VEGF Signaling: Methods and Protocols (Methods in Molecular Biology, 2475) 1071622161, 9781071622162

Table of contents :
Preface
Contents
Contributors
Chapter 1: Opposing Roles of Vascular Endothelial Growth Factor C in Metastatic Dissemination and Resistance to Radio/Chemothe...
1 Introduction
1.1 Increased Aggressiveness of Tumors Depending on Their Metastatic Stage
1.2 A Focus on Lymphatic Vessels: Key Players in Metastatic Spread?
2 The Antagonistic Roles of VEGFC
3 Lymphatic Vessels and VEGFC-Dependent Aggressiveness of BC, RCC, HNSCC, and MDB
3.1 Lymphatic Vessels, VEGFC, and Prognosis in BC
3.2 Lymphatic Vessels, VEGFC, and Prognosis in HNSCC
3.3 Lymphatic Vessels, VEGFC, and Prognosis in MDB
3.4 Lymphatic Vessels, VEGFC, and Prognosis in RCC
4 Conclusions and Treatments´ Options
References
Chapter 2: Identification of VEGF Isoforms in Mouse, Rat, and Zebrafish Using RT-qPCR
1 Introduction
2 Materials
2.1 Nuclease-Free Water
2.2 Tissue Homogenization, Cell Lysis, and RNA Extraction
2.3 RNA Quantification and Purity
2.4 RNA Integrity
2.5 Reverse Transcription
2.6 Validation of qPCR Assays
2.7 qPCR
3 Methods
3.1 Primer Design
3.2 Tissue Homogenization, Cell Lysis, and RNA Extraction
3.3 Quantification of RNA by Fluorimetry
3.4 RNA Spectrophotometry
3.5 Determination of RNA Integrity Using an Agilent Bioanalyzer
3.6 Reverse Transcription
3.7 Validation of qPCR Primer Specificity and Production of Known Copy Number Standards
3.8 qPCR of Experimental Samples
3.9 Data Normalization
4 Notes
References
Chapter 3: Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic Endothelial Cells
1 Introduction
2 Materials
2.1 Culture Medium and Solutions for Maintaining Foreskin and Lymphatic Malformation LEC Cultures
2.2 Enzymatic Digestion
2.3 Antibodies and Buffers
2.4 Equipment
3 Methods
3.1 Sample Collection
3.2 Sample Processing and Enzymatic Digestion
3.3 Sample Staining
3.4 Instrument Set Up and Cell Sorting
3.5 Cell Culture Post FACS Sorting
3.6 Subculturing LECs
3.7 Cryopreservation
4 Notes
References
Chapter 4: Absolute Quantification of Plasma Membrane Receptors Via Quantitative Flow Cytometry
1 Introduction
2 Materials
2.1 Stain Buffer Preparation
2.2 Antibodies and Reagents
3 Methods
3.1 Single-Cell Suspension Preparation
3.2 Fluorophore Panel Design and Controls
3.3 Immunostaining
3.4 Flow Cytometer Setup
3.5 Flow Cytometry Data Acquisition
4 Notes
References
Chapter 5: Measurement of VEGF Content in Exosomes and Subsequent Tumor Tubulogenesis and In Vivo Angiogenesis Functional Assa...
1 Introduction
2 Materials
2.1 Differential Centrifugation
2.2 Nano-flow Cytometry Analysis
2.3 Enzyme-Linked Immunosorbent Assay (ELISA) Detection of VEGF
2.4 Western Blotting Detection of VEGF
2.5 Effect of VEGF-Exosomes on Colony Formation of Tumor Cells
2.6 Coculture (HUVECs/HCC) Assay
2.7 Tubulogenesis Assay of VEGF-Exosome Treated HCC Cells
2.8 Matrigel Plug Assay for Evaluating the Role of VEGF-Exosomes In Vivo Angiogenesis
3 Methods
3.1 Exosomes Isolation by Differential Centrifugation
3.2 Nano-flow Cytometry Analysis of VEGF on Exosomes
3.3 Measurement of VEGF Content in Exosomes by ELISA
3.4 Measurement of VEGF Content in Exosomes by Western Blotting
3.5 Detection of the Effect of VEGF-Exosomes on Colony Formation of Tumor Cells
3.6 Role of HUVECs-Derived Exosomes on Tumor Cells Tubulogenesis By Coculture Assay
3.7 Role of VEGF-Exosomes Derived from Tumor ECs on Tumor Cell Tubulogenesis In Vitro
3.8 Role of Exosomal VEGF in Angiogenesis In Vivo
4 Notes
References
Chapter 6: SH2-Domain Protein Isolation Using Synthetic Phosphorylated Peptides to Study VEGFR2 Signaling
1 Introduction
2 Materials
2.1 Synthetic Peptides
2.2 Peptide Dissolving Reagents
2.3 Magnetic Beads and Reagents
2.4 Cell Culture
2.5 Hypotonic Lysis Buffer and Protein Lysate (See Note 1)
2.6 Determination of Protein Concentration
2.7 Western Blot Reagents
3 Methods
3.1 Peptide Design and Hydration
3.2 Cell Culture
3.3 Optimization of the Cell Lysate Protocol and Hypotonic Lysis Buffer Solution
3.4 Optimization of the Protein Precipitation Protocol
3.4.1 Preparation of Sera-Mag Speed Streptavidin-Coated Beads
3.4.2 Immobilization of Peptides to the Magnetic Beads
3.4.3 Precipitation of Peptide-Binding Proteins
3.4.4 Release of Immobilized Biotinylated Molecules
3.5 Mass Spectrometry
3.6 Determination of Protein Concentration
3.7 Western Blot
4 Notes
References
Chapter 7: Monitoring VEGF-Stimulated Calcium Ion Flux in Endothelial Cells
1 Introduction
2 Materials
2.1 Endothelial Cell Culture
2.2 Calcium Flux Assay
2.3 NFATc2 Localization
2.4 Equipment and Software
3 Methods
3.1 Primary Endothelial Cell Culture
3.2 Preparation of HUVECs and Loading Cells with Fluorescent Dye
3.3 Cytosolic Calcium Flux Assay
3.4 Immunofluorescence Analysis of Endothelial NFATc2 Localization
4 Notes
References
Chapter 8: Co-immunoprecipitation Assays
1 Introduction
2 Materials
2.1 Cell Culture
2.2 Preparation of Cell Extracts
2.3 Pre-clearing the Lysates
2.4 Purification and Isolation of Protein Complexes
2.5 Western Blot for Protein Detection of Co-IP Complexes
3 Methods
3.1 Cell Culture
3.2 Cell Lysis
3.3 Pre-clearing the Lysates (See Note 11)
3.4 Purification and Isolation of Protein Complexes
3.5 SDS-PAGE and Western Blot for Protein Detection of Co-IP Complexes
4 Notes
References
Chapter 9: Using Immortalized Endothelial Cells to Study the Roles of Adhesion Molecules in VEGF-Induced Signaling
1 Introduction
2 Materials
2.1 Nucleofection of siRNA or TAT.Cre
2.2 VEGF-Induced Signaling Assay
3 Methods
3.1 siRNA Knockdown or TAT.Cre-Induced Deletion of Target Genes by Nucleofection
3.1.1 siRNA-Mediated Knockdown of Target Genes in PyMT Immortalized Microvascular ECs
3.1.2 TAT.Cre-Mediated Deletion of Target Genes in PyMT immortalized Microvascular ECs
3.2 Assessing the Effects of Target Genes Knockdown/Deletion on VEGF-Induced Signaling by Western Blot
4 Notes
References
Chapter 10: RNAscope for VEGF-A Detection in Human Tumor Bioptic Specimens
1 Introduction
2 Materials
2.1 Basic Laboratory Supplies
2.2 Histology Laboratory Supplies
2.3 RNAscope Equipment Supplied by Advanced Cell Diagnostic (ACD)
2.4 RNAscope Solutions
2.5 Image Acquisition and Analysis Systems
3 Methods
3.1 Sample Preparation
3.2 RNAscope Protocol
3.3 Image Acquisition
3.4 Image Analysis
4 Notes
References
Chapter 11: Global MicroRNA Profiling of Vascular Endothelial Cells
1 Introduction
2 Materials
2.1 Endothelial Cell Isolation/Culturing
2.2 miRNA Library Preparation
2.3 Gel Size Selection (Only if Library QC Fails, See Subheading 3.8)
2.4 For Troubleshooting with qPCR (Only if Library QC Fails, See Subheading 3.8)
3 Methods
3.1 Endothelial Cell Isolation (See Note 2)
3.2 RNA Extraction (See Note 7)
3.3 Library Preparation: 3′ Ligation
3.4 5′ Ligation
3.5 Reverse Transcription
3.6 Preparation of Beads
3.7 cDNA Cleanup
3.8 Library Amplification
3.9 miRNA Library Pre-sequencing QC
3.10 Real-Time PCR Troubleshooting
3.11 Size Selection of Library
3.12 Library Concentration and Read Allocation
3.13 Sequencing
3.14 Data Analysis
3.14.1 Preprocessing
3.14.2 Obtaining the Annotations for Hairpins and Mature miRNA Sequences
3.14.3 Method 1: Identification of Known Canonical miRNA Species
3.14.4 Method 1: Identification of Potential Novel miRNAs
3.14.5 Method 2, Approach 1: Identification of isomiRs and Preprocessing (miRge2.0, See Note 27)
3.14.6 Method 2, Approach 2: Identification of isomiRs and Preprocessing (Prost!, See Note 30)
3.14.7 Differential Expression Analysis (all Code Is in R)
4 Notes
References
Chapter 12: Endothelial Cell Tube Formation Assay: An In Vitro Model for Angiogenesis
1 Introduction
2 Materials
2.1 Cell Preparation
2.2 Tube Formation Assay
2.3 Visualization and Quantification of Tube Network
3 Methods
3.1 Preparation of Endothelial Cells
3.2 Tube Formation Assay
3.3 Quantification of Network
4 Notes
References
Chapter 13: Use of a Thin Layer Assay for Assessing the Angiogenic Potential of Endothelial Cells In Vitro
1 Introduction
2 Materials
3 Methods
3.1 Thin Layer Tube Formation Assay
3.2 PFA Fixation and Fluorescent Staining
3.3 Image Analysis and Quantification
3.4 RNA Extraction for qPCR
4 Notes
References
Chapter 14: VEGF-A165 -Induced Endothelial Cells Chemotactic Migration and Invasion Assays
1 Introduction
2 Materials
3 Methods
3.1 Chemotaxis Assay
3.2 Invasion Assay
3.3 Analysis
4 Notes
References
Chapter 15: Measuring Mitochondrial Calcium Fluxes in Cardiomyocytes upon Mechanical Stretch-Induced Hypertrophy
1 Introduction
2 Materials
2.1 Cells
2.2 Cyclic Mechanical Stretch
2.3 Mitochondrial Calcium Measurements
3 Methods
3.1 Cyclic Stretch
3.2 Probe Infection to Measure Ca2+ Flux
4 Notes
References
Chapter 16: Simultaneous Measurement of Endothelial Cell Proliferation and Cell Cycle Stage Using Flow Cytometry
1 Introduction
2 Materials
3 Methods
3.1 Carboxyfluorescein Succinimidyl Ester (CFSE) Staining
3.2 Propidium Iodide Staining and Flow Cytometric Analysis
4 Notes
References
Chapter 17: Ex Vivo Mouse Aortic Ring Angiogenesis Assay
1 Introduction
2 Materials
3 Methods
3.1 Harvesting the Aorta and Preparing the Rings (Day 1)
3.2 Embedding the Rings and Treatment with Agents of Interest (Days 2 to 10/12 When the Rings Are Fixed for Staining)
3.3 Immunofluorescent Staining
4 Notes
References
Chapter 18: Retinal Microvasculature-on-a-Chip for Modeling VEGF-Induced Permeability
1 Introduction
2 Materials
2.1 Cell Culture
2.2 Collagen Matrix Preparation and Formation of the Retinal Microvasculature-on-a-Chip
2.3 Permeability Assay
2.4 Immunostaining
2.5 Particular Equipment
3 Methods
3.1 Formation of the Retinal Microvasculature-on-a-Chip
3.2 VEGF-Mediated Induction of Vascular Leakage and Quantitative Assessment of Vascular Permeability
3.2.1 VEGF-Mediated Induction of Vascular Permeability
3.2.2 Leakage Assay
3.2.3 Quantitative Assessment of Permeability
3.3 Immunostaining and Confocal Imaging of the Microvasculature-on-a-Chip
3.3.1 Immunostaining
3.3.2 Automated Confocal Imaging of the Microvasculature-on-a-Chip
4 Notes
References
Chapter 19: Preventing VEGF-Mediated Vascular Permeability by Experimentally Potentiating BBB Characteristics in Endothelial C...
1 Introduction
2 Materials
2.1 Stock Solution Preparation
2.2 Endothelial Cell Culture
2.3 Transendothelial Electrical Resistance (TEER) Assay
2.4 Fluorescein Isothiocyanate (FITC)-Dextran Permeability Assays
2.5 Immunocytochemistry
2.6 Working Stock Preparation for Promoting Barrier Resistance
3 Methods
3.1 Endothelial Cell Maintenance (see Note 13)
3.2 TEER Assay
3.3 FITC-Dextran Permeability Assays
3.4 Immunocytochemistry
4 Notes
References
Chapter 20: The Embryonic Mouse Hindbrain and Postnatal Retina as In Vivo Models to Study Angiogenesis
1 Introduction
1.1 The Mouse Embryo Hindbrain
1.2 The Postnatal Mouse Retina
2 Materials
2.1 Equipment
2.2 Reagents
3 Methods
3.1 Hindbrain Dissection and Fixation
3.2 Retina Dissection and Fixation
3.3 Whole-Mount Staining
3.4 Imaging
3.5 Quantitation of Angiogenesis
4 Notes
References
Chapter 21: Evaluating VEGF-Induced Vascular Leakage Using the Miles Assay
1 Introduction
2 Materials
2.1 Mouse Preparation
2.2 Injections
2.3 Dye Extraction/Quantification
3 Methods
3.1 Mouse Preparation
3.2 Injection of Evans Blue Dye
3.3 Induction of Vascular Leakage
3.4 Quantification of Vascular Leakage
4 Notes
References
Chapter 22: Modulation of VEGFA Signaling During Heart Regeneration in Zebrafish
1 Introduction
2 Cardiac Regeneration in Zebrafish
3 Mechanisms of Revascularization in Cardiac Repair and Regeneration
4 VEGFA and Heart Regeneration
5 Genetic Tools for Modulating VEGFA Signaling During Zebrafish Heart Regeneration
5.1 vegfaa-/-
5.2 Tg(hsp70l:vegfaa121-F17A)
5.3 Tg(hsp70l:loxp-Stop-loxp-mTom-codOptP2A-T46Avegfaa)
5.4 flt1-/-
5.5 nrp1a-/-
5.6 Tg(βactin2:loxP-mTagBFP-STOP-loxP-vegfaa)
6 Concluding Remarks
7 Transgenic Lines Acronyms and Abbreviations
References
Chapter 23: Three-Dimensional Visualization of Blood and Lymphatic Vessels in the Adult Zebrafish Heart by Chemical Clearing
1 Introduction
2 Materials
2.1 Chemical Clearing Reagents
2.2 Solutions
2.3 Antibodies
2.4 Consumables
2.5 Equipment
3 Methods
3.1 Tissue Collection and Preparation
3.2 Tissue Clearing: Fluorescent Reporter Lines
3.3 Tissue Clearing: Immunofluorescent Antibody Staining
3.4 Imaging
4 Notes
References
Chapter 24: Fluorescence-Activated Cell Sorting and Quantitative Real-Time PCR to Reveal VEGF-Expressing Macrophage Population...
1 Introduction
2 Materials
2.1 Needle Stab Injury
2.2 Cell Dissociation and FACS
2.3 RNA/cDNA Prep and qPCR
3 Methods
3.1 Needle Stab Injury
3.2 Cell Dissociation
3.3 FACS Sorting
3.4 RNA and cDNA Preparation
3.5 qPCR
4 Notes
References
Chapter 25: Assessing Molecular Regulation of Vascular Permeability Using a VEGF-Inducible Zebrafish Model
1 Introduction
2 Materials
2.1 Reagents for Generation of Cre recombinase mRNA
2.2 Tricaine (MS-222) Solution
2.3 2.5% Methylcellulose Solution
2.4 Equipment
3 Methods
3.1 Generation of Zebrafish Embryos
3.2 Linearization of pT3TS-Cre Plasmid DNA and Generation of Cre recombinase mRNA
3.3 Cre recombinase mRNA Injections
3.4 Morpholino Injections (Optional)
3.5 1-Phenyl-2-Thiourea (PTU) Treatment to Inhibition of Pigment Formation (Optional)
3.6 Selection of VEGF Transgenic Zebrafish
3.7 Dextran Injections
3.8 Heat Shock Induction of the VEGF Transgene
3.9 Imaging and Quantification of VEGF-Induced Vascular Permeability in Zebrafish
4 Notes
References
Index

Citation preview

Methods in Molecular Biology 2475

Lorna R. Fiedler Caroline Pellet-Many Editors

VEGF Signaling Methods and Protocols Second Edition

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

VEGF Signaling Methods and Protocols Second Edition

Edited by

Lorna R. Fiedler Radcliffe Department of Medicine, University of Oxford, Oxford, Oxfordshire, UK

Caroline Pellet-Many Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK

Editors Lorna R. Fiedler Radcliffe Department of Medicine University of Oxford Oxford, Oxfordshire, UK

Caroline Pellet-Many Department of Comparative Biomedical Sciences Royal Veterinary College London, UK

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-2216-2 ISBN 978-1-0716-2217-9 (eBook) https://doi.org/10.1007/978-1-0716-2217-9 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Illustration Caption: CUBIC clearing and three-dimensional reconstruction of a two-year-old Zebrafish heart from the fluorescent reporter line Kdlr-HsRas-mCherry. This approach reveals a complex, extensive vascular network in high resolution, to clearly visualize and thus more fully understand the vascular responses in the injured heart. Image created by Lorna R. Fiedler. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface Since its discovery and first characterization more than 30 years ago, Vascular Endothelial Growth Factor (VEGF) has retained its position as one of the most potent pro-angiogenic factors, inducing endothelial cell proliferation, migration, and permeability in both physiological and pathological conditions. VEGF is essential for normal cardiovascular and lymphatic development as well as for physiologic vascular homeostasis in diverse cells and tissues, while its dysregulation supports the molecular pathogenesis of many diseases including tumor growth and metastasis, rheumatoid arthritis, heart failure, hypertension, age-related macular degeneration, renal disease, pre-eclampsia, and more. In the past couple of decades, our understanding of VEGF-dependent molecular mechanisms has been greatly refined by the emergence of new technologies such as transcriptomics and gene-editing technologies, which are becoming routinely used tools in many research labs. The wealth of data being generated provides an increasingly more complete though complex picture of the differential roles of VEGF isoforms and their receptors in disease. Clinical translation has been largely unsuccessful, however, recent advances in methodological approaches have brought about an enhanced understanding of the multifaceted role of VEGF in various settings. For example, a newly described role in the modulation of anti-tumor immunity gives insight into why previous approaches have failed clinically and what can be improved, summarized in a review chapter herein. This example highlights that there is still more to learn about VEGF and reinforces its relevance as a potential therapeutic target. This volume provides an updated collection of protocols for manipulating and studying VEGF signaling pathways in vitro and in vivo and aims to present a range of both firmly established and newly emerging technologies. Multiple species are also covered, from mouse to zebrafish to human. Protocols for investigating the role of VEGF and VEGFR isoforms in different settings are presented, e.g., in exosomes, cultured cells, or in tissues, using conventional and more recent techniques, such as fluorescence-activated cell sorting (FACS) and global miRNA profiling, respectively. Robust cell assays essential for the investigation of basic angiogenic mechanisms (proliferation, migration, tubulogenesis) have been updated, and emerging approaches for studying VEGF signaling in more complex cellular systems are a new addition to the previous volume. Later chapters detail a selection of in vivo models, starting with updated protocols for studying angiogenesis in mice. In recent years, zebrafish models have become essential tools for fundamental research into gene function, in part due to the emergence of CRISPR-Cas9 gene editing technology (described in our previous edition). An exhaustive and up to date review of existing lines generated to specifically study VEGF signaling in zebrafish is provided alongside selected zebrafish specific protocols. We hope that this new edition will provide a useful tool for researchers in the vascular biology community and beyond, in understanding the basic biology of VEGF signaling and in translating this research into the clinic. It was a great pleasure to compile this edition and

v

vi

Preface

learn about the latest techniques and the research of our contributors. We would like to thank all the authors for all their contributions and the series editor, John Walker, for his guidance and support. Oxford, UK London, UK

Lorna R. Fiedler Caroline Pellet-Many

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 Opposing Roles of Vascular Endothelial Growth Factor C in Metastatic Dissemination and Resistance to Radio/Chemotherapy: Discussion of Mechanisms and Therapeutic Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christopher Montemagno, Fre´de´ric Luciano, and Gilles Page`s 2 Identification of VEGF Isoforms in Mouse, Rat, and Zebrafish Using RT-qPCR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patricia de Winter 3 Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic Endothelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theresa Connor, Nerida Sleebs, and Zerina Lokmic-Tomkins 4 Absolute Quantification of Plasma Membrane Receptors Via Quantitative Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yingye Fang, Manasi Malik, Sarah K. England, and P. I. Imoukhuede 5 Measurement of VEGF Content in Exosomes and Subsequent Tumor Tubulogenesis and In Vivo Angiogenesis Functional Assays . . . . . . . . . . . Ye Zeng, Xiaoqiang Du, Wenli Jiang, and Yan Qiu 6 SH2-Domain Protein Isolation Using Synthetic Phosphorylated Peptides to Study VEGFR2 Signaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chiara Testini 7 Monitoring VEGF-Stimulated Calcium Ion Flux in Endothelial Cells . . . . . . . . . William R. Critchley, Gareth W. F. Fearnley, Izma Abdul-Zani, Carmen Molina-Paris, Claus Bendtsen, Ian C. Zachary, Michael A. Harrison, and Sreenivasan Ponnambalam 8 Co-immunoprecipitation Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ian M. Evans and Ketevan Paliashvili 9 Using Immortalized Endothelial Cells to Study the Roles of Adhesion Molecules in VEGF-Induced Signaling . . . . . . . . . . . . . . . . . . . . . . . . . James A. G. E. Taylor, Christopher J. Benwell, and Stephen D. Robinson 10 RNAscope for VEGF-A Detection in Human Tumor Bioptic Specimens. . . . . . . Tiziana Annese, Roberto Tamma, and Domenico Ribatti 11 Global MicroRNA Profiling of Vascular Endothelial Cells. . . . . . . . . . . . . . . . . . . . Eloi Schmauch, Anna-Liisa Levonen, and Suvi Linna-Kuosmanen 12 Endothelial Cell Tube Formation Assay: An In Vitro Model for Angiogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mary Kelley, Sara Fierstein, Laura Purkey, and Kathleen DeCicco-Skinner

vii

v ix

1

25

47

61

79

97 113

125

133 143 157

187

viii

13

14

15

16

17 18

19

20

21 22 23

24

25

Contents

Use of a Thin Layer Assay for Assessing the Angiogenic Potential of Endothelial Cells In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James A. E. Lane, Ashton Faulkner, Elizabeth J. T. Finding, Eleanor G. Lynam, and Caroline P. D. Wheeler-Jones VEGF-A165 -Induced Endothelial Cells Chemotactic Migration and Invasion Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caroline Pellet-Many Measuring Mitochondrial Calcium Fluxes in Cardiomyocytes upon Mechanical Stretch-Induced Hypertrophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . Daniela Ramaccini, Carlotta Giorgi, and Michelle L. Matter Simultaneous Measurement of Endothelial Cell Proliferation and Cell Cycle Stage Using Flow Cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eleanor G. Lynam, James A. E. Lane, Elizabeth J. T. Finding, and Caroline P. D. Wheeler-Jones Ex Vivo Mouse Aortic Ring Angiogenesis Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vedanta Mehta and Marwa Mahmoud Retinal Microvasculature-on-a-Chip for Modeling VEGF-Induced Permeability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . He´loı¨se Ragelle, Karen Dernick, Peter D. Westenskow, and Stefan Kustermann Preventing VEGF-Mediated Vascular Permeability by Experimentally Potentiating BBB Characteristics in Endothelial Cells . . . . . . . . . . . . . . . . . . . . . . . Bo Kyoung Kim, Je´re´mie Canonica, Filip Roudnicky, and Peter D. Westenskow The Embryonic Mouse Hindbrain and Postnatal Retina as In Vivo Models to Study Angiogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alessandro Fantin and Christiana Ruhrberg Evaluating VEGF-Induced Vascular Leakage Using the Miles Assay . . . . . . . . . . . James T. Brash, Christiana Ruhrberg, and Alessandro Fantin Modulation of VEGFA Signaling During Heart Regeneration in Zebrafish . . . . Kaushik Chowdhury, Shih-Lei Lai, and Rube´n Marı´n-Juez Three-Dimensional Visualization of Blood and Lymphatic Vessels in the Adult Zebrafish Heart by Chemical Clearing . . . . . . . . . . . . . . . . . . . . . . . . . Lorna R. Fiedler, Paul R. Riley, and Roger Patient Fluorescence-Activated Cell Sorting and Quantitative Real-Time PCR to Reveal VEGF-Expressing Macrophage Populations in the Zebrafish Larvae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrew Herman, Alexander Greenhough, and David B. Gurevich Assessing Molecular Regulation of Vascular Permeability Using a VEGF-Inducible Zebrafish Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luke H. Hoeppner

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

197

205

215

223

229

239

259

275 289 297

313

325

339 351

Contributors IZMA ABDUL-ZANI • School of Molecular & Cellular Biology, University of Leeds, Leeds, UK TIZIANA ANNESE • Department of Basic Medical Sciences, Neurosciences and Sensory Organs, Section of Human Anatomy and Histology, University of Bari Medical School, Bari, Italy CLAUS BENDTSEN • IMED Biotech Unit, AstraZeneca, Cambridge, UK CHRISTOPHER J. BENWELL • Gut Microbes and Health Programme, Quadram Institute Bioscience, Norwich Research Park, Norwich, UK JAMES T. BRASH • UCL Institute of Ophthalmology, University College London, London, UK JE´RE´MIE CANONICA • Pharmaceutical Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland KAUSHIK CHOWDHURY • Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan; Taiwan International Graduate Program in Molecular Medicine, National Yang Ming Chiao Tung University and Academia Sinica, Taipei, Taiwan THERESA CONNOR • Murdoch Children’s Research Institute, The Royal Children’s Hospital, Parkville, VIC, Australia WILLIAM R. CRITCHLEY • School of Molecular & Cellular Biology, University of Leeds, Leeds, UK PATRICIA DE WINTER • Division of Surgery & Interventional Science, Department of Targeted Intervention, University College London, London, UK; qStandard, Middlesex, UK KATHLEEN DECICCO-SKINNER • Department of Biology, American University, Washington, DC, USA KAREN DERNICK • Roche Pharma Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland XIAOQIANG DU • Institute of Biomedical Engineering, West China School of Basic Medical Sciences and Forensic Medicine, Sichuan University, Chengdu, Sichuan, China SARAH K. ENGLAND • Center for Reproductive Health Sciences, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO, USA IAN M. EVANS • Breast Cancer Research, Cancer Stem Cell Team, Institute of Cancer Research, Chester Beatty Laboratories, London, UK YINGYE FANG • Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA; University of Washington, Department of Bioengineering, Seattle, WA, USA ALESSANDRO FANTIN • Department of Biosciences, University of Milan, Milan, Italy ASHTON FAULKNER • Experimental Cardiovascular Medicine, Bristol Medical School, University of Bristol, Bristol, UK GARETH W. F. FEARNLEY • School of Molecular & Cellular Biology, University of Leeds, Leeds, UK LORNA R. FIEDLER • Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, Oxfordshire, UK; MRC Weatherall Institute of Molecular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, Oxfordshire, UK SARA FIERSTEIN • Department of Biology, American University, Washington, DC, USA ELIZABETH J. T. FINDING • Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK

ix

x

Contributors

CARLOTTA GIORGI • Department of Medical Sciences, University of Ferrara, Ferrara, Italy; Laboratory of Technologies for Advanced Therapy (LTTA), Technopole of Ferrara, Ferrara, Italy ALEXANDER GREENHOUGH • Centre for Research in Biosciences, Faculty of Health and Applied Sciences, University of the West of England, Bristol, UK DAVID B. GUREVICH • Department of Biology and Biochemistry, Centre for Therapeutic Innovation, Faculty of Science, University of Bath, Bath, UK MICHAEL A. HARRISON • School of Biomedical Sciences, University of Leeds, Leeds, UK ANDREW HERMAN • Faculty of Life Sciences, School of Cellular and Molecular Medicine, University of Bristol, Bristol, UK LUKE H. HOEPPNER • The Hormel Institute, University of Minnesota, Austin, MN, USA; Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA P. I. IMOUKHUEDE • Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, USA WENLI JIANG • Institute of Biomedical Engineering, West China School of Basic Medical Sciences and Forensic Medicine, Sichuan University, Chengdu, Sichuan, China MARY KELLEY • Department of Biology, American University, Washington, DC, USA BO KYOUNG KIM • Pharmaceutical Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland STEFAN KUSTERMANN • Roche Pharma Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland SHIH-LEI LAI • Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan; Taiwan International Graduate Program in Molecular Medicine, National Yang Ming Chiao Tung University and Academia Sinica, Taipei, Taiwan JAMES A. E. LANE • Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK ANNA-LIISA LEVONEN • A. I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland SUVI LINNA-KUOSMANEN • MIT Computer Science and Artificial Intelligence Laboratory, Cambridge, MA, USA; Broad Institute of MIT and Harvard, Cambridge, MA, USA ZERINA LOKMIC-TOMKINS • Murdoch Children’s Research Institute, The Royal Children’s Hospital, Parkville, VIC, Australia; Faculty of Medicine, Nursing and Health Sciences, Monash University, Clayton, VIC, Australia ´ FREDE´RIC LUCIANO • Institute for Research on Cancer and Aging of Nice (IRCAN), Centre Antoine Lacassagne, University Coˆte d’Azur, CNRS UMR 7284, INSERM U1081, Nice, France; Centre Antoine Lacassagne, Nice, France ELEANOR G. LYNAM • Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK MARWA MAHMOUD • Faculty of Health, Education, Medicine and Social Care, Anglia Ruskin University, Chelmsford, UK MANASI MALIK • Center for Reproductive Health Sciences, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO, USA RUBE´N MARI´N-JUEZ • Centre Hospitalier Universitaire Sainte-Justine Research Centre, Montreal, QC, Canada; Department of Pathology and Cell Biology, University of Montreal, Montreal, QC, Canada MICHELLE L. MATTER • University of Hawaii Cancer Center, Honolulu, HI, USA; Center for Cardiovascular Research John A. Burns School of Medicine, Honolulu, HI, USA

Contributors

xi

VEDANTA MEHTA • Division of Cardiovascular Medicine, Radcliffe Department of Medicine, Wellcome Centre for Human Genetics, Oxford, UK CARMEN MOLINA-PARIS • School of Mathematics, University of Leeds, Leeds, UK CHRISTOPHER MONTEMAGNO • LIA ROPSE, Laboratoire International Associe´, Centre Scientifique de Monaco, Universite´ Coˆte d’Azur, Nice, France; Institute for Research on Cancer and Aging of Nice (IRCAN), Centre Antoine Lacassagne, University Coˆte d’Azur, CNRS UMR 7284, INSERM U1081, Nice, France; De´partement de Biologie Me´ dicale, Centre Scientifique de Monaco, Monaco, Monaco GILLES PAGE`S • LIA ROPSE, Laboratoire International Associe´, Centre Scientifique de Monaco, Universite´ Coˆte d’Azur, Nice, France; Institute for Research on Cancer and Aging of Nice (IRCAN), Centre Antoine Lacassagne, University Coˆte d’Azur, CNRS UMR 7284, INSERM U1081, Nice, France; Centre Antoine Lacassagne, Nice, France KETEVAN PALIASHVILI • Centre for Precision Healthcare, UCL Division of Medicine, University College London, London, UK ROGER PATIENT • Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, Oxfordshire, UK; MRC Weatherall Institute of Molecular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, Oxfordshire, UK CAROLINE PELLET-MANY • Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK SREENIVASAN PONNAMBALAM • School of Molecular & Cellular Biology, University of Leeds, Leeds, UK LAURA PURKEY • Department of Biology, American University, Washington, DC, USA YAN QIU • Institute of Biomedical Engineering, West China School of Basic Medical Sciences and Forensic Medicine, Sichuan University, Chengdu, Sichuan, China ´ HELOI¨SE RAGELLE • Roche Pharma Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland DANIELA RAMACCINI • University of Hawaii Cancer Center, Honolulu, HI, USA; Center for Cardiovascular Research John A. Burns School of Medicine, Honolulu, HI, USA; Department of Medical Sciences, University of Ferrara, Ferrara, Italy; Laboratory of Technologies for Advanced Therapy (LTTA), Technopole of Ferrara, Ferrara, Italy DOMENICO RIBATTI • Department of Basic Medical Sciences, Neurosciences and Sensory Organs, Section of Human Anatomy and Histology, University of Bari Medical School, Bari, Italy PAUL R. RILEY • Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, Oxfordshire, UK STEPHEN D. ROBINSON • Gut Microbes and Health Programme, Quadram Institute Bioscience, Norwich Research Park, Norwich, UK; School of Biological Sciences, University of East Anglia, Norwich Research Park, Norwich, UK FILIP ROUDNICKY • Pharmaceutical Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland CHRISTIANA RUHRBERG • UCL Institute of Ophthalmology, University College London, London, UK ELOI SCHMAUCH • A. I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland; MIT Computer Science and Artificial Intelligence Laboratory, Cambridge, MA, USA; Broad Institute of MIT and Harvard, Cambridge, MA, USA NERIDA SLEEBS • Murdoch Children’s Research Institute, The Royal Children’s Hospital, Parkville, VIC, Australia

xii

Contributors

ROBERTO TAMMA • Department of Basic Medical Sciences, Neurosciences and Sensory Organs, Section of Human Anatomy and Histology, University of Bari Medical School, Bari, Italy JAMES A. G. E. TAYLOR • Gut Microbes and Health Programme, Quadram Institute Bioscience, Norwich Research Park, Norwich, UK CHIARA TESTINI • Department of Medical Cell Biology, Uppsala University, Uppsala, Sweden PETER D. WESTENSKOW • Roche Pharma Research and Early Development, Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd., Basel, Switzerland CAROLINE P. D. WHEELER-JONES • Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK IAN C. ZACHARY • Centre for Cardiovascular Biology and Medicine, University College London, London, UK YE ZENG • Institute of Biomedical Engineering, West China School of Basic Medical Sciences and Forensic Medicine, Sichuan University, Chengdu, Sichuan, China

Chapter 1 Opposing Roles of Vascular Endothelial Growth Factor C in Metastatic Dissemination and Resistance to Radio/Chemotherapy: Discussion of Mechanisms and Therapeutic Strategies Christopher Montemagno, Fre´de´ric Luciano, and Gilles Page`s Abstract Many cancers can be cured by combining surgery with healthy margins, radiation therapy and chemotherapies. However, when the pathology becomes metastatic, cancers can be incurable. The best situation involves “chronicization” of the pathology even for several years. However, most of the time, patients die within a few months. To disseminate throughout the body, cancer cells must enter the vascular network and seed in another organ. However, during the initiation of cancer processes, the tumor is avascular. Later, the production of angiogenic factors causes tumor neovascularization and subsequent growth and spread, and the presence of blood and/or lymphatic vessels is associated with high grade tumors. Moreover, during tumor development, cancer cells enter lymphatic vessels and disseminate via the lymphatic network. Hence, blood and lymphatic vessels are considered as main routes of metastatic dissemination and cancer aggressiveness. Therefore, anti-angiogenic drugs entered in the therapeutic arsenal from 2004. Despite undeniable effects however, they are far from curative and only prolong survival by a few months. Recently, the concepts of angio/lymphangiogenesis were revisited by analyzing the role of blood and lymphatic vessels at the initiation steps of tumor development. During this period, cancer cells enter lymphatic vessels and activate immune cells within lymph nodes to initiate an antitumor immune response. Moreover, the presence of blood vessels at the proximity of the initial nodule allows immune cells to reach the tumor and eliminate cancer cells. Therefore, blood and lymphatic networks have a beneficial role during a defined time window. Considering only their detrimental effects is a concern. Hence, administration of anti-angio/lymphangiogenic therapies should be revisited to avoid the destruction of networks involved in antitumor immune response. This review mainly focuses on one of the main drivers of lymphangiogenesis, the VEGFC and its beneficial and pejorative roles according to the grade of aggressive tumors. Key words VEGFC, Metastasis, Lymphatic vessels, Resistance to treatments

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_1, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

1

2

1

Christopher Montemagno et al.

Introduction

1.1 Increased Aggressiveness of Tumors Depending on Their Metastatic Stage

Metastatic dissemination is responsible for 90% of cancer deaths. In addition to the intrinsic properties of the cancer cells, a key role of the tumor microenvironment in the metastasis process has been proposed. An aberrant vascularization and a hypoxic microenvironment are common characteristics of cancers with high metastatic potential. Thus, the prognosis of patients suffering from the same tumor can differ according to tumor properties and local environment. Unfortunately, the presence of metastases at the time of diagnosis makes the disease incurable in most of the cases. In most solid tumors, the prognosis of non-metastatic disease (M0) is generally good, but metastatic disease (M1) is associated with poorer outcomes. Breast cancers (BC), renal cell carcinoma (RCC), head and neck squamous cell carcinoma (HNSCC) and pediatric medulloblastoma (MDB) are examples of tumors for which the blood vessels were described as a main route of metastatic dissemination. They are also representative of tumors for which prognosis radically changes from a curable to an incurable disease according to the M0/M1 status according to the American Cancer Society (https://www.cancer.org/cancer/breast-cancer). Patients with different genetic subgroups of BC have an increased risk of metastasis which directly correlates with a reduced survival rate from >10 years to 8000  g. 5. Remove 600 μL supernatant taking care not to disturb the pellet and transfer to a 1.5 mL tube. 6. Add an equal volume of freshly prepared 70% ethanol to each sample and vortex to mix. 7. Pipette 700 μL of the sample/ethanol mixture, including any precipitate that may have formed, into an RNeasy spin column (or other silica based column) inserted into a 2 mL collection tube. Cap the tube and centrifuge at >8000  g for 30 s. 8. Discard the flow through. Apply the remainder of the mixture to the same column, repeat the centrifugation and discard the flow through. The RNA in the lysate is now trapped on the filter of the column. 9. Pipette 700 μL of buffer RW1 to the column. Centrifuge at >8000  g for 2 min and discard the flow through (see Note 7). 10. Pipette 500 μL of buffer RPE onto the column. Centrifuge as per step 9 and discard the flow through.

32

Patricia de Winter

11. Repeat step 10 once more. 12. Transfer the spin column to a fresh 2 mL collection tube and centrifuge for 2 min. Leave the spin column with the cap slightly open for a few min to dry the silica membrane (alternatively the column may be placed at 37  C for 2 min). The RPE buffer contains ethanol. Carryover of ethanol can reduce yield and interfere with downstream applications. 13. Place the spin column into a fresh 1.5 mL tube. Add 30 μL of RNase-free water directly onto the center of the silica membrane. Cap the tube and centrifuge for 1 min at maximum speed to recover the RNA. 14. Place RNA samples on ice. 3.3 Quantification of RNA by Fluorimetry

See Note 8. 1. Sum the number of samples and add 3 (2 standards plus one extra) then divide the total by 5 to determine how many mL of Qubit buffer working solution will be required. 2. Prepare a master mix: dilute Qubit RNA BR reagent (or RNA HS reagent if RNA yield is likely to be low) 1:200 with Qubit RNA buffer, i.e., for 2 samples plus the standards prepare 1 mL of master mix by adding 5 μL of reagent to 995 μL of buffer. 3. Pipette 199 μL of master mix into each sample tube and 190 μL of master mix into each standard tube. 4. Pipette 10 μL of standard 1 into the first standard tube. 5. Pipette 10 μL of standard 2 into the second standard tube. 6. Pipette 1 μL of sample into each sample tube. 7. Cap all tubes and vortex to mix. Incubate for 2 min at room temperature. 8. Select the appropriate RNA assay on the Qubit fluorimeter, and read the first standard, then read the second standard. 9. Read each sample and record or save to the Qubit hardware. If using a Qubit 1.0 fluorimeter, the given concentration must be multiplied by 200 to account for the dilution. If using a Qubit 2.0 or above, select calculate sample concentration, 1 μL sample volume, and ng/μL for automated calculation of concentration.

3.4 RNA Spectrophotometry

Determine RNA concentration and purity using Nanodrop spectrophotometer, following manufacturer’s instructions. Pay attention to the absorbance curve, which should be smooth and symmetrical with a peak at 260 nm. Samples with low purity may provide unreliable results. The Nanodrop requires no specialized consumables, and a measure of purity can also be obtained. An absorbance of 1 U at

VEGF Isoform Quantification by RT-qPCR

33

260 nm corresponds to 40 μg of RNA/mL. The ratio of A260/ A280 provides a measure of RNA purity for contamination. For RNA a ratio close to 2 (1.9–2.2) suggests good purity. The ratio of A260/A230 provides a measure of RNA purity with respect to many contaminants such as chaotropic salts (e.g., GITC) and protein, etc. This ratio is sensitive to RNA concentration—low RNA concentrations can give a low A260/A230. There is no spectrophotometric method available that can guarantee that the RNA is completely free of contaminating genomic DNA as the absorbance spectra of these two nucleic acids is indistinguishable, therefore fluorimetry has the advantage here, in that each assay is specific for a nucleic acid, RNA, or DNA. Spectrophotometry cannot determine RNA concentrations lower than a few ng/μL. 3.5 Determination of RNA Integrity Using an Agilent Bioanalyzer

See Note 9. 1. Remove the reagents from the fridge half an hour before they will be required so that they reach room temperature. Switch on the Bioanalyzer and open the software. 2. Slowly fill one of the wells of an electrode cleaning chip with 350 μL RNaseZAP and place it into the Agilent 2100 Bioanalyzer. Close the lid and leave it closed for about 1 min. 3. Meanwhile, fill a second electrode cleaning chip with 350 μL nuclease-free water. 4. Open the lid and remove the electrode cleaning chip containing RNaseZAP and replace it with the one containing water. Close the lid and leave it closed for about 1 min. 5. Open the lid, remove the chip and leave the lid open for the electrode pins to air dry. 6. Vortex mix the dye concentrate for 10 s, centrifuge to collect the tube contents and add 1 μL of dye into the previously prepared aliquot of filtered gel. Protect the dye from light. 7. Vortex the solution well. Centrifuge at 13,000  g for 10 min at room temperature. 8. Put a new RNA chip into the chip priming station. 9. Reverse pipette 9 μL of the gel-dye mix into the well marked with a circled G (third well down in last column). Always pipette to the bottom of the well, not the sides, to prevent bubbles. 10. Set the timer for 30 s, ensure the plunger is set at 1 mL and close the chip priming station. 11. Simultaneously start the timer and press the syringe plunger until it is held by the syringe clip, then wait exactly 30 s and release the clip.

34

Patricia de Winter

12. Wait 5 s, then slowly pull the plunger back to the 1 mL position if it has not returned of its own accord. 13. Open the chip priming station and pipette 9 μL of gel-dye mix into the two remaining wells marked with a G (first and second wells in last column). The RNA chip is now primed. 14. Reverse pipette 5 μL of marker (green lid) in the well marked with a picture of a ladder (fourth well down in last column) and in all 12 sample wells (all wells of columns 1–3). 15. Denature the secondary structure of the ladder by heating to 70  C for 2 min and then reverse pipette 1 μL into the well marked with a picture of a ladder (fourth well down in last column). 16. Reverse pipette 1 μL of RNA into each of the sample wells. Reverse pipette 1 μL of marker (green lid) or sample replicates into each sample well that is not being used. 17. Place the chip into the Agilent chip vortex mixer and mix for 1 min at the setting indicated by the arrow (2400 rpm). 18. Run the chip within 5–10 min to prevent evaporation: place the chip into the Bioanalyzer and close the lid. The software should mirror this action. 19. In the Bioanalyzer 2100 Expert software Instrument Context select assays ) RNA ) Eukaryotic total then click Start. The file will save automatically to the Agilent Bioanalyzer Data folder. The chip will take approximately 30 min to run and the data will appear in real time. During this time you can add sample identities to the run by typing them into the table in the sample tab. 20. Mammalian RNA should exhibit two large peaks corresponding to 18S and 28S rRNA eluting at around 42 s and 47 s, respectively. The baseline should be reasonably flat and small RNAs under 200 nt will not be present as this is the cut off for the RNeasy kit. The marker elutes at about 23 s and its absence indicates a problem with the run. The RIN number calculated by the software indicates the integrity of the RNA with 10 being extremely high integrity and 1 being very degraded. Ensure samples selected for experiments are of similar integrity. 3.6 Reverse Transcription

See Note 10. 1. Thaw reagents, mix and briefly centrifuge to collect the tube contents. 2. Dilute each RNA with nuclease-free water to the maximum concentration permitted by the RT kit using the fluorimetry data. If RNA concentrations are too low to reverse transcribe the maximum amount of RNA, dilute samples to the lowest

VEGF Isoform Quantification by RT-qPCR

35

RNA concentration. Some judgment needs to be applied here—if all samples are high concentration except for one or two it would be more prudent to normalize the high concentration samples to the maximum permitted and load the maximum volume permitted for the low concentration samples. 3. Calculate the total number of reactions required and add at least 2 to this number to account for pipetting loss (a large number of samples, >30, will require more additional volume to account for loss). 4. Prepare a master mix of RT buffer, RT primers, dNTPs, and reverse transcriptase according to the volumes dictated by the kit. Many modern kits do not have the dNTPs as a separate reagent but include them with one of the other reagents such as the buffer. 5. Pipette the recommended amount of master mix into the required number of qPCR tubes or wells. 6. Add the recommended volume of RNA to the wells. 7. If any of the primer pairs for the qPCR step are not intronspanning RT negative reactions will also need to be prepared to check for residual gDNA contamination (this is recommended even of the samples are treated to remove gDNA). Add the same volume of water that was used for the master mix in step 5 to fresh qPCR tubes or wells. 8. Pipette the same volume of RNA that was used in step 6 to the wells containing water. 9. Set the qPCR machine to the correct temperature and time for the reaction, place the tubes or plate into the machine and start the run to convert RNA to cDNA. 10. Remove the cDNA samples from the machine and dilute tenfold with yeast tRNA at 500 ng/mL. Dilution of the cDNA is important in preventing inhibition of the qPCR reaction. I recommend a minimum of fivefold dilution, but I routinely use tenfold. If the cDNA will not be used immediately, place at 5  C overnight or at 20  C in the longer term. 3.7 Validation of qPCR Primer Specificity and Production of Known Copy Number Standards

See Note 11. 1. Centrifuge tubes of primer at max speed for 5 min to pellet the lyophilized oligonucleotides. 2. Reconstitute primers in TE buffer to a stock concentration of 100 mmol/L. Some primer suppliers provide the volume of TE buffer required in microliters that should be added. If not, this can be calculated from the number of nanomoles of primer given in the data specification sheet.

36

Patricia de Winter

Fig. 1 Agarose gel electrophoresis of PCR products to determine size and specificity

3. Vortex the tubes briefly and incubate at room temperature for 5 min to allow the primer to dissolve completely. Vortex again and centrifuge the tubes briefly to collect the contents. 4. Prepare a 10 mmol/L working solution of each primer; pipette 10 μL of stock solution into a labeled 1.5 mL tube and add 90 μL nuclease-free water. Vortex to mix and centrifuge briefly to collect contents. 5. For each primer pair (assay) to be validated prepare a qPCR master mix as follows: pipette a 45 μL aliquot of SYBR Green qPCR 2 mix add 4.5 μL of forward and 4.5 μL of reverse primer working solution and 18 μL of nuclease-free water. The final concentration of each primer in the reaction is 500 nmol/L. 6. Vortex the mastermix, briefly centrifuge to collect the tube contents and pipette 16 μL of mastermix into each of four 100 μL PCR tubes. 7. Add 4 μL of yeast tRNA to the first tube—this is the no template control (NTC) and cap the tube. 8. Add 4 μL of pooled cDNA to the remaining three tubes and cap the tubes. 9. Place the tubes into the qPCR machine and run the following program:

VEGF Isoform Quantification by RT-qPCR

37

95  C for the hot start, at the time recommended by the kit manufacturer. 40 cycles of 95  C denaturation and 60  C combined extension/annealing at the times recommended by the kit manufacturer. Collect fluorescence data in the green channel at the end of the annealing extension step. Melt—ramp from 55 to 95  C 10. When the run is complete open the machine and remove the tubes. Check for amplification in the NTC and cDNA tubes. If amplification is present in the NTC this may indicate primer dimer products or contamination with template. Check the melt curves for a single peak for each assay and for the presence of primer dimer fluorescence, which will usually be an additional peak at a lower temperature. Primer dimers in the NTC are not a problem. Primer dimers in samples interfere with quantification. If samples are free of primer dimers, the assay is acceptable. 11. It is strongly recommended that reactions are not opened in the same room where qPCR reactions are prepared. Take the unopened reactions to a post-PCR room where gels can be run. 12. For each assay, label a 1.5 mL tube with the gene symbol. 13. If there was no amplification in the NTC it need not be run on the 2% agarose gel. If the NTC was positive running it on the gel will identify the presence of primer dimers or contamination. Label an NTC tube if necessary. 14. Combine the PCR products for each assay into the corresponding 1.5 mL tubes and pipette any positive NTC products into their corresponding 1.5 mL tubes. 15. Prepare an aliquot of 1 kb ladder by adding loading buffer to the recommended volume of ladder. 16. Add the recommended volume of loading buffer to each sample tube. 17. Prepare the running buffer: add the DNA electrophoresis stain to the TBE buffer—follow the manufacturer’s recommendation for the correct dilution of stain. Adding the stain to the buffer rather than the gel prevents the formation of a DNA stain front as many stains are electrophoresed along the gel. 18. Immerse the agarose gel into the electrophoresis tank and fill the tank with TBE buffer until the gel is covered. 19. Pipette the 1 kb ladder into a well and the samples into the remaining wells. If there are many samples it may be preferable

38

Patricia de Winter

to load two lanes with ladder or load the ladder in a middle lane. 20. Put on the tank lid, connect the electrodes and run the electrophoresis at 70–80 V for 25–30 min or until the ladder has fully resolved. 21. Transfer the gel to a gel visualizer. An image can be taken with a mobile phone for reference if image software is not available. If there is more than one band or the band is not approximately of the expected size, it suggests that the assay is not specific and primers should be redesigned. Occasionally primer dimer products are visible as a faint additional short fragment on a gel even though they were not visible on the melt curve, if this is the case the assay is acceptable. See Fig. 1 for an example of PCR products on a gel. 22. Excise the specific bands from the gel using a sharp scalpel. Trim away excess gel. Transfer the gel bands to labeled 1.5 mL tubes. 23. Weigh the gel piece and add 3 volumes of gel dissolving buffer from the agarose gel DNA extraction kit to 1 volume of gel (100 mg ¼ approximately 100 μL). 24. Incubate at 50  C for 10 min or until the gel slice has completely dissolved. Mixing the tube contents by vortexing every 2–3 min will speed up the process. The gel must be completely solubilized. 25. Add 1 gel volume of isopropanol to each sample. Mix by inverting the tubes several times. 26. Pipette each sample into a silica spin column provided with the kit. Centrifuge for 1 min at >8000  g. Discard the flow through. 27. Add 500 μL of gel dissolving buffer to the column and centrifuge for 1 min at 8000  g to remove any remaining traces of agarose. Discard the flow through. 28. Add 700 μL of wash buffer to the column and centrifuge at >8000  g. Discard the flow through. 29. Transfer the column to a fresh 1.5 mL tube and centrifuge for 1 min at 17,900  g to dry the membrane. 30. Place the column into a clean 1.5 mL microcentrifuge tube, add 30 μL of TE buffer onto the center of the membrane, incubate on the bench for 4–5 min and then centrifuge for 1 min at 17,900  g. 31. Follow the steps in Subheading 3.4 to quantify the DNA on the Nanodrop but add TE buffer as the blank and select DNA rather than RNA. The Nanodrop is reliable for the

VEGF Isoform Quantification by RT-qPCR

39

quantification of PCR products as they will not contain RNA and are purified. Make a note of the DNA concentration for each assay. 32. Calculate the molecular weight of the amplicon by multiplying the average molecular weight of a base pair, 660 Da, by the amplicon length. A mole is 6.02  1023 molecules (Avogadro’s constant) and also the molecular weight of a substance in grams. The number of copies of amplicon in 1 ng therefore ¼ 6.02  1023/amplicon molecular weight  109. Multiply this by the concentration of purified product obtained from the Nanodrop to calculate the number of copies of amplicon per μL. 33. Dilute the purified products to 1010 copies/μL or 109 copies/ μL if the yield is lower with TE buffer. 34. Dilute 5 μL of this to 5  107 copies/μL with tRNA 500 ng/μL. 35. Pipette 50 μL of this into the first well of an 8-strip tube. Label this 108: the next tenfold dilution will be the highest standard that will be used in the qPCR and 2 μL will be used in a 10 μL reaction, so the total copies per reaction will be 107. 36. Make seven further tenfold dilutions starting with 5 μL of 108 copies per 2 μL and adding 45 μL tRNA. The last tube will the 101 standard and will contain 10 copies in 2 μL. Cap the tubes, vortex to mix and centrifuge briefly to spin down the contents. 37. Return to the qPCR room to prepare the reactions. Some qPCR machines require a reference dye to be added to reactions. If this is the case, reduce the amount of water and replace with the designated volume of reference dye. For each assay prepare a qPCR master mix as follows: pipette 150 μL of SYBR Green qPCR 2 mix and add 15 μL of forward and 15 μL of reverse primer working solution and 60 μL of nuclease-free water. The final concentration of each primer in the reaction is 500 nmol/L. This time the total reaction volume will be 10 μL rather than 20 μL. This reduces the amount of reagents and reduces costs without compromising assay performance. 38. Vortex the master mix, briefly centrifuge to collect the tube contents and pipette 8 μL of mastermix into each of 24 100 μL PCR tubes or 24 wells of a PCR plate 39. Add 2 μL of yeast tRNA to the first three tubes/wells—(this is the NTC). Cap if using tubes to avoid contaminating the reactions.

40

Patricia de Winter

40. Add 2 μL of standard in triplicate to the remaining 21 tubes starting with the lowest standard and cap the tubes or seal the plate once all reactions are pipetted. 41. Place the tubes into the qPCR machine and run the program specified in step 9. Indicate which reactions are NTCs and which are standards in the software and input the concentration of each. 42. Once the run has completed, transform the y-axis to a log scale and examine the amplification curves—they should be parallel with each other with little variation between triplicates at higher concentrations and some degree of variation at the lowest concentrations due to random sampling variation. Analyze each standard curve—all qPCR machines calculate efficiency from the slope and return a value, which may be expressed as a proportion (0–1), a percentage (0–100), or a ratio (1–2). It is generally accepted that an efficiency of >90% is acceptable. A lower efficiency reduces the sensitivity of the assay and should be redesigned or used with caution if expression of the transcript is likely to be low as false negatives may be a problem. If the 10 copies standard does not amplify this may indicate that the sensitivity of the assay is low and only samples that express 100 copies per reaction or more will be quantifiable. Check that the assay is linear over 7 log—this is reflected by the R2 value which should be at least 0.98 but is often above 0.99. 43. Check the melt curves—look for primer dimers in the low concentration standards. These could be a problem in samples if expression is low. If primer dimers are present, consider adding an additional step to subsequent qPCR cycles so that fluorescence measurement takes place at a temperature at which primer dimers have melted and thus cannot interfere with quantification. 44. Standards may be stored at 20  C and will tolerate multiple freeze thaw cycles but care must be taken not to introduce DNases or degradation will occur. 3.8 qPCR of Experimental Samples

See Note 12. The procedures below are performed for each gene assay to be run, including the reference genes. 1. Calculate how many reactions will be required in total (triplicates of the NTC, each standard, and samples) and add 10% to allow for pipetting loss. 2. Prepare a master mix for the given number of samples by multiplying the following by the number calculated in step 1:

VEGF Isoform Quantification by RT-qPCR

41

Fig. 2 Log-transformed amplification curves for a set of standards (purple–pink) (107–101 copies per reaction) and two samples (blue and green) plus an NTC (black). The amplification curves of the samples are parallel with those of the standards implying that samples and standards have amplified with equal efficiency. The NTC has not amplified

5 μL of 2 SYBR Green qPCR mix 0.5 μL of forward primer working solution 0.5 μL or reverse primer working solution 2 μL nuclease-free water. 3. Cap the tube, vortex to mix thoroughly and briefly centrifuge to collect the tube contents. 4. Pipette the master mix into the required number of wells calculated in step 1. 5. In triplicate, pipette 2 μL of tRNA into the NTC wells, 2 μL of each standard into standard wells starting with the lowest concentration, and 2 μL of sample into sample wells. Make note of which reactions are in which wells—a paper template may assist in keeping track. 6. Seal the plate and if recommended by the qPCR machine manufacturer, centrifuge to collect the contents. 7. Place the plate into the machine and run the protocol specified in step 9 of Subheading 3.7. 8. Once the run is completed, check the melt curves for primer dimers and that the melt peak of samples corresponds to that of standards.

42

Patricia de Winter

9. Check log-transformed amplification curves for signs of qPCR inhibition—sample amplification curves should be parallel with those of standards and there should be no crossover. Samples exhibiting inhibition cannot be reliably quantified. See Fig. 2 for an example of amplification curves. 10. Analyze the data to convert Cq to quantities using the machine software and export to a CSV or Excel file. 11. To quantify any qPCR inhibition, data can also usually be exported in LInRegPCR format and opened in this Excel add-in, which is available at this link, as is the instruction document [10]: https://www.medischebiologie.nl/files/? m a i n ¼fi l e s & fi l e N a m e ¼L i n R e g P C R . z i p & description¼LinRegPCR:%20analysis%20of%20quantitative% 20PCR%20data&sub¼LinRegPCR 12. LinRegPCR may be used to determine whether samples should be omitted from analysis due to the presence of qPCR inhibition. Compare the efficiencies of samples with the range of efficiencies calculated for the standards and if these fall below the lowest value, those data may not be reliable. 3.9 Data Normalization

See Note 13. The method below is for manually calculating a normalization factor when stable reference genes can be identified from other data and normalization software will not be used (see Note 13). 1. Select three or four stably expressed reference genes from RNAseq or similar dataset for the samples. Data must be quantities, either relative expression or copies. 2. Paste the expression data into a spreadsheet with gene identifiers as the headers and the samples IDs in the first column (see Table 1). If you have many samples and very high expression (>100,000 copies per sample) for any of the genes, divide them by 10 or 100 and replace the data for those genes with these values. This is for pragmatic reasons: Excel may not be able to calculate a geometric mean if a product is greater than the spreadsheet’s maximum number permissible and an error, #NUM, will be returned. 3. In the column to the right of the data calculate the geometric mean for each sample: ¼Geomean(number 1, [number2], . . .). 4. At the bottom of the same column, calculate the grand geometric mean using the function in step 3. 5. In the column to the right of the geometric mean, calculate the normalization factor: ¼geomean for the sample/grand geomean. This step simply ensures that the normalization factor retains the order of magnitude of expression of the genes of interest following normalization.

VEGF Isoform Quantification by RT-qPCR

43

Table 1 Example of how to set up a spreadsheet to normalize vegf gene of interest data Ref Ref Gene A Gene B

Ref Gene C

GOI Normalized Geomean NormFact copies/rxn copies/rxn

Sample 1048 1 replicate 1

1109384 156

5660.5 0.68

1299

1915

Sample 1110 1 replicate 2

1215560 167

6085.2 0.73

1336

1832

Sample 2610 1 replicate 3

1381900 175

8578.0 1.03

1912

1860

Sample 2113 2 replicate 1

2699638 203

10,501.1 1.26

2501

1987

Sample 3351 2 replicate 2

2331972 211

11,813.9 1.42

2723

1923

Sample 1343 2 replicate 3

2491034 233

9203.2 1.10

2300

2085

Grand geomean

8343.5

6. In the column to the right of the normalization fraction paste in the data for vegf gene (Gene of Interest, GOI). 7. In the column to the right of the GOI data, normalize the gene of interest data: divide the copies per reaction by its corresponding normalization factor to give normalized copies per reaction. Calculation of sample, RT and group means may now be performed.

4

Notes 1. There are many manufacturers of RNA extraction kits, RT kits, and qPCR kits. Here the standard methodology that is common to most kits is described, rather than specifying kits from named manufacturers. 2. If the starting material is tissue, I strongly recommend placing freshly dissected tissue immediately into an RNA stabilizing solution such as RNAlater etc. to prevent RNA degradation. This also permits long-term storage of tissue without RNA degradation and protects it from degradation should accidental thawing occurs. For long-term storage, the supernatant may be removed after 24 h at 5  C and the samples placed at 80  C. 3. Silica-based kits, whether columns or magnetic beads, offer safe, non-hazardous extraction of RNA, which can be

44

Patricia de Winter

automated if sample numbers are large. Here I describe the protocol for RNeasy columns, which is widely used, but this can be adapted for any silica-based method flowing that manufacturer’s recommendation. 4. The most common equipment used for analyzing RNA integrity is the Agilent Bioanalyzer. There are other options, such as the BioRad Experion, Agilent Tapestation, or a denaturing formaldehyde-agarose gel electrophoresis, but here I will describe the Bioanalyzer protocol as that is the most widely used. 5. Yeast tRNA is a carrier and does not participate in the RT or qPCR reactions. It can prevent loss of sample by adhesion to tube walls when the cDNA amount is low. 6. In most cases it is preferable to design primers that span an exon junction so that contaminating gDNA will not be amplified if present. 7. If intron-spanning primers could not be designed for one or more target sequences, a gDNA removal step may be performed on column at this stage, or it may be performed later on the purified RNA. On column will require an incubation period and an additional wash step. Removal from a liquid RNA sample will require DNase and buffer and a clean up to remove the enzyme. Alternatively, the Qiagen Quantitect Reverse Transcription kit contains a gDNA wipeout buffer that is added for 2 min immediately prior to reverse transcription. 8. Fluorimetry measures only RNA concentration, and not other nucleic acids, using an RNA-specific fluorescent dye. It does not provide a measure of RNA purity or integrity. 9. The range of the RNA 6000 Nano assay is 5–500 ng/μL of RNA. It is best to load roughly the same amount of RNA for each sample, although this is not essential if there is less to spare of some samples. If RNA concentration is above 500 ng/μL dilute to the desired concentration with nuclease-free water—a total volume of 3–5 μL is all that is required. The gel is filtered and aliquots of 65 μL are prepared in advance and stored at 5  C until required. A new syringe is provided with each kit and should be changed when the new box of chips is first used. All solutions are reverse pipetted into the Bioanalyzer chip to prevent air bubbles forming (draw up more than required and expel liquid to the first stop only). For RNA, the base plate should be at position C. The RNAs used in any given experiment should be of similar integrity (RNA Integrity Number, RIN). RNA above RIN 7 is accepted as “good” integrity for qPCR.

VEGF Isoform Quantification by RT-qPCR

45

10. Please see Subheading 1 and refs. 6–9 for discussion of RT priming strategies. Reverse transcription is least variable when performed in a qPCR machine as the temperature control is better regulated than in hot blocks, ovens, or water baths. RNA should be maintained on ice. If only a few genes are to be quantified by qPCR, a 10 μL reaction can be prepared rather than a 20 μL reaction. I recommend reverse transcribing 10% of RNA samples in duplicate with a minimum of 2 samples in duplicate if there are fewer than 20. The maximum recommended amount of RNA that can be reverse transcribed differs between kits but is generally 1000–2000 ng. 11. Primers may be stored at room temperature upon receipt. Once reconstituted they should be stored at 20  C long term. As it may not be known which experimental samples express VEGF, it is recommended that a pool of all cDNAs is prepared by taking a small aliquot of each diluted cDNA sample prepared in Subheading 3.6 and combining into a 1.5 mL tube. The Invitrogen e-Gel system with size select precast gels is particularly time saving for purification of DNA as the products can be aspirated directly from the gel and purified using Microclean in a 10 min procedure. If this system is not available, PCR products can be purified from a regular agarose gel or the products from one reaction can be run on a gel and the remaining reactions purified using a silica column kit. Here I describe purification of products from excised gel slices. Reagents for qPCR should be thawed thoroughly before use, vortexed, and briefly centrifuged to collect tube contents. It is not necessary to maintain them on ice as modern mixes are hot-start, but they should be returned to the freezer once reactions have been pipetted. 12. Work surfaces should be cleaned immediately prior to starting qPCR to remove any nucleases or nucleic acids that could contaminate the reactions, Reagents for qPCR, tRNA, standards, samples and primers, should be brought to room temperature, vortexed to mix and centrifuged briefly to bring down tube contents before opening. This is especially important for tubes containing potential template as aerosol contamination of reactions may easily occur—a single copy of contaminating template can produce a positive amplification curve within 40 cycles. Tubes containing template should be opened last, once the master mix is prepared and pipetted. Contamination can be avoided with good practice. It is better to keep all reactions for one gene within one run if possible, however if a second plate is required the standards act as interplate calibrators. The standard curve can be run in duplicate if this permits fitting samples for a gene within one run. A 96 well plate format is assumed, but there are also platforms for

46

Patricia de Winter

100, and 384 reactions in common use in research labs. If the qPCR machine used requires a reference dye to be added to reactions, reduce the amount of water and replace with the designated volume of reference dye. 13. The normalization factor can be calculated manually in Excel [4] but most researchers find it convenient to import their data into reference gene normalization software. geNorm software has been widely used for identification of stable reference genes but it is now only available as part of qBase commercial software upon payment of a license fee. A free trial for 2 weeks is available. Normfinder software for R and Excel is available free of charge here: https://moma.dk/normfinder-software. In practice three stable reference genes generally yield good normalization. If qPCR is being used to validate RNAseq or other genome-wide data obtained from the same samples, the most stable reference genes can be selected from the latter. Stable reference genes are those where expression does not change irrespective of the experimental conditions. A useful means of determining whether reliable normalization has been achieved is to check the gene of interest copy numbers for RT replicates before and after normalization; the normalized values for two replicates should similar whereas non-normalized values may be disparate. References 1. Wittwer CT, Herrmann MG, Moss AA, Rasmussen RP (1997) Continuous fluorescence monitoring of rapid cycle DNA amplification. BioTechniques 22:130–138 2. Morrison TB, Weis JJ, Wittwer CT (1998) Quantification of low-copy transcripts by continuous SYBR Green I monitoring during amplification. BioTechniques 24:954–958. 960, 962 3. Bustin SA et al (2000) Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 25:169–193 4. Vandesompele J et al (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3:0034 5. Bustin SA et al (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55:611–622 6. Lekanne Deprez RH, Fijnvandraat AC, Ruijter JM, Moorman AF (2002) Sensitivity and

accuracy of quantitative real-time polymerase chain reaction using SYBR green I depends on cDNA synthesis conditions. Anal Biochem 307:63–69 7. Nam DK et al (2002) Oligo(dT) primer generates a high frequency of truncated cDNAs through internal poly(A) priming during reverse transcription. Proc Natl Acad Sci 99(9):6152–6156. https://doi.org/10.1073/ pnas.092140899 8. Bustin SA, Benes V, Nolan T, Pfaffl MW (2005) Quantitative real-time RT-PCR – a perspective. J Mol Endocrinol 34(3):597–601 9. Stangegaard M, Dufva IH, Dufva M (2018) Reverse transcription using random pentadecamer primers increases yield and quality of resulting cDNA. BioTechniques 40:649. https://doi.org/10.2144/000112153 10. Ramakers C, Ruijter JM, Deprez RH, Moorman AF (2003) Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci Lett 339(1):62–66

Chapter 3 Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic Endothelial Cells Theresa Connor, Nerida Sleebs, and Zerina Lokmic-Tomkins Abstract Multiparameter fluorescence-activated cell sorting (FACS) procedure separates target cells from a total population of cells by using specific signatures that the target cell expresses on their cell surface. For human lymphatic endothelial cells (LECs) this relates to cell surface expression of the CD34LowCD31HighVEGFR3HighPodoplaninHigh profile that permits their separation from blood vascular endothelial cells and other cells likely to be present in the digested tissue sample. In addition, FACS allows the evaluation of LEC size, volume, granularity, and purity at the time of sorting. Key words FACS, Lymphatic endothelial cell, Cell surface markers, Cell culture

1

Introduction Isolating lymphatic endothelial cells (LECs) facilitates in vitro studies at genetic, epigenetic, lineage and cell biology levels using cell isolates or LEC in in vitro models that assist our understanding of processes governing lymphangiogenesis and lymphatic regression. For diseased lymphatics, this extends to identifying therapeutic targets that may impact the clinical management of lymphatic conditions, such as lymphatic malformations, Kaposiform lymphangioendotheliomas, Gorham-Stout disease, primary lymphoedema, cancer-related lymphoedema and secondary lymphoedema arising from non-cancer causes [1], and providing information to families where genetic testing is warranted [2]. Furthermore, isolated lymphatic malformation LECs can be implanted into mouse xenografts, where they form lymphatic-malformation-like vessels [3], which can also be used for testing therapeutic agents. Baudin and colleagues estimated that, based on podoplanin expression as identified by D2-40 antibody, the lymphatic vessel density in normal adult skin and relative to the skin surface is 3.58
1.31 mm1 [4]. This means that it is to be expected that

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_3, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

47

48

Theresa Connor et al.

the cell yield will not be abundant and that cell isolates are likely to contain a high number of fibroblasts, keratinocytes, and vascular smooth muscle. There is also evidence that LECs are not completely free of CD34 expression when examined by flow cytometry, meaning that they are not that clearly distinguished from human vascular endothelial cells [5]. Similarly, CD34 expression has been reported in other types of lymphatic anomalies [1]. Our understanding regarding the effects of the cell digestion process and use of different types of cell culture media on the profile of the selected markers and whether there is a change of endothelial cell markers expression as cells are passaged over time is poor. Isolating and culturing LECs from diseased and healthy tissue will also facilitate answering these questions on cell lineage identity once grown in vitro. The protocol presented below builds on our previously published methods [5, 6] and it provides updates on the sample digestion method, cell staining, and instrument set up needed to isolate CD34LowCD31HighVEGFR-3HighPodoplaninHigh LECs, which in our lab are derived from neonatal foreskin and pediatric surgically excised lymphatic malformation tissue.

2

Materials All reagents and solutions used in this protocol are prepared and stored as per manufacturer instructions unless specifically stated otherwise. Enzyme solutions used for tissue digestion are prepared on the day of use. Before use, the endothelial cell media, cell culture buffers and enzyme solutions need to be pre-warmed at 37  C for 20 min. The solutions stored at 4  C as per the manufacturer’s instructions are all warmed to room temperature (see Note 1).

