. There can also be an intense autophagic reaction with nascent and degradative‐type autophagic vacuoles detected by electron microscopy on liver biopsies [21].

50:  α1-ANTITRYPSIN DEFICIENCY

647

Table 50.1  Deficiency variants of α1‐antitrypsin Clinical disease Variant

Defect

Site

Liver Lung

Cellular defect

Z S MHeerlen MProcida MMalton MDuarte MMineral Springs Siiyama PDuarte

Single base substitution M1 [Ala213] Single base substitution Single base substitution Single base substitution Single base deletion Unknown Single base substitution Single base substitution Two base substitution

+ − − − + +? − − +?

+ − + + + + + + +

IC accumulation IC accumulation IC accumulation IC accumulation IC accumulation Unknown No function; EC degradation? IC accumulation Unknown

PLowell WBethesda ZWrexham F T I MPalermo MNichinan

Single base substitution Single base substitution Single base substitution Single base substitution Single base substitution Single base substitution Single base deletion Single base deletion and single base substitution Single base substitution Single base substitution Single base substitution Single base substitution Single base substitution S variant defect + single base substitution Single base substitution Single base substitution Single base substitution Single base substitution Single base substitution Single base substitution

Glu342‐Lys Glu264‐Val Pro369‐Leu Leu41‐Pro Phe52 Unknown Gly57‐Glu Ser53‐Phe Arg101‐His Asp256‐Val Asp256‐Val Ala336‐Thre Ser19‐Leu Arg223‐Cys Glu264‐Val Arg39‐Cys Phe51 Phe52 Gly148‐Arg Glu342‐Lys His334‐Asp Lys259‐Ile Lys368‐Glu Pro391‐His S variant site + Ser14‐Phe

− − ? − − − − −

+ + ? − − − − −

IC degradation? Accelerated catabolism? Unknown Unknown Unknown IC degradation Unknown Unknown

− + − − +? −

− +? + − +?

Unknown IC accumulation IC accumulation IC accumulation IC accumulation IC polymerization, degradation; reduced secretion

Ile50‐Asn Ala58‐Asp Glu151‐Lys Phe227‐Cys Thr249‐Ala Lys328‐Glu

− − − − − −

− +? +? +? +? −

IC polymerization, stabilization; reduced secretion IC polymerization, stabilization; reduced secretion None IC polymerization, degradation; reduced secretion IC polymerization, stabilization; altered glycosylation IC stabilization (reduced enzyme activity)

ZAusburg King’s Mpisa Etaurisano Yorzinuovi PiSDonosti PiTijarafe PiSevilla PiCadiz PiTarragona  PiPuerto Real PiValencia

STRUCTURE, FUNCTION, AND PHYSIOLOGY α1‐AT is encoded by a 12.2 kb gene containing seven exons and six introns on human chromosome 14q31‐32.3 (reviewed in [3]). The first three exons and a short 5’ segment of the fourth exon encode variable, tissue‐specific 5’ untranslated regions (UTRs) of the α1‐AT mRNA, which are generated by alternative post‐transcriptional exon splicing [22, 23]. Alternate splicing of the α1‐AT 5’ UTR results in variable inclusion or exclusion of long upstream open reading frames (ORFs) whose presence or absence alters the efficiency of α1‐AT translation [24] and could potentially affect hepatic phenotypes in PIZZ subjects. Most of the fourth exon and the remaining three exons encode the α1‐AT protein sequence. The α1‐AT protein is a single‐chain, 55 kDa polypeptide with 394 amino acids and 3 asparagine‐linked complex carbohydrate side‐chains. Two major serum isoforms differ in the configuration of the carbohydrate side‐chains. α1‐AT is the archetype of the SERPIN family of protease inhibitors [3]. Most SERPINs function as suicide inhibitors of specific target proteases. Others are not inhibitory, but rather form complexes with but do not inactivate their hormone ligands. The reactive site “P1” residue of α1‐AT is the

most important determinant of functional specificity for each SERPIN molecule. This concept was dramatically confirmed by the discovery of α1‐AT Pittsburgh, a variant in which the P1 methionine 358 residue of α1‐AT is replaced by arginine. In this variant, α1‐AT functions as a thrombin inhibitor, and severe bleeding diathesis results [25]. α1‐AT is an inhibitor of serine proteases in general, but under physiologic conditions its targets are probably only neutrophil elastase, cathepsin G, and proteinase 3, proteases released by activated neutrophils because the kinetics of association with these enzymes are dramatically more favorable than with other serine protease [26]. α1‐AT acts competitively by allowing target enzymes to bind directly to its P1 Met residue‐containing reactive loop. The resulting complex contains one molecule of each reactant. The complex of α1‐AT and serine protease is a covalently stabilized structure, resistant to dissociation by denaturing compounds. The interaction between α1‐AT and serine protease is “suicidal” in that the inhibitor is irreversibly modified and no longer able to bind or inactivate enzyme. Studies also show that the irreversible trapping of the target enzyme is mediated by a profound conformational change in α1‐AT, which Carrell and Lomas likened to a “mousetrap, with the active inhibitor circulating in the

648

THE LIVER:  MECHANISM OF DEFICIENCY

metastable stressed‐form and then springing into the stable, relaxed form to lock the complex with its target protease” [27]. The functional activity of α1‐AT in vivo may be regulated by several factors. For one, it may be rendered inactive as an elastase inhibitor by active oxygen products, intermediates of activated neutrophils and macrophages that can oxidize the reactive‐site methionine of α1‐AT [5]. This effect is thought to constitute the basis for increased susceptibility to emphysema in smokers, whether deficient in α1‐AT or not. The α1‐AT molecule may also be inactivated in vivo by proteases including collagenase and pseudomonas elastase [26, 28]. Several studies indicate that α1‐ AT protects experimental animals from the lethal effects of tumor necrosis factor (TNF) [29]. This protective effect is thought to be due to inhibition of the synthesis and release of platelet‐activating factor from neutrophils [30], presumably through the inhibition of neutrophil‐derived proteases. α1‐AT also appears to have functional activities that do not involve inhibition of neutrophil proteases. The C‐terminal fragment of α1‐AT, which can be generated during the formation of a complex with serine protease or during proteolytic inactivation by thiol‐ or metalloproteases, is a potent neutrophil chemoattractant [31]. The predominant site of synthesis of plasma α1‐AT is the liver [32]. Tissue‐specific expression of α1‐AT in human hepatoma cells is directed by structural elements recognized by nuclear transcription factors including HNF‐1α and HNF‐1β, C‐EBP, HNF‐4, and HNF‐3 (reviewed in [3]). Plasma concentrations of α1‐AT increase three‐ to fivefold during the host response to inflammation and/or tissue injury [3]. The source of this additional α1‐AT has always been considered the liver; thus, α1‐AT is classified as a hepatic acute phase reactant. Synthesis of α1‐AT in human hepatoma cells (HepG2, Hep3B) is upregulated by interleukin 6 (IL‐6) but not interleukin 1 (IL‐1) or TNF [33]. Plasma concentrations also increase during oral contraceptive therapy and pregnancy [34]. α1‐AT is expressed in monocytes and macrophages [35] and mRNA has been isolated from multiple tissues in transgenic mice [36, 37], but only in some cases have studies distinguished whether such mRNA is in ubiquitous tissue macrophages or other cell types. For instance, α1‐AT is synthesized in enterocytes and intestinal paneth cells, as determined by studies in intestinal epithelial cell lines, ribonuclease protection assays of human intestinal RNA, and in situ hybridization analysis in cryostat sections of human intestinal mucosa [23, 38]. α1‐AT is also synthesized by pulmonary epithelial cells, with such synthesis less responsive to IL‐6 than to a related cytokine, oncostatin M [39]. The half‐life of α1‐antitrypsin in plasma is approximately 5 days [40]. It is estimated that the daily production rate of α1‐AT is 34 mg kg−1 body weight, with 33% of the intravascular pool degraded daily. Several physiologic factors may affect the rate of α1‐AT catabolism. First, desialylated α1‐AT is cleared from the circulation in minutes [3], probably via hepatic asialoglycoprotein receptor‐mediated endocytosis. Second, α1‐AT in complex with elastase or proteolytically modified is cleared more rapidly than native α1‐AT [41]. Because its ligand specificity is similar to that required for in vivo clearance of serpin‐enzyme complexes (SECs), the SEC receptor may also be involved in the clearance and catabolism of α1‐AT‐elastase and other serpin‐enzyme complexes [42]. The low‐density protein receptor‐related protein

(LRP) can also mediate clearance and catabolism of α1‐AT‐ elastase complexes [43]. α1‐AT diffuses into most tissues and is found in most body fluids [5].

DEFICIENCY VARIANTS OF α1‐ANTITRYPSIN Human α1‐AT variants are classified according to the protease inhibitor (PI1,2) phenotype system as defined by agarose electrophoresis or isoelectric focusing of plasma in polyacrylamide at acid pH [44]. The PI classification assigns a letter to variants, according to migration of the major isoform, using alphabetic order from anode to cathode, or from low to high isoelectric point. The most common normal variant migrates to an intermediate isoelectric point, designated M, and according to PI genotype nomenclature individuals homozygous for this variant are designated PIMM. The most common severe deficiency allelic variant migrates to a high isoelectric point, designated Z, and individuals homozygous for this variant are designated PIZZ. Several null and dysfunctional variants, with absent or reduced serum levels or activity, have been reported, some associated with premature development of emphysema. In addition to Z, other deficiency variants with reduced serum α1‐AT concentrations have been reported (Table  50.1). Some are not associated with clinical disease such as the S variant [45, 46]. Several are associated with premature lung disease (Table 50.1). With respect to other AT variants associated with liver disease, hepatocyte α1‐AT inclusions and liver disease were reported in two persons with MMalton and one with MDuarte [3, 18]. One person with the deficiency variant Siiyama, with emphysema and hepatocyte inclusions but without liver disease, was also reported [47]. Several novel deficiency variants with a tendency to accumulate and polymerize in cell culture models were recently discovered by gene sequencing of subjects with discrepancies between serum α1‐AT concentration and targeted variant genotyping for Z and S variants [48]. Another recent study reported that the Z variant, which is the most common deficiency variant and the variant most often associated with α1‐AT deficiency liver disease, is able to heteropolymerize with S, Mmalton, and Mwurzburg, in cell culture models [49].

MECHANISM OF DEFICIENCY A point mutation results in the substitution of lysine for glutamate 342 [27] and renders the mutant α1‐ATZ molecule impaired in its capacity to transverse the secretory pathway. Newly synthesized α1‐ATZ accumulates predominantly in the ER, and perhaps other as yet undefined compartments that do not stain with conventional markers of the ER [50], with relatively limited proportions reaching the extracellular fluid [51]. This impairment is observed in liver cells, macrophages, transfected cell lines, and iPS (induced pluripotent stem cell)‐derived hepatocyte‐like cells (iHeps) [3, 50]. Site‐directed mutagenesis studies have shown that the single amino acid substitution

50:  α1-ANTITRYPSIN DEFICIENCY

(E342K) is sufficient to produce the defect in secretion [50, 52]. The mutant α1‐ATZ molecule is partially functionally active, having about 50% to 80% of the elastase inhibitory capacity of wild‐type α1‐ATM [53]. There is a modest increase in the rate of in vivo clearance/catabolism of radiolabeled α1‐ATZ compared with wild‐type α1‐ATM when infused into normal individuals, but this difference does not account for the decrease in blood levels of α1‐AT in deficient individuals [41]. The point mutation also renders the α1‐ATZ molecule prone to polymerization and aggregation. Carrell and Lomas have demonstrated the presence of polymers and aggregates in liver biopsy specimens and plasma of patients with homozygous PIZZ α1‐antitrypsin deficiency [27]. Their studies suggested that a “loop‐sheet insertion” mechanism was responsible for the polymerization [54]. Similar polymers have been found in the plasma of patients with the PISiiyama α1‐AT variant and the PIMMalton α1‐AT variant [20, 55]. The mutation in α1‐AT PISiiyama (Ser 53 to Phe) [47], and in α1‐ATPIMMalton (Phe 52 deletion) [18] affect residues that provide a ridge for the sliding movement that opens the A sheet. Thus, these mutations would be expected to interfere with the insertion of the reactive center loop into the gap in the A sheet, and therefore leave the gap in the A sheet available for spontaneous loop‐sheet polymerization. It is indeed interesting that hepatocellular α1‐AT globules have been observed in a few patients with these two variants. Recent observations suggest that the α1‐ATS variant also undergoes loop‐sheet polymerization [19] and that this may account for its retention in the ER, albeit a milder degree of retention than that for α1‐ATZ [45]. Moreover, α1‐ATS can apparently form heteropolymer with α1‐ATZ [19], providing a potential explanation for liver disease in patients with the SZ phenotype [56]. More recent studies by Huntington and colleagues have suggested a different mechanism for polymerogenic and aggregation‐prone properties of ATZ that involves at least two different domain‐swapping phenomena [57–59]. This mechanism seems most consistent with recent characterization of a putative ATZ monomer crystal structure [60]. By themselves, however, these data do not prove that the polymerization of α1‐ATZ results in retention in the ER. In fact, many polypeptides must assemble into oligomeric or polymeric complexes to traverse the ER and reach their destination within the cell, at the surface of the plasma membrane, or into the extracellular fluid [61]. Evidence that polymerization results in the retention of α1‐ATZ in the ER was originally attributed to studies in which the fate of α1‐ATZ was examined after the introduction of additional mutations into the molecule. For instance, Kim et al. introduced a mutation into the α1‐AT molecule at amino acid 51, F51L [62]. This mutation is remote from the Z mutation, E342K, but apparently impedes polymerization and prevents insertion of a synthetic peptide into the gap in the A sheet, implying that the mutation leads to closing of this gap. The double‐mutated F51L α1‐ATZ molecule was less prone to  polymerization and folded more efficiently in vitro than α1‐ATZ. Moreover, the introduction of the F51L mutation partially corrected the intracellular retention properties of α1‐ATZ in microinjected Xenopus oocytes [63] and in yeast [64]. Nevertheless, several lines of evidence suggest that misfolding is the primary defect responsible for impaired secretion/

649

intracellular accumulation of ATZ and, in this conceptualization, polymerization/aggregation is a result of, rather than, the primary defect itself. First, only 18% of the intracellular pool of ATZ is in polymers at steady state in a mammalian cell line model that faithfully recapitulates the intracellular accumulation/fate of ATZ [65]. Second, a naturally occurring variant of ATZ, which has the same E342K mutation as ATZ and also a C‐terminal truncation, accumulates in the ER even though it does not form polymers, suggesting that misfolding is sufficient to lead to intracellular accumulation of ATZ [66]. Third, the results of Sidhar et al. [63] and Kang et al. [64] do not exclude the possibility that diminished secretion of ATZ is partially ­corrected by the second engineered mutation because this second mutation also prevents the primary misfolding defect. Furthermore, in a very interesting study a small molecule that prevents polymerization in vitro does not correct the secretory defect of ATZ in vivo but rather leads to enhanced degradation [67]. Taken together, the data suggests that misfolding is the primary defect and that polymerization is a time‐dependent effect of misfolding and accumulation. This conceptualization is also consistent with the domain‐swapping mechanism of polymerization described by Huntington [57–59] and older studies of folding kinetics by Yu et al. [68] in which polymerization is viewed as a “kinetic” result of delayed folding of monomeric ATZ, and explains how some ATZ gets secreted. The distinction between misfolding or polymerization as the primary inciting event for intracellular accumulation/impaired secretion is important when considering potential therapeutic approaches. If misfolding is the primary event a therapy that prevents polymerization but not misfolding might fail to alter the accumulation and/or the impaired secretion. In either case, however, there is powerful evidence that the tendency for α1‐ATZ to misfold, polymerize and aberrantly accumulate within the secretory pathway is integral to its proteotoxicity and disease‐causing potential. In studies of two families with autosomal dominantly inherited dementia, unique neuronal inclusion bodies were shown to be associated with polymers of a neuron‐specific member of the SERPIN family, neuroserpin [69]. Moreover, the mutation in neuroserpin in one family is homologous to the mutation in the α1‐AT Siiyama allele that is associated with polymerization and inclusions in the ER of liver cells. In one of the most interesting studies on the mechanism for ATZ accumulation in the ER, Nyfeler et al. provided evidence for the lectin ER Golgi intermediate compartment 53 kDa protein (ERGIC‐53) as an export receptor for α1‐antitrypsin [70]. Most importantly, ERGIC‐53 failed to recognize mutant α1‐ATZ. This study did not address whether polymerization prevents α1‐ATZ from being presented to ERGIC‐53 or the altered folding pathway of α1‐ATZ prevents an essential ligand domain from being available to bind to ERGIC‐53.

PATHOGENESIS OF LIVER DISEASE Although there are still many gaps in our understanding of how the deficiency causes liver disease in some of the homozygotes, we know that aberrant intracellular accumulation of α1‐ATZ is

650

THE LIVER:  PATHWAYS FOR INTRACELLULAR DEGRADATION OF MUTANT α1-ATZ

the unifying characteristic of what would be called a gain‐of‐ toxic‐function mechanism. Experimental results in transgenic mice provide the most powerful evidence. Transgenic mice carrying the mutant human α1‐ATZ allele develop periodic acid– Schiff‐positive, diastase‐resistant intrahepatic globules and liver injury [6, 7]. Because there are normal levels of antielastases in these animals, as directed by endogenous genes, the liver injury cannot be attributed to a loss‐of‐function mechanism. The most extensively characterized transgenic mouse model, the PiZ mouse, generated with a human genomic fragment encompassing all of the exons and introns of the human ATZ gene [6], recapitulates the hepatic pathology of the human disease with intrahepatocytic globules reflecting accumulation of misfolded ATZ in the ER, slowly progressing fibrosis, low‐grade inflammation, dysplasia, and increased prevalence of hepatocellular carcinoma [71, 72]. In work with this mouse model over many years we have found that marked progressive nodular regeneration is the dominant hepatic pathological characteristic, strongly resembling liver pathology in the human disease. There is still relatively limited information about how proteinopathy/intracellular α1‐ATZ accumulation leads to liver injury. Although structural and functional alterations in mitochondria and caspase‐3 activation have been observed in liver tissue from the PiZ mouse model and from α1‐antitrypsin deficiency patients [73, 74], mitochondrial dysfunction probably has mostly a cytostatic effect because apoptosis, necrosis, and inflammation are not major pathological characteristics of the liver in α1‐antitrypsin deficiency. The mechanisms responsible for the hepatic fibrotic response to proteinopathy are also not well understood. Genomic analysis of a mouse model with hepatocyte‐specific inducible expression of mutant α1‐ATZ, ideal for elucidating signal transduction pathways that are activated by the proteinopathy, has identified a relatively rich network of downstream targets of the TGFβ pathway, including upregulation of connective tissue growth factor [75], and this pathway is known to play a central role in fibrosis [76]. We have also demonstrated that activation of the NFκB signaling pathway is an important and specific effect of cellular ATZ accumulation [77] and is designed to prevent fibrogenesis by activating collagen‐degrading metalloproteases [78]. Many recent studies have also suggested that fibrosis results from proteinopathies in several other tissues. Lung fibrosis has been described in several rare diseases characterized by proteinopathy in respiratory epithelial cells, including surfactant protein C deficiency and Hermansky–Pudlak syndrome [79, 80]. Similarly myocardial fibrosis has been described for desminopathy that affects cardiomyocytes [81] and skeletal muscle fibrosis in inclusion‐body myositis [82]. Interestingly, by enhancing the degradation of misfolded proteins, autophagy has been shown to mitigate cardiac fibrosis from desminopathy [81] and skeletal muscle fibrosis from inclusion‐body myositis [83] just as it does for hepatic fibrosis in the PiZ model of α1‐antitrypsin deficiency. The liver of the PiZ mouse also shows glycogen depletion [84] and defective ureagenesis [85]. The latter has been attributed to downregulation of hepatocyte nuclear factor‐4α [85]. Because these functional abnormalities are seen clinically in severe forms of liver disease, this report provides additional evidence for the validity of the PiZ mouse as a model of the human disease.

For many years it was conceptually difficult to reconcile the gain‐of‐toxic function mechanism with the observations of Sveger, showing that only a subset of α1‐AT‐deficient homozygotes develop significant liver damage. In 1994, we published a series of experiments investigating the hypothesis that a subset of the PIZZ population is more susceptible to liver injury because of one or more additional inherited traits or environmental factors that exaggerate either the intracellular accumulation of the mutant α1‐ATZ or the cellular pathophysiological consequence of mutant α1‐ATZ accumulation [86]. Furthermore, we hypothesized that these putative genetic/environmental modifiers would affect intracellular degradation of α1‐ATZ or adaptive signaling pathways that are activated by the intracellular accumulation of α1‐ATZ. This hypothetical model was validated by showing a lag in intracellular degradation of α1‐ ATZ in skin fibroblasts from human PIZZ homozygotes already known to have severe liver disease (“susceptible hosts”) as compared to fibroblasts from homozygotes who had no liver disease (“protected hosts”). The fibroblasts had been engineered to express α1‐ATZ using gene transfer by amphotropic recombinant retroviral particles. More recently we found similar results in iPS‐derived hepatocyte cells (iHeps) [50]. The rate of degradation of α1‐ATZ was slower in iHeps from susceptible hosts than in iHeps from protected hosts. Over the years since the 1994 skin fibroblast experiments we have characterized the putative pathways for intracellular ­degradation of α1‐ATZ and the putative adaptive signaling pathways activated by intracellular accumulation of α1‐ATZ (Figure 50.1).

PATHWAYS FOR INTRACELLULAR DEGRADATION OF MUTANT α1‐ATZ Early studies using yeast and mammalian cell lines showed that the proteasomal pathway participates in intracellular disposal of α1‐ATZ by a process that is now known as ER‐associated degradation (ERAD) in which the substrate is extracted retrograde from the ER to the cytoplasm [87, 88]. In fact, α1‐ATZ was one of the first identified substrates of the ERAD pathway [88]. Autophagy was subsequently identified as a second major pathway for disposal of misfolded α1‐ATZ [21]. Autophagy is an intracellular catabolic pathway by which cells digest subcellular structures and cytoplasm to generate amino acids as a survival mechanism (Figure  50.2). It is characterized by double membrane vacuoles called autophagosomes, which fuse with lysosomes for degradation of the internal constituents. Increased numbers of autophagosomes were observed in human fibroblast cell lines engineered for expression of mutant α1‐ATZ, in the liver of PiZ transgenic mice and in liver biopsy specimens from patients with ATD. Definitive evidence for the role of autophagy in ATZ disposal was provided by genetic studies in which α1‐ ATZ disposal was delayed in autophagy‐deficient [Atg5‐null] murine embryonic fibroblast cell lines [89] and Atg6‐null yeast strains [90, 91]. The importance of the autophagic pathway in intracellular ATZ degradation has been further validated by recent studies demonstrating that autophagy enhancer drugs promote intracellular ATZ

50:  α1-ANTITRYPSIN DEFICIENCY

651

Figure 50.1  Conceptual model for mechanisms of proteotoxicity in α1‐AT deficiency liver disease. Cellular factors that determine whether an AT‐deficient individual is protected or susceptible to liver disease. In the susceptible host, there is greater accumulation of misfolded ATZ in the ER because of subtle alternations in putative proteostasis network regulatory mechanisms. Here these proteostasis regulatory mechanisms are envisioned as either ER degradation or signaling pathways that represent cellular protective responses. Reproduced with permission of Cold Spring Harbor Laboratory Press from [122].

Figure 50.2  (a) Intracellular molecules and organelles destined for autophagic degradation (termed “cargo”) are surrounded by a double‐membrane structure termed the “isolation membrane” (IM). The mechanisms regulating formation of the IM and its targeting to cargo remain incompletely defined. The IM matures into the autophagosome vacuole, separating the cargo from other cellular compartments, and then fuses with the lysosome to expose the cargo to degradative enzymes. Image adapted from [123] and available under the terms of the Creative Commons Attribution License, CCBY 4.0.

disposal and attenuate hepatic fibrosis in the PiZ mouse model of ATD in vivo [71]. One of the concepts originating from studies of ATZ disposal in autophagy‐deficient yeast strains is that autophagy becomes particularly important at higher levels of ATZ expression (reviewed in [92]). These results taken together with the structural constraints of the proteasome have led to the supposition that the proteasomal pathway degrades soluble monomeric forms of ATZ whereas autophagy is needed for soluble and

insoluble polymers. However, it is also possible that autophagy plays a role in the disposal of soluble monomeric ATZ that accumulates at levels of expression that exceed the capacity of the proteasome. Another important result from the studies by Kruse et al. in autophagy‐deficient yeast showed that a misfolded fibrinogen variant associated with liver disease in a rare inherited form of hypofibrinogenemia is degraded by autophagy in a manner almost identical to that of misfolded ATZ [91].

652

THE LIVER:  HEPATIC CARCINOMA IN α1-ANTITRYPSIN DEFICIENCY

Pathways for intracellular disposal of ATZ other than the proteasomal system and the canonical macroautophagy system are highly likely. For example, a sortilin‐mediated pathway from Golgi to lysosome has been described to participate in degradation of ATZ in yeast and mammalian cell line models [90, 93]. Another pathway for ATZ disposal which diverges from the canonical autophagy system was recently identified in a powerful Caenorhabditis elegans model of ATD and found to be present in a mammalian cell line model as well [94]. This pathway is particularly interesting because it is suppressed by insulin signaling and when upregulated by knocking down components of the insulin signaling pathway it can completely mitigate ATZ proteotoxicity.

SIGNALING PATHWAYS ACTIVATED BY ACCUMULATION OF α1‐ATZ IN THE ENDOPLASMIC RETICULUM To determine which signaling pathways are activated when α1‐ATZ accumulated in the ER, we developed cell line and mouse model systems with inducible rather than constitutive expression of α1‐ATZ because the latter would potentially permit adaptations that could obscure the primary signaling effects. A series of studies using these kinds of systems have shown that the autophagic response and the nuclear factor κB (NFκB) signaling pathway, but not the unfolded protein response, are a­ ctivated when α1‐ATZ accumulates in the ER [75, 77, 89]. Activation of NFκB appears to be a hallmark of α1‐ATZ ­accumulation [75, 77]. In recent studies in which the PiZ mouse is mated to two different mouse models engineered for absence of NFκB signaling we found more severe inflammation, fibrosis, steatosis, dysplasia, and more hepatocytes with globules [78] indicating that NFκB signaling is intended to protect the liver from the effects of α1‐ATZ accumulation. Interestingly the effect of NFκB signaling pathway under these circumstances is through a small number of downstream target genes: Egr‐1, a transcription factor that is essential for liver regeneration [95]; RGS16, an inhibitor of G‐protein signaling that has been implicated in activation of autophagy [75]; and matrix metalloproteases MMP7 and MMP12. Each of these targets appear to be mediating parts of the hepatic response to α1‐ATZ accumulation: decreased proliferation of hepatocytes with massive ATZ accumulation, known as globule‐containing hepatocytes, due to Egr1‐downregulation; increased autophagy due to RGS16 upregulation; and counteracting of fibrogenesis by MMP7 and MMP12 upregulation. The hepatic transcriptomic analysis of the Z mouse also shows changes in gene expression indicative of TGFβ signaling and this is consistent with the fibrotic response that represents the dominant pathological characteristics of the liver in α1‐­ antitrypsin deficiency [75]. Furthermore we have recently found that accumulation of the α1‐ATZ variant in respiratory epithelial cells elicits fibrosis in the lungs with evidence for fibrosis in the lungs of α1‐antitrypsin deficiency patients with very severe COPD (chronic obstructive pulmonary disease) [96]. The mechanism by which accumulation of misfolded proteins in the ER elicits TGFβ signaling has not been studied.

c‐Jun N‐terminal kinase activation was recently demonstrated in the liver of the PiZ mouse model, as well as in liver specimens from patients with α1‐antitrypsin deficiency [97]. Interestingly this signaling effect leads to increased expression of α1‐ATZ and therefore likely plays a role in amplifying the pathobiology of misfolded protein accumulation.

HEPATIC CARCINOMA IN α1‐ANTITRYPSIN DEFICIENCY Although increased susceptibility to hepatic cancer was shown years ago [4], there have been only a few studies of potential mechanisms for carcinogenesis. Theorizing that hepatocellular hyperproliferation was likely to be involved, Rudnick et  al. investigated the proliferation of liver cells by BrdU labeling in the PiZ mouse model [98]. In these studies hepatocellular proliferation was increased around sevenfold in the PiZ mouse compared to a wild‐type control and this degree of proliferation appeared to reflect the slowly progressing chronic nature of α1‐ antitrypsin deficiency liver disease. Most importantly, by using double immunohistochemical staining it was shown that dividing hepatocytes were almost exclusively the ones that lacked intracellular α1‐ATZ inclusions, called the globule‐devoid hepatocytes. Furthermore it was shown that hyperproliferation of globule‐devoid hepatocytes was driven by the number of adjacent globule‐containing hepatocytes. This last conclusion was based on the observation that the number of globule‐­ containing hepatocytes was markedly increased in male PiZ mice or in testosterone‐treated female PiZ mice and this correlated directly with the degree of hyperproliferation of globule‐ devoid hepatocytes. Taken together these observations led to a theory that hepatocyte hyperproliferation was elicited by crosstalk between globule‐containing and globule‐devoid hepatocytes ([99] and Figure 50.3). The globule‐devoid hepatocytes were viewed as younger cells capable of responding to trans‐acting regenerative signals derived from the globule‐containing hepatocytes. The globule‐containing hepatocytes were viewed as having greater proteotoxic α1‐ATZ accumulation and unable to respond to the existing regenerative signals because of the proteotoxic effect on cell proliferation. Thus, the globule‐containing hepatocytes are sick but not dead and stimulate the regeneration of the globule‐devoid hepatocytes which have a selective proliferative advantage. Interestingly, the replicative defect in the globule‐containing hepatocytes was shown to be relative because these cells could proliferate as well as globule‐ devoid hepatocytes when the regenerative stimulus was particularly powerful, such as in PiZ mice that survived experimental partial hepatectomy [98]. The nature of the differences between the globule‐containing and globule‐devoid cells is not well elucidated. A study by Linblad et al. has suggested that the globule‐devoid hepatocytes have lesser accumulation of ATZ [74] and this would be consistent with younger cells that have had less time to accumulate α1‐ATZ. It is also possible that globule‐devoid hepatocytes are derived from globule‐ containing hepatocytes as they increase capacity for degradation of α1‐ATZ. Several observations militate against this latter

50:  α1-ANTITRYPSIN DEFICIENCY

653

Figure 50.3  Hypothetical model for hepatocarcinogenesis in ATD. Globule‐containing hepatocytes (pale pink) tend to be periportal. They are “sick but not dead” and generate chronic regenerative signals which can only be received effectively in “trans” by globule‐devoid hepatocytes (deep pink). The globule‐devoid hepatocytes tend to be in the centrilobular regions. When regenerative signals are received by globule‐devoid hepatocytes by this crosstalk, it drives mitosis and ultimately carcinogenesis (dark red) in the globule‐devoid regions. Reproduced with permission of John Wiley & Sons from [99].

possibility. The number of globule‐containing hepatocytes decrease with age [98], and Ding et al. [100] showed that transplanted hepatocytes have a selective proliferative advantage that also depends on the number of adjacent globule‐containing hepatocytes in that it was much more evident in male PiZ mice that had significantly more globule‐containing hepatocytes than female PiZ mice. This is associated with enhanced apoptosis of the host hepatocytes, hepatic repopulation with donor hepatocytes, and resolution of the liver fibrosis that occurs in untreated PiZ mice [100].

It is interesting to note that hepatocellular carcinoma develops with aging in male PiZ mice [72] and males with α1‐antitrypsin deficiency were also disproportionately affected by hepatic ­cancer in the autopsy studies of Eriksson et al. [4]. Moreover, in cases of hepatocellular carcinoma associated with α1‐antitrypsin deficiency one observes a staining pattern in which the carcinoma is negative for inclusions but surrounded by adjacent liver cells that are positive for inclusions which is entirely consistent with the carcinogenesis theory proposed by Rudnick and Perlmutter [99].

654

THE LIVER:  TREATMENT

MODIFIERS OF THE HEPATIC PHENOTYPYE OF α1‐ANTITRYPSIN DEFICIENCY Studies designed to identify genetic and environmental modifiers of the liver disease phenotype in human populations have begun to appear in the literature in recent years. In one interesting study, a single nucleotide polymorphism (SNP) in the MAN1B1 gene was found to be statistically over‐represented in a series of infants with end‐stage liver disease [101]. The variant was shown to reduce intracellular levels of the mannosidase [102]. Recent experiments have shown that Man1B1 is actually localized to the Golgi but it plays a role in regulation of protein secretion as a part of the protein quality control network which is recently recognized to be localized in the Golgi [103]. Furthermore those ­experiments have provided a basis for how reduced levels of Man1B1 could theoretically lead to greater intracellular α1‐ATZ ­accumulation [103]. These results for the Man1B1 variant would appear to validate our hypothesis that intracellular degradation pathways are targets of liver disease modifiers but further population studies of this variant would be reassuring [104, 105]. A SNP in the upstream flanking region of the α1‐antitrypsin gene has also been implicated in susceptibility to liver disease [106]. However, the nature of that variant could not be reconciled with how it might affect liver disease susceptibility and its  statistical association with variation in the liver disease ­phenotype was dependent on a questionable classification of population subgroups. Our hypothesis for variation in liver disease susceptibility also identifies signaling pathways that could increase or decrease α1‐ATZ proteotoxicity as potential targets for disease modifiers. As of yet we have not encountered an example of this potential scenario, but we predict that further studies of iHeps from patients with different forms of α1‐antitrypsin deficiency liver disease, in terms of age of onset and type of hepatic pathology, will identify such a mechanism in the near future.

TREATMENT The most important principle in the treatment of α1‐AT ­deficiency is avoidance of cigarette smoking. Cigarette smoking markedly accelerates the destructive lung disease that is associated with α1‐AT deficiency, reduces the quality of life, and ­significantly shortens the longevity of these individuals [107]. There is no specific therapy for α1‐AT deficiency‐associated liver disease. Therefore, clinical care largely involves supportive management of symptoms due to liver dysfunction and for the prevention of complications. Progressive liver dysfunction in α1‐AT‐deficient patients has been treated by orthotopic liver transplantation, with 5‐year survival rates approaching 90% in children and 80% in adults at 1 year and 80% at 5 years [108]. Several novel strategies for treatment of α1‐antitrypsin deficiency liver disease that would obviate the need for organ transplantation and chronic immunosuppression are currently under investigation and at various stages of development. One of the relatively newer strategies targets intracellular degradation ­pathways using autophagy enhancer drugs.

Autophagy was considered an excellent target because it is specifically activated when α1‐ATZ accumulates in cells and it also plays a critical role in intracellular disposal of ATZ. At the time when this approach was first investigated several drugs which could enhance autophagic degradation of other misfolded proteins, such as mutant polyQ proteins that cause Huntington disease, were being described [109]. It had also become apparent that α1‐antitrypsin deficiency liver disease could more frequently have its onset at 50–65 years of age [16] coincident with the decline in autophagy function that is believed to trigger other age‐ dependent degenerative diseases associated with misfolded ­proteins. Hidvegi et al. first investigated the drug carbamazepine (CBZ), which is known for its widespread use in humans as an anticonvulsant and mood stabilizer and found that it enhanced autophagic degradation of α1‐ATZ in mammalian cell line models [71]. Moreover, administration of this drug by oral gavage to the PiZ mouse model over a three‐week period significantly reduced hepatic α1‐ATZ load and hepatic fibrosis in vivo. Because CBZ is already FDA‐approved it could be moved immediately into a phase 2/3 clinical trial for treatment of severe liver disease due to ATD. Several other drugs with autophagy enhancer properties have been identified by high‐throughput screening of drug libraries and these are currently under investigation [92]. A number of relatively new gene therapy strategies are being investigated. One of these involves new methods for silencing gene expression using vectors that are also capable of encoding wild‐type α1‐antitrypsin to address both gain‐ and loss‐of function sequelae of α1‐antitrypsin deficiency, respectively ­ [110, 111]. In one approach, Li et al. utilized adeno‐associated virus harboring short‐hairpin RNA to knockdown endogenous α1‐ATZ expression together with a codon‐optimized wild‐type α1‐antitrypsin transgene cassette [110]. In another approach, Mueller et  al. utilized an adeno‐associated virus harboring microRNA to silence endogenous α1‐ATZ gene expression together with a microRNA‐resistant wild‐type α1‐antitrypsin gene [111]. In each case hepatic α1‐ATZ load was reduced and levels of human α1‐antitrypsin increased in the serum of a transgenic mouse model. However, the effect on liver fibrosis by this strategy was not as compelling and hence further studies are required to test whether more potent and widespread silencing would be more effective. Another study using antisense oligonucleotides by systemic administration to silence α1‐ATZ gene expression had a more impressive effect in reducing hepatic fibrosis in the PiZ mouse model system [112]. Another potential gene therapy approach being investigated is the transfer of genes that activate autophagy and therein reduce α1‐ATZ accumulation and proteotoxicity. Pastore et al have championed this approach using TFEB, a master transcriptional activator of the autophagolysosomal system [113]. Using helper‐dependent adenovirus for systemic delivery of TFEB and targeting of its expression to liver, this approach significantly reduced hepatic α1‐ATZ load and liver fibrosis in the PiZ mouse model. In vitro studies also validated that TFEB reduces cellular α1‐ATZ levels in an autophagy‐dependent manner [113]. Although it will not address the loss‐of‐function mechanisms associated with α1‐antitrypsin deficiency lung disease, gene therapy with TFEB or drugs that target TFEB activation ­constitute exciting potential therapeutic strategies for liver disease associated with α1‐antitrypsin deficiency.

50:  α1-ANTITRYPSIN DEFICIENCY

Ultimately, genomic editing will be considered to definitively correct the genetic defect that causes ATD. The most recent development in this area, CRISPR/Cas‐9‐mediated genome editing, has been used in the PiZ mouse model in vivo, showing reduced α1‐ATZ in the liver and low levels of wild‐type α1AT in the liver [114] and in serum [115]. Several research groups are exploring a “structure‐based” screening strategy that aims to generate peptides to prevent polymerization of the mutant α1‐ATZ with the hypothesis that this would facilitate secretion. A small molecule compound designed against a lateral hydrophobic cavity in α1‐ATZ prevented its polymerization, however, further experiments in a cell line model revealed that this compound enhanced intracellular degradation only, with minimal effect on secretion [67]. These results provide further evidence that it is the misfolding of α1‐ATZ, independent of its tendency to polymerize, which is primarily responsible for impaired secretion. Another small molecule based on a peptide that targets the reactive center loop of α1‐antitrypsin has been designed and introduced in cell line model systems with evidence for improved secretion of α1‐ATZ [116]. However, the efficacy of this peptide in an animal model system for either increasing secretion or reducing liver damage in vivo remains to be tested. It also remains possible that this type of peptide binding changes the conformation of the mutant protein in such a way that both misfolding and polymerization are reduced independently. Chemical chaperones that can non‐selectively facilitate folding of diverse misfolded proteins have also been investigated as a potential therapeutic option for α1‐antitrypsin deficiency liver disease. Glycerol and 4‐phenylbutyric acid (PBA) were found to mediate a robust enhancement in the secretion of α1‐ATZ in a mammalian cell line model and its oral administration in PiZ mice increased blood levels of human α1‐antitrypsin reaching 20–50% of the levels present in PiM mice and normal humans [117]. However, a pilot clinical trial involving 10 patients with α1‐antitrypsin deficiency‐associated liver disease, failed to reveal any significant increase in serum levels of α1‐antitrypsin after 14 days of treatment with PBA [118]. It is not clear why the drug lacked effect but the large doses required are known to be quite challenging to tolerate and so it may be worthwhile to test in the future if newer, more tolerable formulations are developed. Recently, suberoylanilide hydroxamic acid (SAHA), another drug which has many pharmacological similarities to PBA, has been found to enhance secretion of α1‐ATZ in cell line models of α1‐antitrypsin deficiency [119]. However, SAHA has not yet been tested in animal models. Moreover, detailed studies are needed to delineate whether this effect is due to increased synthesis of α1‐ATZ or due to its ability to reduce α1‐ATZ accumulation in cells or both. If increased secretion of α1‐ATZ is because of, or even associated with, increased synthesis, the treatment could produce more rather than less cellular proteotoxicity. Hepatocyte transplantation therapy has also been investigated as a potential treatment for α1‐antitrypsin deficiency. It has been tested in the past as a treatment for several metabolic liver diseases [120]. Compared to orthotopic liver transplantation it has the advantage of being a minimally invasive procedure with little known morbidity, and is considerably less expensive than protein replacement therapy or liver transplantation. Importantly, recent studies have revealed that wild‐type donor hepatocytes

655

can repopulate almost the entire liver of the PiZ mouse model [100]. In the PIZ mouse model, the donor cells replaced both globule‐containing and globule‐devoid cells, indicating that both types of affected hepatocytes have impaired proliferative capacity compared to wild‐type hepatocytes. Because the transplanted hepatocytes have a selective proliferative advantage over α1‐ATZ‐containing endogenous hepatocytes and can substitute for the latter in a diseased liver, this option of therapy may be considered for α1‐antitrypsin deficiency lung and liver disease. Another exciting therapeutic strategy in which genomic editing is combined with hepatocyte transplantation has been tested in a transgenic mouse model of α1‐antitrypsin deficiency. Studies by Yusa et  al. have shown that the mutation in the α1‐antitrypsin gene could be corrected in human induced pluripotent stem (iPS) cells derived from a α1‐antitrypsin deficiency patient using a combination of zinc‐finger nucleases and transposon techniques [121]. Importantly, the corrected iPS cell lines could then be engrafted into the liver of the transgenic mouse model system and, based on the observations of Ding et  al. [100], the corrected cells should expand significantly because they will have a selective proliferative advantage. This strategy, if it proves successful in further preclinical models, has the potential to address both the loss‐ and gain‐of‐function mechanisms of organ damage and the advantage of personalized treatment options without any need for immunosuppression.

REFERENCES   1. Sveger, T. Liver disease in alpha1‐antitrypsin deficiency detected by screening of 200,000 infants. N Engl J Med, 1976;294(24):1316–21.   2. Silverman, E.K. and Sandhaus, R.A. Clinical practice. Alpha1‐antitrypsin deficiency. N Engl J Med, 2009;360(26):2749–57.   3. Teckman, J.H., Qu, D., and Perlmutter, D.H. Molecular pathogenesis of liver disease in alpha1‐antitrypsin deficiency. Hepatology, 1996;24(6):1504–16.   4. Eriksson, S., Carlson, J., and Velez, R. Risk of cirrhosis and primary liver cancer in alpha 1‐antitrypsin deficiency. N Engl J Med, 1986;314(12):736–9.   5. Crystal, R.G. Alpha 1‐antitrypsin deficiency, emphysema, and liver disease. Genetic basis and strategies for therapy. J Clin Invest, 1990;85(5):1343–52.   6. Carlson, J.A., Rogers, B.B., Sifers, R.N. et al. Accumulation of PiZ alpha 1‐antitrypsin causes liver damage in transgenic mice. J Clin Invest, 1989; 83(4):1183–90.   7. Dycaico, M.J., Grant, S.G., Felts, K. et al. Neonatal hepatitis induced by alpha 1‐antitrypsin: a transgenic mouse model. Science, 1988;242(4884):1409–12.   8. Mostafavi, B., Diaz, S., Tanash, H.A., and Piitulainen, E. Liver function in alpha‐1‐antitrypsin deficient individuals at 37 to 40 years of age. Medicine (Baltimore), 2017;96(12):e6180.   9. Silverman, E.K., Province, M.A., Rao, D.C., Pierce, J.A., and Campbell, E.J. A family study of the variability of pulmonary function in alpha 1‐antitrypsin deficiency. Quantitative phenotypes. Am Rev Respir Dis, 1990;142(5): 1015–21. 10. Crystal, R.G. Augmentation treatment for alpha1 antitrypsin deficiency. Lancet, 2015;386(9991):318–20. 11. McElvaney, N.G., Burdon, J., Holmes, M. et  al. Long‐term efficacy and safety of alpha1 proteinase inhibitor treatment for emphysema caused by severe alpha1 antitrypsin deficiency: an open‐label extension trial (RAPID‐ OLE). Lancet Respir Med, 2017;5(1):51–60. 12. Ibarguen, E., Gross, C.R., Savik, S.K., and Sharp, H.L. Liver disease in alpha‐1‐antitrypsin deficiency: prognostic indicators. J Pediatr, 1990; 117(6):864–70. 13. Sharp, H.L., Bridges, R.A., Krivit, W., and Freier, E.F. Cirrhosis associated with alpha‐1‐antitrypsin deficiency: a previously unrecognized inherited disorder. J Lab Clin Med, 1969;73(6):934–9.

656

THE LIVER:  REFERENCES

14. Hadchouel, M. and Gautier, M. Histopathologic study of the liver in the early  cholestatic phase of alpha‐1‐antitrypsin deficiency. J Pediatr, 1976; 89(2):211–5. 15. Sveger, T. The natural history of liver disease in alpha 1‐antitrypsin deficient children. Acta Paediatr Scand, 1988;77(6):847–51. 16. Chu, A.S., Chopra, K.B., and Perlmutter, D.H. Is severe progressive liver disease caused by alpha‐1‐antitrypsin deficiency more common in children or adults? Liver Transpl, 2016;22(7):886–94. 17. Schaefer, B., Mandorfer, M., Viveiros, A. et  al. Heterozygosity for the alpha‐1‐antitrypsin Z allele in cirrhosis is associated with more advanced disease. Liver Transpl, 2018;24(6):744–51. 18. Curiel, D.T., Holmes, M.D., Okayama, H. et al. Molecular basis of the liver and lung disease associated with the alpha 1‐antitrypsin deficiency allele Mmalton. J Biol Chem, 1989;264(23):13938–45. 19. Mahadeva, R., Chang, W.S., Dafforn, T.R. et al. Heteropolymerization of S, I, and Z alpha1‐antitrypsin and liver cirrhosis. J Clin Invest, 1999;103(7): 999–1006. 20. Lomas, D.A., Elliott, P.R., Sidhar, S.K. et al. alpha 1‐Antitrypsin Mmalton (Phe52‐deleted) forms loop‐sheet polymers in vivo. Evidence for the C sheet mechanism of polymerization. J Biol Chem, 1995;270(28):16864–70. 21. Teckman, J.H. and Perlmutter, D.H. Retention of mutant alpha(1)‐antitrypsin Z in endoplasmic reticulum is associated with an autophagic response. Am J Physiol Gastrointest Liver Physiol, 2000;279(5):G961–74. 22. Perlino, E., Cortese, R., and Ciliberto, G. The human alpha 1‐antitrypsin gene is transcribed from two different promoters in macrophages and hepatocytes. EMBO J, 1987;6(9):2767–71. 23. Hafeez, W., Ciliberto, G., and Perlmutter, D.H. Constitutive and modulated expression of the human alpha 1 antitrypsin gene. Different transcriptional initiation sites used in three different cell types. J Clin Invest, 1992;89(4): 1214–22. 24. Corley, M., Solem, A., Phillips, G. et al. An RNA structure‐mediated, posttranscriptional model of human alpha‐1‐antitrypsin expression. Proc Natl Acad Sci USA, 2017;114(47):E10244–53. 25. Owen, M.C., Brennan, S.O., Lewis, J.H., and Carrell, R.W. Mutation of antitrypsin to antithrombin. alpha 1‐antitrypsin Pittsburgh (358 Met leads to Arg), a fatal bleeding disorder. N Engl J Med, 1983;309(12):694–8. 26. Mast, A.E., Enghild, J.J., Nagase, H., Suzuki, K., Pizzo, S.V., ND Salvesen, G. Kinetics and physiologic relevance of the inactivation of alpha 1‐proteinase inhibitor, alpha 1‐antichymotrypsin, and antithrombin III by matrix metalloproteinases‐1 (tissue collagenase), ‐2 (72‐kDa gelatinase/type IV collagenase), and ‐3 (stromelysin). J Biol Chem, 1991;266(24):15810–6. 27. Carrell, R.W. AND Lomas, D.A. Conformational disease. Lancet, 1997; 350(9071):134–8. 28. Vissers, M.C., George, P.M., Bathurst, I.C., Brennan, S.O., and Winterbourn, C.C. Cleavage and inactivation of alpha 1‐antitrypsin by metalloproteinases released from neutrophils. J Clin Invest, 1988;82(2):706–11. 29. Van Molle, W., Libert, C., Fiers, W., and Brouckaert, P. Alpha 1‐acid glycoprotein and alpha 1‐antitrypsin inhibit TNF‐induced but not anti‐Fas‐induced apoptosis of hepatocytes in mice. J Immunol, 1997;159(7):3555–64. 30. Camussi, G., Tetta, C., Bussolino, F., and Baglioni, C. Synthesis and release of platelet‐activating factor is inhibited by plasma alpha 1‐proteinase inhibitor or alpha 1‐antichymotrypsin and is stimulated by proteinases. J Exp Med, 1988;168(4):1293–306. 31. Joslin, G., Griffin, G.L., August, A.M. et  al. The serpin‐enzyme complex (SEC) receptor mediates the neutrophil chemotactic effect of alpha‐1 antitrypsin‐elastase complexes and amyloid‐beta peptide. J Clin Invest, 1992;90(3):1150–4. 32. Hood, J.M., Koep, L.J., Peters, R.L. et al. Liver transplantation for advanced liver disease with alpha‐1‐antitrypsin deficiency. N Engl J Med, 1980; 302(5):272–5. 33. Perlmutter, D.H., May, L.T., and Sehgal, P.B. Interferon beta 2/interleukin 6 modulates synthesis of alpha 1‐antitrypsin in human mononuclear phagocytes and in human hepatoma cells. J Clin Invest, 1989;84(1):138–44. 34. Laurell, C.B. and Rannevik, G. A comparison of plasma protein changes induced by danazol, pregnancy, and estrogens. J Clin Endocrinol Metab, 1979;49(5):719–25. 35. Perlmutter, D.H., Cole, F.S., Kilbridge, P., Rossing, T.H., and Colten, H.R. Expression of the alpha 1‐proteinase inhibitor gene in human monocytes and macrophages. Proc Natl Acad Sci USA, 1985;82(3):795–9. 36. Kelsey, G.D., Povey, S., Bygrave, A.E., and Lovell‐Badge, R.H. Species‐ and tissue‐specific expression of human alpha 1‐antitrypsin in transgenic mice. Genes Dev, 1987;1(2):161–71.

37. Carlson, J.A., Rogers, B.B., Sifers, R.N., Hawkins, H.K., Finegold, M.J., and Woo, S.L. Multiple tissues express alpha 1‐antitrypsin in transgenic mice and man. J Clin Invest, 1988;82(1):26–36. 38. Molmenti, E.P., Perlmutter, D.H., and Rubin, D.C. Cell‐specific expression of alpha 1‐antitrypsin in human intestinal epithelium. J Clin Invest, 1993;92(4):2022–34. 39. Cichy, J., Potempa, J., and Travis, J. Biosynthesis of alpha1‐proteinase inhibitor by human lung‐derived epithelial cells. J Biol Chem, 1997;272(13): 8250–5. 40. Laurell, C.B., Nosslin, B., and Jeppsson, J.O. Catabolic rate of alpha1‐antitrypsin of Pi type M and Z in man. Clin Sci Mol Med, 1977;52(5):457–61. 41. Mast, A.E., Enghild, J.J., Pizzo, S.V., and Salvesen, G. Analysis of the plasma elimination kinetics and conformational stabilities of native, proteinase‐complexed, and reactive site cleaved serpins: comparison of alpha 1‐proteinase inhibitor, alpha 1‐antichymotrypsin, antithrombin II.I., alpha 2‐antiplasmin, angiotensinogen, and ovalbumin. Biochemistry, 1991;30(6):1723–30. 42. Joslin, G., Wittwer, A., Adams, S., Tollefsen, D.M., August, A., and Perlmutter, D.H. Cross‐competition for binding of alpha 1‐antitrypsin (alpha 1 AT)‐elastase complexes to the serpin‐enzyme complex receptor by other serpin‐enzyme complexes and by proteolytically modified alpha 1 A.T. J Biol Chem, 1993;268(3):1886–93. 43. Kounnas, M.Z., Church, F.C., Argraves, W.S., and Strickland, D.K. Cellular internalization and degradation of antithrombin III‐thrombin, heparin cofactor II‐thrombin, and alpha 1‐antitrypsin‐trypsin complexes is mediated by the low density lipoprotein receptor‐related protein. J Biol Chem, 1996; 271(11):6523–9. 44. Pierce, J.A. and Eradio, B.G. Improved identification of antitrypsin phenotypes through isoelectric focusing with dithioerythritol. J Lab Clin Med, 1979;94(6):826–31. 45. Teckman, J.H. and Perlmutter, D.H. The endoplasmic reticulum degradation pathway for mutant secretory proteins alpha1‐antitrypsin Z and S is distinct from that for an unassembled membrane protein. J Biol Chem, 1996; 271(22):13215–20. 46. Long, G.L., Chandra, T., Woo, S.L., Davie, E.W., and Kurachi, K. Complete sequence of the cDNA for human alpha 1‐antitrypsin and the gene for the S variant. Biochemistry, 1984;23(21):4828–37. 47. Seyama, K., Nukiwa, T., Takabe, K., Takahashi, H., Miyake, K., and Kira, S. Siiyama (serine 53 (TCC) to phenylalanine 53 (TTC)). A new alpha 1‐antitrypsin‐deficient variant with mutation on a predicted conserved residue of the serpin backbone. J Biol Chem, 1991;266(19):12627–32. 48. Matamala, N., Lara, B., Gomez‐Mariano, G. et al. Characterization of novel missense variants of SERPINA1 gene causing alpha‐1 antitrypsin deficiency. Am J Respir Cell Mol Biol, 2018;58(6):706–16. 49. Laffranchi, M., Berardelli, R., Ronzoni, R., Lomas, D.A., and Fra, A. Heteropolymerization of alpha‐1‐antitrypsin mutants in cell models mimicking heterozygosity. Hum Mol Genet, 2018;27(10):1785–93. 50. Tafaleng, E.N., Chakraborty, S., Han, B. et al. Induced pluripotent stem cells model personalized variations in liver disease resulting from alpha1‐antitrypsin deficiency. Hepatology, 2015;62(1):147–57. 51. Perlmutter, D.H., Kay, R.M., Cole, F.S., Rossing, T.H., Van Thiel, D., and Colten, H.R. The cellular defect in alpha 1‐proteinase inhibitor (alpha 1‐PI) deficiency is expressed in human monocytes and in Xenopus oocytes injected with human liver mRNA. Proc Natl Acad Sci USA, 1985;82(20): 6918–21. 52. McCracken, A.A., Kruse, K.B., and Brown, J.L. Molecular basis for defective secretion of the Z variant of human alpha‐1‐proteinase inhibitor: secretion of variants having altered potential for salt bridge formation between amino acids 290 and 342. Mol Cell Biol, 1989;9(4):1406–14. 53. Lomas, D.A., Evans, D.L., Stone, S.R., Chang, W.S., and Carrell, R.W. Effect of the Z mutation on the physical and inhibitory properties of alpha 1‐antitrypsin. Biochemistry, 1993;32(2):500–8. 54. Lomas, D.A., Evans, D.L., Finch, J.T., and Carrell, R.W. The mechanism of Z alpha 1‐antitrypsin accumulation in the liver. Nature, 1992;357(6379):605–7. 55. Lomas, D.A., Finch, J.T., Seyama, K., Nukiwa, T., and Carrell, R.W. Alpha 1‐antitrypsin Siiyama (Ser53‐‐>Phe). Further evidence for intracellular loop‐ sheet polymerization. J Biol Chem, 1993;268(21):15333–5. 56. Dafforn, T.R., Mahadeva, R., Elliott, P.R., Sivasothy, P., and Lomas, D.A. A kinetic mechanism for the polymerization of alpha1‐antitrypsin. J Biol Chem, 1999;274(14):9548–55. 57. Yamasaki, M., Li, W., Johnson, D.J., and Huntington, J.A. Crystal structure of a stable dimer reveals the molecular basis of serpin polymerization. Nature, 2008;455(7217):1255–8.

50:  α1-ANTITRYPSIN DEFICIENCY 58. Whisstock, J.C., Silverman, G.A., Bird, P.I. et al. Serpins flex their muscle: I.I. Structural insights into target peptidase recognition, polymerization, and transport functions. J Biol Chem, 2010;285(32):24307–12. 59. Yamasaki, M., Sendall, T.J., Pearce, M.C., Whisstock, J.C., and Huntington, J.A. Molecular basis of alpha1‐antitrypsin deficiency revealed by the structure of a domain‐swapped trimer. EMBO Rep, 2011;12(10):1011–7. 60. Huang, X., Zheng, Y., Zhang, F. et al. Molecular mechanism of Z alpha1‐ antitrypsin deficiency. J Biol Chem, 2016;291(30):15674–86. 61. Hurtley, S.M. and Helenius, A. Protein oligomerization in the endoplasmic reticulum. Annu Rev Cell Biol, 1989;5:277–307. 62. Kim, J., Lee, K.N., Yi, G.S., and Yu, M.H. A thermostable mutation located at the hydrophobic core of alpha 1‐antitrypsin suppresses the folding defect of the Z‐type variant. J Biol Chem, 1995;270(15):8597–601. 63. Sidhar, S.K., Lomas, D.A., Carrell, R.W., and Foreman, R.C. Mutations which impede loop/sheet polymerization enhance the secretion of human alpha 1‐antitrypsin deficiency variants. J Biol Chem, 1995;270(15):8393–6. 64. Kang, H.A., Lee, K.N., and Yu, M.H. Folding and stability of the Z and S(iiyama) genetic variants of human alpha1‐antitrypsin. J Biol Chem, 1997;272(1):510–6. 65. Schmidt, B.Z. and Perlmutter, D.H. Grp78, Grp94, and Grp170 interact with alpha1‐antitrypsin mutants that are retained in the endoplasmic reticulum. Am J Physiol Gastrointest Liver Physiol, 2005;289(3):G444–55. 66. Lin, L., Schmidt, B., Teckman, J., and Perlmutter, D.H. A naturally occurring nonpolymerogenic mutant of alpha 1‐antitrypsin characterized by prolonged retention in the endoplasmic reticulum. J Biol Chem, 2001;276(36):33893–8. 67. Mallya, M., Phillips, R.L., Saldanha, S.A. et al. Small molecules block the polymerization of Z alpha1‐antitrypsin and increase the clearance of intracellular aggregates. J Med Chem, 2007;50(22):5357–63. 68. Yu, M.H., Lee, K.N., and Kim, J. The Z type variation of human alpha 1‐antitrypsin causes a protein folding defect. Nat Struct Biol, 1995;2(5):363–7. 69. Davis, R.L., Shrimpton, A.E., Holohan, P.D., et al. Familial dementia caused by polymerization of mutant neuroserpin. Nature, 1999;401(6751):376–9. 70. Nyfeler, B., Reiterer, V., Wendeler, M.W. et al. Identification of ERGIC‐53 as an intracellular transport receptor of alpha1‐antitrypsin. J Cell Biol, 2008;180(4):705–12. 71. Hidvegi, T., Ewing, M., Hale, P. et  al. An autophagy‐enhancing drug promotes degradation of mutant alpha1‐antitrypsin Z and reduces hepatic fibrosis. Science, 2010;329(5988):229–32. 72. Marcus, N.Y., Brunt, E.M., Blomenkamp, K. et al. Characteristics of hepatocellular carcinoma in a murine model of alpha‐1‐antitrypsin deficiency. Hepatol Res, 2010;40(6):641–53. 73. Teckman, J.H., An, J.K., Blomenkamp, K., Schmidt, B., and Perlmutter, D. Mitochondrial autophagy and injury in the liver in alpha 1‐antitrypsin deficiency. Am J Physiol Gastrointest Liver Physiol, 2004;286(5):G851–62. 74. Lindblad, D., Blomenkamp, K., and Teckman, J. Alpha‐1‐antitrypsin mutant Z protein content in individual hepatocytes correlates with cell death in a mouse model. Hepatology, 2007;46(4):1228–35. 75. Hidvegi, T., Mirnics, K., Hale, P., Ewing, M., Beckett, C., and Perlmutter, D.H. Regulator of G signaling 16 is a marker for the distinct endoplasmic reticulum stress state associated with aggregated mutant alpha1‐antitrypsin Z in the classical form of alpha1‐antitrypsin deficiency. J Biol Chem, 2007;282(38):27769–80. 76. Dooley, S., Hamzavi, J., Ciuclan, L. et  al. Hepatocyte‐specific Smad7 expression attenuates TGF‐beta‐mediated fibrogenesis and protects against liver damage. Gastroenterology, 2008;135(2):642–59. 77. Hidvegi, T., Schmidt, B.Z., Hale, P., and Perlmutter, D.H. Accumulation of mutant alpha1‐antitrypsin Z in the endoplasmic reticulum activates caspases‐4 and ‐12, NFkappaB, and BAP31 but not the unfolded protein response. J Biol Chem, 2005;280(47):39002–15. 78. Mukherjee, A., Hidvegi, T., Araya, P., Ewing, M., Stolz, D.B., and Perlmutter, D.H. NFkappaB mitigates the pathological effects of misfolded alpha1‐antitrypsin by activating autophagy and an integrated program of proteostasis mechanisms. Cell Death Differ, 2019;26(3)455–69. 79. Bridges, J.P., Wert, S.E., Nogee, L.M., and Weaver, T.E. Expression of a human surfactant protein C mutation associated with interstitial lung disease disrupts lung development in transgenic mice. J Biol Chem, 2003;278(52): 52739–46. 80. Young, L.R., Gulleman, P.M., Bridges, J.P. et  al. The alveolar epithelium determines susceptibility to lung fibrosis in Hermansky–Pudlak syndrome. Am J Respir Crit Care Med, 2012;186(10):1014–24. 81. Bhuiyan, M.S., Pattison, J.S., Osinska, H. et al. Enhanced autophagy ameliorates cardiac proteinopathy. J Clin Invest, 2013;123(12):5284–97.

657

  82. Doppler, K., Mittelbronn, M., Lindner, A., and Bornemann, A. Basement membrane remodelling and segmental fibrosis in sporadic inclusion body myositis. Neuromuscul Disord, 2009;19(6):406–11.   83. Nogalska, A., D’Agostino, C., Terracciano, C., Engel, W.K., and Askanas, V. Impaired autophagy in sporadic inclusion‐body myositis and in endoplasmic reticulum stress‐provoked cultured human muscle fibers. Am J Pathol, 2010;177(3):1377–87.   84. Hubner, R.H., Leopold, P.L., Kiuru, M., De, B.P., Krause, A., and Crystal, R.G. Dysfunctional glycogen storage in a mouse model of alpha1‐antitrypsin deficiency. Am J Respir Cell Mol Biol, 2009;40(2):239–47.   85. Piccolo, P., Annunziata, P., Soria, L.R. et al. Down‐regulation of hepatocyte nuclear factor‐4alpha and defective zonation in livers expressing mutant Z alpha1‐antitrypsin. Hepatology, 2017;66(1):124–35.  86. Wu, Y., Whitman, I., Molmenti, E., Moore, K., Hippenmeyer, P., and Perlmutter, D.H. A lag in intracellular degradation of mutant alpha 1‐antitrypsin correlates with the liver disease phenotype in homozygous PiZZ alpha 1‐antitrypsin deficiency. Proc Natl Acad Sci USA, 1994;91(19):9014–8.   87. Qu, D., Teckman, J.H., Omura, S., and Perlmutter, D.H. Degradation of a mutant secretory protein, alpha1‐antitrypsin Z., in the endoplasmic reticulum requires proteasome activity. J Biol Chem, 1996;271(37):22791–5.   88. Werner, E.D., Brodsky, J.L., and McCracken, A.A. Proteasome‐dependent endoplasmic reticulum‐associated protein degradation: an unconventional route to a familiar fate. Proc Natl Acad Sci USA, 1996;93(24):13797–801.   89. Kamimoto, T., Shoji, S., Hidvegi, T. et al. Intracellular inclusions containing mutant alpha1‐antitrypsin Z are propagated in the absence of autophagic activity. J Biol Chem, 2006;281(7):4467–76.   90. Kruse, K.B., Brodsky, J.L., and McCracken, A.A. Characterization of an ERAD gene as VPS30/ATG6 reveals two alternative and functionally distinct protein quality control pathways: one for soluble Z variant of human alpha‐1 proteinase inhibitor (A1PiZ) and another for aggregates of A1PiZ. Mol Biol Cell, 2006;17(1):203–12.   91. Kruse, K.B., Dear, A., Kaltenbrun, E.R. et  al. Mutant fibrinogen cleared from the endoplasmic reticulum via endoplasmic reticulum‐associated protein degradation and autophagy: an explanation for liver disease. Am J Pathol, 2006;168(4):1299–308; quiz 404–5.   92. Wang, Y. and Perlmutter, D.H. Targeting intracellular degradation pathways for treatment of liver disease caused by alpha1‐antitrypsin deficiency. Pediatr Res, 2014;75(1–2):133–9.   93. Gelling, C.L., Dawes, I.W., Perlmutter, D.H., Fisher, E.A., and Brodsky, J.L. The endosomal protein‐sorting receptor sortilin has a role in trafficking alpha‐1 antitrypsin. Genetics, 2012;192(3):889–903.   94. Long, O.S., Benson, J.A., Kwak, J.H. et al. A C. elegans model of human alpha1‐antitrypsin deficiency links components of the RNAi pathway to misfolded protein turnover. Hum Mol Genet, 2014;23(19):5109–22.   95. Liao, Y., Shikapwashya, O.N., Shteyer, E., Dieckgraefe, B.K., Hruz, P.W., and Rudnick, D.A. Delayed hepatocellular mitotic progression and impaired liver regeneration in early growth response‐1‐deficient mice. J Biol Chem, 2004;279(41):43107–16.   96. Hidvegi, T., Stolz, D.B., Alcorn, J.F. et al. Enhancing autophagy with drugs or lung‐directed gene therapy reverses the pathological effects of respiratory epithelial cell proteinopathy. J Biol Chem, 2015;290(50): 29742–57.   97. Pastore, N., Attanasio, S., Granese, B. et al. Activation of the c‐Jun N‐terminal kinase pathway aggravates proteotoxicity of hepatic mutant Z alpha1‐antitrypsin. Hepatology, 2017;65(6):1865–74.  98. Rudnick, D.A., Liao, Y., An, J.K., Muglia, L.J., Perlmutter, D.H., and Teckman, J.H. Analyses of hepatocellular proliferation in a mouse model of alpha‐1‐antitrypsin deficiency. Hepatology, 2004;39(4):1048–55.   99. Rudnick, D.A. and Perlmutter, D.H. Alpha‐1‐antitrypsin deficiency: a new paradigm for hepatocellular carcinoma in genetic liver disease. Hepatology, 2005;42(3):514–21. 100. Ding, J., Yannam, G.R., Roy‐Chowdhury, N. et  al. Spontaneous hepatic repopulation in transgenic mice expressing mutant human alpha1‐antitrypsin by wild‐type donor hepatocytes. J Clin Invest, 2011;121(5): 1930–4. 101. Pan, S., Huang, L., McPherson, J. et al. Single nucleotide polymorphism‐ mediated translational suppression of endoplasmic reticulum mannosidase I modifies the onset of end‐stage liver disease in alpha1‐antitrypsin deficiency. Hepatology, 2009;50(1):275–81. 102. Pan, S., Wang, S., Utama, B. et al. Golgi localization of ERManI defines spatial separation of the mammalian glycoprotein quality control system. Mol Biol Cell, 2011;22(16):2810–22.

658

THE LIVER:  REFERENCES

103. Iannotti, M.J., Figard, L., Sokac, A.M., and Sifers, R.N. A Golgi‐localized mannosidase (MAN1B1) plays a non‐enzymatic gatekeeper role in protein biosynthetic quality control. J Biol Chem, 2014;289(17):11844–58. 104. Chappell, S., Guetta‐Baranes, T., Hadzic, N., Stockley, R., and Kalsheker, N. Polymorphism in the endoplasmic reticulum mannosidase I (MAN1B1) gene is not associated with liver disease in individuals homozygous for the Z variant of the alpha1‐antitrypsin protease inhibitor (PiZZ individuals). Hepatology, 2009;50(4):1315;6. 105. Joly, P., Lachaux, A., Ruiz, M. et al. SERPINA1 and MAN1B1 polymorphisms are not linked to severe liver disease in a French cohort of alpha‐1 antitrypsin deficiency children. Liver Int, 2017;37(11):1608–11. 106. Chappell, S., Hadzic, N., Stockley, R., Guetta‐Baranes, T., Morgan, K., and Kalsheker, N. A polymorphism of the alpha1‐antitrypsin gene represents a risk factor for liver disease. Hepatology, 2008;47(1):127–32. 107. Janus, E.D., Phillips, N.T., and Carrell, R.W. Smoking, lung function, and alpha 1‐antitrypsin deficiency. Lancet, 1985;1(8421):152–4. 108. Kemmer, N., Kaiser, T., Zacharias, V., and Neff, G.W. Alpha‐1‐antitrypsin deficiency: outcomes after liver transplantation. Transplant Proc, 2008;40(5):1492–4. 109. Sarkar, S., Perlstein, E.O., Imarisio, S. et  al. Small molecules enhance autophagy and reduce toxicity in Huntington’s disease models. Nat Chem Biol, 2007;3(6):331–8. 110. Li, C., Xiao, P., Gray, S.J., Weinberg, M.S., and Samulski, R.J. Combination therapy utilizing shRNA knockdown and an optimized resistant transgene for rescue of diseases caused by misfolded proteins. Proc Natl Acad Sci USA, 2011;108(34):14258–63. 111. Mueller, C., Tang, Q., Gruntman, A. et  al. Sustained miRNA‐mediated knockdown of mutant AAT with simultaneous augmentation of wild‐type AAT has minimal effect on global liver miRNA profiles. Mol Ther, 2012;20(3):590–600. 112. Guo, S., Booten, S.L., Aghajan, M. et al. Antisense oligonucleotide treatment ameliorates alpha‐1 antitrypsin‐related liver disease in mice. J Clin Invest, 2014;124(1):251–61. 113. Pastore, N., Blomenkamp, K., Annunziata, F. et al. Gene transfer of master autophagy regulator TFEB results in clearance of toxic protein and correc-

tion of hepatic disease in alpha‐1‐anti‐trypsin deficiency. EMBO Mol Med, 2013;5(3):397–412. 114. Shen, S., Sanchez, M.E., Blomenkamp, K. et al. Amelioration of Alpha‐1 antitrypsin deficiency diseases with genome editing in transgenic mice. Hum Gene Ther, 2018;29(8):861–73. 115. Song, C.Q., Wang, D., Jiang, T. et  al. In vivo genome editing partially restores alpha1‐antitrypsin in a murine model of AAT deficiency. Hum Gene Ther, 2018;29(8):853–60. 116. Alam, S., Wang, J., Janciauskiene, S., and Mahadeva, R. Preventing and reversing the cellular consequences of Z alpha‐1 antitrypsin accumulation by targeting s4A. J Hepatol, 2012;57(1):116–24. 117. Burrows, J.A., Willis, L.K., and Perlmutter, D.H. Chemical chaperones mediate increased secretion of mutant alpha 1‐antitrypsin (alpha 1‐AT) Z: A potential pharmacological strategy for prevention of liver injury and emphysema in alpha 1‐AT deficiency. Proc Natl Acad Sci USA, 2000;97(4):1796–801. 118. Teckman, J.H. Lack of effect of oral 4‐phenylbutyrate on serum alpha‐1‐ antitrypsin in patients with alpha‐1‐antitrypsin deficiency: a preliminary study. J Pediatr Gastroenterol Nutr, 2004;39(1):34–7. 119. Bouchecareilh, M., Hutt, D.M., Szajner, P., Flotte, T.R., and Balch, W.E. Histone deacetylase inhibitor (HDACi) suberoylanilide hydroxamic acid (SAHA)‐mediated correction of alpha1‐antitrypsin deficiency. J Biol Chem, 2012;287(45):38265–78. 120. Fox, I.J., Chowdhury, J.R., Kaufman, S.S. et al. Treatment of the Crigler‐ Najjar syndrome type I with hepatocyte transplantation. N Engl J Med, 1998;338(20):1422–6. 121. Yusa, K., Rashid, S.T., Strick‐Marchand, H. et al. Targeted gene correction of alpha1‐antitrypsin deficiency in induced pluripotent stem cells. Nature, 2011;478(7369):391–4. 122. Perlmutter, D.H. and Silverman, G.A. Hepatic fibrosis and carcinogenesis in alpha1‐antitrypsin deficiency: a prototype for chronic tissue damage in gain‐of‐function disorders. Cold Spring Harb Perspect Biol, 2011;3(3). 123. Juhasz, G. and Neufeld, T.P. Autophagy: a forty‐year search for a missing membrane source. PLoS Biol, 2006;4(2):e36.

51

Pathophysiology of Portal Hypertension Yasuko Iwakiri and Roberto J. Groszmann Section of Digestive Diseases, Yale School of Medicine, New Haven, CT, USA

INTRODUCTION Portal hypertension refers to a pathological increase in blood pressure within the portal system due to obstruction of portal blood flow. The most frequent cause of portal hypertension is liver cirrhosis, and many of lethal complications of cirrhosis such as ascites and gastroesophageal variceal hemorrhage are related to portal hypertension. With less frequency, portal hypertension also occurs in noncirrhotic conditions, such as portal vein thrombosis, chronic schistosomiasis, heart failure, Budd– Chiari syndrome, and idiopathic portal hypertension. As a disorder of portal venous pressure, portal hypertension can be conceptualized using the hydraulic derivation of Ohm’s Law (pressure = flow × resistance) [1], composed of variables grounded in basic vascular biology. Portal hypertension develops in liver cirrhosis as a result of increased intrahepatic vascular resistance caused by multiple pathological events in the sinusoidal circulation, such as structural distortion by fibrosis, microvascular thrombosis, dysfunction of liver sinusoidal endothelial cells (LSECs), and hepatic stellate cell activation [2–5]. LSECs and pericyte‐like hepatic stellate cells are closely associated with one another in these pathological events in paracrine and autocrine manners. Portal hypertension then leads to splanchnic and systemic arterial vasodilation, which in turn contributes to increases in splanchnic blood flow to the portal system and thus further increases in portal pressure despite the formation of portosystemic collaterals [3–7]. This condition aggravates portal ­hypertension and facilitates the hyperdynamic circulation characterized by decreased mean arterial pressure, decreased systemic vascular resistance, increased cardiac index, and decreased

peripheral resistance. The hyperdynamic circulation along with increased blood flow to portosystemic collaterals results in ­clinically devastating complications such as gastroesophageal varices and variceal hemorrhage, hepatic encephalopathy, ascites, and renal failure due to the hepato‐renal syndrome [8–10] (Figure 51.1). This chapter provides current knowledge of the biology of portal hypertension with a focus on cellular and molecular events in different segments of vasculatures (intrahepatic versus extrahepatic circulations) that contribute to the development and perpetuation of portal hypertension and the subsequent development of hyperdynamic circulatory syndrome.

INTRAHEPATIC CIRCULATION An overview Hepatic sinusoids are specialized vascular beds that constitute the hepatic microcirculation. Blockage of sinusoids and the resulting increased hepatic vascular resistance to portal venous flow is the primary cause of portal hypertension. In cirrhosis, blockage of sinusoids is largely a result of massive structural changes in the liver associated with fibrosis/cirrhosis and intrahepatic vasoconstriction [5, 11, 12]. Phenotypic changes in hepatic cells, such as hepatic stellate cells and LSECs, are known to play pivotal roles in increased intrahepatic vascular resistance and have been studied extensively. This section addresses the cellular and molecular mechanisms underlying increased intrahepatic vascular resistance with a focus on hepatic stellate cells, LSECs, and thrombosis.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

660

THE LIVER:  INTRAHEPATIC CIRCULATION

Functional

Structural Varices

Increased resistance Porto-systemic collaterals

Regulation of hepatic stellate cell contraction

PORTAL HYPERTENSION

Vasodilatation

studies demonstrating that activated hepatic stellate cells acquire a myofibroblast‐like phenotype in response to liver injury [20] and exhibit a contractile phenotype [21]. Thus hepatic stellate cells, through perivascular contraction in the sinusoidal microcirculation, are thought to be key contributors to the dynamic and reversible component of portal hypertension in cirrhosis.

Increased flow

Hypotension Central hypovolemia

Increase in CO

Activation NE, VP, A-II

Na and water retention increase venous return to the heart

Figure 51.1  Summary of the pathophysiology of portal hypertension. The increase in hepatic resistance leads to an increase in portal pressure. This leads to a cascade of disturbances in the splanchnic and systemic circulation characterized by vasodilation, sodium and water retention, and plasma volume expansion, which are major players in the pathogenesis of ascites and hepatorenal syndrome. Additionally, these alterations lead to an increase in portal blood inflow that contributes to maintaining and aggravating portal hypertension. Another characteristic feature is the development of portosystemic collaterals, which are responsible for complications such as variceal bleeding and hepatic encephalopathy. CO, cardiac output; NE, norepinephrine; VP, vasopressin; A‐II, angiotensin II; Na, sodium.

Hepatic stellate cell biology

Endothelin signaling is a key pathway that regulates a contractile phenotype of hepatic stellate cells. Endothelin binds to G‐protein coupled receptors, endothelin A (ETA) and endothelin B (ETB), which are generally found on vascular smooth muscle cells and endothelial cells, respectively. Endothelin‐1 (ET‐1) is the major subtype found in liver disease and is preferentially bound to ETA more than the other two subtypes (ET‐2 and ET‐3) [22]. ET‐1 protein levels are elevated in liver injury along with ET‐1 mRNA [23]. While endothelial cells produce the majority of ET‐1 in normal liver, liver injury shifts this production primarily to hepatic stellate cells [24], which also profoundly upregulate ETA and ETB receptors [25, 26], suggesting increased sensitivity to this signal. ET‐1 has been shown to induce contraction of the sinusoidal vasculature in an experimental model [27], and antagonism of ETA has been shown to reduce portal pressure in an animal model of cirrhosis [28]. Smooth muscle cell or hepatic stellate cell contraction is a result of myosin light chain (MLC). A study on the mechanism of ET‐1‐mediated hepatic stellate cell contraction has revealed that ET‐1 leads to MLC phosphorylation through multiple pathways [29]. One is a Ca2+‐dependent pathway in which ET‐1 causes a transient increase in intracellular Ca2+, leading to MLC kinase activation and MLC phosphorylation. Others include activation of the protein kinase C and Rho‐kinase pathways, both leading to MLC phosphorylation.

Fibrosis and architectural alterations

Liver sinusoidal endothelial cell biology

Hepatic fibrosis is thought to be the primary factor that contributes to increased intrahepatic vascular resistance in cirrhotic liver. In a classic study, Bhathal and Grossman demonstrated through vasodilator challenges in the isolated perfused cirrhotic rat liver that 80% of the increased intrahepatic resistance to portal venous flow is attributable to structural changes, while the remaining 20% is due to a reversible, hypercontractile phenotype of the hepatic microcirculation [13]. The key pathway of fibrosis in the liver is proinflammatory signaling that causes activation of hepatic stellate cells and thereby leads to extracellular matrix deposition. Understanding this pathway has led in turn to an understanding of its reversal and the regression of fibrosis, which holds a great therapeutic potential for chronic liver disease and portal hypertension [14, 15].

Fenestration and capillarization

Hepatic stellate cells as pericytes Hepatic stellate cells are thought to play an additional role in portal hypertension beyond fibrosis. Hepatic stellate cells are located in the space of Disse, directly underneath LSECs, and work as hepatic pericytes, perivascular non‐endothelial cells with a variety of functions including regulation of vascular tone through smooth muscle‐like contractility and regulation of endothelial proliferation [16–19]. These notions arise from

LSECs have a distinct phenotype from endothelial cells elsewhere in the liver, as well as elsewhere in the body. Their most distinguishing feature is fenestration with fenestrae being approximately 0.1 microns in size and organized into groups of sieve plates. It is thought that fenestrae facilitate the transport of macromolecules from hepatic sinusoids to the space of Disse, where they can interact with hepatic stellate cells and hepatocytes. Another feature of LSECs distinct from endothelial cells in other organs is the lack of a basement membrane, which allows maximum permeability between the lumen of the sinusoid and the space of Disse [30]. LSECs lose their fenestrae, develop a basement membrane, and become “capillarized” as a consequence of liver fibrosis [31, 32]. Vascular endothelial growth factor (VEGF) is known as a key factor for the maintenance of endothelial fenestrae [33]. Inhibition of VEGF signaling in a transgenic animal model, in which liver‐ specific secretion of a soluble VEGF decoy receptor sequesters endogenous VEGF, caused loss of LSEC fenestration and resulted in portal hypertension and hepatic stellate cell activation independent of hepatic parenchymal damage, with reversal of portal hypertension with restoration of VEGF [34]. This VEGF action on LSEC phenotype is nitric oxide (NO)‐dependent. Accordingly,



51:  Pathophysiology of Portal Hypertension

an NO synthase inhibitor, Nω‐Nitro‐L‐arginine methyl ester hydrochloride (L‐NAME), can lead to the loss of the LSEC phenotype [35]. Besides VEGF, various factors have been shown to alter LSEC fenestration. Collagens in the space of Disse may play a role in maintenance or loss of LSEC fenestration [36]. A role of lipid rafts in regulation of fenestration has also been ­investigated. Using super‐resolution fluorescence microscopy, researchers showed an inverse relationship between areas of lipid rafts and areas of membranes with fenestration in LSECs. They also demonstrated that inhibiting lipid raft formation via 7‐ketocholesterol or actin disruption increased fenestration, and conversely that increasing lipid raft formation with a low concentration of Triton X‐100 decreased fenestration [37].

Crosstalk between LSECs and hepatic stellate cells While phenotypic changes of LSEC is known as a consequence of liver fibrosis/cirrhosis as mentioned above [31, 32], it is also thought that loss of the LSEC phenotype can be permissive for hepatic stellate cell activation [38] and that the communication between LSECs and hepatic stellate cells plays a key role in the pathogenesis of portal hypertension (Figure 51.2) [39]. Research into LSEC and hepatic stellate cell communication demonstrated that an isoform of fibronectin produced by LSECs in a bile duct ligation model of liver damage was able to activate hepatic stellate cells [40], though a subsequent study revealed it to be a factor important for hepatic stellate cell motility but not differentiation to a myofibroblast phenotype [41]. Another study has shown that LSECs communicate with hepatic stellate cells via exosomes containing sphingosine kinase‐1 (SK1) and its product sphingosine‐1 phosphate, providing a signal for hepatic

661

stellate cell migration [42], which is closely connected to their activated phenotype. Nitric oxide produced by LSECs not only regulates vascular tone in the liver, but also plays a key role in the crosstalk between LSECs and hepatic stellate cells. LSECs are able to induce hepatic stellate cell reversion from activation to quiescence via an NO‐dependent mechanism [43], possibly paracrine signaling via the Kruppel‐like factor 2 (KLF2)‐NO‐guanylate cyclase pathway in LSECs [44]. Another study showed that NO donors were able to inhibit proliferation and chemotaxis of activated hepatic stellate cells in response to platelet‐derived growth factor by disrupting its intracellular signaling pathway in prostaglandin E2‐mediated manners [45]. Further, NO inhibited hepatic stellate cell migration via cGMP‐dependent protein kinase (PKG)‐mediated inhibition of the Rac1 pathway [46, 47]. NO signaling has also been shown to induce hepatic stellate cell apoptosis [48] via a caspase‐independent mechanism possibly related to increased mitochondrial oxidative stress and increased mitochondrial membrane permeability, along with a possible lysosomal stress component. These observations may have relevance in diseases such as alcoholic liver injury and non‐alcoholic fatty liver disease, in which loss of the normal LSEC phenotype has been shown to occur before fibrosis [38].

Endothelial dysfunction NO is a critical regulator of normal hepatic vascular tone and portal pressure [49]. The sources of NO in the hepatic vasculature are LSECs and endothelial cells of blood vessels. These cells constitutively express endothelial nitric oxide synthase (eNOS) and produce a low level of NO. Production of NO increases in response to an increase in blood flow via shear

Figure 51.2  LSEC/HSC crosstalk. (a) Injured liver sinusoidal endothelial cells (LSECs) lead to activation of hepatic stellate cells (HSCs). Liver injury leads LSECs to produce the EIIIA isoform of fibronectin, which signals to HSCs through integrin α9β1 to promote motility, which is important for their activated phenotype. LSECs also signal to promote HSC motility through sphingosine kinase 1 – sphingosine‐1‐phosphate (SK1‐S1P)‐ containing exosomes, which adhere to HSCs via fibronectin binding to an integrin receptor. Nitric oxide (NO) production by sinusoidal endothelial cells is important in maintaining HSC quiescence, with a Kruppel‐like factor (KLF) 2 pathway enhancing NO and guanylate cyclase production. (b) NO production from LSECs also inhibiting the Rac/Rho pathway in HSCs. NO also causes HSC apoptosis, and may thereby limit the number of activated HSCs in the liver. Vit A: Vitamin A. Modified from McConnell and Iwakiri [39] and reproduced with permission of Springer Nature. HSC are in green; healthy LSEC are in blue; injured LSEC are in red.

662

THE LIVER:  EXTRAHEPATIC CIRCULATION

stress [50] and can also be increased by VEGF [51]. LSECs in cirrhotic liver express eNOS similarly to LSECs in normal liver. However, eNOS activity is lower under pathologic conditions and NO release is diminished in disease [52]. In addition, LSECs in cirrhosis have a reduced ability to respond to increases in blood flow compared to the healthy state [53]. These adverse changes in LSECs result in decreased production of NO and impaired vasodilation in the hepatic microcirculation in cirrhosis and are important contributors to increased intrahepatic vascular resistance observed in portal hypertension. The regulation of eNOS function involves multiple regulators acting in concert with stimulatory signaling, including its phosphorylation by the protein kinase Akt [54], which is facilitated by G‐protein coupled‐receptor kinase interactor‐1 (GIT1) [55, 56]. The function of eNOS may also be inhibited by binding to caveolin‐1, which can be disrupted by calmodulin [57]. Because endothelial dysfunction leads to increased vascular resistance in the sinusoidal microcirculation and promotes activation of hepatic stellate cells, a pharmacological approach that reverses the dysfunctional LSEC phenotype could be an effective therapeutic strategy. An example is the use of statins. Studies have shown that statins ameliorate portal hypertension in cirrhotic patients [58] as well as experimental models of portal hypertension [59, 60]. These studies demonstrated that statins improve endothelial dysfunction and increase NO bioavailability in the sinusoidal microcirculation. Several potential mechanisms for this increased NO bioavailability have been proposed. One is the ability of statins to inhibit synthesis of isoprenoids, which are critical for membrane anchoring and activation of small GTPases, such as RhoA. Given that RhoA/ Rho‐kinase signaling could decrease eNOS activity [61] and expression [62], statins, by decreasing RhoA activity, could improve eNOS function, thereby enhance NO bioavailability and decrease intrahepatic vascular resistance [60]. Another potential mechanism is that statins increase activity of Akt/protein kinase B, which phosphorylates and activates eNOS, thereby increasing NO bioavailability [59]. In addition to improving endothelial cell dysfunction, statins could target the RhoA/Rho‐kinase pathway in pericytes (e.g. activated hepatic stellate cells) and decrease their contractility, thereby lowering intrahepatic vascular resistance and ameliorating portal hypertension [60].

Pathological angiogenesis Angiogenesis, or the process of new blood vessel formation from pre‐existing vascular beds, has also been implicated in portal hypertension. Hepatic angiogenesis is thought to generate irregular intrahepatic circulatory routes and thus could increase intrahepatic vascular resistance. The Notch1 signaling pathway is known to be important for embryonic vascular development and postnatal vascular remodeling. In the postnatal liver, inhibition of the Notch1 pathway leads to nodular regenerative hyperplasia, a common etiology of noncirrhotic portal hypertension. Notch1 deletion in mice resulted in an abnormal sinusoidal microcirculation, as indicated by de‐differentiation of LSECs, pathological remodeling of the hepatic sinusoidal microvasculature, intussusceptive angiogenesis (also known as splitting angiogenesis), and  dysregulation of ephrinB2/EphB4 and endothelial tyrosine

kinase (i.e. impairment of arterial versus venous specification). Interestingly, animals with lack of Notch1 gene developed portal hypertension even before the onset of nodular regenerative hyperplasia, and this was thought to result from intrahepatic vascular disorders, possibly due to intussusceptive angiogenesis in the liver microcirculation [63]. Angiogenesis has also been shown to be associated with fibrosis progression in the liver [64, 65]. However, this relationship is complex and could be causative or correlative due to involvement of hypoxia, a strong inducer of angiogenesis. Modulation of angiogenesis does not generally provide a predictable effect on liver fibrosis, either [66].

Microvascular thrombosis/platelet activation The study of intrahepatic portal hypertension is evolving to include platelet activation and thrombosis as crucial factors for its pathophysiology. Ian Wanless and others were instrumental contributors in this area, observing what they termed “parenchymal extinction” accounting for fibrosis progression due to intrahepatic vascular thrombosis [67]. Further, while cirrhosis had previously been thought to have a bleeding trend, more advanced physiologic tests to assess coagulation status [68] and systematic studies of bleeding complications [69] have led to an important consensus that hemostatic status in cirrhosis is rebalanced or perhaps even prothrombotic. Platelets are a key player of the typical physiology of vascular thrombosis. Initially, cirrhotic platelets were thought to be dysfunctional and predispose patients to a bleeding trend [70, 71]. More recently, however, it has been found that activity of cirrhotic platelets in hemostasis and thrombosis is potentially preserved [72] or even increased [73], although conflicting data also exist [74]. Generally, the function of platelets in various types of liver injury is quite complex with multiple stage‐specific factors influencing the role of platelets as profibrotic or antifibrotic [75]. Further experiments and clinical trials will continue to expand our knowledge regarding the role of thrombosis and platelets in sinusoidal portal hypertension.

EXTRAHEPATIC CIRCULATION An overview In addition to the liver vasculature, the mesenteric vasculature plays a key role in portal hypertension. Foundational knowledge of the pathophysiology of this vascular bed comes from seminal studies by Groszmann and others [76–96]. These studies demonstrated that in portal hypertension, even with increased hepatic vascular resistance, the splanchnic circulation is hyperdynamic. Splanchnic arterial vasodilation is a key feature of the hyperdynamic circulation, as it can perpetuate increased blood inflow to the portal system and thus exacerbate portal hypertension [4, 5]. Arterial vasodilation is attributable to abnormal cell function in different layers of the vasculature, such as endothelial cells, smooth muscle cells, and the adventitial layer that contains vascular progenitor cells and neuronal termini. This section discusses the mechanisms of arterial vasodilation and collateral vessel formation in the splanchnic and systemic circulations in cirrhosis with portal hypertension.



51:  Pathophysiology of Portal Hypertension

Arterial vasodilation in the splanchnic and systemic circulations

vascular relaxation. The candidates of EDHF include ­arachidonic acid metabolites, epoxyeicosatrienoic acid (EET), potassium ions (K+), components of gap junctions, and hydrogen peroxide [5]. Vasoactive molecules known to be involved in the regulation of vascular tone in cirrhosis are summarized in Figure 51.3. An increase in portal pressure triggers eNOS activation and subsequent NO overproduction in the extrahepatic circulation. Changes in portal pressure are detected at different vascular beds depending on the severity of portal hypertension [51]. An increment increase in portal pressure is sensed first by the intestinal microcirculation and increases VEGF production with a subsequent increase in eNOS levels in the intestinal microcirculation. When portal pressure further increases and reaches a ­certain level, arterial vasodilation develops in the splanchnic circulation (i.e. the mesenteric arteries). In contrast to arterial vasodilation in the intestinal microcirculation where VEGF

Arterial vasodilation NO is probably the most important vasodilator molecule that contributes to excessive vasodilation observed in arteries of the splanchnic and systemic circulations in portal hypertension. Experimental models of portal hypertension with or without cirrhosis have also shown that other vasodilator molecules, such as carbon monoxide (CO), prostacyclin (PGI2), endocannabinoids, and endothelium‐derived hyperpolarizing factor (EDHF), are also induced [5, 7, 97]. The induction of arterial vasodilation despite inhibition of NO, CO, and PGI2 has indicated the presence of other endothelium‐derived vasodilator molecule(s), known as EDHF [98]. As its name implies, the action of EDHF is to hyperpolarize vascular smooth muscle cells, causing

(a)

663

(b)

(c) Cirrhosis

Cirrhosis

Intrahepatic circulation

Splanchnic and Systemic Circulation

Normal

Sinusoidal Endothelial Cells

Endothelial Cells

Agonists CO

TNFα P1

-1 ET

3

BR

ET xx

x

Akt

Ak

t

P

3

GRK2

Shear Stress

Ak

t

ET

eNOS CaM

Hsp90

GTP Cyclooxygenase

1

BR

BH4

P

P1

-1 ET

1

vCa

BH4

eNOS CaM

Hsp90

vCa

IP3 AK1

eNOS

PO4 Hsp90

COX-1

cGMP

BH4

Ca2+

eNOSCa2+

HO-1

COX

CO

PGI2

COX-1

NO sGC

HbCO

Hb

Adrenomedullin VEGF

NO

NO

ETAR sGC cGMP

NO TXA2

Stellate Cells

sGC

AC

cGMP cAMP Smoooth Muscle Cells

Vasoconstriction

TXA2 NO Endothelial hypoactivation

Vasodilatation NO CO

PGI2 Endothelial hyperactivation

Figure 51.3  Vasoactive molecules known to be involved in the regulation of vascular tone in cirrhosis. In the arterial splanchnic and systemic circulation (right panel), agonists such as adrenomedullin, vascular endothelial growth factor (VEGF), and tumor necrosis factor alpha (TNFα) or physical stimuli such as shear stress stimulate Akt, which directly phosphorylates and activates endothelial nitric oxide synthase (eNOS). eNOS requires cofactors such as tetrahydrobiopterin (BH4) for its activity. Heat shock protein 90 (Hsp90) is one of the positive regulators of eNOS. Like NO, carbon monoxide (CO) produced by hemeoxygenase‐1 (HO‐1) causes vasodilation by activating soluble guanylate cyclase (sGC) to generate cyclic guanosine monophosphate (cGMP) in vascular smooth muscle cells. Prostacyclin (PGI2) is synthesized by cyclooxygenase (COX) and elicits smooth muscle relaxation by stimulating adenylate cyclase (AC) and generation of cyclic adenosine monophosphate (cAMP). In the intrahepatic circulation in cirrhosis (left panel), decreased NO and increased thromboxane A2 (TXA2) production in SECs results in a net reduction of vasorelaxation in the intrahepatic circulation. Endothelin‐1 (ET‐1) has dual vasoactive effects, mediating vasoconstriction through binding to endothelin A (ETA) receptors located on HSCs and causing HSC contraction. Binding of ET‐1 to ETB receptor (ETBR) mediates vasodilation through Akt phosphorylation and eNOS phosphorylation in normal liver. In cirrhosis, G‐protein‐coupled receptor kinase‐2 (GRK2), an inhibitor of G protein‐coupled receptor signaling, is upregulated in SECs, leading to the impairment of Akt phosphorylation and a reduction in NO production. An increased production of COX‐1‐derived vasoconstrictor prostanoid TXA2 is also an example of endothelial dysfunction in cirrhosis. Modified from Iwakiri and Groszmann [7] and reproduced with permission of Elsevier.

664

THE LIVER:  HYPERDYNAMIC CIRCULATORY SYNDROME

mediates eNOS upregulation, it is thought that mechanical forces including cyclic strains and shear stress induce eNOS activation and lead to NO overproduction in the splanchnic circulation [51, 92, 93, 99, 100]. Arterial vasodilation in the splanchnic circulation facilitates the systemic circulation to be also hyperdynamic.

anti‐VEGF (rapamycin)/anti‐PDGF (Gleevec) [120], anti‐PlGF [119], apelin antagonist [121], sorafenib [122, 123], and a cannabinoid receptor 2 agonist [124]. However, the reduction of these collaterals does not necessarily decrease portal pressure because it does not substantially change the blood flow to the portal vein. Therefore, the concomitant mitigation of arterial vasodilation is needed to reduce portal pressure.

Hypocontractility A decrease in contractility to vasoconstrictors is also typical of arteries of the splanchnic and systemic circulations in portal hypertension. This hypocontractility occurs largely due to the presence of excessive vasodilator molecules (e.g. NO) in the endothelium, but is to some degree attributable to a decrease in several vasoconstrictive molecules produced in smooth muscle cells and neurons. Those molecules include neuropeptide Y [101], urotensin II [102, 103], angiotensin [104], and bradykinin [105, 106]. In arteries of the splanchnic circulation, it has been shown that vasodilators increase and vasoconstrictors decrease.

Neural factors Neural factors are also thought to be involved in the dysfunction of vascular tone in portal hypertension, especially through the sympathetic nervous system [101, 107, 108]. It is reported that sympathetic nerve atrophy/regression observed in mesenteric arterial beds of portal hypertensive rats leads to vasodilation and/or hypocontractility of those arterial beds [109, 110]. The role of neural factors in decreased contractile responses remains to be fully elucidated.

Structural changes of arteries The thinning of arterial walls is observed in the splanchnic and systemic circulations of rats with cirrhotic livers [111, 112]. While this arterial wall thinning can be a consequence of the hyperdynamic circulation, it may also sustain arterial vasodilation and exacerbate portal hypertension [5, 6]. While NO plays a role at least in part, the molecular mechanisms responsible for arterial wall thinning remain to be understood.

Collateral vessel formation Portosystemic collateral vessels develop in response to an increase in portal pressure. These collateral vessels develop in an attempt to decompress the hypertensive portal system and are formed through the opening of pre‐existing vessels or angiogenesis [113, 114]. However, these collateral vessels are also known to cause serious complications, including variceal bleeding and hepatic encephalopathy [5, 115, 116]. A change in portal pressure is thought to be detected first by the intestinal microcirculatory bed, followed by arteries of the splanchnic circulation [51]. These vascular beds subsequently generate various angiogenic factors, such as VEGF [96, 117, 118] and placental growth factor (PlGF) [119], which promote the formation of portosystemic collaterals. Experimental studies of portal hypertension and cirrhosis have shown that portosystemic collaterals are reduced by 18 to 78% with treatment by anti‐VEGFR2 [100], a combination of

HYPERDYNAMIC CIRCULATORY SYNDROME An overview Excessive arterial vasodilation in the splanchnic and systemic circulations in portal hypertension results in the development of hyperdynamic circulatory syndrome. This syndrome is characterized by increased cardiac index, decreased systemic vascular resistance, and decreased mean arterial pressure and eventually leads to multiple organ failure frequently observed in chronic liver disease. This section discusses multiple organ failure observed in the hyperdynamic circulatory state (Figure 51.4).

The systemic circulation and the heart Although arterial vasodilation in the splanchnic circulation is an essential initiating factor, hyperdynamic circulation does not occur without expansion of plasma volume and the development of portosystemic collaterals [125, 126]. In cirrhotic patients, blood and plasma volumes are elevated, but are not evenly distributed among vascular areas [127]. For example, arterial blood volumes in the heart, lungs, and central arterial tree are decreased compared to those in the splanchnic circulation, resulting in central hypovolemia. Decreased blood volumes together with arterial hypotension lead to baroreceptor activation of potent vasoconstriction systems such as the sympathetic nervous system and the renin angiotensin aldosterone system [128], resulting in water retention and plasma volume expansion. The combination of an expanded plasma volume and a reduction in peripheral vascular resistance leads to an increase in cardiac output. When portal hypertension persists, the heart results in a high cardiac output syndrome: initial compensation occurs according to the degree of individual cardiac output, followed by some degree of cardiac insufficiency. The cardiac index is usually higher than normal (greater than 4 L min−1 m2) but insufficient to maintain arterial pressure in the face of progressive arterial vasodilation [5]. Importantly, high cardiac output failure is reversible once the initial cause leading to the high cardiac output is treated. This reversal has also been observed in patients with cirrhosis after liver transplantation [129, 130].

The hyperdynamic splanchnic circulation As mentioned previously, the hyperdynamic splanchnic circulation is central to the development of the syndrome. Although it  is commonly recognized as a complication of cirrhosis, it should be better conceptualized as a complication of portal



51:  Pathophysiology of Portal Hypertension

Splanchnic Circulation

Renal Circulation

Systemic Circulation

Pulmonary Circulation

665

Brain Circulation

* Acute liver failure Vasodilation

Vasodilation

Vasoconstriction

Initiated by local factors

Vasodilation

Vasodilation

Initiated by local factors

Initiated by local factors

• ↑Blood flow

• Na/H2O retention

• ↑Cardiac output

• Arterial hypoxemia

• ↑Portal pressure

• Hepatorenal syndrome

• High output

• Hepato-pulmonary

• Heart failure

• Brain edema

syndrome

• ↓Peripheral O2 utilization

Figure 51.4  Vasodilation: the source of all evils. *Chronic encephalopathy is associated with a reduced brain blood flow. The mechanism is probably similar to what is observed in the renal circulation. Modified from Iwakiri and Groszmann [5] and reproduced with permission of John Wiley & Sons.

hypertension. The term “portal venous inflow” is often used for splanchnic blood flow entering into the portal system to distinguish it from portal blood flow entering into the liver [77]. Portal hypertension is the only known pathophysiological situation in which portal blood flow entering into the portal system (portal venous inflow) is different from portal blood flow entering into the liver. Portal venous inflow entering into the portal system significantly increases, while portal blood flow entering into the liver decreases, because portal blood escapes into portosystemic collaterals formed as a result of portal hypertension [5]. Arterial vasodilation in the splanchnic vascular bed is thought to contribute to this increased portal venous inflow entering into the portal system and also helps to compensate for the blood that should escape into portosystemic collaterals. Thus, the most accepted concept is that arterial vasodilation starts from the splanchnic vascular bed (i.e. intestinal microcirculation) and then proceeds to the whole splanchnic circulation) and is the key factor that leads to the development of the hyperdynamic circulatory syndrome. The increase in portal pressure itself triggers arterial vasodilation in the splanchnic vascular bed [51, 93].

The hyperdynamic pulmonary circulation The hyperdynamic circulation also affects the lungs. The pulmonary vasodilation is associated with the hepatopulmonary syndrome, one of the most severe complications of chronic liver disease (Figure  51.4). Although the intrinsic mechanism that

triggers this syndrome into its full expression is not fully known, local vasodilation mediated by several endothelial vasodilators, including NO and CO, plays an important role [131]. Local factors in the pulmonary circulation may determine why only some patients develop hepatopulmonary syndrome. High cardiac output may also contribute to the severity of hepatopulmonary syndrome by increasing shear stress in the pulmonary vascular endothelium as well as by shortening a pulmonary and tissue transit time of red blood cells [132, 133].

The renal circulation The hyperdynamic circulatory state indirectly influences the renal circulatory bed (Figure 51.4). It is thought that splanchnic arterial vasodilation in patients with portal hypertension results in redistribution of the blood volume, leading to a reduction in central blood volumes (i.e. a relative hypovolemic state). The kidney responds to this hypovolemic state by renal arterial vasoconstriction, a reduction in glomerular filtration, and retention of sodium and water. The central hypovolemic state activates signals that induce vasoconstrictive and volume retaining neurohumoral conditions, thereby keeping the sodium and water retentive state [134–136]. In patients with compensated cirrhosis, progressive systemic arterial vasodilation leads to an elevation of intravascular volume and cardiac output, so that arterial perfusion pressure can be maintained. As disease conditions progress, cardiac output continues to increase in response to progressive arterial vasodilation. Eventually, cardiac response

666

THE LIVER:  REFERENCES

starts to fail to maintain perfusion pressure, renal blood flow drops, and renal failure develops [137, 138]. This phenomenon is known as hepatorenal syndrome. Treatment of arterial vasodilation improves renal function [139, 140].

The cerebral circulation The effects of hyperdynamic circulatory syndrome on the cerebral circulation are probably the most difficult to define (Figure 51.4). Both an increase and decrease in cerebral blood flow have been described in acute and chronic liver diseases, respectively [141]. An increase in cerebral blood flow has been mainly associated with acute liver failure, which could potentially lead to the development of brain edema [142]. In contrast, cerebral blood flow is decreased in chronic liver disease, and this decrease runs in parallel with the above‐mentioned reduction in renal blood flow, suggesting that the mechanisms of blood flow reductions in chronic liver disease may be similar in these two organs [143].

CONCLUSION The pathogenesis of portal hypertension is complex, because portal hypertension involves not only the hepatic circulation, but also the splanchnic and systemic circulations of the hyperdynamic circulatory state. Because of the disparate conditions of vascular tone in the intrahepatic and extrahepatic circulations (i.e. vasoconstriction in the intrahepatic circulation versus vasodilation in the extrahepatic circulation), the organ/tissue or cell‐ specific modulation of vasodilator or vasoconstrictor molecules is of paramount importance for therapeutic purposes.

REFERENCES 1. Iwakiri, Y., Shah, V., and Rockey, D.C. Vascular pathobiology in chronic liver disease and cirrhosis – current status and future directions. J Hepatol, 2014;61(4):912–24. 2. Bosch, J., Groszmann, R.J., and Shah, V.H. Evolution in the understanding of the pathophysiological basis of portal hypertension: How changes in paradigm are leading to successful new treatments. J Hepatol, 2015;62(1 Suppl):S121–30. 3. Iwakiri, Y. The molecules: mechanisms of arterial vasodilatation observed in the splanchnic and systemic circulation in portal hypertension. J Clin Gastroenterol, 2007;41(3):S288–94. 4. Iwakiri, Y. Pathophysiology of portal hypertension. Clin Liver Dis, 2014;18(2):281–91. 5. Iwakiri, Y. and Groszmann, R.J. The hyperdynamic circulation of chronic liver diseases: from the patient to the molecule. Hepatology, 2006;43(2 Suppl 1):S121–31. 6. Iwakiri, Y. Endothelial dysfunction in the regulation of cirrhosis and portal hypertension. Liver Int, 2012;32(2):199–213. 7. Iwakiri, Y. and Groszmann, R.J. Vascular endothelial dysfunction in cirrhosis. J Hepatol, 2007;46(5):927–34. 8. Tetangco, E.P., Silva, R., and Lerma, E. Portal hypertension: etiology, evaluation, and management. Dis Mon, 2016;62(12):411–26. 9. Cavallin, M., Fasolato, S., Marenco, S. et al. The Treatment of Hepatorenal Syndrome. Dig Dis, 2015;33(4):548–54. 10. Garcia‐Tsao, G., Abraldes, J.G., Berzigotti, A. et  al. Portal hypertensive bleeding in cirrhosis: Risk stratification, diagnosis, and management: 2016

practice guidance by the American Association for the study of liver diseases. Hepatology, 2017;65(1):310–35. 11. Rockey, D.C. Cell and molecular mechanisms of increased intrahepatic resistance and hemodynamic correlates. Sanyal AJ SV, editor. Totowa, NJ: Humana Press Inc; 2005. 12. Pinzani, M., Vizzutti, F. Anatomy and vascular biology of the cells in the portal circulation. Sanyal AJ SV, editor. Totowa, NJ: Humana Press Inc; 2005. 13. Bhathal, P.S. and Grossman, H.J. Reduction of the increased portal vascular resistance of the isolated perfused cirrhotic rat liver by vasodilators. J Hepatol, 1985;1(4):325–37. 14. Lee, Y.A., Wallace, M.C., and Friedman, S.L. Pathobiology of liver fibrosis: a translational success story. Gut, 2015;64(5):830–41. 15. Trautwein, C., Friedman, S.L., Schuppan, D. et al. Hepatic fibrosis: Concept to treatment. J Hepatol, 2015;62(1 Suppl):S15–24. 16. Shepro, D. and Morel, N.M. Pericyte physiology. FASEB, 1993;7(11): 1031–8. 17. Bergers, G. and Song, S. The role of pericytes in blood‐vessel formation and maintenance. Neuro Oncol, 2005;7(4):452–64. 18. Franco, M., Roswall, P., Cortez, E. et al. Pericytes promote endothelial cell survival through induction of autocrine VEGF‐A signaling and Bcl‐w expression. Blood, 2011;118(10):2906–17. 19. LaBarbera, K.E., Hyldahl, R.D., O’Fallon, K.S. et al. Pericyte NF‐kappaB activation enhances endothelial cell proliferation and proangiogenic cytokine secretion in vitro. Physiol Rep, 2015;3(4). 20. Rockey, D.C., Boyles, J.K., Gabbiani, G. et al. Rat hepatic lipocytes express smooth muscle actin upon activation in vivo and in culture. J Submicrosc Cytol Pathol, 1992;24(2):193–203. 21. Rockey, D.C., Housset, C.N., and Friedman, S.L. Activation‐dependent contractility of rat hepatic lipocytes in culture and in vivo. J Clin Invest, 1993;92(4):1795–804. 22. Rockey, D.C. Vascular mediators in the injured liver. Hepatology, 2003;37(1):4–12. 23. Rockey, D.C., Fouassier, L., Chung, J.J. et  al. Cellular localization of endothelin‐1 and increased production in liver injury in the rat: potential for autocrine and paracrine effects on stellate cells. Hepatology, 1998;27(2): 472–80. 24. Shao, R., Yan, W., and Rockey, D.C. Regulation of endothelin‐1 synthesis by endothelin‐converting enzyme‐1 during wound healing. J Biol Chem, 1999;274(5):3228–34. 25. Rothermund, L., Leggewie, S., Schwarz, A. et al. Regulation of the hepatic endothelin system in advanced biliary fibrosis in rats. Clin Chem Lab Med, 2000;38(6):507–12. 26. Yokomori, H., Oda, M., Ogi, M. et al. Enhanced expression of endothelin receptor subtypes in cirrhotic rat liver. Liver, 2001;21(2):114–22. 27. Zhang, J.X., Pegoli, W., Jr., and Clemens, M.G. Endothelin‐1 induces direct  constriction of hepatic sinusoids. Am J Physiol, 1994;266(4 Pt 1): G624–32. 28. Feng, H.Q., Weymouth, N.D., and Rockey, D.C. Endothelin antagonism in portal hypertensive mice: implications for endothelin receptor‐specific signaling in liver disease. Am J Physiol Gastrointest Liver Physiol, 2009;297(1):G27–33. 29. Iizuka, M., Murata, T., Hori, M. et al. Increased contractility of hepatic stellate cells in cirrhosis is mediated by enhanced Ca2+‐dependent and Ca2+‐ sensitization pathways. Am J Physiol Gastrointest Liver Physiol, 2011; 300(6):G1010–21. 30. Wisse, E. An electron microscopic study of the fenestrated endothelial lining of rat liver sinusoids. J Ultrastruct Res, 1970;31(1):125–50. 31. Bhunchet, E. and Fujieda, K. Capillarization and venularization of hepatic sinusoids in porcine serum‐induced rat liver fibrosis: a mechanism to maintain liver blood flow. Hepatology, 1993;18(6):1450–8. 32. Horn, T., Christoffersen, P., and Henriksen, J.H. Alcoholic liver injury: defenestration in noncirrhotic livers – a scanning electron microscopic study. Hepatology, 1987;7(1):77–82. 33. Funyu, J., Mochida, S., Inao, M. et al. VEGF can act as vascular permeability factor in the hepatic sinusoids through upregulation of porosity of endothelial cells. Biochem Biophys Res Commun, 2001;280(2):481–5. 34. May, D., Djonov, V., Zamir, G. et  al. A transgenic model for conditional induction and rescue of portal hypertension reveals a role of VEGF‐mediated regulation of sinusoidal fenestrations. PloS One, 2011;6(7):e21478.



51:  Pathophysiology of Portal Hypertension

35. DeLeve, L.D., Wang, X., Hu, L. et al. Rat liver sinusoidal endothelial cell phenotype is maintained by paracrine and autocrine regulation. Am J Physiol Gastrointest Liver Physiol, 2004;287(4):G757–63. 36. McGuire, R.F., Bissell, D.M., Boyles, J. et al. Role of extracellular matrix in regulating fenestrations of sinusoidal endothelial cells isolated from normal rat liver. Hepatology, 1992;15(6):989–97. 37. Svistounov, D., Warren, A., McNerney, G.P. et al. The relationship between fenestrations, sieve plates and rafts in liver sinusoidal endothelial cells. PloS One, 2012;7(9):e46134. 38. DeLeve, L.D. Liver sinusoidal endothelial cells in hepatic fibrosis. Hepatology, 2015;61(5):1740–6. 39. McConnell, M. and Iwakiri, Y. Biology of portal hypertension. Hepatology Int, 2018;12(Suppl 1):11–23. 40. Jarnagin, W.R., Rockey, D.C., Koteliansky, V.E. et al. Expression of variant fibronectins in wound healing: cellular source and biological activity of the EIIIA segment in rat hepatic fibrogenesis. J Cell Biol, 1994;127(6 Pt 2):2037–48. 41. Olsen, A.L., Sackey, B.K., Marcinkiewicz, C. et  al. Fibronectin extra domain‐A promotes hepatic stellate cell motility but not differentiation into myofibroblasts. Gastroenterology, 2012;142(4):928–37 e3. 42. Wang, R., Ding, Q., Yaqoob, U. et al. Exosome adherence and internalization by hepatic stellate cells triggers sphingosine 1‐phosphate‐dependent migration. J Biol Chem, 2015;290(52):30684–96. 43. Deleve, L.D., Wang, X., and Guo, Y. Sinusoidal endothelial cells prevent rat stellate cell activation and promote reversion to quiescence. Hepatology, 2008;48(3):920–30. 44. Marrone, G., Russo, L., Rosado, E. et  al. The transcription factor KLF2 mediates hepatic endothelial protection and paracrine endothelial‐stellate cell deactivation induced by statins. J Hepatol, 2013;58(1):98–103. 45. Failli, P., De, F.R., Caligiuri, A. et  al. Nitrovasodilators inhibit platelet‐ derived growth factor‐induced proliferation and migration of activated human hepatic stellate cells. Gastroenterology, 2000;119(2):479–92. 46. Routray, C., Liu, C., Yaqoob, U. et al. Protein kinase G signaling disrupts Rac1‐dependent focal adhesion assembly in liver specific pericytes. Am J Physiol Cell Physiol, 2011;301(1):C66–74. 47. Lee, J.S., Kang Decker, N., Chatterjee, S. et  al. Mechanisms of nitric oxide interplay with Rho GTPase family members in modulation of actin membrane dynamics in pericytes and fibroblasts. Am J Pathol, 2005;166(6):1861–70. 48. Langer, D.A., Das, A., Semela, D. et al. Nitric oxide promotes caspase‐independent hepatic stellate cell apoptosis through the generation of reactive oxygen species. Hepatology, 2008;47(6):1983–93. 49. Mittal, M.K., Gupta, T.K., Lee, F.Y. et  al. Nitric oxide modulates hepatic vascular tone in normal rat liver. Am J Physiol, 1994;267(3 Pt 1):G416–22. 50. Shah, V., Haddad, F.G., Garcia‐Cardena, G. et al. Liver sinusoidal endothelial cells are responsible for nitric oxide modulation of resistance in the hepatic sinusoids. J Clin Invest, 1997;100(11):2923–30. 51. Abraldes, J.G., Iwakiri, Y., Loureiro‐Silva, M. et al. Mild increases in portal pressure upregulate vascular endothelial growth factor and endothelial nitric oxide synthase in the intestinal microcirculatory bed, leading to a hyperdynamic state. Am J Physiol Gastrointest Liver Physiol, 2006;290(5):G980–7. 52. Rockey, D.C. and Chung, J.J. Reduced nitric oxide production by endothelial cells in cirrhotic rat liver: endothelial dysfunction in portal hypertension. Gastroenterology, 1998;114(2):344–51. 53. Shah, V., Toruner, M., Haddad, F. et  al. Impaired endothelial nitric oxide synthase activity associated with enhanced caveolin binding in experimental cirrhosis in the rat. Gastroenterology, 1999;117(5):1222–8. 54. Fulton, D., Gratton, J.P., McCabe, T.J. et  al. Regulation of endothelium‐ derived nitric oxide production by the protein kinase Akt. Nature, 1999;399(6736):597–601. 55. Liu, S., Premont, R.T., and Rockey, D.C. G‐protein‐coupled receptor kinase interactor‐1 (GIT1) is a new endothelial nitric‐oxide synthase (eNOS) interactor with functional effects on vascular homeostasis. J Biol Chem, 2012;287(15):12309–20. 56. Liu, S., Premont, R.T., and Rockey, D.C. Endothelial nitric‐oxide synthase (eNOS) is activated through G‐protein‐coupled receptor kinase‐interacting protein 1 (GIT1) tyrosine phosphorylation and Src protein. J Biol Chem, 2014;289(26):18163–74. 57. Michel, J.B., Feron, O., Sacks, D. et al. Reciprocal regulation of endothelial nitric‐oxide synthase by Ca2+‐calmodulin and caveolin. J Biol Chem, 1997;272(25):15583–6.

667

58. Abraldes, J.G., Albillos, A., Banares, R. et al. Simvastatin lowers portal pressure in patients with cirrhosis and portal hypertension: a randomized controlled trial. Gastroenterology, 2009;136(5):1651–8. 59. Abraldes, J.G., Rodriguez‐Vilarrupla, A., Graupera, M. et  al. Simvastatin treatment improves liver sinusoidal endothelial dysfunction in CCl4 cirrhotic rats. J Hepatol, 2007;46(6):1040–6. 60. Trebicka, J., Hennenberg, M., Laleman, W. et al. Atorvastatin lowers portal pressure in cirrhotic rats by inhibition of RhoA/Rho‐kinase and activation of endothelial nitric oxide synthase. Hepatology, 2007;46(1):242–53. 61. Ming, X.F., Viswambharan, H., Barandier, C. et al. Rho GTPase/Rho kinase negatively regulates endothelial nitric oxide synthase phosphorylation through the inhibition of protein kinase B/Akt in human endothelial cells. Mol Cell Biol, 2002;22(24):8467–77. 62. Laufs, U., and Liao, J.K. Post‐transcriptional regulation of endothelial nitric oxide synthase mRNA stability by Rho GTPase. J Biol Chem, 1998; 273(37):24266–71. 63. Dill, M.T., Rothweiler, S., Djonov, V. et  al. Disruption of Notch1 induces vascular remodeling, intussusceptive angiogenesis, and angiosarcomas in livers of mice. Gastroenterology, 2012;142(4):967–77 e2. 64. Corpechot, C., Barbu, V., Wendum, D. et  al. Hypoxia‐induced VEGF and collagen I expressions are associated with angiogenesis and fibrogenesis in experimental cirrhosis. Hepatology, 2002;35(5):1010–21. 65. Ehling, J., Bartneck, M., Wei, X. et  al. CCL2‐dependent infiltrating macrophages promote angiogenesis in progressive liver fibrosis. Gut, 2014;63(12):1960–71. 66. Thabut, D. and Shah, V. Intrahepatic angiogenesis and sinusoidal remodeling in chronic liver disease: new targets for the treatment of portal hypertension? J Hepatol, 2010;53(5):976–80. 67. Wanless, I.R., Wong, F., Blendis, L.M. et al. Hepatic and portal vein thrombosis in cirrhosis: possible role in development of parenchymal extinction and portal hypertension. Hepatology, 1995;21(5):1238–47. 68. Tripodi, A. Hemostasis abnormalities in cirrhosis. Curr Opin Hematol, 2015;22(5):406–12. 69. De Pietri, L., Bianchini, M., Montalti, R. et al. Thrombelastography‐guided blood product use before invasive procedures in cirrhosis with severe coagulopathy: a randomized, controlled trial. Hepatology, 2016;63(2):566–73. 70. Ordinas, A., Escolar, G., Cirera, I. et  al. Existence of a platelet‐adhesion defect in patients with cirrhosis independent of hematocrit: studies under flow conditions. Hepatology, 1996;24(5):1137–42. 71. Rubin, M.H., Weston, M.J., Langley, P.G. et al. Platelet function in chronic liver disease: relationship to disease severity. Dig Dis Sci, 1979;24(3):197–202. 72. Lisman, T., Bongers, T.N., Adelmeijer, J. et  al. Elevated levels of von Willebrand Factor in cirrhosis support platelet adhesion despite reduced functional capacity. Hepatology, 2006;44(1):53–61. 73. Raparelli, V., Basili, S., Carnevale, R. et  al. Low‐grade endotoxemia and platelet activation in cirrhosis. Hepatology, 2017;65(2):571–81. 74. Potze, W., Siddiqui, M.S., Boyett, S.L. et al. Preserved hemostatic status in patients with non‐alcoholic fatty liver disease. J Hepatol, 2016;65(5): 980–7. 75. Chauhan, A., Adams, D.H., Watson, S.P. et al. Platelets: no longer bystanders in liver disease. Hepatology, 2016;64(5):1774–84. 76. Chojkier, M. and Groszmann, R.J. Measurement of portal‐systemic shunting in the rat by using gamma‐labeled microspheres. Am J Physiol, 1981; 240(5):G371–5. 77. Groszmann, R.J., Vorobioff, J., and Riley, E. Splanchnic hemodynamics in portal‐hypertensive rats: measurement with gamma‐labeled microspheres. Am J Physiol, 1982;242(2):G156–60. 78. Vorobioff, J., Bredfeldt, J.E., and Groszmann, R.J. Hyperdynamic circulation in portal‐hypertensive rat model: a primary factor for maintenance of chronic portal hypertension. Am J Physiol, 1983;244(1):G52–7. 79. Sikuler, E., Kravetz, D., and Groszmann, R.J. Evolution of portal hypertension and mechanisms involved in its maintenance in a rat model. Am J Physiol, 1985;248(6 Pt 1):G618–25. 80. Sarin, S.K., Mosca, P., Sabba, C. et  al. Hyperdynamic circulation in a chronic murine schistosomiasis model of portal hypertension. Hepatology, 1991;13(3):581–4. 81. Colombato, L.A., Albillos, A., and Groszmann, R.J. Temporal relationship of peripheral vasodilatation, plasma volume expansion and the hyperdynamic circulatory state in portal‐hypertensive rats. Hepatology, 1992; 15(2):323–8.

668

THE LIVER:  REFERENCES

82. Lee, F.Y., Albillos, A., Colombato, L.A. et al. The role of nitric oxide in the vascular hyporesponsiveness to methoxamine in portal hypertensive rats. Hepatology, 1992;16(4):1043–8. 83. Sieber, C.C. and Groszmann, R.J. Nitric oxide mediates hyporeactivity to vasopressors in mesenteric vessels of portal hypertensive rats. Gastroenterology, 1992;103(1):235–9. 84. Lee, F.Y., Colombato, L.A., Albillos, A. et  al. N omega‐nitro‐L‐arginine administration corrects peripheral vasodilation and systemic capillary hypotension and ameliorates plasma volume expansion and sodium retention in portal hypertensive rats. Hepatology, 1993;17(1):84–90. 85. Sieber, C.C., Lopez‐Talavera, J.C., and Groszmann, R.J. Role of nitric oxide in the in vitro splanchnic vascular hyporeactivity in ascitic cirrhotic rats. Gastroenterology, 1993;104(6):1750–4. 86. Lopez‐Talavera, J.C., Merrill, W.W., and Groszmann, R.J. Tumor necrosis factor alpha: a major contributor to the hyperdynamic circulation in prehepatic portal‐hypertensive rats. Gastroenterology, 1995;108(3):761–7. 87. Sieber, C.C., Lee, F.Y., and Groszmann, R.J. Long‐term octreotide treatment prevents vascular hyporeactivity in portal‐hypertensive rats. Hepatology, 1996;23(5):1218–23. 88. Shah, V., Wiest, R., Garcia‐Cardena, G. et al. Hsp90 regulation of endothelial nitric oxide synthase contributes to vascular control in portal hypertension. Am J Physiol, 1999;277(2 Pt 1):G463–8. 89. Wiest, R., Das, S., Cadelina, G. et al. Bacterial translocation in cirrhotic rats stimulates eNOS‐derived NO production and impairs mesenteric vascular contractility. J Clin Invest, 1999;104(9):1223–33. 90. Wiest, R., Shah, V., Sessa, W.C. et al. NO overproduction by eNOS precedes hyperdynamic splanchnic circulation in portal hypertensive rats. Am J Physiol, 1999;276(4 Pt 1):G1043–51. 91. Iwakiri, Y., Cadelina, G., Sessa, W.C. et al. Mice with targeted deletion of eNOS develop hyperdynamic circulation associated with portal hypertension. Am J Physiol Gastrointest Liver Physiol, 2002;283(5):G1074–81. 92. Iwakiri, Y., Tsai, M.H., McCabe, T.J. et al. Phosphorylation of eNOS initiates excessive NO production in early phases of portal hypertension. Am J Physiol Heart Circ Physiol, 2002;282(6):H2084–90. 93. Tsai, M.H., Iwakiri, Y., Cadelina, G. et al. Mesenteric vasoconstriction triggers nitric oxide overproduction in the superior mesenteric artery of portal hypertensive rats. Gastroenterology, 2003;125(5):1452–61. 94. Wiest, R., Cadelina, G., Milstien, S. et al. Bacterial translocation up‐regulates GTP‐cyclohydrolase I in mesenteric vasculature of cirrhotic rats. Hepatology, 2003;38(6):1508–15. 95. Groszmann, R.J., Garcia‐Tsao, G., Bosch, J. et al. Beta‐blockers to prevent gastroesophageal varices in patients with cirrhosis. N Engl J Med, 2005;353(21):2254–61. 96. Huang, H.C., Haq, O., Utsumi, T. et al. Intestinal and plasma VEGF levels in cirrhosis: the role of portal pressure. J Cell Mol Med, 2012;16(5): 1125–33. 97. Iwakiri, Y. The molecules: mechanisms of arterial vasodilatation observed in the splanchnic and systemic circulation in portal hypertension. J Clin Gastroenterol, 2007;41(10 Suppl 3):S288–94. 98. Feletou, M. and Vanhoutte, P.M. Endothelium‐dependent hyperpolarizations: past beliefs and present facts. Ann Med, 2007;39(7):495–516. 99. Iwakiri, Y. The systemic and splanchnic circulation, in Chronic Liver Failure, Mechanisms and Management, (eds. Gines, P. et  al.), Humana Press, New York, 2011, pp. 305–21. 100. Fernandez, M., Mejias, M., Angermayr, B. et al. Inhibition of VEGF receptor‐2 decreases the development of hyperdynamic splanchnic circulation and portal‐systemic collateral vessels in portal hypertensive rats. J Hepatol, 2005;43(1):98–103. 101. Moleda, L., Trebicka, J., Dietrich, P. et al. Amelioration of portal hypertension and the hyperdynamic circulatory syndrome in cirrhotic rats by ­neuropeptide Y via pronounced splanchnic vasoaction. Gut, 2011;60(8): 1122–32. 102. Trebicka, J., Leifeld, L., Hennenberg, M. et  al. Hemodynamic effects of urotensin II and its specific receptor antagonist palosuran in cirrhotic rats. Hepatology, 2008;47(4):1264–76. 103. Kemp, W., Krum, H., Colman, J. et  al. Urotensin II: a novel vasoactive mediator linked to chronic liver disease and portal hypertension. Liver Int, 2007;27(9):1232–9. 104. Hennenberg, M., Trebicka, J., Kohistani, A.Z. et al. Vascular hyporesponsiveness to angiotensin II in rats with CCl(4)‐induced liver cirrhosis. Eur J Clin Invest, 2009;39(10):906–13.

105. Chu, C.J., Wu, S.L., Lee, F.Y. et al. Splanchnic hyposensitivity to glypressin in a haemorrhage/transfused rat model of portal hypertension: role of nitric oxide and bradykinin. Clin Sci (Lond), 2000;99(6):475–82. 106. Chen, C.T., Chu, C.J., Lee, F.Y. et al. Splanchnic hyposensitivity to glypressin in a hemorrhage‐transfused common bile duct‐ligated rat model of portal hypertension: role of nitric oxide and bradykinin. Hepatogastroenterology, 2009;56(94–95):1261–7. 107. Heinemann, A., Wachter, C.H., Fickert, P. et al. Vasopressin reverses mesenteric hyperemia and vasoconstrictor hyporesponsiveness in anesthetized portal hypertensive rats. Hepatology, 1998;28(3):646–54. 108. Song, D., Liu, H., Sharkey, K.A. et al. Hyperdynamic circulation in portal‐ hypertensive rats is dependent on central c‐fos gene expression. Hepatology, 2002;35(1):159–66. 109. Coll, M., Martell, M., Raurell, I. et al. Atrophy of mesenteric sympathetic innervation may contribute to splanchnic vasodilation in rat portal hypertension. Liver Int, 2010;30(4):593–602. 110. Ezkurdia, N., Coll, M., Raurell, I. et al. Blockage of the afferent sensitive pathway prevents sympathetic atrophy and hemodynamic alterations in rat portal hypertension. Liver Int, 2012;32(8):1295–305. 111. Fernandez‐Varo, G., Ros, J., Morales‐Ruiz, M. et al. Nitric oxide synthase 3‐dependent vascular remodeling and circulatory dysfunction in cirrhosis. Am J Pathol, 2003;162(6):1985–93. 112. Fernandez‐Varo, G., Morales‐Ruiz, M., Ros, J. et al. Impaired extracellular matrix degradation in aortic vessels of cirrhotic rats. J Hepatol, 2007;46(3):440–6. 113. Sumanovski, L.T., Battegay, E., Stumm, M. et al. Increased angiogenesis in portal hypertensive rats: role of nitric oxide. Hepatology, 1999;29(4): 1044–9. 114. Sieber, C.C., Sumanovski, L.T., Stumm, M. et al. In vivo angiogenesis in normal and portal hypertensive rats: role of basic fibroblast growth factor and nitric oxide. J Hepatol, 2001;34(5):644–50. 115. Groszmann, R.J., Kotelanski, B., and Cohn, J.N. Different patterns of porta‐systemic shunting in cirrhosis of the liver studied by an indicator dilution technique. Acta Gastroenterol Latinoam, 1971;3(3):111–6. 116. Bosch, J., Pizcueta, P., Feu, F. et al. Pathophysiology of portal hypertension. Gastroenterol Clin North Am, 1992;21(1):1–14. 117. Fernandez, M., Vizzutti, F., Garcia‐Pagan, J.C. et al. Anti‐VEGF receptor‐2 monoclonal antibody prevents portal‐systemic collateral vessel formation in portal hypertensive mice. Gastroenterology, 2004;126(3):886–94. 118. Geerts, A.M., De Vriese, A.S., Vanheule, E. et al. Increased angiogenesis and permeability in the mesenteric microvasculature of rats with cirrhosis and portal hypertension: an in vivo study. Liver Int, 2006;26(7):889–98. 119. Van Steenkiste, C., Geerts, A., Vanheule, E. et al. Role of placental growth factor in mesenteric neoangiogenesis in a mouse model of portal hypertension. Gastroenterology, 2009;137(6):2112–24 e1–6. 120. Fernandez, M., Mejias, M., Garcia‐Pras, E. et al. Reversal of portal hypertension and hyperdynamic splanchnic circulation by combined vascular endothelial growth factor and platelet‐derived growth factor blockade in rats. Hepatology, 2007;46(4):1208–17. 121. Tiani, C., Garcia‐Pras, E., Mejias, M. et  al. Apelin signaling modulates splanchnic angiogenesis and portosystemic collateral vessel formation in rats with portal hypertension. J Hepatol, 2009;50(2):296–305. 122. Mejias, M., Garcia‐Pras, E., Tiani, C. et al. Beneficial effects of sorafenib on splanchnic, intrahepatic, and portocollateral circulations in portal hypertensive and cirrhotic rats. Hepatology, 2009;49(4):1245–56. 123. Reiberger, T., Angermayr, B., Schwabl, P. et  al. Sorafenib attenuates the portal hypertensive syndrome in partial portal vein ligated rats. J Hepatol, 2009;51(5):865–73. 124. Huang, H.C., Wang, S.S., Hsin, I.F. et al. Cannabinoid receptor 2 agonist ameliorates mesenteric angiogenesis and portosystemic collaterals in cirrhotic rats. Hepatology, 2012;56(1):248–58. 125. Genecin, P., Polio, J., and Groszmann, R.J. Na restriction blunts expansion of plasma volume and ameliorates hyperdynamic circulation in portal hypertension. Am J Physiol Gastrointest Liver Physiol, 1990;259(22):G498–503. 126. Morgan, J.S., Groszmann, R.J., Rojkind, M., and Enriquez, R. Hemodynamic mechanisms of emerging portal hypertension caused by schistosomiasis in the hamster. Hepatology, 1990;11(1):98–104. 127. Moller, S. and Henriksen, J.H. Cardiovascular complications of cirrhosis. Gut, 2008;57(2):268–78. 128. Schrier, R.W. Water and sodium retention in edematous disorders: role of vasopressin and aldosterone. Am J Med, 2006;119(7 Suppl 1):S47–53.



51:  Pathophysiology of Portal Hypertension

129. Aller, R., de Luis, D.A., Moreira, V. et al. The effect of liver transplantation on circulating levels of estradiol and progesterone in male patients: ­parallelism with hepatopulmonary syndrome and systemic hyperdynamic circulation improvement. J Endocrinol Invest, 2001;24(7):503–9. 130. Angus, P.W., Vaughan, R.B., and Chin‐Dusting, J.P. Responses to endothelin‐1 in patients with advanced cirrhosis before and after liver transplantation. Gut, 2004;53(5):773. 131. Fallon, M.B. Mechanisms of pulmonary vascular complications of liver disease: hepatopulmonary syndrome. J Clin Gastroenterol, 2005;39(4 Suppl 2):S138–42. 132. Katsuta, Y., Honma, H., Zhang, X.J. et al. Pulmonary blood transit time and impaired arterial oxygenation in patients with chronic liver disease. J Gastroenterol, 2005;40(1):57–63. 133. Agusti, A.G., Roca, J., Bosch, J. et  al. Effects of propranolol on arterial oxygenation and oxygen transport to tissues in patients with cirrhosis. Am Rev Respir Dis, 1990;142(2):306–10. 134. Moller, S., Henriksen, J.H., and Bendtsen, F. Central and noncentral blood volumes in cirrhosis: relationship to anthropometrics and gender. Am J Physiol Gastrointest Liver Physiol, 2003;284(6):G970–9. 135. Shapiro, M.D., Nicholls, K.M., Groves, B.M. et  al. Interrelationship between cardiac output and vascular resistance as determinants of effective arterial blood volume in cirrhotic patients. Kidney Int, ­ 1985;28(2):206–11.

669

136. Schrier, R.W., Arroyo, V., Bernardi, M. et al. Peripheral arterial vasodilation hypothesis: a proposal for the initiation of renal sodium and water retention in cirrhosis. Hepatology, 1988;8(5):1151–7. 137. Cohn, J.N. Renal hemodynamic alterations in liver disease. Perspect Nephrol Hypertens, 1976;3:255–34. 138. Ruiz‐del‐Arbol, L., Monescillo, A., Arocena, C. et al. Circulatory function and hepatorenal syndrome in cirrhosis. Hepatology, 2005;42(2):439–47. 139. Cohn, J.N., Tristani, F.E., and Khatri, I.M. Systemic vasoconstrictor and renal vasodilator effects of PLV‐2 (octapressin) in man. Circulation, 1968;38(1):151–7. 140. Lenz, K., Hortnagl, H., Druml, W. et  al. Beneficial effect of 8‐ornithin vasopressin on renal dysfunction in decompensated cirrhosis. Gut, 1989;30(1):90–6. 141. Blei, A.T. Monitoring cerebral blood flow. A useful clinical tool in acute liver failure? Liver Transplantation, 2005;11(11):1320–22. 142. Vaquero, J., Chung, C., and Blei, A.T. Cerebral blood flow in acute liver failure: a finding in search of a mechanism. Metab Brain Dis, 2004;19(3–4):177–94. 143. Guevara, M., Bru, C., Gines, P. et al. Increased cerebrovascular resistance in cirrhotic patients with ascites. Hepatology, 1998;28(1):39–44.

52

Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment Yaron Rotman and Devika Kapuria Liver Energy and Metabolism Section, Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, USA

SUMMARY

EPIDEMIOLOGY

Non‐alcoholic fatty liver disease (NAFLD), the excess hepatic accumulation of fat in the form of triglycerides (TG), has become a global epidemic. The disease spans across a spectrum ranging from simple steatosis to non‐alcoholic steatohepatitis (NASH), where the fat accumulation is accompanied by inflammation and injury, and to cirrhosis and hepatocellular carcinoma. NAFLD is associated with other metabolic diseases including type 2 diabetes and coronary artery disease and there is increasing evidence of a link between NAFLD and several cancers. In this chapter, we review the epidemiology of NAFLD, discuss mechanisms of injury and progression, the natural history of the disease, and management options for patients.

Prevalence of NAFLD

DEFINITIONS NAFLD is defined as the presence of excess hepatic TG by imaging or histology in the absence of secondary causes of hepatic fat accumulation [1]. A liver TG content greater than 5.5% is considered abnormal, a threshold that was obtained from population imaging studies [2] and is consistent with data from biochemical assays of autopsy samples [3]. There are two forms of NAFLD, which can only be distinguished histologically. Non‐alcoholic fatty liver (NAFL) is defined as the presence of steatosis without evidence for hepatocellular damage (ballooning), whereas the presence of hepatocyte injury (or ballooning) is required to diagnose NASH. Both NAFL and NASH may present with or without fibrosis.

