Axon Degeneration: Methods and Protocols [1st ed.] 9781071605844, 9781071605851

Table of contents :
Front Matter ....Pages i-xi
Front Matter ....Pages 1-1
Axon Degeneration: Which Method to Choose? (Michael P. Coleman)....Pages 3-12
Front Matter ....Pages 13-13
Axon Degeneration Assays in Superior Cervical Ganglion Explant Cultures (Andrea Loreto, Jonathan Gilley)....Pages 15-24
Microinjection of Superior Cervical Ganglion Neurons for Studying Axon Degeneration (Jonathan Gilley, Andrea Loreto)....Pages 25-39
Assessing Axonal Degeneration in Embryonic Dorsal Root Ganglion Neurons In Vitro (Jung Eun Shin, Yongcheol Cho)....Pages 41-54
Viral Transduction of DRG Neurons (Yo Sasaki)....Pages 55-62
Planning and Analysis of Axon Degeneration Screening Experiments (Lyndah Lovell, John Bramley, William Buchser)....Pages 63-82
A Microfluidic Culture Platform to Assess Axon Degeneration (Yu Yong, Christopher Hughes, Christopher Deppmann)....Pages 83-96
A Schwann Cell–Neuron Coculture System to Study Neuron–Glia Interaction During Axon Degeneration (Elisabetta Babetto)....Pages 97-110
Establishing Myelinating Cocultures Using Human iPSC-Derived Sensory Neurons to Investigate Axonal Degeneration and Demyelination (Alex J. Clark)....Pages 111-129
Front Matter ....Pages 131-131
Preparation of Organotypic Hippocampal Slice Cultures for the Study of CNS Disease and Damage (Claire S. Durrant)....Pages 133-144
Organotypic Culture Assay for Neuromuscular Synaptic Degeneration and Function (Kosala N. Dissanayake, Robert Chang-Chih Chou, Rosalind Brown, Richard R. Ribchester)....Pages 145-157
Intracellular Recordings of Postsynaptic Voltage Responses at the Drosophila Neuromuscular Junction (Vera Valakh, Johanna G. Flyer-Adams)....Pages 159-168
Ex Vivo Studies of Optic Nerve Axon Electrophysiology (Chinthasagar Bastian, Sylvain Brunet, Selva Baltan)....Pages 169-177
Ex Vivo Analysis of Axonal Degeneration Using Sciatic and Optic Nerve Preparations (Rodrigo López-Leal, Felipe A. Court)....Pages 179-189
Measuring Bioenergetic Signatures of Peripheral Nerve Segments by Extracellular Flux Analysis (Bogdan Beirowski)....Pages 191-203
Front Matter ....Pages 205-205
Detection of Neutrophils in the Sciatic Nerve Following Peripheral Nerve Injury (Jon P. Niemi, Jane A. Lindborg, Richard E. Zigmond)....Pages 207-222
Whole-Tissue Immunolabeling and 3D Fluorescence Imaging to Visualize Axon Degeneration in the Intact, Unsectioned Mouse Tissues (Ying Cao, Jing Yang)....Pages 223-232
Transmission Electron Microscopy and Morphometry of the CNS White Matter (Julia M. Edgar, Rebecca Sherrard Smith, Ian D. Duncan)....Pages 233-261
In Vivo Calcium Imaging During Axon Degeneration in Zebrafish (Mauricio E. Vargas)....Pages 263-270
In Vivo Imaging of Anterograde and Retrograde Axonal Transport in Rodent Peripheral Nerves (James N. Sleigh, Andrew P. Tosolini, Giampietro Schiavo)....Pages 271-292
In Vivo Visualization of Moving Synaptic Cargo Complexes within Drosophila Larval Segmental Axons (Rupkatha Banerjee, Joseph A. White II, Shermali Gunawardena)....Pages 293-300
A Chemotherapy-Induced Peripheral Neuropathy Model in Drosophila melanogaster (Martha R. C. Bhattacharya)....Pages 301-310
Models of Axon Degeneration in Drosophila Larvae (E. J. Brace, Aaron DiAntonio)....Pages 311-320
In Vivo Analysis of Glial Immune Responses to Axon Degeneration in Drosophila melanogaster (Mary A. Logan, Sean D. Speese)....Pages 321-338
Back Matter ....Pages 339-340

Citation preview

Methods in Molecular Biology 2143

Elisabetta Babetto Editor

Axon Degeneration Methods and Protocols

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.