2.1 Culture Medium and Solutions for Maintaining Foreskin and Lymphatic Malformation LEC Cultures

1. Endothelial cell media: EGM-2 MV Bullet Kit (Lonza, cc3202) supplemented with 50 ng/mL VEGF-C (R&D, 2179-VC025). The kit contains the following components: human EGF, hydrocortisone, Gentamicin (GA-1000), fetal bovine serum (FBS), VEGF, human FGF-b, R3-IGF-1, and ascorbic acid. 2. Calcium and magnesium-free phosphate-buffered saline (PBS): The PBS is prepared by dissolving 8.752 g NaCl, 1.416 g Na2HPO4  2H2O and 0.395 g KH2PO4 in 1 L of milliQ H2O. The pH is adjusted to 7.4. The PBS is then filtersterilized into 50 mL aliquots and stored at 4  C. Before use, we add 100 U/mL penicillin and streptomycin. 3. Human fibronectin (Sigma-Aldrich, F2006): Dissolve 1 mg of fibronectin in 10 mL of sterile H2O. Store reconstituted solution in 100 μL aliquots at 20  C. On the day of use add

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

49

10 mL of PBS to 100 μL of fibronectin aliquot (to give a final working concentration of 10 μg/mL). 4. StemPro® Accutase® Cell Dissociation Reagent (Life Technologies, Gibco®, A11105-01). 5. 0.4% Trypan Blue solution. 6. Dimethyl sulfoxide (DMSO, Sigma-Aldrich, D4540). 7. Freezing media: 90% FCS/10% DMSO (v/v). 2.2 Enzymatic Digestion

1. Enzymes: dispase II (Roche Applied Biosciences, 4942078001), collagenase type II (Worthington Lab, 4176) and DNAse I (Roche Applied Biosciences, 11284932001). 2. Enzyme media: 0.04% dispase II, 0.25% collagenase II, and 0.01% DNase I in sterile PBS. 10 mL of enzymatic media is required per 1 g of tissue. First, weigh the required weight of dispase and collagenase into a sterile 50 mL tube then add the required volume of PBS and incubate for 30 min with shaking at 37  C to dissolve. Once dissolved, filter-sterilize (0.22 μm filter) dispase/collagenase solution before aseptically adding the required amount of DNAse I in the laminar flow hood.

2.3 Antibodies and Buffers

1. Antibodies: PE-conjugated VEGFR-3 (BioLegend, 356204, clone 9D9F9, clone WM59, used at 1:50 dilution), PE-Cy7-conjugated CD34 (BioLegend, 343516, clone 581, used at 1: 200); APC-conjugated mouse anti-human CD31 (Biolegend, 303116, clone WM59, used at 1:100); Alexa 488-conjugated rat anti-human Podoplanin (BioLegend, 337006, clone NC-80, used at 1:200). 2. Isotype antibodies: PE-conjugated mouse IgG1, κ-isotype (BioLegend, 400112, clone MOPC-21, 1:50 dilution), PE-Cy7-conjugated mouse IgG1, κ-isotype (BioLegend, 400126, clone MOPC-21, 1:200 dilution); APC-conjugated mouse IgG1, κ-isotype (BioLegend, 400120, clone MOPC21, 1:100 dilution); and Alexa 488-conjugated mouse IgG2a, κ-isotype (BioLegend, 400525, clone RATK2758, 1:200 dilution). 3. FACS buffer: 5% FBS/PBS.

2.4

Equipment

All procedures are performed aseptically in Class II laminar flow safety hood equipped with UV light for decontamination. The laminar hood is always decontaminated with UV light and the working space cleaned with 70% ethanol before tissue processing. Dissection equipment is washed in laboratory-grade detergent, soaked in 70% ethanol then autoclaved at 121  C for 20–30 min (depending on the system used). Additional equipment:

50

Theresa Connor et al.

• 25 and 75 cm2 tissue culture flasks and 90 mm Petri dishes. • Sterile 5 and 10 mL tissue culture pipettes. • 15 and 50 mL sterile centrifuge tubes. • 100 and 70 μm cell strainers. • 0.22 μm filtration membranes. • Vacuum pump to aid filtration process. • 3 mL sterile syringe plunger with black rubber tip. • Scalpel handle, scalpel blades, forceps, small scissors with straight ends. • Hemocytometer. • Temperature-controlled water bath or incubator with a shaker. • Centrifuge to hold 15 and 50 mL tubes. • Cell culture incubator with temperature and gas composition control.

3

Methods All procedures are carried out at room temperature using sterile and/or aseptic practices. The endothelial cell media is prepared by warming the 500 mL EGM-2 base media at 37  C and gently adding thawed individual components aseptically added to the EGM-2 media bottle in the laminar flow cell culture hood (Class II biosafety cabinet). The VEGF-C component is added last. The endothelial cell media is then aseptically aliquoted into 50 mL sterile tubes and stored at 4  C until use. Once all components are added to the media, we refer to this media as ‘complete endothelial cell media’ (see Notes 1 and 2).

3.1 Sample Collection

1. Pre-weigh sterile container containing complete endothelial cell media. Tissues collected at the time of surgery will be added to this sterile container containing a known amount of complete endothelial cell media. The amount of media will vary depending on the size of the tissue. Lymphatic malformation samples can be of significant size at the time of collection. For foreskin samples, 5 mL of media is used, for lymphatic malformations, 10–15 mL of media is often needed. 2. Before processing the tissue, weigh the container with the specimen then subtract the weight of the empty container and the volume of media in the container to obtain actual tissue weight. Thereafter, for every 100 mg of tissue, add 1 mL of enzyme media.

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

3.2 Sample Processing and Enzymatic Digestion

51

1. Turn on the decontaminated laminar hood and clean the working space with 70% ethanol. 2. Transfer tissue into a petri dish (90 mm) and using scissors cut tissue into smaller pieces. Using sterile forceps and scalpel mince the tissue into 1–2 mm pieces (see Notes 3 and 4). 3. Transfer minced tissue aseptically into a sterile 50 mL tube and add 1 mL of enzyme media per 100 mg of tissue. Rinse the petri dish with another 1 mL of enzyme media to dislodge any tissue stuck to the dish. This volume comes from the total volume calculated as needed for tissue digestion (see Notes 5– 7 and 10). 4. Incubate the 50 mL tube at 37  C with shaking for 30–90 min. 5. To judge if the incubation is sufficient, simply examine the solution for undigested tissue present (see Notes 6 and 7). 6. Following incubation, pass the cells through a 100 μm cell strainer placed in a sterile 50 mL tube. Using a sterile 3 mL syringe piston with a rubber plunger, grind down the remaining tissue until only traces of the extracellular matrix remain. Tissue that is not well digested will create a big mass of tissue that congeals and cannot be passed through the strainer. 7. Wash the cell strainer with a 1:1 ratio of endothelial cell media to enzyme media to inactivate the enzymes still present in the solution and to wash off cells still adhered to the membrane. 8. Centrifuge the cells at 300  g for 5 min and remove the supernatant. Wash the cells three times in sterile PBS/PenStrep (10 mL per wash). Each time discard the supernatant and gently disrupt the cell pellet before adding new PBS/PenStrep. 9. After the third wash, add complete endothelial cell media to resuspend the cell pellet. This resuspension volume may need to be adjusted if the cell pellet is large, but usually, 1–2 mL of media will suffice. The cell suspension is then passed through a 70 μm strainer to remove cell debris and the membrane washed with 1 mL of complete endothelial cell media. 10. Take a 10 μL sample of cell suspension and resuspend in 90 μL of Trypan Blue and load 10 μL in a hemocytometer and count cells as specific to the hemocytometer in your laboratory (see Note 11). 11. Once the cell count is done, calculate the volume of cell suspension needed to seed 5  105 cells in a 25 cm2 flask, 2  106 cells per 75 cm2 flask, and 5  106 cells in 150 cm2 flasks. Cells can be seeded in 5 mL of complete endothelial cell media if using 25 cm2 flask, 10 mL per 75 cm2 flask, and 20 mL for 150 cm2 flask (see Notes 7–9).

52

Theresa Connor et al.

12. Plate the suspended cells in fibronectin-coated flasks and incubate at 37  C, 5% CO2 and 21% O2 in a humidified incubator. Check cells after 24 h to determine cell attachment (see Note 12). 13. After 24 h, wash away unbound cells (3  5 min PBS/PenStrep) and add new complete endothelial cell media. Return to the incubator. 14. Change the media every second day. When cells are approximately 80% confluent, there are enough cells to undertake a cell sorting procedure (see Note 13). 3.3

Sample Staining

1. Once the cells have reached 80% confluence, aspirate medium and wash cells with PBS/PenStrep. 2. Detach adherent cells by incubating cells with 7 mL of Accutase per 150 cm2 flask for 5 min at 37  C. For a 25 cm2 flask, use 1 mL of Accutase (see Note 14). 3. Inactivate the Accutase by adding three volumes of complete endothelial cell medium. 4. Transfer the cell suspension to a new sterile 50 mL tube. 5. Rinse the flask with 10 mL of PBS/PenStrep to collect the remaining cells. 6. If the cells still adhere to the flask, add another 1 mL (if using 25 cm2 flask) or 7 mL (if using 150 cm2 flask) of Accutase and incubate for a further 1–2 min at 37  C. 7. Rinse the flask with 10 mL of PBS/PenStrep to collect the remaining cells. 8. Centrifuge the cells at 300  g for 5 min. 9. Aspirate supernatant and resuspend the cells in 5 mL of PBS/PenStrep. 10. Centrifuge cells at 300  g for 5 min. 11. Aspirate supernatant then repeat cell wash. 12. Resuspend the cells in 2 mL of FACS buffer (PBS-containing 5% FBS). 13. Count the number of viable cells using Trypan Blue exclusion. Mix 10 μL of cell suspension with 90 μL of 0.4% trypan blue in an Eppendorf tube. Transfer 10 μL of the stained cell suspension to a hemocytometer to count cells. 14. To prepare the antibody solutions for cell staining, the antibody was titrated to dilutions for staining 1  106 cells in a total staining volume of 100 μL sterile 5% FBS/PBS buffer per tissue sample. Antibody stain occurs as a single antibody or “cocktail” stain. Single antibody stain is used to stain 1  105 cells to set up single stain controls. “Cocktail”

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

53

antibody stain is used to stain all cells destined for LEC isolation. Similarly, prepare 100 μL diluted isotype control antibody mixture for staining 10,000 cells per isotype control to facilitate the setting of flow cytometry gates. Antibodies and isotype controls are kept on ice until use. All staining is performed in 5 mL flow cytometry tubes. 15. Before staining with conjugated antibodies and isotype controls, centrifuge cells at 300  g for 2 min, aspirate the supernatant then suspend the cells in 100 μL sterile 5% FBS/PBS buffer per sample. Incubate cells on ice for 20 min. 16. Centrifuge cells at 300  g for 5 min, aspirate FBS/PBScontaining supernatant then resuspend cells in the conjugated antibody cocktail. Also, stain a smaller aliquot of cells (1  105) in a 100 μL isotype control antibody mixture. 17. Incubate cells on ice for 20 min in the dark. 18. Unbound antibody is removed by adding 2 mL of 5% FBS/PBS solution and centrifuging the cells at 300  g for 5 min. 19. Aspirate supernatant and repeat the cell wash. 20. Resuspend the cell pellet for the isotype control and the antibody-stained sample to be sorted in 300 μL of 0.5 mg/ mL propidium iodide/2% FBS/PBS solution. 21. Keep all samples on ice, protected from direct light, until sorting. Avoid excessive lag time between staining and sorting of the cells to help cell viability. 3.4 Instrument Set Up and Cell Sorting

Operation of fluorescent cell sorters should only be done with the appropriate training or with help from a competent FACS operator. 1. To set up the instrument, insert the 100 μm nozzle into the instrument and start the stream. 2. Visually check that the stream is in the correct position and allow the stream to stabilize for 5–30 min, depending on the instrument used. 3. Use alignment beads, such as commercially available fluorescent particles and drop delay beads to ensure cytometer settings are optimal. This may require assistance from trained FACS operators. 4. Run single stain compensation controls with each sort to generate the compensation matrix and to eliminate bleed-through of fluorochromes such as PE into the PE-Cy7 channel. 5. Set up an appropriate gating strategy on the FACS software to exclude debris, doublet and dead cells. Select cell population to sort, for example, CD34LowCD31HighVEGFR-3HighPodoplaninHigh, and assign a location of the collection tube.

54

Theresa Connor et al.

6. Set up sterile 15 mL Falcon collection tubes with approximately 1 mL of complete media in the collection chamber. 7. Sort the LECs on the instrument directly into complete endothelial media using the purity mask sort stringency settings to ensure additional purity to the cell population. The flow rate should be kept reasonably low to increase cell viability. Results of the sorting process, including the gating strategy, are presented in Fig. 1 (see Note 15). 8. Sort the purified human LECs on a multiparameter fluorescence-activated cell sorting instrument into a 15 mL sterile tube containing 2 mL of complete endothelial cell media. 3.5 Cell Culture Post FACS Sorting

1. After the sorting process, centrifuge cells at 300  g for 5 min and aspirate the sorting buffer added to collection media during the sorting process. 2. Resuspend the cells in complete endothelial media as appropriate for the cell flask used. If less than 50,000 cells are sorted, the cells are cultured in a fibronectin-coated 25 cm2 flask. Where more than 50,000 cells are sorted, use a 75 cm2 fibronectin-coated flask to culture the cells. Cells are cultured at 37  C in a 5% CO2, air humidified incubator. Replace the complete endothelial media after 2 days. 3. Thereafter, continue to change complete endothelial media every 2 days. Continue to visually assess the cell culture to identify the presence of contaminating cells. 4. To validate LEC phenotype, cells can also be stained immunohistochemically for PROX-1, VEGFR-3, Podoplanin, CD34, and CD31 expression when the cells are first split for further expansion [3] (see Note 16).

3.6 Subculturing LECs

Once LECs reach 80% confluence, they can be split at a ratio of 1:3 and passaged into 25 or 75 cm2 flasks, depending on the cell number available. 1. After removing the cells from the cell incubator, aspirate the supernatant and aseptically wash the cells three times in PBS/PenStrep. 2. Add 1 mL of warm Accutase® solution to cover the cells. 3. Place the flask in the cell culture incubator for 5 min with the first check for cell detachment at 3–4 min. 4. To check the cells, view the flask under a phase-contrast microscope. If the cells have not detached, tap the flask gently to dislodge the cells. If the cells remain attached, return the flask to the cell culture incubator for a further 2 min. Do not incubate the cells longer than 7 min.

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

55

Fig. 1 Gating strategy for multiparameter flow cytometry cell sorting of human lymphatic malformation LECs subpopulations. (a) Cell gate that plots cell FSC against trigger pulse intending to exclude doublets (b) Gating cells on FSC-A versus FSC-H is another way to define single cell populations and to exclude doublets. (c) Further doublet exclusion is performed when the SSC-A parameter is plotted against the SSC-H parameter thus gating for single cells. (d) Once the doublets are excluded through these three gating strategies, single cells of interest are gated for live cells. In our panel set up propidium iodide is used to label dead cells. (e) Cells of interest are gated on SSC and FSC plot to remove any debris that may arise from the tissue digestion process. (f) Gates depict cell subpopulations containing different expression levels of CD34 and CD31 cell surface markers. Cells that are CD34Low and CD31High and cells that are CD34High and CD31High were gated as starting sort population. (g) Cells that CD34Low CD31High are further sub-gated for expression of podoplanin and VEGFR-3. (h) We also examine cells that CD34High/PosCD31High for expression of podoplanin and VEGFR-3. (i) This plot represents the back gating of cells using CD31High expression to visualize from where sorted cells are derived. (j) Gating populations legend. The instrument set up for plots obtained here are: Piezo amplitude ¼ 3.93; Drop delay ¼ 30.2450; Sort mode: 1.0 drop Pure; Drop envelope: 1.0; Drop Sort objective: Purify; Phase mask: 16/16; Extra coincidence bits: 4; Drop frequency (kHz): 39.10. Cytometer model used is influx v7 sorter; Software: BD FACS TM Software, version 1.0.0.650; ValComp: 5.0. FSC forward scatter, SSC side scatter, A area, H height, PI propidium iodide, CD34 cluster of differentiation 34, CD31 cluster of differentiation 31, Pod/P podoplanin, VEGFR-3/V vascular endothelial growth factor 3

5. After cells have detached add 5 mL of complete endothelial cell media to stop the Accutase® reaction. 6. Pass the cell suspension 2–3 times through a sterile 5 or 10 mL cell culture pipette to gently break the aggregates then transfer cells to a sterile 15 mL centrifuge tube.

56

Theresa Connor et al.

7. Centrifuge the tube at 300  g for 5 min, aspirate the supernatant and resuspend the cells in 15 mL of complete endothelial cell media. 8. Using a sterile pipette, draw the cells up and down in the pipette then transfer 5 mL of the cell suspension to three new 25 cm2 fibronectin-coated flasks. 9. Replace the media every 2 days until the cells reach 80% confluency and can be passaged again (see Notes 17 and 18). 3.7

Cryopreservation

Cryopreservation can be performed usually after a second subculture when enough cells can be obtained to preserve. 1. To cryopreserve the cells, detach the cells as described in Subheading 3.3. Once resuspended in complete endothelial cell media, cells need to be counted using a hemocytometer as described in Subheading 3.2. 2. Centrifuge the cells at 300  g for 5 min, wash the cells in 10 mL of PBS/PenStrep then repeat the centrifuge step. 3. Remove the supernatant and resuspend the cell pellet in 90% FCS/10% DMSO (v/v) freezing media aiming to cryopreserve approximately 100,000 cells per mL of freezing media. 4. The cell/freezing media suspension is aliquoted into cryovials (0.5 mL per cryovial) and stored at 4  C for 1 h, followed by 20  C for 1 h then 80  C overnight before transfer to liquid nitrogen for long-term storage. Alternatively, a control rate freezer can be used to cryopreserve the aliquoted cell/freezing media suspension in cryovials before transfer to liquid nitrogen. 5. Cells are thawed by gently warming the cryovial at 37  C for 1–2 min, enough for the frozen content to turn to a ‘slurry’. This ‘slurry’ is then transferred directly into 5 mL of warm (37  C) endothelial cell media. 6. Centrifuge the cells at 300  g for 5 min, remove the supernatant containing DMSO, then resuspend the cells in fresh 5 mL of complete endothelial cell media. 7. Seed the cells into 25 cm2 fibronectin-coated flasks for 24 h. On the following day, change complete endothelial cell media to remove non-adherent cells (see Notes 19 and 20).

4

Notes 1. All consumables and waste are disposed of after use as specified by our institutional occupational health and safety guidelines. 2. Complete endothelial cell media is used for tissue collection, cell isolation and in vitro cell expansion.

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

57

3. To obtain the best cell yield and viability, it has been our experience that the tissue should be processed upon arrival to the laboratory rather than the following day. Similarly, freezing tissue first then trying to isolate cells also yields poor cell numbers. 4. Mincing tissue to 1–2 mm2 pieces seems to be a critical step as mincing the tissue permits better enzyme digestion. 5. Always prepare fresh enzyme digestive media on the day of use as this aids tissue breakdown, particularly if the tissue sample is fibrotic. 6. The volume of 0.04% dispase II, 0.25% collagenase II, and 0.01% DNase I is prepared based on the tissue weight of the sample being processed. Do not overexpose cells to collagenase and dispase as this impacts cell survival. 7. Lymphatic malformation LEC yields are influenced by the nature of the collected tissue. Thus, tissue containing more sclerosed fibrous tissue yields fewer cells. This may be due to either deleterious effects of prolonged exposure to collagenase II and dispase II and/or due to fewer lymphatic malformation vessels likely to be present in sclerosed tissue. It is not unusual that prolonged time is required to digest sclerosed tissue (60–90 min). Wherever possible, avoid using sclerosed samples. 8. Empirically, the neonatal foreskin aged 6 months. This tissue is rapidly digested by the digestive media. Most neonatal foreskin samples and lymphatic malformation tissue never exposed to sclerosing agents (interventional radiology treatment) require 20–30 min of digestion. 9. The final cell yield will depend on the sample size processed. The usual cell yields per 150 cm2 flask range from 7  105 to 1.2  106 of viable cells. 10. If the lymphatic malformation tissue sample has not been previously treated through interventional radiology, ask the operating surgeon or their assistant to cut the target tissue to about 1 cm  1 cm pieces from the overall tissue mass. The operating surgeon is well versed as to the location of the malformation and their help is immensely valuable in increasing the likelihood of getting lymphatic malformation-rich tissue. The surgeon can provide you with lymphatic malformation tissue most distal to the treated area where the lymphatic malformation has previously been treated with interventional radiology agents. This is an important step since lymphatic malformation tissue that was treated with interventional radiology agents are heavily scarred and difficult to digest with enzymatic media. Furthermore, the final cell yield, in our

58

Theresa Connor et al.

hands, is very low. There is also a risk that lymphatic malformation tissue obtained from such a sample may contain lymphatic malformation LECs that may be phenotypically different to non-treated lymphatic malformation tissue. The tissue is transported to the laboratory on ice. 11. As estimated by trypan blue exclusion, approximately 75–90% of nucleated cells survive tissue digestion. The starting cell count reflects all nucleated cells present in the parent tissue (i.e., keratinocytes, leukocytes, macrophages, connective tissue cells, and blood vessel cells). The cell count at 5–7 days reflects cells that have attached, survived, and proliferated during cell culture. 12. Coating tissue culture flasks with fibronectin improves LEC survival. For 25 cm2 flasks, we use 1 mL of 10 μg/mL fibronectin solution to coat the flask. For 75 cm2 we use 4 mL to coat the flask and for 150 cm2 we use 7 mL to coat the flask. Note that 10 μg/mL fibronectin solution can be reused as the sterile reconstituted solution is stable at 4  C for 1 month. Each flask is coated for a minimum of 30 min before use at room temperature. If expecting a large volume of tissue to process, the flask can also be prepared a day earlier and stored at 4  C until use. 13. Longer times between media changes favor survival of fibroblasts and/or vascular smooth muscle cells, even though these cells may be initially present in very small quantities (2–5%). 14. StemPro® Accutase® Cell Dissociation reagent is used to detach cells from the flask during passaging as it preserves CD34 and CD31 expression and cell viability better than trypsin-based detachment agents. It is used as per the manufacturer’s instructions. 15. If more than 2% of contaminant cells are present, our experience is that the cell sample will need to be sorted again. 16. Podoplanin is not a good sole marker for isolating lymphatic malformation LECs as not all LECs from macrocystic lymphatic malformation express podoplanin (as detected by clone D2-40) [3]. 17. Sorted foreskin LECs become increasingly senescent between passages 6 and 8 and have a reduced capacity to form tubes in vitro. Lymphatic malformation LECs proliferate at an even greater rate than at earlier passages and can grow on top of each other, a characteristic not observed at earlier passages. 18. We do not pool the samples together. This is to reduce the hypothetical HLA incompatibility that may influence cell survival as suggested for HUVECs [4].

Multiparameter Fluorescence-Activated Cell Sorting of Human Lymphatic. . .

59

19. Following cryopreservation, new LEC cultures are passaged at least once before use in experiments. Using LECS straight after cryopreservation negatively impacts LECs migration and capacity to form tubes on Matrigel™. After cryopreservation, a cell loss of 20–30% is usual after 24 h of culture. 20. The advantage of this method over the bead-selection method is that it is faster as there are fewer steps involved, thus decreasing the likelihood of bacterial cell contamination. Furthermore, cell yield and viability are better.

Acknowledgments The authors acknowledge the Baker Foundation and the Royal Children’s Foundation “Women in Science Fellowship” support of Zerina Lokmic-Tomkins and the NHMRC grant AP 1085109 supporting lymphatic malformations flow cytometry and FACS methodology work. References 1. Lokmic Z (2018) Utilizing lymphatic cell markers to visualize human lymphatic abnormalities. J Biophotonics 11(8):e201700117. https://doi. org/10.1002/jbio.201700117 2. Lokmic Z, Hallenstein L, Penington AJ (2017) Parental experience of prenatal diagnosis of lymphatic malformation. Lymphology 50(1):16–26 3. Lokmic Z, Mitchell GM, Koh Wee Chong N, Bastiaanse J, Gerrand YW, Zeng Y, Williams ED, Penington AJ (2014) Isolation of human lymphatic malformation endothelial cells, their in vitro characterization and in vivo survival in a mouse xenograft model. Angiogenesis 17(1): 1–15. https://doi.org/10.1007/s10456-0139371-8

4. Baudin B, Bruneel A, Bosselut N, Vaubourdolle M (2007) A protocol for isolation and culture of human umbilical vein endothelial cells. Nat Protoc 2(3):481–485. https://doi.org/10.1038/ nprot.2007.54 5. Lokmic Z, Ng ES, Burton M, Stanley EG, Penington AJ, Elefanty AG (2015) Isolation of human lymphatic endothelial cells by multiparameter fluorescence-activated cell sorting. J Vis Exp 99:e52691. https://doi.org/10.3791/ 52691 6. Lokmic Z (2016) Isolation, identification, and culture of human lymphatic endothelial cells. Methods Mol Biol 1430:77–90. https://doi. org/10.1007/978-1-4939-3628-1_5

Chapter 4 Absolute Quantification of Plasma Membrane Receptors Via Quantitative Flow Cytometry Yingye Fang, Manasi Malik, Sarah K. England, and P. I. Imoukhuede Abstract Plasma membrane receptors are transmembrane proteins that initiate cellular response following the binding of specific ligands (e.g., growth factors, hormones, and cytokines). The abundance of plasma membrane receptors can be a diagnostic or prognostic biomarker in many human diseases. One of the best techniques for measuring plasma membrane receptors is quantitative flow cytometry (qFlow). qFlow employs fluorophore-conjugated antibodies against the receptors of interest and corresponding fluorophore-loaded calibration beads offers standardized and reproducible measurements of plasma membrane receptors. More importantly, qFlow can achieve absolute quantification of plasma membrane receptors when phycoerythrin (PE) is the fluorophore of choice. Here we describe a detailed qFlow protocol to obtain absolute receptor quantities on the basis of PE calibration. This protocol is foundational for many previous and ongoing studies in quantifying tyrosine kinase receptors and G-protein-coupled receptors with in vitro cell models and ex vivo cell samples. Key words Quantitative flow cytometry, RTK, Protein quantification, Biomarker, Phycoerythrin (PE)

1

Introduction Most plasma membrane receptors are therapeutic targets, and the abundance of plasma membrane receptors is commonly evaluated as clinical biomarkers [1]. Antibody-based proteomic techniques, such as immunohistochemistry and flow cytometry, are often used to measure the abundance of receptors in cancer and other pathologies [2–4]. However, these techniques are only semiquantitative and cannot, therefore, deliver objective, reproducible, and absolute quantification of plasma membrane receptors. Quantitative proteomic techniques can overcome these shortcomings and are much desired for practicing pathology in this new era of data-driven precision medicine. Quantitative flow cytometry (qFlow) is one of the best proteomic techniques for objectively and reproducibly quantifying plasma membrane receptors [5, 6]. qFlow cytometry employing

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_4, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

61

62

Yingye Fang et al.

fluorophore-loaded calibration beads offers a mean of standardization across experiments, instruments, and labs. Among various commonly used fluorophores, phycoerythrin (PE) is superior for precise receptor quantification because PE is one of the brightest organic fluorophores available for flow cytometry (see Note 1) and resistant to pH changes [7]. More significantly, the number of PE molecules per cell can be directly converted to the number of PE-tagged receptors per cell when PE molecules are conjugated to antibodies at a 1:1 protein/fluorophore ratio [8] (see Note 2). Hence, absolute quantification of the receptor of interest can be achieved using PE calibration beads with known numbers of PE molecules per bead. Quantification of plasma membrane receptors, in particular tyrosine kinase receptors (RTK), advances research that correlates such biomarkers to disease progression and drug responses. RTKs are a family of high-affinity ligand-binding plasma membrane receptors that regulate a wide range of essential cellular activities, including cell proliferation, differentiation, and survival. Many RTKs are identified as promising biomarkers or targets in cancers and vascular diseases [9]. Among 20 subfamilies of RTKs, vascular endothelial growth factor receptors (VEGFRs) are widely recognized as promising biomarkers and targets for tumor angiogenesis, vascular diseases [10, 11], and obesity [12]. While much effort has been put into translating VEGFR expression into a reliable indicator for vascular diseases and cancers, a major limitation in these efforts has been the lack of quantitative data for standardizing clinical definitions of biomarkers and predicting individual patient response within large patient cohorts. To enable objective, reproducible, and absolute quantification of membrane-bound RTKs, like VEGFRs, we provide a detailed protocol describing a qFlow approach on the basis of PE calibration. This protocol considers experimental pitfalls and provides optimized solutions that we have summarized based on our work in RTK quantification with in vitro cell cultures, such as human umbilical vascular endothelial cells (HUVECs) [13], HUVEC/ fibroblast coculture [14], patient-derived glioblastoma tumor cell culture [5], and ex vivo cell samples from mice [15–17], human peripheral blood [18] and adipose tissues [19]. These quantification data are useful for establishing systems biology models to predict cell responses to RTK-targeted drugs [20]. This protocol can also be adapted to quantify a variety of transmembrane protein biomarkers in any cellular system (see Note 3).

Absolute Quantification of Plasma Membrane Receptors Via Quantitative Flow Cytometry

2

63

Materials Prepare all solutions using purified water and analytical grade reagents. Diligently follow all waste disposal regulations when disposing of waste materials.

2.1 Stain Buffer Preparation

2.2 Antibodies and Reagents

Stain buffer: Ca2+/Mg2+ free PBS supplemented with 0.5% (w/v) bovine serum albumin (BSA), 0.1% (w/v) sodium azide, pH ¼ 7.4 [21]. Add about 950 mL Ca2+/Mg2+ free PBS to a 1-L glass beaker. Weigh 1 g sodium azide and transfer to the beaker (see Note 4). Stir the PBS solution with a magnetic stirrer until sodium azide is fully dissolved. Adjust pH to 7.4 (7.35–7.45) with sodium hydroxide (NaOH) or hydrochloric acid solution (HCl). Weigh 5 g BSA and transfer to the beaker. Stir the buffer with a magnetic stirrer until BSA is fully dissolved. The final volume is approximately 1 L. Filter stain buffer through 0.22-μm membrane filters to remove bacteria and impurities. Store at 4  C. 1. Cell dissociation solution: Corning® CellStripper® Solution (see Note 5). 2. PE-conjugated monoclonal antibodies against the receptors of interest (e.g., PE-conjugated VEGFR1 antibody). 3. PE calibration bead standard: BD Quantibrite™ Beads (see Note 6). 4. Cell viability dye: SYTOX™ Blue Dead Cell Stain (Invitrogen). 5. Other fluorophore-conjugated antibodies against the cell markers of interest for cell identification purposes (see Note 5). e.g., for quantifying endothelial VEGFR2 expression in biopsy samples, a five-color panel consisting of the basic qFlow fluorescent components (i.e., PE-VEGFR and Sytox Blue) and additional antibodies against endothelial cell markers (i.e., APC-CD146, APC-Cy7-CD31, and FITC-CD45) is recommended because vascular endothelial cells are CD31+CD146+CD45 [22].

3

Methods

3.1 Single-Cell Suspension Preparation

To prepare single-cell suspension from cell culture, researchers should harvest cells after at least one passage has been performed. Check cell confluency and proceed to cell dissociation if the cells are about 80% confluent. Many cell lines can only be cultured for a limited number of passages before the cells change receptor membrane expression. For instance, the numbers of VEGFRs on HUVECs begin to change after six passages [13].

64

Yingye Fang et al.

1. Aspirate off cell culture media from the flask. Add 5 mL Ca2+/ Mg2+ free PBS to the flask and tilt the flask side-to-side to rinse the cells, and then aspirate off the PBS. 2. Add 5 mL Corning® CellStripper® Solution to the attached cells in the T75 flask. Place the flask back in the cell incubator. Check cell dissociation progress at 7 min. Some cell types may require a longer time to be fully dissociated. 3. Add 5 mL Ca2+/Mg2+ free PBS to the flask to neutralize the dissociation media and transfer the 10 mL cell suspension to a 15-mL tube. 4. Centrifuge the cell suspension at 400  g for 5 min at 4  C to bring down the cell pellets. Aspirate off the liquid part and add 10 mL Ca2+/Mg2+ free PBS to resuspend the cells. 5. Take out 10 μL of the cell suspension and count the cells. 6. Centrifuge the cell suspension again at 400  g for 5 min at 4  C and resuspend the cells at ~2–4 million cells per mL in stain buffer. 7. Keep the cell solution on ice until immunostaining. To prepare single-cell suspension from tissue biopsy, researchers should harvest tissue samples following established guidelines and approved protocols. A flow cytometry guide for preparing a single-cell suspension from solid tissues has been described [23]. 1. Mince tissues into 1.0 has been reported for certain individual lots (e.g., PE-CD8, BioLegend, Lot 607) [29]. It is best to consult the manufacturer for antibody- and lot-specific protein/fluorophore ratios. 3. Besides RTKs, G protein-coupled receptors (GPCRs) are another important family of transmembrane receptors that serve as drug targets due to their relevance in the treatment of various diseases, such as inflammatory disorders, metabolic disorders, and cancers [30]. For example, oxytocin receptor is a GPCR that activates myometrial contraction upon binding oxytocin, which is commonly used in labor augmentation. qFlow can be used to quantify genetic variant-associated changes in oxytocin receptor membrane expressions, which have significant implications for patients’ irresponsiveness to oxytocin treatment [31, 32, 33]. 4. For experiments using whole blood samples, an additional 2 mM EDTA should be added at this step to remove metal ions like calcium and magnesium in the blood and therefore prevent blood cells from clustering. Blood cells stored at 4  C show stable expression of cell surface markers and distribution for up to 72 h [34]. Blood samples stored for 30 s, or tilt and rotate for 5 min). 2. Into a new 1.5 mL Eppendorf tube, add 20 μL beads for each experimental condition (see Note 6). Add an excess of beads equivalent to 1–2 samples (see Note 7). 3. Add 1 mL PBS and resuspend the beads by pipetting. 4. Place the tube on a magnet for 1 min and discard the supernatant. 5. Wash with PBS, 3 times, repeating the above steps 3 and 4. 6. Resuspend the beads in PBS, adding 19.8 μL/sample (see Note 8).

Fig. 3 Precipitation of the SH2-domain binding protein using magnetic bead-immobilized peptides. Protein lysate extracted with hypotonic lysis buffer A and C were tested in the precipitation protocol to validate that the different salt concentrations did not affect the interaction between the SH2-domain binding protein and the bead-immobilized peptides. The positive control, PLCγ, was precipitated equally well by pY951-beads with both buffers, while detection was reduced in the Y951 non-phosphorylated peptide precipitate. With both buffers equally efficient, we decided that hypotonic lysis buffer A with lower salt concentration was regarded as more optimal in mass spectrometry analysis

106

Chiara Testini

3.4.2 Immobilization of Peptides to the Magnetic Beads

1. Add 1 nM of the different peptides to 20 μL beads. (a) hVEGFR2 Y951, 5 mg/mL stock. Molecular weight 2188.56 ! 0.1750848 μg ! 0.437712 μL. (b) hVEGFR2 pY951, 5 mg/mL stock. Molecular weight 2268.54 ! 0.1814832 μg ! 0.453708 μL. (c) hVEGFR2 Y1214, 5 mg/mL stock. Molecular weight 2426.72 ! 2.42672 μg ! 0.485344 μL. (d) hVEGFR2 Y1214, 5 mg/mL stock. Molecular weight 2426.72 ! 2.42672 μg ! 0.50134 μL. 2. Incubate 30 min at room temperature using gentle rotation. 3. Separate the peptide-coated beads with a magnet for 2–3 min. 4. Wash the peptide-coated beads 5 times in PBS containing 0.1% BSA.

3.4.3 Precipitation of Peptide-Binding Proteins

1. Add 2 mg protein lysate (see Subheading 3.6 protocol to assess the protein concentration of the lysate) to each sample from Subheading 3.3 to each vial of peptide-bead preparation prepared in Subheading 3.4 (see Notes 5 and 6). 2. Adjust volume to 1 mL with lysis buffer. 3. Incubate 1 h at 4  C on a rotating wheel. 4. Separate the peptide-coated beads with a magnet for 2–3 min. 5. Aspirate and discard the supernatant. 6. Wash three times with lysis buffer. 7. Proceed with Subheading 3.4.4 protocol to assess the protein precipitation efficacy or let the beads dry and proceed with Subheading 3.5 protocol.

3.4.4 Release of Immobilized Biotinylated Molecules

1. Add 25 μL of LDS sample loading buffer with reducing agent in lysis buffer to each tube. 2. Heat the samples to 95  C, 5 min. 3. Vortex very gently. 4. Separate the magnetic beads with a magnet for 2–3 min. 5. Load the supernatant on a 4–12% BisTris polyacrylamide SDS-PAGE gel and proceed with the Western blot protocol (see Subheading 3.7).

3.5 Mass Spectrometry 3.6 Determination of Protein Concentration

See Note 9 and Fig. 4.

1. Make the BSA-standard curve from the BSA stock solution (see Note 10). (a) 1500 μg BSA ¼ 93.75 μL BSA stock solution + 906.25 μL MilliQ water.

Monitoring Endothelial Calcium Flux

107

Fig. 4 Mass spectrometry analysis and western blot validation of the SH2domain proteins binding to VEGFR2 Y1214 (modified from Fig. 3 in the article [5]). (a) Mass spectrometry analysis reported precipitation of approximately 3000 proteins to the different peptide samples. Proteins that were binding equally to the immobilized peptides in all the conditions and protein that did not have a peptide-spectrum match (PSM) > 1 (see Note 11), were excluded, leaving 51 proteins abundant in the analysis of pY1214 peptide sample. Among them, ten proteins were selected for their high PSM number and high area score, both indicators of peptide abundance in the sample precipitate. This selection of pY1214 binding proteins was further reduced excluding all the proteins that have low expression in microvascular ECs, that lack SH2-domains or that are localized in subcellular compartments where they are unlikely to interact with VEGFR2. Growth factor receptor bound-2 (GRB2), p85 subunit of phosphoinositide 30 -kinase (PI3Kp85) and tensin 1 (TNS1) are the 3 proteins that are highly enriched in the pY1214 precipitate. (b) The binding of GRB2, PI3Kp85, and TNS1 to the pY1214 peptide was validated by immunoblotting of the respective HDMEC peptide precipitates. GRB2 and PI3Kp85 enriched presence was confirmed in the pY1214 sample but the presence of TNS1 was not detected. Therefore, TNS1 was excluded from further analysis [5]

108

Chiara Testini

(b) 1000 μg BSA ¼ 62.5 μL stock solution + 937.5 μL MilliQ water. (c) 800 μg BSA ¼ 50 μL stock solution + 950 μL MilliQ water. (d) 600 μg BSA ¼ 37.5 μL stock solution + 962.5 μL MilliQ water. (e) 400 μg BSA ¼ 25 μL stock solution + 975 μL MilliQ water. (f) 200 μg BSA ¼ 12.5 μL stock solution + 987.5 μL MilliQ water. 2. Pipette the following samples in the 96-well plate: (a) 10 μL/well for 2 wells of MilliQ water as blank. (b) 10 μL/well for 2 wells of each BSA-standard curve samples. (c) 9 μL of MilliQ water + 1 μL of each protein lysate samples obtained from Subheading 3.3. 3. Add the protein detection solution according to manufacture instruction (e.g., 200 μL/well for the Pierce BCA Protein Assay). 4. Incubate 30 min at 37  C (or according to manufacture instruction of the protein detection solution used). 5. Using a plate reader, measure the absorbance of each sample as optical density (OD) at 562 nm (or according to manufacture instruction of the protein detection solution used). 6. Determine the protein concentration using the linear equation obtained from the BSA-standard curve. (a) Linear equation y ¼ mx + n (see Note 12). (b) Concentration of the sample in the well ¼ (OD  n)/m. (c) Concentration of the sample lysate ¼ Concentration of the sample in the well * 10 (dilution factor 1:10; 9 μL of MilliQ water + 1 μL of protein lysate sample). 3.7

Western Blot

1. Load 20 μg protein per sample in LDS sample loading buffer from Subheading 3.3 or the supernatant from Subheading 3.4 on a 4–12% BisTris polyacrylamide gel. 2. Run the gel for 50 min, at a constant speed of 200 V. 3. Transfer proteins from the polyacrylamide gel to PVDF membrane for 2 h and 30 min while immersing the tank in ice at a constant speed of 30 V. 4. Rinse the membrane with TBS/0.1% Tween 20. 5. Block the membrane for 1 h with 3% BSA or 5% skimmed milk in TBS/0.1% Tween20 on a shaker at room temperature.

Monitoring Endothelial Calcium Flux

109

6. Incubate with primary antibodies in blocking buffer overnight on a shaker at 4  C. 7. Wash three times for 10 min with TBS/0.1% Tween 20 on a shaker at room temperature. 8. Incubate for 1 h with secondary antibody in blocking buffer on a shaker at room temperature. 9. Wash three times for 10 min with TBS/0.1% Tween 20 on a shaker at room temperature. 10. Incubate in ECL WB detection reagent for 5 min in the dark at room temperature. 11. Eliminate the excess of detection reagent. 12. Visualize the luminescence and record images.

4

Notes 1. The use of hypotonic lysis buffer is recommended when the protocol is followed by mass spectrometry analysis, since high concentrations of lysis detergents interfere with the mass spectrometry. 2. The number of tissue culture dishes used will depend on number of experimental conditions in each experiment. 3. The protocol was performed with cells at 70–80% confluency so that cells were still in an active proliferative status and consequently, proteins involved in the signaling pathways were optimally expressed. 4. Depending on the cell type of choice, proteins of interest, protein localization and mode of interaction, the optimal conditions might be different than the conditions described here. The lysis buffer influences both the quality and quantity of proteins in the lysate, and the binding of the proteins to the peptides. A buffer that is optimal in Subheading 3.3, might not be optimal for Subheading 3.4. Therefore, evaluate which buffer to use for both steps. 5. In the optimization of the protocol in Subheading 3.3, four different lysis buffers (RIPA, buffers A, B, and C) were used therefore, four different experimental conditions. In the final protocol applied in the study, only one buffer (Buffer A) was used, but four different peptides to precipitate the lysate, therefore four experimental conditions were used also. 6. In the optimization of the protocol in Subheading 3.4, two lysis buffers with higher efficiency were used: buffers A and C, and the precipitation with the two control peptides (Y951 and pY951) with both buffers was tested, for a total of four

110

Chiara Testini

experimental conditions. In the final protocol applied in the study, only one buffer (Buffer A) was used, but four different peptides were used to precipitate the lysate, therefore there were four experimental conditions, as well. 7. During the multiple washing of the magnetic beads present in their preparation, some beads will be lost during processing. Add an excess of beads equivalent to 1–2 more samples, to ensure enough beads remain for all your samples. 8. Resuspension of the beads in 19.8 μL/sample of PBS has been calculated considering that the beads are 1% of the initial volume. Therefore, beads need to be resuspended in the remaining 99% of the initial volume of 20 μL, that is 19.8 μL. 9. Mass spectrometry has been performed by the SciLifeLab Clinical Proteomics Mass Spectrometry facility, Solna, Sweden. A detailed protocol of the procedure is described in the original articles of this method [5]. 10. BSA-standard curve solutions can be prepared once and stored at 20  C. 11. The peptide-spectrum match (PSM) indicates the total number of identified peptides matched for a given detected protein. A PSM of 1 means that 1 peptide was detected once for the protein of interest and there is high probability that the detected peptide is a false positive. Therefore, only proteins with a PSM > 1 were included. 12. Use only linear equation where the R2 value is equal or higher than 0.99.

Acknowledgments The author acknowledges the equipment and expert advice supplied by Prof. Lena Claesson-Welsh, Uppsala University, Sweden and the SciLifeLab Clinical Proteomics Mass Spectrometry facility, Solna, Sweden. This method was part of a larger study [5] that was made possible through grants to Prof. Lena Claesson-Welsh from the Swedish Research Council (2015-02375), the Swedish Cancer Foundation (CAN2016/578), and the Knut and Alice Wallenberg Foundation (KAW 2015.0030). KAW also supported LCW with a Wallenberg Scholar grant (2015.0275). References 1. Murakami M (2012) Signaling required for blood vessel maintenance: molecular basis and pathological manifestations. Int J Vasc Med 2012:293641

2. Senger DR et al (1983) Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 219(4587): 983–985

Monitoring Endothelial Calcium Flux 3. Koch S et al (2011) Signal transduction by vascular endothelial growth factor receptors. Biochem J 437(2):169–183 4. Simons M, Gordon E, Claesson-Welsh L (2016) Mechanisms and regulation of endothelial VEGF receptor signalling. Nat Rev Mol Cell Biol 17(10):611–625 5. Testini C et al (2019) Myc-dependent endothelial proliferation is controlled by phosphotyrosine 1212 in VEGF receptor-2. EMBO Rep 20(11):e47845 6. Cunningham SA et al (1997) Interactions of FLT-1 and KDR with phospholipase C gamma: identification of the phosphotyrosine binding

111

sites. Biochem Biophys Res Commun 240(3): 635–639 7. Takahashi T et al (2001) A single autophosphorylation site on KDR/Flk-1 is essential for VEGF-A-dependent activation of PLC-gamma and DNA synthesis in vascular endothelial cells. EMBO J 20(11):2768–2778 8. Matsumoto T et al (2005) VEGF receptor2 Y951 signaling and a role for the adapter molecule TSAd in tumor angiogenesis. EMBO J 24(13):2342–2353 9. Li X et al (2016) VEGFR2 pY949 signalling regulates adherens junction integrity and metastatic spread. Nat Commun 7:11017

Chapter 7 Monitoring VEGF-Stimulated Calcium Ion Flux in Endothelial Cells William R. Critchley, Gareth W. F. Fearnley, Izma Abdul-Zani, Carmen Molina-Paris, Claus Bendtsen, Ian C. Zachary, Michael A. Harrison, and Sreenivasan Ponnambalam Abstract The endothelial response to vascular endothelial growth factor A (VEGF-A) regulates many aspects of animal physiology in health and disease. Such VEGF-A-regulated phenomena include vasculogenesis, angiogenesis, tumor growth and progression. VEGF-A binding to receptor tyrosine kinases such as vascular endothelial growth factor receptor 2 (VEGFR2) activates multiple signal transduction pathways and changes in homeostasis, metabolism, gene expression, cell proliferation, migration, and survival. One such VEGF-A-regulated response is a rapid rise in cytosolic calcium ion levels which modulates different biochemical events and impacts on endothelial-specific responses. Here, we present a series of detailed and robust protocols for evaluating ligand-stimulated cytosolic calcium ion flux in endothelial cells. By monitoring an endogenous endothelial transcription factor (NFATc2) which displays calcium-sensitive redistribution, we can assess the relevance of cytosolic calcium to protein function. This protocol can be easily applied to both adherent and non-adherent cultured cells to evaluate calcium ion flux in response to exogenous stimuli such as VEGF-A. Key words Endothelial cells, Calcium, VEGF-A, VEGFR2, NFATc2, Human umbilical vein endothelial cells (HUVECs)

1

Introduction The calcium ion is a noted second messenger in many biological systems, especially in eukaryotes. Many cellular proteins with calcium-binding properties have specific domains or motifs that can bind calcium ions reversibly. Such binding can trigger conformational changes resulting in changes in protein activity, distribution and/or interactions with other factors. In this way a wide array of biochemical reactions within many biological systems are modulated by the presence or absence of calcium ions. In many eukaryote cells, calcium ions are stored in intracellular, membrane-bound

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_7, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

113

114

William R. Critchley et al.

compartments such as the endoplasmic reticulum (ER). The activation of signal transduction pathways frequently impact on cytosolic calcium ions: this is caused by the activation of inositol-1,4,5triphosphate receptor (IP3R) which is a membrane protein channel located within the ER [1]. The production of IP3 through signaling events at the plasma membrane results in rapid binding of IP3 to IP3R to trigger opening of the membrane channel and rapid movement of calcium ions from the ER to the cytosol [2]. In this way, signal transduction pathways frequently trigger cytosolic calcium ions which then impact on different aspects of cellular physiology and pathophysiology [3]. The phenomenon of angiogenesis is the sprouting of new blood vessels from pre-existing ones [4]. The endothelium is a cell monolayer which lines all blood vessels and is a critical interface between circulating blood and the blood vessel wall. The presence of soluble pro-angiogenic factors in circulating fluids e.g. blood, causes endothelial cells to carry out this unique process of angiogenesis. One such family of pro-angiogenic molecules are the VEGF family comprised of A, B, C, D and placental growth factor, PlGF [5, 6]. VEGF-A is the most intensively studied and is essential for both vasculogenesis and angiogenesis [7]. Deregulation of angiogenesis linked to increased levels of VEGF-A is involved in different pathological states such as diabetic retinopathy and tumor development [4, 7]. VEGF-A binds to the receptor tyrosine kinase VEGFR2 present on endothelial cells to trigger different signaling events [6]. Notably, the canonical mitogen-activated protein kinase (MAPK), p38 MAPK, AKT, and JNK pathways are stimulated by the activation of VEGFR2 [6]. One well-established consequence of VEGF-A binding to human VEGFR2 is autophosphorylation of residue Y1175: this phospho-epitope within the VEGFR2 cytoplasmic tail forms a binding site for a phospholipase, PLCγ1 [8], which is essential for regulating vascular development [9]. PLCγ1 contains Src-homology 2 (SH2) domains which specifically recognize the VEGFR2-pY1175 epitope enabling interaction with activated VEGFR2 [10]. The recruitment of phospholipases such as PLCγ1 triggers hydrolysis of phosphatidylinositol-4,5-bisphosphate (PIP2) located on the cytosolic leaflet of the plasma membrane bilayer. PIP2 breakdown to diacylglycerol (DAG) and inositol1,4,5-trisphosphate (IP3) generates two different types of second messengers which can now modulate a wide range of cellular events. IP3 diffuses rapidly throughout the cell and a major target is the ER-bound IP3R ion channel, which results in rapid calcium ion efflux into the cytosol. DAG binds to protein kinase C (PKC) enzymes, triggering PKC activation and phosphorylation of a wide range of cellular targets, including protein kinases within different signal transduction pathways.

Monitoring Endothelial Calcium Flux

115

The binding of VEGF-A to endothelial VEGFR2 promotes a rapid rise in cytosolic calcium ions [11]. In many immune and hematopoietic-derived tissues including the endothelium, a rise in cytosolic calcium ions leads to activation of a family of transcription factors termed nuclear factor of activated T cells (NFAT) [12]. The rise in cytosolic calcium ions activates a calcium-binding protein called calcineurin, which is a calcium-regulated protein phosphatase. Inactive NFAT is usually phosphorylated on multiple serine and threonine residues and trapped in the cytosol. Upon activation of calcineurin by calcium ions, dephosphorylation of NFAT can occur. Dephosphorylated NFAT is now competent to translocate through nuclear pores into the nucleus. Here, NFAT can now promote gene transcription at multiple gene loci. In endothelial cells, such a pathway has been well established: VEGF-A binding leads to dephosphorylation of NFATc2 (NFAT1) and translocation into the nucleus [13–15]. The functional target of endothelial NFATc2-regulated gene transcription is unclear, but affects cell migration and not proliferation [15]. Such findings suggest that more detailed studies are needed to provide a mechanism for how VEGF-A regulates calcium-regulated cellular responses in endothelial cells. In this chapter, we provide a rapid and effective protocol for monitoring cytosolic calcium ion flux in response to exogenous ligand such as VEGF-A.

2

Materials Buffers and reagents of analytical grade should be used. Working buffers used sterilized autoclaved double-distilled water (purified deionized water with a specific resistance of 18 MΩ/cm2 at 25
C).

2.1 Endothelial Cell Culture

1. Complete endothelial cell growth medium (ECGM; Promocell, Heidelberg, Germany; Cat. No. C-22010): 500 mL endothelial cell basal medium supplemented with 2% (v/v) fetal calf serum, 0.4% (v/v) endothelial growth supplement, 0.1 ng/mL epidermal growth factor (EGF), 1 ng/mL basic fibroblast growth factor (bFGF), 90 μg/mL heparin and 1 μg/mL hydrocortisone; all pre-warmed to 37
C before mixing under sterile conditions. 2. TrypLE™ Express (1), no phenol red. 3. Human umbilical vein endothelial cells (HUVEC) (see Note 1). 4. Improved Neubauer hemocytometer or similar. 5. 0.1% (w/v) porcine skin gelatin in phosphate-buffered saline (PBS) (see Note 2). Phosphate-buffered saline: 140 mM NaCl, 3 mM KCl, 10 mM Na2HPO4, pH 7.3. Phosphate-buffered

116

William R. Critchley et al.

saline tablets can be used by dissolving one tablet in 500 mL double-distilled water and autoclaved on standard wet cycle at 121
C. 6. Tissue-culture grade sterile T75 plastic flasks. 7. Gelatin-coated 13 mm glass coverslips. 8. Superfrost glass slides. 9. Mounting medium suitable for stabilizing fluorescent labeled samples e.g., Fluoromount G mounting medium or similar. 2.2 Calcium Flux Assay

1. Tissue-culture grade 96-well plates. 2. SBS buffer (prepared fresh): 130 mM NaCl, 5 mM KCl, 1.2 mM MgCl2, 8 mM glucose, 10 mM HEPES, 1.5 mM CaCl2, pH 7.4. 3. 20% (w/v) pluronic F-127 in DMSO. 4. Fura-2 AM made to 1 mM in DMSO. 5. VEGF-A165a ligand (Promocell, Heidelberg, Germany; Cat. No. C-64423) supplied as a lyophilized powder. Resuspend at a 1000 stock concentration of 25 μg/mL in sterile PBS pH 7.4 and store at 80
C in 50 μL aliquots.

2.3 NFATc2 Localization

1. Tissue-culture grade sterile 24-well plates. 2. Serum starvation medium (MCDB131/BSA): MCDB131 medium, no glutamine, supplemented with 0.2% (w/v) BSA and 0.22 μm filter sterilized in a sterile environment; pre-warmed to 37
C. 3. Fixative: 3% (w/v) paraformaldehyde (PFA) in PBS containing 0.1 mM CaCl2 and 0.1 mM MgCl2 (see Note 3). 4. Permeabilization buffer: 0.2% (v/v) Triton X-100 in PBS pH 7.4. 5. Blocking buffer: 0.5% (w/v) BSA in PBS pH 7.4. 6. Primary antibody: rabbit monoclonal antibody D43B1 antiNFATc2/NFAT1 (Cell Signaling Technology, Danvers, USA; Cat. No. 5861). 7. Secondary antibody: AlexaFluor594-conjugated donkey antirabbit secondary antibody. 8. 100 stock solution (100 mg/mL) of 40 ,6-Diamidine-20 -phenylindole dihydrochloride (DAPI) in PBS pH 7.4.

2.4 Equipment and Software

1. Programmable fluorescence microplate reader with automated on-board pipettor capable of real-time fluorescence monitoring. The instrument must be capable of simultaneous dual excitation at both 340 and 380 nm while recording emission at 510 nm.

Monitoring Endothelial Calcium Flux

117

2. Inverted digital fluorescence microscope. 3. NIH Image J software with Coloc 2 plugin downloaded from https://imagej.nih.gov/ij/download.html.

3

Methods

3.1 Primary Endothelial Cell Culture

1. Coat T75 tissue-culture flasks with 10 mL of 0.2% porcine skin gelatin solution and disperse so that the entire base of the flask is covered. 2. Incubate at 37
C for at least 1 h. 3. Aspirate gelatin solution and wash flask three times with sterile PBS and allow the flasks to air dry in tissue-culture laminar flow hood. 4. Seed HUVECs into gelatin-coated T75 tissue-culture flask at 5000–10,000 cells per cm2 in ECGM (pre-warmed to 37
C). 5. Incubate at 37
C until approximately 80% confluent. Perform media changes as required (see Note 4). 6. Once HUVEC flask reaches 80% cell confluency, detach cells by aspirating ECGM and replacing with 3 mL TrypLE™ Express per T75 flask. 7. Tip the flask to each side to ensure the trypsin solution is able to cover all of the flask surface. Once the cells are coated with TrypLE™ solution, immediately aspirate and repeat a second time with a further 3 mL TrypLE™. 8. Aspirate solution again and incubate cells at 37
C for up to 5 min, checking for cell detachment every minute. 9. Stop the trypsin activity by adding 1 mL ECGM. Resuspend HUVECs in ECGM and seed new T75 flasks (if passaging) or appropriate tissue-culture plates (for experimental use) at 5000–10,000 cells per cm2. Allow to grow for at least 24 h before performing experiments at approximately 80% confluency (see Note 5).

3.2 Preparation of HUVECs and Loading Cells with Fluorescent Dye

1. Coat 96-well tissue-culture plates with 100 μL gelatin solution to allow coating and drying as described in Subheading 3.1. 2. Detach 80% confluent HUVECs with TrypLE™ Express as described in Subheading 3.1. 3. Count the number of cells per mL using a Neubauer hemocytometer. 4. Seed HUVECs onto gelatin-coated 96-well plates such that they are fully confluent on the planned day of experiment. For example, seeding 25,000 HUVECs per well enables

118

William R. Critchley et al.

experiments to be carried out in 48 h or 19,000 HUVECs per well enables experiments to be carried out in 72 h. 5. Incubate cells at 37
C in tissue culture incubator until wells are 100% confluent. 6. Aspirate media and wash cells twice with 100 μL SBS. 7. Prepare dye loading solution in SBS from stock solutions. Dilute stock solutions of Pluronic acid F-127 1:2000, and Fura-2 AM 1:500 to obtain a final working solution containing 0.01% Pluronic F-127 and 2 μM Fura-2 AM. 8. Load HUVECs with 50 μL per well of Fura-2 AM/Pluronic F-127/SBS (2 μM Fura-2 AM, 0.01% Pluronic F-127) by incubation for 60 min at 37
C. 3.3 Cytosolic Calcium Flux Assay

1. Wash HUVECs with 100 μL SBS. For addition of a single compound or multiple compounds at a single time point, wash a second time with 200 μL of SBS. Incubate cells for 30 min at room temperature to allow complete de-esterification of Fura-2 AM. 2. Prepare compound and/or ligand(s) to be added (e.g., VEGFA) diluted to a 5 stock concentration in a 96-well U-bottom plate. Addition of solutions to the HUVEC cells will result in a 1:5 dilution. Prepare a volume sufficient for 50 μL to be added easily by the machine to each HUVEC well (see Note 6). 3. Switch on an appropriate microplate reader (e.g., FlexStation 2). Open the relevant software (e.g., FlexStation Softmax Pro) and create a new document. 4. Set wavelength to allow measurement of the ratio of emission at 510 nm achieved from excitation of Fura-2 AM dye at 340 nm (calcium-bound state) and 380 nm (calcium-free state). 5. Set the run time for the experiment, the number of readings to take and the interval between each reading. Timings may vary dependent upon stimulus being utilized, source of endothelial cell and expected duration of effect (see Note 7). 6. Specify or select the assay plate format and manufacturer used e.g., 96-well Sarstedt. Clear the program and select which wells in the plate should be read by the instrument. 7. Specify or select the compound source plate format e.g., 96-well U bottom. 8. Input the number, volume and timing of transfers from the compound source (see Note 8). 9. Select which tips to use. Select which column of tips in the rack e.g., column 1, 2, 3 should go into which column of the

Monitoring Endothelial Calcium Flux

119

Fig. 1 VEGF-A stimulation of endothelial cells prompts a significant elevation in cytosolic calcium ions. Data is shown for control (PBS) vs. VEGF-A delivered to HUVECs at three different concentrations. (a) Representative output from real-time monitoring of cytosolic calcium ion flux. (b) Quantification of relative peak magnitude provides an indicator of how much calcium ion levels rose after stimulation. (c) Time to peak magnitude graph represents how rapidly the intracellular calcium ion levels rise after stimulation. (d) The graph of relative curve area provides a single quantitative assessment of the level of calcium ion rise and duration for which the elevation is sustained

compound plate e.g., A-H; 1, 2, 3 and then into which column of the cell plate containing HUVECs e.g., A-H; 1, 2, 3. 10. Load the tips, compound plate and cell plate into the microplate reader. Tips and plates should be loaded with A1 situated at the top left. 11. Initiate the read and once started, click display > reduce > plot to see the graphs being produced in real time. 12. Raw data for change in cytosolic calcium levels over time can be exported and analyzed in data analysis software e.g., OriginPro (OriginLab, US). The data can be expressed as a raw trace displaying the relative peak height and duration of elevation or can be quantified by calculating the peak magnitude, time taken to reach peak magnitude and area under the curve (Fig. 1). 3.4 Immunofluorescence Analysis of Endothelial NFATc2 Localization

1. Grow HUVECs in sterile tissue-culture grade 96-well plates coated with 100 μL per well porcine skin gelatin as described in Subheading 3.1. 2. Aspirate medium and starve the cells for 3 h at 37
C in 90 μL of serum starvation medium. 3. Prepare 10 concentrated VEGF-A stock solutions (0, 2.5, 7.5, 12.5 nM). Make sufficient to allow the addition of 10 μL per well.

120

William R. Critchley et al.

4. Add 10 μL of each stock to each well for a final working concentration of 0.25, 0.75, or 1.25 nM. Incubate for 15 min at 37
C as stimulation period. 5. Aspirate media and fix immediately with 100 μL of pre-warmed fixative followed by incubation for 20 min at room temperature. 6. Wash each well three times with 100 μL of PBS. 7. Permeabilize the cells with fresh permeabilization buffer for 4 min at room temperature. 8. Aspirate and wash each well three times with 100 μL of PBS. 9. Block nonspecific binding sites by incubation for 1 h at room temperature with 100 μL blocking buffer. 10. Aspirate blocking buffer and add 20 μL/well of 1 μg/mL primary antibody (rabbit anti-NFATc2) in PBS/BSA (see Note 9). Incubate overnight (16–20 h) at 4
C in a humidified chamber to prevent drying out. 11. The next day, aspirate the primary antibody and wash the wells three times with 100 μL/well of blocking buffer. 12. Prepare secondary antibody by diluting AlexaFluor594conjugated secondary anti-rabbit antibody to 4 μg/mL in 1% BSA/PBS; also dilute DAPI to 2 μg/mL from stock solution. 13. Add 20 μL/well of solution containing secondary antibody and DAPI and incubate protected from light at room temperature for 2 h. 14. Wash the cells twice each with 100 μL of blocking buffer followed by one final wash with 100 μL double-distilled water. 15. Acquire nuclear DNA (blue) and NFATc2 (red) images for each field of cells using an inverted digital fluorescence microscope. Capture at least 3 fields per well at 20 magnification (Fig. 2). 16. Use the Coloc 2 plugin for NIH ImageJ software to assess nuclear co-localization by determining overlap of DAPI and NFATc2 staining patterns in the blue and red channels respectively (see Note 10). 17. Open ImageJ and load image. Split channels and retain as separate images the blue and red channels. 18. Select Analyze > Co-localization > Coloc 2. Ensure Costes method is selected for threshold regression to automatically set thresholds without user bias. 19. Choose appropriate algorithms for analysis (see Note 11). 20. Set Costes randomizations to 100. 21. Run the co-localization analysis and note the values for Manders’ tM1, Manders’ tM2 and Costes p-value (see Note 12).

Monitoring Endothelial Calcium Flux

121

Fig. 2 NFATc2 redistribution to the cell nucleus following VEGF-A stimulation of endothelial cells. Control or non-stimulated cells display a diffuse cytosolic pattern of NFATc2 staining (red) compared to the much more intense nuclear staining of NFATc2 (red) after VEGF-A stimulation. Nuclear DNA stained using DAPI (blue) (a). Quantification of nuclear co-distribution of NFATc2 determined using NIH ImageJ which can be plotted to demonstrate the redistribution of NFATc2 upon VEGF-A stimulation (b)

4

Notes 1. HUVECs can be obtained from single donor or pooled mixed donor sources from commercial providers. Cells can also be isolated in-house from single donors as described previously [16, 17]. 2. 0.1% (w/v) porcine skin gelatin can be prepared in-house as required. Add the necessary amount of pig skin gelatin powder to sterilized PBS and microwave for 1 min until fully dissolved. Push sterilize through a 0.22 μm syringe filter unit in a clean tissue-culture laminar flow hood. 3. 3% PFA solution can be prepared in advance by dissolving in PBS. Once fully mixed, CaCl2 and MgCl2 are both added to 0.1 mM before push sterilizing through a 0.22 μm syringe unit. 3% PFA can be stored long term as aliquots at 20
C until use. This be thawed and warmed up to room temperature before use. 4. HUVEC growth rate is variable between batches and dependent upon initial seeding density. Media changes must be

122

William R. Critchley et al.

performed every 48 h to maintain optimal cell health and growth until they are ready to be used or passaged (at 80% confluency). Inspect HUVEC growth under the microscope. If confluency is less than 80% replace existing media with fresh ECGM and continue to incubate at 37
C. 5. HUVECs are inherently limited in the number of passages that can be performed. If treated optimally, HUVECs may be maintained until passage 5 without discernible changes in morphology, growth rate or cellular responses. 6. The working concentration of a VEGF-A stimulus is 25 ng/ mL, and this therefore requires a 5 stock solution (in SBS) of 125 ng/mL prepared in the U-bottom multiwell plate. 7. For VEGF-A-induced calcium response in HUVECs, set the total run time to 600 s, and set interval of measurements to every 5 s. This should equate to a total of 121 readings that are collected during the course of a single run. 8. For a single stimulation with VEGF-A, the programmable on-board pipettor should be set to: 1 transfer of 50 μL from compound source at 32 s. 9. The primary antibody must be prepared fresh in a solution of 0.1% (w/v) BSA in PBS immediately before the end of the blocking step. 10. With DAPI as a stain present only in the nucleus, any significant co-localization between the channels is indicative of the nuclear presence of the protein from the red channel. 11. A large number of options are available and may have value depending upon the nature of the experiment. For the purposes of determining the nuclear localization of NFATc2 in HUVECs, select Mander’s Correlation and Costes significance test. Further in-depth information on co-localization methods has been published previously [18]. 12. If you select the red channel as Channel 1 and blue channel as Channel 2 in the analysis settings, then the tM1 value indicates the proportion of the red pixels that co-localize with blue pixels. The tM2 value indicates the reverse (proportion of blue pixels that co-localize with red pixels). As there can be red signal (NFATc2 not translocated to the nucleus) outside of the nucleus but blue DAPI signal is restricted to the nucleus, these values will be different. The Costes p-value indicates the confidence that the localization is real and should be greater than 0.95.

Monitoring Endothelial Calcium Flux

123

Acknowledgments This work was supported by project grants from the British Heart Foundation (S.P., M.A.H., I.Z.), AstraZeneca (S.P., C.M.P., C.B.) and a PhD studentship from the Brunei Government (I.A.Z.). References 1. Sharma A, Elble RC (2020) From orai to E-cadherin: subversion of calcium trafficking in cancer to drive proliferation, anoikisresistance, and metastasis. Biomedicine 8(6): 1 6 9 . h t t p s : // d o i . o r g / 1 0 . 3 3 9 0 / biomedicines8060169 2. Parys JB, Vervliet T (2020) New insights in the IP3 receptor and its regulation. Adv Exp Med Biol 1131:243–270. https://doi.org/10. 1007/978-3-030-12457-1_10 3. Dejos C, Gkika D, Cantelmo AR (2020) The two-way relationship between calcium and metabolism in cancer. Front Cell Dev Biol 8: 573747. https://doi.org/10.3389/fcell. 2020.573747 4. Zecchin A, Kalucka J, Dubois C, Carmeliet P (2017) How endothelial cells adapt their metabolism to form vessels in tumors. Front Immunol 8:1750. https://doi.org/10.3389/ fimmu.2017.01750 5. Bates DO, Beazley-Long N, Benest AV, Ye X, Ved N, Hulse RP, Barratt S, Machado MJ, Donaldson LF, Harper SJ, Peiris-Pages M, Tortonese DJ, Oltean S, Foster RR (2018) Physiological role of vascular endothelial growth factors as homeostatic regulators. Compr Physiol 8(3):955–979. https://doi. org/10.1002/cphy.c170015 6. Shaik F, Cuthbert GA, Homer-Vanniasinkam S, Muench SP, Ponnambalam S, Harrison MA (2020) Structural basis for vascular endothelial growth factor receptor activation and implications for disease therapy. Biomol Ther 10(12): 1 6 7 3 . h t t p s : // d o i . o r g / 1 0 . 3 3 9 0 / biom10121673 7. Apte RS, Chen DS, Ferrara N (2019) VEGF in signaling and disease: beyond discovery and development. Cell 176(6):1248–1264. https://doi.org/10.1016/j.cell.2019.01.021 8. Takahashi T, Yamaguchi S, Chida K, Shibuya M (2001) A single autophosphorylation site on KDR/Flk-1 is essential for VEGF-A-dependent activation of PLC-gamma and DNA synthesis in vascular endothelial cells. EMBO J 20(11):2768–2778 9. Sakurai Y, Ohgimoto K, Kataoka Y, Yoshida N, Shibuya M (2005) Essential role of Flk-1 (VEGF receptor 2) tyrosine residue 1173 in

vasculogenesis in mice. Proc Natl Acad Sci U S A 102(4):1076–1081. https://doi.org/10. 1073/pnas.0404984102 10. Guo D, Jia Q, Song HY, Warren RS, Donner DB (1995) Vascular endothelial cell growth factor promotes tyrosine phosphorylation of mediators of signal transduction that contain SH2 domains. Association with endothelial cell proliferation. J Biol Chem 270(12): 6729–6733 11. Cunningham SA, Tran TM, Arrate MP, Bjercke R, Brock TA (1999) KDR activation is crucial for VEGF165-mediated Ca2+ mobilization in human umbilical vein endothelial cells. Am J Phys 276(1 Pt 1):C176–C181 12. Park YJ, Yoo SA, Kim M, Kim WU (2020) The role of calcium-calcineurin-NFAT signaling pathway in health and autoimmune diseases. Front Immunol 11:195. https://doi.org/10. 3389/fimmu.2020.00195 13. Holmes K, Chapman E, See V, Cross MJ (2010) VEGF stimulates RCAN1.4 expression in endothelial cells via a pathway requiring Ca2+/calcineurin and protein kinase C-delta. PLoS One 5(7):e11435. https://doi.org/10. 1371/journal.pone.0011435 14. Armesilla AL, Lorenzo E, Gomez del Arco P, Martinez-Martinez S, Alfranca A, Redondo JM (1999) Vascular endothelial growth factor activates nuclear factor of activated T cells in human endothelial cells: a role for tissue factor gene expression. Mol Cell Biol 19(3): 2032–2043 15. Fearnley GW, Bruns AF, Wheatcroft SB, Ponnambalam S (2015) VEGF-A isoform-specific regulation of calcium ion flux, transcriptional activation and endothelial cell migration. Biol Open 4(6):731–742. https://doi.org/10. 1242/bio.201410884 16. Jaffe EA, Nachman RL, Becker CG, Minick CR (1973) Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest 52(11):2745–2756. https://doi.org/ 10.1172/JCI107470 17. Howell GJ, Herbert SP, Smith JM, Mittar S, Ewan LC, Mohammed M, Hunter AR, Simpson N, Turner AJ, Zachary I, Walker JH,

124

William R. Critchley et al. Ponnambalam S (2004) Endothelial cell confluence regulates Weibel-Palade body formation. Mol Membr Biol 21(6):413–421. h t t p s : // d o i . o r g / 1 0 . 1 0 8 0 / 09687860400011571

18. Costes SV, Daelemans D, Cho EH, Dobbin Z, Pavlakis G, Lockett S (2004) Automatic and quantitative measurement of protein-protein colocalization in live cells. Biophys J 86(6): 3993–4003. https://doi.org/10.1529/ biophysj.103.038422

Chapter 8 Co-immunoprecipitation Assays Ian M. Evans and Ketevan Paliashvili Abstract Co-immunoprecipitation is a well-established technique for determining whether two proteins interact. It is based on the principle that by pulling down one protein, you will also obtain any other proteins that exist in a complex with that protein. It is a relatively simple technique that does not require expensive reagents or materials. It is however, not without its limitations and some of these will be discussed here along with a step-by-step guide to performing and analyzing co-immunoprecipitation experiments. Key words VEGF, Signal transduction, Protein–protein interaction, p130Cas

1

Introduction Proteins function by interacting with other proteins, whether this is enzymatically (e.g., kinases) or a number of other ways (e.g., ligand–receptor binding). Determining which proteins interact with each other and under which conditions can therefore provide fundamental insights into the function of the proteins being investigated. As a result, there are a wide range of techniques for determining protein-protein interactions which vary considerably in complexity and the requirement for specialized equipment [1]. Of these, co-immunoprecipitation (co-IP) is a widely used and wellestablished technique, based on the principle of using an antibody against a protein of interest that is attached to a relatively large molecule such as an agarose bead (other solid phase supports such as Sepharose or magnetic beads can also be used). The antibody binds to the target of interest and the protein–antibody–agarose complex can then be “precipitated” or pulled down from the solution by centrifugation. In theory any other protein that exists in a complex with the protein of interest will also be pulled down. These can then be identified using techniques such as western blotting or mass spectrometry.