NAFLD is the most common liver disorder in the Western world and its prevalence appears to be increasing [4], concordant with the rise in metabolic syndrome, obesity, and type 2 diabetes. Estimates of disease prevalence vary depending on the method used to detect it. In the NHANES III survey (1988–1994) the prevalence of unexplained aminotransferase elevation, thought to represent NAFLD, was 5.4% [5] while in the 1999–2002 NHANES, the rate of elevated transaminases was 9.8% [6]. More recently, Takyar et al. [7] in an analysis of 3160 “healthy” volunteers reported a 27.9% prevalence of presumed NAFLD, based on elevated transaminases and body mass index (BMI). However, it is well known that normal aminotransferases do not preclude the presence of NAFLD [8]. This was demonstrated when the same 1988–1994 NHANES III cohort was assessed by ultrasonography and the prevalence of steatosis was found to be 21.4% [9]. In a later study using ultrasonography between 2007–2010 [10], 46% of 328 subjects had evidence of steatosis. Finally, using 1H‐magnetic resonance spectroscopy (MRS) in the multiethnic Dallas heart study, steatosis was found in one third of the study cohort [11]. There is a wide variation in the global prevalence and reporting of NAFLD. A recent meta‐analysis [4] that included 57 worldwide studies reported a pooled global NAFLD prevalence of 25.24%. The highest prevalence was reported from South America (30.45%) and the Middle East (31.79%). The lowest prevalence was reported in Africa (13.48%). Asia, Europe, and North America respectively have a pooled NAFLD prevalence of 27.37%, 23.71% and 24.13% [12].

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

A similar trend of increasing NAFLD prevalence is seen in pediatric populations, where prevalence rose from 3.9% in 1988–1994 to 10.7% in 2007–2010 [13]. An autopsy series spanning 1993–2003 estimates the prevalence of pediatric NAFLD as 9.6%, with a prevalence of 38% in obese children [14].

671

Race and ethnicity are major risk factors for NAFLD with a disproportionately higher prevalence in Hispanics and relative protection in African‐Americans compared to non‐Hispanic Caucasians [4, 11]. In Asians, NAFLD is being increasingly ­recognized even in subjects with normal body mass indices.

Prevalence of NASH In contrast to NAFLD, that can be identified reliably through imaging modalities, the diagnosis of NASH remains a histological one. Although several non‐invasive tests and formulas were suggested as a method to differentiate NAFL from NASH (­discussed below), none is sufficiently reliable to allow accurate diagnosis. As a liver biopsy is often not feasible or clinically‐ required in most patient cohorts, estimating the true incidence of NASH can be complex. In a recent meta‐analysis [4], the pooled prevalence of NASH among NAFLD patients that underwent a liver biopsy was 59.1%. However, this is likely overestimating true prevalence due to selection bias. In a study that performed prospective research liver biopsies in all subjects with NAFLD irrespective of the clinical indication 29.9% of subjects with NAFLD were found to have NASH [10]. Extrapolating from these data, the overall prevalence of NASH is estimated at 1.5–6.5% [4]. Consistent with that, in a large cohort of obese subjects who underwent bariatric surgery (expected to be enriched for NAFLD), 7.5% were found to have biopsy‐proven NASH [15].

NAFLD and the metabolic syndrome NAFLD often occurs together with obesity, diabetes, and the metabolic syndrome. There is a clear correlation between BMI and hepatic TG content [11]. The prevalence of obesity in NAFLD subjects is reported at 37–71% [4, 16] and over 90% of severely obese patients undergoing bariatric surgery were found to have NAFLD. Similarly, up to 33% of subjects with NAFLD have been found to have type 2 diabetes, whereas 70% of patients with diabetes have NAFLD; patients with both diabetes and NAFLD have an increased risk of NASH, fibrosis [17], and hepatocellular carcinoma [18]. Almost half of NAFLD and up to 71% of all NASH patients have been found to have the metabolic syndrome [4, 19]. Given the co‐occurrence of NAFLD and metabolic abnormalities, it is difficult to ascertain from cross‐sectional studies whether NAFLD is a consequence of obesity, diabetes, and metabolic syndrome, or vice versa. In a longitudinal study [20], subjects with no NAFLD at baseline were more likely to develop NAFLD after 7 years of follow‐up if they had higher BMI, insulin resistance, or presence of metabolic syndrome at baseline. Similarly, NAFLD at baseline was associated with incident diabetes after 6 years of follow‐up [21]. Thus, a bidirectional association exists between NAFLD and diabetes and the metabolic syndrome.

Additional risk factors Longitudinal studies show a distinctly increased incidence of NAFLD in males compared to females. In women, the incidence of NAFLD is higher in menopausal and post‐menopausal women compared to pre‐menopausal women [20, 22].

GENETIC EPIDEMIOLOGY The prevalence and severity of NAFLD have a strong heritable component, as seen in twin [23] and family [24] studies, and reflected by the strong racial impact [11]. Genome‐wide association studies (GWAS) and in recent years, sequencing approaches, identified several single nucleotide polymorphism (SNPs) that are associated with the disease. Of most interest are variants that associate with NAFLD independently of established risk factors such as obesity or diabetes, reflecting a true hepatic risk.

PNPLA3 The non‐synonymous SNP rs738409 (c.444C>G, p.I148M) in the patatin‐like phospholipase domain containing‐3 (PNPLA3) gene encoding the adiponutrin protein has consistently shown a strong association with NAFLD. Initially identified to be associated in the general population with hepatic TG [25], and liver enzymes [26], PNPLA3‐I148M has been subsequently associated with histological severity of NAFLD and NASH and was shown to promote inflammation, injury and fibrosis in addition to its impact on liver fat content [27, 28]. The high‐ risk allele is also associated with earlier presentation of disease in pediatric patients [27], risk of hepatocellular carcinoma [29], alcoholic liver disease [30], and has also been shown to impact other liver diseases, suggesting it affects a common injury pathway. The mechanism by which the I148M mutation affects hepatic lipid metabolism is not clear. Adiponutrin is present in the endoplasmic reticulum and lipid droplet membranes of hepatocytes and adipose cells and demonstrates both TG hydrolase and acyltransferase activities, although this may not be its physiological function in vivo. The I148M mutation impairs ubiquination, leading to adiponutrin accumulation on hepatic lipid droplets [31] and modulation of the droplet TG and phospholipid composition [32]. Beyond its function in hepatocytes adiponutrin has retinyl‐palmitate lipase activity in stellate cells which may explain its link to hepatic fibrosis [33].

TM6SF2 A SNP in the transmembrane 6 superfamily member 2 (TM6SF2) gene, rs58542926 (c.499A>G, p.E167K), was found to be associated with NAFLD [34]. The minor allele, associated with increased risk of NAFLD, is also associated with decreased serum LDL‐cholesterol and TG and with decreased risk for myocardial infarction [35]. Inhibition of TM6SF2 impairs the lipidation of nascent very low‐density lipoprotein (VLDL) in the liver, resulting in decreased TG secretion and increased cellular TG accumulation [36, 37].

672

THE LIVER:  MECHANISMS OF HEPATIC INJURY AND DEVELOPMENT OF STEATOHEPATITIS

HSD17B13 Several SNPs in 17‐beta hydroxysteroid dehydrogenase 13 (HSD17B13) show a strong association with NAFLD‐­associated liver injury. A splice‐site SNP, rs72613567, generates alternative loss‐of‐function isoforms that associate with protection from hepatocyte injury, inflammation and fibrosis but conversely with increased steatosis in NAFLD subjects [38, 39]. Other loss‐of‐function variants in the HSD17B13 gene have been identified, with a similar protective effect on injury [39, 40]. The gene product is a hepatic lipid‐droplet protein with retinol dehydrogenase activity in vitro and in vivo [39]. Several other genes were associated with histological f­ eatures of NAFLD, including MBOAT7 and GCKR [41]. In addition, epigenetic changes also seem to play a role in disease pathogenesis, with data suggesting that methylation of mitochondrial DNA is associated with histological severity [42].

PATHOGENESIS OF NAFLD MECHANISMS OF FAT ACCUMULATION The hallmark of NAFLD is the accumulation of TG in hepatocytes. The building blocks of TG are fatty acids (FA), which can be delivered to the hepatocyte as non‐esterified fatty acids (NEFA) from adipose tissue lipolysis or as part of chylomicron TG from a dietary source, and can also be formed within the hepatocyte through the de novo lipogenesis pathway. Hepatic TG and FA can be secreted from the liver in VLDL, be oxidized through β‐oxidation or metabolized to other derivatives. Fatty liver reflects an imbalance between these input and output pathways (Figure 52.1).

contributes less than 5% of the total TG synthesis [47]. However, in NAFLD patients, DNL is markedly increased and accounts for nearly a quarter of the FA in hepatic TG [48]. The presence of increased hepatic DNL, an insulin‐driven process, despite hepatic IR suggests the presence of selective IR, either through impairment of specific signaling pathways [49] or as a response to altered nutrient fluxes [50]. Thus, the liver in NAFLD is exposed to increased levels of NEFA derived from dysregulated adipose tissue lipolysis and hepatic lipogenesis.

VLDL excretion The excess influx of NEFA to the liver, and resultant excess formation of hepatic TG are accompanied by increase in VLDL and VLDL‐TG secretion. However, this increase is limited by the ability to secrete ApoB100 and is not sufficient to balance the rate of intrahepatic TG production [51].

β‐oxidation There are conflicting data regarding the role of hepatic mitochondrial β‐oxidation in accumulation of liver TG, where some studies suggest an increase in hepatic FA oxidation in individuals with NAFLD [52] while other studies show a decrease [53]. This discrepance may be due, in part, to difference in study methodologies.

MECHANISMS OF HEPATIC INJURY AND DEVELOPMENT OF STEATOHEPATITIS Two‐hit versus multi‐hit hypotheses

In NAFLD, the majority of FA in hepatic TG are derived from circulating NEFA (59%), while newly‐formed FA contribute 26% and dietary fat 15% [43]. Most circulating NEFA derive from adipocyte lipolysis (82% in the fasting state and 62% in the fed state), suggesting the adipocyte is the main source of hepatic FA. Adipose tissue insulin resistance (IR) is broadly found in NAFLD, manifesting as failure to suppress postprandial lipolysis and leads to increased delivery of NEFA to the liver [44]. Interestingly, the concentration of intrahepatic TG in NAFLD [45] is strongly associated with adipose tissue IR but not with hepatic IR. Furthermore, the degree of adipose IR has been shown to be worse in NASH compared to NAFLD [46], and in NASH patients to be associated with the fibrosis score. Overall, this demonstrates the important role of adipocyte dysfunction and excess delivery of adipocyte‐derived NEFA to the liver in the pathogenesis of hepatic TG accumulation.

Steatosis is common but NASH is only seen in a subset of patients. Similarly, steatosis is relatively easy to generate in murine models but recapitulating NASH in them has been far more difficult. Thus, the mechanisms that underly the transition from steatosis to steatohepatitis need to be elucidated. Initially a two‐hit hypothesis was proposed [54]. The accumulation of hepatic TG was suggested as a first hit, sensitizing the liver to further insults; a second hit would be required to drive the injury process, mainly through induction of lipid peroxidation. Potential suggested second hits were medications, gut‐derived endotoxin, iron overload, and various inducers of oxidative stress. In recent years it has become clear that a two‐hit model is unable to explain the complexity of NAFLD and a more complex, “multi‐hit” or “continuous‐hit” hypothesis has been formed. The major driving process of hepatocyte injury is thought to be lipotoxicity from free FA or their derivatives [55], with subsequent activation of inflammatory responses and fibrosis. Genetic and nutritional factors are modulators of this process.

De novo lipogenesis

Lipotoxicity

De novo lipogenesis (DNL), the formation of FA from acetyl‐ CoA, is highly regulated and key enzymes in the pathway are upregulated in NAFLD. In healthy lean individuals, DNL

As previously discussed, there is an increase in net hepatic FA flux in NAFLD. Shunting of excess FA to form TG is an important defense mechanism against lipotoxicity, which probably

Adipose tissue NEFA



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

(a)

673

VLDL

Acetyl-CoA

Glucose

TG FA

β-oxidation

Glucose

CM

NEFA

(b) VLDL

Acetyl-CoA

Glucose

TG FA

β-oxidation

Glucose

CM

NEFA

Figure 52.1  Mechanisms of hepatic fat accumulation. (a) Fatty acids in the liver reflect a balance between input and disposal. Input pathways include delivery of non‐esterified fatty acids from adipose tissue and triglycerides from chylomicrons and generation of fatty acids from acetyl‐CoA through de novo lipogenesis (double arrow). Disposal includes oxidation or generation of triglycerides. Triglycerides can be secreted in VLDL or hydrolyzed to fatty acids. (b) Adipose tissue insulin resistance in NAFLD causes an increase in delivery of non‐esterified fatty acids to the liver, while hepatic and muscle insulin resistance cause hyperglycemia and increased availability of acetyl‐CoA. De novo lipogenesis is driven by hyperinsulinemia generating a net increase in fatty acids and increased shunting to form triglycerides. Secretion of VLDL particles is increased but not sufficient to compensate. CM, chylomicrons; FA, fatty acids; NEFA, non‐esterified fatty acids; TG, triglycerides; VLDL, very low‐density lipoprotein.

explains the relatively benign phenotype of NAFL. In fact, TG are likely inert and not causing injury by themselves as highlighted by animal studies in which inhibition of TG formation has led to increased liver damage despite the inability to generate steatosis [56]. When this mechanism is overwhelmed, excess FA, especially saturated ones, and downstream metabolites such as ceramides, diacylglycerols, and lysophosphatidylcholine (LPC) accumulate and initiate a cascade of cell injury and death, inflammation, and resultant fibrosis [55].

Oxidative and endoplasmic reticulum stress Oxidative stress has been strongly implicated in the pathogenesis of NASH [57]. The increased hepatic metabolic load drives an adaptive process with increased capacity of hepatic mitochondria for oxidative phosphorylation [58]; however, in NASH this

adaptation is impaired, possibly as a result of lipotoxic injury to the mitochondria, resulting in increased formation of reactive oxygen species (ROS) and generation of oxidative stress [59]. Increased microsomal and peroxisomal oxidation of FA also contributes to the formation of excess ROS [60]. Uncontrolled ROS generate injury to the membranes and DNA and lead to the formation of highly toxic lipid peroxidation products, especially malondialdehyde and 4‐hydroxynonenal [61]. Endoplasmic reticulum (ER) stress has been demonstrated in NAFLD [62], likely due to lipotoxic injury from saturated FA, and leads to activation of the unfolded protein response. As with mitochondria, NASH appears to involve an overwhelming of this adaptive response, and chronic ER stress by itself induces increased DNL and formation of ROS, forming a vicious cycle [63]. When overwhelmed, ER stress activates IkB and c‐Jun N‐terminal kinases, initiating a cascade of proinflammatory and pro‐apoptotic responses.

674

THE LIVER:  NATURAL HISTORY OF NAFLD

Extension of damage beyond the hepatocyte The lipotoxic damage to hepatocytes activates regulated death pathways in NASH, predominantly apoptosis [64], with activation of both intrinsic and extrinsic pathways. In addition, there appears to be a population of stressed cells which undergo sublethal injury [65]; the histological correlate of these cells are the ballooned hepatocytes seen in NASH livers. The apoptotic and ballooned hepatocytes are releasing a myriad of signals including sonic hedgehog [66] and, in addition, these cells are releasing extracellular vesicles [65] containing a variety of signaling molecules including ceramides, CXCL10, TRAIL, miRNA, and mitochondrial DNA [67]. These act as chemokines and damage‐ associated molecular patterns (DAMPs), which, together with pathogen‐associated molecular patterns (PAMPs) derived from gut bacteria, are recognized by pattern recognition receptors (PRRs) including toll‐like receptors (TLR) and NOD‐like receptors (NLR). Of the various TLRs and NLRs, the ones mostly implicated in NASH pathogenesis are TLR4 and the NLRP3 inflammasome, respectively. The DAMP/PAMP‐PRR cascade leads to activation of cells of the innate immune system [68].

The gut microbiota can affect NAFLD through several mechanisms. Disruption or leakage across the gut barrier can lead to increased hepatic exposure to lipopolysaccharide (LPS) and other microbial products that act as PAMPs and activate innate immune responses in the liver, as previously described [75]. Metabolites derived from the interaction of the gut microbiota with gut content such as trimethylamine (TMA), a metabolite of choline, short‐chain fatty acids, amino acid metabolites, and endogenous ethanol, can activate signaling pathways in the liver and modulate NAFLD and NASH. Finally, gut bacteria metabolize and modify the components of the bile acid pool; this in turn can affect the activation of intestinal and hepatic farnesoid X receptor (FXR) and modulate the development of steatosis and steatohepatitis [76]. Furthermore, microbial metabolism of bile acids can modify activation of hepatic NKT cells and their antitumor activity, suggesting a link between the gut microbiome and risk for liver cancer [77]. However, these mechanistic associations have to be confirmed in human studies.

NATURAL HISTORY OF NAFLD Immune cell activation Kupffer cells (KC), the liver‐resident macrophages, are implicated in the initiation of the inflammatory process in NASH. Recognition of PAMP/DAMP by PRRs, as well as a direct effect of free fatty acids leads to KC activation and secretion of cytokines such as tumor necrosis factor (TNF) and interleukin‐1β and chemokines such as C‐C motif ligand 2 (CCL2) and CCL5. These, in turn, lead to recruitment and activation of circulating monocyte‐derived macrophages to the liver, and amplification of the inflammatory process [69]. Furthermore, KC themselves are major producers of ROS, further propagating the damage cascade. Neutrophils are also recruited to the liver by cytokines and DAMPs/PAMPs and further contribute to the inflammatory process through their myeloperoxidase activity [70]. Targeted depletion of KC in mouse models or genetic disruption of TLR4, the CCL2‐CCR2 axis or neutrophil enzymes, all lead to amelioration of experimental NASH, confirming the important role of these innate immune cells [69]. There is also emerging data regarding the role of additional immune cells, including natural killer (NK) and NKT cells, regulatory T cells, and mucosal associated invariant T (MAIT) cells [71].

Microbiome In recent years, multiple studies have demonstrated a difference in the composition of gut microbial populations between subjects with NAFLD and controls, and between subjects with NASH and those with NAFL. However, findings have not been consistent across studies, limiting the ability to generalize those findings and utilize them to establish a plausible mechanism [72]. Stronger evidence supporting the role of dysbiosis in NAFLD comes from animal studies. In a pivotal experiment, the development of fatty liver on a steatogenic diet in germ‐free mice was modulated by the microbiome implanted into these mice [73] and steatosis could be transmitted by co‐housing of mice with different microbiota [74].

NAFLD is a slowly progressing disease, and will affect patients over a span of decades, if not through their entire lifetime. Although often asymptomatic, NAFLD is associated with increased overall and liver‐associated mortality.

Hepatic outcomes of NAFLD Symptoms of NAFLD, if any, are subtle and nonspecific and may include hepatic discomfort and fatigue. More serious symptoms and complications, such as hepatic decompensation, need for liver transplant, and liver‐related mortality, are typically associated with progression to cirrhosis [78]. As such, the ­presence of fibrosis is the main determinant of eventual hepatic outcomes of NAFLD [79–81]. Fibrosis progresses at a variable rate and can even regress spontaneously [82, 83], likely reflecting the highly dynamic and fluctuating nature of the underlying disease. In paired‐biopsy series, the most important predictor of progression of fibrosis is the presence of NASH or inflammation [83, 84]. Thus, the pathophysiological paradigm of injury and inflammation as drivers of fibrosis is consistent with the clinical data. Beyond progression to hepatic decompensation, patients with NAFLD can also develop hepatocellular carcinoma (HCC) [85]. As with other liver diseases, most cases of HCC arise in the context of cirrhosis; however, there is now clear evidence that NAFLD can lead to HCC in the absence of underlying cirrhosis and that the risk of noncirrhotic HCC appears to be higher in NAFLD compared to viral hepatitis [86].

Extrahepatic manifestations Although the presence of NAFLD (and especially NASH) is accompanied by an increase in liver‐related mortality, cardiovascular disease and cancer are still the first and second leading causes of death in these patients. Thus, there is a clear association of NAFLD with extrahepatic manifestations.



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

Cardiovascular disease The presence of fatty liver disease is associated with increased cardiovascular risk. This increased risk spans the spectrum of cardiovascular disease (CVD) from asymptomatic atherosclerosis [87] to increased cardiovascular events [88] and mortality [81, 89]. Given that NAFLD and CVD share risk factors such as obesity and insulin resistance, it is still not clear whether the presence of NAFLD confers an increased risk beyond those risk factors, and whether there is a direct mechanistic effect of NAFLD on atherosclerosis development.

Extrahepatic cancers Beyond the risk of HCC, patients with NAFLD also have higher rates of extrahepatic malignancies. Several studies have shown an increased risk of colorectal adenomas and cancer [90], with more advanced liver disease likely portending a higher risk [91]. An increase in breast cancer [92] as well as in other cancers [93] was also reported to be associated with NAFLD. As with cardiovascular disease risk, it is unclear whether the increased cancer risk associated with NAFLD reflects an association with obesity and diabetes, both known to increase cancer risk, or whether it is independent of them.

Chronic kidney disease Chronic kidney disease is more common in NAFLD and the risk is higher in patients with NASH [94], even after controlling for metabolic risk factors.

EVALUATION OF THE PATIENT WITH NAFLD Clinical features and diagnosis NAFLD is diagnosed when there is evidence of hepatic steatosis on imaging or histology and there are no secondary causes of  steatosis present [1]. Patients with NAFLD are typically asymptomatic, and diagnosis is often incidental. Patients may occasionally present with fatigue or a vague, nondescript pain or discomfort in the right upper quadrant and may have hepatomegaly due to fatty infiltration of the liver. Signs or symptoms of chronic liver disease are generally absent unless patients ­present with advanced liver disease or cirrhosis.

Role of liver biopsy The liver biopsy is an important tool in the diagnosis and management of NAFLD. It is the gold standard and the only reliable method to differentiate NASH from NAFL and also allows for assessment of the multiple components of disease and for accurately staging the degree of fibrosis. There are several scoring schemes that are utilized for semiquantitative assessment of NAFLD histology, most commonly the NASH–CRN scoring system [95]. These scores are appropriate for use in clinical trials but should not be used in lieu of an appropriate histological diagnosis [96]. Despite its usefulness, the use of a liver biopsy is limited by its inherent risk [97], cost and sampling error.

675

Currently, guidelines recommend considering a liver biopsy in patients who are at an increased risk of having advanced fibrosis, as well as in patients who are suspected to have competing etiologies for hepatic steatosis, or to identify the chief cause of liver damage in case of more than one liver disease [1]. Thus, there is an unmet need for non‐invasive tools to identify and stratify NAFLD.

Non‐invasive assessments Assessment of liver steatosis Imaging modalities such as ultrasound or computed tomography (CT) are reasonably sensitive and specific for the detecting the presence of moderate–severe steatosis [98]. For example, for detection of greater than 20% steatosis, ultrasound has 91% sensitivity and 99% specificity [99] but sensitivity markedly decreases for mild steatosis. The degree of hyperechogenicity on ultrasound or the relative or absolute hepatic attenuation on CT can be used for semiquantitative assessment of the degree of steatosis but with limited accuracy. Ultrasound remains the primary modality for diagnosis of hepatic steatosis, due to its wide availability, lack of radiation exposure, and low cost. Recently, the controlled attenuation parameter (CAP), quantifying the degree of ultrasound attenuation in the liver was found to be a sensitive measure for detection and quantification of steatosis [100] In contrast to ultrasound and CT, magnetic resonance imaging (MRI)‐based imaging modalities offer superior sensitivity and specificity for detecting steatosis. An additional advantage of MRI is the ability to accurately quantify liver fat using either MR spectroscopy or MRI proton density fat fraction (MRI‐ PDFF) [101]. While MRI‐based detection methods have a high sensitivity and specificity, cost and limited availability currently do not make them a viable option outside of clinical research. There are several composite scores to calculate the degree of steatosis that are based on serum biomarkers, with or without anthropometric parameters, such as the fatty liver index, the hepatic steatosis index, and the proprietary SteatoTest; these have reasonable accuracy but it is unclear whether they can replace the need for an imaging study in clinical care [102].

Identification of NASH A crucial element in managing patients with NAFLD and in clinical trial recruitment is the differentiation of NAFL from NASH. As detailed above, a liver biopsy is the gold standard for diagnosis, but poses significant limitations in terms of acceptability, risk, and cost. Several blood‐based markers for the diagnosis of NASH have been studied, based on disease pathogenesis. Examples are cytokeratin‐18 fragments as markers of hepatocyte apoptosis [103, 104], oxidized lipid products [104], and several proprietary panels (reviewed in [102]). However, these all suffer from limited accuracy, incomplete validation, and/or cost and are not yet ready for routine clinical application.

Assessment of fibrosis Since fibrosis is the main determinant of mortality and liver‐related complications in NAFLD, a non‐invasive marker of fibrosis would provide risk stratification and allow for selection of patients for

676

THE LIVER:  TREATMENT OPTIONS IN NAFLD

treatment. As with assessment of steatosis, investigated biomarkers include imaging modalities and blood‐based markers [105]. Liver fibrosis is associated with increased hepatic stiffness. Several imaging modalities measure liver elasticity, which is directly related to stiffness, by generating a shear wave in the liver and measuring its velocity. Vibration controlled transient elastography (Fibroscan), shear‐wave elastography, and acoustic radial force imaging (ARFI) are based on ultrasound, while MRI is used for magnetic resonance elastography (MRE). In general, elastography‐based tests provide high accuracy in detecting cirrhosis and advanced fibrosis but are not sufficiently accurate to diagnose the presence of earlier fibrosis (stage 2) [102, 106]. Several blood‐based biomarkers or compound scores were developed to detect fibrosis. Some of these are nonspecific but employ readily available data such as liver enzymes, platelet count, and clinical and anthropometric features. These include scores initially applied to viral hepatitis, such as the AST/ALT ratio, AST/platelet ratio index (APRI) or FIB‐4, as well as scores developed specifically for NAFLD such as the NAFLD fibrosis score (NFS) or BARD score. The accuracy of all of these scores is limited and despite easy availability, they are not sufficient for routine clinical use, though they are useful in epidemiological studies [105].

Other blood‐based biomarkers are more specific and are based on measuring markers of the hepatic fibrotic process and deposition of extracellular matrix. Blood levels of hyaluronic acid, TIMP1, and pro‐collagen III amino terminal peptide (PIIINP and Pro‐C3) are among the promising ones. Several compound and proprietary scores are based on these markers, including the enhanced liver fibrosis (ELF), the FibroSure/ FibroTest, and FibroMeter NAFLD. In general, these specific scores provide greater accuracy than the nonspecific ones, but are currently limited by availability and cost.

TREATMENT OPTIONS IN NAFLD Despite the rise in prevalence and impact of NAFLD and NASH, there are currently no approved therapies for the disease. However, several medications approved for other indications have shown benefit in treating NASH and a plethora of new pharmacological agents are being studied. With the increasing understanding of pathogenic processes, there is a plethora of agents developed to target these processes (Figure 52.2) [107]. Reviewed here are currently available treatments with proven

Figure 52.2  Mechanistic targets for NAFLD/NASH treatment. Increased availability of carbohydrates and non‐esterified fatty acids, together with insulin resistance, drive hepatic accumulation of triglycerides and fatty acids, leading to metabolic and oxidative stress, hepatocyte injury, activation of an inflammatory response and eventually fibrosis. Lifestyle modification and bariatric surgery decrease the excessive caloric load to the liver. GLP‐1 receptor agonists (such as liraglutide) decrease caloric load (through reduction in appetite and weight loss) and improve insulin sensitivity with possible direct effect on the liver. PPAR agonists (like pioglitazone or elafibranor) improve insulin resistance in adipose tissue and/ or muscle, reducing the excess load of glucose and fatty acids to the liver, while also decreasing hepatic lipogenesis. Agonists of FXR (such as obeticholic acid) or FGF‐19 analogs (NGM‐282) decrease bile acid synthesis and hepatic lipogenesis; NGM‐282 may also act as an insulin sensitizer. Inhibitors of acetyl‐coA carboxylase, the first step in hepatic de novo lipogenesis, block hepatic fatty acid synthesis and accumulation. Downstream of hepatic lipid metabolism, vitamin E, a fat‐soluble antioxidant, ameliorates the oxidative stress that results from lipotoxicity while ASK1 inhibitors (such selonsertib) block the ASK1‐JNK pathway through which oxidative stress leads to hepatocyte injury and apoptosis. Cenicriviroc, a C‐C chemokine receptor type 2 (CCR2) and 5 (CCR5) antagonist, blocks the recruitment of inflammatory cells and decreases inflammatory injury. Finally, anti‐fibrotic agents target the fibrotic process itself. ACCi, acetyl‐coA carboxylase inhibitors; ASK1i, apoptosis signal‐regulating kinase 1 inhibitors; CCR, C‐C chemokine receptor; CHO, carbohydrates; FA, fatty acids; FGF, fibroblast growth factor; FXR, farnesoid X receptor; GLP‐1RA, glucagon‐like peptide 1 receptor agonists, NEFA, non‐esterified fatty acids; PPAR, peroxisome proliferator‐activator receptor; TG, triglycerides.



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

677

Table 52.1  Select response rates in clinical trials of NASH Treatment

Duration

Histological responsea (%)

Resolution of NASH (%)

Lifestyle (all) Lifestyle (Subjects with ≥ 5% weight loss)

12 months 12 months

47 82

25 58

Pioglitazone

18 months –  96 weeks 96 weeks 72 weeks 52 weeks 48 weeks

34b–58

47–51

19b 26b (Subjects with ≥ 10% weight loss had 45% fibrosis regression) 44b

43 45 48 43

36 20b 19 35

41b 35 Not reported 26b

Vitamin E Obeticholic acid Elafibranor (120 mg) Liraglutide

Improvement in fibrosis (%)

References [109] [109] [112, 113] [112] [115] [114] [124]

 Histological response criteria varied between trials but typically include at least 2‐point reduction in the NAFLD activity score (NAS). Response percentages calculated by intention‐to‐treat. b  Non‐significant compared to placebo. a

benefit, and those that are in advanced clinical trial stages at the time of writing (Table 52.1).

Reduction of caloric intake Treatment trials of lifestyle intervention, including diet and physical activity, have shown not only reduction in liver TG content but also histological improvement, directly related to the degree of weight loss [108]. Recently Vilar‐Gomez and others [109] reported on the results of a 52‐week lifestyle intervention program in 293 subjects with NASH. Only 30% of participants achieved greater than 5% weight loss, highlighting the difficulty of achieving a sustained effect, the main limitation of lifestyle interventions. However, among those who lost greater than 5% of their weight, resolution of NASH was seen in 58%. Furthermore, in the few subjects who lost greater than 10% of their weight, resolution of NASH was seen in 90% and a reduction in fibrosis stage was seen in 81%. Another evidence, albeit indirect, for the benefit of weight reduction is the marked rate of histological response in the control arms of therapeutic trials, which is typically associated with losing weight [110]. Given the proven effects of lifestyle modification for NASH and its known benefits for associated comorbidities such as diabetes and metabolic syndrome, it should be advocated in any patient with NAFLD, irrespective of pharmacological therapy options. Due to the difficulty of sustaining weight loss with lifestyle interventions alone, bariatric surgery is useful for patients with obesity and has shown benefit in treating diabetes. In the largest series to date [15], histological resolution of NASH was seen in 85% of subjects one year after surgery, and 33% showed improvement in steatosis. Recently, 5‐year follow‐up data was reported from the same cohort [111]. The 85% NASH resolution rate persisted and fibrosis continued to improve between years 1 and 5. Thus, bariatric surgery may be an attractive therapy for a subset of patients with NASH,

Pharmacological treatment PPAR activation The peroxisome proliferator‐activator receptors (PPARs) are a group of nuclear receptors that transcriptionally regulate a wide variety of metabolic and inflammatory processes. Their activation is typically driven by binding various FA and FA derivatives as

ligands. PPARγ is expressed predominantly in the adipose tissue and controls adipogenesis, lipogenesis, and glucose metabolism. Thiazolidinediones (TZDs) are synthetic agonists of PPARγ and are approved for the treatment of diabetes. Pioglitazone, the TZD most commonly used, was studied for the treatment of NASH in several trials. In the PIVENS study, 163 non‐diabetics with NASH were treated with pioglitazone or placebo for 96 weeks. Histological improvement was seen in 34% of subjects treated with pioglitazone, compared to 19% in placebo, but the difference did not reach the pre‐specified statistical significance level [112]. In contrast, in a study including 101 diabetics or pre‐diabetics with NASH, 18 months of pioglitazone treatment were associated with resolution of NASH in 51% of subjects, which was significantly greater than the placebo‐treated arm [113]. Overall, pioglitazone treatment appears to be effective for NASH, especially in subjects with diabetes or pre‐diabetes, but its acceptability may be limited by concerns about cardiovascular risks and by weight gain, which is a common side‐effect. Elafibrinor is a dual PPAR agonist, targeting both PPARα and PPARδ. The phase IIb GOLDEN‐505 trial compared two doses of elafibrinor (80 mg and 120 mg) for 52 weeks to placebo. Treatment with elafibranor was associated with histological improvement, but only in a post hoc analysis in subjects with significant disease activity at baseline [114]. A phase III trial is underway.

Bile acid signaling FXR acts as an intracellular bile acid sensor and when activated, inhibits bile acid synthesis and secretion as well as decreasing lipogenesis. Obeticholic acid (OCA) is a synthetic bile acid that acts as an FXR agonist. The phase IIb FLINT trial compared OCA with placebo in patients with NASH; OCA was superior to placebo in achieving histological improvement (46%) and NASH resolution (22%) [115] but was associated with significant pruritus and a mild increase in LDL‐­cholesterol. A phase III trial to determine efficacy and safety is ongoing and other FXR agonists are in earlier stages of development. Fibroblast growth factor 19 (FGF19) is a peptide hormone that is released in response to FXR activation in the terminal ileum and controls hepatocyte bile acid synthesis through its receptor FGFR4. FGF19 also has insulin‐like effects on hepatic gluconeogenesis and glycogen synthesis. Phase II studies in patients with NAFLD using NGM282, an FGF19 analog, demonstrate

678

THE LIVER:  REFERENCES

efficacy in reducing hepatic fat content [116] and improving NASH histology [117].

Vitamin E Given the evidence for the role of oxidative stress in the pathogenesis of NASH, antioxidants could have therapeutic benefit. Vitamin E (α‐tocopherol), a lipid‐soluble antioxidant, has been studied in several clinical trials. In the previously mentioned PIVENS trial, 43% of non‐diabetic patients with NASH treated with 800 IU/day of natural vitamin E had histological improvement and 36% had resolution of NASH [112] and similar results were seen in the TONIC pediatric trial [118]. Furthermore, of all pharmacological therapies, vitamin E is the only one shown to decrease the rate of clinically meaningful outcomes such as decompensation or need for liver transplant, albeit in a retrospective analysis [119]. However, there have been reports of increased risk of hemorrhagic stroke [120] and prostate cancer [121] in subjects receiving high‐dose vitamin E, and a recent report from a small trial suggests an increase in gastrointestinal bleeding in NAFLD patients, especially in diabetics [122]. Thus, using vitamin E to treat NASH can be considered but requires monitoring and assessment of potential risks.

GLP 1 receptor agonists GLP‐1 (glucagon‐like peptide‐1) is an incretin hormone secreted by enteroendocrine L cells in response to ingested nutrients. It acts predominantly in the fed‐state, modulating and augmenting the insulin response. GLP‐1 potentiates glucose‐stimulated insulin secretion from beta cells, mediated by activation of its  G  protein‐coupled receptor. GLP‐1 receptor agonists (GLP‐1RA) are in clinical use for the treatment of diabetes and obesity and are thought to mediate their effects through a combination of the insulinotropic effect, weight loss, increased energy expenditure, and delayed gastric emptying. In diabetes‐associated NAFLD, GLP‐1RAs lead to reduction in liver enzymes and liver fat content in human and animal studies [123]. In the phase II randomized controlled LEAN trial [124], 52 non‐diabetics with NASH were treated with liraglutide, a GLP‐1RA, or placebo for 48 weeks. Liraglutide provided histological benefit including resolution of NASH in 39%. Larger trials with other GLP‐1RA formulations are ongoing. Given the combined benefit on insulin resistance, weight, liver histology, and beneficial cardiovascular outcome in diabetics, GLP‐1RAs are an attractive option to explore further in the management of NAFLD.

Inhibition of de novo lipogenesis As detailed above, DNL is increased in the liver of subjects with NAFLD and DNL‐derived fatty acids may be drivers of hepatic lipotoxicity. GS‐0976 is a liver‐targeted inhibitor of acetyl‐CoA carboxylase (ACC), the rate‐limiting enzyme in DNL. In a prospective randomized phase 2 study, 100 subjects with likely NASH received GS‐0976 (5 or 20 mg/d) for 12  weeks and were compared with 26 subjects who received ­placebo [125]. GS‐0976 treatment was significantly associated with a dose‐dependent and marked reduction in liver fat ­content, and was generally well‐tolerated, although marked

hypertriglyceridemia was seen in 16% of treated patients. It is still unknown whether GS‐0976 treatment can also lead to histological improvement in NASH.

Other treatments There is a large number of investigative agents currently studied as potential treatments for NASH, targeting various aspects of the disease. Among those are agonists directed at the thyroid hormone receptor β, inhibitors of apoptosis signal‐regulating kinase 1 (ASK‐1), a key mediator of hepatocyte stress response, modulators of gut microbiome, combined agonists of receptors to gut‐derived hormones, an analogue of FGF21, and antifibrotic agents. In addition, there are agents in early stages of development targeting genes that are associated with the disease. Finally, combination treatment with agents targeting different pathways may be needed to achieve meaningful responses in a large proportion of subjects.

CONCLUSION NAFLD has risen to epidemic proportions and directly or indirectly affects the health of many. Increased understanding of genetic and pathogenic mechanisms will hopefully lead to effective and safe therapeutic options.

REFERENCES   1. Chalasani, N., Younossi, Z., Lavine, J.E. et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology, 2018; 67:328–57.  2. Szczepaniak, L.S., Nurenberg, P., Leonard, D. et  al. Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population. Am J Physiol Endocrinol Metab, 2005; 288:E462–8.  3. Kwiterovich, P.O., Jr., Sloan, H.R., and Fredrickson, D.S. Glycolipids and other lipid constituents of normal human liver. J Lipid Res, 1970; 11:322–30.   4. Younossi, Z., Anstee, Q.M., Marietti, M. et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol, 2018; 15:11–20.   5. Clark, J.M., Brancati, F.L., and Diehl, A.M. The prevalence and etiology of elevated aminotransferase levels in the United States. Am J Gastroenterol, 2003; 98:960–7.   6. Ioannou, G.N., Boyko, E.J., and Lee, S.P. The prevalence and predictors of elevated serum aminotransferase activity in the United States in 1999–2002. Am J Gastroenterol, 2006; 101:76–82.   7. Takyar, V., Nath, A., Beri, A. et  al. How healthy are the “Healthy volunteers”? Penetrance of NAFLD in the biomedical research volunteer pool. Hepatology, 2017; 66:825–33.   8. Fracanzani, A.L., Valenti, L., Bugianesi, E. et al. Risk of severe liver disease in nonalcoholic fatty liver disease with normal aminotransferase levels: a role for insulin resistance and diabetes. Hepatology, 2008; 48:792–8.   9. Lazo, M., Hernaez, R., Eberhardt, M.S. et al. Prevalence of nonalcoholic fatty liver disease in the United States: the Third National Health and Nutrition Examination Survey, 1988–1994. Am J Epidemiol, 2013; 178:38–45. 10. Williams, C.D., Stengel, J., Asike, M.I. et  al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle‐ aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology, 2011; 140:124–31. 11. Browning, J.D., Szczepaniak, L.S., Dobbins, R. et al. Prevalence of hepatic steatosis in an urban population in the United States: Impact of ethnicity. Hepatology, 2004; 40:1387–95.



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

12. Younossi, Z.M., Koenig, A.B., Abdelatif, D. et al. Global epidemiology of nonalcoholic fatty liver disease  –  meta‐analytic assessment of prevalence, incidence, and outcomes. Hepatology, 2016; 64:73–84. 13. Welsh, J.A., Karpen, S., and Vos, M.B. Increasing prevalence of nonalcoholic fatty liver disease among United States adolescents, 1988–1994 to 2007–2010. J Pediatr, 2013; 162:496–500. 14. Schwimmer, J.B., Deutsch, R., Kahen, T. et al. Prevalence of fatty liver in children and adolescents. Pediatrics, 2006; 118:1388–93. 15. Lassailly, G., Caiazzo, R., Buob, D. et al. Bariatric surgery reduces features of nonalcoholic steatohepatitis in morbidly obese patients. Gastroenterology, 2015; 149:379–88. 16. Sasaki, A., Nitta, H., Otsuka, K. et al. Bariatric surgery and non‐alcoholic fatty liver disease: current and potential future treatments. Front Endocrinol, 2014; 5:164. 17. Doycheva, I., Patel, N., Peterson, M., and Loomba, R. Prognostic implication of liver histology in patients with nonalcoholic fatty liver disease in diabetes. J Diabetes Complications, 2013; 27:293–300. 18. Anstee, Q.M., Targher, G., and Day, C.P. Progression of NAFLD to diabetes mellitus, cardiovascular disease or cirrhosis. Nat Rev Gastroenterol Hepatol, 2013; 10:330–44. 19. Marchesini, G., Bugianesi, E., Forlani, G. et  al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology, 2003; 37:917–23. 20. Zelber‐Sagi, S., Lotan, R., Shlomai, A. et  al. Predictors for incidence and remission of NAFLD in the general population during a seven‐year prospective follow‐up. J Hepatol, 2012; 56:1145–51. 21. Ma, J., Hwang, S.J., Pedley, A. et al. Bi‐directional analysis between fatty liver and cardiovascular disease risk factors. J Hepatol, 2017; 66:390–7. 22. Kojima, S.‐I., Watanabe, N., Numata, M. et al. Increase in the prevalence of fatty liver in Japan over the past 12 years: analysis of clinical background. J Gastroenterol, 2003; 38:954–61. 23. Loomba, R., Schork, N., Chen, C.H. et al. Heritability of Hepatic Fibrosis and Steatosis Based on a Prospective Twin Study. Gastroenterology, 2015; 149:1784–93. 24. Schwimmer, J.B., Celedon, M.A., Lavine, J.E. et al. Heritability of nonalcoholic fatty liver disease. Gastroenterology, 2009; 136:1585–92. 25. Romeo, S., Kozlitina, J., Xing, C. et al. Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease. Nature Genet, 2008; 40:1461–5. 26. Yuan, X., Waterworth, D., Perry, J.R.B. et  al. Population‐based genome‐ wide association studies reveal six loci influencing plasma levels of liver enzymes. Am J Hum Genet, 2008; 83:520–8. 27. Rotman, Y., Koh, C., Zmuda, J.M. et al. The association of genetic variability in patatin‐like phospholipase domain‐containing protein 3 (PNPLA3) with histological severity of nonalcoholic fatty liver disease. Hepatology, 2010; 52:894–903. 28. Speliotes, E.K., Butler, J.L., Palmer, C.D. et al. PNPLA3 variants specifically confer increased risk for histologic nonalcoholic fatty liver disease but not metabolic disease. Hepatology, 2010; 52:904–12. 29. Hassan, M.M., Kaseb, A., Etzel, C.J. et al. Genetic variation in the PNPLA3 gene and hepatocellular carcinoma in USA: risk and prognosis prediction. Mol Carcinog, 2013; 52: Suppl 1:E139–47. 30. Tian, C., Stokowski, R.P., Kershenobich, D. et  al. Variant in PNPLA3 is associated with alcoholic liver disease. Nature Genet, 2010; 42:21–3. 31. Basuray, S., Smagris, E., Cohen, J.C., and Hobbs, H.H. The PNPLA3 variant associated with fatty liver disease (I148M) accumulates on lipid droplets by evading ubiquitylation. Hepatology, 2017; 66:1111–24. 32. Mitsche, M.A., Hobbs, H.H., and Cohen, J.C. Patatin‐like phospholipase domain‐containing protein 3 promotes transfers of essential fatty acids from triglycerides to phospholipids in hepatic lipid droplets. J Biol Chem, 2018; 293:6958–68. 33. Pirazzi, C., Valenti, L., Motta, B.M. et al. PNPLA3 has retinyl‐palmitate lipase activity in human hepatic stellate cells. Hum Mol Genet, 2014; 23:4077–85. 34. Kozlitina, J., Smagris, E., Stender, S. et al. Exome‐wide association study identifies a TM6SF2 variant that confers susceptibility to nonalcoholic fatty liver disease. Nat Genet, 2014; 46:352–6. 35. Holmen, O.L., Zhang, H., Fan, Y. et  al. Systematic evaluation of coding ­variation identifies a candidate causal variant in TM6SF2 influencing total cholesterol and myocardial infarction risk. Nat Genet, 2014; 46:345–51. 36. Mahdessian, H., Taxiarchis, A., Popov, S. et al. TM6SF2 is a regulator of liver fat metabolism influencing triglyceride secretion and hepatic lipid droplet content. Proc Nat Acad Sci USA, 2014; 111:8913–8.

679

37. Smagris, E., Gilyard, S., Basuray, S. et al. Inactivation of Tm6sf2, a gene defective in fatty liver disease, impairs lipidation but not secretion of very low density lipoproteins. J Biol Chem, 2016; 291:10659–76. 38. Abul‐Husn, N.S., Cheng, X., Li, A.H. et al. A protein‐truncating HSD17B13 variant and protection from chronic liver disease. New Eng J Med, 2018; 378:1096–106. 39. Ma, Y., Belyaeva, O.V., Brown, P.M. et al. HSD17B13 is a hepatic retinol dehydrogenase associated with histological features of non‐alcoholic fatty liver disease. Hepatology, 2019;69(4):1504–19. 40. Kozlitina, J., Stender, S., Hobbs, H.H., and Cohen, J.C. HSD17B13 and chronic liver disease in Blacks and Hispanics. N Engl J Med, 2018; 379:1876–7. 41. Eslam, M., Valenti, L., and Romeo, S. Genetics and epigenetics of NAFLD and NASH: clinical impact. J Hepatol, 2018; 68:268–79. 42. Pirola, C.J., Gianotti, T.F., Burgueno, A.L. et al. Epigenetic modification of liver mitochondrial DNA is associated with histological severity of nonalcoholic fatty liver disease. Gut, 2013; 62:1356–63. 43. Donnelly, K.L., Smith, C.I., Schwarzenberg, S.J. et al. Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J Clin Invest, 2005; 115:1343–51. 44. Sanyal, A.J., Campbell‐Sargent, C., Mirshahi, F. et al. Nonalcoholic steatohepatitis: association of insulin resistance and mitochondrial abnormalities. Gastroenterology, 2001; 120:1183–92. 45. Bril, F., Barb, D., Portillo‐Sanchez, P. et al. Metabolic and histological implications of intrahepatic triglyceride content in nonalcoholic fatty liver disease. Hepatology, 2017; 65:1132–44. 46. Musso, G., Cassader, M., De Michieli, F. et al. Nonalcoholic steatohepatitis versus steatosis: Adipose tissue insulin resistance and dysfunctional response to fat ingestion predict liver injury and altered glucose and lipoprotein metabolism. Hepatology, 2012; 56:933–42. 47. Schwarz, J.M., Neese, R.A., Turner, S. et al. Short‐term alterations in carbohydrate energy intake in humans. Striking effects on hepatic glucose production, de novo lipogenesis, lipolysis, and whole‐body fuel selection. J Clin Invest, 1995; 96:2735–43. 48. Lambert, J.E., Ramos‐Roman, M.A., Browning, J.D., and Parks, E.J. Increased de novo lipogenesis is a distinct characteristic of individuals with nonalcoholic fatty liver disease. Gastroenterology, 2014; 146:726–35. 49. Brown, M.S. and Goldstein, J.L. Selective versus total insulin resistance: a pathogenic paradox. Cell Metab, 2008; 7:95–6. 50. Otero, Y.F., Stafford, J.M., and Mcguinness, O.P. Pathway‐selective insulin resistance and metabolic disease: the importance of nutrient flux. J Biol Chem, 2014; 289:20462–9. 51. Fabbrini, E., Mohammed, B.S., Magkos, F. et al. Alterations in adipose tissue and hepatic lipid kinetics in obese men and women with nonalcoholic fatty liver disease. Gastroenterology, 2008; 134:424–31. 52. Iozzo, P., Bucci, M., Roivainen, A. et al. Fatty acid metabolism in the liver, measured by positron emission tomography, is increased in obese individuals. Gastroenterology, 2010; 139:846–56, 56 e1–6. 53. Naguib, G., Morris, N., Haynes‐Williams, V. et  al. Beta‐oxidation is decreased in non‐alcoholic fatty liver disease: a non‐invasive assessment utilizing palmitate and acetate breath tests. Hepatology, 2017; 66:1043. 54. Day, C.P. and James, O.F.W. Steatohepatitis: a tale of two “hits”? Gastroenterology, 1998; 114:842–5. 55. Neuschwander‐Tetri, B.A. Hepatic lipotoxicity and the pathogenesis of nonalcoholic steatohepatitis: the central role of nontriglyceride fatty acid metabolites. Hepatology, 2010; 52:774–88. 56. Yamaguchi, K., Yang, L., Mccall, S. et  al. Inhibiting triglyceride synthesis  improves hepatic steatosis but exacerbates liver damage and fibrosis in  obese mice with nonalcoholic steatohepatitis. Hepatology, 2007; 45:1366–74. 57. Videla, L.A., Rodrigo, R., Orellana, M. et al. Oxidative stress‐related parameters in the liver of non‐alcoholic fatty liver disease patients. Clin Sci (Lond), 2004; 106:261–8. 58. Sunny, N.E., Parks, E.J., Browning, J.D., and Burgess, S.C. Excessive hepatic mitochondrial TCA cycle and gluconeogenesis in humans with nonalcoholic fatty liver disease. Cell Metab, 2011; 14:804–10. 59. Koliaki, C., Szendroedi, J., Kaul, K. et al. Adaptation of hepatic mitochondrial function in humans with non‐alcoholic fatty liver is lost in steatohepatitis. Cell Metab, 2015; 21:739–46. 60. Bellanti, F., Villani, R., Facciorusso, A. et al. Lipid oxidation products in the pathogenesis of non‐alcoholic steatohepatitis. Free Radic Biol Med, 2017; 111:173–85.

680

THE LIVER:  REFERENCES

61. Seki, S., Kitada, T., Yamada, T. et al. In situ detection of lipid peroxidation and oxidative DNA damage in non‐alcoholic fatty liver diseases. J Hepatol, 2002; 37:56–62. 62. Puri, P., Mirshahi, F., Cheung, O. et al. Activation and dysregulation of the unfolded protein response in nonalcoholic fatty liver disease. Gastroenterology, 2008; 134:568–76. 63. Lebeaupin, C., Vallee, D., Hazari, Y. et al. Endoplasmic reticulum stress signalling and the pathogenesis of non‐alcoholic fatty liver disease. J Hepatol, 2018; 69:927–47. 64. Feldstein, A.E., Canbay, A., Angulo, P. et al. Hepatocyte apoptosis and fas expression are prominent features of human nonalcoholic steatohepatitis. Gastroenterology, 2003; 125:437–43. 65. Ibrahim, S.H., Hirsova, P., and Gores, G.J. Non‐alcoholic steatohepatitis pathogenesis: sublethal hepatocyte injury as a driver of liver inflammation. Gut, 2018; 67:963–72. 66. Rangwala, F., Guy, C.D., Lu, J. et al. Increased production of sonic hedgehog by ballooned hepatocytes. J Pathol, 2011; 224:401–10. 67. Garcia‐Martinez, I., Santoro, N., Chen, Y. et al. Hepatocyte mitochondrial DNA drives nonalcoholic steatohepatitis by activation of TLR9. J Clin Invest, 2016; 126:859–64. 68. Arrese, M., Cabrera, D., Kalergis, A.M., and Feldstein, A.E. Innate immunity and inflammation in NAFLD/NASH. Dig Dis Sci, 2016; 61:1294–303. 69. Schuster, S., Cabrera, D., Arrese, M., and Feldstein, A.E. Triggering and resolution of inflammation in NASH. Nat Rev Gastroenterol Hepatol, 2018; 15:349–64. 70. Xu, R., Huang, H., Zhang, Z., and Wang, F.S. The role of neutrophils in the development of liver diseases. Cell Mol Immunol, 2014; 11:224–31. 71. Byun, J.S. and Yi, H.S. Hepatic immune microenvironment in alcoholic and nonalcoholic liver disease. Biomed Res Int, 2017; 2017:6862439. 72. Sharpton, S.R., Ajmera, V., and Loomba, R. Emerging role of the gut microbiome in nonalcoholic fatty liver disease: from composition to function. Clin Gastroenterol Hepatol.2019;17(2):296–306. 73. Le Roy, T., Llopis, M., Lepage, P. et  al. Intestinal microbiota determines development of non‐alcoholic fatty liver disease in mice. Gut, 2013; 62:1787–94. 74. Henao‐Mejia, J., Elinav, E., Jin, C. et al. Inflammasome‐mediated dysbiosis regulates progression of NAFLD and obesity. Nature, 2012; 482:179–85. 75. Chu, H., Duan, Y., Yang, L., and Schnabl, B. Small metabolites, possible big changes: a microbiota‐centered view of non‐alcoholic fatty liver disease. Gut, 2019;68(2):359–70. 76. Jiang, C., Xie, C., Li, F. et al. Intestinal farnesoid X receptor signaling promotes nonalcoholic fatty liver disease. J Clin Invest, 2015; 125:386–402. 77. Ma, C., Han, M., Heinrich, B. et  al. Gut microbiome‐mediated bile acid metabolism regulates liver cancer via NKT cells. Science, 2018; 360:(6391). 78. Vilar‐Gomez, E., Calzadilla‐Bertot, L., Wai‐Sun Wong, V. et al. Fibrosis severity as a determinant of cause‐specific mortality in patients with advanced nonalcoholic fatty liver disease: a multi‐national cohort study. Gastroenterology, 2018; 155:443–57 e17. 79. Angulo, P., Kleiner, D.E., Dam‐Larsen, S. et al. Liver fibrosis, but no other histologic features, is associated with long‐term outcomes of patients with nonalcoholic fatty liver disease. Gastroenterology, 2015; 149:389–97. 80. Dulai, P.S., Singh, S., Patel, J. et al. Increased risk of mortality by fibrosis stage in nonalcoholic fatty liver disease: systematic review and meta‐analysis. Hepatology, 2017; 65:1557–65. 81. Ekstedt, M., Hagström, H., Nasr, P. et al. Fibrosis stage is the strongest predictor for disease‐specific mortality in NAFLD after up to 33 years of follow‐up. Hepatology, 2015; 61:1547–54. 82. Ekstedt, M., Franzén, L.E., Mathiesen, U.L. et al. Long‐term follow‐up of patients with NAFLD and elevated liver enzymes. Hepatology, 2006; 44:865–73. 83. Singh, S., Allen, A.M., Wang, Z. et al. Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis: a systematic review and meta‐­ analysis of paired‐biopsy studies. Clin Gastroenterol Hepatol, 2015; 13: 643–54.e1–9; quiz e39–40. 84. Argo, C.K., Northup, P.G., Al‐Osaimi, A.M.S., and Caldwell, S.H. Systematic review of risk factors for fibrosis progression in non‐alcoholic steatohepatitis. J Hepatol, 2009; 51:371–9. 85. Younossi, Z.M., Otgonsuren, M., Henry, L. et  al. Association of nonalcoholic fatty liver disease (NAFLD) with hepatocellular carcinoma (HCC) in the United States from 2004 to 2009. Hepatology, 2015; 62:1723–30.

  86. Mittal, S., El‐Serag, H.B., Sada, Y.H. et al. Hepatocellular carcinoma in the absence of cirrhosis in United States veterans is associated with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol, 2016; 14:124–31 e1.   87. Pais, R., Giral, P., Khan, J.F. et al. Fatty liver is an independent predictor of early carotid atherosclerosis. J Hepatol, 2016; 65:95–102.   88. Targher, G., Byrne, C.D., Lonardo, A. et al. Non‐alcoholic fatty liver disease and risk of incident cardiovascular disease: a meta‐analysis. J Hepatol, 2016; 65:589–600.   89. Söderberg, C., Stål, P., Askling, J. et al. Decreased survival of subjects with elevated liver function tests during a 28‐year follow‐up. Hepatology, 2010; 51:595–602.   90. Adams, L.A., Anstee, Q.M., Tilg, H., and Targher, G. Non‐alcoholic fatty liver disease and its relationship with cardiovascular disease and other extrahepatic diseases. Gut, 2017; 66:1138–53.   91. Ahn, J.S., Sinn, D.H., Min, Y.W. et al. Non‐alcoholic fatty liver diseases and risk of colorectal neoplasia. Aliment Pharmacol Ther, 2017; 45:345–53.   92. Kim, G.A., Lee, H.C., Choe, J. et al. Association between non‐alcoholic fatty liver disease and cancer incidence rate. J Hepatol.2017;  93. Hicks, S.B., Mara, K., Larson, J.J. et  al. The incidence of extrahepatic malignancies in nonalcoholic fatty liver disease (NAFLD). Hepatology, 2018; 68:20A–A.   94. Musso, G., Gambino, R., Tabibian, J.H. et al. Association of non‐alcoholic fatty liver disease with chronic kidney disease: a systematic review and meta‐analysis. PLoS Med, 2014; 11:e1001680.   95. Kleiner, D.E., Brunt, E.M., Van Natta, M. et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology, 2005; 41:1313–21.   96. Brunt, E.M., Kleiner, D.E., Wilson, L.A. et al. Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology, 2011; 53:810–20.   97. Takyar, V., Etzion, O., Heller, T. et al. Complications of percutaneous liver biopsy with Klatskin needles: a 36‐year single‐centre experience. Aliment Pharmacol Ther, 2017; 45:744–53.  98. Zhang, Y.N., Fowler, K.J., Hamilton, G. et  al. Liver fat imaging‐a clinical overview of ultrasound, CT, and MR imaging. Br J Radiol, 2018; 91:20170959.   99. Hernaez, R., Lazo, M., Bonekamp, S. et al. Diagnostic accuracy and reliability of ultrasonography for the detection of fatty liver: a meta‐analysis. Hepatology, 2011; 54:1082–90. 100. Karlas, T., Petroff, D., Sasso, M. et al. Individual patient data meta‐analysis of controlled attenuation parameter (CAP) technology for assessing steatosis. J Hepatol, 2017; 66:1022–30. 101. Noureddin, M., Lam, J., Peterson, M.R. et al. Utility of magnetic resonance imaging versus histology for quantifying changes in liver fat in nonalcoholic fatty liver disease trials. Hepatology (Baltimore, Md), 2013; 58:1930–40. 102. Wong, V.W., Adams, L.A., De Ledinghen, V. et al. Noninvasive biomarkers in NAFLD and NASH ‐ current progress and future promise. Nat Rev Gastroenterol Hepatol, 2018; 15:461–78. 103. Kwok, R., Tse, Y.K., Wong, G.L. et al. Systematic review with meta‐analysis: non‐invasive assessment of non‐alcoholic fatty liver disease‐‐the role of transient elastography and plasma cytokeratin‐18 fragments. Aliment Pharmacol Ther, 2014; 39:254–69. 104. Feldstein, A.E., Lopez, R., Tamimi, T.A. et al. Mass spectrometric profiling of oxidized lipid products in human nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. J Lipid Res, 2010; 51:3046–54. 105. Younossi, Z.M., Loomba, R., Anstee, Q.M. et al. Diagnostic modalities for nonalcoholic fatty liver disease, nonalcoholic steatohepatitis, and associated fibrosis. Hepatology, 2018; 68:349–60. 106. Hannah, W.N., Jr. and Harrison, S.A. Noninvasive imaging methods to determine severity of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Hepatology, 2016; 64:2234–43. 107. Rotman, Y. and Sanyal, A.J. Current and upcoming pharmacotherapy for non‐alcoholic fatty liver disease. Gut, 2017; 66:180–90. 108. Promrat, K., Kleiner, D.E., Niemeier, H.M. et al. Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis. Hepatology, 2010; 51:121–9. 109. Vilar‐Gomez, E., Martinez‐Perez, Y., Calzadilla‐Bertot, L. et  al. Weight loss through lifestyle modification significantly reduces features of nonalcoholic steatohepatitis. Gastroenterology, 2015; 149:367–78.e5. 110. Han, M.a.T., Altayar, O., Hamdeh, S. et al. Rates of and factors associated with placebo response in trials of pharmacotherapies for nonalcoholic



52:  Non‐alcoholic Fatty Liver Disease: Mechanisms and Treatment

s­teatohepatitis: systematic review and meta‐analysis. Clin Gastroenterol Hepatol, 2010;17(4):616–29. 111. Lassailly, G., Caiazzo, R., Gnemmi, V. et al. Regression of fibrosis after disappearance of nash in morbidly obese patients: a prospective bariatric surgery cohort with sequential liver biopsies. Hepatology, 2018; 68:44a‐5a. 112. Sanyal, A.J., Chalasani, N., Kowdley, K.V. et al. Pioglitazone, vitamin E, or placebo for nonalcoholic steatohepatitis. New Eng J Med, 2010; 362:1675–85. 113. Cusi, K., Orsak, B., Bril, F. et  al. Long‐term pioglitazone treatment for patients with nonalcoholic steatohepatitis and prediabetes or type 2 diabetes mellitus. Ann Int Med, 2016; 165:305. 114. Ratziu, V., Harrison, S.A., Francque, S. et al. Elafibranor, an agonist of the peroxisome proliferator—activated receptor—α and −δ, induces resolution of nonalcoholic steatohepatitis without fibrosis worsening. Gastroenterology, 2016; 150:1147–59.s 115. Neuschwander‐Tetri, B.A., Loomba, R., Sanyal, A.J. et  al. Farnesoid X nuclear receptor ligand obeticholic acid for non‐cirrhotic, non‐alcoholic steatohepatitis (FLINT): a multicentre, randomised, placebo‐controlled trial. Lancet, 2015; 385:956–65. 116. Harrison, S.A., Rinella, M.E., Abdelmalek, M.F. et al. NGM282 for treatment of non‐alcoholic steatohepatitis: a multicentre, randomised, double‐ blind, placebo‐controlled, phase 2 trial. Lancet, 2018; 391:1174–85. 117. Harrison, S.A., Trotter, J.F., Paredes, A.H. et al. NGM282 rapidly improves NAFLD activity score (NAS) and fibrosis in 12 weeks in patients with biopsy‐confirmed nonalcoholic steatohepatitis (NASH): results of a phase 2 multi‐center dose finding study. Hepatology, 2018; 68:66a‐a.

681

118. Lavine, J.E., Schwimmer, J.B., Van Natta, M.L. et al. Effect of vitamin E or metformin for treatment of nonalcoholic fatty liver disease in children and adolescents. JAMA, 2011; 305:1659. 119. Vilar‐Gomez, E., Vuppalanchi, R., Gawrieh, S. et al. Vitamin E improves transplant‐free survival and hepatic decompensation among patients with NASH and advanced fibrosis. Hepatology.2018. 120. Schürks, M., Glynn, R.J., Rist, P.M. et al. Effects of vitamin E on stroke subtypes: meta‐analysis of randomised controlled trials. BMJ, 2010; 341:c5702. 121. Klein, E.A., Thompson, I.M., Jr., Tangen, C.M. et al. Vitamin E and the risk of prostate cancer: the selenium and vitamin E cancer prevention trial (SELECT). JAMA, 2011; 306:1549–56. 122. Rotman, Y., Lingala, S., and Morris, N. Subtle iron deficiency is common during vitamin E treatment for NAFLD. Hepatology, 2016; 64:571a‐2a. 123. Shao, N., Kuang, H.Y., Hao, M. et al. Benefits of exenatide on obesity and non‐alcoholic fatty liver disease with elevated liver enzymes in patients with type 2 diabetes. Diabetes Metab Res Rev, 2014; 30:521–9. 124. Armstrong, M.J., Gaunt, P., Aithal, G.P. et al. Liraglutide safety and efficacy in patients with non‐alcoholic steatohepatitis (LEAN): a multicentre, double‐blind, randomised, placebo‐controlled phase 2 study. Lancet, 2016; 387:679–90. 125. Loomba, R., Kayali, Z., Noureddin, M. et al. GS‐0976 reduces hepatic steatosis and fibrosis markers in patients with nonalcoholic fatty liver disease. Gastroenterology, 2018; 155:1463–73 e6.

53

Alcoholic Liver Disease Bin Gao1, Xiaogang Xiang1,2, Lorenzo Leggio3, and George F. Koob4 Laboratory of Liver Diseases, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD, USA 2 Department of Infectious Diseases, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China 3 Section on Clinical Psychoneuroendocrinology and Neuropsychopharmacology, National Institute on Alcohol Abuse and Alcoholism and National Institute on Drug Abuse, National Institutes of Health, Bethesda, MD, USA 4 National Institute on Alcohol Abuse and Alcoholism and National Institute on Drug Abuse, National Institutes of Health, Bethesda, MD, USA 1

Excessive alcohol consumption leads a spectrum of liver disorders from simple steatosis (fatty liver) to severe forms of liver injury including steatohepatitis, cirrhosis, and hepatocellular carcinoma [1, 2]. Many proinflammatory mediators, metabolic pathways, transcriptional factors, and epigenetic factors have been identified to contribute to the development and progression of alcoholic liver disease (ALD), which were discussed in detail in the previous edition of this book. In this chapter, we briefly discuss ethanol metabolism, ALD pathogenesis, clinical diagnosis and treatment of ALD, and in more detail recent advances in the pathogenesis and treatment of ALD. We also discuss the neurobiology of alcohol use disorder (AUD) and treatment of this disorder in patients with ALD.

ETHANOL METABOLISM The liver is the major organ for converting ethanol into acetaldehyde via the three enzymatic pathways [3]. The liver expresses high levels of alcohol dehydrogenase (ADH), which is the most important enzyme for converting ethanol to acetaldehyde. Human ADH genes encode at least seven isozymes including ADH1–7 with the major forms of ADH1, ADH2, and ADH3 in human liver. ADH exists in cytosol and oxidizes ethanol into acetaldehyde and reduces NAD+ to NADH, which generates a highly reduced cytosolic environment in cells. Normal liver also expresses high levels of the cytochrome P450 isozyme CYP2E1, which is located in the endoplasmic reticulum and is  highly induced after chronic ethanol consumption. CYP2E1 can oxidize ethanol into acetaldehyde with conversion of NADPH+H++O2 to NADP++2H2O, thereby generating reactive

oxygen species (ROS) and free radicals that induce lipid peroxidation to promote adduct formation with several macromolecules. Finally, another enzyme that can oxidize ethanol is catalase, which is located in the peroxisome. However, this pathway is considered a minor pathway of ethanol oxidation in the liver. The second step of ethanol metabolism is to convert acetaldehyde into acetate by aldehyde dehydrogenase (ALDH) [3]. ALDH also exists in multiple forms with ALDH2 being a key isoform to metabolize acetaldehyde. ALDH2 is expressed at the highest levels in the liver and is located in the mitochondria. Acetate generated from acetaldehyde metabolism in hepatocytes is rapidly secreted into the circulation and is further converted into acetyl coenzyme A (acetyl‐CoA) by acetyl‐ CoA synthetases. Acetyl‐CoA then enters into the citric acid cycle (Krebs cycle) and is finally converted to carbon dioxide and water. Human ADH and ALDH2 genes exist as polymorphisms with the most relevant ALDH2 variants including the ALDH2*1 and ALDH2*2 alleles. The ALDH2*1 allele encodes active acetaldehyde metabolizing enzyme with G at nucleotide position 42421 with corresponding glutamate at amino acid position 487; whereas the ALDH2*2 allele encodes inactive enzyme with A at position 42421 and lysine at position 487. People with homozygous ALDH2*1/1 have high levels of acetaldehyde metabolizing activity in the liver; whereas individuals with homozygous ALDH2*2/2 have very low or undetectable enzyme activities. Individuals with heterozygous ALDH2*2/1 have approximately 80–90% lower enzyme activity because ALDH2 is the isotetramer enzyme and all four subunits of ALDH2 are needed to be normal for keeping its full activity. Individuals with the inactive variants of ALDH2 exhibit acetaldehyde accumulation after

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



53:  Alcoholic Liver Disease

alcohol consumption, and present an acetaldehyde‐mediated “flushing syndrome,” which includes facial flushing, palpitations, drowsiness, and other unpleasant symptoms.

SPECTRUM OF ALD ALD has a broad spectrum of disorders from simple fatty liver (steatosis) to severe forms of steatohepatitis, cirrhosis, and hepatocellular carcinoma (HCC) [1, 2]. Fatty liver occurs in almost all chronic heavy drinkers with early fat accumulation in zone 3 (perivenular) hepatocytes, which can extend to zone 2 and even zone 1 (periportal) hepatocytes when liver damage becomes severe. Steatohepatitis (steatosis and inflammation) and cirrhosis are detected in approximately 20–40% and 8–20% of chronic heavy drinkers, respectively. Chronic excessive drinking can also cause HCC, which occurs mostly after cirrhosis in approximately 3–10% of excessive drinkers. Many patients with chronic ALD are asymptomatic with fully compensated and preserved liver function, but some of them may develop episodes of superimposed alcoholic hepatitis (AH) with obvious jaundice, fever, abdominal pain, anorexia, and weight loss [4]. AH has high short‐term ­mortality especially in those with cirrhosis [4].

GENETIC FACTORS, COMORBIDITY, AND ALD Although almost all heavy drinkers develop fatty liver, only about 35% of them develop advanced ALD, which is probably because other risk factors are involved in the pathogenesis of ALD. Indeed, gender, obesity, dietary factors, drinking patterns, non‐sex‐linked genetic factors, and cigarette smoking have been suggested to affect susceptibility to ALD.

Gender Female sex is a well‐documented risk factor for susceptibility to ALD in rodent models. It is believed that the increased risk among women is probably due to lower levels of gastric ADH, a higher proportion of body fat, and the presence of estrogens.

Obesity and dietary factors Obesity is another important risk factor that accelerates ALD development and progression in alcohol use disorder (AUD). Experimental studies indicate that acute or chronic ethanol feeding and obesity synergistically induced liver injury and inflammation via the activation of endoplasmic reticulum stress, hepatic macrophage and neutrophilic infiltration, and lipotoxicity. For example, acute ethanol gavage induces acute steatohepatitis in obese mice induced by high‐fat diet feeding but not in chow‐fed mice [5]. Acute ethanol gavage upregulates the expression of chemokine (C‐X‐C motif) ligand 1 (CXCL1) mRNA in the liver but not in other organs by 30‐fold in obese mice, which attracts neutrophil accumulation in the liver and subsequently causes liver injury [5]. In addition, intragastric overfeeding of mice with

683

a high-fat diet and coadministration of ­ethanol synergistically induced an elevation of serum alanine aminotransferase (ALT) levels, hepatic steatosis, liver inflammation, and fibrosis via the induction of nitrosative, endoplasmic reticulum, and mitochondrial stress and upregulation of hepatic Toll‐like receptor 4 [6]. The type of fat (saturated versus unsaturated) may also affect the outcome of ALD. Saturated fat has been shown to be protective against ALD, while unsaturated fat is deleterious for ALD in animal models [7]. However, how dietary factors affect ALD  in humans is probably complex and remains poorly characterized.

Drinking patterns During the last five years, one of the major advances in the field of ALD research was the discovery that acute ethanol gavage caused significant neutrophilia, liver injury, and inflammation in chronically ethanol‐fed mice [8–11] and in high‐fat diet‐fed mice [5]. In agreement with these findings from animal models, clinical studies show that individuals with AUD with recent excessive alcohol consumption have higher levels of circulating neutrophils, serum alanine transaminase (ALT), and aspartate transaminase (AST) compared with individuals with AUD without recent drinking [12]. The levels of circulating neutrophils correlate positively with serum ALT and AST in individuals with AUD with recent excessive drinking, suggesting that recent excessive drinking elevates circulating neutrophils and subsequently induces liver injury [12]. In addition, early studies suggest that frequent heavy drinking that begins at an early age is a risk factor for the development of severe forms of ALD compared with episodic or binge drinking [13]. However, a recent prospective cohort study suggests that recent alcohol drinking rather than earlier in life is associated with risk of alcoholic ­cirrhosis [14].

Non‐sex‐linked genetic factors Many non‐sex‐linked genetic factors have been implicated in affecting the susceptibility to advanced ALD. These include variations in genes that encode antioxidant enzymes, cytokines and other inflammatory mediators, and alcohol‐metabolizing enzymes. However, the evidence that links most of these genetic factors to ALD is weak, only a few of these factors seem to be strongly associated with ALD. For example, a single‐nucleotide variant (I148M) in human patatin‐like phospholipase domain‐ containing 3 gene (PNPLA3), known as I148M variant, is one of the best characterized variants and has strong correlations with increased risk of steatosis, fibrosis, and cirrhosis in various types of liver diseases including ALD [15].

Viral hepatitis Viral hepatitis infection is a leading cause of chronic liver disease worldwide. The synergistic effects of excessive alcohol drinking and viral hepatitis B or C infection on liver disease progression have been well documented. It has been reported that excessive alcohol drinking markedly accelerates liver fibrosis and HCC development and progression in patients with viral hepatitis C infection.

684

THE LIVER:  ALCOHOLIC STEATOHEPATITIS (ASH) AND ALCOHOLIC HEPATITIS (AH)

Non‐alcoholic fatty liver disease (NAFLD) Excessive alcohol consumption likely worsens NAFLD progression; while the reports regarding the effects of moderate alcohol drinking on NAFLD have been controversial [16]. Current clinical data are not sufficient to support a recommendation for or against moderate alcohol consumption in NAFLD [16].

Other comorbidities Long‐term alcohol consumption may also accelerate liver disease progression in patients with human immunodeficiency virus infection, hemochromatosis, and so on. A greater understanding of the interaction between alcohol and these comorbid factors could help us design better therapies for the treatment of chronic liver disease.

ALCOHOLIC FATTY LIVER Fat accumulation in the liver (fatty liver) is the earliest response of the liver to acute and chronic alcohol drinking [17]. Fatty liver is diagnosed when the liver fat weight is greater than 5 to 10% of the liver’s weight and is characterized by the accumulation of triglycerides, phospholipids, cholesterol esters, and other types of fat in hepatocytes. Although alcohol drinking causes fat accumulation in the liver, there is no evidence that ethanol and its metabolite acetaldehyde directly participate in biosynthesis of fatty acids (FAs). Acetate, the metabolite of acetaldehyde, can be converted to acetyl‐CoA by acetyl‐CoA synthetases with acetyl‐CoA synthetase 2. Once it is formed, acetyl‐CoA enters the citric acid cycle (Krebs cycle) and contributes to fatty acid biosynthesis. However, how much this acetate–acetyl‐CoA pathway directly contributes to the formation of alcoholic fatty liver is not clear because acetate derived from ethanol metabolism in hepatocytes is rapidly secreted into the circulation and does not maintain high concentrations in hepatocytes, and the liver does not express the mitochondrial acetyl‐CoA synthetase 2 that is a major form for converting acetate into acetyl‐CoA. Emerging evidence suggests that alcohol consumption promotes fat mobilization to the liver and de novo FA biosynthesis but inhibits fat clearance and FA β‐oxidation in the liver, resulting in fat accumulation in the liver [1, 17]. Early studies demonstrated that acute alcohol administration can increase the supply of dietary lipids to the liver from the small intestine by increasing the intestinal lymph flow and output of both dietary and nondietary lipids, but this effect was less evident after chronic alcohol treatment. Recent studies focused on how alcohol consumption increases the supply of lipids from adipose tissues to the liver [18]. It has been shown that alcohol consumption induces lipolysis [19] and adipocyte death [20] and subsequently elevates circulating FAs and their accumulation in the liver. In addition to augmenting FA supply, alcohol can also increase fat accumulation by promoting lipid formation and stimulating de novo FA biosynthesis in hepatocytes via the upregulation of fat‐specific protein 27 (FSP27) and sterol  ­regulatory element‐binding protein 1c (SREBP‐1c), respectively [1, 9, 17]. FSP27 is expressed

at high levels in adipose tissues, playing a critical role in promoting lipid droplet formation. In normal liver, FSP27 is expressed at very low levels but highly upregulated post alcohol consumption, especially acute alcohol consumption [9]. Upregulation of hepatic FSP27 by acute ethanol consumption is attributable to activation of hepatic endoplasmic reticulum stress, which activates the cyclic AMP‐responsive element‐binding protein H (CREBH) and subsequent nuclear translocation of nCREBH [9]. The important role of FSP27 in inducing alcoholic fatty liver is supported by the fact that disruption of the Fsp27 gene prevents the development of alcoholic fatty liver [9]. SREBP‐1c is a master transcription factor for controlling hepatic expression of many lipogenic genes that encode proteins and enzymes to induce FA biosynthesis. Disruption of the Srebp‐1c gene markedly prevents alcoholic fatty liver in mice, suggesting a critical role of SREBP‐1c in inducing alcoholic fatty liver [21]. Alcohol consumption can directly upregulate hepatic SREBP‐1c expression via its metabolite acetaldehyde [22], or indirectly stimulate its expression by modulating multiple factors and pathways that control hepatic expression of SREBP‐1 (see references therein [1, 17]). In contrast to promotion of FA biosynthesis, alcohol attenuates FA β‐oxidation, which is another critical mechanism responsible for the formation of alcoholic fatty liver. Early studies revealed that alcohol treatment increases the ratio of reduced nicotinamide adenine dinucleotide/oxidized nicotinamide adenine dinucleotide in the liver and consequently disrupts the mitochondrial β‐oxidation of FAs and results in fat accumulation in hepatocytes [23]. Recent studies unveiled an additional mechanism by which ethanol inhibits FA β‐oxidation, which is inactivation of peroxisome proliferator‐activated receptor (PPAR)‐α, a nuclear hormone receptor that upregulates expression of a range of genes involved in FA transport and oxidation [24]. Alcohol consumption can also promote hepatic fat accumulation by inhibiting 5’ AMP‐activated protein kinase (AMPK) and deregulating autophagy. AMPK plays a critical role in controlling the activity of several enzymes and transcriptional factors that regulate fat metabolism [25]. For example, AMPK inhibits acetyl‐CoA carboxylase (ACC) and SREBP activity and consequently attenuates FA biosynthesis but enhances its oxidation [25]. Alcohol exposure can suppress AMPK activity and subsequently enhance FA synthesis and reduces FA oxidation, resulting in alcoholic fatty liver [26]. Autophagy plays an important role in clearing lipid droplets in hepatocytes, and chronic alcohol consumption inhibits autophagy, thereby reducing lipid clearance and causing hepatic steatosis [27]. However, acute alcohol intake seems to activate autophagy, which may play a compensatory role in preventing the development of fatty liver during the early stages of alcoholic liver injury [28].

ALCOHOLIC STEATOHEPATITIS (ASH) AND ALCOHOLIC HEPATITIS (AH) Accumulation of fat in hepatocytes is the first response of the liver to alcohol consumption [17]; however, alcohol consumption may also cause hepatocyte damage and inflammation, which is defined as ASH [4]. ASH is a histological concept that



53:  Alcoholic Liver Disease

is characterized by steatosis, hepatocyte death, ballooning degeneration, neutrophilic infiltration, Mallory–Denk bodies, and chicken‐wire fibrosis [4]. Most patients with ASH are asymptomatic for prolonged periods of time and are defined as walking ASH or subclinical ASH, but some patients with ASH develop a significant clinical syndrome, which is defined as AH (see New therapeutic targets for AH) [4]. Over the last four decades, multiple mechanisms have been identified to contribute to ASH caused by alcohol drinking (Figure  53.1). First, alcohol and its metabolite acetaldehyde can directly cause hepatotoxicity by generating ROS and inducing endoplasmic reticulum stress and mitochondrial dysfunction [29]. Second, damaged hepatocytes caused by alcohol release damage‐associated molecular patterns (DAMPs) induce liver inflammation, such as neutrophil infiltration [30]. Third, chronic alcohol consumption induces gut bacterial overgrowth and dysbiosis and increases gut permeability, resulting in elevation of bacteria and their related products in the liver and circulation and subsequent inflammation [31]. AH is a severe form of ALD that presents a form of acute‐on‐ chronic liver failure in individuals with AUD with underlying ALD including ASH and cirrhosis [4, 32, 33]. AH is characterized by an abrupt rise in jaundice (serum bilirubin levels) and liver‐related complications and is often accompanied by fever, abdominal pain, anorexia, and weight loss [33]. In addition, portal hypertension, ascites, encephalopathy, and variceal bleeding

685

are associated with severe cases of AH. Most patients with AH have pre‐existing alcoholic cirrhosis, while some patients with mild underlying liver disease or silent ALD can also develop AH. Diagnosis of AH is based on clinical syndromes with an episode of jaundice with serum bilirubin levels greater than 3 mg dL−1 and AST levels greater than 50 IU mL−1 in chronic AUD with heavy drinking greater than 5 years [33]. Jaundice is an essential syndrome for the diagnosis of AH, but not all episodes of jaundice in AUD with chronic ALD are attributed to AH [4]. For example, severe sepsis, biliary obstruction, diffuse liver cancer, drug‐induced liver injury, and ischemic hepatitis (i.e. due to massive bleeding or cocaine use) can also cause an episode of jaundice and liver decompensation in AUD, which should not be diagnosed as AH [4]. Patients with severe AH especially those with cirrhosis usually have high short‐term mortality with a mortality rate of 20–50% at three months [32]. Liver histology analysis of AH revealed infiltration of a large number of inflammatory cells in the liver with predominant neutrophils and macrophages [30]. Microarray analysis has identified that a large number of inflammatory mediators are upregulated in the liver from AH patients [34]. In addition to massive liver inflammation, a systemic inflammatory response is also a key feature of AH and correlates with disease severity [35]. The inflammation in most AH patients is likely triggered by recent excessive binge drinking, bacterial infection, or both [4]. Recent excessive binge drinking was reported in many

Figure 53.1  Pathogenesis and molecular mechanisms that trigger liver failure, systemic inflammatory response, and multiorgan failure in alcoholic hepatitis. Excessive binge drinking in AUD with underlying ALD induces hepatocellular damage, induces systemic inflammation, and impairs liver regeneration, resulting in liver failure and other complications in alcoholic hepatitis. Modified from [4] and reproduced with permission from Elsevier.

686

THE LIVER:  EMERGING MECHANISMS

patients with AH, and can cause elevation of serum ALT and AST levels as well as circulating neutrophils in AUD [12]. It is possible that excessive binge drinking causes massive hepatocyte damage, which releases a large amount of DAMPs that trigger systemic inflammation in patients with AUD. For example, damaged hepatocytes can release mitochondrial DNA, which can activate various types of inflammatory cells via Toll‐like receptors [11]. Bacterial infection is detected in up to 50% of AH patients and is associated with high mortality in these patients [36]. Pathogen-associated molecular patterns (PAMPs) produced during bacterial infection including endotoxin, can activate macrophages/Kupffer cells to produce a wide variety of inflammatory mediators in AH [35]. Normal liver has the great ability to regenerate after injury or loss of tissue; however, hepatocytes in the injured liver in AH patients seem unable to efficiently regenerate. Instead, liver ­progenitor cells are markedly expanded and accumulated in AH patients, and the number of these cells positively correlates with disease severity. It is likely that these liver progenitor cells are unable to fully differentiate into functional hepatocytes and do not contribute to liver regeneration. Thus, impaired liver regeneration is probably an important mechanism contributing to liver failure in patients with severe AH (Figure 53.1).

ALCOHOLIC LIVER FIBROSIS Fibrosis is a scarring response of the liver to chronic liver damage caused by virtually all etiologies of liver pathology including excessive alcohol drinking. Although all types of chronic liver injury cause liver fibrosis, alcoholic liver fibrosis has a characteristic pattern of pericellular and sinusoidal chicken‐ wire fibrosis, which could be extended to panlobular fibrosis when this characteristic pattern becomes diffused [37]. Liver fibrosis is characterized by excessive accumulation of collagen and other extracellular matrix (ECM) proteins. Most of these ECM proteins are produced by activated hepatic stellate cells (HSCs). Several other types of cells can also produce ECM ­protein, but to a much lesser extent. These cells include portal fibroblasts and bone marrow‐derived myofibroblasts. Activation of HSCs, a key step in developing liver fibrosis, is controlled by many inflammatory mediators and growth factors that are commonly elevated in ALD and other types of liver diseases [37]. For example, transforming growth factor (TGF)‐β is one of the most important mediators that induces HSC transformation; while platelet‐derived growth factor (PDGF) is a critical growth factor that stimulates HSC proliferation. Both TGF‐β and PDGF are elevated in ALD in patients and animal models of alcoholic liver fibrosis [37]. In addition, there are some unique mechanisms that promote alcoholic liver fibrosis [37]. First, acetaldehyde, the first ethanol metabolite, can directly stimulate and maintain HSC activation by stimulating multiple signaling pathways and transcription factors. Second, ALD patients exhibit elevated lipopolysaccharides (LPSs) due to increased gut bacterial growth and gut permeability. LPS is a well‐known mediator that can directly induce HSC activation, thereby playing an important role in promoting alcoholic liver fibrosis. Third, activation of innate immunity is known to play a critical role in

promoting HSC activation and liver fibrosis, but ­activation of some components of innate immunity negatively regulates HSC activation and suppresses liver fibrosis. For example, activation of natural killer (NK) cells can directly kill activated HSCs and produce IFN‐γ that blocks HSC proliferation and induces HSC death. Such an antifibrotic effect of NK cells is markedly suppressed after chronic alcohol consumption, thereby accelerating liver fibrosis [38].

ALCOHOLIC LIVER CANCER Liver cirrhosis derived from all etiologies of liver pathology including alcohol consumption is a major risk factor for liver cancer, mainly HCC. The mechanisms by which cirrhosis promotes HCC are not fully understood. Several hypotheses have been proposed. First, cirrhosis is associated with telomere shortening, which can induce chromosomal instability and mutations of tumor suppressors and oncogenes. Second, cirrhosis is associated with chronic inflammation and elevation of many growth factors and cytokines that stimulate HCC growth. In addition to these common mechanisms, there are some exceptional mechanisms that are believed to specifically contribute to alcoholic liver cancer development. Acetaldehyde, an ethanol metabolite, is considered a carcinogen with mutagenic properties. It is well documented that ALDH2‐deficient individuals have high levels of acetaldehyde after alcohol drinking and are associated with an increased risk of esophageal cancers; however, the ­association between HCC and individuals with AUD with ALDH2 deficiency is less clear. It is known that acetaldehyde is electrophilic and forms an adduct with DNA and interstrand crosslinks, causing DNA damage. In addition, acetaldehyde suppresses DNA repair enzyme activity and subsequently blocks DNA repair. Thus, acetaldehyde not only causes DNA damage but also prevents DNA repair, contributing to hepatocarcinogenesis. Moreover, heavy drinking can induce aberrant DNA methylation including genome‐wide hypomethylation, leading to chromosomal instability and subsequently promoting the development of liver cancer. Excessive alcohol drinking is known to cause broad immune suppression including attenuation of antitumor immunity, which is probably another important mechanism that can accelerate HCC development and progression.

EMERGING MECHANISMS Gut microbiome An important role of gut bacteria‐derived LPS in the development of ALD in animal models was reported more than two ­decades ago. Recent studies suggest that the gut microbiome plays a complex role in the pathogenesis of ALD [31]. Chronic alcohol consumption can cause bacterial overgrowth in the large and small intestines and dysbiosis in humans and animals [31]. In general, changes in gut microbiome post chronic alcohol consumption are complex but are often associated with a ­ decrease in “good” bacteria (such as Lactobacillus spp.) but an



53:  Alcoholic Liver Disease

increase in “bad” bacteria (such as Enterobacteriaceae). These changes could cause pathological bacterial translocation, alterations of intestinal metabolites and bile acid metabolism, resulting in elevation of systemic levels of gut‐derived microbial production and subsequent hepatocellular damage and liver inflammation [31]. In addition, a recent study suggests that gut fungi may also play a role in promoting ALD [39]. Chronic alcohol consumption increases microbiota populations in gut and treatment with antifungal drugs attenuated intestinal fungal overgrowth and ameliorated ALD in mice. Patients with ALD have reduced intestinal fungal diversity and Candida overgrowth and increased systemic exposure and immune response to mycobiota. The levels of these responses correlate with m ­ ortality, suggesting that altered gut mycobiomes contribute to ALD pathogenesis [39]. Moreover, transplantation of intestinal microbiota from a patient with severe AH induced greater liver inflammation and injury in mice than transplantation of these from an individual with AUD without AH [40]. Thus, manipulating the intestinal microbiome (both bacteria and fungi) might be an effective strategy for preventing and treating ALD, and fecal microbiota transplant might be a potential therapeutic option for ALD.

Epigenetics Epigenetics is the investigation of heritable alterations of phenotype (appearance) or gene expression that are caused by mechanisms other than alterations in DNA coding sequences. Epigenetic modifications include DNA methylation, histone modifications, and RNA‐based mechanisms. Alcohol consumption is known to affect metabolism of methionine and thereby DNA methylation [41]. In the liver, homocysteine is methylated to methionine and then S‐adenosylmethionine (SAMe) in a transmethylation reaction that is catalyzed by methionine adenosyltransferase. SAMe is a principal methyl donor in methylation and plays a critical role in inducing DNA and histone methylation. Excessive alcohol consumption reduces hepatic levels of SAMe and DNA and histone methylation, which subsequently upregulates the expression of genes that control the endoplasmic reticulum stress response and ALD [41].

miRNAs miRNAs (small non‐coding RNA molecules with 19–25 nucleotides) play a critical role in controlling gene expression by inducing RNA silencing and by altering post‐transcriptional regulation of gene expression. Many miRNAs have been shown to play important roles in the development of ALD. For example, hepatocyte‐specific miR122 seems to play a role in protecting against ALD by reducing hypoxia‐inducible factor 1 alpha (HIF1α) mRNA. Hepatic miR122 is downregulated in liver samples from patients with ALD and ethanol‐fed mice, thereby exacerbating liver injury and inflammation [42]. In contrast, neutrophil‐specific miR‐223 is upregulated in the liver and serum from ethanol‐fed mice and individuals with AUD compared to their controls [12]. Such elevations likely play a compensatory role in limiting neutrophil activation via inhibition of the IL‐6–p47phox pathway [12].

687

Extracellular vesicles (EVs) EVs are nanometer‐sized, membrane‐bound, extracellular milieu vesicles derived from cells, playing an important role in connecting cell to cell communication and in promoting inflammation in a variety of diseases, including liver diseases. EVs can be grouped as three types based on their cellular biogenesis and sizes: exosomes (40–150 nm diameter in size), microvesicles/ microparticles (MPs; 50–1000 nm diameter), and apoptotic bodies (greater than 500 nm in diameter). EVs exert their functions by delivering their contents from one cell to another. For instance, stressed hepatocytes can activate macrophages and neutrophils via the release of EVs that contain lipids, proteins, chemokines, and nucleic acids (e.g. mitochondrial DNA [mtDNA]), playing a role in the pathogenesis of ALD [43]. Patients with AUD or ALD had an elevated number of EVs in the circulation, which are enriched in mtDNA and correlate with elevated circulating neutrophils and serum levels of ALT [11]. Compared with pair‐fed control mice, mice post chronic or chronic‐plus‐binge ethanol feeding also have elevated circulating EVs, which are enriched in mRNAs, proteins, mtDNAs, and likely promote liver inflammation and injury in ALD [11, 44]. In addition to being an important player in ALD, EVs may be used as potential biomarkers for diagnosis because EVs from ALD contain specific signatures of proteins, RNAs, and DNAs [43].

Adaptive immunity The role of innate immunity in the pathogenesis of ALD has been well documented; however, the involvement of adaptive immunity in ALD remains obscure. It is generally believed that excessive alcohol consumption induces oxidative stress, which subsequently generates lipid peroxidation products, such as malondialdehyde and 4‐hydroxynonenal. These products can modify many proteins and induce formation of protein adducts that can act as neoantigens to induce the activation of adaptive immunity [45]. Activation of adaptive immunity has not been characterized or detected in animal models of ALD, but it was reported in patients with ALD [45]. For example, patients with ALD are associated with increased levels of circulating antibodies against lipid peroxidation adducts and increased numbers of intrahepatic T cells [45]. The infiltration of intrahepatic T cells in ALD is not just bystander activation, and these T cells present a pronounced oligoclonal nature along with ALD‐associated clonotypes as demonstrated in a recent high‐throughput T‐cell receptor sequencing study [46]. If activation of adaptive immunity is a major driver of disease progression in some ALD patients, immunosuppressive therapy may be required to ­effectively treat these patients.

Innate‐like T cells The roles of Kupffer cells/macrophages in the pathogenesis of ALD have been extensively studied over the last two decades. Recent studies suggest that nature killer T (NKT) cells and mucosa‐associated invariant T cells (MAIT) also play a role in the pathogenesis of ALD. Mouse liver lymphocytes contain approximately 30–40% natural killer T (NKT) cells, which are a heterogeneous group of T cells that share properties of both

688

THE LIVER:  CLINICAL DIAGNOSIS AND TREATMENT OF ALD

T  cells and NK cells and rapidly produce a large number of cytokines such as IFN‐γ, IL‐4, and so on. Several recent studies reported that hepatic NKT cells are activated and increased post chronic‐plus‐binge ethanol feeding. These activated NKT cells contribute to Kupffer cell activation and alcoholic liver injury [47, 48]. Of note, human liver lymphocytes contain a very low number of NKT cells but are enriched with MAIT cells, representing 20–50% of intrahepatic T cells in normal human livers. MAIT cells are detected at low levels in commonly used laboratory mouse strains, accounting for less than 1% of intrahepatic T cells in mice. MAIT cells play a key role in host defense against bacterial infection through invariant T‐cell receptors (TCRs) that recognize microbial riboflavin/ vitamin B2 metabolites presented by the major histocompatibility complex class I‐related protein 1. Patients with severe ALD are associated with a marked reduction of MAIT cells [49], which may contribute to an increased risk of bacterial infection in these patients.

Chronic‐plus‐binge ethanol Over the last 5 years, one important advance in the field of ALD study is the introduction of binge ethanol in chronically ethanol‐ fed mice [50] or in high‐fat diet‐fed mice [5]. This chronic‐plus‐ binge ethanol feeding model was initially called the NAAA model [50] and was later also called the Gao‐binge model [51]. Such an ethanol challenge induces more severe alcoholic steatohepatitis with marked elevation of serum ALT and AST and significant hepatic neutrophil infiltration compared with binge ethanol alone, chronic ethanol feeding alone, and high‐fat diet feeding alone [5, 8, 52]. Acute ethanol gavage was also introduced in the chronic intragastric ethanol infusion model, which causes marked hepatic neutrophil infiltration [53]. By using these new animal models, researchers have identified many novel mechanisms that contribute to acute‐on‐chronic liver injury in ALD. These include activation of endoplasmic reticulum stress, activation of inflammatory cells (e.g. neutrophils, Kupffer cells, and NKT cells), upregulation of several factors that promote steatosis (e.g. FSP27), and dysregulation of cell survival/death signaling pathways (e.g. pyroptosis, apoptosis, etc.) [10, 53]. For example, several studies suggest that neutrophils contribute to hepatocellular damage and hepatic injury in this early stage of ALD that is induced by chronic‐plus‐binge ethanol feeding or high‐fat diet‐plus‐binge ethanol feeding [5, 8]. However, neutrophils can also promote liver repair and are vital in the control of bacterial infection. Therefore, neutrophils likely play both beneficial and detrimental roles in the ­pathogenesis of ALD.

CLINICAL DIAGNOSIS AND TREATMENT OF ALD The diagnosis of ALD is based on the combination of clinical and laboratory findings, including a medical history of significant alcohol intake, physical signs of liver disease, supporting laboratory tests of liver disease, non‐invasive liver imaging, and invasive liver biopsy [2, 54]. Documentation of excessive

alcohol drinking and evidence of liver disease are the key diagnostic factors of ALD.

Medical history Denial of excessive alcohol use or underreported alcohol intake in the medical history is a frequent occurrence in AUD patients, which may make diagnosis of ALD challenging in these patients. Therefore, indirect evidence of excessive alcohol use is always considered, such as questionnaires, information from family members, and laboratory tests to reinforce or confirm the suspicion of AUD. Various questionnaires for screening ­excessive alcohol use are used by clinicians, such as the CAGE (cut‐annoyed‐guilty‐eye), MAST (Michigan Alcoholism Screening Test), and AUDIT (Alcohol use disorders identification test) [2]. The CAGE and MAST questionnaires are the most commonly used.

Clinical symptoms and signs Anorexia, nausea and vomiting, upper abdominal pain, malaise, fatigue, dark urine, fever, confusion, and weight loss are the most commonly presenting nonspecific symptoms of ALD. The most common sign of AH is rapid development of jaundice; others include fever, ascites, proximal muscle wasting, tender hepatomegaly, and hepatic bruit. Severe cases often develop hepatic encephalopathy and gastrointestinal bleeding. Patients with mild to moderate ALD are often asymptomatic. In some patients, bilateral parotid gland hypertrophy, muscle wasting, malnutrition, Dupuytren’s contracture sign, and signs of symmetric peripheral neuropathy are suggestive of harmful alcohol consumption. Splenomegaly, gynecomastia, spider angiomas, asterixis, palmar erythema, and digital clubbing are frequently observed in ALD patients with cirrhosis [2]. In decompensated cirrhosis, jaundice, ascites, peripheral edema, and hepatic encephalopathy are always seen in addition to the cirrhotic physical ­findings [2].

Laboratory tests There are no specific laboratory biomarkers for ALD. Elevations of serum enzymes such as AST, ALT, alkaline phosphatase (ALP), and GGT can only provide clues for the diagnosis of ALD. The NIAAA Alcoholic Hepatitis Consortia proposed a working definition of acute AH to include an AST to ALT ratio above 1.5 with AST and ALT less than 400 U L−1 [33]. Clinically, GGT is the most frequently used biomarker to detect previous alcohol consumption, though it can also arise from other etiologies [55]. Routine blood tests and biochemical detection, such as elevations of red blood cell mean corpuscular volume (RBC MCV), white blood cell (WBC) count, GGT, AST, and immunoglobulin A (IgA) may indicate early ALD, while a decrease in albumin, prolonged prothrombin time (PT), elevated total bilirubin (TBIL) level, and low platelet (PLT) count are indications of advanced ALD [56]. Recent studies suggest that a new biomarker of apoptotic and necrotic cell death, the M30/M65 fragment of cytokeratin 18 (CK18), is sensitive to liver injury in ALD patients [57]. In some ALD patients, metabolic abnormalities such as elevated levels of triglycerides and uric acid, hypokalemia, and hyponatremia are frequently observed.



53:  Alcoholic Liver Disease

Liver imaging Ultrasound, computerized tomography, and magnetic resonance imaging can be used to detect the presence of underlying liver disease, but they cannot provide specific information for diagnosing ALD. Ultrasound is routinely used to check hepatocellular carcinoma, biliary obstruction, ascites, splenomegaly, portal hypertension, and portal vein thrombosis. Non‐contrast computerized tomography more easily detects macroscopic fat in the liver in fatty liver disease [58]. Magnetic resonance imaging is the most sensitive and specific imaging modality for detecting steatosis (95% sensitivity, 98% specificity) except for patients with iron overload [59]. Recently, the clinical applications of newer imaging techniques such as transient elastography (TE, Fibroscan®) and magnetic resonance elastography (MRE) improve diagnostic accuracy in quantifying steatosis and fibrosis [60]. Among all imaging modalities, ultrasound is the most widely used due to its low cost. By detecting values of controlled attenuation parameter (CAP value) and liver stiffness, transient elastography (TE, Fibroscan) has demonstrated good accuracy in quantifying steatosis and fibrosis, and it is a significantly cheaper alternative [54].

Liver biopsy Liver biopsy is not routinely recommended for the diagnosis of ALD in most patients in the United States. However, it is believed to be the gold standard for diagnosing and assessing the severity of steatosis and staging of fibrosis, and it is the only modality that is available to distinguish between simple steatosis and steatohepatitis [2]. Liver biopsy is invasive and mostly ­performed percutaneously. A transjugular route of biopsy is required in patients with coagulopathy or ascites. The histologic features of ALD on liver biopsy vary based on the extent and stage of hepatic injury. The typical histological lesions of ALD are large fat droplets and ballooning of hepatocytes including Mallory–Denk bodies and mega‐mitochondria as well as neutrophil infiltration and intrasinusoidal chicken‐like fibrosis [60]. Macrovesicular steatosis is the earliest and most frequently seen pattern of alcohol‐induced liver injury. Alcoholic steatohepatitis is defined by the coexistence of steatosis, hepatocyte ballooning, and neutrophil infiltration [54].

Complications in ALD Cholestasis, fat embolism, and portal hypertension are occasionally observed in patients with severe fatty liver caused by alcohol. Alcoholic ketoacidosis is often observed in patients with chronic alcohol consumption and concurrent malnutrition. Patients with alcoholic cirrhosis or severe AH always have ­complications such as ascites, spontaneous bacterial peritonitis (SBP), variceal bleeding, electrolyte disturbance, hepatorenal syndrome (HRS), hepatic encephalopathy, and HCC [2].

Assessment of ALD There are several prognostic scoring models for assessing ­disease severity and short‐term mortality of ALD, including the Maddrey discriminant function (MDF), the Mayo model for

689

end‐stage liver disease (MELD), the Glasgow alcoholic hepatitis score (GAHS), the Age, bilirubin, INR, creatinine (ABIC) score, the Lille model, and the Child–Turcotte–Pugh (CTP) score [2]. MDF (4.6 × [patient prothrombin time – control prothrombin time] + serum total bilirubin) is the most widely used [2]. ALD patients with a MDF value greater than or equal to 32 is indicative of a high risk of short‐term mortality (30–50% at one month) with improved short‐term clinical outcomes after receiving corticosteroid [60].

TREATMENT OF ALD Alcohol abstinence The backbone therapeutic intervention for patients with ALD is alcohol abstinence. Following abstinence, relapse is a major risk in these patients. Continued alcohol use after diagnosis of ALD has been shown to promote disease progression [2, 54].

Nutritional support Malnutrition is very common in patients with severe ALD, such as deficiencies of vitamin A, vitamin D, vitamin E, thiamine, folate, niacin, pyridoxine, zinc, magnesium, and selenium. A dietary intake of 1.2–1.5 g of protein kg−1 and 35–40 kcal kg−1 body weight is recommended in these patients [60].

Corticosteroids Corticosteroids are recommended for patients with severe AH in the absence of sepsis and infection, which can improve short‐ term survival rates in these patients [2]. However, long‐term follow‐up does not reveal a significant survival difference between those treated with corticosteroids and controls [2].

Pentoxifylline (PTX) PTX is a competitive, nonselective, phosphodiesterase inhibitor that can inhibit tumor necrosis factor and leukotriene synthesis, inflammation, and innate immunity. PTX can also reduce the development of hepatorenal syndrome (HRS) in ASH patients, resulting in improvements in short‐term survival [61].

N‐acetyl cysteine (NAC) NAC has antioxidant activity and is ineffective when used alone in ALD patients. However, a randomized trial showed that the combination of NAC with prednisolone reduced one‐month mortality and incidence of HRS/infection [62].

Artificial liver support system Artificial liver support systems are a therapy option for liver failure associated with severe ALD, which has been controversial for decades. Recently, there was a study that showed that the application of an artificial liver support system in patients with severe liver dysfunction secondary to ALD showed some benefits in a subset of the patients [63].

690

THE LIVER:  NEW THERAPEUTIC TARGETS FOR ALCHOLIC HEPATITIS

Liver transplantation Patients with end‐stage liver disease secondary to alcoholic ­cirrhosis should be considered for liver transplantation. Early liver transplantation, which has been performed for acute severe AH, is discussed in the following section.

NEW THERAPEUTIC TARGETS FOR ALCHOLIC HEPATITIS Although no new drugs for ALD have successfully been developed, many therapeutic targets were recently identified from studies of human ALD samples and animal models, and some of them are currently in clinical trials for the treatment of AH, which are discussed below. Inflammation is considered a critical factor for causing liver damage in AH and has been actively investigated as a therapeutic target for the treatment of AH [64]. For instance, immunosuppressive drug steroids, such as prednisolone, have been used for the treatment of AH for more than five decades, but accumulating data suggest that prednisolone treatment is beneficial for short‐term survival but has no benefits for long‐term survival in AH patients [65]. Steroid drugs have broad immunosuppressive functions but are ineffective for a variety of neutrophil‐mediated disorders (e.g. asthma, septic shock), which is, at least in part, mediated by promoting neutrophil survival. AH is associated with hepatocyte necrosis and infiltration of neutrophils, which may contribute to the ineffectiveness of prednisolone treatment in some AH patients. In addition, steroid treatment is associated with an increased risk of bacterial infection in patients with AH due to its broad immunosuppressive function. More specific targets for inflammation are urgently needed for the treatment of AH. Indeed, a large number of therapeutic targets for inflammation have been recently identified and include inflammatory cytokines and mediators, chemokines and their receptors, gut microbiota and bacterial products, and so on [66]. Some of them are currently being tested in clinical trials for the treatment of AH, including IL‐1 inhibitors, apoptosis signal‐regulating kinase 1 (ASK1) inhibitors, LPS blockers, and probiotics [66]. Preclinical studies demonstrated that IL‐1 plays an important role in promoting ALD in animal models. Thus, IL‐1 inhibitors, which have been approved for the treatment of several types of inflammatory disease due to their good safety profiles and low tendency for adverse side‐effects, are currently being tested in clinical trials for the treatment of AH. ASK1, a serine/threonine signaling kinase, is activated by oxidative stress, and its activation leads to activation of p38 mitogen‐activated kinase (p38 MAPK) and c‐Jun N‐terminal kinase (JNK), playing a role in promoting hepatic inflammation, apoptosis, and fibrosis. ASK1 inhibitors are currently being tested in clinical trials for several types of liver diseases including non‐alcoholic steatohepatitis and AH. In addition, a large number of studies have demonstrated that excessive alcohol consumption induces gut bacterial overgrowth, dysbiosis, and elevation of circulating LPS, which contribute to liver inflammation and injury in AH. Several ongoing clinical trials are targeting these factors to treat AH using probiotics, antibiotics, fecal microbiota transplantation,

hyperimmune bovine colostrum enriched with anti‐LPS antibodies, and so on. Collectively, many inflammatory mediators have been implicated in liver injury and inflammation in patients with severe AH. Many of these ­mediators probably synergistically or additively promote liver inflammation in AH. Clinical trials are needed to identify the inflammatory mediators that play critical roles in the pathogenesis of ALD in patients and can be used as targets for the ­treatment of ALD, which will take years to complete. The immunosuppressive drug prednisolone or other steroids will likely continue to be used for the treatment of severe AH until more specific immunosuppressive drugs are identified. AH is associated not only with hepatocellular injury but also with impairments of liver regeneration. The application of hepatoprotective agents may provide some benefits in ALD therapy to protect against hepatocellular damage and promote liver regeneration. Indeed, the hepatoprotective cytokine IL‐22 is currently being tested in clinical trials for the treatment of AH. IL‐22, which induces STAT3 activation in hepatocytes, has many ­beneficial effects in the liver [67] but likely has minimal side‐effects due to the restricted expression of IL‐22 receptors, which are mainly expressed in epithelial cells but not in immune cells. By targeting hepatocytes, IL‐22 plays an important role in ameliorating hepatocellular damage, promoting liver regeneration, and alleviating liver fibrosis [67]. In addition, IL‐22 treatment may effectively impede bacterial infection and ameliorate kidney injury, two deleterious conditions that often contribute to death in AH patients. IL‐22 therapy is currently being tested in clinical trials for the treatment of patients with severe AH [68]. Because severe ALD and AH are associated with multiple problems, including inflammation, hepatocellular damage, poor liver regeneration, and many complications, combination therapies are likely to be needed to treat these severe disorders [68]. These combination therapies include the inhibition of inflammation, hepatoprotection and stimulation of liver regeneration, and organ‐specific support in multiple organ failure for critically ill patients. Liver transplantation is probably the only potentially life‐saving treatment for many patients with end‐stage ALD and multiple organ failure who do not respond to medical treatment. Indeed, ALD now replaces hepatitis C (HCV) infection as the leading indication for liver transplantation in the United States. Current treatment guidelines require a six‐month period of alcohol abstinence; however, most patients with severe AH do not survive six months. Thus, several centers in Europe and the United States have started early liver transplantation with less than six‐month abstinence, achieving excellent clinical outcomes with great survival rates and low rates of alcohol relapse in highly selected patients [69]. Based on these data, the American Gastroenterological Association Clinical Practices Update Committee gave best practice advice: “Patients with severe AH, particularly those with a MELD score > 26 with good insight into their AUD and good social support should be referred for evaluation for liver transplantation, as the 90‐day mortality rate is very high” [70]. However, this recommendation has not been broadly adopted yet by the majority of transplantation centers and is still challenged by several issues, such as a shortage of donor livers, competition for donor livers by other types of liver diseases, alcohol relapse post transplantation, and potentially spontaneous recovery in these patients after abstinence.



53:  Alcoholic Liver Disease

NEUROBIOLOGY OF ALCOHOL USE DISORDER (AUD) Definitions and conceptual framework for the neurocircuitry of AUD AUD can be defined as a chronically relapsing disorder that is characterized by a compulsion to seek and take the drug (alcohol), loss of control in limiting drug (alcohol) intake, and the emergence of a negative emotional state (e.g. dysphoria, anxiety, irritability [hyperkatifeia]), reflecting a motivational withdrawal syndrome, when access to the drug (alcohol) is prevented [71]. These key elements embrace most of the symptoms of AUD as expressed in moderate to severe AUD in the Diagnostic and Statistical Manual

691

of Mental Disorders, 5th edition (DSM‐5; American Psychiatric Association, 2013). AUD and addiction in general have been ­heuristically framed as a three‐stage cycle: binge/intoxication, withdrawal/negative affect, and preoccupation/anticipation (“craving”). These three stages r­epresent dysregulation in three functional domains (incentive salience/pathological habits, negative emotional states, and executive function, respectively) and are mediated by three major neurocircuitry elements (basal ganglia, extended amygdala, and prefrontal cortex, respectively; Figure  53.2). The three stages are conceptualized as interacting with each other, becoming more intense, and ultimately leading to the pathological state known as addiction [71]. In AUD, a pattern of oral drug taking evolves that is often characterized by binges of alcohol intake that can be daily

Figure 53.2  Conceptual framework for the neurobiological basis of addiction. In the binge/intoxication stage, reinforcing effects of drugs may engage reward neurotransmitters and associative mechanisms in the nucleus accumbens shell and core and then engage stimulus‐response habits that depend on the dorsal striatum. In the withdrawal/negative affect stage, the negative emotional state of withdrawal may engage activation of the extended amygdala. The extended amygdala is composed of several basal forebrain structures, including the bed nucleus of the stria terminalis, central nucleus of the amygdala, and possibly a transition zone in the medial portion (or shell) of the nucleus accumbens. There are major projections from the extended amygdala to the hypothalamus and brainstem. The preoccupation/anticipation (craving) stage involves the processing of conditioned reinforcement in the basolateral amygdala and the processing of contextual information by the hippocampus. Executive control depends on the prefrontal cortex and includes the representation of contingencies, the representation of outcomes, and their value and subjective states (i.e. craving and, presumably, feelings) associated with drugs. The subjective effects, termed “drug craving” in humans, involve activation of the orbitofrontal and anterior cingulate cortices and temporal lobe, including the amygdala. ACC, anterior cingulate cortex; BNST, bed nucleus of the stria terminalis; CeA, central nucleus of the amygdala; DS, dorsal striatum; dlPFC, dorsolateral prefrontal cortex; GP, globus pallidus; HPC, hippocampus; NAC, nucleus accumbens; OFC, orbitofrontal cortex; Thal, thalamus; vlPFC, ventrolateral prefrontal cortex; vmPFC, ventromedial prefrontal cortex. Modified from [125] and reproduced with permission of Springer Nature.

692

THE LIVER:  NEUROBIOLOGY OF ALCOHOL USE DISORDER (AUD)

episodes or prolonged days of heavy drinking and is characterized by a severe emotional and somatic withdrawal syndrome. Many individuals with AUD continue with such a binge/withdrawal pattern for extended periods of time, but some individuals can evolve into an opioid addiction‐like situation in which they must have alcohol available at all times to avoid the negative consequences of abstinence. Here, intense preoccupation with obtaining alcohol (craving) develops that is linked not only to stimuli that are associated with obtaining the drug but also to stimuli that are associated with the aversive motivational state of withdrawal. A pattern ultimately develops in moderate to severe AUD where the drug often is taken to avoid the severe dysphoria and discomfort of abstinence. The thesis argued here, derived largely from fundamental work in neurobiology, is that AUD is a brain neurocircuitry disorder and that neuroadaptations within specific motivational circuits play an important role in defining and perpetuating the disorder. The argument herein is that the excessive engagement of reward circuitry engages high incentive salience for cues and contexts that are conditioned to drug seeking (binge/intoxication stage) and sets up a core deficit of lower reward function and greater activation of brain stress systems (withdrawal/negative affect stage) and significant impairment in executive function, all of which contribute to the compulsive drinking that is associated with AUD. Drug addiction has generally been conceptualized as a ­disorder compulsivity. that involves elements of both impulsivity and ­ Collapsing the cycles of impulsivity and compulsivity yields a composite addiction cycle that consists of three stages – preoccu­ pation/anticipation, binge/intoxication, and withdrawal/negative affect – in which impulsivity often dominates at the early stages and compulsivity dominates at the terminal stages (Figure 53.2). As an

individual moves from impulsivity to compulsivity, a shift occurs from positive reinforcement that drives the motivated behavior to negative reinforcement that drives the motivated behavior [72]. Negative reinforcement can be defined as the process by which the removal of an aversive stimulus (e.g. negative emotional state of  drug withdrawal) increases the probability of a response. Importantly, negative reinforcement is not punishment, although both involve an aversive stimulus. In punishment, the aversive stimulus suppresses behavior, including drug taking (e.g. disulfiram [Antabuse]). Negative reinforcement can perhaps be described in lay terms as reward via relief (i.e. relief reward), such as the removal of pain or in the case of AUD removal of the negative emotional state of acute withdrawal or protracted abstinence. Driving negative reinforcement is a negative emotional state that is a common ­ ­presentation in most individuals with AUD during withdrawal and protracted abstinence. The neurobiological substrates that ­underlie the motivation to seek alcohol will be reviewed herein using the heuristic framework that is outlined above.

Neural substrates for the binge/intoxication stage associated with AUD Alcohol at intoxicating doses has a wide but selective action on neurotransmitter systems in the brain reward systems, based on animal models of the acute reinforcing effects of alcohol, and animal studies that use selective receptor antagonists for specific neurochemical systems, and human positron emission tomography imaging studies. Multiple neurochemical systems are implicated in the acute reinforcing effects of alcohol, including γ‐aminobutyric acid (GABA), opioid peptides, dopamine, serotonin, and glutamate (Figure 53.3).

Figure 53.3  Schematic diagram describing the etiology of addiction based on an early drawing by Dr Loren Parsons. Notice that as positive reinforcement fades as a motivating factor, negative reinforcement is conceptualized to gain importance as driving compulsive drug seeking. Relapse is hypothesized to return one to the addictive process at any point in the cycle, and repeated binges–withdrawals–relapses often speed the trajectory back to compulsive use, presumably through residual changes in neurocircuitry.



53:  Alcoholic Liver Disease

There has been significant work that shows, at least at the pharmacological level, a role for GABA in the intoxicating effects of alcohol [73]. Systemic injections of GABAA receptor antagonists or inverse agonists reverse the motor‐impairing effects of alcohol, the anxiolytic‐like effects of alcohol, and alcohol drinking [73]. Endogenous opioid peptide systems have long been hypothesized to play a role in the reinforcing effects of alcohol. Naltrexone decreases alcohol drinking and self‐ administration in a variety of animal models [73], and these results led to the clinical use of naltrexone in reducing alcohol consumption and preventing relapse. Significant evidence also supports a role for the mesolimbic dopamine system in alcohol reinforcement. Alcohol self‐administration increases extracellular levels of dopamine in the nucleus accumbens in nondependent rats [74]. Such increases occur not only during the actual self‐administration session but also precede the self‐administration session, possibly reflecting the incentive motivational properties of cues that are associated with alcohol [74]. Incentive motivation (i.e. incentive‐salience) is anchored within the construct of conditioned reinforcement and is argued to be a phenomenon by which a previously neutral stimulus acquires incentive value through pairings with a drug of abuse [75]. Systemic injections of dopamine receptor antagonists also decrease responses to alcohol. However, mesolimbic dopamine does not appear to be essential for the acute reinforcing effects of alcohol, since lesions of the mesolimbic dopamine system failed to block operant self‐administration of alcohol [73], suggesting multiple redundant sources of alcohol’s actions that converge on the nucleus accumbens and amygdala. A prominent hypothesis is that when drug addiction progresses from occasional recreational use to compulsive use, drug‐seeking behavior shifts from reward‐driven to habit‐driven behavior to compulsive‐like responding. Three cortico‐basal ganglia circuits form what have been described as cortico‐striatal‐pallidal‐­thalamic loops that process associative, sensorimotor, and emotional information [76]. During this progression to habit‐driven behavior, the control over drug‐seeking behavior also appears to shift from the nucleus accumbens to the dorsal striatum. Furthermore, within the dorsal striatum, there is a shift in function [77].

Neural substrates for the withdrawal/negative affect stage associated with AUD The neurocircuitry and neuropharmacology of the withdrawal/ negative affect stage of the addiction cycle support a conceptual framework that builds on opponent process theory [78] and extends to an allostatic model of brain motivational systems that has been proposed to explain persistent changes in motivation that are associated with dependence in addiction [79, 80]. In this formulation, addiction is conceptualized as a cycle of increasing dysregulation of brain reward/anti‐reward mechanisms that results in a negative emotional state that contributes to the ­compulsive use of drugs. Counteradaptive processes that are part of the normal homeostatic limitation of reward function fail to return within the normal homeostatic range. These counteradaptive processes are hypothesized to be mediated by two mechanisms: within‐system neuroadaptations and between‐­ system neuroadaptations [80].

693

Within‐system neuroadaptations that contribute to the compulsivity associated with the dark side of AUD One prominent hypothesis is that dopamine and opioid peptide reward/incentive motivational systems are compromised in crucial phases of the addiction cycle, such as withdrawal and protracted abstinence. The argument is that a decrease in dopamine and opioid peptide function is hypothesized to lead to lower motivation for non‐drug‐related stimuli and greater sensitivity to cues that are associated with the abused drug (i.e. increase in incentive salience) [81]. Supporting this hypothesis, decreases in activity of the mesolimbic dopamine system and decreases in serotonergic neurotransmission in the nucleus accumbens have been observed during alcohol withdrawal in animal studies [73]. Parallel to these studies is strong evidence that the firing of ventral tegmental area dopamine neurons dramatically decreases during acute withdrawal from alcohol [82]. Imaging studies of humans with addiction have consistently shown long‐lasting decreases in the numbers of dopamine D2 receptors in alcohol‐dependent subjects compared with controls [83]. Additionally, alcohol‐dependent subjects had dramatically lower dopamine release in the striatum in response to a pharmacological challenge with the stimulant drug methylphenidate [83]. These findings suggest an overall reduction of the sensitivity of the dopamine component of reward circuitry to natural reinforcers and other drugs in individuals with addiction. Other within‐system neuroadaptations under this conceptual framework could include greater sensitivity of receptor transduction mechanisms in the nucleus accumbens. Drugs of abuse have acute receptor actions that are linked to intracellular signaling pathways that may undergo adaptations with chronic treatment. In the context of chronic alcohol administration, multiple molecular mechanisms have been hypothesized to counteract the acute effects of alcohol that could be considered within‐­ system neuroadaptations. For example, chronic alcohol decreases GABA receptor function, possibly through downregulation of the α1 subunit, and also decreases the acute inhibition of adenosine reuptake [73]. Whereas acute alcohol activates adenylate cyclase, withdrawal from chronic alcohol decreases cyclic adenosine monophosphate response element binding protein phosphorylation in the amygdala and is linked to decreases in the function of neuropeptide Y (NPY) and increases in anxiety‐like responses during acute alcohol withdrawal [84].

Between‐system neuroadaptations that contribute to the compulsivity associated with the dark side of AUD Another major neuroadaptation that can contribute to the negative emotional state that drives negative reinforcement in the withdrawal/negative affect stage is brain neurocircuits and ­neurochemical systems that are involved in arousal–stress modulation that are engaged within the neurocircuitry of the brain stress systems in an attempt to overcome the chronic presence of the perturbing drug (alcohol) and to restore normal function despite the presence of drug. The neuroanatomical entity that is termed the extended amygdala [85] may represent a common

694

THE LIVER:  NEUROBIOLOGY OF ALCOHOL USE DISORDER (AUD)

anatomical substrate that integrates brain arousal–stress systems with hedonic processing systems to produce the between‐­ system opponent process that is elaborated above. The extended amygdala forms a separate entity within the basal forebrain [86] and has been conceptualized to be composed of several basal forebrain structures, including the bed nucleus of the stria terminalis, central nucleus of the amygdala, sublenticular substantia innominata, and a transition zone in the medial part of the nucleus accumbens (e.g. shell) [85]. The extended amygdala receives numerous afferents from limbic structures, such as the basolateral amygdala and hippocampus, and sends efferents to the medial part of the ventral pallidum and a large projection to the lateral hypothalamus, thus further defining the specific brain areas that interface classical limbic (emotional) structures with the extrapyramidal motor system. The extended amygdala has long been hypothesized to play a key role not only in fear conditioning [87] but also in the emotional component of pain processing [88]. The brain stress system that is mediated by corticotropin‐ releasing factor (CRF) systems in both the extended amygdala and hypothalamic–pituitary–adrenal axis (HPA) are dysregulated by the chronic administration of all major drugs with dependence or abuse potential, with a common response of ­elevated adrenocorticotropic hormone, corticosterone, and amygdala CRF during acute withdrawal from chronic drug administration [80] (Figure 53.3). In animal models of alcohol dependence, extrahypothalamic CRF systems become hyperactive during alcohol withdrawal, with an increase in extracellular CRF in the central nucleus of the amygdala and bed nucleus of the stria terminalis in dependent rats [80]. For example, alcohol withdrawal reliably produces anxiety‐like responses in animal models that can be reversed by CRF receptor antagonists [80].  Indeed, using models of repeated alcohol exposure and ­withdrawal, intermittent alcohol exposure facilitates withdrawal‐ associated anxiety‐like responses, and a CRF1 receptor antagonist prevented the sensitization of withdrawal‐induced anxiety [89]. These results are consistent with a prolonged history of alcohol exposure that produces persistent upregulation of both CRF and CRF1 receptors in the brain [89]. Perhaps even more compelling, a peptide CRF1/CRF2 antagonist, when administered directly in the central nucleus of the amygdala, blocked alcohol self‐administration in alcohol‐ dependent rats. Cellular studies have identified the actions of CRF on GABAergic interneurons in the central nucleus of the amygdala [90]. Systemic injections of small‐molecule CRF1 receptor antagonists also blocked the increase in alcohol intake that was associated with acute withdrawal and protracted abstinence. A CRF receptor antagonist that was administered chronically during the development of dependence blocked the development of compulsive‐like responding for alcohol (for review, see reference [89]). Alcohol addiction has long been associated with dysregulation of the HPA axis [91]. Clinical studies have reported impairments in HPA axis responsivity in AUD [91], even to the point of pseudo‐Cushing’s syndrome, manifested by high levels of cortisol, in individuals with alcohol addiction [92]. However, a more common observation is a blunted cortisol response in individuals with alcohol dependence, again possibly reflecting adaptations to an initial hyperactive cortisol response [91].

Similar effects have been observed in animal models, with a blunted corticosterone response in rats that are made dependent using the chronic intermittent alcohol vapor model [93]. One hypothesis is that activation of the HPA axis can drive neuroadaptive changes in extrahypothalamic CRF systems in the extended amygdala. High corticosterone increases CRF mRNA in the central nucleus of the amygdala and lateral bed nucleus of the stria terminalis and decreases CRF mRNA in the paraventricular nucleus of the hypothalamus. Thus, an initial exposure to high corticosterone, stimulated by moderate to heavy drinking, may stimulate CRF expression in the central nucleus of the amygdala and lateral bed nucleus of the stria terminalis, eventually leading to neuroadaptive changes, including the further sensitization of CRF activation in the extended amygdala and lower HPA function [93]. Consistent with this hypothesis, chronic ­glucocorticoid receptor blockade with mifepristone during the course of alcohol vapor exposure prevented the escalation of alcohol intake and blocked the increase in progressive‐ratio responding for alcohol in dependent animals [94]. Chronic glucocorticoid receptor antagonism also blocked escalated and compulsive alcohol drinking during protracted abstinence in rats with a history of alcohol dependence. These results suggest a critical role for glucocorticoid receptors in the development and maintenance of alcohol dependence. Other brain neurotransmitter or neuromodulatory systems that have pro‐stress actions also converge on the extended amygdala, all of which may contribute to negative emotional states that are associated with drug withdrawal or protracted abstinence [80]. Chronic administration of psychostimulants and opioids has long been hypothesized to activate cyclic adenosine monophosphate response element binding protein, which in turn activates dynorphin in medium spiny neurons that in turn feedback and decrease the activity of ventral tegmental dopamine neurons [95]. Although κ‐opioid receptor agonists suppress nondependent drinking, presumably via aversive stimulus effects, κ‐opioid receptor antagonists block the excessive drinking associated with alcohol withdrawal and dependence, and this effect may be mediated by the shell of the nucleus accumbens [73].  Other neurotransmitter/neuromodulatory systems that comprise the brain stress system in the extended amygdala include ­vasopressin, hypocretin (orexin), substance P, and ­neuroimmune factors. In addition to CRF and dynorphin, there is evidence that norepinephrine, vasopressin, substance P, and hypocretin (orexin) may all contribute to negative emotional states of drug withdrawal, particularly alcohol withdrawal (Figure  53.3). Thus, activation of this pro‐stress, pro‐negative emotional state system is multi‐determined and comprises the neurochemical bases for hedonic opponent processes [96]. Neurotransmitter systems that are implicated in anti‐stress actions include NPY, nociceptin, and endocannabinoids. NPY has powerful orexigenic and anxiolytic effects and has been hypothesized to act in opposition to the actions of CRF in addiction [84]. Nociceptin and synthetic ­nociception receptor agonists have effects on GABA synaptic activity in the central nucleus of the amygdala that are similar to NPY and can block high alcohol consumption in a genetically selected line of rats that is known to be hypersensitive to s­ tressors [97]. Evidence also implicates endocannabinoids in the regulation of affective states, in which reductions of cannabinoid CB1 receptor



53:  Alcoholic Liver Disease

signaling produce anxiogenic‐like behavioral effects [98]. Thus, vulnerability to AUD may involve not only a sensitized stress system but also a hypoactive stress buffer system, and behavioral and pharmacological interventions that block pro‐ stress and stimulate anti‐stress (described in Neural substrates for the preoccupation/anticipation stage associated with AUD) systems may be particularly interesting targets for the treatment of AUD. In summary, two processes are hypothesized to form the neurobiological basis for the withdrawal/negative affect stage: loss of function in the reward systems (within‐system neuroadaptation) and recruitment of the brain stress systems (between‐­ system neuroadaptation) [71]. As dependence and withdrawal develop, brain stress systems, such as CRF, norepinephrine, and dynorphin, are recruited, producing aversive or stress‐like states [99]. The combination of decreases in reward neurotransmitter function and recruitment of brain stress systems provides a powerful motivation for reengaging in drug taking and drug seeking.

695

hypothesized to reflect the neural representations of reward and incentive salience constructs [104]. One parsimonious view of the human imaging data that is consistent with animal model data is that there is a “go” system in the dorsal prefrontal/­ cingulate cortex that drives impulsivity and craving and a “stop” system in ventromedial prefrontal cortex that inhibits impulsivity and craving [105]. In humans, stress and stressors have long been associated with relapse and the vulnerability to relapse [106]. Individuals with addiction are hypersensitive to pain during withdrawal, particularly in the face of negative affect [107]. Indeed, the leading precipitant of relapse is negative emotion/ affect including elements of anger, frustration, sadness, anxiety, and guilt [108]. Studies of “craving” in animal models can be divided into three domains: alcohol seeking that is induced by the drug itself, alcohol priming‐induced reinstatement, alcohol seeking that is induced by stimuli that are paired with drug taking, cue‐ and context‐induced reinstatement, and alcohol seeking that is induced by an acute stressor or a state of stress (i.e. stress‐ induced reinstatement) [109]. Most evidence from animal studies suggests that drug‐induced reinstatement is localized to the Neural substrates for the preoccupation/ medial prefrontal cortex/ventral striatum circuit and mediated anticipation stage associated with AUD by the neurotransmitter glutamate [110]. Neurotransmitter sysThe preoccupation/anticipation or “craving” stage of the addic- tems that are involved in drug‐induced reinstatement involve a tion cycle has long been hypothesized to be a key part of the glutamatergic projection from the frontal cortex to the nucleus neurocircuitry that mediates relapse in humans. Dysregulation accumbens that is modulated by dopamine activity in the frontal of the frontal cortex mediates not only elements of impulsivity cortex (Figure  53.3). In contrast, neuropharmacological and and compulsivity but also protracted abstinence and craving. neurobiological studies that use animal models of cue‐induced Human imaging studies reveal neurocircuitry dysregulation reinstatement involve the basolateral amygdala as a critical subduring the preoccupation/anticipation stage in AUD that strate, with a possible feed‐forward mechanism possibly through includes not only compromises in frontal cortical executive the same prefrontal cortex system that is involved in drug‐ function but also dysregulated substrates that mediate craving. induced reinstatement [109, 111]. Cue‐induced reinstatement Lower frontal cortex activity parallels deficits in executive func- involves dopamine modulation in the basolateral amygdala and tion in neuropsychologically challenging tasks in individuals a glutamatergic projection to the nucleus accumbens from both with AUD with and without Wernicke–Korsakoff’s syndrome the basolateral amygdala and ventral subiculum [111]. Such (for review, see references [100, 101]). In individuals with AUD, reinstatement in alcohol drinking can be blocked by the systhere are impairments in the maintenance of spatial information, temic administration of naltrexone and a selective μ and δ the disruption of decision making, and impairments in behavio- ­opioid receptor antagonist [109]. These results are consistent ral inhibition. Such frontal cortex‐derived executive function with human studies that showed that opioid receptor antagonists disorders in AUD have been linked to deficits in the ability of may blunt the urge to drink that is elicited by the presentation of behavioral treatments to effect recovery from AUD post‐­ alcohol‐related cues in individuals with AUD [112]. In contrast, detoxification [100]. Thus, deficits in the prefrontal cortical the stress‐induced reinstatement of drug‐related responding in control of incentive salience may represent a key mechanism to animal models appears to depend on the activation of both CRF explain individual differences in the vulnerability to AUD, and and norepinephrine in elements of the extended amygdala the excessive attribution of incentive salience to drug‐related (­central nucleus of the amygdala and bed nucleus of the stria cues and residual hypersensitivity of the brain stress systems terminalis) [113]. may perpetuate excessive drug intake, compulsive behavior, and relapse. Compulsivity in AUD: a negative Craving responses to cues in human imaging studies activate reinforcement perspective an overarching cognitive control network in the brain, involving dorsolateral prefrontal, anterior cingulate, and parietal cortices, Compulsivity in AUD can derive from neurocircuitry changes at all of which support a broad range of executive functions [102]. all three stages of the addiction cycle (Figure 53.3). In the binge/ Such studies have led to the hypothesis that a frontal–cingulate– intoxication stage, such changes may involve neurocircuits that parietal–subcortical cognitive control network is consistently include enhanced incentive salience [75] and the engagement of recruited across a range of traditional executive function tasks, pathological habit function [114]. In the withdrawal/negative many of which show deficits in humans with AUDs. Indeed, affect stage, such changes involve the recruitment of neurocirthe most prominent activation by alcohol‐related cues involves cuits that are involved in negative emotional states. In the preoc­ the dorsolateral prefrontal cortex, cingulate cortex, and orbito- cupation/anticipation stage, such changes involve impairments frontal cortex [103]. Such drug cue‐evoked responses are in executive function and the dysregulation of inhibitory control

696

THE LIVER:  TREATMENT OF AUD PATIENTS WITH ALD

that regulates impulsivity [115]. Many of the cue‐ and stress‐ related neuroadaptations persist into protracted abstinence, largely described as dysregulation that persists beyond acute withdrawal [73]. However, negative emotional states of acute and protracted abstinence that are induced by chronic high‐dose alcohol have been largely neglected in the clinical domain. Negative emotional states may strongly impact compulsivity not only by directly driving negative reinforcement but also by enhancing the value of incentive salience, enhancing the value of habit expression, or exacerbating impairments in executive function. Thus, the overall conceptual theme that is argued herein is that moderate to severe AUD represents a break with homeostatic brain regulatory mechanisms that regulate the emotional state of the individual. One hypothesis is that the dysregulation of ­emotion begins with the binge and subsequent acute withdrawal but leaves a residual neuroadaptive trace that allows rapid “re‐addiction” even months and years after detoxification and abstinence [79]. Thus, the emotional dysregulation of alcohol addiction represents more than the simple homeostatic dysregulation of hedonic function; it also represents a dynamic break with the homeostasis of this system that has been termed allostasis [79]. A further argument is that this hypernegative emotional state, termed hyperkatifeia, sensitizes over time, extends into protracted abstinence, and provides the driving force for another source of motivation for prolonging and maintaining addiction, that of negative reinforcement (Figure 53.3).

TREATMENT OF AUD PATIENTS WITH ALD While reductions below harmful drinking levels is beneficial, the cornerstone in the treatment of patients with ALD remains total alcohol abstinence, as abstinence can improve outcome at all stages of liver disease [116]. However, there are only a few ­randomized studies that have investigated behavioral and/or pharmacological treatments in patients with AUD and ALD.

Behavioral treatments Brief interventions are aimed at educating the patient about harmful drinking, increasing motivation to change behavior, and reinforcing skills to address problematic drinking. In primary care settings, studies support the use of brief interventions to reduce drinking [117]. Furthermore, these interventions may synergize with other areas of treatment, such as, compliance to medications and referral to treatment programs, an approach referred to as screening, brief intervention and referral to treat­ ment (SBIRT). Specific psychosocial and behavioral treatments for AUD include 12‐step facilitation, cognitive behavioral therapy (CBT), and motivational enhancement therapy (MET) [116]. Twelve‐step facilitation focuses on abstinence and involves Alcoholics Anonymous meetings. The goal of CBT is to develop coping mechanisms where alcohol is replaced by alcohol‐free circumstances. MET helps patients work through the dilemma by “rolling with the resistance” to change [116]. Specific studies investigating behavioral treatments for AUD in

patients with ALD are very limited. Lieber et  al. [118] conducted a clinical pharmacology study of liver fibrosis and showed that the addition of a brief intervention resulted in a significant reduction in alcohol consumption. Other studies have investigated behavioral platforms, like CBT and MET, as ways to facilitate alcohol abstinence in patients with ALD. A systematic review of these studies was conducted by Khan et al. [119]. A total of 13 papers were selected for a whole sample of 1945 patients. This systematic review indicated that combining medical care and behavioral approaches, like CBT and MET, facilitated alcohol abstinence, a conclusion that supports the importance of multidisciplinary approaches to treat patients with AUD and ALD [119].

Medications Pharmacological approaches to treat AUD patients include medications to treat alcohol withdrawal syndrome (AWS) and those used to help patients reduce their craving, cut their drinking, facilitate abstinence, and prevent relapse. As for the treatment of AWS, it is important to keep in mind that while up to 50% of individuals with AUD present with alcohol withdrawal symptoms after they stop drinking, only a small percentage requires pharmacological treatment. For these patients, benzodiazepines are the gold standard because they represent the only class of medications that reduces the risk of withdrawal seizures and/or delirium tremens [120]. In those individuals with AUD and ALD who develop AWS, a symptom‐triggered schedule with lorazepam or oxazepam is preferred. In fact, these benzodiazepines do not undergo phase I biotransformation; rather, they only undergo glucuronidation, which is preserved even if liver function is compromised [116]. A better understanding of the neurobiology of addiction, reviewed above in this chapter, has played an important role in the development of medications for AUD, as summarized in Table 53.1. In the United States, acamprosate, disulfiram, and naltrexone (oral and intramuscular) are approved by the Food and Drug Administration (FDA) for treatment of AUD. A recent meta‐analysis supports the efficacy of naltrexone and acamprosate, but not disulfiram, for AUD [121]. Furthermore, nalmefene was recently approved in Europe for the treatment of AUD, but it is not approved in the United States. Not only is the number of approved medications for AUD very limited, but also their use in patients with ALD is even more narrow. In fact, disulfiram may cause hepatotoxicity and is not recommended in patients with ALD. Naltrexone’s potential for causing hepatotoxicity also exists, even if it seems to be rare. Acamprosate has not formally been tested in AUD patients with liver disease, but it does not undergo hepatic metabolism and there are no reports of hepatotoxicity [116]. In the past decades, preclinical and clinical studies have ­provided support for some medications as potential novel treatments for AUD. While none of these medications are FDA‐ approved, some of them have shown efficacy for AUD in phase 2/3 trials. Among them, the most promising are baclofen, gabapentin, ondansetron, topiramate, and varenicline, as summarized in Table  53.1. However, formal clinical trials testing these medications in AUD patients with ALD are lacking, except for baclofen. In fact, while general clinical trials testing baclofen



53:  Alcoholic Liver Disease

697

Table 53.1  FDA‐approved medications and others tested in alcohol use disorder patientsa Dosage

Pharmacological target

FDA‐approved medications for alcohol use disorder Acamprosate 666 mg TID Possibly NMDA receptor agonist Disulfiram 250–500 mg QD Inhibition of acetaldehyde dehydrogenase PO: 50 mg QD μ opioid receptor antagonist Naltrexoneb PO or IM IM: 380 mg monthly Not FDA‐approved medications tested for alcohol use disorder Baclofen 10 mg TID; GABAB receptor agonist 80 mg QD max Gabapentin

900–1800 mg QD

Unclear – the most likely mechanism is blockade of voltage‐dependent Ca2+ channels. 5HT3 antagonist

Ondansetron

1–16 mcg /kg−1 BID

Topiramate

300 mg QD

Anticonvulsant multiple targets: −Glutamate/+GABA

Varenicline

2 mg QD

Nicotinic acetylcholine receptor partial agonist

Possible use in alcohol use disorder patients with alcoholic liver disease? Yes (no hepatic metabolism) No (hepatic metabolism; cases of liver toxicity have been reported) With caution (perceptions of liver toxicity limit use in advanced alcoholic liver disease) Yes (minimal hepatic metabolism) Baclofen has been formally tested in clinical studies with alcohol use disorder patients with liver cirrhosis Yes (no hepatic metabolism) Yes, but with caution because liver toxicity has been reported, albeit relationship to ondansetron administration is not determined Yes (partial hepatic metabolism mostly by glucuronidation) In patients with hepatic encephalopathy, use with caution: topiramate‐ related cognitive side‐effects may confound the clinical course and treatment of hepatic encephalopathy Yes (minimal hepatic metabolism)

 Reproduced from [116] with permission of Elsevier.  Nalmefene is not‐FDA approved but was recently approved in Europe for alcohol use disorder. Compared to naltrexone, nalmefene has a longer half‐life and no evidence of hepatotoxicity. FDA, Food and Drug Administration; TID, three times a day; NMDA, N‐methyl‐D‐aspartate; QD, once a day; PO, per os (oral); IM, intramuscular; GABA, gamma‐­ aminobutyric acid; BID, twice a day; HT, serotonin.

a b

have generated conflicting results [122], two independent clinical trials testing baclofen in AUD patients with ALD both ­support its efficacy in treating this specific subpopulation [123, 124]. However, the mechanisms by which baclofen may be efficacious, specifically in AUD patients with ALD, are unknown. Furthermore, although not formally tested in patients with ALD, the other medications mentioned above (especially gabapentin and varenicline), might also be useful in AUD individuals with ALD because they have no evidence of liver toxicity (for additional details, please see Table 53.1). In conclusion, the crucial goal in treating patients with AUD and ALD is to help them to achieve and maintain alcohol abstinence. As briefly summarized above, both behavioral and pharmacological treatments are available and may be effective in helping AUD patients to quit drinking. These treatments remain underutilized, and it is therefore very important that their use is expanded in clinical care, including in hepatology settings where many patients are referred for their ALD. Therefore, it is critical that addiction medicine is integrated into treatment and prescribed by the multidisciplinary teams that care for patients with AUD and ALD.

REFERENCES 1. Gao, B. and Bataller, R. Alcoholic liver disease: pathogenesis and new therapeutic targets. Gastroenterology, 2011;141:1572–85. 2. O’Shea, R.S., Dasarathy, S., McCullough, A.J. et al. Alcoholic liver disease. Hepatology, 2010;51:307–28. 3. Zakhari, S. and Li, T.K. Determinants of alcohol use and abuse: impact of  quantity and frequency patterns on liver disease. Hepatology, 2007; 46;2032–9.

4. Mandrekar, P., Bataller, R., Tsukamoto, H. et al. Alcoholic hepatitis: translational approaches to develop targeted therapies. Hepatology, 2016;64: 1343–55. 5. Chang, B., Xu, M.J., Zhou, Z. et al. Short‐ or long‐term high‐fat diet feeding plus acute ethanol binge synergistically induce acute liver injury in mice: an important role for CXCL1. Hepatology, 2015;62:1070–85. 6. Xu, J., Lai, K.K., Verlinsky, A. et al. Synergistic steatohepatitis by moderate obesity and alcohol in mice despite increased adiponectin and p‐AMPK. J Hepatol, 2011;55:673–82. 7. You, M., Considine, R.V., Leone, T.C. et  al. Role of adiponectin in the ­protective action of dietary saturated fat against alcoholic fatty liver in mice. Hepatology, 2005;42:568–77. 8. Bertola, A., Park, O., and Gao, B. Chronic plus binge ethanol feeding synergistically induces neutrophil infiltration and liver injury in mice: a critical role for E‐selectin. Hepatology, 2013;58:1814–23. 9. Xu, M.J., Cai, Y., Wang, H. et al. Fat‐specific protein 27/CIDEC promotes development of alcoholic steatohepatitis in mice and humans. Gastroenterology, 2015;149:1030–41. 10. Gao, B., Xu, M.J., Bertola, A. et al. Animal models of alcoholic liver disease: pathogenesis and clinical relevance. Gene Expr, 2017;17:173–86. 11. Cai, Y., Xu, M.J., Koritzinsky, E.H. et  al. Mitochondrial DNA‐enriched microparticles promote acute‐on‐chronic alcoholic neutrophilia and hepatotoxicity. JCI Insight, 2017;2. 12. Li, M., He, Y., Zhou, Z. et  al. MicroRNA‐223 ameliorates alcoholic liver injury by inhibiting the IL‐6‐p47phox‐oxidative stress pathway in neutrophils. Gut, 2017;66:705–15. 13. Hatton, J., Burton, A., Nash, H. et  al. Drinking patterns, dependency and life‐time drinking history in alcohol‐related liver disease. Addiction, 2009;104:587–92. 14. Askgaard, G., Gronbaek, M., Kjaer, M.S. et al. Alcohol drinking pattern and risk of  alcoholic liver cirrhosis: a prospective cohort study. J Hepatol, 2015;62:1061–7. 15. Trepo, E., Romeo, S., Zucman‐Rossi, J. et al. PNPLA3 gene in liver d­ iseases. J Hepatol, 2016;65:399–412. 16. Ajmera, V.H., Terrault, N.A., and Harrison, S.A. Is moderate alcohol use in nonalcoholic fatty liver disease good or bad? A critical review. Hepatology, 2017;65:2090–9. 17. Purohit, V., Gao, B., Song, B.J. Molecular mechanisms of alcoholic fatty liver. Alcohol Clin Exp Res, 2009;33:191–205.

698

THE LIVER:  REFERENCES

18. Parker, R., Kim, S.J., and Gao, B. Alcohol, adipose tissue and liver disease: mechanistic links and clinical considerations. Nat Rev Gastroenterol Hepatol, 2018;15:50–9. 19. Zhong, W., Zhao, Y., Tang, Y. et  al. Chronic alcohol exposure stimulates ­adipose tissue lipolysis in mice: role of reverse triglyceride transport in the pathogenesis of alcoholic steatosis. Am J Pathol, 2012;180:998–1007. 20. Sebastian, B.M., Roychowdhury, S., Tang, H. et  al. Identification of a cytochrome P4502E1/Bid/C1q‐dependent axis mediating inflammation in adipose tissue after chronic ethanol feeding to mice. J Biol Chem, 2011;286:35989–97. 21. Ji, C., Chan, C., and Kaplowitz, N. Predominant role of sterol response element binding proteins (SREBP) lipogenic pathways in hepatic steatosis in the murine intragastric ethanol feeding model. J Hepatol, 2006;45:717–24. 22. You, M., Fischer, M., Deeg, M.A. et al. Ethanol induces fatty acid synthesis pathways by activation of sterol regulatory element‐binding protein (SREBP). J Biol Chem, 2002;277:29342–7. 23. Baraona, E. and Lieber, C.S. Effects of ethanol on lipid metabolism. J Lipid Res, 1979;20:289–315. 24. Galli, A., Pinaire, J., Fischer, M. et al. The transcriptional and DNA binding activity of peroxisome proliferator‐activated receptor alpha is inhibited by ethanol metabolism. A novel mechanism for the development of ethanol‐ induced fatty liver. J Biol Chem, 2001;276:68–75. 25. O’Neill, H.M., Holloway, G.P., and Steinberg, G.R. AMPK regulation of fatty acid metabolism and mitochondrial biogenesis: implications for ­obesity. Mol Cell Endocrinol, 2013;366:135–51. 26. You, M., Matsumoto, M., Pacold, C.M. et  al. The role of AMP‐activated protein kinase in the action of ethanol in the liver. Gastroenterology, 2004;127:1798–1808. 27. Dolganiuc, A., Thomes, P.G., Ding, W.X. et  al. Autophagy in alcohol‐ induced liver diseases. Alcohol Clin Exp Res, 2012;36:1301–8. 28. Ding, W.X., Li, M., Chen, X. et al. Autophagy reduces acute ethanol‐induced hepatotoxicity and steatosis in mice. Gastroenterology, 2010;139:1740–52. 29. Nagy, L.E., Ding, W.X., Cresci, G. et al. Linking pathogenic mechanisms of  alcoholic liver disease with clinical phenotypes. Gastroenterology, 2016;150:1756–68. 30. Gao, B. and Tsukamoto, H. Inflammation in alcoholic and nonalcoholic fatty liver disease: friend or foe? Gastroenterology, 2016;150:1704–9. 31. Hartmann, P., Seebauer, C.T., and Schnabl, B. Alcoholic liver disease: the gut microbiome and liver cross talk. Alcohol Clin Exp Res, 2015;39: 763–75. 32. Lucey, M.R., Mathurin, P., and Morgan, T.R. Alcoholic hepatitis. N Engl J Med, 2009;360:2758–69. 33. Crabb, D.W., Bataller, R., Chalasani, N.P. et  al. Standard definitions and common data elements for clinical trials in patients with alcoholic hepatitis: recommendation from the NIAAA alcoholic hepatitis consortia. Gastroenterology, 2016;150:785–90. 34. Affo, S., Dominguez, M., Lozano, J.J. et al. Transcriptome analysis identifies TNF superfamily receptors as potential therapeutic targets in alcoholic hepatitis. Gut, 2013;62:452–60. 35. Michelena, J., Altamirano, J., Abraldes, J.G. et al. Systemic inflammatory response and serum lipopolysaccharide levels predict multiple organ failure and death in alcoholic hepatitis. Hepatology, 2015;62:762–72. 36. Louvet, A., Wartel, F., Castel, H. et al. Infection in patients with severe alcoholic hepatitis treated with steroids: early response to therapy is the key ­factor. Gastroenterology, 2009;137:541–8. 37. Bataller, R. and Gao, B. Liver fibrosis in alcoholic liver disease. Semin Liver Dis, 2015;35:146–56. 38. Jeong, W.I., Park, O., and Gao, B. Abrogation of the antifibrotic effects of natural killer cells/interferon‐gamma contributes to alcohol acceleration of liver fibrosis. Gastroenterology, 2008;134:248–58. 39. Yang, A.M., Inamine, T., Hochrath, K. et  al. Intestinal fungi contribute to development of alcoholic liver disease. J Clin Invest, 2017;127:2829–41. 40. Llopis, M., Cassard, A.M., Wrzosek, L. et  al. Intestinal microbiota contributes to individual susceptibility to alcoholic liver disease. Gut, ­ 2016;65:830–9. 41. Esfandiari, F., Medici, V., Wong, D.H. et al. Epigenetic regulation of hepatic endoplasmic reticulum stress pathways in the ethanol‐fed cystathionine beta synthase‐deficient mouse. Hepatology, 2010;51:932–41. 42. Satishchandran, A., Ambade, A., Rao, S. et al. MicroRNA 122, Regulated by GRLH2, protects livers of mice and patients from ethanol‐induced liver ­disease. Gastroenterology, 2018;154:238–52.

43. Szabo, G. and Momen‐Heravi, F. Extracellular vesicles in liver disease and potential as biomarkers and therapeutic targets. Nat Rev Gastroenterol Hepatol, 2017;14:455–66. 44. Saha, B., Momen‐Heravi, F., Furi, I. et al. Extracellular vesicles from mice with alcoholic liver disease carry a distinct protein cargo and induce ­macrophage activation via heat shock protein 90. Hepatology, 2018;67(5): 1986–2000. 45. Sutti, S., Bruzzi, S., and Albano, E. The role of immune mechanisms in alcoholic and nonalcoholic steatohepatitis: a 2015 update. Expert Rev ­ Gastroenterol Hepatol, 2016;10:243–53. 46. Liaskou, E., Klemsdal Henriksen, E.K., Holm, K. et  al. High‐throughput T‐cell receptor sequencing across chronic liver diseases reveals distinct ­disease‐associated repertoires. Hepatology, 2016;63:1608–19. 47. Mathews, S., Feng, D., Maricic, I. et al. Invariant natural killer T cells contribute to chronic‐plus‐binge ethanol‐mediated liver injury by promoting hepatic neutrophil infiltration. Cell Mol Immunol, 2016;13:206–16. 48. Cui, K., Yan, G., Xu, C. et al. Invariant NKT cells promote alcohol‐induced steatohepatitis through interleukin‐1beta in mice. J Hepatol, 2015; 62:1311–8. 49. Riva, A., Patel, V., Kurioka, A. et al. Mucosa‐associated invariant T cells link intestinal immunity with antibacterial immune defects in alcoholic liver ­disease. Gut, 2017;67(5):918–30. 50. Bertola, A., Mathews, S., Ki, S.H. et al. Mouse model of chronic and binge ethanol feeding (the NIAAA model). Nat Protoc, 2013;8:627–37. 51. Chao, X., Wang, S., Zhao, K. et  al. Impaired TFEB‐mediated lysosome ­biogenesis and autophagy promote chronic ethanol‐induced liver injury and steatosis in mice. Gastroenterology, 2018;155(3):865–79. 52. Wang, W., Xu, M.J., Cai, Y. et al. Inflammation is independent of steatosis in a murine model of steatohepatitis. Hepatology, 2017;66:108–23. 53. Khanova, E., Wu, R., Wang, W. et al. Pyroptosis by caspase11/4‐gasdermin‐ D pathway in alcoholic hepatitis. Hepatology, 2017;67(5):1737–53. 54. European Association for the Study of Liver. EASL clinical practical guidelines: management of alcoholic liver disease. J Hepatol, 2012;57:399–420. 55. Alatalo, P., Koivisto, H., Puukka, K. et  al. Biomarkers of liver status in heavy  drinkers, moderate drinkers and abstainers. Alcohol Alcohol, 2009;44:199–203. 56. Aday, A.W., Mitchell, M.C., Casey, L.C. Alcoholic hepatitis: current trends in management. Curr Opin Gastroenterol, 2017;33:142–8. 57. Ku, N.O., Strnad, P., Bantel, H. et al. Keratins: Biomarkers and modulators of apoptotic and necrotic cell death in the liver. Hepatology, 2016; 64:966–76. 58. Mortele, K.J. and Ros, P.R. Imaging of diffuse liver disease. Semin Liver Dis, 2001;21:195–212. 59. Borra, R.J., Salo, S., Dean, K. et al. Nonalcoholic fatty liver disease: rapid evaluation of liver fat content with in‐phase and out‐of‐phase MR imaging. Radiology, 2009;250:130–6. 60. Stickel, F., Datz, C., Hampe, J. et al. Pathophysiology and management of alcoholic liver disease: update 2016. Gut Liver, 2017;11:173–88. 61. Parker, R., Armstrong, M.J., Corbett, C. et al. Systematic review: pentoxifylline for the treatment of severe alcoholic hepatitis. Aliment Pharmacol Ther, 2013;37:845–54. 62. Nguyen‐Khac, E., Thevenot, T., Piquet, M.A. et  al. Glucocorticoids plus N‐acetylcysteine in severe alcoholic hepatitis. N Engl J Med, 2011 ;365:1781–9. 63. Piechota, M. and Piechota, A. An evaluation of the usefulness of extracorporeal liver support techniques in patients hospitalized in the ICU for severe liver dysfunction secondary to alcoholic liver disease. Hepat Mon, 2016;16:e34127. 64. Wang, H.J., Gao, B., Zakhari, S. et  al. Inflammation in alcoholic liver ­disease. Annu Rev Nutr, 2012;32:343–68. 65. Lieber, S.R., Rice, J.P., Lucey, M.R. et al. Controversies in clinical trials for alcoholic hepatitis. J Hepatol, 2018;68(3):586–92. 66. Xu, M.J., Zhou, Z., Parker, R. et al. Targeting inflammation for the treatment of alcoholic liver disease. Pharmacol Ther, 2017;180:77–89. 67. Kong, X., Feng, D., Mathews, S. et  al. Hepatoprotective and anti‐fibrotic functions of interleukin‐22:therapeutic potential for the treatment of ­alcoholic liver disease. J Gastroenterol Hepatol, 2013;28(1):56–60. 68. Gao, B. and Shah, V.H. Combination therapy: new hope for alcoholic hepatitis? Clin Res Hepatol Gastroenterol, 2015;39(1):S7–11. 69. Kubiliun, M., Patel, S.J., Hur, C. et al. Early liver transplantation for alcoholic hepatitis: ready for primetime? J Hepatol, 2018;68(3):380–2.



53:  Alcoholic Liver Disease

70. Mitchell, M.C., Friedman, L.S., and McClain, C.J. Medical management of  severe alcoholic hepatitis: expert review from the clinical practice updates  committee of the AGA Institute. Clin Gastroenterol Hepatol, 2017;15:5–12. 71. Koob, G.F. and Le Moal, M. Drug abuse: hedonic homeostatic dysregulation. Science, 1997;278:52–8. 72. Koob, G.F. Allostatic view of motivation: implications for psychopathology, in Motivational Factors in the Etiology of Drug Abuse, (eds. R.A. Bevins and M.T. Bardo), (series title: Nebraska Symposium on Motivation, Vol 50). University of Nebraska Press, Lincoln NE, 2004, pp. 1–18. 73. Koob, G.F. Neurocircuitry of alcohol addiction: synthesis from animal ­models, in Alcohol and the Nervous System (eds. E.V. Sullivan and A.  Pfefferbaum), (series title: Handbook of Clinical Neurology, vol 125). Elsevier, Amsterdam, 2014, pp. 33–54. 74. Weiss, F., Lorang, M.T., Bloom, F.E. et al. Oral alcohol self‐administration stimulates dopamine release in the rat nucleus accumbens: genetic and motivational determinants. J Pharmacol Exp Ther, 1993;267:250–8. 75. Robinson, T.E. and Berridge, K.C. The neural basis of drug craving: an incentive‐sensitization theory of addiction. Brain Res Brain Res Rev, 1993;18:247–91. 76. Haber, S.N., Fudge, J.L., and McFarland, N.R. Striatonigrostriatal pathways in primates form an ascending spiral from the shell to the dorsolateral striatum. J Neurosci, 2000;20:2369–82. 77. Lovinger, D.M. and Kash, T.L. Mechanisms of neuroplasticity and ethanol’s effects on plasticity in the striatum and bed nucleus of the stria terminalis. Alcohol Res, 2015;37:109–24. 78. Solomon, R.L. and Corbit, J.D. An opponent‐process theory of motivation. I. Temporal dynamics of affect. Psychol Rev, 1974;81:119–45. 79. Koob, G.F. and Le Moal, M. Drug addiction, dysregulation of reward, and allostasis. Neuropsychopharmacology, 2001;24:97–129. 80. Koob, G.F. and Le Moal, M. Addiction and the brain antireward system. Annu Rev Psychol, 2008;59:29–53. 81. Melis, M., Spiga, S., and Diana, M. The dopamine hypothesis of drug addiction: hypodopaminergic state. Int Rev Neurobiol, 2005;63:101–54. 82. Diana, M., Pistis, M., Carboni, S. et al. Profound decrement of mesolimbic dopaminergic neuronal activity during ethanol withdrawal syndrome in rats: electrophysiological and biochemical evidence. Proc Natl Acad Sci USA, 1993;90:7966–9. 83. Volkow, N.D., Fowler, J.S., Wang, G.J. et  al. Imaging dopamine’s role in drug abuse and addiction. Neuropharmacology, 2009;56(1):3–8. 84. Kyzar, E.J. and Pandey, S.C. Molecular mechanisms of synaptic remodeling in alcoholism. Neurosci Lett, 2015;601:11–9. 85. Heimer, L. and Alheid, G.F. Piecing together the puzzle of basal forebrain anatomy. Adv Exp Med Biol, 1991;295:1–42. 86. Alheid, G.F. and Heimer, L. New perspectives in basal forebrain organization of special relevance for neuropsychiatric disorders: the striatopallidal, amygdaloid, and corticopetal components of substantia innominata. Neuroscience, 1988;27:1–39. 87. LeDoux, J.E. Emotion circuits in the brain. Annu Rev Neurosci, 2000;23:155–84. 88. Neugebauer, V., Li, W., Bird, G.C. et al. The amygdala and persistent pain. Neuroscientist, 2004;10:221–34. 89. Zorrilla, E.P., Logrip, M.L., and Koob, G.F. Corticotropin releasing factor: a key role in the neurobiology of addiction. Front Neuroendocrinol, 2014;35:234–44. 90. Roberto, M., Cruz, M.T., Gilpin, N.W. et al. Corticotropin releasing factor‐ induced amygdala gamma‐aminobutyric acid release plays a key role in ­alcohol dependence. Biol Psychiatry, 2010;67:831–9. 91. Vendruscolo, L.F. and Koob, G.F. Alcohol dependence conceptualized as a stress disorder, in: Oxford Handbook of Stress and Mental Health, (eds. K.  Harkness and E.P. Hayden), Oxford University Press, New York, 2018; in press. 92. Kirkman, S. and Nelson, D.H. Alcohol‐induced pseudo‐Cushing’s disease: a  study of prevalence with review of the literature. Metabolism, 1988;37:390–4. 93. Richardson, H.N., Lee, S.Y., O’Dell, L.E. et al. Alcohol self‐administration acutely stimulates the hypothalamic‐pituitary‐adrenal axis, but alcohol dependence leads to a dampened neuroendocrine state. Eur J Neurosci, 2008;28:1641–53. 94. Vendruscolo, L.F., Barbier, E., Schlosburg, J.E. et al. Corticosteroid‐dependent plasticity mediates compulsive alcohol drinking in rats. J  Neurosci, 2012;32:7563–71.

699

95. Carlezon, W.A., Jr., Nestler, E.J., and Neve, R.L. Herpes simplex virus‐ mediated gene transfer as a tool for neuropsychiatric research. Crit Rev Neurobiol, 2000;14:47–67. 96. Koob, G.F. Neurobiology of addiction, in Textbook of Substance Abuse Treatment, 5th edn, (eds. M. Galanter et  al.), American Psychiatric Publishing, Washington, DC, 2015, pp. 3–24. 97. Economidou, D., Hansson, A.C., Weiss, F. et al. Dysregulation of nociceptin/orphanin FQ activity in the amygdala is linked to excessive alcohol drinking in the rat. Biol Psychiatry, 2008;64:211–8. 98. Serrano, A. and Parsons, L.H. Endocannabinoid influence in drug ­reinforcement, dependence and addiction‐related behaviors. Pharmacol Ther, 2011;132:215–41. 99. Koob, G.F. Alcoholism: allostasis and beyond. Alcohol Clin Exp Res, 2003;27:232–43. 100. Sullivan, E.V. and Pfefferbaum, A. Neurocircuitry in alcoholism: a substrate of disruption and repair. Psychopharmacology, 2005;180:583–94. 101. Oscar‐Berman, M. Function and dysfunction of prefrontal brain circuitry in alcoholic Korsakoff’s syndrome. Neuropsychol Rev, 2012;22:154–69. 102. Niendam, T.A., Laird, A.R., Ray, K.L. et al. Meta‐analytic evidence for a superordinate cognitive control network subserving diverse executive functions. Cogn Affect Behav Neurosci, 2012;12:241–68. 103. Olbrich, H.M., Valerius, G., Paris, C. et al. Brain activation during craving for alcohol measured by positron emission tomography. Aust N Z J Psychiatry, 2006;40:171–8. 104. Jasinska, A.J., Stein, E.A., Kaiser, J. et al. Factors modulating neural reactivity to drug cues in addiction: a survey of human neuroimaging studies. Neurosci Biobehav Rev, 2014;38:1–16. 105. Koob, G.F. and Volkow, N.D. Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry, 2016;3:760–73. 106. Marlatt, G.J. Determinants of relapse: implications for the maintenance of behavioral change, in Behavioral Medicine: Changing Health Lifestyles, (eds. P. Davidson and S. Davidson), Brynner/Mazel, New York, 1980, pp. 410–52. 107. Jochum, T., Boettger, M.K., Burkhardt, C. et al. Increased pain sensitivity in alcohol withdrawal syndrome. Eur J Pain, 2010;14:713–8. 108. Zywiak, W.H., Connors, G.J., Maisto, S.A. et al. Relapse research and the reasons for drinking questionnaire: a factor analysis of Marlatt’s relapse taxonomy. Addiction, 1996;91: S121–30. 109. Martin‐Fardon, R. and Weiss, F. Modeling relapse in animals. Curr Top Behav Neurosci, 2013;13:403–32. 110. McFarland, K. and Kalivas, P.W. The circuitry mediating cocaine‐induced reinstatement of drug‐seeking behavior. J Neurosci, 2001;21:8655–63. 111. Everitt, B.J. and Wolf, M.E. Psychomotor stimulant addiction: a neural systems perspective. J Neurosci, 2002;22:3312–20. 112. Monti, P.M., Rohsenow, D.J., Hutchison, K.E. et al. Naltrexone’s effect on cue‐elicited craving among alcoholics in treatment. Alcohol Clin Exp Res, 1999;23:1386–94. 113. Shaham, Y., Shalev, U., Lu, L. et  al. The reinstatement model of drug relapse: history, methodology and major findings. Psychopharmacology, 2003;168:3–20. 114. Everitt, B.J., Belin, D., Economidou, D. et al. Review. Neural mechanisms underlying the vulnerability to develop compulsive drug‐seeking habits and addiction. Philos Trans R Soc Lond B Biol Sci, 2008;363:3125–35. 115. Jentsch, J.D. and Taylor, J.R. Impulsivity resulting from frontostriatal dysfunction in drug abuse: implications for the control of behavior by reward‐ related stimuli. Psychopharmacology, 1999;146:373–90. 116. Leggio, L. and Lee, M.R. Treatment of alcohol use disorder in patients with alcoholic liver disease. Am J Med, 2017;130:124–34. 117. Bertholet, N., Daeppen, J.B., Wietlisbach, V. et  al. Reduction of alcohol consumption by brief alcohol intervention in primary care: systematic review and meta‐analysis. Arch Intern Med, 2005;165:986–95. 118. Lieber, C.S., Weiss, D.G., Groszmann, R. et al. II. Veterans affairs cooperative study of polyenylphosphatidylcholine in alcoholic liver disease. Alcohol Clin Exp Res, 2003;27:1765–72. 119. Khan, A., Tansel, A., White, D.L. et al. Efficacy of psychosocial interventions in inducing and maintaining alcohol abstinence in patients with chronic liver disease: a systematic review. Clin Gastroenterol Hepatol, 2016;14:191–202. 120. Leggio, L., Kenna, G.A., and Swift, R.M. New developments for the ­pharmacological treatment of alcohol withdrawal syndrome. A focus on non‐benzodiazepine GABAergic medications. Prog Neuropsychopharmacol Biol Psychiatry, 2008;32:1106–17.

700

THE LIVER:  REFERENCES

121. Jonas, D.E., Amick, H.R., Feltner, C. et  al. Pharmacotherapy for adults with alcohol use disorders in outpatient settings: a systematic review and meta‐analysis. JAMA, 2014;311:1889–900. 122. Litten, R.Z., Falk, D.E., Ryan, M.L. et al. Advances in pharmacotherapy development: human clinical studies. Handb Exp Pharmacol, 2018; 248:579–613. 123. Addolorato, G., Leggio, L., Ferrulli, A. et al. Effectiveness and safety of baclofen for maintenance of alcohol abstinence in alcohol‐dependent

patients with liver cirrhosis: randomised, double‐blind controlled study. Lancet, 2007;370:1915–22. 124. Morley, K.C., Baillie, A., Fraser, I. et  al. Baclofen in the treatment of ­alcohol dependence with or without liver disease: multisite, randomised, double‐blind, placebo‐controlled trial. Br J Psychiatry, 2018; 212:362–9. 125. Koob, G.F. and Volkow, N.D. Neurocircuitry of addiction. Neuropsycho­ pharmacology, 2010;35:217–38.

54

Drug‐Induced Liver Injury Lily Dara and Neil Kaplowitz Research Center for Liver Disease, Department of Medicine, Division of Gastrointestinal and Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA

INTRODUCTION Drug‐induced liver injury (DILI) is an increasingly recognized clinical problem encompassing over 50% of cases of acute liver failure (ALF) in the United States [1]. First described as a distinct clinical entity by Hyman Zimmerman in the 1950s, DILI is defined as liver injury with alanine aminotransferase (ALT) more than 3–5 times the upper limit of normal (ULN) or an alkaline phosphatase (ALP) more than twice the ULN with or without elevated bilirubin. DILI remains a diagnostic challenge due to a lack of pathognomonic findings and objective biomarkers and a wide range of presentation. DILI has a wide spectrum of clinical manifestations from acute hepatitis with jaundice to cholestatic DILI to rarer forms of DILI such as nodular regenerative hyperplasia (e.g. azathioprine), sinusoidal obstruction syndrome (e.g. cyclophosphamide), steatohepatitis (e.g. tamoxifen), among others (Table 54.1). DILI can have a rapid or subacute presentation with elevated liver enzymes or occur more gradually and in rare instances present as a more chronic progressive disease [2]. As such, DILI is a diagnosis of exclusion and is suspected when there is temporal association between taking a drug and the onset of liver injury. Despite the various potential presentations of drug hepatotoxicity, most often DILI presents as drug‐induced hepatitis or cholestatic injury or a combination of the two. Most drugs have a unique signature which can be determined by calculating the R‐value. The R‐ value is defined as the ratio of ALT/ULN divided by the ALP/ ULN. It is commonly used as an index to determine the phenotype of liver injury. By convention, an R‐value ≥ 5 suggests a drug‐induced hepatitis, an R‐value ≤ 2 suggests cholestatic DILI, and a value in between (2 < R‐value < 5) suggests a mixed picture  [3]. The phenotype of liver injury also has prognostic

value. Drug‐induced hepatitis, when accompanied by jaundice, results in 10% mortality without transplant. This phenomenon, first observed and reported by Hyman Zimmerman, is often referred to as “Hy’s law”. Drug‐induced cholestasis is a more indolent disease, associated with elevated ALP often due to cholangiocyte damage and may be accompanied by pruritus. Cholestatic DILI is a signature of certain drugs, is more common in the elderly and often resolves more slowly [4]. In the past few decades, hundreds of different drugs and herbal compounds have been reported to cause DILI, with variable latency, patterns of injury, and phenotypic presentation. The underlying mechanism of liver injury and cell death has also been extensively explored in vivo and in vitro using different agents, most notably with the most widely used hepatotoxin, APAP. In this chapter we will focus on the molecular mechanisms and signaling pathways activated in DILI. We will not cover herbal and dietary supplement‐induced liver injury and although these compounds are not studied as systematically, it is likely that the same principles and mechanisms apply.

DIRECT ACTING VERSUS IDIOSYNCRATIC TOXICITY The liver is an important target for drugs because lipophilic drugs are metabolized in the liver by conversion to more aqueous soluble forms which can be excreted. Drug metabolism involves the participation of phase I cytochrome P (Cyp)‐450 system, phase II conjugation, and phase III transporter proteins, which modulate transformation and excretion. Although parent drug accumulation may participate, usually toxic intermediates

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

702

THE LIVER:  ACETAMINOPHEN

Table 54.1  Phenotype and histopathological presentations of DILI Clinical phenotype

Drugs

Hepatocellular injury with non‐specific acute hepatitis on biopsy, zonal necrosis, or spotty necrosis (mild to severe injury)

Acetaminophen, bromfenac, bupropion, carbon tetrachloride, cocaine, diclofenac, halothane, heparin, INH, ipilimumab, ketoconazole, methotrexate, methyldopa, pembrolizumab, phenytoin, propylthiouracil, nivolumab, rifampin, statins, troglitazone Methyldopa, minocycline, nitrofurantoin, statins Amiodarone, methotrexate, tamoxifen Anti‐psychotics, MTP (microsomal triglyceride transfer protein) inhibitors, protease inhibitors Aspirin (Reye’s syndrome), NRTI, tetracycline, valproic acid Allopurinol, amoxicillin/clavulanic acid, carbamazepine, chlorpromazine, hydralazine, methyldopa, penicillamine, phenylbutazone, phenytoin, procainamide, quinidine, sulfasalazine, sulfonamides, sulindac OCP, cyclosporin A, anabolic steroids Chlorpromazine, chlorpropamide, chlorothiazide, rarely others Anabolic steroids, azathioprine, OCPs (Budd–Chiari syndrome), HIV drugs, busulfan, cyclophosphamide, pyrrolizidine alkaloids, 6‐mercaptopurine Busulfan, cyclophosphamide, pyrrolizidine alkaloids

Hepatocellular injury with features of autoimmune hepatitis on biopsy Hepatocellular injury with, steatohepatitis +/− fibrosis on biopsy Hepatocellular injury with macrovesicular steatosis Hepatocellular injury with microvesicular steatosis Cholestasis and inflammation Bland cholestasis Chronic cholestasis with ductopenia Vascular lesions Portal hypertension and peliosis hepatitis, NRH, or SOS Vascular lesions Portal hypertension and sinusoidal obstruction syndrome on biopsy Adenoma Angiosarcoma Hepatocellular carcinoma

participate in mediating toxicity. Certain drugs such as APAP, aspirin, niacin, and chemotherapeutic agents can cause predictable, dose‐dependent injury directly to the liver. This direct and metabolic toxicity is unique to certain drugs as most drugs cause an idiosyncratic, unpredictable, and dose‐independent form of hepatotoxicity often abbreviated IDILI (idiosyncratic DILI). Among the direct hepatotoxins, APAP is the mostly widely studied drug. First introduced by von Mering in 1893, APAP was not widely used until the 1960s [5]. The first hepatotoxicity cases were reported in the same decade by Davidson and Eastham who described two patients with fulminant liver failure and centrilobular hepatic necrosis who subsequently succumbed to APAP toxicity within three days [6]. The toxicity from APAP was shown to be dose dependent and the drug was initially considered safe up to 4 g per 24 hours. More recent studies on healthy volunteers receiving a daily dose of 4 g of APAP have called the concept of a universal safe threshold into question [7]. In addition to dose, fasting, acute illness, concomitant drugs, and metabolic stress can result in liver injury from lower doses of APAP, due to depletion of glutathione (GSH) stores or modulation of biotransformation to toxic intermediary. In contrast to direct toxins, drugs that cause IDILI do so by activating the immune system. The metabolism of drugs can lead to the formation of intermediates which generate haptenic‐peptide antigens. A number of studies have now shown that patients who develop IDILI harbor specific single nucleotide polymorphisms (SNPs) in genes which encode for human leukocyte antigen (HLA) regions. This association strongly suggests a genetically determined immune predisposition in certain individuals leading to an adaptive immune response directed at drug hapten or modified autoantigens. However, most of the identified HLA polymorphisms are relatively common in the general population and thus carry a small hazard ratio for the development of IDILI. Therefore, other contributing factors must be influencing susceptibility to IDILI [8].

Anabolic steroids, oral contraceptives Anabolic steroids, arsenic, copper, polyvinyl chloride, thorotrast Aflatoxin, anabolic steroids, danazol

ACETAMINOPHEN APAP is the single most common cause of ALF in the United States. [9]. Hepatotoxicity from APAP occurs in a dose‐dependent, predictable manner, much like the human injury, in rodents treated with large doses of the drug. Rats are more resistant to APAP; however mice are susceptible and exhibit characteristic histologic findings, congestion around the central vein, followed by vacuolization, and necrosis by 6 hours [5]. Mouse models of APAP toxicity especially C57BL6 mice are commonly used to study APAP DILI and hepatocyte death. However, marked mouse strain differences in susceptibility to APAP have been reported [10]. APAP is the precursor to the toxic metabolite, NAPQI (N‐acetyl‐p‐benzoquinone imine), formed by the direct two electron oxidation of the parent drug by the Cyps [11]. Multiple Cyps can generate NAPQI but the most common enzyme is Cyp2E1. However, Cyp1A2, Cyp3A4, and Cyp2D6 may also participate. NAPQI is highly reactive and covalently binds to intracellular proteins resulting in organelle stress. At nontoxic doses, NAPQI is efficiently detoxified by GSH forming an APAP‐GSH conjugate [12]. GSH synthesis is limited by cysteine availability as the other two amino acids of GSH, glutamate and glycine, are abundant. When GSH is depleted and cysteine supply is limited, the NAPQI is free to attack protein‐ SH groups throughout the cell, with the mitochondria being the key target [13]. This led to the development of the antidote, N‐acetyl‐cysteine (Mucomyst®), which provides cysteine to the liver and is highly effective if administered within 10 hours of overdose in humans and 1.5–2 hours in mice [14]. In the absence of GSH, NAPQI covalently binds intracellular proteins, and APAP–protein adducts can be detected in the liver and serum. The appearance of APAP–protein adducts in serum correlates with hepatotoxicity and serum AST, ALT levels. These adducts have been shown by immunoblot analysis to originate from the liver [15]. This has led to the development of a highly sensitive and specific HPLC‐EC assay for detection of acetaminophen protein adducts (3‐cysteine‐acetaminophen in proteins) in the



54:  Drug‐Induced Liver Injury

serum as specific biomarkers of lysis of hepatocytes in APAP DILI [16, 17]. Despite the excellent correlation between NAPQI‐protein adduct formation and toxicity, no direct causality between adduct formation and hepatocyte necrosis has been demonstrated. The removal of APAP–protein adducts though selective autophagy occurs within 24 hours [18]. NAPQI–protein adducts in the presence of depleted GSH stores result in mitochondrial damage and increase of intracellular peroxides and reactive oxygen species (ROS) and via an iron‐mediated mechanism generate hydroxyl radical (Fenton reaction), as well as, peroxynitrite through the interaction of superoxide with mitochondrial nitric oxide (NO) [19]. Hinson and colleagues have demonstrated the appearance of 3‐nitro‐tyrosine (a marker of nitrogen stress) in mouse models of APAP in the centrilobular zone, mainly in the mitochondria [20]. Furthermore, the development of the nitrated proteins in hepatocytes correlated with the necrotic cell death [20]. The importance of mitochondrial damage in the pathogenesis of APAP‐induced necrosis has been known since the 1980s when electron microscopic examination of livers from acetaminophen treated mice indicated mitochondrial damage [21]. Functionally, APAP has been shown to alter mitochondrial respiration and effect the electron transport chain (ETC) both in vitro and in isolated hepatocytes from in vivo treated animals [22, 23]. Generation of ROS subsequently exacerbates mitochondrial stress and may induce endoplasmic reticulum (ER) stress leading to the activation of intracellular signaling pathways, most importantly the mitogen activated protein kinase (MAPK) pathway [24]. Interference with various MAPK proteins, including mixed lineage kinase protein 3 (MLK3), apoptosis signal inducing kinase (ASK1), mitogen activated protein kinase 4 (MKK4), and C Jun‐N‐terminal kinase (JNK) as well as the JNK binding partner, SH3BP (Sab) markedly protects against APAP induced hepatocyte death [23, 25–29]. Xenobiotic stress, nutrient deprivation, or organelle stress leads to the activation of upstream higher order MAPKs which through cascading phosphorylation events ultimately phosphorylate JNK. Under normal conditions, JNK activation by default is transient, as sustained p‐JNK leads to cell death. When stress signals exceed a certain threshold (such as toxic doses of APAP), p‐JNK interacts with mitochondria by binding to the kinase interacting motif of Sab, a mitochondrial outer membrane protein [30]. The interaction of JNK with Sab leads to the generation of ROS and sustained JNK activation, resulting in a feedforward mechanism. JNK does not enter the mitochondria. However, an intra‐mitochondrial signaling pathway involving the tyrosine phosphatase SH2 phosphatase 1 (SHP1), resulting in dephosphorylation of activated‐Src (proto‐oncogene c‐Src; Src short for sarcoma), which occurs on and requires the platform, DOK4 (an inner membrane protein) has been demonstrated [31]. The deactivation of Src is thought to increase mitochondrial ROS generation by dampening electron transport leading to the buildup of reducing equivalents. The JNK‐ dependent increased mitochondrial ROS then amplifies mitochondrial stress and leads to ROS sensitive mitochondrial permeability transition (MPT) (Figure  54.1). MPT inhibitors such as cyclosporine A inhibit APAP toxicity in vitro and in vivo in mice [32, 33]. Cyclophilin D deficient mice are also been reported to be protected from APAP toxicity and cell death,

703

further consolidating the importance of mitochondria in this form of DILI [34]. Following MPT a combination of collapse of mitochondrial ATP production and release of DNA damaging proteins from the mitochondria result in cell death. In addition to the MAPKs, other kinases and signaling molecules have been implicated in hepatocyte death from APAP. Knockout, knockdown or inhibition of glycogen synthase kinase 3 beta (GSK3β), receptor interacting protein kinase‐1 (RIPK1), and dynamin related protein‐1 (DRP1) have all been shown to abrogate or prevent hepatocyte death from APAP [35–37]. While receptor interacting protein kinase 1 (RIPK1) is thought to participate in APAP toxicity upstream of JNK, the role of receptor interacting protein kinase‐3 (RIPK3) is more controversial [38]. Although APAP‐DILI results in necrotic cell death which is clearly regulated through signaling mechanisms, it is not a form of necroptosis, as mixed lineage kinase domain like (MLKL) knockout does not protect against APAP toxicity [36].

IDIOSYNCRATIC DILI Idiosyncratic DILI (IDILI) remains a major diagnostic challenge. It is a diagnosis of exclusion and common liver disorders such as viral hepatitis, autoimmune hepatitis (AIH), biliary tract disease, ischemic hepatitis, heart failure, and circulatory dysfunction as well as sepsis and alcoholic liver disease must be ruled out. If a high index of suspicion is present for drug toxicity, the more common causes of liver disease are ruled out, and the drug is taken within the correct time frame (latency), and the injury phenotypically resembles the drug’s signature (if known), the probability of an IDILI event is high. A liver biopsy can be helpful if eosinophilic infiltrates, granulomas, or centrilobular necrosis is noted. Since there is no specific test to determine causality and even the liver biopsy is usually nonspecific, the diagnosis of IDILI remains challenging [8]. The exact pathophysiology of IDILI and why it occurs is likely multifactorial involving both drug and pharmacologic factors such as dose and lipophilicity as well as host and individual factors such as genetics, the immune response, and defective adaptive responses. Evidence for the involvement of the immune system in IDILI is strong. For example, certain drugs such as halothane, dihydralazine, and anticonvulsants are associated with an allergic type hypersensitivity and present with rash and eosinophilia. Other notable drugs that have been reported to present with rash and eosinophilia include: trimethoprim/sulfamethoxazole, cefazolin, and ciprofloxacin [39]. Severe allergic skin reactions such as toxic epidermal necrosis (TEN) and Stevens–Johnson syndrome (SJS) have been seen with carbamazepine and phenytoin IDILI and carry a poor prognosis [39]. Some drugs, such as nitrofurantoin and minocycline can cause DILI which mimics autoimmune hepatitis with a positive anti‐nuclear antibody (ANA) and plasma cell infiltrates which are often indistinguishable from AIH on liver biopsy [39]. This has led to the hypothesis that these drugs may be activating AIH in patients with a genetic predisposition [40]. However, the frequencies of AIH‐associated HLA alleles, DRB1*0301 and DRB1*0401, were not increased in patients with AIH type DILI presentation compared to controls [41].

704

THE LIVER:  IDIOSYNCRATIC DILI

Figure 54.1  Signaling pathway in acetaminophen (APAP) induced liver injury. APAP is metabolized by cytochrome P450 2E1 (CYP2E1) to a reactive metabolite called NAPQI N‐acetyl‐p‐benzoquinone imine which depletes glutathione (GSH) followed by covalent binding to intracellular proteins which induces cellular and organelle stress. NAPQI targets mitochondria resulting in the generation of reactive oxygen species (ROS). ROS subsequently activate the mitogen activated protein kinase (MAPK) cascade. Mixed‐lineage protein kinase 3 (MLK3) is activated in the early phase and apoptosis signal‐inducing kinase‐1 (ASK1) in the later phase of injury. MLK3 and ASK1 phosphorylate mitogen activated protein kinase 4 (MKK4) which goes on to phospho‐activate c‐Jun N‐terminal kinase (JNK). Other kinases such as receptor interacting protein kinase 1 (RIPK1), glycogen synthase kinase‐3β (GSK3β) and protein kinase C‐α (PKCα) have also been reported to activate JNK. Activated JNK (p‐JNK) binds to SH3BP (Sab) on the outer mitochondrial membrane. This results in the activation of an intra‐mitochondrial pathway involving a SH2 phosphatase (SHP1) and the docking protein DOK4 (located on the inner membrane of mitochondria) which ultimately results in the dephosphorylation of intra‐mitochondrial Src. The deactivation of Src impairs mitochondrial electron transport, increasing mitochondrial ROS generation. The JNK‐ dependent increased mitochondrial ROS then amplifies mitochondrial stress and sustains p‐JNK in an active form. This eventually leads to mitochondrial permeability transition (MPT) and release of inter‐mitochondrial proteins which cleave nuclear DNA.

Large databases of IDILI patients have looked for evidence of a genetic predisposition to IDILI and interestingly many large genome wide association studies (GWAS) have identified polymorphism in HLA regions in patients with IDILI due to various agents. Since the HLA complex codes for the major histocompatibility complex (MHC) proteins in humans, variants in HLA molecules can result in aberrant antigen presentation and inappropriate activation of an immune response. The HLA‐A, HLA‐B, and HLA‐C genes govern the structure of MHC I molecules, while HLA‐DP, HLA‐DQ, and HLA‐DR control the structure of MHC II. Most identified SNPs are in the region of MHC II molecules. This can affect antigen presentation to T helper cells (CD4+) and thus partly explain why certain individuals develop IDILI.

Further supporting the contribution of immunity to DILI are lymphocyte stimulation tests demonstrating drugs or drug–­ protein adducts can activate peripheral blood lymphocytes from affected individuals [42, 43]. Early studies using the lymphocyte transformation test, a simple in vitro assay based on assessment of lymphocyte proliferative responses in drug‐treated and vehicle control cultures, detected drug‐specific lymphocyte responses in approximately 50% of patients with DILI [44]. Drug‐specific T cells from IDILI patients have been shown to be activated in an HLA restricted manner, also suggesting an adaptive immune pathogenesis [42]. Using healthy donors with known HLA risk alleles for certain drugs (HLA‐B*15:02, HLA‐B*57:01, and HLA‐B*58:01) and donors without the risk mutations, priming of naïve T cells was demonstrated to skew



54:  Drug‐Induced Liver Injury

toward donors expressing the previously identified specific HLA alleles in GWAS studies [45]. T cells were co‐cultured with mature dendritic cells and first treated for eight days with a low dose of the drugs nitroso sulfamethoxazole, carbamazepine, flucloxacillin, oxypurinol, and piperacillin. T cells were re‐stimulated with autologous dendritic cells at higher doses of the drugs and interferon‐γ (IFN‐γ) was measured, and proliferation assays were carried out. Most drugs preferentially activated T cells with risk alleles [45]. However, responses were detected in rare cases in donors not expressing HLA risk alleles, suggesting that certain drugs bind to multiple MHC molecules to ­activate T cells. These panels of peripheral blood mononuclear cells (PBMCs) from donors with diverse HLAs may prove quite useful in predicting the immunogenicity of new drugs [45]. In addition to adaptive immune cells, monocytes derived from patients with IDILI events have also been studied for drug immunogenicity. The monocytes were cultured for 10 days and re‐incubated with the drugs the patients were exposed to and toxicity was measured based on release of lactate dehydrogenase. This assay identified ten drugs as IDILI toxins including amoxicillin‐clavulanate and diclofenac among others [46]. Importantly 13 patients had a recurrence of DILI after inadvertent re‐exposure to a single drug, and the monocyte assay correctly identified 12 of the 13 drugs that caused DILI [46]. Despite clear links between HLA polymorphisms and immune activation from drugs, most people with the at risk alleles do not develop IDILI. In fact, some of these HLA polymorphisms are quite common in the general population [8]. Therefore, in line with the multifactorial pathogenesis of IDILI it is important to recognize that additional contributing factors must be present for IDILI to occur. A failure of clinical adaptation due to a loss of immune‐tolerance in the liver has been hypothesized as one of these factors (see Adaptation). Although the role of the immune system in IDILI appears prominent various preclinical tests systems have been able to demonstrate cellular and organelle stress with known IDILI drugs. IDILI is by definition unpredictable and hepatotoxicity risk is difficult to predict with most drugs. Therefore, it is intriguing that many drugs reported to cause IDILI exhibit danger signals in in vitro cytotoxicity assays. Multiple studies using IDILI drugs have shown stress signals such as GSH depletion, oxidative stress, covalent binding, BSEP inhibition, ATP depletion, mitochondrial dysfunction, ER stress, and cell death using systems such as micropattern co‐cultured primary hepatocytes, hepatoma cell lines, isolated primary mouse mitochondria, human membrane vesicles, iPSC derived hepatocytes, cryopreserved human hepatocytes, and more recently liver organoids [47–52]. Porceddu and colleagues have reported an association between in vitro induction of mitochondrial stress (as measured by cytochrome c release, swelling, and collapse of membrane potential) and the development of IDILI [48]. After screening 124 chemical compounds, 87 of which were known causes of IDILI, a striking correlation was found between the known at risk drugs and mitochondrial toxicity with a sensitivity of 92–94% and high positive predictive value of 82–89% for a hepatotoxic outcome [48]. This evidence of direct stress and toxicity to cell systems does not explain the idiosyncrasy of these compounds but could perhaps be construed as a surrogate marker of IDILI. Additionally, whether these directly toxic

705

read‐outs have any bearing to the clinical scenario in idiosyncratic toxicity, such as predisposing or sensitizing to immunogenic IDILI, remains to be determined.

HYPOTHESES FOR IDILI There are multiple hypotheses for the mechanism of immune activation in patients with susceptible haplotypes. The “hapten/ prohapten hypothesis”, the pharmacological interaction hypothesis or “p‐i hypothesis”, the “altered peptide repertoire ­hypothesis”, the “multiple determinant hypothesis”, and the “inflammatory stress hypothesis”. Interactions between antigen presenting cell (APC) binding grooves and proteins (covalent and non‐covalent) are considered to be important in activation of the adaptive immune system. The hapten hypothesis, the oldest and most commonly cited of the five hypotheses, postulates that the formation of reactive metabolites by phase I enzymes (CYP450 system) and the subsequent binding of these drugs to intracellular proteins results in the formation of “neoantigens” causing organelle stress. Neoantigens presented to immune cells can be perceived as “foreign” in certain individuals with genetic susceptibility and possibly an impaired ability to evoke sufficient compensatory immune tolerance. HLA variants with enhanced affinity for neoantigens result in intensified presentation of the neoantigen to T cells, T‐cell activation, and consequently liver injury [53]. In addition to this adaptive immune response, haptenization, can result in the activation of innate immunity as well (in vitro) [54]. The hapten hypothesis does not fully explain why only a limited number of HLA associations have been observed for DILI events from any given drug. It would be plausible that several potential hapten peptides would be formed after processing to interact with multiple HLA alleles due to the presence of multiple binding sites in proteins. Therefore, alternative mechanisms to the hapten hypothesis have been proposed. The “p‐i hypothesis” postulates that certain drugs directly bind to T‐cell receptors or MHC molecules via non‐covalent interactions resulting in T cell activation without the presence of an antigenic peptide [55]. This hypothesis was developed to explain observed features of drug responsive T‐cell clones that were metabolically inactive and therefore unable to process drugs to form reactive metabolites. Based on the p‐i hypothesis, drugs could also interact with peptides that are already bound to MHC molecules, to activate T cells. It has been proposed that since many modern drugs are designed to interact with cellular receptors or directly to the MHC groove itself, it is not surprising to find a T‐cell receptor or MHC molecule that can bind certain drugs with sufficient affinity to generate an immune response given the diversity of MHC molecules [53]. IDILI from ximelagatran, a potent competitive inhibitor of thrombin, which thus prevents clotting by blocking polymerization of fibrinogen, is an example of the p‐i hypothesis. The drug was withdrawn from the market in 2006 following the observation of IDILI in 8% of patients treated for five weeks. Ximelagatran is not metabolized and does not covalently bind proteins to form neoantigens so the hapten hypothesis is not likely to be relevant to its mechanism of toxicity [56]. However, in vitro studies have

706

THE LIVER:  HLA POLYMORPHISMS

shown a direct inhibition of peptide binding to HLA DRB1*0701 pointing to the p‐i hypothesis as a possible mechanism for immune activation and DILI [53, 56]. The structure of ximelagatran, resembling a short peptide with the peptide bond able to fit in the arginine pocket of thrombin, may be important for the development of an immune‐mediated reaction. Some drugs such as flucloxacillin, a beta lactam antibiotic in the penicillin family used in Europe, can activate the immune system through more than one mechanism. Wuillemin and colleagues demonstrated this by generating flucloxacillin reacting T cells from untreated individuals with HLA‐B*57:01 positive and negative haplotypes to investigate the mechanisms of T cell activation [57]. Flucloxacillin’s metabolite can covalently bind HLA molecules suggesting immune activation by haptenization [57]. However, in individuals positive for the HL‐B*57:01 flucloxacillin was directly responsible for T cell clone activation suggesting a p‐i based stimulation of T cells which is restricted to this HLA haplotype [57]. Direct presentation of a drug to an HLA has also been reported with the anti‐epileptic carbamazepine, which has been associated with Stevens–Johnson syndrome, and with the nucleoside analogue abacavir [58, 59]. The altered peptide repertoire model suggests that small molecules can non‐covalently occupy the sites of peptide binding grooves of MHC proteins leading to an alteration of the specificities of the MHC‐peptide binding and resulting in erroneous presentation of self‐peptides that are not normally recognized by that MHC [60]. This is proposed to be the mechanism for abacavir hypersensitivity. Approximately 5% of Caucasian patients treated with abacavir develop hypersensitivity within six weeks of initial exposure. Abacavir hypersensitivity reactions have been shown to be restricted to the same HLA as flucloxacillin, HLA‐B*5701 (odds ration > 900) [61–63]. In fact, the presence of HLA‐B*5701, HLA‐DR7, and HLA‐DQ3 had a positive predictive value for abacavir hypersensitivity of 100%, and a negative predictive value of 97% [62]. HLA testing for this allele has significantly reduced the number of IDILI events with this drug [64]. Unlike the p‐i model, presentation of abacavir to T cells is dependent on processing factors, which would be expected only in the hapten model. Abacavir can bind the F pocket of the susceptible HLAs thereby altering the repertoire of peptides and ligands that are  bound and presented, rendering innocuous self‐antigens immunogenic [59]. The mechanism for organ sensitivity of ­ HLA‐B*5701, resulting in liver injury with flucloxacillin and skin injury with abacavir is not known. An alternative hypothesis, termed the multiple determinant hypothesis of IDILI, suggests that the concomitant presence of multiple risk factors (HLA polymorphisms, gender, age, underlying illness) could precipitate an IDILI event [65]. Halothane induced IDILI which is more common in older females with prior exposure to halothane is often cited as an example [65, 66]. Inflammatory stress during drug therapy has been suggested by some to contribute to the development of IDILI. Inflammatory stress is thought to alter tissue homeostasis, cellular signaling pathways, and affect drug metabolizing enzymes thereby predisposing to liver injury [67]. On the other hand, minor hepatic injury and cellular stress induced by certain drugs may progress to more serious injury as damaging inflammatory mediators are released. Concurrent inflammation may modify expression of drug transporters including the ABC transporters in hepatocytes

and result in drug accumulation [67]. Inflammatory stress could also predispose to IDILI by preventing adaptation in drug exposed individuals with minor drug disturbances who would develop tolerance under non‐stress conditions. These numerous ways in which underlying stress might affect liver homeostasis and drug metabolism and elimination have led to the inflammatory stress hypothesis. This hypothesis postulates that an episode of inflammation during drug therapy might decrease the threshold for drug toxicity, making a predisposed individual susceptible to IDIL, perhaps with or without the participation of the adaptive immune system [67].

HLA POLYMORPHISMS Since the 1990s associations between HLA polymorphisms and drug reactions have been described [68]. The association of certain HLA alleles with idiosyncratic reactions has long been viewed as a strong indication that IDILI events are mediated by the adaptive immune system. One of the first well‐described associations was discovered during GWAS of patients who developed DILI after receiving flucloxacillin versus those that did not (Table 54.2). Flucloxacillin, a semi‐synthetic penicillin that is commonly used to treat staphylococcal infection in Europe, is a well‐described cause of IDILI. A strong association (OR: 80.6) between cholestatic hepatitis from flucloxacillin and expression of HLA‐B*57:01 was described in a seminal study by Daly and colleagues and later confirmed by others [69]. PBMCs isolated from these patients show immune activation with re‐exposure to the drug and flucloxacillin can activate CD8+ T cells in drug naïve individuals with this allele [42, 57]. Ticlopidine which is a rare cause of cholestatic type IDILI occurs more frequently in the Japanese population. Another MHC class I molecule, HLA‐A*33:03 association was found to be a significant risk factor (OR: 13) for ticlopidine IDILI [70]. Interestingly HLA‐A*33:03 has also been described in patients with terbinafine and fenofibrate DILI [71]. Evidence for genetic predisposition resulting in IDILI has also been reported with amoxicillin‐clavulanate, a commonly prescribed antibiotic in the United States and Europe and the leading cause of IDILI in most registries. Amoxicillin‐clavulanate has been associated with IDILI in patients harboring multiple polymorphisms with  the strongest associations reported with class II HLA DRB1*15:01‐DRB5*0101, DQB1*06:02 (Table 54.2) [72]. Amoxicillin‐clavulanate can cause a severe hepatocellular injury or a more indolent cholestatic IDILI (the latter being more common in the elderly and males). The Spanish DILI registry has noted an increased frequency of class I HLAs, A*3002 and B*1801, in patients with hepatocellular injury compared to controls, while the presence of the DRB1*1501‐DQB1*0602 allele was significantly increased in patients presenting with cholestatic/mixed hepatotoxicity [72–76]. HLA DRB1*07 has been reported to be less common in patients that develop IDILI pointing to the possibility that certain polymorphisms may be protective against IDILI [77]. Up to 5% of IDILI patients suffer from SJS [78]. Carbamazepine SJS and cutaneous reactions has been associated with HLA‐B*15:02 in certain populations such as the Han Chinese population, Thai, and Malay patients (not Caucasians) [79]. In Japanese patients, significant associations



54:  Drug‐Induced Liver Injury

707

Table 54.2  Established HLA associations for drugs that are known to cause IDILI Drug

Phenotype

HLA

Population

Comments

Amoxicillin ‐ clavulanate

Spectrum Cholestatic to hepatocellular

Hepatocellular and mixed

UK UK Spain Spain Spain Spain Belgium–Scotland Belgium Belgium India Caucasians

DRB1*07 is a protective polymorphism, HLADRB1*1501 has been associated with PSC Hazard ration (HR): 0.18–9.3

Anti‐TB drugs Allopurinol

Hepatocellular and mixed

DQA*0102 DRB1*07 A*0201 A*3002 B*18:01 DQB1*0602 DRB1*1501 DRB5*0101 DQB1*0402 DQB1*0201 DQA1*0102 DQB1*0502 B*5801

Carbamazepine

HLA‐B*1502 HLA‐A*3101 B*08

DILI associated with rash and eosinophilia. SNP also associated with SJS/DRESS In SJS/DRESS commonly cholestatic

Clometacin

Spectrum Cholestatic to hepatocellular Hepatocellular

Korea, Han Chinese, Thailand, Japan, Portugal South‐East Asia Japan–Korea–Europe France

Diclofenac Fenofibrate

Hepatocellular Cholestatic

DRB1*13 A*3301

Flupirtine

Hepatocellular

Flucloxacillin

Cholestatic

Lapatinib

Hepatocellular and mixed

Lumiracoxib

Hepatocellular

Minocycline Nevirapine

Hepatocellular Hepatocellular

Pazopanib

Hepatocellular

DRB1*1601‐ DQB1*0502 B* 5701 DRB1*0701 DQB1*0303 DRB1*15 DQA1*0201 DQB1*0202 DRB1*0701 DRB1*1501 DQB1*0602 DRB5*0101 DQA1*0102 B*3502 DRB*01 DRB*0102 B*5801 B*57 01

Ticlopidine

Cholestatic

Tiopronin

Cholestatic

Terbinafine

Cholestatic and mixed

Ximelagatran

Hepatocellular

A*3303 A*3301 B*4403 DRB1*1302 DQB1*0604 Cw*1403 A*33 B*44 DR6 A*3301 DRB1*0701 DQA1*0102

between the haplotype HLA‐B*15:11 and carbamazepine‐ induced SJS/TEN has been reported (OR: 16.3) [80]. Numerous other HLA associations have been reported with IDILI drugs which are outlined in Table 54.2.

DRUG AND PATIENT FACTORS The pathogenesis of IDILI is multifactorial and depends on both host and drug factors. Patient factors such as age, gender, drug metabolism, and mutations in transporter and HLA molecules have all been reported to contribute to toxicity [8]. Drugs

Caucasian Caucasian (Hispanic and non‐Hispanic) Germany

HLA‐DQA1*0102 is protective

90% female, increased IgG in 80% and positive ANA in 60% of patients

HR: 18.7

Caucasian (Europe) Caucasian (Europe) Caucasian (Europe)

B*5701 is associated with abacavir rash also DRB1*15 is associated with reduced risk

Caucasian Caucasian Caucasian Caucasian

DRB1*0701 associated with Ximelagatran HR: 6.9–14.1

Caucasian Caucasian South Africa South Africa Mixed

80% female, 90% ANA+ Low CD4 protective HR:3

Japan Japan Japan Japan Japan Japan Japan Japan Japan Caucasian (Europe) China Caucasian (Europe)

Associated with amoxicillin‐clavulanate Multiple sclerosis HR: 5

Weak association in clinical trials for renal carcinoma Severe cholestatic DILI in Japanese patients more commonly than Caucasians HR:6.7–10.1

Elevated biliary enzymes lasted from 2 months to up to 10 years Same haplotype as with fenofibrate and ticlopidine DQA1*0102 associated with TB meds DRB1*0701 associated with AIH 2 HR:4.4

properties such as dosage and physical chemical properties are very good predictors of drugs which can cause IDILI. Drugs that are both taken at a high dose (greater than 100 mg day−1) and have a calculated octanol‐water partition coefficient (logP) > 3 rendering them lipophilic, have a higher positive predictive value for inducing toxicity [81, 82]. Despite these interesting associations host factors appear to be much more important in the development of IDILI, given that IDILI occurs in only a small subset of individuals exposed to the drug. Age and sex are well known risk factors for IDILI. Women tend to be at higher risk for DILI in many studies, although not all have corroborated this. What has been clearly demonstrated is that female sex is associated with hepatocellular injury, ALF,

708

THE LIVER:  ADAPTATION

and worse outcomes in IDILI [83, 84]. Higher incidence of toxicity in women has been reported with drugs that give an AIH type presentation such as nitrofurantoin and minocycline [39, 85]. Female sex has also been associated with toxicity from halothane, flucloxacillin, isoniazid, chlorpromazine, macrolides, and diclofenac [8]. Age is another determinant of toxicity. Valproic acid and aspirin are common causes of IDILI in the pediatric population (Reye’s syndrome), while toxicity from isoniazid (INH) and amoxicillin‐clavulanate is known to increase with age [86]. INH DILI is five times more common in patients over 50 compared to those under 35 [87]. Interestingly the phenotype of liver injury from amoxicillin‐clavulanate varies depending on age with younger age being associated with hepatocellular damage and older age presenting with a cholestatic or mixed phenotype [83]. Drugs undergo oxidation, hydroxylation, and other reactions mediated by the cytochrome P450 system (phase I) and conjugation/esterification processes such as sulfation, glucuronidation, and so on (phase II) to become more water soluble [8]. Parent drug or reactive metabolites generated with the CYP system can inhibit bile secretion, bile acid uptake, and transport as well as canalicular efflux. All three of these drug detoxification steps have been associated with IDILI. Diclofenac DILI for example, has been associated with CYP2C9 polymorphisms and INH toxicity has been associated with CYP2E1 [88, 89]. Changes in N‐acetylation have also been linked to propensity to drug toxicity. N‐acetyl transferase (NAT) activity modulates susceptibility to INH hepatotoxicity with slow acetylators having a higher incidence of IDILI from INH [90, 91]. Slow acetylation is also a risk factor for sulfonamide toxicity [92]. Contribution of NAT2 mutations to IDILI has been questioned for certain ethnicities [93]. In a recent meta‐analysis, NAT2 slow acetylator genotype was associated with IDILI from anti‐ tuberculosis drugs in East Asians, Indian, and Middle Eastern patients HR: 3.18 (95% CI: 2.49–4.07) but not in Caucasians [93]. In another report, no association between NAT2 mutations and IDILI was found by GWAS in a cohort of Indian patients compared to ethnicity matched controls [94]. Mutations in the metabolic enzymes glutathione S‐ transferases (GST) and manganese superoxide dismutase (MnSOD), have also been shown to be predisposing factors to IDILI [95, 96]. Patients with at least one UGT2B7 mutation are at significantly higher risk for diclofenac IDILI (OR: 8.5) [97]. Additionally mutations in uridine glucuronosyltransferase 2B7 (UGT2B7), CYP2C8, as well as the transporter MRP2 (multidrug associated protein) predispose to the formation and accumulation of reactive diclofenac metabolites and are associated with toxicity [97]. However, it should be emphasized that all of these studies are small and few have been confirmed. The DILIN genetic studies, including exome sequencing found no association with drug metabolism and transport across all drugs, so the conclusions from these rather small studies are tenuous. The rate limiting step in the clearance of lipophilic drugs is their excretion into bile [4]. In humans polymorphisms resulting in decreased expression or activity of the transporters MRP2, BSEP, and MDR1 have been identified [4, 98]. These mutations could lead to decreased drug clearance, affect protein structure, transporter expression, and precipitate drug‐induced cholestasis. For example, MDR1 polymorphisms have been associated

with increased plasma drug concentrations [99]. Certain drugs are substrates for the organic anion‐transporting polypeptides (OATPs) and can therefore bind to and inhibit their function. Examples include: fexofenadine, opioids, digoxin, pravastatin, rifampin, enalapril, and methotrexate [8]. Interestingly Oatp 1b2 knockout mice are resistant to toxicity from phalloidin and microcystin [100, 101]. OATPs are known to take up these toxins as well as the mushroom toxin, amanita. IDILI from some drugs has been associated with mutations in multidrug resistance‐associated proteins (MRPs or ABCC). These basolateral membrane efflux pumps function to efflux conjugated drug metabolites from nucleoside analogue drugs (zidovudine, lamivudine, stavudine), methotrexate, and 6‐metacaptopurine [98, 102]. Bosentan, cyclosporin A, rifamycin, sulindac, glibenclamide, and troglitazone conjugates can inhibit the bile salt export pump (BSEP), preventing bile acid excretion and leading to cholestatic DILI [103–107]. Aside from cyclosporine and estrogen, BSEP inhibitors largely are associated with hepatitis, not cholestasis, possibly suggesting retention of bile acids may aggravate (or sensitize to) immune‐mediated hepatocellular death in IDILI. In addition to BSEP, cyclosporine inhibits other transporters. In rats, cyclosporine inhibits bile salt‐independent bile flow by inhibiting Mrp2 and glutathione secretion [108]. Cyclosporin is also a substrate for MDR1 and can competitively inhibit the ABC transporters. Troglitazone conjugates are excreted into bile via Mrp2. It has been suggested in rat models that this may be an important factor in the pathogenesis of troglitazone induced cholestasis [104]. However, troglitazone‐ induced cholestatic DILI is exceedingly rare. Nevertheless, it remains possible that BSEP inhibition with this drug sensitizes to death receptor killing mediated by the innate and adaptive immune system. Although transporter inhibition occasionally is the main cause of drug‐induced bland cholestasis (e.g. cyclosporine and estrogens), the common occurrence of immuno‐ allergic features with many drugs which are associated with a cholestatic phenotype, including portal tract inflammatory infiltration suggests the cholangiocytes are an important target of the immune response. The protracted course of cholestatic hepatotoxicity and rare progression to ductopenia further argues for immune system targeting of cholangiocytes.

ADAPTATION While advances in genomics studies and the identification of HLA risk alleles have been exciting developments in the field, it is still unclear why only a subset of patients with the risk‐associated HLA alleles experience an IDILI event. Mild liver toxicity from IDILI drugs occurs far more frequently than severe toxicity with jaundice and ALF. In most instances clinical adaptation occurs and mild liver injury subsides despite continuation of the drug [109]. Consequently, a failure of clinical adaptation has been postulated to be a second defect necessary for the occurrence of an IDILI event. This is referred to as defective clinical adaptation or failure to dampen the initiating mechanisms of injury due to diminished adaptive responses [109] (Figure 54.2). The liver is a site of immune privilege and its role in introducing immune tolerance has been demonstrated by studies using



54:  Drug‐Induced Liver Injury

709

Figure 54.2  The pathogenesis of idiosyncratic drug‐induced liver injury (IDILI). Drugs or their metabolites can cause mild injury and cellular stress which can be considered an initial hazard or danger signal. Many drugs that activate the adaptive immune system do so in an HLA restricted manner. The majority of individuals do not have the susceptible HLA and will not develop IDILI. About 1–10% of patients have the at‐risk or susceptible HLA haplotype. Once the immune system is activated the majority of patients with the susceptible HLA will exhibit no injury or mild injury that resolves with continued exposure due to strong immune tolerance and sufficient adaptation. A minority of the susceptible individuals (0.01–1%) will develop clinically significant and overt IDILI. It should be noted that is possible that certain drugs activate the adaptive immune system in a more HLA promiscuous fashion.

porto‐venous application of immunogenic antigens prior to skin pricks, resulting in decreased allergic skin reactions [110]. This effect is due to the unique properties of the liver as a site of immune tolerance as diversion of portal flow from the liver abrogates this protective effect [111, 112]. The mechanisms of immune tolerance include control of antigen presentation, clonal apoptosis of specific T cells, and immune deviation by switching from a T‐helper 2 to a T‐helper 1 predominance in lymphocytes. The detailed mechanisms of the phenomenon in the liver have been extensively reviewed [113]. The most often cited and well described example of adaptation is the study from 1975 in which patients in a psychiatric hospital with a tuberculosis outbreak were treated with INH and followed prospectively with serum liver tests every four weeks for one year [114]. Although 38% of patients developed elevated liver tests during the course of INH therapy, the abnormalities subsided in the majority of patients despite continued treatment (even in patients with hyperbilirubinemia) [114]. Clinical adaptation, in this context may be related to the development and induction of a state of immune tolerance against the drug or its metabolite as an immunogen and overt DILI in certain individuals with predisposition may represent “defective adaptation”. The discovery of mechanistic pathways of IDILI has been hindered by the fact that IDILI events are unpredictable, not dose

related, and therefore irreproducible. Two independent laboratories were able to successfully develop animal models of IDILI by manipulating pathways of immune tolerance providing some evidence that aberrant check point signaling leading to a failure of adaptation may predispose to IDILI events [115, 116]. Metushi et al. examined the effect of knocking out the immune tolerance genes, Casitas B‐lineage lymphoma (Cbl‐b1), and programmed cell death 1 (PD‐1) on IDILI from amodiaquine. Knockout of Cbl‐b1 or PD‐1 resulted in increased liver injury from amodiaquine compared to wild‐type mice. However, the injury resolved despite continued treatment with amodiaquine (adaptation) [115]. The additional blockade of cytotoxic T‐lymphocyte‐­ associate protein 4 (CTLA4), a key inducer of immune‐­tolerance, in PD‐1−/− mice prior to amodiaquine resulted in more severe injury that did not resolve with continuation of the drug. Depletion of CD8+ T cells in this model prevented injury, indicating that IDILI to amodiaquine was mediated by cytotoxic CD8+ T lymphocytes [117]. The halothane hepatitis model was used by Chakaborty et al. to illustrate a similar point [116]. In wild‐type mice halothane treatment resulted in mild and self‐ limited injury that subsided (adaptation). Upon re‐challenge with halothane two weeks later another mild self‐limiting episode of hepatitis occurred. The authors targeted immune tolerance by depleting myeloid derived suppressor cells (MDSCs). MDSC

710

THE LIVER:  REFERENCES

depleted mice were treated with halothane one day after MDSC depletion and then again after two weeks. MDSC depleted mice displayed significant liver injury 10 days after the week two halothane re‐challenge as evidenced by liver cell necrosis, elevated ALT, increased eosinophilia, and T cell responses to halothane metabolites [116]. The mechanism of dampening of injury in the wild‐type mice is due to the significant immune‐suppressive effect of MDSCs on CD4+ and CD8+ T cells which was demonstrated to be dose dependent [116]. In this model injury was mediated by antibody dependent cytotoxicity as pretreatment with anti‐CD4 antibody (not anti‐CD8) abrogated the response to halothane. These studies clearly illustrate a role of the immune checkpoint and adaptive responses in controlling IDILI responses. These studies also show that due to built‐in redundancies, often multiple elements of the pathway need to be inhibited or mutated for an IDILI event to occur [109]. Recently a model of abacavir IDILI was developed using HLA‐B*57:01 transgenic(Tg) humanized mice [118]. When challenged with abacavir alone these mice do not develop liver injury. However simultaneous treatment with CpG‐oligonucleotide (CpG‐ODN), which is a pathogen associated molecular pattern (PAMP) and an immunostimulant activating innate immunity, resulted in IDILI and increased ALTs [118]. This effect was  unique to the HLA‐B*57:01 mice and was not seen in HLA‐B*57:03 animals. IDILI resulted in a mononuclear infiltrate consisting of CD3+CD8+ cells indicating a synergistic immune response between innate and adaptive immune cells [118]. Interestingly after an initial rise in ALT, peaking at day 7, the liver injury subsided and normalized by day 14 of continued drug exposure. CD8+ T cells in HLA‐B*57:01 mice highly expressed the immune tolerance marker programmed cell death receptor 1 (PD1). PD1 aka CD279 is a cell surface receptor that dampens T cell response by promoting self‐tolerance. To our knowledge this is the first time that the link between upregulation of PD‐1 and the development of immune‐tolerance accompanied by normalizing ALT values (clinical adaptation) has been demonstrated. On the other hand, using this model Cardone and colleagues demonstrated that HLA‐B*57:01Tg mice did not develop skin reactions (no liver injury) to abacavir unless CD4+ T cells and regulatory T cells (Tregs) in particular were depleted [119]. They show that Tregs suppress dendritic cell co‐ stimulation, and that CD4+ T cell depletion enhances dendritic cell maturation leading to CD8+ T cells infiltrating the skin, so that depletion of CD4+ T cells impairs immune tolerance. This recent work illustrates a model of idiosyncratic adaptive immune‐mediated drug reaction and the approach will lead to many new insights in host responses which govern the development of IDILI in genetically susceptible individuals.

CONCLUSIONS Hepatotoxicity from drugs and herbal supplements is the most common cause of ALF in the United States. Injury from drugs is divided into two main categories, direct toxicity from drug such as APAP, chemotherapeutic agents, niacin, and several drug related injuries (aspirin, heparin, cholestyramine) and idiosyncratic hepatotoxicity due to as antibiotics, anticonvulsants, and

NSAIDs. APAP‐induced DILI has been studied extensively and is known to be a form of necrotic cell death of hepatocytes due to mitochondrial damage. APAP necrosis involves the activation of multiple signaling pathways including, GSK3β, RIPK1, and the MAPK pathway (ASK1, MLK3, MKK4) culminating in a key role for JNK and its effect on mitochondria. Replenishment of GSH via administration of N‐acetyl cysteine (NAC) can rescue from liver failure in cases of acute overdose, if given in the early hours. IDILI is one of the most common reasons drugs are removed from the market. IDILI is not dose related, though a dose threshold of 50–100 mg has been identified, and by definition is unpredictable. It is a multifactorial disease. Its pathogenesis is impacted by the complex interplay between potentially immunogenic drugs and metabolites and the host’s immune response. Once a drug crosses a “threshold exposure level” which is determined by the drug dosage and lipophilicity it can result in immunogenicity. Host factors such as age, gender, activities of CYPs, drug transporters, and detoxifying enzymes can result in generation of toxic metabolites or retention of immunogenic drug ­compounds causing sufficient exposure to induce an immune response. The main mechanism of injury in IDILI is thought to involve the immune system. Evidence of the immune‐mediated nature of IDILI is multifold including the identification of at‐risk HLA polymorphisms for certain drugs, in vitro evidence of immune activation of T cells isolated from patients with at‐risk HLAs, and clear allergic features such as rash and eosinophilia with certain drugs. Despite a clear association between certain HLAs and IDILI most persons with these variants do not develop overt clinical injury. The majority of exposed individuals at risk for IDILI, do not develop liver injury. In those that do, the injury is often mild and resolves with continued treatment. This is likely due to the liver’s natural propensity to maintain a state of immune tolerance, which results in adaptation to foreign antigens, even in individuals with at‐risk HLA haplotype. This explains why despite many people carrying the at‐risk HLA polymorphisms only a minority of patients exposed to immunogenic drugs develop IDILI. Approaches comparing gene expression profiles in patients that develop injury but subsequently adapt versus those who develop overt injury (non‐adaptors) may help uncover new and relevant pathways in IDILI.

ACKNOWLEDGMENTS This work was supported by NIH grant K08DK109141(LD), R01DK067215 (NK), P30DK048522 (NK).

REFERENCES 1. Abboud, G. and Kaplowitz, N. Drug‐induced liver injury. Drug Saf, 2007;30:277–94. 2. Stine, J.G. and Chalasani, N. Chronic liver injury induced by drugs: a systematic review. Liver Int, 2015;35:2343–53. 3. Chalasani, N.P., Hayashi, P.H., Bonkovsky, H.L., et al. ACG clinical guideline: the diagnosis and management of idiosyncratic drug‐induced liver injury. Am J Gastroenterol, 2014;109:950–66; quiz 67.



54:  Drug‐Induced Liver Injury

  4. Padda, M.S., Sanchez, M., Akhtar, A.J., and Boyer, J.L. Drug‐induced cholestasis. Hepatology, 2011;53:1377–87.  5. Hinson, J.A., Roberts, D.W., and James, L.P. Mechanisms of acetaminophen‐induced liver necrosis. Handb Exp Pharmacol, 2010;369–405.   6. Davidson, D.G. and Eastham, W.N. Acute liver necrosis following overdose of paracetamol. Br Med J, 1966;2:497–9.   7. Watkins, P.B., Kaplowitz, N., Slattery, J.T. et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: a randomized controlled trial. JAMA, 2006;296:87–93.  8. Dara, Liu,.Z.‐X., and Kaplowitz, N. Pathogenesis of idiosyncratic drug induced liver injury, in Liver Pathophysiology, Therapies and Antioxidants, (ed. P. Muriel), Elsevier, Cambridge, 2017, pp. 88–96.   9. Lee, W.M.L., Anne, M., and Todd Stravitz, R. AASLD Position Paper: The Management of Acute Liver Failure: Update 2011. 10. Harrill, A.H., Watkins, P.B., Su, S. et  al. Mouse population‐guided resequencing reveals that variants in CD44 contribute to acetaminophen‐induced liver injury in humans. Genome Res, 2009;19:1507–15. 11. Jollow, D.J., Thorgeirsson, S.S., Potter, W.Z., Hashimoto, M., and Mitchell, J.R. Acetaminophen‐induced hepatic necrosis. VI. Metabolic disposition of toxic and nontoxic doses of acetaminophen. Pharmacology, 1974;12:251–71. 12. Mitchell, J.R., Jollow, D.J., Potter, W.Z., Gillette, J.R., and Brodie, B.B. Acetaminophen‐induced hepatic necrosis. IV. Protective role of glutathione. J Pharmacol Exp Ther, 1973;187:211–7. 13. Tirmenstein, M.A. and Nelson, S.D. Subcellular binding and effects on calcium homeostasis produced by acetaminophen and a nonhepatotoxic regioisomer, 3’‐hydroxyacetanilide, in mouse liver. J Biol Chem, 1989;264:9814–9. 14. Peterson, R.G. and Rumack, B.H. Treating acute acetaminophen poisoning with acetylcysteine. JAMA, 1977;237:2406–7. 15. Pumford, N.R., Hinson, J.A., Benson, R.W., and Roberts, D.W. Immunoblot analysis of protein containing 3‐(cystein‐S‐yl)acetaminophen adducts in serum and subcellular liver fractions from acetaminophen‐treated mice. Toxicol Appl Pharmacol, 1990;104:521–32. 16. Muldrew, K.L., James, L.P., Coop, L. et al. Determination of acetaminophen‐ protein adducts in mouse liver and serum and human serum after hepatotoxic doses of acetaminophen using high‐performance liquid chromatography with electrochemical detection. Drug Metab Dispos, 2002;30:446–51. 17. James, L.P., Alonso, E.M., Hynan, L.S. et al. Detection of acetaminophen protein adducts in children with acute liver failure of indeterminate cause. Pediatrics, 2006;118:e676–81. 18. Ni, H.M., McGill, M.R., Chao, X. et al. Removal of acetaminophen protein adducts by autophagy protects against acetaminophen‐induced liver injury in mice. J Hepatol, 2016;65:354–62. 19. Adamson, G.M. and Harman, A.W. Oxidative stress in cultured hepatocytes exposed to acetaminophen. Biochem Pharmacol, 1993;45:2289–94. 20. Hinson, J.A., Pike, S.L., Pumford, N.R., and Mayeux, P.R. Nitrotyrosine‐ protein adducts in hepatic centrilobular areas following toxic doses of ­acetaminophen in mice. Chem Res Toxicol, 1998;11:604–7. 21. Walker, R.M., Racz, W.J., and McElligott T.F. Scanning electron microscopic examination of acetaminophen‐induced hepatotoxicity and congestion in mice. Am J Pathol, 1983;113:321–30. 22. Donnelly, P.J., Walker, R.M., and Racz, W.J. Inhibition of mitochondrial ­respiration in vivo is an early event in acetaminophen‐induced hepatotoxicity. Arch Toxicol, 1994;68:110–8. 23. Win, S., Than, T.A., Han, D., Petrovic, L.M., and Kaplowitz, N. c‐Jun N‐terminal kinase (JNK)‐dependent acute liver injury from acetaminophen or tumor necrosis factor (TNF) requires mitochondrial Sab protein expression in mice. J Biol Chem, 2011;286:35071–8. 24. Uzi, D., Barda, L., Scaiewicz, V. et al. CHOP is a critical regulator of acetaminophen‐induced hepatotoxicity. J Hepatol, 2013;59:495–503. 25. Gunawan, B.K., Liu, Z.X., Han, D., Hanawa, N., Gaarde, W.A., and Kaplowitz, N. c‐Jun N‐terminal kinase plays a major role in murine acetaminophen hepatotoxicity. Gastroenterology, 2006;131:165–78. 26. Hanawa, N., Shinohara, M., Saberi, B., Gaarde, W.A., Han, D., and Kaplowitz, N. Role of JNK translocation to mitochondria leading to inhibition of mitochondria bioenergetics in acetaminophen‐induced liver injury. J Biol Chem, 2008;283:13565–77. 27. Sharma, M., Gadang, V., and Jaeschke, A. Critical role for mixed‐lineage kinase 3 in acetaminophen‐induced hepatotoxicity. Mol Pharmacol, 2012; 82:1001–7. 28. Nakagawa, H., Maeda, S., Hikiba, Y. et al. Deletion of apoptosis signal‐regulating kinase 1 attenuates acetaminophen‐induced liver injury by inhibiting c‐Jun N‐terminal kinase activation. Gastroenterology, 2008;135:1311–21.

711

29. Zhang, J., Min, R.W.M., Le, K. et  al. The role of MAP2 kinases and p38 kinase in acute murine liver injury models. Cell Death Dis, 2017; 8:e2903. 30. Chambers, J.W., Cherry, L., Laughlin, J.D., Figuera‐Losada, M., and Lograsso, P.V. Selective inhibition of mitochondrial JNK signaling achieved using peptide mimicry of the Sab kinase interacting motif‐1 (KIM1). ACS Chem Biol, 2011;6:808–18. 31. Win, S., Than, T.A., Min, R.W., Aghajan, M., and Kaplowitz, N. c‐Jun N‐­ terminal kinase mediates mouse liver injury through a novel Sab (SH3BP5)‐dependent pathway leading to inactivation of intramitochondrial Src. Hepatology, 2016;63:1987–2003. 32. Kon, K., Kim, J.S., Jaeschke, H., and Lemasters, J.J. Mitochondrial permeability transition in acetaminophen‐induced necrosis and apoptosis of cultured mouse hepatocytes. Hepatology, 2004;40:1170–9. 33. Masubuchi, Y., Suda, C., and Horie, T. Involvement of mitochondrial permeability transition in acetaminophen‐induced liver injury in mice. J Hepatol, 2005;42:110–6. 34. Ramachandran, A., Lebofsky, M., Baines, C.P., Lemasters, J.J., and Jaeschke, H. Cyclophilin D deficiency protects against acetaminophen‐induced oxidant stress and liver injury. Free Radic Res, 2011;45:156–64. 35. Shinohara, M., Ybanez, M.D., Win, S. et  al. Silencing glycogen synthase kinase‐3beta inhibits acetaminophen hepatotoxicity and attenuates JNK activation and loss of glutamate cysteine ligase and myeloid cell leukemia sequence 1. J Biol Chem, 2010;285:8244–55. 36. Dara, L., Johnson, H., Suda, J. et al. Receptor interacting protein kinase 1 mediates murine acetaminophen toxicity independent of the necrosome and not through necroptosis. Hepatology, 2015;Dec; 62:1847–57. 37. Ramachandran, A., McGill, M.R., Xie, Y., Ni, H.M., Ding, W.X., and Jaeschke, H. The receptor interacting protein kinase 3 is a critical early mediator of acetaminophen‐induced hepatocyte necrosis in mice. Hepatology, 2013;58(6):2099–108. 38. Dara, L., Liu, Z.X., and Kaplowitz, N. Questions and controversies: the role of necroptosis in liver disease. Cell Death Discov, 2016;2:16089. 39. Chalasani, N., Bonkovsky, H.L., Fontana, R. et al. Features and outcomes of 899 patients with drug‐induced liver injury: the DILIN prospective study. Gastroenterology, 2015;148:1340–52. 40. Liu, Z.X. and Kaplowitz, N. Immune‐mediated drug‐induced liver disease. Clin Liver Dis, 2002;6:755–74. 41. de Boer, Y.S., Kosinski, A.S., Urban, T.J. et  al. Features of autoimmune hepatitis in patients with drug‐induced liver injury. Clin Gastroenterol Hepatol, 2017;15:103–12 e2. 42. Monshi, M.M., Faulkner, L., Gibson, A. et  al. Human leukocyte antigen (HLA)‐B*57:01‐restricted activation of drug‐specific T cells provides the immunological basis for flucloxacillin‐induced liver injury. Hepatology, 2013;57:727–39. 43. Tsutsui, H., Terano, Y., Sakagami, C., Hasegawa, I., Mizoguchi, Y., and Morisawa, S. Drug‐specific T cells derived from patients with drug‐induced allergic hepatitis. J Immunol, 1992;149:706–16. 44. Maria, V.A., and Victorino, R.M. Diagnostic value of specific T cell ­reactivity to drugs in 95 cases of drug induced liver injury. Gut, 1997;41:534–40. 45. Usui, T., Faulkner, L., Farrell, J. et al. Application of in vitro T cell assay using human leukocyte antigen‐typed healthy donors for the assessment of drug immunogenicity. Chem Res Toxicol, 2018;31:165–7. 46. Benesic, A., Rotter, I., Dragoi, D., Weber, S., Buchholtz, M.L., and Gerbes, A.L. Development and validation of a test to identify drugs that cause ­idiosyncratic drug‐induced liver injury. Clin Gastroenterol Hepatol, 2018; 16(9):1488–94. 47. Khetani, S.R. and Bhatia, S.N. Microscale culture of human liver cells for drug development. Nat Biotechnol, 2008;26:120–6. 48. Porceddu, M., Buron, N., Roussel, C., Labbe, G., Fromenty, B., and Borgne‐ Sanchez, A. Prediction of liver injury induced by chemicals in human with a multiparametric assay on isolated mouse liver mitochondria. Toxicol Sci, 2012;129:332–45. 49. Ware, B.R., Berger, D.R., and Khetani, S.R. Prediction of drug‐induced liver injury in micropatterned co‐cultures containing iPSC‐derived human hepatocytes. Toxicol Sci, 2015;145:252–62. 50. Atienzar, F.A., Novik, E.I., Gerets, H.H. et al. Predictivity of dog co‐culture model, primary human hepatocytes and HepG2 cells for the detection of hepatotoxic drugs in humans. Toxicol Appl Pharmacol, 2014;275:44–61. 51. Morgan, R.E., Trauner, M., van Staden, C.J. et al. Interference with bile salt export pump function is a susceptibility factor for human liver injury in drug development. Toxicol Sci, 2010;118:485–500.

712

THE LIVER:  REFERENCES

52. Proctor, W.R., Foster, A.J., Vogt, J. et  al. Utility of spherical human liver microtissues for prediction of clinical drug‐induced liver injury. Arch Toxicol, 2017;91:2849–63. 53. Grove, J.I. and Aithal, G.P. Human leukocyte antigen genetic risk factors of drug‐induced liver toxicology. Expert Opin Drug Metab Toxicol, 2015;11: 395–409. 54. Megherbi, R., Kiorpelidou, E., Foster, B. et al. Role of protein haptenation in triggering maturation events in the dendritic cell surrogate cell line THP‐1. Toxicol Appl Pharmacol, 2009;238:120–32. 55. Pichler, W.J. Pharmacological interaction of drugs with antigen‐specific immune receptors: the p‐i concept. Curr Opin Allergy Clin Immunol, 2002;2:301–5. 56. Kindmark, A., Jawaid, A., Harbron, C.G. et  al. Genome‐wide pharmacogenetic investigation of a hepatic adverse event without clinical signs of immunopathology suggests an underlying immune pathogenesis. Pharmacogenomics J, 2008;8:186–95. 57. Wuillemin, N., Adam, J., Fontana, S., Krahenbuhl, S., Pichler, W.J., and Yerly, D. HLA haplotype determines hapten or p‐i T cell reactivity to flucloxacillin. J Immunol, 2013;190:4956–64. 58. Wei, C.Y., Chung, W.H., Huang, H.W., Chen, Y.T., and Hung, S.I. Direct interaction between HLA‐B and carbamazepine activates T cells in patients with Stevens‐Johnson syndrome. J Allergy Clin Immunol, 2012;129:1562–9. 59. Ostrov, D.A., Grant, B.J., Pompeu, Y.A. et al. Drug hypersensitivity caused by alteration of the MHC‐presented self‐peptide repertoire. Proc Natl Acad Sci USA, 2012;109:9959–64. 60. Illing, P.T., Vivian, J.P., Dudek, N.L. et al. Immune self‐reactivity triggered by drug‐modified HLA‐peptide repertoire. Nature, 2012;486:554–8. 61. Rauch, A., Nolan, D., Martin, A., McKinnon E., Almeida, C., and Mallal, S. Prospective genetic screening decreases the incidence of abacavir hypersensitivity reactions in the Western Australian HIV cohort study. Clin Infect Dis, 2006;43:99–102. 62. Mallal, S., Nolan, D., Witt, C. et  al. Association between presence of HLA‐B*5701, HLA‐DR7, and HLA‐DQ3 and hypersensitivity to HIV‐1 reverse‐transcriptase inhibitor abacavir. Lancet, 2002;359:727–32. 63. Saag, M., Balu, R., Phillips, E. et al. High sensitivity of human leukocyte antigen‐b*5701 as a marker for immunologically confirmed abacavir hypersensitivity in white and black patients. Clin Infect Dis, 2008;46:1111–8. 64. Hughes, D.A., Vilar, F.J., Ward, C.C., Alfirevic, A., Park, B.K., and Pirmohamed, M. Cost‐effectiveness analysis of HLA B*5701 genotyping in preventing abacavir hypersensitivity. Pharmacogenetics, 2004;14:335–42. 65. Li, A.P. A review of the common properties of drugs with idiosyncratic hepatotoxicity and the “multiple determinant hypothesis” for the manifestation of idiosyncratic drug toxicity. Chem Biol Interact, 2002;142:7–23. 66. Dugan, C.M., MacDonald, A.E., Roth, R.A., and Ganey, P.E. A mouse model of severe halothane hepatitis based on human risk factors. J Pharmacol Exp Ther, 2010;333:364–72. 67. Deng, X., Luyendyk, J.P., Ganey, P.E., and Roth, R.A. Inflammatory stress and idiosyncratic hepatotoxicity: hints from animal models. Pharmacol Rev, 2009;61:262–82. 68. Usui, T. and Naisbitt, D.J. Human leukocyte antigen and idiosyncratic adverse drug reactions. Drug Metab Pharmacokinet, 2017;32:21–30. 69. Daly, A.K., Donaldson, P.T., Bhatnagar, P. et al. HLA‐B*5701 genotype is a major determinant of drug‐induced liver injury due to flucloxacillin. Nat Genet, 2009;41:816–9. 70. Hirata, K., Takagi, H., Yamamoto, M. et al. Ticlopidine‐induced hepatotoxicity is associated with specific human leukocyte antigen genomic subtypes in Japanese patients: a preliminary case‐control study. Pharmacogenomics J, 2008;8:29–33. 71. Nicoletti, P., Aithal, G.P., Bjornsson, E.S. et al. Association of liver injury from specific drugs, or groups of drugs, with polymorphisms in HLA and other genes in a genome‐wide association study. Gastroenterology, 2017;152:1078–89. 72. Hautekeete, M.L., Horsmans, Y., Van Waeyenberge, C. et al. HLA association of amoxicillin‐clavulanate‐‐induced hepatitis. Gastroenterology, 1999; 117:1181–6. 73. Stephens, C., Lopez‐Nevot, M.A., Ruiz‐Cabello, F. et al. HLA alleles influence the clinical signature of amoxicillin‐clavulanate hepatotoxicity. PLoS One, 2013;8:e68111. 74. O’Donohue J., Oien, K.A., Donaldson, P. et  al. Co‐amoxiclav jaundice: clinical and histological features and HLA class II association. Gut, 2000;47:717–20. 75. Lucena, M.I., Andrade, R.J., Fernandez, M.C. et  al. Determinants of the clinical expression of amoxicillin‐clavulanate hepatotoxicity: a prospective series from Spain. Hepatology, 2006;44:850–6.

  76. Lucena, M.I., Molokhia, M., Shen, Y. et al. Susceptibility to amoxicillin‐­ clavulanate‐induced liver injury is influenced by multiple HLA class I and II alleles. Gastroenterology, 2011;141:338–47.   77. Donaldson, P.T., Daly, A.K., Henderson, J. et al. Human leucocyte antigen class II genotype in susceptibility and resistance to co‐amoxiclav‐induced liver injury. J Hepatol, 2010;53:1049–53.   78. Devarbhavi, H., Raj, S., Aradya, V.H. et al. Drug‐induced liver injury associated with Stevens‐Johnson syndrome/toxic epidermal necrolysis: patient characteristics, causes, and outcome in 36 cases. Hepatology, 2016;63:993–9.   79. Chung, W.H., Hung, S.I., Hong, H.S. et al. Medical genetics: a marker for Stevens–Johnson syndrome. Nature, 2004;428:486.   80. Ikeda, H., Takahashi, Y., Yamazaki, E. et  al. HLA class I markers in Japanese patients with carbamazepine‐induced cutaneous adverse reactions. Epilepsia, 2010;51:297–300.   81. Chen, M., Borlak, J., and Tong, W. High lipophilicity and high daily dose of oral medications are associated with significant risk for drug‐induced liver injury. Hepatology, 2013;58:388–96.  82. Kaplowitz, N. Avoiding idiosyncratic DILI: two is better than one. Hepatology, 2013;58:15–7.   83. Lucena, M.I., Andrade, R.J., Kaplowitz, N. et al. Phenotypic characterization of idiosyncratic drug‐induced liver injury: the influence of age and sex. Hepatology, 2009;49:2001–9.  84. Andrade, R.J., Lucena, M.I., Fernandez, M.C. et  al. Drug‐induced liver injury: an analysis of 461 incidences submitted to the Spanish registry over a 10‐year period. Gastroenterology, 2005;129:512–21.   85. Urban, T.J., Nicoletti, P., Chalasani, N. et al. Minocycline hepatotoxicity: clinical characterization and identification of HLA‐B * 35:02 as a risk factor. J Hepatol, 2017;67(1):137–44.   86. Larrey, D. Epidemiology and individual susceptibility to adverse drug reactions affecting the liver. Semin Liver Dis, 2002;22:145–55.   87. Fountain, F.F., Tolley, E., Chrisman, C.R., and Self, T.H. Isoniazid hepatotoxicity associated with treatment of latent tuberculosis infection: a 7‐year evaluation from a public health tuberculosis clinic. Chest, 2005;128:116–23.   88. Aithal, G.P., Day, C.P., Leathart, J.B., and Daly, A.K. Relationship of polymorphism in CYP2C9 to genetic susceptibility to diclofenac‐induced hepatitis. Pharmacogenetics, 2000;10:511–8.   89. Huang, Y.S., Chern, H.D., Su, W.J. et al. Cytochrome P450 2E1 genotype and the susceptibility to antituberculosis drug‐induced hepatitis. Hepatology, 2003;37:924–30.   90. Musch, E., Eichelbaum, M., Wang, J.K., von Sassen, W., Castro‐Parra, M., and Dengler, H.J. Incidence of hepatotoxic side effects during antituberculous therapy (INH, RMP, EMB) in relation to the acetylator phenotype. (author’s transl). Klinische Wochenschrift, 1982;60:513–9.   91. Pande, J.N., Singh, S.P., Khilnani, G.C., Khilnani, S., and Tandon, R.K. Risk factors for hepatotoxicity from antituberculosis drugs: a case‐control study. Thorax, 1996;51:132–6.  92. Rieder, M.J., Shear, N.H., Kanee, A., Tang, B.K., and Spielberg, S.P. Prominence of slow acetylator phenotype among patients with sulfonamide hypersensitivity reactions. Clin Pharmacol Ther, 1991;49:13–7.  93. Cai, Y., Yi, J., Zhou, C., and Shen, X. Pharmacogenetic study of drug‐ metabolising enzyme polymorphisms on the risk of anti‐tuberculosis drug‐ induced liver injury: a meta‐analysis. PLoS One, 2012;7:e47769.   94. Nicoletti, P.D., Goel, H., Eapen, A. et al. Genome‐wide association study (GWAS) to identify genetic risk factors that increase susceptibility to anti‐ tuberculosis drug‐induced liver injury (ATDILI). AASLD Liver Meeting; Washington, DC, 2017.   95. Lucena, M.I., Andrade, R.J., Martinez, C. et al. Glutathione S‐transferase m1 and t1 null genotypes increase susceptibility to idiosyncratic drug‐ induced liver injury. Hepatology, 2008;48:588–96.   96. Huang, Y.S., Su, W.J., Huang, Y.H. et al. Genetic polymorphisms of manganese superoxide dismutase, NAD(P)H:quinone oxidoreductase, glutathione S‐transferase M1 and T1, and the susceptibility to drug‐induced liver injury. J Hepatol, 2007;47:128–34.   97. Daly, A.K., Aithal, G.P., Leathart, J.B., Swainsbury, R.A., Dang, T.S., and Day, C.P. Genetic susceptibility to diclofenac‐induced hepatotoxicity: contribution of UGT2B7:CYP2C8:and ABCC2 genotypes. Gastroenterology, 2007;132:272–81.   98. Boyer, J.L. Bile formation and secretion. Comp Physiol, 2013;3:1035–78.   99. Brinkmann, U. and Eichelbaum, M. Polymorphisms in the ABC drug transporter gene MDR1. Pharmacogenomics J, 2001;1:59–64. 100. Hagenbuch, B. and Stieger, B. The SLCO (former SLC21) superfamily of transporters. Mol Aspects Med, 2013;34:396–412.



54:  Drug‐Induced Liver Injury

101. Lu, H., Choudhuri, S., Ogura, K. et al. Characterization of organic anion transporting polypeptide 1b2‐null mice: essential role in hepatic uptake/ toxicity of phalloidin and microcystin‐L.R. Toxicol Sci, 2008;103:35–45. 102. Pauli‐Magnus, C. and Meier, P.J. Hepatobiliary transporters and drug‐ induced cholestasis. Hepatology, 2006;44:778–87. 103. Fattinger, K., Funk, C., Pantze, M. et al. The endothelin antagonist bosentan inhibits the canalicular bile salt export pump: a potential mechanism for hepatic adverse reactions. Clin Pharmacol Ther, 2001;69:223–31. 104. Kostrubsky, V.E., Vore, M., Kindt, E. et al. The effect of troglitazone biliary excretion on metabolite distribution and cholestasis in transporter‐deficient rats. Drug Metab Dispos, 2001;29:1561–6. 105. Bohme, M., Muller, M., Leier, I., Jedlitschky, G., and Keppler, D. Cholestasis caused by inhibition of the adenosine triphosphate‐dependent bile salt transport in rat liver. Gastroenterology, 1994;107:255–65. 106. Funk, C., Pantze, M., Jehle, L. et al. Troglitazone‐induced intrahepatic cholestasis by an interference with the hepatobiliary export of bile acids in male and female rats. Correlation with the gender difference in troglitazone sulfate formation and the inhibition of the canalicular bile salt export pump (Bsep) by troglitazone and troglitazone sulfate. Toxicology, 2001;167:83–98. 107. Hirano, H., Kurata, A., Onishi, Y. et al. High‐speed screening and QSAR analysis of human ATP‐binding cassette transporter ABCB11 (bile salt export pump) to predict drug‐induced intrahepatic cholestasis. Mol Pharm, 2006;3:252–65. 108. Bramow, S., Ott, P., Thomsen Nielsen, F., Bangert, K., Tygstrup, N., and Dalhoff, K. Cholestasis and regulation of genes related to drug metabolism and biliary transport in rat liver following treatment with cyclosporine A and sirolimus (Rapamycin). Pharmacol Toxicol, 2001;89:133–9. 109. Dara, L., Liu, Z.X., and Kaplowitz, N. Mechanisms of adaptation and progression in idiosyncratic drug induced liver injury, clinical implications. Liver Int, 2016;36(2):158–65.

713

110. Cantor, H.M. and Dumont, A.E. Hepatic suppression of sensitization to antigen absorbed into the portal system. Nature, 1967;215:744–5. 111. Callery, M.P., Kamei, T., and Flye, M.W. The effect of portacaval shunt on delayed‐hypersensitivity responses following antigen feeding. J Surg Res, 1989;46:391–4. 112. Yang, R., Liu, Q., Grosfeld, J.L., and Pescovitz, M.D. Intestinal venous drainage through the liver is a prerequisite for oral tolerance induction. J Pediatric Surg, 1994;29:1145–8. 113. Knolle, P.A. and Gerken, G. Local control of the immune response in the liver. Immunol Rev, 2000;174:21–34. 114. Black, M., Mitchell, J.R., Zimmerman, H.J., Ishak, K.G., and Epler, G.R. Isoniazid‐associated hepatitis in 114 patients. Gastroenterology, 1975;69: 289–302. 115. Metushi, I.G., Hayes, M.A., and Uetrecht, J. Treatment of PD‐1(−/−) mice with amodiaquine and anti‐CTLA4 leads to liver injury similar to idiosyncratic liver injury in patients. Hepatology, 2015;61:1332–42. 116. Chakraborty, M., Fullerton, A.M., Semple, K. et al. Drug‐induced allergic hepatitis develops in mice when myeloid‐derived suppressor cells are depleted prior to halothane treatment. Hepatology, 2015;62:546–57. 117. Mak, A. and Uetrecht, J. The role of CD8 T cells in amodiaquine‐induced liver injury in PD1−/− mice cotreated with anti‐CTLA‐4. Chem Res Toxicol, 2015;28:1567–73. 118. Song, B., Aoki, S., Liu, C., Susukida, T., and Ito, K. An animal model of abacavir‐induced HLA‐mediated liver injury. Toxicol Sci, 2018;162:713–23. 119. Cardone, M., Garcia, K., Tilahun, M.E. et al. A transgenic mouse model for HLA‐B*57:01‐linked abacavir drug tolerance and reactivity. J Clin Invest, 2018;128(7):2819–32.

55

Oxidative Stress and Inflammation in the Liver John J. Lemasters1 and Hartmut Jaeschke2 Departments of Drug Discovery and Biomedical Sciences and Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC, USA 2 Department of Pharmacology, Toxicology and Therapeutics, University of Kansas Medical Center, Kansas City, KS, USA 1

INTRODUCTION The liver is frequently the target of injury by oxidative stress and inflammation, as typified by ischemia/reperfusion, drug‐ induced hepatotoxicity, and both alcoholic and non‐alcoholic steatohepatitis (ASH and NASH) [1–6]. Oxidative stress and inflammation are inherently interrelated, since under various conditions each promotes the other. Oxidative stress is best defined as “a disturbance in the prooxidant–antioxidant balance in favor of the former” [7]. Thus, oxidative stress occurs when formation of prooxidant species increases or when antioxidant defenses become compromised [7, 8].

REACTIVE OXYGEN AND NITROGEN SPECIES Reactive oxygen species The 1‐ and 2‐electron partially reduced forms of O2 are superoxide (O2•−) and hydrogen peroxide (H2O2), respectively. Major sources of these reactive oxygen species (ROS) are NADPH oxidase of Kupffer cells and inflammatory cells, the respiratory chain of mitochondria, xanthine oxidase utilizing xanthine and hypoxanthine generated after ATP degradation, lipoxygenases, cytochrome P450, and flavin monooxygenases, among others. In mitochondria alone, Brand and coworkers identify 11 distinct sites that form O2•− and H2O2 [9]. In the presence of transition metal ions, such as iron and copper, O2•− and H2O2 react to form

the highly reactive 3‐electron reduced form of O2, the hydroxyl radical (•OH), by the iron‐catalyzed Haber–Weiss or Fenton reaction (Figure 55.1) [8]. In addition, transition metals catalyze a lipid peroxidation chain reaction sustained by lipid peroxyl, alkyl, and alkoxy radicals [10]. Beta scission of alkoxy radicals leads in turn to generation of toxic aldehydes like malondialdehyde and 4‐hydroxynonenal. Singlet oxygen (1O2) is an oxidizing and highly reactive ROS that is dioxygen in an excited state. Energy transfer from other excited state molecules generates 1O2. Typically, the source of excitation is a light‐activated photosensitizer, including photosensitizing agents used in photodynamic therapy and porphyrins overproduced in porphyria [11].

Reactive nitrogen species Reactive nitrogen species (RNS) include nitric oxide (•NO) and peroxynitrite (OONO−) [12]. •NO is a signaling molecule ­produced by nitric oxide synthase (NOS), which forms nitrosyl complexes with iron‐sulfur clusters in enzymes like aconitase. •NO also reacts at diffusion‐limited rates with O2•− to form OONO−, a powerful oxidant and nitrating agent (Figure 55.1). OONO− decomposes to a •OH‐like species and also nitrates tyrosyl residues in proteins in a reaction that is catalyzed, at least in part, by iron [12]. In inflammatory cells, myeloperoxidase (MPO) forms hypochlorous acid (HOCl) from H2O2 [13]. HOCl is highly reactive with thiols, amino groups, and methionine of proteins. All of these highly reactive species are important in toxicity to liver and other tissues [8].

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



55:  Oxidative Stress and Inflammation in the Liver

715

Figure 55.1  Scheme of generation of reactive oxygen and nitrogen species and consequent lipid peroxidation. Enhanced formation of prooxidant species by disordered mitochondrial respiration, NADPH oxidase, and other processes and/or impairment of antioxidant defense systems by depletion of glutathione (GSH), NADPH, or antioxidant enzymes leads to oxidative stress and increased net formation of superoxide (O2•−) and hydrogen peroxide (H2O2). Superoxide dismutase converts 2 O2•− to H2O2 and O2 with consumption of 2 H+. Peroxidases and catalase reduce H2O2 to H2O. Alternatively, myeloperoxidase catalyzes the reaction of H2O2 with chloride anion to form hypochlorous acid (HOCl), a potent cytocidal oxidant. In the presence of chelatable iron, O2•− reduces ferric iron (Fe3+) to ferrous iron (Fe2+), which reacts with H2O2 to form the toxic and reactive hydroxyl radical (•OH). Ferrireductases may also reduce Fe3+ to Fe2+ coupled to oxidation of NAD(P)H (or ascorbate). •OH reacts with unsaturated lipids to form alkyl radicals (L•) that initiate an oxygen‐dependent chain reaction generating lipid peroxides (LOOH) and peroxyl radicals (LOO•). Iron also catalyzes a chain reaction producing alkoxyl radicals (LO•) and more LOO• with beta‐scission of LO• leading to formation of reactive aldehydes like malondialdehyde (MDA) and 4‐hydroxynonenal (4HNE). Peroxidases reduce LOOH to the corresponding alcohol (LOH). Nitric oxide synthase catalyzes nitric oxide (•NO) and citrulline formation from arginine. •NO reacts very rapidly with O2•− to form peroxynitrite anion (ONOO−), which decomposes to nitrogen dioxide (•NO2) and •OH. These reactive oxygen, nitrogen, and lipid species also attack proteins and nucleic acids.

ANTIOXIDANT DEFENSES

Glutathione

Superoxide dismutase

The tripeptide glutathione (γ‐L‐glutamyl‐L‐cysteinylglycine, GSH) with a hepatic intracellular concentration of around 10 mM plays a major protective role during oxidative stress by maintaining the sulfhydryl status of proteins and by disposing of electrophiles and oxidants [16]. GSH is synthesized in the ­cytosol in two steps. The first is the rate‐limiting formation of  γ‐glutamylcysteine from glutamate and cysteine, which is followed by synthesis of GSH itself by peptide linkage of γ‐glutamylcysteine to glycine catalyzed by glutathione synthetase. NADPH (E’0 = −320 mV) via glutathione reductase (GR) keeps GSH (E’0 = −230 mV) in a highly reduced state. Consequently, normal hepatic glutathione disulfide (GSSG) concentration is low, and GSH/GSSG ratios are on the order of 200 to 1 [17]. GSH can also be salvaged by the membrane‐­ associated protein gamma glutamyl transferase [16]. GSH has multiple protective roles during oxidative and related stresses. Using glutathione S‐transferase (GST), GSH forms detoxifying adducts with the electrophilic centers of a large variety of compounds, including peroxidized lipids and N‐acetyl‐p‐benzoquinone imine (NAPQI), the toxic byproduct of acetaminophen metabolism. GSTs consist of three superfamilies – cytosolic, mitochondrial, and microsomal – and can comprise up to 10% of cytosolic protein [18]. The peroxidases,

Cells contain a wide variety of antioxidant defenses that directly remove toxic radical species or repair the damage done by oxidative stress. Superoxide dismutase (SOD) accelerates dismutation of 2 O2•− to O2 and H2O2 (Figure 55.1). In humans and other mammals, SOD isoforms are SOD1 in the cytoplasm, SOD2 in the mitochondrial matrix, and SOD3 in the extracellular space. SOD1 and SOD3 contain Cu and Zn in their reactive centers, whereas SOD2 has Mn. Oxidized cytochrome c of the mitochondrial respiratory chain also directly scavenges O2•− to form O2 and reduced cytochrome c, the latter then being re‐oxidized by cytochrome oxidase [14].

Catalase and peroxidase Catalase in peroxisomes converts 2 H2O2 to 2 H2O and O2 with one of the highest turnover numbers of all enzymes (> 106 sec−1). Aquaporins 3 and 8 facilitate transmembrane movement of H2O2 through mitochondrial and plasma membranes but have not been reported to be present in peroxisomes [15]. Perhaps for this reason, peroxidases are more important to eliminate H2O2 outside the peroxisome.

716

THE LIVER:  MITOCHONDRIAL PERMEABILITY TRANSITION

glutathione peroxidase (GPX) and peroxiredoxin (PRX), reduce both hydrogen peroxide and lipid hydroperoxides to water or their corresponding alcohols with different members of each family having different substrate preferences [19]. GPX4 has emerged as particularly important for eliminating lipid hydroperoxides [20]. The GPX‐catalyzed reaction yields GSSG, which is restored to GSH by GR. The PRX reaction creates an oxidized sulfenic acid (S‐OH) moiety at the active site, which is typically reverted to a thiol by thioredoxin (TRX). TRX also reduces oxidized cysteines and disulfide bonds in other oxidatively altered proteins. Oxidized TRX is restored to the reduced state by thioredoxin reductase (TrxR) coupled to oxidation of NADPH. Glutaredoxin (GRX) functions somewhat similarly to TRX in restoring the reduced thiol status of oxidatively modified proteins but in a GSH‐dependent fashion. Sulfiredoxin (SRX) reactivates the more oxidized sulfinic acid (SO2H)‐­modified form of TRX (and other proteins) driven by ATP hydrolysis and oxidation of GSH to GSSG. SRX also participates in the deglutathionylation of some proteins, contributing to the S‐glutathionylation cycle [21]. Importantly, through expression and differential localization of various enzyme isoforms, these thiol‐modifying ­systems are represented both in the mitochondrial matrix and the cytosol [22].

CHELATABLE IRON AND OXIDATIVE STRESS Chelatable versus nonchelatable iron In cells and tissues, iron exists in two forms. “Nonchelatable” iron is sequestered into ferritin, hemosiderin, and iron‐­ containing prosthetic groups of proteins (e.g. heme, iron–sulfur complexes), which are inaccessible to conventional iron chelators like desferal. “Chelatable” iron includes unbound free iron  plus iron loosely attached to membrane surfaces and ­polyanionic metabolites like citrate and ATP. Due to this loose binding, free iron comprises only a small fraction of all chelatable iron. Overall, chelatable iron is estimated to be about 5 μM in hepatocytes [6].

Iron and oxidative stress Although an essential nutrient, excess iron causes acute hepatocellular necrosis after accidental overdose and leads to chronic hepatic injury in hereditary hemochromatosis [23]. High serum iron is also associated with worse outcomes in diabetes, cardiovascular disease, cancer, ASH, and NASH [24–26]. Increases of intracellular chelatable iron promote hepatocellular killing, whereas iron chelators like desferal are cytoprotective in various models of oxidative stress and hypoxia/ischemia. Protection by iron chelation infers a critical role for iron in the pathogenesis of injury, most likely by catalyzing •OH formation and subsequent lipid peroxidation [10–14] (Figure  55.1). More recently, the term ferroptosis has been introduced to describe non‐apoptotic (necrotic) oxidative cell death in association with iron‐­dependent lipid peroxidation [20].

Cellular iron uptake Almost all non‐heme iron in the plasma is bound to transferrin (Tf) as Fe3+. With two iron binding sites per Tf molecule, a plasma Tf concentration of 5–10 μM and 30% iron occupancy, plasma iron is in the range of 3 to 6 μM. The major route of iron uptake into cells is clathrin‐dependent endocytosis, which leads to entry of Tf bound to the Tf receptor (TfR) into the endosomal/lysosomal compartment [27]. As pH drops inside endosomes, Tf releases its bound iron. Subsequent protein sorting leads to recycling of Tf and TfR to the extracellular space and plasma membrane, respectively, leaving Fe3+ relatively concentrated inside the endosome/ lysosomal compartment. Release of iron into the cytosol for cellular needs, such as synthesis of iron‐containing proteins, requires at least two steps. The first is reduction of Fe3+ to Fe2+ by an endosomal ferrireductase identified as the product of the six transmembrane epithelial antigen of the prostate 3 (Steap3) gene [28]. The second step is transport of Fe2+ across the lysosomal/endosomal membrane by divalent metal transporter‐1 (DMT1), an H+/Fe2+ exchanger [29]. However, a channel for Fe2+ and other divalent cations like Ca2+, Mn2+ and Zn2+ named type IV mucolipidosis‐ associated protein (TRPML1) can also mediate Fe2+ release from late endosomes and lysosomes [30]. Autophagy and proteosomal protein degradation also recycles iron for biosynthetic needs. Similarly, heme degradation by heme oxygenase recycles free iron, and heme oxygenase‐2 is at least in part localized to endosomes [31]. In iron overload states, ferritin and hemosiderin accumulate to store iron in an unreactive non‐chelatable form [29].

MITOCHONDRIAL PERMEABILITY TRANSITION Mitochondrial inner membrane permeability During oxidative phosphorylation, respiration drives translocation of protons across the inner membrane to create an electrochemical gradient of protons comprised of a negative inside membrane potential (ΔΨ) and an alkaline inside pH gradient (ΔpH) [32]. The protonmotive force (Δp = ΔΨ – 59 ΔpH) thus formed drives ATP synthesis coupled to reentry of protons through the mitochondrial F1F0‐ATP synthase complex. This chemiosmotic coupling necessitates that the inner membrane be highly impermeable to protons, as well as to other ions and charged metabolites that might collapse Δp. Thus, nearly all metabolite exchange across the inner membrane occurs through specific transporters, such as the adenine nucleotide translocator (ANT), which exchanges ATP for ADP, the phosphate transporter, and numerous other transporter systems for various ­respiratory substrates and additional metabolites. In contrast, the outer membrane is nonspecifically permeable to these ions and metabolites, which move through voltage dependent anion channels (VDAC). VDAC, a somewhat misleading term since VDAC is only weakly selective for anions over cations, remains open within the presumed physiological range of membrane potentials for the outer membrane (±50 mV) and conducts uncharged solutes up 5 kDa in size [32]. Nonetheless, opening and closing of VDAC is emerging as a dynamic regulator of global mitochondrial function [32].



55:  Oxidative Stress and Inflammation in the Liver

Permeability transition pores In the mitochondrial permeability transition (MPT), permeability transition (PT) pores open that make the inner membrane nonselectively permeable to solutes of molecular mass up to around 1500 Da [33]. Ca2+, ROS, OONO− and numerous reactive chemicals promote pore opening, whereas cyclosporin A (CsA), high Mg2+, and pH less than 7 inhibit pore conductance. When PT pores open, mitochondria depolarize and undergo large amplitude swelling driven by colloid osmotic forces, which are the hallmarks of the MPT. Additionally, mitochondria release a variety of solutes from the mitochondrial matrix, including accumulated calcium and pyridine nucleotides. Most importantly, oxidative phosphorylation immediately ceases, and mitochondria begin to futilely hydrolyze ATP. Swelling also leads to rupture of the outer membrane with consequent release of cytochrome c and other proapoptotic factors from the intermembrane space. Patch clamping shows that PT pores have a very large single channel conductance, such that opening of even a single PT pore may be sufficient to cause mitochondrial depolarization and uncoupling [33]. Onset of the MPT is a key event inducing necrotic and apoptotic cell killing after ischemia/reperfusion, oxidative stress, and many instances of hepatotoxicity [6].

Composition of permeability transition pores Despite extensive study, the composition of PT pores remains controversial. In one model, PT pores are comprised of VDAC from the outer membrane, ANT from the inner membrane, cyclophilin D (CypD) from the matrix, and possibly other proteins (Figure  55.2) [34]. However, genetic knockout studies challenge the validity of this model by showing that the MPT still occurs in mitochondria that are deficient in VDAC and ANT [35, 36]. Moreover, in mitochondria from CypD deficient mice, a CsA‐insensitive MPT still occurs but requires greater Ca+ for induction [37]. In newer models, other proteins have been proposed as components/regulators of PT pores, including the phosphate transporter (inner membrane) and subunits of the F1FO‐ATP synthase (inner membrane), but no consensus has emerged (Figure 55.2) [38–40]. An alternative model is that PT pores arise from ­several different proteins after damage and misfolding from oxidative and other stresses. Such misfolded proteins expose hydrophilic surfaces within the membrane bilayer, which leads to protein aggregation at these surfaces to form aqueous channels regulated by CypD and other molecular chaperones (Figure 55.2) [41]. In this way, multiple different proteins can form the conducting core of PT pores.

ISCHEMIA/REPERFUSION The mitochondrial permeability transition in ischemia/reperfusion Acidotic pH during ischemia inhibits PT pores, but as intracellular pH (pHi) increases during reperfusion, PT pores open to induce the MPT with consequent onset of ATP depletion‐ dependent necrotic cell death During ischemia, mitochondria

717

depolarize due to respiratory inhibition from lack of oxygen. After reperfusion, mitochondria initially repolarize as respiration resumes. However, the MPT occurs subsequently, leading to bioenergetic failure (depolarization) and ultimately cell death [42]. Confocal microscopy visualizes PT pore opening from movement of normally impermeant fluorophores like calcein into the mitochondrial matrix space from the cytosol (Figure 55.3). Addition of CsA, an MPT inhibitor, just before reperfusion prevents this membrane permeabilization, bioenergetic failure, and cell death, as does reoxygenation at acidotic pH (Figure 55.3). Reperfusion with the cytoprotective amino acid, glycine, also protects against cell killing, but glycine does not prevent mitochondrial inner membrane permeabilization. Rather, glycine protects downstream of the MPT apparently by preventing breakdown of the plasma membrane permeability barrier after ATP depletion (Figure 55.3) [42, 43]. Overall, MPT onset is the penultimate event in cell death after ischemia/reperfusion. Although promising, CsA has shortcomings as a therapeutic agent to block the MPT, including immunosuppression, nephrotoxicity, and variable bioavailability. Moreover, protection is lost at higher concentrations (≥ 5 μM) of  CsA [44]. Nonimmunosuppressive analogues, such as N‐methyl‐4‐isoleucine cyclosporin (NIM811) and alisporivir, do not lose efficacy at high doses and show better promise as ­therapeutic agents [45, 46].

Minocycline and doxycycline Other drugs blocking the MPT include minocycline and doxycycline. Both are semisynthetic tetracycline derivatives reported to protect against neurodegenerative disease, trauma, and hypoxia–ischemia [46, 47]. These agents block the MPT by inhibiting mitochondrial calcium uptake by the electrogenic mitochondrial calcium uniporter (MCU) [46, 48]. Minocycline and doxycycline may also protect by blocking mitochondrial uptake of prooxidant Fe2+, which is also transported into mitochondria by MCU.

IRON TRANSLOCATION FROM LYSOSOMES INTO MITOCHONDRIA DURING CELLULAR STRESSES Visualization of chelatable iron in hepatocytes Fe2+ stoichiometrically quenches the fluorescence of calcein [49]. As evaluated by Fe2+‐dependent calcein quenching, cytosolic chelatable Fe2+ increases by 100–200 μM after collapse of  the acidic lysosomal pH gradient with bafilomycin, an ­inhibitor of the proton‐pumping vacuolar ATPase (V‐ATPase) (Figure  55.4) [50]. Desferal and starch‐desferal (s‐desferal) prevent this increase (Figure  55.4). Hepatocytes take up ­ ­polysaccharides like dextran and s‐desferal by endocytosis to accumulate in lysosomes. Thus, inhibition of the increase of cytosolic Fe2+ after bafilomycin by s‐desferal signifies a ­lysosomal/endosomal origin of the Fe2+.

718

THE LIVER:  IRON TRANSLOCATION FROM LYSOSOMES INTO MITOCHONDRIA DURING CELLULAR STRESSES

Figure 55.2  Models of the mitochondrial permeability transition (PT) pore. In model (a), the composition of PT pores includes the voltage‐ dependent anion channel (VDAC) from the outer membrane (OM), the adenine nucleotide translocator (ANT) from the inner membrane (IM), and cyclophilin D (CypD) from the matrix space. Other proteins, such as the creatine kinase (CK), hexokinase (HK), and the peripheral benzodiazepine receptor (PBR), may also associate with PT pores. PT pore openers include Ca2+, oxidized pyridine nucleotides (NAD(P)+) and glutathione (GSSG), and reactive oxygen and nitrogen species (ROS, RNS), whereas cyclosporin A (CsA), NIM811 and pH < 7 block pore conductance. In a more recent model (b), PT pores form within F1FO‐ATP synthase dimers at the interface between monomers or in association with c‐rings. A third proposal (c) suggests that oxidative and other damage to integral mitochondrial inner membrane proteins leads to misfolding, which exposes hydrophilic surfaces to the hydrophobic bilayer. These misfolded proteins aggregate at these surfaces to form nascent PT pores. CypD and other chaperones associate to block their conductance, but high matrix Ca2+ acting through CypD leads to pore opening, an effect blocked by cyclosporin A (CsA) and non‐immunosuppressive analogs like NIM811. As misfolded protein clusters exceed the concentration of available chaperones, constitutively open channels form that are independent of Ca2+ and are not inhibited by CsA.

Mitochondrial iron By altering loading conditions, calcein can be selectively loaded into the mitochondria and lysosomes of hepatocytes. After bafilomycin, mitochondrial calcein quenches, whereas lysosomal calcein fluorescence increases, signifying an increase of mitochondrial and a decrease of lysosomal chelatable Fe2+ (Figure  55.4). Fe2+ uptake into mitochondria occurs via MCU, since Ru360, a specific MCU inhibitor, blocks mitochondrial but not cytosolic calcein quenching after bafilomycin (Figure  55.4). s‐Desferal also prevents mitochondrial Fe2+ uptake, implying that Fe2+ taken up by mitochondria originates from lysosomes. Mitochondrial iron loading then sensitizes hepatocytes to MPT onset and cell death after oxidative stress, which desferal, s‐desferal and Ru360 each suppress [50].

Translocation of iron from lysosomes to mitochondria during ischemia During ischemia, cytosolic and mitochondrial chelatable Fe2+ also increase, because of V‐ATPase inhibition caused by ATP depletion [51]. The ATP‐generating glycolytic substrate, fructose, prevents quenching, as do both desferal and s‐desferal. MCU inhibition by Ru360 suppresses the ischemia‐induced increase of mitochondrial but not cytosolic Fe2+. Desferal, s‐desferal and Ru360 before ischemia also decrease ROS formation, MPT onset, and cell killing after reperfusion. ­ ROS generation precedes MPT onset. Desferal and antioxidants block MPT onset after reperfusion, whereas CsA which prevents MPT onset does not block ROS formation [51, 52]. Thus, Fe2+‐dependent ROS formation is driving MPT onset after



55:  Oxidative Stress and Inflammation in the Liver

719

Figure 55.3  Onset of the mitochondrial permeability transition after ischemia/reperfusion to cultured hepatocytes. Calcein‐loaded cultured rat hepatocytes were subjected to 4 hours of ischemia (anoxia at pH 7.4) followed by reperfusion (reoxygenation) at pH 7.4, pH 6.2, pH 7.4 in the presence of cyclosporin A (CsA, 1 μM), or pH 7.4 in the presence of glycine (5 mM). After 4 hours of ischemia, note that mitochondria excluded the fluorescent calcein and appeared as dark voids. After reperfusion at pH 7.4, calcein gained entrance into mitochondria, which was then followed by cell death and release of all calcein. Reperfusion at pH 6.2 or at pH 7.4 in the presence of CsA prevented mitochondrial permeabilization and cell death. Glycine prevented cell death (release of cellular calcein) but not mitochondrial permeabilization.

reperfusion. Overall, iron translocation from lysosomes to ­mitochondria during ischemia is a “first hit” that when followed by a “second hit” of mitochondrial O2•− and H2O2 formation after reperfusion leads to Fe‐dependent OH• generation, MPT onset, and cell death. In this way, chelatable iron is a dynamic variable whose increase within mitochondria promotes ROS formation and hepatocellular injury.

preserve mitochondrial ΔΨ, and improve survival from 40% to 70% [53, 54]. Notably, minocycline protects even when used after resuscitation.

Aldehydes and oxidative stress

Iron‐dependent ROS formation and lipid peroxidation lead to generation of aldehydes like malondialdehyde and 4‐hydroxynonenal by beta‐scission of lipid hydroperoxides (Figure 55.1). Hemorrhagic shock/resuscitation If not metabolically detoxified by aldehyde dehydrogenase‐2 Hemorrhagic shock/resuscitation (HS/R) in mice is a clinically (ALDH2) in mitochondria, such aldehydes produce protein relevant model of hepatic oxidative stress and hypoxia/­ adducts and cell toxicity. Alda‐1, an activator of ALDH2, proreoxygenation injury. After 3 hours of hemorrhage followed by tects in models of ischemia/reperfusion and oxidative stress resuscitation, minocycline, doxycycline, and s‐desferal each [55]. Alda‐1 also protects against alcoholic steatohepatitis in decrease ALT release and hepatic necrosis by around 60%, mice in which acetaldehyde (AcAld) formed by ethanol

720

THE LIVER:  IRON AND OXIDATIVE STRESS IN ACETAMINOPHEN HEPATOTOXICITY

Figure 55.4  Quenching of cytosolic and mitochondrial calcein fluorescence 2 hours after bafilomycin. In cultured mouse hepatocytes, inhibition of the vacuolar proton‐pumping ATPase with bafilomycin (Baf, 50 nM) to collapse the lysosomal pH gradient led to quenching of cytosolic (a) and mitochondrial (b) calcein fluorescence, signifying an increase of chelatable Fe2+ in both compartments, which iron chelators, desferal (DFO, 1 mM) and s‐desferal (sDFO, 1 mM DFO equivalency) prevented. Ru360, an inhibitor of the mitochondrial calcium uniporter, prevent Baf‐induced ­mitochondrial calcein quenching (right lower panel). Bright spots after Baf are lysosomes. In (a), 300 μM free calcein was in the medium.

metabolism must also be detoxified by ALDH2 [56]. After aldehyde stress to hepatocytes, VDAC closes in the mitochondrial outer membrane, and hepatic mitochondria depolarize [57]. These adaptive mitochondrial changes promote selective and more rapid detoxifying oxidation of membrane‐permeant AcAld and other neutral aldehydes. However, VDAC closure also inhibits mitochondrial fatty acid oxidation to cause steatosis. Depolarization also stimulates mitochondrial autophagy (mitophagy). After chronic ethanol exposure, it is proposed that excess mitophagy leads to release of mitochondrial damage‐ associated molecular pattern (mtDAMP) molecules that promote inflammatory and profibrogenic responses, causing hepatitis and fibrosis in ASH [5]. By a similar mechanism, aldehydes formed during lipid peroxidation in NASH and chloracetaldehyde formed during vinyl chloride metabolism may lead to the quite similar hepatic pathologies of ASH, NASH, and toxicant‐associated steatohepatitis (TASH).

IRON AND OXIDATIVE STRESS IN ACETAMINOPHEN HEPATOTOXICITY Acetaminophen metabolism Acetaminophen (APAP) overdose causes fulminant hepatic necrosis and is the leading cause of acute liver failure in the western world [58]. NAPQI produced by cytochrome P450‐dependent oxidation of APAP underlies APAP hepatotoxicity. At therapeutic doses, conjugation of NAPQI with GSH prevents APAP toxicity, but overdose APAP causes GSH exhaustion followed by covalent binding of NAPQI to mitochondrial and other cellular proteins, leading to hepatocellular injury and death [59].

Role of lysosomes In cultured mouse hepatocytes, APAP treatment causes rupture of lysosomes, which then release Fe2+ into the cytosol that is taken up by mitochondria via MCU to promote ROS generation, MPT onset, mitochondrial depolarization, and cell death. Chelation of lysosomal iron with s‐desferal and blocking mitochondrial iron uptake with MCU inhibitors like Ru360 and minocycline protect against injury both in vitro and in vivo, and mice deficient in MCU are protected against APAP hepatotoxicity [50, 60, 61]. Iron chelation also prevents lysosomal rupture after APAP, indicating that an iron‐dependent mechanism is contributing to lysosomal damage. MPT blockers like CsA and NIM811 also protect, but MPT inhibition does not prevent lysosomal rupture. In vivo, deficiency of cyclophilin D, a regulator promoting MPT onset, also protects against APAP‐induced oxidative stress and liver injury [62]. Swelling after MPT induction causes outer membrane rupture and release into the cytosol of intermembrane proteins, such as apoptosis inducing factor (AIF) and endonuclease G, that cause nuclear DNA fragmentation and activation of regulated necrosis [59, 63].

Reactive nitrogen species During APAP hepatotoxicity, mitochondria generate OONO−, as indicated by mitochondrial nitrotyrosine adducts, which also signifies formation of both O2•− and •NO [63]. O2•− is the consequence of mitochondrial oxidative stress, as described above, but the source of •NO is less clear, since each isoform of NOS  –  inducible NOS (iNOS), endothelial NOS (eNOS) and neuronal NOS (nNOS) – can contribute to OONO− formation. Moreover, the mitochondrial respiratory chain can directly reduce nitrite to •NO under certain conditions. Although APAP



55:  Oxidative Stress and Inflammation in the Liver

induces hepatic iNOS gene expression, iNOS deficiency does not protect against APAP toxicity [64]. Rather, nNOS may be the source of •NO for OONO− formation after APAP, since pharmacologic inhibition and genetic deficiency of nNOS ­ delay  hepatoxicity after APAP overdose [65]. In animals ­heterozygous for mitochondrial SOD2, APAP causes greater hepatoxicity with increased OONO− generation and protein carbonyl formation, the latter a biomarker of oxidative stress [66]. Thus, mitochondrial SOD2 is important in removing O2•− and limiting OONO− generation.

c‐Jun N‐terminal kinase and oxidative stress Mitogen‐activated protein kinase (MAPK) signaling occurs in response to a wide variety of physical and chemical stresses. JNKs regulate gene expression by phosphorylation of c‐Jun, a component of the AP‐1 transcription factor, as well as numerous cytoplasmic targets. The liver expresses two isoforms of c‐Jun N‐terminal kinase, JNK1 and JNK2, that are the predominant hepatic effector MAPKs. After APAP and many other hepatic stresses, JNK becomes phosphorylated, signifying activation. Various genetic and pharmacological strategies to inhibit JNK protect against APAP toxicity and also against hepatic ischemia/ reperfusion and other stresses [64, 67–69]. During APAP hepatotoxicity, the initial mitochondrial protein adduct formation triggers a moderate oxidant stress, which causes activation of redox‐sensitive MAPKs, such as apoptosis signal‐regulating kinase 1 (ASK1) and mixed‐lineage kinase 3 (MLK3) [68, 69]. These upstream mitogen‐activated protein kinase kinase kinases (MAP3K) induce a MAPK signaling cascade ultimately leading to phosphorylation of JNK. P‐JNK translocates to mitochondria where it binds and phosphorylates Sab, which leads to inactivation of p‐Src on the inner membrane with consequent inhibition of the respiratory chain causing enhanced ROS generation. The amplified oxidant stress and peroxynitrite formation cause sustained JNK activation and eventually trigger the MPT [64, 68, 69].

INFLAMMATION AFTER LIVER INJURY NADPH oxidase During an acute inflammatory response, activation and hepatic recruitment of neutrophils and monocyte‐derived macrophages occur. Together with the resident macrophages (Kupffer cells), these cells can cause substantial oxidative stress [70]. In particular, NADPH oxidase (NOX2) generates O2•−. NOX2 is a transmembrane protein that uses intracellular NADPH to reduce O2 to toxic O2•− on the opposite side of the membrane. Accordingly, O2•− is released to the extracellular space or to the interior of phagocytic vacuoles and phagolysosomes.

Myeloperoxidase Neutrophils contain high levels of MPO in azurophilic granules, which are released upon activation at the same time as O2•− is formed [71]. MPO utilizes H2O2 formed from dismutation of O2•− to generate the very potent oxidant HOCl [13, 72]. Because

721

MPO is almost exclusively present in neutrophils, HOCl reaction products, such as chlorotyrosine and hypochlorous ­ acid‐modified proteins, can be used as footprint for neutrophil cytotoxicity [72].

Peroxynitrite By contrast, O2•− generated by macrophages is mainly dismutated to H2O2, which is only a mild oxidant. However, when iNOS is induced during inflammation, •NO formation ensues, leading to a near diffusion‐limited reaction with O2•− to form OONO− [73]. This RNS is a very potent oxidant that can be highly cytotoxic. Nitrotyrosine protein adducts are biomarkers for OONO− formation.