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_8, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

125

126

Ian M. Evans and Ketevan Paliashvili

Fig. 1 HUVECS were stimulated with VEGF (25 ng/mL) for the times indicated. Cells were lysed and immunoprecipitated with anti-p130Cas antibody and immunoblotted with antibodies to p130Cas (Control) and IQGAP1. Co-immunopreciptated IQGAP1 was quantified by densitometry and normalized to total immunopreciptated p130Cas; *p < 0.05 versus 0 min time point, n ¼ 3. (Taken from [2])

The main advantage of this technique is its relative simplicity and requires little in the way of specialized reagents or equipment (unless mass spectrometry is involved). As well as determining whether two proteins interact, it can also be used to investigate how this interaction is affected by activating or inhibiting signaling pathways. An example is shown in Fig. 1. In this specific study, we investigated how the interaction between two proteins, IQGAP1 and p130Cas was affected by vascular endothelial growth factor (VEGF) stimulation of human umbilical vein endothelial cells [2]. To do this we stimulated the cells with VEGF for between 0 and 60 min then p130Cas was immunopreciptated followed by western blot analysis for both p130Cas and IQGAP1. The experiment showed that p130Cas interaction with IQGAP1 was VEGF dependent with the most interaction occurring after 30 min of VEGF stimulation. The same basic approach can also be used to determine whether a protein can interact with multiple other proteins. In this case the co-IP is followed by analysis by mass spectrometry. This can reveal otherwise unknown novel interactions between proteins. However, the results should always be interpreted with caution, the inherently sticky nature of antibodies means that the risk of false positives is quite high, and should always be validated by an alternative approach. These risks can be mitigated somewhat by the use of the appropriate IgG control. This should always be of the same species as the immunoprecipitating antibody (i.e., mouse, rabbit, etc.). Although co-IP can indicate whether two proteins interact, it does not provide any information on the nature of the interaction, i.e., whether there is a direct or indirect interaction (with one or more proteins between them). This kind of information can only be obtained using more specialized techniques such as Fo¨rster resonance energy transfer (FRET [3]). Another disadvantage of co-IP is

Co-immunoprecipitation

127

that it can often miss some interactions. This may be due to a number of reasons, either because the interaction is transient, the antibodies used may interfere with the interaction or it may simply be too insensitive to pick up subtle interactions. This last problem can sometimes be overcome by overexpressing the protein of interest, however this can also increase the incidence of false positives. In summary, co-IP is a useful and straightforward technique to identify and analyze protein–protein interactions. However, the user should always be aware of its various limitations.

2 2.1

Materials Cell Culture

1. Cell type: Human umbilical vein endothelial cells (HUVECs), or any other suitable cell type (see Note 1). 2. Culture medium: Endothelial basal medium (EBM, Lonza) supplemented with 10,000 units penicillin and 10 mg streptomycin/mL (P0781, Sigma Life Sciences). 3. Fetal bovine serum (FBS). 4. Human recombinant VEGF-A165 (100-20, Peprotech). 5. Trypsin-EDTA solution 1 (0.25% trypsin, 0.02% EDTA, sterile filtered). 6. Phosphate buffered saline (PBS; 140 mM NaCl, 3 mM KCl, 10 mM phosphate buffer, pH 7.4). 7. 10 cm cell culture dishes.

2.2 Preparation of Cell Extracts

1. NP40 buffer: 150 mM NaCl, 50 mM Tris-HCl pH 8, 0.5% NP40, diluted in PBS. 2. Complete® protease inhibitors (04693116001, Roche). 3. Phosphatase inhibitors I and II (P2850 and P5726, Sigma). 4. Benchtop centrifuge. 5. Sterile cell scrapers. 6. 1.5 mL microcentrifuge tubes.

2.3 Pre-clearing the Lysates

1. Rabbit/mouse IgG agarose. 2. 1.5 mL microcentrifuge tubes. 3. Benchtop centrifuge. 4. Rotating wheel.

2.4 Purification and Isolation of Protein Complexes

1. Primary (immunoprecipitating) antibody (see Note 2). 2. Secondary Antibody, anti-rabbit/mouse IgG (whole molecule)-Agarose, secondary antibody (see Notes 3 and 4).

128

Ian M. Evans and Ketevan Paliashvili

3. IgG control, Rabbit/mouse IgG-Agarose, 2.5 mg/mL (Sigma; see Note 5). 4. 1.5 mL microcentrifuge tubes. 5. NP40 lysis buffer (150 mM NaCl, 50 mM Tris–HCl at pH 8, and 0.5% NP40; see Note 6). 6. 0.6 M lithium chloride diluted in deionized water. 7. Tube rotator. 2.5 Western Blot for Protein Detection of Co-IP Complexes

1. 4–12% sodium dodecyl sulfate (SDS)-Polyacrylamide gel electrophoresis (PAGE) gel. 2. SDS-PAGE apparatus (see Note 7). 3. Transfer apparatus for western blot transfer. 4. Primary antibodies. 5. Anti-rabbit/mouse IgG horseradish peroxidase (HRP) conjugated secondary antibodies. 6. 2 Laemmli buffer. 7. PBS-T: PBS plus 0.1% Tween 20 (see Note 8). 8. Blocking buffer: 5% (w/v) skimmed milk in PBS-T (see Note 9). 9. Polyvinylidene fluoride (PVDF) membrane. 10. Enhanced chemiluminescence (ECL) reagent. 11. Developing Cassette. 12. Autoradiographic films. 13. MOPS-SDS Running buffer. 14. Transfer buffer. 15. Orbital shaker or rocking platform.

3

Methods All procedures are performed on ice unless otherwise indicated.

3.1

Cell Culture

1. Grow HUVECs or other cells of interest under appropriate culture conditions at 37  C and 5% CO2 to 70% confluency in 10 cm cell culture dishes containing 7–10 mL of EBM supplemented with 10% (v/v) FBS and antibiotics (see Note 10). 2. Make the cells quiescent by overnight incubation in 1% (v/v) FBS in EBM. VEGF stimulation at 25 ng/mL for 0–60 min can then be performed in serum-free EBM. 3. Place 10 cm dishes on ice. Carefully aspirate the medium from the dish and add 5–7 mL ice-cold, sterile PBS at the border of the dish. Slightly sway the dish and aspirate PBS. Repeat this

Co-immunoprecipitation

129

washing step twice to completely remove medium. Snap freeze cells immediately by placing the dishes on dry ice and store at 80  C or use straight away for cell lysis. 3.2

Cell Lysis

1. Add 1 mL NP40 lysis buffer containing protease and phosphatase inhibitors to 10 cm dishes and make sure it covers the entire dish. 2. Incubate plates on ice for 15 min. 3. Detach the cell monolayer from the surface of the dish using a cell scraper. 4. Transfer the lysed cell suspension to a 1.5 mL microcentrifuge tubes and centrifuge at 15  g for 15 min at 4  C. Transfer resulting supernatant into pre-cooled fresh 1.5 mL microcentrifuge tubes.

3.3 Pre-clearing the Lysates (See Note 11)

1. Add 20 μL of rabbit/mouse IgG agarose to lysate and rotate slowly for 1 h at 4  C. 2. Centrifuge at 15  g for 30 s at 4  C. 3. Discard the agarose bead pellet and keep supernatant.

3.4 Purification and Isolation of Protein Complexes

1. On ice, add 2 μg of antibody against the protein you wish to immunoprecipitate to the cell lysate and mix by pipetting up and down several times. 2. Incubate overnight at 4  C on the slow setting of a tube rotator. 3. After overnight incubation, add 20 μL of anti-rabbit/Mouse agarose IgG (whole molecule) agarose beads to lysates containing antibodies and 20 μL of rabbit/mouse IgG-agarose beads to IgG control samples without an antibody (to be used as a negative control sample), and rotate slowly for an additional 1 h at 4  C. 4. Centrifuge at 15  g for 30 s at 4  C and carefully remove the supernatant without disturbing the agarose pellet. 5. Resuspend the agarose in 500 μL NP40 lysis buffer and mix on a vortex mixer. Centrifuge at 15  g at 4  C for 30 s and carefully remove the supernatant, without disturbing the agarose pellet. Repeat this washing step three more times to remove nonspecific binding proteins. 6. Do final wash using 0.6 M Lithium chloride. 7. Add 20 μL of 2 Laemmli buffer to agarose pellet and heat at 95  C for 10 min. 8. Centrifuge at 15  g for 30 s and carefully transfer the supernatant (co-IP sample) to a clean 1.5 mL microcentrifuge tube. Avoid agarose carryovers.

130

Ian M. Evans and Ketevan Paliashvili

3.5 SDS-PAGE and Western Blot for Protein Detection of Co-IP Complexes

1. Assemble the gel running chamber and fill it with 1 running buffer. 2. Load 20 μL of co-IP samples including IgG control into the wells (see Note 12). 3. Perform electrophoresis at a current of 180 V until the bromophenol blue in the Laemmli buffer has reached the bottom of the gel. 4. Disassemble the gel running chamber, carefully lift the gel from the cassette and place it into a dish. 5. Incubate gel, filter papers and membrane in blotting buffer. 6. Place the gel on the membrane between two pieces of filter paper soaked with blotting buffer and blot for 1 h at 30 V. 7. Incubate membrane in blocking buffer for 1 h at room temperature, gently shaking on an orbital shaker. 8. Prepare the solution of primary antibody (depending on manufacturer’s recommendations) in blocking buffer (see Note 13). 9. Replace the blocking solution with the primary antibody solution and incubate overnight at 4  C on an orbital shaker. 10. The next day, pour off the primary antibody solution and wash 3 times with PBS-T, rocking for 10 min on an orbital shaker between each PBS-T change. 11. Prepare the solution of secondary antibody (1:10,000) in blocking buffer (see Note 13). 12. Remove the PBS-T wash and add secondary antibody to the membrane. Incubate on an orbital shaker for 1 h at room temperature. 13. Pour off secondary antibody solution and wash 3 times with PBS-T, rocking for 10 min on an orbital shaker between each PBS-T change. 14. Dab the membrane on tissue to remove excess liquid and add ECL reagent ensuring it is evenly covered (see Note 14). 15. Pour off excess ECL reagent and place the membrane in the developing cassette and expose it to an imaging film or chemiluminescent digital imaging system for an appropriate time (usually 30 s-5 min), and develop the film/capture the digital image.

4

Notes 1. This protocol has been optimized for HUVECs but should work for other cell types.

Co-immunoprecipitation

131

2. Many suppliers state the suitability of their antibodies for various applications such as immunoprecipitation. Using antibodies that have been validated for immunoprecipitation can save a lot of trial and error. 3. The species of IgG used for the secondary antibody is decided by which species the primary antibody was generated in. 4. This protocol is optimized for agarose-conjugated secondary antibodies. Other solid phase supports can be used, such as magnetic beads. In this case substitute the centrifugation steps (Subheading 3.4, steps 4 and 5) for separation using a magnetic rack. In our experience, the use of different solid phase supports can alter the Co-IP results. 5. It is important to add the same concentration of IgG control in terms of μg/μL as has been used for the secondary antibody. 6. The original NP40 detergent manufactured by the Shell Chemical Company is no longer available and is now commonly substituted with IGEPAL CA-630 (I8896, Sigma). 7. This protocol is based around the NuPAGE electrophoresis equipment and reagents, but any other electrophoresis systems can be used instead. 8. Phosphate buffered saline can be substituted with Tris Buffered saline (0.05 M Tris–HCl and 0.15 M sodium chloride, pH 7.6) if preferred. 9. 1–5% bovine serum albumin (BSA) can be used instead of milk for blocking preparation of the blocking buffer. 10. In our experience, a confluent 10 cm dish will provide enough sample to obtain a decent amount of immunoprecipitated (reasonably abundant) protein. If your protein of interest is less abundant, it may be best to pool several dishes or use a larger size (e.g., 15 cm). Alternatively, you can overexpress the protein although this may increase the risk of interaction artifacts. 11. The purpose of this step is to remove proteins that bind non-specifically to the solid phase support. 12. It is recommended to include a whole cell lysate in one of the wells. This will act as a positive control for the western blot antibody and will also help confirm that the signal produced by the IP is the same molecular weight as the protein of interest. 13. The exact dilution of primary and secondary antibodies is determined by the manufacturer’s guidelines. 14. Incubation times of the ECL reagent will depend on the manufacturer’s guidelines.

132

Ian M. Evans and Ketevan Paliashvili

References 1. Miura K (2018) An overview of current methods to confirm protein-protein interactions. Protein Pept Lett 25(8):728–733. https://doi.org/10. 2174/0929866525666180821122240 2. Evans IM, Kennedy SA, Paliashvili K, Santra T, Yamaji M, Lovering RC, Britton G, Frankel P, Kolch W, Zachary IC (2017) Vascular endothelial growth factor (VEGF) promotes assembly of the p130Cas interactome to drive endothelial chemotactic signaling and angiogenesis. Mol

Cell Proteomics 16(2):168–180. https://doi. org/10.1074/mcp.M116.064428 3. Margineanu A, Chan JJ, Kelly DJ, Warren SC, Flatters D, Kumar S, Katan M, Dunsby CW, French PM (2016) Screening for proteinprotein interactions using Forster resonance energy transfer (FRET) and fluorescence lifetime imaging microscopy (FLIM). Sci Rep 6:28186. https://doi.org/10.1038/srep28186

Chapter 9 Using Immortalized Endothelial Cells to Study the Roles of Adhesion Molecules in VEGF-Induced Signaling James A. G. E. Taylor, Christopher J. Benwell, and Stephen D. Robinson Abstract The ability to study the role of specific genes in endothelial cell biology is made possible by our ability to modulate their expression through siRNA or knockout technologies. However, many in vitro protocols, particularly those of a biochemical nature, require large numbers of endothelial cells. These types of analyses are encumbered by the need to repeatedly produce and characterize primary endothelial cell cultures and can be greatly facilitated by the use of immortalized microvascular endothelial cells. However, we have found that the manipulation of gene expression in these cells is not always straight forward. Here we describe how we alter gene expression in polyoma middle T antigen immortalized microvascular endothelial cells isolated from wild-type and genetically modified mice to study the role of cell adhesion molecules in downstream assays. Key words Angiogenesis, Endothelial cells, Immortalization, Nucleofection, VEGF-induced signaling

1

Introduction The use of immortalized cells has become an attractive tool in the field of in vitro cell biology. The most obvious benefit of overriding senescence is to achieve unlimited cellular expansion, thus granting the researcher access to otherwise unattainable cell numbers. In turn, rapid generation of larger cell populations broadens the experimental approaches cell biologists can perform, particularly when the cell type in question is unabundant in vivo and/or possesses a relatively limited proliferative capacity in vitro. Overcoming proliferation restrictions also lessens the need for continuous cellline re-derivation, thereby reducing the frequency of this costly and laborious procedure as well as the number of animal donors required (facilitating adherence to 3Rs principles [1]).

James A. G. E. Taylor and Christopher J. Benwell are contributed equally. Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_9, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

133

134

James A. G. E. Taylor et al.

Our own investigations utilize several discrete genetically engineered mouse models, each possessing one or more targeted mutations at the loci of endothelial adhesion receptor genes. Using these models, we hope to decrypt the complex interplay occurring between adhesion receptors of interest, and ultimately determine how an established receptor network coordinates vascular endothelial growth factor (VEGF)-induced angiogenesis in both physiological and pathological scenarios. Like many researchers utilizing a variety of knockout/mutant animal models, the constant isolation and characterization of primary endothelial cell (EC) cultures is both untenable and oft-times unethical. Cell immortalization is therefore a suitable solution. We routinely use polyoma middle T antigen (PyMT) immortalized microvascular endothelial cells (ECs) isolated from mouse lung to study the function of adhesion-related molecules in in vitro angiogenic assays (e.g., VEGF-induced signaling studies). We, and others, have shown PyMT immortalized ECs are a good model for these types of studies [2–6] as transformation does not alter underlying cellular mechanics and phenotypic traits [7]. Nonetheless, we have found that gene silencing in these cells, either by siRNA-mediated knockdown, or by Cre recombinase-induced deletion of floxed alleles, is not always straightforward. In our hands, PyMT immortalized ECs are resistant to: (1) lipid-based transfection methods; and (2) population wide gene excision when tamoxifen is administered in culture to ECs isolated from inducible Cre models (e.g., cdh5.Cre/ERT2 [8] and Pdgfb.Cre/ERT2 [9]). Here, we detail our methods for nucleofection of siRNA or TAT.Cre recombinase to drive targeted gene silencing or excision of floxed genes in PyMT immortalized microvascular ECs respectively. We also briefly describe how we would subsequently use these cells in VEGF-induced signaling assays.

2

Materials Prepare all solutions using ultrapure water (prepared by purifying deionized water, to attain a sensitivity of 18 MΩ-cm at 25
C) and analytical grade reagents. Prepare and store all reagents at 4
C (unless indicated otherwise).

2.1 Nucleofection of siRNA or TAT.Cre

1. Flask coating solution: 0.1% gelatin from porcine skin (Sigma: #1002473556); flasks coated for 30 min at 37
C, 5% CO2 (see Note 1). 2. IMMLEC (Immortalized Mouse Lung Endothelial Cell) media: 1:1 mix of Ham’s F-12 (Sigma: #51651C) and DMEM medium (low glucose) (Sigma: #D6046) supplemented with 10% Fetal bovine serum (FBS), 100 units/mL penicillin/streptomycin (P/S), 0.1 mg/mL heparin.

Generating and Using Immortalised Endothelial Cells

135

3. 0.25% trypsin-EDTA solution. 4. Nucleofection buffer: 200 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 137 mM NaCl, 5 mM KCl, 6 mM D-glucose, 7 mM Na2HPO4. 5. Gene-specific or non-targeting control siRNA resuspended as described by the manufacturer (see Note 2). 6. TAT.Cre recombinase (Sigma: #SCR508). 7. Lonza 4D Core/X Unit Nucleofector™. 8. Lonza 4D Core/X Unit Nucleofector™ cuvettes. 2.2 VEGF-Induced Signaling Assay

1. Flask coating solution: human-plasma (FN) (Sigma: #FC010) in PBS (see Note 3).

fibronectin

2. Serum-free OptiMEM® medium (Thermo Fisher Scientific: #31985070). 3. VEGF-A164 (see Note 4). 4. Tissue culture treated 6-well plates. 5. Electrophoresis sample buffer (ESB): 6.5 mM Tris-HCl, pH 7.4, 60 mM sucrose, 3% SDS.

3

Methods

3.1 siRNA Knockdown or TAT. Cre-Induced Deletion of Target Genes by Nucleofection

Transfection of ECs with either siRNA or TAT.Cre recombinase is performed by electroporation, according to the Lonza general nucleofection protocol, using the Lonza 4D Core/X Unit Nucleofector™. All siRNA-based nucleofection reactions should be ran alongside a non-targeting control pool siRNA reaction for comparison, and target knockdown efficiency should be confirmed (e.g., by western blotting or immunocytochemistry, see Fig. 1).

3.1.1 siRNA-Mediated Knockdown of Target Genes in PyMT Immortalized Microvascular ECs

1. Culture ECs on gelatin-coated tissue culture plasticware until confluent (see Note 1). 2. Wash cells with pre-warmed PBS before detaching using 3 mL 0.25% trypsin-EDTA at 37
C, 5% CO2 in a humidified tissue culture incubator for 5 min. 3. Resuspend the detached cells in pre-warmed IMMLEC media, bringing the final volume to 10 mL. 4. Per nucleofection reaction/cuvette, 1  106 cells are required in a final reaction volume of 100 μL. To achieve this, aliquot the corresponding resuspended volume of ECs, and centrifuge at 300  g for 5 min. 5. Slowly resuspend the cell pellet (containing 1  106 cells) in nucleofection buffer to a total volume of 100 μL (see Note 5). 6. Add this suspension to a pre-determined volume/concentration of siRNA and mix gently by pipetting.

136

James A. G. E. Taylor et al.

Fig. 1 siRNA knock down or TAT.Cre-induced deletion of NRP2 and α5 integrin by nucleofection. (a) siRNAmediated knockdown of neuropilin-2 confirmed by western blot. 1  106 ECs were nucleofected with either a NRP2-specific siRNA pool (siNRP2) or a control siRNA pool (siCTL) using a Lonza 4D Core/X Unit Nucleofector™ under program EO100. NRP2 expression in the ECs lysates was analyzed by western blot alongside expression of heat-shock cognate protein 70 (Hsc70) as loading control to confirm even protein loading. (b) TAT.Cre recombinase-mediated deletion of α5 integrin from α5 integrin floxed (fl/fl) derived ECs confirmed by immunocytochemistry. 1.5  106 ECs were nucleofected with or without TAT.Cre recombinase using a Lonza 4D Core/X Unit Nucleofector™ under program EO100, before fixation and immunostaining for α5 integrin (green) and DAPI staining (blue). Confocal microscopy image on the left shows one EC from control treated sample (without TAT.Cre recombinase addition) and the image on the right shows an EC nucleofected in presence of TAT.Cre recombinase (note the absence of α5 integrin staining on the green channel confirming gene deletion and resulting lack of protein expression). Scale bars ¼ 20 μm

7. Add 100 μL of resuspended cells in nucleofection buffer with siRNA to the cuvette ensuring no bubbles sit at the surface, and place in the Lonza 4D Core/X Unit Nucleofector™. 8. Run the EO100 program as per the Lonza general nucleofection protocol (see Note 6). 9. Carefully add the transfected cell suspension to a pre-coated T25 flask or tissue culture dish containing a suitable volume of pre-warmed IMMLEC media without mixing (see Note 7). 10. Incubate in a humidified tissue culture incubator (37
C, 5% CO2) for a minimum of 24 h before assessing knockdown efficiency by western blotting or by immunocytochemistry (Fig. 1a). 3.1.2 TAT.Cre-Mediated Deletion of Target Genes in PyMT immortalized Microvascular ECs

TAT.Cre recombinase driven excision of floxed genes in PyMT immortalized microvascular ECs follows a similar protocol to using siRNA to effectively delete genes of interest: 1. Culture ECs on gelatin-coated tissue culture plasticware until confluent (see Note 1). 2. Wash cells with pre-warmed PBS before detaching using 3 mL 0.25% trypsin-EDTA at 37
C, 5% CO2 in a humidified tissue culture incubator for 5 min. 3. Resuspend the detached cells in pre-warmed IMMLEC media, bringing the final volume to 10 mL.

Generating and Using Immortalised Endothelial Cells

137

4. Per nucleofection reaction/cuvette, 1.5  106 cells are required in a final reaction volume of 100 μL. To achieve this, aliquot the required resuspended volume of endothelial cells, and centrifuge for 5 min at 300  g until a cell pellet is formed. 5. Slowly resuspend the cell pellet (containing 1.5  106 cells) in nucleofection buffer to a total volume of 100 μL (see Note 5). 6. Add this suspension to 7 μL of TAT.Cre recombinase and mix gently by pipetting. 7. Add 100 μL of resuspended cells in nucleofection buffer with TAT.Cre recombinase to the cuvette ensuring no bubbles sit at the surface, and place in the Lonza 4D Core/X Unit Nucleofector™. 8. Run the EO100 program as per the Lonza general nucleofection protocol (see Note 6). 9. Carefully add the transfected cell suspension to a pre-coated T25 flask or tissue culture dish containing a suitable volume of pre-warmed IMMLEC media without mixing (see Note 7). 10. Incubate in a humidified tissue culture incubator (37
C, 5% CO2) for a minimum of 24 h. 11. Repeat this procedure again 24 h post-TAT.Cre nucleofection before assessing knockdown efficiency by western blotting or by immunocytochemistry (Fig. 1b). 3.2 Assessing the Effects of Target Genes Knockdown/Deletion on VEGF-Induced Signaling by Western Blot

VEGF-A is the best characterized of the VEGF family members and is the most widely researched soluble angiogenic growth factor with essential roles in the regulation of endothelial cell behavior. Binding of VEGF-A to the extracellular domain of vascular endothelial growth factor receptor 2 (VEGFR-2) initiates receptor dimerization and subsequent trans-autophosphorylation of its intracellular tyrosine kinase domain, initiating a cascade of downstream signaling events responsible for mediating a number of important cellular processes such as cell proliferation, migration, and survival. This occurs via the activation and propagation of various signaling networks, such as phospholipase-Cγ (PLCγ)/ protein kinase C (PKC), Ras/Raf/ERK/MAPK, and PI3K/Akt pathways [10]. Studies have demonstrated that numerous endothelial adhesion receptors, such as integrins and members of the neuropilin family, regulate or are regulated by VEGF-induced signaling [11, 12]. The ability to effectively assess which VEGFinduced signaling networks are temporally sensitive to the actions of specific adhesion receptors is of great importance to understanding how endothelial dynamics are controlled during angiogenesis. This can be achieved by incubating cultured endothelial cells, with decreased expression of one or more adhesion receptors of interest following the protocols aforementioned, with VEGF-A over set timepoints followed by quantification of the relative

138

James A. G. E. Taylor et al.

phosphorylation levels of different signaling proteins via western blotting. 1. Culture endothelial cells lacking the adhesion receptor of interest in complete medium as described in Subheading 3.1.1 or 3.1.2, until ~80% confluent in FN pre-coated tissue culture plates. 2. Remove and discard the complete medium before washing the ECs twice with pre-warmed PBS. 3. Incubate endothelial cells in pre-warmed serum-free OptiMEM® for 3 h in a humidified tissue culture incubator (37
C, 5% CO2), to starve the cells of any remaining growth factor stimulation. 4. Following starvation, replace the serum-free OptiMEM® with fresh pre-warmed OptiMEM® medium supplemented with 30 ng/mL VEGF-A, and incubate sequentially for 0, 5, 15, 30, 45, or 60 min in a humidified tissue culture incubator (37
C, 5% CO2) (see Note 8). 5. To stop continued VEGF-A activity once the desired stimulatory period has elapsed, quickly place the culture dishes on ice and immediately remove the VEGF-A supplemented medium, washing twice with ice-cold PBS. Leave in ice-cold PBS until all incubations are complete. 6. Following completion of all incubation timepoints, endothelial cells can be lysed in ESB, quantified and protein expression analysis can be performed by western blotting (see Notes 9–11 and Fig. 2 for summary of Subheading 3.2).

4

Notes 1. Unless otherwise stated, all tissue culture plasticware should be pre-coated with flask coating solution and incubated in a humidified tissue culture incubator (37
C, 5% CO2) for a minimum of 20 min prior to use. Remove excess flask coating solution from the flask immediately prior to use in a biological safety cabinet. 2. Preliminary assays should be performed to determine the siRNA specific concentration required to induce efficient knock down of gene of interest, in addition to the length of time the protein expression remains decreased. 3. Because we are frequently studying the function of Arginylglycylaspartic acid (RGD)-binding integrins, we often use FN coated dishes in our studies, but other extracellular matrices (such as collagen, gelatin, or vitronectin) can be used. When containing plates with FN, use at 2 mg/mL solution diluted in PBS.

Generating and Using Immortalised Endothelial Cells

139

Fig. 2 Schematic representation of the VEGF time-course assay outlined in Subheading 3.2. (1) Culture ECs until ~80% confluent. (2) Remove complete medium before washing twice with pre-warmed PBS. (3) Incubate ECs in pre-warmed serum-free OptiMEM® for a minimum of 3 h. (4) Replace OptiMEM® with fresh pre-warmed OptiMEM® supplemented with 30 ng/mL VEGF-A and incubate each plate for the desired period (sequential 0-, 5-, and 15-min incubations are illustrated here). (5) Following stimulation quickly place the dishes on ice before removing the VEGF-A supplemented OptiMEM® and performing two ice-cold PBS washes. (6) Using ESB (e.g., 100 μL for a 10 cm dish), manually lyse the ECs and collect the lysate for subsequent western blot analysis. (Created with BioRender.com)

4. We use VEGF-A164 manufactured in-house, according to the method of Krilleke et al. [13] but VEGF-A164 can also be sourced commercially. In addition, other growth factors can be used for investigation (e.g., fibroblast growth factor, etc.). 5. Avoid leaving cell suspensions in the nucleofection buffer for longer than 20 min as this will greatly reduce survival following electroporation. 6. We have also used the Lonza Nucleofector II™ apparatus, using program T-005. 7. After removing the cuvette from the Lonza 4D Core/X Unit Nucleofector™ tray, every effort should be made not to disturb the cell suspension. Any vigorous pipetting or mixing will result in cell lysis and poor survival. 8. We have empirically tested this concentration of the growth factor in our cells using the VEGF-A we produce in-house such that we achieve maximal response at this dose; for best practice,

140

James A. G. E. Taylor et al.

we recommend performing a dose response curve in wild-type cells before proceeding to test cells with altered gene expression. 9. When preparing lysates for western blotting, ensure sufficient lysate from each sample is set aside to confirm target deletion of your protein of interest. 10. Following transfer of proteins, nitrocellulose membranes should be sequentially incubated in primary antibodies detecting the phosphorylated state of the signaling receptor of interest, followed by an antibody to detect total receptor expression. 11. All quantification of the phosphorylated and total protein should be made relative to a protein loading control. We frequently use GAPDH or Hsc70 for endothelial lysates. References 1. Sneddon LU, Halsey LG, Bury NR (2017) Considering aspects of the 3Rs principles within experimental animal biology. J Exp Biol 220(Pt 17):3007–3016. https://doi. org/10.1242/jeb.147058 2. Alghamdi AAA, Benwell CJ, Atkinson SJ, Lambert J, Johnson RT, Robinson SD (2020) NRP2 as an emerging angiogenic player; promoting endothelial cell adhesion and migration by regulating recycling of alpha5 integrin. Front Cell Dev Biol 8:395. https://doi.org/ 10.3389/fcell.2020.00395 3. Ni CW, Kumar S, Ankeny CJ, Jo H (2014) Development of immortalized mouse aortic endothelial cell lines. Vasc Cell 6(1):7. https://doi.org/10.1186/2045-824X-6-7 4. Robinson SD, Reynolds LE, Kostourou V, Reynolds AR, da Silva RG, Tavora B, Baker M, Marshall JF, Hodivala-Dilke KM (2009) Alphav beta3 integrin limits the contribution of neuropilin-1 to vascular endothelial growth factor-induced angiogenesis. J Biol Chem 284(49):33966–33981. https://doi. org/10.1074/jbc.M109.030700 5. Steri V, Ellison TS, Gontarczyk AM, Weilbaecher K, Schneider JG, Edwards D, Fruttiger M, Hodivala-Dilke KM, Robinson SD (2014) Acute depletion of endothelial beta3-integrin transiently inhibits tumor growth and angiogenesis in mice. Circ Res 114(1):79–91. https://doi.org/10.1161/ CIRCRESAHA.114.301591 6. Tavora B, Reynolds LE, Batista S, Demircioglu F, Fernandez I, Lechertier T, Lees DM, Wong PP, Alexopoulou A, Elia G, Clear A, Ledoux A, Hunter J, Perkins N,

Gribben JG, Hodivala-Dilke KM (2014) Endothelial-cell FAK targeting sensitizes tumours to DNA-damaging therapy. Nature 514(7520):112–116. https://doi.org/10. 1038/nature13541 7. Atkinson SJ, Gontarczyk AM, Alghamdi AA, Ellison TS, Johnson RT, Fowler WJ, Kirkup BM, Silva BC, Harry BE, Schneider JG, Weilbaecher KN, Mogensen MM, Bass MD, Parsons M, Edwards DR, Robinson SD (2018) The beta3-integrin endothelial adhesome regulates microtubule-dependent cell migration. EMBO Rep 19(7):e44578. https://doi.org/10.15252/embr.201744578 8. Sorensen I, Adams RH, Gossler A (2009) DLL1-mediated Notch activation regulates endothelial identity in mouse fetal arteries. Blood 113(22):5680–5688. https://doi.org/ 10.1182/blood-2008-08-174508 9. Claxton S, Kostourou V, Jadeja S, Chambon P, Hodivala-Dilke K, Fruttiger M (2008) Efficient, inducible Cre-recombinase activation in vascular endothelium. Genesis 46(2):74–80. https://doi.org/10.1002/dvg.20367 10. Cebe-Suarez S, Zehnder-Fjallman A, BallmerHofer K (2006) The role of VEGF receptors in angiogenesis; complex partnerships. Cell Mol Life Sci 63(5):601–615. https://doi.org/10. 1007/s00018-005-5426-3 11. Herzog B, Pellet-Many C, Britton G, Hartzoulakis B, Zachary IC (2011) VEGF binding to NRP1 is essential for VEGF stimulation of endothelial cell migration, complex formation between NRP1 and VEGFR2, and signaling via FAK Tyr407 phosphorylation.

Generating and Using Immortalised Endothelial Cells Mol Biol Cell 22(15):2766–2776. https://doi. org/10.1091/mbc.E09-12-1061 12. Soldi R, Mitola S, Strasly M, Defilippi P, Tarone G, Bussolino F (1999) Role of alphavbeta3 integrin in the activation of vascular endothelial growth factor receptor-2. EMBO J 18(4):882–892. https://doi.org/10.1093/ emboj/18.4.882

141

13. Krilleke D, DeErkenez A, Schubert W, Giri I, Robinson GS, Ng YS, Shima DT (2007) Molecular mapping and functional characterization of the VEGF164 heparin-binding domain. J Biol Chem 282(38):28045–28056. https://doi.org/10.1074/jbc.M700319200

Chapter 10 RNAscope for VEGF-A Detection in Human Tumor Bioptic Specimens Tiziana Annese, Roberto Tamma, and Domenico Ribatti Abstract Different pro-angiogenic factors, such as vascular endothelial growth factor-A (VEGF-A), have been related to microvascular density, clinicopathologic factors, and poor prognosis in many tumors. VEGF-A binds its receptor 2 (VEGFR2) to induce neo-angiogenesis, a constant hallmark of tumor initiation and progression. Based on VEGF-A/VEGFR2 relevance in tumor angiogenesis, several inhibitors were developed. However, the clinical benefits of anti-angiogenic therapies are limited because tumors activate different mechanisms of drug resistance. The need for understanding tumor biology, limitation or failure of anti-angiogenic therapies, and the demand for a personalized therapeutic approach has boosted the search for robust biomarkers for patient stratification as responder or non-responder to anti-VEGF therapies. This chapter presents a detailed protocol to perform chromogenic VEGF-A mRNA detection and quantification in human tumor bioptic specimens using RNAscope technology and RNA-in situ hybridization (ISH) algorithm. RNAscope for VEGF-A detection, even for small amounts, is compatible with precious clinical samples and diagnostic laboratory workflows. Key words Angiogenesis, Cancer diagnosis, Cancer therapy, RNAscope assay, Vascular Endothelial Growth Factor, Whole-slide digital imaging

1

Introduction Vascular endothelial growth factor (VEGF), in the past, also known as vascular permeability factor (VPF), possesses mitogenic and permeability enhancing activities [1]. VEGF is a family of factors, including placental growth factor (PLGF), VEGF-A, VEGF-B, VEGF-C, and VEGF-D (also called c-Fos-induced growth factor, FIGF), and the viral VEGF-Es [2, 3]. VEGF family members explicate their biological functions by binding protein tyrosine kinase receptors (VEGFRs) [4, 5]. VEGF-A is the primary regulator of angiogenesis, vasculogenesis, and endothelial cell growth in normal and pathological conditions, and it is critical for the maintenance of vessel homeostasis

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_10, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

143

144

Tiziana Annese et al.

[6]. It induces endothelial cell proliferation, promotes cell migration, inhibits apoptosis, and provokes blood vessels‘permeabilization. VEGF-A binds to VEGFR2 (also called KDR/Flk-1) and VEGFR1 (Flt-1), which differ considerably in signaling properties. Moreover, VEGF-A binds to non-tyrosine kinase receptors called neuropilins (NRPs) and heparan sulfate [7–9]. Based on UniProt data resources, 17 humans VEGF-A isoforms are described, and additional isoforms seem to exist (Primary accession number: P15692). They are produced by alternative promoter usage, alternative splicing, and alternative initiation. High VEGF-A expression has been demonstrated in many tumors, and correlates with microvascular density, clinicopathologic factors, such as metastasis and poor prognosis [10–14]. It is produced and secreted by tumor and microenvironment cells during hypoxia that induces new vessel formation mainly via activation VEGFR2 expressed by endothelial cells [15]. VEGF-A/VEGFR2 pathway activation stimulates multiple signaling pathways leading to endothelial cell survival, proliferation, migration, invasion, vascular permeability, and vascular inflammation [16]. Based on VEGF-A/VEGFR2 relevance in tumor angiogenesis, several inhibitors were developed, such as monoclonal anti-VEGFA antibody (Bevacizumab), recombinant fusion VEGF-A protein (Aflibercept), monoclonal anti-VEGFR2 antibody (Ramucirumab), and multi-tyrosine kinase inhibitors (i.e., Sunitinib, Sorafenib, etc.) [17]. However, despite the growing list of Food and Drug Administration (FDA) authorized drugs, the anti-angiogenic therapy clinical benefits are limited because tumor cells can adopt alternative ways to grow, such as vascular co-option and mimicry, bypassing the need for tumor angiogenesis, and moreover, can switch between angiogenic and non-angiogenic phenotypes [18, 19]. Validation of robust biomarkers able to stratify patients as responder or non-responder to anti-VEGF therapies is imperative [20, 21]. With the spread of precision medicine, sensitive and specific molecular biomarkers are essential because they enable clinical outcomes in a relatively earlier stage [22, 23]. Biomarkers can be diagnostic, prognostic, or predictive. A diagnostic biomarker allows early disease detection. A prognostic biomarker permits prediction of disease outcomes and risk assessment independent of treatments. Predictive biomarkers provide the prediction treatment responses [24]. Biomarkers also offer potential targets for therapy and knowledge about biological mechanisms associated with diseases [25]. RNA molecules, an essential part of the central dogma of molecular biology, are emerging as cancer biomarkers as they provide genetic and regulatory information that reflects cellular states [26–28]. Although real-time PCR is still reputed as the goldstandard technique for verifying differential gene expression

RNAscope in Human Tumor Biopsies

145

profiles, the tissue’s destruction during RNA extraction prevents RNA analysis in the morphological context [29]. The novel RNA-in situ hybridization method, called RNAscope, has been developed to preserve the morphological context and avoid interference from unwanted cells and tissue elements [30, 31]. For instance, we performed RNAscope experiments to study VEGF-A and VEGFR2 expression in gastric cancer, focusing our analysis on tumor cells lining the tubular glands and the endothelial cells, excluding stroma and inflammatory cells [10]. RNAscope strength is in a probe design and hybridizationbased signal amplification system that simultaneously amplifies signals and suppresses the background [32]. Moreover, it is possible to combine in situ hybridization for RNA/DNA detection with protein labeling to evaluate multiple signals on the same section providing a powerful method that allows the integration of molecular and histopathology data for optimal clinical interpretation [33]. RNAscope is compatible with precious clinical samples and diagnostic laboratory workflows. Even small amounts of RNA at the individual cell level can be detected. Classic molecular biology techniques such as real-time PCR are susceptible to interference from other unwanted cell types, such as non-cancer cells, and other unwanted tissue elements, such as fibrosis and necrosis. However, RNAscope is more expensive than the conventional in situ hybridization method and requires the user to have morphometric skills. Immunohistochemical VEGF-A evaluation constitutes a valuable predictor of chemotherapy response in some carcinomas [34, 35]. However, the clinical benefits of anti-angiogenic therapies are limited because tumors activate different mechanisms of drug resistance. The need for understanding tumor biology, limitation or failure of anti-angiogenic therapies, and the demand for a personalized therapeutic approach has boosted the search for robust biomarkers for patient stratification as responder or non-responder to anti-VEGF therapies. Here, we present a detailed protocol to perform chromogenic VEGF-A mRNA detection and quantification in human tumor bioptic specimens using RNAscope technology and RNA ISH algorithm, as a valid and robust method to improve patient stratification.

2

Materials

2.1 Basic Laboratory Supplies

1. 4% paraformaldehyde in PBS. 2. 15% Saccharose in PBS. 3. 100% ethanol (EtOH). 4. Xylene.

146

Tiziana Annese et al.

5. Paraffin wax 58–60  C. 6. Mold cassettes. 7. Solvent-resistant staining dishes with slide rack insert. 8. Cover glasses. 9. Ammonium hydroxide 1.0 N. 10. Carboy 3 L. 11. Water bath capable of holding temperature 40–100  C. 12. Heating plate. 13. Pipettors and tips, 0.2–5000 μL. 14. Distilled water. 15. Tubes (various sizes). 16. Glass beaker (various sizes). 17. Eppendorf (various sizes). 18. Bottles (various sizes). 19. Fume hood. 20. Graduated cylinder (various sizes). 21. Parafilm. 22. Absorbent paper. 23. Tweezers. 24. Tap water. 25. High Flow Syringe Filter 0.22 μm. 26. Aluminum foil. 27. Thermometer. 2.2 Histology Laboratory Supplies

1. Microtome for Paraffin Sectioning. 2. Glycergel Mounting Agilent Dako).

medium

(Cat.

No.

C0563,

3. Polysine Adhesion Slides (Cat. No. J2800AMNZ, Thermo Scientific). 4. Hematoxylin Solution, Gill I (Cat. No. 1 GHS132, SigmaAldrich). 2.3 RNAscope Equipment Supplied by Advanced Cell Diagnostic (ACD)

1. RNAscope Target Probes: Hs-VEGFA (Cat No. 310061), negative control dapB (Cat No. 310043), positive control Hs-PPIB (peptidylprolyl isomerase B (cyclophilin B, PPIB) Cat No. 313901). 2. RNAscope® 2.5 HD Reagent Kit RED (Cat No. 322350, ACD) including pre-treatments (Hydrogen Peroxide Cat No. 322335; Protease Plus Cat No. 322331; Target Retrieval

RNAscope in Human Tumor Biopsies

147

Reagent; TRR; Cat No. 322000), detection kit (Cat No. 322360, ACD) and wash buffer (WB; Cat No. 310091). 3. HybEZTM Oven (Cat No. 310010). 4. HybEZTM Control Tray with Lid (Cat. No. 310012). 5. HybEZTM Slide Rack (Cat. No. 310014). 6. HybEZTM Slide Humidifying Paper (Cat. No. 310014). 7. ImmEdge Hydrophobic Barrier Pen (Cat. No. 310018). 2.4 RNAscope Solutions

1. Target Retrieval Reagent (TRR): for 1 TRR, add 70 mL of 10 TRR concentrate to 630 mL of distilled water using a graduate cylinder. 2. Wash buffer (WB): for 1 WB, pre-warm 60 mL of 50 WB concentrate at 40  C for 30 min, then add this to 2.94 L of distilled water using a graduate cylinder. Store in a carboy 3 L. 3. Prepare 50% Gill’s I hematoxylin solution: filter Gill’s I hematoxylin solution with 0.22 μm pore size syringe filter, and dilute it at 50% with distilled water, and store at RT protected from light. 4. 0.02% ammonia water: add 1.43 mL of 1 N ammonium hydroxide solution to 250 mL of distilled water in a graduated cylinder.

2.5 Image Acquisition and Analysis Systems

3

1. Aperio Scanscope CS platform (Leica Biosystems). 2. Aperio ImageScope software (Leica Biosystems). 3. Aperio RNA ISH algorithm (Leica Biosystems).

Methods

3.1 Sample Preparation

Tissue samples are fixed in 4% paraformaldehyde at 4  C for a time depending on sample volume (penetration rate 1 mm/h), in a total fixative volume triple that sample volume, and are embedded in paraffin as detailed below. After fixation, cancer biopsy is processed as follows: 1. Put biopsy in a tube filled with a volume of 15% SaccharosePBS (three times the volume of the tissue volume) for 30 s. 2. Refill the tube with fresh 15% Saccharose-PBS and leave the biopsy for 1 h at 4  C (see Note 1). 3. Dehydrate biopsy in a tube filled with 70% EtOH overnight (three times the volume of the sample) (see Note 2). 4. Dehydrate biopsy in a tube filled with 96% EtOH for 1.5 h at 4  C.

148

Tiziana Annese et al.

5. Dehydrate biopsy in a tube filled with fresh 96% EtOH for 1.5 h at 4  C. 6. Dehydrate biopsy in a tube filled with 100% EtOH for 1.5 h at 4  C. 7. Dehydrate biopsy in a tube filled with fresh 100% EtOH for 1.5 h at 4  C. 8. Perform biopsy clarification in a bottle filled with Xylene for 1 h at RT. 9. Perform biopsy clarification in a bottle filled with fresh Xylene for 1 h at RT. 10. Perform biopsy clarification in a bottle filled with fresh Xylene overnight at RT. 11. Immerse biopsy in a glass beaker filled with paraffin wax for 1 h at 60  C. 12. Immerse biopsy in a glass beaker filled with fresh Paraffin wax for 1 h at 60  C. 13. Immerse biopsy in a glass beaker filled with fresh Paraffin wax for 1.5 h at 60  C. 14. Embed sample in mold cassettes destined for cutting or stored at room temperature. 15. Cut and collect 4 μm thick sections on polylysine glass slides (see Note 3). 3.2 RNAscope Protocol

The following steps are based on the protocol provided by ACD. RNAscope: Day1 1. Deparaffinize slides by heating at 60 HybEZTM Oven (see Note 4).



C for 1 h in the

2. Improve deparaffinization by dipping slides in fresh Xylene in a staining dish, 1  5 min, working in a fume hood. 3. Repeat step 2 with fresh Xylene. 4. Dehydrate tissue sections immersing slides in fresh 100% EtOH in a staining dish, 1  1 min, working in a fume hood. 5. Repeat step 4 with fresh 100% EtOH. 6. Place slides on HybEZTM Slide Rack to air-dry for 1 h at RT (see Note 5). 7. Place slide rack with tissue sections into HybEZTM Control Tray to protect slides from light during the next step. 8. Block endogenous peroxidase by incubating tissue sections with 3 drops of Hydrogen Peroxide (pretreat-1 solution) for 10 min in the dark (see Note 6). 9. Wash the slides in distilled water 2  1 min by immersion in a staining dish.

RNAscope in Human Tumor Biopsies

149

10. Equilibrate tissue sections in boiling (98–102  C) distilled water for 10 s. 11. Immerse slides in target retrieval agent (TRR; pretreat-2solution) for 15 min at 98–102  C (see Note 7). 12. Wash the slides in distilled water 2  1 min. 13. Dehydrate slides in 100% EtOH 1  1 min. 14. Air-dry the slides for at least 1 h at RT. 15. Create a hydrophobic barrier with ImmEdge Hydrophobic Barrier Pen around the tissue sections. 16. Leave tissue sections on the bench overnight at RT to facilitate the barrier drying. RNAscope: Day 2

17. Turn on at 40  C the HybEZTM Oven for at least 30 min (see Note 8). 18. Place one humidifying paper, wet with 50 mL of distilled water, in the HybEZTM Control Tray (see Note 9). 19. Incubate tissue sections with 4 drops of protease (pretreat-3 solution) for 30 min at 40  C using the HybEZTM Oven. 20. Wash the slides in a staining dish filled with distilled water 2  1 min at RT. 21. Incubate tissue sections with 4 drops of Hs-VEGF-A, or dapB, or Hs-PPIB probes for 2 h at 40  C using the HybEZTM Oven. 22. Wash the slides in a staining dish filled with wash buffer (WB) 2  2 min at RT. 23. Incubate tissue sections with 4 drops of AMP-1 for 30 min at 40  C using the HybEZTM Oven (see Note 10). 24. Wash the slides in a staining dish filled with WB 2  2 min at RT. 25. Incubate tissue sections with 4 drops of AMP-2 for 15 min at 40  C using the HybEZTM Oven. 26. Wash the slides in a staining dish filled with WB 2  2 min at RT. 27. Incubate tissue sections with 4 drops of AMP-3 for 30 min at 40  C using the HybEZTM Oven. 28. Wash the slides in a staining dish filled with WB 2  2 min at RT. 29. Incubate tissue sections with 4 drops of AMP-4 for 15 min at 40  C using the HybEZTM Oven. 30. Wash the slides in a staining dish filled with WB 2  2 min at RT.

150

Tiziana Annese et al.

31. Incubate tissue sections with 4 drops of AMP-5 for 30 min at RT (see Note 11). 32. Wash the slides in a staining dish filled with WB 2  2 min at RT. 33. Incubate tissue sections with 4 drops of AMP-6 for 15 min at RT. 34. Wash the slides in a staining dish filled with WB 2  2 min at RT (see Note 12). 35. Incubate tissue sections with chromogen by pipetting 61 μL per section of Fast-RED working solution for 10 min at RT (see Note 12). 36. Place slides in the staining dish in the sink and wash slides under flowing tap water for 2 min. 37. Counterstain tissue sections with 50% Gill’s I hematoxylin solution for 2 min to detect nuclei (directly placed on the sections and not by immersion). 38. Wash slides under flowing tap water for 2 min. 39. Immerse slides in 0.02% ammonia water for 30 s to develop blue-violet nuclear staining. 40. Wash slides under flowing tap water for 2 min. 41. Mount slides with 1–2 drops of Glycergel Mounting media (see Note 13). 3.3 Image Acquisition

1. Check under a bright-field microscope the presence and quality of PPIB dots (positive control) and the absence of dapB dots (negative control) to verify your biopsy RNA quality. 2. Scan slides using the whole-slide scanning platform Aperio Scanscope CS at the maximum available magnification (20 at least). 3. Store image at digital high resolution as TIFF.

3.4

Image Analysis

1. With Aperio ImageScope software, select 3–10 annotation fields with an equal area for the analysis at 20 or 40 magnification (see Note 14). 2. Run the Aperio RNA ISH algorithm embedded in the ImageScope and set the parameter for cell identification, signal detection, and scoring criteria for Fast-RED semiquantitative image analysis (Fig. 1). 3. The results will be divided into customized ranges: 0, which includes cells that do not contain dots per cell; 1+, which includes cells containing 1–5 dots per cell; 2+, which includes cells containing 6–20 dots per cell; and 3+, which includes cells containing more than 21 dots per cell.

RNAscope in Human Tumor Biopsies

151

Fig. 1 RNAscope assay for VEGF-A mRNA expression in human gastric carcinoma biopsy. Hela cells are used as a positive method control. If Hela cells staining for the positive control probe PPIB shows perfect red dots, it means that the method has been flawlessly executed and that the reagents are fully functional (a). PPIB red dots on gastric carcinoma sections ensure the good RNA quality of the biopsy (b). Example of VEGF-A mRNA

152

Tiziana Annese et al.

4. Report results in a graph as mean  sd of number or percentage of Fast-RED positive probes per cell in each range and performs statistical analysis (Fig. 1).

4

Notes 1. Biopsies should be kept in Saccharose solution until they sink before proceeding with the next step. 2. It is possible to store the samples in 70% EtOH for years. 3. Tissue sections are collected up to 3 months before the experiment is carried out. If sections are kept for longer than 3 months, the RNA may be degraded due to direct exposure to atmospheric agents even if properly stored. Before sectioning, tissues paraffin-embedded can be stored for years. 4. HybEZTM Oven supplied by ACD is recommended due to its perfect temperature and humidity control. 5. This is an optional stopping point in which the sections could be left air-dry for more time, but we discourage prolonging for more than 3 h. During drying prepare TRR, wash buffer, hematoxylin, and ammonia water solutions. 6. During peroxidase blocking, fill two beakers with 700 mL of distilled water and 700 mL of TRR and boil them at 98–102  C on a hot plate (do not exceed 30 min for TRR). 7. TRR treatment time depends on tumor type. 15 min is the standard, but it is sometimes advisable to increase or reduce the times based on tissue degradation. For instance, when fixed but not embedded samples are used, less time is mandatory [36]. 8. During this time, keep the reagents AMP1–6 from the detection kit at RT and pre-warm probes at 40  C for at least 10 min in a water bath. If Glycergel Mounting medium is used, keep it at 40  C until slides are mounted. 9. Wet paper creates a humidifying chamber that avoids excessive evaporation during probes and AMP1–6 incubation. 10. Tissue sections could be covered with coverslips to favor AMP solution diffusion and reduce the number of drops for each section.

ä Fig. 1 (continued) expression in tumor cells lining the tubular glands (c). Markup image of the same field in (b) after running the RNA ISH algorithm (d). Example of setting for VEGF-A RNAscope assay with Fast-RED chromogen (e). VEGF-A mRNA expression as the number of dots per cell in different gastric carcinoma samples is reported as means  sd (f). Scale bar: a, c–d 25 μm; b 100 μm

RNAscope in Human Tumor Biopsies

153

Fig. 2 VEGF-A mRNA expression in human mastocytoma biopsy. Under a bright-field microscope, the mastocytoma section shows no staining for VEGF-A mRNA, and only the nuclear one is visible (a). The same field evaluated under an epifluorescence microscope shows visible VEGF-A mRNA red dots (b). The overlapped image of nuclear staining and fluorescence shows VEGF-A expression by mast cells (c; arrows; recognizable by their granular cytoplasm). Scale bar: a–c 30 μm

11. Improve AMP-5 time incubation if a more robust signal is needed. 12. During this time and keeping tissue sections in WB, prepare the Fast-RED working solution by mixing 1 μL of Fast-RED A with 60 μL of Fast-RED B, and 61 μL of this mixture are applied for one section. It must be prepared freshly and used within 5 min without exposing it to sunlight or UV rays. 13. Observe slides with a bright-field microscope to evaluate the RNAscope experiment success as being the probe’s red spots signals visible at 20 to decide whether to acquire images or not. Fast-RED chromogen is also visible in fluorescence. Faint signals not visible in the bright-field could be appreciated in fluorescence (Fig. 2). However, the experiment is not successful because the quantitation with the RNA ISH algorithm will not perform. The algorithm works with signals detectable in bright fields and not in fluorescence. Repeat the experiment with new reagents and/or increase the incubation time with AMP-5. 14. The annotation fields for analysis should be selected in such a number and position to represent the whole section. References 1. Ferrara N (2016) VEGF and intraocular neovascularization: from discovery to therapy. Transl Vis Sci Technol 5(2):10. https://doi. org/10.1167/tvst.5.2.10 2. Li X, Eriksson U (2001) Novel VEGF family members: VEGF-B, VEGF-C and VEGF-D.

Int J Biochem Cell Biol 33(4):421–426. https://doi.org/10.1016/s1357-2725(01) 00027-9 3. Shibuya M (2003) Vascular endothelial growth factor receptor-2: its unique signaling and specific ligand, VEGF-E. Cancer Sci 94(9):

154

Tiziana Annese et al.

751–756. https://doi.org/10.1111/j. 1349-7006.2003.tb01514.x 4. Murphy JF, Fitzgerald DJ (2001) Vascular endothelial growth factor induces cyclooxygenase-dependent proliferation of endothelial cells via the VEGF-2 receptor. FASEB J 15(9):1667–1669. https://doi.org/ 10.1096/fj.00-0757fje 5. Shibuya M (2011) Vascular endothelial growth factor (VEGF) and its receptor (VEGFR) signaling in angiogenesis: a crucial target for antiand pro-angiogenic therapies. Genes Cancer 2(12):1097–1105. https://doi.org/10.1177/ 1947601911423031 6. Takahashi H, Shibuya M (2005) The vascular endothelial growth factor (VEGF)/VEGF receptor system and its role under physiological and pathological conditions. Clin Sci (Lond) 109(3):227–241. https://doi.org/10.1042/ CS20040370 7. Soker S, Takashima S, Miao HQ, Neufeld G, Klagsbrun M (1998) Neuropilin-1 is expressed by endothelial and tumor cells as an isoformspecific receptor for vascular endothelial growth factor. Cell 92(6):735–745. https:// doi.org/10.1016/s0092-8674(00)81402-6 8. Gluzman-Poltorak Z, Cohen T, Herzog Y, Neufeld G (2000) Neuropilin-2 is a receptor for the vascular endothelial growth factor (VEGF) forms VEGF-145 and VEGF-165 [corrected]. J Biol Chem 275(24): 18040–18045. https://doi.org/10.1074/jbc. M909259199 9. Zhao W, McCallum SA, Xiao Z, Zhang F, Linhardt RJ (2012) Binding affinities of vascular endothelial growth factor (VEGF) for heparinderived oligosaccharides. Biosci Rep 32(1): 7 1 – 8 1 . h t t p s : // d o i . o r g / 1 0 . 1 0 4 2 / BSR20110077 10. Tamma R, Annese T, Ruggieri S, Marzullo A, Nico B, Ribatti D (2018) VEGFA and VEGFR2 RNAscope determination in gastric cancer. J Mol Histol 49(4):429–435. https:// doi.org/10.1007/s10735-018-9777-0 11. Bais C, Mueller B, Brady MF, Mannel RS, Burger RA, Wei W, Marien KM, Kockx MM, Husain A, Birrer MJ, Group NRGOGO (2017) Tumor microvessel density as a potential predictive marker for bevacizumab benefit: GOG-0218 biomarker analyses. J Natl Cancer Inst 109(11):djx066. https://doi.org/10. 1093/jnci/djx066 12. Scartozzi M, Loretelli C, Galizia E, Mandolesi A, Pistelli M, Bittoni A, Giampieri R, Faloppi L, Bianconi M, Del Prete M, Bianchi F, Belvederesi L, Bearzi I, Cascinu S (2012) Role of vascular endothelial growth factor (VEGF) and VEGF-R

genotyping in guiding the metastatic process in pT4a resected gastric cancer patients. PLoS One 7(7):e38192. https://doi.org/10.1371/ journal.pone.0038192 13. Sopo M, Anttila M, Muukkonen OT, Yl AHS, Kosma VM, Keski-Nisula L, Sallinen H (2020) Microvessels in epithelial ovarian tumors: high microvessel density is a significant feature of malignant ovarian tumors. Anticancer Res 40(12):6923–6931. https://doi.org/10. 21873/anticanres.14716 14. Nico B, Crivellato E, Guidolin D, Annese T, Longo V, Finato N, Vacca A, Ribatti D (2010) Intussusceptive microvascular growth in human glioma. Clin Exp Med 10(2):93–98. https://doi.org/10.1007/s10238-0090076-7 15. Ferrara N (2004) Vascular endothelial growth factor: basic science and clinical progress. Endocr Rev 25(4):581–611. https://doi.org/ 10.1210/er.2003-0027 16. Claesson-Welsh L, Welsh M (2013) VEGFA and tumour angiogenesis. J Intern Med 273(2):114–127. https://doi.org/10.1111/ joim.12019 17. Zirlik K, Duyster J (2018) Anti-Angiogenics: current situation and future perspectives. Oncol Res Treat 41(4):166–171. https://doi. org/10.1159/000488087 18. Pezzella F, Gatter KC (2016) Evidence showing that tumors can grow without angiogenesis and can switch between Angiogenic and Nonangiogenic phenotypes. J Natl Cancer Inst 108(8):djw032. https://doi.org/10. 1093/jnci/djw032 19. Kuczynski EA, Vermeulen PB, Pezzella F, Kerbel RS, Reynolds AR (2019) Vessel co-option in cancer. Nat Rev Clin Oncol 16(8):469–493. https://doi.org/10.1038/s41571-0190181-9 20. Caporarello N, Lupo G, Olivieri M, Cristaldi M, Cambria MT, Salmeri M, Anfuso CD (2017) Classical VEGF, notch and Ang signalling in cancer angiogenesis, alternative approaches and future directions (review). Mol Med Rep 16(4):4393–4402. https://doi. org/10.3892/mmr.2017.7179 21. D’Alessandris QG, Martini M, Cenci T, Capo G, Ricci-Vitiani L, Larocca LM, Pallini R (2015) VEGF isoforms as outcome biomarker for anti-angiogenic therapy in recurrent glioblastoma. Neurology 84(18):1906–1908. h t t p s : // d o i . o r g / 1 0 . 1 2 1 2 / W N L . 0000000000001543 22. Strimbu K, Tavel JA (2010) What are biomarkers? Curr Opin HIV AIDS 5(6):463–466.

RNAscope in Human Tumor Biopsies h t t p s : // d o i . o r g / 1 0 . 1 0 9 7 / C O H . 0b013e32833ed177 23. Hamburg MA, Collins FS (2010) The path to personalized medicine. N Engl J Med 363(4): 3 0 1 – 3 0 4 . h t t p s : // d o i . o r g / 1 0 . 1 0 5 6 / NEJMp1006304 24. Mayeux R (2004) Biomarkers: potential uses and limitations. NeuroRx 1(2):182–188. https://doi.org/10.1602/neurorx.1.2.182 25. Bhattacharya S, Mariani TJ (2009) Array of hope: expression profiling identifies disease biomarkers and mechanism. Biochem Soc Trans 37(Pt 4):855–862. https://doi.org/10. 1042/BST0370855 26. Xi X, Li T, Huang Y, Sun J, Zhu Y, Yang Y, Lu ZJ (2017) RNA biomarkers: frontier of precision medicine for cancer. Noncoding RNA 3 ( 1 ) : 9 . h t t p s : // d o i . o r g / 1 0 . 3 3 9 0 / ncrna3010009 27. Yang YC, Di C, Hu B, Zhou M, Liu Y, Song N, Li Y, Umetsu J, Lu ZJ (2015) CLIPdb: a CLIP-seq database for protein-RNA interactions. BMC Genomics 16:51. https://doi. org/10.1186/s12864-015-1273-2 28. Hu B, Yang YT, Huang Y, Zhu Y, Lu ZJ (2017) POSTAR: a platform for exploring post-transcriptional regulation coordinated by RNA-binding proteins. Nucleic Acids Res 45 (D1):D104–D114. https://doi.org/10. 1093/nar/gkw888 29. Wong ML, Medrano JF (2005) Real-time PCR for mRNA quantitation. BioTechniques 39(1): 7 5 – 8 5 . h t t p s : // d o i . o r g / 1 0 . 2 1 4 4 / 05391RV01 30. Belleri M, Paganini G, Coltrini D, Ronca R, Zizioli D, Corsini M, Barbieri A, Grillo E, Calza S, Bresciani R, Maiorano E, Mastropasqua MG, Annese T, Giacomini A, Ribatti D, Casas J, Levade T, Fabrias G, Presta M (2020) Beta-galactosylceramidase promotes melanoma growth via modulation of ceramide metabolism. Cancer Res 80(22):5011–5023. https://doi.org/10.1158/0008-5472.CAN19-3382

155

31. Tamma R, Ingravallo G, Gaudio F, Annese T, Albano F, Ruggieri S, Dicataldo M, Maiorano E, Specchia G, Ribatti D (2020) STAT3, tumor microenvironment, and microvessel density in diffuse large B cell lymphomas. Leuk Lymphoma 61(3):567–574. https://doi. org/10.1080/10428194.2019.1678154 32. Wang F, Flanagan J, Su N, Wang LC, Bui S, Nielson A, Wu X, Vo HT, Ma XJ, Luo Y (2012) RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14(1):22–29. https://doi. org/10.1016/j.jmoldx.2011.08.002 33. Annese T, Tamma R, De Giorgis M, Ruggieri S, Maiorano E, Specchia G, Ribatti D (2020) RNAscope dual ISH-IHC technology to study angiogenesis in diffuse large B-cell lymphomas. Histochem Cell Biol 153(3): 185–192. https://doi.org/10.1007/s00418019-01834-z 34. Siddiqui GK, Maclean AB, Elmasry K, Wong te Fong A, Morris RW, Rashid M, Begent RH, Boxer GM (2011) Immunohistochemical expression of VEGF predicts response to platinum based chemotherapy in patients with epithelial ovarian cancer. Angiogenesis 14(2): 155–161. https://doi.org/10.1007/s10456010-9199-4 35. Bendardaf R, El-Serafi A, Syrjanen K, Collan Y, Pyrhonen S (2017) The effect of vascular endothelial growth factor-1 expression on survival of advanced colorectal cancer patients. Libyan J Med 12(1):1290741. https://doi.org/10. 1080/19932820.2017.1290741 36. Girolamo F, Errede M, Longo G, Annese T, Alias C, Ferrara G, Morando S, Trojano M, Kerlero de Rosbo N, Uccelli A, Virgintino D (2019) Defining the role of NG2-expressing cells in experimental models of multiple sclerosis. A biofunctional analysis of the neurovascular unit in wild type and NG2 null mice. PLoS One 14(3):e0213508. https://doi.org/10. 1371/journal.pone.0213508

Chapter 11 Global MicroRNA Profiling of Vascular Endothelial Cells Eloi Schmauch, Anna-Liisa Levonen, and Suvi Linna-Kuosmanen Abstract MicroRNA sequencing (miRNA-seq) enables the detection and characterization of the cell miRNome, including miRNA isoforms (isomiRs) and novel miRNA species. In roughly half of the cases, the most abundant isomiR in the cells is not the reference miRNA given in miRBase, which highlights the importance of isomiR-specific analysis. Here, we describe a gel-free protocol for global miRNA profiling in vascular endothelial cells and the main steps of the subsequent data analysis with two alternative analysis methods. In addition to endothelial cells, the protocol is suitable for other cell and tissue types and has been successfully used to obtain miRNA-seq data from human cardiac tissue, plasma, pericardial fluid, and biofluid exosomes. Key words microRNA, miRNA-seq, Endothelial cell, isomiR, Tissue, Biofluid, Exosome

1

Introduction MicroRNA sequencing (miRNA-seq) offers an advanced research approach for both high-throughput miRNA expression analysis and novel miRNA discovery. Typical workflow starts from RNA extraction, followed by library preparation (adapter ligation, reverse transcription, library amplification), and concludes with sequencing and data analysis (Fig. 1) [1]. miRNA-seq has facilitated the discovery of novel miRNA species and miRNA isoforms (isomiRs) and shed light into context-specific miRNA expression [2, 3]. However, bias and contaminations introduced during library preparation can hamper the accuracy of the measured sequences and their expression values. This sequencing bias is protocol-dependent, and its primary cause is adapter ligation, complemented with adapter dimer formation, and gel-based library size selection [4–7]. Given the importance of data accuracy and quality on downstream data analysis, it is essential to choose RNA extraction and library preparation methods carefully and ensure proper handling of RNA and library samples at all times. The protocol presented here utilizes several parallel strategies to facilitate

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_11, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

157

158

Eloi Schmauch et al.

Fig. 1 Workflow for miRNA-seq library preparation. Adapters are ligated to the 30 and 50 ends of miRNAs, utilizing the hydroxyl and phosphate groups mature miRNAs possess, and thus minimizing the background from other RNA species. The ligated miRNAs are reverse transcribed to cDNA using primers with integrated Unique Molecular Identifiers (UMIs). During the process, a universal sequence that is later used in the index assignment is also added. No libraries are prepared from adapter-dimers. After reverse transcription the library is cleaned using a magnetic bead-based method. The library is amplified using a universal forward primer and indexing reverse primers. After final cleanup, the library goes through quality check, and sequencing. Figure is created with BioRender.com

preparation of robust and accurate miRNA-specific libraries: The protocol (a) utilizes miRNA-specific adapter ligation to eliminate biases and background contaminants; (b) uses modified oligonucleotides to prevent adapter-dimerization; (c) does not require size selection with gel extraction; (d) inserts Unique Molecular Identifiers (UMIs) to enable unbiased and accurate quantification of

Global MicroRNA Profiling of Vascular Endothelial Cells

159

Fig. 2 Posttranscriptional modifications of miRNAs give rise to an important diversity of miRNA reads. (a) miRNA processing consists of several steps. First, a miRNA gene is transcribed into a primary miRNA (pri-miRNA), which is then cleaved by Drosha into a pre-miRNA. Further cleavage and modifications give finally rise to the mature miRNA(s). Variations of the process and modifications of the mature form create isomiRs. (b) The miRNome is composed of a very diverse set of isomiRs. With respect to the canonical form, modifications can be additions or deletions in the 30 (30 isomiR) or 50 (50 isomiR) strand, or nucleotide substitutions within the sequence (polymorphic isomiR). Additions can be either templated (same nucleotide as the pre-miRNA, highlighted in orange) or non-templated (highlighted in blue). (c) Due to high sequence diversity resulting from isomiRs, the choice of data analysis method affects the obtained results. IsomiR analysis results in a detailed and complex view while canonical analysis yields simpler results, which are easier to grasp but may potentially hide important information and lead to misinterpretation of the results

different mature miRNAs; and (e) contains internal controls to allow fast and efficient troubleshooting, in case quality control fails. Numerous methods have been established for the processing and analysis of miRNA-seq data [8–15]. However, the consensus as to which exact methods and parameters to use has not been established, especially for isomiRs [16, 17]. While miRNA-seq processing is similar to mRNA-seq processing, key differences exist: (a) miRNA size introduces new biases, which are corrected in mRNA sequencing as the long transcripts are randomly sampled and aligned, (b) read duplicates are expected and represent the biological reality, and (c) the established existence and significance of isomiRs (Fig. 2) raises the question of how to identify them [18– 22]. The first step of miRNA-seq data analysis is preprocessing that includes adapter removal and quality filtering. Known, novel miRNA species and isomiRs are then identified using alignmentbased methods, and finally, model-based processing of miRNA expression is used for differential expression analysis to identify

160

Eloi Schmauch et al.