Leukocyte activation It is important to recognize that activation and recruitment of inflammatory cells into the liver does not automatically mean that the cells generate ROS and other oxidants to cause toxicity [72]. Most proinflammatory mediators like cytokines and chemokines increase cell surface expression of adhesion molecules (e.g. CD11b) and prime inflammatory cells for enhanced ROS formation but do not directly trigger an oxidative stress [74]. To cause cytotoxicity, a primed neutrophil located in the sinusoid needs to receive a signal to extravasate and to adhere to a specific target through the β2 integrin CD11b/CD18, which triggers a prolonged adherence‐dependent oxidant stress. For  resident macrophages (Kupffer cells), ROS formation is triggered by the engagement complement receptor during phagocytosis of opsonized debris or after binding of systemically generated activated complement factors such a C5a [74]. Adherence of neutrophils to target cells allows formation of ROS and release of proteolytic and cytotoxic proteins in close proximity to the targets [72]. This means that oxidants such as H2O2 and HOCl can diffuse into target cells to cause increased intracellular oxidative stress and formation of hypochlorite‐ modified proteins. This neutrophil‐derived oxidant stress is normally insufficient to directly cause cell death through lipid peroxidation. Instead, oxidant stress causes cellular and mitochondrial dysfunction, which eventually leads to cell necrosis [74]. In general, a neutrophil does not attack healthy cells. Targets are for the most part stressed and compromised cells, which are much more susceptible to injury than healthy cells [72].

NECROSIS AND INITIATION OF STERILE INFLAMMATION Damage‐associated molecular patterns Although it was recognized for some time that acute liver injury can be enhanced by an innate immune response, the concept of sterile inflammation came only recently into focus [75, 76]. Sterile inflammation means that no pathogens or their products such as endotoxin are involved in the pathogenesis. Instead, the initiating event is cellular necrosis that releases cell contents

722

THE LIVER:  NECROSIS AND INITIATION OF STERILE INFLAMMATION

into the circulation. Among all the molecules being released, a selected number termed damage‐associated molecular patterns (DAMPs) are recognized by pattern recognition receptors, principally toll‐like receptors (TLRs), located on immune cells like macrophages [77]. Examples of DAMPs include the nuclear protein high mobility group box 1 (HMGB1), which binds to TLR4 and the receptor for advanced glycation end products (RAGE), as well as nuclear DNA fragments and mitochondrial DNA, which activate TLR9. Ligand binding to TLRs induces transcriptional activation of proinflammatory cytokine genes, including tumor necrosis factor‐alpha (TNF‐α), interleukin‐1α (IL‐1α), and others. However, for interleukins like IL‐1β and IL‐18, an inactive pro‐form is first made, which requires subsequent proteolytic activation by active caspase‐1, a part of the NACHT, LRR, and PYD domains‐ containing protein 3 (NALP3) inflammasome complex [4, 75, 76]. This complex consists of NALP3 and apoptosis‐associated speck‐like protein containing a CARD (ASC), which binds

pro‐caspase‐1 through its caspase‐recruiting domain (CARD) and then autocatalytically activates it [4, 75, 76]. When necrotic cells release ATP, another DAMP, purinergic P2X receptor 7 (P2X7) binding leads to activation of the NALP3 inflammasome complex. Once caspase‐1 cleaves pro‐IL1β or pro‐IL‐18, macrophages release the mature cytokines, which bind to their respective receptors on inflammatory cells, e.g. the interleukin‐1 receptor (IL‐1R) for IL‐1β and IL‐1α. The consequence is activation and recruitment of inflammatory cells into the liver. It  is generally believed that most cytokines and chemokines exert their effects through neutrophils and monocyte‐derived macrophages (Figure 55.5).

Sterile inflammation in ischemia‐reperfusion The prototype of a sterile inflammatory response that aggravates liver injury is hepatic ischemia/reperfusion injury, which occurs clinically during liver surgery using the Pringle maneuver

Figure 55.5  Sterile inflammatory response after acute liver injury. The scheme shows the different mechanisms of a sterile inflammatory response after hepatic ischemia–reperfusion (a), acetaminophen hepatotoxicity (b) and obstructive cholestasis (c). After hepatic ischemic injury, necrotic hepatocytes release damage‐associated molecular patterns (DAMPs), which activate Kupffer cells and neutrophils. The oxidant stress generated by these inflammatory cells aggravates liver injury (a). After an acetaminophen overdose, reactive metabolite formation and an intracellular oxidant stress cause necrosis, which also triggers the release of DAMPs. However, activated leukocytes act to remove necrotic cellular debris and promote the recovery (b). During obstructive cholestasis, cholangiocytes release osteopontin, which is cleaved by matrix metalloproteinase 2 (MMP2). Cleaved osteopontin together with bile acid‐induced chemokines promote neutrophil chemotaxis and a subsequent neutrophil‐mediated injury (c). See text for details. Abbreviations: CCR 2, chemokine receptor type 2; HMGB1, high mobility group box 1; MCP‐1, monocyte chemoattractant protein 1; MIP‐2, macrophage inflammatory protein 2; mtDNA, mitochondrial DNA; TLR, toll‐like receptor; RAGE, receptor for advanced ­glycation end products.



55:  Oxidative Stress and Inflammation in the Liver

(warm  ischemia), liver transplantation (cold and warm ischemia), and hemorrhagic shock [78]. Ischemic stress and injury leads to an initial necrotic cell death, which causes release of DAMPs, including HMGB1 and DNA fragments. Extensive activation of the complement cascade also occurs. During early reperfusion, complement factors induce a Kupffer cell‐mediated oxidant stress, which further aggravates the ischemic cell injury [79]. DAMP‐induced cytokine and chemokine formation and ­activated complement factors also recruit neutrophils into the liver. These neutrophils can extensively aggravate the initial injury. Multiple lines of direct ­evidence support the critical role of o­ xidant stress in the mechanisms of neutrophil cytotoxicity [79]. During the neutrophil‐ mediated injury phase, hepatic neutrophils are fully activated and generate enhanced reactive o­ xygen. Additionally, neutrophils generate HOCl, as shown by chlorotyrosine staining of hepatocytes. Blocking neutrophil cytotoxicity through CD18 antibodies or pharmacological inhibition of NADPH oxidase eliminates oxidant stress and can markedly decrease injury. However, despite this strong experimental ­evidence for a critical contribution of reactive species to the pathophysiology of hepatic ischemia/reperfusion, it appears unlikely that oxidative stress generated by inflammatory cells kills cells due to lipid peroxidation causing plasma membrane failure [79]. Rather, diffusion of extracellularly generated oxidants disturbs hepatocellular homeostasis and causes mitochondrial oxidative stress and dysfunction, which ultimately leads to the MPT and cell necrosis (Figure 55.5a).

Other pathways stimulated by oxidative stress Although the protective effects of numerous antioxidant interventions support the involvement of oxidative stress in the ­overall pathophysiology, mechanistically these studies can be difficult to interpret, since multiple different sources of ROS and other oxidant species (e.g. intracellular sources, Kupffer cells, neutrophils, etc.), contribute at various times to the oxidative stress. Moreover, these self‐amplifying events are strongly dependent on each other, making it difficult to draw solid mechanistic conclusions. In addition, oxidative stress affects cytokine and chemokine formation through the nuclear transcription ­factor NF‐kB [2]. Oxidative stress can also induce adaptive mechanisms through the KEAP1‐Nrf2 pathway, which ­regulates important antioxidant defense systems, including glutathione synthesis and MnSOD levels in mitochondria [80]. The induction of protective adaptive mechanisms by a limited oxidant stress is also the basis of the extensively studied ischemic preconditioning intervention. Despite expanding insights into mechanisms of the sterile inflammatory response and oxidative stress during hepatic ischemia/reperfusion injury, no therapeutic interventions have been developed for clinical use. One reason is that hepatic ischemia/reperfusion injury is less of a problem today than ­previously due to improved surgical and organ procurement techniques. An exception to this trend is the increased vulnerability to ischemia/reperfusion injury of fatty livers, in which the inflammatory response and oxidative stress may be less important for injury than microcirculatory disturbances and an ­elevated susceptibility to mitochondrial dysfunction [81].

723

ACETAMINOPHEN‐INDUCED LIVER INJURY Neutrophils APAP overdose triggers mitochondrial dysfunction and ­extensive necrosis [82]. Both in animals and humans, necrosis precipitates release of DAMPs, including HMGB1, mtDNA, and nuclear DNA fragments, which activate macrophages and neutrophils to form cytokines and chemokines through binding to TLR9 (mtDNA; nDNA), TLR4 (HMGB1), and RAGE (HMGB1) [4, 75, 76]. Because of this proinflammatory response, rapid recruitment of neutrophils and monocytes into the liver ensues. Whether these infiltrating leukocytes contribute to an oxidative stress and aggravate liver injury is controversial [4, 83]. Most studies supporting a role of neutrophils and ­monocyte‐derived macrophages investigate various inflammatory mediators and extrapolate from the results that neutrophils or monocytes cause the injury, but plasma levels of proinflammatory mediators indicate that only limited quantities are produced during APAP hepatotoxicity in both mice and humans. Thus, cytokines such as TNF‐α, IL‐1β, or IL‐1α may be insufficient to activate neutrophils under these conditions at all [4, 84]. Indeed, orders of magnitude more IL‐1β than endogenously generated is required to enhance hepatic neutrophil accumulation [84]. Moreover, blocking β2 integrins (CD18) with antibodies and use of mice deficient of CD18, the critical adhesion molecule for an adherence‐dependent oxidant stress by neutrophils, do not decrease injury after APAP overdose [4, 83]. In addition, direct evidence is lacking for a neutrophil‐dependent oxidant stress during APAP‐induced liver injury, and infiltrating neutrophils from these livers are not activated when assessed for CD11b expression, ROS formation and phagocytosis capacity. Furthermore, pharmacological NOX2 inhibition and genetical deletion, which prevent oxidant stress from inflammatory cells, fail to prevent APAP‐induced oxidant stress and have no effect on injury. Together these data demonstrate that neutrophils do not participate in the injury phase after APAP overdose [4, 83]. The only direct evidence for a role of neutrophils comes from neutropenia experiments, which show some protection but only when the animals are pretreated with neutropenia‐inducing antibodies for 24 hours or longer [85]. No protection is observed when neutrophils are eliminated from the circulation several hours before the injury [86]. The reason for this differential effect of neutropenia is that antibody‐tagged inactive neutrophils line the hepatic sinusoids after prolonged neutropenia and trigger extensive phagocytosis by Kupffer cells. The massive removal of cell debris activates Kupffer cells, and mediators generated during this process cause a pre‐conditioning effect in hepatocytes with upregulation of metallothionein and other protective genes. Thus, the beneficial effect of prolonged neutropenia is caused by a preconditioning effect and not the absence of neutrophils [4, 83]. Taken together, these data indicate that neutrophils do not extend APAP hepatotoxicity during the injury phase. Measurement of the activation status of circulating neutrophils in APAP overdose patients confirms the animal data and show no activation during the injury phase. However, extensive activation occurs during regeneration suggesting a function of the cells during the recovery phase (Figure 55.5b) [87].

724

THE LIVER:  THERAPEUTIC IMPLICATIONS

Macrophages Monocytes are also recruited into the liver during APAP toxicity. Very early infiltration of these cells may cause a transient increase in injury, but most monocytes are recruited into necrotic areas at the end of the injury phase (> 12 hours in mice) [88, 89]. Monocyte chemoattractant protein 1 (MCP1) generated by dying hepatocytes and monocyte‐derived macrophages already in the necrotic area are responsible for recruitment [88]. Animals deficient of C‐C chemokine receptor type 2 (CCR2), the receptor for MCP1 on monocytes, have delayed regeneration of the necrotic areas, indicating that these macrophages are important for the recovery phase [88]. More recently, Kupffer cells are shown to also be involved in regeneration, and deficiency of both Kupffer cells and recruited monocytes in the liver prevents recovery after APAP toxicity [89]. In addition to the clearance of cellular debris, macrophages generate angiogenic factors, which are critical for the restoration of healthy tissue [89]. Studies of explanted livers from APAP‐induced acute liver failure patients show predominantly pro‐regenerative macrophages and only a limited number of neutrophils in centrilobular areas of necrosis. Like mice, humans generate large amounts of MCP‐1 to recruit these monocytes into the liver [90]. Together these data indicate that Kupffer cells, monocyte‐ derived macrophages, and neutrophils contribute to hepatic recovery after severe APAP‐induced liver injury in mice and humans by removal of cell debris and likely also be producing angiogenic mediators (Figure 55.5b) [4, 83]. Mice deficient of a functional NADPH oxidase in all phagocytes do not show delayed recovery after APAP overdose. Thus, hepatic regeneration after APAP overdose appears to be independent of an NADPH oxidase‐derived oxidative stress [87]. Overall, direct evidence that neutrophils, monocytes, or Kupffer cells aggravate APAP‐induced liver injury is very limited.

OBSTRUCTIVE CHOLESTASIS Rodent models Early mechanistic studies of liver injury during obstructive cholestasis focus on in vitro experiments using glycine‐­conjugated bile acids to induce apoptosis in rat hepatocytes, but concerns are raised that the species (mainly glycine‐conjugated bile acids) and high concentrations of bile acids used are not relevant to the pathophysiology in rodents [91, 92]. The observation that biliary infarcts (focal necrosis) after bile duct ligation (BDL) is almost completely eliminated in mice deficient in critical adhesion molecules for neutrophil cytotoxicity (CD18, ICAM‐1) leads to the hypothesis that cholestatic injury is an inflammatory process [93]. In support, substantial numbers of neutrophils accumulate in bile infarcts, which also stain positive for chlorotyrosine, a biomarker for HOCl [93]. Because mice deficient in TLR4, the receptor for endotoxin, are not protected against BDL‐induced liver injury, the mechanism of injury appears to involve a sterile inflammatory response (Figure 55.5c) [94]. In the absence of severe necrosis and release of DAMPs, the question arises concerning the initiating mechanism that triggers proinflammatory mediator formation. Follow‐up studies

identified that the major bile acids in mice are taurocholate, tauromuricholate, and muricholate. While none of these bile acids causes cell death of mouse hepatocytes at concentrations up to 5 mM, they each trigger formation of the neutrophil chemotactic factor macrophage inflammatory protein 2 (MIP‐2) and cell surface expression of ICAM‐1 dependent on the transcription factor early growth response protein 1 (Egr‐1) [94, 95]. Additional experiments indicate that cholangiocyte‐derived osteopontin, which is cleaved by matrix metalloproteinases to a potent neutrophil chemoattractant, is responsible for the initial neutrophil recruitment and injury [96]. This inflammatory injury process is extended by bile acid‐stimulated MIP‐2 formation [94, 95]. Thus, when the bile duct is obstructed and biliary pressure increases, cholangioles rupture to leak bile into the parenchyma. Biliary osteopontin causes an initial neutrophil activation and recruitment into the area of bile leakage, which is amplified by MIP‐2 generated by bile acid‐stimulated hepatocytes in the same area. The consequence is a prolonged inflammatory injury and development of areas of focal necrosis (bile infarcts) typical for BDL (Figure 55.5c) [97].

Human cholestasis An important question is whether the mechanism in mice also applies to human cholestasis. Human bile acids are generally more glycine‐conjugated and thus more hydrophobic and potentially cytotoxic. In cholestatic patients, serum bile acids, especially glycochenodeoxycholic acid, increase from 2.8 μM in control to 22 μM, but bile acids in the bile remain the same independent of cholestasis (2–4 mM range) [98]. Importantly, when primary human hepatocytes are exposed to these levels of bile acids individually or in combination, only the high biliary concentrations cause necrotic cell death [98]. In addition, human‐relevant bile acids cause moderate chemokine formation in human hepatocytes, which may be responsible for recruitment of neutrophils into the areas of necrosis [95]. Although necrotic biliary infarcts occur in cholestatic patients, apoptotic cell death is minimal as assessed by biomarkers such as caspase‐cleaved cytokeratin‐18 [98]. Whether inflammatory cells aggravate the injury in patients remains to be investigated. Together, these recent studies demonstrate that the liver injury caused by obstructive cholestasis involves bile acids and a sterile inflammatory response, but there are differences in the mechanism between the murine model and patients [99, 100]. Whereas the injury in mice is purely a neutrophil‐mediated injury facilitated by bile acid‐induced local chemokine formation, necrotic cell death in human cholestasis is directly caused by the more hydrophobic and cytotoxic bile acids in human bile. The relevance of the subsequent sterile inflammatory response for the injury is unclear. Overall, these findings emphasize the importance of translational studies.

THERAPEUTIC IMPLICATIONS Given the improved understanding of the mechanisms of sterile inflammation, it is tempting to assume that the individual mediators and cells involved are potential therapeutic targets to be



55:  Oxidative Stress and Inflammation in the Liver

exploited to protect against injury. However, several fundamental issues must be considered before of any target is selected. First, targets identified in animal models need to be validated in human disease, since animal models may not accurately reproduce human pathophysiology and since humans and animals can differ in immune responses. Second, a beneficial effect of a sterile inflammatory response is removal of necrotic tissue, and the leukocytes involved in this inflammatory process still have a vital host defense function. Thus, interference with leukocyte recruitment to prevent aggravation of the initial injury may negatively impact recovery and compromise host defense function. Such unintended detrimental effects may not be detected in short‐term animal experiments, which are mainly focused on the injury phase. In humans, recovery and long‐term survival are the most important outcome measures of any therapy. Thus, the efficacy and side‐effects of an anti‐inflammatory therapeutic approach after an acute injury must be very carefully evaluated. Nonetheless, given the self‐amplifying nature of the sterile inflammatory response, targeting its signaling mechanisms remains a promising therapeutic strategy with the goal of minimizing additional injury without impairing recovery and host defenses.

ACKNOWLEDGMENTS This work was supported, in part, by Grants AA012916, AA021191, AA025379, DK070195, DK073336 and DK102142 from the National Institutes of Health. Imaging facilities were  supported, in part, by P30 CA138313, GM103542 and 1S10OD018113.

REFERENCES 1. Du, K., Ramachandran, A., and Jaeschke, H. Oxidative stress during acetaminophen hepatotoxicity: sources, pathophysiological role and therapeutic potential. Redox Biol, 2016;10:148–56. 2. Konishi, T. and Lentsch, A.B. Hepatic ischemia/reperfusion: mechanisms of tissue injury, repair, and regeneration. Gene Expr, 2017;17(4):277–87. 3. Schuster, S., Cabrera, D., Arrese, M., and Feldstein, A.E. Triggering and resolution of inflammation in NASH. Nat Rev Gastroenterol Hepatol, 2018;15(6):349–64. 4. Woolbright, B.L. and Jaeschke, H. Role of the inflammasome in acetaminophen‐induced liver injury and acute liver failure. J Hepatol, ­ 2017;66(4):836–48. 5. Zhong, Z. and Lemasters, J.J. A unifying hypothesis linking hepatic adaptations for ethanol metabolism to the proinflammatory and profibrotic events of alcoholic liver disease. Alcohol Clin Exp Res, 2018;42(11):2072–89. 6. Lemasters, J.J. Hepatotoxicity due to mitochondrial injury, in Drug‐Induced Liver Disease, (eds. N. Kaplowitz, and L.D. DeLeve), Elsevier, Amsterdam, 2013, pp. 85–100. 7. Sies, H., Berndt, C., and Jones, D.P. Oxidative stress. Annu Rev Biochem, 2017;86:715–48. 8. Kehrer, J.P. and Klotz, L.O. Free radicals and related reactive species as mediators of tissue injury and disease: implications for health. Crit Rev Toxicol, 2015;45(9):765–98. 9. Wong, H.S., Dighe, P.A., Mezera, V., Monternier, P.A., Brand, M.D. Production of superoxide and hydrogen peroxide from specific mitochondrial sites under different bioenergetic conditions. J Biol Chem, 2017; 292(41):16804–9. 10. Minotti, G., and Aust, S.D. Redox cycling of iron and lipid peroxidation. Lipids, 1992;27(3):219–26.

725

11. Bissell, D.M., Anderson, K.E., and Bonkovsky, H.L. Porphyria. New Engl J Med, 2017;377(9):862–72. 12. Pacher, P., Beckman, J.S., and Liaudet, L. Nitric oxide and peroxynitrite in health and disease. Physiol Rev, 2007;87(1):315–424. 13. Klebanoff, S.J. Myeloperoxidase: friend and foe. J Leukoc Biol, 2005;77(5):598–625. 14. Pereverzev, M.O., Vygodina, T.V., Konstantinov, A.A., and Skulachev, V.P. Cytochrome c, an ideal antioxidant. Biochem Soc Trans, 2003 ;31(6):1312–5. 15. Miller, E.W., Dickinson, B.C., and Chang, C.J. Aquaporin‐3 mediates ­hydrogen peroxide uptake to regulate downstream intracellular signaling. Proc Natl Acad Sci USA, 2010;107(36):15681–6. 16. Yuan, L. and Kaplowitz, N. Glutathione in liver diseases and hepatotoxicity. Mol Aspects Med, 2009;30(1–2):29–41. 17. Giustarini, D., Colombo, G., Garavaglia, M.L. et  al. Assessment of glutathione/glutathione disulphide ratio and S‐glutathionylated proteins in human blood, solid tissues, and cultured cells. Free Radic Biol Med, 2017;112:360–75. 18. Board, P.G. and Menon, D. Glutathione transferases, regulators of cellular metabolism and physiology. Biochim Biophys Acta, 2013;1830(5):3267–88. 19. Rhee, S.G. and Kil, I.S. Multiple functions and regulation of mammalian peroxiredoxins. Annu Rev Biochem, 2017;86:749–75. 20. Yang, W.S. and Stockwell, B.R. Ferroptosis: death by lipid peroxidation. Trends Cell Biol, 2016;26(3):165–76. 21. Zhang, J., Ye, Z.W., Singh, S., Townsend, D.M., and Tew, K.D. An evolving understanding of the S‐glutathionylation cycle in pathways of redox regulation. Free Radic Biol Med, 2018;120:204–16. 22. Lu, J. and Holmgren, A. The thioredoxin antioxidant system. Free Radic Biol Med, 2014;66:75–87. 23. Powell, L.W., Seckington, R.C., and Deugnier, Y. Haemochromatosis. Lancet, 2016;388(10045):706–16. 24. Basuli, D., Stevens, R.G., Torti, F.M., and Torti, S.V. Epidemiological ­associations between iron and cardiovascular disease and diabetes. Front Pharmacol, 2014;5:117. 25. Manz, D.H., Blanchette, N.L., Paul, B.T., Torti, F.M., abd Torti, S.V. Iron and cancer: recent insights. Ann N Y Acad Sci, 2016;1368(1):149–61. 26. Kohgo, Y., Ohtake, T., Ikuta, K., Suzuki, Y., Torimoto, Y., and Kato, J. Dysregulation of systemic iron metabolism in alcoholic liver diseases. J Gastroenterol Hepatol, 2008;23 Suppl 1:S78–81. 27. Gammella, E., Buratti, P., Cairo, G., and Recalcati, S. The transferrin receptor: the cellular iron gate. Metallomics, 2017;9(10):1367–75. 28. Ohgami, R.S., Campagna, D.R., McDonald, A., and Fleming, M.D. The Steap proteins are metalloreductases. Blood. 2006;108(4):1388–94. 29. Anderson, G.J. and Frazer, D.M. Current understanding of iron homeostasis. Am J Clin Nutr, 2017;106(6):1559s–66s. 30. Dong, X.P., Cheng, X., Mills, E. et al. The type IV mucolipidosis‐associated protein TRPML1 is an endolysosomal iron release channel. Nature, 2008;455(7215):992–6. 31. West, A.R. and Oates, P.S. Subcellular location of heme oxygenase 1 and 2 and divalent metal transporter 1 in relation to endocytotic markers during heme iron absorption. J Gastroenterol Hepatol, 2008;23(1):150–8. 32. Lemasters, J.J. Evolution of voltage‐dependent anion channel function: From molecular sieve to governator to actuator of ferroptosis. Front Oncol, 2017;10:1–4. 33. Kim, J‐S., He, L., and Lemasters, J.J. Mitochondrial permeability transition: a common pathway to necrosis and apoptosis. Biochem Biophys Res Commun, 2003;304(3):463–70. 34. Halestrap, A.P. and Brenner, C. The adenine nucleotide translocase: a central component of the mitochondrial permeability transition pore and key player in cell death. Curr Med Chem, 2003;10(16):1507–25. 35. Kokoszka, J.E., Waymire, K.G., Levy, S.E. et al. The ADP/ATP translocator is not essential for the mitochondrial permeability transition pore. Nature, 2004;427(6973):461–5. 36. Krauskopf, A., Eriksson, O., Craigen, W.J., Forte, M.A., and Bernardi, P. Properties of the permeability transition in VDAC1(−/−) mitochondria. Biochim Biophys Acta, 2006;1757(5–6):590–5. 37. Basso, E., Fante, L., Fowlkes, J., Petronilli, V., Forte, M.A., and Bernardi, P. Properties of the permeability transition pore in mitochondria devoid of Cyclophilin D. J Biol Chem, 2005;280(19):18558–61. 38. Bernardi, P. Why F‐ATP synthase remains a strong candidate as the mitochondrial permeability transition pore. Front Physiol, 2018;9:1543.

726

THE LIVER:  REFERENCES

39. He, J., Carroll, J., Ding, S., Fearnley, I.M., and Walker, J.E. Permeability transition in human mitochondria persists in the absence of peripheral stalk subunits of ATP synthase. Proc Natl Acad Sci USA, 2017;114(34):9086–91. 40. Jonas, E.A., Porter, G.A., Jr., Beutner, G., Mnatsakanyan, N., and Alavian, K.N. Cell death disguised: the mitochondrial permeability transition pore as  the c‐subunit of the F(1)F(O) ATP synthase. Pharmacol Res, 2015;99:382–92. 41. He, L. and Lemasters, J.J. Regulated and unregulated mitochondrial permeability transition pores: a new paradigm of pore structure and function? FEBS Lett, 2002;512(1–3):1–7. 42. Qian, T., Nieminen, A.L., Herman, B., and Lemasters, J.J. Mitochondrial permeability transition in pH‐dependent reperfusion injury to rat hepatocytes. Am J Physiol, 1997;273(6 Pt 1):C1783–92. 43. Nishimura, Y. and Lemasters, J.J. Glycine blocks opening of a death channel in cultured hepatic sinusoidal endothelial cells during chemical hypoxia. Cell Death Differ, 2001;8(8):850–8. 44. Nazareth, W., Nasser, Y., and Crompton, M. Inhibition of anoxia‐induced injury in heart myocytes by cyclosporin A. J Mol Cell Cardiol, 1991;23:1351–4. 45. Waldmeier, P.C., Feldtrauer, J.J., Qian, T., and Lemasters, J.J. Inhibition of the mitochondrial permeability transition by the nonimmunosuppressive cyclosporin derivative NIM811. Mol Pharmacol, 2002;62(1):22–9. 46. Theruvath, T.P., Zhong, Z., Pediaditakis, P. et al. Minocycline and N‐methyl‐ 4‐isoleucine cyclosporin (NIM811) mitigate storage/reperfusion injury after rat liver transplantation through suppression of the mitochondrial permeability transition. Hepatology, 2008;47(1):236–46. 47. Garrido‐Mesa, N., Zarzuelo, A., and Galvez, J. Minocycline: far beyond an antibiotic. Br J Pharmacol, 2013;169(2):337–52. 48. Schwartz, J., Holmuhamedov, E., Zhang, X., Lovelace, G.L., Smith, C.D., and Lemasters, J.J. Minocycline and doxycycline, but not other tetracycline‐ derived compounds, protect liver cells from chemical hypoxia and ischemia/ reperfusion injury by inhibition of the mitochondrial calcium uniporter. Toxicol Appl Pharmacol, 2013;273(1):172–9. 49. Breuer, W., Epsztejn, S., Millgram, P., and Cabantchik, I.Z. Transport of iron and other transition metals into cells as revealed by a fluorescent probe. Am J Physiol, 1995;268(6 Pt 1):C1354–C61. 50. Uchiyama, A., Kim, J.S., Kon, K. et al. Translocation of iron from lysosomes into mitochondria is a key event during oxidative stress‐induced hepatocellular injury. Hepatology, 2008;48(5):1644–54. 51. Zhang, X. and Lemasters, J.J. Translocation of iron from lysosomes to ­mitochondria during ischemia predisposes to injury after reperfusion in rat hepatocytes. Free Radic Biol Med, 2013;63:243–53. 52. Kim, J.S., Wang, J.H., and Lemasters, J.J. Mitochondrial permeability transition in rat hepatocytes after anoxia/reoxygenation: role of Ca2+‐dependent mitochondrial formation of reactive oxygen species. Am J Physiol Gastrointest Liver Physiol, 2012;302(7):G723–31. 53. Czerny, C., Kholmukhamedov, A., Theruvath, T.P. et  al. Minocycline decreases liver injury after hemorrhagic shock and resuscitation in mice. HPB Surg, 2012;2012:259512. 54. Kholmukhamedov, A., Zhang, X., Schwartz, J., and Lemasters, J.J. Lysosomal iron release promotes hemorrhage/resuscitation‐induced liver injury. Toxicol Sci, 2012;125 (Suppl.):537. 55. Zhang, T., Zhao, Q., Ye, F., Huang, C.Y., Chen, W.M., and Huang, W.Q. Alda‐1, an ALDH2 activator, protects against hepatic ischemia/reperfusion injury in rats via inhibition of oxidative stress. Free Radic Res, 2018;52(6):629–38. 56. Zhong, W., Zhang, W., Li, Q. et al. Pharmacological activation of aldehyde dehydrogenase 2 by Alda‐1 reverses alcohol‐induced hepatic steatosis and cell death in mice. J Hepatol, 2015;62(6):1375–81. 57. Zhong, Z., Ramshesh, V.K., Rehman, H. et al. Acute ethanol causes hepatic mitochondrial depolarization in mice: role of ethanol metabolism. PLoS One, 2014;9(3):e91308. 58. Fontana, R.J. Acute liver failure including acetaminophen overdose. Med Clin North Am, 2008;92(4):761–94. 59. Ramachandran, A. and Jaeschke, H. Acetaminophen toxicity: novel insights into mechanisms and future perspectives. Gene Expr, 2018;18(1):19–30. 60. Hu, J., Kholmukhamedov, A., Lindsey, C.C., Beeson, C.C., Jaeschke, H., and Lemasters, J.J. Translocation of iron from lysosomes to mitochondria during acetaminophen‐induced hepatocellular injury: protection by starch‐desferal and minocycline. Free Radic Biol Med, 2016;97:418–26. 61. Hu, J., Nieminen, A.L., Ward, D.M., Klatt, S.C., and Lemasters, J.J. Important role of the mitochondrial Ca2+ uniporter (MCU) but not divalent

metal trans‐porter 1 (DMT1) or mitoferrin2 (Mfrn2) in acetaminophen (APAP)‐induced hepatotoxicity in mice. Hepatology, 2018;68 (Suppl.):201A. 62. Ramachandran, A., Lebofsky, M., Baines, C.P., Lemasters, J.J., and Jaeschke, H. Cyclophilin D deficiency protects against acetaminophen‐induced ­oxidant stress and liver injury. Free Radic Res, 2011;45(2):156–64. 63. Cover, C., Mansouri, A., Knight, T.R. et al. Peroxynitrite‐induced mitochondrial and endonuclease‐mediated nuclear DNA damage in acetaminophen hepatotoxicity. J Pharmacol Exp Ther, 2005;315(2):879–87. 64. Saito, C., Lemasters, J.J., and Jaeschke, H. c‐Jun N‐terminal kinase modulates oxidant stress and peroxynitrite formation independent of inducible nitric oxide synthase in acetaminophen hepatotoxicity. Toxicol Appl Pharmacol, 2010;246(1–2):8–17. 65. Agarwal, R., Hennings, L., Rafferty, T.M. et  al. Acetaminophen‐induced hepatotoxicity and protein nitration in neuronal nitric‐oxide synthase knockout mice. J Pharmacol Exp Ther, 2012;340(1):134–42. 66. Ramachandran, A., Lebofsky, M., Weinman, S.A., and Jaeschke, H. The impact of partial manganese superoxide dismutase (SOD2)‐deficiency on mitochondrial oxidant stress, DNA fragmentation and liver injury during acetaminophen hepatotoxicity. Toxicol Appl Pharmacol, 2011;251(3):226–33. 67. Uehara, T., Bennett, B., Sakata, S.T. et al. JNK mediates hepatic ischemia reperfusion injury. Journal of Hepatology, 2005;42(6):850–9. 68. Win, S., Than, T.A., Zhang, J., Oo, C., Min, R.W.M., and Kaplowitz, N. New insights into the role and mechanism of c‐Jun‐N‐terminal kinase signaling in the pathobiology of liver diseases. Hepatology, 2018;67(5):2013–24. 69. Du, K., Xie, Y., McGill, M.R., and Jaeschke, H. Pathophysiological significance of c‐jun N‐terminal kinase in acetaminophen hepatotoxicity. Exp Opin Drug Metab Toxicol, 2015;11(11):1769–79. 70. Jaeschke, H. Reactive oxygen and mechanisms of inflammatory liver injury: present concepts. J Gastroenterol Hepatol, 2011;26(1):173–9. 71. Yin, C. and Heit, B. Armed for destruction: formation, function and trafficking of neutrophil granules. Cell Tissue Res, 2018;371(3):455–71. 72. Jaeschke, H. Mechanisms of liver injury. II. Mechanisms of neutrophil‐ induced liver cell injury during hepatic ischemia‐reperfusion and other acute inflammatory conditions. Am J Physiol Gastrointest Liver Physiol, 2006;290(6):G1083–8. 73. Jorens, P.G., Matthys, K.E., and Bult, H. Modulation of nitric oxide synthase activity in macrophages. Mediators Inflamm, 1995;4(2):75–89. 74. Jaeschke, H. and Hasegawa, T. Role of neutrophils in acute inflammatory liver injury. Liver Int, 2006;26(8):912–9. 75. Kubes, P. and Mehal, W.Z. Sterile inflammation in the liver. Gastroenterology, 2012;143(5):1158–72. 76. Woolbright, B.L. and Jaeschke, H. The impact of sterile inflammation in acute liver injury. J Clin Transl Res, 2017;3(1):170–88. 77. Takeuchi, O. and Akira, S. Pattern recognition receptors and inflammation. Cell, 2010;140(6):805–20. 78. Jaeschke, H. Molecular mechanisms of hepatic ischemia‐reperfusion injury and preconditioning. Am J Physiol Gastrointest Liver Physiol, 2003;284(1):G15–26. 79. Jaeschke, H. and Woolbright, B.L. Current strategies to minimize hepatic ischemia‐reperfusion injury by targeting reactive oxygen species. Transplant Rev, 2012;26(2):103–14. 80. Ma, Q. Role of nrf2 in oxidative stress and toxicity. Annu Rev Pharmacol Toxicol, 2013;53:401–26. 81. Farrell, G.C., Teoh, N.C., and McCuskey, R.S. Hepatic microcirculation in fatty liver disease. Anat Rec, 2008;291(6):684–92. 82. Jaeschke, H., McGill, M.R., and Ramachandran, A. Oxidant stress, mitochondria, and cell death mechanisms in drug‐induced liver injury: ­ lessons learned from acetaminophen hepatotoxicity. Drug Metab Rev, ­ 2012;44(1):88–106. 83. Jaeschke, H., Williams, C.D., Ramachandran, A., and Bajt, M.L. Acetaminophen hepatotoxicity and repair: the role of sterile inflammation and innate immunity. Liver Int, 2012;32(1):8–20. 84. Williams, C.D., Farhood, A., and Jaeschke, H. Role of caspase‐1 and interleukin‐1beta in acetaminophen‐induced hepatic inflammation and liver injury. Toxicol Appl Pharmacol, 2010;247(3):169–78. 85. Liu, Z.X., Han, D., Gunawan, B., and Kaplowitz, N. Neutrophil depletion protects against murine acetaminophen hepatotoxicity. Hepatology, 2006;43(6):1220–30. 86. Cover, C., Liu, J., Farhood, A. et  al. Pathophysiological role of the acute inflammatory response during acetaminophen hepatotoxicity. Toxicol Appl Pharmacol, 2006;216(1):98–107.



55:  Oxidative Stress and Inflammation in the Liver

87. Williams, C.D., Bajt, M.L., Sharpe, M.R., McGill, M.R., Farhood, A., and Jaeschke, H. Neutrophil activation during acetaminophen hepatotoxicity and repair in mice and humans. Toxicol Appl Pharmacol, 2014;275(2):122–33. 88. Dambach, D.M., Watson, L.M., Gray, K.R., Durham, S.K., and Laskin, D.L. Role of CCR2 in macrophage migration into the liver during acetaminophen‐ induced hepatotoxicity in the mouse. Hepatology, 2002;35(5):1093–103. 89. You, Q., Holt, M., Yin, H., Li, G., Hu, C.J., and Ju, C. Role of hepatic resident and infiltrating macrophages in liver repair after acute injury. Biochem Pharmacol, 2013;86(6):836–43. 90. Antoniades, C.G., Quaglia, A., Taams, L.S. et al. Source and characterization of hepatic macrophages in acetaminophen‐induced acute liver failure in humans. Hepatology, 2012;56(2):735–46. 91. Woolbright, B.L. and Jaeschke, H. Novel insight into mechanisms of cholestatic liver injury. World J Gastroenterol, 2012;18(36):4985–93. 92. Faubion, W.A., Guicciardi, M.E., Miyoshi, H. et al. Toxic bile salts induce rodent hepatocyte apoptosis via direct activation of Fas. J Clin Invest, 1999;103(1):137–45. 93. Gujral, J.S., Farhood, A., Bajt, M.L., and Jaeschke, H. Neutrophils aggravate acute liver injury during obstructive cholestasis in bile duct‐ligated mice. Hepatology, 2003;38(2):355–63.

727

94. Allen, K., Jaeschke, H., and Copple, B.L. Bile acids induce inflammatory genes in hepatocytes: a novel mechanism of inflammation during obstructive cholestasis. Am J Pathol, 2011;178(1):175–86. 95. Cai, S.Y., Ouyang, X., Chen, Y. et  al. Bile acids initiate cholestatic liver  injury by triggering a hepatocyte‐specific inflammatory response. JCI insight, 2017;2(5):e90780. 96. Yang, M., Ramachandran, A., Yan, H.M. et  al. Osteopontin is an initial mediator of inflammation and liver injury during obstructive cholestasis after bile duct ligation in mice. Toxicol Lett, 2014;224(2):186–95. 97. Woolbright, B.L., Antoine, D.J., Jenkins, R.E., Bajt, M.L., Park, B.K., and Jaeschke, H. Plasma biomarkers of liver injury and inflammation demonstrate a lack of apoptosis during obstructive cholestasis in mice. Toxicol Appl Pharmacol, 2013;273(3):524–31. 98. Woolbright, B.L., Dorko, K., Antoine, D.J. et al. Bile acid‐induced necrosis in primary human hepatocytes and in patients with obstructive cholestasis. Toxicol Appl Pharmacol, 2015;283(3):168–77. 99. Li, M., Cai, S.Y., and Boyer, J.L. Mechanisms of bile acid mediated ­inflammation in the liver. Mol Aspects Med, 2017;56:45–53. 100. Woolbright, B.L. and Jaeschke, H. Therapeutic targets for cholestatic liver injury. Exp Opin Ther Targets, 2016;20(4):463–75.

56

The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury Shi‐Ying Cai, Man Li, and James L. Boyer The Liver Center, Yale University School of Medicine, New Haven, CT, USA

INTRODUCTION Bile acids are metabolites of cholesterol. They are amphipathic molecules because they contain a hydrophobic nuclear steroid ring and a hydrophilic carboxyl side‐chain. There are many different forms of bile acids in mammals, varying on the length of the side‐chain, the conjugates on the C‐terminus of the side‐ chain, and the position and configuration of hydroxy groups on the nuclear steroid ring [1]. Because of these variations, different bile acids possess different physiochemical properties with some being more hydrophobic than others. Primary bile acids are synthesized in hepatocytes and are mostly conjugated with glycine or taurine in humans and rodents, whereas secondary bile acids are generated by gut microbes which remove hydroxyl groups or alter their configuration on the nuclear steroid ring. In general, secondary bile acids are more hydrophobic and cytotoxic than primary bile acids. Bile acids were initially recognized as detergents for facilitating the absorption of lipids and fat‐soluble vitamins in our digestion system. Recently, they have also been identified as signaling molecules that modulate gene expression through specific receptors on the cell membranes and nucleus that are involved in many physiological processes. Therefore, bile acids play a very important role in health and disease, and their homeostasis is tightly regulated. Normally, the bile acid pool undergoes an enterohepatic circulation with around 5% lost into the feces daily. Bile acid transporters play a pivotal role in mediating the uptake and excretion of conjugated bile acid across cell membranes in the liver and ileum [2]. When bile formation is impaired due to primary injury of hepatocytes or obstruction of the bile duct, bile acids accumulate in the liver  and systemic circulation, resulting in the syndrome of cholestasis.

There are many causes of cholestasis including genetic and developmental defects, as well the effects of drugs, viral hepatitis, alcoholic liver disease, primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), pregnancy, metabolic syndrome, and bile duct obstruction from gallstones or tumors [3–5]. Many of these disorders become chronic leading eventually to biliary cirrhosis and the need for liver transplantation. Whatever the cause, the hepatic accumulation of bile acids is common to all. Nevertheless, the mechanism as to how bile acids injure the liver has remained elusive and controversial. Bile acids are potent detergents and earlier studies suggested that bile acids caused liver injury by their direct cytotoxic effects. However, these studies used millimolar concentrations of toxic bile acids to treat hepatocytes in culture [6–8], and such high levels of toxic bile acids are not pathophysiologically ­relevant as they have never been observed in vivo in the serum or livers of cholestatic patients or animals [9]. Later, it was proposed that bile acids injured the liver by inducing apoptosis in hepatocytes [10–16], and that the apoptotic body triggered inflammation that further exacerbated the liver injury [17, 18]. This hypothesis was based on in vitro observations obtained by exposing rat hepatocytes to the toxic bile acid, GCDCA. However, GCDCA is not a major endogenous bile acid in the rat  [19]. Most importantly, evidence for apoptosis has not been found in cholestatic livers from humans and rodents, nor in mouse and human hepatocyte cultures when treated with appropriate species‐specific major endogenous bile acids ­ at  pathophysiological relevant concentrations [20–24]. Furthermore, apoptosis normally does not trigger an inflammatory response. In contrast, most recent studies indicate that cholestatic liver injury results from an inflammatory response where pathophysiological levels of bile acids induce the

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



56:  The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury

production of proinflammatory mediators in hepatocytes that attract immune cells and initiate inflammation in the liver [20, 24]. In this chapter, we review recent advances in this research.

CHOLESTATIC HEPATOCYTES RELEASE PROINFLAMMATORY MEDIATORS Hepatocytes are the major source of bile formation. They take up bile acids from the blood and excrete them into the bile lumen via specific membrane transport systems (see Chapter 31). This process depends primarily on the sodium‐dependent taurocholate cotransporting polypeptide (NTCP/SLC10A1) on the basolateral membrane and the bile salt export pump (BSEP/ ABCB11) on the canalicular membrane, respectively. Deficiency of NTCP/Ntcp in a patient and Slc10a1−/− mice respectively leads to hypercholanemia but without apparent liver injury [25, 26]. This suggests that if bile acids are only elevated in the serum but not in the liver as a result of deficiency of NTCP/ Ntcp, liver injury should not occur. In contrast, mutations or polymorphisms in the BSEP (ABCB11), result in bile acid accumulation in hepatocytes and the onset of cholestasis as seen in patients with progressive familial intrahepatic cholestasis type 2 (PFIC2), benign recurrent intrahepatic cholestasis (BRIC2), and some cases of intrahepatic cholestasis of pregnancy (ICP) [27]. Together, these observations suggest that bile acids must enter hepatocytes in order to cause cholestatic liver injury and that this injury is not mediated through a plasma membrane receptor on hepatocytes. Recently, Allen et  al. first proposed that cholestatic liver injury is mediated through an inflammatory response, because the major endogenous bile acid in mouse, taurocholic acid (TCA), stimulated the expression of proinflammatory genes in mouse hepatocyte cultures. These included chemokines and adhesion molecules, for example, Ccl2, Ccl5, Cxcl1, Cxcl2, Cxcl10, Icam‐1, and Vcam1 [20]. Release of these chemoattractants resulted in the hepatic recruitment of immune cells that triggered inflammatory tissue injury. This hypothesis is supported by our recent studies where we confirmed that under pathophysiologically relevant concentrations of TCA (25–200 μM), a panel of chemokines was expressed in a dose‐ and time‐ dependent manner in isolated mouse hepatocyte cultures [24]. We found that TCA continued to induce chemokine production in these cells when treated for a period of 48 hours. These findings help explain why persistence in elevated levels of bile acids continues to sustain liver inflammation in cholestatic patients. In contrast, TCA did not substantially stimulate the expression of these chemokines in cultures of mouse cholangiocytes, or liver non‐parenchymal cells (including Kupffer cell, stellate cell, and sinusoidal endothelium). In addition, knockdown of Ntcp or blocking the uptake of bile acids using biotinylated bile acid analogs significantly reduced bile acid induced expression of chemokines in mouse hepatocytes, explaining why hepatocytes but not other liver cells that don’t express Ntcp are so susceptible to bile acids. This is also consistent with the observations in the NTCP/Ntcp‐deficient patient and mice mentioned earlier. Furthermore, the chemoattractants released in the culture medium from bile acid treated mouse hepatocytes stimulated

729

neutrophil migration in a transwell experiment, further supporting the concept that bile acid induction of proinflammatory genes in hepatocytes is pathologically important in cholestatic liver injury. Liver injury was also greatly diminished after bile duct ligation (BDL) or cholic acid feeding in Ccl2‐deficient mice [24]. Most importantly, GCDCA, the major endogenous bile acid in humans, significantly stimulated the expression of chemokine CCL2, CCL15, CCL20, CXCL1, and IL‐8 in primary human hepatocyte cultures, indicating that human and mouse share the same mechanistic response to bile acids [24]. To better understand how bile acids stimulate the expression of proinflammatory genes in cholestatic liver, Allen et al. examined the functional roles of the bile acid nuclear receptor (farnesoid X receptor) FXR/NR1H4, the transcription factor early growth response factor‐1 (EGR‐1), and the toll‐like receptor (TLR) 4 signal transduction pathway. Although FXR plays a key role in maintaining bile acid homeostasis by regulating the expression of genes involved in bile acid synthesis and transport, deficiency of Fxr did not reduce hepatic expression of Cxcl2 and Icam‐1, nor cholestatic liver injury after bile duct ligation in mice. Cholestatic liver injury was also not affected in Tlr4‐deficient mice after bile duct ligation when compared to their wild‐type experimental controls, nor was hepatic neutrophil infiltration altered either [20]. These observations not only indicate that Tlr4 does not play a direct role in stimulating hepatic inflammation in cholestasis but also suggest that lipopolysaccharide (LPS) is not a contributor to liver injury in these mouse models [20]. In contrast, they found that knockout of Egr‐1 significantly attenuated cholestatic liver injury in mice after bile duct ligation [28]. Deficiency of Egr‐1 also reduced TCA induction of some proinflammatory genes in mouse ­hepatocyte cultures, suggesting that Egr‐1 regulates bile acid stimulated expression of proinflammatory genes at least in part [20]. To explain how bile acid causes Egr‐1 activation, they speculated that MAPKs may be activated by bile acids because they found that bile acid induction of Egr‐1 expression was associated with ERK activation in mouse hepatocytes [20, 29]. However, the details of this activation remain to be determined. Additional studies found that TLR9 played an important role in proinflammatory gene induction in hepatocytes [24]. TLR9 is a membrane‐bound receptor that senses DNA preferably from mitochondria, bacteria, and viruses. When activated, TLR9 triggers an innate immune response via signaling cascades that lead to a proinflammatory cytokine response. In cholestatic hepatocytes, bile acids cause mitochondrial damage and endoplasmic reticulum (ER) stress, as evidenced by changes in the mitochondrial membrane potential and the leak of mitochondrial and ER  proteins into the cytosol. These abnormalities were also observed in other cultured cells overloaded with bile acids [30, 31]. Bile acid‐induced leakage of DNA from damaged mitochondria could then activate TLR9 that in turn stimulates the expression of proinflammatory genes. Indeed, deficiency of Tlr9 diminished TCA induction of chemokine Cxcl2 in mouse hepatocytes [24]. Reduced liver injury was also found in Tlr9‐ deficient mice after bile duct ligation [24, 32]. However, how intracellular mitochondrial DNA activates Tlr9 remains to be studied. In addition, deficiency of Tlr9 did not fully eliminate bile acid induction of chemokines in mouse hepatocytes, indicating that there are also Tlr9‐independent signaling pathways

730

THE LIVER:  THE RESPONSE OF CHOLANGIOCYTES TO CHOLESTASIS

involved in this regulation. One such pathway may involve Ca2+ signaling because cyclosporine A, a well‐known inhibitor in this pathway greatly repressed bile acid induction of chemokines in mouse hepatocytes [24]. Future studies may address these questions.

THE RESPONSE OF CHOLANGIOCYTES TO CHOLESTASIS Cholangiocytes are epithelial cells that line the lumen of the intra‐ and extrahepatic bile ducts. Intrahepatic cholangiocytes are heterogenous in morphology as well as function. In rodents, the cholangiocytes that are greater than or less than 15 μm in diameter are categorized into large and small cholangiocytes, respectively. Large cholangiocytes, which are more columnar, express various transporters such as the secretin receptor, the water channel aquaporin, the cystic fibrosis transmembrane conductance regulator (CFTR/ABCC7), and the Cl− /HCO3− anion exchanger 2 (AE2/SLC4A2), whereas the expression of these proteins is absent under normal conditions in small cholangiocytes. Functionally, large but not small cholangiocytes are responsible for secretion of an aqueous fluid rich in bicarbonate in response to meal‐induced excretion of the gastrointestinal hormones secretin, vasoactive intestinal peptide, and bombesin, which in humans contributes about 25% of the daily bile production [2]. Large cholangiocytes are also more susceptible to liver damage and proliferate extensively after bile duct ligation. In contrast, small cholangiocytes are more resistant to liver damage and possess characteristics of stem/progenitor cells that can differentiate into large cholangiocytes during bile duct injury and regeneration [33]. Under physiological conditions cholangiocytes are constantly exposed to high concentrations of bile acids in the millimolar range in the bile. However, unlike hepatocytes, these cells do not normally show signs of injury and there is little evidence to support the notion that bile acids directly initiate an inflammatory response in normal cholangiocytes. This is demonstrated by recent studies, where treatment of isolated mouse cholangiocytes with pathophysiological concentrations of major endogenous bile acids present in serum (25–200 μM) did not stimulate proinflammatory cytokine production [24]. It is likely that the intracellular concentrations of bile acids in cholangiocytes do not reach levels that would trigger an inflammatory response as seen in hepatocytes, because the abundant expression of the basolateral membrane bile acid efflux transporter OSTα–OSTβ in cholangiocytes can effectively allow diffusion of bile acids back into the blood. In addition, cholangiocytes also secrete a layer of mucous that is rich in bicarbonate. This “biliary HCO3− umbrella” maintains an alkaline pH near the apical surface of cholangiocytes as a barrier to prevent the protonation of glycine‐conjugated bile salts and minimize the uptake of bile acids from bile. The pKa of glycine bile acid conjugates are around 4, so that an alkaline bile will maintain these bile acids in a charged non‐protonated state, thus preventing their uncontrolled diffusion across the apical membranes into the cholangiocytes [34]. In line with this hypothesis, a 20 to 40 nm‐thick extracellular, juxta membranous layer of glycocalyx has been identified on

the apical membrane of human and mouse biliary epithelium. It is also supported by studies showing that bile acid uptake and toxicity were dependent on pH and the key HCO3− exporter AE2 in immortalized human cholangiocytes [35, 36]. The potential role of the biliary HCO3− umbrella has also been implicated in the pathogenesis of PBC where defective expression of AE2 and impaired HCO3− secretion has been reported. Deletion of the Ae2a,b gene in mice also causes a PBC‐like phenotype, including portal inflammation with CD8(+) and CD4(+) T ­lymphocyte infiltration around damaged bile ducts and altered gene ­expression profiles in isolated cholangiocytes [37–39]. Besides bile acid transporters, cholangiocytes also express bile acid receptors, including nuclear receptors FXR and the vitamin D receptor, as well as plasma‐membrane bound G protein‐coupled receptors, such as the Takeda G protein‐­coupled receptor 5 (TGR5, Gpbr‐1, M‐BAR) and sphingosine 1‐phosphate receptor 2 (S1PR2) [40]. Whereas both unconjugated and conjugated bile acids are ligands of TGR5, only conjugated bile acids have been shown to activate S1PR2 [41]. In the liver, TGR5 has been detected in large and small cholangiocytes, Kupffer cells, and sinusoidal endothelial cells, but not hepatocytes [42]. TGR5 has been associated with PSC as mutations in TGR5 have been identified in PSC patients [43]. Interestingly, the effect of TGR5 activation in cholangiocytes is dependent on its subcellular localization. In ciliated human cholangiocyte H69 cell lines, activation of TGR5 resulted in colocalization with inhibitory Gα(i) proteins, leading to lower intracellular cAMP levels and attenuated cell proliferation. However, the opposite effect was observed in non‐ciliated H69 cells, where TGR5 is localized on the apical plasma membrane and colocalized with stimulatory Gα(s) proteins upon activation [44]. Studies using Tgr5−/− mice demonstrated that deficiency of Tgr5 leads to reduced bile duct proliferation in response to BDL or cholic acid feeding, and that in primary mouse cholangiocytes, Tgr5 induces cell proliferation via ROS production and subsequent activation of Src kinase, EGF receptor, and ERK1/2 phosphorylation [45]. In addition, studies in partially hepatectomized and Mdr2−/− (Abcb4−/−) mice with sclerosing cholangitis‐like bile duct injury showed that Tgr5 activation enhances choleresis by increasing intracellular cAMP levels and consequently Cl− and HCO3− secretion mediated by CFTR and AE2, respectively [46, 47]. Recently, Wang et  al. demonstrated that the expression of S1PR2 mRNA was upregulated in mouse liver after BDL for two weeks, and that TCA induced cell proliferation by activating S1PR2 and the ERK1/2‐AKT signaling pathway in isolated mouse cholangiocytes. S1PR2 knockout mice had significant reductions in serum bile acid and ALP levels as well as liver fibrosis after two‐week BDL where diminished cyclooxygenase 2 mRNA expression was also found in the knockout livers and cholangiocytes [48]. However, it remains unclear whether these mice had an attenuated inflammatory response in hepatocytes to account for this improvement as opposed to an absence of cyclooxygenase 2 mRNA induction [48]. In response to inflammatory insults such as TNF‐α, IL‐6, or LPS stimulation, cholangiocytes become reactive and start to release large amounts of proinflammatory mediators, including neutrophil chemoattractant IL‐8 and the epithelial cell‐derived neutrophil activating protein (ENA‐78). Fibrotic mediators that



56:  The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury

activate hepatic stellate cells such as TGF‐β2 and Ccl2 are also  stimulated and further augment biliary inflammation and fibrosis [49]. In addition, cholangiocytes express and secret osteopontin (OPN), a multifunctional glycophosphoprotein that recruits neutrophils, macrophages, and natural killer T cells by binding to integrin receptors expressed on these inflammatory cells [50]. In BDL mice, both the expression of osteopontin in cholangiocytes and its cleaved form in bile are significantly induced, presumably by cholangiocyte stress that results from the pressure in the biliary system after obstruction of the bile duct. In addition, Opn−/− mice showed a delayed inflammatory response after BDL, as neutrophil infiltration was dramatically reduced and bile infarcts were nearly absent 1 but not 3 days after BDL, suggesting that osteopontin plays a role in attracting neutrophils only at the very early stages of liver injury in this model [51]. In humans, high levels of IL‐8 in the bile and cholangiocytes were detected in PSC patients compared to other non‐PSC patients. Increased expression of IL‐8 protein on intrahepatic bile duct epithelium was observed as the disease advanced. Treatment with IL‐8 in primary human cholangiocytes cultures also induced cell proliferation and production of profibrotic genes, suggesting that IL‐8 may be involved in the pathogenesis of PSC [52].

THE ROLE OF IMMUNE CELLS IN CHOLESTATIC LIVER INJURY The liver sinusoids feature a capillary system lined with highly fenestrated sinusoidal endothelial cells, which allows liver cells to directly receive blood supply from both the systemic circulation via the hepatic artery and the gastrointestinal tract via the portal vein, making the liver constantly exposed to a massive load of antigens derived from nutrients or microbiota. Therefore, while it must remain tolerant to harmless antigens, the hepatic immune system also needs to respond rapidly to fight infection, limit tissue damage, and facilitate tissue regeneration. The hepatic sinusoids are enriched with several types of immune cells, such as the liver resident macrophages (Kupffer cells), neutrophils, T cells, natural killer cells, and dendritic cells that are capable of pathogen sensing, phagocytosis, cytotoxicity, cytokine release, and antigen presentation [53]. Neutrophils are a key component of the innate immune system. They are the most abundant type of white blood cells in mammals and circulate in the peripheral blood in normal conditions. In response to microbial infection as well as tissue damage, neutrophils act as one of the first‐responders of inflammatory cells that migrate through the blood vessels, then through interstitial tissue, toward the site of inflammation following chemical signals in a process known as chemotaxis. In the process of sterile inflammation, molecules termed damage associated molecular patterns (DAMPs), such as high‐mobility group box‐1 (HMGB1) protein, heat shock proteins, ATP, nuclear and mitochondrial DNA, are released from stressed or damaged cells. These DAMPs then bind and activate their respective receptors, which in turn initiate the expression of proinflammatory mediators that recruit cytotoxic cells to the injured sites [54].

731

Neutrophils directly kill hepatocytes through production of large quantities of cytotoxic reactive oxygen species (ROS) such as hypochlorous acid, a potent oxidant generated via myeloperoxidase, as well as proinflammatory cytokines. Excessive activation and infiltration of neutrophils to the hepatic parenchyma leads to liver injury, as observed in many liver diseases, such as viral hepatitis, non‐alcoholic fatty liver disease, alcoholic liver disease, liver fibrosis/cirrhosis, and other causes of liver failure [55, 56]. Accumulating studies reviewed above ­support the hypothesis that neutrophils are the principal cause of hepatocyte toxicity in the early stages of cholestatic liver injury. After bile duct ligation in mice, neutrophils are the predominant infiltrating immune cells observed during the acute phase of liver injury. They can be detected within 8 hours after BDL and reach maximum levels around 2–3 days, mainly in areas of injured hepatocytes and in surrounding sinusoids [57, 58]. Elevated myeloperoxidase activity, which indicates increased number of neutrophils, has also been found in the liver of bile duct ligated rats [59, 60]. Neutrophil depleted‐mice displayed much less hepatocyte injury when subjected to α‐naphthyl ­isothiocyanate (ANIT)‐induced intrahepatic cholestasis [61]. In Mdr2−/− mice, a well‐established model for cholestasis, marked hepatic neutrophil infiltration is observed following elevation of proinflammatory cytokines but prior to detectable histologic and biochemical evidence of liver cell injury [62]. These findings strongly suggest that neutrophils are activated and recruited to the hepatic parenchyma by proinflammatory mediators induced by high levels of bile acids, where they target and kill stressed or injured hepatocytes. Intercellular adhesion molecule‐1 (ICAM‐1) is a cell adhesion molecule that mediates neutrophil extravasation by interacting with β2‐integrins expressed on neutrophil cell surface [63]. A growing body of evidence indicates a strong correlation between the expression levels of ICAM‐1 and the degree of cholestatic injury in both humans and in animal models. In healthy humans, ICAM‐1 is only expressed at low levels in the endothelium of some portal vessels and sinusoidal lining cells, and is not detected in hepatocytes. However, in patients with obstructive cholestasis, ICAM‐1 is not only upregulated on sinusoidal endothelial and Kupffer cells but also induced on the sinusoidal membrane of hepatocytes in areas of parenchymal cell injury [64]. Increased ICAM‐1 expression was also detected on the endothelium of micro vessels in chronic cholangitis patients with complete bile duct obstruction [64]. Compared with the wild‐type control, mice deficient of Icam‐1 or β2 integrin CD18 had dramatically attenuated liver necrosis as well as hepatic neutrophil accumulation after BDL [57, 65]. The high mobility of neutrophils observed during chemotaxis requires rapid cytoskeletal reorganization. Our studies found that the cytoskeletal proteins ezrin‐radixin‐moesin (ERM) and Na+/H+ exchanger regulatory factor‐1 (NHERF‐1, also known as ERM‐binding phosphoprotein 50 [EBP50]) participate in neutrophil extravasation in cholestatic liver injury [66]. In the liver, expression of ERM proteins and NHERF‐1 are detected beneath plasma membranes of hepatocytes and biliary epithelial cells [67–69]. They possess multiple protein binding sites that are capable of tethering membrane proteins to underlying F‐actin network in microvilli‐like membrane projections. These “docking” structures anchor and partially embrace leukocytes,

732

THE LIVER:  THE ROLE OF IMMUNE CELLS IN CHOLESTATIC LIVER INJURY

including neutrophils, to enhance firm leukocyte adhesion and initiate leukocyte transmigration across endothelial and epithelial cells such as hepatocytes [70, 71]. Our results show that in wild‐type mice, NHERF‐1 recruits ERM proteins, ICAM‐1 and F‐actin into a macromolecule complex that is increased in the liver at the plasma membranes after BDL and participates in neutrophil trans‐endothelial and ‐hepatocyte migration induced by BDL. In contrast, mice deficient in NHERF‐1 display reduced levels of activated ERM and ICAM‐1 protein in both the liver and hepatocytes. Compared with wild‐type controls, Nherf‐1−/− mice exhibit significantly reduced neutrophil infiltration in the liver accompanied with attenuated liver necrosis and lower serum alanine aminotransferase levels after BDL. These findings suggest that NHERF‐1 plays a key role in the formation of ICAM‐1/ERM/NHERF‐1 macromolecule complexes that participate in neutrophil mediated liver injury in cholestasis. The liver contains about 80% of all body macrophages that can be divided into Kupffer cells, the tissue‐resident and self‐ renewing macrophages located in the sinusoids, and macrophages originate from circulating blood monocytes or myeloid precursors that are recruited to the liver during injury or after Kupffer cell depletion [72]. Conventionally, macrophages have been assigned as “M1” or “M2” subsets. The M1 macrophages can be differentiated by interferon gamma (IFNγ) or toxins such as bacterial LPS. Activated M1 macrophages produce proinflammatory mediators such as tumor necrosis factor (TNF)‐α, IL‐1β, and ROS, which promote liver inflammation and injury as the disease progresses. The other subset, M2 macrophages, which are alternatively activated by IL‐4 and IL‐13, release IL‐4, IL‐10, and IL‐13 that have anti‐inflammatory effects. However, recent studies show that during liver injury, liver macrophages are highly plastic as “mixed” macrophage phenotypes that simultaneously express pro‐ and anti‐inflammatory mediators have also been discovered. These “mixed” macrophages can adapt rapidly from a proinflammatory to an anti‐inflammatory phenotype in response to changes in the hepatic microenvironment [72, 73]. The broad spectrum of subtypes of hepatic macrophages exert multiple functions in liver immunity, including phagocytosis of dying cells and cell debris, initiation of an immune response in other liver cells such as hepatocytes, antigen presentation, and immune cell recruiting [74]. Populational and functional alterations of macrophages and monocytes have been implicated in cholestatic liver diseases and animal models. Increased peribiliary M1‐ and M2‐like monocyte‐derived macrophages have been found in patients with stage 4 PSC as well as in the liver of Mdr2−/− mice compared to normal livers [75]. Kupffer cells isolated from BDL mice exhibit delayed clearance of bacteria and higher levels of IL‐10 and reciprocally lower levels of IL‐12 production in response to LPS stimulation [76]. In cholestatic patients, monocytes obtained from jaundiced patients demonstrated reduced IL‐1β and IL‐6 release in response to endotoxin compared with controls [77]. Monocytes and macrophages abundantly express TGR5 that can be activated by both conjugated and unconjugated bile acids, with the rank order of potency (EC50) at TLCA (0.33 μM) greater than LCA (0.53 μM) greater than DCA (1.01 μM) greater than CDCA (4.43 μM) greater than CA (7.72 μM) [78, 79]. Studies using primary human macrophages showed that TLCA inhibits the LPS‐induced expression of

proinflammatory cytokines TNF‐α, IL‐6, IL‐12, and IFNβ without affecting the expression of the anti‐inflammatory cytokine IL‐10, resulting in a macrophage phenotype with an increased IL‐10/IL‐12 ratio as well as a suppressed basal phagocytic activity [80]. A recent analysis by Wammers et al. found that in the presence of LPS, TLCA treatment modulates the expression of a broad spectrum of genes in primary human macrophages, with downregulation of proinflammatory genes involved in phagocytosis, pathogen interaction, and recruitment of immune cells, as well as upregulation of genes involved in wound healing, cell differentiation, and anti‐inflammatory signaling. TLCA treatment also blocked LPS‐induced, macrophage‐dependent natural killer (NK) cell migration in vitro [81]. In the liver, TGR5 has been identified in Kupffer cells and is upregulated in rats after BDL [82]. Deletion of TGR5 in mice cause higher aspartate aminotransferase (AST) levels after cholic acid feeding, increased liver necrosis and serum CCL2 (MCP‐1) levels 2 or 3 days after BDL, further demonstrating a role for TGR5 in protection of the cholestatic liver [45, 46]. The anti‐inflammatory effect of TGR5 in macrophages is mediated by inhibition of NF‐κB and JNK signaling pathways [78], as well as through inhibition of NLRP3 inflammasome ­activation as discussed in more detail below. However, findings on the role of Kupffer cells obtained from animal models of cholestatic liver injury remain controversial. One study showed that administration of gadolinium chloride, a Kupffer cell inhibitor, attenuated liver injury and fibrosis in BDL rats, indicating that Kupffer cells promote BDL‐induced liver injury [83]. In contrast, mice treated with liposome‐­ encapsulated dichloromethylene diphosphonate or alendronate for Kupffer cell depletion had augmented liver injury, as well as decreased hepatocyte regeneration and liver fibrosis than control mice 7 or 10 days after BDL, suggesting that Kupffer cells have a protective role for hepatocyte injury and promote cell survival, regeneration, and fibrosis in cholestasis [84, 85]. In addition, these studies showed that Kupffer cells from mice at 6 hours but not at 24 hours after BDL released more IL‐6 that suppressed liver injury, whereas no significant differences in liver histology and ALT levels were found 24 hours after surgery. These contrary observations might be related to the heterogeneity and functional complexity of Kupffer cells in cholestatic liver injury. Further investigation is needed to clarify whether Kupffer cells have different functions at different phases to ­promote or protect from liver injury in cholestasis. Increasing evidence shows that other immune cells also contribute to cholestatic liver injury. T cell infiltration in the liver has been detected in both mice and rats subjected to BDL as well as in cholestatic Mdr2−/− mice [58, 60]. IL‐17, a proinflammatory and fibrogenic cytokine secreted mainly by TH17 cells, has been shown to participate in cholestasis. Expansion of IL‐17 positive cells or TH17 cells were identified in the livers of patients with PBC and of mice after BDL [86–88]. Significantly higher levels of serum IL‐17 were also reported in pregnant women with intrahepatic cholestasis of pregnancy and in mice after BDL [89]. IL‐17 promotes hepatic inflammation during cholestasis by enhancing bile acid induced expression of proinflammatory cytokines in hepatocytes. Neutralization of IL‐17 with anti‐IL17 antibody significantly reduced BDL‐induced liver necrosis, proinflammatory cytokine production, and



56:  The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury

neutrophil infiltration in mice 9 or 14 days after BDL [86, 90]. In contrast, hepatic natural killer cells and invariant natural killer T cells have been shown to suppress cholestatic liver injury by stimulating anti‐inflammatory or repressing proinflammatory cytokines production in Kupffer cells [91, 92]. Recently another type of innate immune cells, mast cells, have also been implicated in cholestasis. Both PSC patients and Mdr2−/− mice have increased biliary expression of histamine receptors. In mast cell‐deficient mice subjected to BDL or in Mdr2−/− mice treated with inhibitors of histamine release or ­histamine receptors, attenuated liver damage, biliary proliferation, and fibrosis were observed compared with wild‐type or untreated mice, suggesting that mast cells and histamine could  be novel therapeutic targets in cholestatic liver diseases [93–95].

THE ROLE OF THE INFLAMMASOME IN CHOLESTATIC LIVER INJURY Inflammasomes are cytosolic multiprotein complexes that detect signals from injured cells and pathogens known as DAMPs and PAMPs leading to the secretion of IL‐1 family members. These complexes consist of NLRP3 protein, apoptosis‐associated speck‐like protein containing a caspase‐activating recruitment domain (ASC) and caspase 1. Upon stimulation, they assemble to activate caspase‐1 which then proteolytically activates cytokine IL‐1β and IL‐18. IL‐1β then amplifies the inflammatory response by further stimulating production of inflammatory cytokines. Activation of inflammasomes has been seen primarily in alcoholic hepatitis, NASH, chronic HCV, ischemia‐reperfusion injury, and acetaminophen toxicity [96]. However, the role of inflammasomes in the pathogenesis of cholestatic liver injury is less clear. Proteins that make up the inflammasome are prominently expressed in Kupffer cells and in liver sinusoidal endothelial cells but are essentially absent from normal hepatic parenchymal cells. When mouse hepatocytes were treated with a pan‐caspase inhibitor z‐VAD‐FMK, bile acid induction of proinflammatory cytokines was not repressed [24]. However, cholestatic liver injury was altered in caspase 1 knockout mice after 7 days of BDL as evidenced by reduced plasma liver enzymes but increased liver fibrosis (manuscript in preparation), suggesting that the inflammasome plays a role in the pathogenesis of cholestatic liver injury. Cholestasis is commonly associated with sepsis, and LPS is a major activator of the inflammasome in Kupffer cells. Recent studies suggest that pathophysiologic levels of bile acids can inhibit NLRP3 inflammasome activation in isolated macrophages via the TGR5–cAMP–PKA axis and phosphorylation of NLRP3 on Ser 291. When mice are subjected to LPS‐induced sepsis or alum induced peritoneal inflammation and treated with TLCA or the bile acid receptor TGR5 agonist INT‐777, IL‐1β and IL‐18 were significantly reduced in Nlrp3 wild‐type mice but not in Nlrp3−/− mice. These findings suggest that bile acids significantly constrain NLRP3 inflammasome‐related inflammation  [97], as suggested by several earlier studies [82, 98]. However completely opposite effects have been seen with chenodeoxycholic acid, which is reported to activate the NLRP3

733

inflammasome in LPS sensitized isolated macrophages and Kupffer cells [99]. It needs to be stressed that high non‐pathophysiologic concentrations of this unconjugated bile acid were used in these in vitro assays in this study. Such high concentrations of bile acids are never seen in serum or liver in cholestatic animal models or in humans with cholestasis [9]. A more recent study also demonstrated that deoxycholic acid and chenodeoxycholic acid activated both signal 1 and 2 of the NLRP3 inflammasome in a mouse model of LPS‐induced sepsis but also used concentrations of bile acids above pathophysiologic levels [100]. Physiologically relevant levels of endogenous bile acids such as taurocholic acid did not produce these effects in mouse macrophages as we have recently confirmed in the non‐parenchymal cell fraction from mouse livers [24]. Very little is known about the role of inflammasome in human cholestatic liver disease. Tian and colleagues showed that Galectin 3, NLRP3, and the adaptor protein ASC and the downstream activation of caspasae‐1 and IL‐1β were all upregulated in liver tissue from patients with primary biliary cholangitis. Galectin‐3 (Gal3), is a pleiotropic lectin produced by monocytes and macrophages and was thought to be a key activator of the inflammasome and result in an IL‐17 immune response characteristic of the inflammatory response in PBC [101]. However, much further study is needed before the role of the inflammasome in cholestatic liver injury in animal models and in patients with cholestasis can be clarified.

CONCLUSION In the past few years, the hypothetical mechanism of bile acid‐ induced liver injury has evolved from direct hepatic cytotoxicity to hepatocyte apoptosis and now to inflammation‐mediated injury. As seen in almost all diseases, the immune response plays a very important role in their development and progression. This response is an attempt to resolve the initial insult(s) to the tissue, for example, pathogen invasion, homeostasis change,

Figure 56.1  A schematic view of bile acid exacerbated cholestatic liver injury. Regardless of the initial insult(s) to the liver, once bile ­formation is impaired, bile acids (BA) accumulate in the liver. This ­triggers an inflammatory response and results in further hepatic injury. The initial insults include genetic and developmental defects, as well the effects of drugs, viral hepatitis, alcoholic liver disease, metabolic syndrome, pregnancy, primary biliary cholangitis (PBC), primary sclerosing cholangitis (PSC), and bile duct obstruction from gallstones or tumors.

734

THE LIVER:  REFERENCES

Table 56.1  Bile acids stimulate variable responses in different liver cell types Cell type

Stimulus/ treatment

Response

Mediator/pathway

References

Mouse hepatocytes

TCA (200 μM), DCA (200 μM), CDCA (200 μM), bile Major endogenous bile acids (25–200 μM) BDL model

Induction of proinflammatory cytokines

MAPKs/Egr1

[20, 21]

Release of proinflammatory cytokines, neutrophil chemotaxis Secretion of osteopontin for immune cell recruitment Cell proliferation Activation of COX‐2

ER stress, mitochondrial damage, Tlr9 activation Excessive pressure in the biliary system/MMPs TGR5‐ROS/cSrc‐EGFR‐ MEK‐ERK1/2 S1PR2/ERK1/2/NF‐kB DAMPs CXC/CCL chemokines Adhesion molecules Cytoskeletal proteins TGR5/NF‐kB/JNK/ Inflammasome

[24]

Mouse and human hepatocytes Mouse cholangiocytes

TLCA (25μM) TCA (100μM) Mouse and human neutrophils

BDL model

Activation, chemotaxis, and cytotoxicity

Mouse, rabbit, and human monocytes/ macrophages

BDL model TCA, TCDCA, GCDCA, TLCA, CDCA (10–50 μM) BDL model

Production of pro‐ and/or anti‐inflammatory cytokines

Mouse TH cells Mouse NK and invariant NK T cells

BDL model

Mouse mast cells

BDL & Mdr2−/− mouse models

Production of pro‐inflammatory and fibrotic cytokine IL‐17 Stimulation of anti‐ or suppression of pro‐inflammatory cytokines produced in Kupffer cells Production of histamine that induces focal necrosis, biliary hyperplasia, and fibrosis

and so on. However, when unresolved, the immune response will persist and cause progressive harm to the tissue. In this chapter, we reviewed the role of inflammation in cholestatic liver injury. As illustrated in Figure 56.1, regardless of the initial cause of the hepatic injury, once bile flow is compromised, bile acids then accumulate in hepatocytes. In hepatocytes, high levels of bile acids induce the release of proinflammatory mediators (Table  56.1), for example, chemokines that trigger the inflammatory response, including neutrophil activation and hepatic infiltration. Recruited neutrophils kill the stressed hepatocytes and further stimulate inflammatory response from other liver cells and immune cells, such as cholangiocytes, Kupffer cells, T cells, and mast cells. In addition, necrotic hepatic cells may also release DAMPs to further inflame the liver [102, 103]. When the initial insult(s) is not resolved or bile acid homeostasis is not reestablished, cholestatic liver injury will persist and ultimately progress. This scheme provides a current mechanistic explanation for cholestatic liver injury under pathophysiological conditions, although further details remain to be elucidated.

REFERENCES 1. Hofmann, A.F. and Hagey, L.R. Bile acids: chemistry, pathochemistry, biology, pathobiology, and therapeutics. Cell Mol Life Sci, 2008;65(16):2461–23. 2. Boyer, J.L. Bile formation and secretion. Compr Physiol, 2013; 3(3):1035–78. 3. Hirschfield, G.M. Genetic determinants of cholestasis. Clin Liver Dis, 2013; 17(2):147–59. 4. Keitel, V., Droge, C., Stepanow, S. et al. Intrahepatic cholestasis of pregnancy (ICP): case report and review of the literature. Z Gastroenterol, 2016; 54(12):1327–33. 5. Pollheimer, M.J., Fickert, P., and Stieger, B. Chronic cholestatic liver ­diseases: clues from histopathology for pathogenesis. Mol Aspects Med, 2014; 37:35–56.

[45, 48, 51]

[24, 57, 58, 62, 64–66] [46, 76, 78, 80–82, 98, 104] [86, 87] [91, 92] [93–95]

6. Scholmerich, J., Becher, M.S., Schmidt, K. et al. Influence of hydroxylation and conjugation of bile salts on their membrane‐damaging properties  – ­studies on isolated hepatocytes and lipid membrane vesicles. Hepatology, 1984;4(4):661–6. 7. Attili, A.F., Angelico, M., Cantafora, A., Alvaro, D., and Capocaccia, L. Bile acid‐induced liver toxicity: relation to the hydrophobic‐hydrophilic balance of bile acids. Med Hypotheses, 1986;19(1):57–69. 8. Galle, P.R., Theilmann, L., Raedsch, R., Otto, G., and Stiehl, A. Ursodeoxycholate reduces hepatotoxicity of bile salts in primary human hepatocytes. Hepatology, 1990;12(3 Pt 1):486–91. 9. Woolbright, B.L. and Jaeschke, H. Novel insight into mechanisms of cholestatic liver injury. World J Gastroenterol, 2012;18(36):4985–93. 10. Kwo, P., Patel, T., Bronk, S.F., and Gores, G.J. Nuclear serine protease activity contributes to bile acid‐induced apoptosis in hepatocytes. Am J Physiol, 1995;268(4 Pt 1):G613–21. 11. Patel, T., Bronk, S.F., and Gores, G.J. Increases of intracellular magnesium promote glycodeoxycholate‐induced apoptosis in rat hepatocytes. J Clin Invest, 1994;94(6):2183–92. 12. Roberts, L.R., Kurosawa, H., Bronk, S.F. et  al. Cathepsin B contributes to  bile salt‐induced apoptosis of rat hepatocytes. Gastroenterology, 1997;113(5):1714–26. 13. Takikawa, Y., Miyoshi, H., Rust, C. et al. The bile acid‐activated phosphatidylinositol 3‐kinase pathway inhibits Fas apoptosis upstream of bid in rodent hepatocytes. Gastroenterology, 2001;120(7):1810–7. 14. Usechak, P., Gates, A., and Webster, C.R. Activation of focal adhesion kinase and JNK contributes to the extracellular matrix and cAMP‐GEF mediated survival from bile acid induced apoptosis in rat hepatocytes. J Hepatol, 2008;49(2):251–61. 15. Webster, C.R., Usechak, P., and Anwer, M.S. cAMP inhibits bile acid‐ induced apoptosis by blocking caspase activation and cytochrome c release. Am J Physiol Gastrointest Liver Physiol, 2002;283(3):G727–38. 16. Webster, C.R. and Anwer, M.S. Cyclic adenosine monophosphate‐mediated protection against bile acid‐induced apoptosis in cultured rat hepatocytes. Hepatology, 1998;27(5):1324–31. 17. Canbay, A., Feldstein, A.E., Higuchi, H. et  al. Kupffer cell engulfment of apoptotic bodies stimulates death ligand and cytokine expression. Hepatology, 2003;38(5):1188–98. 18. Malhi, H., Guicciardi, M.E., and Gores, G.J. Hepatocyte death: a clear and present danger. Physiol Rev, 2010;90(3):1165–94. 19. Kinugasa, T., Uchida, K., Kadowaki, M., Takase, H., Nomura, Y., and Saito, Y. Effect of bile duct ligation on bile acid metabolism in rats. J Lipid Res, 1981;22(2):201–7.



56:  The Role of Bile Acid‐Mediated Inflammation in Cholestatic Liver Injury

20. Allen, K., Jaeschke, H., and Copple, B.L. Bile acids induce inflammatory genes in hepatocytes: a novel mechanism of inflammation during obstructive cholestasis. Am J Pathol, 2011;178(1):175–86. 21. Zhang, Y., Hong, J.Y., Rockwell, C.E., Copple, B.L., Jaeschke, H., and Klaassen, C.D. Effect of bile duct ligation on bile acid composition in mouse serum and liver. Liver Int, 2012;32(1):58–69. 22. Woolbright, B.L., Dorko, K., Antoine, D.J. et al. Bile acid‐induced necrosis in primary human hepatocytes and in patients with obstructive cholestasis. Toxicol Appl Pharmacol, 2015;283(3):168–77. 23. Woolbright, B.L., Antoine, D.J., Jenkins, R.E., Bajt, M.L., Park, B.K., and Jaeschke, H. Plasma biomarkers of liver injury and inflammation demonstrate a lack of apoptosis during obstructive cholestasis in mice. Toxicol Appl Pharmacol, 2013;273(3):524–31. 24. Cai, S.Y., Ouyang, X., Chen, Y. et  al. Bile acids initiate cholestatic liver injury by triggering a hepatocyte‐specific inflammatory response. JCI Insight, 2017;2(5):e90780. 25. Slijepcevic, D., Kaufman, C., Wichers, C.G. et al. Impaired uptake of conjugated bile acids and hepatitis b virus pres1‐binding in na(+) ‐taurocholate cotransporting polypeptide knockout mice. Hepatology, 2015; 62(1):207–19. 26. Vaz, F.M., Paulusma, C.C., Huidekoper, H. et  al. Sodium taurocholate cotransporting polypeptide (SLC10A1) deficiency: conjugated hypercholanemia without a clear clinical phenotype. Hepatology, 2015; 61(1):260–7. 27. Kubitz, R., Droge, C., Stindt, J., Weissenberger, K., and Haussinger, D. The bile salt export pump (BSEP) in health and disease. Clin Res Hepatol Gastroenterol, 2012;36(6):536–53. 28. Kim, N.D., Moon, J.O., Slitt, A.L., and Copple, B.L. Early growth response  factor‐1 is critical for cholestatic liver injury. Toxicol Sci, 2006;90(2):586–95. 29. Allen, K., Kim, N.D., Moon, J.O., and Copple, B.L. Upregulation of early growth response factor‐1 by bile acids requires mitogen‐activated protein kinase signaling. Toxicol Appl Pharmacol, 2010;243(1):63–7. 30. Spivey, J.R., Bronk, S.F., and Gores, G.J. Glycochenodeoxycholate‐induced lethal hepatocellular injury in rat hepatocytes. Role of ATP depletion and cytosolic free calcium. J Clin Invest, 1993;92(1):17–24. 31. Denk, G.U., Kleiss, C.P., Wimmer, R. et  al. Tauro‐beta‐muricholic acid restricts bile acid‐induced hepatocellular apoptosis by preserving the mitochondrial membrane potential. Biochem Biophys Res Commun, 2012;424(4):758–64. 32. Gabele, E., Muhlbauer, M., Dorn, C. et al. Role of TLR9 in hepatic stellate cells and experimental liver fibrosis. Biochem Biophys Res Commun, 2008;376(2):271–6. 33. Carpino, G., Cardinale, V., Onori, P. et al. Biliary tree stem/progenitor cells in glands of extrahepatic and intraheptic bile ducts: an anatomical in situ  study yielding evidence of maturational lineages. J Anat, 2012; 220(2):186–99. 34. Beuers, U., Maroni, L., and Elferink, R.O. The biliary HCO(3)(‐) umbrella: experimental evidence revisited. Curr Opin Gastroenterol, 2012; 28(3):253–7. 35. Hohenester, S., Wenniger, L.M., Paulusma, C.C. et al. A biliary. Hepatology, 2012;55(1):173–83. 36. Maillette de Buy Wenniger, L.J., Hohenester, S., Maroni, L., van Vliet, S.J., Oude Elferink, R.P., and Beuers, U. The cholangiocyte glycocalyx stabilizes the “biliary hco3 umbrella”: an integrated line of defense against toxic bile acids. Dig Dis, 2015;33(3):397–407. 37. Prieto, J., Qian, C., Garcia, N., Diez, J., and Medina, J.F. Abnormal expression of anion exchanger genes in primary biliary cirrhosis. Gastroenterology, 1993;105(2):572–8. 38. Medina, J.F., Martinez, A., Vazquez, J.J., and Prieto, J. Decreased anion exchanger 2 immunoreactivity in the liver of patients with primary biliary cirrhosis. Hepatology, 1997;25(1):12–7. 39. Salas, J.T., Banales, J.M., Sarvide, S. et al. Ae2a,b‐deficient mice develop antimitochondrial antibodies and other features resembling primary biliary cirrhosis. Gastroenterology, 2008;134(5):1482–93. 40. Nagahashi, M., Yuza, K., Hirose, Y. et al. The roles of bile acids and sphingosine‐1‐phosphate signaling in the hepatobiliary diseases. J Lipid Res, 2016;57(9):1636–43. 41. Deutschmann, K., Reich, M., Klindt, C. et  al. Bile acid receptors in the biliary tree: TGR5 in physiology and disease. Biochim Biophys Acta, ­ 2018;1864(4 Pt B):1319–25.

735

42. Keitel, V. and Haussinger, D. Perspective: TGR5 (Gpbar‐1) in liver physiology and disease. Clin Res Hepatol Gastroenterol, 2012;36(5):412–9. 43. Hov, J.R., Keitel, V., Schrumpf, E., Haussinger, D., and Karlsen, T.H. TGR5 sequence variation in primary sclerosing cholangitis. Dig Dis, 2011; 29(1):78–84. 44. Masyuk, A.I., Huang, B.Q., Radtke, B.N. et al. Ciliary subcellular localization of TGR5 determines the cholangiocyte functional response to bile acid signaling. Am J Physiol Gastrointest Liver Physiol, 2013;304(11):G1013–24. 45. Reich, M., Deutschmann, K., Sommerfeld, A. et al. TGR5 is essential for bile acid‐dependent cholangiocyte proliferation in vivo and in vitro. Gut, 2016;65(3):487–501. 46. Pean, N., Doignon, I., Garcin, I. et al. The receptor TGR5 protects the liver from bile acid overload during liver regeneration in mice. Hepatology, 2013;58(4):1451–60. 47. Baghdasaryan, A., Claudel, T., Gumhold, J. et al. Dual farnesoid X receptor/ TGR5 agonist INT‐767 reduces liver injury in the Mdr2‐/‐ (Abcb4‐/‐) mouse cholangiopathy model by promoting biliary HCO(‐)(3) output. Hepatology, 2011;54(4):1303–12. 48. Wang, Y., Aoki, H., Yang, J. et al. The role of sphingosine 1‐phosphate receptor 2 in bile‐acid‐induced cholangiocyte proliferation and cholestasis‐ induced liver injury in mice. Hepatology, 2017;65(6):2005–18. 49. Pinto, C., Giordano, D.M., Maroni, L., and Marzioni, M. Role of inflammation and proinflammatory cytokines in cholangiocyte pathophysiology. Biochim Biophys Acta, 2018;1864(4 Pt B):1270–8. 50. Wen, Y., Jeong, S., Xia, Q., and Kong, X. Role of osteopontin in liver diseases. Int J Biol Sci, 2016;12(9):1121–8. 51. Yang, M., Ramachandran, A., Yan, H.M. et al. Osteopontin is an initial mediator of inflammation and liver injury during obstructive cholestasis after bile duct ligation in mice. Toxicol Lett, 2014;224(2):186–95. 52. Zweers, S.J., Shiryaev, A., Komuta, M. et al. Elevated interleukin‐8 in bile of patients with primary sclerosing cholangitis. Liver Int, 2016;36(9):1370–7. 53. Heymann, F. and Tacke, F. Immunology in the liver – from homeostasis to disease. Nat Rev Gastroenterol Hepatol, 2016;13(2):88–110. 54. Kubes, P. and Mehal, W.Z. Sterile inflammation in the liver. Gastroenterology, 2012;143(5):1158–72. 55. Xu, R., Huang, H., Zhang, Z., and Wang, F.S. The role of neutrophils in the development of liver diseases. Cell Mol Immunol, 2014;11(3):224–31. 56. Marques, P.E., Amaral, S.S., Pires, D.A. et al. Chemokines and mitochondrial products activate neutrophils to amplify organ injury during mouse acute liver failure. Hepatology, 2012;56(5):1971–82. 57. Gujral, J.S., Farhood, A., Bajt, M.L., and Jaeschke, H. Neutrophils aggravate acute liver injury during obstructive cholestasis in bile duct‐ligated mice. Hepatology, 2003;38(2):355–63. 58. Georgiev, P., Jochum, W., Heinrich, S. et al. Characterization of time‐related changes after experimental bile duct ligation. Br J Surg, 2008;95(5):646–56. 59. Demirbilek, S., Akin, M., Gurunluoglu, K. et al. The NF‐kappaB inhibitors attenuate hepatic injury in bile duct ligated rats. Pediatr Surg Int, 2006;22(8):655–63. 60. Yu, D., Cai, S.Y., Mennone, A., Vig, P., and Boyer, J.L. Cenicriviroc, a cytokine receptor antagonist, potentiates all‐trans retinoic acid in reducing liver injury in cholestatic rodents. Liver Int, 2018;38(6):1128–38. 61. Kodali, P., Wu, P., Lahiji, P.A., Brown, E.J., and Maher, J.J. ANIT toxicity toward mouse hepatocytes in vivo is mediated primarily by neutrophils via CD18. Am J Physiol Gastrointest Liver Physiol, 2006;291(2):G355–63. 62. Cai, S.Y. and Boyer, J.L. The role of inflammation in the mechanisms of bile acid‐induced liver damage. Dig Dis, 2017;35(3):232–4. 63. Futosi, K., Fodor, S., and Mocsai, A. Reprint of neutrophil cell surface receptors and their intracellular signal transduction pathways. ­ Int Immunopharmacol, 2013;17(4):1185–97. 64. Gulubova, M., Vlaykova, T., Manolova, I., Hadjipetkov, P., and Popharitov, A. Implication of adhesion molecules in inflammation of the common bile duct in patients with secondary cholangitis due to biliary obstruction. Hepatogastroenterology, 2008;55(84):836–41. 65. Gujral, J.S., Liu, J., Farhood, A., Hinson, J.A., and Jaeschke, H. Functional importance of ICAM‐1 in the mechanism of neutrophil‐induced liver injury in bile duct‐ligated mice. Am J Physiol Gastrointest Liver Physiol, 2004;286(3):G499–G507. 66. Li, M., Mennone, A., Soroka, C.J. et al. Na(+) /H(+) exchanger regulatory factor 1 knockout mice have an attenuated hepatic inflammatory response  and are protected from cholestatic liver injury. Hepatology, 2015;62(4):1227–36.

736

THE LIVER:  REFERENCES

67. Li, M., Wang, W., Soroka, C.J. et  al. NHERF‐1 binds to Mrp2 and regulates  hepatic Mrp2 expression and function. J Biol Chem, 2010; ­ 285(25):19299–307. 68. Fouassier, L., Duan, C.Y., Feranchak, A.P. et  al. Ezrin‐radixin‐moesin‐­ binding phosphoprotein 50 is expressed at the apical membrane of rat liver epithelia. Hepatology, 2001;33(1):166–76. 69. Wang, W., Soroka, C.J., Mennone, A. et al. Radixin is required to maintain apical canalicular membrane structure and function in rat hepatocytes. Gastroenterology, 2006;131(3):878–84. 70. Barreiro, O., Yanez‐Mo, M., Serrador, J.M. et  al. Dynamic interaction of VCAM‐1 and ICAM‐1 with moesin and ezrin in a novel endothelial docking structure for adherent leukocytes. J Cell Biol, 2002;157(7):1233–45. 71. Carman, C.V., Jun, C.D., Salas, A., and Springer, T.A. Endothelial cells ­proactively form microvilli‐like membrane projections upon intercellular adhesion molecule 1 engagement of leukocyte LFA‐1. J Immunol, 2003;171(11):6135–44. 72. Tacke, F. Targeting hepatic macrophages to treat liver diseases. J Hepatol, 2017;66(6):1300–12. 73. Calmus, Y. and Poupon, R. Shaping macrophages function and innate immunity by bile acids: mechanisms and implication in cholestatic liver diseases. Clin Res Hepatol Gastroenterol, 2014;38(5):550–6. 74. Ju, C. and Tacke, F. Hepatic macrophages in homeostasis and liver diseases: from pathogenesis to novel therapeutic strategies. Cell Mol Immunol, 2016;13(3):316–27. 75. Guicciardi, M.E., Trussoni, C.E., Krishnan, A. et al. macrophages contribute to the pathogenesis of sclerosing cholangitis in mice. J Hepatol, 2018;69(3):676–86. 76. Abe, T., Arai, T., Ogawa, A. et  al. Kupffer cell‐derived interleukin 10 is responsible for impaired bacterial clearance in bile duct‐ligated mice. Hepatology, 2004;40(2):414–23. 77. Chowdhury, A.H., Camara, M., Martinez‐Pomares, L. et  al. Immune dysfunction in patients with obstructive jaundice before and after endoscopic retrograde cholangiopancreatography. Clin Sci, 2016;130(17):1535–44. 78. Perino, A. and Schoonjans, K. TGR5 and immunometabolism: insights from physiology and pharmacology. Trends Pharmacol Sci, 2015;36(12):847–57. 79. Guo, C., Chen, W.D., and Wang, Y.D. TGR5, not only a metabolic regulator. Front Physiol, 2016;7:646. 80. Haselow, K., Bode, J.G., Wammers, M. et al. Bile acids PKA‐dependently induce a switch of the IL‐10/IL‐12 ratio and reduce proinflammatory ­capability of human macrophages. J Leukoc Biol, 2013;94(6):1253–64. 81. Wammers, M., Schupp, A.K., Bode, J.G. et  al. Reprogramming of pro‐ inflammatory human macrophages to an anti‐inflammatory phenotype by bile acids. Sci Rep, 2018;8(1):255. 82. Keitel, V., Donner, M., Winandy, S., Kubitz, R., and Haussinger, D. Expression and function of the bile acid receptor TGR5 in Kupffer cells. Biochem Biophys Res Commun, 2008;372(1):78–84. 83. Zandieh, A., Payabvash, S., Pasalar, P. et al. Gadolinium chloride, a Kupffer cell inhibitor, attenuates hepatic injury in a rat model of chronic cholestasis. Hum Exp Toxicol, 2011;30(11):1804–10. 84. Gehring, S., Dickson, E.M., San Martin, M.E. et al. Kupffer cells abrogate cholestatic liver injury in mice. Gastroenterology, 2006;130(3):810–22. 85. Osawa, Y., Seki, E., Adachi, M. et  al. Role of acid sphingomyelinase of  Kupffer cells in cholestatic liver injury in mice. Hepatology, 2010;51(1):237–45.

86. O’Brien, K.M., Allen, K.M., Rockwell, C.E., Towery, K., Luyendyk, J.P., and Copple, B.L. IL‐17A synergistically enhances bile acid‐induced inflammation during obstructive cholestasis. Am J Pathol, 2013; ­ 183(5):1498–1507. 87. Licata, L.A., Nguyen, C.T., Burga, R.A. et al. Biliary obstruction results in PD‐1‐dependent liver T cell dysfunction and acute inflammation mediated by Th17 cells and neutrophils. J Leukoc Biol, 2013;94(4):813–23. 88. Lan, R.Y., Salunga, T.L., Tsuneyama, K. et  al. Hepatic IL‐17 responses in  human and murine primary biliary cirrhosis. J Autoimmun, 2009;32(1):43–51. 89. Kirbas, A., Biberoglu, E., Ersoy, A.O. et al. The role of interleukin‐17 in intrahepatic cholestasis of pregnancy. J Matern Fetal Neonatal Med, 2016;29(6):977–81. 90. Zhang, S., Huang, D., Weng, J. et al. Neutralization of interleukin‐17 attenuates cholestatic liver fibrosis in mice. Scand J Immunol, 2016;83(2):102–8. 91. Cheng, C.W., Duwaerts, C.C., Rooijen, N., Wintermeyer, P., Mott, S., and Gregory, S.H. NK cells suppress experimental cholestatic liver injury by an interleukin‐6‐mediated, Kupffer cell‐dependent mechanism. J Hepatol, 2011;54(4):746–52. 92. Duwaerts, C.C., Sun, E.P., Cheng, C.W., van R.N., and Gregory, S.H. Cross‐activating invariant NKT cells and kupffer cells suppress cholestatic liver injury in a mouse model of biliary obstruction. PLoS One, 2013;8(11):e79702. 93. Jones, H., Hargrove, L., Kennedy, L. et al. Inhibition of mast cell‐secreted histamine decreases biliary proliferation and fibrosis in primary sclerosing cholangitis Mdr2(‐/‐) mice. Hepatology, 2016;64(4):1202–16. 94. Hargrove, L., Kennedy, L., Demieville, J. et al. Bile duct ligation‐induced biliary hyperplasia, hepatic injury, and fibrosis are reduced in mast cell‐­ deficient Kit(W‐sh) mice. Hepatology, 2017;65(6):1991–2004. 95. Kennedy, L., Hargrove, L., Demieville, J. et al. Blocking H1/H2 histamine receptors inhibits damage/fibrosis in Mdr2(‐/‐) mice and human cholangiocarcinoma tumorigenesis. Hepatology, 2018’68(3):1042–56. 96. Szabo, G. and Petrasek, J. Inflammasome activation and function in liver disease. Nat Rev Gastroenterol Hepatol, 2015;12(7):387–400. 97. Guo, C., Xie, S., Chi, Z. et al. Bile acids control inflammation and metabolic disorder through inhibition of NLRP3 inflammasome. Immunity, 2016;45(4):802–16. 98. Kawamata, Y., Fujii, R., Hosoya, M. et  al. A G protein‐coupled receptor responsive to bile acids. J Biol Chem, 2003;278(11):9435–40. 99. Gong, Z., Zhou, J., Zhao, S. et al. Chenodeoxycholic acid activates NLRP3 inflammasome and contributes to cholestatic liver fibrosis. Oncotarget, 2016;7(51):83951–63. 100. Hao, H., Cao, L., Jiang, C. et  al. Farnesoid X receptor regulation of the NLRP3 inflammasome underlies cholestasis‐associated sepsis. Cell Metab, 2017;25(4):856–67. 101. Tian, J., Yang, G., Chen, H.Y. et  al. Galectin‐3 regulates inflammasome activation in cholestatic liver injury. FASEB J, 2016;30(12):4202–13. 102. Woolbright, B.L. and Jaeschke, H. Therapeutic targets for cholestatic liver injury. Expert Opin Ther Targets, 2016;20(4):463–75. 103. Woolbright, B.L. and Jaeschke, H. The impact of sterile inflammation in acute liver Injury. J Clin Transl Res, 2017;3(1):170–88. 104. Reich, M., Klindt, C., Deutschmann, K., Spomer, L., Haussinger, D., and Keitel, V. Role of the G Protein‐coupled bile acid receptor TGR5 in liver damage. Dig Dis, 2017;35(3):235–40.

57

Toll‐like Receptors in Liver Disease So Yeon Kim1 and Ekihiro Seki1,2 Division of Digestive and Liver Diseases, Department of Medicine, Cedars‐Sinai Medical Center, Los Angeles, CA, USA 2 Department of Biomedical Sciences, Cedars‐Sinai Medical Center, Los Angeles, CA, USA 1

INTRODUCTION Toll‐like receptors (TLRs) are the evolutionary conserved pattern recognition receptors (PRRs) that sense microbe‐derived pathogen‐associated molecular patterns (PAMPs) (e.g. lipopolysaccharide [LPS)]) and damage‐associated molecular patterns (DAMPs) from damaged host tissues. Toll was originally ­identified in Drosophila as the regulatory gene for dorsoventral patterning during embryonic development [1]. Toll was then determined to be a key molecule for antifungal innate immunity in Drosophila. Subsequently human Toll was identified and is now recognized as toll‐like receptor 4 (TLR4) [1]. TLR4 was discovered as a receptor for LPS, (aka endotoxin), a Gram‐­ negative bacterial cell wall component, through the identification of the P712H mutation of TLR4 cytoplasmic domain in the LPS unresponsive C3H/HeJ mouse strain [1]. To date, more than 10 TLR family members have been identified in mammals (10 TLRs in humans and 12 TLRs in laboratory mice) [1]. All TLRs contain leucine‐rich repeats in the extracellular domain, which are crucial for sensing specific molecular patterns. By the binding of corresponding ligands to TLRs, their downstream signaling is activated and induces the innate immune response, including production of inflammatory cytokines and reactive oxygen species (ROS). TLRs also play a role in bridging the innate and adaptive immunity. Many studies for TLRs in infectious diseases were conducted due to the nature of TLRs for sensing microbial products. Liver and gut are anatomically ­connected by the portal vein. When the intestinal barrier is disrupted, the functions lead to increased intestinal permeability, translocating gut‐derived LPS to the liver, where TLR4 is ­activated. As such, TLRs play a major role in liver disease ­progression. Additionally, damaged liver cell‐derived DAMPs participate in the progression of non‐infectious inflammatory

disease through the activation of TLRs. This chapter will discuss the functions and downstream signaling of TLRs, the role of TLR signaling in various liver disease, and the gut–liver axis in liver diseases.

TLR LIGANDS, RECEPTORS, AND DOWNSTREAM SIGNALING Among TLRs, TLR1, 2, 4, 5, and 6 are expressed on the cell surface and TLR3, 7/8, and 9 are present within endosomes (Figure 57.1) [1]. TLR1, 2, 4, 5, and 6 extracellularly bind to bacterial products and TLR3, 7/8, and 9 are the intracellular sensors for nucleotides. In conjunction with co‐receptors CD14 and MD2, TLR4 recognizes LPS, a Gram‐negative bacterial cell wall component, also known as endotoxin (Table  57.1) [1]. TLR2 is mainly activated by Gram‐positive bacterial components, such as bacterial peptidoglycan and lipoproteins. TLR2 heterodimerizing with TLR1 senses triacyl lipoprotein [1]. Heterodimerized TLR2 with TLR6 binds to diacyl lipoprotein [1]. TLR5 is activated by bacterial flagellin [1]. Intracellular TLR3 recognizes virus‐produced double‐stranded RNA and synthetic polyinosinic–polycytidylic acid (poly I:C) [1]. TLR7 was initially identified as the receptor for synthetic compounds imiquimod and R‐848 [1]. Later, single‐stranded RNA derived from viruses, such as influenza A virus, HIV, and hepatitis C virus (HCV) were determined as natural ligands for TLR7/8 [1]. TLR9 is the receptor for unmethylated CpG motif‐containing DNA [1]. Methylation on CpG motifs is common in bacteria, but not in vertebrates. This is the reason TLR9 can discriminate the pathogens from the host. These receptors are also activated by endogenous host‐derived molecules, which are released from

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

738

THE LIVER:  TLR LIGANDS, RECEPTORS, AND DOWNSTREAM SIGNALING

damaged tissues. TLR2 and TLR4 can be activated by nuclear protein high mobility group box‐1 (HMGB1), hyaluronan, and Hsp60 [2]. S100A8/A9 are also possible ligands for TLR4. Saturated fatty acid, palmitate can activate TLR4. A previous study reported that liver‐derived protein fetuin‐A bridges between palmitate and TLR4 and contributes to TLR4 activation [3]. Recent studies showed that palmitate can bind to TLR4 but cannot form the TLR4 homodimer that is required for activation of canonical TLR4‐mediated myeloid differentiation factor 88 (MyD88) and TRIF (Toll/IL‐1 receptor domain containing adaptor inducing interferon‐β)‐dependent pathways [4,5]. Instead, palmitate can induce NOX2‐dependent ROS production through the monomeric TLR4 [5]. TLR7 can interact with GU‐rich single‐stranded RNA, including microRNA (e.g. miR‐21, miR‐29a, Let‐7b) [6]. In addition to bacterial DNA,

intracellular TLR9 can recognize host‐derived mitochondrial DNA. Because the endosymbiotic hypothesis states that mitochondria originated from bacteria, mitochondrial DNA contains unmethylated CpG DNA similar to bacterial DNA [7]. TLR activation initiates activation of two major intracellular signaling pathways – MyD88 and TRIF‐dependent pathways. MyD88 is a common adaptor molecule for all TLRs, except for TLR3. Upon TLR activation, TLR intracellular domain recruits MyD88, forming the complex with IRAK1, IRAK4, TRAF6, and TAK1 (Figure  57.1) [1, 2]. The K63 polyubiquitination chain linked to TRAF6 binds to TAK1, which is mediated through TAB2. TAK1 is then phosphorylated. Phosphorylated TAK1 phosphorylates IKKα/IKKβ in the IKK complex ­comprising IKKα/IKKβ and NEMO. Subsequently, IκBα that is bound to cytosolic NF‐κB is phosphorylated, K48‐ubiquitinated, finally

Figure 57.1  Schematic overview of toll‐like receptor (TLR) signaling pathways. TLR1/2, TLR2/6, TLR4, and TLR5 are expressed on the cell surface and sense triacyl lipopeptides, diacyl lipopeptides, LPS and flagellin, respectively. TLR3, TLR7/8, and TLR9 are located in the endosome and sense dsRNA, ssRNA, and CpG‐DNA, respectively. All TLRs except for TLR3 utilize MyD88 to activate NF‐κB and p38/JNK. TIRAP is required for the TLR2 and TLR4‐MyD88 signaling. TLR3 uses TRIF to activate TBK1/IKKε. TLR4 requires both TRAM and TLR4 internalization for activation of TRIF‐dependent pathway. Activated TRIF‐dependent pathways activate IRF‐3 for IFN‐β induction. TLR7/8 and TLR9 require the complex of MyD88/IRAK1/IRF7/IKKα for induction of IFN‐α.



57:  Toll‐like Receptors in Liver Disease

739

Table 57.1  Toll‐like receptors (TLRs), ligands, endogenous ligands, and localization Receptors

Ligands (PAMPs)

Ligands (DAMPs, synthetic molecules)

Localization

TLR1 TLR2

Triacyl lipoprotein Bacterial peptidoglycan, lipopeptide, lipoprotein

Cell surface Cell surface

TLR3

Viral‐derived dsRNA

TLR4

LPS

β‐defensin‐3 HSP60, 70, Gp96 HMGB1, serum amyloid A Hyaluronic acid Antiphospholipid antibodies Host‐derived mRNA poly I:C HMGB1, fibronectin EDA Fibrinogen, HSP60,70,72 Gp96, S100A8, S100A9 Serum amyloid A Oxidized LDL Saturated fatty acids Hyaluronic acid fragments Heparan sulfate fragments Antiphospholipid antibodies Paclitaxel

TLR5 TLR6 TLR7/ TLR8 (human)

Bacterial flagellin Diacyl lipoprotein ssRNA (HCV, HIV, influenza)

TLR9

Unmethylated CpG‐DNA (bacteria, virus, protozoa)

degraded. Freed NF‐κB will translocate into the nucleus and induces transcription of inflammatory genes. In TLR2 and TLR4 activation, an adaptor TIRAP (Toll/IL‐1 receptor domain containing adaptor protein, aka Mal) bridges between TLR cytoplasmic domain and MyD88 [1]. IRF5 is involved in the activation of MyD88‐dependent pathway [2]. In dendritic cells, TLR7 and TLR9 signaling forms the complex IRAK1, TRAF6, IKKα, and MyD88, and induces interferon (IFN)α production through IRF7 [1, 2]. TLR3 and TLR4 can activate the TRIF‐dependent pathway. TLR4 and TRIF are linked by an adaptor TRAM (TRIF‐related adaptor molecule) [1]. The TRIF‐dependent pathway induces IFN‐β through IRF3 activation, which is important for viral immunity [2]. In TLR4 signaling, the activation of TRIF‐ dependent pathways requires the internalization of TLR4 [1]. The TRIF‐dependent pathway is also crucial for TLR4‐­mediated IRF1 transcription and activation, which is important for the activation of inflammasome and IL‐1β [8].

TLRS AND THE GUT–LIVER AXIS IN ALCOHOLIC LIVER DISEASE Among liver diseases, the relationship between TLR4 and the gut–liver axis has initially been identified in alcoholic liver disease (ALD). Ingested alcohol and/or systemic blood alcohol reach the intestine and damage the intestinal epithelial barriers, which increases intestinal permeability [9]. This is known as “leaky gut”. The leaky gut promotes translocation of gut bacteria‐derived LPS to the liver via the portal vein (Figure  57.2). Alcohol abuse also increases the total abundance of intestinal bacteria and changes their composition by decreasing beneficial bacteria (e.g. Lactobacillus, Bacteroides) [10–12]. Translocated bacterial LPS activates TLR4 in Kupffer cells to produce TNFα,

GU‐rich microRNA Imiquimod, resiquimod (R848) Mitochondrial DNA Self‐denatured nuclear DNA IgG‐chromatin complex

Endosome Cell surface Endosome

Cell surface Cell surface Endosome Endosome

IL‐1β, and ROS. These inflammatory mediators promote hepatic steatosis, hepatocyte death, inflammation, and hepatic stellate cell (HSC) activation in the liver, which results in alcohol‐ induced fatty liver, steatohepatitis, and fibrosis (Figure  57.2). Indeed, systemic endotoxin levels were increased in mice and  humans after binge alcohol ingestions [13]. Moreover, mice  deficient in TLR4, CD14, and LPS‐binding protein and mice with depletion of gut bacteria by orally administered non‐ absorbable antibiotics showed reduced ethanol‐induced liver injury and steatosis, indicating the pivotal role played by gut‐ derived LPS and hepatic TLR4 signaling in the development of ALD. TLR2 and TLR9 also play a role in ALD since mice deficient in TLR2 and TLR9 are resistant for ALD [14]. TLR9 could be activated by translocated bacterial DNA because bacterial DNA was found in the blood after binge alcohol drinking in humans [13]. Fungi and their products also play a role in ALD. Alcohol consumption increased intestinal fungi and their products β‐D‐glucan in blood [15]. Mice with antifungal treatment or with deletion of dectin‐1, a pattern recognition receptor for β‐D‐glucan, showed reduced alcohol‐induced steatosis and injury, indicating the detrimental effect of intestinal fungi in ALD development [15].

TLR4 SIGNALING‐MEDIATED HSC ACTIVATION AND LIVER FIBROSIS The single nucleotide polymorphism (SNP) analysis for HCV patients with cirrhosis revealed that the TLR4 SNPs (TLR4 D299G and T399I) were associated with a reduced degree of cirrhosis in HCV patients [16], suggesting the role of TLR4 signaling in fibrosis progression. In the liver, TLR4 signaling induces proinflammatory cytokine production in Kupffer cells

740

THE LIVER:  TLRS AND NASH DEVELOPMENT

Figure 57.2  Toll‐like receptor (TLR)‐mediated liver disease caused by DAMPs and PAMPs. High‐fat diet, excessive ethanol intake, acetaminophen, ischemia‐reperfusion, viral hepatitis, and carcinogens induce hepatocyte damage. Damaged hepatocytes release damaged‐associated molecular patterns (DAMPs), such as HMGB1, and extracellular vesicles (EVs) that contain mitochondrial DNA. A high‐fat diet and excessive alcohol affect the composition of intestinal microbiota and increase intestinal permeability by disrupting intestinal epithelial barrier functions. Intestine‐derived pathogen‐associated molecular patterns (PAMPs), such as lipopolysaccharide (LPS), translocate to the liver via portal veins. In the liver, translocated gut‐derived PAMPs and/or hepatocyte‐derived DAMPs activate neutrophils, Kupffer cells, and hepatic stellate cells (HSCs) through TLR2, 4, 9, and RAGE. These cells produce ROS and proinflammatory cytokines, such as TNF‐α, IL‐1, and chemokines, leading to hepatocyte damage and liver inflammation. Activated HSCs produce extracellular matrix collagens. These mediators promote various liver injury. In contrast, TLR3 and TLR7 negatively regulate liver disease progression.

and promotes inflammation. HSCs are the precursors of collagen‐producing hepatic myofibroblasts that express TLR4. In liver fibrosis, TLR4 signaling in HSC is relatively important for HSC activation and liver fibrosis progression compared to that of Kupffer cells [17]. However, Kupffer cells are still important as the major source of TGF‐β, a potent fibrogenic cytokine, in liver fibrosis. TLR4 signaling can enhance TGF‐β signaling activation and type I collagen production in HSCs by reducing the expression of BMP and activin membrane‐bound inhibitor (Bambi) and miR‐29 [17, 18]. Bambi is an endogenous inhibitor for TGF‐β receptor. Bambi expression is transcriptionally regulated by the NF‐κBp50 homodimer in conjunction with histone deacylase 1 (HDAC1) in HSCs [19]. miR‐29 is a negative regulator for various collagen genes. During liver fibrosis progression or in cirrhotic patients, plasma endotoxin and bacterial DNA levels were increased, suggesting the role of bacterial translocation in human liver fibrosis [20, 21]. Indeed, gut‐sterilized mice by oral administration of non‐absorbable antibiotics showed reduced liver fibrosis development [17]. It appears that TLR2 plays an important role in regulation of intestinal

epithelial tight junction integrity and intestinal permeability through production of TNFα in the development of liver fibrosis [22]. Although translocated LPS promotes liver fibrosis progression, germ‐free mice that ubiquitously do not have commensal bacteria displayed aggravated fibrosis development compared to mice with normal commensal bacteria [23]. This suggests that commensal bacterial contain beneficial bacteria that prevent fibrosis progression and fibrotic bacteria are increased in disease conditions.

TLRS AND NASH DEVELOPMENT Non‐alcoholic fatty liver disease (NAFLD) is a hepatic manifestation of metabolic syndrome. The spectrum of NAFLD ranges from simple steatosis (NAFL) to steatosis with hepatocyte ballooning, inflammation, and fibrosis as referred to as non‐alcoholic steatohepatitis (NASH). Because high‐caloric high‐fat



57:  Toll‐like Receptors in Liver Disease

diets are associated with the development of NAFLD, these diets induced changes in the composition of intestinal microbiome and intestinal epithelial barrier functions, which increased gut permeability, causing so‐called “metabolic endotoxemia” [24]. These chronically increased systemic LPS levels activate hepatic TLR4 signaling. Because free fatty acid (FFA) is an a­ ctivator of TLR4 signaling [5], hyperlipidemia should be associated with TLR4 activation. While TLR4 in Kupffer cells and hepatocytes are important for inflammatory cytokine production and steatosis during NAFLD progression [25], TLR4 in HSCs promotes HSC activation and fibrosis [25]. Both MyD88 and TRIF‐ dependent pathways promote hepatic steatosis, in which hepatocytes and Kupffer cells play a role [25,26]. Interestingly, these adaptor proteins show distinct functions in inflammation and fibrosis in HSCs during NASH [25]. MyD88 in HSCs contributes to inflammatory cytokine production and fibrogenic response whereas TRIF tends to inhibit HSC activation [25,26]. As such, TRIF−/− mice showed less hepatics steatosis but more fibrosis under choline‐deficient amino acid‐defined diet [25]. TLR2 also played a role in NASH development. In contrast to TLR4, TLR2 was relatively important in Kupffer cells [27]. Inflammasome is activated by TLR2 and palmitic acid in Kupffer cells, but not in HSCs during the NASH development.

HMGB1 AND LIVER DISEASES Non‐infectious sterile inflammation can also activate TLR signaling in the liver. TLR4 is activated by not only LPS, but also endogenous host‐derived molecules. HMGB1, a nuclear protein, is translocated from nucleus to cytosol and further released extracellularly when hepatocytes get damaged. In the field of liver research, HMGB1 was initially described as an activator for TLR4 in ischemia reperfusion (I/R) liver injury [28]. Deletion of TLR4 and inhibition of HMGB1 with neutralizing antibodies showed similar protective effects for I/R liver injury in mice [28], suggesting that HMGB1 mediates liver injury likely via TLR4. As an adaptor molecule for TLR4, TRIF is relatively more important than MyD88 in I/R liver injury [29]. The TRIF‐ dependent pathway promotes I/R liver injury by producing type I IFN through IRF3 [29, 30]. Both hepatocytes and liver macrophages contribute to TLR4‐mediated I/R liver injury [31]. HMGB1 release requires hepatocyte TLR4. However, not only TLR4, but RAGE (receptor for advanced glycation end product) – another receptor for HMGB1 – and TLR9 also ­participate in the progression of HMGB1‐mediated I/R liver injury [32, 33]. Acetaminophen (APAP) causes acute liver failure, which presents massive hepatocyte necrosis and neutrophil recruitment. This is another acute sterile liver injury. Following APAP intoxication, HMGB1 was seen in blood before ALT elevation [34]. Necrotic hepatocytes are the major source of HMGB1. Hepatocyte‐derived HMGB1 promoted APAP‐induced liver injury by recruiting neutrophils (Figure  57.2). Expression of RAGE, but not TLR4, in neutrophils is crucial for HMGB1‐ mediated APAP liver injury [35]. Neutralizing antibodies for HMGB1 and Glycyrrhizin, a natural sweetener extracted from Chinese herbal medicine that can inhibit HMGB1 activity, inhibited APAP‐induced liver injury [36, 37].

741

In the setting of chronic liver disease, HMGB1 has been shown to be relocated from nucleus to cytosol in hepatocytes in liver fibrosis and ALD in mice and humans [38–40]. The ­experiments using mice with deletion of HMGB1 in hepatocytes concluded that hepatocyte HMGB1 has a detrimental effect in both liver fibrosis and ALD. Taken together, in most cases of sterile liver injury and inflammation, released HMGB1 induces hepatocytotoxicity. It is conceivable that blocking the signaling activated by HMGB1 can reduce liver damage and inflammation caused by sterile injury.

MITOCHONDRIAL DNA‐MEDIATED LIVER DISEASES THROUGH TLR9 Besides bacterial DNA, mitochondrial DNA can be a ligand for TLR9. Elevated serum mitochondrial DNA content has been observed in various liver diseases, including NASH, ALD, and APAP‐induced acute liver injury, suggesting the role of TLR9 recognition of mitochondrial DNA in these liver diseases [41–43]. The studies further demonstrated an increase in mitochondrial DNA content within serum extracellular vesicles, including exosomes, in patients with NASH and heavy alcohol drinkers [41, 42]. These exosomal mitochondrial DNAs are derived from hepatocytes. In NASH progression, endosomal recognition of mitochondrial DNA by TLR9 activates Kupffer cells, inducing proinflammatory cytokine production and aggravating liver injury [41]. In human alcoholic hepatitis, the role of neutrophils is crucial. A mouse acute‐on‐ chronic ethanol feeding model demonstrated the significant accumulation of neutrophils to liver parenchyma and showed the critical role of neutrophils in disease progression. In APAP‐induced sterile liver injury, neutrophils and their derived oxidative stress also play a crucial role. This neutrophil recruitment to the liver is mitochondrial DNA‐dependent and TLR9‐ dependent [14, 42, 44].

ROLE OF TLR3 AND TLR7 IN LIVER DISEASE In contrast to most TLRs that facilitate liver diseases, TLR3 and TLR7 activation could prevent liver disease progression. TLR3 is a receptor for double‐stranded RNA or poly I:C. TLR3 activation leads to IFN‐γ secretion in NK cells and it induces HSC cell cycle arrest and apoptosis via activation of STAT1 [45]. TLR3 signaling‐mediated NK cell accumulation and activation attenuate liver fibrosis via incapacitating activated HSCs. However, in advanced stages of liver fibrosis and in ALD, TLR3 activation of NK cells cannot eliminate HSCs [46, 47]. Intriguingly, TLR3−/− mice are protective against liver fibrosis. TLR3 is not only important for NK cell‐killing of HSCs, but it also activates HSCs by inducing proinflammatory IL‐17A from hepatic γδ T cells in liver fibrosis [48]. TLR3‐mediated fibrogenesis may occur in advanced fibrosis and ALD. This may be an explanation of a paradoxical role for TLR3 between early and advanced fibrosis.

742

THE LIVER:  ROLE OF TLRS IN VIRAL HEPATITIS

By contrast, TLR7 has been reported as being protective against liver fibrosis and NAFLD in mice. TLR7 signaling induces production of type I IFNs (and proinflammatory cytokines through NF‐κB). The main producer of type I IFNs is hepatic plasmacytoid dendritic cells that highly express TLR7. TLR7‐mediated type I IFNs induce antifibrotic interleukin‐1 receptor antagonist (IL‐1ra) in Kupffer cells, suppressing HSC activation and fibrosis [49]. In NAFLD, TLR7 promotes hepatocyte autophagy and reduces insulin‐like growth factor 1 levels through its secretion from hepatocytes, which suppresses NAFLD progression [50].

TLRS AND LIVER CANCER DEVELOPMENT IL‐6 plays a central role in hepatocellular carcinoma (HCC) initiation and HCC cell proliferation. IL‐6 production is mediated through MyD88 in the chemically‐induced HCC mouse model [51]. In HCC patients, increased TLR4 and TLR9 expression is correlated with a poor prognosis [52]. It is suggested that TLRs play a pivotal role in HCC development. In mouse models of HCC, mice deficient in TLR4, TLR9, and MyD88 were resistant to chemically‐ and genetically‐induced HCC [53–55]. Gut‐derived LPS is presumed to be the ligand for TLR4 in HCC development. In fact, gut‐sterilized mice, generated by orally administering non‐absorbable antibiotics, and germ‐free mice showed reduced HCC whereas chronically low‐dose LPS‐treated mice had increased HCC development [54]. These findings indicate that the intestinal microbiome, TLR4, and TLR9 contribute to HCC development. Most HCCs are developed in a fibrotic liver environment in humans. In the fibrotic condition, HSCs produce epiregulin in a TLR4‐­ mediated NF‐κB‐dependent manner, and epiregulin drives HCC development [54]. HMGB1, an endogenous ligand for TLR4, is overexpressed in human HCC. Hypoxia induces translocation of HMGB1 from the nucleus to the cytosol in HCC cells. In the hypoxic condition, HMGB1 plays a role in the promotion of HCC invasion and metastasis via TLR4 and RAGE, both HMGB1 receptors [56]. Hypoxia induces TLR9 overexpression and TLR9 can be activated by HMGB1 via binding to mitochondrial DNA [57]. Diminished hepatocyte autophagy, often observed in advanced NAFLD and aged people, also contributes to HMGB1 translocation, which promotes liver tumor development [58]. For this HMGB1 translocation, NRF2 and caspase‐11/caspase‐1/gasdermin D activation is required. HMGB1 could be important in autophagy‐suppressing NAFLD‐associated HCC [58]. During HCC development, ductular reaction and progenitor cell proliferation are often observed and suggested to play a role. Hepatocyte HMGB1 contributes to ductular reaction and progenitor cell proliferation through the post‐transcriptional modifications of HMGB1 [59]. Disulfide HMGB1 is proinflammatory and promotes ductular cell/hepatic progenitor cell proliferation in a RAGE‐dependent and ERK‐ and CREB‐phosphorylation‐ dependent manner [59]. Thus, in HMGB1‐mediated HCC, hypoxia, HMGB1 translocation, TLR4, RAGE, TLR9, and ductular reactions play roles.

A high‐fat diet (HFD) feeding is known to change the composition of the gut microbiome, increasing the number of Gram‐ positive Clostridium in the intestine. Clostridium‐mediated HCC development is mediated through TLR2 [60]. Clostridium can convert primary bile acids to the secondary bile acids  – deoxycholic acids – that increase levels of circulating deoxycholic acids. Translocation of these toxic, secondary bile acids to the liver causes HSC senescence, induction of senescence‐associated secretory phenotype (SASP) [60], and the SASP‐mediated HCC development in mice with the Ras mutation‐inducing chemical (7,12‐dimethylbenz(a)anthracene [DMBA]) and HFD. Overall, these studies suggested that both intestine‐derived factors and DAMPs contribute to HCC promotion via TLRs or other related signaling receptors, and that i­nhibition of TLR signaling or HMGB1 function could be therapeutic options for HCC. However, as anticancer strategies, TLR activation could enhance tumor immunity (discussed below).

ROLE OF TLRS IN VIRAL HEPATITIS While adaptive immunity plays a major role in the development of hepatitis B and C, both HBV and HCV‐derived PAMPs can stimulate innate immune signaling. Intriguingly, HBV and HCV have multiple mechanisms to evade the host immune response. This is the reason why these viruses cause chronic infection. Both HBV and HCV infect and replicate in hepatocytes. In chronic HBV infection, TLR2, TLR7, and TLR9 expression is reduced in immune cells. Additionally, HBV antigen and HBV polymerase can inhibit TLR3 and TLR9‐mediated IFN production (Figure 57.3) [61]. While HBV is a DNA virus, like other viruses, HBV produces single‐stranded RNA to translate its coding proteins. The single‐stranded RNA produced could activate TLR7. However, the expression of TLR7 and TLR7 signaling activation are impaired in chronic HBV infection and immune response against HBV, such as IFN‐α production, is defective [62]. By these mechanisms, HBV has potential to infect chronically. During HCV infection, HCV core and NS3 proteins can be recognized by TLR2 in immune cells. TLR7 senses HCV single‐stranded RNA in plasmacytoid dendritic cells for antigen presentation. For HCV antigen‐presentation, the cell–cell interaction between plasmacytoid dendritic cells and hepatocytes is required. The major pattern recognition receptors for HCV in hepatocytes is the retinoic acid inducible gene I (RIG‐I) and melanoma differentiation‐associated protein 5 (MDA5). HCV genome contains 5′ triphosphate (5′ ppp) and uridine/adenosine rich sequences (polyA/U) that are recognized by RIG‐I [63]. MDA5 can sense long double‐stranded RNA (greater than 1 kb) [63]. Although the expression is low, TLR3 can be activated by double‐stranded RNA produced during HCV replication. These receptors activate NF‐κB and IRF3/7 to induce IFN stimulated genes (ISGs) and type I and III IFNs to eradicate HCV (Figure 57.3). However, this IFN‐inducible signaling is immediately inhibited by HCV‐associated molecules [63]. HCV NS3/4A serine proteases disrupt IPS‐1 and TRIF to inhibit IFN‐ inducing signaling (Figure 57.3). The NS4B interferes with the STING‐TBK1‐IRF3 signaling pathway by binding to STING.



57:  Toll‐like Receptors in Liver Disease

743

Figure 57.3  Hepatitis B virus (HBV) and hepatitis C virus (HCV) regulate innate immune responses. ssRNA produced during HBV infection can activate TLR7 to induce type I IFN production for HBV clearance. However, the HBV viral components can inhibit TLR3, TLR7, and TLR9 signaling to prevent viral eradication. HCV can activate TLR2, TLR3, TLR7 and cytosolic RIG‐I and MDA5. HCV NS3/4A and NS3 inhibits TRIF and STING to suppress type I IFN‐β production. HCV NS3/4A cleaves off IPS‐1 at C508, resulting in prevention of IRF‐3‐mediated type I IFN production. HCV NS5A is associated with HCC progression through upregulation of TLR4.

Thus, HCV can activate innate immune signaling that induce antiviral IFN production, but concomitantly blocks this signaling by disrupting the essential signaling molecules, which allows for chronic HCV infection in hepatocytes. In HCV‐mediated HCC development, TLR4 plays a major role. HCV‐NS5A protein enhances nanog and yes‐associated protein (YAP)‐mediated HCC initiation by increasing ­hepatocyte TLR4 expression [64, 65]. This TLR4‐dependent HCV‐mediated hepatocarcinogenesis is enhanced by additional alcohol abuse or HFD feeding [64, 65]. The therapeutic potential of TLR agonists for viral hepatitis are discussed below.

expression are increased [66], suggestive of their roles. Moreover, TLR4 expression is correlated with the degree of fibrosis [66]. The significant role of TLR7 and type I IFN have also been suggested using mice with overexpression of IFNγ that recapitulate human PBC [67]. These studies strongly ­suggested the contribution of TLR signaling in PBC. PSC is often associated with inflammatory bowel disease (IBD), in particular, ulcerative colitis. Because there is a close association between IBD and gut microbiome, some level of contribution of the gut microbiome in PSC is suggested but it has not been studied well.

AUTOIMMUNE LIVER DISEASE

THERAPEUTIC PERSPECTIVE

Autoimmune hepatitis, primary biliary cholangitis (PBC), and primary sclerosing cholangitis (PSC) are three major autoimmune liver diseases. These autoimmune liver diseases are characterized by the presence of serum autoantibodies (e.g. antinuclear antibodies [ANA]) and high serum IgG levels. In addition, B cell autoimmunity plays a role. In PBC, B cells stimulated with CpG‐B produce more IgM and antimitochondrial antibodies (AMA), suggestive of the role played by TLR9 and intestine‐derived bacterial DNA in the progression of PBC [66]. In animal studies, repeated administration of poly I:C developed a PBC‐like phenotype characterized by elevated AMA and ANA in their sera [66]. In PBC livers, TLR3 and type I IFN

Several TLR4 antagonists have been tested in clinical trials for treating sepsis and agonists for TLR3, TLR7, and TLR9 for bacterial and viral infection. Target diseases for TLR‐related agents are not limited to inflammatory and infectious diseases but include malignant tumors. Because strong TLR activation induces harmful tissue and organ damage, optimization of safe doses of TLR ligands is critical. Optimized doses of TLR ligands may have the potential to induce antibacterial and antiviral mediators, such as IFNs and other cytokines for clearance of pathogens, without detrimental side‐effects. In chronic liver disease, such as viral hepatitis and liver malignancy, immune systems are often suppressed. Using adequate doses of TLR

744

THE LIVER:  REFERENCES

agonists to boost the immune system might enhance the ­immunotherapeutic drugs as adjuvants. TLR ligands can prime tumor‐specific T cells via dendritic cell/macrophage maturation and activation, which could boost the effectiveness of tumor ablation therapies and immune checkpoint inhibitors. Eritoran, a pharmacological TLR4 antagonist, has been shown to protect I/R liver injury through inhibition of HMGB1‐ mediated inflammatory signaling by interfering with the HMGB1 and TLR4 interaction [68]. This can also inhibit hepatic HMGB1 release. I/R liver injury was significantly ameliorated with eritoran [68]. TAK‐242, another TLR4 antagonist, has originally targeted sepsis, but has the potential to treat TLR4‐mediated liver damage. This agent has been reported to protect I/R liver injury in the cardiac death animal model [69]. This agent inhibited the intracellular domain of TLR4, and ­successfully reduced liver tissue injury by administrating prior to cardiac death. Preclinical studies using animal models demonstrated that TLR9 antagonists inhibited I/R liver injury and NAFLD development [41, 70]. TLR7 and TLR9 agonists are considered treatments for HBV  and HCV infection through their capacity to upregulate intrahepatic type I IFNs levels that can inhibit hepatitis virus replication. GS‐9620, a TLR7 agonist, has been tested on HBV patients. While there was no significant decline of hepatitis B surface antigen levels, GS‐9620 improved T cells and NK cell response by producing type I IFN from dendritic cells and B cells, which might treat HBV infection [71]. The therapeutic effects of classical TLR7 agonists imiquimod and resiquimod on mouse models of liver fibrosis and NAFLD have been shown [49, 50]. These agonists induce proinflammatory cytokine ­production in addition to anti‐inflammatory effects. As such, for clinical use, some modifications are needed. In addition to the pro‐tumorigenic role of TLRs as we discussed earlier, agonists for TLR7 (resiquimod, imiquimod) and TLR9 (CpG‐oligonucleotides) have anti‐tumorigenic effects via enhancing dendritic cell activity on tumor antigen presentation [72, 73]. Moreover, Poly I:C, a TLR3 agonist, enhances NK cell tumoricidal activity [74, 75]. In addition, Bacillus Calmette– Guerin (BCG), a tuberculosis vaccine – an attenuated strain of Mycobacterium, and monophosphoryl lipid A – enhance tumor immunity via TLR2/4 [76–78].

CONCLUSION Liver is continuously exposed to exogenous and endogenous gut‐derived materials through the portal vein, but it rarely causes liver inflammation. Liver immunity is tightly regulated to avoid detrimental inflammation. However, once serious infections occur, activated TLR signaling contributes to host defense. If the TLR signaling is activated excessively, it causes unfavorable organ and tissue damage. The tightly regulated immune balance is critical for maintenance of liver tissue homeostasis. Expression and activation of TLRs can affect liver disease progression. This is also associated with a prognosis of various liver diseases. Although we need further mechanistic studies to fully understand the underlying molecular mechanisms of liver diseases mediated by TLRs, TLRs and their associated and

downstream molecules have the potential to be targets for the treatment of liver diseases. Currently, clinical trials for testing antagonists for TLRs are limited to sepsis. Some trials testing TLR agonists are now for treating viral hepatitis and cancers. Preclinical studies have widely been performed and shown the pivotal roles of TLR signaling in various liver diseases, such as I/R liver injury, ALD, NAFLD/NASH, and liver fibrosis. Although many of them have not been done for clinical trials, TLR agonists and antagonists have potential to be tested further. Existing agents may still have detrimental side‐effects. For example, TLR antagonists may cause immunosuppression, which may cause severe secondary infection. Excessive use of TLR agonists may induce tissue damage due to overt inflammation. As such, we still need further improvement of TLR‐­ targeting agents to reduce detrimental side‐effects. Some of them are under development and further investigations of these novel agents on liver diseases are needed [79, 80].

ACKNOWLEDGMENTS Conflicts of interest The authors have no conflicts of interest to declare.

Financial support This work is supported by NIH grants R01DK085252 and R21AA025841, and a Winnick Research award from Cedars‐ Sinai Medical Center.

REFERENCES 1. O’Neill, L.A., Golenbock, D., and Bowie, A.G. The history of Toll‐like ­receptors ‐ redefining innate immunity. Nat Rev Immunol, 2013;13(6):453–60. 2. Yang, L. and Seki, E. Toll‐like receptors in liver fibrosis: cellular crosstalk and mechanisms. Front Physiol, 2012;3:138. 3. Pal, D., Dasgupta, S., Kundu, R. et  al. Fetuin‐A acts as an endogenous ligand  of TLR4 to promote lipid‐induced insulin resistance. Nat Med, 2012;18(8):1279–85. 4. Lancaster, G.I., Langley, K.G., Berglund, N.A. et al. Evidence that TLR4 is not a receptor for saturated fatty acids but mediates lipid‐induced inflammation by reprogramming macrophage metabolism. Cell Metab, 2018;27(5):1096–110. 5. Kim, S.Y., Jeong, J.M., Kim, S.J. et  al. Pro‐inflammatory hepatic macrophages generate ROS through NADPH oxidase 2 via endocytosis of ­monomeric TLR4‐MD2 complex. Nat Commun, 2017;8(1):2247. 6. Chen, X., Liang, H., Zhang, J., Zen, K., and Zhang, C.Y. microRNAs are ligands of Toll‐like receptors. RNA, 2013;19(6):737–9. 7. Oka, T., Hikoso, S., Yamaguchi, O. et al. Mitochondrial DNA that escapes from autophagy causes inflammation and heart failure. Nature, 2012;485(7397):251–5. 8. Zhong, Z., Liang, S., Sanchez‐Lopez, E. et  al. New mitochondrial DNA synthesis enables NLRP3 inflammasome activation. Nature, 2018; ­ 560(7717):198–203. 9. Seki, E. and Schnabl, B. Role of innate immunity and the microbiota in liver  fibrosis: crosstalk between the liver and gut. J Physiol, 2012; 590(3):447–58. 10. Hartmann, P., Chen, W.C., and Schnabl, B. The intestinal microbiome and the leaky gut as therapeutic targets in alcoholic liver disease. Front Physiol, 2012;3:402.



57:  Toll‐like Receptors in Liver Disease

11. Hartmann, P., Seebauer, C.T., and Schnabl, B. Alcoholic liver disease: the  gut microbiome and liver cross talk. Alcohol Clin Exp Res, 2015; 39(5):763–75. 12. Ferrere, G., Wrzosek, L., Cailleux, F. et al. Fecal microbiota manipulation prevents dysbiosis and alcohol‐induced liver injury in mice. J Hepatol, 2017;66(4):806–15. 13. Bukong, T.N., Cho, Y., Iracheta‐Vellve, A. et al. Abnormal neutrophil traps and impaired efferocytosis contribute to liver injury and sepsis severity after binge alcohol use. J Hepatol, 2018;69(5):1145–54. 14. Roh, Y.S., Zhang, B., Loomba, R., and Seki, E. TLR2 and TLR9 contribute to alcohol‐mediated liver injury through induction of CXCL1 and neutrophil infiltration. Am J Physiol Gastrointest Liver Physiol, 2015;309(1):G30–41. 15. Yang, A.M., Inamine, T., Hochrath, K. et  al. Intestinal fungi contribute to development of alcoholic liver disease. J Clin Invest, 2017;127(7):2829–41. 16. Huang, H., Shiffman, M.L., Friedman, S. et al. A 7 gene signature identifies the risk of developing cirrhosis in patients with chronic hepatitis C. Hepatology, 2007;46(2):297–306. 17. Seki, E., De Minicis, S., Osterreicher, C.H. et al. TLR4 enhances TGF‐beta signaling and hepatic fibrosis. Nat Med, 2007;13(11):1324–32. 18. Roderburg, C., Urban, G.W., Bettermann, K. et  al. Micro‐RNA profiling reveals a role for miR‐29 in human and murine liver fibrosis. Hepatology, 2011;53(1):209–18. 19. Liu, C., Chen, X., Yang, L., Kisseleva, T., Brenner, D.A., and Seki, E. Transcriptional repression of the transforming growth factor beta (TGF‐beta) pseudoreceptor BMP and activin membrane‐bound inhibitor (BAMBI) by nuclear factor kappaB (NF‐kappaB) p50 enhances TGF‐beta signaling in hepatic stellate cells. J Biol Chem, 2014;289(10):7082–91. 20. Chan, C.C., Hwang, S.J., Lee, F.Y. et  al. Prognostic value of plasma endotoxin levels in patients with cirrhosis. Scand J Gastroenterol, ­ 1997;32(9):942–6. 21. Frances, R., Benlloch, S., Zapater, P. et al. A sequential study of serum bacterial DNA in patients with advanced cirrhosis and ascites. Hepatology, 2004;39(2):484–91. 22. Hartmann, P., Haimerl, M., Mazagova, M., Brenner, D.A., and Schnabl, B. Toll‐like receptor 2‐mediated intestinal injury and enteric tumor necrosis factor receptor I contribute to liver fibrosis in mice. Gastroenterology, 2012;143(5):1330–40 e1. 23. Mazagova, M., Wang, L., Anfora, A.T. et  al. Commensal microbiota is  hepatoprotective and prevents liver fibrosis in mice. FASEB J, 2015;29(3):1043–55. 24. Cani, P.D., Amar, J., Iglesias, M.A. et al. Metabolic endotoxemia initiates obesity and insulin resistance. Diabetes, 2007;56(7):1761–72. 25. Yang, L., Miura, K., Zhang, B. et al. TRIF differentially regulates hepatic steatosis and inflammation/fibrosis in mice. Cell Mol Gastroenterol Hepatol, 2017;3(3):469–83. 26. Miura, K., Kodama, Y., Inokuchi, S. et al. Toll‐like receptor 9 promotes steatohepatitis by induction of interleukin‐1beta in mice. Gastroenterology, 2010;139(1):323–34. 27. Miura, K., Yang, L., van Rooijen, N., Brenner, D.A., Ohnishi, H., and Seki, E. Toll‐like receptor 2 and palmitic acid cooperatively contribute to the development of nonalcoholic steatohepatitis through inflammasome ­activation in mice. Hepatology, 2013;57(2):577–89. 28. Tsung, A., Sahai, R., Tanaka, H. et al. The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia‐reperfusion. J Exp Med, 2005;201(7):1135–43. 29. Zhai, Y., Shen, X.D., O’Connell, R. et  al. Cutting edge: TLR4 activation mediates liver ischemia/reperfusion inflammatory response via IFN ­regulatory factor 3‐dependent MyD88‐independent pathway. J Immunol, 2004;173(12):7115–9. 30. Zhai, Y., Qiao, B., Gao, F. et al. Type I, but not type II, interferon is critical in liver injury induced after ischemia and reperfusion. Hepatology, 2008;47(1):199–206. 31. Nace, G.W., Huang, H., Klune, J.R. et al. Cellular‐specific role of toll‐like receptor 4 in hepatic ischemia‐reperfusion injury in mice. Hepatology, 2013;58(1):374–87. 32. Zeng, S., Dun, H., Ippagunta, N. et al. Receptor for advanced glycation end product (RAGE)‐dependent modulation of early growth response‐1 in hepatic ischemia/reperfusion injury. J Hepatol, 2009;50(5):929–36. 33. Bamboat, Z.M., Balachandran, V.P., Ocuin, L.M., Obaid, H., Plitas, G., and DeMatteo, R.P. Toll‐like receptor 9 inhibition confers protection from liver ischemia‐reperfusion injury. Hepatology, 2010;51(2):621–32.

745

34. Antoine, D.J., Williams, D.P., Kipar, A. et  al. High‐mobility group box‐1 protein and keratin‐18, circulating serum proteins informative of acetaminophen‐induced necrosis and apoptosis in vivo. Toxicol Sci, ­ 2009;112(2):521–31. 35. Huebener, P., Pradere, J.P., Hernandez, C. et  al. The HMGB1/RAGE axis  triggers neutrophil‐mediated injury amplification following necrosis. J Clin Invest, 2015;125(2):539–50. 36. Lundback, P., Lea, J.D., Sowinska, A. et al. A novel high mobility group box 1 neutralizing chimeric antibody attenuates drug‐induced liver injury and postinjury inflammation in mice. Hepatology, 2016;64(5):1699–710. 37. Yan, T., Wang, H., Zhao, M. et  al. Glycyrrhizin protects against acetaminophen‐induced acute liver injury via alleviating tumor necrosis factor alpha‐mediated apoptosis. Drug Metab Dispos, 2016;44(5):720–31. 38. Ge, X., Arriazu, E., Magdaleno, F. et al. High mobility group box‐1 drives fibrosis progression signaling via the receptor for advanced glycation end‐products in mice. Hepatology, 2018;68(6):2380–404. 39. Ge, X., Antoine, D.J., Lu, Y. et  al. High mobility group box‐1 (HMGB1) participates in the pathogenesis of alcoholic liver disease (ALD). J Biol Chem, 2014;289(33):22672–91. 40. Laursen, T.L., Stoy, S., Deleuran, B., Vilstrup, H., Gronbaek, H., and Sandahl, T.D. The damage‐associated molecular pattern HMGB1 is elevated in human alcoholic hepatitis, but does not seem to be a primary driver of inflammation. APMIS, 2016;124(9):741–7. 41. Garcia‐Martinez, I., Santoro, N., Chen, Y. et al. Hepatocyte mitochondrial DNA drives nonalcoholic steatohepatitis by activation of TLR9. J Clin Invest, 2016;126(3):859–64. 42. Cai, Y., Xu, M.J., Koritzinsky, E.H. et  al. Mitochondrial DNA‐enriched microparticles promote acute‐on‐chronic alcoholic neutrophilia and hepatotoxicity. JCI Insight, 2017;2(14). 43. McGill, M.R., Staggs, V.S., Sharpe, M.R., Lee, W.M., Jaeschke, H., and Acute Liver Failure Study Group. Serum mitochondrial biomarkers and damage‐associated molecular patterns are higher in acetaminophen overdose patients with poor outcome. Hepatology, 2014;60(4):1336–45. 44. He, Y., Feng, D., Li, M. et al. Hepatic mitochondrial DNA/Toll‐like receptor 9/MicroRNA‐223 forms a negative feedback loop to limit neutrophil overactivation and acetaminophen hepatotoxicity in mice. Hepatology, ­ 2017;66(1):220–34. 45. Jeong, W.I., Park, O., Radaeva, S., and Gao, B. STAT1 inhibits liver fibrosis in mice by inhibiting stellate cell proliferation and stimulating NK cell ­cytotoxicity. Hepatology, 2006;44(6):1441–51. 46. Jeong, W.I., Park, O., and Gao, B. Abrogation of the antifibrotic effects of natural killer cells/interferon‐gamma contributes to alcohol acceleration of liver fibrosis. Gastroenterology, 2008;134(1):248–58. 47. Jeong, W.I., Park, O., Suh, Y.G. et  al. Suppression of innate immunity (­natural killer cell/interferon‐gamma) in the advanced stages of liver fibrosis in mice. Hepatology, 2011;53(4):1342–51. 48. Seo, W., Eun, H.S., Kim, S.Y. et al. Exosome‐mediated activation of toll‐like receptor 3 in stellate cells stimulates interleukin‐17 production by gammadelta T cells in liver fibrosis. Hepatology, 2016;64(2):616–31. 49. Roh, Y.S., Park, S., Kim, J.W., Lim, C.W., Seki, E., and Kim, B. Toll‐like receptor 7‐mediated type I interferon signaling prevents cholestasis‐ and hepatotoxin‐induced liver fibrosis. Hepatology, 2014; 60(1):237–49. 50. Kim, S., Park, S., Kim, B., and Kwon, J. Toll‐like receptor 7 affects the pathogenesis of non‐alcoholic fatty liver disease. Sci Rep, 2016;6:27849. 51. Naugler, W.E., Sakurai, T., Kim, S. et  al. Gender disparity in liver cancer due  to sex differences in MyD88‐dependent IL‐6 production. Science, 2007;317(5834):121–4. 52. Eiro, N., Altadill, A., Juarez, L.M. et  al. Toll‐like receptors 3, 4 and 9 in hepatocellular carcinoma: relationship with clinicopathological characteristics and prognosis. Hepatol Res, 2014;44(7):769–78. 53. Song, I.J., Yang, Y.M., Inokuchi‐Shimizu, S., Roh, Y.S., Yang, L., and Seki, E. The contribution of toll‐like receptor signaling to the development of liver fibrosis and cancer in hepatocyte‐specific TAK1‐deleted mice. Int J Cancer, 2018;142(1):81–91. 54. Dapito, D.H., Mencin, A., Gwak, G.Y. et  al. Promotion of hepatocellular carcinoma by the intestinal microbiota and TLR4. Cancer Cell, 2012;21(4):504–16. 55. Miura, K., Ishioka, M., Minami, S. et al. Toll‐like receptor 4 on macrophage promotes the development of steatohepatitis‐related hepatocellular ­carcinoma in mice. J Biol Chem, 2016;291(22):11504–17.

746

THE LIVER:  REFERENCES

56. Yan, W., Chang, Y., Liang, X. et  al. High‐mobility group box 1 activates caspase‐1 and promotes hepatocellular carcinoma invasiveness and metastases. Hepatology, 2012;55(6):1863–75. 57. Tohme, S., Yazdani, H.O., Liu, Y. et  al. Hypoxia mediates mitochondrial ­biogenesis in hepatocellular carcinoma to promote tumor growth through HMGB1 and TLR9 interaction. Hepatology, 2017;66(1):182–97. 58. Khambu, B., Huda, N., Chen, X. et  al. HMGB1 promotes ductular reaction  and tumorigenesis in autophagy‐deficient livers. J Clin Invest, ­ 2018;128(6):2419–35. 59. Hernandez, C., Huebener, P., Pradere, J.P., Antoine, D.J., Friedman, R.A., and Schwabe, R.F. HMGB1 links chronic liver injury to progenitor responses and hepatocarcinogenesis. J Clin Invest, 2018;128(6):2436–51. 60. Yoshimoto, S., Loo, T.M., Atarashi, K. et al. Obesity‐induced gut microbial metabolite promotes liver cancer through senescence secretome. Nature, 2013;499(7456):97–101. 61. Ma, Z., Zhang, E., Yang, D., and Lu, M. Contribution of Toll‐like receptors to the control of hepatitis B virus infection by initiating antiviral innate responses and promoting specific adaptive immune responses. Cell Mol Immunol, 2015;12(3):273–82. 62. Sepehri, Z., Kiani, Z., Alavian, S.M., Arababadi, M.K., and Kennedy, D. The link between TLR7 signaling and hepatitis B virus infection. Life Sci, 2016;158:63–9. 63. Bang, B.R., Elmasry, S., and Saito, T. Organ system view of the hepatic innate immunity in HCV infection. J Med Virol, 2016;88(12):2025–37. 64. Machida, K., Tsukamoto, H., Mkrtchyan, H. et al. Toll‐like receptor 4 mediates synergism between alcohol and HCV in hepatic oncogenesis involving stem cell marker Nanog. Proc Natl Acad Sci USA, 2009;106(5):1548–53. 65. Chen, C.L., Tsukamoto, H., Liu, J.C. et  al. Reciprocal regulation by TLR4  and TGF‐beta in tumor‐initiating stem‐like cells. J Clin Invest, 2013;123(7):2832–49. 66. Moritoki, Y., Lian, Z.X., Ohsugi, Y., Ueno, Y., and Gershwin, M.E. B cells and autoimmune liver diseases. Autoimmun Rev, 2006;5(7):449–57. 67. Bae, H.R., Hodge, D.L., Yang, G.X. et al. The interplay of type I and type II interferons in murine autoimmune cholangitis as a basis for sex‐biased ­autoimmunity. Hepatology, 2018;67(4):1408–19. 68. McDonald, K.A., Huang, H., Tohme, S. et al. Toll‐like receptor 4 (TLR4) antagonist eritoran tetrasodium attenuates liver ischemia and reperfusion injury through inhibition of high‐mobility group box protein B1 (HMGB1) signaling. Mol Med, 2015;20:639–48.

69. Shao, Z., Jiao, B., Liu, T., Cheng, Y., Liu, H., and Liu, Y. TAK‐242 treatment ameliorates liver ischemia/reperfusion injury by inhibiting TLR4 signaling pathway in a swine model of Maastricht‐category‐III cardiac death. Biomed Pharmacother, 2016;84:495–501. 70. Shaker, M.E., Trawick, B.N., and Mehal, W.Z. The novel TLR9 antagonist COV08–0064 protects from ischemia/reperfusion injury in non‐steatotic and steatotic mice livers. Biochem Pharmacol, 2016;112:90–101. 71. Boni, C., Vecchi, A., Rossi, M. et al. TLR7 agonist increases responses of hepatitis B virus‐specific T cells and natural killer cells in patients with chronic hepatitis B treated with nucleos(T)ide analogues. Gastroenterology, 2018;154(6):1764–77. 72. Drobits, B., Holcmann, M., Amberg, N. et al. Imiquimod clears tumors in mice independent of adaptive immunity by converting pDCs into tumor‐­ killing effector cells. J Clin Invest, 2012;122(2):575–85. 73. Nierkens, S., den Brok, M.H., Garcia, Z. et  al. Immune adjuvant efficacy of  CpG oligonucleotide in cancer treatment is founded specifically upon  TLR9 function in plasmacytoid dendritic cells. Cancer Res, 2011; 71(20):6428–37. 74. Ebihara, T., Azuma, M., Oshiumi, H. et  al. Identification of a polyI: C‐inducible membrane protein that participates in dendritic cell‐mediated natural killer cell activation. J Exp Med, 2010;207(12):2675–87. 75. Shime, H., Matsumoto, M., Oshiumi, H. et al. Toll‐like receptor 3 signaling converts tumor‐supporting myeloid cells to tumoricidal effectors. Proc Natl Acad Sci USA, 2012;109(6):2066–71. 76. Hoffman, E.S., Smith, R.E., and Renaud, R.C., Jr. From the analyst’s couch: TLR‐targeted therapeutics. Nat Rev Drug Discov, 2005;4(11):879–80. 77. Paavonen, J., Naud, P., Salmeron, J. et al. Efficacy of human papillomavirus (HPV)‐16/18 AS04‐adjuvanted vaccine against cervical infection and precancer caused by oncogenic HPV types (PATRICIA): final analysis ­ of  a  double‐blind, randomised study in young women. Lancet, 2009; 374(9686):301–14. 78. Lehtinen, M. and Paavonen, J. Sound efficacy of prophylactic HPV vaccination: basics and implications. Oncoimmunology, 2012;1(6):995–6. 79. Shinchi, H., Crain, B., Yao, S. et al. Enhancement of the immunostimulatory activity of a TLR7 ligand by conjugation to polysaccharides. Bioconjug Chem, 2015;26(8):1713–23. 80. Hayashi, T., Crain, B., Yao, S. et al. Novel synthetic toll‐like receptor 4/MD2 ligands attenuate sterile inflammation. J Pharmacol Exp Ther, 2014; 350(2):330–40.

PART FIVE: LIVER CANCER

58

Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models Sun Young Yim1, Jae‐Jun Shim2, Bo Hwa Sohn3, and Ju‐Seog Lee3 Department of Internal Medicine, Korea University College of Medicine, Seoul, Korea Department of Internal Medicine, Kyung Hee University College of Medicine, Seoul, Korea 3 Department of Systems Biology, The University of Texas M.D. Anderson Cancer Center, Houston, TX, USA 1 2

INTRODUCTION Hepatocellular carcinoma (HCC) represents 75% of cases of primary liver cancer [1], which is the seventh most common cancer globally [2]. Despite the implementation of surveillance programs for at‐risk populations, 30–60% of HCC tumors are detected at an advanced stage [3], resulting in a dismal prognosis (5‐year survival rates of 0–10%) [1]. Currently, sorafenib and regorafinib are the only approved molecular targeted therapies for HCC [4, 5]. Therefore, improved treatments are still needed, and much remains to be discovered in clinical and experimental studies. Hepatocarcinogenesis is a complex and multistep process involving the accumulation of genetic changes and resulting in altered expression of cancer‐related genes, such as oncogenes and tumor suppressor genes, and their related molecular signaling pathways. To understand such complex processes, it is important to have good experimental models that accurately recapitulate the steps of hepatocarcinogenesis in humans. In this review, we will briefly summarize the currently available experimental models for studying HCC. Currently, genomic studies of human HCC are progressing, and systematic analyses of genomic data are providing insights into the biology and pathogenesis of HCC. Many HCC genome studies have catalogued potential driver genes in HCC. TERT encodes a rate‐limiting catalytic subunit of telomerase that is essential to maintain telomere length and plays a pivotal role in stem cells, aging, and cancer [6, 7]. Although expression of TERT is mostly repressed in somatic cells, except for self‐renewing cells such as stem cells [8], 70–90% of cancer cells stably express this enzyme, which is reactivated during tumorigenesis and is necessary for

unlimited proliferation of cancer cells [9–11]. Whole‐genome sequencing studies showed that TERT is one of most frequently mutated genes in HCC. TERT promoters have been found to be mutated in more than 50% of HCC tissue samples examined, making them the most frequently occurring single‐nucleotide ­ mutations observed in HCC [12, 13]. Frequent alterations are also known to occur in other key cancer genes such as TP53, CTNNB1, ARID1A, ARID2, NFE2L2, KEAP1, and cell cycle‐related genes such as CCND1 and CDKN2A in HCC [14–17]. Since genomic characteristics of HCC closely reflect biology and clinical outcomes [18–20], we will also describe how accurately experimental models recapitulate human HCC at the genome level.

EXPERIMENTAL MODELS OF LIVER CANCER As HCC arises from normal hepatocytes or progenitor cells through the stepwise accumulation of genetic alterations, it is necessary to have experimental models that accurately recapitulate each step of carcinogenesis. Establishing HCC models mimicking the human condition are essential both for understanding tumor biology and for preclinical therapeutic studies. It is important to choose the most appropriate models to address particular questions during hepatocarcinogenesis (Figure 58.1). Rodents are widely accepted models of HCC because of their similarity with humans. Advantages include rapid, reproducible tumor induction and the possibility of studying progression of tumors from early to late stage. Cell lines are equally important models as much of our understanding of the molecular basis of

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

750

THE LIVER:  EXPERIMENTAL MODELS OF LIVER CANCER

Figure 58.1  Schematic diagrams showing various models for studying hepatocellular carcinoma.

HCC derives from the study of established cell lines that are explanted from human and mouse tumors.

Cell line models HCC cell lines have been widely used for modeling HCC to study molecular mechanisms because they carry most of genetic and epigenetic alterations that arose in the tumor from which they were derived. Importantly, cell lines provide an unlimited supply of homogeneous material and are very easy to handle. Cell lines have been used as test platforms to investigate functional roles of novel oncogene and tumor suppressor candidates as well as novel mutations in well‐known cancer genes that were discovered from many genome‐sequencing projects [14, 21–31]. Although many studies showed that telomerase is frequently reactivated in HCC [32–34], the genetic and molecular basis of telomerase reactivation has only recently been uncovered. Somatic mutations at two hotspots in the telomerase promoter were identified in 60% of patients with HCC [13]. The functional roles of these mutations have been validated in cell line models [35–37]. These mutations created high‐affinity binding sites for ETS/TCF transcription factors, increased

telomerase promoter activity, and induced expression of telomerase transcription. Cell lines are also essential for hypothesis‐ driven studies. Despite the extensive use of HCC cell line models for functional validations in various experiments, it is not clear how well these cell lines recapitulate the biological features of primary HCC and response to treatment. Recent generation of genomic data from cell lines and primary HCC allow good assessment of how similar these cell lines are to primary HCC at a molecular level. Several studies have shed light on how the genomic profiles, molecular subtypes, and heterogeneity of tumors compare with cell lines, and the results can be used to identify the cell lines that best model the genomic features of particular subtypes of HCC. Systematic comparison of gene expression profile data from HCC cell lines revealed global similarities in gene expression patterns between cancer cell lines and primary HCC [38, 39]. National Cancer Institute proliferation (NCIP) subtypes were first discovered using unsupervised analysis of genome‐ level expression data from human HCC tissues; subtype A represents tumors with poor prognoses [18]. When HCC cell lines are stratified according to NCIP gene expression signatures, 6 of the 18 HCC cells are classified as subtype A (Figure  58.2a).



58:  Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models

751

Figure 58.2  Similarity of hepatocellular carcinoma (HCC) cell lines to primary HCC. National Cancer Institute proliferation (NCIP) signature (a) and Hoshida signature (b) in human HCC cell lines.

Hoshida et  al. classified HCC tumors into three molecular ­subtypes (S1, S2, and S3) [40]. S1 and S2 subtypes are characterized by poor prognosis, high cell proliferation, and stem cell‐like characteristics, while S3 subtype is associated with better prognosis and tumors are typically well differentiated. Direct comparison of gene expression data showed that many of unique gene expression patterns for each subtype of primary HCC are well conserved in cell lines (Figure 58.2b). In recent study, Qiu et al. carried out more comprehensive analysis of newly established HCC cell lines with their matched primary tumors at the genomic level [41]. They demonstrated that HCC cell lines retained the vast majority of the genomic landscape of primary HCC. Only a small number of non‐silent mutations were detected during the establishment of cell lines. More importantly, no driver mutations or copy number alterations were additionally accumulated in cell lines during establishment of cell lines. Taken together, these studies suggest that many HCC cell lines may well recapitulate biological and molecular characteristics of primary tumors if they are properly maintained. Because cell lines serve as models to study connecting genomic alterations in response to treatments, many investigators have generated data sets that can link genomic profiles to the pharmacologic responses of cell lines. This approach was pioneered with data sets from NCI 60 cell lines, a collection of 60 cancer cell lines [42]. Unfortunately, HCC cell lines were not included in the NCI 60 cells. The Cancer Cell Line Encyclopedia (CCLE) project included 947 cancer cell lines representing 36 different cancers, 27 of which were liver cancer cells [43].

It  characterized cell lines for gene expression, copy number alterations, and somatic mutations of selected genes which were coupled with the pharmacological profiles of 24 anticancer drugs. Analysis of integrated data uncovered many interesting genetic, lineage, and gene expression pattern associated with sensitivity to anticancer drugs. Additional pharmacogenomics data are available from the Genomics of Drug Sensitivity in Cancer (GDSC) database (www.cancerRxgene.org) [44, 45]. These data are freely available without any restriction. The GDSC database currently contains drug‐sensitivity data describing the response to 265 anticancer drugs across 1074 cancer cell lines. To uncover molecular features correlated with response to anticancer drugs, the drug‐sensitivity data of cell lines are ­integrated with genomic data sets including somatic mutations of cancer genes, copy number alterations, and gene expression data. Currently data from 19 liver cancer cell lines are available through the web portal.

Non‐mouse models Animal models that recapitulate human physiology and clinical settings have been crucial for understanding hepatocarcinogenesis and improving the treatment of HCC. However, finding animal HCC models that are faithfully analogous to the human condition has been a great challenge to most of the investigators trying to uncover the mechanism of hepatocarcinogenesis in humans. The perfect animal model would reproduce the natural history, etiology, and pathology of human HCC and would not

752

THE LIVER:  EXPERIMENTAL MODELS OF LIVER CANCER

only allow the molecular mechanisms of HCC development to be uncovered over time but also allow the examination and ­evaluation of potential novel therapeutic approaches in the ­preclinical setting. Unfortunately, currently available animal models are far from ideal for all purposes. Since each model provides only limited characteristics of HCC development in humans, one should be aware of such limitations during investigation.

Zebrafish The zebrafish has been an excellent animal model for developmental biologist due to the optical clarity and easy manipulation of the embryos [46]. Its liver is related to human liver at physiological and genetic levels. Its hepatic cellular composition, function, signaling, and response to injury are all similar to those in human liver. Genes are also well conserved between human and zebrafish. These similarities makes zebrafish a useful system to study the basic mechanisms of liver disease, including liver cancer. Early studies showed that exposure of zebrafish to various carcinogens, such as 7,12‐dimethylbenzanthracene (DMBA), N‐methyl‐N′‐nitro‐N‐nitrosoguanidine (MNNG), N‐ethyl‐N‐ nitrosourea (ENU), and N‐nitrosodiethylamine, can induce formation of various type of cancers, including liver cancer [47]. Recent development in transgenic technology for zebrafish have made it more suitable for studying liver cancer. Similar to mouse model [48], overexpression of HBX in liver of zebrafish induced hepatic steatosis, which in turn led to liver degeneration [49]. The histology of transgenic zebrafish displayed steatosis, lobular inflammation, and balloon degeneration, similar to non‐ alcoholic steatohepatitis (NASH) in humans. However, unlike mouse model [50], HBX alone could not induce development of HCC. HCC development in zebrafish requires inactivation of the tp53 tumor suppressor gene [51]. MYC is one of most frequently amplified genes in human HCC and zebrafish has two MYC orthologs, myca and mycb. Overexpression of myca or mycb under inducible conditions is sufficient to develop HCC in zebrafish [52]. Introduction of tp53‐null mutation into the myca transgenic fish significantly accelerated tumor progression. The malignant status of hepatocytes was dependent on continued expression of myca since withdrawal of myca overexpression resulted in a rapid regression of liver tumors even in a tp53‐null background. As human HCC is more prevalent in men than in women, sex disparity is an important issues in properly understanding hepatocarcinogenesis [53]. Some transgenic zebrafish models showed similar sex disparity, for example zebrafish expressing krasV12 in liver showed sex disparity in development of HCC [54]. Similar patterns of sex disparity were also observed in the xmrk (fish ortholog of human epidermal growth factor receptor EGFR) transgenic model [55], indicating that zebrafish may reflect hormone‐related hepatocarcinogenesis in humans. Conserved similarity of zebrafish HCC to human HCC is supported by genome‐wide analysis of tumors. By using microarrays containing a small number of coding genes (1861 zebrafish unique genes), comparative analysis of microarray data from zebrafish HCC with those from four human tumor types revealed that zebrafish HCC is most similar to human HCC [56]. Furthermore, differentially expressed genes in zebrafish HCC were highly conserved in human HCC. Comparable levels of

similarity between human HCC and zebrafish HCC have also been observed in transgenic zebrafish models. Analysis of gene expression data of HCC from krasV12, xmrk, and Myc transgenic models showed that gene expression signatures significantly correlated with human HCC. Overall, around 50% of human HCCs share at least one of the three gene expression signatures identified in zebrafish [55].

Woodchuck Because the etiological roles of viral infection with hepatitis B virus (HBV) and hepatitis C virus (HCV) in hepatocarcinogenesis are well established through many epidemiological and clinical studies [57], it is important to have good animal models that can model virus‐mediated HCC development in humans. Hepatocellular adenoma in woodchuck was first reported in the early 1900s from zoos [58]. Woodchuck with hepatocellular adenoma were infected with woodchuck hepatitis virus (WHV), which is similar in structure and life cycle to human HBV [59]. Like HBV, WHV infection caused chronic hepatitis that led to development of HCC in 2–4 years after infection. Since the life cycle of WHV is similar to HBV, this model has been used for preclinical evaluation of antiviral drugs for treatment of HBV infection [60]. However, because cirrhosis does not typically develop in this model, it is not considered to be a fully accurate animal model for human HCC. Genomic profiling of WHV‐infected liver showed reduced response of type 1 interferon, T‐cell exhaustion, inhibition of cytokine signaling, and accumulation of neutrophils that are typically observed in HBV‐infected human liver [61]. At the HCC stage, their gene expression patterns are highly similar to poor prognostic NCIP subtype A [18]. Their tumors are also similar to Hoshida’s S2 subclass, in which activation of MYC and AKT signaling was associated with elevated levels of the hepatic progenitor markers AFP and EPCAM and relative suppression of IFN target genes [40].

Rabbit Although there are no practical rabbit models for studying HCC, VX2 tumor has been used quite extensively for studying imaging and treatment methods in the area of interventional oncology. VX2 tumor is an anaplastic squamous cell carcinoma derived from a virus‐induced papilloma in rabbits [62]. This tumor model is widely used for studying liver cancer because its hepatic artery blood supply is similar to that of human liver tumors and transplanted tumors in rabbit are large enough to be imaged [63, 64]. Furthermore, rabbits are large enough for effective catheter manipulation. However, VX2 tumor has been poorly characterized, because it does not grow in culture. For use, the tumor must be grown from animal to animal. It grows quickly, rapidly develops a necrotic core, and kills the host within weeks. Since its similarity to human HCC at the genomic or molecular level has not been evaluated, its relevance to human HCC is questionable.

Mouse models Due to the physiological, molecular, and genetic similarities to humans, the small size, large number of offspring, short lifespan, and low cost when compared to larger animals for the same



58:  Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models

experiment, mouse has become the preferred animal model for studying HCC. Mice are invaluable tools in uncovering the molecular mechanisms of hepatocarcinogenesis by introducing genetic alterations observed from human HCC that are discovered by sequencing cancer genomes [14, 21–31].

Chemically induced models The genotoxic drug diethylnitrosamine (DEN) has been extensively used to induce HCC in mice. DEN is a DNA alkylating agent leading to the formation of mutagenic DNA adducts. DEN is metabolized to α‐hydroxylnitrosamine first and further generates an electrophilic ethyldiazonium ion that causes DNA damage by reacting with DNA bases [65]. The process is oxygen‐ and NADPH‐dependent and is mediated by the cytochrome P450 family in hepatocytes. When injected into young mice with actively proliferating hepatocytes, it can activate certain oncogenes by generating somatic mutations. In addition, oxidative stress induced by reactive oxygen species (ROS) during DEN metabolization is known to contribute to hepatocarcinogenesis as they cause DNA, protein, and lipid damage [66]. Promotion of HCC development can be achieved by treating mice with ­carbon tetrachloride or phenobarbital [67]. These chemical models generated genetically distinct HCCs. Mutations activated in H‐Ras are most frequently observed in DEN‐ induced HCC, while mutations in Ctnnb1 are more frequent in DEN‐induced and phenobarbital‐promoted HCC [68]. Peroxisome proliferators are xenobiotics that can induce hepatomegaly when fed to mice and enlarged livers frequently develop HCC after a long latency period as a response to long‐ term exposure to these compounds [69]. Typical peroxisome proliferators include methyl clofenapate, ciprofibrate, fenofibrate, clofibrate, and Wy‐16,643 [70]. Peroxisome proliferators activate PPARα, which is a receptor regulating the expression of genes involved in cell proliferation and apoptosis [71]. It is ­currently unknown if long‐term exposure to peroxisome proliferators is equally carcinogenic to humans. When peroxisome proliferators were fed to mice humanized for PPARα on a long‐ term basis, a hepatocarcinogenic effect was not observed [72], suggesting that peroxisome proliferator‐mediated hepatocarcinogenesis could be a very species‐specific process.

Genetically engineered mouse models Genetically engineered mouse (GEM) models recapitulate the complex multistep process of hepatocarcinogenesis so that researchers can understand it and design therapeutic experiments. They are highly useful for assessing the impacts of driver oncogene alone or in combination with other driver oncogenes or tumor suppressors. GEM dramatically facilitate the functional validation of genetic alterations observed in the human HCC genome. Since a large number of HCC GEM models have already been covered by others, we will only discuss small ­number of representative mouse models and their similarity to human HCC at the genomic level. The MYC oncogene is known to be activated in more than 50% of human cancers by multiple mechanisms, and its activation is  frequently associated with poor prognosis and unfavorable ­outcome [73]. MYC plays a central role in multiple oncogenic process by regulating cell proliferation, apoptosis, and metabolism

753

and has been a key therapeutic target for treatment of many cancers, including HCC. However, directly targeting MYC as a therapeutic method has proven to be a challenge for many decades because of the undruggable nature of its molecular activity [74]. MYC is frequently amplified in HCC (around 20% of HCC according to the Cancer Genome Atlas (TCGA) study) [30]. GEM expressing MYC under albumin promoter develop HCC after a long period of latency [75, 76]. Coactivation of E2F1 with MYC significantly accelerates HCC development [76, 77]. Interestingly, a large number of HCC tumors in MYC and E2F1/MYC GEM models show activation of CTNNB1 [77], suggesting CTNNB1 might be an important interacting partner of MYC for development and progression of HCC. TGFA is one of ligands of EGFR and the TGFA–EGFR–RAS– MAPK signaling pathway is commonly upregulated in HCC [78]. GEM mice expressing TGFA under the inducible MT‐1 promoter develop HCC [79]. As expectedly, coactivation of MYC with TGFA significantly accelerates HCC development. Within 40 weeks of age, all of the mice developed HCC (100%) [75]. The role of FGF19 in HCC development was first discovered by oncogenomic screening for potential drivers of HCC in mice [80]. Later studies showed that FGF19 is frequently amplified in HCC genome and its activation is associated with worse survival of HCC patients [30, 81]. Expression of FGF19 in skeletal muscle in GEM mice led to the occurrence of HCC in 50% of the mice after long latency [82]. Interestingly, the incident rate of HCC in FGF19 GEM mice is higher in female mice than in male. More interestingly, Ctnnb1 is frequently activated in HCC tumors in the female mice while its activation is not observed in HCC tumors in the male mice. CTNNB1 encodes β‐catenin, which is a subunit of the cadherin protein complex on the cellular surface that acts as a signaling molecule in the WNT pathway [83]. When WNT signaling is absent, cytosolic β‐catenin protein levels are low because of phosphorylation‐dependent ubiquitination and degradation that is orchestrated by the AXIN complex, which is composed of AXIN, APC, CK1, and GSK3. When WNT ligands bind to a Frizzled receptor and its co‐receptor LRP5/6, WNT induces a receptor complex formation, which relocates AXIN to the plasma membrane, resulting in stabilization of β‐catenin. Stabilized β‐catenin forms a complex with the TCF/LEF and increases transcription of genes involved in cell growth. CTNNB1 is one of most frequently mutated genes in HCC. Aberrant activation of β‐catenin has been observed in 20–30% of HCC patients [24, 29, 30]. Most mutated residues are at phosphorylation sites or near to phosphorylation sites, preventing phosphorylation of β‐catenin. Thus, β‐catenin is constitutively activated by mutations. Interestingly, previous studies have shown that mutations in β‐catenin are almost mutually exclusive with mutations in TP53 [21]. These observations strongly suggest that HCC with a β‐catenin mutation may represent a clinically distinct subtype of HCC. Intriguingly, in GEM mice, introduction of activated mutants of β‐catenin or overexpression of wild‐type β‐catenin induces only hepatomegaly and is not enough for development of HCC [84, 85], suggesting that β‐catenin alone is insufficient to initiate tumorigenesis in liver. However, follow‐up studies in these GEM mice showed that β‐catenin collaborates with other signaling pathways to contribute to hepatocarcinogenesis. β‐Catenin was shown to cooperate with activated Hras to induce hepatocarcinogenesis. Mutations in both

754

THE LIVER:  GENOMIC RESEMBLANCE OF MOUSE MODELS TO HUMAN HCC

β‐catenin and Hras cause HCC development at an incidence of 100% [86], supporting the notion that β‐catenin activation is one of the hits that may be critical to the development of HCC, but that additional aberrations are necessary to initiate and promote tumorigenesis. The SV40 T antigen induces oncogenic transformation of normal cells, including hepatocytes, by inactivating the tumor suppressor genes p53 and Rb and interacting with a number of signaling proteins such as HSC70, CBP/p300, CUL7, IRS1, FBXW7, and BUB1 [87]. In GEM mice expressing the SV40 T‐antigen under hepatic promoters, HCC developed after a short period of latency (4–12 weeks) and metastasis to the lungs is frequently observed [88, 89]. Since tumor progression is very rapid, this model is considered to have some differences from human HCC tumors, which progress more slowly. The Hippo pathway was first discovered in Drosophila and is evolutionarily well conserved. All core components of the pathway identified to date have one or more mammalian orthologs, including MST1/2 (Hippo), SAV1 (Salvador, also known as WW45), LATS1/2 (Warts), MOB1 (Mats), YAP1 and its paralog TAZ (Yorkie), and TEAD1/2/3/4 (Scalloped) [90]. When Hippo signaling is active, YAP1/TAZ is phosphorylated (serine 127 residue in human YAP1) by LATS1/2 and sequestered by 14‐3‐3 proteins in cytoplasm. When Hippo signaling is reduced or absent, unphosphorylated YAP1/TAZ enters into the nucleus and increases transcriptional activation of genes involved in proliferation and survival. Hence, the normal function of the Hippo pathway is to repress growth and when Hippo signaling is ­attenuated, tissue overgrowth occurs. Because the Hippo pathway is required for restricting cell growth and proliferation, as well as to induce programmed cell death, many members of the pathway are known to be involved in tumor development. The first clear clue as to the involvement of the Hippo pathway in HCC came from identification of YAP1 as a potential oncogene in mouse HCC [91]. This was supported by the finding that ectopic overexpression of YAP1 in mouse liver led to the development of HCC [92]. This oncogenic function of YAP1 is further supported by the tumor suppressor function of its upstream regulators, which inhibit YAP1 activity by phosphorylating it. Mst1/2 and Sav1 knockout in liver leads to the development of HCC [93, 94]. The results from both genetically modified mouse models clearly demonstrated that the Hippo pathway is a key tumor suppressor in liver. In good agreement with data from mouse models, recent studies have shown that activation of the YAP1 and TAZ oncogenes in HCC is significantly associated with shorter survival rate, higher recurrence rate, and resistance to chemotherapy [93, 95–97], indicating the importance of YAP1/TAZ in the development of HCC. ARID1A and ARID2 are also frequently mutated in HCC (in up to 20% of cases) [27, 29, 30, 98]. They belong to the AT‐rich interaction domain (ARID) family, which contains 7 subfamilies and 15 members. ARID genes are characterized by a 100‐amino‐ acid DNA‐binding ARID domain. ARID1A associates with several other proteins to form the BRG1‐associated factor (BAF) complexes, a subfamily of a switch/sucrose‐nonfermentable (SWI/SNF) chromatin remodeling complex [99]. This complex uses the energy from ATP to mobilize nucleosomes by sliding, ejecting, and inserting histone octamers, thereby regulating DNA accessibility to other cellular machineries involved in transcription, DNA replication, and repair. Most cancer‐associated

mutations in ARID1A appear to be loss‐of‐function mutations; nonsense or frameshift rather than missense mutations in ARID1A are the dominant forms in many cancers, including HCC, suggesting that ARID1A is a tumor suppressor. ARID2 is a member of the polybromo‐associated BRG1‐associated factor (PBAF) complex, another SWI/SNF complex involved in ligand‐dependent transcriptional activation by nuclear receptors. Although mutations in ARID2 are less common than those in ARID1A, most mutations in ARID2 are also loss‐of‐function mutations, as seen in ARID1A [30]. An HCC mouse model for ARID1A has recently been generated but showed an unexpected and complicated phenotype. In contrast to the general notion of ARID1A as a tumor suppressor, as demonstrated in the colon cancer model [100], deletion of Arid1a in mouse liver protected against development of HCC [101], suggesting that Arid1a is necessary for the initiation of tumorigenesis in hepatocytes. The critical activity of Arid1a in tumor initiation appears to be related to its transcriptional regulation of the CYP450 family, which oxidize metabolites and generate ROS in hepatocytes. High expression of Arid1a in hepatocytes promoted tumorigenesis by increased CYP450‐ mediated production of ROS. In contrast to its tumor‐promoting activity during tumor initiation, deletion of Arid1a accelerated HCC tumor progression and metastasis in late stage of HCC development, further indicating the complicated roles for Arid1A in HCC development. Metastasis‐suppressive activity is related to a decrease in global chromatin accessibility and reduced expression of genes that inhibit metastasis. Although these opposite roles of Arid1a at different stages of HCC development are interesting, it is not surprising to find that epigenetic regulators have highly context‐specific functions as they play critical roles in remodeling chromatin that can support the actions of both oncogenic and tumor‐suppressive networks.

GENOMIC RESEMBLANCE OF MOUSE MODELS TO HUMAN HCC Although mouse models are extensively used in cancer research, no one mouse model fits all purposes, and each model can only recapitulate part of the process of hepatocarcinogenesis in humans. Furthermore, genomic studies have identified molecularly distinct subtypes of HCC with different clinical outcomes [18–20, 102–105]. Therefore, establishing the molecular and clinical resemblance of mouse HCC models to these subtypes of human HCC will enable the appropriate selection of mouse models for use in investigations of the functional roles of newly discovered cancer genes and validation of potential therapeutic targets. There have been several studies of systematic comparison of mouse models to human HCC [18, 103, 106]. In a recent study, genomic data from nine mouse HCC models were integrated and analyzed together with genomic data from human HCC [107] to identify the mouse models that best resembled subtypes of human HCC and determine the clinical relevance of each model (Figure  58.3). Mst1/2‐knockout, Sav1‐knockout, and SV40 T antigen mouse models best recapitulated the poor prognostic subtypes of human HCC, whereas the Myc transgenic model best resembled human HCCs with more favorable ­prognoses. The Myc model was also significantly associated with  ­activation of β‐catenin. E2f1, E2f1/Myc, E2f1/Tgfa, and



58:  Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models

755

Figure 58.3  Summary of stratification of mouse tumors according to clinically defined human genomic signatures.

diethylnitrosamine (DEN)‐induced models were heterogeneous and were unequally split into poor and favorable prognoses. Hepatic stem cell HCCs have been identified by several studies. Analysis with genomic signatures showed that two mouse HCC models with disrupted Hippo pathway (Mst1/2‐knockout and Sav1‐knockout) are most similar to the HS subtype. Significant association of mouse models with the HS subtype is also supported by their similarity to previously established EPCAM‐positive human HCC [108] and newly discovered IDH‐like human HCC tumors [30], whose underlying biology closely reflects stem cell characteristics by inhibiting hepatic differentiation of hepatic stem cells through production of 2‐ hydroxyglutarate [109]. This observation is in good agreement with a previous study showing that Yap1 reprogramed mature hepatocytes in adult mice into progenitor‐like cells that could trans‐differentiate into biliary epithelial cells [110]. Two mouse models are also characterized by high probability of recurrence after treatment as reflected in high recurrence risk scores among examined models, further supporting their similarity to the HS subtype as previous studies have demonstrated that HCC with stem cell characteristics have the poorest clinical outcomes among all HCCs [30, 103, 108]. Genome‐wide comparison of mouse models with human HCC would be the best way to identify the most appropriate mouse models for various subtypes of human HCC. Finding such an association would significantly facilitate preclinical studies for functional validation of new cancer genes and testing new drugs for therapeutic efficacy.

FUNDING This study was supported in part by the Duncan Cancer Prevention Research Seed Funding Program at MD Anderson Cancer Center (2016 cycle), the MD Anderson Sister Institute

Network Fund (2012 and 2016 cycles), and the National Institutes of Health through Cancer Center Support Grant P30 CA016672.

REFERENCES  1. Forner, A., Llovet, J.M., and Bruix, J. Hepatocellular carcinoma. Lancet, 2012;379(9822):1245–55.   2. Torre, L.A., Bray, F., Siegel, R.L. et al. Global cancer statistics, 2012. CA Cancer J Clin, 2015;65(2):87–108.  3. Bruix, J. and Sherman, M. American Association for the Study of Liver Disease. Management of hepatocellular carcinoma: an update. Hepatology, 2011;53(3):1020–2.   4. Llovet, J.M., Ricci, S., Mazzaferro, V. et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med, 2008;359(4):378–90.   5. Bruix, J., Qin, S., Merle, P. et al. Regorafenib for patients with hepatocellular carcinoma who progressed on sorafenib treatment (RESORCE): a randomised, double‐blind, placebo‐controlled, phase 3 trial. Lancet, 2017; 389(10064):56–66.   6. Smogorzewska, A. and de Lange, T. Regulation of telomerase by telomeric proteins. Annu Rev Biochem, 2004;73:177–208.  7. Bryan, T.M., Sperger, J.M., Chapman, K.B., and Cech, T.R. Telomerase reverse transcriptase genes identified in Tetrahymena thermophila and Oxytricha trifallax. Proc Natl Acad Sci U S A, 1998;95(15):8479–84.   8. Blasco, M.A. Telomeres and human disease: ageing, cancer and beyond. Nat Rev Genet, 2005;6(8):611–22.   9. Shay, J.W. and Bacchetti, S. A survey of telomerase activity in human cancer. Eur J Cancer, 1997;33(5):787–91. 10. Kim, N.W., Piatyszek M.A., Prowse, K.R. et  al. Specific association of human telomerase activity with immortal cells and cancer. Science, 1994;266(5193):2011–15. 11. Harley, C.B. Telomerase and cancer therapeutics. Nat Rev Cancer, 2008; 8(3):167–79. 12. Totoki, Y., Tatsuno, K., Covington, K.R. et  al. Trans‐ancestry mutational landscape of hepatocellular carcinoma genomes. Nat Genet, 2014; 46(12):1267–73. 13. Nault, J.C., Mallet, M., Pilati, C. et al. High frequency of telomerase reverse‐ transcriptase promoter somatic mutations in hepatocellular carcinoma and preneoplastic lesions. Nat Commun, 2013;4:2218.

756

THE LIVER:  REFERENCES

14. Lee, J.S. The mutational landscape of hepatocellular carcinoma. Clin Mol Hepatol, 2015;21(3):220–9. 15. Lee, J.S. Genomic profiling of liver cancer. Genomics Inform, 2013;11(4): 180–5. 16. Marquardt, J.U., Andersen, J.B., and Thorgeirsson, S.S. Functional and genetic deconstruction of the cellular origin in liver cancer. Nat Rev Cancer, 2015;15(11):653–67. 17. Cancer Genome Atlas Research Network. Comprehensive and integrative genomic characterization of hepatocellular carcinoma. Cell, 2017;169(7): 1327–41. 18. Lee, J.S., Chu, I.S., Heo, J. et al. Classification and prediction of survival in hepatocellular carcinoma by gene expression profiling. Hepatology, 2004;40(3):667–76. 19. Sohn, B.H., Shim, J.J., Kim, S.B. et  al. Inactivation of Hippo pathway is significantly associated with poor prognosis in hepatocellular carcinoma. Clin Cancer Res, 2016;22(5):1256–64. 20. Kim, S.M., Leem, S.H., Chu, I.S. et al. Sixty‐five gene‐based risk score classifier predicts overall survival in hepatocellular carcinoma. Hepatology, 2012;55(5):1443–52. 21. Ahn, S.M., Jang, S.J., Shim, J.H. et al. Genomic portrait of resectable hepatocellular carcinomas: implications of RB1 and FGF19 aberrations for patient ­stratification. Hepatology, 2014;60(6):1972–82. 22. Cleary, S.P., Jeck, W.R., Zhao, X. et al. Identification of driver genes in hepatocellular carcinoma by exome sequencing. Hepatology, 2013;58(5):1693–702. 23. Fujimoto, A., Totoki, Y., Abe, T. et al. Whole‐genome sequencing of liver cancers identifies etiological influences on mutation patterns and recurrent mutations in chromatin regulators. Nat Genet, 2012;44(7):760–4. 24. Guichard, C., Amaddeo, G., Imbeaud, S. et al. Integrated analysis of somatic mutations and focal copy‐number changes identifies key genes and pathways in hepatocellular carcinoma. Nat Genet, 2012;44(6):694–8. 25. Huang, J., Deng, Q., Wang, Q. et al. Exome sequencing of hepatitis B virus‐ associated hepatocellular carcinoma. Nat Genet, 2012;44(10):1117–21. 26. Kan, Z., Zheng, H., Liu, X. et al. Whole‐genome sequencing identifies recurrent mutations in hepatocellular carcinoma. Genome Res, 2013;23(9):1422–33. 27. Schulze, K., Imbeaud, S., Letouze, E. et al. Exome sequencing of hepatocellular carcinomas identifies new mutational. Nat Genet, 2015;47(5):505–11. 28. Sung, W.K., Zheng, H., Li, S. et al. Genome‐wide survey of recurrent HBV integration in hepatocellular carcinoma. Nat Genet, 2012;44(7):765–9. 29. Totoki, Y., Tatsuno, K., Covington, K.R. et  al. Trans‐ancestry mutational landscape of hepatocellular carcinoma genomes. Nat Genet, 2014;46(12): 1267–73. 30. Cancer Genome Atlas Research Network. Comprehensive and integrative genomic characterization of hepatocellular carcinoma. Cell, 2017;169(7): 1327–41.e23. 31. Lee, J.S. Exploring cancer genomic data from the cancer genome atlas project. BMB Rep, 2016;49(11):607–11. 32. Nakayama, J., Tahara, H., Tahara, E. et al. Telomerase activation by hTRT in human normal fibroblasts and hepatocellular carcinoma. Nat Genet, 1998;18(1):65–8. 33. Kojima, H., Yokosuka, O., Imazeki, F. et al. Telomerase activity and telomere length in hepatocellular carcinoma and chronic liver disease. Gastroenterology, 1997;112(2):493–500. 34. Miura, N., Horikawa, I., Nishimoto, A. et al. Progressive telomere shortening and telomerase reactivation during hepatocellular. Cancer Genet Cytogenet, 1997;93(1):56–62. 35. Borah, S., Xi, L., Zaug, A.J. et al. Cancer. TERT promoter mutations and telomerase reactivation in urothelial cancer. Science, 2015;347(6225):1006–10. 36. Horn, S., Figl, A., Rachakonda, P.S. et  al. TERT promoter mutations in familial and sporadic melanoma. Science, 2013;339(6122):959–61. 37. Huang, F.W., Hodis, E., Xu, M.J. et  al. Highly recurrent TERT promoter mutations in human melanoma. Science, 2013;339(6122):957–9. 38. Lee, J.S. and Thorgeirsson, S.S. Functional and genomic implications of global gene expression profiles in cell lines from human hepatocellular cancer. Hepatology, 2002;35(5):1134–43. 39. Staib, F., Krupp, M., Maass, T. et  al. CellMinerHCC: a microarray‐based expression database for hepatocellular carcinoma. Liver Int, 2014;34(4): 621–31. 40. Hoshida, Y., Nijman, S.M., Kobayashi, M. et  al. Integrative transcriptome analysis reveals common molecular subclasses of human hepatocellular carcinoma. Cancer Res, 2009;69(18):7385–92. 41. Qiu, Z., Zou, K., Zhuang, L. et al. Hepatocellular carcinoma cell lines retain the genomic and transcriptomic. Sci Rep, 2016;6:27411.

42. Shoemaker, R.H. The NCI60 human tumour cell line anticancer drug screen. Nat Rev Cancer, 2006;6(10):813–23. 43. Barretina, J., Caponigro, G., Stransky, N. et  al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug. Nature, 2012;483(7391):603–7. 44. Yang, W., Soares, J., Greninger, P. et  al. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic. Nucleic Acids Res, 2013; 41(Database issue):D955–61. 45. Iorio, F., Knijnenburg, T.A., Vis, D.J. et al. A landscape of pharmacogenomic interactions in cancer. Cell, 2016;166(3):740–54. 46. Cox, A.G. and Goessling, W. The lure of zebrafish in liver research: regulation of hepatic growth in development and regeneration. Curr Opin Genet Dev, 2015;32:153–61. 47. Lu, J.W., Ho, Y.J., Yang, Y.J. et al. Zebrafish as a disease model for studying human hepatocellular carcinoma. World J Gastroenterol, 2015;21(42): 12042–58. 48. Kim, J.Y., Song, E.H., Lee, H.J. et  al. HBx‐induced hepatic steatosis and apoptosis are regulated by TNFR1‐ and NF‐kappaB‐dependent pathways. J Mol Biol, 2010;397(4):917–31. 49. Shieh, Y.S., Chang, Y.S., Hong, J.‐R., et al. Increase of hepatic fat accumulation by liver specific expression of hepatitis B. Biochim Biophys Acta, 2010; 1801(7):721–30. 50. Kim, C.M., Koike, K., Saito, I. et al. HBx gene of hepatitis B virus induces liver cancer in transgenic mice. Nature, 1991;351(6324):317–20. 51. Lu, J.W., Yang, W.Y., Tsai, S.‐M. et al. Liver‐specific expressions of HBx and src in the p53 mutant trigger. PLoS One. 2013;8(10):e76951. 52. Sun, L., Nguyen, A.T., Spitsbergen, J.M., and Gong, Z. Myc‐induced liver tumors in transgenic zebrafish can regress in tp53 null mutation. PLoS One, 2015;10(1):e0117249. 53. Ruggieri, A., Barbati, C., Malorni, W., and Malorni, W. Cellular and molecular mechanisms involved in hepatocellular carcinoma gender. Int J Cancer, 2010;127(3):499–504. 54. Yang, Q., Yan, C., Yin, C., and Gong, Z. Serotonin activated hepatic stellate cells contribute to sex disparity in hepatocellular carcinoma. Cell Mol Gastroenterol Hepatol, 2017;3(3):484–99. 55. Yang, Q., Yan, C.A., and Gong, Z. Activation of liver stromal cells is associated with male‐biased liver tumor initiation in xmrk and Myc transgenic zebrafish. Sci Rep, 2017;7(1):10315. 56. Lam, S.H., Wu, Y., Vega, V.B. et al. Conservation of gene expression signatures between zebrafish and human liver. Nat Biotechnol, 2006;24(1):73–5. 57. El‐Serag, H.B. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology, 2012;142(6):1264–73.e1. 58. Habermann, R.T., Williams, F.P., Jr., and Eyestone, W.H. Spontaneous hepatomas in two woodchucks and a carcinoma of the testis in a badger. J Am Vet Med Assoc, 1954;125(931):295–8. 59. Tennant, B.C., Toshkov, I.A., Peek, S.F. et al. Hepatocellular carcinoma in the woodchuck model of hepatitis B virus infection. Gastroenterology, 2004;127(5 Suppl 1):S283–93. 60. Liaw, Y.F. Hepatitis B virus replication and liver disease progression: the impact of antiviral therapy. Antivir Ther, 2006;11(6):669–79. 61. Fletcher, S.P., Chin, D.J., Ji, Y. et al. Transcriptomic analysis of the woodchuck model of chronic hepatitis B. Hepatology, 2012;56(3):820–30. 62. Shope, R.E. and Hurst, E.W. Infectious papillomatosis of rabbits: with a note on the histopathology. J Exp Med, 1933;58(5):607–24. 63. Ko, Y.H., Pedersen, P.L., and Geschwind, J.F. Glucose catabolism in the rabbit VX2 tumor model for liver cancer: characterization and targeting hexokinase. Cancer Lett, 2001;173(1):83–91. 64. Kuszyk, B.S., Boitnott, J.K., Choti, M.A. et al. Local tumor recurrence following hepatic cryoablation: radiologic‐histopathologic. Radiology, 2000; 217(2):477–86. 65. Verna, L., Whysner, J., and Williams, G.M. N‐Nitrosodiethylamine mechanistic data and risk assessment: bioactivation. Pharmacol Ther, 1996;71(1–2):57–81. 66. Qi, Y., Chen, X., Chan, C.‐Y., et al. Two‐dimensional differential gel electrophoresis/analysis of diethylnitrosamine. Int J Cancer, 2008;122(12):2682–8. 67. Waxman, D.J. and Azaroff, L. Phenobarbital induction of cytochrome P‐450 gene expression. Biochem J, 1992;281(Pt 3):577–92. 68. Chen, B., Liu, L., Castonguay, A. et al. Dose‐dependent ras mutation spectra in N‐nitrosodiethylamine induced mouse liver. Carcinogenesis, 1993; 14(8):1603–8. 69. Reddy, J.K., Rao, S., and Moody, D.E. Hepatocellular carcinomas in acatalasemic mice treated with nafenopin, a hypolipidemic peroxisome proliferator. Cancer Res, 1976;36(4):1211–17.



58:  Experimental Models of Liver Cancer: Genomic Assessment of Experimental Models

70. Misra, P. and Reddy, J.K. Peroxisome proliferator‐activated receptor‐alpha activation and excess energy. Biochimie, 2014;98:63–74. 71. Hasmall, S.C., James, N.H., Macdonald, N. et  al. Suppression of mouse hepatocyte apoptosis by peroxisome proliferators: role of PPARalpha and TNFalpha. Mutat Res, 2000;448(2):193–200. 72. Morimura, K., Cheung, C., Ward, J.M., Reddy, J.K., and Gonzalez, F.J. Differential susceptibility of mice humanized for peroxisome proliferator‐ activated receptor alpha to Wy‐14,643‐induced liver tumorigenesis. Carcinogenesis, 2006;27(5):1074–80. 73. Meyer, N. and Penn, L.Z. Reflecting on 25 years with MYC. Nat Rev Cancer, 2008;8(12):976–90. 74. Huang, H., Weng, H., Zhou, H., and Qu, L. Attacking c‐Myc: targeted and combined therapies for cancer. Curr Pharm Des, 2014;20(42):6543–54. 75. Thorgeirsson, S.S., and Santoni‐Rugiu, E. Transgenic mouse models in carcinogenesis: interaction of c‐myc with transforming. Br J Clin Pharmacol, 1996;42(1):43–52. 76. Conner, E.A., Lemmer, E.R., Sanchez, A., Factor, V.M., and Thorgeirsson, S.S. E2F1 blocks and c‐Myc accelerates hepatic ploidy in transgenic mouse models. Biochem Biophys Res Commun, 2003;302(1):114–20. 77. Calvisi, D.F., Conner, E.A., Ladu, S. et al. Activation of the canonical Wnt/ beta‐catenin pathway confers growth advantages in c‐Myc/E2F1 transgenic mouse model of liver cancer. J Hepatol, 2005;42(6):842–9. 78. Breuhahn, K., Longerich, T., and Schirmacher, P. Dysregulation of growth factor signaling in human hepatocellular carcinoma. Oncogene, 2006;25(27): 3787–800. 79. Jhappan, C., Stahle, C., Harkins, R.N. et  al. TGF alpha overexpression in transgenic mice induces liver neoplasia and abnormal. Cell, 1990;61(6): 1137–46. 80. Sawey, E.T., Chanrion, M., Cai, C. et al. Identification of a therapeutic strategy targeting amplified FGF19 in liver. Cancer Cell, 2011;19(3):347–58. 81. Miura, S., Mitsuhashi, N., Shimizu, H. et  al. Fibroblast growth factor 19 expression correlates with tumor progression and poorer prognosis of hepatocellular carcinoma. BMC Cancer, 2012;12:56. 82. Nicholes, K., Guillet, S., Tomlinson, E. et al. A mouse model of hepatocellular carcinoma: ectopic expression of fibroblast. Am J Pathol, 2002;160(6): 2295–307. 83. Nusse, R. and Clevers, H. Wnt/beta‐catenin signaling, disease, and emerging therapeutic modalities. Cell, 2017;169(6):985–99. 84. Tan, X., Apte, U., Micsenyi, A. et  al. Epidermal growth factor receptor: a novel target of the Wnt/beta‐catenin pathway. Gastroenterology, 2005;129(1): 285–302. 85. Cadoret, A., Ovejero, C., Saadi‐Kheddouci, S. et al. Hepatomegaly in transgenic mice expressing an oncogenic form of beta‐catenin. Cancer Res, 2001;61(8):3245–9. 86. Harada, N., Oshima, H., Katoh, M. et al. Hepatocarcinogenesis in mice with beta‐catenin and Ha‐ras gene mutations. Cancer Res, 2004;64(1):48–54. 87. Ali, S.H. and DeCaprio, J.A. Cellular transformation by SV40 large T antigen: interaction with host proteins. Semin Cancer Biol, 2001;11(1):15–23. 88. Held, W.A., Mullins, J.J., Kuhn, N.J. et al. T antigen expression and tumorigenesis in transgenic mice containing a mouse major urinary protein/SV40 T antigen hybrid gene. EMBO J, 1989;8(1):183–91. 89. Dubois, N., Bennoun, M., Allemand, I. et al. Time‐course development of differentiated hepatocarcinoma and lung metastasis transgenic mice. J Hepatol, 1991;13(2):227–39. 90. Saucedo, L.J. and Edgar, B.A. Filling out the Hippo pathway. Nat Rev Mol Cell Biol, 2007;8(8):613–21.

757

  91. Zender, L., Spector, M.S., Xue, W. et  al. Identification and validation of oncogenes in liver cancer using an integrative oncogenomic approach. Cell, 2006;125(7):1253–67.   92. Dong, J., Feldmann, G., Huang, J. et al. Elucidation of a universal size‐control mechanism in Drosophila and mammals. Cell, 2007;130(6):1120–33.   93. Lu, L., Li, Y., Kim, S.M. et al. Hippo signaling is a potent in vivo growth and tumor suppressor pathway in the mammalian liver. Proc Natl Acad Sci U S A, 2010;107(4):1437–42.   94. Zhou, D., Conrad, C., Xia, F. et  al. Mst1 and Mst2 maintain hepatocyte quiescence and suppress hepatocellular carcinoma development through inactivation of the Yap1 oncogene. Cancer Cell, 2009;16(5):425–38.   95. Park, Y.Y., Sohn, B.H., Johnson, R.L. et  al. Yes‐associated protein 1 and transcriptional coactivator with PDZ‐binding motif activate the mammalian target of rapamycin complex 1 pathway by regulating amino acid transporters in hepatocellular carcinoma. Hepatology, 2016;63(1):159–72.   96. Sohn, B.H., Shim, J.J., Kim, S.B. et al. Inactivation of Hippo pathway is significantly associated with poor prognosis in hepatocellular carcinoma. Clin Cancer Res, 2016;22(5):1256–64.   97. Simile, M.M., Latte, G., Demartis, M.I. et al. Post‐translational deregulation of YAP1 is genetically controlled in rat liver cancer and determines the fate and stem‐like behavior of the human disease. Oncotarget, 2016; 7(31):49194–216.   98. Li, M., Zhao, H., Zhang, X. et al. Inactivating mutations of the chromatin remodeling gene ARID2 in hepatocellular. Nat Genet, 2011;43(9):828–9.   99. Wilson, B.G. and Roberts, C.W. SWI/SNF nucleosome remodellers and cancer. Nat Rev Cancer, 2011;11(7):481–92. 100. Mathur, R., Alver, B.H., San Roman, A.K. et  al. ARID1A loss impairs enhancer‐mediated gene regulation and drives colon cancer in mice. Nat Genet, 2017;49(2):296–302. 101. Sun, X., Wang, S.C., Wei, Y. et al. Arid1a has context‐dependent oncogenic and tumor suppressor functions in liver. Cancer Cell, 2017;32(5):574–89. 102. Lee, J.S., Chu, I.S., Mikaelyan, A. et al. Application of comparative functional genomics to identify best‐fit mouse models to study human cancer. Nat Genet, 2004;36(12):1306–11. 103. Lee, J.S., Heo, J., Libbrecht, L. et al. A novel prognostic subtype of human hepatocellular carcinoma derived from hepatic progenitor cells. Nat Med, 2006;12(4):410–16. 104. Woo, H.G., Lee, J.H., Yoon, J.H. et al. Identification of a cholangiocarcinoma‐like gene expression trait in hepatocellular carcinoma. Cancer Res, 2010;70(8):3034–41. 105. Woo, H.G., Park, E.S., Cheon, J.H. et al. Gene expression‐based recurrence prediction of hepatitis B virus‐related human hepatocellular carcinoma. Clin Cancer Res, 2008;14(7):2056–64. 106. Lee, J.S., Chu, I.S., Mikaelyan, A. et al. Application of comparative functional genomics to identify best‐fit mouse models to study human cancer. Nat Genet, 2004;36(12):1306–11. 107. Yim, S.Y., Shim, J.J., Shin, J.H. et al. Integrated genomic comparison of mouse models reveals their clinical resemblance to human liver cancer. Mol Cancer Res, 2018;16(11):1713–23. 108. Yamashita, T., Ji, J., Budhu, A. et al. EpCAM‐positive hepatocellular carcinoma cells are tumor‐initiating cells with stem/progenitor cell features. Gastroenterology, 2009;136(3):1012–24. 109. Saha, S.K., Parachoniak, C.A., Ghanta, K.S. et  al. Mutant IDH inhibits HNF‐4 alpha to block hepatocyte differentiation and promote biliary cancer. Nature, 2014;513(7516):110–14. 110. Yimlamai, D., Christodoulou, C., Galli, G.G. et al. Hippo pathway activity influences liver cell fate. Cell, 2014;157(6):1324–38.

59

Epidemiology of Hepatocellular Carcinoma Hashem B. El‐Serag Department of Medicine, Baylor College of Medicine and Michael E. DeBakey Veterans Affairs Medical Center, Houston, TX, USA

TRENDS IN HEPATOCELLULAR CARCINOMA Global epidemiology Liver cancer, of which hepatocellular carcinoma (HCC) is the dominant variety, represents the fourth leading cause of cancer deaths worldwide. Globally, it ranks sixth in cancer incidence and second in the absolute years of life lost. The incidence and mortality of liver cancer vary by region. The age‐standardized incidence rates of liver cancer are highest in Asia Pacific, East Asia, and central sub‐Saharan Africa and lowest in southern Latin America and tropical Latin America. Liver cancer ranks first in cancer deaths in Egypt and Thailand, whereas it ranks 14th in Ukraine and Poland [1]. The geographic map of the incidence and mortality of HCC has changed over time. Regions that traditionally have lower incidence of HCC, such as North America and certain areas of Europe, have seen increased incidence, and regions that are traditionally high risk, such as Japan and China, have shown declining rates. This observation is likely due to changes in exposure to HCC risk factors, including chronic hepatitis C virus (HCV) or hepatitis B virus (HBV) infection, heavy alcohol consumption, diabetes, and non‐alcoholic fatty liver disease (NAFLD) [2, 3]. In low‐risk countries, increases in the rates of obesity, diabetes, and NAFLD threaten gains in reducing the incidence of HCC, whereas in high‐risk countries, aggressive HBV vaccination and or aflatoxin‐reduction programs and the use of antivirals to treat hepatitis B and C infections have reduced HCC incidence [2, 4]. Environment could play a role in regional and temporal variations in HCC incidence. This hypothesis is built in part on rates of HCC in migrant populations. Migrants moving from low‐risk African, Asian, or South American countries to Western Europe

had increased risk of HCC compared with their originating ­countries [5]. Among Chinese expats, HCC risk decreased compared with that of Chinese population; however, the risk was still higher than that of the host country. Even within similar geographic regions there can be wide variations, such as the 75.5% difference in age‐standardized incidence rate observed between persons from Bhopal and Dindigul, India and from Qidong City, China [2]. The environmental factors contributing to these differences in HCC risk have yet to be clearly identified.

Sex differences in rates of HCC In general, HCC is more prevalent in men than women (Figure 59.1). For men, liver cancer was the most common cancer diagnosis in 11 countries and cause of cancer death in 40 countries. By contrast, Mongolia was the only country in which liver cancer was the most common cancer diagnosis for women, and liver cancer was the most common cause of cancer death for women in only five countries [1]. Typically, the rate of liver cancer is two to three times higher in men than women [6]. There are some exceptions; several countries have HCC rates in women that approach those of men, but the risk factors associated with the increased incidence in women have yet to be identified [4]. Notably, there is no correlation between gender disparity and the level of HCC risk of the geographical region [2]. There are several hypotheses, including differences in behaviors, alcohol consumption, immune responses, and epigenetics [7–9], regarding the source of this disparity. One particularly strong hypothesis involves differences in endogenous sex hormones. A prospective cohort study of more than 9000 men in Taiwan found that the relative risk of HCC increased between 50% and 400% for men in upper tertile of testosterone level compared with the middle and lowest tertiles [10, 11]. Several studies of CAG repeats in exon

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



59:  Epidemiology of Hepatocellular Carcinoma

759 Male:Female ratio

Thailand China Japan Philippines Singapore Italy France Switzerland Uganda Spain Croatia Germany Austria United States Brazil Czech Republic Slovenia Slovakia Latvia Costa Rica New Zealand Finland Australia Russia Poland United Kingdom Denmark Colombia India Ecuador Canada Estonia Sweden Iceland Malta Israel Norway The Netherlands

2.7:1 3.4:1 3.0:1 2.9:1 3.8:1 3.0:1 5.5:1 4.8:1 1.3:1 3.6:1 2.8:1 2.6:1 3.1:1 3.2:1 2.3:1 2.5:1 3.4:1 2.7:1 3.1:1 1.6:1 2.7:1 2.5:1 2.9:1 2.6:1 2.0:1 2.2:1 2.6:1 1.3:1 2.4:1 1.3:1 2.7:1 2.1:1 2.0:1 2.5:1 2.6:1 2.4:1 2.2:1 2.6:1 30

20

10 Male Age-Adjusted Rate

0

10 Female Age-Adjusted Rate

20

Figure 59.1  Age‐adjusted incidence rates of hepatocellular carcinoma in men and women separately. The men:women incidence ratios are also shown. Data from [1].

1 of the androgen receptor and its role in HCC have been conducted. In one study, men with chronic HBV infection who had fewer CAG repeats (≤20) and increased testosterone levels had a fourfold increase in HCC risk compared with men with more repeats [12]. A similar study was performed in women both with and without chronic HBV. Those with two androgen receptor alleles with more than 23 repeats had a higher risk of HCC compared with those who had two short alleles or a short and long allele, and the risk was increased further when the woman was an HBV carrier [9]. Transgenic mouse studies investigating the role of the androgen receptor in HBV infections demonstrated that increased androgen pathway signaling can increase HBV gene transcription [13], indicating that the combination of chronic HBV infection, a substantial risk factor for HCC, and sex hormone levels could contribute to the apparent gender disparity in the incidence of HCC.

United States trends in incidence In the United States, the number of cancer‐related deaths attributed to HCC is rising rapidly. One study of the United States Cancer Statistics Registry reported the age‐adjusted incidence rate increased by 2.3/100 000 persons between 2000 and 2012.

When broken down by year, these changes in incidence increased approximately 4.5% between 2000 and 2009 and began plateauing in 2010 [14]. The Centers for Disease Control have reported similar increases in mortality through 2015. Further investigation of the registry showed that the populations experiencing the greatest increase in incidence were men aged 55–64 years and in individuals of Hispanic, African American, and white ethnicity [15]. Notably, the overall age‐adjusted incidence rates for HCC have skyrocketed among persons of Hispanic ethnicity, surpassing even that of persons of Asian race (Figure 59.2). The trend is amplified among the Hispanic population born in the United States. The higher rates of HCC in this population have been attributed to a higher incidence of HCV [16] as well as increased rates of alcoholic liver disease, NAFLD, metabolic syndrome, and diabetes [17, 18], all of which are risk factors for developing HCC. Among Hispanic patients with chronic HCV or NAFLD, the risk of progression to cirrhosis and HCC is higher. This observation has been attributed in part to a genetic predisposition to cirrhosis, such as that associated with the phospholipase domain‐containing 3 gene (PNPLA3) ­polymorphism (discussed later) [19].

760

THE LIVER:  RISK FACTORS FOR HCC 16

Age-adjusted rate per 100,000

14 12 Hispanic

10

White

8

Black API

6

AI/AN

4 2

12

11

20

10

20

09

20

08

20

07

20

06

20

05

20

04

20

03

20

02

20

01

20

20

20

00

0

Year

Figure 59.2  Yearly age‐adjusted incidence rates of hepatocellular carcinoma in the United States between 2000 and 2012, broken down by race and ethnicity. API, Asian Pacific Islanders; AI/AN, American Indian or Alaska Native. Modified from [14].

RISK FACTORS FOR HCC Hepatitis B virus HBV is a double‐stranded DNA virus that preferentially infects hepatocytes. Infections can be acute or chronic, the latter being a risk factor for HCC. The risk of the infection turning chronic increases with earlier age at infection. Among adults who are infected via sexual contact, needle stick, or transfusion, fewer than 10% develop chronic infections, whereas among infants infected by vertical transmission from HBV‐infected mothers, 80–90% develop chronic infections [20]. With chronic infections, HBV integrates into the host genome. This integration is a precursor to hepatocarcinogenesis. HBV induces chronic necroinflammatory disease, a continuous cycle of hepatocyte necrosis and regeneration, which promotes mutations in liver cells and leads to HCC [20, 21]. When evaluating HCC tissue from individuals with chronic HBV, HBV DNA is integrated into the genome in nearly all cases  [2]. Worldwide, chronic HBV accounts for 50% of HCC cases  [20]. However, this percentage varies by country. In areas with endemic HBV infections, such as Asia and sub‐Saharan Africa, HBV accounts for more than 70% of HCC cases [22]. By contrast, HBV only accounts for 10–15% of HCC cases in the United States [23]. It is estimated that 350–400 million individuals have a chronic HBV infection – approximately 5% of the global population [24]. In the United States, approximately 0.33% of the population are thought to have chronic HBV based on data from the National Health and Nutrition Examination Surveys between 1988 and 1994. Those analyses showed that there was a higher prevalence of chronic HBV carriers among men compared with women and among non‐Hispanic blacks compared with non‐ Hispanic whites and Mexican Americans [25]. Most individuals with chronic HBV have inactive infections. However, 10–30% develop active infections. Cirrhosis develops in approximately 1–2% of those with chronic active infections. Among those with cirrhosis, 1–6% per year will progress to HCC. The time between infection and HCC development is estimated to be in the range of 30–50 years [20].

Relative risk of HCC from HBV The relative risk of HCC from HBV is consistently quite high. One meta‐analysis looked at case–control and prospective studies evaluating individuals positive for HBsAg, a marker of HBV infection. When individuals who were HBsAg positive were compared with those who were HBsAg negative, the authors found that the lifetime relative risk of HCC was 15–20 [26]. When evaluating case–control studies from across the globe, the estimated odds ratios of developing HCC ranged from 5 to 65  among individuals with chronic HBV. This observation is supported by prospective studies of HBV carriers, who had relative risks for HCC of 5–103 compared with noncarriers [2].

Absolute risk of HCC The World Health Organization estimates that the lifetime risk of developing HCC in individuals carrying HBV ranges from 10 to 25% [2]. Based on cohort studies, the incidence of HCC in patients with HBV seems to vary based on the level of cirrhosis. Those with inactive disease had an HCC incidence of 0.02–0.2 per 100 person‐years, those with active chronic infections but no cirrhosis had an incidence of 0.3–0.6 per 100 person‐years, and those with compensated cirrhosis had an incidence of 2.2–3.7 per 100 person‐years [27, 28]. Much of our information on the risk of HCC among individuals with HBV comes from areas where HBV is endemic and the individuals evaluated were infected in infancy or early childhood [29]. In a multicenter study from the United States, where most individuals are infected later in life, the annual incidence was estimated to be 0.42%. However, more than 50% of the evaluated cohort were Asian Pacific Islanders [30]. Another recent study evaluated a cohort of more than 8000 primarily male patients with chronic HBV infections identified from the United States Veterans Health Administration (VHA) database between 2001 and 2013. Among that cohort, the annual incidence rate of HCC was 0.65% for American Pacific Islanders, 0.57% for whites, and 0.40% for African Americans [31].



59:  Epidemiology of Hepatocellular Carcinoma

Determinants of risk associated with HBV Increased risk Numerous risk factors for HCC have been identified for individuals with chronic HBV infections. Demographics (e.g. male sex, older age, Asian or African ancestry, family history of HCC), clinical features (e.g. cirrhosis), viral factors (e.g. high HBV replication levels, HBV genotype, infection duration, coinfection), and environmental exposures (e.g. aflatoxin exposure, alcohol and tobacco use) all contribute to the development of HCC [27]. With respect to demographics, a 90 000‐participant prospective cohort study showed that male patients with chronic HBV have a higher risk of both cirrhosis and HCC than female patients [32]. Age has been shown to play a role in HCC risk, both in general and based on race. Studies conducted in East Asia showed a marked increased risk of HCC beyond 40 years of age [31]. Studies in South Africa from the 1970s and 1980s reported that African black patients with chronic HBV could develop HCC at less than 40 years of age [33, 34], and these data were confirmed in a recent study evaluating HCC in multiple African countries [35]. Data from the VHA cohort described above showed that HCC risk increased with age. In that cohort, adjusted hazard ratios for HCC were 1.97 for individuals 40–49 years of age, 3.00 for those 50–59 years of age, and 4.02 for those older than 60 years of age compared with individuals younger than 40 years of age. Notably, the risk of HCC was extremely low in individuals younger than 40 years of age, including those who were African American [31], indicating that in the United States, African American race may not be as high a risk factor for early HCC in the absence of other risk factors. Cirrhosis is the primary clinical risk factor for HCC. In patients with chronic HBV, 70–90% of HCC cases arise from cirrhotic liver [23]. As mentioned earlier, the incidence of HCC in patients with HBV increases substantially with the progression of cirrhosis. This observation was supported by the results of the VHA cohort study, which found an adjusted hazard ratio of 3.69 for patients with cirrhosis compared with those without cirrhosis [31]. Several viral factors have been shown to increase the risk of HCC. The effects of viral load have been investigated in a prospective cohort of HBsAg‐positive participants in a Taiwanese study. The authors found a proportional increase in the incidence of cirrhosis and HCC with increasing serum levels of HBV DNA [36]. HBV genotype also factors into HCC risk. In studies of Asian populations, the association between genotype and severe liver disease, cirrhosis, and HCC is greater for genotype C than genotype B. Among studies out of North America and Western Europe, the incidences of severe liver disease and HCC were greater for genotype D than genotype A [37]. Population‐based studies have shown that the risk of HCC associated with genotype C is higher than that of genotypes A2, Ba, Bj, and D [38]. When looking deeper into the viral genetics, studies have shown that specific mutations in the basal core promotor and precore region affect the incidence of HCC. In the case of the former, mutations at T1762 and A1764 have been shown to increase incidence [39]. In the case of the latter, infection with the G1896A mutation is associated with decreased incidence [40]. In addition to the virus itself, coinfection with

761

other viruses, such as HCV and HIV, can have substantial effects on risk. The extent of the effect is somewhat controversial. With respect to HCV coinfection, older meta‐analyses support an additive effect on HCC risk [41, 42]. Studies conducted more recently, including those in countries where coinfection is uncommon, have reported sub‐additive effects on HCV [26]. Previous studies have noted a negative correlation between levels of HBsAg and anti‐HCV antibody in serum, indicating viral interference [42, 43]. This observation could account for the sub‐additive effects observed. Environmental factors that increase risk include both behavioral factors (alcohol and tobacco) and exogenous exposures. Alcohol is hepatotoxic but in itself is not mutagenic. Heavy alcohol consumption can compound damages already incurred from chronic HBV infection and is thought to have a greater than additive effect on the development of HCC [44]. For example, a population cohort study in Korea noted that the relative risk of HCC increased significantly with increasing daily alcohol intake [45]. The effect of cigarette use is smaller, but a significant positive interaction has been found between smoking and the risk of HCC in individuals with HBV [46, 47]. One known exogenous environmental risk factor for HCC is aflatoxin B1, a mycotoxin generated by Aspergillus sp. that is found in food items stored under warm, damp conditions. Aflatoxin is known to mutate the p53 tumor suppressor, and this mutation is often found in HCC tumors from individuals with HBV infections who live in aflatoxin‐endemic regions [8]. When evaluating the risk of HCC based on aflatoxin exposure and HBV infection, the HCC risk for aflatoxin exposure alone is fourfold greater, for chronic HBV carriers it is sevenfold greater, and for aflatoxin and HBV combined it is 60‐fold greater than unexposed individuals, indicating a synergistic association between the two [2]. All of these risk factors have been translated into scoring systems to identify individuals with chronic HBV for HCC surveillance. The American Association for the Study of Liver Disease devised a surveillance system based on cirrhosis, family history of HCC, age, and race. Specifically, the guidelines recommend surveillance for all patients with cirrhosis, for Asian men older than 40 years and Asian women older than 50 years in the absence of cirrhosis, and at younger ages for Africans and African Americans [48, 49]. Several other scoring systems developed for Asian patients with HBV infections have been validated and yielded high 3‐ to 10‐year negative predictive values for HCC development. These scoring systems are based on age, sex, cirrhosis, viral load, HBsAg status, and alanine aminotransferase levels [50–53]. The REACH‐B score is highly effective at identifying at‐risk patients who are not cirrhotic. The GAG‐HC and CU‐HCC scoring systems incorporate cirrhosis and HBV core promoter mutations into the calculations and could potentially be applicable for non‐Asian populations [3]. However, these scoring systems are not adequate for predicting HCC risk in patients with HBV who are successfully undergoing antiviral therapy [54, 55]. Among patients with HBV on nucleoside analog therapy, age, sex, cirrhosis, liver stiffness, platelet count, and diabetes are known HCC risk factors [56–58]. Notably, pretreatment viral load, HBsAg status and quantity, and aminotransferase level are not predictive of HCC risk for patients treated with nucleoside analogs [36, 59, 60]. To address the unmet need for an HCC risk scoring system

762

THE LIVER:  RISK FACTORS FOR HCC

in this population, the Cirrhosis, Age, Male sex, Diabetes (CAMD) score was developed. The model was based on data from Taiwanese patients in the national healthcare database who received entecavir or tenofovir to treat chronic HBV and was externally validated using data extracted from the Hong Kong national healthcare database [61]. This system has a high level of accuracy and has been validated in Caucasian and Asian populations [62].

Decreased risk The most effective way to decrease the risk of HCC associated with HBV infection is to not get the infection in the first place. The development of an HBV vaccine has made excellent strides toward reducing HBV infection rates in endemic areas, which could reduce the incidence of HCC long term [63]. A large randomized controlled trial in China found that HBV vaccination reduced the risk of developing primary liver cancer by 84% over 25 years [3]. This effect is starting to be seen in other areas where vaccination programs have been implemented [63]. The 30‐year report following the implementation of a universal HBV vaccination program in Taiwan found that sex‐adjusted HCC incidence declined 80% and sex‐adjusted HCC mortality declined 92% in cohorts born after the vaccination program began compared with those born before it began [64]. Vaccination programs have been implemented in most Asian and Eastern European countries, Spain, and many sub‐Saharan countries. Many of the countries that implemented their programs in the 1980s, such as China, Singapore, and Spain are seeing reductions in the prevalence of HBsAg that are similar to those seen in Taiwan, and the HCC incidence in these countries is expected to decrease at similar rates [2, 63]. Many of the countries that have implemented vaccination programs are places where aflatoxin exposure is also endemic. Aflatoxin abatement programs have the potential to further decrease HCC risk. One such program was implemented in Qidong, China, an area with one of the highest rates of liver cancer. This program was initiated at approximately the same time as the government began a slow rollout of an HBV vaccination program. In cohorts with reduced aflatoxin exposure and minimal vaccination ­penetrance, the liver cancer mortality decreased by 45% and age‐ specific liver cancer incidence decreased 1.2‐ to 4‐fold. In cohorts where reduced aflatoxin exposure was combined with higher vaccine penetrance, the age‐specific liver cancer incidence decreased 14‐fold, indicating that a combination of aflatoxin reduction and HBV vaccination has the potential to make meaningful reductions in HCC incidence in these areas [65, 66]. Although HBV vaccination can prevent future cases of HCC, the primary intervention in patients with HBV is antiviral treatment. Randomized controlled trials have provided evidence that antiviral treatment can achieve sustained reductions in HBV DNA levels as well as improve liver function and histology, but the evidence that these agents meaningfully affect long‐term clinical outcomes or risk of HCC is limited [3]. The primary drugs used to treat chronic HBV infection are nucleoside a­nalogs (NAs). The five NAs with regulatory agency approval – ­lamivudine, telbivudine, adefovir, entecavir, and tenofovir disoproxil  –  can suppress HBV DNA levels in nearly all patients [67]. Lamivudine treatment is associated with reduced progression to cirrhosis, lower Child–Pugh scores, and a 50% reduction in the rate of HCC

[68]. However, because lamivudine is associated with drug resistance, the first‐line therapies are currently entecavir and tenofovir disoproxil [67]. Increasing evidence suggests that NA treatment can reduce but not eliminate the medium‐term risk of HCC [56, 69–71]. A meta‐analysis evaluating cohort studies and clinical trials that investigated NAs, primarily lamivudine, and HCC risk found that HCC incidence was significantly lower in NA‐treated patients compared with untreated controls. Furthermore, patients with lamivudine‐resistant HBV had a significantly higher risk of HCC compared with NA‐naïve patients, and patients with lamivudine‐resistant HBV who achieved virological response following rescue therapy did not have a significant reduction in HCC risk. Collectively, these data suggest that long‐term viral suppression is critical for HCC risk reduction [71]. In a randomized controlled trial investigating lamivudine versus placebo in ­ Taiwanese patients with chronic HBV and cirrhosis or advanced fibrosis, HCC incidence was significantly lower in the lamivudine group. However, this trial was terminated early due to the significant response (median treatment duration, 32.4 months) [68]. There are fewer studies evaluating risk reduction in the newer NAs. A retrospective nationwide study out of Taiwan identified 21 595 matched patients with chronic HBV who either received NA therapy (lamivudine or entecavir) or were untreated. Among patients who received entecavir, the 7‐year HCC incidence was significantly lower than in untreated patients (7.3% vs. 22.7%, respectively) [72]. A Japanese study comparing entecavir‐treated patients with a matched historical control of untreated patients found significantly lower 5‐year HCC incidence rates for the entecavir‐treated group (3.7% vs. 13.7%, respectively) [73]. A study investigating entecavir or tenofovir treatment in Caucasian patients with chronic HBV found that 5‐year HCC risk reduced in patients with cirrhosis following treatment, but the overall risk was still higher than that in patients without cirrhosis. They also found that all cases of HCC in their cohort developed in patients who began NA treatment after the age of 50 years [7]. In a multinational European cohort study of clinical outcomes of patients with chronic HBV treated with entecavir, patients who had a virologic response were 71% less likely to develop hepatic decompensation or HCC or to die than those who did not have a response, but this effect was only significant in patients with cirrhosis. However, this study had a small sample size and short follow‐up, so strong conclusions cannot be drawn from these data [74]. In addition to NAs, interferons can be used to treat HBV, although they are less used due to side‐effects and associated patient compliance. Several meta‐analyses have evaluated the effect of interferon treatment on HCC risk in patients with chronic HBV, and they suggest that interferon treatment reduces HCC incidence in those patients who can sustain a virological response [75–77]. Overall, the data support that patients with chronic HBV who receive treatment have a lower incidence of HCC. However, these treatments do not eliminate the virus, so continuous treatment is likely needed to sustain the effect until new treatments capable of viral eradication are developed [3].

Hepatitis C virus HCV is a positive‐sense RNA virus that primarily infects liver tissues [20]. There are several HCV genotypes and subtypes with different ethnic and geographic distributions [27, 78].



59:  Epidemiology of Hepatocellular Carcinoma

An estimated 75–85% of HCV infections become chronic based on the continued presence of HCV RNA, an HCV biomarker, in serum [16]. The association between chronic HCV infection and HCC has been firmly established [20]. Unlike HBV, which integrates in the genome to initiate tumorigenesis, HCV does not integrate into the genome and is unlikely to initiate tumorigenesis itself. Because as much as 90% of HCV‐associated HCC cases are preceded by cirrhosis, it is likely that HCV promotes tumorigenesis through the repetitive killing of hepatocytes by HCV and the subsequent cellular regeneration that leads to cirrhosis. HCV infection can also alter lipid metabolism, leading to liver steatosis [79]. When considering the burden of HCV in HCC incidence, estimates vary based on geography, largely due to the geographic differences in prevalence [22]. It is estimated that HCV accounts for approximately 20% of HCC cases worldwide [8] and 40–50% of new HCC cases in the United States [23]. A model based on the population of HCV‐infected individuals in the United States estimated that the number of HCV‐associated HCC cases increased by 130% when comparing the cases occurring between 1990 and 1999 with those occurring between 2000 and 2009. The model projected that the number of cases would increase an additional 50% for cases occurring between 2010 and 2019 [80]. Similar trends were seen in a cohort of veterans diagnosed with HCV. In that study, the prevalence of HCC increased 19‐fold between 1996 and 2006 [81]. This observation could be due to the high prevalence of HCV infection found in persons born between 1945 and 1965 [4]. Accordingly, the numbers of HCV‐associated HCC are anticipated to continue to rise and to peak in 2020 [63].

Relative and absolute risk of HCC from HCV The estimates of relative risk of HCC in persons with HCV compared with those who are uninfected vary in cohort studies between 10 and 20 [23, 41, 63, 82–84]. HCV‐related HCC cases typically arise from patients with advanced fibrosis or cirrhosis [82, 83]. The annual incidence of HCC in patients with HCC‐induced cirrhosis ranges from 0.5 to 10% [85, 86]. At 25–30 years post infection, the incidence of cirrhosis ranges from 15 to 35% [87, 88].

Determinants of risk associated with HCV Increased risk There are several known risk factors for HCC in patients with HCV, including sex, race, HCV genotype, coinfections with HBV or HIV, insulin resistance, obesity, diabetes, and alcohol consumption [27, 85, 86, 89]. When compared with non‐ Hispanic white patients, Hispanic patients have a higher risk of cirrhosis and HCC development and African American patients have a lower risk. The detection of HCV RNA in serum, which indicates viremia, is a strong risk factor for HCC. Some studies reported correlations between viral load and HCC; however, other studies were unable to reproduce this correlation [3]. The HCV genotype represents a substantial risk factor for HCC in patients with HCV. A meta‐analysis of studies investigating genotype 1b as a risk factor for HCC found that patients infected with genotype 1b had a 78% greater risk of HCC than those infected with other genotypes and a 60% greater risk among patients who

763

also had cirrhosis [90]. However, genotype 3 has been consistently and strongly associated with a higher risk of both cirrhosis and HCC in studies based in the United States [86]. Patients with genotype 3 infections also have the highest risk of cirrhosis and HCC, with incidence rates of 30 per 1000 person‐years and 7.9 per 1000 person‐years, respectively [86, 91]. Genotype 1 infections tend to respond well to antiviral treatments compared with other genotypes [8], whereas genotype 3 is more difficult to treat, so genotype 3 is considered the greater risk [86, 91]. Patients with HCV/HIV coinfections have been shown to progress to cirrhosis and decompensated liver disease faster than those with an HCV mono‐infection. Further, the risk increases as immunosuppression increases, indicating that uncontrolled HIV infection exacerbates the pathological damage caused by HCV [92, 93]. Patients with HCV who are insulin resistant have a higher risk of hepatic steatosis, advanced fibrosis, and HCC. Notably, several retrospective and prospective studies showed that patients with HCV have an approximately 68% increased incidence of diabetes compared with uninfected persons [94]; however, there is little evidence of a synergistic effect between HCV and diabetes with respect to HCC risk [94, 95]. The implications of insulin resistance and diabetes in hepatocellular carcinoma risk are discussed in detail later. Environmental factors can also affect the risk of HCC. Studies have demonstrated a synergistic effect between heavy alcohol consumption and HCV infection [44, 96]. Patients with HCV who are also heavy drinkers have a twofold increased risk of HCC compared with persons who are heavy drinkers alone [44]. A similar relationship has been identified between cigarette smoking and HCC risk in subgroups of patients with HCV [46]. There is some evidence of a synergistic effect between HCV and aflatoxin on HCC risk, but the scope of this issue is limited because HCV infections are less common in aflatoxin‐endemic areas [2, 97].

Decreased risk As mentioned above, the cohort of persons born between 1945 and 1965 who have a high prevalence of HCV infections have contributed to the rise in HCV‐related HCC [4, 63]. The main factor that will likely decrease HCC incidence in this cohort is sustained virologic response (SVR) achieved via directly acting antiviral (DAA) treatments [98]. SVR is the primary modifier of HCC risk among patients with HCV. SVR has been shown to slow the progression of liver disease and reduce cirrhosis and HCC risk [95]. A meta‐analysis of interferon‐based therapies, the predecessor to DAAs, found that SVR following treatment reduced the risk of HCC 76% at all stages of liver disease compared with non‐responders. A similar reduction was found when analyzing only those patients with cirrhosis. Although the risk was reduced in patients with SVR, the rates did not revert to baseline levels, especially for patients with cirrhosis. These data indicate that virus eradication with antiviral treatment is an important part of managing HCC risk [99]. Interferon‐based treatments were eventually phased out because of their mediocre SVR rates and disabling side‐effects. By contrast, DAAs achieve greater than 90% SVR rates and have comparatively few side‐effects [100]. DAAs have been shown to be at least

764

THE LIVER:  RISK FACTORS FOR HCC

equivalent to interferon with respect to HCC risk. A systematic review comparing the two treatment regimens found no differences between treatments in both HCC occurrence and recurrence risk following treatment [101]. Studies overwhelmingly show that de novo HCC risk is reduced by 50–80% in patients who achieve SVR following DAA treatment [102–104]. However, HCC risk remains

relatively high for these patient populations and others, including patients with diabetes and those who achieved SVR at an older age, those who have high α‐fetoprotein levels, and those with low platelet counts [95, 105–107] (Figure  59.3). Accordingly, these patient populations are likely to require continued HCC surveillance over time [108]. A study by Kanwal et al. [108] found that the risk of HCC reduced 76% in patients

With cirrhosis Overall Age Younger than 65 65 year and older Race White African American Hispanics Diabetes No Yes Alcohol use No Yes Drug use No Yes Previous HCV treatment No Yes Without cirrhosis Overall Age Younger than 65 65 year and older Race White African American Hispanics Diabetes No Yes Alcohol use No Yes Drug use No Yes Previous HCV treatment No Yes FIB-4 3.25 0.1

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

5.0

Incidence rate (per 100 person year)

Figure 59.3  The cumulative incidence and determinates of hepatocellular carcinoma following directly acting antiviral (DAA)‐related sustained virologic response (SVR) among Veterans Administration (VA) patients with hepatitis C virus (HCV). The incidence rates are shown according to several demographic and clinical features for patients with and without cirrhosis at baseline. Modified from [108] with permission of Elsevier.



59:  Epidemiology of Hepatocellular Carcinoma

who achieved SVR following DAA treatment. Among patients who achieved SVR, the annual incidence of HCC was 0.9%, and the highest rate of 1.0–2.2% was seen among patients who had cirrhosis [108]. These incidence rates are at or below the threshold for cost‐effective HCC surveillance [48, 49, 108]. Another consideration is the whether the performance can be translated into real‐world settings. The high efficacy and reduced side‐ effects seen with DAA regimens is likely to result in high effectiveness in real‐world settings [3]. One notable feature of DAA treatment is that it has been shown to achieve SVR in patients with advanced cirrhosis, those who were heavy drinkers, and those with HIV coinfection, all of which independently increase HCC risk [102–104]. However, the impact of DAAs will depend largely on the identification of individuals with chronic HCV infections, many of whom are asymptomatic and do not realize they are infected, and the accessibility and penetration of DAA treatment in diagnosed patients [109]. In the United States, an estimated 45–70% of persons infected with HCV are unaware of their infection status [110]. Further, in the United States and Europe, fewer than 20% of patients with HCV obtain treatment, either because they have not been screened for HCV or because of treatment costs [111]. A modelling study investigated the reduction in the number of HCC cases that would occur over a 10‐year period at current treatment rates and if half or all persons with HCV were treated. The study estimated that the number of HCC cases would reduce by 5% at the current treatment rate, by 30% if half were treated, and by 60% if all were treated [80]. To achieve these gains, more of the population needs to be screened, and those with HCV need to be diagnosed and treated [63]. In an effort to reduce the incidence of HCV‐associated liver disease and HCC in the United States, the Centers for Disease Control implemented a screening initiative that recommends testing all people born between 1945 and 1965 for HCV [112]. However, the success of the initiative hinges on patient consent, buy‐in by physicians, and the ability of patients to afford the expensive treatments [113]. Overall, DAAs have the potential to substantially reduce HCV‐associated HCC risk if the barriers to identification of persons with HCV and the subsequent access to treatment can be addressed.

765

because of differences in mean alcohol consumption and interactions with other risk factors [2].

Relative and absolute risk of HCC from heavy alcohol consumption Heavy alcohol consumption is estimated to have an increased relative risk for HCC of 1.5‐ to 3‐fold [114]. A meta‐analysis of studies investigating associations between alcohol intake and liver cancer found a relative risk of liver cancer to be 1.16 for heavy drinking (i.e. more than three drinks per day). There was a linear relationship between risk and the amount of alcohol consumed, with the risk increasing by 46% with a daily consumption of 50 g ethanol and by 66% with a daily consumption of 100 g [117].

Determinants of risk Approximately 13–23% of cases of HCC can be linked to heavy alcohol consumption [118, 119]. In a cohort study of patients with HCC who attended United States Veterans Administration (VA) hospitals, the cases of HCC associated with heavy alcohol consumption decreased from 21.9% to 15.7% between 2005 and 2010 [18]. The risk of HCC from heavy alcohol consumption varies by race and gender. The population‐attributable fraction was 25.6% for white, 30.1% for Hispanic, and 18.5% for black populations [118]. Similarly, the odds ratios of HCC were 9.5 for Hispanic and 3.6 for black populations [119]. For gender, the odds ratios were estimated to be 4.41–7.5 for males and 3.34–6.4 for females [118, 119]. Other comorbid conditions, such as HBV or HCV coinfection and obesity are likely risk factors as well [120]. The only known way to decrease the risk of HCC associated with heavy alcohol consumption is to cease drinking. A meta‐ analysis investigating the effects of alcohol cessation on HCC risk estimated that the risk of HCC declined by 6–7% per year, although the authors noted the estimate is uncertain due to heterogeneity among studies. Based on their estimate, they calculated that it would take 23 years of abstinence (95% confidence interval 14, 70 years) to reach the same risk of HCC as those who never drank alcohol [121].

Alcoholic liver disease

Metabolic syndrome, diabetes, and obesity

Studies have consistently shown an association between heavy alcohol consumption and risk of HCC [114–116]. The definition of heavy alcohol consumption and the duration of heavy drinking that puts individuals at risk varies highly between studies, with values ranging from 240 to 560 g/week and from greater than 1 year to greater than 5 years, respectively. Some studies do not even include a definition of the duration of heavy drinking [96]. It is also unclear if alcohol is carcinogenic or if carcinogenesis is secondary to cirrhosis development. Regardless of the outstanding questions, alcohol consumption contributes to liver disease and therefore contributes to HCC [23]. Generally, there are some disparities seen between studies conducted in areas with a low rate of HCC and those in areas with a high rate of HCC. In the former, studies tend to show correlations between alcohol consumption and HCC; in the latter, studies tend to be less conclusive. This observation could be

Metabolic syndrome is a cluster of risk factors related to insulin resistance, obesity, cardiovascular disease, blood pressure, and glucose metabolism. Metabolic syndrome itself has been linked to several cancers [122], and several risk factors associated with metabolic syndrome independently increase the risk of HCC [123]. Increasing evidence suggests that metabolic derangements associated with metabolic syndrome could be an etiologic factor for HCC in the absence of cirrhosis [124]. The mechanism underlying the role of metabolic syndrome is unclear. However, the changes in lipid metabolism associated with the syndrome could contribute to steatosis, NAFLD, and non‐alcoholic steatohepatitis (NASH), all of which are precursors to HCC [2, 23]. With the prevalence of metabolic syndrome as well as obesity and diabetes, major contributors to metabolic syndrome, rising dramatically in the United States, the prevalence of related cancers, including HCC, is increasing [5, 125, 126].

766

THE LIVER:  RISK FACTORS FOR HCC

HCC risk associated with diabetes Type 2 diabetes was estimated to increase risk of HCC two‐ to threefold in a meta‐analysis of case–control and cohort studies. This analysis included various study designs from several ­geographic populations, and the results from included studies consistently showed positive correlations [94]. These numbers were corroborated in other systematic reviews [127, 128]. However, there are several outstanding questions that muddy data interpretation. First, it is unclear if the liver damage caused by diabetes is the source of the HCC risk or if the diabetes independently increases risk beyond that associated with liver damage. Second, there are cases in which diabetes is caused by chronic liver disease prior to the development of HCC (reverse causality); the effects of these cases on risk of HCC are unknown [23]. A recent meta‐analysis investigated cohort studies that only evaluated patients with chronic liver to reduce the inclusion of reverse causality cases. This analysis found that risk of HCC increased 1.5‐ to 2‐fold in patients with diabetes [129]. Collectively, these data support that type 2 diabetes independently increases the risk of HCC [94, 129]. The treatment of type 2 diabetes with metformin has been shown to decrease the risk of HCC, whereas treatment with insulin or sulfonylurea can increase the risk of HCC [130]. Further, the duration of diabetes may be associated with an incremental increase in the risk of HCC [131]. Flemming et al. devised a predictive HCC risk model for patients who are ­cirrhotic based on the National Liver Transplantation Waitlist database. The model evaluated diabetes status in addition to age, race, cirrhotic etiology, sex, and liver dysfunction severity to predict 1‐year HCC risk [132].

HCC risk associated with obesity Most, but not all, studies have reported modest increases in relative risk of HCC in persons with obesity. A systematic review of studies investigating associations between body mass index (BMI) and HCC found positive associations in seven cohort studies, with relative risks ranging from 1.4 to 4.1, no association in two cohort studies, and an inverse association for one subgroup in one cohort study [133]. A meta‐analysis of observational studies evaluating excess body weight and liver cancer risk found that persons who were overweight (25–30 kg m−2) and obese (>30 kg m−2) had increases in liver cancer risk of 17% and 89%, respectively [134]. A case–control study investigated the effect of obesity in early adulthood (i.e. mid‐20s to mid‐40s) on HCC risk in patients with HCC matched with healthy controls. This study found that persons who were obese in early adulthood had an estimated odds ratio for HCC of 2.6 for all participants. The estimated odds ratios for men and women were 2.3 and 3.6, respectively [135]. A cohort study evaluating the effect of abdominal obesity and weight gain on HCC risk found a threefold higher risk of HCC in participants in the highest tertile of waist‐to‐hip and waist‐to‐height ratio compared with those in the lowest tertile. Similarly, participants in the highest tertile of weight gain had a 2.48‐fold higher risk of HCC compared with those in the lowest tertile [136]. Similar to type 2 diabetes, the mechanisms by which obesity increases the risk of liver disease are unknown. Obesity could induce physiological changes or changes on the molecular level that lead to HCC,

or chronic liver disease could induce metabolic changes that lead to obesity [23]. Further studies are needed to clarify these mechanisms.

HCC risk associated with metabolic syndrome One meta‐analysis evaluated cohort and case–control studies investigating the association between metabolic syndrome and HCC risk, and included four studies with a total of 829 651 participants. The estimated relative risk of HCC in participants with metabolic syndrome was 1.81 compared with healthy participants [137].

Determinants of risk Increased risk When compared with the relative risks associated with viral hepatitis infections, the risks associated with type 2 diabetes, obesity, and metabolic syndrome are quite low. However, in more developed countries, the prevalence of these conditions is much higher than that of viral hepatitis [2]. In 2008, approximately 6.4% of the global population was estimated to have diabetes. Further, in developing countries, the rate at which diabetes prevalence is increasing far exceeds that in developed countries. Based on projection models, the prevalence of diabetes is ­estimated to increase by 69% in developing countries and by 20% in developed countries [138]. Parallel trends have been observed for the prevalence of obesity based on increases in BMI [139]. Given these trends, it is likely that the number of HCC cases related to metabolic syndrome, type 2 diabetes, and obesity will increase in the future [2]. Several factors have been identified that affect HCC risk in persons with obesity. As mentioned earlier, waist‐to‐hip ratio, a measure of abdominal obesity, increases the risk of HCC [136]. The visceral adipose tissue associated with obesity alters the levels of cytokines in the body, making obesity a low‐grade inflammatory disorder. These alterations in cytokine levels are thought to contribute to HCC tumorigenesis and progression and are likely to represent a risk factor for HCC [140]. Obesity can lead to NAFLD and the subsequent development of NASH, which increases the risk of HCC [141]. Another potential risk factor is viral hepatitis infection. In the study of obesity in early adulthood described earlier, an independent synergistic interaction on HCC risk was observed between obesity and viral hepatitis infection [135].

Decreased risk One of the features associated with metabolic syndrome is abnormal cholesterol and triglycerides levels [142]. Statins are a class of drugs primarily used to lower cholesterol to prevent cardiovascular disease [2, 143]. Increasing evidence suggests that statins could be potential anticancer agents based on their ability to inhibit angiogenesis and metastasis and enhance apoptosis [144, 145]. Population studies out of Taiwan have reported HCC risk reductions of 47–56% with statin treatment [146, 147]. In similar studies of Taiwanese patients with HBV and HCV, HCC risk decreased up to 66–67% in those treated with statins ­compared with those not taking statins, and a dose–response relationship was observed [148, 149]. A similar Taiwanese



59:  Epidemiology of Hepatocellular Carcinoma

study investigating the effects of individual statin medications found that simvastatin, lovastatin, and atorvastatin reduced HCC risk by 30–48% [150]. These associations were not observed in early studies conducted in Denmark and the United States, areas with low rates of HCC [151, 152]. However, more recent studies out of the United States and Sweden found significant inverse associations between statin use and HCC risk in the general population [153–155] and in subgroups, such as men with diabetes [156] and men with HCV infection [157]. Two meta‐analyses conducted in parallel found similar results, with a relative risk of 0.58 [158] and an adjusted odds ratio of 0.63 [159] for statin users compared with non‐statin users. Notably, the meta‐analysis including three randomized clinical trials found no significant benefit (adjusted odds ratio, 0.95) when only the trials were analyzed [159]; however, none of the trials had sufficient power to detect changes in HCC development, so this finding may be somewhat misleading [63, 106]. With respect to type 2 diabetes, use of the antidiabetic medication metformin could be used to reduce HCC risk. Metformin is a first‐line treatment for type 2 diabetes used to reduce serum glucose and insulin levels [2]. Three meta‐analyses investigating the effects of metformin on cancer risk were performed in parallel. One meta‐analysis evaluated observational studies and randomized trials looking at the effects of metformin and other hypoglycemic drugs on cancer. Meta‐analysis of the nine studies investigating liver cancer uncovered odds ratios of 0.34 when metformin was compared with no metformin, 0.09 when metformin was compared with other drugs, and 0.56 when ­metformin was compared with sulfonylureas [160]. Another meta‐analysis evaluated seven studies looking at the effect of metformin use on HCC. This analysis found a relative risk of 0.24 when metformin use was compared with non‐users [161]. The third meta‐analysis evaluated studies investigating the effect of metformin and other diabetes treatment on the risk of HCC. This analysis revealed an odds ratio of 0.50 for metformin use, 1.62 for sulfonylurea use, and 2.61 for insulin use. No effect was seen for thiazolidinedione use [130]. However, the results of these meta‐analyses should be interpreted cautiously, largely because metformin is prescribed early in disease progression whereas the others are prescribed after metformin ­failure once the disease has progressed. This difference can lead to an overestimation of risk reduction by metformin [2]. An alternate interpretation of the data is that lack of glycemic control for long periods and more severe disease could increase the risk of HCC [23].

Non‐alcoholic fatty liver disease NAFLD is caused by the accumulation of excess triglycerides in the liver. This accumulation occurs in the absence of heavy alcohol consumption and leads to steatosis. One consistent feature of nearly all NAFLD cases is insulin resistance, a risk factor associated with metabolic syndrome. Accordingly, NAFLD is thought to be a hepatic manifestation of metabolic syndrome [162]. Worldwide, NAFLD is becoming a leading cause of chronic liver disease [163]. The prevalence of NAFLD in the United States has doubled over the last two decades up to 30% in some segments of the adult general population [164–166]. Among patients with NAFLD, 20–30% are estimated to develop

767

NASH and concomitant necroinflammation and fibrosis, which progresses to cirrhosis in 10–20% of cases [162, 167]. NAFLD is now a leading cause of cirrhosis, and 30–40% of new HCC cases are associated with metabolic disorders [118, 119, 164– 166]. NASH is the second‐leading etiology for liver transplants related to HCC in the United States [168]. The increasing number of HCC cases due to NASH‐related cirrhosis could offset the reductions in HCV‐related HCC expected after 2020 [63]. One pressing threat to reductions in HCC incidence is the fact that patients are often unaware they have NASH‐related cirrhosis because the cirrhosis can be well compensated for years. These asymptomatic patients go undiagnosed, all the while, the risk of developing HCC remains [23]. Additionally, a small proportion of patients with NAFLD and/or NASH develop HCC in the absence of cirrhosis [169].

Relative and absolute risk of HCC from NAFLD Both the etiology and presence of cirrhosis affect HCC risk. A systematic analysis of studies investigating the links between NAFLD or NASH and HCC risk found that the HCC risk of the included studies varied from 0 to 38% with follow‐ups of 5–10 years, with most studies having small‐ to medium‐sized cohorts from tertiary care settings. The authors found that NAFLD or NASH patients lacking cirrhosis had minimal HCC risk. By contrast, patients with NASH‐related cirrhosis had incidences ranging from 2.4 to 12.8%. However, the appropriateness of performing subgroup analysis and extrapolating these data to the general population is limited due to the nature of the included studies. Notably, HCC risk was found to be substantially lower for patients with NASH‐related cirrhosis compared with patients with HCV‐associated cirrhosis [170]. A large cohort study including 296 707 patients with NAFLD and an equal number of matched controls without NAFLD from 130 VHA facilities found that patients with NAFLD had a higher annual HCC risk than controls (hazard ratio 7.62) [171]. These data support an association between HCC risk and NAFLD.

Determinants of HCC risk for NAFLD Increased risk Most of the existing epidemiological studies of NAFLD and HCC risk lacked the sample size to perform adequate subgroup analyses and identify high‐risk groups, or the power to arrive at accurate estimates [170]. An exception is the VHA cohort study, where the HCC risk for patients with NAFLD ranged from 1.6 to 23.7 per 1000 person‐years, with the highest risk among older Hispanic patients with cirrhosis. However, the authors also found that approximately 20% of patients with NAFLD and HCC had no evidence of cirrhosis [171]. Some case–control studies found that the prevalence of diabetes and obesity were higher for patients with NAFLD‐related HCC compared with patients with other chronic liver diseases [170]. A subgroup analysis from a meta‐analysis investigating the associations of PNPLA3 polymorphisms with liver fibrosis, HCC risk, and HCC prognosis found that patients with NASH‐associated cirrhosis and the PNPLA3 polymorphism had a higher risk of developing HCC than those who lacked the polymorphism (odds ratio 1.67), indicating that there could be a genetic component to NASH‐associated HCC development [172].

768

THE LIVER:  ACKNOWLEDGMENTS

Decreased risk Weight loss can reduce NASH severity [173]; however, to date, no studies have directly addressed whether treatment of NAFLD or NASH can reduce the risk of HCC. One study found that bariatric surgery reduced the risk of HCC in persons who are obese and have the PNPLA3 polymorphism [174], but more studies are needed to clarify the effect in the context of NAFLD and NASH.

HCC in the absence of cirrhosis While most HCC develops in the presence of cirrhosis, a subset of cases can occur in the absence of cirrhosis. In the case of HBV‐related HCC, approximately 10–30% of cases arise in the absence of cirrhosis [175, 176]. HCC in the absence of cirrhosis is an even rarer occurrence in patients with HCV, with one study finding only 3% of HCC cases lacking cirrhosis [175]. A study of HCC risk in patients with HCV who were taking peginterferon maintenance therapy found that the few HCC cases developing in the absence of cirrhosis did so in the presence of advanced fibrosis [177]. In general, HCV‐associated HCC occurring in the absence of cirrhosis tends to affect those with bridging hepatic fibrosis [3]. By contrast, NAFLD‐ and NASH‐ related HCC can develop with little to no fibrosis [178, 179]. NAFLD‐related HCC represents the largest proportion of HCC cases occurring in the absence of advanced fibrosis or cirrhosis [169]. Because HCC surveillance is centered around cirrhosis, these cases tend to be caught later. Patients with alcohol‐ or HCV‐related HCC are significantly more likely to have received HCC surveillance within three years before their HCC diagnosis than those with NAFLD‐related HCC. Similarly, patients with NAFLD are less likely to receive HCC‐specific treatments than those with HCV‐related HCC [169]. Given the available data, HCC screening in patients with NAFLD‐related advanced fibrosis and cirrhosis may be warranted, particularly among Hispanic patients and patients with metabolic syndrome.

Autoimmune hepatitis Autoimmune hepatitis is an autoimmune disorder in which a T cell‐mediated immune response to liver autoantigens erodes hepatic tissues, leading to fibrosis and cirrhosis [180, 181]. Patients can have an array of presentations from asymptomatic to severe acute hepatitis, and in some cases can progress to fulminant hepatic failure [182, 183]. Among patients with autoimmune hepatitis, approximately 30% present with cirrhosis at the time of diagnosis [184, 185]. Some studies have linked autoimmune hepatitis with HCC risk, but the magnitude of that risk varies [186–189]. One meta‐analysis evaluated 25 cohort studies investigating HCC risk and risk factors among patients with autoimmune hepatitis. The pooled incidence of HCC among all patients with autoimmune hepatitis was 3.06 per 1000 patient‐ years, and among those who had cirrhosis at diagnosis was 10.07 per 1000 patient‐years. Further, 98.9% of patients with autoimmune hepatitis who developed HCC had cirrhosis before or at the time of diagnosis. However, the risk of HCC was still lower for patients with autoimmune hepatitis than for those with viral hepatitis. Given the high incidence of HCC in patients with

autoimmune hepatitis and cirrhosis, HCC surveillance in those patients who progress to cirrhosis may be warranted [190].

CONTRIBUTION OF RISK FACTORS TO THE BURDEN OF HCC Cirrhosis is by and large the most significant risk factor and the foundation for most HCC surveillance programs. Because patients with compensated cirrhosis can have their condition go undetected for years, HCC surveillance is often underused [3]. When looking at the contribution of risk factors on HCC burden, we must consider the population‐attributable fraction, a combination measure of the prevalence of a condition and the risk estimate. This measure provides the estimated proportion of disease cases that can be eliminated by addressing the risk factor. For example, in the United States, viral hepatitis infections are much rarer than NAFLD. Despite their higher HCC risk estimates, the population‐attributable fraction of HBV and HCV are lower than that of NAFLD [118]. As of 2013, the population‐attributable fractions for obesity and diabetes were 36.6%, for HCV was 22.4% and for HBV was 6.3% in the United States [118]. Because the prevalence of obesity and diabetes is on the rise in developed and developing countries [138, 139], we anticipate that the population‐attributable fractions of obesity, ­diabetes, metabolic syndrome, and NAFLD will increase in the coming years. Notably, despite the lower population‐attributable fraction, HCV is still the leading cause of HCC and the leading etiology in patients with HCC undergoing liver transplants. However, NAFLD is the fastest growing cause of HCC, with the number of NASH‐associated liver transplants increasing by 11.8‐fold between 2002 and 2016 [191]. In the VHA cohort investigating temporal trends in NAFLD‐related HCC, the proportion of NAFLD‐related HCC cases remained relatively unchanged between 2005 and 2010, whereas that of HCV‐ related HCC cases increased significantly [18]. The differences seen between population‐attributable fraction and epidemiological reports are likely due to the lag between changes in the risk factors of a population and the time of HCC development, which can take decades to manifest [192, 193]. Generally, the population‐attributable fraction is used to identify the targets for ­prevention programs [3]. The temporal lag associated with HCC population‐attributable fraction can give clinicians and public health workers the opportunity to identify candidates for HCC surveillance, including those who do not fit within traditional risk‐assessment models, and intervene with risk factors projected to cause a higher proportion of HCC cases, essentially changing the trajectory of disease over time.

ACKNOWLEDGMENTS GRANT SUPPORT: This material is based upon work supported by Cancer Prevention & Research Institute of Texas grant (RP150587). The work is also supported in part by the Center for Gastrointestinal Development, Infection and Injury (NIDDK P30 DK 56338).



59:  Epidemiology of Hepatocellular Carcinoma

REFERENCES   1. Global Burden of Disease Cancer Center. Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability‐adjusted life‐years for 32 cancer groups, 1990 to 2015: A systematic analysis for the global burden of disease study. JAMA Oncol, 2017;3(4):524–48.   2. McGlynn, K.A., Petrick, J.L., and London, W.T. Global epidemiology of hepatocellular carcinoma: an emphasis on demographic and regional variability. Clin Liver Dis, 2015;19(2):223–38.   3. Singal, A.G. and El‐Serag, H.B. Hepatocellular carcinoma from epidemiology to prevention: translating knowledge into practice. Clin Gastroenterol Hepatol, 2015;13(12):2140–51.  4. Petrick, J.L., Braunlin, M., Laversanne, M. et  al. International trends in liver  cancer incidence, overall and by histologic subtype, 1978–2007. Int J Cancer, 2016;139(7):1534–45.   5. Arnold, M., Razum, O., and Coebergh, J.W. Cancer risk diversity in non‐ western migrants to Europe: an overview of the literature. Eur J Cancer, 2010;46(14):2647–59.   6. Dorak, M.T. and Karpuzoglu, E. Gender differences in cancer susceptibility: an inadequately addressed issue. Front Genet, 2012;3:268. 6A. Ferlay, J., Parkin, D.M., Curado, M.P. et al. Cancer Incidence in Five Continents, Volumes I–X. IARC CANCERBase No. 10 [Internet]. http://ci5.iarc.fr   7. Papatheodoridis, G.V., Idilman, R., Dalekos, G.N. et al. The risk of hepatocellular carcinoma decreases after the first 5 years of entecavir or tenofovir in Caucasians with chronic hepatitis B. Hepatology, 2017;66(5):1444–53.   8. Franceschi, S., El‐Serag, H.B., Forman, D., Newton, R., and Plummer, M. Infectious agents, in Cancer Epidemiology and Prevention, 4th edn. (eds. M. Thun, M.S. Linet, J.R. Cerhan, C.A. Haiman, and D. Schottenfeld), Oxford University Press, Oxford, 2017, p. 446.   9. Yu, M.W., Yang, Y.C., Yang, S.Y. et  al. Androgen receptor exon 1 CAG repeat length and risk of hepatocellular carcinoma in women. Hepatology, 2002;36(1):156–63. 10. Yuan, J.M., Ross, R.K., Stanczyk, F.Z. et al. A cohort study of serum testosterone and hepatocellular carcinoma in Shanghai, China. Int J Cancer, 1995;63(4):491–3. 11. Yu, M.‐W. and Chen, C.‐J. Elevated serum testosterone levels and risk of hepatocellular carcinoma. Cancer Res, 1993;53(4):790–4. 12. Yu, M.W., Cheng, S.W., Lin, M.W. et  al. Androgen‐receptor gene CAG repeats, plasma testosterone levels, and risk of hepatitis B‐related hepatocellular carcinoma. J Natl Cancer Inst, 2000;92(24):2023–8. 13. Yeh, S.H. and Chen, P.J. Gender disparity of hepatocellular carcinoma: the roles of sex hormones. Oncology, 2010;78(Suppl 1):172–9. 14. White, D.L., Thrift, A.P., Kanwal, F., Davila, J., and El‐Serag, H.B. Incidence of hepatocellular carcinoma in all 50 United States, from 2000 through 2012. Gastroenterology, 2017;152(4):812–20. e5. 15. US Cancer Statistics Working Group. U.S. Cancer Statistics Data Visualizations Tool, based on November 2017 submission data (1999–2015): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; 2018 (updated June 2018). www.cdc.gov/cancer/dataviz 16. Armstrong, G.L., Wasley, A., Simard, E.P. et al. The prevalence of hepatitis C virus infection in the United States, 1999 through 2002. Ann Intern Med, 2006;144(10):705–14. 17. Kallwitz, E.R., Daviglus, M.L., Allison, M.A. et al. Prevalence of suspected nonalcoholic fatty liver disease in Hispanic/Latino individuals differs by heritage. Clin Gastroenterol Hepatol, 2015;13(3):569–76. 18. Mittal, S., Sada, Y.H., El‐Serag, H.B. et al. Temporal trends of nonalcoholic fatty liver disease–related hepatocellular carcinoma in the veteran affairs population. Clin Gastroenterol Hepatol, 2015;13(3):594–601. e1. 19. Kulik, L. and El‐Serag, H.B. Epidemiology and management of hepatocellular carcinoma. Gastroenterology, 2019;156(2):477–91.e1 20. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Biological agents. Volume 100 B. A review of human carcinogens. IARC Monogr Eval Carcinog Risks Hum, 2012;100(Pt B):1–441. 21. Tabor, E. Tumor suppressor genes, growth factor genes, and oncogenes in hepatitis B virus‐associated hepatocellular carcinoma. J Med Virol, 1994; 42(4):357–65. 22. de Martel, C., Maucort‐Boulch, D., Plummer, M., and Franceschi, S. World‐ wide relative contribution of hepatitis B and C viruses in hepatocellular ­carcinoma. Hepatology, 2015;62(4):1190–200.

769

23. Massarweh, N.N. and El‐Serag, H.B. Epidemiology of hepatocellular carcinoma and intrahepatic cholangiocarcinoma. Cancer Control, 2017;24(3): 1073274817729245. 24. McGowan, C.E., Edwards, T.P., Luong, M.U., and Hayashi, P.H. Suboptimal surveillance for and knowledge of hepatocellular carcinoma among primary care providers. Clin Gastroenterol Hepatol, 2015;13(4):799–804. 25. McQuillan, G.M., Coleman, P.J., Kruszon‐Moran, D. et  al. Prevalence of hepatitis B virus infection in the United States: the National Health and Nutrition Examination Surveys, 1976 through 1994. Am J Public Health, 1999;89(1):14–18. 26. Cho, L.Y., Yang, J.J., Ko, K.P. et al. Coinfection of hepatitis B and C viruses and risk of hepatocellular carcinoma: systematic review and meta‐analysis. Int J Cancer, 2011;128(1):176–84. 27. El‐Serag, H.B. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology, 2012;142(6):1264–73. e1. 28. Fattovich, G., Bortolotti, F., and Donato, F. Natural history of chronic hepatitis B: special emphasis on disease progression and prognostic factors. J Hepatol, 2008;48(2):335–52. 29. Lok, A.S. and McMahon, B.J. Chronic hepatitis B: update 2009. Hepatology, 2009;50(3):661–2. 30. Gordon, S.C., Lamerato, L.E., Rupp, L.B. et al. Antiviral therapy for chronic hepatitis B virus infection and development of hepatocellular carcinoma in a US population. Clin Gastroenterol Hepatol, 2014;12(5):885–93. 31. Mittal, S., Kramer, J.R., Omino, R. et al. Role of age and race in the risk of hepatocellular carcinoma in veterans with hepatitis B virus infection. Clin Gastroenterol Hepatol, 2018;16(2):252–9. 32. Evans, A.A., Chen, G., Ross, E.A. et al. Eight‐year follow‐up of the 90,000‐ person Haimen City cohort: I. Hepatocellular carcinoma mortality, risk ­factors, and gender differences. Cancer Epidemiol Biomark Prev, 2002; 11(4):369–76. 33. Kew, E. and Marcus, R. Some characteristics of Mozambican Shangaans with primary hepatocellular cancer. South African Med J, 1977;51(10): 306–9. 34. Kew, M.C. and Macerollo, P. Effect of age on the etiologic role of the hepatitis B virus in hepatocellular carcinoma in blacks. Gastroenterology, 1988;94(2):439–42. 35. Yang, J.D., Mohamed, E.A., Aziz, A.O. et al. Characteristics, management, and outcomes of patients with hepatocellular carcinoma in Africa: a multicountry observational study from the Africa Liver Cancer Consortium. Lancet Gastroenterol Hepatol, 2017;2(2):103–11. 36. Chen, C.‐J., Yang, H.‐I., Su, J. et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA, 2006; 295(1):65–73. 37. Huang, Y. and Lok, A.S. Viral factors and outcomes of chronic HBV infection. Am J Gastroenterol, 2011;106(1):93–5. 38. McMahon, B.J. Natural history of chronic hepatitis B. Clin Liver Dis, 2010;14(3):381–96. 39. Liu, C.J., Chen, B.F., Chen, P.J. et al. Role of hepatitis B viral load and basal core promoter mutation in hepatocellular carcinoma in hepatitis B carriers. J Infect Dis, 2006;193(9):1258–65. 40. Yang, H.I., Yeh, S.H., Chen, P.J. et al. Associations between hepatitis B virus genotype and mutants and the risk of hepatocellular carcinoma. J Natl Cancer Inst, 2008;100(16):1134–43. 41. Shi, J., Zhu, L., Liu, S., and Xie, W.F. A meta‐analysis of case‐control studies on the combined effect of hepatitis B and C virus infections in causing hepatocellular carcinoma in China. Br J Cancer, 2005;92(3):607–12. 42. Donato, F., Boffetta, P., and Puoti, M. A meta‐analysis of epidemiological studies on the combined effect of hepatitis B and C virus infections in causing hepatocellular carcinoma. Int J Cancer, 1998;75(3):347–54. 43. Thomas, D.L., Astemborski, J., Rai, R.M. et al. The natural history of hepatitis C virus infection: host, viral, and environmental factors. JAMA, 2000; 284(4):450–6. 44. Donato, F., Tagger, A., Gelatti, U. et  al. Alcohol and hepatocellular carcinoma: the effect of lifetime intake and hepatitis virus infections in men and women. Am J Epidemiol, 2002;155(4):323–31. 45. Jee, S.H., Ohrr, H., Sull, J.W., and Samet, J.M. Cigarette smoking, alcohol drinking, hepatitis, B., and risk for hepatocellular carcinoma in Korea. J Natl Cancer Inst, 2004;96(24):1851–6. 46. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Personal habits and indoor combustions. Volume 100 E. A review of human carcinogens. IARC Monogr Eval Carcinog Risks Hum, 2012;100(Pt E):1–538.

770

THE LIVER:  REFERENCES

47. Chuang, S.C., Lee, Y.C., Hashibe, M. et  al. Interaction between cigarette smoking and hepatitis B and C virus infection on the risk of liver cancer: a meta‐analysis. Cancer Epidemiol Biomark Prev, 2010;19(5):1261–8. 48. Sarasin, F.P., Giostra, E., and Hadengue, A. Cost‐effectiveness of screening for detection of small hepatocellular carcinoma in western patients with Child‐Pugh class A cirrhosis. Am J Med, 1996;101(4):422–34. 49. Arguedas, M.R., Chen, V.K., Eloubeidi, M.A., and Fallon, M.B. Screening for hepatocellular carcinoma in patients with hepatitis C cirrhosis: a cost‐ utility analysis. Am J Gastroenterol, 2003;98(3):679–90. 50. Abu‐Amara, M., Cerocchi, O., Malhi, G. et al. The applicability of hepatocellular carcinoma risk prediction scores in a North American patient ­population with chronic hepatitis B infection. Gut, 2016;65(8):1347–58. 51. Wong, V.W.‐S., Chan, S.L., Mo, F. et al. Clinical scoring system to predict hepatocellular carcinoma in chronic hepatitis B carriers. J Clin Oncol, 2010;28(10):1660–5. 52. Yuen, M.‐F., Tanaka, Y., Fong, D.Y.‐T. et  al. Independent risk factors and predictive score for the development of hepatocellular carcinoma in chronic hepatitis B. J Hepatol, 2009;50(1):80–8. 53. Yang, H.‐I., Yuen, M.‐F., Chan, H.L.‐Y. et al. Risk estimation for hepatocellular carcinoma in chronic hepatitis B (REACH‐B): development and validation of a predictive score. Lancet Oncol, 2011;12(6):568–74. 54. Kim, W.R., Loomba, R., Berg, T. et al. Impact of long‐term tenofovir disoproxil fumarate on incidence of hepatocellular carcinoma in patients with chronic hepatitis B. Cancer, 2015;121(20):3631–8. 55. Ahn, J., Lim, J.K., Lee, H.M. et  al. Lower observed hepatocellular carcinoma incidence in chronic hepatitis B patients treated with entecavir: results of the ENUMERATE study. Am J Gastroenterol, 2016;111(9):1297–304. 56. Hsu, Y.‐C., Wu, C.‐Y., Lane, H.‐Y. et al. Determinants of hepatocellular carcinoma in cirrhotic patients treated with nucleos(t)ide analogues for chronic hepatitis B. J Antimicrob Chemother, 2014;69(7):1920–7. 57. Papatheodoridis, G.V., Dalekos, G.N., Yurdaydin, C. et  al. Incidence and predictors of hepatocellular carcinoma in Caucasian chronic hepatitis B patients receiving entecavir or tenofovir. J Hepatol, 2015;62(2):363–70. 58. Raffetti, E., Fattovich, G., and Donato, F. Incidence of hepatocellular carcinoma in untreated subjects with chronic hepatitis B: a systematic review and meta‐analysis. Liver Int, 2016;36(9):1239–51. 59. Yang, H.‐I., Lu, S.‐N., Liaw, Y.‐F. et al. Hepatitis B e antigen and the risk of hepatocellular carcinoma. N Engl J Med, 2002;347(3):168–74. 60. Tseng, T.C., Liu, C.J., Yang, H.C. et al. High levels of hepatitis B surface antigen increase risk of hepatocellular carcinoma in patients with low HBV load. Gastroenterology, 2012;142(5):1140–9. e3. 61. Hsu, Y.C., Ho, H.J., Lee, T.Y. et al. Temporal trend and risk determinants of hepatocellular carcinoma in chronic hepatitis B patients on entecavir or tenofovir. J Viral Hepatitis, 2018;25(5):543–51. 62. Kim, M.N., Hwang, S.G., Rim, K.S. et al. Validation of PAGE‐B model in Asian chronic hepatitis B patients receiving entecavir or tenofovir. Liver Int, 2017;37(12):1788–95. 63. El‐Serag, H.B. and Kanwal, F. Epidemiology of hepatocellular carcinoma in the United States: where are we? Where do we go? Hepatology, 2014; 60(5):1767–75. 64. Chiang, C.J., Yang, Y.W., You, S.L., Lai, M.S., and Chen, C.J. Thirty‐year outcomes of the national hepatitis B immunization program in Taiwan. JAMA, 2013;310(9):974–6. 65. Sun, Z., Chen, T., Thorgeirsson, S.S. et al. Dramatic reduction of liver cancer incidence in young adults: 28 year follow‐up of etiological interventions in an endemic area of China. Carcinogenesis, 2013;34(8):1800–5. 66. Chen, J.‐G., Egner, P.A., Ng, D. et al. Reduced aflatoxin exposure presages decline in liver cancer mortality in an endemic region of China. Cancer Prev Res, 2013;6(10):1038–45. 67. World Health Organization. Guidelines for the Prevention, Care and Treatment of Persons with Chronic Hepatitis B Infection. World Health Organization, Geneva, Switzerland, 2015, p. 136. 68. Liaw, Y.F., Sung, J.J., Chow, W.C. et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med, 2004;351(15): 1521–31. 69. Sung, J., Tsoi, K., Wong, V., Li, K., and Chan, H. Meta‐analysis: treatment of hepatitis B infection reduces risk of hepatocellular carcinoma. Aliment Pharmacol Ther, 2008;28(9):1067–77. 70. Wong, G.L.H., Chan, H.L.Y., Mak, C.W.H. et al. Entecavir treatment reduces hepatic events and deaths in chronic hepatitis B patients with liver cirrhosis. Hepatology, 2013;58(5):1537–47.

71. Papatheodoridis, G.V., Lampertico, P., Manolakopoulos, S., and Lok, A. Incidence of hepatocellular carcinoma in chronic hepatitis B patients receiving nucleos(t)ide therapy: a systematic review. J Hepatol, 2010;53(2): 348–56. 72. Wu, C.‐Y., Lin, J.‐T., Ho, H.J. et al. Association of nucleos (t) ide analogue therapy with reduced risk of hepatocellular carcinoma in patients with chronic hepatitis B  –  a nationwide cohort study. Gastroenterology, 2014; 147(1):143–51. e5. 73. Hosaka, T., Suzuki, F., Kobayashi, M. et al. Long‐term entecavir treatment reduces hepatocellular carcinoma incidence in patients with hepatitis B virus infection. Hepatology, 2013;58(1):98–107. 74. Zoutendijk, R., Reijnders, J.G., Zoulim, F. et al. Virological response to entecavir is associated with a better clinical outcome in chronic hepatitis B patients with cirrhosis. Gut, 2013;62(5):760–5. 75. Sung, J.J., Tsoi, K.K., Wong, V.W., Li, K.C., and Chan, H.L. Meta‐analysis: treatment of hepatitis B infection reduces risk of hepatocellular carcinoma. Aliment Pharmacol Ther, 2008;28(9):1067–77. 76. Miyake, Y., Kobashi, H., and Yamamoto, K. Meta‐analysis: the effect of interferon on development of hepatocellular carcinoma in patients with chronic hepatitis B virus infection. J Gastroenterol, 2009;44(5):470–5. 77. Yang, Y.F., Zhao, W., Zhong, Y.D. et al. Interferon therapy in chronic hepatitis B reduces progression to cirrhosis and hepatocellular carcinoma: a meta‐ analysis. J Viral Hepat, 2009;16(4):265–71. 78. Mohd Hanafiah, K., Groeger, J., Flaxman, A.D., and Wiersma, S.T. Global epidemiology of hepatitis C virus infection: new estimates of age‐specific antibody to HCV seroprevalence. Hepatology, 2013;57(4):1333–42. 79. Alter, H.J. and Seeff, L.B. Recovery, persistence, and sequelae in hepatitis C virus infection: a perspective on long‐term outcome. Semin Liver Dis, 2000;20(1):17–35. 80. Davis, G.L., Alter, M.J., El‐Serag, H., Poynard, T., and Jennings, L.W. Aging of hepatitis C virus (HCV)‐infected persons in the United States: a multiple cohort model of HCV prevalence and disease progression. Gastroenterology, 2010;138(2):513–21, 21.e1–6. 81. Kanwal, F., Hoang, T., Kramer, J.R. et al. Increasing prevalence of HCC and cirrhosis in patients with chronic hepatitis C virus infection. Gastroenterology, 2011;140(4):1182–8.e1. 82. De Mitri, M.S., Poussin, K., Baccarini, P. et al. HCV‐associated liver cancer without cirrhosis. Lancet, 1995;345(8947):413–15. 83. Haydon, G.H., Jarvis, L.M., Simmonds, P., and Hayes, P.C. Association between chronic hepatitis C infection and hepatocellular carcinoma. Lancet, 1995;345(8954):928–9. 84. Mittal, S. and El‐Serag, H.B. Epidemiology of hepatocellular carcinoma: consider the population. J Clin Gastroenterol, 2013;47(Suppl):S2–6. 85. El‐Serag, H.B., Kramer, J., Duan, Z., and Kanwal, F. Racial differences in the progression to cirrhosis and hepatocellular carcinoma in HCV‐infected veterans. Am J Gastroenterol, 2014;109(9):1427. 86. Kanwal, F., Kramer, J.R., Ilyas, J., Duan, Z., and El‐Serag, H.B. HCV genotype 3 is associated with an increased risk of cirrhosis and hepatocellular cancer in a national sample of US Veterans with HCV. Hepatology, 2014; 60(1):98–105. 87. Freeman, A.J., Dore, G.J., Law, M.G. et al. Estimating progression to cirrhosis in chronic hepatitis C virus infection. Hepatology, 2001;34(4 Pt 1): 809–16. 88. Poynard, T., Bedossa, P., and Opolon, P. Natural history of liver fibrosis progression in patients with chronic hepatitis C. The OBSVIRC, METAVIR, CLINIVIR, and DOSVIRC groups. Lancet, 1997;349(9055):825–32. 89. Chang, K.C., Wu, Y.Y., Hung, C.H. et al. Clinical‐guide risk prediction of hepatocellular carcinoma development in chronic hepatitis C patients after interferon‐based therapy. Br J Cancer, 2013;109(9):2481–8. 90. Raimondi, S., Bruno, S., Mondelli, M.U., and Maisonneuve, P. Hepatitis C virus genotype 1b as a risk factor for hepatocellular carcinoma development: a meta‐analysis. J Hepatol, 2009;50(6):1142–54. 91. Kattakuzhy, S., Levy, R., Rosenthal, E. et al. Hepatitis C genotype 3 disease. Hepatol Int, 2016;10(6):861–70. 92. Clifford, G.M., Rickenbach, M., Polesel, J. et al. Influence of HIV‐related immunodeficiency on the risk of hepatocellular carcinoma. AIDS, 2008; 22(16):2135–41. 93. Kramer, J.R., Giordano, T.P., and El‐Serag, H.B. Effect of human immunodeficiency virus and antiretrovirals on outcomes of hepatitis C: a systematic review from an epidemiologic perspective. Clin Gastroenterol Hepatol, 2007;5(11):1321–8.e7.