Fig. 3 Overview of the data analysis pipelines. The first steps consist of preprocessing, selection of the reference genome and miRNA annotation. Then, two main analysis methods can be used. In the miRBasecentric, canonical analysis (Method 1), the user can either discover novel miRNAs or map known miRNA species. For the isomiR analysis (Method 2), two interchangeable approaches are shown: fast isomiR identification with MiRGE and comprehensive analysis that identifies ambiguous isomiRs with Prost!. In all of these approaches, the user obtains a miRNA species count matrix that can then be inputted to DESeq2 for differential expression analysis

condition-specific trends. A simple straightforward method for detecting expression is to map reads corresponding to known canonical reference miRNAs in miRBase. This method assembles subsets of the isoform reads for a given miRNA (depending on the allowed mismatches) and focuses on known miRNA species. Novel miRNAs can then be discovered and characterized with a separate analysis step. Another method for detecting miRNA expression is isomiR-specific analysis that yields the most complete and accurate description of the studied miRNome, at the cost of more complex results. Here, we provide protocols for both miRBase-centric and isomiR-specific data analysis with two different options described for isomiR identification (Fig. 3). We also describe model-based processing of RNA expression for differential expression analysis to identify condition-specific trends.

Global MicroRNA Profiling of Vascular Endothelial Cells

2

161

Materials This protocol utilizes Qiaseq miRNA library preparation kit (Qiagen) that is meant for Illumina Next Generation Sequencing (NGS) systems (see Note 1).

2.1 Endothelial Cell Isolation/Culturing

HBSS/PBS (must contain Calcium), store at room temperature (RT). Collagenase (type IV, 3 mg/10 mL), store at 4  C or 20  C. Fibronectin (10 mg/mL) and 0.0 5% gelatin in PBS, store at 4  C. Fetal bovine serum (FBS), store at 4  C (temporarily). EGM growth medium: EBM Basal Medium (Lonza), store at 4  C (temporarily). EGM SingleQuots Supplements (Lonza), store at 20  C. NALGENE Cryo 1  C Freezing Container or similar. FBS + 10% dimethyl sulfoxide (DMSO). Freezing tubes.

2.2 miRNA Library Preparation

QIAseq miRNA Library Kit (Qiagen). Agilent High Sensitivity DNA Kit (Agilent, recommended) OR equipment and consumables to run a 6% polyacrylamide gel electrophoresis (PAGE), Tris Borate EDTA (TBE) gel. Qubit dsDNA HS Assay Kit and Qubit Assay Tubes (Thermo Fisher Scientific). Illumina NGS Sequencer.

2.3 Gel Size Selection (Only if Library QC Fails, See Subheading 3.8)

6% PAGE TBE gel. 5 GelPilot DNA Loading Dye or similar. 25 bp DNA Ladder. SYBR Gold Nucleic Acid Gel Stain. Gel Breaker Tubes. 3 M Sodium Acetate, pH 5.2 Linear Acrylamide.

2.4 For Troubleshooting with qPCR (Only if Library QC Fails, See Subheading 3.8)

miScript SYBR Green PCR Kit (2 QuantiTect® SYBR® Green PCR Master. Mix) or similar.

162

3

Eloi Schmauch et al.

Methods

3.1 Endothelial Cell Isolation (See Note 2)

1. Prepare collagenase (3 mg/10 mL sterile PBS); 10–15 mL per umbilical cord. 2. Coat T25 flasks with fibronectin/gelatin and incubate for 10 min at 37  C. Remove excess fibronectin/gelatin prior to use. Prepare 1–2 flasks per umbilical cord. 3. Place a sterile PBS bottle in a 37  C water bath for collagenase incubation. 4. Prepare EGM with 10% FBS, and HBSS/PBS with 10% FBS. For each cord, prepare 10 mL of both (especially long or thick cords may require more). 5. On a drape in a laminar flow cabinet, wipe the umbilical cord clean. 6. Cut one end of the cord and gently milk/massage blood out of the vein. 7. Insert a feeding tube about 5 cm into the vein (largest of the three vessels) and tie strings around the end to hold the tube firmly in place. 8. Cut the other end of the cord and rinse with PBS through the feeding tube using a syringe until PBS rinse appears clear of blood. Remove excess PBS by massaging. Tie the end tightly (see Note 3). 9. Place a 0.2 μm filter on the syringe and fill the vein with filtered collagenase solution through the feeding tube. Avoid pushing air into the vein (see Note 4). 10. Place the cord into a centrifuge bottle filled with pre-warmed PBS. Make sure the cord is completely submerged in the PBS. Incubate the bottle for 10 min at 37  C in a water bath. Add 5 mL of growth medium +10% FBS to a 50 mL tube (1 tube per cord) (see Note 5). 11. Remove the collagenase from the vein using a clean syringe and wash the vein with PBS. Massage the cord while doing this. Collect both collagenase solution and PBS rinse into the 50 mL tube that has 5 mL of growth medium. 12. Centrifuge for 10 min at 300–500  g. Remove supernatant and resuspend the cells to 5 mL growth medium +10% FBS, then place the suspension in the coated T25 flask(s). 13. The next day, remove media and wash the cells with PBS. 14. Add 5 mL fresh growth medium with supplements and 10% FBS. 15. Change medium every 2 days. Use growth medium with normal supplements (no FBS excess).

Global MicroRNA Profiling of Vascular Endothelial Cells

163

16. When cells are confluent, passage into three T25 or one T75. 17. Change fresh EGM every 2 days until cells are confluent. 18. Proceed to Subheading 3.2 or store cells in liquid nitrogen following steps 19–25. 19. Add 250 mL isopropanol into NALGENE Cryo 1  C Freezing Container. 20. Remove media from cells and wash with PBS. 21. Cover the bottom of the flask with trypsin (about 2 mL for T75), incubate for no longer than 1–2 min at 37  C and add 4 mL EGM to deactivate trypsin (see Note 6). 22. Collect cells immediately into a centrifuge tube and centrifuge cells at 300–500  g for 5 min. 23. Remove supernatant, resuspend cells with FBS + 10% DMSO (1 mL/T25, 3 mL/T75), and divide into freezing tubes (1 mL/tube). 24. Place the tubes in the tube wells of Freezing Container and leave the Container in 70  C overnight. 25. Transfer the tubes to liquid nitrogen for long-term storage. 3.2 RNA Extraction (See Note 7)

1. Seed 180,000 HUVECs/well in 6-well plates precoated with fibronectin/gelatin for 10 min at 37  C, and culture in EGM medium (2 mL/well) until confluent. 2. Extract RNA using any method compatible with downstream library preparation (see Note 8). 3. Determine RNA concentration and purity with a spectrophotometer. Pure RNA has an A260:A280 ratio of 1.9–2.1 in 10 mM Tris–HCl, pH 7.5 (see Note 9). 4. Assess the RNA Integrity Number (RIN) using Agilent 2100 Bioanalyzer according to manufacturers’ instructions. The RIN should ideally be equal to or greater than 8, but the library preparation is possible from lower RIN values (see Note 10).

3.3 Library Preparation: 30 Ligation

1. From the QIAseq miRNA Library Kit, thaw 30 Adapter, 30 Buffer, 2 Ligation Activator and nuclease-free water at room temperature (RT). Mix by flicking the tubes, centrifuge briefly and keep at RT. 2. Thaw RNA on ice, mix gently and centrifuge briefly to collect all liquid from the sides of the tubes to the bottom. Return sample(s) to ice. 3. Dilute 30 Adapter in nuclease-free water if needed, depending on the total RNA yield (RNA input):

164

Eloi Schmauch et al.

100 or 500 ng

Use undiluted

10 ng

Dilute 1:5

1 ng

Dilute 1:10

Biofluid/exosome

Dilute 1:5

Centrifuge briefly, mix by pipetting and centrifuge briefly again. Do not vortex. 4. Prepare 30 ligation master mix on ice adding components in the listed order (see Note 11): 30 Adapter

1 μL Adapter sequence: AACTGTAGGCACCATCAAT

RI

1 μL

0

1 μL

0

3 Buffer

2 μL

2 ligation activator

10 μL

Total volume

15 μL

3 Ligase

Centrifuge briefly, mix the viscous reaction slowly by pipetting (at least 15–20 times), and centrifuge again. Do not vortex. Correct preparation of the ligation is crucial for success. Prepare 10% excess when preparing master mix for more than 1 sample. 5. Add RNA to 30 ligation master mix. Centrifuge briefly, mix the viscous reaction by slowly pipetting 15–20 times and centrifuge briefly again. Do not vortex. 6. Incubate (in a thermocycler): 28  C for 1 h 65  C for 20 min 4  C hold at least 5 min (important). While incubating, thaw 50 ligation reagents (50 Adapter and 5 Buffer) at RT. Mix by flicking the tube and centrifuge briefly. Keep at RT until needed. 0

7. Proceed immediately to 50 ligation using the entire 20 μL of 30 ligation as the starting material.

Global MicroRNA Profiling of Vascular Endothelial Cells

3.4

165

1. Dilute 50 Adapter if needed, depending on the RNA input:

50 Ligation

100 or 500 ng

Use undiluted

10 ng

Dilute 1:2.5

1 ng

Dilute 1:5

Biofluid/exosome

Dilute 1:2.5

Briefly centrifuge, mix by pipetting (12 times), and centrifuge briefly again. 2. Prepare 50 ligation reaction on ice by adding components directly to the 30 ligation reaction tube. Add the components in the listed order: 30 ligation reaction (already in the tube)

20 μL

Nuclease-free water

15 μL

50 Buffer

2 μL

RI

1 μL

0

1 μL

0

5 Adapter

1 μLa

Total volume

40 μL

5 Ligase

a

Adapter Sequence: GTTCAGAGTTCTACAGTCCGACGATC

Briefly centrifuge, mix the viscous reaction by pipetting slowly up and down (at least 15–20 times), and centrifuge briefly again. Do not vortex. Correct preparation of the ligation is crucial for success. Prepare 10% excess when preparing master mix for more than 1 sample. 3. Incubate: 28  C for 30 min. 65  C for 20 min 4  C hold at least 5 min. While incubating, thaw reverse transcription reagents (RT initiator, RT Buffer, RT Primer) at RT. Mix by flicking the tube and centrifuge briefly. Keep at RT until needed. 4. Proceed immediately to Reverse Transcription using the entire 40 μL of 50 ligation as the starting material.

166

Eloi Schmauch et al.

3.5 Reverse Transcription

1. Add 2 μL of RT Initiator to 50 ligation reaction (40 μL). Centrifuge briefly, mix by pipetting 15–20 times and centrifuge again. 2. Incubate (in a thermocycler): 75  C 2 min 70  C 2 min 65  C 2 min 60  C 2 min 55  C 2 min 37  C 5 min 25  C 5 min 4  C hold at least 5 min (important) or until setup of the reverse transcription reaction. 3. Dilute RT Primer if needed, depending on the RNA input: 100 or 500 ng

Use undiluted

10 ng

Dilute 1:5

1 ng

Dilute 1:10

Biofluid/exosome

Dilute 1:5

Briefly centrifuge, mix by pipetting (12 times), and centrifuge briefly again. 4. Prepare reverse transcription reaction by adding components directly to the 50 ligation reaction tube on ice. Add the components in the listed order: 50 ligation reaction + RT initiator (already in the tube)

42 μL

RT primer

2 μL

Nuclease-free water

2 μL

RT buffer

12 μL

RI

1 μL

RT enzyme

1 μL

Total volume

60 μL

Briefly centrifuge, mix the reaction by pipetting slowly up and down 12 times, and centrifuge briefly again. Do not vortex. Prepare 10% excess when preparing master mix for more than 1 sample.

Global MicroRNA Profiling of Vascular Endothelial Cells

167

5. Incubate (in a thermocycler): 50  C for 1 h 70  C for 15 min 4  C hold at least 5 min (important). While incubating, prepare beads for cDNA cleanup following protocol in Subheading 3.6. 6. Proceed to cDNA Cleanup using the entire 60 μL of reverse transcription reaction as the starting material for the cleanup. 3.6 Preparation of Beads

1. Vortex Beads and Bead Binding Buffer thoroughly to ensure homogenous suspension. Do not centrifuge (see Note 12). 2. Add 400 μL of beads to 2 mL tube, which is enough for both “cDNA cleanup” and “library cleanup” for one sample. Centrifuge briefly and immediately separate beads on a magnet stand (see Note 13). 3. When beads have fully separated, remove and discard supernatant. A small amount of supernatant may remain in the tube. 4. Remove the tube from the magnet stand, and carefully pipet 150 μL of Bead Binding Buffer onto the beads. Vortex thoroughly to resuspend the beads completely. Centrifuge briefly and place immediately to magnet stand. 5. When beads have fully separated, remove as much supernatant as possible. Discard supernatant. 6. Remove tube from the magnetic stand, and carefully pipet 400 μL of Bead Binding Buffer onto the beads. Vortex thoroughly to resuspend the beads completely. After preparation, the beads need to be placed on ice. Beads may be stored at 4  C for up to 1 week, if not used immediately.

3.7

cDNA Cleanup

1. Prepare fresh 80% ethanol using nuclease-free water. 2. Centrifuge the cDNA tube(s) and mix the beads thoroughly. 3. Add 143 μL of beads to each cDNA tube. Vortex 3 s and centrifuge briefly (See Note 14). 4. Incubate for 5 min at RT. 5. Separate the beads on a magnet stand (4 min). If the supernatant is not completely clear after 6 min, proceed with the cleanup. 6. Discard the supernatant and continue with the beads on the magnet stand. 7. Add 200 μL of 80% ethanol, remove it immediately and discard the ethanol wash.

168

Eloi Schmauch et al.

8. Repeat the wash (step 7). Remove all traces of ethanol after the second wash: Centrifuge, return the tubes to the magnet stand and use 10 μL pipette to remove any remaining ethanol. 9. Leave the tubes on the magnet stand and air-dry at RT for 10 min (see Note 15). 10. Remove the tubes from the stand, add 17 μL nuclease-free water to each tube to elute DNA. 11. Resuspend by pipetting or vortexing. Incubate 2 min at RT and centrifuge briefly. 12. Separate beads on magnetic stand (2 min). Ensure that the beads have fully separated before proceeding. If the supernatant is not completely clear after 6 min, proceed with the protocol. 13. Transfer 15 μL of eluted DNA to new tube. DNA can be stored at 20  C in a constant-temperature freezer, if protocol is not continued immediately. The entire 15 μL eluate is the starting material for the library amplification. 3.8 Library Amplification

1. Choose index primer for every sample and note the index in a summary table. 2. Thaw Library Buffer, Forward Primer and Index Primer(s) at RT. Mix by flicking the tubes and centrifuge briefly. 3. Prepare the library amplification reaction on ice: cDNA eluate

15 μL

Library buffer

16 μL

DNA polymerase

3 μL

Forward primer

2 μL

Index primer

2 μL

Nuclease-free water

42 μL

Total volume

80 μL

Briefly centrifuge, pipet slowly to mix (15–20 times), and centrifuge briefly again. Do not vortex. Prepare 10% excess when preparing master mix for more than 1 sample.

Global MicroRNA Profiling of Vascular Endothelial Cells

169

4. Place the tube in a thermal cycler and run the following program: 15 min

95  C

Denaturation

15 s

95  C

Annealing

30 s

60  C

Extension

15 s

72  C

Hold Cyclea:

a

Number of cycles is based on original RNA input

Hold

2 min

72  C

Hold

>5 min

4  C (important)

Original RNA input, total RNA

Cycle #

500 ng

13

100 ng

16

10 ng

19

1 ng

22

Serum/plasma

22

5. Mix the beads (from Subheading 3.6) thoroughly. Add 75 μL of beads to a fresh tube. One tube per amplification reaction. As before, ensure that the beads are thoroughly suspended at all times: Work quickly and resuspend beads immediately before use. 6. Centrifuge briefly the 80 μL amplification reaction and transfer 75 μL to the tube containing the beads. Vortex for 3 s and centrifuge briefly. 7. Incubate 5 min at RT. 8. Separate beads on a magnet stand (4 min). Ensure that the supernatant is clear before proceeding. If the supernatant is not completely clear after 6 min, proceed with the cleanup. 9. Transfer 145 μL of the supernatant to a new tube (use LoBind Tubes) and discard the tube containing the beads. Continue with the supernatant. 10. Add 130 μL beads to the supernatant. Vortex 3 s and centrifuge briefly. 11. Incubate 5 min at RT. 12. Separate beads on a magnet stand (4 min). Ensure that the supernatant is clear before proceeding. If the supernatant is not completely clear after 6 min, proceed with the cleanup.

170

Eloi Schmauch et al.

13. Discard the supernatant and keep the beads. Do not remove the tube(s) from the magnet stand. 14. Add 200 μL of 80% ethanol. Turn the tube(s) on the magnetic stand to move the beads. Remove and discard the ethanol wash immediately. 15. Repeat the ethanol wash. Remove all ethanol after the second wash. Centrifuge the tube(s) briefly and return to magnet stand. Remove any residual ethanol with 10 μL pipette. 16. Air-dry the beads for 10 min at RT on the magnet stand. As before, ensure that all residual ethanol has evaporated, and the pellet is completely dry before proceeding to next step. Humidity may extend the drying time. 17. Remove the tube(s) from the magnet stand, add 17 μL of nuclease-free water to the tube(s) to elute DNA. 18. Pipet or vortex carefully to resuspend the beads. Incubate 2 min at RT. Centrifuge briefly. 19. Separate beads on a magnet stand (2 min). Ensure that the beads have fully separated before proceeding. If the supernatant is not completely clear after 6 min, proceed with the protocol. 20. Transfer 15 μL of the DNA eluate (i.e., miRNA sequencing library) to a new tube. 21. Proceed to next step or store the library at 20  C in a constant-temperature freezer. 3.9 miRNA Library Pre-sequencing QC

This protocol gives two options for library QC: Agilent Bioanalyzer 2100 and polyacrylamide gel electrophoresis. Option 1: Agilent Bioanalyzer 2100 1. Use 1 μL of the miRNA sequencing library to analyze the sample using a High Sensitivity DNA chip on Agilent bioanalyzer following manufacturer’s instructions. The library size is expected to be approximately 180 bp. Option 2: Polyacrylamide gel electrophoresis 1. Prepare a 6% PAGE TBE gel, use a 25 bp ladder and load 3 μL of the miRNA sequencing library on the gel. 2. Run the gel at 120 V for 1 h or until the dye front has reached the bottom. 3. The library size should be approximately 173 bp. If no library is detected, assess the integrated reaction controls using real-time PCR to determine if there is a technical or a sample issue (Subheading 3.10). If an additional large peak is observed at approximately 157 bp for bioanalyzer and 150 for PAGE gel (adapter-primer) or other

Global MicroRNA Profiling of Vascular Endothelial Cells

171

large bands are noted, perform gel excision on the remaining library to select the library of interest (Subheading 3.11). If no significant undesired bands are observed, proceed to determining library concentration and read allocation (Subheading 3.12) (see Note 16). 3.10 Real-Time PCR Troubleshooting

The library kit contains integrated reaction controls to monitor the critical steps of the workflow. If Library QC (Subheading 3.9) is unsuccessful, these controls help to locate the step where the library preparation failed (technical issue) or if the problem was in the sample itself. The protocol given is for Master mix of the miScript SYBR Green PCR Kit. If other master mix is used, check recommended PCR program and adjust accordingly. 1. Thaw PCR Master Mix, miC30 (assessment of 30 ligation performance), miC50 (assessment of 50 ligation performance), and miCRT (assessment of reverse transcription performance) Primer Assays at RT. Mix by flicking the tubes (no vortex), centrifuge briefly, and keep at RT. Do not use Universal Primer as the individual assays contain both forward and reverse primers. 2. Dilute 1 μL of sequencing library (from Subheading 3.8) as indicated: Number of library amplification cycles

Dilution

13

1 μL + 4 μL water

16

1 μL + 49 μL water

19

1 μL + 499 μL water

22

Step 1: 1 μL + 49 μL water step 2: 1 μL of step 1 + 99 μL water ! use for qPCR

Mix by flicking the tube (no vortex), centrifuge briefly, and keep at RT. 3. Prepare a master mix for each sample: 96well

Rotor-Disc 100

2 SYBR Green PCR master mix 20 μL

50 μL

40 μL

Nuclease-free water

12 μL

36 μL

28 μL

Diluted library amplification

4 μL

4 μL

4 μL

Total volume

36 μL

90 μL

72 μL

Component

384well

Mix gently but thoroughly.

172

Eloi Schmauch et al.

4. Pipet master mix into three individual wells for each sample: 9 μL for 384-well plates, 22.5 μL for 96-well plates, 18 for Rotor-Disc 100. 5. Pipet one of the three control primers into one of each sample’s three wells (i.e., 1 well per control primer assay): 1 μL for 384-well plates, 2.5 μL for 96-well plates, 2 μL for RotorDisc 100. 6. Seal the plate or disc tightly. 7. Centrifuge 384/96-well plate for 1 min at 1000  g at RT to remove bubbles. 8. Run RT-PCR program (melt curve analysis not required): 15 min

95  C

Denaturation

15 s

94  C

Annealing

30 s

55  C

Extensiona

30 s

70  C fluorescence data collection

Activation Cycling (35 cycles)a

a

Adjust ramp rate, number of cycles, and fluorescence detection step according to the specific requirements of the machine and software used

9. Analyze the data: Set baseline: Use the “Linear View” of the amplification plot to determine the earliest visible amplification. Set the baseline from cycle 2 to two cycles before the earliest visible amplification. Reduce the number of cycles used to calculate the baseline if template amount is high. Define threshold: Use a logarithmic amplification plot to set the threshold so that the log-linear range of the curve can be easily identified. Using the “Log View” of the amplification plot, place the threshold above the background signal but within the lower half of the log-linear range of the amplification plot. The threshold should never be set in the plateau phase. The absolute position of the threshold is less critical than its consistent position across PCR runs. Ensure that the baseline and threshold settings are the same across all runs associated with the same experiment to allow comparison of results. 10. Export the Ct values and interpret the results: If all Ct values are less than 28, the individual reaction steps have been performed correctly. This might mean that the problem was the sample itself. If the Ct values for some or all are greater than 28, either the respective step of library preparation has failed, or the sample has been compromised. Ensure that the RT-PCR was performed correctly.

Global MicroRNA Profiling of Vascular Endothelial Cells

3.11 Size Selection of Library

173

1. Prepare a 6% PAGE TBE gel. 2. Take the remaining Library sample (from Subheading 3.8) and adjust the volume of the library to 24 μL using nuclease-free water. 3. Add 6 μL of 5 Loading dye and mix thoroughly. 4. Distribute the mixture across three lanes of the 6% PAGE TBE gel. 5. Run at 120 V until the dye has reached the bottom of the cassette (for about 1 h). 6. Remove the gel from the cassette and stain with 1x SYBR Gold for 10 min. 7. Excise the correct sized band containing the library of choice. For miRNAs it is approximately 173 bp. It is important to cut carefully and transfer the piece without touching other parts of the gel. 8. Place each excised band in a 0.5 mL Gel Breaker tube in a 2 mL tube and centrifuge 2 min at max speed. 9. Soak the debris in 250 μL of 0.3 M Sodium Acetate. 10. Rotate for at least 2 h in RT. 11. Transfer eluate and gel debris to a Corning Costar Spin-X Centrifuge Tube Filter column and centrifuge for 2 min at max speed. 12. Recover eluate and add 1 μL of Linear Acrylamide and 750 μL of 100% ethanol. 13. Vortex and incubate at 80  C for at least 1 h. 14. Centrifuge for 30 min at 14,000  g at 4  C. 15. Remove supernatant carefully without disturbing the pellet. 16. Wash the pellet with 500 μL of freshly made 80% ethanol. Use nuclease-free water. 17. Centrifuge for 30 min at 14,000  g at 4  C. 18. Remove alcohol and air-dry the pellet for 10 min at 37  C. 19. Resuspend pellet in 15 μL nuclease-free water. 20. Proceed to next step or store the library at 20  C in a constant-temperature freezer.

3.12 Library Concentration and Read Allocation

1. Keep the Library on ice. 2. It is recommended to use a Qubit Fluorimeter to determine the library concentration for more accurate results. Use 2 μL of the Library sample on a Qubit Fluorimeter according to the manufacturer’s instructions. 3. Determine the molarity of each sample in nM using the following equation meant for miRNA library:

174

Eloi Schmauch et al.


ðX ng=μLÞ 106 =ð112, 450Þ ¼ Y nM 4. Dilute Libraries to 4 nM using nuclease-free water. If library is of good quality, but low concentration, 2 nM can be used. 5. For multiplexing, combine libraries in equimolar amounts and mix well. Allocate 5–10 million reads per sample. 6. Proceed to next step or store the library at 20  C in a constant-temperature freezer. 3.13

Sequencing

The prepared libraries can be sequenced using an Illumina system (MiSeq® Personal Sequencer, NextSeq 500/550, HiSeq® 1000, HiSeq 1500, HiSeq 2000, HiSeq 2500, HiSeq 4000, NovaSeq™ 6000, and GAIIx). 1. For instructions to sample dilution and pooling, refer to “Standard Normalization Method” protocol in the system-specific Illumina document (i.e., MiSeq/NextSeq/etc.: Denature and Dilute Libraries Guide). 2. Load libraries and set up the sequencing run using systemspecific Illumina document (i.e., MiSeq/NextSeq/etc.: “System User Guide”). Recommended final library concentration is 10 pM for MiSeq and 1.2 pM for NextSeq. 3. Set up sequencing run by selecting “FASTQ Only” and choose “TruSeq Small RNA” from the Sample Prep Kit dropdown menu. Recommended protocol is 75 bp single read. A 50 bp single read can also be used, however, that will exclude UMIs.

3.14

Data Analysis

3.14.1

Preprocessing

1. If the input data are BCL files, use the bcl2fastq command to convert them to fastq files. Else, skip this step (see Note 17). 2. If the miRNA-seq library includes UMIs, extract them and remove reads containing UMI duplicates (see Note 18). 3. Remove adapters from the reads using fastx_clipper, use the -a argument to provide the adapter sequence (see Note 19). (a) fastx_clipper -a [Adapter sequence] -i input_reads.fastq -o clipped_reads.fastq -v -Q33 4. Filter by read length (see Note 20): (a) seqtk comp clipped_reads.fastq | awk ’{ if (($2 >¼ 17) & & ($2 selected-sequencesnames.list (b) seqtk subseq clipped_reads.fastq selected-sequences-names.list > short_reads.fastq 5. Remove low quality reads using fastq_quality_filter (see Note 21): (a) fastq_quality_filter -q 30 -p 90 -i short_reads.fastq -o filtered_reads.fastq -v -Q33

Global MicroRNA Profiling of Vascular Endothelial Cells

175

6. Control the quality of the processed reads using fastqc. Good sequence quality should be observed, a peak of length distribution at 22 nucleotides, and no adapters nor hits in overrepresented sequences. Per base sequence content and GC distribution are expected to be non-homogeneous. (a) fastqc filtered_reads.fastq -o QC_output_folder/ 7. Convert the filtered fastq files to fasta files using fastq_to_fasta: (a) fastq_to_fasta -i filtered_reads.fastq -o reads.fa -v -Q33 3.14.2 Obtaining the Annotations for Hairpins and Mature miRNA Sequences

1. Download mature.fa and hairpin.fa from miRBase (http:// www.mirbase.org/ftp.shtml) [23], place them in the same folder. 2. Create a hsa_mature.fa and hsa_hairpin.fa file that contains the mature sequences and hairpins (precursors) for the Homo sapiens species only (see Note 22): (a) makeblastdb -in hairpin.fa -parse_seqids -dbtype nucl (b) awk ’/sapiens/ {print hs_hairpin_names_1

$1}’

hairpin.fa

>

(c) cut -c 2- hs_hairpin_names_1 > hs_hairpin_names_2 (d) while read line; do blastdbcmd -db hairpin.fa -entry $line >>hsa_hairpin.fa; done < hs_hairpin_names_2 (e) makeblastdb -in mature.fa -parse_seqids -dbtype nucl (f) awk ’/sapiens/ {print hs_mature_names_1

$1}’

mature.fa

>

(g) cut -c 2- hs_mature_names_1 > hs_mature_names_2 (h) while read line; do blastdbcmd -db mature.fa -entry $line >>hsa_mature.fa; done < hs_mature_names_2 3.14.3 Method 1: Identification of Known Canonical miRNA Species

1. Collapse the reads using mirdeep2 [14] mapper.pl command (see Note 23): (a) mapper.pl reads.fa -c -j -m -s collapsed_reads 2. Map reads against precursor and known mature miRNA species and count them using mirdeep2 quantifier.pl command. The results are found in the output directory expression_analyses_sample1/output/miRNA_expressed.csv. For multiple samples, run step 1. and 2. for each of them, separately. (a) quantifier.pl -p hsa_hairpin.fa -m hsa_mature.fa -r collapsed_reads -t hsa -d -j -y sample1 3. If multiple samples are used, create a file named list_samples that contains one sample name on each line. Then rename and merge the multiple count csv files that were previously outputted (see Note 24):

176

Eloi Schmauch et al.

(a) mkdir counts (b) while read line (c) do l

cp expression_analyses_${line}/miRNA_expressed.csv counts/${line}_ counts.csv

(d) done < list_samples (e) cd counts (f) ls *.csv > list_files (g) echo "" > counts.csv (h) awk ’{print $1}’ sample1_counts.csv >> counts.csv (i) while read line (j) do l

var1¼$(echo $line | cut -f1 -d.)

l

echo $var1 > column

l

awk ’{print $2}’ $line >> column

l

paste counts.csv column > counts_2.csv

l

cat counts_2.csv > counts.csv

(k) done < list_files. 4. Load counts.csv containing miRNA counts for each sample (R code): (a) counts ¼ read.csv("counts.csv", sep ¼ ’\t’) 5. Remove duplicate miRNA rows (R code, see Note 25): (a) counts ¼ counts[!duplicated(counts$mirna_name),] 3.14.4 Method 1: Identification of Potential Novel miRNAs

1. Download the reference genome and build the index with bowtie [24]: (a) bowtie-build genome.fa genome 2. Map reads to the reference genome with miRdeep2 mapper.pl command: (a) mapper.pl reads.fa -c -j -m -s -p genome -s collapsed_reads_2. fa -t alignment.arf 3. Run the mirRdeep2 pipeline (see Note 26): (a) miRDeep2.pl collapsed_reads_2.fa genome.fa alignment. arf hsa_mature.fa mouse_mature.fa hsa_hairpins.fa -t Human 2 > report.log

3.14.5 Method 2, Approach 1: Identification of isomiRs and Preprocessing (miRge2.0, See Note 27)

1. Install bowtie2.0, samtools [25], RNAFold [26], and conda [27]. 2. Install miRge2.0 [28]: (a) conda config --add channels defaults (b) conda config --add channels conda-forge

Global MicroRNA Profiling of Vascular Endothelial Cells

177

(c) conda config --add channels bioconda (d) conda install mirge 3. Download the miRge library for your species of interest as instructed in https://github.com/mhalushka/miRge 4. Create a text file named samplesfile, where each row specifies the location of one sample fasta file. 5. Run miRge2.0 (see Note 28): (a) miRge2.0 annotate -s samplesfile -d miRBase -pb [path to bowtie binary] -lib [path to downloaded miRge libraries] -sp human 6. Load mapped.csv (R code): (a) counts ¼ read.csv("mapped.csv", sep ¼ ’\t’) 7. Remove sequences that have less than a certain number of counts in total (R code, see Note 29): (a) counts ¼ [(rowSums(counts) >¼ 25),] 3.14.6 Method 2, Approach 2: Identification of isomiRs and Preprocessing (Prost!, See Note 30)

1. Download and install Java version 8 (see Note 31): (a) sudo apt update (b) sudo apt install openjdk-8-jdk openjdk-8-jre 2. Download and install BBMap (BBMap – Bushnell B. – sourceforge.net/projects/bbmap/): (a) cd $HOME (b) tar xzf BBMap_XX.XX.tar.gz (c) export PATH ¼ $PATH:$HOME/bbmap 3. Install Prost! [15] (see Note 32): (a) pip install prost --user 4. Download the reference genome for your species of interest and build the BBMap database (see Note 33): (a) bbmap.sh k ¼ 7 path ¼ BBMap/Homo_sapiens.GRCh38. dna_sm.toplevel ref ¼ Homo_sapiens.GRCh38.dna_sm. toplevel.fa.gz 5. Create samples_filelist, a text file listing the input fasta files and their sample names. Each line contains the path to the processed sample fasta file and the sample identifier after a space (one line per sample). 6. Create a prost.config file and modify the parameters to fit your needs (path to miRBase mature.fa and hairpin.fa file, samples_filelist, the created genome database, change chosen species to hsa for human) (see Note 34).

178

Eloi Schmauch et al.

7. Run Prost! (see Note 35): (a) prost. 8. Load the isomiR count matrix with the annotation matrix (R code): (a) counts ¼ read.csv("output_uncompressed_isomirs.tsv", sep ¼ ’\t’) (b) annotation ¼ read.csv("output_compressed_by_annotation. tsv", sep ¼ ’\t’) 9. Subset the isomiR dataframe to contain only species that have annotations i.e., are identified as actual isomiRs (R code, see Note 36): (a) counts ¼ counts[(as.vector(counts.species$Anno_idx) !¼ "")),] 3.14.7 Differential Expression Analysis (all Code Is in R)

1. The R package DESeq2 [29] is used for differential expression analysis. Install DESeq2 on R (see Note 37): (a) if (!requireNamespace("BiocManager", quietly ¼ TRUE)) l

install.packages("BiocManager")

(b) BiocManager::install("DESeq2") (c) library("DESeq2") 2. Create an R count matrix from the loaded dataframe (see Note 38): (a) count_matrix ¼ counts[,c(2:20)] 3. Create a normalized count matrix (in reads per million, RPM): (a) normalized_count_matrix ¼ 1000000 * (count_matrix / colSums(count_matrix)) 4. Create a dataframe called colldata that contains the count matrix column names as row names and has at least one column that recapitulates the phenotype(s)/condition(s). The column type is factor (see Note 39). 5. Select miRNAs/isomiRs for which average of normalized expression per sample is above a chosen threshold (see Note 40). Create a list of species identifiers. (a) selected_species ¼ (rowSums(normalized_count_matrix) / ncol(normalized_count_matrix)) >¼ 20 6. Create the DESeq2 dataset: (a) dds ¼ DESeqDataSetFromMatrix(countData ¼ as.matrix (count_matrix), colData ¼ coldata, design ¼ ~ phenotype) 7. Run the DESeq2 model: (a) dds ¼ DESeq(dds) 8. Subset the dataset with the selected species:

Global MicroRNA Profiling of Vascular Endothelial Cells

179

(a) dds ¼ dds[selected_species,] 9. Have the same identifier for the DESeq2 data and our dataset: (a) row.names(dds) [selected_species,])

¼

row.names(count_matrix

10. Run the DESeq2 analysis: (a) res ¼ results(dds) 11. Print a summary of the differential expression results (see Note 41): (a) summary(res)

4

Notes 1. Data quality is of extreme importance to the downstream data analysis. Therefore, it is crucial to choose the RNA extraction and library preparation protocols with care. This protocol utilizes Qiaseq miRNA library preparation kit that is meant for Illumina Next Generation Sequencing (NGS) systems. Other commercial and non-commercial options exist and can be used to achieve similar results. However, the benefit of this kit over others is that it does not require gel purification, excision, and elution, which shortens the required hands-on time and the length of the whole protocol significantly, especially when handling larger sample sets (the kit enables preparation and multiplexing of up to 96 samples). Shorter handling time improves data quality. Also, gel purification and excision are common sources of contamination. If the purpose of the study is to study small RNAs in general, this kit is not suitable, because it enriches for mature miRNAs and can therefore cause bias. As mature miRNAs contain both a 30 hydroxyl group and a 50 phosphate group, it allows for specific adapter ligation to both miRNA ends and enables mature miRNA-specific reverse transcription and library preparation, while minimizing background from other RNA species. 2. Although it is important to work carefully and diligently, it is also advisable to work as fast as possible for the best cell yield. For the best number and quality of cells, the isolation should be done within 4 h of the cord collection. 3. Before cutting the other end open, push PBS into the cord. Allow the pressure to build up, and place the cord on the dry drape, so any puncture holes in the vein can be spotted. Cut the other end of the cord and rinse with PBS until no blood is visible. Separate the cord into smaller portions if holes are found. It is crucial to locate any possible holes before collagenase treatment.

180

Eloi Schmauch et al.

4. After filling the vein with collagenase, pinch the feeding tube before removing the syringe, as pressure build-up in the cord may push the collagenase out. 5. Do not exceed the incubation time for it may detach other cell types from the wall resulting in fibroblast and smooth muscle cell contamination. A contamination like this is easy to spot under microscope based on cell morphology. Endothelial cells grow in a cobblestone-like single layer as opposed to more spindle-shaped appearance of fibroblasts and smooth muscle cells. 6. Endothelial cells are sensitive to trypsin treatment. Therefore, incubation time should be kept to minimum. Cells detach faster when the flask is placed to 37  C. After incubation, cells will detach when the flask is tapped gently. However, once the EGM is added to deactivate trypsin, the cells can attach to the flask again and should thus be quickly removed from the coated cell culture flask to a centrifuge tube. 7. Protocol given here is a general protocol for Human Umbilical Vein Endothelial Cells (HUVECs). Grow and treat cells according to protocol that is specific to the study and cell type used. The library preparation protocol is suitable for other cell and tissue types. We have successfully produced miRNA-seq data with this protocol from human cells, tissue, biofluids, and biofluid exosomes. The Library preparation protocol starts from total RNA containing miRNAs and does not require enrichment of small RNAs. The recommended starting amount of total RNA is 100 ng for cell and tissue samples, but you may use as little as 1 ng. Less than 1 ng is not recommended, as it may result in extra peaks in Library QC and lower data quality. If needed, pool samples together for sufficient yield. When working with serum/plasma samples the recommended starting amount of total RNA is 5 μL of the RNA eluate when 200 μL of serum/plasma has been processed using the miRNeasy Serum/Plasma Kit, miRNeasy Serum/Plasma Advanced Kit or similar. When working with other biofluid samples the recommended starting amount of total RNA is 5 μL of the RNA eluate when 600 μL of biofluid has been processed using the miRNeasy Kit, miRNeasy Advanced Kit or similar. When working with exosomal RNA samples the recommended starting amount of total RNA is 5 μL of the RNA eluate when 500 μL of plasma/serum or 1500 μL biofluid has been processed using the exoRNeasy Kit or similar. Cell growth media (especially the supplements) typically contain miRNAs. If you think this might affect your results, check the availability of products (suitable for your sample type) that have been depleted from miRNAs. In our studies,

Global MicroRNA Profiling of Vascular Endothelial Cells

181

we have not observed significant miRNA contamination arising from the use of normal growth media and supplements. 8. Ribonucleases (RNases) are very stable, active, and potent enzymes that are difficult to inactivate. It is crucial to maintain RNase-free environment. Hands and dust particles are the most common sources of RNase contamination. Never handle reagents or RNA samples without latex or vinyl gloves. This helps to prevent RNase contamination from the skin surface and from dusty laboratory equipment. Keep the RNA workspace clean and do not bring any cross-contaminants to the space. Change gloves frequently and keep tubes closed whenever possible. Keep purified RNA on ice. 9. This step is meant for RNA isolated from cells and fresh/frozen tissues. The spectral properties of nucleic acids depend on pH, and therefore, it is recommended to measure absorbance in 10 mM Tris–HCl, pH 7.5, instead of RNase-free water. This step is not useful for total RNA derived from fluids or exosomes. 10. This step is meant for the total RNA extracted from cells or fresh/frozen tissue to ensure the best data quality. However, for RNA derived from fluids or exosomes this step is not useful. We recommend the use of an automated analysis system, such as the Agilent 2100 Bioanalyzer, for determining RIN. For samples with lower RIN values than 8, increase the sequencing reads allocated per sample as, in addition to miRNAs, reads will be spend on RNA degradation products. The same applies to FFPE-derived samples, which commonly have low RIN values. 11. Keep all enzymes in all steps of the protocol in the freezer until needed. Keep them on ice at all times when used and return to freezer immediately after use. It is crucial to add all the reagents in all steps of the protocol in the indicated order and mix well. Some of the reagents/reactions are very viscous. Be extremely careful with these to ensure proper volume and thorough mixing! Failure to follow these instructions is a common cause for library QC failure and unsuccessful library preparation. 12. It is crucial that the beads are homogenous. Therefore, work quickly and resuspend beads thoroughly immediately before use. In case of delay, vortex beads again. Unsuccessful bead preparation can lead to formation of prominent unwanted side products and lead to failure of library QC. 13. The bead buffer is viscous. Beads for 4 samples (1.6 mL) can be prepared in a single 2 mL tube. 14. Remember to ensure that the beads are thoroughly mixed at all times. Work quickly and resuspend the beads immediately before use. In case of delay, vortex beads again.

182

Eloi Schmauch et al.

15. Before continuing to next step, check that the bead pellets appear to be completely dry and all residual ethanol has evaporated. Ethanol can hinder the amplification efficiency in the subsequent reactions. Evaporation time depends on humidity and should be extended if required. Remember to protect the samples from contamination while they are drying. 16. Size of the undesired band should exceed 25% of the height of the miRNA peak to be considered significant. 17. Follow Illumina’s bcl2fastq guide (https://support.illumina. com/sequencing/sequencing_software/bcl2fastq-conver sion-software.html) for this step and use a sample sheet for the barcode indexes when the data is multiplexed. 18. PCR duplicates are not a source of strong bias in miRNA-seq [4]. Nevertheless, UMIs can still be useful to control for such bias. We advise to keep the reads with duplicate UMIs and count them to assess the bias, which should be low if the experiment is successful. We have generated data both with and without UMIs. 19. Fastx_clipper, fastq_quality_filter and fastq_to_fasta are from fastx_toolkit http://hannonlab.cshl.edu/fastx_toolkit/index. html 20. Code adapted from https://www.biostars.org/p/374959/. It uses seqtk (https://github.com/lh3/seqtk). 17 and 25 are the minimum and maximum length that we use for filtering. 21. We use 30 and 90 for the -q and -p parameter respectively, meaning that 90% of the bases need to have a quality of at least 30 for the read to be included. These parameters can be modified for more/less stringent quality filtering. 22. To create such files for another species, replace the “sapiens” with the name of the species of interest. We use the command makeblastdb from BLAST+ [30]. 23. Mirdeep2 is the most used pipeline for canonical miRNA analysis. In this case, the mapper command is used to collapse and process the reads, but not for mapping per se. 24. The code here is an example of a method to merge the multiple count csv files that are outputted. Replace “sample1” by the name of the first sample. 25. The Mirdeep2 software is very permissive and will count several times multiple aligning reads. As the software counts reads associated with each miRNA and with each hairpin sequence, the counts.csv file contains several rows for the same miRNA species. Some miRBase hairpins are identical to one another while being labeled as different hairpins, leading to several lines with the same counts being outputted by mirdeep2. For hairpins with different sequences that are associated with the same

Global MicroRNA Profiling of Vascular Endothelial Cells

183

mature miRNA, values can differ slightly. Removal of duplicate rows will result in the most accurate count values. One can also manually remove duplicate hairpins, and sort out which hairpins are to be included. It would also make sense to take only the highest counts for each miRNA species. The processing must be done the same way for all samples, to avoid bias in the differential expression analysis steps. 26. The mouse_mature.fa file contains mature miRNA sequences from closely related species (here mouse) to the species of interest (here human). Generate the file the same way hsa_mature.fa was previously generated. 27. miRge2.0 is simple to use, and outputs data that can be used immediately, with many options. Ready to use libraries exist for human, mouse, fruitfly, nematode, rat, and zebrafish. You can construct custom libraries for other species using miRge-build. pl. Documentation and more complete installation instructions can be found at https://github.com/mhalushka/miRge. 28. Replace human by your species of interest, if necessary. Other parameters can be found in the documentation. We use the standard parameters. Results are outputted in miRge_ouputs folder, where an annotation report pdf contains a summary of the analysis and provides useful QC information. miRcounts. csv contains the count matrix on the canonical level analysis (gives similar results than mirdeep2 counts matrix) and mapped.csv contains the actual isomiR count table. 29. This filter needs to be implemented to aid the downstream analysis. We use a filter of 25 counts in total (for 30 samples), that will remove very lowly expressed, lower confidence sequences, for which no conclusion can be drawn. The actual filter for biological significance will be much higher and performed later. Keep the filter below average of 1 read/sample, which will already remove most of the noise, but prevent premature discarding of lowly expressed species. 30. Prost! yields results with a larger complexity, which are thus more comprehensive. It allows for example the identification of ambiguous isomiRs which can be associated with several miRBase sequences. 31. The software was developed for Java 7 and does not work on all Java versions. We use Java 8. To switch between Java versions on Linux, use the command sudo update-alternatives --config java. 32. We recommend the setup of a Python 2.7 conda environment specifically for Prost! and the use of source activate [environment name] to activate that environment before running Prost!

184

Eloi Schmauch et al.

33. We use the Ensembl [31] reference genome: ftp://ftp. ensembl.org/pub/release-102/fasta/homo_sapiens/dna/ Homo_sapiens.GRCh38.dna_sm.toplevel.fa.gz 34. We use the example provided in the documentation (https:// raw.githubusercontent.com/uoregon-postlethwait/prost/ master/prost.config.example) and keep all other parameters as is. Specific tweaking can be useful for custom analysis. 35. The output directory stipulated in the config files, [output] _uncompressed_isomirs.tsv, contains the isomiR count table. The isomiR annotation and the canonical level data can be found in [output]_compressed_by_annotation.tsv. A precise description of all of Prost! outputs can be found at https:// prost.readthedocs.io/en/latest/output.html. 36. The Anno_idx column in the count dataframe links the isomiRs to canonical/miRBase miRNA, for which information is found in the annotation dataframe, containing one line for each annotation index in the Anno_idx column). 37. For further downstream analysis and statistical assessment of the differential expression analysis, consult the DESeq2 vignette: https://www.bioconductor.org/packages/devel/ bioc/vignettes/DESeq2/inst/doc/DESeq2.html. Other differential expression analysis frameworks such as EdgeR [32] or Limma [33] can be used instead of DESeq2. 38. Replace 2 and 20 by the respective first and last column indexes corresponding to the columns that contain the actual counts of the dataset. 39. In our case the column is named “phenotype.” For this analysis, we have two different phenotypes and thus a binary factor for that column (e.g., treatment and control). 40. We use a threshold of 20 RPM/sample on average. This is a relatively high threshold, which we apply both for isomiR analysis and canonical miRNAs. This not only increases the power by reducing the number of species, but also permits higher confidence in the results, as all differentially expressed species have relatively high expression and selected isomiR are unlikely to be random sequencing errors. The threshold can be adapted based on your biological questions, especially for isomiR analysis. A higher threshold means a higher chance of biological significance and lower threshold will yield more isomiR species. 41. The res object is a dataframe containing results for differential expression analysis, including logFC and FDR corrected p-values, for each miRNA species. The dds object contains the expression dataset processed through DESeq2 model. Further options, transformations and statistical analysis are described in the DESeq2 vignette.

Global MicroRNA Profiling of Vascular Endothelial Cells

185

Acknowledgements This work was supported by the Doctoral Program in Molecular Medicine (DPMM) of the University of Eastern Finland [to ES], the Academy of Finland [grant no 342074, to SLK], the Finnish ˜Foundation for Cardiovascular Research [to SLK], the Sigrid JusO lius Foundation [to SLK], the Orion Research Foundation [to SLK], and the Yrj- Jahnsson Foundation [to SLK]. References 1. Zhang N, Hu G, Myers TG, Williamson PR (2019) Protocols for the analysis of microRNA expression, biogenesis, and function in immune cells. Curr Protoc Immunol 126:e78. https://doi.org/10.1002/cpim.78 2. Fasolo F, Di Gregoli K, Maegdefessel L, Johnson JL (2019) Non-coding RNAs in cardiovascular cell biology and atherosclerosis. Cardiovasc Res 115:1732–1756. https://doi. org/10.1093/cvr/cvz203 3. Gholaminejad A, Zare N, Dana N et al (2021) A meta-analysis of microRNA expression profiling studies in heart failure. Heart Fail Rev 26:997–1021. https://doi.org/10. 1007/s10741-020-10071-9 4. Wong RKY, MacMahon M, Woodside JV, Simpson DA (2019) A comparison of RNA extraction and sequencing protocols for detection of small RNAs in plasma. BMC Genomics 20:446. https://doi.org/10.1186/s12864019-5826-7 5. Giraldez MD, Spengler RM, Etheridge A et al (2018) Comprehensive multi-center assessment of small RNA-seq methods for quantitative miRNA profiling. Nat Biotechnol 36: 746–757. https://doi.org/10.1038/nbt. 4183 6. Coenen-Stass AML, Magen I, Brooks T et al (2018) Evaluation of methodologies for microRNA biomarker detection by next generation sequencing. RNA Biol 15:1133–1145. https://doi.org/10.1080/15476286.2018. 1514236 7. Belair CD, Hu T, Chu B et al (2019) Highthroughput, efficient, and unbiased capture of small RNAs from low-input samples for sequencing. Sci Rep 9:2262. https://doi.org/ 10.1038/s41598-018-38458-7 8. Baras AS, Mitchell CJ, Myers JR et al (2015) miRge - a multiplexed method of processing small RNA-Seq data to determine microRNA entropy. PLoS One 10:e0143066. https://doi. org/10.1371/journal.pone.0143066

9. Muller H, Marzi MJ, Nicassio F (2014) IsomiRage: from functional classification to differential expression of miRNA isoforms. Front Bioeng Biotechnol 2:38. https://doi.org/10. 3389/fbioe.2014.00038 10. Sun Z, Evans J, Bhagwate A et al (2014) CAP-miRSeq: a comprehensive analysis pipeline for microRNA sequencing data. BMC Genomics 15:423. https://doi.org/10.1186/ 1471-2164-15-423 11. Stocks MB, Moxon S, Mapleson D et al (2012) The UEA sRNA workbench: a suite of tools for analysing and visualizing next generation sequencing microRNA and small RNA datasets. Bioinformatics 28:2059–2061. https:// doi.org/10.1093/bioinformatics/bts311 12. Pantano L, Estivill X, Martı´ E (2010) SeqBuster, a bioinformatic tool for the processing and analysis of small RNAs datasets, reveals ubiquitous miRNA modifications in human embryonic cells. Nucleic Acids Res 38:e34. https:// doi.org/10.1093/nar/gkp1127 13. Rueda A, Barturen G, Lebro´n R et al (2015) sRNAtoolbox: an integrated collection of small RNA research tools. Nucleic Acids Res 43: W467–W473. https://doi.org/10.1093/ nar/gkv555 14. Friedl€ander MR, Mackowiak SD, Li N et al (2012) miRDeep2 accurately identifies known and hundreds of novel microRNA genes in seven animal clades. Nucleic Acids Res 40: 37–52. https://doi.org/10.1093/nar/ gkr688 15. Desvignes T, Batzel P, Sydes J et al (2019) miRNA analysis with prost! Reveals evolutionary conservation of organ-enriched expression and post-transcriptional modifications in threespined stickleback and zebrafish. Sci Rep 9: 3913. https://doi.org/10.1038/s41598019-40361-8 16. Zhong X, Pla A, Rayner S (2019) Jasmine: a Java pipeline for isomiR characterization in miRNA-Seq data. Bioinformatics 36:

186

Eloi Schmauch et al.

1933–1936. https://doi.org/10.1093/bioin formatics/btz806 17. Kesharwani RK, Chiesa M, Bellazzi R, Colombo GI (2019) CBS-miRSeq: a comprehensive tool for accurate and extensive analyses of microRNA-sequencing data. Comput Biol Med 110:234–243. https://doi.org/10. 1016/j.compbiomed.2019.05.019 18. Sheu-Gruttadauria J, Xiao Y, Gebert LF, MacRae IJ (2019) Beyond the seed: structural basis for supplementary microRNA targeting by human Argonaute2. EMBO J 38:e101153. h t t p s : // d o i . o r g / 1 0 . 1 5 2 5 2 / e m b j . 2018101153 19. Telonis AG, Magee R, Loher P et al (2017) Knowledge about the presence or absence of miRNA isoforms (isomiRs) can successfully discriminate amongst 32 TCGA cancer types. Nucleic Acids Res 45:2973–2985. https://doi. org/10.1093/nar/gkx082 20. Karali M, Persico M, Mutarelli M et al (2016) High-resolution analysis of the human retina miRNome reveals isomiR variations and novel microRNAs. Nucleic Acids Res 44: 1525–1540. https://doi.org/10.1093/nar/ gkw039 21. Guo L, Chen F (2014) A challenge for miRNA: multiple isomiRs in miRNAomics. Gene 544: 1–7. https://doi.org/10.1016/j.gene.2014. 04.039 22. Neilsen CT, Goodall GJ, Bracken CP (2012) IsomiRs – the overlooked repertoire in the dynamic microRNAome. Trends Genet 28: 544–549. https://doi.org/10.1016/j.tig. 2012.07.005 23. Kozomara A, Birgaoanu M, Griffiths-Jones S (2019) miRBase: from microRNA sequences to function. Nucleic Acids Res 47: D155–D162. https://doi.org/10.1093/nar/ gky1141 24. Langmead B, Salzberg SL (2012) Fast gappedread alignment with Bowtie 2. Nat Methods 9: 357–359. https://doi.org/10.1038/nmeth. 1923

25. Li H, Handsaker B, Wysoker A et al (2009) The sequence alignment/map format and SAMtools. Bioinformatics 25:2078–2079. https://doi.org/10.1093/bioinformatics/ btp352 26. Lorenz R, Bernhart SH, Ho¨ner Zu Siederdissen C et al (2011) ViennaRNA Package 2.0. Algorithms Mol Biol 6:26. https://doi.org/ 10.1186/1748-7188-6-26 27. Gru¨ning B, Dale R, Sjo¨din A et al (2018) Bioconda: sustainable and comprehensive software distribution for the life sciences. Nat Methods 15:475–476. https://doi.org/10.1038/ s41592-018-0046-7 28. Lu Y, Baras AS, Halushka MK (2018) miRge 2.0 for comprehensive analysis of microRNA sequencing data. BMC Bioinformatics 19: 275. https://doi.org/10.1186/s12859-0182287-y 29. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:550. https://doi.org/10.1186/s13059014-0550-8 30. Camacho C, Coulouris G, Avagyan V et al (2009) BLAST+: architecture and applications. BMC Bioinformatics 10:421. https://doi.org/ 10.1186/1471-2105-10-421 31. Birney E, Andrews TD, Bevan P et al (2004) An overview of Ensembl. Genome Res 14: 925–928. https://doi.org/10.1101/gr. 1860604 32. Robinson MD, McCarthy DJ, Smyth GK (2010) edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26:139–140. https://doi.org/10.1093/bioinformatics/ btp616 33. Ritchie ME, Phipson B, Wu D et al (2015) Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 43:e47. https://doi.org/ 10.1093/nar/gkv007

Chapter 12 Endothelial Cell Tube Formation Assay: An In Vitro Model for Angiogenesis Mary Kelley, Sara Fierstein, Laura Purkey, and Kathleen DeCicco-Skinner Abstract Endothelial cell tube formation assay is one of the most widely used and reliable methods for studying in vitro angiogenesis. Endothelial cells plated over a basement membrane extract and subjected to angiogenic factors in conditioned medium, form a rapid and quantifiable tube network within hours. Tube formation is sustained for 18–24 h, after which time apoptosis occurs and tube networks disintegrate. The tube network can be imaged using a phase contrast microscope, or upon Calcein-AM treatment, a fluorescence/confocal microscope. This assay has several advantages, namely: ease of set up, the ability to test numerous angiogenic/anti-angiogenic factors simultaneously, quick network formation, ability to view live or fixed tube networks, and quantifiability. To ensure successful results and limit variability, proper selection of basement membrane extracts and endothelial cells is necessary, and conditions must be optimized. In summary, this assay is a useful method for screening potential angiogenic/anti-angiogenic factors as well as identifying critical mechanisms and signaling pathways underlying angiogenic-related pathologies. Key words Angiogenesis, Human umbilical vein endothelial cells (HUVEC), 3B-11, Tube formation, Endothelial, Basement membrane extract (BME)

1

Introduction Blood vessels play a critical role in the delivery of oxygen, inflammatory cells, and nutrients to tissues, as well as waste removal [1]. Angiogenesis, the formation of new blood vessels from existing ones, is vital to many physiological functions including wound healing, skeletal growth, reproduction, and fetal development [2]. However, dysregulated angiogenesis has been implicated in the pathogenesis of numerous disorders including diabetic retinopathy, rheumatoid arthritis, traumatic brain injury, and tumor growth/metastasis [3]. All blood vessels in the body are lined by endothelial cells [4]. These polarized cells contact the basement membrane containing extracellular matrix. When angiogenic stimuli bind to receptors

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_12, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

187

188

Mary Kelley et al.

on endothelial cells, a series of steps is initiated including disruption of the basement membrane, endothelial cell migration, invasion, proliferation, and differentiation into capillaries [4]. A variety of signaling molecules play important roles in regulating angiogenesis including vascular endothelial growth factor (VEGF), plateletderived growth factor (PDGF) and fibroblast growth factor (FGF) family members, among others [5]. Endothelial cell tube formation can be modeled in vitro using a tube formation assay, first established 40 years ago [6, 7]. In the absence of basement membrane extract, endothelial cells can take 4–6 weeks to form a tube network [8]. However, as Kubota first showed, exposing endothelial cells to a reconstituted matrix of basement membrane components accelerates the ability of endothelial cells to form in vitro capillary-like structures in a manner similar to an in vivo environment [9]. There are a variety of factors in basement membrane extracts that can encourage tube formation of endothelial cells. Most commercially available basement membrane extracts contain a mixture of components including laminin, collagen IV, entactin, and heparin sulfate proteoglycan. In the tube formation assay, endothelial cells are combined with conditioned medium containing suspected angiogenic/anti-angiogenic factors and plated over reduced growth factor basement membrane extract. Endothelial cells begin to align quickly and form extensive tube networks which can be imaged and quantified within hours [10]. This assay has been used to study processes ranging from cancer progression, development, and reproduction to traumatic brain injury. As described below, this technique should be optimized, and investigators should carefully select endothelial cells, basement membrane extracts and conditions that most closely resemble the model or disease being studied.

2

Materials All cells, media, solutions, and equipment should be kept sterile at all times. The preparation of these materials and execution of the assay should be performed under a sterile tissue culture hood to prevent contamination.

2.1

Cell Preparation

1. 37
C incubator with 5% CO2. 2. Complete supplemented medium; Dulbecco’s Modified Eagle Medium (DMEM) with 10% Fetal Bovine Serum (FBS), 1% Penicillin-Streptomycin (10,000 U/mL), and 1% GlutaMAX™ (100). 3. 3B-11 endothelial cells (see Note 1) 4. Starvation medium; DMEM with 0.2% FBS, 1% PenicillinStreptomycin (10,000 U/mL), and 1% GlutaMAX™ (100).

Global MicroRNA Profiling of Vascular Endothelial Cells

189

5. Calcein-AM; final concentration of 2 μg/mL, protect from light (** this is optional and should be used if fluorescence imaging is desired). 6. 1  Phosphate magnesium free.

Buffered

Saline

(PBS),

calcium/

7. Trypsin-EDTA, 0.25%. 8. Hemocytometer. 9. Trypan blue solution, 0.4%. 10. 100-micron cell strainer. 11. T-75 tissue culture treated flasks. 2.2 Tube Formation Assay

1. 24-well cell culture plate (see Note 2). 2. 10 μL, 200 μL and 1 mL pipette tips (see Note 2). 3. Container to hold cell culture plate and ice. 4. Basement Membrane Extract (BME) (see Note 3). 5. Conditioned medium from primary or immortalized cells. 6. 4% paraformaldehyde in PBS (if cell fixation is desired).

2.3 Visualization and Quantification of Tube Network

1. Fluorescence/Confocal Microscope if Calcein-AM is used. Otherwise, phase contrast microscope (see Fig. 1 for representative images from a phase contrast vs. fluorescence microscope). 2. NIH Image J with Angiogenesis Analyzer plugin [11].

3

Methods All steps of this experiment except imaging should be conducted under a sterile tissue culture hood.

3.1 Preparation of Endothelial Cells

1. Start a new culture of 3B-11 cells (or another endothelial cell line such as HUVEC) from liquid nitrogen at least 1 week prior to the start of the assay. Cells should be grown and passaged at least twice prior to assay (see Note 1). The last cell passage must be completed 2 days prior to the start of the assay. 2. The day before starting the assay, confluency of the 3B-11 endothelial cells should be approximately 80–90%. Remove DMEM from the T-75 flask and wash cells with PBS. Serum starve the cells by adding starvation medium to flasks and culturing cells in the 37
C incubator with 5% CO2 for 18–24 h. 3. The day of the assay add Calcein-AM to the flask of 3B-11 cells at a final concentration of 2 μg/mL (see Notes 4 and 5) and place cells back into the incubator for 30–45 min. Note that

190

Mary Kelley et al.

Fig. 1 Representative images using phase contrast microscope (left) or labeling cells with Calcein-AM followed by imaging with a fluorescence microscope (right)

this step is optional and used when fluorescence microscopy of the tube network is desired. For normal phase contrast imaging, skip this step and proceed to step 5. 4. After incubation, remove the medium with the Calcein-AM from the flask and wash the 3B-11 cells with 5 mL of PBS. 5. Add 1 mL of Trypsin with EDTA to the T-75 flask containing 3B-11 cells and wait 5 min, or longer if necessary, to allow for complete cell detachment. Add complete DMEM to a final volume of 10 mL. Pipette the cell suspension through a 100-micron strainer into a 50 mL conical tube. This will remove clumps of cells. 6. Count the filtered cells using a hemocytometer. Centrifuge cells at 180  g for 5 min and remove the supernatant, leaving the pellet undisturbed. Resuspend the pellet in the calculated volume needed to have 100,000 cells in 10 μL of medium (107 cells/mL) (see Note 6). 3.2 Tube Formation Assay

1. Place your 24-well plate into a container filled with ice. Make sure the plate is level in the ice. Remove the BME from the refrigerator and keep it on ice. 2. Add 250 μL of BME into each well that is being used for the assay (see Notes 2 and 7). Place the plate into the incubator for a minimum of 30 min to allow the BME to solidify. 3. Add 10 μL of the 3B-11 endothelial cells to 250–300 μL of the conditioned medium being tested. Make sure conditioned medium was centrifuged (180  g for 5 min) to remove dead cells prior to using it in this assay. Resuspend the cell/ conditioned medium mixture thoroughly (see Note 8). Add all of the cell mixture into the appropriate well in the 24-well

Global MicroRNA Profiling of Vascular Endothelial Cells

191

Fig. 2 Images of tube formation networks throughout the image and analysis. (a) Tube formation of 3B-11 endothelial cells visualized by Calcein-AM staining, and incubated with primary murine cell conditioned medium to induce tube formation. Image taken using a fluorescence microscope. (b) Image of (a) when all settings from (Subheading 3.3 step 2) were applied and saved. (c) Image of (a) once analysis has been completed. Blue labels mesh, yellow traces segments/tubes and red encircles nodes

plate. Repeat this step for each sample. Make sure you include technical replicates for each condition (triplicate per plate). Return the cell culture plate to the incubator and incubate undisturbed for 2 h. Inspect the plate under an inverted light microscope after 2 h to confirm that endothelial cells are starting to align, and a network is beginning to form. 4. Between 6 and 12 h the plate should be viewed to determine when the network is at peak formation. Once that has been identified, remove the plate from the incubator (see Note 9). 5. To capture images, use either a fluorescence scope with a GFP filter (if Calcein-AM was added) or a phase contrast microscope (Fig. 1). Make sure that you capture images of each well that depict a thorough representation of the network of tubes, mesh, and segments (see Note 10 and Fig. 2a). 6. After imaging, remove sample solution and wash each well with 1 mL of PBS. Remove the medium carefully as to avoid disturbing the network. If cell fixation is desired, add 1 mL of 4% Paraformaldehyde to each well and incubate at room temperature for 15 min. Remove paraformaldehyde carefully and wash again with 1 mL of PBS. After washing, leave 1 mL of PBS in each well. Wrap the cell culture plate in aluminum foil and store at room temperature in a dark location. 3.3 Quantification of Network

1. On a MAC or PC computer download ImageJ from NIH by going to https://imagej.nih.gov/ij/download.html and

192

Mary Kelley et al.

download the angiogenesis analyzer from http://image.bio. methods.free.fr/ImageJ/?Angiogenesis-Analyzer-forImageJ&artpage¼2-6. Save the angiogenesis analyzer to the desktop. Open ImageJ on the computer, go to plugins_macros_Install and install the angiogenesis analyzer.txt from the desktop. 2. Open the file with the image that needs to be analyzed. Choose the following settings: (a) Image_type_16-bit. (b) Process_enhance_contrast_saturated “Normalize”.

¼

0.4%,

click

(c) Process_smooth. (d) Image_adjust_threshold_click “Dark Background” and Auto. Then manually adjust the threshold. Do this by moving the top bar to zero and the bottom bar slowly to the right or left. Stop directly before background noise starts to appear. Click apply. (e) Process_binary_make binary. (f) Process_noise_despeckle. (g) Edit_invert. (h) Image_type_RGB Color. (i) Save the image (Fig. 2). 3. Open the saved image and click the “spider web” icon on the toolbar and choose “Analyze HUVEC Phase Contrast”. This will take 3–4 min and it will produce the saved image with colored lines indicating mesh, nodes, and tubes (Fig. 2b, c). A spreadsheet will appear with the numerical values for all measurements. Repeat for all images that need to be analyzed (see Note 11).

4

Notes Also refer to Fig. 3 for troubleshooting. 1. There are several different endothelial cell lines, derived from different models (mouse, rat, human, etc.). For best comparisons one should stay within their model. For optimal tube formation, the endothelial cell passage number should be between 2 and 8. Higher passage numbers will not form tube networks as quickly or extensively. Researchers should also avoid using cells immediately from liquid nitrogen, as cells need at least two passages to acclimate to CO2 and temperature to ensure maximal tube formation. Be careful not to overgrow cells. Healthy endothelial cells grow extremely quickly. One

Global MicroRNA Profiling of Vascular Endothelial Cells

193

Fig. 3 Common pitfalls and solutions for tube formation assay

may need to passage cells at 1:10 to 1:50 to avoid cells being confluent the next day. Each confluent T-75 flask will generate enough endothelial cells to set up two 24-well plates. 2. This assay can also be completed in a 12- or 96-well plate, if desired. If using a 12-well plate, the BME and conditioned media should be adjusted to 500 μL, and 20 μL of endothelial cells should be used to accommodate for larger wells. If using a 96-well plate, the volume of BME and conditioned medium should be adjusted to 60 μL and 2.5 μL of endothelial cells should be used to adjust for the smaller well. All cell culture plates, and pipette tips should be kept in a freezer prior to executing the assay. 3. We recommend calling BME manufacturers to check BME lot numbers available for purchase. BME should have a protein concentration that is a minimum of 10 mg/mL. BME should be kept frozen at 20
C and defrosted in refrigerator 1–2 days prior to assay. One should use reduced growth factor BME for best results. The basement membrane extract solidifies at room temperature quickly. Keeping the BME cold will prevent premature solidification.

194

Mary Kelley et al.

4. This should be done in the sterile tissue culture hood with the lights off. Calcein-AM is sensitive to light. 5. To save time while the Calcein-AM is labeling the cells during incubation, execute steps 1 and 2 under Tube Formation Assay (Subheading 3.2). The solidified BME can stay in the incubator until the cells are counted and the assay is ready to be performed. 6. Cell number will need to be optimized. Typically, concentrations of 7.5  106 to 107 cells/mL work well for our model. If the conditioned medium contains high concentrations of angiogenic factors, you may need fewer cells. In contrast, having a lower concentration of angiogenic factors will require increasing your cell number to form an extensive tube network. 7. Try not to make bubbles when pipetting the BME into the cell culture wells. Depress the pipette only to the first stop. Bubbles will obstruct the scope and camera from viewing the tube formation. Try not to poke the BME with the end of the pipette tip, dispense the solution slightly above where the BME is solidifying. Once all samples are added, lightly shake the plate back and forth, then left to right (a “T-shake”) to ensure equal cell distribution. 8. Conditioned medium can be collected from primary or immortalized cells prior to the day of the assay and stored at 80
C. Centrifuge the conditioned medium at 180  g for 5 min. Use the supernatant in this assay as dead cells will pellet. If many samples are being prepared for the assay, it is recommended to prepare all samples (10 μL of 3B-11 + the conditioned medium tested) prior to adding any samples to the plate. The cells can settle at the bottom of the 1.5 mL tube fairly quickly. We suggest resuspending the mixture immediately prior to plating into each well. We also recommend doing a T-shake to evenly distribute the cells across the BME. If the cells are not resuspended, pockets of clumped cells will start to form networks, and be indistinguishable when imaging due to congestion in one area. Additionally, the volume of conditioned medium can be optimized depending on the concentrations of angiogenic factors present. Typically, we find that 250–300 μL of conditioned media per sample work well for a 24-well plate. 9. Do not let the network continue to incubate after peak network formation has been determined. It is recommended that you always have a positive and a negative control well in your experiment to help determine this point. To determine peak network formation, it is recommended to check the plate every 1–2 h within the 6–12 h incubation window. At the beginning of the assay, you should only see individual cells equally dispersed in each well. As the plate incubates the network will start

Global MicroRNA Profiling of Vascular Endothelial Cells

195

to form branches of segments, nodes, and mesh, this should look similar to a spiderweb or a honeycomb would look like. Using your positive control, view the whole well to see if the surface area of the BME is covered with branches/straight lines (segments), mesh (honeycomb shapes) or nodes (points intersecting the segments). It will not be at peak formation when sections of BME have no network created, when limited nodes have formed and/or when very small branches have formed for the positive control. If the network is passed peak formation, the network will start separating from the BME and float freely in the media. Eventually the network will fully deteriorate, and cells will go through apoptosis. 10. It is recommended to take 4–5 images of each well in different locations within the well. During image analysis, analyze all 4 images and take the average of the number of mesh, nodes, and tubes/segments formed. Segments/tubes are represented by the straight/branching lines. Mesh corresponds to the regions enclosed by segments that can look similar to honeycombs, whereas nodes are points where the segments diverge. In Fig. 2c, blue outlines mesh, yellow traces segments, and red encircles nodes. 11. After analyzing images and determining the quantitative values associated with the number of nodes, tubes, and mesh, there are many ways to determine statistical significance. To analyze the data, depending on your conditions, statistical tests such as a standard T-test, one-way ANOVA or two-way ANOVA may be appropriate to use.

Acknowledgments This work was supported by grant #2R15CA152907-02. U.S. Department of Health & Human Services, National Institutes of Health (NIH). References 1. Eelen G, Treps L, Li X, Carmeliet P (2020) Basic and therapeutic aspects of angiogenesis updated. Circ Res 127:310–329 2. Yoo SY, Kwon SM (2013) Angiogenesis and its therapeutic opportunities. Mediat Inflamm 2013:127170 3. Folkman J (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1:27–31 4. Kru¨ger-Genge A, Blocki A, Franke RP, Jung F (2019) Vascular endothelial cell biology: an update. Int J Mol Sci 20:4411

5. Zhao Y, Adjei AA (2015) Targeting angiogenesis in cancer therapy: moving beyond vascular endothelial growth factor. Oncologist 20: 660–673 6. Folkman J, Haudenschild C (1980) Angiogenesis in vitro. Nature 288:551–556 7. Folkman J, Haudenschild C (1980) Angiogenesis by capillary endothelial cells in culture. Trans Ophthalmol Soc U K 100:346–353 8. Maciag T, Kadish J, Wilkins L, Stemerman MB, Weinstein R (1982) Organizational behavior of

196

Mary Kelley et al.

human umbilical vein endothelial cells. J Cell Biol 94:511–520 9. Kubota Y, Kleinman HK, Martin GR, Lawley TJ (1988) Role of laminin and basement membrane in the morphological differentiation of human endothelial cells into capillary-like structures. J Cell Biol 107:1589–1598

10. DeCicco-Skinner KL et al (2014) Endothelial cell tube formation assay for the in vitro study of angiogenesis. J Vis Exp (91):e51312 11. Carpentier G (2012) ImageJ contribution: angiogenesis analyzer. ImageJ News

Chapter 13 Use of a Thin Layer Assay for Assessing the Angiogenic Potential of Endothelial Cells In Vitro James A. E. Lane, Ashton Faulkner, Elizabeth J. T. Finding, Eleanor G. Lynam, and Caroline P. D. Wheeler-Jones Abstract Angiogenesis is essential for wound healing and regeneration and plays a significant role in several pathologies including cancer and atherosclerosis. In vitro assays offer simple and powerful tools for investigating the regulation of the angiogenic functions of primary endothelial cells (ECs) before moving to in vivo studies. The classic in vitro two-dimensional angiogenesis assay utilizes Basement Membrane Extract (BME) to study the differentiation and sprouting of ECs over a 24-h period. The protocol described here details a thin layer BME adaptation of the angiogenesis assay requiring significantly less BME and carried out in 96-well plates, allowing for a larger data yield at a greatly reduced cost, while maintaining the robustness of an assay used extensively over the past three decades. Key words Angiogenesis, Tube formation, Endothelial cell, Endothelial Colony Forming Cells, Basement membrane extract, VEGF

1

Introduction Angiogenesis, the sprouting of new blood vessels from preexisting vasculature, is crucial for embryonic development, organ growth and physiological processes such as wound healing and regeneration [1, 2]. Numerous factors are involved, including cytokines, integrins, angiopoietins, junctional adhesion molecules and growth factors, such as vascular endothelial growth factor (VEGF). It is well established that dysregulated angiogenesis contributes to cancer, atherosclerosis, inflammatory and immune disorders, as well as many other disease pathologies [3, 4]. As such, understanding the pro- and anti-angiogenic signals and pathways that regulate angiogenesis both in health and disease remains a key area of investigation for identifying new potential therapeutic targets [5, 6]. The angiogenesis assay, also known as the “tube” formation assay is a widely published in vitro assay used to interrogate the molecular

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_13, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

197

198

James A. E. Lane et al.

basis of angiogenesis in ECs. First described by Kubota in 1988 [7], the assay involves plating ECs at an optimized density onto a layer of BME. These ECs sprout and differentiate to form tube-like structures over a period of 2–24 h, which can be easily quantified by light microscopy. BME is crucial to this process [8] and expensive commercial products such as Geltrex™ or Matrigel™, which contain extracellular matrix proteins isolated from EngelbrethHolm-Swarm tumor, work especially well. Traditionally however, a large volume of between 50 and 200 μL/cm2 BME per well of a 24-well plate is used to carry out this assay. This makes it expensive and precludes subsequent RNA extraction and high-resolution microscopy analysis without BME depolymerization using commercial cell recovery solutions. Additionally, proteins within the BME and components of cell culture media bound by thick layers of BME are a significant source of autofluorescence background during imaging. We developed a thin layer angiogenesis assay [9] which we have used extensively for assessing the angiogenic potential of human umbilical vein ECs (HUVEC) [10]. Here we describe the use of this assay for quantifying tubulogenesis by endothelial colony forming cells (ECFCs) in particular, but the assay is equally applicable to any primary human ECs. The assay is performed in 96-well plates, allowing for high-throughput, and utilizing only 2 μL (6.25 μL/cm2) of BME per well, as well as requiring fewer cells than the standard 24-well plate format. This simple and easy to use assay generates comparable data with a greater yield and allows for direct high-resolution immunofluorescent imaging and transcriptional analysis, while being 8–32 fold relatively more cost effective.

2

Materials 1. Geltrex™ lactose dehydrogenase elevating virus (LDEV)-Free reduced growth factor basement membrane matrix. 2. Basal Medium (BM): Endothelial Basal Medium 2 (EBM™-2) (PromoCell). 3. Complete growth medium: Endothelial Cell Growth Medium (EGM™-2) (PromoCell): EBM™-2 complemented with 20% FBS, epidermal growth factor 5 ng/mL, basic fibroblast growth factor 10 ng/mL, insulin-like growth factor 20 ng/ mL, vascular endothelial growth factor-A 165 (VEGF) 0.5 ng/ mL, ascorbic acid 1 μg/mL, hydrocortisone 0.2 μg/mL (see Note 1). 4. Tissue culture flask 80 cm2, sterilized, polystyrene, filter capped (T75).

Thin Layer Angiogenesis Assay for Assessing ECFC Tubulogenesis

5. Tissue culture polystyrene.

96-well

plate,

flat

bottom,

199

sterilized,

6. Trypsin-EDTA: 0.25% trypsin, 0.02% EDTA. 7. Human recombinant VEGF-A. 8. ECFCs or other primary human endothelial cells (e.g. HUVECs, human dermal blood ECs (HDBECs)). 9. Phosphate Buffered Saline (PBS): 140 mM NaCl, 3 mM KCl, 10 mM phosphate buffer, pH 7.4. 10. 4% Paraformaldehyde (PFA) in PBS. 11. Hoechst 33342 fluorescent stain. 12. Alexa Fluor® 488 phalloidin. 13. Sterile 250 μL repeat dispenser tip insert (Starlab). 14. Minimum Essential Medium (MEM) with Hanks’ salts, L-glutamine, Phenol Red (Gibco) complemented with Penicillin/Streptomycin 50 U/mL. 15. Commercial RNA extraction kit such as RNeasy (Qiagen) or TRIzol™.

3

Methods

3.1 Thin Layer Tube Formation Assay

1. Place an aliquot of Geltrex™ at 4  C overnight or on ice for at least 3 h to allow the solution to thaw. 2. Culture primary ECFCs in a T75 flask in complete growth medium until 80–90% confluent (see Note 2) in readiness for the assay. 3. To serum-deprive the cells, remove supernatant from the flask and add 10 mL of BM. Incubate cells for 1–24 h in a humidified tissue culture incubator (37  C, 5% CO2), depending on specific experimental settings. 4. Add 2 μL of Geltrex™ to each well of a 96-well flat-bottomed cell culture plate with both the plate and the Geltrex™ on ice (see Note 3). Make sure to add the solution to the center of the well and spread it evenly across the surface of the well using the flat end of a sterile 250 μL repeat dispenser tip insert (see Note 4). 5. Place the plate in the incubator for 30 min to allow the BME to fully set. 6. Wash ECFC monolayer with pre-warmed sterile PBS to fully remove cell supernatant. Aspirate excess PBS. 7. Detach cells using pre-warmed trypsin-EDTA solution (1 mL/ T75 flask) by incubating (37  C, 5% CO2) for 1–5 min.

200

James A. E. Lane et al.

8. Resuspend cells in 1–5 mL BM and count using a hemocytometer. 9. Plate 5,000 ECFCs (100,000 cells/mL) (see Note 5) in each well, immediately followed by the addition of suitable controls and treatments diluted in BM (see Note 6). We routinely use VEGF (25 ng/mL) as a positive control. A total volume of 100 μL per well works well in our experience. We combine 50 μL of cell suspension in BM with 50 μL of 2 treatment/ control also in BM. For example, for VEGF treatment we use 50 μL of 50 ng/mL stock to achieve a final concentration of 25 ng/mL. 10. Incubate the plate for 16 h (overnight) in a humidified tissue culture incubator (37  C, 5% CO2) (see Note 7). 11. Image immediately, using an inverted brightfield microscope, to avoid collapse of tube-like structures. Alternatively, cells may be fixed for later analysis (see Subheading 3.2). 3.2 PFA Fixation and Fluorescent Staining

1. Remove supernatant and gently wash the cells once with PBS at room temperature (RT) (see Note 8). 2. Visualize the wells using bright-field microscopy to check for any disruption to tube-like EC networks. If any of the wells have significantly deteriorated networks, then mark these wells and do not include them in the quantification process. 3. Replace the PBS with 50 μL of 4% PFA per well and leave for 10 min at RT. 4. Remove PFA and wash twice with PBS. 5. Add 25 μL total volume of Alexa Fluor® 488 phalloidin (1:20 dilution in PBS, 50 Units/mL) supplemented with Hoechst 33342 fluorescent stain (final concentration: 1 μg/mL) and leave for 30 min at RT covered in foil on an orbital shaker with gentle agitation (see Note 9). 6. Remove staining solution and add 100 μL/well PBS for imaging, or 250 μL/well if storing the plate for bulk imaging later.

3.3 Image Analysis and Quantification

1. To capture phase-contrast images we use a Leica DMIRB inverted microscope (10 magnification) and for fluorescent imaging, a Leica SP5 confocal microscope (20 magnification); other similar microscopes are equally suitable (Fig. 1). Phase-contrast images are taken at the center of each well (using the meniscus as a marker and to increase contrast) whereas representative fluorescent images are taken of fixed cells. All images are saved as tiff files. Mean number of branches per high powered field can be counted manually using ImageJ software (https://imagej.nih.gov/ij/download.html).

Thin Layer Angiogenesis Assay for Assessing ECFC Tubulogenesis

201

Fig. 1 Tubulogenesis assay using ECFCs plated on 2 μL Geltrex™ BME in a 96-well plate. Cells were exposed to control medium or medium containing VEGF (25 ng/mL) for 16 h. Cells were stained with Alexa Fluor® 488 phalloidin (green) and Hoechst 33342 fluorescent stain (blue). Images were acquired using a Leica SP5 confocal microscope (20 magnification)

2. For automated quantification, images can be analyzed using the angiogenesis macro (http:// http://image.bio.methods. free.fr/ImageJ/?%20Angiogenesis-Analyzer-for-ImageJ) for ImageJ. Once the Angiogenesis Analyzer plugin is installed, open the images using ImageJ and click on the “Network Analysis Menu” drop down menu. Choose the “Analyze Fluo” option for fluorescent images and “Analyze Phase Contrast” option for phase contrast images. The software will generate various data parameters, including the number of branches, length of branches and nodes. 3. For live cell imaging (see Fig. 2), images are acquired using a fluorescent microscope (10 objective) fitted with a heated (37  C) closed imaging stage and supplied with 5% CO2. 3.4 RNA Extraction for qPCR

1. Due to the significantly reduced volume of BME used in the thin layer assay, RNA extraction can be performed directly without the use of digestion steps. However, it is recommended to carry out the assay in 24-well plates using 10 μL/ well of Geltrex™ and 25,000 ECFCs per well (or 50,000 HUVECs per well) to yield sufficient RNA to carry out subsequent qPCR analyses. 2. At the peak of network formation (typically around 16 h, depending on ECs and specific treatment), carefully remove supernatant from the wells and wash once with RT PBS. 3. Visualize the wells under bright-field microscopy to check for any disruption to tube-like networks. If any of the wells have

202

James A. E. Lane et al.

Fig. 2 Live cell imaging of a thin layer angiogenesis assay over 24 h performed using HUVEC at an initial seeding density of 10,000 cells per well (25,000 cells/cm2) in basal medium supplemented with VEGF (25 ng/ mL). Images at time zero and at 4, 8 and 16 h are shown and were acquired using a Leica SP5 confocal microscope (10 magnification) fitted with a heated (37  C) closed imaging stage supplied with 5% CO2. The 16-h time point had the largest number of branches and greatest effect size for many measurements, including branch length and number of nodes

significantly deteriorated networks, then mark these wells and do not include them in the RNA extraction process. 4. The remaining cells can be lysed directly in the well and the RNA extracted using most commercial extraction kits following the manufacturer’s tissue culture extraction instructions.

4

Notes 1. Depending on the primary EC type used in the assay other growth/experimental media may be more appropriate. 2. Several rounds of passaging can significantly affect the extent of tube formation measured in this assay (see Fig. 3) and is

Thin Layer Angiogenesis Assay for Assessing ECFC Tubulogenesis

203

Fig. 3 Images of ECFC networks obtained after 16 h incubation of low passage (P5) or replicatively senescent (P15) ECFCs on 2 μL Geltrex™ BME in a 96-well plate. Cells were stained with Alexa Fluor® 488 phalloidin (green) and Hoechst 33342 fluorescent stain (blue). Images were acquired using a Leica SP5 confocal microscope (20 magnification)

dependent upon cell type. For example, ECFCs beyond passage 8 and HUVEC beyond passage 4 should not be used in this assay. 3. Geltrex™ can set rapidly at room temperature so ensure that pipette tips are pre-cooled and always keep the 96-well plate and aliquot of BME on ice to prevent it from solidifying. Once set, Geltrex™ should not be used again. It is helpful, when preparing a full 96-well plate assay, to add Geltrex™ to a maximum of 24 wells at a time, before distributing the gel across the wells evenly with a repeat dispenser tip insert. 4. Any sterile, flat, ~200 μL tip-sized consumable will work for spreading out the Geltrex™ evenly across the well. Only apply light pressure when distributing the BME. 5. Seeding density is of paramount importance for this assay and varies with EC type used. If too many cells are plated, they will form a monolayer; in contrast, too few cells will fail to form a tube-like network. In addition, even when plating the adequate number of cells, they need to be seeded evenly across the BME, otherwise the network will be patchy making quantification unreliable. Make sure to use a robust pipetting technique to resuspend the cells. Additionally, try gently shaking the plate on a flat surface or use a cell strainer to avoid/disperse cell clumps. 6. We recommend using six technical repeats to account for any loss of wells due to damaged networks and technical variation.

204

James A. E. Lane et al.

7. It is advised to carry out an initial series of 24-h time-lapse experiments (see Subheading 3.3, step 3 and Fig. 2) to optimize for a timepoint with the largest effect of the positive control (VEGF; 25 ng/mL) versus untreated controls, due to the high variability between EC types and between isolates of ECFCs from individuals. 8. Tube-like networks are easily disturbed before fixation. Make sure to wash wells very slowly, running the PBS down the side of the well to reduce the force experienced by the fragile networks. Failure to do so can cause large areas of cell networks to detach. Try using Ethanol fixation instead of PFA if this technique leads to poor cell membrane staining. 9. Traditional primary and secondary immunocytochemistry methods using fluorophore-conjugated antibodies can be used instead of the F-actin binding phalloidin. However, these may not stain the tubular networks as extensively and will require additional incubation steps and washes that increase the risk of disruption to the EC networks. References 1. Chung A, Ferrara N (2011) Developmental and pathological angiogenesis. Annu Rev Cell Dev Biol 27:563–584 2. Johnson K, Wilgus T (2014) Vascular endothelial growth factor and angiogenesis in the regulation of cutaneous wound repair. Adv Wound Care 3(10):647–661 3. Eelen G, Treps L, Li X et al (2020) Basic and therapeutic aspects of angiogenesis updated. Circ Res 127:310–329 4. Carmeliet P (2005) Angiogenesis in life, disease and medicine. Nature 438:932–936 5. Bouis D, Kusumanto Y, Meijer C et al (2006) A review on pro- and anti-angiogenic factors as targets of clinical intervention. Pharmacol Res 53(2):89–103 6. Carmeliet P, Jain K (2011) Molecular mechanisms and clinical applications of angiogenesis. Nature 473:298–307

7. Kubota Y, Kleinman HK, Martin GR et al (1988) Role of laminin and basement membrane in the morphological differentiation of human endothelial cells into capillary-like structures. J Cell Biol 107:1589–1598 8. Staton C, Reed M, Brown N (2009) A critical analysis of current in vitro and in vivo angiogenesis assays. Int J Exp Pathol 90(3):195–221 9. Faulkner A, Purcell R, Hibbert A et al (2014) A thin layer angiogenesis assay: a modified basement matrix assay for assessment of endothelial cell differentiation. BMC Cell Biol 15:41 10. Faulkner A, Lynam E, Purcell R et al (2020) Context-dependent regulation of endothelial cell metabolism: differential effects of the PPARβ/δ agonist GW0742 and VEGF-A. Sci Rep 10(1):7849

Chapter 14 VEGF-A165 -Induced Endothelial Cells Chemotactic Migration and Invasion Assays Caroline Pellet-Many Abstract In vitro assays of endothelial cell migration have led to critical insights into the mechanisms of angiogenesis. The transwell assay, or modified Boyden chamber assay was developed to investigate chemotaxis, which corresponds to the directional migration of cells in response to a chemoattractant gradient. It is a reliable and convenient assay that does not require expensive equipment. In the modified Boyden chamber assay, two compartments are separated with a porous membrane through which cells can migrate. The lower compartment contains the chemoattractant, creating a gradient by diffusing into the upper chamber containing the cells. Adherent cells will migrate through the membrane and remain on the lower side of the membrane, where they can finally be fixed, stained, and counted. Key words Chemotaxis, VEGF-A165, Endothelial cells, Migration, Transwell

1

Introduction Angiogenesis corresponds to the generation of new blood vessels from pre-existing ones. This process involves several essential steps initiated by the proteolytic degradation of the extracellular matrix (ECM) followed by chemotactic migration toward an unvascularized region, proliferation of endothelial cells (EC), formation of a lumen and finally, maturation and stabilization of the newly formed vessel via the recruitment of perivascular cells [1]. In response to environmental cues, EC leave their quiescent state and acquire a migratory profile to invade ischemic regions of tissue, toward chemoattractant gradients. One of the most potent molecules responsible for the migratory function of EC, along with fibroblast growth factor (FGF) and angiopoietins, is vascular endothelial growth factor (VEGF) [2], initially called vascular permeability factor (VPF) because of its ability to render microvessels hyperpermeable to plasma proteins [3]. The heparin binding, but also diffusible isoform, VEGF-A165, is regarded as the most abundant

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_14, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

205

206

Caroline Pellet-Many

[4] and biologically active isoform and is able to induce proliferation and migration of EC in vitro and blood vessel permeabilization in vivo [5]. Cell migration falls into 4 different categories: chemotaxis corresponds to the directional response cells toward a gradient of chemoattractant in a fluid phase; it can be either positive if the movement is toward a higher concentration of the chemical involved, or negative if the organism or cell movement occurs in the opposite direction. This process is called haptotaxis when the chemoattractant molecules are attached to a support, like the surface of a dish or the extracellular matrix, which would be the case for example, for the larger heparin-binding isoform VEGF-A206. Necrotaxis is a special type of chemotaxis that occurs when the chemoattractant molecules are released from necrotic or apoptotic cells. Finally, chemokinesis is the response of a cell to a chemical that causes the cell to modulate its movement by speeding it up, slowing it down or changing its direction in a random manner (non-directional). Several assays have been developed to assess the migration of endothelial cells. One of the most basic experiments consists of scraping a confluent adherent layer of EC with a fine tip or scalpel. This reproduces the endothelium denudation of vessels following angioplasty. EC situated at the wound edges migrate laterally to repopulate the bare inner surface of the artery. Although this technique presented some drawbacks in early stages, with issues relating to the reproducibility of the wound size and damage inflicted to the plasticware, now gentler and highly reproducible scratch wounds can be produced with commercially available specialized equipment (WoundMaker™, Essen Bioscience). The lateral migration of EC culminates with complete wound closure after a few days. These classical non-directional chemokinetic experiments are used to determine the effect of treatments with drug compounds or cytokines on overall migration. In order to investigate chemotaxis, migration of a cell population in a directional manner, a gradient of chemoattractant needs to be set up. The Boyden chamber assay, was originally developed by Stephen Boyden for the analysis of leukocyte chemotaxis [6]. The assay is based on a chamber of two medium-filled compartments separated by a microporous membrane. Cells are seeded in the upper compartment and allowed to migrate through the pores of a membrane into the lower compartment comprising chemotactic molecules. After an appropriate incubation time depending on the cell type, the membrane between the two compartments is fixed and stained, and the cells that have migrated to the lower side of the membrane are counted. Therefore, the Boyden chamber-based cell migration assay has also been called a filter membrane migration assay, transwell migration assay, or chemotaxis assay. This type of transwell chemotactic assay is commonly used to assess primary

VEGF-A165-Driven Endothelial Cell Chemotaxis and Invasion

207

endothelial cell migration through a porous polyethylene terephtalate (PET) membrane that minimizes non-specific binding of compounds and small molecules. Membranes are available with different pore sizes and the diameter of the cell is to be taken into account when setting up the assay: smaller pore size results in a greater challenge for the migrating cell. Most eukaryotic animal cells measure between 10 and 30 μm in diameter and can readily migrate through 3–12 μm pores. It is to be noted that the transfilter gradient cannot be prolonged over a lengthy period of time (more than a few hours); therefore, it is necessary to optimize cell seeding density and gradient conditions to obtain meaningful and statistically relevant results (if only few cells migrate, it is likely to be associated with high variation). Additionally, the ability of endothelial cells to invade extracellular matrix can also be assessed via invasion assays. As mentioned earlier, the first step during sprouting angiogenesis relies on the partial degradation of the basement membrane on which the endothelium sits [7]. Similar to the chemotaxis assay, the endothelial cell invasion assay relies on a chemotactic gradient created across two compartments separated by a porous membrane. However, the invasive ability of the cells is tested by coating the inside of the transwell with an extracellular matrix solution that closely mimics the basement membrane of blood vessels. Endothelial cells, attracted by the chemoattractant at the bottom of the well, will degrade and invade the layer of extracellular matrix, migrate through the pores and remain at the lower side of the well where they will remain adherent. As for the transwell assay, after the appropriate incubation time, those cells are fixed, stained, and counted. The following protocols have been optimized for human umbilical vein endothelial cells (HUVEC) migrating toward a VEGF-A165 gradient but can be applied to other endothelial cell types and used with other chemoattractants (e.g., fibroblast growth factor, angiopoietins or sphingosine 1 phosphate). Additionally, cells can be genetically manipulated (using e.g., siRNA or adenovirus transduction systems) or pre-treated with chemical agents such as small molecule inhibitors. These approaches can be used to investigate the role of a particular signaling molecule or pathway, or for drug discovery screening purposes.

2

Materials 1. Human umbilical vein endothelial cells (HUVEC, up to passage 5) (see Note 1). 2. Endothelial basal medium (EBM, Lonza).

208

Caroline Pellet-Many

3. Penicillin-Streptomycin solution (10,000 units penicillin and 10 mg/mL streptomycin), sterile-filtered. 4. Fetal Bovine Serum (FBS). 5. Complete EBM: Basal EBM complemented with 10% FBS and Penicillin-Streptomycin (dilution of stock concentration 1: 100, final concentration 1%). 6. Cell culture grade 0.25% Trypsin-EDTA solution or non-enzymatic cell dissociation buffer (see Note 2). 7. Tissue culture grade sterile phosphate buffered saline (PBS; 140 mM NaCl, 3 mM KCl, 10 mM phosphate buffer, pH 7.4). 8. Human recombinant VEGF-A165 (100–20, Peprotech). 9. Tissue culture treated 24-well plates. 10. Transwell inserts for 24-well plate: clear PET membrane with 8 μm pore size (353,097, Corning). 11. Collagen stock solution (1 mg/mL) for migration assay (see Note 3). 12. Growth factor reduced Matrigel® or other basement membrane extract solution. 13. Reastain Quick-Diff Kit or Crystal Violet solution (0.5% in 25% Methanol solution diluted in water) (see Note 4). 14. Glass slides and cover slips. 15. DPX mounting medium. 16. Micro-pipettes and sterile pipette tips. 17. Forceps. The preparation and setup of the chambers is performed in a sterile tissue culture hood.

3

Methods

3.1 Chemotaxis Assay

1. HUVEC are grown to 80–90% confluence in complete EBM in a humidified cell culture incubator (37
C, 95% air/5% CO2). 2. Endothelial cells are serum-starved overnight before the assay in EBM supplemented with antibiotics and 0.5% FBS (see Note 5). 3. Coat the membranes of the transwells with 0.01% collagen overnight at 4
C (1/10 dilution of collagen stock solution in PBS, 0.75 mL at the bottom of a 24-well plate and 0.25 mL in the transwell). 4. The day after, the excess collagen solution is aspirated and transwells are left to dry in the culture hood, while cells are prepared for the assay (see Note 6).

VEGF-A165-Driven Endothelial Cell Chemotaxis and Invasion

209

5. Detach the cells from tissue culture flask using trypsin (0.25%) or a non-enzymatic cell dissociation buffer and resuspend the cells in complete medium. 6. Count the number of cells using an automated cell counter or hemocytometer. 7. Pellet the appropriate number of cells according to the number of transwells needed by centrifugation (50  g for 10 min), knowing that the final seeding density is 1  105 cells per insert. 8. Remove carefully complete medium and resuspend pelleted HUVEC in EBM supplemented with antibiotics and 0.5% FBS (1  105 cells per insert/0.5 mL). 9. While cells are being centrifuged, prepare medium with chemoattractant (VEGF-A125 to a final concentration of 25 ng/ mL) for the bottom of the well (750 μL per well). 10. Place the insert in the well containing the chemoattractant and immediately (see Note 7) seed 0.5 mL of cell solution containing a total of 1  105 cells inside the well and allow the cells to migrate for 4 h in a humidified tissue culture incubator (37
C, 95% air/5% CO2) (see Note 8). 11. After 4 h, gently swipe the upper surface of the transwell membranes with a cotton bud to remove non-migrated cells, be careful to leave the bottom of the membrane untouched as cells are easily dislodged (see Note 9). 12. Stain the transwell inserts with the Reastain Quick-Diff kit: fill a fresh 24 well plate with the three Reastain solutions (DiffQuik fixative reagent, a methanol-based fixing agent, the red Diff-Quik solution I, an eosinophilic stain and the blue DiffQuik solution II, an basophilic stain) and dip the previously swiped and rinsed inserts sequentially for around 1 min in each solution. 13. Finally rinse the transwells by dipping and agitating them in a container filled with double-distilled water and air-dry (see Note 10). 14. To continue with the analysis, cells can be counted within the insert under the microscope. Alternatively, the membrane can be carefully cut with a scalpel and mounted onto a glass slide with DPX and a coverslip for longer-term room temperature storage. 3.2

Invasion Assay

For the invasion assay, we modify the transwell migration assay described in Subheading 3.1 by coating the insert with Matrigel (or other ECM solution of your choice) (Fig. 1).

210

Caroline Pellet-Many

Fig. 1 Endothelial Cell Migration and Invasion assays workflow. Created with BioRender.com

1. Thaw the Matrigel on ice the night before the assay. 2. Precoat the inside of the well with 100 μL of a 1:5 Matrigel solution (Matrigel is diluted in cold EBM) (see Note 11). 3. Leave the Matrigel solution to polymerize for 1–2 h by placing the coated inserts in the 24-well plate inside a humidified tissue culture incubator (37
C, 95% air/5% CO2). 4. Discard the excess solution by pipetting or aspirating the side of the well, avoiding scratching the layer of matrix deposited at the bottom of the well.

VEGF-A165-Driven Endothelial Cell Chemotaxis and Invasion

211

5. Proceed to step 5 of Subheading 3.1. Follow similar steps as described in Subheading 3.1 but leave the endothelial cells to invade and migrate for 16–28 h instead. 3.3

Analysis

The cells that traversed the membrane can finally be counted under bright field microscopy (200 magnification) using an eyepieceindexed graticule. Cells can be counted manually or digitally. 1. Separate the area of the membrane into 4 equal parts using a fine permanent marker pen. 2. Manually count 8 fields per insert (2 per quadrats) (see Note 12). 3. Calculate the mean and use this as the count for that insert. 4. For digital counting, capture 8 fields per insert as digital images. 5. Analyze with Fiji using the cell counter tool in Fiji (see Note 13). The following images and bar graph were obtained following HUVEC migration toward a VEGF-A125 gradient and shows that chemotaxis is dependent on VEGF-A165 signaling via the B1 domain of Neuropilin 1 (Fig. 2) [8]. For this specific experiment, cells were previously infected for 48 h with adenoviral vectors encoding for green fluorescent protein (GFP) as control, wildtype Neuropilin 1, a co-receptor for VEGF (NRP1 WT) or a dominant negative form of NRP1 with a single point mutation within the VEGF-binding domain of NRP1 (B1 mutant). Figure 2a shows HUVEC migrated to the bottom of the transwell after 4 h and Fig. 2b represents the quantification of the migrated HUVEC per microscopic field.

4

Notes 1. This assay has been used for studying the migration of many primary cells and cell lines. However, depending on cell type, different seeding densities may be used (between 5  104 to 106 cells per insert), the chemoattractant concentration can also be adjusted (for example, for VEGF-A165, between 10 ng/mL and 100 ng/mL) as well as the timing of the assay (the longer the cells are left to migrate, the more will do so and subtle differences between control and treated cells might be lost if the assay is left to run for too long). 2. Depending on the cell line and/or the signaling pathway to be investigated, the use of a non-enzymatic dissociation buffer is preferred as trypsin might result in the cleavage of cell-surface receptors.

212

Caroline Pellet-Many

Fig. 2 (a) Photos of microscopic fields showing migrated HUVEC at the bottom of the PET membrane. A clear dark purple staining shows nuclei of migrated HUVEC that are counted. Eight similar fields are counted per insert at 200 magnification. (b) Bar graph representing the number of migrated HUVEC per microscopic field (average of 3 experiments, treatment performed as duplicates) in response to VEGF-A165 and in function of adenoviral overexpression. The overexpression of the dominant negative mutant NRP1 mutant with a non-functional VEGF-binding domain results in impaired migration of HUVEC in response to VEGF-A165

3. Collagen solution (1 mg/mL) should be diluted tenfold with sterile water to obtain a working concentration of 0.01%. 4. Reastain Quick-Diff Kit comprises three solutions including a fixative, a red (Xanthene dye (Eosin Y)) and blue (Thiazine dye, methylene blue and azure A) staining solutions. It is relatively expensive therefore, as an alternative, a simple Crystal violet solution can also be used to stain the cells. Once reconstituted,

VEGF-A165-Driven Endothelial Cell Chemotaxis and Invasion

213

the Crystal Violet solution should be stored as a solution in the dark and used within 2 months. 5. Serum-starving the cells for 5 h in EBM supplemented with antibiotics and 0.1% FBS can be done as an alternative. 6. Use sterile forceps to grab the upper side of the transwell, where it is slightly larger and pour excess collagen back in well or aspirate the solution inside the well. Leave to dry bottom-side up on paper while preparing the cells for seeding. 7. You may add the solution with the chemoattractant to the bottom of all the wells and then add the inserts one by one and immediately seed the cells. Alternatively, all inserts can be added to the 24-well plate and the solution with the chemoattractant can be added subsequently by inserting the pipette carefully to the side of the insert, taking care of not introducing any air bubble. 8. With this setup, only a few endothelial cells are able to migrate through the 8 μm pores without the stimulation of VEGFA165. Ideally, there should be a minimum of threefold increase in migration, with numbers of cells high enough so there is little variation between inserts. These conditions need to be established beforehand depending on the cell type and the experimental design. 9. With one hand hold the upper side of the transwell with forceps and pour the medium inside the well into liquid waste disposal. Wash transwell insert by dipping in a large beaker filled with sterile 1 PBS at room temperature to remove debris and non-attached cells. Carefully holding the insert with your fingers, taking care not to touch the bottom of the well, wipe the inside of the transwell with the wet tip of the cotton swab (use PBS). Be consistent in your method, apply the same pressure and wipe for the same length of time (five clockwise and five anticlockwise movements are enough to remove all non-migrated cells). 10. Alternatively, fixing with 70% Ethanol or Cold Methanol into a well of a 24-well plate or 0.2% crystal violet or hematoxylin stain can also be used (10–15 min incubation at room temperature). 11. It is recommended to use pre-cooled pipette tips (by keeping them on ice for example) to obtain optimal delivery of Matrigel to the insert. Matrigel will polymerize quickly at room temperature. Additionally, we found that cutting the tip of the 200 μL pipette tip will help with taking up the viscous solution. 12. For manual counting, by separating the membrane in 4 equal quadrats, we avoid counting the same microscopic field twice

214

Caroline Pellet-Many

and cover a more complete and representative area of the membrane. 13. Cells can be counted directly on dry inserts held inside the 24-well plate or membranes can be gently cut away from the insert with a scalpel and mounted on glass slides with DPX mounting medium for longer storage at room temperature and imaging. Alternatively, digital images can be taken and analyzed with Fiji (http://fiji.sc/). The cell counter tool (https:// imagej.nih.gov/ij/plugins/cell-counter.html) can be used to count the migrated cells and to avoid counting the same cell twice as a marker is placed on the cells that were counted. Once the plugin has been downloaded, restart Fiji/ImageJ and open your image (no need to change to grayscale). Go to Plugins ! Analyze ! cell counter, a new window will appear. Click “Initialize” and then select “Type 1” (we are only counting one type of cells). You will be prompted to choose your marker option (find a color that contrast well for better visualization). Start clicking on the nuclei of all migrated cells, those will be counted and will appear in the counter frame within the “Cell Counter” window, next to “Type 1”. By clicking “results” a table will be generated and numbers can then be exported/ saved for statistical analysis using the program of your choice. You can also export the image with the markers added for future reference. References 1. Risau W (1997) Mechanisms of angiogenesis. Nature 386(6626):671–674. https://doi.org/ 10.1038/386671a0 2. Lamalice L, Le Boeuf F, Huot J (2007) Endothelial cell migration during angiogenesis. Circ Res 100(6):782–794. https://doi.org/10. 1161/01.RES.0000259593.07661.1e 3. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N (1989) Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 246(4935):1306–1309. https://doi. org/10.1126/science.2479986 4. Neufeld G, Cohen T, Gitay-Goren H, Poltorak Z, Tessler S, Sharon R, Gengrinovitch S, Levi BZ (1996) Similarities and differences between the vascular endothelial growth factor (VEGF) splice variants. Cancer Metastasis Rev 15(2):153–158. https://doi. org/10.1007/BF00437467

5. Neufeld G, Cohen T, Gengrinovitch S, Poltorak Z (1999) Vascular endothelial growth factor (VEGF) and its receptors. FASEB J 13(1):9–22 6. Boyden S (1962) The chemotactic effect of mixtures of antibody and antigen on polymorphonuclear leucocytes. J Exp Med 115:453–466. https://doi.org/10.1084/jem.115.3.453 7. Davis GE, Senger DR (2005) Endothelial extracellular matrix: biosynthesis, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ Res 97(11): 1093–1107. https://doi.org/10.1161/01. RES.0000191547.64391.e3 8. Herzog B, Pellet-Many C, Britton G, Hartzoulakis B, Zachary IC (2011) VEGF binding to NRP1 is essential for VEGF stimulation of endothelial cell migration, complex formation between NRP1 and VEGFR2, and signaling via FAK Tyr407 phosphorylation. Mol Biol Cell 22(15):2766–2776. https://doi.org/10.1091/ mbc.E09-12-1061

Chapter 15 Measuring Mitochondrial Calcium Fluxes in Cardiomyocytes upon Mechanical Stretch-Induced Hypertrophy Daniela Ramaccini, Carlotta Giorgi, and Michelle L. Matter Abstract Calcium Ca2+ regulation is a key component of numerous cellular functions. In cardiomyocytes, Ca2+ regulates excitation-contraction coupling and influences signaling cascades involved in cell metabolism and cell survival. Prolonged dysregulation of mitochondrial Ca2+ leads to dysfunctional cardiomyocytes, apoptosis and ultimately heart failure. VEGF promotes cardiomyocyte contractility by increasing calcium transients to control the strength of the heartbeat. Here, we describe a method to measure mitochondrial Ca2+ fluxes in human ventricular cardiomocytes after inducing stretch-mediated hypertrophy in vitro. Key words Cardiomyocyte, Cyclic stretch, Hypertrophy, AC16, Calcium, Mechanotransduction

1

Introduction Cardiac function is dependent upon mitochondrial Ca2+ fluxes for ATP production, mitochondrial metabolism, muscle contraction and cell survival [1]. In cardiomyocytes, mitochondria are located near the sarcoplasmic reticulum (SR) that acts as a Ca2+ reservoir. SR release of Ca2+ is rapidly taken up by mitochondria to promote mitochondrial ATP synthesis activation to regulate cellular energy metabolism [2]. Prolonged changes in Ca2+ flux may lead to abnormal mitochondrial activity thereby promoting decreased energy production, changes in reactive oxygen species (ROS) and cardiomyocyte apoptosis [3], all of which contribute to cardiovascular disease [4]. Cardiomyocytes subjected to in vitro mechanical stretch recapitulate hypertrophic cellular responses in vivo [5]. We previously reported that cyclic mechanical stretch-induced hypertrophy of primary adult rat cardiomyocytes promotes vascular endothelial growth factor (VEGF) secretion by upregulating NFkB signaling [6, 7]. In cultured rat cardiomyocytes VEGF signals to regulate cardiomyocyte contractility through modifying Ca2+ flux

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_15, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

215

216

Daniela Ramaccini et al.

Fig. 1 Calcium Flux analysis after cyclic mechanical stretch: human ventricular cardiomyocytes are plated on flexible-bottom 6-well plates where each well has been coated with laminin. Cells are allowed to attach to the matrix and then infected with a mitochondrial specific Aequorin probe. Cells are then subjected to mechanical stretch at an extension level of 10% at 30 cycles/min to induce hypertrophy. At various time points cells are trypsinized and replated onto laminin-coated black 96-well plates for 4 h. Each well is then lyzed and mixed with 5 mM 100
native coelenterazine for 1.5 h. To measure calcium levels, plates are read on luminescence microplate reader.

[8]. Here, we describe a method to measure mitochondrial calcium changes in cardiomyocytes that have undergone hypertrophy through mechanical force in vitro (Fig. 1).

2

Materials All solutions are for tissue culture use and should be prepared using sterilized ultrapure water.

2.1

Cells

1. AC16 cells (see Note 1) are used at a density of 0.5
106 cells/ mL. 2. AC16 culture medium: Sterile DMEM/F12 containing 2 mM L-Glutamine, 12.5% FBS and 1
PenicillinStreptomycin Solution (see Note 1).

Stretch-Induced Mitochondrial Calcium Flux Analysis

217

3. Phosphate-buffered saline (PBS). 4. Trypsin. 5. Laminin membrane coating solution: Laminin (15 μg/mL final concentration diluted in PBS). For 6-well plates: Prepare 12 mL Laminin solution (2 mL/well): to 11.82 mL 1
PBS add e.g., 180 μL laminin from a stock solution of 1 mg/mL. Mix by inverting tube. 2.2 Cyclic Mechanical Stretch

1. Flexcell FX-4000 (V4.0) Tension system (Flexcell® International Corporation, Burlington, NC). 2. 6-well flexible-bottom tissue culture plates (BioFlex UF-400 IU; (Flexcell® International Corporation, Burlington, NC). 3. A CO2 incubator with 5% CO2 is required for culturing cells during mechanotransduction FlexCell experiments.

2.3 Mitochondrial Calcium Measurements

1. Adenovirus mitochondrial aequorin wild type (Ad-mtAEQwt) probe, kindly gifted by Prof. Pinton P. at the University of Ferrara, Italy who generated this probe. The stock is concentrated 5,11
1010 PFU/mL. 2. Laminin-coated 96-well black flat bottom plates and two clear 96-well flat bottom plates. 3. Krebs Ringer buffer (KRB): Prepare as below and adjust pH to 7.4 with NaOH. Store at 4  C for a maximum of 3 days. MW (g/mol)

g in 1 L

Final concentration (mM)

NaCl

58.44

7.89

135

KCl

74.55

0.373

5

KH2PO4

136.09

0.014

0.4

MgSO4

120.37

0.120

1

HEPES

238.30

4.76

20

Glucose

180.16

1

5.5

4. Modified KRB: Add 0.5 mL of 1 M CaCl2 to 500 mL of KRB for a final CaCl2 concentration of 1 mM. This solution may be stored at 4  C for 3 days (see Note 2). 5. 2
Ca2+ mobilizing solution (CMS): For AC16 cells the agonist required consists of a combination of Bradykinin:Histamine (1:1 ratio) both at a final concentration of 100 μM, dissolved in freshly prepared modified KRB (see Note 3). 6. 4
Lysis solution: 400 mL of Triton X-100 and 1 mL of 1 M CaCl2 mixed in 100 mL Milli-Q water. Mix by inverting and store at 4  C for up to 1 month (see Note 4).

218

Daniela Ramaccini et al.

7. Native coelenterazine solution 100
powder: dissolve in pure Methanol for a final concentration of the stock solution of 500 mM (see Note 5). 8. Luminescent microplate reader.

3 3.1

Methods Cyclic Stretch

1. Prepare laminin solution to coat 6-well flexible-bottom stretch plates, non-stretch control plates and black flat bottom 96-well plates (see Note 6). 2. Vortex to mix the solution. Add 2 mL laminin solution to each well of the 6-well plates and 100 μL to each well of the 96-well plates. Be sure to cover the entire surface by gently rocking each plate. 3. Incubate at 37  C for 1 h and wash twice with 1
PBS. 4. Air dry in a sterile tissue culture hood. Plates may now be stored at 4  C for up to 1 week, sealed with parafilm to avoid contamination. 5. Plate 1.0
106 AC16 cells per 2 mL of medium into each well of the laminin-coated 6-well flexible-bottom tissue culture plates and into each well of non-stretch control 6-well plates (see Note 7).

3.2 Probe Infection to Measure Ca2+ Flux

1. Infect cells with the Ad-mtAEQwt, which allows for the measurement of mitochondrial Ca2+. Prepare a mix for the total number of wells considering 1 μL of adenovirus per 2 mL of normal growth media, for a final concentration of 5.0
104 PFU/mL. Begin measurements at 36–48 h post infection (see Note 8). 2. 24 h after infection remove media from each well and replace with fresh media. 3. Insert plates into the Flexcell FX 4000 plate holder. Move plate holder apparatus into the incubator at 5% CO2 and 37  C. Place non-stretch control plates in the same incubator but not in the Flexcell FX plate holder. 4. Set parameters on the Flexcell FX-4000 to simulate a hypertrophic heart waveform at an extension level of 10%, 30 cycles/ min for 24 h [7] (see Note 9). 5. Non-stretched cardiomyocytes are cultured as control under the identical experimental conditions and in the same incubator but not exposed to cyclic stretch. 6. At 24 h timepoint release the cells by trypsinization: remove media and wash cells with 1 mL/well of PBS. 7. Remove PBS and add 500 μL/well of trypsin 1
.

Stretch-Induced Mitochondrial Calcium Flux Analysis

219

8. Incubate plate for 5 min at 37  C with 5% CO2. 9. Collect cells with fresh media into 5 mL conical tubes and centrifuge for 3 min at 200
g. Resuspend cells in 3 mL of fresh media. Seed cells onto the laminin-coated 96-well black flat bottom plates at 40,000 cells/well. 10. Incubate cells for 4 h at 37  C with 5% CO2. 11. Wash cells once with 100 μL modified KRB. 12. Add to each well 5 μM 100
native coelenterazine (50 μL/ well) and incubate for 1.5 h at 37  C with 5% CO2 (see Note 10). 13. During incubation, prepare two different 96-well plates: one with 50–100 μL of 2
CMS per well; the other with 100–200 μL per well of 4
lysis buffer. 14. Carefully wash cells twice with 100 μL modified KRB. 15. Place 96-well plates into the plate reader and begin recording luminescent values at 30 s, after stimuli and after lysis (see Note 11). (a) Record Basal levels for at least 30 s. (b) Inject 50 μL 2
CMS, record for 2 min. (c) Inject 100 μL 4
Lysis solution, record for 2 min. 16. Export results into an Excel-compatible format. After subtraction of the background measured at the start of the measurement (Basal Levels), luminescent values may be converted into Ca2+ concentration by using the equation below [9].   n1  n1 L L þ
K TR  1 L max
λ L max
λ   Ca2þ ðM Þ ¼ n1 L KR 
KR L max
λ L ¼ light intensity at sampling time. Lmax ¼ total light emitted at sampling time. KR ¼ constant for calcium-bound state. KTR ¼ constant for calcium-unbound state. λ ¼ rate constant for aequorin consumption at saturating [Ca2+]. n ¼ number of Ca2 + binding site.

220

4

Daniela Ramaccini et al.

Notes 1. The AC16 cell line is a human ventricular cardiomyocyte cell line that recently became commercially available and is suitable for cardiac biology and function studies. This cell line proliferates in culture and differentiates when cultured in mitogen free media. These cells allow for the evaluation of cardiomyocyte function. We stipulate the specific media required for these AC16 cells and also note that these experiments require phenol red free media. It is also possible to use cardiac human induced pluripoptent stem cells for these experiments. 2. Ca2+ may bind to the glass, therefore the solution should be prepared in a polypropylene tube in order to avoid inaccuracies during experiments due to reduced Ca2+ concentration. 3. The concentration of the stimuli should be doubled because during injection it will be diluted in the volume already present within each well (in our case 50 μL per well of modified KRB). In order to be sure that the plate reader will collect and inject the same equal volume (in our case 50 μL per well), and thus the same concentration of stimuli into each well, we suggest making a total solution that is double the volume per well. We prepare for a 96-well plate a solution of 10 mL CMS 2
, and we use a multichannel pipette set at 100 μL. This will ensure the injection of the same volume and concentration into each well throughout the experiment. If using different cell types, choose the appropriate agonist. Common Ca2+ perturbating agents can be found in the published protocol [9]. 4. For the 4
Lysis solution: prepare a 4
solution after considering the volume injected and the dilution in each well. In our case, after administration of the 2
CMS we have 100 μL of volume per well and we set the injection of the lysis solution at 100 μL. Therefore, we always prepare a 4
lysis solution and we add 200 μL per well. 5. Coelenterazine is extremely sensitive to light, therefore it should be stored in small aliquots at 80  C, sealed with parafilm, in order to avoid methanol evaporation, and covered with aluminum foil to protect from light. For use either take out an aliquot from the freezer or prepare a fresh mix solution and add 50 μL per well. Always keep the plate covered with aluminum foil to block light. 6. It is important to use the black flat bottom 96-well plates, which allow for more precise accurate readings from each well. 7. 6-well flexible-bottom stretch plates will also be used to plate cells for non-stretch controls.

Stretch-Induced Mitochondrial Calcium Flux Analysis

221

8. We use the aequorin chimera specific for mitochondrial calcium measurements. There are other organelle specific probes [10]. To infect the aequorin probe pipette into cells with media. 9. The flex-bottom plates consist of a membrane that will be stretched upon attaching a vacuum system to the plate holder. In this manner the membrane can be deformed to represent healthy or hypertrophic heart function based upon the number of cycles, amplitude, and frequency of the waveform of strain, which are available in the Flexwell software. 10. It is important to always work in the dark and cover plates with aluminum foil due to coelenterazine’s sensitivity to light. The aequorin active form is obtained after the addition of the prosthetic group, coelenterazine. When calcium binds the 3 high-affinity Ca2+ binding sites of the aequorin, a photon is emitted [10]. 11. Set reading time at 1000 ms. Mitochondrial calcium uptake and release is a fast and dynamic event of about 100 ns so that slower reading time will not allow for analysis of the Ca2+ kinetics.

Acknowledgments M.L. Matter is supported by a grant from the National Institutes of Health (R01HD091162). C. Giorgi is supported by the Italian Association for Cancer Research (IG-19803) and the Progetti di Rilevante Interesse Nazionale (PRIN20177E9EPY). C. Giorgi is grateful for local funds from University of Ferrara and A-ROSE. References 1. Walsh C, Barrow S, Voronina S et al (2009) Modulation of calcium signalling by mitochondria. Biochim Biophys Acta Bioenerg 1787: 1374–1382. https://doi.org/10.1016/j. bbabio.2009.01.007 2. Frederick RL, Shaw JM (2007) Moving mitochondria: establishing distribution of an essential organelle. Traffic 8:1668–1675. https:// doi.org/10.1111/j.1600-0854.2007. 00644.x 3. Balaban RS, Bose S, French SA et al (2003) Role of calcium in metabolic signaling between cardiac sarcoplasmic reticulum and mitochondria in vitro. Am J Physiol Cell Physiol 284: C285–C293. https://doi.org/10.1152/ ajpcell.00129.2002 4. Ramaccini D, Montoya-Uribe V, Aan F et al (2021) Mitochondrial function and

dysfunction in cardiomyopathy. Front Cell Dev Biol 8:624216 https://doi.org/10. 3389/fcell.2020.624216 5. Sadoshima J, Izumo S (1993) Mechanotransduction in stretch-induced hypertrophy of cardiac myocytes. J Recept Res 13(1–4):777–794. h t t p s : // d o i . o r g / 1 0 . 3 1 0 9 / 107998993090736 6. Leychenko A, Konorev E, Jijiwa M et al (2011) Stretch-induced hypertrophy activates NFkBmediated VEGF secretion in adult cardiomyocytes. PLoS One 6(12):e29055. https://doi. org/10.1371/journal.pone.0029055 7. Matter ML (2015) Induction of VEGF secretion in cardiomyoctes by mechanical stretch. Methods Mol Biol 1332:67–74. https://doi. org/10.1007/978-1-4939-2917-7_5

222

Daniela Ramaccini et al.

8. Rottbauer W, Just S, Wessels G et al (2005) VEGF–PLCγ1 pathway controls cardiac contractility in the embryonic heart. Genes Dev 19:1624–1634. https://doi.org/10.1101/ gad.1319405 9. Kim HK, Kang YG, Jeong SH et al (2018) Cyclic stretch increases mitochondrial biogenesis in a cardiac cell line. Biochem Biophys Res

Commun 505(3):768–774. https://doi.org/ 10.1016/j.bbrc.2018.10.003 10. Bonora M, Giorgi C, Bononi A et al (2013) Subcellular calcium measurements in mammalian cells using jellyfish photoprotein aequorinbased probes. Nat Protoc 8:2015–2118. https://doi.org/10.1038/nprot.2013.127

Chapter 16 Simultaneous Measurement of Endothelial Cell Proliferation and Cell Cycle Stage Using Flow Cytometry Eleanor G. Lynam, James A. E. Lane, Elizabeth J. T. Finding, and Caroline P. D. Wheeler-Jones Abstract Endothelial cell proliferation rate is an important indicator of vascular health. Being able to detect the rate of endothelial cell proliferation, or cell cycle disturbances after intervention is a valuable tool for analysing any beneficial or detrimental effects of treatments in vitro. Here, we describe a straightforward flow cytometric-based method of proliferation and cell cycle tracking that can be performed on human endothelial cells in culture over several days. Key words Flow cytometry, Proliferation, Propidium Iodide, Cell cycle, VEGF

1

Introduction The proliferation of endothelial cells is essential for angiogenesis and vascular repair. Understanding how certain interventions affect endothelial cell proliferation is crucial for developing treatments for cardiovascular diseases and tumorigenesis. Many methods of proliferation measurement rely on determining the number of cells in a sample or the total DNA content of a sample. The interpretation of such results can be skewed by cells that have gone, or are going through apoptosis and/or necrosis and can be misleading, depending on where in the cell cycle the cells in a sample are arresting. Another commonly employed method of proliferation measurement uses bromodeoxyuridine (BrdU) incorporation, as BrdU is an analogue of thymidine [1]. This method, however, relies on tagging replicating DNA and will give misleading results if cells are becoming arrested after the S-phase of the cell cycle [2]. The simple method described here allows for the measurement of in vitro endothelial cell proliferation over several days and gives the opportunity to examine the percentage of cells in G0/G1, S or G2/M phases of the cell cycle [3]. This allows the user to identify

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_16, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

223

224

Eleanor G. Lynam et al.

where cells may be arresting and therefore implicate pathways that may be altered by an intervention. Carboxyfluorescein succinimidyl ester (CFSE) is a fluorescent dye which enters the cell, binds to intracellular molecules such as lysine residues and is retained within the cell during proliferation. Its signal can be tracked using flow cytometry to determine whether cells have undergone proliferation, as each generation of cells harbors less fluorescence than the previous generation [4, 5]. When this method is combined with a DNA dye, such as propidium iodide (PI), the DNA content per cell can be measured, giving an estimation of where the cell lies in the cell cycle [2, 5]. With these combined methods, it is possible to determine if cells in culture have been proliferating, and if they have not, where in the cycle the cells are becoming arrested. Therefore, this method gives the user a clearer idea of the molecular pathways that may be affected in a specific experimental setting.

2

Materials 1. Complete M199 growth medium: M199 medium. 20 μg/mL endothelial cell growth factor (Sigma-Aldrich, catalogue number E1388). 20% (v/v) Fetal Bovine Serum (FBS). Penicillin/streptomycin (100 Units/mL). 2. Reduced M199 growth medium: M199 medium. 1% (v/v) FBS. Penicillin/streptomycin (100 Units/mL). 3. Human Dermal Microvascular Endothelial Cells. 4. Gelatin solution. 5. CSFE (Thermo Fisher Scientific, catalogue number: C34554). 6. Recombinant Human VEGF-A (R & D systems, catalogue number: 293-VE-010). 7. Paraformaldehyde (PFA). 8. 0.05% Trypsin-EDTA. 9. Phosphate buffered saline (PBS). 10. Ethanol (EtOH). 11. RNase A. 12. Tween 20. 13. Triton X-100.

Simultaneous Measurement of Endothelial Cell Proliferation and Cell Cycle. . .

225

14. Bovine Serum Albumin (BSA). 15. Propidium iodide (PI) (Sigma-Aldrich, catalogue number: P4170).

3

Methods 1. Coat the wells of a 6-well tissue culture plate (see Note 1) with 1 mL of a 1% gelatin solution in PBS. 2. Aspirate the excess gelatin and immediately plate endothelial cells at a density of 200,000 cells per well in 1.5 mL (see Note 1) of complete M199 growth medium.

3.1 Carboxyfluorescein Succinimidyl Ester (CFSE) Staining

1. The following day (see Note 2) label the cells with CFSE. Dilute the CSFE in pre-warmed PBS, adding a volume of 1 mL per well with a final concentration of 1 μM (see Note 3) for 20 min in a humidified tissue culture incubator (37  C/5% CO2). 2. After incubation wash cells twice with complete M199 growth medium and add fresh M199 growth medium (see Note 4). 3. To generate a reading of peak fluorescence control to compare to the experimental sample cell peaks, a well of non-proliferated cells or “baseline” cells should be used. This “baseline” sample should be collected 10 min after addition of the dye to allow time for dye incorporation but not for proliferation to occur (see Note 5). 4. Collect baseline cells as described in step 6 and keep fixed and washed baseline cells at 4  C in the dark until analysis. 5. After the cells have been incubated with CSFE, the varying experimental treatments (e.g., VEGF 10–25 ng/mL diluted in reduced M199 growth medium) can be added (see Note 6). 6. Collect cells after desired experimental treatment duration (see Note 7). Collection of cells: (a) Wash the monolayers twice with pre-warmed PBS and then add 1 mL of 0.05% Trypsin-EDTA per well of a 6-well plate for 2 min in a humidified tissue culture incubator (37  C/5% CO2). (b) After 2 min, pipette the trypsin solution over the well to remove any remaining adherent cells. (c) Transfer the samples to 15 mL falcon tubes containing 2 mL of complete M199 growth medium. (d) Centrifuge samples at 180  g for 5 min.

226

Eleanor G. Lynam et al.

(e) Wash pellets twice by resuspending in 1 mL of room temperature PBS and re-centrifuging at 180  g for 5 min. (f) Resuspend cells in 2 mL of 3% PFA (diluted in PBS) and leave for 15 min at room temperature. 3.2 Propidium Iodide Staining and Flow Cytometric Analysis

To perform cell cycle analysis of the samples, the following propidium iodide (PI) staining protocol is used (see Note 8). 1. Centrifuge samples at 180  g for 5 min. 2. Re-suspend pellets in 350 μL of 50 μg/mL propidium iodide plus 50 μL of 10 mg/mL RNAse A. 3. Transfer samples to clean flow cytometry tubes. 4. Run and record samples through a flow cytometer. 5. Gate samples for whole, single cells using forward versus side scatter profile. 6. For CFSE expression use 488 nm excitation and emission filters and plot signal strength over cell number (Fig. 1). 7. For PI signal, use excitation at 535 nm and fluorescence emission maximum at 617 nm. Cells should be compared by PI signal on the x-axis and cell count on the y-axis (Fig. 2). Present PI data as proportion of cells in different stages of the cycle as a percentage of cells in G0/1, S phase, or M/G2 phase according to PI signal.

4

Notes 1. The endothelial cell density and type of plate can be altered, bearing in mind that samples containing low numbers of cells are not best suited for this kind of analysis. 2. Cells can be stained with CFSE as soon as they are attached, so the timing of the staining can be altered to best suit the experiment. 3. If the CFSE signal is too weak or too strong the concentration of CFSE may need to be optimized for different experimental conditions. 4. It is important after adding CFSE to monitor the health of the cells daily as they may react to the dye and high concentrations of CFSE are cytotoxic. 5. Collect a well of cells that have not received CFSE so that there is a positive and negative control when optimizing flow cytometry settings.

Simultaneous Measurement of Endothelial Cell Proliferation and Cell Cycle. . .

227

Fig. 1 Flow cytometric analysis of the amount of proliferation in an in vitro sample after a defined experimental time. Created using BioRender. Step 1: Dilute the CSFE in pre-warmed PBS, adding a volume of 1 mL per well to incorporate the CFSE into the cells. Step 2: After staining, cells are cultured in reduced growth medium with or without growth factor or other molecules being tested, according to the specific experimental design. During this time cells will divide and fluorescence will be halved following each cell division. Step 3: Following the period specified by the experimental design (depending on the specific factor used for stimulation), fix cells and stain with propidium iodide. Proceed to flow cytometric analysis. Step 4: Example of signal from CFSE in sample cells (green) compared to VEGF-treated cells (red) and baseline cells (blue). Baseline cells show one strong peak of fluorescence as all of the cells in the sample have not had a chance to replicate. The sample cells depict a pattern of normal endothelial division with different populations within the sample at different peaks in fluorescence, according to the number of divisions the cells have been through. The VEGF-treated cells will have the biggest peak of cells at low fluorescence, demonstrating that the majority of the cells have been through multiple rounds of division

6. To measure proliferation rate using VEGF as a positive control, VEGF (25 ng/mL) diluted in serum-reduced medium is added to the cells after a 1 hour period of serum deprivation (1% serum). 7. Collect cells to perform CFSE proliferation analysis between 1 and 4 days post-labeling. Signal may fade to below detectable levels after long periods so this should be taken into account when considering assay duration. 8. It is important to have enough cells per sample (>10,000) during PI staining so that accurate and statistically relevant percentages may be recorded.

228

Eleanor G. Lynam et al.

Fig. 2 Stages of the cell cycle with corresponding graph representing propidium iodide intensity at different stages of the cell cycle. Created using BioRender. As PI dyes DNA content, the number of chromosomes affects the signal strength per cell. Cells in G1/0 will have one set of chromosomes only and so will show as one peak in the fluorescence signal by flow cytometry. As cells are preparing for mitosis the amount of DNA will increase and so the PI signal will be more than that of the cells peaked at G1/G0. Once cells have two complete sets of chromosomes ready for division they will have roughly double the PI signal of the cells in the G1/0 peak References 1. Mokry´ J, Nĕmecek S (1995) Immunohistochemical detection of proliferative cells. Sb Ved Pr Lek Fak Karlovy Univerzity Hradci Kralove 38(3):107–113 2. Fleisig H, Wong J (2012) Measuring cell cycle progression kinetics with metabolic labeling and flow cytometry. J Vis Exp (63):e4045 3. Kawamoto K, Herz F, Wolley RC, Hirano A, Koss LG (1980) Flow cytometric analysis of the DNA content in cultured human brain tumor

cells. Virchows Arch B Cell Pathol Incl Mol Pathol 35(1):11–17 4. Lyons AB, Parish CR (1994) Determination of lymphocyte division by flow cytometry. J Immunol Methods 171(1):131–137 5. Ueckert JE, Nebe von-Caron G, Bos AP, ter Steeg PF. (1997) Flow cytometric analysis of lactobacillus plantarum to monitor lag times, cell division and injury. Lett Appl Microbiol 25(4):295–299

Chapter 17 Ex Vivo Mouse Aortic Ring Angiogenesis Assay Vedanta Mehta and Marwa Mahmoud Abstract The ex vivo aortic ring assay is one of the most widely used protocols to study sprouting angiogenesis. It is a highly adaptable method that can be utilized to investigate the effects of different growth factors, smallmolecule drugs, and genetic modifications on vascular sprouting in a physiologically relevant setting. In this chapter we describe a simple and optimized protocol for investigating vascular sprouting in the mouse aortic ring model. The protocol describes the harvesting and embedding of the aortic rings in a collagen matrix, treatment of the rings with agents of interest, and the visualization and quantification of the vascular sprouts. Key words Angiogenesis, Ex vivo, Aortic ring, Vascular sprouting, Mouse

1

Introduction Angiogenesis refers to the growth of new vessels from pre-existing vasculature via either intussusceptive growth or sprouting [1]. Angiogenesis is key in embryonic and postnatal vascular growth, and in pathological angiogenesis, where it is important for establishing and maintaining adequate tissue perfusion. It is a complex, tightly regulated, process which is balanced by an interplay of proteins which act either to stimulate (e.g., vascular endothelial growth factor (VEGF)) or to inhibit (e.g., thrombospondins and endostatin) blood vessel growth [2]. A deficiency in angiogenesis resulting in insufficient vascular growth, maintenance and pathological malformations can cause severe ischemic diseases, including myocardial disease and neurodegenerative disorders, while exaggerated vascular growth is associated with cancer and inflammatory disorders [3]. Many experimental assays have been established to decipher the mechanisms regulating angiogenesis, with the aim of developing therapies that can control angiogenesis to achieve favorable tissue regeneration or inhibit exaggerated vascular growth.

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_17, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

229

230

Vedanta Mehta and Marwa Mahmoud

An obvious advantage of examining angiogenesis in vivo is that it allows the study of the contribution of several cell types in a physiologically relevant environment which includes the effects of blood flow. Some of the most commonly used in vivo angiogenesis assays include the Matrigel implant assay and the chick embryo chorioallantoic membrane (CAM) assay [4–7]. However, in vivo assays are relatively expensive and the results are often difficult to decipher due to the involvement of inflammation and the difficulty in discriminating the responses of the cells that make up the blood vessels versus the paracrine effects of other surrounding tissues/ cells. On the other hand, in vitro assays which use isolated endothelial cells only allow the study of the endothelial cell response, and paracrine interactions between endothelial cells and other perivascular cells such as smooth muscle cells, pericytes, and fibroblasts, are not taken into account. Thus the data generated using these assays are limited and not very “translatable” or physiologically relevant. Ex vivo assays were developed to bridge the gap between in vivo and in vitro models, providing a more physiologically relevant model by allowing the study of a mixed population of native cells that interact through paracrine mechanisms in a chemically defined environment. One of the most widely used ex vivo angiogenesis assays is the aortic ring model. First established in 1990 by Nicosia and Ottinetti using rat tissue [8, 9], it has now been adapted to other species including: mouse, chicken, and porcine tissue [10– 13]. This assay allows easy quantification of vascular sprouting, the examination of intercellular interactions with accessory supportive cells, excludes inflammatory components, and is fast and inexpensive to establish. This assay is also highly adaptable, allowing the study of the effects of different conditions on vascular sprouting. Recent adaptations to the aortic ring assay have included co-culturing human mesenchymal stromal cells (MSCs) with developing rat aortic ring cultures to study the contribution of the MSC population to vascular sprout formation and growth [14, 15]. The aortic ring assay has also been adapted to study sprouting angiogenesis in the context of atherosclerosis, using tissue from transgenic mouse models and human plaques [16, 17]. Herein we describe a simplified and optimized procedure for the mouse aortic ring angiogenesis assay which can be utilized to examine the effect of different conditions, such as different growth factors, small-molecule drugs, and genetic manipulations (through the use of transgenic mice). This assay provides a quick and inexpensive method to study the growth of microvessels in a physiologically relevant setting, which encompasses the involvement of supporting cells and the formation of tubule structures which develop over a time course similar to in vivo conditions. The assay comprises three main steps, the first of which is “harvesting the aorta and preparing the rings” which is followed by “embedding

Mouse Aortic Ring Assay

231

Fig. 1 Images of aortic rings embedded in a collagen gel and treated with VEGF-A165 (30 ng/mL) for 7 days in the presence or absence of Marimastat (which inhibits angiogenesis). Rings were stained with Dylight 594 labeled Isolectin B4 and imaged on a fluorescent microscope

the rings and treatment with agents of interest.” The aortic rings are embedded in a collagen matrix which allows three-dimensional growth of the developing vascular sprouts and the evaluation of key steps of angiogenesis including matrix degradation and cell migration and proliferation following the addition of growth factors or agents such as siRNAs for modulation of gene expression (Fig. 1). In this assay, the microvessel sprouting occurs over a period of 10–12 days and begins with the migration of fibroblasts from the aortic rings. Endothelial sprouts first become visible after approximately 4–5 days of culture. As the assay progresses the endothelial sprouts gradually grow and mature, recruiting smooth muscle cells and pericytes to their periphery and form a lumenized vessel. After approximately 10–12 days, the micro-vessel outgrowths start to regress and matrix degradation occurs as the vessels are reabsorbed. Treatments with agents of interest therefore need to take place in the first 10–12 days of the assay, prior to microvascular regression. The final step in the assay is the “visualization and quantification of the vascular sprouts.” The effects on angiogenesis can be quantified using widely-available image analysis software (e.g., ImageJ) to assess various properties such as network radial growth, branch numbers, and nodes. The neovessels express endothelial markers such as von Willibrand factor (vWF), CD31 and can bind to the Griffonia Simplicifolia isolectin-B4, while surrounding support cells such as pericytes and smooth muscle cells express the proteoglycan Neural/glial antigen 2 (NG2) and α-smooth muscle actin. This allows for easy visualization and cell-specific characterization of the sprouts by whole-mount immunostaining (Fig. 2).

232

Vedanta Mehta and Marwa Mahmoud

Fig. 2 Representative image of a vascular sprout from a wild-type aortic ring treated with VEGF-A165 (30 ng/ mL) for 6 days. Endothelial cells are labeled with Dylight 594 conjugated Isolectin B4 (IB4), smooth muscle cells are labeled with FITC conjugated anti-alpha smooth muscle actin antibody, and nuclei are counterstained with DAPI. Images were taken at 40 magnification

2

Materials Curved forceps. Microdissection/iridectomy scissors. Scalpel. Falcon tubes. Petri-dishes. Opti-MEM™ Reduced Serum Medium, GlutaMAX™ (OptiMEM). (Thermofisher Scientific, 51985034). Penicillin-Streptomycin solution (Pen/Strep). 10 Dulbecco’s Modified Eagle Medium (DMEM) 12800017 Thermo Fisher Scientific (prepared by dissolving 1 sachet in 95 mL of autoclaved distilled water and filter sterilizing using a 0.22 μm stericup vacuum filtration system).

Mouse Aortic Ring Assay

233

Collagen type I (from rat tail) 3 mg/mL. 5 N NaOH. 96-well plate. 1 mL syringe. 19G and 26G needles. Glass slides. Cover slips. Fetal bovine serum (FBS). Vascular Endothelial Growth Factor A165 (VEGF). 4% Formalin. DAKO Serum free blocking solution. DyLight 594 labeled GSL-Isolectin B4 (Vector Biolabs). PBS containing 1 mM CaCl2 and 1 mM MgCl2. Diluent (PBS supplemented with 1 mM CaCl2, MgCl2 and MnCl2 and 0.1% Tween-20). Triton X-100. Tissue Culture Incubator (37  C/5% CO2).

3

Methods

3.1 Harvesting the Aorta and Preparing the Rings (Day 1)

1. Cull the mouse by asphyxiation in a CO2 chamber (see Note 1) and remove the skin on the ventral side. Cut the rib cage on both sides and cut open the thoracic cavity by cutting through the diaphragm. Remove the heart, lungs, and esophagus. The aorta will be running ventrally, adjacent to the trachea to the left side of the spine (Fig. 3). 2. Cut gently along the length of the aorta without applying any stretch or tension, clamping the free end with a pair of forceps until the abdominal end is reached (until the level of the kidneys). Cut it at the abdominal end and put it immediately into a 15 mL tube of medium (Opti-MEM supplemented with 100 U/mL of Penicillin and 100 μg/mL of Streptomycin) at room temperature. 3. Place the aorta in a petri-dish (with a few drops of medium to avoid the sample drying out), and dissect out the surrounding fat, connective tissue and branching vessels under a dissection microscope using a pair of fine forceps and microdissection scissors (see Note 2). 4. Flush out blood from the aortic lumen by inserting a 26G needle and syringe through, using 1 mL of Opti-MEM with Pen/Strep.

234

Vedanta Mehta and Marwa Mahmoud

Fig. 3 Dissecting the mouse aorta. Arrows indicate the aorta running ventrally along the spine

5. Cut the aorta into 0.5 mm rings with a sterile scalpel and ensure the lumen of each ring is wide open. If the edges of the ring are stuck, it can be opened gently with a pair of forceps. However, attention should be paid to not damaging the endothelium in the process (see Notes 3 and 4). 6. Transfer the rings (20–25 per mouse) to a 10 cm tissue culture dish containing Opti-MEM with Pen/Strep. Incubate the rings overnight in a humidified tissue culture incubator (37  C/5%CO2). 3.2 Embedding the Rings and Treatment with Agents of Interest (Days 2 to 10/12 When the Rings Are Fixed for Staining)

1. For a single 96-well plate, add 0.5 mL of 10 DMEM to 3.13 mL of tissue culture-grade water, on ice. 2. Add 1.37 mL collagen type I (from rat tail) to this mix on ice. 3. Adjust the pH by adding 10 μL of 5 N NaOH and giving the tube a gentle swirl. The mix should now appear pink. If it is not pink after the initial addition, another 10 μL of 5 N NaOH should be added. 4. Add 50 μL of the collagen mix to a 96-well plate, doing so in few wells at a time to prevent the mix from gelatinizing. Immediately add an aortic ring to the well (see Notes 5 and 6). The 96-well plate should be maintained at room temperature; however the collagen master mix must be kept on ice at all times. 5. Once all aortic rings have been embedded, leave the plate undisturbed at room temperature for 10–15 min to set.