59:  Epidemiology of Hepatocellular Carcinoma

  94. El‐Serag, H.B., Hampel, H., and Javadi, F. The association between diabetes and hepatocellular carcinoma: a systematic review of epidemiologic evidence. Clin Gastroenterol Hepatol, 2006;4(3):369–80.   95. Arase, Y., Kobayashi, M., Suzuki, F. et al. Effect of type 2 diabetes on risk for malignancies includes hepatocellular carcinoma in chronic hepatitis C. Hepatology, 2013;57(3):964–73.   96. Hutchinson, S.J., Bird, S.M., and Goldberg, D.J. Influence of alcohol on the progression of hepatitis C virus infection: a meta‐analysis. Clin Gastroenterol Hepatol, 2005;3(11):1150–9.  97. Kuang, S.Y., Lekawanvijit, S., Maneekarn, N. et  al. Hepatitis B 1762T/1764A mutations, hepatitis C infection, and codon 249 p53 mutations in hepatocellular carcinomas from Thailand. Cancer Epidemiol Biomarkers Prev, 2005; 14(2):380–4.  98. Chhatwal, J., Kanwal, F., Roberts, M.S., and Dunn, M.A. Cost‐­ effectiveness and budget impact of hepatitis C virus treatment with sofosbuvir and ledipasvir in the United States. Ann Intern Med, 2015;162(6):397–406.   99. Morgan, R.L., Baack, B., Smith, B.D. et al. Eradication of hepatitis C virus infection and the development of hepatocellular carcinoma: a meta‐analysis of observational studies. Ann Intern Med, 2013;158(5 Pt 1):329–37. 100. Kish, T., Aziz, A., and Sorio, M. Hepatitis C in a new era: a review of current therapies. P T, 2017;42(5):316–29. 101. Waziry, R., Hajarizadeh, B., Grebely, J. et al. Hepatocellular carcinoma risk following direct‐acting antiviral HCV therapy: A systematic review, meta‐ analyses, and meta‐regression. J Hepatol, 2017;67(6):1204–12. 102. Huang, A.C., Mehta, N., Dodge, J.L., Yao, F.Y., and Terrault, N.A. Direct‐ acting antivirals do not increase the risk of hepatocellular carcinoma recurrence after local‐regional therapy or liver transplant waitlist dropout. Hepatology, 2018;68(2):449–61. 103. Ioannou, G.N., Green, P.K., and Berry, K. HCV eradication induced by direct‐acting antiviral agents reduces the risk of hepatocellular carcinoma. J Hepatol, 2018;68(1):25–32. 104. Pol, S. Lack of evidence of an effect of Direct Acting Antivirals on the recurrence of hepatocellular carcinoma. J Hepatol, 2016;65:734–40. 105. El‐Serag, H.B., Kanwal, F., Richardson, P., and Kramer, J. Risk of hepatocellular carcinoma after sustained virological response in Veterans with hepatitis C virus infection. Hepatology, 2016;64(1):130–7. 106. Singal, A.K., Singh, A., Jaganmohan, S. et al. Antiviral therapy reduces risk of hepatocellular carcinoma in patients with hepatitis C virus‐related cirrhosis. Clin Gastroenterol Hepatol, 2010;8(2):192–9. 107. Chang, K.C., Hung, C.H., Lu, S.N. et al. A novel predictive score for hepatocellular carcinoma development in patients with chronic hepatitis C after sustained response to pegylated interferon and ribavirin combination therapy. J Antimicrob Chemother, 2012;67(11):2766–72. 108. Kanwal, F., Kramer, J., Asch, S.M. et al. Risk of hepatocellular cancer in HCV patients treated with direct‐acting antiviral agents. Gastroenterology, 2017;153(4):996–1005. e1. 109. Thrift, A.P., El‐Serag, H.B., and Kanwal, F. Global epidemiology and burden of HCV infection and HCV‐related disease. Nat Rev Gastroenterol Hepatol, 2017;14(2):122. 110. Volk, M.L., Tocco, R., Saini, S., and Lok, A.S. Public health impact of antiviral therapy for hepatitis C in the United States. Hepatology, 2009; 50(6):1750–5. 111. Kanwal, F. and El‐Serag, H.B. Hepatitis C virus treatment: the unyielding chasm between efficacy and effectiveness. Clin Gastroenterol Hepatol, 2014;12(8):1381–3. 112. Smith, B.D., Beckett, G.A., Yartel, A. et  al. Previous exposure to HCV among persons born during 1945–1965: prevalence and predictors, United States, 1999–2008. Am J Public Health, 2014;104(3):474–81. 113. Kanwal, F., Lok, A.S., and El‐Serag, H.B. CDC and USPSTF 2012 recommendations for screening for hepatitis C virus infection: overview and take‐ home messages. Clin Gastroenterol Hepatol, 2013;11(3):200–3. 114. Grant, B.F., Stinson, F.S., Dawson, D.A. et  al. Prevalence and co‐occurrence of substance use disorders and independent mood and anxiety disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Arch Gen Psychiatry, 2004;61(8):807–16. 115. Alcohol drinking. IARC Working Group, Lyon, 13‐20 October 1987. IARC Monogr Eval Carcinog Risks Hum, 1988;44:1–378. 116. World Cancer Research Fund/American Institute for Cancer Research. Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. American Institute for Cancer Research, Washington, D.C., 2007, pp. 1–537.

771

117. Turati, F., Galeone, C., Rota, M. et al. Alcohol and liver cancer: a systematic review and meta‐analysis of prospective studies. Ann Oncol, 2014; 25(8):1526–35. 118. Welzel, T.M., Graubard, B.I., Quraishi, S. et  al. Population‐attributable fractions of risk factors for hepatocellular carcinoma in the United States. Am J Gastroenterol, 2013;108(8):1314. 119. Makarova‐Rusher, O.V., Altekruse, S.F., McNeel, T.S. et  al. Population attributable fractions of risk factors for hepatocellular carcinoma in the United States. Cancer, 2016;122(11):1757–65. 120. Palmer, W.C. and Patel, T. Are common factors involved in the pathogenesis of primary liver cancers? A meta‐analysis of risk factors for intrahepatic cholangiocarcinoma. J Hepatol, 2012;57(1):69–76. 121. Heckley, G.A., Jarl, J., Asamoah, B.O., and G‐Gerdtham, U. How the risk of liver cancer changes after alcohol cessation: a review and meta‐analysis of the current literature. BMC Cancer, 2011;11:446‐. 122. Esposito, K., Chiodini, P., Colao, A., Lenzi, A., and Giugliano, D. Metabolic syndrome and risk of cancer: a systematic review and meta‐analysis. Diabetes Care, 2012;35(11):2402–11. 123. Farrell, G. Insulin resistance, obesity, and liver cancer. Clin Gastroenterol Hepatol, 2014;12(1):117–19. 124. Welzel, T.M., Graubard, B.I., Zeuzem, S. et  al. Metabolic syndrome increases the risk of primary liver cancer in the United States: a study in the SEER‐Medicare database. Hepatology, 2011;54(2):463–71. 125. Flegal, K.M., Carroll, M.D., Kit, B.K., and Ogden, C.L. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999–2010. JAMA, 2012;307(5):491–7. 126. Flegal, K.M., Kruszon‐Moran, D., Carroll, M.D., Fryar, C.D., and Ogden, C.L. Trends in Obesity Among Adults in the United States, 2005 to 2014. JAMA, 2016;315(21):2284–91. 127. Wang, P., Kang, D., Cao, W., Wang, Y., and Liu, Z. Diabetes mellitus and risk of hepatocellular carcinoma: a systematic review and meta‐analysis. Diabetes/Metab Res Rev, 2012;28(2):109–22. 128. Wang, C., Wang, X., Gong, G. et al. Increased risk of hepatocellular carcinoma in patients with diabetes mellitus: a systematic review and meta‐analysis of cohort studies. Int J Cancer, 2012;130(7):1639–48. 129. Chen, J., Han, Y., Xu, C., Xiao, T., and Wang, B. Effect of type 2 diabetes mellitus on the risk for hepatocellular carcinoma in chronic liver diseases: a meta‐analysis of cohort studies. Eur J Cancer Prev, 2015;24(2):89–99. 130. Singh, S., Singh, P.P., Singh, A.G., Murad, M.H., and Sanchez, W. Anti‐diabetic medications and the risk of hepatocellular cancer: a systematic review and meta‐analysis. Am J Gastroenterol, 2013;108(6):881–91; quiz 92. 131. Chen, H.P., Shieh, J.J., Chang, C.C. et al. Metformin decreases hepatocellular carcinoma risk in a dose‐dependent manner: population‐based and in vitro studies. Gut, 2013;62(4):606–15. 132. Flemming, J.A., Yang, J.D., Vittinghoff, E., Kim, W.R., and Terrault, N.A. Risk prediction of hepatocellular carcinoma in patients with cirrhosis: the ADRESS‐HCC risk model. Cancer, 2014;120(22):3485–93. 133. Saunders, D., Seidel, D., Allison, M., and Lyratzopoulos, G. Systematic review: the association between obesity and hepatocellular carcinoma–epidemiological evidence. Aliment Pharmacol Ther, 2010;31(10):1051–63. 134. Larsson, S.C. and Wolk, A. Overweight, obesity and risk of liver cancer: a meta‐analysis of cohort studies. Br J Cancer, 2007;97(7):1005–8. 135. Hassan, M.M., Abdel‐Wahab, R., Kaseb, A. et al. Obesity early in adulthood increases risk but does not affect outcomes of hepatocellular carcinoma. Gastroenterology, 2015;149(1):119–29. 136. Schlesinger, S., Aleksandrova, K., Pischon, T. et  al. Abdominal obesity, weight gain during adulthood and risk of liver and biliary tract cancer in a European cohort. Int J Cancer, 2013;132(3):645–57. 137. Jinjuvadia, R., Patel, S., and Liangpunsakul, S. The association between metabolic syndrome and hepatocellular carcinoma: systemic review and meta‐analysis. J Clin Gastroenterol 2014;48(2):172–7. 138. Shaw, J.E., Sicree, R.A., and Zimmet, P.Z. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract, 2010;87(1):4–14. 139. James, W.P. The epidemiology of obesity: the size of the problem. J Intern Med, 2008;263(4):336–52. 140. Zhao, J. and Lawless, M.W. Stop feeding cancer: pro‐inflammatory role of visceral adiposity in liver cancer. Cytokine, 2013;64(3):626–37. 141. Yu, J., Shen, J., Sun, T.T., Zhang, X., and Wong, N. Obesity, insulin resistance, NASH and hepatocellular carcinoma. Semin Cancer Biol, 2013;23(6 Pt B): 483–91. 142. Grundy, S.M. Metabolic syndrome: a multiplex cardiovascular risk factor. J Clin Endocrinol Metab, 2007;92(2):399–404.

772

THE LIVER:  REFERENCES

143. Ott, C. and Schmieder, R.E. The role of statins in the treatment of the metabolic syndrome. Curr Hypertens Rep, 2009;11(2):143–9. 144. Chan, K.K., Oza, A.M., and Siu, L.L. The statins as anticancer agents. Clin Cancer Res, 2003;9(1):10–9. 145. Gazzerro, P., Proto, M.C., Gangemi, G. et al. Pharmacological actions of statins: a critical appraisal in the management of cancer. Pharmacol Rev, 2012;64(1):102–46. 146. Chiu, H.F., Ho, S.C., Chen, C.C., and Yang, C.Y. Statin use and the risk of liver cancer: a population‐based case‐control study. Am J Gastroenterol, 2011;106(5):894–8. 147. Leung, H.W., Chan, A.L., Lo, D., Leung, J.H., and Chen, H.L. Common cancer risk and statins: a population‐based case‐control study in a Chinese population. Exp Opin Drug Safety, 2013;12(1):19–27. 148. Tsan, Y.T., Lee, C.H., Wang, J.D., and Chen, P.C. Statins and the risk of hepatocellular carcinoma in patients with hepatitis B virus infection. J Clin Oncol, 2012;30(6):623–30. 149. Tsan, Y.T., Lee, C.H., Ho, W.C. et al. Statins and the risk of hepatocellular carcinoma in patients with hepatitis C virus infection. J Clin Oncol, 2013; 31(12):1514–21. 150. Lai, S.W., Liao, K.F., Lai, H.C. et al. Statin use and risk of hepatocellular carcinoma. Eur J Epidemiol, 2013;28(6):485–92. 151. Friis, S., Poulsen, A.H., Johnsen, S.P. et al. Cancer risk among statin users: a population‐based cohort study. Int J Cancer, 2005;114(4):643–7. 152. Marelli, C., Gunnarsson, C., Ross, S. et  al. Statins and risk of cancer: a retrospective cohort analysis of 45,857 matched pairs from an electronic medical records database of 11 million adult Americans. J Am Coll Cardiol, 2011;58(5):530–7. 153. Friedman, G.D., Flick, E.D., Udaltsova, N. et al. Screening statins for possible carcinogenic risk: up to 9 years of follow‐up of 361,859 recipients. Pharmacoepidemiol Drug Safety, 2008;17(1):27–36. 154. McGlynn, K.A., Divine, G.W., Sahasrabuddhe, V.V. et  al. Statin use and risk of hepatocellular carcinoma in a U.S. population. Cancer Epidemiol, 2014;38(5):523–7. 155. Bjorkhem‐Bergman, L., Backheden, M., and Soderberg Lofdal, K. Statin treatment reduces the risk of hepatocellular carcinoma but not colon cancer‐ results from a nationwide case‐control study in Sweden. Pharmacoepidemiol Drug Safety, 2014;23(10):1101–6. 156. El‐Serag, H.B., Johnson, M.L., Hachem, C., and Morgana, R.O. Statins are associated with a reduced risk of hepatocellular carcinoma in a large cohort of patients with diabetes. Gastroenterology. 2009;136(5):1601–8. 157. Khurana, V., Saluja, A., Caldito, G., Fort, C., and Schiff, E. Statins are protective against hepatocellular cancer in patients with hepatitis C virus infection: half a million U.S. veterans’ study. Gastroenterology, 2005;128(4):A714. 158. Pradelli, D., Soranna, D., Scotti, L. et al. Statins and primary liver cancer: a meta‐analysis of observational studies. Eur J Cancer Prev, 2013;22(3):229–34. 159. Singh, S., Singh, P.P., Singh, A.G., Murad, M.H., and Sanchez, W. Statins are associated with a reduced risk of hepatocellular cancer: a systematic review and meta‐analysis. Gastroenterology, 2013;144(2):323–32. 160. Franciosi, M., Lucisano, G., Lapice, E. et al. Metformin therapy and risk of cancer in patients with type 2 diabetes: systematic review. PloS One, 2013;8(8):e71583‐e. 161. Zhang, H., Gao, C., Fang, L., Zhao, H.C., and Yao, S.K. Metformin and reduced risk of hepatocellular carcinoma in diabetic patients: a meta‐analysis. Scand J Gastroenterol, 2013;48(1):78–87. 162. Vernon, G., Baranova, A., and Younossi, Z.M. Systematic review: the epidemiology and natural history of non‐alcoholic fatty liver disease and non‐­alcoholic steatohepatitis in adults. Aliment Pharmacol Ther, 2011;34(3):274–85. 163. Stepanova, M., De Avila, L., Afendy, M. et al. Direct and indirect economic burden of chronic liver disease in the United States. Clin Gastroenterol Hepatol, 2017;15(5):759–66. e5. 164. Williams, C.D., Stengel, J., Asike, M.I. et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle‐aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology, 2011;140(1):124–31. 165. Adams, L.A. and Lindor, K.D. Nonalcoholic fatty liver disease. Ann Epidemiol, 2007;17(11):863–9. 166. Wattacheril, J. and Chalasani, N. Nonalcoholic fatty liver disease (NAFLD): is it really a serious condition? Hepatology, 2012;56(4):1580–4. 167. Ahmed, A., Wong, R.J., and Harrison, S.A. Nonalcoholic fatty liver disease review: diagnosis, treatment, and outcomes. Clin Gastroenterol Hepatol, 2015;13(12):2062–70.

168. Wong, R.J., Cheung, R., and Ahmed, A. Nonalcoholic steatohepatitis is the most rapidly growing indication for liver transplantation in patients with hepatocellular carcinoma in the US. Hepatology, 2014;59(6):2188–95. 169. Mittal, S., El‐Serag, H.B., Sada, Y.H. et al. Hepatocellular carcinoma in the absence of cirrhosis in United States veterans is associated with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol, 2016;14(1):124–31. e1. 170. White, D.L., Kanwal, F., and El‐Serag, H.B. Association between nonalcoholic fatty liver disease and risk for hepatocellular cancer, based on systematic review. Clin Gastroenterol Hepatol, 2012;10(12):1342–59 e2. 171. Kanwal, F., Kramer, J.R., Mapakshi, S. et al. Risk of hepatocellular cancer in patients with non‐alcoholic fatty liver disease. Gastroenterology, 2018; 155(6):1828–37.e2. 172. Singal, A.G., Manjunath, H., Yopp, A.C. et al. The effect of PNPLA3 on fibrosis progression and development of hepatocellular carcinoma: a meta‐ analysis. Am J Gastroenterol, 2014;109(3):325–34. 173. Marchesini, G., Petta, S., and Dalle Grave, R. Diet, weight loss, and liver health in nonalcoholic fatty liver disease: pathophysiology, evidence, and practice. Hepatology, 2016;63(6):2032–43. 174. Burza, M.A., Pirazzi, C., Maglio, C. et  al. PNPLA3 I148M (rs738409) genetic variant is associated with hepatocellular carcinoma in obese individuals. Dig Liver Dis, 2012;44(12):1037–41. 175. Yang, J.D., Kim, W.R., Coelho, R. et al. Cirrhosis is present in most patients with hepatitis B and hepatocellular carcinoma. Clin Gastroenterol Hepatol, 2011;9(1):64–70. 176. Chayanupatkul, M., Omino, R., Mittal, S. et al. Hepatocellular carcinoma in the absence of cirrhosis in patients with chronic hepatitis B virus infection. J Hepatol, 2017;66(2):355–62. 177. Lok, A.S., Everhart, J.E., Wright, E.C. et al. Maintenance peginterferon therapy and other factors associated with hepatocellular carcinoma in patients with advanced hepatitis C. Gastroenterology, 2011;140(3):840–9; quiz e12. 178. Paradis, V., Zalinski, S., Chelbi, E. et  al. Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant liver fibrosis: a pathological analysis. Hepatology, 2009;49(3):851–9. 179. Yasui, K., Hashimoto, E., Komorizono, Y. et al. Characteristics of patients with nonalcoholic steatohepatitis who develop hepatocellular carcinoma. Clin Gastroenterol Hepatol, 2011;9(5):428–33; quiz e50. 180. Mieli‐Vergani, G., Vergani, D., Czaja, A.J. et al. Autoimmune hepatitis. Nat Rev Dis Primers, 2018;4:18017. 181. Liberal, R., Krawitt, E.L., Vierling, J.M. et al. Cutting edge issues in autoimmune hepatitis. J Autoimmun, 2016;75:6–19. 182. Zachou, K., Muratori, P., Koukoulis, G.K. et al. Autoimmune hepatitis – current management and challenges. Aliment Pharmacol Ther, 2013;38(8): 887–913. 183. Vierling, J.M. Autoimmune hepatitis and overlap syndromes: diagnosis and management. Clin Gastroenterol Hepatol, 2015;13(12):2088–108. 184. Czaja, A.J. and Freese, D.K. Diagnosis and treatment of autoimmune hepatitis. Hepatology, 2002;36(2):479–97. 185. Manns, M.P., Czaja, A.J., Gorham, J.D. et al. Diagnosis and management of autoimmune hepatitis. Hepatology, 2010;51(6):2193–213. 186. Montano‐Loza, A.J., Carpenter, H.A., and Czaja, A.J. Predictive factors for hepatocellular carcinoma in type 1 autoimmune hepatitis. Am J Gastroenterol, 2008;103(8):1944–51. 187. Gronbaek, L., Vilstrup, H., and Jepsen, P. Autoimmune hepatitis in Denmark: incidence, prevalence, prognosis, and causes of death. A nationwide registry‐based cohort study. J Hepatol, 2014;60(3):612–17. 188. Yeoman, A.D., Al‐Chalabi, T., Karani, J.B. et al. Evaluation of risk factors in the development of hepatocellular carcinoma in autoimmune hepatitis: implications for follow‐up and screening. Hepatology, 2008;48(3):863–70. 189. EASL Clinical Practice Guidelines: autoimmune hepatitis. J Hepatol, 2015;63(4):971–1004. 190. Tansel, A., Katz, L.H., El‐Serag, H.B. et al. Incidence and determinants of hepatocellular carcinoma in autoimmune hepatitis: a systematic review and meta‐analysis. Clin Gastroenterol Hepatol, 2017;15(8):1207–17.e4. 191. Younossi, Z., Stepanova, M., Ong, J.P. et al. Non‐alcoholic steatohepatitis is the fastest growing cause of hepatocellular carcinoma in liver transplant candidates. Clin Gastroenterol Hepatol, 2019;17(4):748–75. 192. Balakrishnan, M. and El‐Serag, H.B. Editorial: NAFLD‐related hepatocellular carcinoma  –  increasing or not? With or without cirrhosis? Aliment Pharmacol Ther, 2018;47(3):437–8. 193. El‐Serag, H.B. and Kanwal, F. Obesity and hepatocellular carcinoma: hype and reality. Hepatology, 2014;60(3):779–81.

60

Mutations and Genomic Alterations in Liver Cancer Jessica Zucman‐Rossi1,2 and Jean‐Charles Nault1,3 Centre de Recherche des Cordeliers, Sorbonne Université, Inserm, Université de Paris, Functional Genomics of Solid Tumors Laboratory, Paris, France 2 Hôpital Europeen Georges Pompidou, Assistance Publique Hôpitaux de Paris, Paris, France 3 Service d’hépatologie, Hôpital Jean Verdier, Hôpitaux Universitaires Paris‐Seine‐Saint‐Denis, Assistance Publique Hôpitaux de Paris, Bondy, France 1

INTRODUCTION Liver cancer delineates a heterogeneous group of solid tumors that include mainly epithelial tumors. As in other solid cancers, tumor development is the result of an accumulation of genetic and epigenetic alterations. The mechanism by which these genomic alterations are accumulated in liver cells and contribute to the process of carcinogenesis is closely related to the natural history of the disease and origin of the cancer cell. In adults, liver tumors are a heterogeneous group of cancers. Hepatocellular carcinoma (HCC) and cholangiocarcinoma (CCK) are the two major types, derived from the malignant transformation of hepatocytes and cholangiocytes, respectively [1, 2]. Among these two types of tumors, pathologists have identified distinct phenotypic subtypes of tumors with, for example, fibrolamellar carcinoma as a specific HCC subtype and mixed hepatocholangiocarcinoma, which exhibits both hepatocyte and cholangiocyte differentiation [3]. In childhood, liver cancers are rare, but the most frequent is hepatoblastoma, an embryonal derived tumor that develops in normal liver, and, rarely, hepatocellular carcinoma, usually developing in cirrhosis due to severe metabolic disease or viral infection. Understanding the genomic alterations that initiate these various subtypes of tumors is a major objective in deciphering the mechanisms of hepatocarcinogenesis and then translating this knowledge into clinical care of patients [4]. Indeed, if cancer cells in the liver had oncogene addiction, these defects could constitute good candidate targets for therapy. Genomic alterations could also be transformed into biomarkers that are useful for diagnosis and prognosis assessment. Following the recent

development of high‐throughput techniques we are now able to sequence the whole coding sequence of a tumor (WES) or the complete whole‐genome sequence (WGS), which includes the noncoding sequences, and all the mRNA transcripts expressed in a tumor (RNAseq). Using these techniques we can catalogue the mutations and chromosome aberrations in each tumor to identify somatic mutations that occur only in the tumor and are not found in the non‐tumor, Somatic mutations and chromosomal aberrations are genomic defects that occur in tumor cells. Among them, we have to discriminate between mutations that are drivers for the carcinogenesis process and mutations that are passengers [5]. These cancer driver mutations are usually clonally selected during the evolution of cancer cells because they have functional consequences that promote cell survival or cell proliferation and consequently tumor development. Cancer driver genes are usually defined by their recurrence in specific tumor types, or more broadly across tumors, and importantly by their functional consequences, with mutations altering cell signaling pathways or metabolism, ­proliferation, invasiveness, etc. Some “cancer driver genes” are oncogenes, and their activation leads to tumor promotion; others are tumor suppressor genes, which control cell proliferation and cell maintenance, and need to be inactivated to promote carcinogenesis. In contrast, “passenger mutations” are also acquired by tumor cells but they have no consequences in terms of the mechanism of carcinogenesis. Tumor cell development is closely under the control of cells from the microenvironment, but currently, no recurrent genomic alterations have been identified in immune and stromal cells in the liver. Finally, germline mutations that are present in all the cells of the body could occur in genes predisposing to tumor development; they frequently alter

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

774

THE LIVER:  GENOMIC PROFILE OF HEPATOCELLULAR CARCINOMA

tumor suppressor genes with one mutation in the germline and the other in the tumor. Here, we will review the profiles of mutations and genomic alterations in the major subtypes of liver tumors.

Early genomic alterations in hepatocellular carcinogenesis

GENOMIC PROFILE OF HEPATOCELLULAR CARCINOMA

Genetic predisposition to HCC

The most frequent malignant tumors in the liver are hepatocellular carcinoma, a serious disease that is the third most common case of cancer death worldwide [6]. As in other solid tumors, hepatocyte transformation in malignant cells and tumors is the result of a multistep process. The majority of HCCs develop following (i) exposure to various risk factors, including mainly hepatitis B or C (HBV or HCV) viral infection, excessive alcohol intake, metabolic syndrome, hemochromatosis, and other genetic metabolic diseases; (ii) the development of a chronic liver disease (in 80–90% of cases, HCC is diagnosed in a ­cirrhotic liver); and (iii) cirrhosis, in which dysplastic nodules are the lesions most at risk of malignant transformation to HCC. After transformation, small HCCs usually progress to more advanced tumors and can acquire an aggressive phenotype with vascular invasion and intrahepatic metastasis; more rarely, extrahepatic metastasis may occur. During this process, genomic and epigenomic alterations progressively accumulate in hepatocytes

Risk factors

A family history of HCC is rarely described but familial aggregation of HCC cases in Asia in association with HBV or HCV infection suggest that genetic factors could modulate the risk of  HCC occurrence [7. 8]. Only few rare mendelian genetic ­diseases predispose to the development of HCC. It is mainly metabolic inherited diseases such as glycogenosis type 1a (G6PC), hemochromatosis (HFE), Wilson disease (ATP7B), tyrosinemia type 1 (FAH), porphyria acute intermittent (HMBS), cutanea tarda (UROD) and alpha 1 antitrypsin deficiency (SERPINA1) that cause cirrhosis and lead to secondary HCC [9]. Several genome‐wide association studies (GWAS) have been conducted in Asia to identify genetic polymorphisms that could predispose to HCC in the population of patients with tumors related to HBV or HCV infection (cases) compared to individuals from the general population or infected patients without HCC (controls) [10–15]. Single nucleotide polymorphisms

Malignant transformation

Initiation

Tumor progression

Accumulation of genetic and epigenetic alterations

Risk factors Hepatitis B and C Alcohol, Obesity Aflatoxin B1 Hemochromatosis

Cirrhosis

and cancer‐driving events that give a functional advantage in term of cell survival or proliferation are positively selected in the clonal tumor evolution (Figure 60.1).

Low grade dysplastic nodule

TERT prom

Low grade dysplastic nodule

TERT prom

TP53, CTNNB1 ARID1A, RPS6KA3, NFE2L2….

TP53, RB1, SF3B1, G3 subgroup

Hepatocellular carcinoma

Advanced HCC

Telomerase reactivation

95%

Dedifferentiation Proliferation

)

al V2 ion , AA t r se BV l in s (H a r Vi esi en tag

mu

Normal liver 5%

CTNNB1 exon 3 mutation

Hepatocellular adenoma

TERT promoter mutations

Figure 60.1  Genetic and genomic alterations in hepatocellular carcinoma in cirrhotic and noncirrhotic liver according to time of evolution.



60:  Mutations and Genomic Alterations in Liver Cancer

(SNP) associated with increased HCC risk were located in ­various genes: DDX18, KIF1B, DEPDC5, MICA, GRIK1, and STAT4. However, only STAT4 had currently been validated in a case–control study conducted in a Western country. Other GWAS analyses performed in patients with non‐alcoholic fatty liver disease identified two SNPs in the patatin‐like phospholipase domain‐containing 3 (PNPLA3) and in transmembrane 6  superfamily 2 (TM6SF2), two genes involved in lipid ­metabolism. These two polymorphisms were further validated as being associated with HCC risk in patients with alcohol or non‐alcoholic liver disease [16–19].

Viral insertional mutagenesis HBV is one of the most common causes of chronic liver diseases and HCC worldwide, with a high incidence and prevalence of the disease in Asian and African countries. HBV promotes HCC development indirectly through cirrhosis but also has a direct oncogenic effect due to action of viral oncoproteins such as HBx and mechanisms of viral insertional mutagenesis [20, 21]. Direct mechanisms of carcinogenesis explain the development of HBV‐related HCC on noncirrhotic liver. HBV is a DNA virus that stays in covalently closed circular DNA in the nucleus but is also inserted in human DNA and consequently can induce insertional mutagenesis [21–24]. When HBV inserts near a cancer gene, it can modify its expression or function through a cis transactivation mechanism and promote HCC development by proliferation of clonal transformed hepatocytes. Recurrent viral insertion in cancer genes leading to their overexpression has been described in TERT, CCNA2, CCNE1, and MLL4. Interestingly, only a sequence of the virus including the HBx region is found as recurrently inserted in human HCC, suggesting an important functional role of this viral region that includes several promoter/enhancer elements in insertional mutagenesis. Recently, other viruses have been described in human HCC, with recurrent clonal insertion of adeno‐associated virus type 2 (AAV2) in the human tumor genome [25]. AAV2 is a defective DNA virus that requires infection by a helper virus to produce new infective virions [26–28]. AAV2 is considered to be a nonpathogenic virus and frequent antibodies against AAV2 are found in adults, suggesting a high prevalence of the infection in humans. In human HCC, recurrent viral insertions of the 3′ inverse tandem region (ITR) of the virus have been identified in cancer genes such as TERT, CCNA2, CCNE1, and TNFSF10 [25, 29–32]. These viral insertions lead to an overexpression of the inserted genes that promotes malignant transformation. Interestingly, most of these HCC developed on normal liver without known risk factors. These data demonstrate that AAV2 infection is a new etiology of HCC developing in normal liver in humans.

TERT promoter mutation is the earliest and most frequent somatic event in HCC Telomerase is a key enzyme involved in cell survival that prevents telomere shortening after cell division and subsequent senescence (see Chapter  74). In humans, telomerase is not expressed in mature hepatocytes. However, during the process

775

of hepatocarcinogenesis, telomerase activation is required to ensure perpetual proliferation and to maintain telomere length in malignant hepatocytes. In HCC telomerase expression in identified in 90–95% of tumors. Telomerase is activated mainly by mutation in the promoter of TERT, which codes for telomerase. These mutations are identified in 40–60% of HCCs. They are somatic and clustered mainly at two hot‐spots, located 124 and 146 base‐pairs 5′ from the ATG initiating telomerase ­traduction [1, 29, 33, 34]. TERT promoter mutation creates a binding site (TTCCGA) for transcription factors of the ETS/ TCF family, such as GBP. Recently, a germline TERT promoter mutation has been identified in a patient with HCC, but the role of TERT in a genetic predisposition to HCC development remains to be explored [29]. Other mechanisms that activate telomerase have been identified in HCCs. These include viral insertions of HBV or AAV2 within the promoter, amplification of the gene, and chromosomal translocations putting TERT gene expression under the control of a strong hepatocyte promoter, such as hepatocyte transporters (SLC12A7 and SLC7A2), albumin (ALB), alcohol dehydrogenase (ADH), and hydroxysteroid 17‐beta dehydrogenases (HSD17B) [29, 31]. Ultimately, only 5–10% of HCCs do not express telomerase, and these tumors could use an alternative mechanism of telomere maintenance based on DNA recombination [35]. During the sequence of hepatocarcinogenesis that starts from exposure to various risk factors, chronic liver disease occurrence, and cirrhosis followed by HCC development, TERT promoter mutations have been identified within cirrhosis in 6% of low‐grade dysplastic nodules and 19% of high‐grade dysplastic nodules [36]. This frequency of mutation increased to 60% in early and more advanced HCC, whereas mutations in common other cancer driver genes frequently altered in HCC were not mutated in cirrhotic dysplastic nodules. Consequently, TERT promoter mutation is the earliest recurrent mutation recurrently identified in precancerous cirrhotic lesions.

Mutational signatures result from environmental exposure The accumulation of somatic mutations in tumor cells can be the result of various aggressive genotoxic factors and/or due to a defect in DNA damage repair. The Sanger Institute has developed a powerful method that aims to identify “mutational signatures” that originate from various specific mutational processes in tumor cells [37]. This method is based on the analysis of the six types of possible nucleotide mutations in the DNA, by taking into account the DNA context in which these mutations occur. Ninety‐six possible combinations have been recorded and the number of mutations in each category are counted using whole exome sequencing or whole genome sequencing data in tumors. In HCC, according to the COSMIC nomenclature, five ubiquitous signatures have been identified [34, 33, 38]. These correspond to signatures 16 and 12, which are specific to the liver, signature 4, which is related to tobacco or other adduct events, and signatures 1 and 5, which are related to aging. Five sporadic signatures were also identified: signature 24 is specific to aflatoxin B1 exposure, signature 22 is specific to aristolochic

776

THE LIVER:  GENOMIC PROFILE OF HEPATOCELLULAR CARCINOMA

Ubiquitous mutational signatures C>A

Signature 1 Clock-like: related to age Deamination of methylcytosine

Signature 5 Clock-like: related to age

C>G

C>T

T>A

T>C

T>G

0.2 0.1 0.0

C>A

C>G

C>T

T>A

T>C

T>G

C>A

C>G

C>T

T>A

T>C

T>G

0.10 0.05 0.00

Signature 4 Exposure to polycyclic aromatic hydrocarbons (including tobacco)

Signature 12 Liver specific

0.10 0.05 0.00

C>A

C>G

C>T

T>A

T>C

T>G

C>A

C>G

C>T

T>A

T>C

T>G

0.10 0.05 0.00

Signature 16 Liver specific Male, alcohol, tobacco

0.10 0.05 0.00

Sporadic mutational signatures Signature 22 Aristocholic acid

C>A

C>G

C>T

T>A

T>C

T>G

C>A

C>G

C>T

T>A

T>C

T>G

C>A

C>G

C>T

T>A

T>C

T>G

C>A

C>G

C>T

T>A

T>C

T>G

0.2 0.1 0.0

Signature 24 Aflatoxin B1

0.2 0.1 0.0

Signature 17 and 23 Unexplained

0.2 0.1 0.0

0.2 0.1 0.0

Figure 60.2  Various mutational signatures identified in hepatocellular carcinoma.

exposure, signature 6 occurs in exceptional cases with microsatellite instability, and signatures 17 and 23 remain unexplained (Figure 60.2).

Signaling pathways and cell process recurrently altered in HCC by genomic alterations Various signaling pathways and mechanisms of cellular maintenance have been shown to be altered by gene mutations and

chromosomal alterations [29, 33, 34, 39–45]. The most important functionally are the following: • Telomere maintenance through telomerase reactivation has been identified as a key mechanism required to avoid ­telomere shortening during uncontrolled cell proliferation. The different mechanisms leading to telomerase reactivation have been described in the section entitled TERT ­promoter mutation is the earliest and most frequent somatic event in HCC.



60:  Mutations and Genomic Alterations in Liver Cancer

• The WNT/β‐catenin pathway is one of the main signaling pathway aberrantly activated in HCC. In inactivated state, β‐catenin is phosphorylated by an inhibitory complex ­composed of AXIN1/APC and GSK3B. Phosphorylation of β‐catenin induces its degradation by the proteasome. Activating mutations of CTNNB1 (coding for the β‐catenin) are observed in around 20–40% of HCCs. They impair the phosphorylation of β‐catenin and protect it from degradation by the proteasome. HCCs with CTNNB1 mutations have a specific transcriptional program with overexpression of target genes of these pathways such as GLUL, LGR5, and REG3A. CTNNB1‐mutated HCCs are characterized at the histological level by intratumor cholestasis and good differentiation. Inactivating mutations of AXIN1 are observed in around 15% of the cases. They are exclusive from CTNNB1 mutations and harbor a transcriptomic profile that is different from that of HCCs harboring mutations of CTNNB1. • Control of the cell cycle can be altered by the disruption of tumor suppressor mechanisms in the TP53 and CDKN2A/RB1 pathways, which leads to uncontrolled progression in the cell cycle and breakdown of DNA repair machinery. Mutations of TP53 are observed in 20–50% of HCCs and are more frequently observed in HBV‐related HCCs. Inactivating mutations/deletions of CDKN2A and RB1 are described in around 5–20% of HCCs. Recently, a new subgroup of HCC accounting for 7% of all tumors was identified. These are associated with a poor prognosis and characterized by an overexpression of the cyclin genes CCNA2 and CCNE1 induced by viral insertions and chromosomal rearrangements. Interestingly, these tumors harbor a specific structural rearrangement, with  several tandem duplications and templated insertions, frequently activating TERT promoter. • Epigenetic modifier genes: Recurrent genetic alterations have been identified in genes modeling the epigenome, including mutations in partners of chromatin remodeling machinery such as ARID1A (8–20%) or ARID2 (5–15%) and in members of the family of histone methyltransferase genes such as MLL, MLL2, MLL3, and MLL4. • The oxidative stress pathway is activated in a subset of HCCs by activating mutations of the transcription factor NRF2 (encoded by NFE2L2, 5–10% of mutations) or inactivating mutations of the inhibitor of the pathway KEAP1 (2–10%). Activation of NRF2/KEAP1 pathway controlled a detoxification program protecting the tumor cells from the toxicity of reactive oxygen species. • Genes highly expressed are frequently mutated and somatic mutations in coding for albumin, fibrinogen, cytochrome P450 enzymes, and alcohol dehydrogenase are observed specifically in liver tumors such as HCCs. These genes are highly expressed in mature hepatocytes and classical markers of hepatocyte differentiation. High rates of insertions/deletions have been identified in these hepato‐specific genes resulting from collision between the replication and transcription machineries. Mutations could be the result of accidents of the cell machinery. The functional consequence of these mutations in the process of carcinogenesis, if any, remains to be proven. • The RAS/RAF/MAP kinase and AKT/mTor pathways are activated by rare genetic alterations (20 pre‐treatment were delisted as opposed to 35% with MELD 60% having detectable HBV DNA at the time of transplant, high rates of HBsAg seroconversion and viral suppression were possible. Oral nucleosides have had a major impact in the management of patients with HBV‐related cirrhosis. Antiviral therapy has been shown to improve outcomes in decompensated cirrhosis, especially with early treatment initiation [71]. Entecavir and tenofovir are recommended as preferred first‐line agents in patients with decompensated cirrhosis [72].

Currently, an individualized approach to use of HBIg is recommended [73]. HBIg during the anhepatic phase and for 5–7 days or no HBIg is reasonable in low‐risk patients. Combination antiviral therapy with a nucleoside analog and HBIg may be preferred in patient at high risk of progressive disease. This would include the presence of drug resistance, high DNA at time of transplant, HDV‐ and HIV‐coinfected patients and questionable medication compliance.

Transmission of HBV infection by donor grafts De novo acquisition of HBV infection from allografts previously infected with HBV (hepatitis B core antibody (HBcAb)‐ positive, HBsAg‐negative) into liver HBsAg‐negative liver recipients occurs in up to 75%, but varies with the HBV immune status of the recipient. A systematic review showed risk of de novo HBV infection to be higher in HBV‐naive recipients compared to recipients with anti‐HBc‐ and/or anti‐HBs‐positive recipients (48% vs. 15%) [74]. Antiviral prophylaxis with a nucleoside (preferably entecavir or tenofovir) has been shown to be effective in preventing de novo infection and should be started immediately after transplant. Use of HBIg is not necessary. There is no data that the anti‐HBs status of an HBsAg‐ negative, HBcAb‐positive donor impacts the risk of HBV transmission to recipients.

CONCLUSION The advent of highly effective DAA therapy has revolutionized the peri‐OLT management of chronic HBV and HCV. As access to DAA therapy expands, the number of patients with HBV and HCV waitlisted for transplant will continue to fall. Furthermore, the use of post‐OLT DAA therapy will lead to decreased graft loss in HBV‐ and HCV‐positive recipients and provide potential benefit to waitlisted patients without virus‐induced ESLD who stand to benefit from the expanded use of HBV‐ and HCV‐positive donor organs.

REFERENCES 1. Forns, X., Garcia‐Retortillo, M., Serrano, T. et al. Antiviral therapy of patients with decompensated cirrhosis to prevent recurrence of hepatitis C after liver transplantation. J Hepatol, 2003;39:389–96. 2. Iacobellis, A., Siciliano, M., Perri, F. et al. Peginterferon alfa‐2b and ribavirin in patients with hepatitis C virus and decompensated cirrhosis: a controlled study. J Hepatol, 2007;46:206–12. 3. Charlton, M., Gane, E., Manns, M.P. et al. Sofosbuvir and ribavirin for treatment of compensated recurrent hepatitis C virus infection after liver transplantation. Gastroenterology, 2015;148:108–17. 4. Manns, M., Samuel, D., Gane, E.J. et al. Ledipasvir and sofosbuvir plus ribavirin in patients with genotype 1 or 4 hepatitis C virus infection and advanced liver disease: a multicentre, open‐label, randomised, phase 2 trial. Lancet Infect Dis, 2016;16:685–97. 5. Berenguer, M. Systematic review of the treatment of established recurrent hepatitis C with pegylated interferon in combination with ribavirin. J Hepatol, 2008;49:274–87. 6. Reau, N., Kwo, P.Y., Rhee, S. et al. MAGELLAN‐2: safety and efficacy of glecaprevir/pibrentasvir in liver or renal transplant adults with chronic hepatitis C genotype 1–6 infection. J Hepatol, 2017;66:S90–S91.



69:  BIOLOGICAL PRINCIPLES AND CLINICAL ISSUES UNDERLYING LIVER TRANSPLANTATION

  7. Saxena, V., Khungar, V., Verna, E.C. et  al. Safety and efficacy of current direct‐acting antiviral regimens in kidney and liver transplant recipients with hepatitis C: results from the HCV‐TARGET Study. Hepatology, 2017;66:1090–101.   8. Kim, W.R., Lake, J.R., Smith, J.M. et al. OPTN/SRTR 2016 Annual Data Report: Liver. Am J Transplant, 2018;18(Suppl 1):172–253.   9. Goldberg, D., Ditah, I.C., Saeian, K. et  al. Changes in the prevalence of hepatitis C virus infection, nonalcoholic steatohepatitis, and alcoholic liver disease among patients with cirrhosis or liver failure on the waitlist for liver transplantation. Gastroenterology, 2017;152:1090–9 e1091. 10. Chazouilleres, O., Kim, M., Combs, C. et  al. Quantitation of hepatitis C virus RNA in liver transplant recipients. Gastroenterology, 1994;106: 994–9. 11. Gane, E.J., Portmann, B.C., Naoumov, N.V. et  al. Long‐term outcome of hepatitis C infection after liver transplantation. N Engl J Med, 1996;334:815–20. 12. Gane, E.J., Naoumov, N.V., Qian, K.P. et al. A longitudinal analysis of hepatitis C virus replication following liver transplantation. Gastroenterology, 1996;110:167–77. 13. Berenguer, M., Prieto, M., Rayon, J.M. et  al. Natural history of clinically compensated hepatitis C virus‐related graft cirrhosis after liver transplantation. Hepatology, 2000;32:852–8. 14. Berenguer, M., Ferrell, L., Watson, J. et al. HCV‐related fibrosis progression following liver transplantation: increase in recent years. J Hepatol, 2000;32:673–84. 15. Wiesner, R.H., Sorrell, M., and Villamil, F., International Liver Transplantation Society Expert P: Report of the first International Liver Transplantation Society expert panel consensus conference on liver transplantation and hepatitis C. Liver Transpl, 2003;9:S1–9. 16. Ross, A.S., Bhan, A.K., Pascual, M. et al. Pegylated interferon alpha‐2b plus ribavirin in the treatment of post‐liver transplant recurrent hepatitis C. Clin Transplant, 2004;18:166–73. 17. Sharma, P., Marrero, J.A., Fontana, R.J. et al. Sustained virologic response to therapy of recurrent hepatitis C after liver transplantation is related to early virologic response and dose adherence. Liver Transpl, 2007;13:1100–8. 18. Panel AIHG: Hepatitis C guidance: AASLD‐IDSA recommendations for testing, managing, and treating adults infected with hepatitis C virus. Hepatology, 2015;62:932–54. 19. Mandorfer, M., Kozbial, K., Schwabl, P. et al. Sustained virologic response to interferon‐free therapies ameliorates HCV‐induced portal hypertension. J Hepatol, 2016;65:692–699. 20. Pascasio, J.M., Vinaixa, C., Ferrer, M.T. et al. Clinical outcomes of patients undergoing antiviral therapy while awaiting liver transplantation. J Hepatol, 2017;67:1168–76. 21. Poordad, F., Schiff, E.R., Vierling, J.M. et  al. Daclatasvir with sofosbuvir and ribavirin for hepatitis C virus infection with advanced cirrhosis or post‐ liver transplantation recurrence. Hepatology, 2016;63:1493–505. 22. Foster, G.R., Irving, W.L., Cheung, M.C. et al. Impact of direct acting antiviral therapy in patients with chronic hepatitis C and decompensated cirrhosis. J Hepatol, 2016;64:1224–31. 23. Belli, L.S., Duvoux, C., Berenguer, M. et al. ELITA consensus statements on the use of DAAs in liver transplant candidates and recipients. J Hepatol, 2017;67:585–602. 24. Facciorusso, A., Del Prete, V., Turco, A. et  al. Long‐term liver stiffness assessment in hepatitis C virus patients undergoing antiviral therapy: results from a 5‐year cohort study. J Gastroenterol Hepatol, 2018;33:942–9. 25. Mauro, E., Crespo, G., Montironi, C. et al. Portal pressure and liver stiffness measurements in the prediction of fibrosis regression after sustained virological response in recurrent hepatitis C. Hepatology, 2018;67:1683–94. 26. Lens, S., Alvarado‐Tapias, E., Marino, Z. et al. Effects of all‐oral anti‐viral therapy on HVPG and systemic hemodynamics in patients with hepatitis C virus‐associated cirrhosis. Gastroenterology, 2017;153:1273–83 e1271. 27. Burchill, M.A., Roby, J.A., Crochet, N. et  al. Rapid reversal of innate immune dysregulation in blood of patients and livers of humanized mice with HCV following DAA therapy. PLoS One, 2017;12:e0186213. 28. Bruix, J., Takayama, T., Mazzaferro, V. et al. Adjuvant sorafenib for hepatocellular carcinoma after resection or ablation (STORM): a phase 3, randomised, double‐blind, placebo‐controlled trial. Lancet Oncol, 2015;16:1344–54. 29. Conti, F., Buonfiglioli, F., Scuteri, A. et al. Early occurrence and recurrence of hepatocellular carcinoma in HCV‐related cirrhosis treated with direct‐acting antivirals. J Hepatol, 2016;65:727–33.

933

30. Reig, M., Marino, Z., Perello, C. et al. Unexpected high rate of early tumor recurrence in patients with HCV‐related HCC undergoing interferon‐free therapy. J Hepatol, 2016;65:719–26. 31. ANRS Collaborative Study Group on Hepatocellular Carcinoma. Lack of evidence of an effect of direct‐acting antivirals on the recurrence of hepatocellular carcinoma: data from three ANRS cohorts. J Hepatol, 2016; 65:734–40. 32. Terrault, N.A., McCaughan, G.W., Curry, M.P. et  al. International Liver Transplantation Society Consensus Statement on Hepatitis C Management in Liver Transplant Candidates. Transplantation, 2017;101:945–55. 33. Salazar, J., Saxena, V., Kahn, J.G. et al. Cost‐effectiveness of direct‐acting antiviral treatment in hepatitis C‐infected liver transplant candidates with compensated cirrhosis and hepatocellular carcinoma. Transplantation, 2017;101:1001–8. 34. Curry, M.P., Forns, X., Chung, R.T. et al. Sofosbuvir and ribavirin prevent recurrence of HCV infection after liver transplantation: an open‐label study. Gastroenterology, 2015;148:100–7 e101. 35. Yoshida, E.M., Kwo, P., Agarwal, K. et al. Persistence of virologic response after liver transplant in hepatitis C patients treated with ledipasvir/sofosbuvir plus ribavirin pretransplant. Ann Hepatol, 2017;16:375–81. 36. Picciotto, F.P., Tritto, G., Lanza, A.G. et al. Sustained virological response to antiviral therapy reduces mortality in HCV reinfection after liver transplantation. J Hepatol, 2007;46:459–65. 37. Kwo, P.Y., Mantry, P.S., Coakley, E. et al. An interferon‐free antiviral regimen for HCV after liver transplantation. N Engl J Med, 2014;371:2375–82. 38. Charlton, M., Everson, G.T., Flamm, S.L. et al. Ledipasvir and sofosbuvir plus ribavirin for treatment of HCV infection in patients with advanced liver disease. Gastroenterology, 2015;149:649–59. 39. Fontana, R.J., Brown, R.S., Jr., Moreno‐Zamora, A. et al. Daclatasvir combined with sofosbuvir or simeprevir in liver transplant recipients with severe recurrent hepatitis C infection. Liver Transpl, 2016;22:446–58. 40. Terrault, N.A., Berenguer, M., Strasser, S.I. et  al. International Liver Transplantation Society Consensus Statement on Hepatitis C Management in Liver Transplant Recipients. Transplantation, 2017;101:956–67. 41. Saab, S., Niho, H., Comulada, S. et al. Mortality predictors in liver transplant recipients with recurrent hepatitis C cirrhosis. Liver Int, 2005;25:940–5. 42. Curry, M.P., O’Leary, J.G., Bzowej, N. et al. Sofosbuvir and Velpatasvir for HCV in Patients with Decompensated Cirrhosis. N Engl J Med, 2015; 373:2618–28. 43. Leroy, V., Dumortier, J., Coilly, A. et  al. Efficacy of sofosbuvir and daclatasvir in patients with fibrosing cholestatic hepatitis C after liver transplantation. Clin Gastroenterol Hepatol, 2015;13:1993–2001 e1991–1992. 44. Salcedo, M., Prieto, M., Castells, L. et al. Efficacy and safety of daclatasvir‐ based antiviral therapy in hepatitis C virus recurrence after liver transplantation. Role of cirrhosis and genotype 3. A multicenter cohort study. Transpl Int, 2017;30:1041–50. 45. Levitsky, J., Verna, E.C., O’Leary, J.G. et al. Perioperative ledipasvir/sofosbuvir for HCV in liver‐transplant recipients. N Engl J Med, 2016;375:2106–8. 46. Saab, S., Rheem, J., Jimenez, M. et al. Curing hepatitis C in liver transplant recipients is associated with changes in immunosuppressant use. J Clin Transl Hepatol, 2016;4:32–8. 47. Raschzok, N., Schott, E., Reutzel‐Selke, A. et al. The impact of directly acting antivirals on the enzymatic liver function of liver transplant recipients with recurrent hepatitis C. Transpl Infect Dis, 2016;18:896–903. 48. Northup, P.G., Argo, C.K., Nguyen, D.T. et al. Liver allografts from hepatitis C positive donors can offer good outcomes in hepatitis C positive recipients: a US National Transplant Registry analysis. Transpl Int, 2010;23:1038–44. 49. Saab, S., Ghobrial, R.M., Ibrahim, A.B. et al. Hepatitis C positive grafts may be used in orthotopic liver transplantation: a matched analysis. Am J Transplant, 2003;3:1167–72. 50. Reese, P.P., Abt, P.L., Blumberg, E.A., and Goldberg, D.S. Transplanting hepatitis C‐positive kidneys. N Engl J Med, 2015;373:303–5. 51. Chhatwal, J., Samur, S., Bethea, E.D. et al. Transplanting hepatitis C virus‐ positive livers into hepatitis C virus‐negative patients with preemptive antiviral treatment: a modeling study. Hepatology, 2018;67(6):2085–95. 52. Trotter, P.B., Summers, D.M., Ushiro‐Lumb, I. et  al. Use of organs from hepatitis C virus‐positive donors for uninfected recipients: a potential cost‐ effective approach to save lives? Transplantation, 2018;102:664–72. 53. Goldberg, D.S., Abt, P.L., Blumberg, E.A. et al. Trial of transplantation of HCV‐infected kidneys into uninfected recipients. N Engl J Med, 2017; 376:2394–5.

934

THE LIVER:  REFERENCES

54. Schlendorf, K.H., Zalawadiya, S., Shah, A.S. et al. Early outcomes using hepatitis C‐positive donors for cardiac transplantation in the era of effective direct‐ acting anti‐viral therapies. J Heart Lung Transplant, 2018;37(6):763–9. 55. Mutimer, D. Review article: hepatitis B and liver transplantation. Aliment Pharmacol Ther, 2006;23:1031–41. 56. Davies, S.E., Portmann, B.C., O’Grady, J.G. et al. Hepatic histological findings after transplantation for chronic hepatitis B virus infection, including a unique pattern of fibrosing cholestatic hepatitis. Hepatology, 1991;13:150–7. 57. Samuel, D., Muller, R., Alexander, G. et al. Liver transplantation in European patients with the hepatitis B surface antigen. N Engl J Med, 1993;329:1842–7. 58. Faria, L.C., Gigou, M., Roque‐Afonso, A.M. et al. Hepatocellular carcinoma is associated with an increased risk of hepatitis B virus recurrence after liver transplantation. Gastroenterology, 2008;134:1890–9; quiz 2155. 59. Terrault, N.A. and Vyas, G. Hepatitis B immune globulin preparations and use in liver transplantation. Clin Liver Dis, 2003;7:537–50. 60. Burbach, G.J., Bienzle, U., Neuhaus, R. et al. Intravenous or intramuscular anti‐HBs immunoglobulin for the prevention of hepatitis B reinfection after orthotopic liver transplantation. Transplantation 1997;63:478–80. 61. Yilmaz, N., Shiffman, M.L., Todd Stravitz, R. et  al. Prophylaxsis against recurrance of hepatitis B virus after liver transplantation: a retrospective analysis spanning 20 years. Liver Int, 2008;28:72–8. 62. Hooman, N., Rifai, K., Hadem, J. et al. Antibody to hepatitis B surface antigen trough levels and half‐lives do not differ after intravenous and intramuscular hepatitis B immunoglobulin administration after liver transplantation. Liver Transpl, 2008;14:435–42. 63. Zuckerman, J.N. Review: hepatitis B immune globulin for prevention of hepatitis B infection. J Med Virol, 2007;79:919–21. 64. Ghany, M.G., Ayola, B., Villamil, F.G. et al. Hepatitis B virus S mutants in liver transplant recipients who were reinfected despite hepatitis B immune globulin prophylaxis. Hepatology, 1998;27:213–22.

65. Perrillo, R.P., Wright, T., Rakela, J. et  al. A multicenter United States‐ Canadian trial to assess lamivudine monotherapy before and after liver transplantation for chronic hepatitis B. Hepatology, 2001;33:424–32. 66. Markowitz, J.S., Martin, P., Conrad, A.J. et al. Prophylaxis against hepatitis B recurrence following liver transplantation using combination lamivudine and hepatitis B immune globulin. Hepatology, 1998;28:585–9. 67. Han, S.H., Ofman, J., Holt, C. et al. An efficacy and cost‐effectiveness analysis of combination hepatitis B immune globulin and lamivudine to prevent recurrent hepatitis B after orthotopic liver transplantation compared with hepatitis B immune globulin monotherapy. Liver Transpl, 2000;6:741–8. 68. Naoumov, N.V., Lopes, A.R., Burra, P. et al. Randomized trial of lamivudine versus hepatitis B immunoglobulin for long‐term prophylaxis of hepatitis B recurrence after liver transplantation. J Hepatol, 2001;34:888–94. 69. Fung, J., Cheung, C., Chan, S.C. et al. Entecavir monotherapy is effective in suppressing hepatitis B virus after liver transplantation. Gastroenterology, 2011;141:1212–19. 70. Fung, J., Wong, T., Chok, K. et al. Long‐term outcomes of entecavir monotherapy for chronic hepatitis B after liver transplantation: Results up to 8 years. Hepatology, 2017;66:1036–44. 71. Jang, J.W., Choi, J.Y., Kim, Y.S. et al. Long‐term effect of antiviral therapy on disease course after decompensation in patients with hepatitis B virus‐ related cirrhosis. Hepatology, 2015;61:1809–20. 72. Terrault, N.A., Bzowej, N.H., Chang, K.M. et  al. AASLD guidelines for treatment of chronic hepatitis B. Hepatology, 2016;63:261–83. 73. Terrault, N.A., Lok, A.S.F., McMahon, B.J. et  al. Update on prevention, diagnosis, and treatment of chronic hepatitis B: AASLD 2018 hepatitis B guidance. Hepatology, 2018;67:1560–99. 74. Cholongitas, E., Papatheodoridis, G.V., and Burroughs, A.K. Liver grafts from anti‐hepatitis B core positive donors: a systematic review. J Hepatol, 2010;52:272–9.

70

Time for the Elimination of Hepatitis C Virus as a Global Health Threat John W. Ward1,2, Alan R. Hinman1, and Harvey J. Alter3 Task Force for Global Health, Decatur, GA, USA Centers for Disease Control and Prevention, Atlanta, GA, USA 3 Department of Transfusion Medicine, Clinical Center, National Institutes of Health, Bethesda, MD, USA 1 2

INTRODUCTION Disease elimination and eradication, the ultimate goals of disease control, can achieve large health benefits and ensure global health equity. HCV meets the benchmarks of a major public health problem whose elimination is feasible [1–3]. In 2016, the World Health Organization released the Global Health Sector Strategy on Viral Hepatitis 2016–2021: Towards Ending Viral Hepatitis, which set goals to be achieved by 2030 that would indicate progress toward the elimination of hepatitis C virus (HCV) as a public health threat. The goals are 90% reduction in global incidence and 65% reduction in global mortality by 2030 [4]. If those goals were met, then estimated global HCV incidence would decrease from 1.75 million per year in 2015 to 175 000 per year in 2030; and global mortality would decrease from approximately 400 000 deaths with HCV as the underlying cause in 2015 to 140 000 per 100 000 population in 2030 [5]. In 2015, the World Health Organization (WHO) estimated that 71 million people were infected with HCV globally. In 2016, the World Health Assembly (WHA) passed a resolution of endorsement supporting the WHO goals and adding HCV to the select group of diseases targeted for global elimination [6]. Biologically, HCV is a virus that is easily transmitted via parenteral exposures to contaminated blood with low infectivity through other routes. Global initiatives have greatly reduced transmission in healthcare settings, increasing the proportion of cases attributable to injection drug use. The latency from HCV infection to onset of severe disease is several decades, providing ample time for early diagnosis and treatment to prevent premature mortality. Tests for anti‐HCV and HCV RNA reliably detect HCV infection. New all‐oral therapies for HCV, a major advance

in medicine, cure over 90% of people who complete treatment [7]. The HCV elimination goals provide the impetus for countries to put in place their own comprehensive HCV prevention programs. The development of such programs is just beginning, and the knowledge and experience gained from past elimination efforts (e.g. smallpox and polio) can be employed to successfully eliminate HCV. In this chapter, we will review the health burden of HCV infection, the process for development of HCV elimination goals, the successes and challenges in the implementation of interventions to reach these goals, field studies illustrating the essential components of effective elimination programs, and the key research that will accelerate progress toward elimination of HCV infection.

THE PUBLIC HEALTH PROBLEM OF HCV INFECTION Diseases must be of sufficient magnitude to warrant the prioritization of resources for a dedicated elimination program. HCV is a substantial public health problem globally: in 2015, an estimated 71 million people (1% of the global population) were infected with HCV [5]. HCV is present in almost all countries. However, HCV varies greatly at the national level, reflecting differences in risk exposures and the importance of reliable data to reveal local patterns of transmission and burden of disease. First, based on genetic differences, HCV is classified into seven genotypes which differ in frequency by country [7]. Certain countries have high prevalence of HCV infection, including Mongolia (6.7%), Egypt (6.4%), Pakistan (3.8%),

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

936

THE LIVER:  THE PROCESS FOR SETTING HCV ELIMINATION GOALS

Russia (2.9%), and Romania (2.5%) [8]. Prevalence can also vary at the sub‐national level. For example, in Egypt, a country of high prevalence, the prevalence is higher in Lower Egypt (8.2%) compared with Upper Egypt (2.2%) [9–11]. In the United States, the estimated 1.67% anti‐HCV prevalence in 2010 varied by state, ranging from a high of 3.34% (Oklahoma) to a low of 0.71% (Illinois) [12]. Globally, the population with highest prevalence of HCV infection is persons who inject drugs (PWIDs). In 2017, of an estimated 15.6 million people with a history of drug injection, 8.1 million (52%) have been infected with HCV [13, 14]; 58% of PWIDs with HCV have a history of incarceration. As a result, in 2012, an estimated 2.2 million incarcerated persons (26%) have been infected with HCV [15]. In the United States, an estimated 1.5 million people who are currently injecting drugs or did so in the past are infected with HCV [14]. In 2010, an estimated 3.5 million people were infected with HCV in the United States [16]. HCV is a leading cause of mortality from infectious disease in the United States [17]. Deaths from HCV surpassed the number of deaths from HIV in 2007 and in 2014, deaths from HCV exceeded the number of deaths from the 60 other reportable infectious diseases in the United States [17, 18]. In 2016, a total of 18, 153 deaths (4.5 deaths/100 000 population) were associated with HCV infection [19]. In 2015, a total of 1.7 million new HCV infections occurred globally, with 40% attributed to unsafe injection in healthcare settings and about one‐third attributable to current injecting‐ drug use [5]. Exposures in healthcare settings include receipt of unscreened blood and blood products from HCV‐infected donors and procedures with nonsterile injection equipment [5, 7]. The proportion of HCV attributable to healthcare and injection drug use varies globally. Receipt of unscreened blood and poor infection control are the major modes of transmission in low and middle income countries. PWIDs account for most new infections in high and some middle income countries. Other routes of HCV transmission are less common. Children born to HCV‐infected mothers have a 6–14% risk of HCV infection [20, 21]. Other incidental parenteral contacts with blood, intranasal inhalation of drugs, unregulated tattooing, and ritual scarification are associated with the spread of HCV [7]. Sexual transmission of HCV is rare except among certain populations of HIV‐infected men who have sex with men ­ (MSM) [22]. Prevention is important in these populations. However, achieving HCV elimination goals hinges on preventing the most common routes of transmission from unscreened blood and blood products and nonsterile injections in healthcare and community settings.

THE PROCESS FOR SETTING HCV ELIMINATION GOALS The development of the WHO strategy for HCV elimination followed a series of policies and actions globally. As new data revealed mounting mortality from HCV infection, there were progressively urgent calls for improvements in HCV prevention, care, and treatment. The licensure and demonstrated effectiveness of antiviral therapy in curing HCV infection added to the

international call for action to apply this new prevention tool to more effectively prevent HCV transmission and disease. From 2000 through 2010, HCV‐related mortality increased 10% globally, whereas mortality from other infectious diseases, including HIV, malaria, and TB, declined during this time period [23]. In 2010, 2014, and 2016, the WHA passed three resolutions recognizing HCV and other forms of viral hepatitis as global public health problems [2, 24, 25]. In 2010, the WHA called on WHO and member states to take greater action to improve viral hepatitis prevention diagnosis and treatment. In 2011, the first‐generation oral antiviral therapies that required continued use of interferon‐based therapies for HCV infection were licensed in the United States by the US Food and Drug Administration (FDA). In 2012, WHO released the first action plan to prevent viral hepatitis and began to assist countries with local prevention planning [26]. At the time, only 37% of 126 member countries had national plans for viral hepatitis prevention [27]. In 2014, the WHA called for continued improvements in HCV prevention and emphasized the importance of prevention measures to protect PWIDs [24]. Importantly, the WHA requested WHO examine the feasibility of setting goals for elimination of HBV and HCV infection. In 2014, the first all‐ oral therapies for HCV infection were licensed in the United States by the FDA. To respond to the WHA request, WHO convened numerous stakeholder consultations with member states, organizations in the United Nations system, and other multilateral agencies, donor and development agencies, civil society, nongovernmental organizations, scientific and technical institutions and networks, and the private sector. WHO also commissioned models to estimate the impact of various prevention strategies on t­ ransmission and burden of disease. While this process was underway in 2015, the United Nations, in the Sustainable Development Goals, called for a global response to combat viral hepatitis [28]. In 2016, WHO released the WHO Global Health Sector Strategy on Viral Hepatitis 2016–2021: Towards Ending Viral Hepatitis [4]. In the global strategy, WHO set global targets for the elimination of HCV as a public health threat, defined as a 90% reduction in the incidence of HCV infection and a 65% reduction in HCV‐related mortality by 2030. A discussion regarding setting and communicating elimination goals follows below. Countries are encouraged to set more ambitious national goals based on the local assessment of burden of disease, the populations affected, the capacity of clinical care and public health systems, and the resources available for mobilization. The 2016 WHA resolution endorsed the strategy and elimination goals. WHO has set performance targets for 2020 and 2030 for the key prevention interventions, including HCV diagnosis and treatment (Figure 70.1). In parallel with global actions, steps were taken in the United States to strengthen the policy foundation for HCV prevention. In 2010, the US Institute of Medicine (IOM) reported findings from an expert panel that the national prevention capacity for HCV and other forms of viral hepatitis was inadequate and recommended improvements in viral hepatitis surveillance and prevention services [29]. The IOM called for the US government to draft a national action plan. In May 2011, the national action plan for viral hepatitis prevention was released by Assistant Secretary of Health Dr. Howard Koh [30]. The national action plan was updated in 2014 and 2017



70:  Time for the Elimination of Hepatitis C Virus as a Global Health Threat

937

Incidence of HCV infection Incidence rate (per 100 000) WHO region

Best estimate

Uncertainty interval

31.0

22.5–54.4

309

222–544

6.4

5.9–7.0

63

59–69

Eastern Mediterranean Region

62.5

55.6–65.2

409

363–426

European Region

61.8

50.3–66.0

565

460–603

South-East Asia Region

14.8

12.5–26.9

287

243–524

Western Pacifc Region

6.0

5.6–6.6

111

104–124

23.7

21.3–28.7

1 751

1 572–2 120

African Region

Map key

Total number (000)

Region of the Americas

Global

Best estimate

Uncertainty interval

Figure 70.1  Incidence of HCV infection in the general population by WHO region, 2015. Reproduced from [5] with permission of the World Health Organization.

(https://www.hhs.gov/hepatitis/index.html). In 2017, the National Academy of Sciences, formerly the IOM, released two reports recommending the global HCV elimination goals be adopted as national goals for the United States. The National Academies proposed a set of actions to build the prevention and clinical capacity to achieve these goals [31].

SETTING A GOAL FOR ELIMINATION OF HCV AS A PUBLIC HEALTH THREAT The terms “eradication” and “elimination” are sometimes used interchangeably, as both convey information about the intended scope of disease control. The concept of disease eradication originated with Edward Jenner, who developed the first smallpox vaccine. Jenner wrote in 1801 “It now becomes too manifest to admit

of controversy, that the annihilation of the Small Pox, the most dreadful scourge of the human species, must be the final result of this practice” [32]. However, organized efforts to eradicate ­diseases from humans did not begin until the early and mid‐­ twentieth century. Yellow fever (1915–1977), yaws (1954–1967), malaria (1955–1969), and smallpox (1955–1980) were the first diseases targeted for eradication [32]. Of these, only smallpox has been eradicated as certified by the WHA in 1981 [33–35]. The successful campaign to eradicate smallpox sparked interest in targeting other diseases for eradication, including dracunculiasis (i.e. guinea worm) in 1986 and polio in 1988; both of these efforts are ongoing [1, 36]. The terms eradication and elimination both convey information about the intended scope of disease control. Decades of academic discussion including conferences in Berlin (1997) Atlanta (1998), and Frankfurt (2010) [4, 36] resulted in general acceptance of standard definitions for disease eradication and

938

THE LIVER:  DETERMINANTS FOR SETTING FEASIBLE TARGETS FOR HCV ELIMINATION

elimination. Eradication is the permanent reduction to zero of the worldwide incidence of infection caused by a specific agent as a result of deliberate efforts. Elimination of transmission (also referred to as interruption of transmission) is the mean reduction to zero of the incidence of infection caused by a specific pathogen in a defined geographical area, with minimal risk of reintroduction, as a result of deliberate efforts. Continued actions to prevent re‐establishment of transmission may be required [37]. De Serres et al. have pointed out that zero incidence is essentially unattainable in the absence of eradication because of the continued risk of importation with limited subsequent spread [38]. In practice, the term elimination has been used to describe different targets  –  specific levels of control (e.g. reduction of neonatal tetanus deaths to 90% declines in incidence [100, 101]. Because of reductions in both HCV transmission and disease, treatment of PWIDs for HCV is highly cost effective or cost saving [102, 103]. Globally, in 2010, an estimated 27 sets of syringes and needles were exchanged per user, per year, a rate far below the WHO targets of 200 and 300 of such exchanges for the years 2020 and 2030, respectively [13, 104, 105]. A study of PWIDs in North America, Australia, and the Netherlands from 1985 through 2011 found an overall incidence rate of 22.6 per 100 person‐years of observation (PYO), with declines in incidence among PWIDs observed over this period from 24.6/100 PYO to 18.8/100 PYO [97, 106]. The greatest declines in HCV incidence were observed in Australia and the Netherlands, two countries that began in the 1980s to expand access to SSPs and OST. In contrast, HCV risk behaviors and incidence among PWID are high in North America, where access to SSPs and OST is limited. From 2010 through 2016, HCV incidence in the United States has increased threefold, temporally associated with increases in the injection of prescription opioids and heroin [19]. Increases in HCV incidence are greatest in states with no or few SSPs [107]. The recent rise in HCV transmission in the United States might cause some to question the feasibility of reducing HCV incidence by 90% from 2015 to 2030. This questioning is a desired outcome of the disease elimination process [1, 2, 33–35]. The sense of urgency created by setting limited elimination goal leads to an examination of prevention capacity. Indeed the recent rise in HCV transmission in the United States adds to the challenges of reaching elimination targets. However, the new infections are not the result of changes in the virus or new modes of HCV transmission that increase spread of HCV. The new infections in the United States are among PWIDs [19]. Evidence from other countries demonstrate that strong prevention programs ­sustained over time result in low HCV incidence among this risk population [106]; the availability of curative HCV therapies is the new intervention that can further decrease transmission. The ­incidence trends in the United States are the result of a poor HCV prevention infrastructure for this population. With sufficient improvements in prevention capacity, which are totally feasible for a  high‐income country such as the United States, HCV incidence  will decline among PWIDs and elimination goals can be achieved.

Increasing the proportion of infected persons diagnosed with HCV HCV testing is the essential step in a plan to reduce the prevalence of HCV and subsequent mortality. However, few silent HCV carriers have been diagnosed with HCV, and even fewer have received treatment. WHO estimates that in 2015, a total of 71 million people were infected with HCV globally [6, 8, 104], of whom only 14 million (20%) have been diagnosed [104]. Reaching the HCV elimination goals will require diagnosing 90% of HCV‐infected persons [1]. The first step on the care continuum, HCV testing, must be scaled up to potentiate increased access to treatment. National



70:  Time for the Elimination of Hepatitis C Virus as a Global Health Threat

plans guided by local epidemiologic data must include policies prioritizing testing for populations at greatest risk (e.g. persons receiving unscreened blood donations and PWID) [5, 6]. The Centers for Disease Control (CDC), the United State Preventive Services Task Force (USPSTF), professional societies, and WHO recommend routine HCV testing of certain populations based on risk exposures, settings, and demographic characteristics indicative of increased prevalence of HCV infection [108–111]. HCV testing and treatment of PWIDs is highly cost effective or cost saving in low, middle, and high‐income ­countries [99, 112–117]. Similar returns on investment are seen for HCV testing and treatment of particular demographic subpopulations, including age and birth cohorts, PWIDs, and the incarcerated. In most countries, targeting multiple populations for HCV testing is needed to fully capture the universe of persons at risk for HCV infection. Testing of individuals for HCV based on risk behaviors (e.g. recipients of unscreened blood, injection drug use) is a core strategy; HCV testing of incarcerated populations augments risk‐based testing. However, many people with exposures in the distant past do not recall drug‐use risk behaviors, or are unaware of risks in healthcare settings or of perinatal transmission [115, 118, 119]. High incidence of HCV in earlier decades has resulted in a cohort of chronically infected individuals who are becoming increasingly ill as their infection progresses, manifesting as chronic liver disease. Cohorts of older adults who have been infected for decades are at highest risk for severe liver disease. In the United States, 81% of the 3.5 million people living with HCV in 2010 were born during 1945–1965; of these, one in four has clinical evidence of severe fibrosis or cirrhosis [118–120]. In other countries, different birth cohorts are disproportionately affected by HCV based on different epidemiologic characteristics and different policies for controlling risk (e.g. initiation of HCV screening in blood banks) [116, 121, 122]. In the United States, the CDC and the USPSTF recommend a one‐time test for HCV for people born during 1945–1965 [109, 110]. Accordingly, a birth cohort strategy is one population‐based approach that is cost effective while making HCV testing broadly available [123]. Other strategies include testing all people in a particular setting, such as correctional facilities, emergency departments, and all people receiving inpatient services. Testing all adults for HCV can also be cost effective and readily implemented [124, 125]. In 2018, WHO recommended routine HCV testing in settings with a ≥2% or ≥5% of HCV antibody seroprevalence in the general population [126]. The release of a policy to guide HCV testing increases the magnitude of HCV testing [123]. Evidence supports the use of certain strategies to implement policies and increase access to HCV testing for target populations. Reflex testing whereby specimens positive for HCV antibody are immediately tested for HCV RNA improves diagnosis of current HCV infection [127]. The ordering of HCV tests by clinicians is improved by professional education, and electronic reminders that prompt testing. The return of evaluation data can also motivate improvements in clinicians’ testing practices [128–131]. State‐based mandates for HCV testing by clinicians also can be evaluated as a means to increase testing [132].

941

INCREASING HCV TREATMENT AND CURE AFTER DIAGNOSIS Of HCV‐infected individuals who complete 8–12 weeks of FDA‐approved antiviral therapies, more than >90% will achieve a sustained virologic response (SVR), with no evidence of HCV on testing 12–24 weeks after therapy [7, 133–137]. As of May 2018, the FDA and the European Medicines Agency (EMA) together have approved 13 antivirals and several fixed‐dose combinations [126]. HCV therapies are safe, are active against all genotypes of HCV, and can effectively treat people with cirrhosis, HIV, and other previously hard‐to‐treat patients. HCV treatment has a substantial impact on reducing the risks of HCV mortality. For patients with advanced liver disease, virologic cure reduces the risk of mortality and HCC by 78% and 84%, respectively [138–140]. In a large meta‐analysis, achievement of SVR reduces the risk of HCC and all‐cause mortality by 80% and 75%, respectively, compared with people who did not achieve SVR; virologic cure also results in a 56% decline in mortality from extrahepatic manifestations [140]. HCV testing and treatment is highly cost effective and cost saving for many populations [141–147]. However, the economic benefits of HCV treatment are sensitive to the cost of HCV therapies. The original market prices of HCV medications (US$84 000–$96 000 per course) resulted in payers placing restrictive criteria for approval of patients for therapy based the stage of liver fibrosis, the prescribing provider specialty, and level of sobriety from alcohol and other drugs [141]. Since 2014, prices of HCV medications have decreased and restrictions have been removed or loosened in some US states. The declines are result of lawsuits by patients, competition with licensure of additional antiviral therapies, and data supporting shortening therapeutic courses from 8 to 12 weeks of treatment [142–144]. At a pricing of less than $60 000 per curative course in 2015, HCV therapy is cost saving in the United States [145]. The initial market price of HCV medication also posed ­challenges globally, resulting in many countries placing restrictions on treatment. However, the costs of HCV therapies have declined globally. Over 100 countries can now access generic medicines for US$200 (or less) per curative treatment [104]; the prices to obtain generic HCV medications render HCV treatment cost saving in countries [146, 147]. In 2017, 62% of HCV‐ infected persons lived in countries where generic medicines are available [104]. The availability of new curative antiviral therapies has not yet resulted in large decreases in HCV globally. To reach the WHO 2030 goal, the number of cures must be large enough to reduce global HCV prevalence by 7% annually [148]. In 2016, only 10 countries, (the United States, Egypt, Japan, Australia, the Netherlands, France, Spain, Germany, Iceland, and Qatar) had treated ≥7% of their HCV‐infected population. Other countries have not made similar progress; 44 countries have treated 11 μM is

considered necessary for reducing emphysema risk [39]. ­ Subsequently, phase I and phase II trials were conducted using rAAV1 to improve muscle transduction [40]. Both trials, resulted in minor plasma levels of ATM. Antibodies against AAV1 were observed and moderate infiltration with reactive T lymphocytes at the injection site was noted in muscle biopsies [40]. An initial clinical trial of rAAV‐mediated liver‐directed gene therapy for factor IX deficiency (hemophilia B), using rAAV2, resulted in the appearance of plasma factor IX for only a few weeks [41]. In a subsequent landmark clinical trial in hemophilia B patients, using portal vein infusion of rAAV8 vectors, 1–6% of normal plasma factor IX levels were achieved for a median duration of 3.2 years [42]. Prolonged persistence of the episomal vector reflects the slow turnover of hepatocytes, particularly in liver‐based monogenic diseases that are not associated with hepatocellular injury. Serum alanine aminotransferase (ALT) levels increased 7–10 weeks after rAAV injection in a high proportion of patients and was ameliorated by prednisolone administration, suggesting injury to some of the transduced hepatocytes from an adaptive immune response against rAAV. Currently, after 35 years of investigation of rAAVs, there is great enthusiasm among clinicians and investigators for clinical application of these vectors. Indeed, since 2004, clinical trials of monogenic liver diseases have used exclusively rAAV vectors (http://clinicaltrial.gov/). Diseases targeted in recently ­completed, continuing, or upcoming clinical trials using rAAV vectors include hemophila A, hemophilia B, α1‐antitrypsin deficiency, acute intermittent porphyria, homozyogous familial hypercholesterolemia, ornithine transaminase deficiency, mucopolysaccharidosis VI (lysosomal enzyme arylsulfatase B deficiency), and Crigler–Najjar syndrome type 1.

984

THE LIVER:  NONVIRAL VECTORS

In view of current enthusiasm for this vector and numerous ongoing and upcoming clinical trials, it is important to consider certain unanswered questions regarding this technology. For example, most circular DNAs introduced into cells in vivo have a limited duration of expression, but rAAVs, which are also episomal circular DNAs, persist for years in non‐dividing cells without undergoing epigenetic silencing. As the ITRs are the only common elements among the various rAAV genomes, these are likely to be involved in persistence of transgene expression, but the mechanism remains to be explored. The persistence issue is important, because the 3.2‐year median duration of therapeutic level of factor IX in the recent clinical trial [42] would not be optimal for most monogenic liver diseases, especially in view of the current immunological barrier to readministration of rAAV vectors. In fact, many patients are currently excluded from clinical trial because of preexisting immunity against wild‐ type AAV. In terms of transduction e­ fficiency, there seems to be a marked difference between studies in animal models and clinical trials. The factor IX blood levels obtained in the recent hemophilia B clinical trial were therapeutically sufficient in many of the subjects [42]. However, a similar level of transduction efficiency would not be sufficient to treat inherited metabolic disorders, such as Wilson disease, which would require transduction of a greater proportion of hepatocytes [43]. The characteristic delay in attaining the transgene expression plateau after rAAV administration is a hurdle against application of this vector for management of metabolic emergencies, such as acute liver failure in Wilson disease [43]. Although wild‐type AAV infection is generally considered to be non‐cytotoxic, apoptosis has been reported after high input doses of rAAV vectors [44]. An even more serious concern is related to oncogenic potential of rAAVs, which is hotly debated. A small but growing number of studies on rAAV‐mediated gene transfer into mice have reported insertional mutagenesis and hepatocellular carcinoma (HCC) [45]. In three separate mouse studies, rAAV integrations were found within a 6000 bp region of a locus termed RNA imprinted and accumulated in nucleus (Rian) on mouse chromosome 12. Of these integrations, 57% were present within  an intronic microRNA (Mir341) and 43% were elsewhere within Rian. Targeted insertion by homologous recombina­tion of an rAAV genome containing an exogenous promoter/enhancer into the Rian locus, outside Mir341, resulted in HCC with nearly 100% penetrance [46]. The human ortholog of Rian, maternally expressed 8 gene (MEG8), is located in a cluster of imprinted genes on chromosome 14q32.3, termed the Dlk1‐Dio3 domain. This domain contains three protein‐coding genes, Dlk1, Rtl1, and Dio3, expressed from the paternally inherited chromosome and several imprinted large and small noncoding RNA genes expressed from the maternally inherited homolog. Several noncoding RNAs and protein‐ coding genes within the Dlk1‐Dio3 domain have been linked to HCC in human subjects and mouse models [47]. Thus, although presence of wild‐type AAV genomes in human tissues is not associated with increased cancer risk, further work is warranted regarding the oncogenic potential of recombinant rAAV.

Simian virus 40‐based vectors Recombinant simian virus 40 (SV40) is a non‐enveloped DNA virus of the papova family with a 5.2 kb circular double‐stranded

genome. The large (Tag) and small (tag) T antigens, that are expressed by differential splicing of a single RNA transcript, are required for transcription of the viral structural genes, VP1, VP2, and VP3. Tag is the most immunogenic protein of SV40 and is able to impart an immortalizing effect on the cell. In the recombinant viral genome, the Tag gene is replaced by a target transgene. Viral particles are generated by transfecting the recombinant genome into COS‐7 cells that provide Tag in trans. Because the recombinant virus lacks the Tag gene, it cannot replicate [48], and its immunogenicity is markedly reduced [49]. SV40 is a small virus and the capacity of the recombinant SV40 to accommodate exogenous DNA is limited to 4.7 kb. SV40 vectors can be concentrated to high titers. Recombinant SV40 appears to integrate into the host chromosomal DNA. The ability of these vectors to infect nondividing cells makes them attractive for liver‐directed gene therapy [50].

NONVIRAL VECTORS Although viral vectors have been widely utilized in gene therapy research, owing to their efficient delivery of transgenes into cells, nonviral gene transfer methods can be potentially less toxic and easier to standardize. Three major categories of synthetic nonviral delivery systems for systemic delivery of nucleic acids to tissues have been studied extensively [51]: (i) Lipid‐based delivery systems such as lipid‐encapsulated nucleic acids or cationic lipid–nucleic acid complexes (lipoplex) provided the initial proof‐of‐principle for systemic transgene delivery. (ii) Polyplex systems consist of complexes formed by the addition of nucleic acid to a polycation, such as poly‐l‐lysine, polyethylenimine (PEI), polyglucosamines, lipopolyamines, and cationic peptides, generating efficient water‐soluble delivery systems. (iii) In the lipopolyplex delivery systems, DNA compacted with polycations are encapsulated or complexed with lipids, reducing the final particle size and protecting the nucleic acid from nuclease degradation. Cationic lipid transfecting agents, such as lipofectamine, DOTAP(1‐oleoyl‐2‐[6‐[(7‐nitro‐2‐1,3‐benzoxadiazol‐4‐yl) amino]hexanoyl]‐3‐trimethylammonium propane), and cationic polymers, including poly‐l‐lysine and PEI, efficiently transfer nucleic acids across the plasma membrane into the cytoplasm [52]. However, they are cytotoxic because of their large particle size [53] and the high positive zeta potential required for their uptake [51], which is neutralized by plasma proteins [54]. Cytotoxicity of cationic‐based delivery systems is reduced by incorporation of poly(ethylene glycol) (PEG), which stabilizes, prevents aggregation, reduces binding to serum proteins, and maintains the small size required for endocytosis [55]. However, shielding of the cationic charge by PEG reduces the transfection efficiency. To make the delivery system cell‐type specific, ligands are utilized to promote receptor‐mediated delivery. For example, asialoorosomucoid (ASOR) [56] and galactose [57, 58] have been conjugated to polylysine, lipopolyamines, or PEI for targeting to the asialoglycoprotein receptor (ASGPR) on hepatocytes. Lipid‐ based delivery systems have utilized galactocerebrosides as the targeting moiety for ASGPR [59]. Receptor‐mediated ligand targeting has also been exploited using transferrin, folate, and



73:  Liver‐Directed Gene Therapy

cell‐specific antibodies conjugated to the polycations or liposomes [60]. These ligand‐targeted systems increase hepatocyte‐directed gene delivery both in vitro [59, 60] and in vivo [56, 61]. Size of the targeting complex is important in hepatocyte specificity, because larger particles are cleared by Kupffer cells [62]. Nucleic acids fully enclosed in nanocapsules [63] containing hyaluronan or ASOR have been used for targeted delivery to liver sinusoidal cells [64] or hepatocytes [65] without significant uptake by other cells. Cargo delivered via receptor‐mediated endocytosis is naturally translocated to lysosomes, where a large fraction of it is degraded. Disruption of the translocation of endosomes to lysosomes, for example, by transient depolymerization of microtubules, prolongs transgene expression, but the DNA remains compartmented in cytoplasmic vesicular pools, resulting in low levels of transgene expression [66]. Destabilization of endosomal vesicles, for example, by using proton sponges (e.g. PEI) or disruptor peptides incorporated into the delivery vehicles promote endosomal release of the nucleic acid into the cytoplasm [67]. Another approach to bypass the endosomal pathway employs proteoliposomes containing the naturally galactose‐terminated F‐glycoprotein of the Sendai virus envelope [68], which binds specifically to the hepatocyte cell surface ASGPR. The fusogenic activity of the F‐protein results in deposition of the contents of the proteoliposome directly into the cytosol, bypassing the endosomal pathway, thereby enhancing transgene expression. Rapid clearance of the gene delivery vehicle from the plasma by hepatocytes reduces exposure of other tissues to the gene transfer vehicle, markedly reducing its immunogenicity [69]. Translocation of the transgenes from the cytosol to the nucleus can be enhanced by incorporating nuclear localization signal peptides on the plasmid constructs [70]. However, an important limitation of nonviral gene transfer methods is that expression of episomal DNA is inherently transient. To achieve transgene integration for long‐term expression, especially for the treatment of inherited diseases, a transposon system, termed “Sleeping Beauty” [71], has been used. Sleeping Beauty promotes transgene integration at a TA dinucleotide site via a cut‐and‐paste mechanism [72]. Typically, the system ­consists of a plasmid containing two transcription units, one expressing the enzyme, Sleeping Beauty transposase, and the other expressing the gene of interest. The transcription unit expressing the gene of interest is flanked by inverted/direct repeat sequences that are cleaved by the expressed transposase, followed by insertion of the transgene into the host genome. The  Sleeping Beauty transposition system has been used in UGT1A1‐deficient jaundiced Gunn rats, resulting in long‐term amelioration of hyperbilirubinemia [69]. The Sleeping Beauty system has been also introduced to the liver by hydrodynamic delivery [73] for persistent expression of coagulation factor IX [74], factor VIII [75], β‐glucuronidase or α‐l‐iduronidase [76], and fumarylacetoacetate hydrolase (FAH) deficiency [77]. In all cases, significant correction of the disease phenotype was observed. However, this method delivers the cargo nonspecifically to hepatocytes, as well as LSECs, and Kupffer cells [78], leading to immune response causing loss of phenotypic correction, as observed with factor VIII [75] and α‐l‐iduronidase [76]. In contrast, by coupling the Sleeping Beauty transposon system with targeted cell‐type specific delivery, long‐term persistent

985

factor VIII levels and phenotypic correction of the bleeding diathesis were achieved [79]. In these studies, Sleeping Beauty‐ mediated integration was found to occur preferentially into nontranscribed and intronic regions of the genome [80]. In a clinical trial for treatment of advanced B‐cell malignancies, chimeric antigen receptors (CAR) recognizing tumor‐associated antigen (so‐called “CAR‐T” cells) are expressed by nonviral transfer of the Sleeping Beauty system into T cells ex vivo [81]. Systemic nonviral delivery of therapeutic RNAs: Nonviral carriers, particulary lipid nanoparticles have been used successfully to transfer RNA‐based therapeutics in mouse models of monogenic liver‐based disorders, such as primary hyperoxaluria (PH) types 1, 2, and 3. PHs are autosomal recessive disorders caused by overproduction of oxalate by hepatocytes. Excess oxalate excretion by the kidneys leads to calcium oxalate precipitation in the kidney and eventually to end‐stage renal disease. PH1, PH2, and PH3 are caused by mutations in enzymes that divert the metabolic pathway away from oxalate production, namely alanine:glyoxylate aminotransferase (AGXT), glyoxylate reductase/hydroxypyruvate reductase (GRHPR) ­ gene, and 4‐hydroxy‐2‐oxoglutarate aldolase 1 (HOGA1) genes, respectively. One promising strategy to treat all three types of PH is by downregulating expression of glycolate oxidase (GO), which controls the conversion of glycolate to glyoxylate, or lactate dehydrogenase (LDH), which is responsible for converting glyoxylate to oxalate. Lipid nanoparticle‐based delivery to the liver of dicer substrate RNAs that are converted by cellular dicer enzyme to siRNAs to inhibit the expression of GO or LDH resulted in marked reduction of urinary oxalate excretion in mouse models of PH1 and PH2 [82, 83].

Gene editing As precise duplication of genomic DNA is vital to the survival of a species, eukaryotes have evolved elaborate and redundant mechanisms to limit the effects of mutagens and exogenous genetic invaders. Homology directed repair machinery is used to recombine DNA sequences between sister‐chromatids during germ cell development or, in rarer instances, to repair damaged DNA sequences. This process is critical for genetic diversity, as well as maintenance of genomic integrity. Transformation protocols originally studied in lower eukaryotes [84] were subsequently used to develop strategies for gene targeting in higher organisms. Homologous pairing and DNA strand exchange can be harnessed to correct a mutation, create a mutation or insert a gene at a specific site in the genome.

Homology directed repair and non‐homologous end joining Insertion of exogenous DNA into the genome by activation of endogenous homology directed repair (HDR) mechanisms requires the presence of an undamaged DNA strand to be used as a template to faithfully restore the original DNA sequence in a “seamless” manner (see Figure 73.3). It begins with trimming of DNA break ends, after which the undamaged DNA strand aligns by sequence homology with the ends of the damaged strands forming a “bridge” of correct sequence spanning the damaged site. The bridging undamaged strand serves as a template to

986

THE LIVER:  NONVIRAL VECTORS

Site-directed DNA break

Target site on chromosomal DNA

Nuclease monomer

Zinc finger protein recognizing right hand homology

5’ 3’

3’ 5’

Nuclease monomer Zinc finger protein recognizing left hand homology 5’ 3’

Left hand homology

3’ 5’

Right hand homology

Homology-directed recombination (HDR) template

Non-homologous End-joining (NHEJ) • Insertion • Deletions • Substitution (“indel”)

Homologous recombination Useful for:

Useful for:

• Site-specific insertion • Gene repair

• Knockout • Random mutation

Figure 73.3  Homology‐based gene editing enhanced by targeted DNA break. Site‐specific DNA break is generated by nucleases fused to proteins that are engineered to recognize specific DNA sequences (such as ZFN or TALEN). Alterrnatively, a nuclease, such as Cas9 could be guided to specific genomic DNA sequences by a guide RNA, which also activates the nuclease (CRISPR‐Cas9). As an example, this figure shows two ZFN molecues hybridizing to genomic DNA in such a way that the fused nuclease (e.g. Fok1) becomes activated by dimerization and generates double‐ stranded DNA break at the target site. In the presence of a homology directed recombination template, which consists of a DNA of interest, flanked by sequences homologous to the genomic DNA, two possible outcomes can occur: (i) homology directed recombination (HDR) resulting in repair of a genetic lesion or insertion of a DNA of interest to a target site, or (ii) non‐homologous end joining (NHEJ) that can introduce insertions, deletions, or substitutions (“indel”) at the DNA break repair site. HDR is useful in site‐specific gene insertion or mutation repair, whereas NHEJ is useful in generating knockouts or random mutations at target sites.

restore the correct sequence in the damaged DNA strand [85]. The HDR mechanism can be utilized to modify genomic DNA by flanking an exogenous “donor” DNA sequence with homology arms at both ends for insertion at a precise genomic location (see Figure  73.3). Although very inefficient, this method has been used successfully since the 1980s to generate knockout and knockin mice and other animal strains by targeting and screening ES cells in vitro to identify successful recombinants before proceeding with production of mutant embryos. Despite the low frequency of homologous recombination in the unmodified genome in vivo, attempts have been made to use this approach to gene editing. As DNA–RNA hybridization is stronger than DNA–DNA hybridization, investigators have tested RNA– DNA chimera complexed with lactosylated polyethyleneimine to increase the efficiency of homology‐based gene repair with limited success in factor IX‐deficient mice [86] and UGT1A1‐deficient jaundiced Gunn rats [87]. After it was shown that rAAV genomes could serve as a donor template for HDR, promoterless DNA sequences consisting of open reading frames of genes of

interest, flanked by DNA sequences complementary to the desired insertion site, were used to generate rAAV vectors. To take advantage of the strong endogenous albumin promoter, rAAV8 vectors were designed to insert the HDR donor sequence into the mouse albumin locus immediately after (i.e. 3′ to) the translational stop codon [88]. Despite very low frequency of gene editing, administration of these so‐called “GeneRide” AAV8 vectors resulted in therapeutically effective levels of factor IX in hemophilia B mice [88] and Ugt1a1 knockout jaundiced mouse model of Crigler–Najjar syndrome type 1 [89]. The GeneRide vector was also used in mice transgenic for a mutant human SERPINA1 gene that expresses misfolded variant of α1‐antitrypsin (PiZ), which is associated with human α1‐antitrypsin deficiency disease [90]. The HDR donor DNA contained the open reading frame for wild‐type α1‐antitrypsin (PiM), as well as shRNA to disrupt the expression of PiZ. The procedure resulted in a very low level of gene editing, but the few hepatocytes that were gene edited were relieved of the proteotoxic load and proliferated extensively by replacing the unedited mutation‐carrying host hepatocytes [90].



73:  Liver‐Directed Gene Therapy

Genomic targeting platforms: ZNF, TALEN, and CRISPR DNA breaks can be mediated by directing nucleases to desired genomic sites by DNA sequence‐recognizing proteins, such as zinc finger proteins and transcription activator‐like effector, or by guide RNAs such as clustered, regularly interspaced, short, palindromic repeats (CRISPR). The development of chimeric DNA endonucleases fused to engineered specific sequence‐recognizing proteins, such as zinc finger nuclease (ZFN) [91], transcription activator‐like effector nuclease (TALEN) [92], and the naturally occurring CRISPR‐Cas9 system [93, 94], greatly facilitated precision genome engineering. It was quickly recognized that these systems could be used not only for targeted DNA breaks but also to deliver other biological functions to specific genomic locations. A rapidly growing tool box is now available for indel‐mediated gene knockouts, HDR‐mediated gene repair/restoration/insertion, transcriptional regulation, epigenetic modifications, and single base pair editing. What started out as targeted endonucleases is now revolutionizing the field of precision genome engineering. ZFNs and TALENs are both chimeric proteins consisting of specific DNA sequence binding domains fused to nucleases. For both systems, the desired genomic target sequence is analyzed by design algorithms to predict sets of zinc fingers or DNA binding motifs that are cloned together, fused to the desired effector domain. ZNFs and TALENs generally display high sequence specificity but are time‐consuming to design, build, and test. Also these proteins do not lend themselves to multiplexing of targets. In contrast, the CRISPR system is easy to set up for multiple targets. Derived from a bacterial defense system that degrades foreign phage DNA, it comprises Cas9, a double‐ stranded nuclease, and two short RNA sequences: one 20–21 bp seed sequence that recognizes specific DNA sequences and one structural element that activates the nuclease. The two RNAs can be combined into one synthetic small guide RNA (sgRNA). By expressing multiple sgRNA with different seed sequences, multiple targets can be cleaved simultaneously. It is also possible to inactivate the two nuclease domains in Cas9 by inserting point mutations and to fuse various functional domains to the C‐terminus without impacting sgRNA‐mediated targeting, effectively turning it into a targeted delivery platform. The ease of design and high efficiency has made CRISPR the current “favorite” in the fields of genome editing and genome ­modification. Here we will give an overview of some common applications.

Precision gene knockout The most common application of ZFN, TALEN, and CRISPR utilize the original endonuclease capable forms of these systems to generate double‐stranded DNA breaks (DSB) at precise locations. This activates the DSB repair machinery which can utilize two different pathways for DNA repair: non‐homologous end joining (NHEJ) or homology directed repair (HDR). As NHEJ always dominates over HDR, targeted DSBs are highly efficient in disrupting genes, especially by targeting exon/intron junctions, which often results in frame shifts and/or incorrect splicing, both of which give rise to nonfunctional and often unstable RNA. Targeted insertions, deletions, or substitutions (indel) has rapidly become the tool of choice for generating knockout animal strains.

987

NHEJ has also been used to create loss‐of‐function mutations for the treatment of several liver‐based genetic diseases. In a mouse model of familial hypercholesteremia (FH) knocking out the gene for the LDLR antagonist PCSK9 using CRISPR, permanently reduced plasma cholesterol levels [95]. Disruption of the HPD gene in vivo by CRISPR targeting in a mouse model of lethal hereditary tyrosinemia type I converted the disease to hereditary tyrosinemia type III, resulting in asymptomatic mice. Importantly, the gene‐edited cells exhibited a growth advantage and repopulated the liver over time, amplifying the beneficial effects of the treatment [96].

HDR‐mediated gene repair and location‐specific gene insertion Targeted DNA break has greatly increased the efficiency and specificity of targeted HDR, enabling gene targeting of increasing complexity. For precise genome modification, a donor DNA with flanking homology sequences spanning the break site is provided in excess to increase the frequency of HDR (see Figure  73.3). Methods to shift the equilibrium from NHEJ toward HDR are being introduced, such as cell cycle synchronization, chemical intervention, and transient inhibition of the NHEJ using RNA interference [85]. An early example of nuclease‐mediated genetic repair of an inherited liver‐based metabolic disease was in a hemophilia B mouse model. AAV8 vectors were used in neonatal mice to express ZFNs targeting the endogenous mouse factor IX gene and to provide an HDR donor DNA to express wild‐type human factor IX. This resulted in successful HDR in ~3% of the mouse genome, restoring functional factor IX expression from the natural factor IX promoter. Human factor IX levels in the mouse plasma were high enough to ameliorate the hemophilia phenotype [97]. This result was reproduced in adult mice by another group, demonstrating the feasibility of genetic restoration in fully grown liver [98].

Dealing with off‐target nuclease activity The Achilles heel of ZNFs, TALENs, and CRISPR is the potential for off‐target DNA cleavage due to incomplete sequence specificity. Expressed genes are more vulnerable to off‐target DNA cleavage because of the open conformation of actively transcribed chromatin. Potential biological consequences of unintended cleavage of genomic DNA include inactivation of tumor suppressors, impacting master regulators such as transcription factors or epigenetic modifiers, deactivating metabolic enzymes, activating apoptosis, and so on. Our understanding of DNA binding motifs and sgRNA target recognition is incomplete, and coupled with the fact that even single base differences can affect sequence specificity of DNA binding, great care must be taken when designing and testing ZFNs, TALENs, and CRISPR. It is hoped that advanced polymerase chain reaction (PCR) techniques, as well as whole genome sequencing will provide information concerning the frequency of off‐target breaks offering avenues for reducing this risk.

Gene therapy for neoplastic diseases Cancer gene therapy presents specific challenges. The goal is to  eliminate tumor cells while minimizing injury to normal

988

THE LIVER:  NONVIRAL VECTORS

host cells. These strategies include: (i) killing or inhibiting the growth of tumor cells, (ii) inducing an immune response against tumor cells, (iii) reducing vascular supply to tumors, and (iv) enhancing the effect of conventional therapies, such as chemotherapy and radiotherapy. Ablating tumor cells has been attempted by transferring “suicide genes,” such as the gene for herpes simplex virus thymidine kinase (HSV‐TK), which converts a prodrug, ganciclovir, to its active phosphate derivative [99]. Cytosine deaminase [100] and purine nucleoside phosphorylase (which converts fludarabine to a diffusible toxic metabolite) [101] are other genes that are being used for this purpose. p53, a sentinel gene of the cell cycle, has been used to induce apoptosis in tumor cells [102]. Clearly, it is not possible to transfer genes to all cells even in a well‐circumscribed tumor. Therefore, the efficacy of tumor kill must depend on the bystander effect that kills neighboring cells, as well as the host immune response that may kill cancer cells both in the vicinity and at remote locations. Bystander effect typically depends on the exchange of toxic agents between neighboring cells. For example, ganciclovir phosphate generated in a cell by the action of HSV‐TK may diffuse to surrounding cells, thereby extending the zone of cell killing beyond HSV‐TK transduced cells. In other cases, tumor cells undergoing apoptosis following expression of the wild‐type p53 may deliver a “kiss of death” to neighboring cells [103]. The therapeutic ratio of cancer gene therapy could be increased by limiting expression of the therapeutic gene to cancer cells. This has been attempted by conjugating the therapeutic DNA to a monoclonal antibody directed against AF‐20, a 180 kDa tumor‐specific cell surface glycoprotein expressed in hepatoma cells [104], or by using tumor‐specific promoters (e.g. α‐fetoprotein or carcinoembryonic antigen) to drive expression of the gene of interest. An alternative approach utilizes E1B‐mutant adenoviruses that are capable of replicating in cells that lack active p53, which includes many tumor cells [105]. However, although these mutant adenoviruses have been shown to kill tumor cells effectively, their replication may not always depend on the absence of p53. Another effective approach is based on expressing angiostatin or endostatin that inhibit neovascularization, which is needed for growth of both primary and metastatic tumors [106, 107]. The therapeutic effect of cancer gene therapy has been expanded beyond the transduced cells, both locally and at distant metastatic sites by cancer immunotherapy targeting “neoantigens” expressed by many tumors, which are native or altered proteins that are not expressed by normal adult cells [108]. Releasing these neoantigens by expressing toxic proteins or by physical injury (e.g. irradiation or ultrasound), followed by expression or administration of various cytokines (e.g. FMS‐like tyrosine kinase ligand, Flt3L), can evoke a host immune response against tumor cells. The tumor vaccination approach has been validated in mouse models of both ectopic and diffuse orthotopic hepatocellular carcinoma, using a combination of radiotherapy, radio‐inducible suicide gene therapy, and Flt3L expression using adenoviral vectors [109, 110]. Molecular identification of tumor‐ associated antigens, such as those from melanoma cells, may permit gene transfer‐based tumor vaccination. Isolation and genetic manipulation of antigen‐presenting cells may permit the induction of a potent immune response against tumor cells. Large tumors often produce cytokines, such as TGFβ or IL10, that may

suppress immune response or stimulate the development of immune suppressor Treg cells [111]. “Debulking” the tumor by surgery, radiotherapy, chemotherapy, or gene therapy may ­augment the effect of immunotherapy by reducing production of these immunosuppressors. Recently, oncolytic viruses that preferentially replicate in tumor cells, have expanded the role of cancer gene therapy as both tumoricidal and immune adjuvant therapy. These viruses exhibit selective replication in tumor cells with low levels of protein kinase R (PKR) and dysfunctional type I interferon (IFN) signaling. In addition, the combination of direct tumor cell lysis, release of soluble tumor antigens, and danger‐associated molecular pattern (DAMP) ligands after oncolytic virotherapy induce strong antitumoral immunity, and convert “cold” immunosuppressed tumor microenvironment into “hot” immunologically activated tumors [112]. The FDA has approved Talimogene laherparepvec (T‐VEC), a genetically modified herpes simplex virus type I for the treatment of melanoma. In clinical trials, virotherapy with T‐VEC increased T‐cell infiltration into tumors and improved the efficacy of anti‐PD1 immunotherapy [113]. With the approval of anti‐PD1 therapy for HCC, several preclinical and clinical studies with oncolytic viruses have been initiated for treatment of hepatocellular carcinoma [114, 115], including reovirus [116] and M1 alphavirus [115]. Combination of targeted agents, such as, MEK inhibitors [117] and oncolytic viruses along with immune checkpoint blockade have shown promise in the treatment of hepatocellular carcinoma. In view of the aforementioned considerations, gene therapy is likely to be effective in combination with chemotherapy or radiotherapy [118]. Irradiation of tumors has been shown to enhance the number of tumor cells that can be transduced using recombinant viruses. Expression of suicide genes, driven by radiation‐sensitive promoters, may be enhanced by irradiation [119]. On the other hand, transferring genes, such as ATM (mutated in ataxia telangiectasia), that increase the sensitivity of cells to irradiation may increase the efficacy of radiotherapy in killing tumor cells. Gene therapy for the prevention or treatment of infectious liver diseases. These approaches can be divided into (i) DNA‐ based vaccination [120], and (ii) interference with the viral life cycle by delivery of antisense RNAs, ribozymes, or DNA ribonucleases (see later). Ribozymes, antisense RNAs, or dominant negative proteins can also be expressed within the target cells by gene transfer. Interestingly, both sense and antisense RNAs can inhibit the replication of HBV [121, 122]. DNA vaccination [123] has the potential advantage over conventional vaccines in that the pure antigen is generated in vivo, which eliminates the possibility of contamination with microbes or other organic material. Furthermore, the DNA itself may serve as an adjuvant, enhancing the immune response. In this respect, the immunogenicity of antigens contained in viral vectors may turn into an advantage, permitting an immune response to poorly immunogenic molecules. Many antigens are presented poorly by antigen‐presenting cells. In these cases, the antigenic peptides may be expressed directly in the antigen‐presenting cells by gene transfer. Another approach is based on the expression of single‐ chain antibodies or antibody fragments within the cells vulnerable to viral infections by gene transfer [124]. This method, termed “intracellular vaccination,” could render cells resistant to infection by specific viruses.



73:  Liver‐Directed Gene Therapy

CONCLUSIONS After four decades of intensive research and some serious initial setbacks, there is a surge of interest in the clinical application of gene therapy. After many years of being “just around the corner”, gene therapy has extended beyond the bounds of academic endeavor and has attracted serious commercial interest. The number of companies included in the Alliance for Regenerative Medicine has increased steeply from 69 in 2014 to 255 in 2018 (www.risingtidebio.com). Although multiple hurdles remain to be overcome, not the least of which is the very high cost, there is reason to believe that the application of gene therapy for a wide variety of liver diseases will continue to expand.

ACKNOWLEDGMENT Supported by NIDDK grants 1RO1 DK092469‐010, 1 R01 DK100490‐01A1. The authors acknowledge the important contributions by many investigators in the field of liver‐directed gene therapy that could not be cited here because of limited space.

REFERENCES 1. Greeve, J., Jona, V.K., Roy Chowdhury, N. et al. Hepatic gene transfer of the catalytic subunit of the apolipoprotein B mRNA editing enzyme, APOBEC‐1, leads to reduction of LDL in normal and Watanabe heritable hyperlipidemic rabbits. J Lipid Res, 1996;37:2001–17. 2. Borel, F., Tang, Q., Gernoux, G. et  al. Survival advantage of both human hepatocyte xenografts and genome‐edited hepatocytes for treatment of α‐1 antitrypsin deficiency. Mol Ther, 2017;25:2477–89. 3. Roy Chowdhury, J., Grossman, M., Gupta, S. et al. Long‐term improvement of hypercholesterolemia after ex vivo gene therapy in LDLR‐deficient rabbits. Science, 1991;254:1802–5. 4. Grossman, M., Rader, D.J., Muller, D.W. et al. A pilot study of ex vivo gene therapy for homozygous familial hypercholesterolaemia. Nat Med, 1995;1:1148–54. 5. Zhou, H., Dong, X., Kabarriti, R. et al. Single liver lobe repopulation with wildtype hepatocytes using regional hepatic irradiation cures jaundice in Gunn rats. PLoS One, 2012;7(10):e46775. 6. Peterson, E.A., Polgar, Z., Devakanmalai, G.S. et al. Genes and pathways promoting long‐term liver repopulation by ex vivo hYAP‐ERT2 transduced hepatcytes and treatment of jaundice Gunn rats. Hepatology Communications, 2018. doi:10.1002/hep4.1278. 7. Trobridge, G.D., Miller, D.G., Jacobs, M.A. et al. Foamy virus vector integration sites in normal human cells. Proc Natl Acad Sci U S A, 2006;103:1498–503. 8. Hacein‐Bey‐Abina, S., Garrigue, A., Wang, G.P. et al. Insertional oncogenesis in 4 patients after retrovirus‐mediated gene therapy of SCID‐X1. J Clin Invest, 2008;118:3132–42. 9. Kuhn, E.J. and Geer, P.K. Genomic insulators: connecting properties to mechanism. Curr Opin Cell Biol, 2003;15:259–65. 10. Jaffe, H.A., Danel, C., and Longenecker, G. Adenovirus‐mediated in vivo gene transfer and expression in normal rat liver. Nat Genet, 1992;1:372–8. 11. Takahashi, M., Ilan, Y., Sengupta, K. et al. Induction of tolerance to recombinant adenoviruses by injection into newborn rats: long‐term amelioration of hyperbilirubinemia in Gunn rats. J Biol Chem, 1996;271:26536–42. 12. Horwitz, M.S. Adenovirudae and their replication, in Virology (eds. B.N. Fields and D.M. Knipe), Raven Press, New York, 1990, pp. 1679–721. 13. Yang, Y., Li, Q., Ertl, H.C. et al. Cellular and humoral immune responses to viral antigen create barriers to lung‐directed gene therapy with recombinant adenoviruses. J Virol, 1995;67:2004–15. 14. Morrel, N., O’Neil, W., Rice, K. et al. Administration of helper‐dependent adenoviral vectors and sequential delivery of different vector serotype for

989

long‐term liver‐directed gene transfer in baboon. Proc Natl Acad Sci U S A, 1999;96:12816–21. 15. Ilan, Y., Droguett, G., Roy Chowdhury, N. et al. Insertion of the adenoviral E3 region into a recombinant viral vector prevents antiviral humoral and cellular immune responses and permits long‐term gene expression. Proc Natl Acad Sci U S A, 1997;94:2587–92. 16. Kay, M.A., Meuse, L., Gown, A.M. et al. Transient immunomodulation with anti‐CD40 ligand antibody and CTLA4Ig enhances persistence and secondary adenovirus‐mediated gene transfer into mouse liver. Proc Natl Acad Sci U S A, 1997;94:4686–91. 17. Ilan, Y., Attavar, P., Takahashi, M. et  al. Induction of central tolerance by intrathymic inoculation of adenoviral antigens into the host thymus permits long‐term gene therapy in Gunn rats. J Clin Invest, 1996;98:2640–7. 18. Ilan, Y., Prakash, R., Davidson, A. et al. Oral tolerization to adenoviral antigens permits long‐term gene expression using recombinant adenoviral vectors. J Clin Invest, 1997;99:1098–106. 19. Schnepp, B.C., Jensen, R.L., Clark, K.R. et al. Infectious molecular clones of adeno‐associated virus isolated directly from human tissues. J Virol, 2009;83:1456–64. 20. Gao, G., Lu, Y., Calcedo, R. et al. Biology of AAV serotype vectors in liver‐ directed gene transfer to nonhuman primates. Mol Ther, 2006;13:77–87. 21. Grimm, D., Lee, J.S., Wang, L. et al. In vitro and in vivo gene therapy vector evolution via multispecies interbreeding and retargeting of adenoassociated viruses. J Virol, 2008;82:5887–911. 22. Lisowski, L., Dane, A.P., Chu, K. et al. Selection and evaluation of clinically relevant AAV variants in a xenograft liver model. Nature, 2014;506:382–6. 23. Berns, K.I. and Giraud, C. Biology of adeno‐associated virus. Curr Top Microbiol Immunol, 1996;218:1–23. 24. Samulski, R.J., Zhu, X., Xiao, X. et al. Targeted integration of adeno‐associated virus (AAV) into human chromosome 19. EMBO J, 1991;10:3941–50. 25. Yakinoglu, A.O., Heilbronn, R., Burkle, A. et  al. DNA amplification of adeno‐associated virus as a response to cellular genotoxic stress. Cancer Res, 1988;48:3123–9. 26. Yakobson, B., Hrynko, T.A., Peak, M.J. et al. Replication of adeno‐associated virus in cells irradiated with UV light at 254 nm. J Virol, 1989;63:1023–30. 27. Su, H., Lu, R., Chang, J.C. et al. Tissue‐specific expression of herpes simplex virus thymidine kinase gene delivered by adeno‐associated virus inhibits the growth of human hepatocellular carcinoma in athymic mice. Proc Natl Acad Sci U S A, 1997;94:13891–6. 28. Hosel, M., Huber, A., Bohlen, S., et al. Autophagy determines efficiency of liver‐directed gene therapy with adeno‐associated viral vectors. Hepatology, 2017;66:252–65. 29. Jiang, H., Lillicrap, D., Patarroyo‐White, S. et al. Multiyear therapeutic benefit of AAV serotypes 2, 6, and 8 delivering factor VIII to hemophilia A mice and dogs. Blood, 2006;108:107–15. 30. Bortolussi, G., Zentillin, L., Vaníkova, J. et al. Life‐long correction of hyperbilirubinemia with a neonatal liver‐specific AAV‐mediated gene transfer in a lethal mouse model of Crigler–Najjar syndrome. Hum Gene Ther, 2014;25:844–55. 31. Mingozzi, F., Liu, Y.‐L., Dobrzynski, E. et al. Induction of immune tolerance to coagulation factor IX antigen by in vivo hepatic gene transfer. J Clin Invest, 2003;111:1347–56. 32. Ziegler, R.J., Lonning, S., Armentano, D. et  al. AAV2 vector harboring a liver‐restricted promoter facilitates sustained expression of therapeutic levels of alpha‐galactosidase A and the induction of immune tolerance in Fabry mice. Molec Ther, 2004;9:231–40. 33. Ashley, S.N., Nordin, J.M.L., Buza, E.L., Greig, J.A., and Wilson, J.M. Adeno‐associated viral gene therapy corrects a mouse model of argininosuccinic aciduria. Mol Genet Metab, 2018;125:241–50. 34. Song, S. and Lu, Y. Gene delivery of alpha‐1‐antitrypsin using recombinant adeno‐associated virus (rAAV). Methods Mol Biol, 2018;1826:183–96. 35. Baloula, V., Fructuoso, M., Kassis, N. et  al. Homocysteine‐lowering gene therapy rescues signaling pathways in brain of mice with intermediate hyperhomocysteinemia. Redox Biol, 2018;19:200–9. 36. Brooks, E.D., Landau, D.J., Everitt, J.I. et al. Long‐term complications of glycogen storage disease type 1a in the canine model treated with gene replacement therapy. J Inherit Metab Dis, 2018;4:965–7. 37. Greig, J.A., Nordin, J.M.L., White, J.W. et al. Optimized adeno‐associated viral‐mediated human factor VIII gene therapy in cynomolgus macaques. Hum Gene Ther, 2018. doi:10.1089/hum.2018.080. [Epub ahead of print]. 38. Jimenez, V., Jambrina, C., Casana, E. et al. FGF21 gene therapy as treatment for obesity and insulin resistance. EMBO Mol Med, 2018;10:e8791.

990

THE LIVER:  REFERENCES

39. Brantly, M.L., Spencer, L.T., Humphries, M. et al. Phase I trial of intramuscular injection of a recombinant adeno‐associated virus serotype 2 alpha1‐ antitrypsin (AAT) vector in AAT‐deficient adults. Hum Gene Ther, 2006;17: 1177–86. 40. Flotte, T.R., Trapnell, B.C., Humphries, M. et al. Phase 2 clinical trial of a recombinant adeno‐associated viral vector expressing alpha1‐antitrypsin: interim results. Hum Gene Ther, 2011;22:1239–47. 41. Manno, C.S., Pierce, G.F., Arruda, V.R. et al. Successful transduction of liver in hemophilia by AAV‐Factor IX and limitations imposed by the host immune response. Nat Med, 2006;12:342–7. 42. Nathwani, A.C., Reiss, U.M., Tuddenham, E.G. et al. Long‐term safety and efficacy of factor IX gene therapy in hemophilia B. N Engl J Med, 2014;371:1994–2004. 43. Roy‐Chowdhury, J. and Schilsky, M.L. Gene therapy for correction of copper metabolism in Wilson disease: a “golden” opportunity using rAAV on the 50th anniversary of discovery of the virus. J Hepatol, 2016;64:265–7. 44. Ulusoy, A., Sahin, G., Bjorklund, T. et al. Dose optimization for long‐term rAAV‐mediated RNA interference in the nigrostriatal projection neurons. Mol Ther, 2009;7:1574–84. 45. Chandler, R.J., LaFave, M.C., Varshney, G.K. et al. Vector design influences hepatic genotoxicity after adeno‐associated virus gene therapy. J Clin Invest, 2015;125:870–80. 46. Wang, P.R., Xu, M., Toffanin, S. et al. Induction of hepatocellular carcinoma by in vivo gene targeting. Proc Natl Acad Sci U S A, 2012;109:11264–9. 47. Lim, L., Balakrishnan, A., Huskey, N. et  al. MicroRNA494 within an ­oncogenic microRNA megacluster regulates G1/S transition in liver tumorigenesis through suppression of mutated in colorectal cancer. Hepatology, 2014;59:202–15. 48. Lane, D.P. and Crawford, L.V. T antigen is bound to a host protein in SV40‐ transformed cells. Nature, 1979;278:261–3. 49. DeCaprio, J.A., Ludlow, J.W., Figge, J. et  al. SV40 large tumor antigen forms a specific complex with the product of the retinoblastoma susceptibility gene. Cell, 1988;54:275–83. 50. Kondo, R., Feitelson, M.A., and Strayer, D.S. Use of SV40 to immunize against hepatitis B surface antigen: implications for the use of SV40 for gene transduction and its use as an immunizing agent. Gene Ther, 1998;5:575–82. 51. Yin, H., Kanasty, R.L., Eltoukhy, A.A. et  al. Non‐viral vectors for gene‐ based therapy. Nat Rev Genetics, 2014;15:541–55. 52. Subramanian, A., Ranganathan, P., and Diamond, S.L. Nuclear targeting peptide scaffolds for lipofection of nondividing mammalian cells. Nat Biotechnol, 1999;17:873–7. 53. Fischer, D., Bieber, T., Li, Y. et al. A novel non‐viral vector for DNA delivery based on low molecular weight, branched polyethylenimine: effect of molecular weight on transfection efficiency and cytotoxicity. Pharm Res, 1999;16:1273–79. 54. Li, S., Rizzo, M.A., Bhattacharya, S. et al. Characterization of cationic lipid– protamine–DNA (LPD) complexes for intravenous delivery. Gene Ther, 1998;5:930–7. 55. MacLachlan, I., Cullis, P., and Graham, R.W. Progress towards a synthetic virus for systemic gene therapy. Curr Opin Mol Ther, 1999;1:252–59. 56. Findeis, M.A., Wu, C.H. and Wu, G.Y. Ligand‐based carrier systems for delivery of DNA to hepatocytes. Methods Enzymol, 1994;247:341–51. 57. Remy, J.S., Kichler, A., Mordvinov, V. et  al. Targeted gene transfer into hepatoma cells with lipopolyamine‐condensed DNA particles presenting galactose ligands: a stage toward artificial viruses. Proc Natl Acad Sci U S A, 1995;92:1744–8. 58. Perales, J.C., Grossmann, G.A., Molas, M. et al. Biochemical and functional characterization of DNA complexes capable of targeting genes to hepatocytes via the asialoglycoprotein receptor. J Biol Chem, 1997;272:7398–407. 59. Bandyopadhyay, P., Kren, B.T., Ma, X. et al. Enhanced gene transfer into HuH‐7 cells and primary rat hepatocytes using targeted liposomes and polyethylenimine. BioTechniques, 1998;25:282–92. 60. Cotten, M. and Wagner, E. Receptor‐mediated gene delivery strategies, in Development of Human Gene Therapy (ed. T. Friedman), Cold Spring Harbor Laboratory Press, New York, 1999, pp. 261–77. 61. Roy Chowdhury, N., Wu, C.H., Wu, G.Y. et al. Fate of DNA targeted to the liver by asialoglycoprotein receptor‐mediated endocytosis in vivo. Prolonged persistence in cytoplasmic vesicles after partial hepatectomy. J Biol Chem, 1993;268:11265–71. 62. Bartlett, D.W. and Davis, M.E. Physicochemical and biological characterization of targeted, nucleic acid‐containing nanoparticles. Bioconjug Chem, 2007;18:456–68.

63. Kren, B.T., Unger, G.M., Reding, M.T., and Steer, C.J. Targeted nanocapsules for liver cell‐type delivery of plasmids in vivo. Mol Ther, 2006;13(Suppl. 1):s414–5. 64. Fraser, J.R., Laurent, T.C., and Laurent, U.B. Hyaluronan: its nature, distribution, functions and turnover. J Intern Med, 1997;242:27–33. 65. Stockert, R.J., Haimes, H.B., Morell, A.G. et al. Endocytosis of asialoglycoprotein–enzyme conjugates by hepatocytes. Lab Invest, 1980;43:556–63. 66. Bommineni, V.R., Roy Chowdhury, N., Wu, G.Y. et al. Depolymerization of hepatocellular microtubules after partial hepatectomy. J Biol Chem, 1994;269:25200–5. 67. Godbey, W.T., Wu, K.K., Hiraski, G.J. et  al. Improved packing of poly(ethylenimine)/DNA complexes increases transfection efficiency. Gene Ther, 1999;6:1380–8. 68. Ramani, K., Hassan, Q., Venkaiah, B., Hasnain, S.E., and Sarkar, D.P. Site‐ specific gene delivery in vivo through engineered Sendai viral envelopes. Proc Natl Acad Sci U S A, 1998;95:11886–90. 69. Wang, X., Sarkar, D.P., Mani, P. et al. Proteoliposomes containing Sendai virus F‐glycoprotein for hepatocyte‐targeted gene therapy of the Gunn rat model of Crigler–Najjar syndrome type 1. Gastroenterology, 2008;134: W1862. 70. Zanta, M.A., Belguise‐Valladier, P., and Behr, J.‐P. (1998). Gene delivery: a single nuclear localization signal peptide is sufficient to carry DNA to the cell nucleus. Proc Natl Acad Sci U S A, 1998;96:91–6. 71. Ivics, Z., Hackett, P.B., Plasterk, R.H. et  al. Molecular reconstruction of Sleeping Beauty, a Tc1‐like transposon from fish, and its transposition in human cells. Cell, 1997;91:501–10. 72. Izsvak, Z., Ivics, Z., and Plasterk, R.H. Sleeping Beauty, a wide host‐range transposon vector for genetic transformation in vertebrates. J Mol Biol, 2000;302:93–102. 73. Bell, J.B., Podetz‐Pedersen, K.M., Aronovich, E.L. et al. Preferential delivery of the Sleeping Beauty transposon system to livers of mice by hydrodynamic injection. Nat Protoc, 2007;2:3153–65. 74. Mikkelsen, J.G., Yant, S.R., Meuse, L. et al. Helper‐independent Sleeping Beauty transposon‐transposase vectors for efficient nonviral gene delivery and persistent gene expression in vivo. Mol Ther, 2003;8:654–65. 75. Ohlfest, J.R., Frandsen, J.L., Fritz, S. et al. Phenotypic correction and long‐ term expression of factor VIII in hemophilic mice by immunotolerization and nonviral gene transfer using the Sleeping Beauty transposon system. Blood, 2005;105:2691–8. 76. Aronovich, E.L., Bell, J.B., Belur, L.R. et al. Prolonged expression of a lysosomal enzyme in mouse liver after Sleeping Beauty transposon‐mediated gene delivery: implications for non‐viral gene therapy of mucopolysaccharidoses. J Gene Med, 2007;9:403–15. 77. Wilber, A., Wangensteen, K.J., Chen, Y. et al. Messenger RNA as a source of transposase for Sleeping Beauty transposon‐mediated correction of hereditary tyrosinemia type I. Mol Ther, 2007;15:1280–7. 78. Kren, B.T., Ghosh, S.S., Linehan, C.L. et al. Hepatocyte‐targeted delivery of Sleeping Beauty mediates efficient transposition in vivo. Gene Ther Mol Biol, 2003;7:231–40. 79. Kren, B.T., Unger, G.M., Trossen, A.A. et al. Long‐term FVIII expression via Sleeping Beauty in liver sinusoidal endothelial cells of transgenic mice. Mol Ther, 2007;15:s341. 80. Yant, S.R., Wu, X., Huang, Y. et al. High‐resolution genome‐wide mapping of transposon integration in mammals. Mol Cell Biol, 2005;25:2085–94. 81. Kebriaei, P., Huls, H., Singh, H. et  al. Adoptive therapy using Sleeping Beauty gene transfer system and artificial antigen presenting cells to manufacture T cells expressing CD19‐specific chimeric antigen receptor. Blood, 2014;124:311. 82. Martin‐Higueras, C., Luis‐Lima, S., and Salido, E. Glycolate oxidase is a safe and efficient target for substrate reduction therapy in a mouse model of primary hyperoxaluria type 1. Mol Ther, 2016;24:719–25. 83. Lai, C., Pursell, N., Gierut, J. et  al. Specific inhibition of hepatic lactate dehydrogenase reduces oxalate production in mouse models of primary hyperoxaluria. Molec Ther, 2018;26:1983–95. 84. Capecchi, M.R. Altering the genome by homologous recombination. Science, 1989;244:1288–92. 85. Pawelczak, K.S., Gavande, N.S., VanderVere‐Carozza, P.S., and Turchi, J.J. Modulating DNA repair pathways to improve precision genome engineering. ACS Chem Biol, 2018;13(2):389–96. 86. Kren, B.T., Bandyopadhyay, P., and Steer, C.J. In vivo site‐directed mutagenesis of the factor IX gene by chimeric RNA/DNA oligonucleotides. Nat Med, 1998;4:285–90.



73:  Liver‐Directed Gene Therapy

87. Roy Chowdhury, J., Huang, T.J., Kesari, K. et  al. Molecular basis for the  lack of bilirubin‐specific and 3‐methylcholanthrene‐inducible UDP‐ glucuronosyltransferase activities in Gunn rats. The two isoforms are encoded by distinct mRNA species that share an identical single base ­deletion. J Biol Chem, 1991;266:18294–8. 88. Barzel, A., Paulk, N.K., Shi, Y. et al. Promoterless gene targeting without nucleases ameliorates haemophilia B in mice. Nature, 2015;517(7534): 360–4. 89. Porro, F., Bortolussi, G., Barzel, A. et  al. Promoterless gene targeting ­without nucleases rescues lethality of a Crigler–Najjar syndrome mouse model. EMBO Mol Med 2017;9:1346–55. 90. Borel, F., Tang, Q., and Gernoux, G. Survival advantage of both human hepatocyte xenografts and genome‐edited hepatocytes for treatment of α‐1 antitrypsin deficiency. Mol Ther, 2017;25:2477–89. 91. Klug, A. The discovery of zinc fingers and their applications in gene regulation and genome manipulation. Annu Rev Biochem, 2010;79:213–31. 92. Zhang, M., Wang, F., Li, S. et al. TALE: a tale of genome editing. Prog Biophys Mol Biol, 2014;114:25–32. 93. Jinek, M., Chylinski, K., and Fonfara, I. A programmable dual‐RNA‐guided DNA endonuclease in adaptive bacterial immunity. Science, 2012;337: 816–21. 94. Mali, P., Yang, L., Esvelt, K.M. et al. RNA‐guided human genome engineering via Cas9. Science, 2013;339:823–6. 95. Ding, Q., Strong, A., Patel, K.M., et  al. Permanent alteration of PCSK9 with in vivo CRISPR‐Cas9 genome editing. Circ Res, 2014;115:488–92. 96. Pankowicz, F.P., Barzi, M., Legras, X. et  al. Reprogramming metabolic ­pathways in vivo with CRISPR/Cas9 genome editing to treat hereditary tyrosinaemia. Nat Commun, 2016;30:7:12642. 97. Li, H., Haurigot, V., Doyon, Y. et al. In vivo genome editing restores haemostasis in a mouse model of haemophilia. Nature, 2011;475:217–21. 98. Anguela, X.M., Sharma, R., Doyon, Y. et al. Robust ZFN‐mediated genome editing in adult hemophilic mice. Blood 2013;122:3283–7. 99. Kokoris, M.S., Sabo, P., Adman, E.T. et al. Enhancement of tumor ablation by a selected HSV‐1 thymidine kinase mutant. Gene Ther, 1999;6: 1415–26. 100. Mullen, C.A., Coale, M.M., Lowe, R. et al. Tumors expressing the cytosine deaminase suicide gene can be eliminated in vivo with 5‐fluorocytosine and induce protective immunity to wild type tumor. Cancer Res, 1994;54: 1503–6. 101. Mohr, L., Shankara, S., Yoon, S.‐K. et al. Gene therapy of hepatocellular carcinoma in vitro and in vivo in nude mice by adenoviral transfer of the Escherichia coli purine nucleoside phosphorylase gene. Hepatology, 2000;31:606–4. 102. Roth, J.A., Nguyen, D., Lawrence, D.D. et al. Retrovirus‐mediated wild‐ type p53 gene transfer to tumors of patients with lung cancer. Nat Med, 1996;2:985–91. 103. Frank, D.K., Frederick, M.J., Liu, T.J. et al. Bystander effect in the adenovirus‐mediated wild‐type p53 gene therapy model of human squamous cell carcinoma of the head and neck. Clin Cancer Res, 1998;4:2521–28. 104. Mohr, L., Schauer, J.I., Boutin, R.H. et al. Targeted gene transfer to hepatocellular carcinoma cells in vitro using a novel monoclonal antibody‐based gene delivery. Hepatology, 1999;29:82–9. 105. Harada, J.N. and Berk, A.J. p53‐independent and ‐dependent requirements for E1B‐55K in adenovirus type 5 replication. J Virol, 1999;73:5333–44. 106. Tanaka, T., Cao, Y., Folkman, J. et al. Viral vector‐targeted anti‐angiogenic gene therapy utilizing an angiostatin complementary DNA. Cancer Res, 1998;58:3362–9.

991

107. Blezinger, P., Wang, J., Gondo, M. et  al. Systemic inhibition of tumor growth and tumor metastases by intramuscular administration of the endostatin gene. Nat Biotechnol, 1999;17:343–8. 108. Paillard, F. Immunosuppression mediated by tumor cells: a challenge for immunotherapeutic approaches. Hum Gene Ther, 2000;11:657–8. 109. Kawashita, Y., Deb, N.J., Garg, M., et al. An autologous in situ tumor vaccination approach for hepatocellular carcinoma. 1. Flt3 ligand gene transfer increases antitumor effects of a radio‐inducible suicide gene therapy in an ectopic tumor model. Rad Res, 2014;182:191–200. 110. Kawashita, Y., Deb, N.J., Garg, M. et al. An autologous in situ tumor vaccination approach for hepatocellular carcinoma. 2. Tumor‐specific immunity and cure after radio‐inducible suicide gene therapy and systemic CD40‐ligand and Flt3‐ligand gene therapy in an orthotopic tumor model. Rad Res, 2014;182:201–10. 111. Fakhrai, H., Dorigo, O., Shawler, D.L. et  al. Eradication of established intracranial rat gliomas by transforming growth factor β antisense gene therapy. Proc Natl Acad Sci U S A, 1996;93:2909–14. 112. Dyer, A., Baugh, R., Chia, S.L. et al. Turning cold tumours hot: oncolytic virotherapy gets up close and personal with other therapeutics at the 11th Oncolytic Virus Conference. Cancer Gene Ther, 2019;26(3–4):59–73. doi:10.1038/s41417‐018‐0042‐1. Epub 2018 Sep 4. 113. Kohlhapp, F.J. and Kaufman, H.L. Molecular pathways: mechanism of action for talimogene laherparepvec, a new oncolytic virus immunotherapy. Clin Cancer Res, 2016;22:1048–54. 114. Yoo, S.Y., Badrinath, N., Woo, H.Y., and Heo J. Oncolytic virus‐based immunotherapies for hepatocellular carcinoma. Mediators Inflamm, 2017:5198798. doi:10.1155/2017/5198798. Epub 2017 Apr 20. 115. Zhang, H., Li, K., Lin, Y. et al. Targeting VCP enhances anticancer activity of oncolytic virus M1 in hepatocellular carcinoma. Sci Transl Med, 2017;9:404. 116. Samson, A., Bentham, M.J., Scott, K. et al. Oncolytic reovirus as a combined antiviral and anti‐tumour agent for the treatment of liver cancer. Gut, 2018;67:562–73. 117. Bommareddy, P.K., Aspromonte, S., Zloza, A. et  al. MEK inhibition enhances oncolytic virus immunotherapy through increased tumor cell killing and T cell activation. Sci Transl Med, 2018;10:471. 118. Heise, C., Sampson‐Johannes, A., Williams, A. et al. ONYX‐015, an E1B gene‐attenuated adenovirus, causes tumor‐specific cytolysis and anti‐ tumoral efficacy that can be augmented by standard chemotherapeutic agents. Nat Med, 1997;3:639–45. 119. Kawashita, Y., Ohtsuru, A., Kaneda, Y. et al. Regression of hepatocellular carcinoma in vitro and in vivo by radiosensitizing suicide gene therapy under the inducible and spatial control of radiation. Hum Gene Ther, 1999;10:1509–19. 120. Encke, J., zu Putlitz, J., and Wands, J.R. DNA vaccines. Intervirology, 1999;42:117–24. 121. zu Putlitz, J. and Wands, J.R. Specific inhibition of hepatitis B virus replication by sense RNA. Antisense Nucleic Acid Drug Dev, 1999;9:241–52. 122. zu Putlitz, J., Wieland, S., Blum, H.E. et al. Antisense RNA complementary to hepatitis B virus specifically inhibits viral replication. Gastroenterology, 1998;115:702–13. 123. Encke, J., zu Putlitz, J., Geissler, M. et al. Genetic immunization generates cellular and humoral immune responses against the nonstructural proteins of the hepatitis C virus in a murine model. J Immunol, 1998;161:4917–23. 124. zu Putlitz, J., Skerra, A., and Wands, J.R. Intracellular expression of a cloned antibody fragment interferes with hepatitis B virus surface antigen secretion. Biochem Biophys Res Commun, 1999;255:785–91.

74

Telomeres and Telomerase in Liver Generation and Cirrhosis Sonja C. Schätzlein1 and K. Lenhard Rudolph2 Spark@FLI, Leibniz Institute on Aging – Fritz Lipmann Institute (FLI), Jena, Germany Research Group on Stem Cell Aging, Leibniz Institute on Aging – Fritz Lipmann Institute (FLI), Jena, Germany

1 2

INTRODUCTION The liver has a tremendous capacity to regenerate in response to injury. In the context of chronic liver disease, hepatocytes can regenerate for 20–40 years before the proliferative capacity declines, which coincides with the development of cirrhosis and organ failure. On a molecular level, telomere shortening has been linked to hepatocyte regenerative failure and studies on telomerase‐deficient mice provided experimental proof that ­telomere shortening causes impairment in liver regeneration by induction of DNA damage checkpoints and hepatocyte senescence. At early disease stages, hepatocytes readily reenter the cell cycle and participate in liver regeneration. In contrast, at late disease stages the induction of hepatocyte senescence results in the activation of stellate cells, increased fibrosis, and cirrhosis formation. There is evidence that subsets of hepatocytes exhibit stem cell‐like features including increases in Wnt expression and elevated activity of the telomerase enzyme, which can synthesize telomeres de novo. These stem cell‐like hepatocytes harbor increased self‐renewal capacity and are the origin of liver regeneration during homeostasis in mice. In human cirrhosis, an increased activation of stem cell marker positive hepatocytes occurs at late stages of the disease and marks an increasing risk for the development of organ failure and hepatocarcinogenesis. Together, these findings support a model of cirrhosis development according to which telomere shortening and the exhaustion of stem cell‐like hepatocytes lead to the progression of chronic liver disease toward cirrhosis and disease complications associated with it.

TELOMERE SHORTENING LIMITS THE PROLIFERATIVE CAPACITY OF HUMAN CELLS It was first recognized by Leonard Hayflick that the proliferative capacity of human cells is limited to a finite number of cell divisions [1]. When human fibroblasts are grown in cell culture, proliferation ceases after 50–70 cell divisions and cells enter permanent cell cycle arrest, which is referred to as replicative senescence. Senescent fibroblasts show typical morphological alterations including an enlarged cytoplasm and an increased activity of senescence‐associated β‐galactosidase. Using nuclear transplantation experiments Hayflick recognized that the cell nucleus carries the memory that limits the proliferative capacity of human cells. However, it took 30 more years for scientists to  disclose the molecular mechanism responsible for this phenomenon. Today, it is known that telomere shortening is the underlying cause limiting the proliferative capacity of primary human cells including hepatocytes [2, 3]. Telomeres form the ends of human chromosomes [4] and consist of small tandem DNA repeats (TTAGGGn in human cells). Telomeres do not encode for a protein product. The main function of telomeres is to cap the chromosomal ends. Telomere capping is essential to distinguish the chromosome ends from DNA breaks within the chromosome. Human telomeres are 5–15 kb long [5]. It has been shown that a minimum telomere length of 84 base pairs (14 TTAGGG repeats) is required for telomere capping [6]. To fulfill capping, telomeres

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



74:  Telomeres and Telomerase in Liver Generation and Cirrhosis

form tertiary structures, such as telomere loops or G‐quadruplexes [7]. The tertiary structure of telomeres is stabilized by telomere binding proteins that specifically bind to telomeric DNA [8]. There is evidence that alterations in the expression of telomere binding proteins can be the cause of diminished organ maintenance and increased cancer formation [9, 10]. Telomeres shorten during each round of cell division because of (i) the end replication problem of DNA polymerase, and (ii) the processing of telomeres during the cell cycle [11, 12]. Human telomeres shorten at a rate of 50–100 base pairs per cell division [13]. Experiments on re‐expression of the enzyme telomerase have proven that telomere shortening is the underlying cause of senescence limiting proliferation of human cells [14]. Telomerase synthesize telomere sequences de novo [15]. The enzyme consists of two essential components: (i) the telomerase reverse transcriptase (TERT) is the catalytic subunit of the enzyme [16–18]; (ii) the telomerase RNA component (TERC) serves as template for the synthesis of telomere sequence [19–21]. In addition, there is evidence that processing of TERC and TERT is required for the generation of a functional enzyme complex [22, 23]. In human cells and tissues, TERC is expressed ubiquitously. In contrast, telomerase activity is only detectable in embryonic tissues but is repressed in most somatic tissues after birth [24]. In the adult organism, high levels of telomerase are restricted to germline stem cells. Low levels of telomerase activity remain detectable in somatic stem cells of adult tissues and in activated lymphocytes [25]. The postnatal suppression of telomerase activity correlates with a repression in the expression of Tert (the catalytic subunit of the enzyme), which remains

993

detectable in germ cells and some somatic stem cells in adult humans [26]. Studies on human Tert (hTert) promoter activity revealed that hTert expression is reactivated in hepatocytes in response to regeneration stimuli in reporter mice [27]. These findings suggest that transient hTert activation may contribute to the maintenance of regenerative capacity of hepatocytes during chronic liver diseases. In mice, increases in mTert expression were identified in a subpopulation of stem cell‐like hepatocytes that represent the origin of liver regeneration during homeostasis in mice, and are required for unperturbed liver regeneration in response to injury [28].