Mouse Aortic Ring Assay

235

Fig. 4 Representative phase contrast images of collagen-embedded aortic rings at different time-points following treatment with PDGF-BB (50 ng/mL)

6. Incubate the plate in a humidified tissue culture incubator (37  C/5% CO2) for 1 h to fully gelatinize. 7. Feed each embedded aortic ring with 150 μL of Opti-MEM supplemented with 2.5% FBS and penicillin-streptomycin, and 30 ng/mL VEGF (or an appropriate growth factor). If a specific pathway is desired to be blocked, an inhibitor may also be added to the medium. The addition of VEGF may be omitted in negative controls. 8. Change the medium on days 3, 5, 7, 9, 11 by removing 130 μL and adding 150 μL of fresh medium, without disturbing the collagen gel (Fig. 4). 3.3 Immunofluorescent Staining

1. Aspirate the culture medium and wash the wells gently (so as to not dislodge the gel with the embedded ring) with PBS containing 1 mM CaCl2 and 1 mM MgCl2. 2. Fix the rings in 100 μL of 4% formalin for 30 min at room temperature.

236

Vedanta Mehta and Marwa Mahmoud

3. Remove the fixative and permeabilize the rings by adding 50 μL of 0.25% Triton X-100 in PBS (containing CaCl2 and MgCl2) to each well. Leave for 15 min at room temperature undisturbed, Aspirate the solution, and repeat this step with 50 μL of fresh 0.25% Triton X-100 in PBS. 4. Thereafter, block with DAKO serum free blocking solution for 30 min at room temperature, undisturbed (see Note 7). 5. Prepare fresh diluent. Dilute DyLight 594 labeled GSL-Isolectin B4 in the diluent (1:100) and add 30 μL to each well. Incubate with rings overnight at 4  C in the dark (see Note 8). 6. Next morning wash the rings three times for 15 min each wash, with PBS containing CaCl2 and MgCl2 and image under a fluorescent microscope. 7. To proceed with further imaging using a confocal microscope, detach the collagen gel from the well with a 19G needle. Pinch the top of the gel gently with curved forceps and place it on a glass slide. Add a drop of mounting medium. 8. Carefully place a cover slip over the gel (avoiding any air bubbles). Placement of the cover slip will result in the flattening of the sample. 9. Seal the edges with nail varnish and image on a confocal microscope. 10. Microscopic images of the angiogenic sprouts can be analyzed on Image J for various parameters, such as, network area and number of branch points.

4

Notes 1. It is preferable to use younger mice (less than 8 weeks of age) as they give more reproducible results and more vascular sprouts, however mice that are 8–12 weeks of age also give good results. It is not recommended to use mice older than 18 weeks of age. 2. It is important to remove as much fat and connective tissue as possible, without damaging the smooth muscle cell layer or the endothelial lining. Smooth muscle and endothelial integrity can be preserved by ensuring the vessel is not stretched during the fat removal. 3. It is also important to keep the size of the rings as consistent as possible for accurate comparison between experimental samples. 4. The lower half of the thoracic aorta gives more reproducible sprouting.

Mouse Aortic Ring Assay

237

5. It is advisable to embed a few rings at a time. We routinely handle 5–6 rings at a time. This way, the gel does not have time to set before the ring is placed in the well. A recommended way to embed the rings is to add 40 μL of collagen to the well, then embed the ring in the well and top it up with an extra 10 μL of collagen. The collagen under the ring must not begin to set before it is topped up. 6. Orientate the rings to look like circles in the well when observed from the top of the plate (rather than on their sides in which case, they would appear like flat rectangles). 7. We use this reagent is because it does not require matching with the antibodies and is compatible with any secondary reagent system. It also generally blocks non-specific staining (unlike serum blocks which just target non-specific secondary antibody binding) so can be especially useful when staining types of tissues which are susceptible to high background. 8. Specific fluorescent immunostaining of the aortic rings may be performed using either a fluorophore-conjugated primary antibody (such as αSMA-FITC) for an overnight duration in the dark at 4  C or a primary antibody for an overnight incubation at 4  C, followed by a 2–3 h incubation with a fluorophoreconjugated secondary antibody, at room temperature and in the dark. All antibodies should be dissolved in the specified diluent. References 1. Risau W (1997) Mechanisms of angiogenesis. Nature 386(6626):671–674. https://doi.org/ 10.1038/386671a0 2. Iruela-Arispe ML, Dvorak HF (1997) Angiogenesis: a dynamic balance of stimulators and inhibitors. Thromb Haemost 78(1):672–677 3. Carmeliet P (2005) Angiogenesis in life, disease and medicine. Nature 438(7070): 9 3 2 – 9 3 6 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / nature04478 4. Passaniti A, Taylor RM, Pili R, Guo Y, Long PV, Haney JA, Pauly RR, Grant DS, Martin GR (1992) A simple, quantitative method for assessing angiogenesis and antiangiogenic agents using reconstituted basement membrane, heparin, and fibroblast growth factor. Lab Investig 67(4):519–528 5. Baker JH, Huxham LA, Kyle AH, Lam KK, Minchinton AI (2006) Vascular-specific quantification in an in vivo Matrigel chamber angiogenesis assay. Microvasc Res 71(2):69–75. https://doi.org/10.1016/j.mvr.2006.01.002

6. Ribatti D, Urbinati C, Nico B, Rusnati M, Roncali L, Presta M (1995) Endogenous basic fibroblast growth factor is implicated in the vascularization of the chick embryo chorioallantoic membrane. Dev Biol 170(1):39–49. https://doi.org/10.1006/dbio.1995.1193 7. Ejaz S, Seok KB, Woong LC (2004) A novel image probing system for precise quantification of angiogenesis. Tumori 90(6):611–617 8. Nicosia RF (2009) The aortic ring model of angiogenesis: a quarter century of search and discovery. J Cell Mol Med 13(10):4113–4136. https://doi.org/10.1111/j.1582-4934.2009. 00891.x 9. Nicosia RF, Ottinetti A (1990) Growth of microvessels in serum-free matrix culture of rat aorta. A quantitative assay of angiogenesis in vitro. Lab Investig 63(1):115–122 10. Zhu WH, Iurlaro M, MacIntyre A, Fogel E, Nicosia RF (2003) The mouse aorta model: influence of genetic background and aging on bFGF- and VEGF-induced angiogenic sprouting. Angiogenesis 6(3):193–199. https://doi.

238

Vedanta Mehta and Marwa Mahmoud

org/10.1023/B:AGEN.0000021397. 18713.9c 11. Devy L, Blacher S, Grignet-Debrus C, Bajou K, Masson V, Gerard RD, Gils A, Carmeliet G, Carmeliet P, Declerck PJ, Noel A, Foidart JM (2002) The pro- or antiangiogenic effect of plasminogen activator inhibitor 1 is dose dependent. FASEB J 16(2):147–154. https:// doi.org/10.1096/fj.01-0552com 12. Auerbach R, Muthukkaruppan V (2012) The chick embryo aortic arch assay. In: Zudaire E, Cuttitta F (eds) The textbook of angiogenesis and lymphangiogenesis: methods and applications. Springer, Dordrecht. https://doi.org/ 10.1007/978-94-007-4581-0_8 13. Stiffey-Wilusz J, Boice JA, Ronan J, Fletcher AM, Anderson MS (2001) An ex vivo angiogenesis assay utilizing commercial porcine carotid artery: modification of the rat aortic ring assay. Angiogenesis 4(1):3–9. https:// doi.org/10.1023/a:1016604327305 14. Iqbal F, Gratch YS, Szaraz P, Librach CL (2017) The aortic ring co-culture assay: a

convenient tool to assess the angiogenic potential of mesenchymal stromal cells in vitro. J Vis Exp (127):56083. https://doi.org/10.3791/ 56083 15. Iqbal F, Szaraz P, Librach M, Gauthier-FisherA, Librach CL (2017) Angiogenic potency evaluation of cell therapy candidates by a novel application of the in vitro aortic ring assay. Stem Cell Res Ther 8(1):184. https:// doi.org/10.1186/s13287-017-0631-1 16. Moulton KS, Vakili K, Zurakowski D, Soliman M, Butterfield C, Sylvin E, Lo KM, Gillies S, Javaherian K, Folkman J (2003) Inhibition of plaque neovascularization reduces macrophage accumulation and progression of advanced atherosclerosis. Proc Natl Acad Sci U S A 100(8):4736–4741. https://doi.org/10. 1073/pnas.0730843100 17. Aplin AC, Nicosia RF (2019) The plaqueaortic ring assay: a new method to study human atherosclerosis-induced angiogenesis. Angiogenesis 22(3):421–431. https://doi. org/10.1007/s10456-019-09667-z

Chapter 18 Retinal Microvasculature-on-a-Chip for Modeling VEGF-Induced Permeability He´loı¨se Ragelle, Karen Dernick, Peter D. Westenskow, and Stefan Kustermann Abstract Relevant human in vitro models of the retinal microvasculature can be used to study the role of disease mediators on retinal barrier dysfunction and assess the efficacy of early drug candidates. This chapter describes an organ-on-a-chip model of the retinal microvasculature that allows for facile quantification of barrier permeability in response to leakage mediators, such as Vascular Endothelial Growth Factor (VEGF), and enables screening of VEGF-induced permeability inhibitors. This chapter also presents an automated confocal imaging method for the visualization of endothelial tube morphology as an additional measure of barrier integrity. Key words Organ-on-a-chip, Microvasculature, Blood-retinal barrier, Advanced in vitro model, 3D model, Vascular permeability, VEGF-induced leakage, High content imaging

1

Introduction The blood-retinal barrier (BRB) controls the retinal environment by regulating fluxes of ions and molecules and maintains the retina as a relatively immune privileged site [1, 2]. The BRB is composed of two distinct components: the outer BRB which controls fluxes from the extraretinal choriocapillaries and is regulated by the retinal pigmented epithelium cells and the inner BRB that is regulated by highly specialized retinal vascular endothelial cells (characterized by tight junctions and limited transcytosis), pericytes, astrocytes, Mueller glia, and smooth muscle cells [3]. Impairment of inner BRB function, as observed in diabetic retinopathy (DR), leads to abnormal cytokine and immune cell infiltration in the retina, excessive fluid accumulation, and vision loss [4, 5].

He´loı¨se Ragelle and Karen Dernick contributed equally to this work. Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_18, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

239

240

He´loı¨se Ragelle et al.

Vascular Endothelium Growth Factor (VEGF)—one of the main contributors to increased permeability and vascular leakage observed in DR—is elevated in the vitreous and retina of diabetic patients, and anti-VEGF-based therapies are currently the standard of care for DR [6]. Relevant human in vitro models of BRB biology can help interrogate the role of VEGF and other disease mediators on barrier properties, study signaling pathways, improve clinical translation, and support investigations in the field of ophthalmology drug development. Microphysiological systems or organ-on-chip models are of particular interest for the production of predictive models for ocular diseases, notably in the field of BRB biology [7, 8]. Traditional in vitro models of the inner BRB have been based on 2-dimensional (2D) transwell cultures, which might have limited biological relevance due to a lack of critical signals, such as cell– matrix interaction and shear stress induced by flow [8]. The presented method describes a human retinal microvasculature-on-a-chip model based on the Mimetas OrganoPlate® microfluidic platform that integrates immortalized human retinal microvascular endothelial cells (hRMVECs), an extracellular matrix, and perfusion. Under these conditions, hRMVECs recapitulate physiological barrier function characterized by the formation of tight and adherens junctional complexes and a low apparent permeability. These features were not observed in hRMVECs in 2D transwell culture [9]. In addition, the retinal microvasculatureon-a-chip enables the investigation of the effects of leakage mediators on barrier properties and is amenable to the screening of potential therapeutic modalities [9, 10]. In this method chapter, we describe how to use the retinal microvasculature-on-a-chip model to quantify the effect of VEGF on endothelial permeability (Fig. 1a–d) and to test the efficacy of inhibitors of VEGF-induced permeability, using a dextran leakage assay. Using the presented model, we show that Aflibercept (a VEGF-neutralizing antibody used to treat wet Age-related Macular Degeneration; AMD) and Wortmannin (an inhibitor of phosphoinositide 3-kinase) could revert VEGF-induced permeability to the level of untreated controls, in line with clinical data on Aflibercept efficacy in wet AMD and Wortmannin anticipated effect on regulating cell–cell junctions. In addition, the chapter details a high-resolution confocal imaging method to visualize tube morphology and sub-cellular structures, such as cell–cell junctions, in 3D and in an automated fashion at a medium scale throughput (Fig. 2a, b). One challenge for automated image acquisition is the potential for imprecision in the manufacturing process of complex microplates, such as the OrganoPlate®, that can lead to shifts in chip alignment and result in drift when imaging is done automatically. To solve this issue, we have developed an imaging method that first uses a scan at low

Retinal Microvasculature-on-a-Chip

241

Fig. 1 Quantitative assessment of vascular permeability. (a) To quantify permeability of the endothelial barrier within the retinal microvasculature-on-a-chip, a fluorescent tracer (dextran 20 kDa or 70 kDa) was perfused through the open lumen of the endothelial tubes by adding it to the medium channel (top compartment). Fluorescence intensities in both the medium channel and matrix channel were quantified over time to calculate the leakage score (L ), which was the ratio of average intensity of fluorescent signal in the matrix channel (IMATRIX red box) to the average intensity of fluorescent signal in the medium channel (IMEDIUM, yellow box). Scale bar 400 μm. (b) In the case of a tight endothelial barrier (hTERT-hRMVECs untreated control), the tracer was retained in the medium channel (top). Upon treatment with VEGF, the tracer diffused into the matrix during the measurement (leaky barrier, middle and bottom). Each panel represents a discrete time point during the experiment, where t300 indicates t ¼ 300 s. Scale bar 400 μm. (c) Leakage scores were plotted as a function of perfusion time for untreated and VEGF-treated (100 ng/mL and 200 ng/mL) endothelialized tubes (n ¼ 5). (d) The apparent permeability (Pa) within the retinal microvasculature-on-a-chip was quantified and compared for untreated samples and VEGF-treated samples (100 ng/mL and 200 ng/mL) using a 20 kDa dextran tracer. (e) Relative apparent permeability values for VEGF positive control (100 ng/mL), Wortmannin, Aflibercept, and GF 109203X samples in the presence of VEGF (100 ng/mL). Pa values were normalized to the untreated control (n ¼ 3)

magnification (called prescan) that is coupled with an analysis sequence to identify the fields of interest for an automated second scan at a higher magnification (rescan). We used high-resolution confocal imaging to detect the effects of Interleukin-1 beta (IL-1β) treatment on endothelial cells leading to leaky endothelial tubes (Fig. 2 c). Taken together, the presented retinal microvasculatureon-a-chip model is useful for modeling relevant components of retinal barrier dysfunction and potential intervention strategies making it a promising model for early testing of novel drug candidates.

242

He´loı¨se Ragelle et al.

Fig. 2 Automated confocal imaging of the retinal microvasculature-on-a-chip. A prescan at low magnification (5) is first performed to define the ROI for an automated rescan at higher magnification (20). (a) Plate microstructures are autofluorescent in the DAPI channel (light gray) and differences in intensity values (yellow) are used to define a threshold to separate the dark endothelial tube region. (b) In a next step, position property Centroid Y is utilized to identify the endothelial tube region. (c) Correct identification of the ROI by the analysis software (yellow box). (d) The analysis sequence calculates the position of the imaging fields for the rescan at

Retinal Microvasculature-on-a-Chip

2 2.1

243

Materials Cell Culture

1. Immortalized human retinal microvascular endothelial cells (hTERT-hRMVECs, Angio-Proteomie, see Note 1). 2. Endothelial Cell Growth Medium-2 BulletKit (EGM-2, Lonza). 3. TrypLE Express Enzyme 1, no phenol red. 4. Hank’s Balanced Salt Solution supplemented with calcium and magnesium (HBSS+/+). 5. Phosphate Buffer Saline (PBS). 6. Starvation medium: EGM-2 medium without VEGF and supplemented with 0.5% Fetal Bovine Serum (FBS) and all other factors from the BulletKit.

2.2 Collagen Matrix Preparation and Formation of the Retinal Microvasculature-ona-Chip

1. Collagen I Rat Tail 5 mg/mL (Cultrex). 2. HEPES 1M (Gibco). 3. NaHCO3: Prepare a 37 mg/mL solution in sterile DNAse/ RNAse-free deionized water by weighing 370 mg of NaHCO3 and adding 10 mL water. Sterile filter. 4. OrganoPlate® 2-lane (9603-400-B, Mimetas BV). 5. Perfusion rocker OrganoFlow® and plate stand (Mimetas BV).

2.3 Permeability Assay

1. 20 kDa Fluorescein isothiocyanate-dextran (FITC-dextran, Sigma) and 70 kDa Tetramethylrhodamine isothiocyanatedextran (TRITC-dextran, Sigma). 2. Dextran solutions: Prepare dextran stock solutions (25 mg/ mL) by weighing 50 mg of dextran in 2 mL DNAse/RNAsefree deionized water. Sterile filter. The dextran stock solutions can be kept in the dark at 4  C for several weeks. 3. Vascular Endothelial Growth Factor (VEGF, R&D Systems), Interleukin-1 beta (IL-1β, R&D systems), Wortmannin (Merck), Aflibercept, and GF 109203X (Toronto Research Chemicals).

ä Fig. 2 (continued) higher magnification (yellow squares numbered 1–4). Scale bars (a–d) 500 μm. (e) Stitched images of the rescanned fields at higher magnification (20). Maximum projection of a 10 μm Z-stack of the bottom region of the tube treated with IL-1β for 24 h (1 ng/mL). Cell nucleus was labeled with NucBlue (blue) and hRMVEC were labeled with CD31 (green) and VE-Cadherin (red). Scale bar 500 μm. (f) Three-dimensional reconstructions of a tight (untreated, left) and a leaky (IL-1β treated, right) tube obtained after the rescan. Cell nucleus was labeled with NucBlue (blue) and hRMVEC were labeled with CD31 (green). Size of the 3D frame: height 220 μm, width 600 μm, depth 600 μm

244

2.4

He´loı¨se Ragelle et al.

Immunostaining

1. Fixing solution: 4% (v/v) paraformaldehyde in HBSS+/+: Dilute from 16% paraformaldehyde solution (Alfa Aesar). 2. Permeabilizing solution: 0.3% (v/v) Triton X-100 (Sigma) in deionized water. 3. Blocking solution: 0.5% (v/v) donkey serum in PBS. Prepare fresh and discard after use. 4. Primary antibodies for the staining of adherens junctions and endothelial cell marker are anti-VE Cadherin (Abcam, ab33168, 1:200) and anti-CD31 (Abcam, ab24590, 1:200), respectively. Respective dilutions are indicated in brackets. 5. Donkey anti-mouse and donkey anti-rabbit secondary antibodies (ThermoFisher, 1:1000). 6. NucBlue Fixed ThermoFisher).

2.5 Particular Equipment

Cell

ReadyProbes

Reagent

(DAPI,

1. Confocal microscope. We used an Opera Phenix® High Content Imager (Perkin Elmer) equipped with 5 (air) and 20 (water) objectives and light source/filter sets suited for brightfield (transmission), DAPI (excitation 405 nm, emission 435–480 nm), FITC (excitation 488 nm, emission 500–550 nm), and TRITC (excitation 561 nm, emission 570–630 nm) imaging. 2. Harmony software (version 4.9, Perkin Elmer). 3. Fixed region building block as add-on for the image analysis within Harmony (Perkin Elmer). 4. Automated cell counter or hemocytometer.

3

Methods

3.1 Formation of the Retinal Microvasculature-ona-Chip

1. Place the OrganoPlate® 2-lane at 4  C for minimum 15 min. 2. Prepare collagen I matrix solution on ice by adding 400 μL collagen I (5 mg/mL), 50 μL HEPES (1 M), and 50 μL NaHCO3 (37 mg/mL). Final collagen, HEPES, and NaHCO3 concentrations in the solution are 4 mg/mL, 100 mM, and 3.7 mg/mL, respectively (see Note 2). 3. Before pipetting, place the end of the pipette tip in the eppendorf tube, which contains the matrix solution, that is on ice for 10 s (see Note 2). Using an automatic dispenser, such as MultiPette E3 (Eppendorf) with 0.1 mL Combitips Advanced (Eppendorf), pipette the volume of matrix solution needed to fill one column of the OrganoPlate® (e.g.,16 wells). Add 1.9 μL of matrix solution to each matrix inlet. Monitor that the channel is correctly filled with the matrix by the presence of a black line in the observation window. Fill one column (16 wells) at a time and repeat for the remaining columns.

Retinal Microvasculature-on-a-Chip

245

4. Let the matrix polymerize at 37  C for 45 min. 5. While the matrix is polymerizing, harvest hTERT-HRMVEC according to the manufacturer’s protocol (see Note 3). 6. After centrifugation for 5 min at 500  g, minimize the volume to resuspend cells to ensure that the cell suspension is concentrated enough (e.g., for cells grown in a T175 flask use 200–400 μL). 7. For counting, take 1 μL of cell suspension, add 9 μL of PBS (1: 10 dilution of the original cell suspension) and 10 μL TrypanBlue. 8. Count the cells using a hemocytometer or automated cell counter. 9. Adjust the volume of the cell suspension to have a final concentration of 30  106 cells/mL. 10. Fill each cell inlet with 2 μL of cell suspension to obtain a final density of 60,000 cells per channel. If the medium channel is correctly filled with cells, the black line in the observation window is not visible anymore. 11. Add 30 μL PBS in each observation window. 12. Place the plate on the side using the plate stand (matrix channels facing downwards) for 60 min at 37  C to allow cells to sediment against the support matrix. 13. Add 50 μL of EGM-2 in each cell inlet and place the plate on the side for another 2 h at 37  C. 14. Add 50 μL of EGM-2 in each cell outlet (see Note 3). 15. Place plate horizontally on OrganoFlow® rocker (Program 1 with 2 inclination and 8 min interval between two rotation cycles). 16. One hour later add 50 μL of HBSS+/+ in each gel inlet. 17. Place plate horizontally on the rocker and incubate overnight. 18. The day after seeding, replace medium in each cell inlet and outlet. Do not change HBSS+/+ in the gel inlet. 19. Tubes are ready 48 h after seeding (see Note 4). 3.2 VEGF-Mediated Induction of Vascular Leakage and Quantitative Assessment of Vascular Permeability 3.2.1 VEGF-Mediated Induction of Vascular Permeability

1. Starve cells for 24 h by replacing growth medium with 50 μL starvation medium in each of the gel inlets, cell inlets, and cell outlets. 2. Prepare VEGF inhibitor solutions by diluting the inhibitor in starvation medium (final concentrations are 10 μM GF 109203X, 0.1 μM Wortmannin, and 1 μg/mL Aflibercept). Remove medium in the wells and add 50 μL of the inhibitor solution in each of the gel inlets, cell inlets, and cell outlets. For the VEGF controls and untreated controls, add 50 μL starvation medium without inhibitors. Incubate for 30 min just before adding VEGF.

246

He´loı¨se Ragelle et al.

3. Add 5 μL VEGF solution (100 ng/mL final concentration) in each of the gel inlets, cell inlets, and cell outlets. For the untreated controls, add 5 μL starvation medium. Incubate at 37  C on the rocker for 24 h (see Note 5). 3.2.2 Leakage Assay

The steps below describe the workflow for image acquisition and analysis of barrier integrity using the Opera Phenix® high content imager. The leakage assay can be performed on other high content analysis (HCA) instruments and the methodology and terminology will have to be adapted accordingly (see Note 6). To calibrate the instrument and leakage assay, steps 1–10 should be carried out using a reference well and should be saved on Harmony as Experiment for future measurements (see Note 7). Set up as below and save the method. Once the method is saved, in future experiments go directly to step 11. 1. In Setup, select OrganoPlate® 2-lanes as the type of plate to image (see Note 8). 2. Set the Autofocus to Two Peak (default) for microplates (see Note 9). 3. Select 5 air objective (see Note 10). 4. Set Imaging mode to non-confocal to collect the fluorescence signal from the whole tube depth (Z-direction). To reduce image file size, the pixels can be clustered by using 2  2 binning. 5. The barrier integrity assay described in this protocol uses the brightfield channel for tube identification and two fluorescent channels for quantifying leakage. In channel selection, define channel 1 as brightfield, channel 2 as FITC, and channel 3 as TRITC and assign appropriate excitation and emission filter settings to each of the fluorescent channels (see Note 11). 6. Determine exposure settings for each of the channels on a reference well: add 30 uL of dextran solution to the cell inlet and cell outlet of the endothelial tube (0.1 mg/mL FITCdextran and 0.1 mg/mL TRITC-dextran final concentrations) and add 30 uL medium without tracer to the gel inlet (see Note 12). Load plate into the instrument, select the reference area. Select the brightfield channel and set the height such that the image is in focus by taking snapshots at various heights. Then, adjust light source power and exposure time for each fluorescent channel to obtain a good signal but avoid saturation (see Note 13). 7. If you are using more than one fluorophore, review excitation and emission spectra carefully to minimize crosstalk (see Note 14).

Retinal Microvasculature-on-a-Chip

247

8. Once you are satisfied with the setting for the reference area, expand layout selection to the wells of interest for your assay. 9. Set up time series: 5 min intervals over the course of 20 min (see Note 15). 10. Save your experiment settings. 11. When the setup is ready and just prior to performing the permeability assay, prepare the dextran working solution: Dilute 28 μL of FITC-dextran stock solution and 28 μL of TRITC-dextran stock solution in 7 mL starvation medium (0.1 mg/mL FITC-dextran and 0.1 mg/mL TRITC-dextran final concentrations). Discard the working solution after use. 12. Under the hood, add 30 μL of starvation medium in the matrix inlet. 13. Rapidly add 30 μL of dextran working solution in both the cell inlet and cell outlet. Remove the lid and seal the plate with a plate seal. Be quick as the diffusion starts as soon as dextran is added to the perfusion channel. 14. Load the plate in the instrument. 15. Add plate name in the Run Experiment tab and start the image acquisition (see Note 16). 3.2.3 Quantitative Assessment of Permeability

Image analysis can be done either after image acquisition in the Image Analysis tool or, once set up, in parallel with the measurement by selecting the respective analysis protocol as online job when setting up the image acquisition protocol. A detailed overview of the parameters used can be found in Table 1 (Analysis Sequence called Mimetas Leakage Assay). 1. In the Image Analysis tab, load your experiment. Select one well in the brightfield channel and load the building block named Fixed Region. Define two rectangular regions of interest (ROIs) of the same size, one in the medium compartment (source) and one in the matrix compartment (receiver). Verify in a few wells across the plate that the ROIs are well positioned in the middle of the channels (see Fig. 1a). If necessary, apply a correction factor along the X- and/or Y-axis to correct for well offsets and to position ROIs similarly in all wells (see Note 17). 2. Calculate the fluorescence mean intensities in the medium compartment (IMEDIUM) and in the matrix compartment (IMATRIX) for both fluorophores by using the building block termed Calculate Intensities. 3. For each well and time point, calculate the leakage score (L) defined as the ratio of average intensity of fluorescent signal in the matrix compartment (IMATRIX) to the average intensity of fluorescent signal in the medium compartment (IMEDIUM) by

248

He´loı¨se Ragelle et al.

Table 1 Analysis sequence “Mimetas leakage assay” Input Input Image

Method

Output

Flatfield Correction: None Brightfield Correction Stack Processing: Individual Planes Min. Global Binning: Dynamic

ABB: Fixed Regions Channel: Brightfield (2) X Mid: 500 X Width: 900 Y1 Mid: 380 Y1 Height: 100 Y2 Mid: 570 Y2 Height: 100 Y3 Mid: 1 Y3 Height: 1 Y Shift with Column: -15 Y Shift with Row: 0 Output Population: Fixed Regions Calculate Intensity Properties

Channel: FITC Population: Fixed Regions Region: Region1

Property Prefix: Intensity Method: Medium FITC Standard Mean

Calculate Intensity Properties (2)

Channel: FITC Population: Fixed Regions Region: Region2

Method: Property Prefix: Intensity Standard Matrix FITC Mean

Calculate Intensity Properties (3)

Channel: TRITC Population: Fixed Regions Region: Region1

Method: Property Prefix: Intensity Standard Medium TRITC Mean

Calculate Intensity Properties (4)

Channel: TRITC Population: Fixed Regions Region: Region2

Property Prefix: Intensity Method: Matrix TRITC Standard Mean

Define Results

Results Method: List of Outputs Method: Formula Output Formula: a/b Population Type: Objects Variable a: Fixed Regions - Intensity Matrix FITC Mean Mean Variable b: Fixed Regions - Intensity Medium FITC Mean Mean Output Name: leakage score FITC Method: Formula Output Formula: a/b Population Type: Objects Variable a: Fixed Regions - Intensity Matrix TRITC Mean Mean Variable b: Fixed Regions - Intensity Medium TRITC Mean Mean (continued)

Retinal Microvasculature-on-a-Chip

249

Table 1 (continued) Input

Method

Output

Output Name: leakage score TRITC Object Results Population: Fixed Regions: None

adding the formula below in the define results building block (see Note 18): L ¼ I MATRIX =I MEDIUM 4. Go to the analysis tab and copy the summary table of leakage score over time in an excel file. 5. Calculate the permeability coefficient (Pa) using the following formula: P a ðcm=sÞ ¼ dL=dt  V MATRIX  1=A where L is the leakage score, VMATRIX is the volume of the matrix (4.54  104 cm3), and A is the surface area of the endothelial tube in contact with the matrix (1.21  102 cm2) (see Note 19). The leakage score is calculated for each individual time point and a linear regression is fitted through these ratios to calculate the slope, dL/dt [11]. An example of a plot of leakage score as a function of time for a leaky and tight vessel is shown in Fig. 1c. 3.3 Immunostaining and Confocal Imaging of the Microvasculature-ona-Chip 3.3.1 Immunostaining

Confocal imaging at high magnification can provide further information about cell morphology as well as on expression and 3D spatial distribution of cellular markers following treatment with VEGF. The procedure described below is done on wells that have not been used for the leakage assay. 1. Wash cells by adding 100 μL HBSS+/+ in the cell inlet and 50 μL HBSS+/+ in the gel inlet and cell outlet. 2. Fix cells by adding 30 μL of 4% paraformaldehyde solution in each of the gel inlets, cell inlets, and cell outlets for 15 min at room temperature. 3. Rinse twice by adding 100 μL HBSS+/+ in the cell inlet and 50 μL HBSS+/+ in the gel inlet and cell outlet. 4. Permeabilize cells by adding 30 μL of 0.3% Triton X-100 in each of the gel inlets, cell inlets, and cell outlets. Incubate 10 min at room temperature. 5. Rinse twice by adding 100 μL HBSS+/+ in the cell inlet and 50 μL HBSS+/+ in the gel inlet and cell outlet.

250

He´loı¨se Ragelle et al.

6. Incubate with 0.5% donkey serum as blocking solution at room temperature for 1 h. Use 50 μL in each of the gel inlets, cell inlets, and cell outlets. 7. Dilute antibodies in 0.5% donkey serum and incubate at room temperature for 2 h or at 4  C overnight (20 μL in cell inlet and in cell outlet). Change plate orientation during incubation or use the rocker to allow for efficient distribution of antibodies. 8. Rinse three times with 0.5% donkey serum and add secondary antibodies (diluted 1:1000 in 0.5% donkey serum) for 1 h at 4  C (20 μL in cell inlet and in cell outlet). Change plate orientation during incubation or use the rocker. 9. Rinse twice by adding 100 μL HBSS+/+ in the cell inlet and 50 μL HBSS+/+ in the gel inlet and cell outlet. 10. Incubate with NucBlue following supplier’s recommendations to stain for nuclei. 3.3.2 Automated Confocal Imaging of the Microvasculature-on-aChip

This method describes our methodology to mitigate drift during automatic imaging that can occur due to imprecision in the plate manufacturing process. Analysis parameters for the automated confocal imaging sequence are detailed in Table 2 (Mimetas Prescan Analysis Sequence). 1. For the prescan, set up an image acquisition protocol with the 5 air objective as described previously in steps 1–4 of “Leakage Assay” 3.2.2. 2. The OrganoPlate® microstructures (phase guide and walls) are autofluorescent in the DAPI channel and can be leveraged to identify the endothelial tube region. Set channel 1 as DAPI, load the plate, and take a snapshot of a selected field in order to adjust height and exposure time as described previously (see steps 6 and 7 of “Leakage Assay” 3.2.2). Make sure all substructures appear in light gray (Fig. 2a). Save the prescan experiment. 3. Run a prescan test measurement on a few wells. The images from the prescan test will be used to generate the analysis sequence to identify the ROI for the rescan. 4. In the Image Analysis section, select one of the test images and display the DAPI channel. The microstructures appear in light gray while the matrix and endothelial tube regions appear in dark. Turn on the Show Intensity option (right click on the image) and compare the intensities of the microstructures (light gray) with the intensities of the darker matrix and endothelial regions (dark gray). Define a threshold suited to separate the endothelial tube from the microstructures (e.g., 500, Fig. 2a). Select the building block termed Find Image Region, choose the DAPI channel, and set the absolute

Retinal Microvasculature-on-a-Chip

251

Table 2 Analysis sequence “Mimetas Prescan” Input

Method

Output

Find Image Region Channel: DAPI ROI: None

Method: Absolute Threshold Lowest Intensity: 0 Highest Intensity: 500 Split into Objects Area: >0 px2

Output Population: Image Region Output Region: Image Region

Calculate Position Properties

Population: Image Region Region: Image Region

Method: Standard Centroid Position in Image

Property Prefix: Image Region

Calculate Morphology Properties

Population: Image Region Region: Image Region

Method: Standard Area Roundness

Property Prefix: Image Region

Select Population

Population: Image Region

Method: Filter by Property Output Population: Image Region Centroid Y in Image Region Selected Image [μm]: >200 Image Region Centroid Y in Image [μm]: 400000 Boolean Operations: F1 and F2 and F3

Determine Well Layout

Population: Image Region Selected Region: Image Region

Method XY: Multiple Objects Rescan Magnification: 20x Max No of Fields: Object Margin: 10 μm Field Overlap: 2% Method Z: None

Define Results

Results Method: List of Outputs Population: Image Region Number of Objects Object Results Population: Image Region: None Population: Image Region Selected: None

Input Image

Flatfield Correction: None Brightfield Correction Stack Processing: Individual Planes Min. Global Binning: Dynamic

Output Population: Well Layout

252

He´loı¨se Ragelle et al.

threshold of the highest intensity below the threshold value. Only the dark regions will be selected for further analysis. 5. To isolate the endothelial tube region from the other dark regions, select the building block called Calculate Position Properties and choose the option Centroid Position in Image. Check the values of the Centroid in Y for the endothelial tube region compared to the other dark regions in different wells of the test run by looking at the properties of the selected region. Select minima and maxima values that separate the endothelial tube region from the others (Fig. 2b, see Note 20). 6. Add the Select Population building block and select the method called Filter by Property. Choose Centroid in Y and set the thresholds respectively above and below the maxima and minima values determined in step 5. The endothelial tube region is highlighted in yellow (Fig.2c, see Note 21). 7. Add the building block termed Determine Well Layout and select the output population of step 6 as Population. Select Multiple Objects as Method XY and define the Rescan Magnification as 20 (see Note 22). Define the overlap of the individual imaging fields for a later stitching of the images for analysis (usually 2–5% when the 5 objective is used). Verify that the rescan well layout shown in the image field covers the endothelial tube region for all wells in the test run (Fig. 2d). If this is not the case, adjust the thresholds as described in step 5. 8. Save the Analysis Sequence as Mimetas Prescan. 9. Go back to the Setup tab, load the prescan experiment from step 2, ensure that all wells of interest are included, and add the analysis sequence Mimetas Prescan as online job, then save the experiment. 10. Select one of the test wells to set up the rescan and acquire a prescan image by pressing the Test button. 11. Select the 20 water objective and switch to confocal mode. Make sure that you position the image acquisition field in the endothelial tube region (see Note 23). 12. Select the fluorescence channels according to your fluorophores. In our case channel 1 is DAPI (ex 405 nm, em 435–480 nm), channel 2 is AF647 (ex 640 nm, em 650–760 nm), and channel 3 is AF555 (ex 561 nm, em 570–630 nm). Find the cells at the bottom of the tube by taking snapshots at different heights. Set up laser power and exposure time as described in steps 6–7 of “Leakage Assay” 3.2.2. 13. To image the full tube, set the height of the plate bottom as first plane and add the number of planes and distance so that

Retinal Microvasculature-on-a-Chip

253

the Overall Height matches the diameter of the tube (220 μm). The number of planes in the stack depends on user preferences but be aware that a recommended minimum plane distance exists for every objective (see respective objective lens specification in the Opera Phenix® User Manual). 14. In the tab Define Layout, at least one well needs to be selected. It is not relevant which wells or fields are selected, as they will be defined automatically after the prescan by the Mimetas Prescan analysis sequence. 15. Save the protocol. If you used a water-based objective for setting up the rescan, do not forget to eject the plate and clean the plate bottom to remove any water droplets to prevent focusing errors. Load your plate and start the run. 16. In the section Image Analysis of the harmony software, the quality of the tubes can be inspected visually by selecting individual fields and setting the Stack Processing in the Input Image field of the Analysis Sequence to 3D analysis. Images can be exported and movies generated.

4

Notes 1. We have observed variation in permeability and barrier function between custom and commercially available immortalized hRMVEC batches. We recommend testing and comparing multiple cell batches. 2. For the collagen matrix preparation, it is important that all reagents and the matrix solution remain cold. For this, remove reagents from the fridge at the last minute. Place the reagents as well as the eppendorf tube that will contain the matrix solution on ice and allow the end of the tip to cool before pipetting the matrix solution. When preparing the matrix solution, add NaHCO3 last. Mix carefully by pipetting up and down several times and avoid bubbles. HEPES and NaHCO3 solutions can be kept for several months at 4  C. Replace collagen I solution maximum 2 months after opening the vial as older collagen I solutions lead to the formation of weaker matrix which negatively impacts permeability values. 3. In the chip and while harvesting the cells before seeding, make sure to use endothelial cell medium containing low FBS percentage, such as EGM-2 medium that contains 2% FBS. Medium with higher FBS concentrations result in migration of cells in the matrix channel, destabilization of the tubes, and a leaky barrier. 4. As a quality control for proper tube formation, check wells individually using a brightfield microscope, mark wells with

254

He´loı¨se Ragelle et al.

matrix overflow and/or incomplete filling of matrix channels or endothelial channels, and exclude them from further analysis. 5. We recommend a minimum of triplicates per condition. Five replicates is ideal. 6. The respective field application specialist can aid you in adapting the workflow to the platform of choice. It is recommended that researchers new to HCA consult the field application specialist or other professionals familiar with the instrument for help in setting up the assay. 7. Diffusion starts from the moment the dextran solution is added to the medium channel. Before adding the dextran solution, make sure the microscope is ready and the protocol loaded on the Harmony interface to minimize time between dextran addition and acquisition. 8. The OrganoPlate® is based on a standard 384-well plate format with a 150 μm thick glass bottom and is therefore compatible with standard laboratory imaging equipment. The 2-lane OrganoPlate® comprises 96 microfluidic lanes, the equivalent of 4 wells of a standard 384-well plate. In each lane, one well is used for imaging. Therefore, only 96 wells of the 384-well plate are of interest for imaging. These can either be selected using the layout of a standard 384-well plate or by defining in Harmony a new plate type that comprises only the wells of interest for imaging (e.g., 6 columns corresponding to the observation windows of the OrganoPlate®). 9. The laser-based autofocus system of the Opera Phenix® relies on the detection of reflections of the laser beam at surfaces (peaks) while the autofocus system moves upwards. These reflections are caused by changes in the refractive index at the different surfaces. The focusing is done in the center of the image. If the measurement hits the phase guide, the standard autofocus type for microplates (“two peak”) might fail. For imaging at low magnifications, the risk of hitting the phase guide is quite low and can be bypassed by adjusting the plate type definition (to make sure that the phase guide is not in the center of the image field). For higher magnification objectives, it can be beneficial to switch to “one peak” autofocus (for microplates only available from Harmony version 4.9 onwards). 10. Make sure the field of view covers the area of interest. To assess barrier integrity, it is advisable to image the complete tissue area (whole length of the tubes separated by the phase guide). We use the 5 air objective (NA 0.16) and image one field in the center of each well of interest. 11. Please consult the technical specifications of the HCA instrument to determine if the light source and excitation/emission

Retinal Microvasculature-on-a-Chip

255

filters installed are appropriate for use with the selected fluorophore combination prior to labeling cells or performing experiments. 12. We used 20 kDa and 70 kDa dextrans as we are interested in monitoring fluxes of macromolecules across the barrier. To determine the concentration of the dextran solution for your assay, make calibration curves of fluorescence intensity in the medium and in the matrix channels as a function of dextran concentration. Permeability measurements should occur at a dextran concentration range where fluorescence intensity is proportional to dextran concentration. More than one tracer can be used simultaneously when different fluorescent labels are chosen. In this case make sure that the combination is in line with light source/filter settings of your instrument and take appropriate measures to avoid crosstalk (see Note 14). 13. The permeability of fluorescent tracers from the medium channel to the matrix channel is measured as a functional readout of barrier integrity. The maximum signal intensity is expected right after addition of the fluorophore in tight tubes. In leaky tubes, the fluorescence intensity in the medium channel is decreasing over time while the signal is increasing in the matrix channel until steady state is reached. Choose exposure time and power of the light source suited to cover the dynamic range of this process. 14. Crosstalk can occur if the spectra of several fluorophores in one sample have a significant overlap. In a multichannel application this can lead to situations where the signal recorded in one fluorescence channel is partially coming from the fluorophore of another fluorescence channel. It is recommended to use a fluorescence spectra viewer tool to select fluorophores with minimal overlap at given excitation/emission filter combination (e.g., Fluorescence SpectraViewer of Thermo Fisher Scientific). 15. For quantitative assessment of permeability of the barrier, it is necessary to run a time kinetic. The measurement needs to be started immediately after addition of the fluorophore. The leakage score is calculated for each individual timepoint and a linear regression is fitted through these ratios to calculate the slope. 16. As a quality control for proper image acquisition, save an overview of the plate image in the green and red channels for each time point. Right click on the plate in the measurement layout, select Overview—Plate and Well realistic, then select the individual fluorescence channels separately. Check the wells individually and make sure no artifact is observed (e.g., bubble formation in the perfusion channel; dextran solution not

256

He´loı¨se Ragelle et al.

homogeneously distributed in the perfusion channel). These artifacts might impact ROI values measured by the Opera Phenix® and result in misleading Pa values. 17. The in-build analysis software of standard HCA instruments often focuses on the detection of individual cells and might not allow selection of fixed regions. Here, we used a Fixed region building block that was provided as an extension to the buildin analysis tools. Please be aware that the grid of the microfluidic structures of the OrganoPlate® is sometimes not exactly aligned with the plate frame and therefore a shift in the X- and Y-position can be observed throughout the plate. This requires a shift correction in X- and Y-direction related to the rows and columns. Please refer to your technical field specialist for more information on fixed region analysis and shift correction options. Alternatively, the images obtained with the HCA instrument can be exported and analyzed in a separate image analysis software (e.g., ImageJ). 18. As a quality control for Pa assessment, inspect images for signs of premature leakage and check the value of the leakage score at t0. If the images suggest tracer permeability at t0 and the leakage score is high at t0 (0.5 or above), the barrier should be considered leaky and discarded, even though the Pa might appear low on account of a low slope due to minimum variation between t0 and t20. 19. The new format of the 2-lane OrganoPlate® (9605-400B) has slightly different dimensions than the older plate format (9603-400B) and the VMATRIX of the new format (5.45  104 cm3) is different from the VMATRIX of the previous format (4.54  104 cm3), which was used in this and in previous publications [9, 12]. Please be aware that also volumes and/or cell density might have to be adjusted when using the new plate format as channel dimensions are slightly different. 20. Be aware that the parameter Centroid in Y alone might not be sufficient for discrimination (especially if a shift in direction of the Y-axis of the grid of the microfluidic substructure of the OrganoPlate® compared with the plate frame is seen). If necessary, additional parameters can be included using the Calculate Morphology building block. Check again for the values of the endothelial tube region versus the others and select the parameter (e.g., Area and/or Roundness) that discriminates them best. 21. If additional parameters for discrimination were necessary in step 5 (such as Area and/or Roundness), include the respective selection and thresholds as well in the Select Population step.

Retinal Microvasculature-on-a-Chip

257

22. When choosing a higher magnification objective for the rescan, you have to consider the working distance (WD) of the objective (this can be found in the technical specifications). The WD describes the distance between the objective lens and the focal point. For imaging the whole tube, the WD needs to be bigger than the sum of the tube height (220 μm) and the plate bottom thickness (150 μm). 23. The Harmony software allows selection of an overview image of a test measurement as background of the well layout (select the measured fields in the well pane, right click on the well layout and choose Background for Well in the context menu). If you then switch to a high-magnification objective, the background image of the low-magnification measurement is still visible and fields of interest can be easily selected.

Acknowledgments The authors would like to thank Achim Kirsch for technical assistance in setting up imaging and image analysis workflows. References 1. Shakib M, Cunha-Vaz JG (1966) Studies on the permeability of the blood-retinal barrier: IV. Junctional complexes of the retinal vessels and their role in the permeability of the bloodretinal barrier. Exp Eye Res 5:229-IN16 2. Palm E (1947) On the occurrence in the retina of conditions corresponding to the “bloodbrain barrier”. Acta Ophthalmol 25:29–35 3. Chow BW, Gu C (2017) Gradual suppression of transcytosis governs functional blood-retinal barrier formation. Neuron 93:1325–1333.e3 4. Cunha-Vaz J, Bernardes R, Lobo C (2011) Blood-retinal barrier. Eur J Ophthalmol 21 (suppl 6):S3–S9 5. Frey T, Antonetti DA (2011) Alterations to the blood-retinal barrier in diabetes: cytokines and reactive oxygen species. Antioxid Redox Signal 15:1271–1284 6. Usui Y, Westenskow PD, Murinello S et al (2015) Angiogenesis and eye disease. Annu Rev Vis Sci 1:155–184 7. Haderspeck JC, Chuchuy J, Kustermann S et al (2019) Organ-on-a-chip technologies that can

transform ophthalmic drug discovery and disease modeling. Expert Opin Drug Discov 14: 47–57 8. Ragelle H, Goncalves A, Kustermann S et al (2020) Organ-on-a-chip technologies for advanced blood-retinal barrier models. J Ocul Pharmacol Ther 36:30–41 9. Ragelle H, Dernick K, Khemais S et al (2020) Human retinal microvasculature-on-a-chip for drug discovery. Adv Healthc Mater 9: e2001531 10. Roudnicky F, Kim BK, Lan Y et al (2020) Identification of a combination of transcription factors that synergistically increases endothelial cell barrier resistance. Sci Rep 10:3886 11. van Duinen V, van den Heuvel A, Trietsch SJ et al (2017) 96 perfusable blood vessels to study vascular permeability in vitro. Sci Rep 7: 18071 12. van Duinen V, Zhu D, Ramakers C et al (2019) Perfused 3D angiogenic sprouting in a highthroughput in vitro platform. Angiogenesis 22: 157–165

Chapter 19 Preventing VEGF-Mediated Vascular Permeability by Experimentally Potentiating BBB Characteristics in Endothelial Cells Bo Kyoung Kim, Je´re´mie Canonica, Filip Roudnicky, and Peter D. Westenskow Abstract Difficulties with poor reproducibility and translatability of animal model-based research, along with increased efforts to abide by the 3Rs tenet of animal welfare, are driving demand for more relevant human cellular systems. This is especially true for central nervous system (CNS) vasculatures with specialized properties and barriers, namely the blood-brain and blood-retinal barriers (BBB and BRB, respectively) which are difficult to model in vitro. The BBB and BRB protect neurovascular units by regulating nutrient homeostasis, maintaining local ion levels, protecting against exposure from circulating toxins and pathogens, and restricting passage of peripheral immune factors. In this manuscript, we will describe transgenic and pharmacological-based protocols to generate relevant BBB and BRB models both from human pluripotent stem cell-derived endothelial cells (hPSC-ECs) and from primary human umbilical vein endothelial cells (HUVECs). When followed, researchers can expect to generate well-characterized, anatomical and functional BBB and BRB EC monolayers in 36–48 h that are stable up to 90 h. The ability to generate more relevant BBB and BRB EC cultures will improve drug discovery efforts and inform future therapies for neurovascular disorders. Key words Vascular endothelial growth factor (VEGF), Endothelial cell barrier, Human pluripotent stem cell-derived endothelial cells, Transcription factors, RepSox, Small molecule inhibitor, Bloodbrain barrier, Blood-retinal barrier, Claudin-5

1

Introduction BBB and BRB breakdown is implicated in a wide range of devastating pathological conditions, including Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis, age-related macular degeneration, and diabetic retinopathy, respectively. Claudin-5 is highly enriched in CNS-ECs, and dose-dependent expression is a

Filip Roudnicky and Peter D. Westenskow are co-corresponding and contributed equally. Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_19, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

259

260

Bo Kyoung Kim et al.

modifying factor in neurovascular disorders [1, 2]. Improving CNS-ECs properties in pathological blood vessels deficits could improve patient outcomes in the diseases listed above. Furthermore, understanding how to better control blood barriers integrity can inform drug delivery approaches to the CNS. Therefore, developing a faithful in vitro blood barrier model is an urgent need for neuropharmaceutical research. Detailed descriptions of the rationale behind the transgenic and pharmacological assays are outlined in our previously published studies [3, 4]. Molecular and functional properties of ECs within blood vessels vary greatly depending on their vascular origin and microenvironment [5]. The organotypic profiles of vasculatures diverge during development and are driven by cell–cell communication within vascular niches. Individual properties of endothelial cells are important for delivering energy substrates and oxygen into organ systems according to their specific needs. In the CNS, strictly controlled barriers are needed to protect neurons, astrocytes, and glia from peripheral immune factors and circulating pathogens. There are two blood barriers in the retina, the outer and inner BRB. The outer barrier is generated by retinal pigment epithelium cells, a pentalaminar basement membrane and fenestrated extraretinal vasculature (the choriocapillaris). The inner barrier is generated and maintained by endothelial cells and glia. The focus of this manuscript is on the inner BRB only; please note that all subsequent references to the BRB are for the inner BRB only. In both the BBB and BRB, cell adhesion molecules and junctions are upregulated, compared with non-CNS vasculatures, to generate the physical barrier between the periphery and CNS [6, 7]. While the BBB and BRB are formed in different organ systems, they share strikingly similar overlapping anatomical and molecular profiles and do not need to be generated from different protocols. The dynamic regulation of blood barriers is accomplished by dynamically controlling expression of tight junctional proteins including Claudin-5 and Occludin. We reported previously that relevant BBB and BRB cultures can be generated by forcing expression of a cocktail of transcription factors in immature ECs (ETS1+SOX18+SOX7 and TAL1 or LEF1) [3] or with small molecule TGF-β inhibitors [3]. In particular, RepSox (a TGF-β/ALK5 inhibitor) potently upregulates Claudin-5, a tight junction protein that is highly expressed in the CNS-ECs [2]. What sets RepSox apart from other inhibitors is that it induces more CLDN5 expression (+50%) than SB431542 (+1%) after an 8 h incubation period [4]. RepSox-induced CLDN5 upregulation also correlates with increased barrier integrity compared with SB431542 [3]. In this manuscript, we outline the protocols used in the work cited above to differentiate hPSC-ECs and human primary ECs into CNS-like ECs using both transgenic and pharmacological approaches.

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

261

Using the protocols outlined in this manuscript, researchers can expect to reproducibly generate CNS-like ECs in 36–48 h that remain stable up to 90 h. When deciding which cell type to use (hPSC-ECs or primary ECs) the reader should consider the advantages and disadvantages of both. One major advantage to using hPSCs is that hPSCs are much more amenable to genomic editing approaches (in the hPSC stage), cryopreservation, and repeated passaging than primary ECs are. However, the use of hPSCs requires more technical expertise than primary ECs do, and differentiating hPSCs to ECs requires more time and effort than it takes to obtain primary EC lines and to maintain them. Finally, the protocol described here can be adapted to facilitate drug screening (by using miniaturized culture formats), and for other EC lines including primary human retinal microvascular cells or primary human brain microvascular cells [4]. The ability to work with robust and adaptable BBB and BRB models may enable CNSEC-based studies and even inform future therapies for neurovascular disorders.

2

Materials

2.1 Stock Solution Preparation

1. Adenovirus-encoding transcription factors: ETS1, SOX18, SOX7, TAL1, and LEF1 constructs are aliquoted into small microfuge tubes. Store at
80  C until ready to use (see Note 1). 2. RepSox (Tocris; 3742): Reconstitute in sterile dimethyl sulfoxide (DMSO) (Sigma; S-002-M) at 10 mM concentration and aliquot into small microfuge tubes. Store tubes at
20  C until ready to use (see Note 2). 3. SB431542 (Tocris; 1614): Reconstitute in sterile DMSO at 10 mM concentration and aliquot into small microfuge tubes. Store tubes at
20  C until ready to use (see Note 3). 4. VEGF-A recombinant protein (PEPROTECH): Reconstitute VEGF-A recombinant (100 μg/mL stock concentration) in phosphate buffer saline without calcium and magnesium (PBS
/
) (Gibco) supplemented with 0.1% (w/v) bovine serum albumins (BSA) (Sigma; A1933) and aliquot solutions into small microfuge tubes. Flash freeze aliquots in liquid nitrogen. Store at
80  C until ready to use (see Note 4). 5. 40 kDa Fluorescein isothiocyanate (FITC)-dextran solution (Sigma; 46944): Dissolve FITC-dextran powder in sterile distilled water at 25 mg/mL and store FITC-dextran solution aliquots in light-shielded 1.5 mL microfuge tubes at
20  C until ready to use (see Note 5). 6. Fibronectin solution: Reconstitute Fibronectin, e.g., Corning® Fibronectin (Corning; 356008) (1 mg/mL) in phosphate

262

Bo Kyoung Kim et al.

buffer saline with calcium and magnesium (PBS+/+) (Gibco) and aliquot solutions into small microfuge tubes at
20  C until ready to use (see Note 6). 2.2 Endothelial Cell Culture

1. VascuLife VEGF Endothelial Medium Complete Kit (Lifeline Cell Technology; LL-0003) or EGMTM-2 Endothelial Cell Growth Medium-2 BulletKitTM (Lonza; CC-3162) for hPSC-ECs or primary ECs, respectively (see Note 7). 2. 1 mg/mL Fibronectin solution. 3. Cell dissociation solution, e.g., Accutase (StemCell Technologies; 07920). 4. 100 mm Tissue Culture-treated Culture Dish (Corning). 5. Phosphate buffer saline without calcium and magnesium (PBS
/
) (Gibco). 6. Hemocytometer or automated cell counter, e.g., TC20™ Automated Cell Counter (BioRad; 1450102). 7. Trypan Blue solution, e.g., Trypan Blue Dye, 0.40% solution (BioRad; 1450021). 8. 50 mL Falcon tubes (Falcon). 9. Micropipettes.

2.3 Transendothelial Electrical Resistance (TEER) Assay

1. Voltmeter to measure transendothelial resistance, e.g., cellZscope from nanoAnalytics system for real-time measurements (see Note 8). 2. HTS Transwell 24-well format, e.g., HTS Transwell® 24-well Permeable Support (Corning; 3378). 3. A 24-well plate, e.g., Costar® 24-well Clear Multiple Well Plates (Corning; 3524). 4. Sterile forceps. 5. 1 mg/mL Fibronectin solution. 6. Phosphate buffer saline without calcium and magnesium (PBS
/
) (Gibco). 7. Hemocytometer or automated cell counter, e.g., TC20™ Automated Cell Counter (BioRad; 1450102). 8. Trypan Blue solution, e.g., Trypan Blue Dye, 0.40% solution (BioRad; 1450021). 9. Micropipettes.

2.4 Fluorescein Isothiocyanate (FITC)Dextran Permeability Assays

1. 25 mg/mL solution.

Fluorescein

isothiocyanate

(FITC)-dextran

2. Transwell 96-well format, e.g., HTS Transwell® 96-well Permeable Support (Corning; 3391). 3. 1 mg/mL Fibronectin solution.

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

263

4. Microplate reader (see Note 9). 5. Hemocytometer or automated cell counter, e.g., TC20™ Automated Cell Counter (BioRad; 1450102). 6. Trypan Blue solution, e.g., Trypan Blue Dye, 0.40% solution (BioRad; 1450021). 7. Micropipettes. 2.5 Immunocytochemistry

1. 4% (v/v) Paraformaldehyde Solution. 2. Blocking solution: SuperBlock™ (PBS) Blocking Buffer (Thermo; 37515) supplemented with 0.3% (v/v) Triton X-100 (Sigma; X100) (see Note 10). 3. Washing buffer: PBS
/
supplemented with 0.05% (v/v) TWEEN 20 (Sigma; 93773). 4. Claudin-5 Polyclonal Antibody (Invitrogen; 34-1600). 5. CD31, Endothelial Cell Antibody (DAKO; GA61061-2). 6. Goat anti-Rabbit IgG (H+L), Alexa Fluor® 647 (Invitrogen; A27040). 7. Goat anti-Mouse IgG (H+L), Alexa Fluor® 488 (Invitrogen; A28175). 8. Nuclear counterstaining fluorescent dye, e.g., NucBlue™ Fixed Cell ReadyProbes™ Reagent (Thermo; R37606). 9. Phosphate Buffer Saline (PBS) supplemented with calcium and magnesium (PBS+/+). 10. Imaging plate, e.g., Falcon® 96-well Black/Clear Flat Bottom TC-treated Imaging Microplate (Corning; 353219). 11. 1 mg/mL Fibronectin solution. 12. Plate sealer, e.g., PlateSeal (Perkin Elmer; 1450-462). 13. Fluorescence microscope (see Note 11). 14. Hemocytometer or automated cell counter, e.g., TC20™ Automated Cell Counter (BioRad; 1450102). 15. Trypan Blue solution, e.g., Trypan Blue Dye, 0.40% solution (BioRad; 1450021). 16. Micropipettes.

2.6 Working Stock Preparation for Promoting Barrier Resistance

1. 10 concentrated adenoviral cocktail solution for treatment: Prepare 10 concentrated adenoviral quadruple cocktail at 800 MOI (200 MOI of ETS1 + SOX18 + SOX7 and 200 MOI of TAL1 or LEF1) with complete media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium. This concentration is 10 concentrated solution for treatment (see Note 12). The final working concentration is 80 MOI in total.

264

Bo Kyoung Kim et al.

2. 10 concentrated RepSox solution for treatment: Dilute 10 mM RepSox or SB431542 at 100 μM with complete media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium. This concentration is 10 concentrated solution for treatment. The final working concentration is 10 μM.

3

Methods

3.1 Endothelial Cell Maintenance (see Note 13)

1. Coat 100 mm dish with 6 mL of fibronectin (25 μg/mL final concentration) for 2 h at room temperature. 2. After the 2 h incubation, remove fibronectin solution just before plating cells and the dishes will be ready to use without washing steps. 3. Wash the cells with PBS
/
twice and add 6 mL of Accutase or equivalent solution for cell dissociation. 4. After 7 min incubation with Accutase or equivalent solution for cell dissociation at room temperature, add 14 mL of VascuLife VEGF Endothelial Medium or EGMTM-2 Endothelial Cell Growth Medium and centrifuge the collected cell suspension in a 50 mL Falcon tube at 300  g for 10 min. 5. Resuspend cell pellet with 5 mL of fresh complete media and count cell number with automated cell counter or hemocytometer. 6. Plate hPSC-ECs or primary ECs at a density of 20,000 cells per cm2 in complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium, respectively. 7. Exchange media every other day until cells reach 80–90% confluence. 8. Passage hPSC-ECs or primary ECs at a density of 20,000 cells per cm2 with complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium.

3.2

TEER Assay

1. Coat HTS Transwell® 24-well permeable supports with 100 μL of fibronectin (25 μg/mL final concentration) for 2 h at room temperature. After the 2 h incubation, remove fibronectin solution just before plating cells and the supports will be ready to use without washing steps (see Note 8). 2. Add 950 μL of complete media to bottom stainless steel electrodes pot (basolateral chamber) in cellZscope system (see Note 8).

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

265

3. Take the maintained ECs in 100 mm dish, wash the cells with PBS
/
twice and add 6 mL of Accutase or equivalent solution for cell dissociation. 4. After 7 min incubation with Accutase or equivalent solution for cell dissociation at room temperature, add 14 mL of VascuLife VEGF Endothelial Medium or EGMTM-2 Endothelial Cell Growth Medium and centrifuge the collected cell suspension in a 50 mL Falcon tube at 300  g for 10 min. 5. Resuspend cell pellet with 5 mL of fresh complete media and count cell number with automated cell counter or hemocytometer. 6. Place fibronectin-coated transwell permeable supports into stainless steel electrodes pots including 950 μL of complete media, and seed hPSC-ECs or human primary ECs on fibronectin-coated HTS Transwell® 24-well permeable supports at 120,000 cells/membrane and incubate for 48 h at 37  C. Total media volume for upper chamber is 350 μL (see Notes 14 and 15). 7. Start real-time TEER recording (see Note 8). 8. After 48 h at 37  C, pause the real-time measurement setting in cellZscope, and carefully transfer each transwell permeable supports to empty 24-well plate using sterile forceps. 9. Slowly remove old media from upper chamber from each transwell and add fresh 350 μL of complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium (see Note 16). 10. Remove old media from stainless steel electrodes pot (basolateral chamber) and add fresh 950 μL of complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium. 11. Carefully move transwell permeable supports back to stainless steel electrodes pot from empty 24-well plate. 12. Add 35 μL of freshly prepared 10 concentrated viral cocktail (800 MOI) or 10 concentrated RepSox solution (100 μM) to each well at final concentration of 80 MOI or 10 μM respectively, to induce ECs resistance for 36 h at 37  C. At this time point ECs exhibit the highest resistance values (see Notes 3 and 17). 13. Restart real-time TEER recording (see Note 19). RepSox and quadruple transcription factor combinations, which increases expression of Claudin-5, improves barrier properties, as exemplified by potentiation of TEER values (Figs. 1a, c and 3).

RepSox

B

DMSO

RepSox

12 18 24 30 36 42 48

C Empty

50

0

D M SB SO 43 15 42 R ep So x

0

100

D

SO

6

50

*

M

0

100

150

D

50

*

SB

100

150

w/ VEGF-A FITC-dextran permeability (ctrl=100%)

150

Permeability FITC-dextran (DMSO = 100 %)

200

Run Time (h)

Empty

ETS1+SOX18+SOX7+LEF1 ETS1+SOX18+SOX7+TAL1

ETS1+SOX18+SOX7+LEF1 ETS1+SOX18+SOX7+TAL1

w/o VEGF-A

6

12 18 24 30 36 42 48

0

ET S1 +S O ET S1 X18 +S Em +S pt O O y X X1 8+ 7+L EF SO 1 X7 +T AL 1

Run Time (h)

50

0

8 E SO + SO m p ty X X1 8+ 7+L E SO F1 X7 +T AL 1

0

50

SO X1

80

100

S1 +

100

100

** **

ET

120

Permeability FITC-dextran (DMSO = 100 %)

140

** *

w/ VEGF-A 150

S1 +

160

150

ET

180

Permeability FITC-dextran (DMSO = 100 %)

Relative TEER value [%]

SB431542

w/o VEGF-A

42 ep So x

SB431542

R

DMSO

15

A

43

Bo Kyoung Kim et al.

Relative TEER value [%]

266

Fig. 1 Pharmacological and transgenic approaches improve EC resistance and limit transcellular permeability. (a) Time-course measurements of TEER in hPSC-ECs treated with 10 μM RepSox, 10 μM SB431542 (TGF-β inhibitor), or DMSO (vehicle). Exogenous addition of RepSox (10 μM) potentiated TEER values up to +56% in hPSC-ECs after 48 h compared to vehicle controls (DMSO) (0.001%) while the exogenous addition of SB431542 (10 μM), which does not increase expression of Claudin-5, reduced TEER (
25%). (b) FITC-dextran permeability assays performed 48 h after treating hPSC-ECs with 10 μM RepSox, 10 μM SB431542, or vehicle (left) and co-treatment of 50 ng/mL VEGF-A (right). RepSox significantly reduced FITC-dextran permeability (
40%) (left). VEGF-A induced permeability was prevented by RepSox-treated condition (right). (c) Time-course measurements of TEER in hPSC-ECs transduced with either ETS1+SOX18+SOX7+TAL1 or ETS1+SOX18+SOX7+LEF1 cocktail (80 MOI). Both quadruple transcription factor combinations (ETS1+SOX18+SOX7+TAL1 or LEF1) improve barrier resistance in hPSC-ECs up to 57% and 52% respectively, at 36 h post-transduction compared to empty vector-treated condition. (d) FITC-dextran permeability assays performed 48 h after transducing hPSCECs with ETS1+SOX18+SOX7+TAL1 or LEF1 cocktail (80 MOI) (left) and the co-treatment of 50 ng/mL VEGF-A (right). FITC-dextran assay confirmed that both quadruple combination including TAL1 or LEF1 significantly reduced paracellular permeability by
60% and
47%, respectively; also VEGF-A-induced barrier permeability was significantly suppressed (
46%, and
54%). Permeability assays and voltmeter resistance measurements were performed in triplicate. Data in columns represent mean  SD. Lines represent means, and shadows represent SD. *P < 0.05, **P < 0.01, ***P < 0.001

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

3.3 FITC-Dextran Permeability Assays

267

1. Coat HTS Transwell® 96-well permeable supports with 50 μL of fibronectin (25 μg/mL final concentration) for 2 h at room temperature. After the 2 h incubation, remove fibronectin solution just before plating cells and the supports will be ready to use without washing steps. 2. Add 235 μL of complete media to bottom chamber of HTS Transwell® 96-well. 3. Take the maintained ECs in 100 mm dish, wash the cells with PBS
/
twice, and add 6 mL of Accutase or equivalent solution for cell dissociation. 4. After 7 min incubation with Accutase or equivalent solution for cell dissociation at room temperature, add 14 mL of VascuLife VEGF Endothelial Medium or EGMTM-2 Endothelial Cell Growth Medium and centrifuge the collected cell suspension in a 50 mL Falcon tube at 300  g for 10 min. 5. Resuspend cell pellet with 5 mL of fresh complete media and count cell number with automated cell counter or hemocytometer. 6. Fill the desired number of wells from a Transwell® 96-well plates with 235 μL of complete media, and place fibronectincoated transwell permeable supports into those wells. Seed hPSC-ECs or human primary ECs on the upper surface of the fibronectin-coated HTS Transwell® 96-well permeable supports at 10,000 cells/membrane. Add VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium (75 μL) to the supports and incubate for 48 h at 37  C (see Notes 14 and 15). 7. After 48 h at 37  C, carefully transfer the transwell permeable support plate to the receiver plate (receiver plates are provided from the HTS Transwell® 96-well Permeable Support kit set). 8. Slowly remove old media from upper chamber from each transwell and add fresh 75 μL of complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium (see Note 16). 9. Remove old media from bottom chamber and add fresh 235 μL of complete culture media either VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium. 10. Carefully transfer transwell permeable supports back to bottom chamber plate including 235 μL of complete media from receiver plate. 11. Add 7.5 μL of freshly prepared 10 concentrated viral cocktail (800 MOI) or RepSox solution to each upper chamber at final concentration of 80 MOI or 10 μM respectively, to induce ECs resistance for 36 h at 37  C. At this time point ECs exhibit the highest resistance values (see Notes 3 and 17).

268

Bo Kyoung Kim et al.

12. At the day of measurement (24–36 h post-treatment), refresh the media in both upper (75 μL) and lower chambers (235 μL) prior to starting (see Notes 18 and 19). 13. Add 10 μL of 40 kDa (25 mg/mL final concentration) FITCdextran to the upper chamber. 14. After 30 min of incubation at 37  C with 5% CO2, remove the transwell support from the plate, and measure the fluorescence signal from the bottom chamber with a plate reader (excitation 485 nm, emission 535 nm) (see Note 9). RepSox and quadruple transcription factor combinations, which increases expression of Claudin-5, improves barrier properties, as exemplified by suppressed FITC-dextran signal from bottom chamber (Figs. 1b, d and 3). 3.4 Immunocytochemistry

1. Coat imaging plate with 100 μL of fibronectin (25 μg/mL final concentration) for 2 h at room temperature. After the 2 h incubation, remove fibronectin solution just before plating cells and the plates will be ready to use without washing steps. 2. Plate hPSC-ECs or human primary ECs on fibronectin-coated imaging plate at 12,000 cells/well with the VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium and incubate for 48 h at 37  C. 3. After 48 h at 37  C, refresh 100 μL of the medium with VascuLife VEGF Endothelial Medium or EGM™-2 Endothelial Cell Growth Medium. 4. Add freshly prepared 10 concentrated viral cocktail (800 MOI) or RepSox solution to each well at final concentration of 80 MOI or 10 μM respectively, to induce ECs resistance for 36 h at 37  C (see Notes 3 and 17). 5. After 36 h, wash the plate two times with cold PBS+/+ and fix cells with 4% (v/v) PFA for 20 min at room temperature (see Note 19). 6. Rinse cells twice with PBS+/+ and add blocking buffer containing 0.3% (v/v) Triton-X for 1 h at 4  C on a shaker (see Note 10). 7. Add 5 μg/mL primary antibodies diluted in blocking buffer and incubate overnight at 4  C on a shaker (see Note 20). 8. The next day wash cells three times with washing buffer, PBS-T (PBS
/
+ 0.05% (v/v) TWEEN 20) and add secondary antibodies 2 μg/mL diluted in blocking buffer. 9. After 3 h incubation at 4  C on a shaker, wash three times with PBS
/
and add 0.2 μg/mL DAPI at the end of the washing step. 10. Wash the imaging plate with PBS
/
twice and add 100 μL of PBS+/+.

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

CLDN5

Relative Arbitrary Unit (Normalized to DMSO)

150

**

B DMSO SB431542 RepSox

100

50

0

CLDN5

150

Relative Arbitrary Unit (Normalized to DMSO)

A

*** **

269

Empty ETS1+SOX18+SOX7+LEF1 ETS1+SOX18+SOX7+TAL1

100

50

CLDN5

CD31

Merge

RepSox SB431542

DMSO

C

ET S1 + ET SO S1 X1 +S 8+ Em S O X1 OX p 8+ 7+ ty SO LE X7 F1 +T AL 1

D M SB SO 43 15 42 R ep So x

0

TAL1 mix LEF1 mix

Empty

D

Fig. 2 Pharmacological and transgenic approaches alter expression of Claudin-5 protein expression around EC cell junctional complexes. (a, b) Quantification of the fluorescent intensity from Immunocytochemistry images for each treatment condition. RepSox-treated condition showed +26% increased signal intensity compared to DMSO-treated condition. Also both quadruple combination including TAL1 or LEF1 increased signal intensity by up to +20% and +18% compared to empty vector-treated condition. (c, d) Representative images for each treatment condition. Data in columns represent mean  SD. **P < 0.01, ***P < 0.001

270

Bo Kyoung Kim et al.

11. Cover the plate with plate sealer. 12. Image the prepared samples with fluorescence microscope (see Note 11). 13. Use open-source image analysis tool for further analysis, e.g., Image J or CellProfiler (see Note 21). RepSox and quadruple transcription factor combinations increases expression of Claudin-5, that correlates with improved barrier properties. SB431542 and empty adenovirus are negative controls, which do not increase expression of Claudin-5 (Fig. 2).

4

Notes 1. Adenovirus serotype Ad5 vectors expressing individually transcritption factors were generated by SIRION Biotech [8]. 2. Weigh 5 mg of RepSox and dissolve it in 1.74 mL of DMSO to make 10 mM stock solution. 3. 10 μM SB431542 can be used as a negative control for BBB and BRB induction [4]. Running the experiments in triplicate is strongly recommended. Dissolve 5 mg of SB431542 in 1.3 mL of DMSO to make 10 mM stock solution. 4. Avoid freeze-thaw cycles. Once an aliquot of VEGF-A has been thawed, keep the vial at 4  C. We do not recommend to use VEGF-A kept at 4  C for more than a week. Reconstitute 50 μg of VEGF-A lyophilized recombinant proteins in 500 μL of suspension buffer composed of phosphate buffer saline without calcium and magnesium (PBS
/
) supplemented with 0.1% (w/v) bovine serum albumins (BSA). Aliquot VEGF-A solutions into small microfuge tubes and store at
80  C until ready to use. 5. Weigh 25 mg of FITC-dextran powder and dissolve it in 1 mL of sterile water. 1 mL of 25 mg/mL FITC-dextran solution is sufficient for a 96-well plate format assay. 6. Reconstitute 5 mg of lyophilized fibronectin in 5 mL of PBS+/+. 7. We used VascuLife VEGF Endothelial Medium Complete Kit for hPSC-ECs [9, 10] and EGMTM-2 Endothelial Cell Growth Medium-2 BulletKitTM for primary EC line, e.g., HUVECs. This protocol can also apply to primary retinal or brain microvascular endothelial cells [4]. 8. Researchers can use any voltmeter to measure transendothelial resistance using standard protocols [11]. In this study, we used the cellZscope from nanoAnalytics system for real-time measurements. If using an automated voltmeter, adjust settings to record resistance values every 30 min over the course of 48–72 h. Depending on the insert type, and manufacturer of

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

271

the transwell support, media volumes for apical and basolateral chamber might vary. User should check media volume requirement before starting the assay. We also recommend measuring basal TEER value before starting treatment. 9. We used a Microplate Reader: Infinite® M1000 PRO (Tecan). 10. We used SuperBlock™ (PBS) Blocking Buffer (Thermo; 37515) as a blocking buffer. Alternatively, 5% BSA diluted in PBS
/
or 10% goat serum can be used for this step. 11. We used an Operetta CLS™ high-content analysis system (Perkin Elmer) and Harmony software (version 4.9, Perkin Elmer) for imaging and changed the settings on the device to detect the following: Alexa Fluor 647 (Ex 650 nm/Em 665 nm), Alexa Fluor 488 (Ex 490 nm/Em 525 nm) and DAPI (Ex 350 nm/Em 470 nm). 12. Multiplicity of infection (MOI) refers to plaque forming units (PFU) of virus used for infection/number of cells. The user should determine experimentally the optimal MOI for ECs when using adenovirus from different suppliers. For the cell line transductions used in this paper, Multiplicity of Infection (MOI) was calculated following this formula: Viral particle volume treated ðμLÞ ¼

MOI  the number of cells plated   Titre virus particles μL

13. Refer to these references for protocols to convert SA001 hPSCs into EC-fates [9, 10]. 14. For TEER assay or FITC-dextran permeability experiments represented in Fig. 3, human umbilical vein endothelial cells (HUVECs) were incubated under starvation conditions (0.5% FBS and no VEGF-A, supplemented in the basal EC medium e.g. EBMTM-2 Endothelial Cell Growth Medium) prior to compound treatments in order to enhance the effect of VEGF-A-induced vascular permeability and its recovery. 15. Cell density is important. From start to completion of the experiment, these protocols require 96 h. Therefore, cells should be seeded appropriately to avoid overgrowth. Thus, cell density needs to be determined empirically in each lab depending on the type of ECs used. 16. It is critical not to touch the bottom of transwell semipermeable support when you refresh media in order to avoid high well-to-well variation. 17. OPTIONAL: Add VEGF-A to the upper chamber of the transwell to a final concentration of 50 ng/mL without refreshing the media. You may choose to prepare the VEGF-

272

Bo Kyoung Kim et al.

A

B

Untreated VEGF-A (50 ng/mL)

✱✱✱

VEGF-A + anti-IgG (20Pg/mL)

120

Relative TEER value [%] (24 h Post-treatment)

VEGF-A + RepSox (10PM)

100

80

✱✱✱

150

✱✱✱ ✱✱✱

✱✱✱ ✱✱✱ ✱✱✱

100

50

(5 0

U nt

Ig G

ti-

an

VE

G

VE

FA

G

FA

Run Time (h)

+

48

FA

Treatment

36

G

24

+

VEGF-A 50 ng/mL

12

VE

0

re at

60

ed ng /m tiL) VE VEG (20 P G F g -A F/m A L) (1 + 0 R P ep g/ m So L) x (1 0 PM )

0

an

Relative TEER value [%]

✱✱✱

VEGF-A + anti-VEGF-A (10Pg/mL)

Fig. 3 RepSox treatment rescues VEGF-A-induced endothelial barrier breakdown. (a) Time-course measurements of TEER using the CellZscope technology in HUVECs cultured on transwell filters. To induce endothelial barrier breakdown, HUVECs were treated (upper and lower chambers) with 50 ng/mL of VEGF-A for 24 h, and TEER values were measured over time. Following 24-h VEGF-A treatment, 20 μg/mL of anti-IgG, 10 μg/mL of anti-VEGF-A, and 10 μM RepSox compounds were added to the culture wells, and TEER values recording continued. (b) Quantification of TEER values at 24 h after compound treatments. All relative TEER values were normalized to TEER values from untreated condition at starting point (Run Time ¼ 0). Treatment of mature HUVECs monolayers cultured on transwell with VEGF-A decreased the endothelial barrier integrity over time (
25%), as observed by the decrease in TEER when compared to untreated control cells. RepSox reversed the EC monolayer permeability induced by VEGF-A and increased endothelial barrier integrity by +66% at 24-h post-VEGF-A treatment. The experimental condition in which HUVECs were treated with an anti-VEGF-A antibody served as a positive control for VEGF-A-induced endothelial barrier breakdown rescue. Anti-VEGF-A antibody represented 33% higher TEER value than that of VEGF-A-treated HUVECs at 24-h post-VEGF-A treatment. Anti-VEGF-A treatment reverted back TEER values of ECs monolayers to values of untreated control cells or to values before the addition to the cell culture of VEGF-A. Addition of RepSox 24 h after VEGF-A treatment not only reversed VEGF-A-induced endothelial barrier breakdown but improved endothelial barrier integrity in HUVECs in a direction towards those seen in CNS vasculatures. Also, TEER values were significantly increased by 17% and 21% when compared to untreated control cells and anti-VEGF-A-treated HUVECs, respectively. Values are means  SD, n ¼ 6–12 per time-point measurements. ***P < 0.001

A stock to a 10 solution (500 ng/mL) beforehand to avoid disturbing the cells with larger volume of the media. Minimum triplicate per condition recommended. 18. Repeat the medium change procedure described above steps 7–10. 19. Timing is important, and in our hands, ECs treated with either the transgenic or the pharmacological protocols display the highest barrier resistances at 36 or 48 h post-treatment, respectively (Fig. 1a, c).

Potentiating BBB Properties to Reduce VEGF-Mediated Pathological Vascular Permeability

273

20. In this book chapter, we used Claudin 5 Polyclonal Antibody (Invitrogen; 34-1600), CD31, Endothelial Cell (DAKO; GA61061-2) for primary antibody, Goat anti-Rabbit IgG (H +L), Alexa Fluor® 647 (Invitrogen; A27040), Goat antiMouse IgG (H+L), Alexa Fluor® 488 (Invitrogen; A28175) for secondary antibody, and Nuclear counterstaining fluorescent dye, e.g., NucBlue™ Fixed Cell ReadyProbes™ Reagent (Thermo; R37606) for nuclear counterstaining. 21. Use image analysis tool such as Image J or CellProfiler for analyzing the data obtained from images. Download Image J software from: https://imagej.nih.gov/ij/ or CellProfiler from: https://cellprofiler.org/releases.

Acknowledgments We gratefully acknowledge Martine Kapps, Nicole Soder, Silke Zimmerman, Pamela Strassburger, and Olivier Partouche for technical support. Disclosures F. Hoffmann-La Roche Ltd supported this study. F. R., B.K.K., J.C., and P.D.W. are employees of F. Hoffmann-La Roche. B.K.K. is supported by a Roche doctoral fellowship. J.C., is supported by Roche postdoctoral fellowship.

References 1. Greene C, Kealy J, Humphries M, Gong Y, Hou J, Hudson N, Cassidy L, Martiniano R, Shashi V, Hooper S (2018) Dose-dependent expression of claudin-5 is a modifying factor in schizophrenia. Mol Psychiatry 23(11): 2156–2166 2. Greene C, Hanley N, Campbell M (2019) Claudin-5: gatekeeper of neurological function. Fluids Barriers CNS 16(1):3 3. Roudnicky F, Kim BK, Lan Y, Schmucki R, Ku¨ppers V, Klaus C, Martin G, Patsch C, Mark B, Claas Aiko M (2020) Identification of a combination of transcription factors that synergistically increases endothelial cell barrier resistance. Sci Rep 10(1):3886 4. Roudnicky F, Zhang JD, Kim BK, Pandya NJ, Lan Y, Sach-Peltason L, Ragelle H, Strassburger P, Gruener S, Lazendic M, Uhles S, Revelant F, Eidam O, Sturm G, Kueppers V, Christensen K, Goldstein LD, Tzouros M, Banfai B, Modrusan Z, Graf M, Patsch C, Burcin M, Meyer CA, Westenskow PD, Cowan CA (2020) Inducers of the

endothelial cell barrier identified through chemogenomic screening in genome-edited hPSC-endothelial cells. Proc Natl Acad Sci 117:19854 5. Aird WC (2012) Endothelial cell heterogeneity. Cold Spring Harb Perspect Med 2(1): a006429 6. Abbott NJ, Patabendige AA, Dolman DE, Yusof SR, Begley DJ (2010) Structure and function of the blood–brain barrier. Neurobiol Dis 37(1):13–25 7. Runkle EA, Antonetti DA (2011) The bloodretinal barrier: structure and functional significance. Methods in molecular biology (Clifton, NJ) 686:133–148 8. Jager L, Hausl MA, Rauschhuber C, Wolf NM, Kay MA, Ehrhardt A (2009) A rapid protocol for construction and production of highcapacity adenoviral vectors. Nat Protoc 4(4): 547 9. Christensen K, Roudnicky F, Burcin M, Patsch C (2019) Monolayer generation of vascular

274

Bo Kyoung Kim et al.

endothelial cells from human pluripotent stem cells. Methods Mol Biol 1994:17–29 10. Patsch C, Challet-Meylan L, Thoma EC, Urich E, Heckel T, O’Sullivan JF, Grainger SJ, Kapp FG, Sun L, Christensen K (2015) Generation of vascular endothelial and smooth muscle cells from human pluripotent stem cells. Nat Cell Biol 17(8):994–1003

11. Theile M, Wiora L, Russ D, Reuter J, Ishikawa H, Schwerk C, Schroten H, Mogk S (2019) A simple approach to perform TEER measurements using a self-made volt-amperemeter with programmable output frequency. J Vis Exp (152):e60087

Chapter 20 The Embryonic Mouse Hindbrain and Postnatal Retina as In Vivo Models to Study Angiogenesis Alessandro Fantin and Christiana Ruhrberg Abstract Angiogenesis, the growth of new blood vessels from pre-existing ones, is a fundamental process for organ development, exercise-induced muscle growth, and wound healing, but is also associated with different diseases such as cancer and neovascular eye disease. Accordingly, elucidating the molecular and cellular mechanisms of angiogenesis has the potential to identify new therapeutic targets to stimulate new vessel formation in ischemic tissues or inhibit pathological vessel growth in disease. This chapter describes the mouse embryo hindbrain and postnatal retina as models to study physiological angiogenesis and provides detailed protocols for tissue dissection, sample staining and analysis. Key words Angiogenesis, Hindbrain, Retina

1

Introduction Angiogenesis is a tightly regulated process involving a range of cell types and signaling pathways, and therefore the choice of the right experimental model is critical to uncover molecular and cellular mechanisms that are of physiological relevance. Many different experimental settings have been exploited, ranging from the simplest in vitro models that examine specific parameters implicated in angiogenesis, to complex in vivo models that recapitulate key features of pathological conditions with a prominent neovascular component, such as murine models of solid cancers or ischemic eye pathology. Yet, much knowledge of cellular and molecular mechanisms controlling vascular growth and patterning has been provided by studies of vascular growth in developing organs. In particular, several recent studies into the mechanisms of angiogenesis utilized the mouse embryo hindbrain (e.g., [1–11]) and the perinatal mouse retina (e.g., [9–17]).

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_20, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

275

276

Alessandro Fantin and Christiana Ruhrberg

1.1 The Mouse Embryo Hindbrain

The embryonic central nervous system is vascularized by angiogenic sprouting in a highly stereotypical fashion [1, 3, 18–21]. Vascularization of the developing mouse brain starts between 9.5 and 10.0 days post coitum (dpc); during that time, the limiting membrane on the brain surface is degraded focally and capillary sprouts extend radially from the perineural vascular plexus into the neuroectodermal tissue in response to vascular endothelial growth factor (VEGF) A secreted by neuroprogenitors [22–24]. The capillary sprouts consist of solid cords of cells composed of a leading tip cell that extends filopodia and 1 or 2 trailing stalk cells [1, 13, 25]. Next, capillary sprouts elongate towards the ventricular surface, reaching the subventricular zone (SVZ) of the brain with its neuroprogenitors by 10.5 dpc (Fig. 1b). Blood vessel sprouts then change direction by turning approximately 90 to grow laterally. Subsequently, these vessels branch and anastomose with adjacent sprouts to form a vascular plexus that supplies the SVZ and is termed the subventricular vascular plexus (SVP) [1, 3, 19, 20]. Between 10.5 and 12.5 dpc, the SVP can be visualized easily when the roof plate is opened and the hindbrain is flat-mounted and imaged from the dorsal side (Fig. 1a–d) [1, 3, 19, 20]. The main advantage of the hindbrain as a model system for angiogenesis is that growth follows a spatially and temporally welldefined sequence of events [20]. Moreover, the hindbrain can be easily dissected and used for whole-mount immunostaining, ligand binding assays, and mRNA in situ hybridization techniques, with the 2D planar nature of the nascent SVP greatly facilitating imaging of the growing blood vessels [20]. In addition, blood vessel density in the hindbrain can be readily quantified in flat-mounts as the number of microvessel branchpoints and intersections when it is imaged from the ventricular side (Fig. 1a–c, f), or as the number of vessel sprouts that invade the hindbrain from the perineural plexus when it is imaged from the ventral side (Fig. 1e) [1, 3, 19, 20]. Finally, this model system allows the study of mice lacking specific genes implicated in vessel development if they survive the period of vasculogenesis, even if the mutation causes lethality after 12.5 dpc [20]. For example, the hindbrain model has been used to describe angiogenesis defects in mice expressing only the soluble VEGF120 isoform of VEGF-A, but lacking the heparin/neuropilin (NRP) binding VEGF164 isoform; these mice die within the first 2 weeks after birth (Fig. 3a, b) [26]. In addition, the hindbrain analysis allowed the characterization of mice lacking the VEGF164 receptor NRP1, which die between 10.5 and 14.5 dpc depending on the genetic background [11, 27–29]. In cases where embryonic lethality in constitutive knockout mice occurs prior to hindbrain angiogenesis, modern genetic techniques, such as constitutive or tamoxifen-inducible Cre/lox technology, can be used to allow conditional gene deletion only in selected cell lineages or just prior to hindbrain angiogenesis [30].

Mouse Models of Physiological Angiogenesis

277

Fig. 1 (a) Schematic of a 12.5 dpc hindbrain dissection. The embryonic brain is shown in gray. Cut the embryo head just above the forearms (step 1) and remove the face (step 2). Sever the skin and the pial membrane of the roof plate covering the fourth ventricle in the dorsal side of the head (step 3) and continue to cut toward the midbrain (step 4) and the spinal cord (step 5), allowing the not-yet-fused developing neural tube to unfold. After removal of the head mesenchyme and the pial membrane from the ventral side of the neural tube (step 6), separate the midbrain (step 7) and beginning of the spinal cord (step 8) from the unfurled hindbrain. (b–g) IB4 staining of the SVZ in 10.5 (b), 11.5 (c), and 12.5 dpc (d) hindbrains using a 10 objective. (e) High magnification image of the SVZ of a 10.5 dpc hindbrain with a 63 objective. (f) IB4 staining of the pial side of a 11.5 dpc hindbrain. (g) Visualization of the vascular intersections (blue dots) by AngioTool in the 12.5 dpc hindbrain shown in (d). Scale bars: 25 μm (b–d, f, g), 50 μm (e)

1.2 The Postnatal Mouse Retina

Three different types of vascular networks support eye development: choroid, hyaloid, and retinal vessels. While the first two networks form during embryogenesis to support eye specification and growth, the retinal vasculature appears only after birth,

278

Alessandro Fantin and Christiana Ruhrberg

Fig. 2 (a) Schematic of a P7 retina dissection. Grab the cornea with the forceps (step 1) and cut along the border between the sclera and the iris (white dots) using the spring scissors (step 2). After dissecting the cornea, remove the iris (step 3) and pull the lens and the hyaloid vasculature out with the forceps (step 4). Carefully peel away from the retina the outer layers of the eye, which comprise the sclera and the retinal pigmented epithelium (rpe; step 5). Make four radial incisions (step 6) and position the retina with the inner side facing upward. (b) IB4 staining of a P7 retina imaged with a stereomicroscope with 1.6 magnification. (c) Confocal imaging of the same retina shown in (b) with a 10 objective. (d) High magnification imaging of the retina vascular front with a 40 objective. Scale bars: 1 mm (b), 200 μm (c), 50 μm (d)

concomitant with hyaloid vessel regression (reviewed in [31, 32]). Retinal vessels emerge from the optic nerve head at the center of the retina and extend outward to form a two-dimensional planar network, variably called the primary, superficial or inner vascular plexus (Fig. 2); these vessels follow a gradient of VEGF-A deposited by retinal astrocytes that migrate in front of the nascent vasculature [13, 33]. The more mature vessels near the optic nerve head begin to remodel after postnatal day (P) 3 and form an alternating radial pattern of arteries and veins that are connected by capillary beds. At around P8, the primary vascular plexus has reached the

Mouse Models of Physiological Angiogenesis

279

periphery of the retina. Subsequently, vessels start sprouting downwards to form two deeper plexi that will provide oxygen and nutrients to the internal retinal layers. In this process, angiogenic sprouts emerge from veins, venules, and capillaries near veins to penetrate the retina. Initially, the vascular sprouts grow perpendicular to the primary plexus along Mu¨ller cell processes, but turn laterally when they reach the inner or outer boundary of the inner nuclear layer. In this fashion, they establish the deeper vascular plexus and then the intermediate vascular plexus. By the third week of age, all three networks are fully established in mice [31, 32]. Similar to the embryonic hindbrain, an advantage of the developing mouse retina as a model system for angiogenesis is the planar nature of its primary vascular plexus, because it facilitates the visualization of angiogenic sprouting in flat-mounts. Also similar to the embryonic hindbrain, retinal angiogenesis proceeds in a stereotypical and well-defined sequence of events, as described above. In addition, remodeling in the center of the primary plexus allows the study of vascular maturation in the same preparation used to study sprouting at the vascular front. Thus, some vessel segments increase in diameter as they remodel into arteries and veins, while other vessel segments are pruned to reinforce optimal blood flow in vessels that are retained. Pruning is particularly evident in the vicinity of arteries, where capillary free zones emerge. This process can occur via migration and relocalization of endothelial cells into feeder vessels or via selective endothelial cell apoptosis [34]. Even though the retina provides a useful model system to study arteriovenous remodeling, this feature also reduces its suitability for quantitative analysis of capillary branching complexity, as areas containing only capillaries are small and interspersed between arteries and veins (Fig. 2b, c). The fact that the retinal vasculature develops postnatally can also pose a disadvantage when the targeting of genes implicated in vessel development leads to embryonic lethality (e.g., the Vegfa, Vegfr1, Vegfr2, and Nrp1 genes). However, the development of constitutive or tamoxifen-inducible conditional knockout technology can overcome this limitation, as for the hindbrain model [30]. As an advantage over the hindbrain model, the retina is more accessible to experimental manipulation, such as the local injection of drugs, inhibitors, or growth factors treatments [13]. Furthermore, the postnatal retina can be explanted and cultured for live imaging or ex vivo pharmacological treatments [35].

280

2 2.1

Alessandro Fantin and Christiana Ruhrberg

Materials Equipment

1. 96-well flat-bottom tissue culture plate. 2. Benchtop centrifuge. 3. Benchtop orbital shaker. 4. Benchtop tube roller. 5. Confocal laser scanning microscope. 6. Dissection instruments: standard curved forceps, Watchmaker forceps (no. 5 or 55), standard surgical scissors, and spring scissors. 7. Electrical tape. 8. Epifluorescence stereomicroscope equipped with a digital camera. 9. Falcon tubes, 50 mL. 10. Glass bottles. 11. Glass coverslips, 22  55 mm. 12. Glass slides. 13. Fiber-optic gooseneck lamps. 14. Parafilm. 15. Plastic cell culture dishes, 60 mm diameter. 16. Plastic Pasteur pipettes. 17. Round-bottomed reagent tubes, 2.0 mL. 18. Stereomicroscope equipped with a digital camera. 19. Tabletop balance. 20. Water bath.

2.2

Reagents

1. Absolute methanol. 2. Alexa Fluor 633-conjugated streptavidin (Life Technologies, cat. no. S32354). It is possible to choose an alternative fluorophore-conjugated streptavidin. 3. Biotinylated IB4 from Bandeiraea simplicifolia BS-I (Sigma, cat. no. L2140). 4. Blocking solution (0.1% Triton X-100 (v/v) and 10% normal goat serum (v/v) in PBS). 5. Mowiol/DABCO antifade (0.6% Mowiol 4-88 (w/v) in 25% glycerol with 2.5% DABCO (w/v)). 6. Paraformaldehyde (PFA); dissolve at 4% (w/v) in PBS to yield formaldehyde fixative. 7. PBS. 8. PBT (0.1% Triton X-100 (v/v) in PBS).

Mouse Models of Physiological Angiogenesis

281

9. Rising methanol gradient (25%, 50%, and 75% (vol/vol) absolute methanol in PBS, then absolute methanol). 10. SlowFade antifade reagent kit (Life Technologies, cat. no. S2828).

3

Methods

3.1 Hindbrain Dissection and Fixation

To obtain mouse embryos of defined gestational ages, mice are mated in the evening, and the morning of vaginal plug formation is counted as 0.5 dpc. The females are culled at the desired dpc (usually 11.5 dpc for sprouting phase analysis or 12.5 dpc for analysis of the mature SVP) by cervical dislocation and the uterine horns harvested and placed into a plastic dish with ice-cold PBS. 1. Under a dissecting stereomicroscope, use forceps to remove each individual embryo sac from the uterus and rupture the yolk sac. 2. Using the forceps, cut the embryo head just above the forearms (Fig. 1a, step 1) and remove the face (Fig. 1a, step 2) [20]. 3. Sever the skin and the pial membrane of the roof plate covering the fourth ventricle in the dorsal side of the head (Fig. 1a, step 3) and continue to cut toward the midbrain (Fig. 1a, step 4) and the spinal cord (Fig. 1a, step 5), allowing the not-yet-fused developing neural tube to unfold. 4. After removal of the head mesenchyme and the pial membrane from the ventral side of the neural tube (Fig. 1a, step 6), separate the midbrain (Fig. 1a, step 7) and beginning of the spinal cord from the unfurled hindbrain (Fig. 1a, step 8) [20]. 5. Using a plastic Pasteur pipette transfer the samples to a 2 mL plastic tube and fix in 1 mL of formaldehyde at 4  C for 2 h on a benchtop tube roller. Increasing the opening of the pipette by cutting off a portion of the tip will prevent any damage to the unfixed hindbrains. 6. After fixation, wash the hindbrains in PBS and either process them for whole-mount immunostaining or store them for a short term in PBS at 4  C. For storage longer than a few days, it is recommended to dehydrate the hindbrains through a rising methanol gradient to store them at 20  C for up to 3 months [20].

3.2 Retina Dissection and Fixation

Mouse pups of the desired age (usually between P4 and P7 for primary plexus analysis) are culled by cervical dislocation. 1. Immediately remove the eyes and fix for 5 min in formaldehyde at room temperature (RT) in a plastic dish. 2. The fixative is then replaced with PBS for further dissection.

282

Alessandro Fantin and Christiana Ruhrberg

3. Grab the cornea with the forceps (Fig. 2a, step 1) and cut along the border between the sclera and the iris using spring scissors (Fig. 2a, step 2). 4. After dissecting the cornea, remove the iris (Fig. 2a, step 3) and pull the lens out with the forceps; this action normally also removes any remaining fragments of iris, the ciliary body, the hyaloid vasculature, and the vitreous humor (Fig. 2a, step 4). 5. Ensure to completely remove the hyaloid vessels that might still be connected to the optic nerve head region before fixing the retina. 6. Carefully peel the sclera and retinal pigmented epithelium (rpe) away from the retina (Fig. 2a, step 5). 7. To flat-mount the dissected retina, make four radial incisions using spring scissors (Fig. 2a, step 6), position the retina with the inner side facing upward (Fig. 2a), and remove excess PBS from the dish using a plastic Pasteur pipette. If the retina is not completely opened up, flatten the unfolded lobes using curved forceps. 8. Fix the tissue by slowly dropping cold 100% methanol directly on the center of the retina for a few seconds until the tissue appears white and stiff. 9. Using curved forceps, transfer the retina to a 96-well plate already containing 200 μL of methanol. 10. Seal the plate with parafilm and incubate at 20  C for 2 h. Retinas can be stored at 20  C for many months, providing methanol in the wells is replenished regularly, as it evaporates over time even in the freezer. 3.3 Whole-Mount Staining

1. For whole-mount staining, the samples are washed in PBS twice at RT. 2. Hindbrains can be pooled into one 2 mL plastic tube (maximal four hindbrains per tube) and processed in a minimum volume of 300 μL in subsequent steps. 3. Retinas are instead transferred with curved forceps into the wells of a fresh 96-well plate previously filled with PBS and should be kept in a volume of 100 μL to cover the sample throughout the staining procedure. 4. During all subsequent incubations, tubes containing hindbrains are placed on a benchtop tube roller, while plates containing retinas are placed on an orbital shaker and agitated at the lowest speed available. 5. Incubate for 30 min to 1 h at RT in the blocking solution to permeabilize and block against unspecific interactions.

Mouse Models of Physiological Angiogenesis

283

6. Incubate in the blocking solution containing a 1:200 dilution of biotinylated IB4 overnight at 4  C (see Note 2). 7. Wash samples three times in PBT at RT for 15 min each. 8. Incubate overnight at 4  C or for 2 h at RT with Alexaconjugated streptavidin in blocking solution (see Notes 2–4). 9. Wash samples as described above and post-fix in formaldehyde at 4  C for at least 30 min. 3.4

Imaging

1. To image the fluorescently labeled hindbrains, place two layers of electrical tape on a glass slide, and cut squares to create pockets large enough to accommodate a single hindbrain [20]. 2. Using a Pasteur pipette, transfer the hindbrains to the pockets. 3. Remove excess PBS and mount in SlowFade (Molecular Probes) underneath a coverslip with the ventricular side facing up (see Notes 4 and 5). 4. On the contrary, place retinas directly on a slide with the inner side facing up using curved forceps. 5. Mount using Mowiol/DABCO antifade and a coverslip (see Notes 4 and 5). Fluorescent images for the embryonic hindbrains are best recorded in maximal intensity projections of z-stacks performed with a laser scanning confocal microscope (Figs. 1b–e and 3a, b). High magnification imaging is required to visualize small diameter structures such as tip cells and their filopodia (Fig. 1e). Turning the slide upside down will allow visualization of blood vessels entering the brain from the pial side (Fig. 1f). Postnatal retinas can be imaged using a fluorescent stereomicroscope at 1.6 magnification to capture the whole retina in a single picture (Figs. 2b and 3c, d), while confocal imaging is required to obtain more detailed images of the vasculature including tip cell filopodia (Fig. 2c, d).

3.5 Quantitation of Angiogenesis

For each hindbrain, the number of vessel intersections or radial vessels can be easily quantified manually in images of the ventricular or pial side, respectively (e.g., Fig. 1b, e). For the retina, low magnification images allow the quantification of radial vascular expansion from the center towards the periphery of the eye by measuring the distance between the optic nerve head in the center of the retina and the vascular front of each segment of the mounted tissue (red arrow in Fig. 2a). Several freely available or commercial software may be used for automated quantitation of these parameters, such as AngioTool (Fig. 1f) [36], ImageJ, Imaris (Bitplane), or Volocity (Improvision).

284

4

Alessandro Fantin and Christiana Ruhrberg

Notes 1. Figure 3 shows an example of angiogenesis defects in a hindbrain and retina with perturbed VEGF-A gradients relative to littermate wild-type controls. Here, the mutant phenotype is due to the expression of only the soluble VEGF120 isoform at the expense of a mixture of VEGF-A isoforms that include VEGF120 as well as the larger, heparin/NRP-binding VEGFA isoforms. Only the heterozygous mutant retina is shown because the homozygous mutation is lethal at birth in the C57/Bl6 background (see Subheading 1.1). 2. This protocol can be used for immunostaining the hindbrain or retina with many different antibodies, either alone or in combination, and is compatible with histochemical detection

Fig. 3 (a, b) IB4 staining of a Vegfa120/120 homozygous mutant (b) and littermate control (a) hindbrain at 11.5 dpc. (c, d) IB4 staining of a Vegfa+/120 heterozygous mutant (d; see Note 1) and littermate control (c) retina at P4. Scale bars: 500 μm (a, b), 1 mm (c, d)

Mouse Models of Physiological Angiogenesis

285

methods (e.g., HRP-conjugated streptavidin), provided that endogenous HRP activity is blocked at the beginning of the staining procedure, and the appropriate blocking reagents and secondary antibodies are selected (see for example ref. [10, 20, 37]). 3. If fluorescent speckles are seen, the reason could be that the streptavidin forms precipitates. To avoid this, always centrifuge the streptavidin in a refrigerated benchtop centrifuge at top speed before use to remove precipitates. 4. Weak staining might be due to insufficient tissue penetration by IB4 or photobleaching. Increase incubation time, optimize IB4 concentration, keep samples in the dark (e.g., wrap containers in tin foil), and mount samples in antifade solution. The choice of fluorochromes that are excited and emit at high wavelengths (red or far red) will allow improved detection of IB4 in deeper layers of whole-mounted tissues. 5. Trapping of air bubbles in mounted sections can occur when placing the coverslip. Lower the coverslip carefully onto the samples to avoid trapping air bubbles; if bubbles persist, slowly remove the coverslip and repeat the mounting procedure.

Acknowledgments We thank Marcus Fruttiger and Shalini Jadeja for teaching us the retina dissection technique. This work was supported by a Medical Research Council grant (MR/N011511/1) to C. Ruhrberg. References 1. Ruhrberg C, Gerhardt H, Golding M, Watson R, Ioannidou S, Fujisawa H, Betsholtz C, Shima DT (2002) Spatially restricted patterning cues provided by heparin-binding VEGF-A control blood vessel branching morphogenesis. Genes Dev 16(20): 2684–2698 2. Gerhardt H, Ruhrberg C, Abramsson A, Fujisawa H, Shima D, Betsholtz C (2004) Neuropilin-1 is required for endothelial tip cell guidance in the developing central nervous system. Dev Dyn 231(3):503–509 3. Fantin A, Vieira JM, Gestri G, Denti L, Schwarz Q, Prykhozhij S, Peri F, Wilson SW, Ruhrberg C (2010) Tissue macrophages act as cellular chaperones for vascular anastomosis downstream of VEGF-mediated endothelial tip cell induction. Blood 116(5):829–840. https://doi.org/10.1182/blood-200912-257832

4. Fantin A, Vieira JM, Plein A, Denti L, Fruttiger M, Pollard JW, Ruhrberg C (2013) NRP1 acts cell autonomously in endothelium to promote tip cell function during sprouting angiogenesis. Blood 121(12):2352–2362. https://doi.org/10.1182/blood-201205-424713 5. Graupera M, Guillermet-Guibert J, Foukas LC, Phng LK, Cain RJ, Salpekar A, Pearce W, Meek S, Millan J, Cutillas PR, Smith AJ, Ridley AJ, Ruhrberg C, Gerhardt H, Vanhaesebroeck B (2008) Angiogenesis selectively requires the p110alpha isoform of PI3K to control endothelial cell migration. Nature 453(7195): 6 6 2 – 6 6 6 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / nature06892 6. Lu X, Le Noble F, Yuan L, Jiang Q, De Lafarge B, Sugiyama D, Breant C, Claes F, De Smet F, Thomas JL, Autiero M, Carmeliet P, Tessier-Lavigne M, Eichmann A (2004) The

286

Alessandro Fantin and Christiana Ruhrberg

netrin receptor UNC5B mediates guidance events controlling morphogenesis of the vascular system. Nature 432(7014):179–186 7. Tammela T, Zarkada G, Nurmi H, Jakobsson L, Heinolainen K, Tvorogov D, Zheng W, Franco CA, Murtomaki A, Aranda E, Miura N, Yla-Herttuala S, Fruttiger M, Makinen T, Eichmann A, Pollard JW, Gerhardt H, Alitalo K (2011) VEGFR-3 controls tip to stalk conversion at vessel fusion sites by reinforcing Notch signalling. Nat Cell Biol 13(10):1202–1213. https://doi.org/10. 1038/ncb2331 8. Plein A, Fantin A, Denti L, Pollard JW, Ruhrberg C (2018) Erythro-myeloid progenitors contribute endothelial cells to blood vessels. Nature 562(7726):223–228. https://doi. org/10.1038/s41586-018-0552-x 9. Fantin A, Herzog B, Mahmoud M, Yamaji M, Plein A, Denti L, Ruhrberg C, Zachary I (2014) Neuropilin 1 (NRP1) hypomorphism combined with defective VEGF-A binding reveals novel roles for NRP1 in developmental and pathological angiogenesis. Development 141(3):556–562. https://doi.org/10.1242/ dev.103028 10. Fantin A, Schwarz Q, Davidson K, Normando EM, Denti L, Ruhrberg C (2011) The cytoplasmic domain of neuropilin 1 is dispensable for angiogenesis, but promotes the spatial separation of retinal arteries and veins. Development 138(19):4185–4191. https://doi.org/ 10.1242/dev.070037 11. Fantin A, Lampropoulou A, Gestri G, Raimondi C, Senatore V, Zachary I, Ruhrberg C (2015) NRP1 regulates CDC42 activation to promote filopodia formation in endothelial tip cells. Cell Rep 11(10):1577–1590. https:// doi.org/10.1016/j.celrep.2015.05.018 12. Benedito R, Roca C, Sorensen I, Adams S, Gossler A, Fruttiger M, Adams RH (2009) The notch ligands Dll4 and Jagged1 have opposing effects on angiogenesis. Cell 137(6): 1124–1135. https://doi.org/10.1016/j.cell. 2009.03.025 13. Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A, Abramsson A, Jeltsch M, Mitchell C, Alitalo K, Shima D, Betsholtz C (2003) VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 161(6):1163–1177. https:// doi.org/10.1083/jcb.200302047 14. Wang Y, Nakayama M, Pitulescu ME, Schmidt TS, Bochenek ML, Sakakibara A, Adams S, Davy A, Deutsch U, Luthi U, Barberis A, Benjamin LE, Makinen T, Nobes CD, Adams RH (2010) Ephrin-B2 controls VEGF-induced angiogenesis and lymphangiogenesis. Nature

465(7297):483–486. https://doi.org/10. 1038/nature09002 15. Hellstrom M, Phng LK, Hofmann JJ, Wallgard E, Coultas L, Lindblom P, Alva J, Nilsson AK, Karlsson L, Gaiano N, Yoon K, Rossant J, Iruela-Arispe ML, Kalen M, Gerhardt H, Betsholtz C (2007) Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445(7129): 7 7 6 – 7 8 0 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / nature05571 16. Liyanage SE, Fantin A, Villacampa P, Lange CA, Denti L, Cristante E, Smith AJ, Ali RR, Luhmann UF, Bainbridge JW, Ruhrberg C (2016) Myeloid-derived vascular endothelial growth factor and hypoxia-inducible factor are dispensable for ocular neovascularization--brief report. Arterioscler Thromb Vasc Biol 36(1): 1 9 – 2 4 . h t t p s : // d o i . o r g / 1 0 . 1 1 6 1 / ATVBAHA.115.306681 17. Raimondi C, Fantin A, Lampropoulou A, Denti L, Chikh A, Ruhrberg C (2014) Imatinib inhibits VEGF-independent angiogenesis by targeting neuropilin 1-dependent ABL1 activation in endothelial cells. J Exp Med 211(6):1167–1183. https://doi.org/10. 1084/jem.20132330 18. Bar T (1983) Patterns of vascularization in the developing cerebral cortex. Ciba Found Symp 100:20–36 19. Vieira JM, Schwarz Q, Ruhrberg C (2007) Selective requirements for NRP1 ligands during neurovascular patterning. Development 134(10):1833–1843. https://doi.org/10. 1242/dev.002402 20. Fantin A, Vieira JM, Plein A, Maden CH, Ruhrberg C (2013) The embryonic mouse hindbrain as a qualitative and quantitative model for studying the molecular and cellular mechanisms of angiogenesis. Nat Protoc 8(2): 418–429. https://doi.org/10.1038/nprot. 2013.015 21. Tata M, Ruhrberg C, Fantin A (2015) Vascularisation of the central nervous system. Mech Dev 138(pt 1):26–36. https://doi.org/10. 1016/j.mod.2015.07.001 22. Breier G, Albrecht U, Sterrer S, Risau W (1992) Expression of vascular endothelial growth factor during embryonic angiogenesis and endothelial cell differentiation. Development 114:521–532 23. Raab S, Beck H, Gaumann A, Yuce A, Gerber HP, Plate K, Hammes HP, Ferrara N, Breier G (2004) Impaired brain angiogenesis and neuronal apoptosis induced by conditional homozygous inactivation of vascular endothelial growth factor. Thromb Haemost 91(3):

Mouse Models of Physiological Angiogenesis 595–605. https://doi.org/10.1160/TH0309-0582 24. Haigh JJ, Morelli PI, Gerhardt H, Haigh K, Tsien J, Damert A, Miquerol L, Muhlner U, Klein R, Ferrara N, Wagner EF, Betsholtz C, Nagy A (2003) Cortical and retinal defects caused by dosage-dependent reductions in VEGF-A paracrine signaling. Dev Biol 262(2):225–241. pii:S0012160603003567 25. Farrell CL, Risau W (1994) Normal and abnormal development of the blood-brain barrier. Microsc Res Tech 27(6):495–506 26. Carmeliet P, Ng YS, Nuyens D, Theilmeier G, Brusselmans K, Cornelissen I, Ehler E, Kakkar VV, Stalmans I, Mattot V, Perriard JC, Dewerchin M, Flameng W, Nagy A, Lupu F, Moons L, Collen D, D’Amore PA, Shima DT (1999) Impaired myocardial angiogenesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat Med 5(5): 495–502 27. Schwarz Q, Gu C, Fujisawa H, Sabelko K, Gertsenstein M, Nagy A, Taniguchi M, Kolodkin AL, Ginty DD, Shima DT, Ruhrberg C (2004) Vascular endothelial growth factor controls neuronal migration and cooperates with Sema3A to pattern distinct compartments of the facial nerve. Genes Dev 18(22): 2822–2834. https://doi.org/10.1101/gad. 322904 28. Kawasaki T, Kitsukawa T, Bekku Y, Matsuda Y, Sanbo M, Yagi T, Fujisawa H (1999) A requirement for neuropilin-1 in embryonic vessel formation. Development 126(21):4895–4902 29. Jones EA, Yuan L, Breant C, Watts RJ, Eichmann A (2008) Separating genetic and hemodynamic defects in neuropilin 1 knockout embryos. Development 135(14):2479–2488. https://doi.org/10.1242/dev.014902

287

30. Nagy A (2000) Cre recombinase: the universal reagent for genome tailoring. Genesis 26(2): 99–109 31. Fruttiger M (2007) Development of the retinal vasculature. Angiogenesis 10(2):77–88. https://doi.org/10.1007/s10456-0079065-1 32. Hofmann JJ, Luisa Iruela-Arispe M (2007) Notch expression patterns in the retina: an eye on receptor-ligand distribution during angiogenesis. Gene Expr Patterns 7(4):461–470. https://doi.org/10.1016/j.modgep.2006. 11.002 33. Carmeliet P, Tessier-Lavigne M (2005) Common mechanisms of nerve and blood vessel wiring. Nature 436(7048):193–200 34. Hughes S, Chang-Ling T (2000) Roles of endothelial cell migration and apoptosis in vascular remodeling during development of the central nervous system. Microcirculation 7(5): 317–333 35. Sawamiphak S, Ritter M, Acker-Palmer A (2010) Preparation of retinal explant cultures to study ex vivo tip endothelial cell responses. Nat Protoc 5(10):1659–1665. https://doi. org/10.1038/nprot.2010.130 36. Zudaire E, Gambardella L, Kurcz C, Vermeren S (2011) A computational tool for quantitative analysis of vascular networks. PloS One 6(11): e27385. https://doi.org/10.1371/journal. pone.0027385 37. Pitulescu ME, Schmidt I, Benedito R, Adams RH (2010) Inducible gene targeting in the neonatal vasculature and analysis of retinal angiogenesis in mice. Nat Protoc 5(9): 1518–1534. https://doi.org/10.1038/nprot. 2010.113

Chapter 21 Evaluating VEGF-Induced Vascular Leakage Using the Miles Assay James T. Brash, Christiana Ruhrberg, and Alessandro Fantin Abstract Before the endothelial mitogenic activity of the Vascular Endothelial Growth Factor A (VEGF) was described, VEGF had already been identified for its ability to induce vascular leakage. VEGF-induced vascular leakage has been most frequently studied in vivo using the Miles assay, a simple yet invaluable technique that has allowed researchers to unravel the molecular mechanisms underpinning vascular leakage both for VEGF and other permeability inducing agents. In this protocol, a mouse is intravenously injected with Evans Blue dye before VEGF is administered locally via intradermal injection. VEGF promotes vascular leak of serum proteins in the dermis, enabling Evans Blue-labeled albumin extravasation from the circulation and subsequent accumulation in the skin. As the volume of dye extravasation is proportional to the degree of vascular leak, it can be quantified as a proxy measurement of VEGF-induced vascular leakage. Key words VEGF, Miles assay, Vascular permeability, Vascular leakage, Evans blue, Edema

1

Introduction All blood vessels are lined with an endothelial monolayer that regulates fluid and solute exchange between the circulation and vascularized tissues. Several proteins, such as histamine or the vascular endothelial growth factor A (VEGF), disrupt the endothelial barrier, thereby increasing vascular leakage and causing tissue edema [1]. First applied in the 1950s to guinea pigs [2], the Miles assay has emerged as the principal technique for studying vascular leakage in vivo [3]. Now invariably performed on mice, especially to exploit their potential for genetic manipulation, the Miles assay quantifies vascular leakage through the assessment of Evans Bluelabeled albumin extravasation from the dermal vasculature. In the initial steps of the procedure, Evans Blue dye is introduced into the mouse circulation by intravenous injection. Subsequently, VEGF or another permeability inducing agent is administered to the skin of

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_21, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

289

290

James T. Brash et al.

the mouse via intradermal injection, thereby promoting endothelial barrier disruption and leakage of Evans Blue dye into the skin. In principle, the quantity of Evans Blue dye that accumulates in the skin is proportional to the degree of vascular leakage. Thus, after the mouse has been culled, Evans Blue dye can be extracted from the skin and its abundance measured. Comparing the abundance of Evans Blue dye extracted from VEGF-injected skin versus vehicleinjected skin in the same animal provides a normalized readout of VEGF-induced vascular leakage, which can then be compared between experimental groups to determine how a variable influences vascular leakage. As the Miles assay requires relatively few reagents and consists of only a small number of steps, it is often favored for its simplicity compared to other techniques that measure vascular leakage. Moreover, the Miles assay can be used to compare vascular leakage between different strains of genetically modified mice or coupled with pharmacological treatments to identify novel inhibitors or molecular mediators of vascular leakage [4–6].

2

Materials All solutions should be prepared in a laminar flow cabinet and sterilized by passing through a 22 μm filter using a syringe.

2.1 Mouse Preparation

Electric trimmer. Isoflurane. Weighing scale for rodents.

2.2

Injections

Pyrilamine maleate: 4 μg/μL in 0.9 % saline. Evans Blue dye: 1% Weight/Volume Percentage Concentration (w/v) in 0.9% saline. VEGF solution: 2.5 ng/μL recombinant mouse VEGF-A164 in phosphate buffered saline (PBS). 1 mL and 350 μL syringes. 30G needles. Mouse restrainer.

2.3 Dye Extraction/ Quantification

Deionized formamide. Cork/polystyrene board. Scalpels. Spectrophotometer.

VEGF-Induced In Vivo Vascular Leakage

3

291

Methods

3.1 Mouse Preparation

See Note 1 1. Mice must be shaved 24–48 h before step 1 in Subheading 3.2. Anesthetize the first mouse using isoflurane (3% isoflurane for anesthetic induction, 1.5–2% isoflurane for anesthetic maintenance). Do not move onto the next step until the mouse is unresponsive to external stimuli, such as a gentle tail pinch. 2. Using an electric trimmer, shave both flanks of the anesthetized mouse, taking special care not to cause any skin damage that will disturb subsequent analysis. 3. Weigh the anesthetized mouse and return it to its cage for it to regain consciousness. 4. Repeat steps 1–3 for each mouse to be studied. When using male mice, return each to an individual cage to prevent fighting after recovery and therefore skin damage and irritation; always use a cage that contains bedding from the home cage.

3.2 Injection of Evans Blue Dye

1. For each mouse to be studied, prepare two 1 mL syringes with 30G needles in a sterile laminar flow cabinet. Load one syringe with 10 μL sterile pyrilamine maleate per gram of mouse body weight (as recorded in Subheading 3.1, step 3); then load the second syringe with 100 μL sterile 1% Evans Blue dye (w/v sterile PBS). 2. Administer pyrilamine maleate to the first mouse by intraperitoneal injection (Fig. 1a). To do so, scruff the mouse and tilt it downwards to expose the lower abdomen. 3. Insert the needle into the lower quadrant of the mouse, avoiding the midline, where the bladder resides, and then inject the pyrilamine maleate solution. 4. Move the mouse to a 37  C heat chamber for 10 min to promote vasodilation, and repeat steps 2–4 for the remaining mice. 5. Move the first mouse into a restrainer. Position the restrainer containing the mouse under a suitable light source and rub the tail with 70% ethanol to enable tail vein visualization. 6. Once the tail vein is readily observable, inject 100 μL Evans blue dye into the tail vein (Fig. 1b, see Note 2). 7. After removing the needle, immediately apply pressure to the injection site to prevent blood loss. After 10 s, return the mouse to its cage. 8. Repeat steps 5–7 for the remaining mice. 9. Allow Evans Blue to circulate for 30–60 min.

292

James T. Brash et al.

Fig. 1 Main steps of the Miles assay for vascular leakage. (a) Intraperitoneal injection of pyrilamine maleate: scruff the mouse and tilt it downwards; insert the needle into the lower quadrant of the mouse abdomen, away from the midline. (b) Intravenous injection of Evans Blue dye: align the needle with the tail vein and insert; gently administer 100 μL Evans Blue dye. If the dye does not freely enter the circulation, cease the injection and attempt to inject at a second site further up the tail. (c) Intradermal injection of VEGF or vehicle solutions: pull the flank skin and insert the needle at a 15 angle; gently administer 20 μL of VEGF solution and withdraw the needle, observing a raised bump in the skin. Repeat the injection at two additional sites before turning the mouse and administering three intradermal injections of vehicle solution on the opposing flank. (d) After cervical dislocation, dissect the mouse skin to reveal Evans Blue accumulation at injection sites and proceed with skin sampling for Evans blue dye extraction and measurement

3.3 Induction of Vascular Leakage

1. For each mouse, prepare two 300 μL syringes with a 30G needle in a sterile laminar flow hood. 2. Load one syringe with 100 μL sterile PBS and the second syringe with 100 μL sterile VEGF in PBS (2.5 ng/μL). 3. Anesthetize the first mouse (3% isoflurane for anesthetic induction, 1.5–2% for anesthetic maintenance). Do not move to step 4 until the mouse is unresponsive to external stimuli such as a gentle tail pinch. 4. Administer 20 μL of VEGF solution into the dermis of the first flank of the mouse by intradermal injection (Fig. 1c, see Note 3). 5. Repeat the injection at a further two adjacent dermal sites, approximately 1 cm apart. 6. Record the location of each injection site on a piece of paper. 7. Turn the mouse to reveal the second flank. 8. Administer 20 μL PBS by intradermal injection. 9. Repeat the injection at a further two sites, approximately 1 cm apart. 10. Record the locations of each injection site on a piece of paper. 11. Return the mouse to its cage and monitor it as it recovers from anesthesia. 12. Repeat steps 2–11 for each mouse.

VEGF-Induced In Vivo Vascular Leakage

3.4 Quantification of Vascular Leakage

293

1. Twenty minutes after the intradermal injections, cull each mouse by cervical dislocation. 2. Place each dead mouse onto a cork/polystyrene board and pin their feet down. 3. Make an incision between the mouse abdomen and chest using blunt scissors, and then use forceps and a scalpel to peel the flank skin away from the mouse (Fig. 1d). 4. Pin down the flank skin and use a scalpel to scrape away any fat connected to the skin. 5. Using forceps and a scalpel, excise the skin around the injection sites, where Evans Blue dye has accumulated (see Note 4). Records of injection sites will aid in this step, especially for the PBS control. 6. Place each skin sample into a labeled 1.5 mL reagent tube and dry in the open tube overnight at 55  C. 7. In a fume cupboard, add 250 μL deionized formamide to each tube and incubate with a closed lid overnight at 55  C to extract the Evans Blue dye from the skin. 8. Centrifuge samples at 10,000  g for 40 min. 9. In a fume cupboard, transfer 100 μL of supernatant (Evans Blue dye-containing formamide) from each tube to an individual well of a transparent 96 flat-well plate. 10. Measure the absorbance of each Evans Blue dye-containing formamide solution with a spectrophotometer at 620 nm using a reference wavelength of 740 nm. 11. For each mouse, average the absorbance readings for the triplicate injections (i.e., 3  VEGF injections into the same flank, 3  PBS injections into the same flank) and calculate the fold difference in absorbance for the VEGF-injected flank versus the PBS-injected flank of the same mouse.

4

Notes 1. Mice should be at least 8 weeks of age. Where possible, littermate mice should be used across experimental groups. Alternatively, mice in different experimental groups should be genetic background strain-, age-, weight-, and gender-matched. Do not use more than 6 mice per experimental session to allow sufficient time for each step. To identify a significant difference in moderate to severe leakage across an experiment, 5 mice per experimental group are typically required. 2. Performing the tail vein injection is the most difficult step of this procedure, as it needs to be performed with similar

294

James T. Brash et al.

efficiency in all the animals to ensure that results between different animals can be reliably compared. Researchers should first clearly identify the tail vein they wish to inject. Visualizing the length of the tail vein is aided by adequate light and rubbing 70% ethanol onto the tail. Once the vein is clearly visible, researchers should angle the mouse so that the vein is facing upwards, and then extend the tail so that it runs parallel to the mouse body. Close to the tip of the tail, researchers should align the needle with the vein then slowly enter the vein towards the tail base and with the needle bevel facing upwards. If the needle has successfully entered the vein, the Evans Blue solution will inject with little resistance. If the researcher feels resistance when attempting to inject, or if Evans Blue dye accumulates at the injection site instead of entering the circulation, the researcher should withdraw the needle and attempt the injection again at a higher location on the tail (i.e., closer to the tail base). 3. To perform intradermal injections, the researcher should pull back the skin of the mouse so that the skin is tight at the injection site; it is important to not pinch skin near the injection site because this may promote unwanted leakage. Intradermal injections should be performed at approximately 15 to the skin. There should be a degree of resistance as the needle enters the dermis; if the needle slides quickly into the skin, the researcher may have passed the dermal layer and would be delivering the solution subcutaneously. A raised bump in the skin indicates a successful intradermal injection, while a flat injection site likely indicates that the injection load was delivered beneath the skin. 4. Cut similarly sized regions of skin from each injection site, irrespective of the apparent Evans Blue dye accumulation at each site. This helps to normalize for Evans Blue dye that has not leaked but still resides within the skin vessels. To facilitate equal sampling, a punch biopsy blade may be used.

Acknowledgments This work was supported by a Medical Research Council grant (MR/N011511/1) to C. Ruhrberg and a British Heart Foundation PhD studentship to J.T. Brash (FS/13/59/30649).

VEGF-Induced In Vivo Vascular Leakage

295

References 1. Plein A, Fantin A, Ruhrberg C (2014) Neuropilin regulation of angiogenesis, arteriogenesis, and vascular permeability. Microcirculation 21(4):315–323. https://doi.org/10.1111/ micc.12124 2. Miles AA, Miles EM (1952) Vascular reactions to histamine, histamine-liberator and leukotaxine in the skin of guinea-pigs. J Physiol 118(2): 228–257. https://doi.org/10.1113/jphysiol. 1952.sp004789 3. Brash JT, Ruhrberg C, Fantin A (2018) Evaluating vascular hyperpermeability-inducing agents in the skin with the miles assay. J Vis Exp 136: 57524. https://doi.org/10.3791/57524 4. Fantin A, Lampropoulou A, Senatore V, Brash JT, Prahst C, Lange CA, Liyanage SE, Raimondi C, Bainbridge JW, Augustin HG, Ruhrberg C (2017) VEGF165-induced vascular

permeability requires NRP1 for ABL-mediated SRC family kinase activation. J Exp Med 214(4): 1049–1064. https://doi.org/10.1084/jem. 20160311 5. Eliceiri BP, Paul R, Schwartzberg PL, Hood JD, Leng J, Cheresh DA (1999) Selective requirement for Src kinases during VEGF-induced angiogenesis and vascular permeability. Mol Cell 4(6):915–924. https://doi.org/10.1016/ s1097-2765(00)80221-x 6. Aman J, van Bezu J, Damanafshan A, Huveneers S, Eringa EC, Vogel SM, Groeneveld AB, Vonk Noordegraaf A, van Hinsbergh VW, van Nieuw Amerongen GP (2012) Effective treatment of edema and endothelial barrier dysfunction with imatinib. Circulation 126(23): 2728–2738. https://doi.org/10.1161/ CIRCULATIONAHA.112.134304

Chapter 22 Modulation of VEGFA Signaling During Heart Regeneration in Zebrafish Kaushik Chowdhury, Shih-Lei Lai, and Rube´n Marı´n-Juez Abstract Over the last decades, myocardial infarction and heart failure have accounted every year for millions of deaths worldwide. After a coronary occlusion, the lack of blood supply to downstream muscle leads to cell death and scarring. To date, several pro-angiogenic factors have been tested to stimulate reperfusion of the affected myocardium, VEGFA being one of the most extensively studied. Given the unsuccessful outcomes of clinical trials, understanding how cardiac revascularization takes place in models with endogenous regenerative capacity holds the key to devising more efficient therapies. Here, we summarize the main findings on VEGFA’s role during cardiac repair and regeneration, with a particular focus on zebrafish as a regenerative model. Moreover, we provide a comprehensive overview of available tools to modulate Vegfa expression and action in zebrafish regeneration studies. Understanding the role of Vegfa during zebrafish heart regeneration may help devise efficient therapies and circumvent current limitations in using VEGFA for therapeutic angiogenesis approaches. Key words VEGFA, Cardiac regeneration, Coronary occlusion, Zebrafish, Heart failure, Revascularization therapy

1

Introduction Cardiovascular diseases, particularly myocardial infarction (MI), remain the leading cause of morbidity and mortality worldwide [1]. MI occurs when a coronary artery is occluded, which prevents the supply of blood and nutrients to a portion of the myocardial wall, causing tissue damage. Adult mammalian cardiomyocytes (CMs) cannot replenish the damaged tissue; instead, fibrotic repair takes over, leading to scar deposition, pathophysiological remodeling, and, eventually, heart failure [2, 3]. Recent studies have shown that adult human CMs retain a limited capacity to proliferate, indicating the possibility of promoting cardiac regeneration in MI patients via the stimulation of endogenous CM proliferation [4– 7]. With this motivation, a considerable effort has been directed toward identifying and designing new therapeutic strategies to

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_22, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

297

298

Kaushik Chowdhury et al.

stimulate cardiac regeneration in humans. As part of these efforts, animal models with endogenous regenerative capacity have proven to be valuable tools in identifying mechanisms regulating cardiac repair [8, 9]. It is becoming increasingly clear that achieving successful cardiac regeneration requires the coordinated responses of different tissues and cell types. Studies in humans and animal models indicate that efficient restoration of the coronary (cardiac) vascular network is key to efficient regeneration [10–14]. Among the pro-angiogenic factors that play a role in cardiac regeneration, vascular endothelial growth factor (VEGF) has been extensively studied due to its ability to stimulate angiogenesis [15]. Motivated by results in animal models, several clinical trials tested the effects of VEGF as a treatment for MI patients with no success [14, 16, 17]. The timing and dosage of VEGF administration have proven to be critical factors to take into consideration [14, 18, 19]. Indeed, recent studies have shown that after cardiac injury, vegfaa is rapidly upregulated in the zebrafish heart and that both loss- and gain-offunction approaches impair its regenerative capacity [12, 20, 21]. Investigating an animal model with the intrinsic capacity for heart regeneration may help in the design and refinement of the therapeutic approaches. Here, we focus on the zebrafish as a model for heart regeneration with a special focus on available tools to modulate Vegfa signaling in adult animals.

2

Cardiac Regeneration in Zebrafish Zebrafish possess a remarkable capacity to regenerate their adult hearts upon different types of insults [22–26]. Initial wound response triggers inflammation and the rapid recruitment of immune cells to the injured area [27, 28]. Chemical ablation of macrophages prior to injury delays leukocyte recruitment, compromises revascularization, and impairs CM proliferation leading to increased scarring [28]. Moreover, it was recently shown that macrophages contribute to collagen deposition in both mouse and zebrafish hearts after cardiac damage [29]. The cardiac endothelium (endocardium and coronary vasculature) also responds early after cardiac injury. Border zone endocardium upregulates several different factors upon injury [21, 30– 34]. Early studies by Kikuchi et al. identified that endocardial retinoic acid synthesizing enzyme aldh1a2 is rapidly upregulated after injury [30]. Similarly, the injured endocardium upregulates TgBAC(vegfaa:EGFP) a few hours after injury, suggesting a role in vascular regeneration [20]. Other studies showed that components of Notch and Wnt signaling pathways are also expressed by regenerating endocardial cells [32–34]. Interestingly, most of these molecules are well-known regulators of vasculogenesis and

VEGFA in Zebrafish Heart Regeneration

299

angiogenesis, further supporting the notion that regenerating endocardial cells regulate coronary regeneration. Using the ventricular resection model, early studies observed the first coronary endothelial cells (cECs) at 7–10 days postamputation (dpa) [35, 36]. More recently, using the cryoinjury model, it was shown that cECs start to proliferate within hours after cardiac injury and cover the whole injured area within 7 days post cryoinjury (dpci) (Fig. 1a, b) [12]. Using this injury model allows detailed analysis of coronary regeneration and identified Apelin, Cxcl12/Cxcr4, Vegf, and Hif pathways as regulators of cardiac revascularization [21]. Recent reports have highlighted the importance of cardiac lymphatics in the regulation of cardiac regeneration in zebrafish. While the coronary network in zebrafish develops around 6–8 weeks post-fertilization [37], cardiac lymphatics do not invade the ventricle until 12–14 weeks of age [38, 39]. This delay between coronary and lymphatic development seems to be recapitulated during regeneration [38–40]. Mechanistically, alterations in Cxcl12/Cxcr4 and Vegfc signaling have been shown to impair cardiac lymphatic development and regeneration in zebrafish [38, 39]. The epicardium is at the interphase between CMs, endothelial cells (ECs), and fibroblasts, and can express a broad range of factors during regeneration, making it the focus of numerous studies [41]. Following an injury, epicardium-derived cells (EPDCs) proliferate and cover the damaged region [25, 35]. Lineage tracing studies revealed that EPDCs differentiate into fibroblasts and perivascular cells after cardiac injury in zebrafish [42]. The role of the epicardium during heart regeneration has been reviewed elsewhere [41, 43]. EPDCs are not the only contributors to fibroblasts during regeneration. The zebrafish heart has an endogenous population of fibroblast [44, 45]. Recent studies show that endogenous fibroblasts participate in the regenerative response by producing extracellular matrix and regulating CM proliferation [45]. CMs are the building blocks of the heart muscle and have been the main focus of the cardiac regeneration field. Landmark studies by Kikuchi et al. and Jopling et al. showed that pre-existing CMs are the primary source of new CMs during heart regeneration [46, 47]. Since then, the field has endeavored to identify cellular and molecular mechanisms to stimulate innate CM regeneration. Several studies have identified mitogens with the ability to stimulate CM proliferation. These topics have been reviewed elsewhere [48, 49]. Although less explored in zebrafish, nerves have also been shown to play a role in heart regeneration. Mahmoud et al. showed that reducing cardiac innervation by the overexpression of Sema3aa diminished CM proliferation and led to scarring [50]. Overall, successful regeneration is a complex process that requires a tightly coordinated multi-tissue response. Different cell populations can regulate cardiac regeneration by providing either

300

Kaushik Chowdhury et al.

Fig. 1 Coronary regeneration in zebrafish. (a) Sagittal section of a 7 dpci TgBAC(etsrp:EGFP) adult heart stained for EGFP expression (green, endothelial cells), Mef2 (red, CM nuclei), PCNA (white, proliferation marker), and DNA (blue, cell nuclei). High magnification images show an intra-ventricular sprouting coronary (a0 , yellow arrowhead) and proliferating CMs in the injured area in close proximity to regenerated coronaries (a00 , cyan arrowheads). (b) Wholemount image of a Tg(flt1:Mmu.Fos-EGFP) ventricle at 7 dpci displaying a new coronary network covering the injured area. Transgene expression is specifically detected in coronary vessels. Yellow dotted lines delineate injured areas. Ba bulbus arteriosus, V ventricle. Scale bars: 100 μm

signaling cues or structural support. Despite the fact that different cell types and molecules have been shown to regulate regeneration, the crosstalk among them might be more important than their individual actions. Future research will help to define how these responses are collectively regulated to support heart regeneration.

3

Mechanisms of Revascularization in Cardiac Repair and Regeneration The increased ability to grow coronary arteries from neighboring vessels (termed collaterals) in patients correlates with a decreased extent of injury upon MI [11, 51, 52]. Current invasive therapies such as stenting and coronary artery bypass grafting improve cardiac repair by reperfusion of the injured area. However, many patients are not eligible for or responsive to these invasive therapies and require alternative approaches, including therapeutic angiogenesis [53]. Therapeutic angiogenesis seeks to stimulate collateral formation with different treatment modalities, including the administration of growth factors, miRNAs, DNA, and cells [17]. Despite promising results of therapeutic angiogenesis in animal models, clinical trials have shown no or minimal effect in humans [14, 17, 54]. Heart regeneration requires effective neovascularization, which can be achieved by either angiogenesis or vasculogenesis. Neovascularization implicates the reactivation of mechanisms involved in

VEGFA in Zebrafish Heart Regeneration

301

the early stages of coronary development [55]. Among the different cell types present in the heart, ECs (endocardium and coronary vasculature) are the most abundant cell type in mice, humans [56], and zebrafish [57]. The endocardium is a source of cECs during development in mice and zebrafish [37, 58, 59] and recent studies suggest that this phenomenon is also recapitulated in adult mice after MI [60, 61]. Similarly, studies in medaka, a fish species unable to regenerate its heart, showed that they develop vessel-like structures of endocardial origin that support cardiac repair following injury [28]. Therefore, the endocardium may be a source of cECs after cardiac damage in non-regenerative models, making it an attractive therapeutic target. In the adult rodent and zebrafish heart, cardiac injury induces angiogenic expansion of pre-existing coronaries, which is believed to be the primary mechanism of neovascularization [12, 21, 34, 62] and the target of therapies for acute MI [63, 64]. Moreover, recent studies in neonatal mice suggest an alternative mode of neovascularization where single arterial cECs mobilize and reassemble to form new collaterals after cardiac injury [10]. Overall, due to the relevance of angiogenic revascularization to cardiac regeneration and its potential therapeutic applications, several studies aimed to identify mechanisms to stimulate coronary regeneration. Early studies in zebrafish implicated different signaling pathways in the regulation of revascularization during heart regeneration. Lepilina and colleagues reported that Fibroblast growth factor (Fgf) signaling components are upregulated following resection injury [35]. Accordingly, overexpression of a dominant-negative form of fgfr1 blocked the regenerative capacity of the zebrafish heart, impairing CM, epicardial and coronary regeneration [35]. Furthermore, Platelet-derived growth factor (Pdgf) signaling has also been shown to be essential for heart regeneration in zebrafish. After ventricular resection, pdgfrβ expression is upregulated in the injured area, and chemical inhibition of Pdgfr led to impaired epicardial proliferation and coronary formation into the injury [36]. Based on these and other studies, coronary network replenishment was proposed as a late event taking place at the same time as CM regeneration (7–10 dpa). More recent studies using the cryoinjury model, where the damaged region is retained, enabled a more detailed characterization of coronary regeneration. Zebrafish hearts show a remarkable capacity to rapidly revascularize following cardiac injury, displaying coronary sprouting and proliferation as early as 15 h post cryoinjury (hpci) [12]. In the same study, it was shown that inducible overexpression of a dominant-negative form of vegfaa dampened coronary revascularization. Notably, blocking early revascularization was sufficient to reduce CM proliferation and scar clearance [12]. Conversely, promoting and stabilizing cardiac vasculature by poly I:C in the non-regenerative medaka heart increased CM

302

Kaushik Chowdhury et al.

replenishment and reduced scarring [28]. These studies also indicated that macrophages play an essential role in cardiac revascularization in fish, similar to neonatal mice [65]. Recently, Marı´n-Juez et al. showed that coronary revascularization takes place in zebrafish by a process termed coronary-endocardial anchoring [21]. In this process, coronary vessels regenerate superficially and intraventricularly toward the activated endocardium, forming a vascular scaffold available for regenerating CMs to replenish the injured area (Fig. 1a0 , a00 ). Different signaling pathways regulate these two revascularization modalities. Superficial regenerating coronaries are guided by epicardial Cxcl12/Cxcr4 signaling in response to hypoxia, while intra-ventricular coronaries are controlled by endocardial Vegfa signaling. Interestingly, in the same report, it was shown that superficial revascularization is metabolically regulated via Apelin and Pgc1α [21]. Overall, these studies highlight the importance of the coronary vasculature during regeneration and point toward a role beyond its classical function as a transport network for blood and nutrients.

4

VEGFA and Heart Regeneration Multiple growth factors are involved in regulating vascular development. Different Vascular Endothelial Growth Factor (VEGF) family members are well-known regulators of various aspects of vascular development and pathology [66]. Among them, VEGFA is a master endothelial regulator known to be instrumental in the coordination of angiogenesis, arteriogenesis, vasculogenesis, and endothelial proliferation and differentiation [14, 66–68]. Vegfa mutant mice die before birth due to severe vascular defects [69]. Knockout of different VEGFA isoforms (VEGF164 and VEGF188) in mice leads to impaired postnatal myocardial angiogenesis and ischemic cardiomyopathy [70]. Conversely, increased administration of VEGFA induces the formation of new, but unstable vessels in the adult mouse heart [71, 72]. Due to its potency and ability to regulate different vascular processes, VEGFA has received extensive attention as a potential candidate for developing new therapeutic angiogenesis approaches. However, all these attempts have failed to translate to treatments for MI patients. The timing and dosage of VEGFA have been proposed and experimentally tested to be critical factors for devising such approaches [14, 16, 18, 66]. Recent data in zebrafish showed that vegfaa expression is upregulated early after cardiac injury [12]. Moreover, chemical and genetic blockade of Vegfa signaling efficiently halt vascular growth and tissue regeneration in both zebrafish fin and cardiac regeneration models [12, 73]. Recent studies using gain-of-function approaches showed that CM-specific overexpression of vegfaa in zebrafish stimulates

VEGFA in Zebrafish Heart Regeneration

303

coronary development and impairs regeneration [20]. These findings further support the notion that the timing, dosage, and duration of Vegfa administration are critical to successful heart regeneration. Using a TgBAC(vegfaa:EGFP) line, recent reports have shown that EPDCs upregulate this transgene during development, homeostasis, and regeneration in zebrafish [20]. vegfaa mutants display disorganized and thinner coronary vessels [12], similar to mouse mutants [70], suggesting that EPDCs could be regulating coronary network patterning and maturation via Vegfa in zebrafish. Interestingly, TgBAC(vegfaa:EGFP) expression is also detected in the regenerating endocardium [20, 21]. The combined action of endocardial Vegfa and epicardial Cxcl12/Cxcr4 signaling has been proposed to regulate coronary network regeneration in zebrafish [21] and mice [10]. VEGFA can bind and act via different receptors, including VEGFR1, VEGFR2, NRP1, and NRP2. Recent studies in zebrafish showed that nrp1a mutants have impaired coronary regeneration and epicardial activation [74]. These data suggest that Vegfa might regulate cardiac revascularization, at least in part, via Nrp1a in zebrafish. Further studies will be required to address the specific role of Vegfa receptors during cardiac regeneration. Overall, using the zebrafish model might identify critical factors that regulate coronary revascularization during heart regeneration. Besides the issues with the timing and dosage of administration mentioned above, one of the main drawbacks of pro-angiogenic therapies is their reliance on a single factor. Using an adult vertebrate model with endogenous regenerative capacity will enable us to define the temporal and spatial dynamics of different angiogenic factors and interactions between different signaling pathways.