TELOMERE SHORTENING IN HUMAN AGING, LIVER DISEASE, AND CIRRHOSIS Telomere shortening occurs in almost all tissues and organs during human aging [5, 25]. In addition, there is evidence that chronic diseases accelerating the rate of cell turnover also increase the rate of telomere shortening as documented in chronic liver disease. Cell division rates are low and there is limited telomere shortening in healthy liver during aging [29–31]. However, chronic liver diseases accelerate the rate of hepatocyte turnover, which leads to increases in telomere shortening in liver of affected patients compared to age‐matched control patients without liver disease (Figure 74.1) [29, 32–34]. Aside the evidence on telomerase expression in activated hepatocytes and in subpopulations of stem cell‐like hepatocytes, the level of

20 to 40 years injury chronic damage

Hepatocytes (wnt and tert high) Senescent Hepatocytes (SASP secretion and loss of proliferative competition) Dysplastic Nodules (Terct Activation)

Activated Stellate Cell+ Fibrosis

Senescence Associated Secretory Phenotype (SASP)

Figure 74.1  The telomere concept of cirrhosis formation. Chronic liver diseases stimulate hepatocyte proliferation and liver regeneration. In principal, all hepatocytes can regenerate the liver but there is also evidence for stem cell‐like hepatocytes with increased Wnt or telomerase expression. These stem cell‐like hepatocytes may represent a reserve population, and activation of stem and progenitor cells occurs in cirrhosis. However, these cells cannot prevent the overall decrease in telomere length and the induction of hepatocyte senescence at the end stage of chronic liver ­disease. As a consequence, the regenerative capacity declines leading to stellate cell activation, an increase in fibrosis, and the development of ­cirrhosis. Senescent hepatocytes promote the activation of stellate cells and the selection of dysplastic nodules by increasing secretion of inflammatory cytokines (SASP) and by limiting the proliferative competition of normal, nontransformed hepatocytes thus enabling the dominance of dysplastic nodules.

994

THE LIVER:  SENESCENCE AND CRISIS CHECKPOINTS IN RESPONSE TO TELOMERE SHORTENING

telomerase activity is not sufficient to prevent (i) telomere shortening, and (ii) the induction of hepatocyte senescence and regenerative exhaustion at the end stage of chronic liver disease. Cirrhosis evolves at the end stage of chronic liver disease and is characterized by critical telomere shortening [29, 31–33], a decline in cell proliferation [34, 35], and an upregulation of senescence markers (Sa‐βGal and p21) [29, 36–41]. These data support the concept that telomere shortening and senescence inhibit hepatocyte proliferation, thus promoting the evolution of cirrhosis at the end stage of chronic liver disease (Figure 74.1). In agreement with this hypothesis, there is genetic evidence that telomere shortening influences human aging and organ maintenance. It was shown that telomerase mutations in humans lead to premature telomere shortening, impaired organ maintenance, and reduced survival. The first clinical example was ­dyskeratosis congenita (DKC). The autosomal dominant form of this disease is caused by mutations in the RNA component of telomerase [42]. Another form of the disease is caused by mutations in the dyskerin gene, which is necessary for the processing of small nuclear RNAs. Dyskerin also processes the RNA component of telomerase and DKC patients with dyskerin mutations show reduced levels of TERC expression [43]. DKC patients have abnormally short telomeres and shortened life expectancy due to the development of bone marrow failure and other d­ isease pathologies such as intestinal failure, cirrhosis, and cancer [44]. Of note, telomerase mutations have been linked to cirrhosis formation in patients with spontaneous cirrhosis associated with chronic liver disease [45, 46]. Together, these studies indicate that human telomere reserves are limited and that a heterozygous mutation of telomerase can have detrimental effects on organ homeostasis and survival early in life and also represent a risk factor for cirrhosis formation in chronic liver disease. Interestingly, it was shown that somatic mutations in the hTert promoter, leading to an activation of hTert expression, counter‐balance loss of function mutations in individuals with inherited (germline) loss‐of‐function mutations in telomerase genes [47]. These findings support the concept that telomere shortening represents a context of growth inhibition driving the selection of hTert promoter mutations, which also occur at high frequency in cirrhosis‐associated development of hepatocellular carcinoma (HCC) (see Chapter 60 ‘Mutations and Genomic Alterations in Liver Cancer’).

DNA DAMAGE RESPONSES (DDR) TO SHORTENED TELOMERES INDUCE GROWTH ARREST, APOPTOSIS, AND CHROMOSOMAL INSTABILITY It has been shown that telomeres protect chromosomal ends from being recognized as DNA damage by two different mechanisms: (i) Telomeres form loop structures or G‐quadruplexes preventing the recognition of telomere ends as chromosome brakes; (ii) Telomeres bind shelterin proteins that prevent the activation of DNA damage response at chromosome ends. If telomeres lose capping function, chromosome ends are recognized as DNA breaks, which induce DDRs including the ATM/ATR‐dependent activation of p53 as well as the induction of DNA repair responses that re‐ligate DNA breaks, such as non‐homologous end joining (NHEJ) or alternative

end joining (AltEJ). The activation of end joining repair pathways results in the formation of chromosomal fusions, which can trigger chromosomal breakage, chromosomal instability, and aneuploidy if cells continue to divide in the p­ resence of fused chromosomes – a scenario that leads to cancer initiation in aged tissues with short telomeres (see later; for review, see [48]). In response to telomere shortening, telomere uncapping appears to evolve in two steps: first, telomeres become too short for the formation of tertiary structures (loops and quadruplexes). Consequently, unfolded telomeres are recognized as DNA breaks leading to activation of AMT/p53 pathways. At this stage of telomere dysfunction, shelterin proteins continue to bind to short stretches of telomere sequences that remain at the end of chromosomes even though the telomeres are unfolded. This first stage of telomere dysfunction has been referred to as intermediate stage of telomere dysfunction as the activation of DNA repair responses and the formation of chromosomal fusions remain suppressed at this stage [49]. However, in response to further cell division the telomeric sequences can be completely lost, which then also leads to the induction repair response and chromosomal fusions. This second stage of telomere dysfunction aggravates DNA damage by induction of breakage and ­religation of fused chromosomes in each round of cell division [50, 51]. This process has been shown to lead to induction of tumor initiation, especially in the context of abrogated DNA damage checkpoints, such as p53 mutations (for review, see [48]). Fusion‐bridge‐breakage cycles also aggravate tissue atrophy by elevating DNA damage and DDRs. Mechanistically, the formation of chromosomal fusions in response to telomere dysfunction is activated by end resection of dysfunctional ­ ­telomeres and the inhibition of this pathway can rescue chromosomal fusions, tissue atrophy, and the life span of telomerase‐ deficient mice with dysfunctional telomeres [52].

SENESCENCE AND CRISIS CHECKPOINTS IN RESPONSE TO TELOMERE SHORTENING Senescence and crisis represent the main checkpoints limiting cell proliferation and cell survival in response to telomere dysfunction [53]. It has been shown that a subset of dysfunctional telomeres is sufficient to induce replicative senescence [54]. The main characteristic of senescence is the permanent loss of cell proliferation. The induction of senescence depends on the induction of p53‐ and Rb‐checkpoints [55]. When these checkpoints are abrogated, cells can bypass the senescence checkpoint and continue to divide with critically short telomeres. In consequence, a second checkpoint, which is referred to as crisis, limits the survival of cells with dysfunctional telomeres. Aggravation of telomere shortening leads to complete telomere uncapping and chromosomal fusions, which in turn induce chromosomal instability and the activation of p53‐independent DNA damage responses, and cell death (see earlier, and [55]). The molecular signaling pathways that induce crisis are less well understood than senescence. However, it has been shown that chromosomal fusion and prolonged mitotic arrest trigger telomere deprotection and cell death in crisis [56].



74:  Telomeres and Telomerase in Liver Generation and Cirrhosis

Rare cells that bypass the crisis checkpoint (1 per 107 in human fibroblasts) must activate a mechanism of telomere ­stabilization. Studies on human fibroblast have shown that two‐ thirds of surviving clones show a spontaneous reactivation of telomerase  –  the enzyme synthesizing telomeres de novo [55, 57]. The remaining one‐third of the clones activates an alternative mechanism of telomere elongation (ALT). The molecular basis of ALT is not yet completely understood, but it involves the activation of DNA repair pathways and epigenetic alterations that lead to telomere exchange between different chromosomes and a highly heterogeneous telomere length [58]. The high prevalence of activating hTert promoter mutations in human HCCs indicates that the stable reactivation of telomerase is the main route of cell immortalization during hepatocarcinogenesis. The mechanisms that underlie the age‐associated linear increase in DNA mutations in normal liver are not yet fully understood [59]. In addition, it is conceivable that replication‐ associated DNA mutation [60] increases in chronic liver diseases that increase the rates of hepatocyte turnover [60]. Both types of DNA mutation could contribute to telomerase promoter activation during hepatocarcinogenesis (see Chapter  60 ‘Mutations and Genomic Alterations in Liver Cancer’ for a detailed review of Tert promoter mutation in HCC). Senescence and crisis checkpoints act as tumor suppressor checkpoints thereby limiting the proliferative capacity of genetically unstable cells with dysfunctional telomeres. As a downside, the same checkpoints compromise the regenerative reserve of tissues and organs harboring short telomeres as a consequence of aging or chronic disease, such as in chronic liver diseases. Aside the evidence that telomere dysfunction‐ induced checkpoints represent a causal factor for cirrhosis formation, the induction of these checkpoints can also increase the risk of cancer formation by different mechanisms: (i) there is evidence that senescent cells promote the growth of tumor cells in a cell non‐autonomous manner by secreting inflammatory cytokines and growth factors – which is referred to as the senescence‐associated secretory phenotype (SASP) [61, 62]; (ii) the crisis checkpoint in response to telomere dysfunction is associated with high levels of chromosomal instability, which can lead to induction of genetic lesions promoting cancer formation [48]; and (iii) both checkpoints (senescence and crisis) can increase the selection of genetically altered, pre‐tumorous cells by inhibiting proliferative competition of nontransformed hepatocytes (for a review of the concept, see [63]).

STEM CELL‐LIKE HEPATOCYTES IN LIVER REGENERATION The telomere hypothesis of cirrhosis formation indicates that cirrhosis is the consequence of telomere shortening, hepatocyte senescence, and impaired regenerative reserve (Figure 74.1). In response to liver injury, almost all hepatocytes can reenter the cell cycle, but the number of hepatocytes contributing to liver regeneration decreases in the context of telomere shortening [64]. While these studies suggested that the overall capacity of hepatocyte regeneration declines in response to telomere shortening, it was also shown that the liver contains a subpopulation

995

of stem cell‐like hepatocytes fueling hepatocyte replacement during homeostasis. Labeling of Wnt‐active hepatocytes (via an  Axin2‐reporter) identified diploid hepatocytes expressing Axin2‐activity and Tbx3 (a marker of hepatocyte progenitors) that surround the central vein and give rise to polyploidy, Tbx3‐ negative hepatocytes, throughout the entire liver lobe during organ homeostasis [65]. In response to liver injury, a subset of cells near bile ducts show an induction of Wnt‐expression (Lgr5‐reporter) and contribute to the regeneration of both hepatocytes and bile ducts [66]. Also Sox9‐positive oval cell progenitor cells were shown to have the potential to give rise to hepatocytes and bile duct cells in organoid cultures but showed little capacity to produce hepatocytes in vivo [67]. Aside the presence of stem cell‐like cells in liver, it was shown that differentiated liver cells harbor plasticity to transdifferentiate into another differentiated cell type, for example, hepatocytes into cholangiocytes and vice versa [68, 69]. Together, both the proliferation of stem cell‐like liver cell and the transdifferentiation of differentiated liver cells into another differentiated cell type may contribute to liver regeneration. The relevance of these process for liver regeneration in chronic liver diseases and the exhaustion of these processes during cirrhosis formation in humans remains incompletely understood. There is evidence that stem and progenitor cells become increasingly activated in the late stage of chronic liver diseases and during the development of cirrhosis [70, 71] suggesting that these stem and progenitor cells may represent a backup system to ensure regeneration when regenerative capacities decline. Given the evidence that telomere shortening limits liver regeneration in chronic liver diseases, it will be important to delineate how different subpopulations of stem and progenitor cells in liver (that can contribute to liver regeneration) are affected by telomere shortening and how this impinges on the decline in the capacity of liver regeneration in the context of cirrhosis formation. It was demonstrated that a subpopulation of hepatocytes in mice are marked by high expression levels of mTert coinciding with increased telomerase activity [28]. These mTert‐high hepatocytes locate in a scattered pattern throughout the liver lobes and exhibit an increased capacity to regenerate new ribbons of hepatocytes during homeostasis. Of note, the depletion of mTert‐high hepatocytes accelerates fibrosis formation in response to injury in mouse models [28]. Another mouse model revealed evidence that the hTert promoter is activated in hepatocytes during liver regeneration [27]. Together, these studies indicate that telomerase activity can counter‐balance telomere loss in subpopulations of highly regenerative hepatocytes in the context of chronic liver disease. However, these mechanisms seem to be limited in human chronic liver disease given the fact that telomere shortening and hepatocyte senescence increase at the end stage of chronic liver diseases transiting into cirrhosis development (see earlier).

TARGETING TELOMERE TO DEVELOP ANTI‐CIRRHOSIS THERAPIES As outlined earlier, telomere shortening and hepatocyte senescence represent underlying causes of regenerative exhaustion and cirrhosis development at the end stage of chronic liver

996

THE LIVER:  TARGETING TELOMERE TO DEVELOP ANTI‐CIRRHOSIS THERAPIES

diseases (Figure  74.1). In response to partial hepatectomy, ­telomere shortening reduces the number of regenerating hepatocytes in mTerc−/− mice with shortened telomeres compared to mTerc+/+ mice with long telomere reserves [64]. Hepatocytes with short telomeres express markers of senescence, fail to reenter the cell cycle, and thus are blocked from contributing to liver regeneration [64]. These data provided a proof of concept that telomere shortening limits liver regeneration by reducing the number of hepatocytes reentering the cell cycle. In response to chronic liver injury, telomere dysfunction accelerated the activation of stellate cells, fibrotic scarring, and the induction of steatosis [72]. Growth‐restricted tissue environments can increase the selection of cancerous clones in different tissues (for review, see [63]) and the increase in senescent hepatocytes in cirrhosis [29] could drive the selection of malignant clones by increasing the expression of inflammatory, pro‐growth signals

(see earlier, and [73]. Given that the end stage of cirrhosis can often not be reverted and transplantation therapies remain restricted to a small group of the patients, the development of molecular therapies that aim to improve liver regeneration and the progression of cirrhosis, represent an unmet medical need in hepatology. Several lines of evidence indicate that telomere ­stabilizing therapies or the inhibition of tissue destructive DDRs could be explored as therapeutic targets in this regard.

Transient telomerase activation When hTert is over‐expressed in adult human fibroblast or fetal hepatocytes, telomerase is activated, leading to telomere stabilization and immortal proliferation ([2, 3] and Figure  74.2). Today, it is well accepted that the suppression of hTert expression is responsible for telomere shortening and the limited life

Telomere Structure

N O

N N

N

Therapeutic Interventions

O

Exo1 Endresection

Telomere Elongation

Telomerase

N N

DNA Damage Response

N

N

Therapeutic Interventions

O

p53

CIN

Senescence

Figure 74.2  Telomere targeting anti‐cirrhosis therapies. Telomere shortening leads to loss of telomere capping function. In response to critical telomere shortening, 3D structures of telomeres, such as t‐loops, can no longer form. Uncapped chromosomal ends elicit DNA damage responses involving the induction of p53‐dependent cell senescence and DNA repair pathways. DNA repair at dysfunctional telomeres results in chromosomal fusions, further increasing DNA damage, chromosomal instability, and tissue destruction. Two molecular targets could be explored as future therapies to prevent cirrhosis progression: (i) The transient activation of telomerase could restore telomere length and regenerative capacity of hepatocytes; (ii) The inhibition of repair pathways could prevent the formation of chromosomal fusions and the aggravation of DNA damage and tissue destruction.



74:  Telomeres and Telomerase in Liver Generation and Cirrhosis

span of human cells. Of note, the reactivation of hTert expression does not lead to transformation of human cells into cancer cells [74]. However, there are reports that extensive culture of TERT‐transduced cells can lead to chromosomal instability, gene mutations, and transformation of human cells [75, 76]. Together, these studies indicate that a transient reactivation of telomerase could be employed as a therapeutic approach to improve the regenerative capacity of hepatocytes and to prevent the progression of cirrhosis at the end stage of chronic liver diseases (Figure 74.2). Indeed, transient activation of telomerase is sufficient to improve liver regeneration and to prevent induction of fibrosis in telomerase‐deficient mice with shortened telomeres [72]. Pharmacological or molecular approaches for nontoxic, transient activation of telomerase remain yet to be developed, and the consequences of transient telomerase activation on liver tumor development need to be investigated. The high prevalence of activating mutations of the hTert‐promoter in hepatocellular carcinoma (see Chapter 60 ‘Mutations and Genomic Alterations in Liver Cancer’) indicates that telomere shortening‐induced inhibition of hepatocyte proliferation increases the selection of telomerase‐expressing hepatocytes at the cirrhosis–carcinoma transition. Theoretically, the ­transient activation of telomerase could alleviate impairment in liver regeneration thus reducing the selection of mutant, pre‐­ tumorigenic hepatocytes at the cirrhosis stage. In line with this hypothesis, studies on mouse models revealed experimental ­evidence that the loss of proliferative competition of hematopoietic stem and progenitor cells increases the selection of pre‐­ leukemic cell clones and leukemia development [77]. It is also possible that transient activation of telomerase would lower the accumulation of senescent hepatocytes in the liver thus decreasing the p­ ro‐tumorigenic effects of the SASP (see earlier, and [73]). In ­contrast to potential anti‐tumor effects, the transient activation of telomerase could increase the growth capacity of premalignant cell clones that occur in increasing frequency in cirrhosis. In sum, it will be important to carefully evaluate the net outcome of transient telomerase activation on liver regeneration, hepatocarcinogenesis, and organism survival in humanized mouse models of telomere shortening. A potential problem in the development of pro‐regenerative telomerase therapies may include the irreversibility of senescence. If a high load of senescent cell is already present in cirrhosis, telomerase activation may fail to reactivate the proliferative capacity of these cells, which may limit the overall induction of liver regeneration in response to such therapies. However, studies in late generation telomerase‐deficient mice provided a proof of concept that reactivation of telomerase is successful in reverting late stage of massive tissue atrophy induced by telomere dysfunction [78]. These data suggest that telomerase mediated activation of proliferative competent cells could be successful in reversing late‐stage cirrhosis. Aside the role of hepatocyte senescence in driving regenerative exhaustion and cirrhosis development, a protective role of stellate cell senescence in limiting fibrotic scarring has been revealed in a mouse model of repeated massive rounds of tissue damage [79]. Unlashing senescence of fibrotic cells could have the unwanted effect of increasing fibrosis. However, it remains to be investigated whether stellate cell senescence occurs in human cirrhosis, which usually evolves as a consequence of

997

chronic low‐level tissue damage and may thus not involve the strong mitotic activity of chemically induced cirrhosis in mouse models, which may be a prerequisite for the induction of stellate cell senescence.

Inhibition of tissue destructive DDRs The induction of DDRs limits proliferation of cells with ­dysfunctional telomeres by induction of cellular senescence or  crisis checkpoints (see earlier). The induction of these ­checkpoints inhibits the formation of tumors that can originate for genetically unstable cells with dysfunctional telomeres. However, DDR‐induced DNA repair pathways also contribute to the formation of chromosomal fusions, which aggravate DNA damage, and chromosomal instability (Figure  74.2). These repair responses can thus enhance the progression of tissue atrophy and cancer initiation in tissues with dysfunctional ­ ­telomeres, as in cirrhosis. Studies in mouse models provided a proof of concept that the selective inhibition of p21‐dependent senescence or Puma‐dependent apoptosis checkpoints can prolong tissue maintenance in the context of telomere dysfunction without increasing the evolution of chromosomal instability and cancer initiation [80, 81]. Also the inhibition of Exo1‐dependent end resection blocked chromosomal fusion formation and prolonged tissue maintenance and chromosomal stability in ­ ­telomere dysfunctional mice ([52] and Figure 74.2). The development and preclinical testing of inhibitors targeting these ­tissue‐destructive checkpoint or repair responses remains yet to be conducted. Such studies should include the analysis of long‐ lived animal models to assess the risk of tumor formation of such approaches. If this risk is low, the therapeutic approach could be promising in patients with late‐stage cirrhosis who are not amenable for liver transplantation therapies.

CONCLUSIONS Together, telomere shortening appears to have a prominent role in driving the regeneration failure of hepatocytes and the exhaustion hepatic stem and progenitor cells at late stages of chronic liver disease and during cirrhosis development. The adverse effects of telomere dysfunction on organ regeneration and disease progression are mediated by DNA damage checkpoints and repair responses. Targeting the tissue‐destructive effects of these DDRs could guide the development of future therapies aiming to prolong liver regeneration and organ function in patients who cannot be treated by transplantation therapies.

REFERENCES 1. Hayflick, L. and Moorhead, P.S. The serial cultivation of human diploid cell strains. Exp Cell Res, 1961;25(3):585–621. 2. Bodnar, A.G., Ouellette, M., Frolkis, M. et  al. Extension of life‐span by ­introduction of telomerase into normal human cells. Science, 1998;279(5349): 349–52. 3. Wege, H., Chui, M.S., Le, H.T., Strom, S.C., and Zern, M.A. In vitro expansion of human hepatocytes is restricted by telomere‐dependent replicative aging. Cell Transplant, 2003;12(8):897–906.

998

THE LIVER:  REFERENCES

4. Blackburn, E.H. Structure and function of telomeres. Nature, 1991;350(6319): 569–73. 5. Allsopp, R.C., Chang, E., Kashefi‐Aazam, M. et al. Telomere shortening is associated with cell division in vitro and in vivo. Exp Cell Res, 1995;220(1):194–200. 6. Capper, R., Britt‐Compton, B., Tankimanova, M. et  al. The nature of telomere fusion and a definition of the critical telomere length in human cells. Genes Dev, 2007;21(19):2495–508. 7. Griffith, J.D., Comeau, L., Rosenfield, S. et al. Mammalian telomeres end in a large duplex loop. Cell, 1999;97(4):503–14. 8. de Lange, T. Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev, 2005;19(18):2100–10. 9. Munoz, P., Blanco, R., Flores, J.M., and Blasco, M.A. XPF nuclease‐dependent telomere loss and increased DNA damage in mice overexpressing TRF2 result in premature aging and cancer. Nat Gen, 2005;37(10):1063. 10. Hartmann, K., Illing, A., Leithäuser, F. et al. Gene dosage reductions of Trf1 and/or Tin2 induce telomere DNA damage and lymphoma formation in aging mice. Leukemia, 2016;30(3):749. 11. Lingner, J., Cooper, J.P., Cech, T. Telomerase and DNA end replication: no longer a lagging strand problem? Science, 1995;269(5230):1533–5. 12. Levy, M.Z., Allsopp, R.C., Futcher, A.B., Greider, C.W., and Harley, C.B.  Telomere end‐replication problem and cell aging. J Mol Biol, 1992;225(4):951–60. 13. Harley, C.B., Futcher, A.B., and Greider, C.W. Telomeres shorten during ageing of human fibroblasts. Nature, 1990;345(6274):458–60. 14. Bodnar, A.G., Ouellette, M., Frolkis, M. et  al. Extension of life‐span by  introduction of telomerase into normal human cells. Science, 1998;279(5349):349–52. 15. Greider, C.W. and Blackburn, E.H. Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell, 1985;43(2 Pt 1):405–13. 16. Meyerson, M., Counter, C.M., Eaton, E.N. et al. hEST2, the putative human telomerase catalytic subunit gene, is up‐regulated in tumor cells and during immortalization. Cell, 1997;90(4):785–95. 17. Nakamura, T.M., Morin, G.B., Chapman, K.B. et  al. Telomerase catalytic subunit homologs from fission yeast and human. Science, 1997;277(5328): 955–9. 18. Weinrich, S.L., Pruzan, R., Ma, L. et al. Reconstitution of human telomerase with the template RNA component hTR and the catalytic protein subunit hTRT. Nat Gen, 1997;17(4):498–502. 19. Greider, C.W. and Blackburn E.H. A telomeric sequence in the RNA of Tetrahymena telomerase required for telomere repeat synthesis. Nature, 1989;337(6205):331–7. 20. Feng, J., Funk, W.D., Wang, S.S. et  al. The RNA component of human ­telomerase. Science, 1995;269(5228):1236–41. 21. Yu, G.‐L., Bradley, J.D., Attardi, L.D., and Blackburn, E.H. In vivo alteration of telomere sequences and senescence caused by mutated Tetrahymena ­telomerase RNAs. Nature, 1990;344(6262):126. 22. Chen, L., Roake, C.M., Freund, A. et  al. An activity switch in human telomerase based on RNA conformation and shaped by TCAB1. Cell, ­ 2018;174(1):218–30.e13. doi: 10.1016/j.cell.2018.04.039 23. Venteicher, A.S., Meng, Z., Mason, P.J., Veenstra, T.D., and Artandi, S.E. Identification of ATPases pontin and reptin as telomerase components essential for holoenzyme assembly. Cell, 2008;132(6):945–57. 24. Wright, W.E., Piatyszek, M.A., Rainey, W.E., Byrd, W., and Shay, J.W. Telomerase activity in human germline and embryonic tissues and cells. Dev Genet, 1996;18(2):173–9. 25. Jiang, H., Ju, Z., and Rudolph, K.L. Telomere shortening and ageing. Z Gerontol Geriatr, 2007;40(5):314–24. 26. Kolquist, K.A., Ellisen, L.W., Counter, C.M. et al. Expression of TERT in early premalignant lesions and a subset of cells in normal tissues. Nat Gen, 1998;19(2):182. 27. Sirma, H., Kumar, M., Meena, J.K. et al. The promoter of human telomerase reverse transcriptase is activated during liver regeneration and hepatocyte proliferation. Gastroenterology, 2011;141(1):326–37.e3. 28. Lin, S., Nascimento, E.M., Gajera, C.R. et  al. Distributed hepatocytes expressing telomerase repopulate the liver in homeostasis and injury. Nature, 2018;556(7700):244. 29. Wiemann, S.U., Satyanarayana, A., Tsahuridu, M. et al. Hepatocyte telomere shortening and senescence are general markers of human liver cirrhosis. FASEB J, 2002;16(9):935–42.

30. Verma, S., Tachtatzis, P., Penrhyn‐Lowe, S. et al. Sustained telomere length in hepatocytes and cholangiocytes with increasing age in normal liver. Hepatology, 2012;56(4):1510–20. 31. Aikata, H., Takaishi, H., Kawakami, Y. et al. Telomere reduction in human liver tissues with age and chronic inflammation. Exp Cell Res, 2000;256(2): 578–82. 32. Kitada, T., Seki, S., Kawakita, N., Kuroki, T., and Monna, T. Telomere shortening in chronic liver diseases. Biochem Biophys Res Commun, 1995;211(1):33–9. 33. Urabe, Y., Nouso, K., Higashi, T. et  al. Telomere length in human liver ­diseases. Liver, 1996;16(5):293–7. 34. Delhaye, M., Louis, H., Degraef, C. et al. Relationship between hepatocyte proliferative activity and liver functional reserve in human cirrhosis. Hepatology, 1996;23(5):1003–11. 35. Delhaye, M., Louis, H., Degraef, C. et al. Hepatocyte proliferative activity in human liver cirrhosis. J Hepatol, 1999;30(3):461–71. 36. Sasaki, M., Ikeda, H., Yamaguchi, J., Nakada, S., and Nakanuma, Y. Telomere shortening in the damaged small bile ducts in primary biliary cirrhosis reflects ongoing cellular senescence. Hepatology, 2008;48(1): 186–95. 37. Plentz, R.R., Park, Y.N., Lechel, A. et al. Telomere shortening and inactivation of cell cycle checkpoints characterize human hepatocarcinogenesis. Hepatology, 2007;45(4):968–76. 38. Lunz, J.G., 3rd, Tsuji, H., Nozaki, I., Murase N., and Demetris, A.J. An inhibitor of cyclin‐dependent kinase, stress‐induced p21Waf‐1/Cip‐1, mediates hepatocyte mito‐inhibition during the evolution of cirrhosis. Hepatology, 2005;41(6):1262–71. 39. Sasaki, M., Ikeda, H., Haga, H., Manabe, T, and Nakanuma, Y. Frequent ­cellular senescence in small bile ducts in primary biliary cirrhosis: a possible role in bile duct loss. J Pathol, 2005;205(4):451–9. 40. Wagayama, H., Shiraki, K., Sugimoto, K. et  al. High expression of p21WAF1/CIP1 is correlated with human hepatocellular carcinoma in patients with hepatitis C virus‐associated chronic liver diseases. Human Pathol, 2002;33(4):429–34. 41. Crary, G.S. and Albrecht, J.H. Expression of cyclin‐dependent kinase inhibitor p21 in human liver. Hepatology, 1998;28(3):738–43. 42. Vulliamy, T., Marrone, A., Goldman, F. et al. The RNA component of telomerase is mutated in autosomal dominant dyskeratosis congenita. Nature, 2001;413(6854):432–5. 43. Mitchell, J.R., Wood, E., and Collins, K. A telomerase component is defective in the human disease dyskeratosis congenita. Nature, 1999;402(6761):551. 44. Dokal, I. Dyskeratosis congenita in all its forms. Br J Haematol, 2000;110(4):768–79. 45. Hartmann, D., Srivastava, U., Thaler, M. et al. Telomerase gene mutations are associated with cirrhosis formation. Hepatology, 2011;53(5):1608–17. 46. Calado, R.T., Brudno, J., Mehta, P. et al. Constitutional telomerase mutations are genetic risk factors for cirrhosis. Hepatology, 2011;53(5):1600–7. 47. Maryoung, L., Yue, Y., Young, A. et  al. Somatic mutations in telomerase ­promoter counterbalance germline loss‐of‐function mutations. J Clin Invest, 2017;127(3):982–6. 48. Gunes, C., Avila, A.I., and Rudolph, K.L. Telomeres in cancer. Differentiation, 2018;99:41–50. 49. Van Ly, D., Low, R.R.J., Frölich, S. et  al. Telomere‐loop dynamics in ­chromosome end protection. bioRxiv, 2018:279877. 50. Hackett, J.A., Feldser, D.M., and Greider, C.W. Telomere dysfunction increases mutation rate and genomic instability. Cell, 2001;106(3):275–86. 51. Rudolph, K.L., Millard, M., Bosenberg, M.W., and DePinho, R.A. Telomere dysfunction and evolution of intestinal carcinoma in mice and humans. Nat Gen, 2001;28(2):155–9. 52. Schaetzlein, S., Kodandaramireddy, N.R., Ju, Z. et  al. Exonuclease‐1 ­deletion impairs DNA damage signaling and prolongs lifespan of telomere‐ dysfunctional mice. Cell, 2007;130(5):863–77. 53. Shay, J.W. and Wright, W.E. Telomeres are double‐strand DNA breaks ­hidden from DNA damage responses. Mol Cell, 2004;14(4):420–1. 54. Zou, Y., Sfeir, A., Gryaznov, S.M., Shay, J.W., and Wright, W.E. Does a ­sentinel or a subset of short telomeres determine replicative senescence? Mol Biol Cell, 2004;15(8):3709–18. 55. Wright, W.E. and Shay, J.W. The two‐stage mechanism controlling cellular senescence and immortalization. Exp Gerontol, 1992;27(4):383–9. 56. Hayashi, M.T., Cesare, A.J., Rivera, T., and Karlseder, J. Cell death during crisis is mediated by mitotic telomere deprotection. Nature, 2015; 522(7557):492.



74:  Telomeres and Telomerase in Liver Generation and Cirrhosis

57. Shay, J.W. and Wright, W.E. Quantitation of the frequency of immortalization of normal human diploid fibroblasts by SV40 large T‐antigen. Exp Cell Res, 1989;184(1):109–18. 58. Reddel, R.R. and Bryan, T.M. Alternative lengthening of telomeres: dangerous road less travelled. The Lancet, 2003;361(9372):1840–1. 59. Blokzijl, F., De Ligt, J., Jager, M. et al. Tissue‐specific mutation accumulation in human adult stem cells during life. Nature, 2016;538(7624):260. 60. Tomasetti, C., Li, L., and Vogelstein, B. Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science, 2017;355(6331): 1330–4. 61. Pribluda, A., Elyada, E., Wiener, Z. et al. A senescence‐inflammatory switch from cancer‐inhibitory to cancer‐promoting mechanism. Cancer Cell, 2013; 24(2):242–56. 62. Campisi, J. Cancer, aging and cellular senescence. In Vivo, 2000; 14(1):183–8. 63. DeGregori, J. Adaptive Oncogenesis: A New Understanding of How Cancer Evolves Inside Us, Harvard University Press, Cambridge, MA, 2018. 64. Satyanarayana, A., Wiemann, S.U., Buer, J. et  al. Telomere shortening impairs organ regeneration by inhibiting cell cycle re‐entry of a subpopulation of cells. EMBO J, 2003;22(15):4003–13. 65. Wang, B., Zhao, L., Fish, M., Logan, C.Y., and Nusse, R. Self‐renewing diploid Axin2(+) cells fuel homeostatic renewal of the liver. Nature, 2015;524(7564):180–5. 66. Huch, M., Dorrell, C., Boj, S.F. et al. In vitro expansion of single Lgr5+ liver stem cells induced by Wnt‐driven regeneration. Nature, 2013;494(7436): 247–50. 67. Tarlow, B.D., Finegold, M.J., and Grompe, M. Clonal tracing of Sox9+ liver progenitors in mouse oval cell injury. Hepatology, 2014;60(1):278–89. 68. Kopp, J.L., Grompe, M., and Sander, M. Stem cells versus plasticity in liver and pancreas regeneration. Nat Cell Biol, 2016;18(3):238–45. 69. Schaub, J.R., Huppert, K.A., Kurial, S.N. et  al. De novo formation of the biliary system by TGFβ‐mediated hepatocyte transdifferentiation. Nature, 2018;557(7704):247.

999

70. Stueck, A.E. and Wanless, I.R. Hepatocyte buds derived from progenitor cells repopulate regions of parenchymal extinction in human cirrhosis. Hepatology, 2015;61(5):1696–707. 71. Van Haele, M. and Roskams, T. Hepatic progenitor cells: an update. Gastroenterol Clin, 2017;46(2):409–20. 72. Rudolph, K.L., Chang, S., Millard, M., Schreiber‐Agus, N., and DePinho, R.A. Inhibition of experimental liver cirrhosis in mice by telomerase gene delivery. Science, 2000;287(5456):1253–8. 73. Laberge, R.‐M., Sun, Y., Orjalo, A.V. et al. MTOR regulates the pro‐tumorigenic senescence‐associated secretory phenotype by promoting IL1A translation. Nat Cell Biol, 2015;17(8):1049. 74. Harley, C.B. Telomerase is not an oncogene. Oncogene, 2002;21(4): 494–502. 75. Schreurs, M.W., Hermsen, M.A., Geltink, R.I. et al. Genomic stability and functional activity may be lost in telomerase‐transduced human CD8+ T lymphocytes. Blood, 2005;106(8):2663–70. 76. Serakinci, N., Guldberg, P., Burns, J.S. et  al. Adult human mesenchymal stem cell as a target for neoplastic transformation. Oncogene, 2004;23(29): 5095–8. 77. Bilousova, G., Marusyk, A., Porter, C.C., Cardiff, R.D., and DeGregori, J. Impaired DNA replication within progenitor cell pools promotes leukemogenesis. PLoS Biol, 2005;3(12):e401. 78. Jaskelioff, M., Muller, F.L., Paik, J.‐H. et al. Telomerase reactivation reverses tissue degeneration in aged telomerase‐deficient mice. Nature, 2011; 469(7328):102. 79. Krizhanovsky, V., Yon, M., Dickins, R.A. et al. Senescence of activated stellate cells limits liver fibrosis. Cell, 2008;134(4):657–67. 80. Choudhury, A.R., Ju, Z., Djojosubroto, M.W. et al. Cdkn1a deletion improves stem cell function and lifespan of mice with dysfunctional telomeres without accelerating cancer formation. Nat Genet, 2007;39(1):99–105. 81. Sperka, T., Song, Z., Morita, Y. et  al. Puma and p21 represent cooperating checkpoints limiting self‐renewal and chromosomal instability of somatic stem cells in response to telomere dysfunction. Nat Cell Biol, 2012;14(1):73–9.

75

Toxins and Biliary Atresia Michael Pack1,2 and Rebecca G. Wells1,3,4 Departments of Medicine (GI Division) and Cell and Developmental Biology and 3 Pathology and Laboratory Medicine, Perelman School of Medicine, and 4 Bioengineering, School of Engineering and Applied Sciences, University of Pennsylvania, Philadelphia, PA, USA 1 2

INTRODUCTION Biliary atresia (BA) is a rapidly progressive cholangiopathy of neonates notable for pronounced fibro‐obliteration of the bile ducts, particularly extrahepatic bile ducts, which in the majority of patients results in end‐stage liver disease. It is uniquely diag­ nosed in neonates, sparing mothers and older children. New clinical evidence suggests that the onset of BA is in utero and that, although BA patients are born apparently healthy, all who go on to develop the disease have detectable elevations in con­ jugated bilirubin within the first few days after birth [1]. This suggests that the neonatal bile duct is uniquely susceptible to injury and to a fibrotic response. The etiological agent of BA, however, is unknown. Although multiple large‐scale genetic studies have identified potential disease modifiers, there is no evidence that BA is a primary genetic disease [2]. Rather, most sources suggest that the cause is environmental [3]. The identification of the isofla­ vonoid biliatresone  –  thought to be responsible for a BA‐like disease in calves and lambs born to livestock who consumed biliatresone‐containing plants while pregnant  –  led to the ­demonstration that short‐term exposure to a toxin could cause selective extrahepatic biliary damage in larval zebrafish [4] (Figures 75.1 and 75.2). The combination of the livestock and zebrafish data provided a proof of principle that a toxin could cause bile duct damage in neonates, sparing mothers. Biliatresone is unlikely to be consumed by pregnant women, and no toxin with similar effects has been identified. The bilia­ tresone story, however, raises the possibility that such toxins exist and are significant contributors to the burden of disease in humans. This chapter reviews the data suggesting that BA has an envi­ ronmental cause with a particular emphasis on the evidence that toxins contribute to other human biliary diseases and to other

forms of human organ fibrosis. There are several plausible mechanisms whereby a toxin could mediate a complex BA‐like disease, and biliatresone may provide important clues about the identity and mechanisms of toxins in BA.

TOXIC CAUSES OF BILIARY INJURY Cholangiocytes are metabolically active cells that have impor­ tant roles in normal liver physiology including modification of  bile, antigen presentation, and immune signaling [5, 6]. Cholangiocyte injury in response to environmental exposures can disrupt or aberrantly activate these processes, in some cases leading to fibrotic duct destruction. While biliatresone is the only known environmental agent linked to a naturally occurring biliary fibrosis model [4, 7, 8], toxicants and medicinal drugs that target cholangiocytes are well described [9, 10]. None of these agents cause a BA‐like syndrome, but understanding the mechanisms of injury associated with these exposures may offer insights into the pathogenesis of BA and other biliary diseases.

BILIATRESONE Beginning in 1964, there were four naturally occurring large‐ scale outbreaks of BA in Australian livestock. All were associ­ ated with severe droughts leading to unusual pasturing and ingestion of plants in the Dysphania genus, strongly suggesting a toxic etiology [11]. Chemical fractionation of Dysphania littoralis and D. glomulifera ssp. glomulifera collected during one outbreak, guided by an in vivo zebrafish bioassay, led to the  identification of the previously unknown isoflavonoid biliatresone as the responsible agent [4]. Additional studies ­

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



75:  Toxins and Biliary Atresia

Liver histology

Liver grossly

2 week old

1 week old

Extrahepatic biliary tree

1001

Figure 75.1  Histology and gross pictures from two minimally symptomatic lambs in a flock in 2013 in which nearly 50% of lambs developed a BA‐like disease. The extrahepatic bile ducts (EHBD) are fibrotic and obstructed, while the livers show minimal damage. Size bar, 100 μm. Source: Reprinted from [7] under the terms of the Creative Commons Attribution Non‐Commercial License CC BY‐NC.

(a)

(b)

(c)

OCH3

O

OH

O OCH3

O Biliatresone

Figure 75.2  Biliatresone structure and activity in zebrafish larvae. (a) Biliatresone structure as determined by mass spectroscopy. Red arrow points to electrophilic alpha‐methylene ketone moiety. (b, c) Confocal projections through the extrahepatic biliary system and a segment of the liver of 5‐day post‐fertilization zebrafish larvae. (b) Control larva showing normal gallbladder and intrahepatic bile ducts. Round structures in the gall­ bladder are epithelial cell nuclei. (c) Larva treated with biliatresone for 24 hours showing destruction of the gallbladder with sparing of the intrahe­ patic bile ducts. GB, gallbladder; IHD, intrahepatic bile ducts; bil, biliatresone.

confirmed that biliatresone was toxic to mammalian cholangio­ cytes grown as 3D spheroids and to neonatal mouse bile ducts in explant culture [4], and that synthetic biliatresone had similar biological effects [12]. Initial biochemical studies showed that biliatresone was a strong electrophile that covalently bound reduced glutathione, amino acids, and nucleic acids, thus pro­ viding insight into a mechanistic basis for toxicity [13]. Subsequent studies in the zebrafish model and mammalian cholangiocyte culture systems highlighted the important role of antioxidant defenses in the cholangiocyte response to bilia­ tresone [7, 8]. Biliatresone lowered hepatic glutathione levels in both zebrafish larvae and mammalian cholangiocytes and glu­ tathione depletion sensitized the extrahepatic cholangiocytes (EHC) of zebrafish larvae to low doses of biliatresone that are normally not toxic, leading to injury of the normally resist­ ant intrahepatic cholangiocytes (IHC) in larvae treated with

standard doses of the toxin. Use of a transgenic fish that expresses an in vivo glutathione redox sensor showed that sus­ ceptibility to biliatresone correlates with the basal redox status of cholangiocytes. Specifically, EHC are more oxidized than IHC and hepatocytes before treatment with biliatresone, and they are further oxidized when injury is first detected, whereas the resistant IHC and hepatocytes were not. Highlighting the  importance of antioxidant defenses in the cholangiocyte response to biliatresone, glutathione depletion had a ­comparable effect on the sensitivity of mammalian cholangiocytes to the toxin. Furthermore, treatment with N‐acetylcysteine, a glu­ tathione precursor, blocked toxicity in both the zebrafish and cholangiocyte spheroid models. A similar albeit less pronounced effect was seen in response to treatment with the Nrf2‐activator sulforaphane (which upregulates transcriptional responses to oxidant stress) in both systems.

1002

THE LIVER:  3,5‐DIETHOXYCARBONYL‐1,4‐DIHYDRO‐COLLIDINE

Transcriptional profiling experiments in zebrafish and mam­ malian cholangiocytes identified other cellular stress responses and signaling pathways that are activated by biliatresone [7, 8]. In the zebrafish model, activation of heat shock pathways, autophagy, and ER stress responses were prominent at the earli­ est stages of injury. In cholangiocytes, an important role for the depletion of the transcription factor Sox17 was identified [8]. Remarkably, heterozygous mutation of Sox17 was linked to bil­ iary injury in mice [14]. Ongoing studies in zebrafish and mammalian models should provide fresh insight into the mechanism of biliatresone toxic­ ity. In particular, it will be important to understand biliatresone metabolism and how this affects its ability to cross the placenta and gain access to the fetal biliary system. Studies in the zebrafish model suggest that secretion of biliatresone or one of its metabolites into bile is required for toxicity [4], which is not surprising as dietary isoflavones undergo extensive enterohe­ patic circulation. Determining how this affects maternal metab­ olism and placental transfer of biliatresone is essential to understanding the impact of fetal exposures to this and other xenobiotics on the biliary system.

4,4′‐METHYLENEDIANILINE The most extensively characterized toxicant associated with bil­ iary injury in humans is 4,4′‐methylenedianiline (MDA). Accidental or occupational exposure to MDA, which is used in the manufacture of epoxy resins, polyurethanes, and other indus­ trial products, causes reversible cholangitis and cholestatic liver disease [15, 16]. In rodents, acute exposure to MDA causes severe intrahepatic cholangitis that is associated with biliary and periportal necrosis and reduced bile flow [17, 18], particularly in larger interlobular ducts compared with small ducts and ductules. In animals with longer exposures to higher doses of MDA, chol­ angitis progresses to periportal biliary fibrosis. Extrahepatic duct and hepatocellular injury and fibrosis also occur with longer exposure to MDA in animal models [19]; however, this has not been reported in humans exposed to MDA. Metabolic studies have provided insight into the mechanism of MDA‐mediated biliary toxicity. Parabiosis experiments con­ ducted in rodents indicate that injury arises from cholangiocyte exposure to MDA or one of its metabolites that are present in bile [19]. Several lines of evidence argue that glutathionylation of MDA and its subsequent biliary secretion is an important mechanism for detoxification: first, glutathione conjugates are the initial MDA metabolites identified in bile of MDA‐treated rodents; second, toxicity is more pronounced in female rats, which have reduced secretion of glutathionylated MDA into bile compared to males [20]; third, and most significant, MDA tox­ icity is enhanced by pharmacological depletion of glutathione, presumably because it increases secretion of other, more toxic metabolites in bile [21]. Whether enhanced MDA toxicity caused by glutathione depletion also results from changes in cholangiocyte stress responses, as has been shown for bilia­ tresone [7, 8], has not been examined. The cell biological mechanisms of MDA‐mediated cholangio­ cyte injury are not fully understood. There is strong evidence, however, supporting mitochondrial toxicity as a cause of ATP

depletion and tight junction abnormalities, as shown by ultras­ tructural analyses and changes in transepithelial resistance, para­ cellular flow, and cation selectivity [20]. Other studies suggest alterations in ECM and coagulation pathways. Mitochondrial toxicity argues that MDA induces oxidative stress although, unlike biliatresone, MDA treatment does not reduce total hepatic GSH, presumably because it is not as strong an oxidant. Thus, even though glutathione depletion worsens toxicity of both MDA and biliatresone [7, 8], it is likely that biliatresone and MDA work, at least in part, through distinct mechanisms. This may also contribute to their respective predilections for EHC versus IHC.

ALPHA‐NAPHTHYLISOTHIOCYANATE Alpha‐naphthylisothiocyanate (ANIT) is a second well‐charac­ terized toxicant that induces biliary cell injury [22–24]. Like MDA, ANIT causes injury to cholangiocytes lining rodent inter­ lobular ducts within several hours of an acute exposure. Chronic exposure to ANIT (weeks duration) leads to biliary fibrosis. Injury to hepatocytes following acute exposure is minimal and develops significantly later than biliary injury, likely secondary to the initial duct damage. Ultrastructural and histological analyses show that ANIT, like MDA, alters cholangiocyte microvilli and tight junctions, suggesting a shared injury mechanism [24]. The metabolism of  the two compounds is similar in that both are glutathio­ nylated. Unlike MDA, however, the ANIT‐glutathione conju­ gate secreted into bile is toxic, as it readily dissociates, thereby allowing free ANIT to enter cholangiocytes [25–27]. Consistent with this model, pharmacologic depletion of glutathione reduces ANIT toxicity. In contrast, glutathione depletion enhances tox­ icity of MDA and biliatresone [7, 8, 28]. Transcriptional profiling experiments in rat cholangiocytes showed that ANIT exposure led to increased expression of genes involved in the ER stress/unfolded protein response and glutathione metabolism and antioxidant responses, similar to biliatresone‐treated zebrafish larvae; however, the magnitude of the responses was less in ANIT‐ than in biliatresone‐treated ani­ mals [7, 29]. Like MDA, ANIT does not lower total hepatic glu­ tathione, presumably because it is not as strong an oxidant as biliatresone. Nonetheless, treatment with an Nrf2 activator blunts ANIT toxicity [30], similar to its effect on biliatresone toxicity, although this may involve upregulation of genes involved in ANIT metabolism rather than glutathione‐mediated antioxidant responses, as GSH depletion rescues rather than enhances ANIT toxicity.

3,5‐DIETHOXYCARBONYL‐1,4‐DIHYDRO‐ COLLIDINE 3,5‐Diethoxycarbonyl‐1,4‐dihydro‐collidine (DDC) is a third xenobiotic with well‐characterized biliary toxicity [9, 31]. Chronic feeding of DDC causes progressive injury to large and mid‐size interlobular bile ducts, leading to a ductular reaction that culmi­ nates in an onion‐skin pattern of portal fibrosis resembling patho­ logical changes in sclerosing cholangitis. Plasticine casts of the



75:  Toxins and Biliary Atresia

intrahepatic biliary network in DDC‐treated animals showed seg­ mental bile duct strictures and duct dilation. DDC does not cause extrahepatic duct obstruction although wall thickening and neutro­ philic infiltration of the extrahepatic ductal network can be detected in histological analyses. DDC‐mediated hepatocyte injury is initially limited compared to its early effects on the biliary system, although DDC‐mediated biliary injury likely arises from early effects on both hepatocytes and cholangiocytes [31–33]. DDC treatment is associated with reduced levels of glu­ tathione in bile. This was initially attributed to reduced expres­ sion of the Mrp2 canalicular transporter, although levels of total hepatic glutathione are increased in Mrp2 mutant mice [33]. A more likely explanation is that oxidative stress induced by changes in porphyrin metabolism leads to decreased hepatic glu­ tathione. Consistent with this idea, DDC induces hepatic nuclear translocation of Nrf2, an important transcriptional regulator of antioxidant responses, as shown by western blot analyses of hepatic lysates [34]. While this likely reflects Nrf2 activation in response to altered porphyrin metabolism in hepatocytes, it is conceivable that the combined effects of porphyrins and low glu­ tathione in bile trigger Nrf2 activation in cholangiocytes.

Drug‐induced cholangiocyte injury Biliary toxicity is an uncommon side effect of exposure to a large number of drugs. Injury typically manifests as a cholesta­ sis syndrome with histological evidence of a mixed periportal infiltrate (neutrophils, eosinophils, and lymphocytes), cholan­ giocyte injury (without apoptosis), and evidence of bile stasis [31, 35]. A ductular reaction and bile duct destruction occur with prolonged exposure. By definition, a bile duct to portal tract ratio of T in exon 26 with decreased expression of ABCB1 in the duodenum of patients with the T allele (variant) compared to those with the C allele (wild type) [69]. However, a subsequent study reported

that this effect might be due to the non‐synonymous SNP 2677G>T/A [70], which is frequently linked to C3435T. In contrast to these earlier reports, Gerloff and colleagues reported no changes in digoxin clearance between Caucasian patients carrying the variant or the wild‐type allele [71]. Elevated ABCB1 expression was reported in Japanese and Caucasian patients carrying the variant allele (3435T) [72, 73]. The non‐synonymous SNP 2677G>T/A is also the subject of conflicting data [74]. Kimchi‐Sarfaty and colleagues analyzed the role of synonymous mutations in protein folding and function [75]. Syn­o­ nymous SNPs (3435C>T, 1236C>T) and non‐synonymous 2677G>T in the ABCB1 gene sequence result in a protein with altered drug and inhibitor interactions, without a change in expression levels possibly due to altered protein folding related to a change in the rhythm of translation [75]. Based on our current knowledge, overall drug bioavailability is only moderately influenced by ABCB1 polymorphisms, as compared to variants of the drug metabolizing enzymes (CYP family). Although the findings for ABCC1 showing a high rate of polymorphisms are similar to those for ABCB1, studies on ABCG2 report a significant effect of polymorphisms on ABCG2 expression and function [76, 77]. The correlation of ABC transporter genetic variants to treatment outcomes is gradually being clarified, yet the overall ­picture is still puzzling, as much of the published data are conflicting. Nevertheless, many studies reporting correlation between SNPs and clinical outcome indicate the necessity to pursue further investigations. The Pharmacogenomics Research Network plays a crucial role in the development of this complex field [78]. One of their various goals is to understand how genetic variation in membrane transporters contributes to variation in drug transport.

CONCLUSION ABC transporters are active efflux transporters that play a major role in the metabolism of endogenous compounds and drugs. Six ABC transporters have been localized to the apical membrane of the hepatocyte, including ABCB1, B4, B11, C2, and the heterodimer ABCG5/G8. Five additional transporters have been localized to the basolateral membrane, including ABCC1, C3, C4, C5, and C6. Beside these 11 ABC transporters, we used a recent database that was established based upon a quantitative proteomic profiling of subcellular organelles, to generate a map with the subcellular compartments in which rat liver ABC transporters are most enriched. Although this map is constructed from rat orthologs of human ABC proteins, this is a valuable tool to further study candidate proteins in human hepatocytes. The similarity in gene expression profiles of normal hepatocytes and multidrug‐resistant cells is striking. Once a neoplasm develops, the particular gene expression pattern renders it well equipped to resist chemotherapeutic treatments. ABC transporters again play key roles in these mechanisms. Many of them, especially ABCB1, have been correlated with poor prognosis in cancers such as AML, NSCLC, and ovarian cancer. However, the majority of clinical trials evaluating ABCB1 inhibitors have



76:  The Dual Role of ABC Transporters in Drug Metabolism and Resistance to Chemotherapy

failed to reach a positive endpoint. Several questions have been raised and should be further investigated to understand the ­reasons for this failure. Similarly, the clinical relevance of the preclinical models has been questioned. This is an ongoing debate. Though most research indicates that cancer cell lines resemble primary tumors at the genomic level, at the transcriptomic level this is not the case for many cancer types. However, to the best of our knowledge, this debate is not as intense with regard to hepatocellular carcinoma. Several studies have shown the clinical relevance of HCC cell lines at both the genomic and transcriptomic levels. We propose that this cancer type serves as a blueprint for studying intrinsic multidrug resistance. HCC cell lines grown in an adapted culture environment, close to in vivo parameters, may constitute a valuable tool in the drug development pipeline to assess strategies to increase cancer drug response. It is now clear that genetic polymorphisms within transporters and drug‐metabolizing enzymes are clinically relevant. Advances have been made in the understanding of the kinetic interplay between transporters, both SLCs and ABCs, and phase I and II metabolizing enzymes. The Pharmacogenomics Research Network plays an important role in this field and is leading the way toward personalized medicine. For further details, please see Chapter 49, and also Chapters 5 and 23.

REFERENCES 1. Perland, E., and Fredriksson, R. Classification systems of secondary active transporters. Trends Pharmacol Sci, 2017;38:305–15. 2. Rives, M.L., Javitch, J.A., and Wickenden, A.D. Potentiating SLC transporter activity: emerging drug discovery opportunities. Biochem Pharmacol, 2017;135:1–11. 3. Patel, M., Taskar, K.S., and Zamek‐Gliszczynski, M.J. Importance of hepatic transporters in clinical disposition of drugs and their metabolites. J Clin Pharmacol, 2016;56(Suppl 7):S23–39. 4. Nelson, D.R. The cytochrome p450 homepage. Hum Genomics, 2009;4:59–65. 5. Daly, A.K. Pharmacogenetics of the cytochromes P450. Curr Topics Med Chem, 2004;4:1733–44. 6. Coughtrie, M.W. Ontogeny of human conjugating enzymes. Drug Metab Lett, 2015;9:99–108. 7. Nakata, K., Tanaka, Y., Nakano, T. et  al. Nuclear receptor‐mediated transcriptional regulation in Phase I, II, and III xenobiotic metabolizing systems. Drug Metab Pharmacokinet, 2006;21:437–57. 8. Deeley, R.G., Westlake, C., and Cole, S.P. Transmembrane transport of endo‐ and xenobiotics by mammalian ATP‐binding cassette multidrug resistance proteins. Physiol Rev, 2006;86:849–99. 9. Pal, D., and Mitra, A.K. MDR‐ and CYP3A4‐mediated drug–drug interactions. J Neuroimmune Pharmacol, 2006;1:323–39. 10. Zhou, S., Yung Chan, S., Cher Goh, B. et al. Mechanism‐based inhibition of cytochrome P450 3A4 by therapeutic drugs. Clin Pharmacokinet, 2005;44:279–304. 11. Katoh, M., Nakajima, M., Yamazaki, H., and Yokoi, T. Inhibitory effects of CYP3A4 substrates and their metabolites on P‐glycoprotein‐mediated transport. Eur J Pharm Sci, 2001;12:505–13. 12. Catania, V.A., Sanchez Pozzi, E.J., Luquita, M.G. et  al. Co‐regulation of expression of phase II metabolizing enzymes and multidrug resistance‐associated protein 2. Ann Hepatol, 2004;3:11–17. 13. Jigorel, E., Le Vee, M., Boursier‐Neyret, C., Parmentier, Y., and Fardel, O. Differential regulation of sinusoidal and canalicular hepatic drug transporter expression by xenobiotics activating drug‐sensing receptors in primary human hepatocytes. Drug Metab Dispos, 2006;34:1756–63.

1013

14. Nies, A.T., Schwab, M., and Keppler, D. Interplay of conjugating enzymes with OATP uptake transporters and ABCC/MRP efflux pumps in the elimination of drugs. Expert Opin Drug Metab Toxicol, 2008;4:545–68. 15. Robey, R.W., Pluchino, K.M., Hall, M.D. et  al. Revisiting the role of ABC transporters in multidrug‐resistant cancer. Nat Rev Cancer, 2018;18, 452–64. 16. Suchy, F.J., and Ananthanarayanan, M. Bile salt excretory pump: biology and pathobiology. J Pediatr Gastroenterol Nutr, 2006;43(Suppl 1):S10–16. 17. Stieger, B., Meier, Y., and Meier, P. J. The bile salt export pump. Pflugers Arch, 2007;453:611–20. 18. Oude Elferink, R.P., and Paulusma, C.C. Function and pathophysiological importance of ABCB4 (MDR3 P‐glycoprotein). Pflugers Arch, 2007;453:601–10. 19. Nies, A.T., and Keppler, D. The apical conjugate efflux pump ABCC2 (MRP2). Pflugers Arch, 2007;453:643–59. 20. Wang, J., Sun, F., Zhang, D.W. et al. Sterol transfer by ABCG5 and ABCG8: in vitro assay and reconstitution. J Biol Chem, 2006;281:27894–904. 21. Wang, J., Zhang, D.W., Lei, Y. et al. Purification and reconstitution of sterol transfer by native mouse ABCG5 and ABCG8. Biochemistry, 2008;47:5194–204. 22. Durmus, S., Hendrikx, J.J., and Schinkel, A.H. Apical ABC transporters and cancer chemotherapeutic drug disposition. Adv Cancer Res, 2015;125:1–41. 23. Guminski, A.D., Balleine, R.L., Chiew, Y.E. et al. MRP2 (ABCC2) and cisplatin sensitivity in hepatocytes and human ovarian carcinoma. Gynecol Oncol, 2006;100:239–46. 24. Materna, V., Stege, A., Surowiak, P., Priebsch, A., and Lage, H. RNA interference‐triggered reversal of ABCC2‐dependent cisplatin resistance in human cancer cells. Biochem Biophys Res Commun, 2006;348:153–7. 25. Roelofsen, H., Muller, M., and Jansen, P. L. Regulation of organic anion transport in the liver. Yale J Biol Medicine, 1997;70:435–45. 26. Flens, M.J., Zaman, G.J., van der Valk, P.G.L., et al. Tissue distribution of the multidrug resistance protein. Am J Pathol, 1996;148:1237–47. 27. Chen, Z.S., Guo, Y., Belinsky, M.G., Kotova, E., and Kruh, G.D. Transport of bile acids, sulfated steroids, estradiol 17‐beta‐D‐glucuronide, and leukotriene C4 by human multidrug resistance protein 8 (ABCC11). Mol Pharmacol, 2005;67:545–57. 28. Bera, T.K., Lee, S., Salvatore, G., Lee, B., and Pastan, I. MRP8, a new ­member of ABC transporter superfamily, identified by EST database mining and gene prediction program, is highly expressed in breast cancer. Mol Med, 2001;7:509–16. 29. Yabuuchi, H., Shimizu, H., Takayanagi, S., and Ishikawa, T. Multiple splicing variants of two new human ATP‐binding cassette transporters, ABCC11 and ABCC12. Biochem Biophys Res Commun, 2001;288:933–9. 30. Zelcer, N., Reid, G., Wielinga, P. et al. Steroid and bile acid conjugates are substrates of human multidrug‐resistance protein (MRP) 4 (ATP‐binding cassette C4). Biochem J, 2003;371:361–7. 31. Szakacs, G., Paterson, J.K., Ludwig, J.A., Booth‐Genthe, C., and Gottesman, M.M. Targeting multidrug resistance in cancer. Nat Rev Drug Discov, 2006;5:219–34. 32. Christoforou, A., Mulvey, C. M., Breckels, L.M. et al. A draft map of the mouse pluripotent stem cell spatial proteome. Nat Commun, 2016;7:8992. 33. Foster, L.J., de Hoog, C.L., Zhang, Y. et al. A mammalian organelle map by protein correlation profiling. Cell, 2006;125:187–99. 34. Itzhak, D.N., Tyanova, S., Cox, J., and Borner, G.H. Global, quantitative and dynamic mapping of protein subcellular localization. Elife, 2016;5. 35. Della Valle, M.C., Sleat, D.E., Zheng, H. et al. Classification of subcellular location by comparative proteomic analysis of native and density‐shifted lysosomes. Mol Cell Proteomics, 2011;10:M110.006403. 36. Jadot, M., Boonen, M., Thirion, J. et al. Accounting for protein subcellular localization: a compartmental map of the rat liver proteome. Mol Cell Proteomics, 2017;16:194–212. 37. Gai, J., Ji, M., Shi, C. et al. FoxO regulates expression of ABCA6, an intracellular ATP‐binding‐cassette transporter responsive to cholesterol. Int J Biochem Cell Biol, 2013;45:2651–9. 38. Ile, K.E., Davis, W., Jr., Boyd, J.T. et al. Identification of a novel first exon of the human ABCA2 transporter gene encoding a unique N‐terminus. Biochim Biophys Acta, 2004;1678:22–32. 39. Vulevic, B., Chen, Z., Boyd, J.T. et  al. Cloning and characterization of human adenosine 5′‐triphosphate‐binding cassette, sub‐family A, transporter 2 (ABCA2). Cancer Res, 2001;61:3339–47. 40. Zhou, C., Zhao, L., Inagaki, N. et  al. ATP‐binding cassette transporter ABC2/ABCA2 in the rat brain: a novel mammalian lysosome‐associated membrane protein and a specific marker for oligodendrocytes but not for myelin sheaths. J Neurosci, 2001;21:849–57.

1014

THE LIVER:  REFERENCES

41. Chapuy, B., Koch, R., Radunski, U. et al. Intracellular ABC transporter A3 confers multidrug resistance in leukemia cells by lysosomal drug sequestration. Leukemia, 2008;22:1576–86. 42. Efferth, T., Gillet, J.P., Sauerbrey, A. et al. Expression profiling of ATP‐binding cassette transporters in childhood T‐cell acute lymphoblastic leukemia. Mol Cancer Ther, 2006;5:1986–94. 43. Steinbach, D., Gillet, J.P., Sauerbrey, A. et al. ABCA3 as a possible cause of drug resistance in childhood acute myeloid leukemia. Clin Cancer Res, 2006;12:4357–63. 44. Henne, W.M. Organelle remodeling at membrane contact sites. J Struct Biol, 2016;196:15–19. 45. Pfisterer, S.G., Peranen, J., and Ikonen, E. LDL‐cholesterol transport to the endoplasmic reticulum: current concepts. Curr Opin Lipidol, 2016;27:282–7. 46. Akiyama, M., Sugiyama‐Nakagiri, Y., Sakai, K. et  al. Mutations in lipid transporter ABCA12 in harlequin ichthyosis and functional recovery by ­corrective gene transfer. J Clin Invest, 2005;115:1777–84. 47. Sakai, K., Akiyama, M., Sugiyama‐Nakagiri, Y. et  al. Localization of ABCA12 from Golgi apparatus to lamellar granules in human upper epidermal keratinocytes. Exp Dermatol, 2007;16:920–6. 48. Hashimoto, K. and Khan, S. Harlequin fetus with abnormal lamellar granules and giant mitochondria. J Cutan Pathol, 1992;19:247–52. 49. Ueda, K., Clark, D.P., Chen, C.J. et al. The human multidrug resistance (mdr1) gene. cDNA cloning and transcription initiation. J Biol Chem, 1987;262:505–8. 50. Gottesman, M.M., Lavi, O., Hall, M.D., and Gillet, J.P. Toward a better understanding of the complexity of cancer drug resistance. Annu Rev Pharmacol Toxicol, 2016;56:85–102. 51. Gillet, J.P., Efferth, T., and Remacle, J. Chemotherapy‐induced resistance by ATP‐ binding cassette transporter genes. Biochim Biophys Acta, 2007;1775:237–62. 52. Shaffer, B.C., Gillet, J.P., Patel, C. et al. Drug resistance: still a daunting challenge to the successful treatment of AML. Drug Resist Updat, 2012;15:62–9. 53. Patel, C., Stenke, L., Varma, S. et al. Multidrug resistance in relapsed acute myeloid leukemia: evidence of biological heterogeneity. Cancer, 2013;119: 3076–83. 54. Fletcher, J.I., Williams, R.T., Henderson, M.J., Norris, M.D., and Haber, M. ABC transporters as mediators of drug resistance and contributors to cancer cell biology. Drug Resist Updat, 2016;26:1–9. 55. Vitale, M., Rezzani, R., Rodella, L. et al. HLA class I antigen and transporter associated with antigen processing (TAP1 and TAP2) down‐regulation in high‐grade primary breast carcinoma lesions. Cancer Res, 1998;58:737–42. 56. Nelson‐Rees, W.A. The identification and monitoring of cell line specificity. Prog Clin Biol Res, 1978;26:25–79. 57. Gillet, J.P., Varma, S., and Gottesman, M.M. The clinical relevance of cancer cell lines. J Natl Cancer Inst, 2013;105:452–8. 58. Gillet, J.P., Calcagno, A.M., Varma, S. et  al. Redefining the relevance of established cancer cell lines to the study of mechanisms of clinical anti‐­ cancer drug resistance. Proc Natl Acad Sci U S A, 2011;108:18708–13. 59. Lukk, M., Kapushesky, M., Nikkila, J. et al. A global map of human gene expression. Nat Biotechnol, 2010;28:322–4. 60. Barretina, J., Caponigro, G., Stransky, N. et  al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature, 2012;483:603–7. 61. Neve, R.M., Chin, K., Fridlyand, J. et al. A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer Cell, 2006;10:515–27.

62. Gillet, J.P., Andersen, J.B., Madigan, J.P. et al. A gene expression signature associated with overall survival in patients with hepatocellular carcinoma suggests a new treatment strategy. Mol Pharmacol, 2016;89:263–72. 63. Lamb, J., Crawford, E.D., Peck, D. et al. The Connectivity Map: using gene‐ expression signatures to connect small molecules, genes, and disease. Science, 2006;313:1929–35. 64. Giacomini, K.M., Yee, S.W., Mushiroda, T. et al. Genome‐wide association studies of drug response and toxicity: an opportunity for genome medicine. Nat Rev Drug Discov, 2017;16:1. 65. McLean, C., Wilson, A., and Kim, R.B. Impact of transporter polymorphisms on drug development: is it clinically significant? J Clin Pharmacol, 2016;56(Suppl 7):S40–58. 66. Ieiri, I., Takane, H., and Otsubo, K. The MDR1 (ABCB1) gene polymorphism and its clinical implications. Clin Pharmacokinetics, 2004;43: 553–76. 67. Cascorbi, I. Role of pharmacogenetics of ATP‐binding cassette transporters in the pharmacokinetics of drugs. Pharmacol Ther, 2006;112:457–73. 68. Choudhuri, S. and Klaassen, C.D. Structure, function, expression, genomic organization, and single nucleotide polymorphisms of human ABCB1 (MDR1), ABCC (MRP), and ABCG2 (BCRP) efflux transporters. Int J Toxicol, 2006;25:231–59. 69. Hoffmeyer, S., Burk, O., von Richter, O. et al. Functional polymorphisms of the human multidrug‐resistance gene: multiple sequence variations and correlation of one allele with P‐glycoprotein expression and activity in vivo. Proc Natl Acad Sci U S A, 2000;97:3473–8. 70. Kim, R.B., Leake, B.F., Choo, E.F. et al. Identification of functionally variant MDR1 alleles among European Americans and African Americans. Clin Pharmacol Ther, 2001;70:189–99. 71. Gerloff, T., Schaefer, M., Johne, A. et al. MDR1 genotypes do not influence the absorption of a single oral dose of 1 mg digoxin in healthy white males. Br J Clin Pharmacol, 2002;54:610–16. 72. Nakamura, T., Sakaeda, T., Horinouchi, M. et  al. Effect of the mutation (C3435T) at exon 26 of the MDR1 gene on expression level of MDR1 messenger ribonucleic acid in duodenal enterocytes of healthy Japanese subjects. Clin Pharmacol Ther, 2002;71:297–303. 73. Siegmund, W., Ludwig, K., Giessmann, T. et al. The effects of the human MDR1 genotype on the expression of duodenal P‐glycoprotein and disposition of the probe drug talinolol. Clin Pharmacol Ther, 2002;72:572–83. 74. Wolking, S., Schaeffeler, E., Lerche, H., Schwab, M., and Nies, A.T. Impact of genetic polymorphisms of ABCB1 (MDR1, P‐Glycoprotein) on drug disposition and potential clinical implications: update of the literature. Clin Pharmacokinetics, 2015;54:709–35. 75. Kimchi‐Sarfaty, C., Oh, J.M., Kim, I.W. et al. A “silent” polymorphism in the MDR1 gene changes substrate specificity. Science, 2007;315:525–8. 76. Cole, S.P. Targeting multidrug resistance protein 1 (MRP1, ABCC1): past, present, and future. Annu Rev Pharmacol Toxicol, 2014;54:95–117. 77. Hira, D. and Terada, T. BCRP/ABCG2 and high‐alert medications: biochemical, pharmacokinetic, pharmacogenetic, and clinical implications. Biochem Pharmacol, 2018;147:201–10. 78. Relling, M.V., Krauss, R.M., Roden, D.M. et  al. New pharmacogenomics research network: an open community catalyzing research and translation in precision medicine. Clin Pharmacol Ther, 2017;102:897–902.

77

Stem Cell‐Derived Liver Cells: From Model System to Therapy Helmuth Gehart1 and Hans Clevers1,2 Oncode Institute, Hubrecht Institute, Royal Netherlands Academy of Arts and Sciences (KNAW) and University Medical Centre (UMC) Utrecht, Utrecht, The Netherlands 2 Princess Máxima Centre for Paediatric Oncology, Utrecht, The Netherlands 1

INTRODUCTION The increasing incidence of liver disease with organ replacement as the only therapeutic option has created a significant shortage of donor material. This scarcity imposes dangerously long wait­ ing times for patients and creates high demand for a very finite resource. Not only transplantation surgeons, but also the pharma­ ceutical industry and scientific community rely on primary liver tissue. Due to a lack of suitable alternatives, pharmacological testing and disease modeling is widely performed with primary hepatocytes. However, fast deterioration in culture, inability to proliferate in vitro, and striking batch to batch variability make primary hepatocytes a less than ideal assay platform. Recently, stem cell technology has stepped up to close the gap between supply and demand for liver cells. In particular, two stem cell sources have shown great potential to substitute for primary liver cells in vitro and eventually in vivo: induced pluripotent stem cells and adult tissue stem cells. In this chapter we discuss the individual advantages and disadvantages of each technology, highlight established applications, and examine the remaining hurdles for replacing primary liver tissue in drug development, disease modeling, and regenerative medicine.

SOURCES OF LIVER STEM CELLS ES/iPS cells Pluripotent stem cells (embryonic stem cells [ESCs] and induced pluripotent stem cells [iPSCs]) have the potential to generate all cell types of our body. Human ESCs are highly

expandable and can be guided to differentiate to a multitude of different tissues by mimicking fetal development in vitro. However, the use of human ESCs comes with significant ethical concerns. This ethical limitation has been overcome with the development of iPSCs: mature cells (e.g. fibroblasts) can regain pluripotency by forced expression of “Yamanaka factors” (Oct4, Sox2, Klf4, and c‐Myc) [1] (Figure  77.1a). While pluripotent cells have high proliferation rates and the potential to differenti­ ate into any cell type, they also suffer from limitations: complex and often inefficient differentiation processes, generation of cells with fetal characteristics, and limited genetic stability [2]. When reprogramming cells with the goal to generate hepato­ cytes, both the cell of origin and the reprogramming method have to be carefully chosen. The ability of iPSCs to differentiate to hepatocytes shows not only high donor to donor variability, but also dependence on the original cell type. For example, hepato­ cyte differentiation from peripheral blood cell‐derived iPSCs is more efficient than from dermal fibroblast‐derived lines [3]. This is most likely a result of “epigenetic memory,” where some char­ acteristics of the cell type of origin are insufficiently erased in the dedifferentiation process. Originally, retroviruses were used to permanently introduce Yamanaka factors in the course of repro­ gramming [1]. While this method is efficient, it is not compatible with regenerative medicine. Thus, several alternatives have been developed over the years: nonviral vectors [4], direct RNA or protein transfection [5], and use of nonintegrating viruses [6]. Once the iPSC line is established, it has to be differentiated towards the hepatocyte fate. Most protocols do so in a stepwise manner that mimics embryonic development of the liver (Figure 77.1b). The first step simulates induction of definitive endoderm and foregut endoderm by exposure of iPSCs to activin

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

1016

THE LIVER:  SOURCES OF LIVER STEM CELLS

Figure 77.1  Generation of hepatocytes from iPSCs and adult tissue stem cells. (a) Schematic procedure of iPSC culture establishment and generation of liver organoids from adult liver stem cells. (b) Schematic outline of the hepatocyte differentiation process for both technologies. PBMC, peripheral blood mononucleated cell; Dexa, dexamethason; DAPT, N‐[N‐(3,5‐Difluorophenacetyl)‐L‐alanyl]‐S‐phenylglycine‐t‐butyl ester, a Notch inhibitor.

A, bone morphogenic protein 4 (BMP4), and fibroblast growth factor (FGF) [7, 8]. Subsequently, cells need to be differentiated towards the liver fate. To this end, cells are stimulated with a combination of retinoic acid (or BMP4) and FGF10 [7]. Finally, the hepatoblast/hepatocyte fate is induced by hepatocyte growth factor (HGF) in combination with oncostatin M (OSM) and glu­ cocorticoid stimulation [7, 8]. In addition to the full dedifferentiation and redifferentiation protocol described earlier, a process has been described that ded­ ifferentiates the cell of origin (e.g. a fibroblast) only partially to a multi‐potent progenitor before induction of hepatic fate [9]. Even though this approach is slightly faster and safer, due to avoidance of pluripotency, the resulting induced hepatocytes (iHeps) are comparable to those generated with “conventional” protocols. While certain functions of adult hepatocytes, such as albumin production, reach comparable levels in induced hepatocytes

(iHeps), they still retain fetal markers, such as alpha‐fetoprotein (AFP) [10]. More importantly, some essential hepatocyte func­ tions, such as cytochrome (CYP) activity, lack inducibility and comparable function [11]. This issue is widely ascribed to the fact that iHeps resemble fetal or neonatal hepatocytes. In fact, several hepatocyte functions, such as CYP3A4 activity or biliru­ bin glucuronidation, are minimal at birth and only acquired later in adult life; supposedly in response to nutritional, oxygenation, or even bacterial colonization stimuli [11]. Thus, many current approaches to improve iHep maturity aim at providing these postnatal stimuli. Indeed, vitamin K, bacterial‐derived litho­ cholic acid, and oxygenation have been shown to promote meta­ bolic maturation in iHeps [8, 11]. Though these findings are important steps in the right direction, the main driver or combi­ nation of drivers that convert iHeps into bona fide adult hepato­ cytes has yet to be identified.



77:  Stem Cell‐Derived Liver Cells: From Model System to Therapy

Not only hepatocyte but also biliary cells have been gener­ ated from iPSCs. Since the biliary epithelium and hepatocytes originate from the same common progenitor (the hepatoblast), only the last steps of differentiation differ from the protocol described earlier. Instead of HGF and OSM, iPSCs at the hepa­ toblast stage are exposed to retinoic acid, activin A, and FGF10 [12]. Subsequently, cells are transferred into a 3D gel (Matrigel) and stimulated with epidermal growth factor (EGF). The result­ ing cells express markers of cholangiocytes and perform func­ tions such as bile acid transfer, alkaline phosphatase activity, and γ‐glutamyl‐transpeptidase activity [12]. The quality of liver tissue derived from pluripotent stem cells has continually increased over the last decade. Yet, some limita­ tions still remain: immaturity, high variability in differentiation efficiency of iPSC lines, and the significant investment of time and resources that is necessary to generate hepatocytes from a somatic cell. Nevertheless, the technology holds great promise for personalized medicine, drug discovery, and disease modeling.

Adult somatic stem cells Adult somatic stem cells are multi‐potent but fate restricted. Due to their fate restriction, these cells have to traverse a signifi­ cantly shorter “differentiation distance” to turn into mature, functional cells. Adult stem cell differentiation does not mimic fetal development, but rather regeneration of mature tissue. As a result, the protocols are significantly faster and generate cells with adult instead of fetal characteristics. Finally, in contrast to ES and iPS cells, adult somatic stem cells show high genetic stability in culture and eliminate the risk of teratoma formation due to their lack of pluripotency [13]. However, their fate restric­ tion implies that an adult somatic liver stem cell has to be directly isolated from a liver. So, their derivation is significantly more invasive when compared to iPSCs, which can be estab­ lished from a small piece of skin. For many protocols, a needle biopsy provides enough material to start an adult stem cell cul­ ture, but others may require a small resection. In contrast to ES/iPS cells, which are grown in 2D, epithelial adult tissue stem cells are mostly grown in 3D (in Matrigel or a similar gel matrix). There they form so‐called organoids, small structures that consist of stem cells and their more differentiated progeny. Organoids mimic the cellular organization of their tissue of origin and can be repeatedly (often indefinitely) passaged and expanded [13, 14]. They grow in complex, but defined media that mimic the regenerative state of the individual tissue. Commonly, these media stimulate Wnt signaling, inhibit differentiation sig­ nals by blocking TGFβ, and drive proliferation with EGF [13, 14]. In contrast to iPSCs, which will produce both epithelium and mesenchyme upon differentiation, adult stem cell‐derived orga­ noids are exclusively epithelial. In fact, many of the components in organoid growth medium serve to mimic the mesenchymal contribution to the signaling environment. During expansion, organoids contain a high percentage of stem cells that undergo rapid proliferation similar to the native regenerative response of the respective tissue. To generate tissue that represents the organ in its functional resting state, differentiation is usually needed. By withdrawing Wnt signals, releasing the block on TGFβ, and pro­ viding tissue‐specific stimuli (e.g. block or stimulation of Notch signaling) proliferating organoids can be turned into functional

1017

differentiated structures that are highly enriched for mature cells and mimic adult tissue in structure and function. Human liver organoids are derived from small wedge‐ or nee­ dle biopsies (Figure 77.1a). Organoids arise from Epcam+ (thus biliary) cells with surprising efficiency. Roughly one‐third of all Epcam+ cells will be able to form an organoid [13]. This high rate of organoid formation indicates that liver organoid culture does not exclusively expand a rare population of stem cells, but rather utilizes the inherent potential of almost all intrahepatic biliary cells to return to a bi‐potential progenitor or stem cell state. It is possible, however, that specific Epcam+ subpopula­ tions give rise to organoids with differing longevity and differen­ tiation potential. Organoid formation is induced by a combination of Wnt stimulation, TGFβ inhibition, FGF signaling, and strong activation of cAMP signal cascades [13]. In addition to soluble stimuli a suitable 3D matrix, such as Matrigel, is essential for successful expansion of the stem cell culture. In expansion con­ ditions, the growing organoids show little hepatocyte or cholan­ giocyte function, which makes differentiation to either fate necessary. For hepatocyte fate specification Wnt stimulation is withdrawn and Notch signals are blocked. At the same time, glu­ cocorticoid and BMP signaling are stimulated and FGF10 is switched to FGF19 [13] (Figure 77.1b). This change in growth conditions induces gradual differentiation of the organoid cul­ ture within 11 days. During this period, the culture stops prolif­ erating and acquires hepatocyte functions such as albumin secretion, cytochrome activity, and bile acid production [13]. The resulting hepatocytes are mature in function (e.g. Cyp3A4 activity) and lack fetal markers, such as AFP or fetal cytochromes. Whereas the maturity of hepatocytes in organoids is very good, differentiation of organoids is not uniform. The differentiation protocol enriches for hepatocytes, but also cholangiocytes and some remaining progenitors can be found dispersed between hepatocytes. Additional improvements to the differentiation pro­ tocol that promote uniformity of differentiation would further increase the suitability of the system for cell‐type specific assays. Similar to ES/iPS cells and even primary hepatocytes, the meta­ bolic activity of differentiated organoid cultures varies from donor to donor. Whether this variability reflects metabolic diver­ sity among individual humans remains to be shown. Since adult liver stem/progenitor cells are bi‐potent, they can also be differentiated to the biliary fate. This process requires a far less complex medium than hepatocyte differentiation, prob­ ably due to the innate tendency of biliary‐derived organoids to form cholangiocytes. By withdrawing Wnt and cAMP signals, organoids increase their expression of biliary markers, such as CK7 and CK19. This approach has been successfully used to model Alagille syndrome in vitro [13]. Adult stem cell‐derived liver organoids have the potential to excel at a wide variety of applications. The technique combines the advantages of high expansion potential, high genetic stabil­ ity, and generation of functionally mature cells. However, the method is still limited by the need for tissue biopsies for culture establishment and hepatocyte differentiation protocols are not yet as efficient as desired. Due to the recent development of the technology, the first studies to utilize the method are only begin­ ning to appear [13, 15]. Whereas these reports are promising, a wider body of literature will be needed to showcase the potential of the method and the range of its applications.

1018

THE LIVER:  IN VITRO APPLICATIONS OF STEM CELL‐DERIVED LIVER CELLS

IN VITRO APPLICATIONS OF STEM CELL‐ DERIVED LIVER CELLS Primary human hepatocytes are the current gold standard for liver in vitro assays. However, their quality varies highly. This problem is further exacerbated by the lack of hepatocyte prolif­ eration in vitro and the low metabolic stability of primary cells in prolonged assays. Replacement of primary cells with cell lines overcomes limitations of proliferation and stability. However, the metabolism of predominantly tumor‐derived cell lines does not reflect that of primary tissue [16]. Likewise, animals could provide a steady source of tissue, but due to species differences animal testing flags less than 50% of drugs that cause drug‐ induced liver injury (DILI) in humans [17]. Thus, an expandable human system with high culture stability, reproducibility, and a metabolic profile that matches primary hepatocytes would com­ bine the advantages of the currently employed techniques, while avoiding their pitfalls. Both ES/iPS‐ and adult stem cell‐derived hepatocytes do fit these requirements. Indeed, several proof‐of‐ principle studies showcase the use of these technologies for dis­ ease modeling and drug safety testing [13, 15, 18–26].

Pharmacological safety testing Pharmaceutical compounds and their metabolites can directly damage liver tissue. DILI is among the most common causes for liver injury and has led to a significant amount of post‐market drug withdrawals. Thus, exclusion of DILI is an important mile­ stone in drug development. In the past, the fetal nature of ES/ iPS‐derived hepatocytes resulted in rather poor correlation with primary hepatocytes in drug toxicity assays [23]. However, recent improvements to iHep maturity have increased both sensi­ tivity and classification accuracy significantly. By addition of vitamin K and lithocholic acid Avior et al. showed correct clas­ sification of 12 pharmacological compounds and a significant better correlation of TC50 values of iHeps and primary hepato­ cytes (0.94) than HepG2 and primary hepatocytes (0.62) [11]. Berger et  al. were able to increase significantly the functional maturity and longevity of iHeps by co‐culturing and micro‐pat­ terning iHeps and 3T3‐J2 cells, which results in an increased ability to identify DILI [25]. While short incubations (24 h) are preferred in experiments with primary hepatocytes, because of their rapid loss of functionality, the use of stem cell‐derived hepatocytes allows for much longer incubation times. This not only improves assay sensitivity, but also enables evaluation of chronic drug exposure [21, 23]. While the previously discussed assays focus on intrinsic toxicity of pharmacological compounds, stem cell‐derived hepatocytes can also be used to study patient‐ specific (idiosyncratic) reactions. Idiosyncratic DILI happens in a minority of patients who carry mutations in metabolic enzymes or suffer from a specific genetic disease. Idiosyncratic drug responses cannot be readily tested in primary hepatocytes, since liver tissue from a specific patient group (e.g. poor CYP2D6 metabolizers) is not available in the necessary quantities. This limitation can be overcome by establishing living biobanks of stem cells that represent the spectrum of genetic variation in the human population. Several recent studies have aimed to demon­ strate that the metabolic characteristics of an individual are

maintained in stem cell‐derived hepatocytes. Takayama et  al. showed that metabolic capacity and drug responsiveness corre­ late between iHeps and primary hepatocytes that served as ­starting material for the iPSC culture [24]. Patients with Alpers– Huttenlocher syndrome, a mitochondrial disease, are specifically sensitive to DILI by valproic acid. This effect could be success­ fully modeled in vitro in iPSC‐derived hepatocytes [22]. In the same vein, it was demonstrated that iHeps derived from α1‐­antitrypsin (A1AT)‐deficient patients are more susceptible to acetaminophen, which reflects clinical reality [26]. These studies clearly demonstrate that stem cell‐derived hepat­ ocytes have a wide range of applications in studying DILI. However, despite these encouraging results, certain differences in metabolism remain (e.g. lack of glucuronidation of acetaminophen in ES/iPS cell‐derived hepatocytes [27]) that warrant further improvement of differentiation methods. In this context adult stem cell‐derived hepatocytes could bring us, owing to their mature nature, a significant step forward. However, due to the recent development of the technology, studies using adult stem cell‐ derived hepatocytes for DILI prediction are not yet available.

Disease modeling Stem cell‐derived hepatocytes and biliary cells will not only find application in safety testing but may soon play crucial roles in the early phases of drug discovery. Our ability to screen pharma­ ceutical compounds for treatment of genetic liver diseases has been hampered for the longest time by the absence of suitable model systems. Many monogenetic defects, such as A1AT‐­ deficiency, Wilson disease, Crigler–Najjar syndrome, or Alagille syndrome have severe symptoms but low incidence rates in the general population. This makes the procurement of primary hepatocytes from these patients for drug development or research an almost impossible task. As a result, the pharmaceutical indus­ try and scientific community had to rely on cell lines or animal models, which have very limited ability to replicate the in vivo situation in humans. With the advent of stem cell technology, iPSC‐ or adult stem cell cultures can be established with high success rates from patients of all ages. This gives us the opportu­ nity to generate disease‐specific living biobanks that provide access to stem cell‐derived hepatocytes from a wide range of individual patients. This development will significantly facilitate research and drug development for rare diseases by overcoming the inherent limitation of available tissue. Since stem cell‐ derived hepatocytes reflect the phenotype and complete genetic makeup of the patient they were isolated from, the technology will enable us to study the often divergent phenotypic presenta­ tion in individuals with the same genetic disorder. A1AT‐deficiency (ATD) is caused by a mutation in the SERPINA1 gene. This mutation causes the protein to mis‐fold, which prevents it from being correctly secreted. Lack of A1AT in circulation results in indiscriminate neutrophil elastase activ­ ity throughout the patient’s body, which causes severe tissue damage. Additionally, mutant A1AT can aggregate in granules within hepatocytes, which eventually leads to the death of the cell. Indeed, iHeps from ATD patients show the characteristic accumulation of A1AT after differentiation [18]. This model has also been successfully applied to drug discovery. In a proof‐of‐ concept study, Choi et  al. screened more than 3000 approved



77:  Stem Cell‐Derived Liver Cells: From Model System to Therapy

drugs in iPSC‐derived hepatocytes from ATD patients and iden­ tified five compounds that could reliably decrease intracellular A1AT levels in iHeps from four patients [18]. Likewise, ATD has been modeled in adult stem cell‐derived liver organoids. Differentiated hepatocyte organoids from ATD patients accu­ mulate A1AT in granules, secrete significantly less A1AT, and lose their ability to inhibit neutrophil elastase [13]. In addition to ATD, several other monogenetic hepatocyte dis­ eases have been successfully modeled in stem cell‐derived cells: copper export defects in iHeps from Wilson disease patients [28], glycogen and lipid accumulation in patients with glu­ cose‐6‐phosphate deficiency [29], and deficient LDL‐C uptake in a patient with familial hypercholesterolemia [30]. Not only hepatocyte, but also biliary defects can be recapitulated in vitro. Two studies demonstrated that polycystic liver disease and cystic fibrosis can be successfully modeled in iPSC‐derived cholangio­ cytes and rescued with verapamil or VX809, respectively [12, 31]. Conversely, Alagille syndrome has been modeled in adult stem cell‐derived biliary cells. Liver organoids of a patient with Alagille syndrome, an inherent defect in Notch‐signaling that causes biliary paucity, displayed defects in cell–cell contact and reduced survival after differentiation to cholangiocytes [13]. Stem cell‐derived liver cells are not only useful in modeling genetic defects, but also find application in studying infectious dis­ eases. As multiple studies have shown, iHeps support entry and reproduction of both hepatitis B and hepatitis C virus particles [20, 32]. When infected, iHeps mount an innate interferon response and react to antiviral drugs. Likewise, iPSC‐derived hepatocytes were successfully infected with malaria parasites. Whereas, P. berghei, P. yoelii, and P. vivax could infect cells already at the hepatoblast stage, further maturation of iHeps was necessary to become susceptible to P. falciparum [19]. Additionally, plasmo­ dium infection could be treated in iHeps with primaquine, a pro‐ drug, whose activation by hepatic CYP2D6 is necessary for its action. The use of stem cell‐derived liver cells for the study of hepatotropic liver infections gives us the unique opportunity to study host–pathogen interaction in a multitude of different genetic host backgrounds. Thus, future drugs can be tailored to pathogen and host alike.

Cancer research A third major area of application for liver stem cell technology lies in the study of liver cancer. In the past, cell lines have proven inadequate to represent the wide spectrum of malignancies that arise in the liver. Thus, there is a need for in vitro systems that can propagate primary liver tumors in a state that mimics the in vivo situation. The same growth conditions that support the gen­ eration of liver organoids from healthy adult liver stem cells can also be used to grow liver tumoroids from small biopsies of malignant tissue [15]. Similar systems have already been suc­ cessfully established for colon, pancreatic, breast, prostate, and bladder cancer [33]. Broutier et al. have established tumor orga­ noids from hepatocellular carcinoma (HCC), cholangiocarci­ noma (CC), and combined HCC/CC (CHC) tumors. They showed that tumoroids maintain the characteristics of the origi­ nal tumor including mutational profile, marker expression, and transcriptional profile in culture. Thus, tumoroids, like normal liver organoids, are suitable for drug screens to tailor cancer

1019

therapy to the need of the individual patient. In a proof‐of‐­ principle screen, the authors identified ERK as a new therapeu­ tic target for a subset of liver cancers [15]. This innovation opens the door for generation of liver tumor biobanks that will significantly improve our understanding of common events in the wide spectrum of liver malignancies.

STEM CELL‐DERIVED LIVER CELLS IN REGENERATIVE MEDICINE Regenerative medicine has always been a key aspect in the development of stem cell‐derived liver tissue. The promise of eliminating waiting lists for allogeneic organ transplantation or even the possibility of generating autologous, in vitro grown tis­ sue for specific patients is a strong driving force in the field. In contrast to in vitro applications where maturity and functional­ ity are the most important aspects, safety, genetic stability, and the ability to engraft damaged liver tissue are the main criteria for regenerative efforts. This means that the most common method of generating iPSCs (viral stable integration of “Yama­ naka factors”) cannot be employed. Fortunately, alternative methods have been developed, but their efficiency is compara­ bly low [3–5]. Furthermore, the inherent genetic instability dur­ ing the reprogramming step necessitates thorough screening of derived iPSC lines for chromosomal aberrations and potentially oncogenic mutations [2, 34]. Adult stem cell‐derived cultures do not need ectopic introduction of genetic elements and display high genetic stability [13]. This feature gives them an inherent advantage for regenerative applications, but their utility in regenerative medicine will depend on their ability to sufficiently repopulate livers. The ultimate goal of stem cell‐based liver therapy is autolo­ gous transplantation of healthy liver tissue. Treating patients with their own cells abolishes the need for immunosuppression and should reduce dramatically the chance of graft rejection. For certain diseases with environmental causes, isolation, exp­ ansion, and re‐transplantation of liver cells could be sufficient. For others, especially genetic disorders, the correct function of the cell has first to be restored in vitro. Both iPSC and adult stem cell technologies have been shown to be fully compatible with genome engineering using CRISPR/Cas9 or TALENs [18, 35]. Thus, both approaches present valid options to combine autolo­ gous transplantation with genome editing. A multitude of different transplantation models, culturing methods, and engraftment readouts makes the comparison of transplantation success among different stem cell systems sur­ prisingly difficult. Nevertheless, it is safe to say that to this day no stem cell system outperforms primary mature hepatocytes in terms of engraftment and repopulation ability. However, some systems have come close. Zhu et al. demonstrated extensive proliferation of iHeps from partially dedifferentiated human fibroblasts after transplantation into immunosuppressed fumarylacetoacetate hydrolase (FAH)‐ deficient animals [9]. These mice accumulate a toxic metabo­ lite, which induces severe liver damage when not treated with 2‐(2‐nitro‐4‐trifluoro‐methylbenzoyl)1,3‐cyclohexedione (NTBC). Thus, liver damage and selective growth advantage for

1020

THE LIVER:  REFERENCES

the graft can be induced by NTBC withdrawal. After 9 months, 2% of the liver had been replaced by transplanted cells, which was a significant improvement over previous studies. However, it has to be noted that adult human hepatocytes can reach engraftment levels of more than 90% in the same model of chronic liver damage [36]. Carpentier et  al. reported 15% engraftment in immunodeficient Mup‐uPA mice 100 days after transplantation of human iHeps [37]. The recipient mice trans­ genically express urinary plasminogen activator (uPA) under the major urinary protein (Mup) promoter, which causes slow con­ tinuous necrosis of host hepatocytes to give the transplant a selective advantage. Interestingly, while iHeps showed clear signs of immaturity at the time of transplantation, significant maturation of the graft occurred in situ [37]. More recently, Yang et  al. reported 5% engraftment of iHeps derived from patients with familial hypercholesterolemia in gamma‐irradi­ ated livers of immunosuppressed mice after 21 days [38]. This level of engraftment was sufficient to test the effects of simvas­ tatin and proprotein convertase subtilisin/kexin type 9 (PCSK9) antibodies on LDL‐C clearance by human liver cells in vivo. Not only iPSC‐derived iHeps but also hepatocytes generated from adult stem cells have been successfully transplanted in ani­ mals. Using a CCl4‐based chemical liver damage model, Huch et al. showed stable engraftment for more than 100 days [13]. However, because of the acute nature of the damage model, the proliferative capacity of the graft has not been assessed. More transplantation studies in standardized damage models will be needed to compare the performance of adult stem cell‐derived hepatocytes directly with the longer established and therefore significantly larger body of iHep literature. Finally, also stem cell‐derived biliary cells have been recently successfully transplanted. Sampaziotis et al. have shown that chol­ angiocyte organoids, grown from adult stem cells of the extrahe­ patic bile duct, are able to engraft and repair damaged gall bladders [39]. This opens a very promising new cell source for surgical treatment of common bile duct disorders such as biliary atresia.

FUTURE CHALLENGES Stem cell‐derived liver cells have come a long way from the first differentiation attempts with ESCs to today’s highly refined expansion and differentiation protocols for iPSCs and adult stem cells. These improvements have already made stem cell‐ derived liver cells an attractive option for drug safety testing, disease modeling, and drug development. Further improve­ ments in maturity of iHeps made from iPSCs and more efficient differentiation protocols for hepatocytes derived from adult stem cells could render the in vitro use of primary hepatocytes in the near future obsolete. To enter clinical practice, both tech­ nologies have to overcome several hurdles: Safety concerns of iPSC‐iHep transplantation have to be addressed and engraft­ ment needs still to be optimized. While safety is less of a con­ cern for adult stem cell‐based technologies, the expansion of the initial organoid graft to therapeutically relevant levels has yet to be demonstrated. Inevitably, both technologies will have to prove their efficacy in short‐ and long‐term clinical trials to finally replace or reduce the need for whole organ transplanta­ tion and abolish waiting lists once and for all.

REFERENCES 1. Takahashi, K. and Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell, 2006;126(4):663–76. 2. Laurent, L.C., Ulitsky, I., Slavin, I. et al. Dynamic changes in the copy num­ ber of pluripotency and cell proliferation genes in human ESCs and iPSCs during reprogramming and time in culture. Cell Stem Cell, 2011;8(1): 106–18. 3. Kajiwara, M., Aoi, T., Okita, K. et al. Donor‐dependent variations in hepatic differentiation from human‐induced pluripotent stem cells. Proc Natl Acad Sci U S A, 2012;109(31):12538–43. 4. Yu, J., Hu, K., Smuga‐Otto, K. et al. Human induced pluripotent stem cells free of vector and transgene sequences. Science, 2009;324(5928):797–801. 5. Plews, J.R., Li, J., Jones, M. et al. Activation of pluripotency genes in human fibroblast cells by a novel mRNA based approach. PloS One, 2010;5(12): e14397. 6. Nishimura, K., Sano, M., Ohtaka, M. et al. Development of defective and persistent Sendai virus vector: a unique gene delivery/expression system ideal for cell reprogramming. J Biol Chem, 2011;286(6):4760–71. 7. Touboul, T., Hannan, N.R., Corbineau, S. et  al. Generation of functional hepatocytes from human embryonic stem cells under chemically defined conditions that recapitulate liver development. Hepatology, 2010;51(5): 1754–65. 8. Si‐Tayeb, K., Noto, F.K., Nagaoka, M. et al. Highly efficient generation of human hepatocyte‐like cells from induced pluripotent stem cells. Hepatology, 2010;51(1):297–305. 9. Zhu, S., Rezvani, M., Harbell, J. et al. Mouse liver repopulation with hepato­ cytes generated from human fibroblasts. Nature, 2014;508(7494):93–7. 10. Baxter, M., Withey, S., Harrison, S. et al. Phenotypic and functional analyses show stem cell‐derived hepatocyte‐like cells better mimic fetal rather than adult hepatocytes. J Hepatol, 2015;62(3):581–9. 11. Avior, Y., Levy, G., Zimerman, M. et al. Microbial‐derived lithocholic acid and vitamin K2 drive the metabolic maturation of pluripotent stem cells‐ derived and fetal hepatocytes. Hepatology, 2015;62(1):265–78. 12. Sampaziotis, F., de Brito, M.C., Madrigal, P. et al. Cholangiocytes derived from human induced pluripotent stem cells for disease modeling and drug validation. Nat Biotechnol, 2015;33(8):845–52. 13. Huch, M., Gehart, H., van Boxtel, R. et al. Long‐term culture of genome‐­ stable bipotent stem cells from adult human liver. Cell, 2015; 160(1‐2):299–312. 14. Sato, T., Stange, D.E., Ferrante, M. et al. Long‐term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett’s epi­ thelium. Gastroenterology, 2011;141(5):1762–72. 15. Broutier, L., Mastrogiovanni, G., Verstegen, M.M. et  al. Human primary liver cancer‐derived organoid cultures for disease modeling and drug screen­ ing. Nat Med, 2017;23(12):1424–35. 16. Wilkening, S., Stahl, F., and Bader A. Comparison of primary human hepato­ cytes and hepatoma cell line Hepg2 with regard to their biotransformation properties. Drug Metab Dispos, 2003;31(8):1035–42. 17. Olson, H., Betton, G., Robinson, D. et  al. Concordance of the toxicity of pharmaceuticals in humans and in animals. Regul Toxicol Pharmacol, 2000;32(1):56–67. 18. Choi, S.M., Kim, Y., Shim, J.S. et al. Efficient drug screening and gene cor­ rection for treating liver disease using patient‐specific stem cells. Hepatology, 2013;57(6):2458–68. 19. Ng, S., Schwartz, R.E., March, S. et al. Human iPSC‐derived hepatocyte‐like cells support Plasmodium liver‐stage infection in vitro. Stem Cell Rep, 2015;4(3):348–59. 20. Sa‐Ngiamsuntorn, K., Wongkajornsilp, A., Phanthong, P. et al. A robust model of natural hepatitis C infection using hepatocyte‐like cells derived from human induced pluripotent stem cells as a long‐term host. Virology J, 2016;13:59. 21. Holmgren, G., Sjogren, A.K., Barragan, I. et al. Long‐term chronic toxicity testing using human pluripotent stem cell‐derived hepatocytes. Drug Metab Dispos, 2014;42(9):1401–6. 22. Li, S., Guo, J., Ying, Z. et al. Valproic acid‐induced hepatotoxicity in Alpers syndrome is associated with mitochondrial permeability transition pore opening‐dependent apoptotic sensitivity in an induced pluripotent stem cell model. Hepatology, 2015;61(5):1730–9. 23. Szkolnicka, D., Farnworth, S.L., Lucendo‐Villarin, B. et al. Accurate predic­ tion of drug‐induced liver injury using stem cell‐derived populations. Stem Cells Transl Med, 2014;3(2):141–8.



77:  Stem Cell‐Derived Liver Cells: From Model System to Therapy

24. Takayama, K., Morisaki, Y., Kuno, S. et al. Prediction of interindividual dif­ ferences in hepatic functions and drug sensitivity by using human iPS‐ derived hepatocytes. Proc Natl Acad Sci U S A, 2014;111(47):16772–7. 25. Ware, B.R., Berger, D.R., and Khetani, S.R. Prediction of drug‐induced liver injury in micropatterned co‐cultures containing iPSC‐derived human hepato­ cytes. Toxicol Sci, 2015;145(2):252–62. 26. Wilson, A.A., Ying, L., Liesa, M. et  al. Emergence of a stage‐dependent human liver disease signature with directed differentiation of alpha‐1 antit­ rypsin‐deficient iPS cells. Stem Cell Rep, 2015;4(5):873–85. 27. Sengupta, S., Johnson, B.P., Swanson, S.A. et  al. Aggregate culture of human embryonic stem cell‐derived hepatocytes in suspension are an improved in vitro model for drug metabolism and toxicity testing. Toxicol Sci, 2014;140(1):236–45. 28. Zhang, S., Chen, S., Li, W. et al. Rescue of ATP7B function in hepatocyte‐ like cells from Wilson’s disease induced pluripotent stem cells using gene therapy or the chaperone drug curcumin. Human Mol Genet, 2011; 20(16):3176–87. 29. Rashid, S.T., Corbineau, S., Hannan, N. et al. Modeling inherited metabolic disorders of the liver using human induced pluripotent stem cells. J Clin Invest, 2010;120(9):3127–36. 30. Cayo, M.A., Cai, J., DeLaForest, A. et al. JD induced pluripotent stem cell‐ derived hepatocytes faithfully recapitulate the pathophysiology of familial hypercholesterolemia. Hepatology, 2012;56(6):2163–71.

1021

31. Ogawa, M., Ogawa, S., Bear, C.E. et al. Directed differentiation of cholan­ giocytes from human pluripotent stem cells. Nat Biotechnol, 2015;33(8): 853–61. 32. Shlomai, A., Schwartz, R.E., Ramanan, V. et al. Modeling host interactions with hepatitis B virus using primary and induced pluripotent stem cell‐derived hepatocellular systems. Proc Natl Acad Sci U S A, 2014;111(33):12193–8. 33. Drost, J. and Clevers, H. Organoids in cancer research. Nat Rev Cancer, 2018;18(7):407–18. 34. Liang, G. and Zhang, Y. Genetic and epigenetic variations in iPSCs: potential causes and implications for application. Cell Stem Cell, 2013;13(2):149–59. 35. Schwank, G. and Clevers, H. CRISPR/Cas9‐mediated genome editing of mouse small intestinal organoids. Methods Mol Biol, 2016;1422:3–11. 36. Azuma, H., Paulk, N., Ranade, A. et al. Robust expansion of human hepato­ cytes in Fah−/−/Rag2−/−/Il2rg−/− mice. Nat Biotechnol, 2007;25(8):903–10. 37. Carpentier, A., Tesfaye, A., Chu, V. et al. Engrafted human stem cell‐derived hepatocytes establish an infectious HCV murine model. J Clin Invest, 2014;124(11):4953–64. 38. Yang, J., Wang, Y., Zhou, T. et al. Generation of human liver chimeric mice with hepatocytes from familial hypercholesterolemia induced pluripotent stem cells. Stem Cell Rep, 2017;8(3):605–18. 39. Sampaziotis, F., Justin, A.W., Tysoe, O.C. et al. Reconstruction of the mouse extrahepatic biliary tree using primary human extrahepatic cholangiocyte organoids. Nat Med, 2017;23(8):954–63.

78

Extracellular Vesicles and Exosomes: Biology and Pathobiology Gyongyi Szabo1 and Fatemeh Momen‐Heravi2 1 2

Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Department of Medicine, University of Massachusetts Medical School, Worcester, MA, USA

EXTRACELLULAR VESICLES AND EXOSOMES: DEFINITION AND BIOGENESIS Extracellular vesicles (EVs) are a heterogenous population of vesicles shed by almost all cell types that can be detected in cir­ culation and cellular microenvironment [1]. Extracellular vesi­ cles can be classified based on the mode of biogenesis and size to exosomes, microvesicles, and apoptotic bodies. Exosomes are the smallest (30–100 nm) and the most characterized subpopula­ tion of EVs. Exosomes are generated through the endosomal pathway which matures into late endosomes that produce intra­ cellular vesicles in multivesicular bodies (MVBs) [2]. Exosomes originate from activation of late‐endosomal pathways, and their release is controlled by RAB guanosine triphosphatases, which are involved with endosome recycling, vesicular trafficking, and plasma membrane fusion of the exosomes [3, 4]. Late‐endoso­ mal pathways are ubiquitin‐independent and are dependent on multivesicular bodies, which after maturation are directed to lys­ osomes for degradation or transported to the plasma membrane with which they fuse for sorting and releasing of exosomes [5]. In these pathways, ALG‐2‐interacting protein X (ALIX) binds to the exosomal cargoes, which differentiate between lysosomal recycling pathways and exosomal sorting pathways. Rab5 regu­ lates endocytic trafficking and cargo sequestration and RAB27a, RAB27b, and RAB35 play primary roles in the late‐endosome docking and fusion to the plasma membrane [4, 6]. Exosome and EV contents may include different molecules including onco­ genes, phosphoproteins, cytok­ines, growth factors, tumor sup­ pressors, microRNAs, mRNAs, and  DNA sequences that can modulate recipient cell response (Figure 78.1).

Microvesicles (also called shedding vesicles, shedding microvesicles, or microparticles) are approximately 100–1000 nm in diameter and originate from the outward budding of the plasma membrane. The Rho signaling pathway, actin motor ­proteins, GTP‐binding protein ARF6, and the compartments of cytoskeleton play roles in the formation of microvesicles [7, 8]. Stress responses, increase in intracellular Ca2+ and lipotoxicity in hepatocytes can induce microvesicle release; however, knowledge is limited on MV biogenesis [9]. Apoptotic vesicles are a subpopulation of EVs that range from 100–2000 nm in diameter and are generated by the bleb­ bing of the plasma membrane of cells undergoing apop­ tosis. Larger apoptotic vesicles (1000–5000 nm) are referred to as apoptotic bodies and contain fragmented nuclei as well as ­fragmented cytoplasmic organelles [10].

EVs AND EXOSOMES AS EMERGING BIOMARKERS IN LIVER DISEASES EVs exhibit significant potential to be used as biomarkers with clinical utility. As the stable repertoire of different biomolecules with disease‐specific signature, EVs reduce biological noise, and have the potential to serve as liquid biopsies in clinical set­ tings. EVs’ profile can exhibit a snapshot of cells in different disease stages and completely noninvasive. With the optimiza­ tion of isolation and characterization protocols, the functional and physiological roles of EVs in different disease models is being appreciated and studied. However, methodological and analytical challenges exist in translating EV‐based diagnosis and molecular targeting. The transition from bench to bed

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



78:  Extracellular Vesicles and Exosomes: Biology and Pathobiology

1023

cholangiocytes hepatocytes

stellate cells

Kupffer cells macrophages

sinusoidal endothelial cells

immune cells Extracellular Vesicles (EV):

EV Cargo • nucleic acids • proteins • lipids

Exosomes 50–100 (50–150 nm) Microvesicles 100–1,000 (150–1,000 nm)

Biomarkers

Communication

Therapeutic Applications

• EV cargo is characteristic of the cell type and activation state • microRNA, lncRNA • mRNA • proteins • viral or bacterial nucleic acid

• cell-to-cell communication • inter-organ communication • cancer metastases

• EV as a delivery vehicle • EV from stem cells

Figure 78.1  Exosome and EV contents may include different molecules including oncogenes, phosphoproteins, cytokines, growth factors, tumor suppressors, microRNAs, mRNAs, and DNA sequences that can modulate recipient cell response.

requires developing standardized isolation and characterization protocols, and validation in a well‐characterized patient popula­ tion. Several reports have shown alterations in the rate of pro­ duction and cargo of EVs and these characteristics position EVs as a new, promising class of biomarkers with strong diagnostic potential in the context of personalized medicine [11, 12]. EVs carry diverse molecular cargoes including nucleic acids, proteins, and lipids, which provide a snapshot of the parental cell at the time of secretion [13]. Due to the presence of the coat­ ing lipid bilayer, EV cargoes appear to be protected against degrading enzymes such as nucleases and proteases. In search of biomarkers, EVs/exosomes were found in various body flu­ ids, including plasma, serum, saliva, urine, and bile [13] (Table 78.1). The biomarker potential of EVs and exosomes is based on the observation that the EV cargo changes in response to cellular activation, stress, or disease state. Thus, investigators focus on the characterization of EV cargo with the goal of iden­ tifying unique signatures of disease‐specific liver pathology, prognostic factors, and/or indicators of response to therapy.

EVs IN CELL‐TO‐CELL AND INTER‐ORGAN COMMUNICATION IN THE LIVER EVs can act as biological carriers and function as important mediators of intercellular communications between the different types of liver cells [13]. This involves secretion of EVs by a par­ ent cell and uptake of EVs by target cells most likely through receptor‐mediated endocytosis or independent of membrane receptors [13, 14]. Increasing evidence suggests that EVs can influence the function of the recipient cell. In vivo, it was shown

that EVs isolated from wild‐type mice transfer microRNA‐155 in the exosome content to various organs in miR‐155‐deficient mice after an intravenous injection [15]. The majority of miR‐155 from the exosomes accumulated in the recipient mice in the liver and the EV‐packaged miR‐155 was detected in both hepatocytes and liver mononuclear cells indicating broad uptake of circulat­ ing EVs by different cell types [15]. Another study showed that EVs isolated from mice with alcoholic liver d­ isease, when trans­ ferred to naive mice, result in functional changes in the liver [16]. Specifically, EVs derived from mice with alcoholic liver disease induced recruitment of infiltrating M1‐type macrophages to the liver that was associated with increased expression of MCP‐1 in the recipient livers. Mechanistically, it was shown that Hsp90 enriched in the cargo of EVs from ALD mice could mediate the biologic effect via MCP‐1 induction in hepatocytes [16]. Exosomes and extracellular vesicles can be taken up by differ­ ent cell types. In vitro, hepatocyte‐derived EVs were shown to be taken up by macrophages and these EVs have a biological effect on the target cell. Hepatocyte‐derived exosomes were enriched in miR‐122, a hepatocyte‐specific microRNA and these hepato­ cyte‐derived exosomes effectively transferred miR‐122 to mac­ rophages that normally do not express this miRNA [17]. Importantly, we found that macrophages primed with exosomes derived from alcohol through hepatocytes increased TNFα pro­ duction in response to LPS stimulation and this was mediated via the exosome‐transferred hepatocyte‐derived miR‐122 [17]. Another study showed that alcohol also increases exosome production in monocytes/macrophages. Monocyte‐derived exosomes after monocyte treatment with alcohol were enriched in miR‐27a and had a functional effect when trans­ ferred to ­ alcohol‐naive monocytes. Specifically, exosomes from alcohol‐exposed monocytes can modulate the phenotype of naive monocytes to differentiate into M2‐like macrophages

1024

THE LIVER:  EVs IN LIVER DISEASES

Table 78.1  EVs as potential biomarkers in different liver diseases Type of liver disease

Species

Biofluid platform

Content of extracellular vesicles

Nucleic acid‐based biomarkers Early alcoholic steatohepatitis

Mouse

Plasma

Increased in miRNAs: let7f, miR‐29a, and miR‐340

Alcoholic hepatitis

Mouse

Increase in miR‐155, miR‐122

Alcoholic hepatitis

Human

Serum, plasma Serum

HCC HCC Liver fibrosis

Human Human Mouse

Serum Serum Serum

Liver fibrosis Alcoholic hepatitis

Human Mouse

Serum Serum

Acute liver injury (acetaminophen or thioacetamide‐induced liver injury) Hepatitis C

Rat

Serum

Human

Serum

Protein‐based biomarkers Liver injury

Human

Hepatitis C HCC and hepatitis C cirrhosis HCC

Human Human Human

Plasma, serum Serum Plasma Serum

HCC Acute liver injury (d‐galactosamine)

Human Rat

Serum Urine

Increased in miR‐122

Control group

Refs

Chronic liver injury, including bile duct ligation, non‐alcoholic steatohepatitis, obese mice Pair‐fed mice

[49] [50]

Healthy individuals, patients own baseline before binge alcohol consumption Healthy individuals Healthy individuals, patients with HBV Control mice

[35] [51] [52]

HCC Control mice

[40] [22]

Control rats

[53]

Increases in miRNAs: miR‐122, miR‐134, miR‐424, miR 629

Healthy individuals

[54]

Increase in sPTPRG

Subjects without liver injury

[55]

Increase in soluble CD81 Increase in Hep par 1 Increase in AnnexinV+ EpCAM+ CD147+ t SMAD3 Increase in CD26, SLC3A1, CD81, CD10

Healthy individuals Hepatitis C cirrhosis without HCC HCC, cirrhosis, healthy

[56] [57] [58]

Healthy individuals Control rats

[59] [60]

Decrease in miR‐718 Increase in miR‐21 Increase in Ccn2; decrease in Twist1, miR‐214 Increase in miR‐125a Increase in miRNAs: miR‐122, miR‐192, miR‐30a Increase in miRNAs: miRNA‐122a, miRNA192, miRNA193a

with increased expression of CD206, CD163, and IL‐10 and this effect is primarily mediated by the exosome‐transferred miR‐27a [18]. In vivo studies showed that EVs derived from the circulation of mice with ALD have a functional effect in naive recipient mice. Intravenous administration of ALD‐EVs to healthy mice resulted in increased expression of the monocyte recruiting cytokine, MCP‐1, in hepatocytes. Furthermore, mice that received the ALD‐EVs had an increase in liver macrophages that expressed markers of the pro‐inflammatory, M1, phenotype and a decrease in the M2‐type macrophages [16]. These results indicate that circulating EVs in ALD have a functional role in maintaining inflammation and recruitment of macrophages to the liver.

EVs IN LIVER DISEASES EVs in viral hepatitis In hepatitis C virus (HCV) infection, HCV viral RNA was found in exosomes isolated from patients with chronic HCV infection and infected hepatocytes were shown to release exosomes that contain viral RNA [19]. It has been shown that this HCV RNA in the exosomes is replication competent and exosomes can transmit HCV infection to naive noninfected hepatocytes. The HCV RNA was found to be in a replication ready complex with miR‐122, a host factor in HCV replication, stabilized by Hsp90

[17]

in the exosome cargo [19]. Interestingly, exosomes were also found to mediate cell‐to‐cell transmission of IFN‐α‐induced viral immunity [20]. In HBV infection, exosomes isolated from HBV‐infected patients contained viral components and induced HBV infection in naive hepatocytes [21]. This study also found that HBV exosomes modulate functions of NK cells thereby undermining host antiviral responses [21].

EVs in alcoholic and non‐alcoholic fatty liver disease Both animal models of alcoholic liver disease and human serum from patients with alcoholic liver disease showed an increase in the number of circulating extracellular vesicles compared to controls [22]. This increase in EVs was mostly represented by increased serum levels of exosomes (defined as 40–150 mm in this study) and minimal increase in the microvesicle fraction. The source of these EVs found in the circulation was mostly likely from different cell types in the liver and potentially from other organs. In vitro, exposure of primary hepatocytes, Kupffer cells, liver macrophages or liver mononuclear cells resulted in increased secretion of exosomes and microvesicles to cell supernatants [18, 22]. Most impor­ tantly, it has been shown that patients with alcoholic hepatitis also have increased numbers of circulating exosomes ­compared to normal healthy controls [17]. The cargo of exosomes in ALD was shown to contain specific miRNA signatures with  enrichment of miR‐199, miR‐30a, and miR‐122 in the



78:  Extracellular Vesicles and Exosomes: Biology and Pathobiology

exosomes both in the mouse model and in human alcoholic hepatitis [18, 22]. In alcoholic liver disease in mice, it was shown that the protein cargo of exosomes or EVs is also different compared to control exosomes. Proteomic analysis of EVs from mice with ALD revealed that proteins involved in cell movement, inflammation, were differently expressed in ALD EVs compared to control EVs. Furthermore, there was a cluster of uniquely expressed ­proteins in the ALD EVs that were not abundant in control EVs indicating that the EV cargo could be exploited as a potential biomarker of disease [16]. Another study found increased mito­ chondrial DNA content in circulating microparticles in patients with excessive drinking as well as in mice after acute‐on‐chronic alcohol administration [23]. This study also showed that the ethanol‐induced elevation in mitochondrial DNA‐enriched ­ microparticles originated from hepatocytes and  was linked to induction of hepatic neutrophil leukocyte infiltration. Studies in non‐alcoholic fatty liver disease in mice found a significant increase in EVs in the blood and liver represented by both microvesicles and exosomes [24]. They found enrichment of miR‐122 and miR‐192 in the blood. Hepatocyte‐derived EVs in nonalcoholic steatohepatitis (NASH) promote activation of endothelial cells (EC) via delivery of miR‐128‐3p [25]. EVs in NASH also contain mitochondrial DNA, which is implicated in macrophage activation [26]. In ANSHJ and the metabolic syn­ drome EVs can also participate in inter‐organ communication. It has been shown that adipocyte‐derived EVs can modulate metabolic dysregulation [27].

EVs in liver cholangiopathies and  drug‐induced liver injury Pioneering work from the LaRusso group showed the presence of exosomes in the bile and described regulatory mechanisms by which these exosomes influence cholangiocyte function [28]. Other studies showed that cholangiocyte‐derived exosomes are enriched in lncRNA H19, which promotes cholestatic liver injury both in mice and humans [29]. A recent report showed EVs and exosomes are also involved in drug‐induced liver injury (DILI). In vivo subtoxic acetaminophen (APAP) expo­ sure in mice resulted in elevations in circulating exosomal albu­ min mRNA [30]. In primary hepatocytes, APAP increased the miR‐122 content in exosomes without causing hepatocyte ­cytotoxicity [30]. Another study demonstrated that exosomes derived from mice with APAP‐induced liver injury promoted toxicity in recipient mice as well as in isolated hepatocytes [31]. These results indicate that in drug‐induced liver injury hepato­ cyte‐derived exosomes contribute to disease progression and liver injury.

EVs in liver fibrosis Investigations in animal models indicate that exosomes and EVs are also derived from liver endothelial cells (EC). Furthermore, the EC‐derived exosomes were internalized by hepatic stellate cells and induced their migration [32]. It was shown that this was sphingosine 1‐phosphate dependent and that SK1 levels were increased in serum exosomes of mice with experimental fibrosis. In another study, ethanol treatment induced EC‐derived EV pro­ duction and these EVs were enriched in miRNA-106b and the

1025

long non‐coding RNAs (lncRNA) HOTAIR, and MALAT1. The EC‐derived exosomes induced enhanced vascularization bioac­ tivity via these lncRNAs [33]. It has also been shown that serum EVs from normal mice, but not from mice with liver fibrosis, ameliorated liver fibrosis in a CCL4‐induced model of liver fibrosis [34]. Normal EVs decreased hepatic ­stellate cell activa­ tion and a specific cluster of microRNAs was present in the EV from formal but not from fibrotic mice.

EVs in liver cancer Different studies documented the role of EVs in the pathogene­ sis and metastasis of hepatocellular carcinoma (HCC) [35, 36]. EV‐mediated miRNA transfer was identified as a mechanistic contributor to the tumor spread [37]. EVs derived from HCC modulate transforming growth factor‐β‐activated kinase‐1 expression and promote anchorage‐independent growth in the recipient cells [38]. HCC‐derived EVs that harbor heat shock proteins were able to increase the antitumor effect of natural killer (NK) cells and increase tumor immunogenicity [39]. EVs originated from metastatic HCCs carried a substantial amount of pro‐oncogenic RNA and proteins, including MET protoonco­ gene, caveolins, and S100 proteins [40]. Interestingly, EVs orig­ inated from motile HCC cell lines could substantially promote the migratory and invasive abilities of nonmotile MIHA cells, a nontumorigenic immortalized human hepatocyte cell line. EVs isolated from HCCs modulated PI3K/AKT and MAPK signal­ ing pathways in MIHA and promoted secretion of active metal­ loproteinases [40].

CHALLENGES AND PROMISES IN TRANSLATIONAL EV BIOMARKER DISCOVERY Biomarker discovery for early detection, monitoring of disease progression, and evaluation of response to therapies is an active line of research in many disciplines including liver disease. Although liver biopsy is considered to be the gold standard for diagnosis, staging, and monitoring of liver diseases, because of its invasive nature, hepatic enzymes profiles such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, and γ‐glutamyltranspeptidase (GGT) have been used to monitor the extent of liver injury [13, 41]. As those enzymatic changes are nonspecific and lack diagnostic accuracy, there is an unmet need for new noninvasive biomark­ ers. EVs provide a promising utility to be used as “liquid biopsy” for biomarker discovery due to multiple reasons. First, the amount of EVs is usually increased in the disease state, and EVs reflect the snapshot of the transcriptomic and proteomic land­ scape of disease effector cells at the time of release [42]. This characteristic makes them an ideal platform to provide a snap­ shot of the cells in the context of causal disease pathways rather than the stochastic type of biomarkers. Second, due to the ­presence of lipid bilayer membrane the nucleic acid and prot­ eomic content of EVs is protected from enzymatic degradation such as proteases and nucleases and EVs can tolerate adverse environmental and chemical conditions such as extreme pH [43]. Third, EVs offer a significant statistical advantage in

1026

THE LIVER:  REFERENCES

reducing biological matrix complexity and assay background noise, which can lead to the detection of low abundance micro­ molecules with higher sensitivity and specificity [22, 44]. Several proteins and nucleic acid markers have been proposed as diagnostic and prognostic markers for liver diseases in pre­ clinical models and clinical studies as outlined in Table  78.1. The potential utility of EVs positions them as one of the pro­ mising biomarkers for future use in personalized medicine. However, there remain significant challenges in isolation, repro­ ducibility, and establishment of normative controls in different populations and different disease stages.

CHALLENGES AND PROMISES IN TRANSLATIONAL EV BIOMARKER DISCOVERY: SUBPOPULATIONS, ISOLATION TECHNIQUES, AND DIVERSE FUNCTIONS EVs have been successfully recovered from different biofluids including saliva, cerebrospinal fluid (CSF), serum, plasma, urine, saliva, amniotic fluid, pancreatic duct fluid, and breast milk and cultured media [45]. However, increasing evidence suggests that the currently identified EV populations are not homogenous and further understanding of the heterogeneity and function of the different types and sizes of EVs remains a chal­ lenge [46]. Different enrichment methods have been used to iso­ late EVs including the ultracentrifugation, antibody‐coated magnetic bead isolation, size exclusion and filtering, microflu­ idic devices, polymeric precipitation technologies, and porous nanostructures. Methods of EV isolation and processing are cru­ cial steps that can affect the downstream signature of isolated EVs. The choice of the isolation method should be determined based on the type of biomarker, diagnostic purpose, type of bio­ fluid, EV subclass, and clinical setting. In many cases, a combi­ nation of these methods as well as optimization based on the type of biofluid is desired. In most cases, there is a trade‐off between yield and purity and the choice is dependent on the specific question of interest, clinical application, and impor­ tance of cost‐effectiveness versus complexity. With the emer­ gence of the concept of precision medicine and increased understanding of EV subtypes, a method of isolation that can distinguish between different subtypes of EVs is required. Moreover, obtaining EV samples that are devoid of protein‐ aggregate, lipoproteins, and other debris is desirable.

LOOKING FORWARD: EVs AND EXOSOMES IN LIVER DISEASE THERAPY Exosomes are being actively explored as vehicles in therapeutic delivery [47]. The liver is an ideal target because of uptake of EVs by virtually all cell types in the liver. In mice, intravenous administration of EVs was shown to result in rapid EV delivery to the liver and uptake in hepatocytes as well as in liver mono­ nuclear cells (LMNCs) [15]. EVs can be reliably loaded with

small RNAs (miRNAs) or small molecule drugs for therapeutic delivery. In addition, EVs derived from mesenchymal stem cell (MSC) promise tissue repair in liver injury. EVs derived from human liver stem cells were shown to restore arginosuccinate synthase deficiency in mice [48]. Finally, EVs are actively investigated as future therapeutic tools in HCC.

CONCLUSIONS In summary, increasing evidence demonstrates that circulating exosomes and extracellular vesicles can serve as biomarkers in various liver diseases. The unique miRNA, lncRNA, mRNA, and protein cargo of exosomes and EVs provide promise for disease‐specific biomarkers. Furthermore, in addition to serving as biomarkers, exosomes and EVs have an important functional role in the liver in intercellular communication as well as in inter‐organ signaling. Uptake of EVs can modulate the function and phenotype of target cells, specifically of hepatocytes, mac­ rophages, and stellate cells in the liver, indicating an important role for EVs in various types of liver disease. Future studies in this area will provide better understanding of the biology, ­disease association, and therapeutic utility of exosomes and EVs in liver diseases.

REFERENCES 1. Colombo, M., Raposo, G., and Thery, C. Biogenesis, secretion, and intercel­ lular interactions of exosomes and other extracellular vesicles. Annu Rev Cell Dev Biol, 2014;30:255–89. 2. Hessvik, N.P. and Llorente, A. Current knowledge on exosome biogenesis and release. Cell Mol Life Sci, 2018;75(2):193–208. 3. Raiborg, C. and Stenmark, H. The ESCRT machinery in endosomal sorting of ubiquitylated membrane proteins. Nature, 2009;458(7237):445–52. 4. Ostrowski, M., Carmo, N.B., Krumeich, S. et al. Rab27a and Rab27b control different steps of the exosome secretion pathway. Nat Cell Biol, 2010;12(1):19–30; Suppl 1–13. 5. de Gassart, A., Geminard, C., Hoekstra, D., and Vidal, M. Exosome secretion: the art of reutilizing nonrecycled proteins? Traffic, 2004;5(11):896–903. 6. Hurley, J.H. and Odorizzi, G. Get on the exosome bus with ALIX. Nat Cell Biol, 2012;14(7):654–5. 7. Li, B., Antonyak, M.A., Zhang, J., and Cerione, R.A. RhoA triggers a specific signaling pathway that generates transforming microvesicles in ­ ­cancer cells. Oncogene, 2012;31(45):4740–9. 8. Charras, G.T., Hu, C.K., Coughlin, M., and Mitchison, T.J. Reassembly of contractile actin cortex in cell blebs. J Cell Biol, 2006;175(3):477–90. 9. Hirsova, P., Ibrahim, S.H., Krishnan, A. et al. Lipid‐induced signaling causes release of inflammatory extracellular vesicles from hepatocytes. Gastroenterology, 2016;150(4):956–67. 10. Crescitelli, R., Lasser, C., Szabo, T.G. et al. Distinct RNA profiles in sub­ populations of extracellular vesicles: apoptotic bodies, microvesicles and exosomes. J Extracell Vesicles, 2013;2. 11. Zocco, D., Ferruzzi, P., Cappello, F., Kuo, W.P., and Fais, S. Extracellular vesicles as shuttles of tumor biomarkers and anti‐tumor drugs. Front Oncol, 2014;4:267. 12. Verma, M., Lam, T.K., Hebert, E., and Divi, R.L. Extracellular vesicles: potential applications in cancer diagnosis, prognosis, and epidemiology. BMC Clin Pathol, 2015;15:6. 13. Szabo, G. and Momen‐Heravi, F. Extracellular vesicles in liver disease and potential as biomarkers and therapeutic targets. Nat Rev Gastroenterol Hepatol, 2017;14(8):455–66. 14. Masyuk, A.I., Masyuk, T.V., and LaRusso, N.F. Exosomes in the pathogen­ esis, diagnostics and therapeutics of liver diseases. J Hepatol, 2013;59(3): 621–5.



78:  Extracellular Vesicles and Exosomes: Biology and Pathobiology

15. Bala, S., Csak, T., Momen‐Heravi, F. et al. Biodistribution and function of extracellular miRNA‐155 in mice. Sci Rep, 2015;5:10721. 16. Saha, B., Momen‐Heravi, F., Furi, I. et al. Extracellular vesicles from mice with alcoholic liver disease carry a distinct protein cargo and induce mac­ rophage activation through heat shock protein 90. Hepatology, 2018; 67(5):1986–2000. 17. Momen‐Heravi, F., Bala, S., Kodys, K., and Szabo, G. Exosomes derived from alcohol‐treated hepatocytes horizontally transfer liver specific miRNA‐122 and sensitize monocytes to LPS. Sci Rep, 2015;5:9991. 18. Saha, B., Momen‐Heravi, F., Kodys, K., and Szabo, G. MicroRNA cargo of extracellular vesicles from alcohol‐exposed monocytes signals naive mono­ cytes to differentiate into M2 macrophages. J Biol Chem, 2016; 291(1):149–59. 19. Bukong, T.N., Momen‐Heravi, F., Kodys, K., Bala, S., and Szabo, G. Exosomes from hepatitis C infected patients transmit HCV infection and contain replication competent viral RNA in complex with Ago2‐miR122‐ HSP90. PLoS Pathogens, 2014;10(10):e1004424. 20. Li, J., Liu, K., Liu, Y. et al. Exosomes mediate the cell‐to‐cell transmission of IFN‐alpha‐induced antiviral activity. Nat Immunol, 2013;14(8):793–803. 21. Yang, Y., Han, Q., Hou, Z. et al. Exosomes mediate hepatitis B virus (HBV) transmission and NK‐cell dysfunction. Cell Mol Immunol, 2017; 14(5):465–75. 22. Momen‐Heravi, F., Saha, B., Kodys, K. et al. Increased number of circulat­ ing exosomes and their microRNA cargos are potential novel biomarkers in alcoholic hepatitis. J Transl Med, 2015;13:261. 23. Cai, Y., Xu, M.J., Koritzinsky, E.H. et  al. Mitochondrial DNA‐enriched microparticles promote acute‐on‐chronic alcoholic neutrophilia and hepato­ toxicity. JCI Insight, 2017;2(14). 24. Povero, D., Eguchi, A., Li, H. et al. Circulating extracellular vesicles with specific proteome and liver microRNAs are potential biomarkers for liver injury in experimental fatty liver disease. PLoS One, 2014;9(12):e113651. 25. Povero, D., Eguchi, A., Niesman, I.R. et al. Lipid‐induced toxicity stimulates hepatocytes to release angiogenic microparticles that require Vanin‐1 for uptake by endothelial cells. Sci Signal, 2013;6(296):ra88. 26. Garcia‐Martinez, I., Santoro, N., Chen, Y. et al. Hepatocyte mitochondrial DNA drives nonalcoholic steatohepatitis by activation of TLR9. J Clin Invest, 2016;126(3):859–64. 27. Ferrante, S.C., Nadler, E.P., Pillai, D.K. et al. Adipocyte‐derived exosomal miRNAs: a novel mechanism for obesity‐related disease. Pediatr Res, 2015;77(3):447–54. 28. Masyuk, A.I., Huang, B.Q., Ward, C.J. et  al. Biliary exosomes influence cholangiocyte regulatory mechanisms and proliferation through interaction with primary cilia. Am J Physiol Gastrointest Liver Physiol, 2010;299(4): G990–9. 29. Li, X., Liu, R., Huang, Z. et al. Cholangiocyte‐derived exosomal long non­ coding RNA H19 promotes cholestatic liver injury in mouse and humans. Hepatology, 2018;68(2):599–615. 30. Holman, N.S., Mosedale, M., Wolf, K.K., LeCluyse, E.L., and Watkins, P.B. Subtoxic alterations in hepatocyte‐derived exosomes: an early step in drug‐ induced liver injury? Toxicol Sci, 2016;151(2):365–75. 31. Cho, Y.E., Seo, W., Kim, D.K. et al. Exogenous exosomes from mice with acetaminophen‐induced liver injury promote toxicity in the recipient hepato­ cytes and mice. Sci Rep, 2018;8(1):16070. 32. Wang, R., Ding, Q., Yaqoob, U. et al. Exosome adherence and internalization by hepatic stellate cells triggers sphingosine 1‐phosphate‐dependent migra­ tion. J Biol Chem, 2015;290(52):30684–96. 33. Lamichhane, T.N., Leung, C.A., Douti, L.Y., and Jay, S.M. Ethanol induces enhanced vascularization bioactivity of endothelial cell‐derived extracellular vesicles via regulation of microRNAs and long non‐coding RNAs. Sci Rep, 2017;7(1):13794. 34. Chen, L., Chen, R., Kemper, S. et al. Therapeutic effects of serum extracel­ lular vesicles in liver fibrosis. J Extracell Vesicles, 2018;7(1):1461505. 35. Sugimachi, K., Matsumura, T., Hirata, H. et al. Identification of a bona fide microRNA biomarker in serum exosomes that predicts hepatocellular ­carcinoma recurrence after liver transplantation. Br J Cancer, 2015;112(3): 532–8. 36. Yang, N., Li, S., Li, G. et al. The role of extracellular vesicles in mediating progression, metastasis and potential treatment of hepatocellular carcinoma. Oncotarget, 2017;8(2):3683–95. 37. Tu, T., Budzinska, M.A., Maczurek, A.E. et  al. Novel aspects of the liver microenvironment in hepatocellular carcinoma pathogenesis and develop­ ment. Int J Mol Sci, 2014;15(6):9422–58.

1027

38. Wei, J.X., Lv, L.H., Wan, Y.L. et al. Vps4A functions as a tumor suppressor by regulating the secretion and uptake of exosomal microRNAs in human hepatoma cells. Hepatology, 2015;61(4):1284–94. 39. Lv, L.H., Wan, Y.L., Lin, Y. et al. Anticancer drugs cause release of exosomes with heat shock proteins from human hepatocellular carcinoma cells that elicit effective natural killer cell antitumor responses in vitro. J Biol Chem, 2012;287(19):15874–85. 40. Zheng, J., Zhou, Z., Xu, Z. et al. Serum microRNA‐125a‐5p, a useful bio­ marker in liver diseases, correlates with disease progression. Mol Med Rep, 2015;12(1):1584–90. 41. Kim, W.R., Flamm, S.L., Di Bisceglie, A.M., and Bodenheimer, H.C.; Public Policy Committee of the American Association for the Study of Liver Disease. Serum activity of alanine aminotransferase (ALT) as an indicator of health and disease. Hepatology, 2008;47(4):1363–70. 42. Jia, S., Zocco, D., Samuels, M.L. et al. Emerging technologies in extracel­ lular vesicle‐based molecular diagnostics. Expert Rev Mol Diagn, 2014;14(3):307–21. 43. Nawaz, M., Camussi, G., Valadi, H. et al. The emerging role of extracellular vesicles as biomarkers for urogenital cancers. Nat Rev Urol, 2014;11(12): 688–701. 44. Boukouris, S. and Mathivanan, S. Exosomes in bodily fluids are a highly stable resource of disease biomarkers. Proteomics Clin Appl, 2015;9(3‐4): 358–67. 45. Momen‐Heravi, F. Isolation of extracellular vesicles by ultracentrifugation. Methods Mol Biol, 2017;1660:25–32. 46. Willms, E., Cabanas, C., Mager, I., Wood, M.J.A., and Vader, P. Extracellular vesicle heterogeneity: subpopulations, isolation techniques, and diverse functions in cancer progression. Front Immunol, 2018;9:738. 47. Borrelli, D.A., Yankson, K., Shukla, N. et al. Extracellular vesicle therapeu­ tics for liver disease. J Control Release, 2018;273:86–98. 48. Herrera Sanchez, M.B., Previdi, S., Bruno, S. et  al. Extracellular vesicles from human liver stem cells restore argininosuccinate synthase deficiency. Stem Cell Res Ther, 2017;8(1):176. 49. Eguchi, A., Lazaro, R.G., Wang, J. et al. Extracellular vesicles released by hepatocytes from gastric infusion model of alcoholic liver disease contain a microRNA barcode that can be detected in blood. Hepatology, 2017;65(2): 475–90. 50. Bala, S., Petrasek, J., Mundkur, S. et al. Circulating microRNAs in exosomes indicate hepatocyte injury and inflammation in alcoholic, drug‐induced, and inflammatory liver diseases. Hepatology, 2012;56(5):1946–57. 51. Wang, H., Hou, L., Li, A. et al. Expression of serum exosomal microRNA‐21 in human hepatocellular carcinoma. Biomed Res Int, 2014;2014:864894. 52. Chen, L., Chen, R., Kemper, S., Charrier, A., and Brigstock, D.R. Suppression of fibrogenic signaling in hepatic stellate cells by Twist1‐dependent micro­ RNA‐214 expression: role of exosomes in horizontal transfer of Twist1. Am J Physiol Gastrointest Liver Physiol, 2015;309(6):G491–9. 53. Motawi, T.K., Mohamed, M.R., Shahin, N.N., Ali, M.A.M., and Azzam, M.A. Time‐course expression profile and diagnostic potential of a miRNA panel in exosomes and total serum in acute liver injury. Int J Biochem Cell Biol, 2018;100:11–21. 54. Zhang, S., Ouyang, X., Jiang, X. et  al. Dysregulated serum microRNA expression profile and potential biomarkers in hepatitis C virus‐infected patients. Int J Med Sci, 2015;12(7):590–8. 55. Moratti, E., Vezzalini, M., Tomasello, L., Giavarina, D., and Sorio, C. Identification of protein tyrosine phosphatase receptor gamma extracellular domain (sPTPRG) as a natural soluble protein in plasma. PLoS One, 2015;10(3):e0119110. 56. Welker, M.W., Reichert, D., Susser, S. et al. Soluble serum CD81 is elevated in patients with chronic hepatitis C and correlates with alanine aminotrans­ ferase serum activity. PLoS One, 2012;7(2):e30796. 57. Brodsky, S.V., Facciuto, M.E., Heydt, D. et  al. Dynamics of circulating microparticles in liver transplant patients. J Gastrointestin Liver Dis, 2008;17(3):261–8. 58. Julich‐Haertel, H., Urban, S.K., Krawczyk, M. et  al. Cancer‐associated ­circulating large extracellular vesicles in cholangiocarcinoma and hepatocel­ lular carcinoma. J Hepatol, 2017;67(2):282–92. 59. Fu, Q., Zhang, Q., Lou, Y. et al. Primary tumor‐derived exosomes facilitate metastasis by regulating adhesion of circulating tumor cells via SMAD3 in liver cancer. Oncogene, 2018. 60. Conde‐Vancells, J., Rodriguez‐Suarez, E., Gonzalez, E. et al. Candidate bio­ markers in exosome‐like vesicles purified from rat and mouse urine samples. Proteomics Clin Appl, 2010;4(4):416–25.

79

Integrated Technologies for Liver Tissue Engineering Tiffany N. Vo1, Amanda X. Chen2, Quinton B. Smith1, Arnav Chhabra3, and Sangeeta N. Bhatia1,3,4,5,6 Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA, USA Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA 3 Harvard‐MIT Department of Health Sciences and Technology, Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Boston, MA, USA 4 Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, USA 5 David H. Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA, USA 6 Howard Hughes Medical Institute, Chevy Chase, MD, USA 1 2

INTRODUCTION Tissue engineering approaches utilizing strategic combinations of cells, biomaterial scaffolds, and bioactive factors are promising alternatives to whole‐organ transplantation for the augmentation or replacement of liver function. Additionally, convergence of biomedical technologies across diverse disciplines has enabled development of novel platforms and tools for elucidating liver biology. This chapter reviews the latest advances in liver tissue engineering for the treatment and understanding of liver disease, including cell therapy, in vitro culture models, organ‐on‐chip platforms, bioprinting, and three‐dimensional ­implantable constructs.

CURRENT THERAPIES AND CHALLENGES Liver failure is a significant health problem and major global burden, accounting for approximately 2% of all deaths annually in the United States, and represents a multibillion‐dollar economic burden [1]. Fulminant and chronic liver disease arise from varied etiologies, including drug‐mediated toxicity, hepatotropic infections, excess lipid accumulation, and metabolic genetic ­disorders [2]. The only therapy shown to improve mortality is partial or full orthotopic liver transplantation. However, the

availability of transplantable donor tissue is insufficient, and the demand for donor organs continues to be exacerbated by a growing obesity epidemic and opioid crisis. Excitingly, the advent of accessible direct‐acting antiviral therapies may enable the use of donor organs with detectable hepatitis C viral load, though a full evaluation of the impact of coinfections and treatment regimens across patient genotypes remains to be examined [3]. Due to the need for methods to support liver function prior to or instead of transplantation, a variety of acellular liver support devices were developed [4]. However, these nonbiological systems were unable to recapitulate the complete repertoire of liver functions, including carbohydrate, protein, and lipid metabolism, synthesis and secretion of blood and bile components, and detoxification. Subsequently, biomedical engineering enabled cell‐based technologies such as bioartificial liver devices, cell transplantation, and tissue‐engineered constructs (Figure 79.1).

Cell‐based therapeutics Extracorporeal bioartificial liver devices (BALs) are temporary support devices that can process the blood of a patient with liver failure. A hollow fiber device is the most common BAL configuration and was inspired by hemodialysis equipment. The range of BAL devices also include flat plate devices, perfusion bed or scaffold systems, and suspension bioreactors. Across these device designs, which are reviewed in detail elsewhere [5] and discussed later in this chapter, challenges with mass

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



79:  Integrated Technologies for Liver Tissue Engineering

1029

Figure 79.1  Technologies for treatment and study of liver disease. Hepatocytes can be sourced from freshly isolated or cryopreserved livers, and can also be derived from progenitor or stem cell populations. They can be cultured in different 2D or 3D configurations to promote liver‐specific functions. Technologies such as bioprinting, biomaterials, and microfluidic platforms enable the development of 3D models for implantation. The therapeutic potential of implants can be assessed using a variety of preclinical animal models.

transfer, scalability, and hepatocyte phenotype stability have hindered clinical translation. Aside from extracorporeal liver support, several technologies have attempted to replace damaged or diseased tissue in vivo. Studies detailing successful transplantation of adult hepatocytes in rodent models showed potential, though clinical trials reported poor engraftment and an underwhelming increase in disease‐free survival [2].

Cell source A significant challenge common to engineered cell‐based therapies, such as BAL devices and cell transplantation, is cell sourcing. The availability of fresh or cryopreserved donor hepatocytes is limited and isolated populations are difficult to expand and maintain in vitro. Stem and progenitor cells are being investigated as an alternative cell source due to their robust proliferative capacity and pluripotency, though the ability to achieve complete differentiation and phenotypic stability is challenging [2, 6]. Immortalized primary cell lines or hepatoma cell lines are also problematic, owing to their aberrant function and tumorigenic potential. Regardless of the cell source, there are major concerns regarding the sustained stability of hepatic phenotype ex vivo, which is a requirement for the development of any cell‐based liver therapy.

In vitro models The development of in vitro culture platforms to improve hepatic tissue engineering is an active area of research. It is possible that signals from the native stroma, including cell–cell interactions, cell–matrix interactions, and trophic factors that control hepatic phenotype may be used to recapitulate an optimized artificial microenvironment. This task is no small feat, considering the liver’s highly organized organ‐level structure  –  consisting of hepatic cords sandwiched between the hepatic artery, portal vein, sinusoids and biliary network. Furthermore, on the cellular level, dynamic heterotypic cell–cell and cell–matrix interactions involve hepatocytes, nonparenchymal cells, and their surrounding extracellular matrix (ECM). Thus, it is no surprise that the hepatic phenotype has remained elusive in long‐term cultures ex vivo.

Two‐dimensional culture models Hepatocytes are acutely sensitive to their microenvironment, evidenced by their decline in phenotype ex vivo. Manipulation of the microenvironment through control of ECM scaffolds, presentation of soluble cues, and co‐culture with nonparenchymal cells has been found to influence hepatocyte phenotypic stability [7].

1030

THE LIVER:  DYNAMIC 3D MODELS

The impact of ECM on hepatocyte culture has been extensively studied. Notably, hepatocytes can be cultured on a collagen‐coated tissue culture plate, or in a traditional collagen “sandwich” format. These culture formats lead to an improvement in cell morphology and function for weeks, though aspects of function relating to drug metabolism are not maintained long term. Hepatocyte function has also been improved by co‐culture with nonparenchymal liver cells [8]. A study consisting of primary hepatocytes, Kupffer cells, liver sinusoidal endothelial cells (LSECs), and hepatic stellate cells was shown to recapitulate some aspects of organ‐level response not captured by monocultures [9]. Additionally, hepatocytes can be supported by cell types from outside the native liver microenvironment; for example, fibroblasts, which are not explicitly present in the liver, emerged as a support population for the ex vivo culture of hepatocytes. Specifically, Khetani and Bhatia controlled heterotypic and homotypic cell–cell interactions in a primary hepatocyte and fibroblast co‐culture through soft lithography techniques. Liver‐ specific functions, including difficult‐to‐stabilize drug metabolism‐related enzymes, were maintained for several weeks [10]. These two‐dimensional (2D) hepatic cultures are scalable and relatively inexpensive, leading to their common use for ADME/ TOX (absorption, distribution, metabolism, and excretion/toxicity) screening. However, hepatocytes adopt a flat morphology in most 2D cultures, even when overlaid with collagen in sandwich cultures. Many have posited that this deviation from normal three‐dimensional (3D) cytoarchitecture leads to discrepancies in hepatic function and stability [11]. In the following sections, we highlight examples of strategies used to sustain hepatic function in culture, with an emphasis on recent advances.

lethal liver failure model [16]. In these platforms, self‐organization occurs over the course of days. To attain a higher degree of control of the 3D microenvironment, methods from photolithography were adapted to produce techniques for rapid, reproducible patterning. For example, Stevens et al. demonstrated rapid microscale organization of spheroids and showed compatibility with various cell combinations and hydrogel biomaterial scaffolds. Control of architecture in these tissue‐engineered scaffolds was shown to impact function post implantation [17].

DYNAMIC 3D MODELS Despite myriad improvements in model systems, in vitro liver cultures still lack many of the architectural and functional features of in vivo tissues, spatial distribution of soluble factors, and the non‐steady state forces and systemic factors provided by blood circulation. Even in configurations where multiple liver cell types are co‐cultured to enhance tissue‐specific signaling and functionality, the effects of multiorgan interactions, ambient oxygen tension gradients, and cellular trafficking through tissues are not captured. Factors such as oxygen and hormone gradients, nutrients, matrix composition, and the distribution of nonparenchymal cells are known to regulate functions of hepatocytes along a sinusoid. The liver is also supplied with antigen‐ rich blood from the gastrointestinal tract and is being constantly surveyed by antigen‐presenting cells and lymphocytes. As such, advancements in microfluidic and 3D technologies have fueled various platforms to improve biological fidelity in order to comprehensively model liver physiology.

Three‐dimensional culture models

Microfluidic platforms

Multiple comparisons indicate that 2D cultures are morphologically and functionally inferior to 3D cultures built to mimic the  native tissue microenvironment [11]. This discrepancy is thought to be partially due to the unique manner in which chemical and mechanical signals are presented to cells in a 3D matrix, as reviewed in depth elsewhere [12]. The scaffold materials used to host these 3D cell constructs can consist of natural and/ or synthetic biomaterials, some of which are compatible with emerging bioprinting techniques [13]. Three‐dimensional spheroid culture evolved as a method for controlling cellular composition and interactions [14]. Spheroids are commonly encapsulated in polymeric biomaterials or cultured in suspension (discussed later). The design of spheroid‐laden scaffolds offers a seemingly endless parameter space, and has yielded a number of successful studies that can be broadly categorized into stem or progenitor cell‐derived organoids or primary cell‐derived spheroids. Stem and progenitor cells can be obtained from a range of tissues, and typically have a large proliferative capacity. Huch et al. used stem cells to generate hepatic organoids from a single Lgr5+ cell source, which could be transplanted into diseased mouse models and also differentiated into hepatocyte‐like cells [15]. Stem cells have also been combined with other stromal and endothelial cells to mimic early organogenesis; specifically, Takebe et al. reported on the ectopic transplantation of vascularized, self‐organized liver “buds,” which were able to rescue a

Historically, improvements in the semiconductor manufacturing industry for the purpose of microelectronics led to the miniaturization of biological culture and analysis. Techniques such as photolithography and replica molding were adapted to control microscale architecture and create 3D microfluidic structures with hollow channels for liquid perfusion [18, 19]. Advances in these microengineering techniques have allowed for the development of in vitro platforms that recapitulate complex 3D mechanical cues such as flow and shear stress, and organ‐level functions that are not present in static 2D cultures.

Stand‐alone bioreactors A number of bioreactors have been developed for liver cultures, including those that recapitulate complex 3D microenvironments, which have been covered elsewhere [20, 21], to configurations that provide microfluidic flow over a flat plate of cultured hepatocytes. The latter has been shown to improve nutrient and oxygen delivery, but makes it challenging to decouple autocrine versus paracrine cues. Hollow fibers are culture systems that consist of fibers fixed into a module with cells seeded on the outside of the fibers and media delivered through the lumen. They have been shown to shield hepatocytes from shear stresses associated with perfusion. These bioreactors typically exhibit superior mass transport into the interstitium of the device.



79:  Integrated Technologies for Liver Tissue Engineering

Stirred bioreactors can be used to generate spheroids of consistent sizes. Additionally, they sustain cultures for weeks and provide the added benefit of convective mass transfer and nearly homogeneous dissolved oxygen concentrations in the culture. Lastly, microfabricated 3D reactors allow for the precise control over local perfusion of capillary bed‐sized tissue structures that mimic in vivo architecture. The precise control over flow rates that these devices offer is necessary to study shear‐dependent systems such as the interplay between hepatocytes and LSECs during liver regeneration.

Integrated plug‐and‐play While stand‐alone bioreactors provide a configurable model system to study the liver in isolation, interactions between the liver, the immune system, and other organs mediate a large number of disease processes. Since the liver’s hepatic portal system and biliary tree interface with a number of other organs, liver function is often disrupted even when the root cause of the problem is the perturbation of another organ. Multiorgan systems help capture this phenomenon by embedding varied tissue models around preformed vasculature and connecting them through microfluidic channels that represent blood vessels [22]. In these devices, metabolites, soluble factors, and proteins from the liver can deliver cues to other organ systems such as the heart, and various readouts can be monitored in real time through a dynamic repertoire of physiologically relevant biomarkers. Microfluidic devices also allow for the integration of circulating immune cells to monitor the interaction between the immune system and liver cells under flow conditions. It has been demonstrated that the adhesion and migration of circulating monocytes triggers repolarization of tissue‐resident macrophages in a liver‐on‐a‐chip device [23].

Drug development applications Even though disease modeling in liver‐on‐a‐chip devices is in its early stages, one area where the field has made significant headway is in studying drug metabolism. The pharmaceutical pipeline is clogged by drugs that pass the preclinical (in vitro and animal testing) phase but fail after several years of human trials. Approximately 90% of the drugs that make it to human trials fall under this category and end up costing pharmaceutical companies hundreds of millions of dollars per drug. For drug studies, in particular, animal models are often not predictive of human outcomes, due to species‐specific drug metabolism pathways such as those regulated by cytochrome P450 enzymes. Additionally, given their ability to integrate multiple organ systems, liver‐on‐a‐chip systems have been shown to more faithfully model human ADME/TOX and capture adverse drug reactions. These physiologically coupled devices have predicted the formation of toxic metabolites by the liver, which have downstream deleterious effects on other organ systems. For example, Viravaidya et  al. characterized the conversion of naphthalene into its toxic metabolite with a liver–lung–fat tissue chip, in which they observed depleted glutathione levels in the downstream lung chamber caused by accumulated compounds in the fat chamber [24]. Coupling

1031

organ systems offers preclinical analysis of ADME/TOX and of cancer drug metabolism [25, 26], which is typically limited to animal models.

THREE‐DIMENSIONAL IMPLANTABLE CONSTRUCTS While 2D technologies have proved promising in a wide array of applications, the advent of 3D printing opened a door to the development of large‐scale engraftable tissues that can augment liver function for translation in a clinical setting [27]. Yet these efforts still require a functioning vascular system that can meet the metabolic demand of engineered tissues. It is estimated that cells must be within 150–200 μm of a capillary structure or risk necrosis due to the lack of oxygen, nutrients, and waste transport. To this end, advances in liver tissue engineering hinge upon the ability to generate vasculature as well as other complex network structures. Here, we discuss recent advances in biomaterial design and strategies for building multiscale architectures in implantable constructs.

Biomaterial design The ECM serves as a structurally supportive scaffold, providing instructive chemical cues that support hepatic function and promote de novo vasculogenesis. Biomaterial approaches aiming at directing the assembly and stability of the cellular microarchitecture and vascular tree within engineered tissues need to support the bidirectional integrin‐mediated signaling between cells and the surrounding ECM. Early work focused on the use of porous biodegradable scaffolds such as poly (lactic‐co‐glycolic acid) or poly (l‐lactic acid) as planar‐like substrates for hepatocyte attachment [28]. Given the importance of 3D conformation for cellular function, moisture‐retaining polymeric biomaterial scaffolds known as hydrogels represent a powerful engineering tool to recapitulate tissue‐specific ECMs in a physiologically relevant manner [29]. Natural biomaterial approaches utilize fibronectin, collagen, gelatin, chitosan, cellulose, and glycosaminoglycans such as hyaluronic acid to study vascular assembly and liver function in 3D matrices because they are present in the native vascular and liver microenvironment. In particular, collagen I hydrogels have been extensively used in the mechanistic study of vasculogenesis, uncovering the roles of integrin‐mediated vacuole and subsequent lumen formation. Like collagen I, fibrin also supports the self‐assembly of patent vessel structures in vitro and has been extensively studied to direct vascularization pre‐implantation. However, natural scaffolds have limited clinical use due to their xenogeneic origin and batch‐to‐batch variation. Furthermore, decoupling the role of matrix characteristics on cell behavior, including stiffness, adhesion, and degradation kinetics is difficult, as these systems have limited tunability. Synthetic hydrogels such as polyethylene glycol (PEG) and dextran, on the other hand, are biocompatible polymers that can be functionalized by conjugation of bioactive moieties, addressing the limitation of  reproducibility and the ability to define matrix properties.

1032

THE LIVER:  ANIMAL MODELS FOR ASSESSING THERAPEUTIC EFFICACY

Beyond architecture and bioinspired characteristics, matrix design has also evolved to generate biomaterials with inducible, stimuli‐responsive, and dynamic properties [2, 30].

Network structure: the control of architecture in implantable liver grafts Whole‐organ decellularization Perfusion‐based liver decellularization methodologies have led to the formation of scaffolds that maintain liver ECM composition (i.e. collagen IV, fibronectin, and laminin), growth factor presentation (HGF and bFGF), and vascular/biliary architecture [31, 32]. These decellularized scaffolds are not only amenable to fetal liver cell engraftment and supporting ductal/hepatocyte lineage maturation [31], but also support mature vascular cells and hepatocytes, which show key gene expression and protein synthesis functionality in vivo [33]. Re‐cellularization strategies in which both adult and natal parenchymal cells are introduced in the biliary tract and portal vein can result in up to 95% cell repopulation [34]. Importantly, decellularized livers have been shown to engraft and support interspecies biocompatibility in immunocompetent models, showing limited rejection or inflammation [35]. However, this strategy is limited by the availability of liver tissue.

In a similar manner, biological materials can be layered sequentially as sheaths through microextrusion, a process involving mechanical or pneumatic forces to expel cell‐laden or acellular constructs. This technique is advantageous in the spatial assembly of cells embedded within natural or synthetic ECM in that it does not require temperatures that are outside physiologically permissive ranges. Exploiting differences in reversible thermal gelation has proven successful in a range of 3D printing applications [38]. In one example, fugitive Pluronic F127 synthetic polymer ink and cell‐laden gelatin methacrylate (GelMA) hydrogels allowed for concurrent printing of cells and sacrificial vascular elements [40]. Building upon this strong technological foundation, large vascularized constructs exceeding 1 cm in thickness were printed alongside cell‐laden inks composed of gelatin‐fibrin composites, opening avenues for additional biomaterial inks [41]. The described technologies also show great potential for generating other network structures. The liver contains a complex ductal network lined by biliary epithelial cells, termed cholangiocytes, that aid in the modification and removal of hepatocyte‐secreted bile. Several disease models argue the necessity of an intact biliary network in liver regeneration [42, 43], but this has not yet been explored for implantable liver constructs.

Three‐dimensional printing vascular architecture Engineering the hierarchical structure of the vasculature will help to mimic the organization of blood vessels in vitro, as resident vessels vary in density, alignment, and tortuosity as they meet the metabolic demand of surrounding tissue. To this end, microfabrication techniques have allowed the precise control of cell–cell and cell–ECM interactions, guiding vascular architecture [36]. The advent of 3D printing technology has transformed our ability to mimic vascular architecture in vitro. Building upon principles from layered stereolithography, biological materials including cells, instructive growth factors, and polymeric materials can be arranged with precision in both lateral and axial directions [37]. Utilization of imaging modalities including computed topography or magnetic resonance imaging allows for tissue‐specific reconstruction of vascular networks through sequential mapping of 2D slices. Bioprinting of sacrificial or fugitive inks has served as a promising alternative to traditional inkjet printing modalities. These strategies in particular circumvent the issues of toxicity imposed by heating or acoustic frequency by printing acellular conduits prior to cellular introduction. This was elegantly ­demonstrated through the fabrication of sacrificial, thermally extruded carbohydrate glass lattices, which could be embedded in a range of cell‐laden natural and synthetic biomaterials, including Matrigel, fibrin, and PEG‐based hydrogels [38, 39]. Importantly, both the introduction and removal of the sugar lattices with media had no effect on the crosslinking scheme (i.e. ionic, enzymatic, photopolymerization) or bulk properties of the various materials. To this end, 3D printed perfusable channels were able to better support hepatocyte albumin secretion and urea synthesis in agarose gels, compared to bulk encapsulation, demonstrating the necessity of vascular architecture for supporting parenchymal tissue.

ANIMAL MODELS FOR ASSESSING THERAPEUTIC EFFICACY Experimental animal models remain an important factor in the development of engineered liver therapies. Despite extensive research efforts, the complex interplay of multicellular interactions and signaling cascades of liver regeneration has not been fully understood, and thus cannot be replicated in vitro. In order to assess therapeutic efficacy and safety, a number of large and small animal models have been developed, and are covered in detail elsewhere [44–46]. As shown in Figure 79.1, modes of injury fall into several major categories: surgical, chemical, dietary, and genetic. Surgical models include partial hepatectomy and bile duct ligation. While less clinically relevant, these models provide a well‐defined injury stimulus by which to assess hepatocyte proliferation in engineered grafts in response to regenerative cues. However, the regenerative stimulus is short‐lived, making it difficult to assess graft function over time. Alternatively, chemical injury models rely on hepatotoxins such as carbon tetrachloride, acetaminophen, or thioacetamide to induce centrilobular or periportal necrotic lesions in the liver. These injuries are more representative of human liver pathophysiology, but are highly variable and difficult to reproduce depending on the model species and age, toxin dose, and mode of administration. While tissue‐engineered liver constructs have been successfully tested using these established models, there are aspects of human liver biology that have not yet been replicated. Efforts in recent years have sought to develop models that accurately reflect human metabolic disease, enable understanding of human‐ specific liver biology, and evaluate efficacy and safety in a human context.



79:  Integrated Technologies for Liver Tissue Engineering

1033

Humanized mouse models

Chimeric xenotransplants

Study and expansion of human hepatocytes in mice for drug testing and cell therapy have historically been hampered by low engraftment rates. Advances in gene modification technologies have enabled improved immunodeficient transgenic mouse models with robust host liver injury to promote a selective growth advantage for human hepatocytes [47]. The first system, albumin‐uroplasminogen activator (uPA), causes hepatotoxicity through constitutive expression of uPA under an albumin promoter in mouse hepatocytes [48]. The other system is a model of hereditary tyrosinemia I, in which a genetic knockout of fumarylacetoacetate hydrolase (FAH) leads to toxic accumulation of fumarylacetoacetate [49]. Whereas the uPA mice experience injury throughout their lifetime, leading to a narrow window for engraftment, the health and initiation of liver injury in FAH‐knockout mice can be controlled through administration of a drug, 2‐(2‐nitro‐4‐­ trifluoro‐methylbenzoyl)‐1,3‐cyclohexanedione. Engraftment levels upwards of 70% have been observed in both models, and, in some cases, greater than 90% when the FAH knockout is combined with severe immunodeficiency. Inducible‐injury transgenic models have also been developed, which activate hepatotoxic transgenes following small‐molecule drug administration [50, 51]. The TK‐NOG model causes hepatocyte ablation after ganciclovir activation of herpes simplex type 1 thymidine kinase transgene under an albumin promoter. Similarly, hepatocyte death in the AFC8 model is caused by AP20187 induction of a fusion protein consisting of FK506‐ binding protein and caspase 8 under an albumin promoter. Both models been successfully used to achieve high levels of human chimerism.

One exciting prospect with the advancements in genetic engineering technologies is the possibility of generating functional human organs for transplantation. Xenotransplantation, or the transplantation of nonhuman organs to humans, has great potential to address the current severe organ shortage by leveraging the breeding efficiency of animals, and their similar morphological and physiological characteristics to humans. While still in the preliminary stages, genetically modified pigs with inactivated porcine endogenous retroviruses have been successfully produced to prevent cross‐species immunological rejection [55]. In another example, researchers have reported success in producing human–porcine and human–ovine chimera embryos with human stem cells [56], taking the first steps towards production of human organs inside animals. Overall, experimental animal models remain a critical part of the evaluation of engineered liver tissues. The convergence of new technologies, as well as the novel insights gained from current studies on human liver injury and regeneration will enable the design of more efficacious and scalable therapies.

Extrahepatic/intracorporeal humanized models In transgenic models, the ability to achieve high levels of humanization coupled with normal function and morphological characteristics makes them useful for studying human liver‐­ specific biology, such as hepatotropic infectious diseases and species‐specific drug metabolism. Interestingly, the soluble regenerative signals produced by mouse liver injury creates a repopulation advantage that can also be used for the selective expansion of human hepatocytes in extrahepatic grafts [52]. Engineered liver tissues implanted in ectopic sites such as the subcutaneous space or mesenteric fat‐pad provide several benefits. They are easily produced compared to transgenic models and accessible for both surgical manipulation and noninvasive imaging. Ectopic grafts also bypass issues present in the native liver such as high portal pressures and microenvironmental cues that may lower engraftment rates during transplantation of high quantities of hepatocytes. Lastly, although distant from the native liver environment, ectopic implants allow for decoupling of human liver regeneration mechanisms from the animal host. Recent work has demonstrated that extrahepatic human liver grafts with defined cell composition and microarchitecture can integrate with the host vasculature [53], perform human‐specific metabolic and secretory functions [54], and even improve survival after liver injury [16].

NEW READOUTS FOR LIVER FUNCTION As new technologies for creating liver therapies and preclinical models become more complex and widespread, real‐time, longitudinal, noninvasive, and more scalable readouts are needed for evaluation. Advances in and integration of disparate fields have enabled new sensors for monitoring long‐term function, assessing therapeutic efficacy, and elucidating biology. For longitudinal and real‐time monitoring, imaging modalities such as magnetic resonance spectroscopy, bioluminescence imaging, and radioactive labeling provide ways to gauge liver status and metabolic function noninvasively over time. In addition to traditional blood protein measurements using host biomarkers, nanotechnology also has enabled new ways to diagnose liver disease. For example, nano‐sized synthetic protease‐sensitive activity markers have successfully been employed to noninvasively detect liver fibrosis with high signal‐to‐noise ratio using urinary detection [57]. For evaluation and elucidating biology, soft epidermal electronics or electronically integrated tissues provide ways to analyze tissue characteristics and augment existing function. Electrochemical sensors, force‐sensitive cantilevers, and electrosensitive materials have been integrated within tissue culture platforms and benchtop organ‐on‐chip systems or engineered artificial tissues to track cell signaling [58]. Although preliminary in many ways, these new readouts can help improve translation of therapies, refine nonhuman experimental models, and ultimately bridge the gap between in vitro data and clinical translation.

CONCLUSION Tissue engineering holds great potential for the development of clinically effective liver therapies. Progress in the fields of biomaterial science, microfabrication, regenerative medicine,

1034

THE LIVER: REFERENCES

bioengineering, genetic engineering, and developmental biology have enabled novel in vitro platforms and implantable tissues that recapitulate the structural complexity and functional axes of the liver. Concurrently, advances in the physical sciences have led to improved tools and readouts for functional assessment, longitudinal monitoring, and enhanced characteristics. Although challenges remain for overall translation, integrated technologies for liver tissue engineering provide a positive outlook for future treatment and study of liver disease.

REFERENCES 1. Kim, W.R., Brown, R.S., Terrault, N.A., and El‐Serag, H. Burden of liver disease in the United States: summary of a workshop. Hepatology, 2002;36:227–42. 2. Bhatia, S.N., Underhill, G.H., Zaret, K.S., and Fox, I.J. Cell and tissue engineering for liver disease. Sci Transl Med, 2014;6:245sr2. 3. Bushyhead, D. and Goldberg, D. Use of hepatitis C‐positive donor livers in liver transplantation. Curr Hepatol Rep, 2017;16:12–17. 4. Carpentier, B., Gautier, A., and Legallais, C. Artificial and bioartificial liver devices: present and future. Gut, 2009;58:1690–702. 5. Allen, J.W., Hassanein, T., and Bhatia, S.N. Advances in bioartificial liver devices. Hepatology, 2001;34:447–55. 6. Hansel, M.C., Davila, J.C., Vosough, M. et al. The use of induced pluripotent stem cells for the study and treatment of liver diseases. Curr Protoc Toxicol, 2016;67:14.13.1–14.13.27. 7. Godoy, P., Hewitt, N.J., Albrecht, U. et al. Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non‐parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch Toxicol, 2013;87:1315–530. 8. Guguen‐Guillouzo, C., Clément, B., Baffet, G. et  al. Maintenance and reversibility of active albumin secretion by adult rat hepatocytes co‐cultured with another liver epithelial cell type. Exp Cell Res, 1983;143:47–54. 9. Bale, S.S., Geerts, S., Jindal, R., and Yarmush, M.L. Isolation and co‐culture of rat parenchymal and non‐parenchymal liver cells to evaluate cellular interactions and response. Sci Rep, 2016;6:25329. 10. Khetani, S.R. and Bhatia, S.N. Microscale culture of human liver cells for drug development. Nat Biotechnol, 2008;26:120–6. 11. Yuasa, C., Tomita, Y., Shono, M. et al. Importance of cell aggregation for expression of liver functions and regeneration demonstrated with primary cultured hepatocytes. J Cell Physiol, 1993;156:522–30. 12. Griffith, L.G. and Swartz, M.A. Capturing complex 3D tissue physiology in vitro. Nat Rev Mol Cell Biol, 2006;7:211–24. 13. Liaw, C.‐Y., Ji, S. and Guvendiren, M. Engineering 3D hydrogels for personalized in vitro human tissue models. Adv Healthc Mater, 2018;7. 14. Fennema, E., Rivron, N., Rouwkema, J. et al. Spheroid culture as a tool for creating 3D complex tissues. Trends Biotechnol, 2013;31:108–15. 15. Huch, M., Dorrell, C., Boj, S.F. et al. In vitro expansion of single Lgr5+ liver stem cells induced by Wnt‐driven regeneration. Nature, 2013;494:247–50. 16. Takebe, T., Sekine, K., Enomura, M. et  al. Vascularized and functional human liver from an iPSC‐derived organ bud transplant. Nature, 2013;499: 481–4. 17. Stevens, K.R., Ungrin, M.D., Schwartz, R.E. et al. InVERT molding for scalable control of tissue microarchitecture. Nat Commun, 2013;4:1847. 18. Bhatia, S.N. and Ingber, D.E. Microfluidic organs‐on‐chips. Nat Biotechnol, 2014;32:760–72. 19. Huh, D., Hamilton, G.A., and Ingber, D.E. From 3D cell culture to organs‐ on‐chips. Trends Cell Biol, 2011;21:745–54. 20. Ebrahimkhani, M.R., Neiman, J.A.S., Raredon, M.S.B. et  al. Bioreactor technologies to support liver function in vitro. Adv Drug Deliv Rev, 2014;69–70:132–57. 21. Powers, M.J., Domansky, K., Kaazempur‐Mofrad, M.R. et al. A microfabricated array bioreactor for perfused 3D liver culture. Biotechnol Bioeng, 2002;78:257–69. 22. Skardal, A., Murphy, S.V., Devarasetty, M. et al. Multi‐tissue interactions in an integrated three‐tissue organ‐on‐a‐chip platform. Sci Rep, 2017;7:8837.

23. Gröger, M., Rennert, K., Giszas, B. et  al. Monocyte‐induced recovery of inflammation‐associated hepatocellular dysfunction in a biochip‐based human liver model. Sci Rep, 2016;6:21868. 24. Viravaidya, K., Sin, A., and Shuler, M.L. Development of a microscale cell culture analog to probe naphthalene toxicity. Biotechnol Prog, 2004;20: 316–23. 25. Choucha Snouber, L., Bunescu, A., Naudot, M. et al. Metabolomics‐on‐a‐ chip of hepatotoxicity induced by anticancer drug flutamide and Its active metabolite hydroxyflutamide using HepG2/C3a microfluidic biochips. Toxicol Sci, 2013;132:8–20. 26. Tatosian, D.A. and Shuler, M.L. A novel system for evaluation of drug mixtures for potential efficacy in treating multidrug resistant cancers. Biotechnol Bioeng, 2009;103:187–98. 27. Damania, A., Jain, E., and Kumar, A. Advancements in in vitro hepatic ­models: application for drug screening and therapeutics. Hepatol Int, 2014; 8:23–38. 28. Kim, S.S., Utsunomiya, H., Koski, J.A. et al. Survival and function of hepatocytes on a novel three‐dimensional synthetic biodegradable polymer scaffold with an intrinsic network of channels. Ann Surg, 1998;228:8–13. 29. Drury, J.L. and Mooney, D.J. Hydrogels for tissue engineering: scaffold design variables and applications. Biomaterials, 2003;24:4337–51. 30. Tibbitt, M.W., Rodell, C.B., Burdick, J.A., and Anseth, K.S. Progress in material design for biomedical applications. Proc Natl Acad Sci U S A, 2015;112:14444–51. 31. Baptista, P.M., Siddiqui, M.M., Lozier, G. et  al. The use of whole organ decellularization for the generation of a vascularized liver organoid. Hepatology, 2011;53:604–17. 32. Soto‐Gutierrez, A., Zhang, L., Medberry, C. et al. A whole‐organ regenerative medicine approach for liver replacement. Tissue Eng Part C Methods, 2011;17:677–86. 33. Uygun, B.E., Soto‐Gutierrez, A., Yagi, H. et al. Organ reengineering through development of a transplantable recellularized liver graft using decellularized liver matrix. Nat Med, 2010;16:814–20. 34. Ogiso, S., Yasuchika, K., Fukumitsu, K. et al. Efficient recellularisation of decellularised whole‐liver grafts using biliary tree and foetal hepatocytes. Sci Rep, 2016;6:35887. 35. Mazza, G., Rombouts, K., Rennie Hall, A. et al. Decellularized human liver as a natural 3D‐scaffold for liver bioengineering and transplantation. Sci Rep, 2015;5:13079. 36. Baranski, J.D., Chaturvedi, R.R., Stevens, K.R. et al. Geometric control of vascular networks to enhance engineered tissue integration and function. Proc Natl Acad Sci U S A, 2013;110:7586–91. 37. Murphy, S.V. and Atala, A. 3D bioprinting of tissues and organs. Nat Biotechnol, 2014;32:773–85. 38. Hinton, T.J., Jallerat, Q., Palchesko, R.N. et al. Three‐dimensional printing of complex biological structures by freeform reversible embedding of suspended hydrogels. Sci Adv, 2015;1:e1500758. 39. Miller, J.S., Stevens, K.R., Yang, M.T. et al. Rapid casting of patterned vascular networks for perfusable engineered three‐dimensional tissues. Nat Mater, 2012;11:768–74. 40. Kolesky, D.B., Truby, R.L., Gladman, A.S. et al. 3D bioprinting of vascularized, heterogeneous cell‐laden tissue constructs. Adv Mater Weinheim, 2014;26:3124–30. 41. Kolesky, D.B., Homan, K.A., Skylar‐Scott, M.A., and Lewis, J.A. Three‐ dimensional bioprinting of thick vascularized tissues. Proc Natl Acad Sci U S A, 2016;113:3179–84. 42. Raven, A., Lu, W.‐Y., Man, T.Y. et al. Cholangiocytes act as facultative liver stem cells during impaired hepatocyte regeneration. Nature, 2017;547: 350–4. 43. Ferreira‐Gonzalez, S., Lu, W.‐Y., Raven, A. et al. Paracrine cellular senescence exacerbates biliary injury and impairs regeneration. Nat Commun, 2018;9:1020. 44. Forbes, S.J. and Newsome, P.N. Liver regeneration – mechanisms and models to clinical application. Nat Rev Gastroenterol Hepatol, 2016;13:473–85. 45. Palmes, D. and Spiegel, H.‐U. Animal models of liver regeneration. Biomaterials, 2004;25:1601–11. 46. Tan, A.K.Y., Loh, K.M., and Ang, L.T. Evaluating the regenerative potential and functionality of human liver cells in mice. Differentiation, 2017; 98:25–34. 47. Grompe, M. and Strom, S. Mice with human livers. Gastroenterology, 2013;145:1209–14.



79:  Integrated Technologies for Liver Tissue Engineering

48. Rhim, J.A., Sandgren, E.P., Palmiter, R.D., and Brinster, R.L. Complete reconstitution of mouse liver with xenogeneic hepatocytes. Proc Natl Acad Sci U S A, 1995;92:4942–6. 49. Azuma, H., Paulk, N., Ranade, A. et  al. Robust expansion of human  ­hepatocytes in Fah‐/‐/Rag2‐/‐/I12rg‐/‐ mice. Nat Biotechnol, 2007; 25:903–10. 50. Washburn, M.L., Bility, M.T., Zhang, L. et al. A humanized mouse model to study hepatitis C virus infection, immune response, and liver disease. Gastroenterology, 2011;140:1334–44. 51. Hasegawa, M., Kawai, K., Mitsui, T. et  al. The reconstituted “humanized liver” in TK‐NOG mice is mature and functional. Biochem Biophys Res Commun, 2011;405:405–10. 52. Demetriou, A.A., Whiting, J.F., Feldman, D. et  al. Replacement of liver function in rats by transplantation of microcarrier‐attached hepatocytes. Science, 1986;233:1190–2.

1035

53. Chen, A.A., Thomas, D.K., Ong, L.L. et al. Humanized mice with ectopic artificial liver tissues. Proc Natl Acad Sci U S A, 2011;108:11842–7. 54. Stevens, K.R., Scull, M.A., Ramanan, V. et al. In situ expansion of e­ ngineered human liver tissue in a mouse model of chronic liver disease. Sci Transl Med, 2017;9. 55. Niu, D., Wei, H.‐J., Lin, L. et al. Inactivation of porcine endogenous retrovirus in pigs using CRISPR‐Cas9. Science, 2017;357:1303–7. 56. Wu, J., Platero‐Luengo, A., Sakurai, M. et al. Interspecies chimerism with mammalian pluripotent stem cells. Cell, 2017;168:473–86.e15. 57. Kwong, G.A., von Maltzahn, G., Murugappan, G. et al. Mass‐encoded synthetic biomarkers for multiplexed urinary monitoring of disease. Nat Biotechnol, 2013;31:63–70. 58. Zhou, Q., Patel, D., Kwa, T. et al. Liver injury‐on‐a‐chip: microfluidic co‐ cultures with integrated biosensors for monitoring liver cell signaling during injury. Lab Chip, 2015;15:4467–78.

80

Pluripotent Stem Cells and Reprogramming: Promise for Therapy James A. Heslop and Stephen A. Duncan Department of Regenerative Medicine and Cell Biology, Medical University of South Carolina, Charleston, SC, USA

INTRODUCTION The liver is a functionally complex organ with multiple critical roles [1]. As such, the development of a cell‐based system to model these functions in health and disease has been an on‐ going priority for basic scientists, pharmaceutical companies, and clinicians alike. Primary human hepatocytes retain hepatocyte function during the first few hours of culture. However, these cells are ­difficult to maintain, and under normal culture conditions lose hepatocyte characteristics as the cells dedifferentiate [2]. Approaches have now been described that extend the differentiated state, but performing genetic manipulations with such cells remains a challenge [3–7]. Moreover, gaining access to hepatocytes from patients with a given liver disease can be difficult to source and the material is often limited. The small window of usability, donor variability, and limited access to rare genotypes have led researchers to use different culture systems as models for hepatocyte function. Established hepatoma cell lines may be expanded indefinitely and are amenable to genetic manipulation, opening up the potential for material‐intensive experiments. However, transformed cells commonly have genetic aberrations and lack many of the key attributes of a hepatocyte. Consequently, the information derived from these cells has well‐accepted caveats and, as cancer cells, cannot be used as a basis for cell‐replacement therapy. Therefore, there is a need for a physiologically relevant and expandable model of hepatic function, which is diploid and can  be genetically manipulated. Pluripotent stem cell‐derived

hepatocytes fulfill these criteria. Pluripotent stem cells are defined by their capacity to form the three developmental germ layers from which all the cells of the body are formed: ectoderm, endoderm, and mesoderm. Human embryonic stem cells (ESCs) are derived from the inner cell mass of the blastocyst and were first isolated in 1998 [8], adapting previously established protocols for deriving mouse ESCs [9–12]. In 2007, adult human somatic cells were reprogrammed to a pluripotent state by overexpressing transcription factors with known roles in hESCs [13, 14]. These induced pluripotent stem cells (iPSCs) display all of the major hallmarks of a hESCs and, importantly, faithfully capture the genotype of the donor. As pluripotent stem cells, hESCs and iPSCs can be expanded and reproducibly differentiated to a hepatic phenotype. Moreover, by deriving iPSCs from patients with genetic conditions of interest, it is possible to recapitulate the patient’s disease in culture. The introduction of disease‐causing mutations to pluripotent cells with wildtype genetic backgrounds is also possible using genome editing techniques [15–17]. Genome editing is a useful tool when attempting to study extremely rare diseases or when comparing the impact of a range of mutations within the same background genotype. Moreover, the same techniques have been used to correct disease‐causing mutations; thus, gene‐corrected iPSCs provide important control cells and  are also a potential isogenic transplant option for cell‐­ replacement therapy [18, 19]. In this chapter, we discuss how pluripotent stem cells have been used to address liver disease and the potential applications of this model in future work.

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



80:  Pluripotent Stem Cells and Reprogramming: Promise for Therapy

GENERATING THE CELLS OF THE LIVER FROM PLURIPOTENT STEM CELLS In order to investigate liver function, the pluripotent cells must first be differentiated into the cell type of interest (Figure 80.1). Building on knowledge derived from mouse studies, the most common differentiation protocols mimic the known stages and  signaling cascades that drive liver development during embryogenesis. Hepatocytes account for many of the functions associated with the liver. As such, the majority of published protocols for deriving liver cells have focused on the generation of hepatocyte‐like cells [20–24]. In general, these protocols follow three stages: definitive endoderm formation, hepatic specification, and hepatocyte maturation. The resulting cells exhibit many hallmarks of primary hepatocytes, including albumin secretion and APOB production. Notwithstanding the acquisition of these and many other hepatocyte functions, they do not completely recapitulate the entire repertoire of functions found in the adult liver. For example, they commonly display a reduced capacity for xenobiotic metabolism, and in this regard resemble fetal, rather than adult, liver cells [25, 26]. Enhancing the maturity of hepatocyte‐like cells is a significant focus of the field, particularly for use in toxicological screens during drug discovery [27]. In recent years, several publications have reported cholangiocyte differentiation protocols [28–31]. In general, bipotent hepatic progenitor cells are first derived using the early stages of established hepatocyte differentiation protocols. These cells are then directed to a biliary fate using lineage‐specific differentiation cues. The resulting cells display several cholangiocyte‐ specific functions, including bile acid transfer and alkaline phosphatase activity [28, 31].

1037

Stellate and liver endothelial cells can also be generated from pluripotent stem cells. First, pluripotent cells are differentiated into mesodermal cells. Cell‐specific markers and optimized maturation conditions are then used to derive populations of functional stellate‐like and liver endothelial‐like cells [32–34]. Furthermore, pluripotent stem cell-derived monocyte‐like cells are a potential model for Kupffer cell function [35]. The described cell types account for the majority of conditions associated with the liver, and thus researchers now have an extensive pluripotent stem cell‐derived toolkit at their disposal in the search for novel therapies for liver disease. An alternative to deriving individual populations of specific liver cell types is the organoid culture system. This system was initially described using pluripotent stem cell-derived hepatocytes alongside established stromal and endothelial cell lines [36, 37]. Remarkably, when cultured together, these cells formed a ­vascularized structure that resembles the liver bud, which is an embryonic structure consisting of multiple cell types that ­ultimately develops into the liver. The authors have subsequently shown that established cell lines can be replaced using pluripotent stem cell‐derived cultures, representing stromal and endothelial progenitor cell populations [38]. When cultured together, the progenitor populations mature and form a vascularized liver bud structure. The entirely pluripotent stem cell-derived organoid model increases the capacity for scaled‐ up culture, potentially facilitates high‐throughput disease ­modeling screens, and, in the long term, enhances the potential for isogenic cell transplant therapy [39]. Outside of hepatocyte‐like cells, pluripotent stem cell-derived nonparenchymal cells of the liver remain in their infancy. The further use and development of these protocols, as seen during the formative years of hepatocyte‐like cell differentiation protocols,

Figure 80.1  Schematic representation of the pluripotent stem cell-derived cell “toolbox” available to researchers and how they can be used to identify novel therapeutic options for liver disease.

1038

THE LIVER:  DERIVING NEW THERAPEUTIC OPTIONS USING PLURIPOTENT STEM CELLS

will no doubt enhance the capacity of these cells and allow for more in‐depth analyses in single and co‐culture systems. Here we describe how pluripotent stem cell-derived cell types have been used to discover novel information, and discuss the hurdles the field must overcome to realize their full potential.

MODELING LIVER DISEASE USING PLURIPOTENT STEM CELLS The first liver diseases that were modeled using iPSC‐derived hepatocytes were monogenic [40]. In these studies, somatic cells derived from patients were reprogrammed to iPSCs and expanded. Following differentiation into hepatocyte‐like cells, disease‐specific assays were used to determine whether the pathophysiology associated with the patient’s liver could be replicated in culture. The range of diseases amenable to modeling is limited by whether the disease has a known etiology, whether the phenotype is reproducible in pluripotent‐derived cells, and whether a suitable assay can be performed to confirm the disease phenotype. Due to the continuing efforts to improve differentiation protocols, many diseases fit these criteria. To date, reports have  shown the successful recapitulation of Alper’s disease, Tangier disease, Wilson’s disease, alpha‐1‐antitrypsin deficiency, familial hypercholesterolemia, glycogen storage disease, and ­ Niemann–Pick type C disease in pluripotent stem cell-derived hepatocytes [19, 41–47]. Moreover, such liver disease models are not limited to hepatocytes, with pluripotent stem cell-derived cholangiocytes successfully recapitulating aspects of Alagille syndrome, polycystic liver disease, and cystic fibrosis [28, 31]. Together, these disease models demonstrate the unique capacity of pluripotent stem cells – to take an often‐rare inherited disorder and allow for the recapitulation of the disease phenotype in an inexhaustible and highly reproducible culture system. Pluripotent stem cell-derived liver cells can also be used to predict disease severity. Identifying treatment regimens for patients with de novo mutations of unknown liability represents a significant challenge in the clinic. By comparing iPSC‐hepatocytes derived from patients with alpha‐1‐antitrypsin deficiency mutations of known severity, studies have been able to recapitulate the impact of these variations in culture [48]. This concept could be employed within the clinic, to ensure that patients with mutations of unknown severity are provided a treatment that is the most compatible with disease progression. Such capacity would be useful across a range of liver conditions, and, in the longer term, may also direct the investigation of specific mutations which predispose individuals to liver diseases. The use of pluripotent stem cell-derived liver cells for the investigation of chronic disease states is less well‐established. The long‐term culture and multicellular models required to study the progression of chronic diseases remain in development. Notwithstanding these challenges, some studies have successfully modeled specific aspects of chronic liver disease using pluripotent‐derived liver cells. By supplementing pluripotent stem cell-derived hepatocytes with fatty acids, researchers have successfully modeled non‐alcoholic fatty liver disease (NAFLD) [49, 50]. These treatments led to lipid droplet formation and

induced the expression of genes associated with NAFLD, allowing the respective groups to derive novel information regarding disease progression [49, 50]. By co‐culturing pluripotent stem cell-derived stellate‐like cells with an established hepatic cell line, one study reported the induction of a fibrotic gene expression profile and the secretion of pro‐collagen type I in response to pro‐fibrogenic stimuli [34]. As these and other chronic liver diseases present an ever‐increasing clinical burden, a better understanding of how these conditions develop will be a crucial part of developing strategies to prevent and alleviate disease progression. Pluripotent stem cell-derived hepatocytes have also been used in genome‐wide association studies (GWAS) [51, 52]. GWAS uses large cohorts of patients with a complex human trait to define allelic variations of functional significance. Two recent studies using GWAS were able to uncover novel information about lipid metabolism, validating the pluripotent‐based system for use in future studies [51, 52].

DERIVING NEW THERAPEUTIC OPTIONS USING PLURIPOTENT STEM CELLS Pluripotent stem cells have not only allowed researchers to investigate the mechanisms of disease states, but also to identify new targets for therapeutic intervention. The studies of Alpers and Tangier’s diseases are two examples of how mechanistic investigation of disease models can result in therapeutically valuable information [42, 44]. Patients with Alpers disease are commonly treated with valproic acid for neurological symptoms, but are known to have an increased susceptibility to valproic acid‐induced hepatotoxicity. Using iPSC‐ hepatocytes carrying the disease mutation, the authors were able to identify superoxide flashes in the mitochondria as the cause of the increased toxicity [44]. Moreover, they were able to alleviate this phenotype in vitro using small molecules. Tangier disease is a condition manifesting in abnormal lipoprotein and triglyceride plasma concentrations. Analyses of the transcriptional profile of Tangier disease iPSC‐hepatocytes revealed ­significantly elevated expression of mRNAs encoding angiopoietin‐related protein 3 (ANGPTL3) [42]. ANGPTL3 is known to impact the concentration of plasma triglycerides and therefore may contribute to the abnormal levels observed in patients with Tangier disease. Taken together, mechanistic evaluation of these diseases identified novel therapeutic targets which could result in the improved treatment of these conditions in the clinic. Another method of deriving clinically relevant information is through the use of high‐throughput screens. Mechanistic studies of disease focus on systematically identifying new targets and rely on current understanding; conversely, phenotypic screens of drug libraries can identify compounds which target unknown or previously inaccessible parts of the disease process. This approach has been successfully used to identify novel treatments for familial hypercholesterolemia and alpha‐1‐antitrypsin deficiency [45, 53]. Following the establishment and characterization of the familial hypercholesterolemia model, pluripotent stem cell-derived hepatocytes were screened using a  repurposed drug library for compounds which could lower



80:  Pluripotent Stem Cells and Reprogramming: Promise for Therapy

APOB levels – a core protein required for the formation of cholesterol [43, 53]. This system identified cardiac glycosides as a class of drugs that could reproducibly lower cholesterol in vitro. The authors confirmed that this was also true in humanized mouse models and clinical samples. The identification of drugs to alleviate alpha‐1‐antitrypsin deficiency used high‐throughput imaging to identify the intracellular accumulation of misfolded alpha‐1‐antitrypsin protein [45]. High‐throughput screening of iPSC‐hepatocytes was able to identify five drugs which significantly reduced the accumulation of misfolded protein. These high‐throughput screens were possible as the disease state had a phenotype that was amenable to a high‐throughput screen. Ensuring that a greater range of diseases are compatible with such screens will no doubt lead to an increase in the translational output of iPSC‐hepatocytes. Importantly, follow‐up studies investigating an identified compound’s mechanism of action can be challenging, particularly when the drug target is not immediately evident. Difficulty in unravelling the mechanism of action could delay the translatability of the discovery to the clinic.

PLURIPOTENT STEM CELLS FOR CELL REPLACEMENT THERAPY Another potential use of pluripotent stem cell-derived liver cells is as a source of cells for transplant therapy. Transplantation of primary hepatocytes has been shown to improve outcome associated with a number of liver diseases [54, 55]; however, the large number of cells required to achieve efficacy is prohibitive as primary human hepatocytes are hard to source and do not sufficiently proliferate in culture. The indefinite expansion and high purity differentiation capacity of pluripotent cells could theoretically provide the numbers of cells required to perform cell‐replacement therapy. Moreover, by generating iPSCs from patients with genetic ­conditions, genome editing techniques can be used to correct ­disease‐causing mutations, before expanding the cells for differentiation and transplant [18, 19, 45]. Such treatment pipelines would negate issues regarding transplant rejection, as the ­corrected cells come from the eventual recipient. Regardless of the therapeutic potential held by pluripotent stem cell-derived cells, many hurdles remain before clinical application is achievable [56]. The most significant hurdle for cell‐replacement therapy is the low engraftment, survival, and expansion efficiency of pluripotent stem cell-derived hepatocytes, when compared to primary cells using the same techniques [57–62]. Various models have been utilized to study engraftment, including different animal strains with varying disease or injury types. The most successful engraftment rates are seen in acute chemical injury models [23, 24, 57, 60, 63, 64], suggesting that the associated inflammatory and proliferative microenvironment may be more amenable to engraftment. Alternative solutions include extrahepatic transplantation of liver “seeds” – small organoid structures in degradable hydrogels [65]. This approach would be of particular use in architecturally damaged liver tissues associated with chronic diseases and could feasibly be employed using the aforementioned liver bud system derived from pluripotent cells [38, 39].

1039

Furthermore, there is continued concern regarding the oncogenic potential of pluripotent stem cell-derived cells. In their pluripotent state, the cells form teratomas upon transplantation. As such, any cells which have not fully undergone differentiation can form unwanted and potentially harmful cell growths [66–69]. As a further concern, during culture, and particularly during reprogramming and gene editing, cells are known to pick up chromosomal aberrations and mutations [56, 70–75]. Therefore, careful monitoring of cell karyotype and mutations in known oncogenic regions, such as P53, is required to minimize tumorigenic potential.

ALTERNATIVES TO PLURIPOTENT CELLS By introducing hepatic‐enriched transcription factors, researchers have been able to transdifferentiate fibroblasts directly into hepatocyte‐like cells [76–78]. These cells appear to have maturity similar to that of pluripotent stem cell-derived hepatocytes. However, the primary advantage of these techniques is the reduced time and cost of transdifferentiation compared with reprogramming, maintaining, and differentiating iPSCs. As such, comparable hepatic disease models can be rapidly generated from patient sources with significantly reduced resources, time, and financial outlay, when compared to pluripotent stem cells. Importantly, transdifferentiation could provide a method to derive theoretically safer hepatocyte‐like cells for use in cell‐ replacement therapy. The direct reprogramming method removes the need for a pluripotent expansion stage and therefore reduces the adverse capacity to form teratomas and other cancerous growths. Transdifferentiation of patient or donor cells directly to hepatocytes also reduces the time required in vitro, reducing the mutagenic selection pressures associated with culture conditions [72, 73]. Interestingly, one group reported transdifferentiated hepatocytes could repopulate a mouse liver at relatively high efficiency (30%) [76]; however, a recent study comparing pluripotent stem cell-derived and transdifferentiated cells from the same donor resulted in similar repopulation capacities (0–5%) [79]. As such, despite the potential advantages of transdifferentiation, engraftment remains a hurdle that all hepatocyte‐like cells are currently unable to reproducibly overcome. Furthermore, in general, primary hepatocytes are particularly difficult to culture as they currently cannot be maintained or expanded to the levels required for transplant. Expandable cultures of transdifferentiated hepatocytes have relied on further genetic manipulation that would potentially enhance oncogenic potential following transplantation [76, 77]. Thus, producing the numbers of compatible cells required for engraftment‐based therapies could be prohibitively difficult using available culture systems.

CONCLUSIONS AND FUTURE PERSPECTIVES Pluripotent stem cell‐based studies offer researchers a unique capacity to investigate liver function in health and disease. The

1040

THE LIVER:  REFERENCES

manipulability and inexhaustible supply of high‐purity liver cell populations has allowed researchers to study mechanisms and diseases which are not feasible in other cell models. Despite the contribution that pluripotent stem cells have made to liver‐based studies, it is also essential to address the limitations of the model. Following differentiation, pluripotent‐ derived liver cells acquire many of the functions associated with primary liver cells; however, the expression and functional capacity are often at significantly reduced levels. For example, proteomic and transcriptomic studies have shown that pluripotent‐derived hepatocytes are most similar to fetal hepatocytes, with limited xenobiotic and metabolic functions [25, 26]. Consequently, pluripotent‐derived hepatocytes remain suboptimal for toxicological screens and other experiments requiring full metabolic competency. Indeed, the lack of maturity is likely the most significant factor contributing to the poor engraftment and expansion seen following transplantation. Furthermore, significant variation in hepatocyte‐like cell maturity exists between donors [52, 80–82], which limits the investigation of subtle phenotypes across populations. Because of this, sustained efforts are being made to enhance the maturity and reduce the inter‐donor variation of pluripotent‐ derived cells. Approaches have included screening for new maturation compounds, growth factors, substrates, and polymers, alongside the development of multicellular and 3D culture systems [3, 7, 32, 36, 83–90]. All of these methodologies demonstrate improvements in phenotype maturity and stability; however in vivo‐level functionality has yet to be acquired, and the increased complexity and cost often reduces the capacity to scale‐up experiments as required. As with all available culture models, the pluripotent‐derived model is not perfect; however, while these are important considerations, the lack of a fully matured phenotype does not preclude the use of the pluripotent model to establish novel information. Indeed, pluripotent cells represent an extremely useful tool for researchers looking to identify and pull apart novel aspects of liver function and disease, but, where possible, novel findings should be confirmed using primary cells and animal models. Despite the caveats the differentiation of pluripotent stem cells to liver cell fates offers a valuable system for discovery and could have direct therapeutic benefits in the near future.

REFERENCES 1. Godoy, P., Hewitt, N.J., Albrecht, U. et al. Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non‐parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch Toxicol, 2013;87:1315–530. 2. Heslop, J.A., Rowe, C., Walsh, J. et  al. Mechanistic evaluation of primary human hepatocyte culture using global proteomic analysis reveals a selective dedifferentiation profile. Arch Toxicol, 2017;91:439–52. 3. Roth, A., Maher, S.P., Conway, A.J. et al. A comprehensive model for assessment of liver stage therapies targeting Plasmodium vivax and Plasmodium falciparum. Nat Commun, 2018;9:1837. 4. Shan, J., Schwartz, R.E., Ross, N.T. et al. Identification of small molecules for human hepatocyte expansion and iPS differentiation. Nat Chem Biol, 2013;9:514–20. 5. Bell, C.C., Dankers, A.C.A., Lauschke, V.M. et al. Comparison of hepatic 2D sandwich cultures and 3D spheroids for long‐term toxicity applications: a multicenter study. Toxicol Sci, 2018;162:655–66.

6. Ware, B.R., Durham, M.J., Monckton, C.P., and Khetani, S.R. A cell culture platform to maintain long‐term phenotype of primary human hepatocytes and endothelial cells. Cell Mol Gastroenterol Hepatol, 2018;5:187–207. 7. Khetani, S.R. and Bhatia, S.N. Microscale culture of human liver cells for drug development. Nat Biotechnol, 2008;26:120–6. 8. Thomson, J.A., Itskovitz‐Eldor, J., Shapiro, S.S. et al. Embryonic stem cell lines derived from human blastocysts. Science, 1998;282:1145–7. 9. Martin, G.R. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci U S A, 1981;78:7634–8. 10. Bradley, A., Evans, M., Kaufman, M.H., and Robertson, E. Formation of germ‐line chimaeras from embryo‐derived teratocarcinoma cell lines. Nature, 1984;309:255–6. 11. Evans, M.J. and Kaufman, M.H. Establishment in culture of pluripotential cells from mouse embryos. Nature, 1981;292:154–6. 12. Lerou, P. and Daley, G. Therapeutic potential of embryonic stem cells. Blood Rev, 2005;19:321–31. 13. Takahashi, K., Tanabe, K., Ohnuki, M. et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 2007;131: 861–72. 14. Yu, J., Vodyanik, M.A., Smuga‐Otto, K. et al. Induced pluripotent stem cell lines derived from human somatic cells. Science, 2007;318:1917–20. 15. Ding, Q., Regan, S.N., Xia, Y. et al. Enhanced efficiency of human pluripotent stem cell genome editing through replacing TALENs with CRISPRs. Cell Stem Cell, 2013;12:393–4. 16. Ran, F.A., Hsu, P.D., Wright, J. et  al. Genome engineering using the CRISPR‐Cas9 system. Nat Protoc, 2013;8:2281–308. 17. Mali, P., Yang, L., Esvelt, K.M. et al. RNA‐guided human genome engineering via Cas9. Science, 2013;339:823–6. 18. Yusa, K., Rashid, S.T., Strick‐Marchand, H. et al. Targeted gene correction of α1‐antitrypsin deficiency in induced pluripotent stem cells. Nature, 2011;478:391–4. 19. Maetzel, D., Sarkar, S., Wang, H. et al. Genetic and chemical correction of cholesterol accumulation and impaired autophagy in hepatic and neural cells derived from Niemann‐Pick Type C patient‐specific iPS cells. Stem Cell Rep, 2014;2:866–80. 20. Si‐Tayeb, K., Noto, F.K., Nagaoka, M. et al. Highly efficient generation of human hepatocyte‐like cells from induced pluripotent stem cells. Hepatology, 2010;51:297–305. 21. Hay, D.C., Zhao, D., Fletcher, J. et al. Efficient differentiation of hepatocytes from human embryonic stem cells exhibiting markers recapitulating liver development in vivo. Stem Cells, 2008;26:894–902. 22. Agarwal, S., Holton, K.L., and Lanza, R. Efficient differentiation of functional hepatocytes from human embryonic stem cells. Stem Cells, 2008;26:1117–27. 23. Hannan, N.R.F., Segeritz, C.‐P., Touboul, T., and Vallier, L. Production of hepatocyte‐like cells from human pluripotent stem cells. Nat Protoc, 2013;8:430–7. 24. Cai, J., Zhao, Y., Liu, Y. et al. Directed differentiation of human embryonic stem cells into functional hepatic cells. Hepatology, 2007;45:1229–39. 25. Godoy, P., Schmidt‐Heck, W., Natarajan, K. et al. Gene networks and transcription factor motifs defining the differentiation of stem cells into hepatocyte‐like cells. J Hepatol, 2015;63:934–42. 26. Rowe, C., Gerrard, D.T., Jenkins, R. et al. Proteome‐wide analyses of human hepatocytes during differentiation and dedifferentiation. Hepatology, 2013; 58:799–809. 27. Kia, R., Sison, R.L.C., Heslop, J. et al. Stem cell‐derived hepatocytes as a predictive model for drug‐induced liver injury: are we there yet? Br J Clin Pharmacol, 2013;75:885–96. 28. Sampaziotis, F., de Brito, M.C., Madrigal, P. et al. Cholangiocytes derived from human induced pluripotent stem cells for disease modeling and drug validation. Nat Biotechnol, 2015;33:845–52. 29. De Assuncao, T.M., Sun, Y., Jalan‐Sakrikar, N. et al. Development and characterization of human induced pluripotent stem cell‐derived cholangiocytes. Lab Invest, 2015;95:684–96. 30. Dianat, N., Dubois‐Pot‐Schneider, H., Steichen, C. et al. Generation of functional cholangiocyte‐like cells from human pluripotent stem cells and HepaRG cells. Hepatology, 2014;60:700–14. 31. Sampaziotis, F., de Brito, M.C., Geti, I. et  al. Directed differentiation of human induced pluripotent stem cells into functional cholangiocyte‐like cells. Nat Protoc, 2017;12:814–27.



80:  Pluripotent Stem Cells and Reprogramming: Promise for Therapy

32. Koui, Y., Kido, T., Ito, T. et  al. An in vitro human liver model by iPSC‐ derived parenchymal and non‐parenchymal cells. Stem Cell Rep, 2017; 9:490–8. 33. Coll, M., Perea, L., Boon, R. et al. Directed Differentiation of IPS Cells to Hepatic Stellate Cells. J Hepatol, 2016;64:S710. 34. Coll, M., Perea, L., Boon, R. et al. Generation of hepatic stellate cells from human pluripotent stem cells enables in vitro modeling of liver fibrosis. Stem Cell, 2018;23:101–13.e7. 35. Buchrieser, J., James, W., and Moore, M.D. Human induced pluripotent stem cell‐derived macrophages share ontogeny with MYB‐independent tissue‐ resident macrophages. Stem Cell Rep, 2017;8:334–45. 36. Takebe, T., Zhang, R.‐R., Koike, H. et al. Generation of a vascularized and functional human liver from an iPSC‐derived organ bud transplant. Nat Protoc, 2014;9:396–409. 37. Camp, J.G., Sekine, K., Gerber, T. et al. Multilineage communication regulates human liver bud development from pluripotency. Nature, 2017; 546:533–38. 38. Takebe, T., Sekine, K., Kimura, M. et al. Massive and reproducible production of liver buds entirely from human pluripotent stem cells. Cell Rep, 2017;21:2661–70. 39. Takebe, T., Wells, J.M., Helmrath, M.A., and Zorn, A.M. Organoid center strategies for accelerating clinical translation. Cell Stem Cell, 2018;22:806–9. 40. Pournasr, B. and Duncan, S.A. Modeling inborn errors of hepatic metabolism using induced pluripotent stem cells. Arterioscler Thromb Vasc Biol, 2017;37:1994–9. 41. Rashid, S.T., Corbineau, S., Hannan, N. et al. Modeling inherited metabolic disorders of the liver using human induced pluripotent stem cells. J Clin Invest, 2010;120:3127–36. 42. Bi, X., Pashos, E.E., Cuchel, M. et al. ATP‐binding cassette transporter A1 deficiency in human induced pluripotent stem cell‐derived hepatocytes abrogates HDL biogenesis and enhances triglyceride secretion. EBioMedicine, 2017;18:139–45. 43. Cayo, M.A., Cai, J., DeLaForest, A. et al. JD induced pluripotent stem cell‐ derived hepatocytes faithfully recapitulate the pathophysiology of familial hypercholesterolemia. Hepatology, 2012;56:2163–71. 44. Li, S., Guo, J., Ying, Z. et al. Valproic acid‐induced hepatotoxicity in alpers syndrome is associated with mitochondrial permeability transition pore opening‐dependent apoptotic sensitivity in an induced pluripotent stem cell model. Hepatology, 2015;61:1730–9. 45. Choi, S.M., Kim, Y., Shim, J.S. et al. Efficient drug screening and gene correction for treating liver disease using patient‐specific stem cells. Hepatology, 2013;57:2458–68. 46. Zhang, S., Chen, S., Li, W. et al. Rescue of ATP7B function in hepatocyte‐like cells from Wilson’s disease induced pluripotent stem cells using gene therapy or the chaperone drug curcumin. Hum Mol Genet, 2011;20:3176–87. 47. Ghodsizadeh, A., Taei, A., Totonchi, M. et al. Generation of liver disease‐ specific induced pluripotent stem cells along with efficient differentiation to functional hepatocyte‐like cells. Stem Cell Rev, 2010;6:622–32. 48. Tafaleng, E.N., Chakraborty, S., Han, B. et al. Induced pluripotent stem cells model personalized variations in liver disease resulting from α1‐antitrypsin deficiency. Hepatology, 2015;62:147–57. 49. Lyall, M.J., Cartier, J., Thomson, J.P. et  al. Modelling non‐alcoholic fatty liver disease in human hepatocyte‐like cells. Philos Trans R Soc Lond B Biol Sci, 2018;373. 50. Graffmann, N., Ring, S., Kawala, M.‐A. et al. Modeling nonalcoholic fatty liver disease with human pluripotent stem cell‐derived immature hepatocyte‐ like cells reveals activation of PLIN2 and confirms regulatory functions of peroxisome proliferator‐activated receptor alpha. Stem Cells Dev, 2016; 25:1119–33. 51. Warren, C.R., O’Sullivan, J.F., Friesen, M. et al. Induced pluripotent stem cell differentiation enables functional validation of GWAS variants in metabolic disease. Cell Stem Cell, 2017;20:547–57.e7. 52. Pashos, E.E., Park, Y., Wang, X. et al. Large, diverse population cohorts of hiPSCs and derived hepatocyte‐like cells reveal functional genetic variation at blood lipid‐associated loci. Cell Stem Cell, 2017;20:558–70.e10. 53. Cayo, M.A., Mallanna, S.K., Di Furio, F. et al. A drug screen using human iPSC‐derived hepatocyte‐like cells reveals cardiac glycosides as a potential treatment for hypercholesterolemia. Cell Stem Cell, 2017;20:478–89.e5. 54. Iansante, V., Mitry, R.R., Filippi, C., Fitzpatrick, E., and Dhawan, A. Human hepatocyte transplantation for liver disease: current status and future perspectives. Pediatr Res, 2018;83:232–40.

1041

55. Fagiuoli, S., Daina, E., D’Antiga, L., Colledan, M., and Remuzzi, G. Monogenic diseases that can be cured by liver transplantation. J Hepatol, 2013;59:595–612. 56. Heslop, J.A., Hammond, T.G., Santeramo, I. et al. Concise review: workshop review: understanding and assessing the risks of stem cell‐based therapies. Stem Cells Transl Med, 2015;4:389–400. 57. Yiangou, L., Ross, A.D.B., Goh, K.J., and Vallier, L. Human pluripotent stem cell‐derived endoderm for modeling development and clinical applications. Cell Stem Cell, 2018;22:485–99. 58. Ang, L.T., Tan, A.K.Y., Autio, M.I. et al. A roadmap for human liver differentiation from pluripotent stem cells. Cell Rep, 2018;22:2190–205. 59. Woo, D., Kim, S., Lim, H. et al. Direct and indirect contribution of human embryonic stem cell–derived hepatocyte‐like cells to liver repair in mice. Gastroenterology, 2012;142:602–11. 60. Liu, H., Kim, Y., Sharkis, S., Marchionni, L., and Jang, Y.‐Y. In vivo liver regeneration potential of human induced pluripotent stem cells from diverse origins. Sci Transl Med, 2011;3:82ra39. 61. Basma, H., Soto‐Gutiérrez, A., Yannam, G.R. et  al. Differentiation and transplantation of human embryonic stem cell‐derived hepatocytes. ­ Gastroenterology, 2009;136:990–9. 62. Zhang, L., Shao, Y., Li, L. et al. Efficient liver repopulation of transplanted hepatocyte prevents cirrhosis in a rat model of hereditary tyrosinemia type I. Sci Rep, 2016;6:31460. 63. Tolosa, L., Caron, J., Hannoun, Z. et  al. Transplantation of hESC‐derived hepatocytes protects mice from liver injury. Stem Cell Res Ther, 2015;6:246. 64. Chen, Y., Li, Y., Wang, X. et al. Amelioration of hyperbilirubinemia in Gunn rats after transplantation of human induced pluripotent stem cell‐derived hepatocytes. Stem Cell Rep, 2015;5:22–30. 65. Stevens, K.R., Scull, M.A., Ramanan, V. et  al. In situ expansion of engineered human liver tissue in a mouse model of chronic liver disease. Sci Transl Med, 2017;9:eaah5505. 66. Payne, C.M., Samuel, K., Pryde, A. et al. Persistence of functional hepatocyte‐like cells in immune‐compromised mice. Liver Int, 2011;31:254–62. 67. Andrews, P.W., Matin, M.M., Bahrami, A.R. et  al. Embryonic stem (ES) cells and embryonal carcinoma (EC) cells: opposite sides of the same coin. Biochem Soc Trans, 2005;33:1526. 68. Hong, S.G., Winkler, T., Wu, C. et  al. Path to the Clinic: Assessment of iPSC‐based cell therapies in vivo in a nonhuman primate model. Cell Rep, 2014;7:1298–309. 69. Fujikawa, T., Oh, S.‐H., Pi, L. et al. Teratoma formation leads to failure of treatment for type I diabetes using embryonic stem cell‐derived insulin‐producing cells. Am J Pathol, 2005;166:1781–91. 70. Haapaniemi, E., Botla, S., Persson, J., Schmierer, B., and Taipale, J. CRISPR–Cas9 genome editing induces a p53‐mediated DNA damage response. Nat Med, 2018;24:927–30. 71. Weissbein, U., Plotnik, O., Vershkov, D., and Benvenisty, N. Culture‐induced recurrent epigenetic aberrations in human pluripotent stem cells. PLOS Genet, 2017;13:e1006979. 72. Mayshar, Y., Ben‐David, U., Lavon, N. et al. Identification and classification of chromosomal aberrations in human induced pluripotent stem cells. Cell Stem Cell, 2010;7:521–31. 73. Ben‐David, U., Mayshar, Y., and Benvenisty, N. Large‐scale analysis reveals acquisition of lineage‐specific chromosomal aberrations in human adult stem cells. Cell Stem Cell, 2011;9:97–102. 74. Ruiz, S., Diep, D., Gore, A. et al. Identification of a specific reprogramming‐ associated epigenetic signature in human induced pluripotent stem cells. Proc Natl Acad Sci U S A, 2012;109:16196–201. 75. Kang, X., Yu, Q., Huang, Y. et al. Effects of integrating and non‐integrating reprogramming methods on copy number variation and genomic stability of human induced pluripotent stem cells. PLoS One, 2015;10:e0131128. 76. Du, Y., Wang, J., Jia, J. et al. Human hepatocytes with drug metabolic function induced from fibroblasts by lineage reprogramming. Cell Stem Cell, 2014;14:394–403. 77. Huang, P., Zhang, L., Gao, Y. et al. Direct reprogramming of human fibroblasts to functional and expandable hepatocytes. Cell Stem Cell, 2014;14:370–84. 78. Pournasr, B., Asghari‐Vostikolaee, M.H., and Baharvand, H. Transcription factor‐mediated reprograming of fibroblasts to hepatocyte‐like cells. Eur J Cell Biol, 2015;94:603–10. 79. Gao, Y., Zhang, X., Zhang, L. et al. Distinct gene expression and epigenetic signatures in hepatocyte‐like cells produced by different strategies from the same donor. Stem Cell Rep, 2017;9:1813–24.

1042

THE LIVER:  REFERENCES

80. Kajiwara, M., Aoi, T., Okita, K. et al. Donor‐dependent variations in hepatic differentiation from human‐induced pluripotent stem cells. Proc Natl Acad Sci U S A, 2012;109:12538–43. 81. Heslop, J.A., Kia, R., Pridgeon, C.S. et  al. Donor‐dependent and other ­nondefined factors have greater influence on the hepatic phenotype than the starting cell type in induced pluripotent stem cell derived hepatocyte‐like cells. Stem Cells Transl Med, 2017;6:1321–31. 82. Ortmann, D. and Vallier, L. Variability of human pluripotent stem cell lines. Curr Opin Genet Dev, 2017;46:179–85. 83. Hay, D.C., Pernagallo, S., Diaz‐Mochon, J.J. et  al. Unbiased screening of polymer libraries to define novel substrates for functional hepatocytes with inducible drug metabolism. Stem Cell Res, 2011;6:92–102. 84. Shan, J., Schwartz, R.E., Ross, N.T. et al. Identification of small molecules for human hepatocyte expansion and iPS differentiation. Nat Chem Biol, 2013;9:514–20. 85. Brafman, D.A., Phung, C., Kumar, N., and Willert, K. Regulation of ­endodermal differentiation of human embryonic stem cells through integrin‐ ECM interactions. Cell Death Differ, 2013;20:369–81.

86. Mallanna, S.K., Cayo, M.A., Twaroski, K., Gundry, R.L., and Duncan, S.A. Mapping the cell‐surface N‐glycoproteome of human hepatocytes reveals markers for selecting a homogeneous population of iPSC‐derived hepatocytes. Stem Cell Rep, 2016;7:543–56. 87. Korostylev, A., Mahaddalkar, P.U., Keminer, O. et al. A high‐content small molecule screen identifies novel inducers of definitive endoderm. Mol Metab, 2017;6:640–50. 88. Ogawa, S., Surapisitchat, J., Virtanen, C. et  al. Three‐dimensional culture and cAMP signaling promote the maturation of human pluripotent stem cell‐ derived hepatocytes. Development, 2013;140:3285–96. 89. Gieseck III, R.L., Hannan, N.R.F., Bort, R. et  al. Maturation of induced pluripotent stem cell derived hepatocytes by 3D‐culture. PLoS One, 2014;9:e86372. 90. Rashidi, H., Luu, N.‐T., Alwahsh, S.M. et  al. 3D human liver tissue from pluripotent stem cells displays stable phenotype in vitro and supports compromised liver function in vivo. Arch Toxicol, 2018;92:3117–29.

81

Chromatin Regulation and Transcription Factor Cooperation in Liver Cells Ido Goldstein Institute of Biochemistry, Food Science and Nutrition, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Rehovot, Israel

INTRODUCTION Liver cells constantly receive a multitude of signals in the form of hormones, cytokines, metabolites, etc. Integrating these signals and translating them to provide a coherent and effective response is critical for proper liver function and homeostasis. Many of these signals converge to regulate gene expression via altering rate of transcription. Recent years have seen tremendous advances in our understanding of hepatic gene expression regulation and how it affects hepatic functions. This chapter will summarize recent advances in the emerging field of liver genomics (focusing on transcription and chromatin genomics) as well as on mechanisms of transcription factor (TF) cooperation within the chromatin environment.

REGULATION OF GENE TRANSCRIPTION Transcriptional regulation is a multi‐step process involving TFs,  coactivators, histone‐modifying enzymes, chromatin‐ remodeling complexes, chromatin structural proteins, noncoding RNAs, general transcription factors (GTFs), and RNA polymerase II (RNAP II). The interplay between these factors occurs at the chromatin fiber and is affected by both DNA sequence and chromatin architecture (Figure 81.1). Regions in DNA that mediate transcriptional regulation are termed cis‐regulatory elements. The two prominent types of cis‐ regulatory elements are promoters and enhancers. Confusingly, the term promoter is used in the literature to describe two different types of cis‐regulatory elements – the core promoter and the promoter‐proximal region. The core promoter is a ~35 bp region

residing mostly upstream of the transcription start site (TSS). It harbors DNA sequences facilitating the binding of the basal transcription machinery (RNAP II and GTFs). While sequences of core promoters may vary across genes, the principal purpose of this region is uniform – to initiate RNAP II‐mediated gene transcription [1]. Conversely, the promoter‐proximal region refers to the sequences immediately upstream of the core promoter. There is no clearcut definition of what is considered “proximal” to the promoter and studies have used anywhere between 200 and 5000 bp upstream of the TSS. In essence, the proximal promoter region fulfils most of the characteristics of an enhancer as outlined below (excluding certain histone modification profiles) and could therefore be considered as such. In contrast to the core promoter that serves to initiate transcription, enhancers are regions that determine whether a certain gene  is transcribed and at which rate. Enhancers contain DNA sequences that bind context‐specific TFs (i.e. TFs expressed selectively in a certain tissue or TFs that become active only upon a certain signal). Thereby, while promoters partake in directing the basal transcription machinery to the TSS and initiating transcription, enhancers provide the tissue‐specific and signal‐responsive pattern of gene expression and are the principal effectors of transcriptional regulation [2]. Functionally, enhancers are defined as DNA elements that fulfil these demands: upon activation, they increase the transcription of a linked gene. They can function even at long distances from the core promoter and their function is not dependent on orientation to the core promoter (thus, they can reside either upstream or downstream of the TSS) [2]. Both promoters and enhancers have distinct DNA sequence and chromatin determinants. The core promoter is characterized by DNA elements that mediate the recruitment of the basal transcription machinery [1]. The chromatin environment of the core

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.

1044

THE LIVER:  HOW THE GENOMIC REVOLUTION RESHAPED LIVER RESEARCH

Figure 81.1  Transcriptional regulation. Gene transcription is initiated by binding of RNAP II and GTFs to the core promoter. Regulation of RNAP II recruitment to the core promoter is mediated by TFs that are expressed in a tissue‐specific manner and respond to extracellular and intrinsic cues. TFs bind specific DNA sequences termed motifs located within enhancer regions. Binding of an activating TF (often together with coactivators) and cooperation between TFs (see Figure 81.2) leads to several events culminating in enhanced gene transcription: increase in chromatin accessibility due to chromatin‐remodeling complexes; increase in certain histone modifications due to histone‐modifying enzymes; enhancer–promoter contact due to looping factors, and increased eRNA transcription.

promoter and its proximal region is characterized by enrichment of RNAP II binding, accessible (“open”) chromatin and distinct histone modifications (post‐translational modifications of histone tails). The principal histone modification of promoter regions is H3K4 trimethylation (H3K4me3). Enhancer elements contain a mixture of TF‐binding sequences termed motifs. Variability of motif composition between enhancers is vast and is the basis for context‐dependent gene expression. Most liver enhancers contain motifs for the liver lineage‐determining factors forkhead box A (FoxA) and hepatocyte nuclear factor 4 (HNF4) as well as for CCAAT/enhancer‐ binding protein (CEBP) [3]. Distal enhancers (defined as not directly upstream to the core promoter) reside in accessible chromatin and are marked by H3K4me1, H3K4me2, and H3K427 acetylation (H3K27ac). In addition, the binding of p300 and CBP histone‐modifying enzymes serve as an efficient enhancer marker [2]. RNAP II transcribes enhancer DNA to noncoding RNAs termed enhancer RNAs (eRNAs). eRNAs also accurately mark active enhancers [2]. Altering the activity of context‐specific TFs is the chief channel by which extracellular and cell‐intrinsic inputs regulate gene transcription. Upon activation by an upstream signal, activating TFs bind their motif on DNA and recruit several protein components that promote an “active” chromatin environment and increase transcription: histone acetyl transferases (HATs) that covalently mark histones with acetyl marks associated with gene activation, chromatin‐remodeling complexes that de‐compact chromatin, and structural proteins that mediate promoter– enhancer contacts via looping (Figure  81.1). Cellular signals can also affect repressor TFs that mediate the opposite processes (i.e. chromatin compaction, histone deacetylation, and histone methylation), leading to decreased transcription.

HOW THE GENOMIC REVOLUTION RESHAPED LIVER RESEARCH Technologies involving high‐throughput sequencing to profile the different aspects of transcriptional regulation have dramatically

changed the field. A mainstay of genome‐wide TF profiling is chromatin immunoprecipitation followed by sequencing (ChIP‐ seq). In ChIP‐seq, a chromatin‐resident protein is precipitated by an antibody along with crosslinked DNA which is then deeply sequenced. consequently, the regions in the genome bound by the protein (commonly termed “binding sites”) are highly enriched in the sequencing results. ChIP‐seq is used to profile the genome‐ wide binding profile of a TF or the occupancy of modified histones. Because the accessibility of chromatin to TF binding is an important factor in transcriptional regulation, profiling chromatin accessibility in a genome‐wide manner has become very popular. In principle, chromatin is partially digested with a nuclease (DNase, transposase, etc.) and accessible regions are more readily liberated from compacted chromatin and are sequenced more f­ requently in the subsequent step of high‐throughput sequencing. Clearly, high‐ throughput sequencing of steady state RNA levels (RNA‐seq) or actively transcribed RNA (GRO‐seq) are instrumental in transcriptomic profiling. These and other genomic t­echniques discussed here are summarized in Table 81.1. The liver has been a focus of genomic research from its onset, resulting in many insights pertaining to liver biology. In the genomic era, liver features and liver‐related pathways are being re‐addressed with modern tools and re‐evaluated, leading to ­surprising discoveries. An example of a genomic viewpoint reshaping our understanding of liver transcriptional networks deals with the genome‐wide effects of thyroid hormone [4]. Thyroid hormone is central to regulation of systemic metabolism of cholesterol, triglycerides, and carbohydrates, with the liver mediating many of its effects [5]. Following activation by triiodothyronine (T3, the active form of the hormone), thyroid hormone receptor beta (TRβ) regulates genes in hepatocytes that exert many of the hormone’s effects [5]. The prevalent model of TRβ‐dependent gene regulation was that prior to hormone activation, TRβ is bound to DNA in a repressive complex together with corepressors. Following binding to T3, the receptor was thought to undergo a conformational change that favors association with coactivators and displacement of corepressors, thereby leading to gene induction [5]. However, a study mapping the genome‐wide effects of thyroid hormone revealed a more  complex scenario [4]. Using mouse models mimicking



81:  Chromatin Regulation and Transcription Factor Cooperation in Liver Cells

1045

Table 81.1  Common genomic techniques used to profile chromatin and transcriptional regulation Technique

Full name

Principle

RNA‐seq GRO‐seq ChIP‐seq

RNA sequencing Global run‐on assay sequencing Chromatin immunoprecipitation sequencing DNase sequencing Assay for transposase‐accessible chromatin sequencing Bivariate genomic footprinting

RNA is isolated, cDNA is synthesized and sequenced Actively transcribed RNA is labeled, immunoprecipitated, and sequenced Chromatin is crosslinked to fixate chromatin‐bound proteins (usually TFs or modified histones) that are immunoprecipitated. Co‐precipitated DNA is sequenced Chromatin is partially digested with DNase, accessible chromatin is isolated and sequenced Transposase inserts sequencing adapters to accessible chromatin regions which are isolated, amplified, and sequenced DNase‐seq or ATAC‐seq data is analyzed for differences in footprint depth and motif‐flanking accessibility between two experimental conditions

DNase‐seq ATAC‐seq BaGFoot

hypo‐ and hyperthyroidism, the authors found many T3‐induced genes that were accompanied with nearby increases in chromatin accessibility (measured by DNase‐seq, Table 81.1). Profiling TRβ binding (via ChIP‐seq, Table 81.1) revealed both preexisting TRβ‐binding sites (i.e. TRβ is bound prior to T3 treatment) and, surprisingly, a sizable group of T3‐dependent TRβ‐binding sites. These results show that, from a genomic perspective, at least two modes of action are plausible; one aligns with the prevalent model in which TRβ is bound to chromatin regardless of hormone. The second mechanism suggests that, at a subset of enhancers, thyroid hormone promotes both TRβ binding and increases in chromatin accessibility. These two models are not mutually exclusive and it is possible that each mode of action is responsible for a different set of TRβ‐regulated genes and for a different metabolic outcome of thyroid hormone. Another example of how a genomic perspective can lead to re‐ evaluation of liver biology was provided by a study examining the transcriptional regulation of the liver’s circadian clock. The authors mapped the genome‐wide landscape of mouse liver eRNAs and their altering levels across the ~24 hour circadian cycle [6]. Mapping transcribed RNA (by GRO‐seq, Table 81.1), the authors found thousands of hepatic eRNAs showing altered levels throughout the circadian cycle. These “circadian eRNAs” marked enhancers involved in circadian gene expression and were in phase with nearby circadian genes; that is, the levels of both eRNA and the RNA of the linked gene were correlated. Examining the TF motifs in enhancers marked by circadian eRNAs revealed the dominant TF at each circadian phase that directs gene expression at the given phase [6]. Collectively, these findings show that dynamic enhancer activity can be efficiently measured by eRNA levels and is indicative of circadian gene expression. Although ChIP‐seq is an invaluable tool in genomic studies of TF function, it has a major pitfall. ChIP‐seq is a biased approach where the experiments are focused on one or a few TFs. A more general tool, termed genomic footprinting, has been used to predict TF binding in an unbiased manner. Upon binding to DNA, TFs protect the motif they bind from being digested by a nuclease (e.g. DNase) while the surrounding sequence is more readily digested [7]. Because all TFs bind DNA, it was assumed that all TFs leave a footprint and therefore all TF‐binding events can be measured by the presence of a motif footprint in genome‐wide chromatin accessibility assays (DNase‐seq and ATAC‐seq, Table 81.1). However, assaying TF footprints in mouse liver and in other cell types revealed that 80% of motifs do not show a detectable footprint [8]. Although the reason for that observation is unclear, it became clear that footprinting alone is insufficient in detecting the activity of most TFs. To better predict TF activity

from chromatin accessibility assays, a computational tool was designed, termed “bivariate genomic footprinting” (BaGFoot, Table  81.1). This method does not rely on absolute values of accessibility that are difficult to reliably determine. Rather, it measures the change between two experimental conditions. BaGFoot measures a change in two variables, the first is footprint depth and the second is chromatin accessibility surrounding the motif. Measuring the change rather than absolute values in addition to measuring two variables led to better prediction of TF activity. To exemplify the advantages of the method, the change in the two variables was measured following fasting in mouse liver. Of the 20 TFs known to regulate the fasting response in liver [9], BaGFoot was able to narrow down the list to three main TFs regulating it at the genome‐wide level. Thus, an unbiased way to assay changes in liver chromatin led to a better understanding of the major TFs regulating the fasting response, a finding that was further utilized to investigate fuel production in the liver following fasting [10]. Taken together, the three examples show how a genome‐wide perspective combined with advanced computational tools and integrated analyses of different aspects of transcriptional regulation can advance liver research.

HEPATIC ENHANCERS – NODES OF DYNAMIC TRANSCRIPTION FACTOR CROSSTALK A major conclusion arising from genomic studies is that TF binding is biased toward enhancer regions and that often, multiple TFs are bound in a given enhancer at a given biological state. Indeed, numerous analyses of hepatic enhancers reveal groups of motifs that are commonly enriched in different types of enhancers. Examining the entirety of liver enhancers reveals the high occurrence of FoxA, HNF, and CEBP motifs [3], while subtypes of enhancers activated following a certain signal give a more context‐dependent set of motifs (for examples see [6, 10–12]). These motif co‐occurrences are also observed when comparing ChIP‐seq profiles of different TFs. For example, extensive overlap was seen in liver between liver X receptor (LXR), peroxisome proliferator‐activated receptor alpha (PPARα), retinoid X receptor (RXR), CEBPα, and HNF4α [13] as well as between farnesoid X receptor (FXR), retinoic acid receptor alpha (RARα), CEBPβ, and HNF4α [14]. Overlap of TF binding in space and time suggests a high degree of TF crosstalk in regulating liver functions. Evidence supporting this notion came from a study examining TF cooperation by using

1046

THE LIVER:  MECHANISMS OF TRANSCRIPTION FACTOR COOPERATION

a massively parallel reporter assay in mouse liver. In that setting, synthetic DNA regulatory elements containing various combinations of TF motifs led to stronger gene expression than motifs containing either one or two types of TF motifs (the total number of motifs per enhancer was kept the same across groups) [15]. The liver is constantly bombarded by endocrine, cytokine, metabolite, and neuronal signals, many of which affect gene expression. Nevertheless, the healthy liver is able to produce a coherent response to these signals and maintain homeostasis. Cooperation between TFs at the chromatin template as suggested by the findings described above is an efficient way to integrate and converge different signals and to produce a relevant context‐dependent outcome. Below I detail the various mechanisms and emerging concepts of TF cooperation as a prevalent mechanism to integrate extracellular and intrinsic

signals in liver cells. Although the suggested mechanisms are universal and some were initially described in other cell types, I will focus on studies examining them in liver cells.

MECHANISMS OF TRANSCRIPTION FACTOR COOPERATION Heterodimerization Probably the most straight forward manner by which TFs can cooperate is direct protein–protein interaction. Heterodimerization refers to two different TFs that interact physically and together bind DNA as a dimer, each to its own motif (Figure 81.2a). Within

Figure 81.2  Mechanisms of transcription factor cooperation. TFs cooperate within enhancers to promote gene expression by several mechanisms. (a) Heterodimerization – two TFs physically interact and together bind DNA, each to its own motif. (b) Tethering – two TFs physically interact; one binds its motif while the other is brought to the enhancer by the interaction with its partner TF. (c) Coactivator – A TF binds a scaffolding protein with no specific DNA‐binding capacity that mediates recruitment of regulatory components. (d) Squelching – A TF interacts and isolates its partner from the enhancer, leading to gene repression. (e) Assisted loading – following a signal, a TF binds the enhancer and recruits regulatory components leading to enhancer activation (increased accessibility and enrichment of certain histone marks). In the presence of a second signal activating a different TF, the second TF would bind more efficiently to the enhancer, leading to synergistic gene expression. (f) TF cascade – the first TF induces a gene encoding the second TF; following gene translation, the second TF becomes active and initiates a distinct gene program.



81:  Chromatin Regulation and Transcription Factor Cooperation in Liver Cells

the nuclear receptor superfamily of TFs there are several prominent cases of heterodimerization events relevant to liver biology. RXR is an obligate heterodimerization partner for several nuclear receptors: FXR, LXR, PPARs, RAR, pregnane X receptor (PXR), vitamin D receptor (VDR), and constitutive androstane receptor (CAR). These interactions are essential to liver physiology and are the basis for several hepatic gene expression programs [16]. Heterodimerization of RXR with different partners at the same enhancer leads to different outcomes depending on the partner. Indeed, evidence in liver suggest that RXR partners bind the same region under different circumstances, sometimes with opposite downstream effects on gene expression [13, 17]. Another prominent example of heterodimerization is the bZIP superfamily of TFs. bZIPs readily heterodimerize, leading to specific gene programs. Various cases of heterodimerization within bZIPs have been documented [18], many of which are relevant to liver biology [19, 20]. Activator protein 1 (AP‐1) is the prototypical bZIP heterodimer. It can consist of various heterodimerization partners, one from the Jun group of proteins and the other from the Fos group. The choice of dimerization partners has implications for AP‐1 downstream gene program. An example of how different AP‐1 proteins affect liver biology was discovered following the observation that the levels of Fra‐1 (a Fos protein) are reduced in the livers of mice fed a high‐fat diet [20]. Reintroduction of Fra‐1 to the liver results in reversal of diet‐induced non‐alcoholic fatty liver disease (NAFLD). The effect of Fra‐1 on liver lipid metabolism is through its repression of the lipid regulator PPARγ. Conversely, when AP‐1 comprises c‐Fos, PPARγ is induced, resulting in NAFLD. Thus, different heterodimers can lead to opposite effects with detrimental health implications. The choice and availability of heterodimerization partners is a potential regulatory step affecting gene programs. In the case of obligate partners such as RXR, the regulatory step might be the levels of available partner. For example, if under certain conditions, RXR favors dimerization with a certain partner, the other RXR partners are effectively deactivated because they cannot bind DNA without RXR. In contrast to obligate partners, modular heterodimerization such as in the case of bZIPs may be a mechanism for signal integration. Namely, each partner is activated by a different signal and only in the presence of the two signals can the partners dimerize and promote a transcriptional program reflecting the two upstream signals. Lastly, heterodimerization might provide tissue specificity and signal responsiveness of gene expression if one partner is only expressed in certain tissues while the second is activated by a certain signal.

Tethering, coactivators, and squelching The concept of tethering also requires protein–protein interaction between TFs. However, in tethering, only one TF is bound to DNA while the other is brought to the enhancer region via its interaction with the motif‐bound TF (Figure 81.2b). This mechanism is thought to allow complexity in gene regulation programs because TFs can exert an effect without the presence of their motif in the enhancer. Although tethering is not a new model, its biological relevance remains controversial. Still, several evidence suggest that a tethering mode of action exists in certain scenarios [21].

1047

A similar mechanism of action is seen in the case of coactivators (i.e. proteins that do not bind DNA directly but form a complex with TFs and augment their activity) (Figure  81.2c). The term coactivator is most commonly ascribed to non‐TF proteins that lack enzymatic activity and promote TF activity by recruiting regulatory and structural proteins (such as histone‐modifying enzymes and looping proteins that mediate promoter–enhancer contacts). The prominent liver‐related coactivators are PPARγ coactivator 1 (PGC1), cAMP responsive element‐binding protein regulated transcription coactivator 2 (CRTC2), and the steroid receptor family of coactivators (SRCs). These proteins play crucial roles in transcriptional regulation in response to nutritional status and have been extensively reviewed elsewhere [22–24]. Intriguingly, tethering could lead to deactivation of a TF via disrupting TF–coactivator interaction. A report examining hepatic autophagy found a role for cAMP responsive element‐ binding protein (CREB) in promoting autophagy during fasting. Under fed conditions, FXR was found to suppress hepatic autophagy via its interaction with CREB. FXR interrupted with a CREB‐dependent autophagic gene program by directly binding CREB, leading to the dissociation of its coactivator CRTC2 [25]. A different case of repressive tethering was reported. There, HNF6 tethered the Rev‐erbα repressor, leading to gene repression [26]. Gene repression can also be mediated by “squelching,” whereby the interacting TF deactivates its partner by removing it from DNA (Figure 81.2d) [27].

Assisted loading and facilitated repression The mechanisms of TF crosstalk described earlier all depend on direct protein–protein interactions. An alternative model, termed “dynamic assisted loading,” has been suggested whereby TF cooperativity is indirectly achieved via the recruitment of chromatin remodelers and modifiers by one TF, thereby making the enhancer more accessible to other TFs (Figure  81.2e) [28]. Several biological systems have been shown to involve assisted loading (reviewed in [7]). Specifically, three studies portray assisted loading as a major mechanism for hepatic transcriptional regulation. The first report dealt with the highly tissue‐specific binding patterns of glucocorticoid receptor (GR). Although GR is ubiquitously expressed in virtually every tissue, its genome‐wide DNA‐binding profile varies considerably between cell types. Comparing five different cell types, the authors found that 83% of hepatic GR‐binding sites are specific to liver and only 0.5% of sites are common among the five examined cell types [12]. This led the authors to suggest that the tissue‐specific pattern of GR binding is determined by TFs differently expressed among tissues. These TFs mediate chromatin remodeling of potential GR‐binding sites, thereby enabling GR binding. Indeed, using ChIP‐seq and DNase‐seq, the report shows that CEBPβ, a liver‐ enriched TF, facilitates the recruitment of GR via assisted loading to a subset of sites in liver while it does not affect the other GR‐binding patterns [12]. This report exemplifies that the activity of ubiquitously expressed TFs such as GR is modulated and fine‐tuned by tissue‐restricted TFs, thereby providing tissue‐ specific transcriptional programs. Two recent studies dealing with the hepatic responses to fasting and to inflammation revealed another function for assisted

1048

THE LIVER:  MECHANISMS OF TRANSCRIPTION FACTOR COOPERATION

loading  –  integration of extracellular signals into a dedicated, context‐dependent transcriptional response. During fasting, the liver dynamically responds to hormonal/metabolic cues, and supplies fuel to extrahepatic tissues. The major hormones elevated during fasting are glucagon and cortisol (corticosterone in rodents). Although many TFs are known to participate in this response [9], a systematic view of the TF network controlling it and of TF crosstalk during fasting was lacking. Aiming to examine this response in a genome‐wide manner without relying on a specific TF, a study profiled changes in chromatin accessibility following fasting in mouse liver [10]. Fasting led to massive reorganization of hepatic chromatin, showing thousands of hepatic enhancers that dynamically respond to fasting. Two prominent TFs occupying many of these enhancers, CREB and GR, synergistically induced a gene program promoting glucose production following glucagon and corticosterone. ChIP‐seq profiles and imaging techniques revealed that GR assists the loading of CREB on a subset of enhancers, thus facilitating synergistic gene expression. This cooperation led to augmented glucose production, an essential hepatic feature during fasting [10]. Thus, during fasting, assisted loading serves to integrate two hormonal cues to facilitate glucose production. Another process where extracellular signals are integrated to elicit a transcriptional response is inflammation. Proinflammatory cytokines (IL‐1β and IL‐6) rapidly induce the expression of hepatic “acute phase” genes that curb infection and tissue damage [29]. Although heavily studied and with clear implications for innate immunity and chronic inflammatory disorders, the regulatory events producing these hepatic immune responses remained elusive. To resolve these questions, the response of primary hepatocytes to combinatorial cytokine treatments was assessed at a few genomic levels – TF binding, histone modifications, and gene expression [11]. These revealed that following IL‐1β stimulus, hepatic NF‐κB is activated and “primes” a discrete group of enhancers; that is, it leads to a chromatin state transition whereby the enhancer is in a more “active” state. STAT3, which is activated by IL‐6, can bind most of its binding sites without priming by NF‐κB. However, 20% of STAT3‐binding sites only show efficient STAT3 binding following NF‐κB priming, resulting in highly synergistic expression of nearby acute phase genes. Conversely, the binding of STAT3 to other enhancers is inhibited by NF‐κB. Thus, these TFs crosstalk in an enhancer‐specific manner, thereby allowing a bifurcated output; some genes are synergistically induced by the two cytokines while others are antagonized. This is another example of signal integration by assisted loading whereby two signal‐activated TFs cooperate to produce a context‐specific gene program. The three studies described above examine assisted loading from a genome‐wide perspective. A recent study examined the transcriptional regulation of FGF21, a hepatokine with overreaching effects on systemic metabolism [30]. Although the term assisted loading was not mentioned, the study presents evidence that is in line with the assisted loading model. In agreement with previous studies, the authors describe that FGF21 is induced both during fasting and following a glucose challenge. This regulation is mediated by the unlikely alliance of the fasting‐related TF PPARα and carbohydrate‐responsive element‐ binding protein (ChREBP), a feeding‐related TF. Both TFs are needed for FGF21 induction as shown by knockout mouse

models. Importantly, ChREBP binding to the promoter‐proximal region of FGF21 is facilitated by PPARα. PPARα acts to increase chromatin accessibility and histone acetylation, a probable mechanism mediating ChREBP binding and eventually increasing RNAP II occupancy and gene transcription [30]. Although the assisted loading model was mostly attributed to gene induction, a recent report describes a similar mechanism that acts to repress genes, thereby setting the hepatic circadian rhythm [31]. The circadian clock and its rhythm are mediated by transcriptional activators (RAR‐related orphan receptor alpha [RORα] and BMAL1) and the Rev‐erbα repressor. A commonly accepted model posits that RORα and Rev‐erbα compete for binding and thus mediate rhythmicity. A recent study portrays a different scenario termed “facilitated repression.” RORα and its coactivator SRC2 bind circadian enhancers and recruit chromatin‐remodeling complexes that in turn lead to increased chromatin accessibility. Then, Rev‐erbα binds these regions more readily and initiates chromatin compaction, bringing about gene repression. These cycles of TF binding mediating rhythmic chromatin compaction and decompaction result in the circadian rhythm of gene expression [31]. A study dealing with the transcriptional control of hepatic autophagy suggests a mechanism of repression that is in line with a facilitated repression mode of action. The authors describe an antagonistic relationship between the autophagy‐ promoting effects of PPARα during fasting and the autophagy‐ inhibitory actions of FXR during feeding [17]. PPARα binds next to autophagy‐related genes, leading to gene induction, while FXR inhibits the expression of these genes. Importantly, the authors show that this antagonism between the two TFs is manifest at the chromatin template in autophagy‐related enhancers. Intriguing evidence distinctly show that while PPARα promotes a chromatin environment favoring transcription (increased p300 binding and H4 acetylation), FXR “shuts down” the enhancer by recruitment of corepressors and increasing the H3K27me3 repressive mark. Thus, FXR may facilitate repression of autophagy genes via an unknown repressor. The assisted loading and facilitated repression modes of action hold a significant advantage over heterodimerization or tethering. It does not require dedicated protein‐interaction domains because there is no need for physical interaction between the cooperating TFs. Therefore, the evolutionary constraints on establishing a cooperative mode of action between two TFs is significantly lower. Thus, I trust that assisted loading and facilitated repression will emerge as a principal mechanism for TF cooperation.

TF cascades A TF cascade is an indirect type of cooperation between TFs. In a TF cascade, one TF induces the expression of a second TF by promoting the transcription of its encoding gene. Thus, after a few hours from the initial signal activating the first TF, the amount of the second TF is increased, affecting secondary gene programs (Figure  81.2f). Such cascades are prevalent in liver physiology. A TF cascade with far‐reaching effects on liver biology was mentioned earlier whereby AP‐1 can lead to either induction or repression of PPARγ, a TF that promotes lipogenesis [20]. In another case, GR was shown to induce the expression



81:  Chromatin Regulation and Transcription Factor Cooperation in Liver Cells

of PPARα [32, 33]. It seems that this cascade is manifest during fasting as evidence shows that corticosterone activates GR after 10 hours of fasting in mice. This results in induction of PPARα and an increase in its activity during prolonged fasting [10]. However, the possibility of an assisted loading model between GR and PPARα is also plausible in light of evidence showing that PAPRα activation promotes GR binding at some sites in the genome [34]. For more cases of hepatic TF cascades the readers are referred to more focused reviews [9, 35].

SUMMARY The liver is constantly presented with various extracellular cues. Metabolite, immune, endocrine, xenobiotic, bacterial, and neural signals all affect hepatic gene expression. Altering gene expression patterns allow the liver to dynamically respond to these signals and maintain homeostasis. Indeed, many liver‐related pathologies such as obesity, diabetes, NAFLD, liver cancer, and chronic inflammation were shown to involve dysregulation of gene expression. Thus, understanding the mechanisms of TF cooperation and signal integration in the liver is expected to lead to novel treatment modalities.

REFERENCES 1. Smale, S.T. and Kadonaga, J.T. The RNA polymerase II core promoter. Annu Rev Biochem, 2003;72:449–79. 2. Kim, T.K. and Shiekhattar, R. Architectural and functional commonalities between enhancers and promoters. Cell, 2015;162(5):948–59. 3. Iwafuchi‐Doi, M., Donahue, G., Kakumanu, A. et al. The pioneer transcription factor FoxA maintains an accessible nucleosome configuration at enhancers for tissue‐specific gene activation. Mol Cell, 2016;62(1):79–91. 4. Grontved, L., Waterfall, J.J., Kim, D.W. et al. Transcriptional activation by the thyroid hormone receptor through ligand‐dependent receptor recruitment and chromatin remodelling. Nat Commun, 2015;6:7048. 5. Mullur, R., Liu, Y.Y., and Brent, G.A. Thyroid hormone regulation of metabolism. Physiol Rev, 2014;94(2):355–82. 6. Fang, B., Everett, L.J., Jager, J. et al. Circadian enhancers coordinate multiple phases of rhythmic gene transcription in vivo. Cell, 2014;159(5):1140–52. 7. Goldstein, I. and Hager, G.L. Dynamic enhancer function in the chromatin context. Wiley Interdiscip Rev Syst Biol Med, 2018;10(1). doi:10.1002/ wsbm.1390. 8. Baek, S., Goldstein, I., and Hager, G.L. Bivariate genomic footprinting detects changes in transcription factor activity. Cell Rep, 2017;19(8):1710–22. 9. Goldstein, I. and Hager, G.L. Transcriptional and chromatin regulation during fasting – the genomic era. Trends Endocrinol Metab, 2015;26(12):699–710 10. Goldstein, I., Baek, S., Presman, D.M. et  al. Transcription factor assisted loading and enhancer dynamics dictate the hepatic fasting response. Genome Res, 2017;27(3):427–39. 11. Goldstein, I., Paakinaho, V., Baek, S., Sung, M.H., and Hager, G.L. Synergistic gene expression during the acute phase response is characterized by transcription factor assisted loading. Nat Commun, 2017;8(1):1849. 12. Grontved, L., John, S., Baek, S. et al. C/EBP maintains chromatin accessibility in liver and facilitates glucocorticoid receptor recruitment to steroid response elements. EMBO J, 2013;32(11):1568–83. 13. Boergesen, M., Pedersen, T.A., Gross, B. et al. Genome‐wide profiling of liver X receptor, retinoid X receptor, and peroxisome proliferator‐activated

1049

receptor alpha in mouse liver reveals extensive sharing of binding sites. Mol Cell Biol, 2012;32(4):852–67. 14. Dubois‐Chevalier, J., Dubois, V., Dehondt, H. et al. The logic of transcriptional regulator recruitment architecture at cis‐regulatory modules controlling liver functions. Genome Res, 2017;27(6):985–96. 15. Smith, R.P., Taher, L., Patwardhan, R.P. et al. Massively parallel decoding of mammalian regulatory sequences supports a flexible organizational model. Nat Genet, 2013;45(9):1021–8. 16. De Cosmo, S. and Mazzoccoli, G. Retinoid X receptors intersect the molecular clockwork in the regulation of liver metabolism. Front Endocrinol, 2017;8:24. 17. Lee, J.M., Wagner, M., Xiao, R. et  al. Nutrient‐sensing nuclear receptors coordinate autophagy. Nature, 2014;516(7529):112–15. 18. Rodriguez‐Martinez, J.A., Reinke, A.W., Bhimsaria, D., Keating, A.E., and Ansari, A.Z. Combinatorial bZIP dimers display complex DNA‐binding specificity landscapes. Elife, 2017;6. 19. Thomsen, M.K., Bakiri, L., Hasenfuss, S.C. et al. JUNB/AP‐1 controls IFN‐ gamma during inflammatory liver disease. J Clin Invest, 2013;123(12): 5258–68. 20. Hasenfuss, S.C., Bakiri, L., Thomsen, M.K. et al. Regulation of steatohepatitis and PPARgamma signaling by distinct AP‐1 dimers. Cell Metab, 2014;19(1):84–95. 21. Ratman, D., Vanden Berghe, W., Dejager, L. et al. How glucocorticoid receptors modulate the activity of other transcription factors: a scope beyond tethering. Mol Cell Endocrinol, 2013;380(1–2):41–54. 22. Stashi, E., York, B., and O’Malley, B.W. Steroid receptor coactivators: servants and masters for control of systems metabolism. Trends Endocrinol Metab, 2014;25(7):337–47. 23. Altarejos, J.Y. and Montminy, M. CREB and the CRTC co‐activators: sensors for hormonal and metabolic signals. Nat Rev Mol Cell Biol, 2011;12(3): 141–51. 24. Finck, B.N. and Kelly, D.P. PGC‐1 coactivators: inducible regulators of energy metabolism in health and disease. J Clin Invest, 2006;116(3):615–22. 25. Seok, S., Fu, T., Choi, S.E. et al. Transcriptional regulation of autophagy by an FXR‐CREB axis. Nature, 2014;516(7529):108–11. 26. Zhang, Y., Fang, B., Emmett, M.J. et al. Gene regulation. Discrete functions of nuclear receptor Rev‐erbalpha couple metabolism to the clock. Science, 2015;348(6242):1488–92. 27. Schmidt, S.F., Larsen, B.D., Loft, A., and Mandrup, S. Cofactor squelching: artifact or fact? BioEssays, 2016;38(7):618–26. 28. Voss, T.C. and Hager, G.L. Dynamic regulation of transcriptional states by chromatin and transcription factors. Nat Rev Genet, 2014;15(2):69–81. 29. Bode, J.G., Albrecht, U., Haussinger, D., Heinrich, P.C., and Schaper, F. Hepatic acute phase proteins – regulation by IL‐6‐ and IL‐1‐type cytokines involving STAT3 and its crosstalk with NF‐kappaB‐dependent signaling. Eur J Cell Biol, 2012;91(6–7):496–505. 30. Iroz, A., Montagner, A., Benhamed, F. et  al. A specific ChREBP and PPARalpha cross‐talk is required for the glucose‐mediated FGF21 response. Cell Rep, 2017;21(2):403–16. 31. Zhu, B., Gates, L.A., Stashi, E. et al. Coactivator‐dependent oscillation of chromatin accessibility dictates circadian gene amplitude via REV‐ERB loading. Mol Cell, 2015;60(5):769–83. 32. Steineger, H.H., Sorensen, H.N., Tugwood, J.D. et al. Dexamethasone and insulin demonstrate marked and opposite regulation of the steady‐state mRNA level of the peroxisomal proliferator‐activated receptor (PPAR) in hepatic cells. Hormonal modulation of fatty‐acid‐induced transcription. Eur J Biochem, 1994;225(3):967–74. 33. Lemberger, T., Staels, B., Saladin, R. et  al. Regulation of the peroxisome proliferator‐activated receptor alpha gene by glucocorticoids. J Biol Chem, 1994;269(40):24527–30. 34. Ratman, D., Mylka, V., Bougarne, N. et al. Chromatin recruitment of activated AMPK drives fasting response genes co‐controlled by GR and PPARalpha. Nucleic Acids Res, 2016;44(22):10539–53. 35. Goldstein, I. and Hager, G.L. The three Ds of transcription activation by glucagon: direct, delayed, and dynamic. Endocrinology, 2018;159(1):206–16.

82

Drug Interactions in the Liver Guruprasad P. Aithal1,2 and Gerd A. Kullak‐Ublick3,4 Nottingham Digestive Diseases Centre, School of Medicine, University of Nottingham, Nottingham, UK NIHR Nottingham Biomedical Research Centre, Nottingham University Hospitals NHS Trust and the University of Nottingham, Nottingham, UK 3 University Hospital Zurich and University of Zurich, Zurich, Switzerland 4 Mechanistic Safety, Chief Medical Office and Patient Safety, Novartis Global Drug Development, Basel, Switzerland 1 2

INTRODUCTION The liver plays a key role in drug metabolism and elimination. The unique microarchitecture of the liver allows blood plasma to enter the perisinusoidal space through the fenestrated liver sinusoidal endothelium. The absorptive area of hepatocytes is augmented by microvilli that extend into this space. Drug transporters in the liver include members of the solute carrier family (SLCs) which mediate the influx or bidirectional transport of drugs, as well as members of the ATP‐binding cassette family (ABC transporters), which mediate the efflux of drugs and their metabolites into bile or into sinusoidal blood for subsequent renal excretion [1].

HEPATOCELLULAR DRUG UPTAKE Drugs were long thought to enter hepatocytes via passive diffusion across the basolateral hepatocyte membrane. This concept was first questioned when the hepatocellular uptake of bile salts was shown to be a carrier‐mediated, sodium‐dependent process [2]. Molecular cloning of the involved transport protein in 1991 identified the sodium‐taurocholate cotransporting polypeptide (NTCP) (gene symbol SLC10A1) [3]. NTCP transports bile salts and statins [4–7] and is also the hepatocellular uptake mechanism for hepatitis B and D virus [8, 9]. It interacts with the myristoylated N‐terminal pre‐S1 domain of the large envelope glycoprotein [10], which inhibits the transport function of NTCP [11]. Bile salts also enter hepatocytes by sodium‐independent uptake, mediated by members of the organic anion transporting polypeptide (OATP) superfamily [12, 13]. OATPs are hepatic uptake transporters for bile salts and also numerous drugs [12].

OATP1A2 was the first OATP cloned from human liver, ­however it is expressed mainly in neurons of the frontal cortex and ­hippocampus, and at a lower level in brain capillary endothelial cells [14, 15]. Human hepatocytes show strong expression of  three OATPs, namely OATP1B1 (SLCO1B1), OATP1B3 (SLCO1B3), and OATP2B1 (SLCO2B1) [16]. The key role of OATP1B1 in hepatic drug uptake became evident from studies showing that plasma statin levels increase in the presence of OATP1B1 inhibitors such as cyclosporin A or gemfibrozil [17–19]. The rs4149056 single nucleotide polymorphism (SNP) in the SLCO1B1 gene causes a Val174Ala amino acid substitution in exon 6 and is the strongest genetic risk factor for the onset of myopathy caused by simvastatin, with an odds ratio of 16.9 in CC homozygotes (2.1% population frequency) as compared with TT homozygotes (73% population frequency) [20]. Clinically relevant drug–drug interactions due to inhibition of OATPs are also caused by tyrosine kinase inhibitors (TKIs). OATP1B1 is inhibited by pazopanib and nilotinib with IC50 ­values of 3.89 ± 1.21 and 2.78 ± 1.13 μM, respectively [21]. Pazopanib has a boxed warning for hepatotoxicity in the US Food and Drug Administration (FDA) label, but the mechanism of hepatotoxicity is not related to inhibition of the uptake transporter OATP1B1. TKIs have to enter hepatocytes to induce hepatotoxicity, and pazopanib uptake is mediated by the organic cation transporter 1 (OCT1, SLC22A1), as shown in vitro in the human embryonic kidney cell line HEK293 stably transfected with human OCT1 [22]. The TKI erlotinib is a potent competitive inhibitor of OATP2B1 (Ki = 41 nM) [23] and erlotinib uptake is selectively mediated by OATP2B1 [24]. A combined genetic deficiency of both OATP1B1 and OATP1B3 is the cause of the human Rotor syndrome, an ­autosomal recessive disorder characterized by conjugated hyperbilirubinemia, coproporphyrinuria, and practically absent hepatic

The Liver: Biology and Pathobiology, Sixth Edition. Edited by Irwin M. Arias, Harvey J. Alter, James L. Boyer, David E. Cohen, David A. Shafritz, Snorri S. Thorgeirsson, and Allan W. Wolkoff. © 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.



82:  Drug Interactions in the Liver

BIOTRANSFORMATION OF DRUGS

uptake of anionic diagnostic agents [25]. In normal human liver, a substantial fraction of bilirubin conjugated in hepatocytes is excreted back into blood by the multidrug resistance protein 3 (MRP3, ABCC3) and subsequently reabsorbed in downstream hepatocytes by OATP1B1 and OATP1B3 (“hepatocyte hopping” model) [25] (Figure 82.1). Knockout mice lacking both hepatocellular Oatp1a and Oatp1b transporters exhibit conjugated hyperbilirubinemia and markedly decreased hepatic uptake of the model substrates methotrexate and fexofenadine [26]. Given the role of human OATP1B1/1B3 as bilirubin uptake transporters, drug–drug interactions at the basolateral entry site of hepatocytes may lead to a reduced clearance of such endogenous substrates. For example, high‐dose intravenous silibinin treatment of patients with hepatitis C infection can increase serum bilirubin [27, 28]. Accordingly, in vitro studies with OATP1B1, OATP1B3, and OATP2B1 in a heterologous expression system have shown inhibition of these transporters by silibinin [29]. In addition to the role of OATPs, uptake of drugs into hepatocytes is also mediated by members of the organic anion transporter (OAT) family [30]. OAT2 (SLC22A7) mediates the uptake of entecavir [31] and tolbutamide [32], a clinical probe substrate for the cytochrome P450‐dependent (CYP) enzyme CYP2C9, as well as the efflux of glutamate from hepatocytes [33]. From the nucleoside transporter families SLC28 and SLC29, the concentrative nucleoside transporter CNT1 and the equilibrative nucleoside transporter ENT1 mediate the uptake of ribavirin into hepatocytes [34]. The FDA provides further guidance on in vitro experimental approaches to evaluate the interaction potential between investigational drugs, transporters, and metabolizing enzymes [35].

Biotransformation of drugs within hepatocytes occurs by phase I and phase II metabolism. Phase I comprises mainly oxidation reactions (sometimes reduction or hydrolysis), notably aromatic or aliphatic hydroxylation, oxygenation at carbon, nitrogen or sulfur atoms and N‐ and O‐dealkylation. Phase I reactions are catalyzed by CYP enzymes located in the endoplasmic reticulum. Phase I metabolites usually have only minor structural differences from the parent drug but can exhibit very different pharmacological actions. Phase II metabolism involves conjugation of a drug or its phase I metabolite with endogenous molecules, such as glucuronic acid or sulfate, thereby making the product more polar and usually abolishing its pharmacological activity. The wide spectrum of drugs metabolized by CYP enzymes inevitably confers a risk for drug–drug interactions at the level of the metabolizing enzyme. Drugs can compete directly at the active site, as exemplified by the interaction between the centrally acting muscle relaxant tizanidine and the selective serotonin reuptake inhibitor fluvoxamine. Tizanidine is mainly metabolized by CYP1A2, which is strongly expressed in the liver. Fluvoxamine is a potent inhibitor of CYP1A2 and other CYP enzymes and increases the mean exposure (AUC, area under the curve) to tizanidine by a factor of 32.6 and the maximum plasma concentration (Cmax) 12.1‐fold. The elimination half‐life of tizanidine is prolonged almost threefold from 1.5 to 4.3 hours [36]. Based on this drug–drug interaction, the combination of these two drugs is contraindicated and potentially life‐threatening. Another drug which increases exposure of

Hepatocyte

Sinusoid

1051

Canaliculus

ABCC3 ABCC2 BG

BG OATP1B1/3

B L O O D

BG

C

UCB

UGT1A1

B I L E

ABCC3 ABCC2

BG BG UCB

OATP1B1/3

BG

UGT1A1

Figure 82.1  Hepatocyte hopping cycle. Unconjugated bilirubin (UCB) enters the hepatocytes via passive diffusion (OATP1B1/3 may also have a role). Conjugation with glucuronic acid by uridine 5′‐diphosphoglucuronosyltransferase (UGT) 1A1 to bilirubin glucuronides (BG) takes place in the endoplasmic reticulum. BG is effluxed into biliary canaliculi by ABCC2. A substantial proportion of the intracellular BG is rerouted by ABCC3 to the blood, from which it can be taken up by downstream hepatocytes via OATP1B1/3 transporters. “Hepatocyte hopping” distributes the biliary excretion load of bilirubin glucuronides across the liver lobule. This prevents saturation of the biliary excretion capacity in upstream hepatocytes.

1052

THE LIVER:  BIOTRANSFORMATION OF DRUGS

tizanidine through inhibition of CYP1A2 is the antibiotic ciprofloxacin. Ciprofloxacin greatly elevates plasma concentrations of tizanidine and dangerously increases its hypotensive and sedative effects. Despite the known risk of this drug combination, an analysis of Swiss claims data showed that 199 of a total of 524 797 patients analyzed were co‐prescribed tizanidine with  ciprofloxacin in 2014–2015, resulting in a significantly increased frequency of outpatient visits to physicians [37]. This is an example of a preventable medication error that incurs significant avoidable costs and health hazards to patients through a drug–drug interaction. Another example is the interaction between cerivastatin and gemfibrozil, which was the cause of 12 of 31 fatalities in the United States attributed to cerivastatin‐ induced rhabdomyolysis [38]. Gemfibrozil inhibits both the uptake transporter OATP1B1 as well as the metabolizing enzyme CYP2C8, which are required for cerivastatin clearance from the circulation [39]. The drug combination led to reduced clearance of cerivastatin and increased plasma Cmax levels, thus resulting in severe muscle injury. Cerivastatin was subsequently withdrawn from the market. In addition to direct steric interference of drugs at the level of  the enzyme, interactions can also occur through induction of  gene expression and subsequent gain‐of‐enzyme activity. Potent enzyme inducers include carbamazepine, phenytoin, and rifampicin. The herb St. John’s wort (Hypericum perforatum) is an inducer of CYP3A4, CYP2C19, CYP2C9 as well as the P‐ glycoprotein (P‐gp) transporter MDR1 (multidrug resistance gene product 1, ABCB1). The mechanism of enzyme induction by drugs frequently involves activation of the nuclear receptor pregnane X receptor (PXR), which binds to regulatory elements in the gene promoters of its target genes as a heterodimer with the retinoid X receptor, RXR [40]. Xenobiotics can act as ligands of PXR, thereby activating transcription of genes regulated by PXR. In a clinical study, the administration of St. John’s wort extract to eight healthy male volunteers during 14 days resulted in an 18% decrease in exposure of the MDR1 substrate digoxin after a single digoxin dose (0.5 mg), a 1.4‐ and 1.5‐fold

increased expression of duodenal P‐glycoprotein/MDR1 and CYP3A4, respectively, and a 1.4‐fold increase in the functional activity of hepatic CYP3A4 [14C]erythromycin breath test [41]. This is an example of how relatively small changes in the expression levels of enzymes and transporters can have a major impact on drug pharmacokinetics and consequently safety of the substrate drugs. Acetaminophen (paracetamol)‐induced liver injury can be enhanced by induction of CYP2E1, which produces the hepatotoxic metabolite N‐acetyl‐p‐benzoquinone imine (NAPQI). CYP2E1 is induced by chronic alcohol consumption and also in non‐alcoholic fatty liver disease, explaining why these conditions are associated with an increased risk of acetaminophen hepatotoxicity [42, 43]. CYP2E1 expression is also induced by the tuberculostatic agents isoniazid (INH) and rifampicin, which is one of the mechanisms hypothesized to explain enhanced hepatotoxicity when these drugs are used in combination to treat tuberculosis. INH is metabolized by N‐acetyltransferase 2 to acetylhydrazine and hydrazine, which are further oxidized to reactive metabolites by CYP2E1 (Figure 82.2). To what degree induction of CYP2E1 increases INH toxicity remains controversial. Rifampicin is a ligand of PXR and a strong inducer of CYP3A4 expression. As shown in human hepatocytes, it also increases CYP2E1 enzymatic activity and mRNA expression [44], thus providing a possible additional mechanism for the increase in INH hepatotoxicity associated with rifampicin treatment. An alternative hypothesis based on a PXR‐humanized mouse model proposes that the combination of INH and rifampicin leads to accumulation of the endogenous hepatotoxin protoporphyrin IX in the liver through PXR‐mediated upregulation of aminolevulinic synthase (the rate‐limiting enzyme in porphyrin biosynthesis) [45]. The role that hepatic CYP2E1 plays in INH‐induced hepatotoxicity by generating free radicals is an example of metabolic activation by CYP enzymes. Metabolism of drugs in phase I and II reactions can lead to the formation of reactive metabolites, which are a known risk factor for the onset of hepatotoxicity

Protoporphyrin IX (Toxic)

Aminolevulinic acid

Aminolevulinic acid synthase

Pregnane X receptor

Rifampicin

Isoniazid Hydrolysis

N-Acetyl transferase

Acetylisoniazid

Isoniazid hydrazine (Toxic)

Hydrolysis

Acetylhydrazine N-Acetyl transferase

Diacetylhydrazine (Non-toxic)

CYP2E1

N-hydroxy-acetylhydrazine Acetyldiazene (Toxic)

Figure 82.2  Mechanisms underlying antituberculosis drug‐induced liver injury. Antituberculosis drug‐induced liver injury has been attributed to key steps in the isoniazid metabolic pathway. N‐acetyl transferase 2 is responsible for metabolism of isoniazid to acetyl isoniazid, which in turn is hydrolyzed to acetyl hydrazine. The latter could be oxidized by CYP2E1 to form N‐hydroxy‐acetyl hydrazine, which further dehydrates to yield acetyldiazene, a toxic metabolite.



82:  Drug Interactions in the Liver

[46]. Formation of reactive metabolites can be assessed in vitro by adduct formation and covalent protein binding. Covalent binding of reactive metabolites to cellular proteins can lead to alteration of function or location of the target protein, or to the formation of immunogenic haptens, which can trigger a downstream immune response [47]. To estimate the clinical risk of  hepatotoxicity, the bioactivation potential is determined by  glutathione‐trapping assays, mechanism‐based cytochrome P450 (CYP)‐inactivation screens, or covalent‐binding assessment using radiolabeled compounds [48]. The detection of stable detoxification products such as glutathione adducts or dihydrodiols in the metabolic pattern can indicate metabolic activation, as can time‐dependent inhibition of an enzyme, which predicts the formation of reactive metabolites in >90% of cases. However, compounds that form reactive metabolites do not always lead to time‐dependent inhibition. Reactive metabolites formed by CYP2C9, CYP1A2, and other selected enzymes have a higher likelihood of being associated with clinical observations [49]. The nonsteroidal anti‐inflammatory drug (NSAID) diclofenac can cause severe hepatotoxicity due to formation of reactive quinone imines by CYP2C9 and CYP3A4 and activation to acyl glucuronides by UDP‐glucuronyl transferase (UGT) 2B7 [50]. The formation of acyl glucuronides by glucuronidation of the carboxylic acid moiety is also seen with ibuprofen and naproxen, although both belong to the safer NSAIDs from a hepatic perspective [50]. Excretion of diclofenac acyl glucuronide is mediated by MRP2 (ABCC2). An SNP associated with reduced activity of the transporter could increase the accumulation of the reactive metabolite within the hepatocyte, hence increasing the hepatotoxicity [50] (Figure 82.3). Fatal hepatotoxicity leading to market withdrawal has been seen with lumiracoxib and

troglitazone, both of which form quinine metabolites [51–53]. Other hepatotoxic drugs that form reactive metabolites include clozapine (iminium ion), amiodaquine, flutamide, zafirlukast, carbamazepine, sulfamethoxazole, tamoxifen, tolcapone, terbinafine, nefazodone, felbamate, and halothane.

DRUG INTERACTIONS WITH HEPATOCELLULAR EFFLUX TRANSPORTERS Bile salts are effluxed from hepatocytes into the canaliculus against a steep concentration gradient by the bile salt export pump (BSEP) [54]. There is no backup transporter for the canalicular export of bile salts and an inherited inactivation of this transporter leads to progressive familial intrahepatic cholestasis type 2 [55]. Inhibition of BSEP by drugs or drug metabolites may lead to drug‐induced cholestasis [56–58]. A well‐­characterized BSEP inhibitor is the endothelin receptor antagonist bosentan, approved for pulmonary hypertension but with a boxed warning for hepatotoxicity [59]. Cyclosporine A, which can lead to drug‐induced cholestasis in clinical routine, is also a potent BSEP inhibitor [56, 60–62]. Inhibition of BSEP leads to an intracellular elevation of bile salts, which may translate into elevated levels of serum bile salts [59]. As bile salts have detergent properties  [63] they can damage mitochondria [64, 65], thus leading to cytotoxicity and liver injury [66, 67]. In the case of the antidiabetic drug troglitazone, its major metabolite, troglitazone sulfate, has a high potential to competitively inhibit BSEP and accumulate in hepatocytes [68]. As there is currently no firm evidence that interaction of a

Basolateral/ sinusoidal membrane DCF

1053

Canalicular membrane 5– OH DCF

DCF

CYP2C8 UGT2B7

ABCC3 DCF-AG

ABCC2 DCF-AG

DCF-AG

Figure 82.3  Effect of interindividual variability of drug metabolism and clearance excretion on adverse drug reactions. Diclofenac (DCF) undergoes extensive first‐pass metabolism in humans. The predominant metabolite is 4′‐hydroxydiclofenac (OH‐DCF). In addition, a role for CYP2C8 in the metabolism of diclofenac to 5-hydroxydiclofenac (5-OH DCF) has been demonstrated. Diclofenac is also metabolized by UGT2B7. Diclofenac acyl glucuronide (DCF-AG) undergoes biliary excretion via ABCC2 (MRP2). As both diclofenac acyl glucuronide and benzoquinone imines derived from 5‐hydroxydiclofenac modify proteins covalently, this might lead to hepatotoxicity in susceptible individuals. Allelic variants of UGT2B7, CYP2C8, and ABCC2, which may predispose to the formation and accumulation of reactive diclofenac metabolites, are associated with diclofenac hepatotoxicity. Sinusoidal efflux of diclofenac acyl glucuronide is dependent on ABCC3 (MRP3) [91]. Compromised ABCC3 function can enhance injury in the gastrointestinal tract after DCF treatment [92].

1054

THE LIVER:  CLINICAL IMPLICATION

new chemical entity in an in vitro assay reliably predicts an increased risk of drug‐induced cholestasis, the FDA does not recommend routine testing of BSEP interaction during drug ­ development [69]. The European Medicines Agency (EMA) ­recommends interaction testing of drugs with BSEP during development [70]. In cases of elevated liver enzymes (alanine aminotransferase or alkaline phosphatase) during clinical trials, testing for inhibition of BSEP by the compound is critical for understanding the mechanism of drug‐induced liver injury [71]. The preferred assay for such cases has been defined by the International Transporter Consortium [72]. It has to be emphasized that lack of interaction of the parent drug with BSEP does not rule out the interaction with one or more metabolites. As drug metabolites are substrates of MRP2 (ABCC2) [73], this canalicular export system also constitutes a risk factor for drug‐induced cholestatic liver disease. Variants of this transporter have been associated with drug‐induced liver injury [74, 75]. Moreover, troglitazone decreases expression of both human MRP2 and BSEP in the liver of PXB chimeric mice [76]. When BSEP function is impaired, basolateral efflux systems (MRP3 and MRP4) are potential salvage systems to lower the burden of bile salts and drug metabolites for hepatocytes. Hence, these two transporters are additional potential susceptibility factors for drug‐induced cholestasis [77]. The breast cancer resistance protein (BCRP, ABCG2) is another ATP‐binding cassette transporter expressed at the canalicular domain of hepatocytes. Several in vivo studies have shown that both BCRP and MDR1 play a role in determining the pharmacokinetics of substrate drugs such as the TKI sunitinib. Rats treated with the MDR1 inhibitor PSC833 (valspodar) or the BCRP inhibitor pantoprazole show a significant increase in the AUC and a markedly reduced biliary excretion of sunitinib [78]. Accordingly, sunitinib was found to induce severe hepatotoxicity in a 73‐year‐old Japanese woman, in whom the plasma exposure of sunitinib and its major active metabolite N‐desethyl sunitinib was extremely high. Genotyping of seven SNPs potentially relevant to the pharmacokinetics of sunitinib revealed a 421C>A substitution in the ABCG2 gene, resulting in a homozygous 421 amino acid genotype [79]. This polymorphism is associated with low BCRP expression and activity [80]. The multidrug resistance gene product 3 (MDR3, ABCB4) is a  phosphatidylcholine transporter expressed at the canalicular membrane of hepatocytes. It translocates phosphatidylcholine ­ from the inner to the outer leaflet of the lipid bilayer. Phospholipids are an essential lipid component of bile that solubilize cholesterol in phospholipid–cholesterol vesicles. Genetic deficiency of the

ABCB4 gene causes a spectrum of cholestatic liver diseases that extend from transient neonatal cholestasis and progressive familial intrahepatic cholestasis type 3 in children, to adult biliary cirrhosis, low‐phospholipid‐associated cholelithiasis syndrome, intrahepatic cholestasis of pregnancy, and drug‐induced cholestasis [81, 82]. MDR3 is inhibited by certain drugs such as the antifungal agent itraconazole, resulting in cholestatic liver injury. In itraconazole‐treated rats, the biliary concentration of phospholipids was drastically reduced due to inhibition of MDR3‐mediated biliary phospholipid excretion [83]. Phospholipids are critically required in bile for the formation of so‐called mixed micelles, which consist of phospholipids, cholesterol, and bile salts. The inhibition of biliary phospholipid excretion impairs the formation of mixed micelles and thereby increases the toxic effect of bile salts towards the biliary epithelium. In MDR3‐expressing LLC‐PK1 cells, the MDR3 mediated efflux of [14C]phosphatidylcholine was decreased by itraconazole, further confirming that MDR3 transport function is inhibited by itraconazole. These data were reproduced in LLC‐PK1 cells stably transfected with NTCP, BSEP, MDR3, and ABCG5/ABCG8 – a polarized cell system that serves as a model for canalicular lipid secretion. The antifungal azoles posaconazole, itraconazole, and ketoconazole inhibited MDR3‐mediated phosphatidylcholine secretion, whereas amoxicillin‐clavulanate and troglitazone did not [84]. The antifungal azoles also inhibited BSEP‐mediated taurocholate excretion. The combined inhibition of MDR3 and BSEP represents a dual mechanism by which azoles cause drug‐induced liver injury in susceptible patients.

CLINICAL IMPLICATION Pharmacokinetic drug interactions through the stages of absorption, distribution, metabolism, and/or elimination influence both efficacy and adverse effects of a therapeutic regimen. Drugs that induce or inhibit the membrane transporter P‐glycoprotein lead to altered bioavailability in the intestine [85]. These are essential considerations at the time of adding a new medication into an ongoing pharmacological intervention or initiating a combination therapy, especially when the therapeutic index of the drugs involved is low. A common example is the drug interaction between immunosuppressants and antimicrobials which can lead to reduced efficacy and increased adverse effects. These considerations are an integral part of the choice of drugs and, therefore, of the monitoring plan in the clinical management of patients with organ transplantation (Table 82.1) [86, 87].

Table 82.1  Potential pharmacokinetic interactions between immunosuppressants and antifungals Immunosuppressants Cyclosporine Tacrolimus Sirolimus Everolimus Antifungals Ketoconazole Fluconazole Itraconazole Voriconazole Posaconazole

Substrate (CYP)

Inhibitor (CYP)

Type of inhibition of 3A4

P‐glycoprotein substrate

3A4 3A4 3A4/3A5 3A4/3A5/2C8

3A4 3A4 Not significant No data

Competitive Competitive Not applicable No data

Yes Yes Yes Yes

3A4 3A4 3A4 3A4