5 Genetic Tools for Modulating VEGFA Signaling During Zebrafish Heart Regeneration Here, we highlight the genetic reagents currently available to modulate Vegfa signaling during heart regeneration in adult zebrafish. These lines are summarized in Table 1. 5.1

vegfaa2/2

vegfa is duplicated in zebrafish as vegfaa and vegfab. vegfaa encodes Vegfaa-121 and -165, whereas vegfab encodes Vegfab-171 and -210 [78]. Mutations in vegfaa lead to severe vascular defects during development, including the defective differentiation of arteries and the sprouting of intersegmental vessels [76, 79]. Conversely, vegfab mutants display mild vascular phenotypes in vertebral arteries and are viable until adulthood [80]. Although the loss of Vegfaa is lethal, vegfaa mutants can be rescued and raised to adulthood after injecting vegfaa mRNA at embryonic stages [12, 76]. Adult vegfaa mutants display alterations in coronary

304

Kaushik Chowdhury et al.

Table 1 Mutant and transgenic lines to modulate Vegfa signaling during heart regeneration in adult zebrafish Line name

Description

Advantages

Limitations

TALEN mutant Vegfaa LoF model

Global KO of vegfaa

Lack of tissue specificity [75]

vegfabbns92

CRISPR mutant Vegfab LoF model

Global KO of vegfab

Lack of tissue specificity [76]

Tg(hsp70l:vegfaa121F17A)bns100

Transgenic line Inducible LoF HS inducible expression of dnVegfaa-121

Lack of tissue specificity [12] Early induction after cryoinjury

Tg(hsp70l:loxp-Stop-loxpmTom-codOptP2AT46Avegfaa)bns288

Transgenic line Cre-Lox system Conditional HS inducible expression of dnVegfaa-121

Inducible LoF Reliance on strong and [21] efficient Allows Cre/CreERT2 lines temporal and spatial control

flt1bns29

CRISPR mutant mFlt1 and sFlt1 LoF model

Global KO of both sflt1 and mflt1

Lack of tissue specificity [77]

mflt1 fh390

ENU mutant mFlt1 LoF model

Global KO of mflt1

Lack of tissue specificity [76]

nrp1asa1485

ENU mutant Nrp1a LoF model

Global KO of nrp1a

Lack of tissue specificity [74]

Tg(βactin2:loxPmTagBFP-STOP-loxPvegfaa) pd262

Transgenic line Cre-Lox system Conditional HS inducible expression of Vegfaa

Inducible GoF Reliance on strong and [20] efficient Allows Cre/CreERT2 lines temporal and spatial control

vegfaa

bns1

Refs.

GoF gain-of-function, LoF loss-of-function, HS heatshock

network formation that present as abnormal vascular patterning and thinner coronaries [12]. Despite this phenotype, vegfaa mutants do not show obvious defects in vascular integrity and differentiation. However, these mutants have significantly reduced cardiac function as detected by echocardiography, similar to mouse mutants [81]. Surprisingly, the vegfaa/ mutant hearts revascularize after cardiac injury, although this response is delayed at early time points. Based on these observations and transcriptomic analyses, it was hypothesized that adult vegfaa mutants deploy compensatory mechanisms as reported at embryonic stages [75]. Future studies using this mutant should help to elucidate the role of vegfaa during the development of the coronary network in zebrafish.

VEGFA in Zebrafish Heart Regeneration

305

5.2 Tg(hsp70l: vegfaa121-F17A)

To activate downstream signaling, Vegfa dimerizes before binding to its receptors. Based on this property, previous studies found that mutations in key residues affect the ability of VEGFA to bind VEGFR2, therefore potentially exerting a dominant-negative effect upon the dimerization of mutant to wild-type molecules [82]. Most of these residues are highly conserved in zebrafish and have been experimentally tested to be key for Vegfa function in vivo [76]. Among them, the mutation of phenylalanine 17 to alanine (F17A) reduces affinity for VEGFR2 by 90-fold [82]. In zebrafish, expression of a dominant-negative form of Vegfaa by mutating the same residues leads to the inactive heterodimer formation of wildtype Vegfaa and dn-Vegfaa, failing to stimulate angiogenesis via Vegfr2 [76]. Based on its potential to act as a dominant-negative form of Vegfaa, Marı´n-Juez et al. developed a Tg(hsp70l:vegfaa121F17A) line that overexpresses Vegfaa121-F17A upon heat-shock [12]. After induction, Tg(hsp70l:vegfaa121-F17A) fish display impaired angiogenesis at early developmental [12, 83] and adult stages [12, 84]. Overexpression of this dominant-negative form of Vegfaa blocked revascularization, resulting in reduced CM proliferation during heart regeneration in zebrafish. Importantly, inhibition of early sprouting with this transgenic line led to scar retention and impaired CM replenishment, indicating that the rapid revascularization of the injury is required to achieve cardiac regeneration [12]. Moreover, recent studies using an adult zebrafish model for pathological vascular remodeling have shown that Tg(hsp70l:vegfaa121-F17A) expression prevents hypoxia-induced vascular intussusception [84].

5.3 Tg(hsp70l:loxpStop-loxp-mTomcodOptP2AT46Avegfaa)

It is equally important to understand whether superficial and intraventricular coronaries regulate different aspects of CM regeneration. Cardiac cryoinjury elicits a heat-shock like response and induces the expression of genes under the control of hsp70l regulatory sequences [12], precluding the use of heat-shock inducible lines to stimulate gene expression exclusively at later stages after injury. To circumvent this limitation, a Tg(hsp70l:loxp-Stop-loxpmTom-codOptP2A-T46Avegfaa) (hereafter Tg(hsp70l:LSL-dnvegfaa)), was recently developed [21]. This form of Vegfaa also acts as a dominant-negative and blocks angiogenesis in zebrafish [76]. Taking advantage of the HOTcre system [85], when combined with a CreERT2 line, this Tg(hsp70l:LSL-dnvegfaa) line allows temporal and spatial control of dnVegfaa overexpression. This line was used with a Tg(myl7:CreERT2) line that allows efficient recombination in CMs upon tamoxifen administration [47]. Using this approach, interfering with early revascularization was avoided and allowed investigation of the role that intraventricular revascularization has in CM repopulation [21]. Although the HOTcre system allows tissue-specific overexpression of transgenes in a time-controlled manner, it relies on

306

Kaushik Chowdhury et al.

efficient CreERT2 lines. In the case of secreted factors like Vegfaa, this limitation can be resolved using a CreERT2 line driving recombination in cells neighboring the cells of interest. Although the number of tissue-specific CreERT2 lines available is increasing, new lines capable of strong and efficient recombination in adult animals are required. 5.4

flt12/2

The FLT1 gene encodes for both mFLT1 (VEGFR1), a membranebound tyrosine kinase receptor, and sFLT1, a soluble form of the receptor lacking the transmembrane and tyrosine kinase domains [86]. While mFLT1 mediates VEGF signaling, sFLT1 functions as a potent VEGFA decoy receptor [86]. Flt1 KO mice display severe endothelial overgrowth and embryonic lethality [87]. Similarly, zebrafish flt1 mutants (i.e., lacking both membrane and soluble forms of the receptor) display vascular over-sprouting at embryonic stages [77]. However, zebrafish mflt1 mutants do not display any apparent phenotypes, indicating that vascular over-sprouting is caused by sFlt1 deficiency [77]. Unlike mouse mutants, zebrafish flt1 mutants survive to adulthood. Similar to the embryonic phenotypes, adult flt1 mutants exhibit increased revascularization [21]. During heart regeneration, these mutants display a significant increase in the number and length of intra-coronary sprouts, supporting a role for endocardial Vegfaa during revascularization in zebrafish [21]. It is worth mentioning that Matsuoka and colleagues also generated a Tg(hsp70l:sflt1) line. Using this overexpression line, they were able to efficiently block developmental angiogenesis in zebrafish [77]. Although this transgenic line has not been used for adult cardiac regeneration studies, overexpression of sFlt1 should allow the interruption of coronary sprouting, similar to Vegfaa121-F17A.

5.5

nrp1a2/2

Neuropilins are single transmembrane non-tyrosine kinase co-receptors with the potential to bind different ligands [88– 93]. The Neuropilin family is composed of two isoforms, NRP1 and NRP2, and regulates a wide range of processes, including vascular patterning and angiogenesis. Namely, NRP1 was initially identified as a regulator of angiogenesis via VEGF [92–94]. Studies in zebrafish using knockdown approaches showed that Vegfa and Nrp1 regulate vascular development synergistically [95]. Lowe and colleagues found that the expression of different neuropilins changes during heart regeneration in zebrafish [74]. Specifically, zebrafish nrp1a mutants display reduced coronary coverage and impaired epicardial response during heart regeneration [74]. These results indicate that zebrafish cardiac revascularization is regulated, at least in part, via Nrp1a acting as a co-receptor for Vegfaa. Whether Nrp1a regulates the endothelial and epicardial

VEGFA in Zebrafish Heart Regeneration

307

responses to injury upon binding to any of its other ligands remains to be determined. 5.6 Tg(βactin2:loxPmTagBFP-STOP-loxPvegfaa)

6

So far, most zebrafish studies have developed and utilized loss-offunction approaches to investigate the role of Vegfaa signaling during heart regeneration in zebrafish. In a recent study, Karra and colleagues investigated the role of Vegfaa during development and regeneration [20]. The authors developed a TgBAC(vegfaa: EGFP) reporter line and found EGFP expression in the epicardium of uninjured and regenerating animals, as well as in endocardial cells after injury [20]. To further test the role of Vegfaa, they developed a Tg(βactin2:loxP-mTagBFP-STOP-loxP-vegfaa) line (hereafter Tg (βactin2:LBSL-vegfaa)) that combined with a Tg(myl7:CreERT2) line allowed CM overexpression of Vegfaa. Using this gain-offunction approach, they observed that the overexpression of Vegfaa led to hypervascularization, cardiomegaly, and increased CM proliferation [20]. Despite the stimulation of vascular growth and CM proliferation, Tg(myl7:CreERT2); Tg(βactin2:LBSL-vegfaa) fish failed to regenerate after the induction of Vegfaa overexpression. Interestingly, although heart regeneration was impaired in Vegfaa overexpressing hearts, CM proliferation was increased throughout the whole ventricle. Similarly, previous studies suggested that VEGFA can stimulate CM cell cycle re-entry [96].

Concluding Remarks Therapeutic angiogenesis has been extensively studied as an alternative approach to treat MI patients. After promising results in animal models, a broad range of factors and treatment modalities have been tested in clinical trials, with no apparent effects on humans. Importantly, when most of these trials were designed and conducted, the knowledge gained from animals with endogenous cardiac regenerative capacity was extremely limited. Since the pioneering study by Poss et al. was first published [22], the zebrafish cardiac regeneration field has been continuously expanding. Recent findings in zebrafish indicate that coronary vessels, beyond their role as a passive transport system, are active regulators of the regenerative response. Over the last years, a number of studies have highlighted the ability of different vascular beds to regulate tissue development, homeostasis, and regeneration [97]. This evidence open new therapeutic avenues where endothelial factors could be utilized to stimulate tissue regeneration. Being a potent mitogen, VEGFA has been, and still is, considered a factor with promising therapeutic applications. Indeed, studies using the zebrafish model have identified Vegfa signaling to be key in the regulation of cardiac regeneration. Still, the major questions that contributed to the failure of clinical trials over the last decades remain: When and

308

Kaushik Chowdhury et al.

where should VEGFA be administered? At what concentrations? In combination with which other factors? In the time of the genome engineering revolution, the zebrafish model holds a privileged position to help answer these and other important questions in the path toward heart regeneration therapies.

7

Transgenic Lines Acronyms and Abbreviations hsp70l: heatshock cognate 70-kd protein like (regulatory sequence used as a promoter) vegfaa121-F17A: vegfaa121 variant where the codon coding for Phenylalanine 17 was mutagenized to code for Alanine (F17A) mTom: N-terminal membrane tagged version of tdTomato reporter codOptP2A: codon optimized P2A self-cleaving viral peptide sequence T46Avegfaa: vegfaa121 variant where the codon coding for Threonine 46 was mutagenized to code for Alanine (T46A) STOP: transcription stop sequence myl7: myosin light chain 7 (regulatory sequence used as a promoter) βactin2: beta 2 actin (regulatory sequence used as a promoter) mTagBFP: blue fluorescent protein tag etsrp: ETS1-related protein (regulatory sequence used as a promoter) flt1: fms related receptor tyrosine kinase 1 (regulatory sequence used as a promoter) Mmu.Fos: minimal mouse Fos promoter (regulatory sequence used as a promoter) nrp1a: neuropilin 1a HOTcre: Heat induced Cre-based genetic switch

Acknowledgments We thank Michelle Collins, Stephanie Larrive´e-Vanier, Hadil El-Sammak, and Armaan Mehra for comments on the manuscript and discussions. The research in Lai group is supported by the Ministry of Science and Technology (MOST 108-2320-B-001032-MY2) and the IBMS/Academia Sinica (IBMS-CRC108P03) in Taiwan. Kaushik Chowdhury is the recipient of the AS-TIGP Research Progress Fellowship. The Marı´n-Juez lab is currently supported by the Canadian Institutes of Health Research (PJT-178037).

VEGFA in Zebrafish Heart Regeneration

309

References 1. Benjamin EJ et al (2019) Heart disease and stroke statistics-2019 update: a report from the American Heart Association. Circulation 139(10):e56–e528 2. van den Borne SW et al (2010) Myocardial remodeling after infarction: the role of myofibroblasts. Nat Rev Cardiol 7(1):30–37 3. Laflamme MA, Murry CE (2011) Heart regeneration. Nature 473(7347):326–335 4. Beltrami AP et al (2001) Evidence that human cardiac myocytes divide after myocardial infarction. N Engl J Med 344(23):1750–1757 5. Bergmann O et al (2009) Evidence for cardiomyocyte renewal in humans. Science 324(5923):98–102 6. Mollova M et al (2013) Cardiomyocyte proliferation contributes to heart growth in young humans. Proc Natl Acad Sci 110(4): 1446–1451 7. Polizzotti BD et al (2015) Neuregulin stimulation of cardiomyocyte regeneration in mice and human myocardium reveals a therapeutic window. Sci Transl Med 7(281):281ra45 8. Sahara M, Santoro F, Chien KR (2015) Programming and reprogramming a human heart cell. EMBO J 34(6):710–738 9. Uygur A, Lee RT (2016) Mechanisms of cardiac regeneration. Dev Cell 36(4):362–374 10. Das S et al (2019) A unique collateral artery development program promotes neonatal heart regeneration. Cell 176(5):1128–1142 11. Habib GB et al (1991) Influence of coronary collateral vessels on myocardial infarct size in humans. Results of phase I thrombolysis in myocardial infarction (TIMI) trial. The TIMI Investigators. Circulation 83(3):739–746 12. Marin-Juez R et al (2016) Fast revascularization of the injured area is essential to support zebrafish heart regeneration. Proc Natl Acad Sci U S A 113(40):11237–11242 13. Sabia PJ et al (1992) An association between collateral blood flow and myocardial viability in patients with recent myocardial infarction. N Engl J Med 327(26):1825–1831 14. Taimeh Z et al (2013) Vascular endothelial growth factor in heart failure. Nat Rev Cardiol 10(9):519–530 15. Carmeliet P (2000) Mechanisms of angiogenesis and arteriogenesis. Nat Med 6(4):389–395 16. Molin D, Post MJ (2007) Therapeutic angiogenesis in the heart: protect and serve. Curr Opin Pharmacol 7(2):158–163 17. Robich MP et al (2011) Myocardial therapeutic angiogenesis: a review of the state of

development and future obstacles. Expert Rev Cardiovasc Ther 9(11):1469–1479 18. Patterson C, Runge MS (2000) Therapeutic myocardial angiogenesis via vascular endothelial growth factor gene therapy: moving on down the road. Circulation 102(9):940–942 19. Svensson EC et al (1999) Efficient and stable transduction of cardiomyocytes after intramyocardial injection or intracoronary perfusion with recombinant adeno-associated virus vectors. Circulation 99(2):201–205 20. Karra R et al (2018) Vegfaa instructs cardiac muscle hyperplasia in adult zebrafish. Proc Natl Acad Sci U S A 115(35):8805–8810 21. Marin-Juez R et al (2019) Coronary revascularization during heart regeneration is regulated by epicardial and endocardial cues and forms a scaffold for cardiomyocyte repopulation. Dev Cell 51(4):503–515 e4 22. Poss KD, Wilson LG, Keating MT (2002) Heart regeneration in zebrafish. Science 298(5601):2188–2190 23. Wang J et al (2011) The regenerative capacity of zebrafish reverses cardiac failure caused by genetic cardiomyocyte depletion. Development 138(16):3421–3430 24. Chablais F et al (2011) The zebrafish heart regenerates after cryoinjury-induced myocardial infarction. BMC Dev Biol 11(1):21 25. Gonzalez-Rosa JM et al (2011) Extensive scar formation and regression during heart regeneration after cryoinjury in zebrafish. Development 138(9):1663–1674 26. Schnabel K et al (2011) Regeneration of cryoinjury induced necrotic heart lesions in zebrafish is associated with epicardial activation and cardiomyocyte proliferation. PLoS One 6(4):e18503 27. Bevan L et al (2020) Specific macrophage populations promote both cardiac scar deposition and subsequent resolution in adult zebrafish. Cardiovasc Res 116(7):1357–1371 28. Lai SL et al (2017) Reciprocal analyses in zebrafish and medaka reveal that harnessing the immune response promotes cardiac regeneration. Elife 6:e25605 29. Simoes FC et al (2020) Macrophages directly contribute collagen to scar formation during zebrafish heart regeneration and mouse heart repair. Nat Commun 11(1):600 30. Kikuchi K et al (2011) Retinoic acid production by endocardium and epicardium is an injury response essential for zebrafish heart regeneration. Dev Cell 20(3):397–404

310

Kaushik Chowdhury et al.

31. Koth J et al (2020) Runx1 promotes scar deposition and inhibits myocardial proliferation and survival during zebrafish heart regeneration. Development 147(8):dev186569 32. Munch J et al (2017) Notch signalling restricts inflammation and serpine1 expression in the dynamic endocardium of the regenerating zebrafish heart. Development 144(8):1425–1440 33. Zhao L et al (2019) Endocardial Notch signaling promotes cardiomyocyte proliferation in the regenerating zebrafish heart through Wnt pathway antagonism. Cell Rep 26(3): 546–554 e5 34. Zhao L et al (2014) Notch signaling regulates cardiomyocyte proliferation during zebrafish heart regeneration. Proc Natl Acad Sci U S A 111(4):1403–1408 35. Lepilina A et al (2006) A dynamic epicardial injury response supports progenitor cell activity during zebrafish heart regeneration. Cell 127(3):607–619 36. Kim J et al (2010) PDGF signaling is required for epicardial function and blood vessel formation in regenerating zebrafish hearts. Proc Natl Acad Sci U S A 107(40):17206–17210 37. Harrison MRM et al (2015) Chemokineguided angiogenesis directs coronary vasculature formation in zebrafish. Dev Cell 33(4): 442–454 38. Gancz D et al (2019) Distinct origins and molecular mechanisms contribute to lymphatic formation during cardiac growth and regeneration. Elife 8:e44153 39. Harrison MR et al (2019) Late developing cardiac lymphatic vasculature supports adult zebrafish heart function and regeneration. Elife 8: e42762 40. Vivien CJ et al (2019) Vegfc/d-dependent regulation of the lymphatic vasculature during cardiac regeneration is influenced by injury context. NPJ Regen Med 4:18 41. Cao J, Poss KD (2018) The epicardium as a hub for heart regeneration. Nat Rev Cardiol 15(10):631–647 42. González-Rosa JM, Peralta M, Mercader N (2012) Pan-epicardial lineage tracing reveals that epicardium derived cells give rise to myofibroblasts and perivascular cells during zebrafish heart regeneration. Dev Biol 370(2):173–186 43. Cao Y, Cao J (2018) Covering and re-covering the heart: development and regeneration of the epicardium. J Cardiovasc Dev Dis 6(1):3 44. Lafontant PJ et al (2013) Cardiac myocyte diversity and a fibroblast network in the junctional region of the zebrafish heart revealed by transmission and serial block-face scanning electron microscopy. PLoS One 8(8):e72388

45. Sanchez-Iranzo H et al (2018) Transient fibrosis resolves via fibroblast inactivation in the regenerating zebrafish heart. Proc Natl Acad Sci U S A 115(16):4188–4193 46. Jopling C et al (2010) Zebrafish heart regeneration occurs by cardiomyocyte dedifferentiation and proliferation. Nature 464(7288): 606–U168 47. Kikuchi K et al (2010) Primary contribution to zebrafish heart regeneration by gata4(+) cardiomyocytes. Nature 464(7288):601–U162 48. Gonzalez-Rosa JM, Burns CE, Burns CG (2017) Zebrafish heart regeneration: 15 years of discoveries. Regeneration (Oxf) 4(3): 105–123 49. Tzahor E, Poss KD (2017) Cardiac regeneration strategies: staying young at heart. Science 356(6342):1035–1039 50. Mahmoud AI et al (2015) Nerves regulate cardiomyocyte proliferation and heart regeneration. Dev Cell 34(4):387–399 51. Helfant RH, Vokonas PS, Gorlin R (1971) Functional importance of the human coronary collateral circulation. N Engl J Med 284(23): 1277–1281 52. Kim EK et al (2016) A protective role of early collateral blood flow in patients with ST-segment elevation myocardial infarction. Am Heart J 171(1):56–63 53. Seiler C et al (2013) The human coronary collateral circulation: development and clinical importance. Eur Heart J 34(34):2674–2682 54. Gupta R, Tongers J, Losordo DW (2009) Human studies of angiogenic gene therapy. Circ Res 105(8):724–736 55. Tian X, Pu WT, Zhou B (2015) Cellular origin and developmental program of coronary angiogenesis. Circ Res 116(3):515–530 56. Pinto AR et al (2016) Revisiting cardiac cellular composition. Circ Res 118(3):400–409 57. Patra C et al (2017) The zebrafish ventricle: a hub of cardiac endothelial cells for in vitro cell behavior studies. Sci Rep 7(1):2687 58. Tian X et al (2014) De novo formation of a distinct coronary vascular population in neonatal heart. Science 345(6192):90–94 59. Wu B et al (2012) Endocardial cells form the coronary arteries by angiogenesis through myocardial-endocardial VEGF signaling. Cell 151(5):1083–1096 60. Dube KN et al (2017) Recapitulation of developmental mechanisms to revascularize the ischemic heart. JCI Insight 2(22):e96800 61. Miquerol L et al (2015) Endothelial plasticity drives arterial remodeling within the

VEGFA in Zebrafish Heart Regeneration endocardium after myocardial infarction. Circ Res 116(11):1765–1771 62. He L, Lui KO, Zhou B (2020) The formation of coronary vessels in cardiac development and disease. Cold Spring Harb Perspect Biol 12(5): a037168 63. van Royen N et al (2009) A critical review of clinical arteriogenesis research. J Am Coll Cardiol 55(1):17–25 64. Potente M, Gerhardt H, Carmeliet P (2011) Basic and therapeutic aspects of angiogenesis. Cell 146(6):873–887 65. Aurora AB et al (2014) Macrophages are required for neonatal heart regeneration. J Clin Invest 124(3):1382–1392 66. Apte RS, Chen DS, Ferrara N (2019) VEGF in signaling and disease: beyond discovery and development. Cell 176(6):1248–1264 ´ lvarez-Aznar A, Muhl L, Gaengel K (2017) 67. A VEGF receptor tyrosine kinases: key regulators of vascular function. In: Current topics in developmental biology. Elsevier, Amsterdam, pp 433–482 68. Hudlicka O, Brown M, Egginton S (1992) Angiogenesis in skeletal and cardiac muscle. Physiol Rev 72(2):369–417 69. Ferrara N et al (1996) Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380(6573): 439–442 70. Carmeliet P et al (1999) Impaired myocardial angiogenesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF 164 and VEGF 188. Nat Med 5(5):495–502 71. Dor Y et al (2002) Conditional switching of VEGF provides new insights into adult neovascularization and pro-angiogenic therapy. EMBO J 21(8):1939–1947 72. Zangi L et al (2013) Modified mRNA directs the fate of heart progenitor cells and induces vascular regeneration after myocardial infarction. Nat Biotechnol 31(10):898–907 73. Bayliss PE et al (2006) Chemical modulation of receptor signaling inhibits regenerative angiogenesis in adult zebrafish. Nat Chem Biol 2(5): 265–273 74. Lowe V et al (2019) Neuropilin 1 mediates epicardial activation and revascularization in the regenerating zebrafish heart. Development 146(13):dev174482 75. Rossi A et al (2015) Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature 524(7564): 230–233

311

76. Rossi A et al (2016) Regulation of Vegf signaling by natural and synthetic ligands. Blood 128(19):2359–2366 77. Matsuoka RL et al (2016) Radial glia regulate vascular patterning around the developing spinal cord. Elife 5:e20253 78. Bahary N et al (2007) Duplicate VegfA genes and orthologues of the KDR receptor tyrosine kinase family mediate vascular development in the zebrafish. Blood 110(10):3627–3636 79. Koenig AL et al (2016) Vegfa signaling promotes zebrafish intestinal vasculature development through endothelial cell migration from the posterior cardinal vein. Dev Biol 411(1): 115–127 80. Matsuoka RL et al (2017) CNS-resident progenitors direct the vascularization of neighboring tissues. Proc Natl Acad Sci U S A 114(38): 10137–10142 81. May D et al (2008) Transgenic system for conditional induction and rescue of chronic myocardial hibernation provides insights into genomic programs of hibernation. Proc Natl Acad Sci U S A 105(1):282–287 82. Muller YA et al (1997) Vascular endothelial growth factor: crystal structure and functional mapping of the kinase domain receptor binding site. Proc Natl Acad Sci U S A 94(14): 7192–7197 83. Walton EM et al (2018) Cyclopropane modification of trehalose dimycolate drives granuloma angiogenesis and mycobacterial growth through Vegf signaling. Cell Host Microbe 24(4):514–525 e6 84. Ali Z et al (2019) Intussusceptive vascular remodeling precedes pathological neovascularization. Arterioscler Thromb Vasc Biol 39(7): 1402–1418 85. Hesselson D et al (2009) Distinct populations of quiescent and proliferative pancreatic betacells identified by HOTcre mediated labeling. Proc Natl Acad Sci U S A 106(35): 14896–14901 86. Kendall RL, Thomas KA (1993) Inhibition of vascular endothelial cell growth factor activity by an endogenously encoded soluble receptor. Proc Natl Acad Sci U S A 90(22): 10705–10709 87. Fong GH et al (1995) Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376(6535): 66–70 88. Ball SG et al (2010) Neuropilin-1 regulates platelet-derived growth factor receptor signalling in mesenchymal stem cells. Biochem J 427(1):29–40

312

Kaushik Chowdhury et al.

89. Prud’homme GJ et al (2016) Neuropilin-1 is a receptor for extracellular miRNA and AGO2/ miRNA complexes and mediates the internalization of miRNAs that modulate cell function. Oncotarget 7(42):68057–68071 90. West DC et al (2005) Interactions of multiple heparin binding growth factors with neuropilin-1 and potentiation of the activity of fibroblast growth factor-2. J Biol Chem 280(14):13457–13464 91. Pellet-Many C et al (2011) Neuropilin-1 mediates PDGF stimulation of vascular smooth muscle cell migration and signalling via p130Cas. Biochem J 435(3):609–618 92. Gu C et al (2003) Neuropilin-1 conveys semaphorin and VEGF signaling during neural and cardiovascular development. Dev Cell 5(1): 45–57

93. Kitsukawa T et al (1997) Neuropilinsemaphorin III/D-mediated chemorepulsive signals play a crucial role in peripheral nerve projection in mice. Neuron 19(5):995–1005 94. Kawasaki T et al (1999) A requirement for neuropilin-1 in embryonic vessel formation. Development 126(21):4895–4902 95. Lee P et al (2002) Neuropilin-1 is required for vascular development and is a mediator of VEGF-dependent angiogenesis in zebrafish. Proc Natl Acad Sci U S A 99(16): 10470–10475 96. Laguens R et al (2002) Entrance in mitosis of adult cardiomyocytes in ischemic pig hearts after plasmid-mediated rhVEGF165 gene transfer. Gene Ther 9(24):1676–1681 97. Rafii S, Butler JM, Ding BS (2016) Angiocrine functions of organ-specific endothelial cells. Nature 529(7586):316–325

Chapter 23 Three-Dimensional Visualization of Blood and Lymphatic Vessels in the Adult Zebrafish Heart by Chemical Clearing Lorna R. Fiedler, Paul R. Riley, and Roger Patient Abstract Unlike humans, the zebrafish can repair and regenerate its heart following injury. Understanding the molecular and physiological mechanisms of heart regeneration is critical in developing pro-regenerative strategies for clinical application. The cardiac lymphatic and non-lymphatic vasculature both respond to injury in zebrafish and are instrumental in driving optimal repair and regeneration. However, progress has been impeded by an inability to obtain high resolution images to clearly visualize and thus to fully understand the vascular responses in the injured heart and how this might intersect with successful repair and regeneration in humans. In this chapter, we describe a chemical clearing approach using Clear Unobstructed Brain/Body Imaging Cocktails and Computational analysis (CUBIC), for obtaining high resolution images of the adult zebrafish heart. This approach permits three-dimensional reconstruction of cardiac vasculature throughout the entire organ. By applying CUBIC methodology to tissues from transgenic zebrafish reporter lines or in conjunction with immunofluorescent staining, optical slices can be be generated, negating the need for standard tissue processing and sectioning procedures and yielding higher resolution images. The resultant images enable a holistic view of the coronary blood and lymphatic vasculature during heart injury and regeneration. Herein, we describe our protocol for visualizing vessels in the adult zebrafish heart using these approaches. Key words Chemical clearing, CUBIC, Heart, Lymphatics, Blood vessels, Vasculature, Zebrafish

1

Introduction The lymphatic and blood vessels of the heart are dynamic structures that respond to locally released factors when the heart is damaged during a heart attack (myocardial infarction; MI). This cardiac event can be reproduced in zebrafish using a liquid nitrogen cooled probe that is applied directly to the exposed ventricle [1–3]. In mice, MI is modeled more commonly by occluding the coronary vessel that supplies blood to the ventricle, by means of tying a suture around the vessel (ischemia). This can be left as a permanent occlusion or released to model a second wave of injury, reperfusion

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_23, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

313

314

Lorna R. Fiedler et al.

injury, that occurs clinically when the coronary artery is unblocked and blood flow restored [4]. In humans, following MI, the damaged tissue is removed by incoming and tissue-resident immune cells and replaced with a collagen-rich scar, which in regenerative and non-regenerative species, is derived from both fibroblasts and immune cells [5]. This extracellular matrix deposition serves to repair the wound and maintain tissue integrity but is non-contractile. Cardiac muscle is non-regenerative, and as such the scar persists and becomes more mature and stiffer over time, leading to sustained stress on the remaining muscle. This eventually leads to progressive cardiomyocyte drop-out, persistent fibrosis, pathological remodeling, and ultimately heart failure. Unlike mammals, adult zebrafish fully regenerate the heart following injury. Of note, mice initially possess the capacity for full regeneration at 1 day old (P1, postnatal) but by P7 this potential is lost and, as in adults, scarring persists [6, 7]. Lymphangiogenesis is a feature of the cardiac injury response in rodents and humans, and in mice, this is critical for clearing excess tissue fluid (oedema) and immune cells to draining lymph nodes [8–11]. In zebrafish with genetically disrupted lymphatics, hearts became non-regenerative, along with impaired clearance of necrotic tissue and sustained inflammation [12], highlighting a critical role for the cardiac lymphatics in the regenerative response [12]. Heart injury also stimulates angiogenesis, which is essential for restoration of blood flow to the injured myocardium, and similarly to the lymphatic vasculature, these responses differ between regenerative and non-regenerative species [13]. In zebrafish, blocking neo-angiogenesis of coronary vessels also impairs regeneration, highlighting the importance of the non-lymphatic vasculature in the injury response [14]. While stimulation of these processes presents an opportunity for therapeutic intervention, a better understanding of the cellular and molecular processes involved is critical to fully realize this potential [13, 15]. A key utility of zebrafish is in understanding the processes that govern repair and regeneration in the heart, and to explore the molecular mechanisms involved, with the aim of applying this knowledge to development of pro-regenerative therapeutics for treating MI in the clinic. However, progress has been hampered by difficulties in clearly visualizing the cardiovascular tissues and specific cell types in the adult setting due to tissue density and opacity. This includes the blood and lymphatic vasculature, which consequently prohibits a deep understanding of their role in cardiac repair and regeneration. While standard immunohistochemical and immunofluorescent staining approaches are well established, the resolution of these is limiting. Chemical clearing is an alternative approach to render the heart (or any tissue of interest) transparent. When combined with transgenic fluorescent reporter zebrafish lines or antibody-based immunofluorescent staining and confocal

Visualizing Vasculature in the Adult Zebrafish Heart

315

laser scanning microscopy (CLSM) to capture stepwise optical slices, this is a powerful approach that can be used to clearly visualize cells, lineages, tissues, or whole organs. The acquired images can then be used to reconstruct the structures of interest, in this case the cardiac vasculature, in their entirety at an unprecedented resolution. Several such reagents have been developed, with CUBIC being one of the most efficient and one that serves to preserve fluorescent signals [16–21]. Here, we describe our protocol for optical clearing and visualization of the coronary blood and lymphatic vasculature in the adult zebrafish heart, using both fluorescent reporter lines (Figs. 1 and 2) and immunofluorescent staining (Fig. 3), with CLSM and maximum intensity projections.

2

Materials

2.1 Chemical Clearing Reagents

See Note 1. 1. CUBIC-L (T3740; TCI Chemicals; N-butyldiethanolamine +10% Triton X-100).

10%

2. CUBIC-R+ (M) (T3741; TCI Chemicals; 30% N-methylnicotinamide +45% antipyrine, buffered with 0.5% (v/w) N-butyldiethanolamine, pH ~10). 2.2

Solutions

1. Phosphate buffered saline (PBS). 2. Autoclaved ddH2O. 3. Ringer’s solution + Heparin (final concentration 50 units/mL; H4784, Sigma). 4. Fixing solution: 4% paraformaldehyde (PFA) in PBS. 5. Wash buffer: PBS + 0.1% Tween-20. 6. Blocking buffer: 10% normal goat serum (NGS) + 0.1% Tween-20 + 0.5% Tx-100 in PBS. 7. Antibody dilution buffer: 1% normal goat serum +0.5% Tx-100 + 0.1% Tween-20 + 0.01% Thimerosal (SBR0018, Sigma, see Note 2). 8. 50% CUBIC-L solution: CUBIC-L diluted in autoclaved ddH2O. 9. 50% CUBIC-R+(M) solution: CUBIC-R+(M) diluted in autoclaved ddH2O.

2.3

Antibodies

1. Primary Antibodies: Prox1 (AB5475, Merck-Millipore, rabbit polyclonal serum) diluted 1:500 in antibody dilution buffer.

316

Lorna R. Fiedler et al.

Fig. 1 Visualizing the coronary vasculature of the zebrafish heart. Zebrafish coronary vasculature was visualized by CUBIC clearing of hearts from the fluorescent reporter lines Kdrl-HsRas-mCherry, that permits visualization of venous and arterial endothelial cells, and Fli1a-eGFP that is a pan-endothelial cell marker. Maximum intensity projections are shown. (a) Venous and arterial endothelial cells were visualized using the fluorescent reporter line Kdrl-HsRas-mCherry. The image shows a heart from a 2-year-old zebrafish, where an extensive vascular system can be seen. (b) The pan-endothelial cell marker Fli1a-eGFP was used to demarcate lymphatic and non-lymphatic endothelial cells. The image shows a heart from an 8-month-old zebrafish, where a less extensive vascular system can be seen. (c), (d) Higher magnification images of the white boxed regions in parts (a) and (b), respectively. The white scale bar indicates 200 μm. Ve ventricle, At atrium, BA bulbus arteriosus

Fig. 2 Visualizing the coronary vasculature of the zebrafish heart in optical slices. (a) An optical section from the same heart shown in Fig. 1—maximum intensity projection of a 2-year-old zebrafish heart from the reporter line Kdrl-HsRas-mCherry. (b) A higher magnification image of the white boxed region. The white scale bar indicates 200 μm. Ve ventricle

Visualizing Vasculature in the Adult Zebrafish Heart

317

Fig. 3 Visualizing Prox1-positive lymphatic networks in the zebrafish heart. (a) Lymphatics in the zebrafish heart at 11 months of age were visualized by immunostaining with antibodies to Prox1 (green), followed by chemical clearing. Maximum intensity projections show the presence of surface Prox1 positive vessel structures that are extensively extended around the ventricle. (b) shows an optical slice from the surface of the heart shown in (a), with the white boxed regions showing progressively higher magnification in (c) and (d). Another optical slice, towards the center of the heart, is shown in (e), with progressively higher magnification of the white boxed region shown in (f) and (g). BA bulbus arteriosus, Ve ventricle. White scale bar indicates 200 μm

2. Secondary antibodies: Alexa-488 conjugated Goat anti-rabbit IgG (H + L) (Invitrogen, supplied at 2 mg/mL) diluted 1:500 in antibody dilution buffer. 2.4

Consumables

1. 35 mm dishes with glass coverslip bottom (#1.5, 81218-200, Ibidi). 2. Glass coverslips. 3. 35 mm petri dishes. 4. Plastic Pasteur pipettes. 5. 1.5 mL autoclaved Eppendorf tubes. 6. 50 mL Falcon tubes. 7. Tin foil.

2.5

Equipment

1. Dissecting microscope. 2. Rotator at room temperature or 37  C. 3. Confocal laser scanning microscope (CLSM), e.g., inverted Zeiss 880 confocal microscope. 4. Dissecting instruments.

318

3

Lorna R. Fiedler et al.

Methods

3.1 Tissue Collection and Preparation

1. Excise hearts and place into Ringer’s solution + Heparin in a 35 mm petri dish. 2. Transfer hearts to a 1.5 mL autoclaved Eppendorf tube containing 1 mL fresh Ringer’s solution + Heparin and rotate at room temperature. 3. Once all hearts have been collected, place into fresh 1.5 mL Eppendorf tubes containing approximately 500 μL fixing solution and incubate overnight in the fridge. 4. The following day, rotate samples at room temperature for 1 h (see Note 3). 5. Using a plastic Pasteur pipette, remove fixing solution and discard. 6. Briefly wash hearts to remove excess fixing solution by adding approximately 1 mL PBS using a fresh pipette and discarding the solution immediately. Repeat twice. 7. Add 1 mL PBS and rotate samples at room temperature for 1 h. 8. Discard solution and replace with fresh PBS. Rotate samples at room temperature for 1 h. 9. Repeat step 7 twice.

3.2 Tissue Clearing: Fluorescent Reporter Lines

Fluorescent reporter lines or antibody-based immunofluorescent staining can be used to visualize cells or structures of interest. This section describes a protocol for clearing hearts obtained from the fluorescent reporter lines Kdrl-HsRas-mCherry, that permits visualization of venous and arterial endothelial cells [22] and Fli1aeGFP, a pan-endothelial cell marker [23] (see Fig. 1). For antibody staining, in this case using antibodies to Prox1, a marker of lymphatic endothelial cells (see Fig. 3), proceed to Subheading 3.3. 1. Remove PBS and replace with 50% CUBIC-L solution, 300–500 μL per tube (see Note 4). 2. Rotate overnight at room temperature. 3. Remove 50% CUBIC-L solution and replace with 500 μL per tube 100% CUBIC-L solution. 4. Rotate for 2–5 days at room temperature or at 37  C, replacing with fresh 100% CUBIC-L solution after 1, 2, and 4 days (see Note 5). 5. Remove 100% CUBIC-L solution washed hearts briefly twice with PBS (1 mL per tube). 6. Add 1 mL PBS per tube and rotate samples for 90 min at room temperature. 7. Wash similarly twice more (see Note 6).

Visualizing Vasculature in the Adult Zebrafish Heart

319

8. Add 50% CUBIC-R+ (M) and rotate overnight at room temperature (see Note 7). 9. Remove and discard the solution and replace with 100% CUBIC-R+ (M) (see Note 8). 10. Rotate at room temperature for at least 2 days, covering the tubes with tin foil to protect from the light (see Notes 9 and 10). 11. Proceed to Subheading 3.4 for imaging. 3.3 Tissue Clearing: Immunofluorescent Antibody Staining

Here, we used primary and secondary antibody steps, but a directly conjugated primary antibody can also be used if desired. 1. Remove PBS and replace with 50% CUBIC-L solution, 300–500 μL per tube (see Note 4). 2. Rotate overnight at room temperature. 3. Remove 50% CUBIC-L solution and replace with 500 μL per tube 100% CUBIC-L solution. 4. Rotate for 2–5 days at room temperature or at 37  C, replacing with fresh 100% CUBIC-L solution after 1, 2, and 4 days (see Note 5). 5. Remove 100% CUBIC-L solution washed hearts briefly twice with PBS (1 mL per tube). 6. Add 1 mL PBS per tube and rotate samples for 90 min at room temperature. 7. Wash similarly twice more (see Note 6). 8. To block non-specific binding sites, remove PBS and add 1 mL blocking buffer. Rotate for 2 h at room temperature. 9. Remove blocking buffer and replace with primary antibody solution (see Note 11). 10. Rotate at room temperature for 3–4 days. 11. Rinse hearts twice briefly with wash buffer (1 mL per tube), before adding another 1 mL and rotating for 90 min at room temperature. 12. Remove wash buffer, add another 1 mL and rotate for 90 min at room temperature. 13. Repeat. 14. Add secondary antibody solution and rotate at room temperature for 2–3 days, covering the tubes with tin foil to protect from the light. 15. Wash hearts similarly to after primary antibody (steps 11–13) then incubate in 50% CUBIC-R+ (M) overnight, rotating at room temperature (see Note 7).

320

Lorna R. Fiedler et al.

16. Remove and discard the solution and replace with 100% CUBIC-R+ (M) (see Note 8). 17. Rotate at room temperature for at least 2 days (see Notes 9 and 10). 3.4

Imaging

1. Remove hearts from the Eppendorf tube and transfer to a glass-bottomed 35 mm dish, placing the heart on the glassed area. 2. Add a generous drop of 100% CUBIC-R+ (M) from the tube; enough to cover the entire heart. 3. Place a glass coverslip on top to hold the heart in place. 4. Obtain whole organ images on an inverted confocal laser scanning microscope (we used a Zeiss 880) at 20 magnification, scanning the full depth of the heart at 5–10 μm steps (z-stacks), using the tiling function to capture the entire x-y length and breadth of the heart (see Notes 12 and 13). 5. Optical slices can be viewed and exported (see Figs. 2 and 3), and maximum intensity projections of confocal stacks reconstructed from the tiled and z-stack acquisitions using, e.g., Imaris or Zeiss confocal software (see Figs. 1 and 3).

4

Notes 1. CUBIC-R+ (M) is equivalent to CUBIC-R+ (but with the addition of a buffering agent). CUBIC-R+ is also sometimes referred to as CUBIC-RA [18, 24]. 2. Antibody solutions are incubated with samples at room temperature for several days, and a preservative such as Thimerosal is required to inhibit growth of bacterial or fungal contaminants. We recommend resuspending 100 mg Thimerosal (30376, Sigma) in 1 mL autoclaved ddH2O and diluting 1: 1000 to achieve a final concentration of 0.01%. Sodium azide is a common preservative that can also be used if desired. 3. Eppendorf tubes can be placed into a 50 mL falcon and the falcon placed on the rotator. This is also more convenient when samples need to be protected from light by wrapping the falcon tube in tin foil. 4. CUBIC-L is used for delipidation and decoloring tissue. Note that tissues will only be partially transparent at the end of this stage. 5. This step can be performed at room temperature or at 37  C; the latter will be faster. For small hearts, 2–3 days is sufficient (e.g., 2 days at 37  C or 3 days at room temperature), while

Visualizing Vasculature in the Adult Zebrafish Heart

321

large hearts should be incubated up to 5 days in CUBIC-L before proceeding to the next step. 6. At this stage, the tissue will be semi-transparent, and can be difficult to see, especially when hearts are very small. To avoid losing the sample, wait approximately 10 s to allow hearts to sink to the bottom of the tube before removing any solutions. Be careful not to remove all the liquid; leave a small volume undisturbed at the bottom of the tube. It is also advisable to check the heart has not been pulled into the plastic pipette by holding the pipette up to the light and turning—the tissue can generally be observed more easily in this manner than in the Eppendorf tube. If needed, the full liquid volume from the Eppendorf tube can be taken up into the pipette and examined in the light to confirm the sample is still present, before placing back into the tube and resuming the protocol. 7. CUBIC-R+ (M) is optimized for further decolorization (further to CUBIC-L) and RI matching, while also preserving the fluorescent signal. 8. At this stage, the tissue will be more fully transparent and even more difficult to see than following CUBIC-L incubation stages. Handle the tissue and solution changes as for Note 6. It can be generally assumed that the hearts are still present if the tissue and solution changes are handled as suggested, even if they cannot easily be seen. Once placed into the imaging dish, the hearts are generally more visible. 9. The manufacturer’s protocol states >1 day is sufficient; however, we have found that 2 days or more results in improved transparency of the tissue. If time is imperative, 1 day is sufficient, but better results will be obtained if the tissue is incubated for longer. Full transparency should be achieved within 2–3 days. Tissues can also be imaged and then placed back into fresh 100% CUBIC-R+ (M) and incubated further if transparency is not sufficient. 10. At this stage, the tissues can be stored in 100% CUBIC-R+ (M) for up to 1 month on the bench, wrapped in tin foil to protect from light. Imaging is usually carried out within 1–2 weeks but the fluorescent signal is maintained up to 1 month. Do not store samples in the fridge as the solution will become cloudy and precipitates will appear. The sample is stable at room temperature. 11. Optimal dilutions should be empirically determined for any antibody. A good starting point is to use the recommended dilution for immunofluorescent staining and to titrate down to a lower dilution as needed. If a range is given in the antibody datasheet, the lowest dilution is more likely to be required.

322

Lorna R. Fiedler et al.

12. Since imaging at high resolution can be very time consuming and access to equipment may be limited, we often perform a lower resolution scan with larger steps in the first instance and then repeat the scan at higher resolution/smaller steps on selected samples. 13. Owing to their small size, CLSM is suitable for imaging zebrafish hearts. For larger samples, e.g., adult mouse heart, lightsheet imaging is required.

Acknowledgments The authors would like to acknowledge OxStem, the funder of this project, and the Wolfson Imaging Centre, MRC Weatherall Institute at the University of Oxford, in particular Dr. Jana Koth for her valuable input. References 1. Chablais F, Veit J, Rainer G, Jaz´win´ska A (2011) The zebrafish heart regenerates after cryoinjury-induced myocardial infarction. BMC Dev Biol 11:21 2. González-Rosa JM, Martı´n V, Peralta M et al (2011) Extensive scar formation and regression during heart regeneration after cryoinjury in zebrafish. Development 138(9):1663–1674 3. Schnabel K, Wu CC, Kurth T, Weidinger G (2011) Regeneration of cryoinjury induced necrotic heart lesions in zebrafish is associated with epicardial activation and cardiomyocyte proliferation. PLoS One 6(4):e18503 4. De Villiers C, Riley PR (2020) Mouse models of myocardial infarction: comparing permanent ligation and ischaemia-reperfusion. Dis Model Mech 13(11):dmm046565 ˜ es FC, Cahill TJ, Kenyon A, 5. Simo Gavriouchkina D, Vieira JM, Sun. (2020) Macrophages directly contribute collagen to scar formation during zebrafish heart regeneration and mouse heart repair. Nat Commun 11(1):600 6. Porrello ER, Mahmoud AI, Simpson E, Hill JA et al (2011) Transient regenerative potential of the neonatal mouse heart. Science 331(6020): 1078–1080 7. Porrello ER, Mahmoud AI, Simpson E et al (2013) Regulation of neonatal and adult mammalian heart regeneration by the miR-15 family. Proc Natl Acad Sci U S A 110(1):187–192 8. Klotz L, Norman S, Vieira JM, Masters M et al (2015) Cardiac lymphatics are heterogeneous

in origin and respond to injury. Nature 522(7554):62–67 9. Henri O, Pouehe C, Houssari M, Galas L et al (2016) Selective stimulation of cardiac lymphangiogenesis reduces myocardial edema and fibrosis leading to improved cardiac function following myocardial infarction. Circulation 133(15):1484–1497 10. Nilsson JC, Nielsen G, Groenning BA et al (2001) Sustained postinfarction myocardial oedema in humans visualised by magnetic resonance imaging. Heart 85(6):639–642 11. Vieira JM, Norman S, Villa Del Campo C et al (2018) The cardiac lymphatic system stimulates resolution of inflammation following myocardial infarction. J Clin Invest 128(8): 3402–3412 12. Vivien CJ, Pichol-Thievend C, Boon Sim C, Smith JB et al (2019) Vegfc/d-dependent regulation of the lymphatic vasculature during cardiac regeneration is influenced by injury context. NPJ Regen Med 4:18 13. Lupu IE, De Val S, Smart N (2020) Coronary vessel formation in development and disease: mechanisms and insights for therapy. Nat Rev Cardiol 17(12):790–806 14. Marin-Juez R et al (2016) Fast revascularization of the injured area is essential to support zebrafish heart regeneration. Proc Natl Acad Sci U S A 113:11237–11242 15. Klaourakis K, Vieira JM, Riley PR (2021) The evolving cardiac lymphatic vasculature in development, repair and regeneration. Nat Rev Cardiol 18(5):368–379

Visualizing Vasculature in the Adult Zebrafish Heart 16. Susaki EA, Tainaka K, Perrin D et al (2014) Whole-brain imaging with single-cell resolution using chemical cocktails and computational analysis. Cell 157(3):726–739 17. Kubota SI, Takahashi K, Nishida J et al (2017) Whole-body profiling of cancer metastasis with single-cell resolution. Cell Rep 20(1):236–250 18. Tainaka K, Murakami TC, Susaki EA et al (2018) Chemical landscape for tissue clearing based on hydrophilic reagents. Cell Rep 24(8): 2196–2210.e9 19. Tian T, Yang Z, Li X (2021) Tissue clearing technique: recent progress and biomedical applications. J Anat 238(2):489–507 ´ et al (2017) High20. Fre´taud M, Rivie`re L, Job E resolution 3D imaging of whole organ after clearing: taking a new look at the zebrafish testis. Sci Rep 7:43012

323

21. Tainaka K, Kubota SI, Suyama TQ et al (2014) Whole-body imaging with single-cell resolution by tissue decolorization. Cell 159(4): 911–924 22. Bertrand JY, Chi NC, Santoso B, Teng S, Stainier DY, Traver D (2010) Haematopoietic stem cells derive directly from aortic endothelium during development. Nature 464:108– 111 23. Lawson ND, Weinstein BM (2002) In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev Biol 248:307– 318 24. Susaki EA, Shimizu C, Kuno A et al (2020) Versatile whole-organ/body staining and imaging based on electrolyte-gel properties of biological tissues. Nat Commun 11(1):1982

Chapter 24 Fluorescence-Activated Cell Sorting and Quantitative Real-Time PCR to Reveal VEGF-Expressing Macrophage Populations in the Zebrafish Larvae Andrew Herman, Alexander Greenhough, and David B. Gurevich Abstract The transparent, genetically tractable zebrafish is increasingly recognized as a useful model to both live image and uncover mechanistic insight into cell interactions governing tissue homeostasis, pathology, and regeneration. Here, we describe a protocol for the isolation of macrophages from zebrafish wounds using fluorescence-activated cell sorting (FACS), and the identification of specific pro-angiogenic macrophage populations that express high levels of vascular endothelial growth factor (vegf) using quantitative real-time PCR (qPCR). The cell dissociation and FACS sorting techniques have been optimized for immune cells and successfully used to isolate other fluorescently marked populations within the wound such as neutrophils and endothelial cells. More broadly, this protocol can be easily adapted to other contexts where identification of pro-angiogenic immune cells is transformative for understanding, from development to pathologies such as infection, cancer, and diabetes. Key words Zebrafish, Wounding, VEGF, Immune cells, Macrophages, FACS, Quantitative PCR

1

Introduction Wound repair and regeneration is a complex and highly integrated process, dependent on interactions and collaborations between many cell types in order to restore damaged tissue [1]. Innate immune cells play significant roles throughout wound healing, and macrophages in particular are vital for functions such as protecting tissue against infection, controlling the tissue inflammatory status, and co-ordinating the activity of other cell types such as endothelial cells to drive re-vascularization [2, 3]. To facilitate this broad spectrum of functions, macrophages maintain plasticity that allows them to select and switch between differing phenotypic “states” throughout tissue repair [4, 5]. Isolation of macrophages at specific timepoints during the wound healing process represents a powerful approach to identify the expression profile and changing

Lorna R. Fiedler and Caroline Pellet-Many (eds.), VEGF Signaling: Methods and Protocols, Methods in Molecular Biology, vol. 2475, https://doi.org/10.1007/978-1-0716-2217-9_24, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

325

326

Andrew Herman et al.

functionality of these phenotypic states, to better understand how these cells interact with the surrounding tissue during repair. The zebrafish is a tractable and highly versatile vertebrate model of tissue repair, possessing essentially all components of mammalian tissue but with the added advantage of being translucent and genetically tractable [6, 7]. This combination of strengths makes the zebrafish ideal for elucidating mechanistic insight by non-invasively observing tissue repair processes such as inflammation and angiogenesis. Recently developed transgenic reporter lines allow for the visualization of macrophages [8–10] and their identification as pro- or anti-inflammatory cells [11–13]. These transgenic lines provide the key tools necessary for isolating macrophages based on fluorescence profile to unravel how their function changes between phenotypic states and throughout tissue repair, particularly with regard to the control over angiogenesis. In this chapter, we provide a detailed description of how to induce wounds in zebrafish double transgenic Tg(mpeg1: mCherry); TgBAC(tnfa:GFP) larvae [9, 12], which mark macrophages in red and pro-inflammatory cells in green. Next, we describe how to dissociate larval tissue in a manner optimized for immune cells, and how to best FACS sort fluorescently labeled proand anti-inflammatory macrophage phenotypes from the resultant single cell suspension. We also explain how to use these cells for RNA extraction and qPCR, with the specific example of identifying which macrophage phenotype corresponds to the Vegf-expressing, pro-angiogenic population. The protocol described here can be utilized for numerous other contexts where identification of proand anti-angiogenic immune cell populations is vital, such as zebrafish models of infectious disease (reviewed [14]), cancer (reviewed [15]), and pathologies of impaired healing such as diabetes (reviewed [16]). With this protocol, we have also succeeded in isolating neutrophils and endothelial cells from cell sorting of dissected wounds taken from respectively Tg(mpx:GFP) [17] and Tg( fli1:GFP) [18] transgenic larval fish, which respectively label neutrophils or endothelial cells in green (data not shown). We subsequently used the qPCR approaches described here on these cells, providing valuable mechanistic insight into how macrophages, neutrophils, and endothelial cells interact with each other during tissue repair [19]; however, numerous other downstream applications can be performed on these purified cell populations, e.g., RNAseq analyses [20] and proteomics [21].

2

Materials

2.1 Needle Stab Injury

1. Wild-type larval fish (control), as well as larvae from the transgenic reporter line Tg(mpeg1:mCherry); TgBAC(tnfa:GFP), 4 days post fertilization (dpf) (see Note 1).

FACS Sorting of Pro-Inflammatory, VEGF Expressing Macrophages

327

2. Incubator set to 28.5
C. 3. Stereomicroscope. 4. Microscope slide. 5. 30-gauge needles. 6. Plastic Pasteur pipettes. 7. 90 mm plastic petri dishes. 8. 0.4% 3-amino benzoic acidethylester (tricaine) stock solution. Dissolve 400 mg tricaine with 97.9 mL ddH2O. Adjust to pH 7.2–7.4 with Tris-HCl (pH 9). Store solution at 4
C. 9. E3 buffer (embryo water). Prepare a 100 stock in advance. 14.61 g NaCl, 0.63 g KCl, 2.43 g mM CaCl2, and 1.99 g MgSO4, mix well in 1 L of deionized water. Can be stored at room temperature. Prepare 1 working solution prior to fish breeding and egg storage. 2.2 Cell Dissociation and FACS

1. Dissociation solution. Prepare a stock solution of 1 mg/mL of Collagenase type II, dissolve in ddH2O and store at 20
C. Immediately prior to use, thaw collagenase stock and prepare (20 mg/mL) Collagenase solution in 0.05% Trypsin-EDTA (also stored at 20
C) (see Note 2). 2. Stop/resuspension solution. Prepare Hank’s balanced salt solution (HBSS) (Ca2+, Mg2+ free), supplemented with 2.5% fetal bovine serum (FBS). Store solution at 4
C. 3. Scalpel blades. 4. Fine tipped forceps. 5. Microcentrifuge tubes, 1.5 mL. 6. Refrigerated tabletop microcentrifuge. 7. Water bath set to 32
C. 8. Pipettes and filtered tips (P1000, P200, P10). 9. Ice. 10. Sterile disposable 40 μm cell strainers (adaptable to 50 mL Falcon tubes). 11. 50 mL Falcon tubes. 12. Becton Dickinson InFlux cell sorter. 13. Sterile Falcon 5 mL Polypropylene test tubes (Catalog number 352063) for quality control of particles and for zebrafish samples. 14. Spherotech Calibration 8 peak beads (3–3.4 μm) for laser alignment and QC (Catalog number RCP-30-5A 3). 15. BD Biosciences Accudrop fluorescent beads 6 (μm diameter) for sort drop delay calculation (Catalog number 345249).

328

Andrew Herman et al.

Table 1 Forward and reverse primer sequences for vascular growth factor aa (vegfaa) and elongation factor 1a (ef1a, used as reference gene) Gene name

Forward primer 50 -30

Reverse primer 50 -30

Elongation factor 1α (ef1α)

CTTCTCAGGCTGACTG TGC

CCGCTAGCATTACCC TCC

Vascular endothelial growth factor aa (vegfaa)

AAAAGAGTGCG TGCAAGACC

AGCACCTCCATAG TGACGTT

16. Cell viability dye, e.g., Propidium Iodide (PI). Prepare 400 stock of PI using Sigma Aldrich 1.0 mg/mL in H2O, and use at 2.5 μg/mL final concentration (see Note 3). 17. Milty Zerostat anti-static remover gun (see Note 4). 2.3 RNA/cDNA Prep and qPCR

1. RNase-free workstation. 2. RNase-eliminating solution (such as RNaseZAP). 3. RNase-free filtered pipette tips and microcentrifuge tubes. 4. RNeasy Micro kit (Qiagen), with DNase I. 5. 100% ethanol, ACS grade. 6. Nuclease- and DNA-free water. 7. Refrigerated microcentrifuge capable of at least 10,000  g. 8. Ice. 9. Nanodrop ND-1000 spectrophotometer (Thermo Scientific) to assess RNA concentration and purity. 10. cDNA Synthesis Kit (e.g., Thermo Fisher Maxima First Strand cDNA Synthesis Kit). 11. PCR and qPCR machine (e.g., Agilent MX3005P QPCR cycler). 12. qPCR Primers with a Tm of 56
C (Table 1). 13. Optical 96-well reaction plates with transparent sealing. 14. Real-time PCR kit (e.g., Qiagen QuantiTect SYBR Green PCR kit). 15. 1% agarose gel (stained with ethidium bromide at 0.5 μg/mL).

3

Methods

3.1 Needle Stab Injury

1. Store eggs and larvae in 1 working solution of embryo water. At 4dpf, remove larvae from the incubator and anesthetize by adding tricaine solution (approx. 1 mL per 30 mL embryo water, see Note 5).

FACS Sorting of Pro-Inflammatory, VEGF Expressing Macrophages

329

2. Transfer individual larvae in a drop of liquid onto a microscope slide under a stereomicroscope using a plastic Pasteur pipette (see Note 6). 3. Maneuver larvae into appropriate lateral lying position and remove excess liquid using the plastic pipette (see Note 7). 4. Perform a needle stick injury into the dorsal somites opposite the cloaca using the 30-gauge needle held at a 75
angle (see Note 8 and Fig. 1). 5. Immediately transfer the injured larvae into a petri dish containing clean embryo water. Once injured larvae are observed to be moving, place back into incubator to fully recover (see Note 9). 3.2

Cell Dissociation

1. Remove appropriately staged injured larvae from the incubator and anesthetize by adding tricaine solution (approx. 1 mL per 15 mL embryo water). 2. Transfer anesthetized larvae in a drop of the tricaine containing solution onto a microscope slide under a stereomicroscope using a plastic Pasteur pipette. Load up to 10 larvae in this manner. 3. Make two incisions on the trunk of each larval using a scalpel blade, dissecting the wound away from the head and tail (see Note 10). 4. Transfer all wounded tissue of interest into a 1.5 mL microcentrifuge tube containing 300 μL of chilled dissociation solution using fine tipped forceps. Samples should be kept chilled on ice until tissue collection is completed (normally 50 wounds per sample). 5. Incubate in a water bath at 32
C for approx. 20 min, gently pipetting up and down every 3–5 min using a P200 pipette to promote tissue disruption and dissociation. 6. While incubation is in progress, chill resuspension solution on ice. Add 4 mL of resuspension solution to a 50 mL Falcon tube. 7. Once cells are fully dissociated, stop the reaction by adding dissociated cells to the 50 mL Falcon tube containing 4 mL of resuspension solution. Rinse the 1.5 mL microcentrifuge tube with another 1 mL of chilled resuspension solution and reunite with the remainder of cells. 8. Gently pipette the homogenized material 15–20 times against the bottom of the 50 mL Falcon tube to minimize cell clumping. 9. Centrifuge the dissociated cells at 300  g for 10 min at 4
C. 10. Discard the supernatant and resuspend the pelleted cells in 4 mL resuspension solution, on ice.

330

Andrew Herman et al.

Fig. 1 Macrophages expressing mpeg1 (red), together with tnfα-expressing pro-inflammatory cells (green), accumulate at the sight of needle stab injury. (a) Schematic with boxed area showing location of needle stab injury at the dorsal somites above the cloaca. This area is subsequently excised to isolate the injury region for cell sorting. (b, c) Maximum intensity projection through a representative fluorescent Z-stack of a laterally mounted 5 dpf larvae, uninjured or at 1 day post injury, taken using a confocal microscope. An enrichment of macrophages is seen in the injury site compared to the ventral side of the fish, with some of these macrophages expressing the pro-inflammatory marker tnfα (yellow)

FACS Sorting of Pro-Inflammatory, VEGF Expressing Macrophages

331

11. Pass the 4 mL of resuspended cells through a 40 μm cell strainer placed into a clean 50 mL Falcon tube. Rinse the first 50 mL Falcon tube with another 1 mL of chilled resuspension solution and add this to the strainer. 12. Centrifuge the dissociated cells at 300  g for 10 min at 4
C. 13. Discard the majority of the supernatant, leaving approx. 500 μL of resuspension solution. 14. Resuspend cells in the remaining 500 μL of resuspension solution, resulting in a maximum concentration of 5–10 million cells/mL. 3.3

FACS Sorting

1. Prepare cell sorter as per manufacturers’ guidance (see Note 11). 2. Once the instrument has been calibrated, proceed to cell sorting, set up to FACS sort samples at 4
C using a 100 μm nozzle at 21 psi. 3. Add viability dye (e.g., Propidium Iodide in a 1:400 dilution) to the cell suspension to identify the live cell population. 4. Determine optimal excitation voltages and gating strategy using non-fluorescent (“no stain”) (see Note 12) and single fluorophore controls (see Note 13). 5. Use the above gating strategy to define the double positive population (see Fig. 2). 6. Collect single positive and double positive populations directly into 500 μL lysis buffer (RLT buffer from RNeasy Micro kit) in 1.5 mL eppendorf tubes.

3.4 RNA and cDNA Preparation

1. For RNA extraction, use RNeasy Micro kit and proceed according to manufacturer’s protocol (see Note 14). 2. During the procedure, perform on-column treatment with DNase I provided in the RNeasy Micro kit to remove DNA contaminants. 3. At the end of the procedure, elute RNA extracted from sorted fluorescent cells in 10 μL of RNase-free water. Transfer to a fresh Eppendorf tube and store at 80
C. 4. Prior to cDNA synthesis, measure RNA quality and quantity using a Nanodrop spectrophotometer. Normalize all samples to the same concentration by diluting with RNase-free water. 5. Using equal concentrations of each RNA sample, synthesize cDNA using the Maxima First Strand cDNA Synthesis Kit according to manufacturer’s protocol (see Notes 15 and 16). 6. Store synthesized cDNA (final volume of 20 μL) at 80
C.

332

Andrew Herman et al.

Fig. 2 Fluorescence-activated cell sorting of tnfα+ve and tnfαve macrophages at 24 h post needle stab injury. (a–d) Representative flow cytometry data showing gating strategy for sorting macrophages from unwounded control fish. (e-h) Representative flow cytometry data showing gating strategy for sorting macrophages from wounded fish, with wounds harvest at 24 h post injury. Cells are first gated for expected size (Forward and Side Scatter), singlets, and viability (Propidium Iodide). Live, single cells of expected size are subsequently gated on GFP and mCherry expression to identify tnfα and mpeg expression, respectively, to determine tnfα+ve and tnfαve macrophage populations. (i) Representative qRT-PCR showing equal levels of expression of housekeeping gene ef1α and differential levels of expression of vegfaa between tnfα+ve and tnfαve macrophage populations following 40 rounds of amplification

FACS Sorting of Pro-Inflammatory, VEGF Expressing Macrophages

3.5

qPCR

333

1. Dilute each cDNA sample 1:10 in nuclease-free water to obtain the final concentration used in the qRT-PCR reaction (see Note 17). 2. Prepare the SYBR Green mix for each condition (target and control samples at each of the different time points). For each condition, prepare three replicates for each cDNA sample, together with data normalization using expression of a reference gene (see Note 18). Calculate 7.5 μL SYBR Green master mix (2), 1.5 μL primer mix (100 nM final concentration for forward and reverse primers), and 4 μL nuclease-free water per well (see Note 19). Prepare the SYBR Green master mix in an Eppendorf tube for the number of wells required, plus an excess (two extra volumes) to allow for pipetting error. 3. Put 13 μL of SYBR Green mix in each well and then add 2 μL of cDNA. 4. Perform real-time PCR using the following qPCR program: (a) 95
C for 15 min (initial denaturation). (b) 94
C for 15 s; 56
C for 30 s; 72
C for 30 s ! 40 cycles. (c) Perform dissociation (melting curves) analysis using a final step of 95
C for 30 s; 56
C for 30 s; followed by gradual temperature increases to 95
C (approximately 10 min). (d) 4
C, hold. 5. Analyze the qPCR results (see Notes 20 and 21). 6. Run qPCR products on a 1% agarose gel (stained with ethidium bromide) to verify quality and specificity of the qPCR reactions (see Fig. 2i).

4

Notes 1. There are numerous published zebrafish transgenic reporter lines that mark macrophages, such as mpeg1 [9], mfap4 [8], and cfms [10]. Furthermore, numerous transgenic reporter lines exist that mark pro-inflammatory cells, such as il1β [11] and tnfα [12]. In this protocol, we have chosen the mpeg1 marker to identify wound macrophages, and overlaid the tnfα marker to help separate pro-inflammatory macrophages (which express both markers) from anti-inflammatory macrophages (which express mpeg1 only). Other markers could be used to segregate macrophage phenotypes, including markers of antiinflammatory macrophages such as spp1 [13]. 2. The composition of the dissociation solution is critical for maximizing the efficiency of cell dissociation process and cell survival. Through testing numerous combinations of reagents from different suppliers we have identified that Collagenase

334

Andrew Herman et al.

from Sigma (Catalog number C8176) and Trypsin-EDTA from Thermo Fisher (Catalog number MT25051CI) are the optimum reagents for wound tissue dissociation, especially for immune cells. 3. PI was used as a viability marker in these studies and dead cells were detected using the violet (405 nm) laser in the 610/20 BP detector (via the 600 nm LP filter). It is critical to use a viability dye to exclude dead or dying cells from the sorted populations and there are many choices, depending on the FACS configuration and the fluorescent markers used in the experiment. For FACS instruments without a violet laser, one could use DRAQ7 (Biostatus) as an alternative (detected with the red laser e.g. 640 nm, through the 750 nm LP filter). 4. Using this static gun on collection tubes prior to sorting improves yields during sorting, as it minimizes drops (containing cells) striking the tube wall instead of the reservoir of buffer. 5. This ratio of tricaine to embryo water has been optimized to anesthetize larvae at 4dpf. Early larvae are less sensitive and may require more tricaine (approximately 1:20), while older larvae are more sensitive and require less (approximately 1:50). Assessment of heartbeat and blood flow is critical throughout the wounding procedure—if either of these processes stop, fish must be immediately recovered in fresh embryo water until heartbeat and blood flow is restored. 6. Have all equipment necessary for wounding ready and easily accessible by this step, to allow for wounding experiments to be performed as rapidly as possible while maintaining reproducibility of injuries. 7. Minimizing the amount of water surrounding the larvae also minimizes the “bolus” within which the larvae may move or float in, making the stab procedure easier and more reproducible. However, it is important to maintain a thin layer of liquid on the fish, as allowing the larvae to dry out will result in much higher mortality—if the larvae appear to be drying out, immediately transfer to fresh embryo water. 8. It is critical to avoid injury to the notochord, as this will affect survival of injured larvae. To maintain a consistent 2-somite block of injury, hold the bevel facing the fish and perform the stab as a single wound. 9. When recovering the wounded larvae into clean embryo water, make sure these larvae “sink” to the bottom of the dish by gently pipetting embryo water onto them. Fish that remain floating at the water surface may dry out sufficiently to cause mortality. It should be noted that other injury approaches such as wounding using tungsten needles [22], tail fin amputation

FACS Sorting of Pro-Inflammatory, VEGF Expressing Macrophages

335

[23], or injuries using a laser microablation system [24, 25] can also be used in this pipeline. 10. In order to harvest similar amounts of macrophages for control unwounded fish, perform a single cut as per wounded fish at the anterior most point (near the cloaca) and collect the entire tail region. Perform all downstream steps as per wounds, with an increase in agitation to assist in tissue dissociation. 11. For fluidics and laser stabilization, choose appropriate sheath pressure and drop drive frequency, optimize and align lasers and calculate sort drop delay. Use Spherotech Rainbow Calibration 8 peak beads for laser alignment and QC, BD Biosciences Accudrop fluorescent beads for drop delay calculation. Optimize sort stream deflection for 1.5 mL Eppendorf tubes. Clean sample line with 70% IMS (5 min wash at high flow rate). Set chiller to 4
C (sample port and collection tube holder). “Sterilize” the cell sort chamber w/UV light treatment for 30 min. Sort mode set to 1.0 drop purity for these studies. Based on drop drive frequency: ensured that event rate was