Axon Degeneration Methods and Protocols

Edited by

Elisabetta Babetto Department of Biochemistry, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA; Department of Pharmacology and Toxicology, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA; Hunter James Kelly Research Institute, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA

Editor Elisabetta Babetto Department of Biochemistry Jacobs School of Medicine and Biomedical Sciences University at Buffalo Buffalo, NY, USA Department of Pharmacology and Toxicology Jacobs School of Medicine and Biomedical Sciences University at Buffalo Buffalo, NY, USA Hunter James Kelly Research Institute Jacobs School of Medicine and Biomedical Sciences University at Buffalo Buffalo, NY, USA

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-0584-4 ISBN 978-1-0716-0585-1 (eBook) https://doi.org/10.1007/978-1-0716-0585-1 © Springer Science+Business Media, LLC, part of Springer Nature 2020 Chapters 2, 3 and 20 are licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). For further details see license information in the chapters. This work is subject to copyright. All rights are reserved 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. 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 The study of axon degeneration is as fascinating as it is challenging. First and foremost, axons are extremely fragile structures because of their slender shape and extraordinary length, sometimes spanning half of our human body. Especially in the peripheral nervous system, long axons undergo constant mechanical stresses due to movement that results in tension and compression forces. Moreover, they are often physically inaccessible, particularly in the central nervous system, where they are surrounded by bones, connective tissue, and membranes. Finally, a high level of compartmentalization separates the biological responses of the axon from the other portions of the neuron (cell body, dendrites, and synapse), yet all the structures belong to the same cell, the neuron. Indeed, as axons and neurons are very complex structures, with multifaceted functions, it is understandable that the prospect of modifying their genome, injecting dyes, or simply injuring them in a selective manner seems daunting, intimidating, or even unachievable. Yet, they are the fundamental units that functionally connect different regions of our bodies. Ambulation can occur only if the brain and muscles communicate via signals propagated along axons. Likewise, temperature, touch, and pain can be “sensed” only if the skin is innervated and sensory stimuli are communicated to the central nervous system via axons. Because of their elongated and thin shape, axons are extremely vulnerable to insults and are therefore affected in a myriad of acute and chronic conditions. Their injury is the cause of many symptoms of such conditions. Moreover, their close proximity and dependence on glia makes them additionally vulnerable to pathological circumstances in which glial biology is altered. The long and thin structures crossing several regions of our body can be studied from different point of views, structurally and functionally. Experimental mechanical axon transection is a popular method, because it produces stereotypical molecular changes and it is temporally defined. Several protocols in this book describe how to study these changes and characterize the events occurring during these initial steps of axon degeneration. As the degeneration process progresses, culminating in structural axon breakdown, other correlated glial responses occur that together define the process of Wallerian degeneration. The book comprises methods to study these changes, as well as alteration of the myelin structure. Additionally, changes to the most distal nerve terminal and to the neuromuscular junction are discussed in detail. In sum, this book is a collection of classical as well as innovative methods used to investigate axon degeneration with a particular focus on addressing the common challenges encountered while performing these procedures. Particular attention is devoted to the study of axon loss in several model organisms, as each poses unique challenges and provides powerful advantages. The result is a coordinated effort of multiple excellent authors, leaders in the field, describing their routinely used techniques, with the hope of facilitating the application and further development of these protocols, which will help the scientific community tackle important questions regarding axon degeneration. Buffalo, NY, USA

Elisabetta Babetto

v

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

PART I

INTRODUCTION

1 Axon Degeneration: Which Method to Choose? . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael P. Coleman

PART II

v ix

3

IN VITRO ASSAYS

2 Axon Degeneration Assays in Superior Cervical Ganglion Explant Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Andrea Loreto and Jonathan Gilley 3 Microinjection of Superior Cervical Ganglion Neurons for Studying Axon Degeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Jonathan Gilley and Andrea Loreto 4 Assessing Axonal Degeneration in Embryonic Dorsal Root Ganglion Neurons In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Jung Eun Shin and Yongcheol Cho 5 Viral Transduction of DRG Neurons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Yo Sasaki 6 Planning and Analysis of Axon Degeneration Screening Experiments . . . . . . . . . . 63 Lyndah Lovell, John Bramley, and William Buchser 7 A Microfluidic Culture Platform to Assess Axon Degeneration . . . . . . . . . . . . . . . 83 Yu Yong, Christopher Hughes, and Christopher Deppmann 8 A Schwann Cell–Neuron Coculture System to Study Neuron–Glia Interaction During Axon Degeneration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Elisabetta Babetto 9 Establishing Myelinating Cocultures Using Human iPSC-Derived Sensory Neurons to Investigate Axonal Degeneration and Demyelination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Alex J. Clark

PART III 10

11

EX VIVO ASSAYS

Preparation of Organotypic Hippocampal Slice Cultures for the Study of CNS Disease and Damage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Claire S. Durrant Organotypic Culture Assay for Neuromuscular Synaptic Degeneration and Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Kosala N. Dissanayake, Robert Chang-Chih Chou, Rosalind Brown, and Richard R. Ribchester

vii

viii

12

13 14

15

Contents

Intracellular Recordings of Postsynaptic Voltage Responses at the Drosophila Neuromuscular Junction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vera Valakh and Johanna G. Flyer-Adams Ex Vivo Studies of Optic Nerve Axon Electrophysiology . . . . . . . . . . . . . . . . . . . . Chinthasagar Bastian, Sylvain Brunet, and Selva Baltan Ex Vivo Analysis of Axonal Degeneration Using Sciatic and Optic Nerve Preparations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rodrigo Lo pez-Leal and Felipe A. Court Measuring Bioenergetic Signatures of Peripheral Nerve Segments by Extracellular Flux Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bogdan Beirowski

PART IV 16

17

18

19

20

21

22

23 24

159 169

179

191

IN VIVO ASSAYS

Detection of Neutrophils in the Sciatic Nerve Following Peripheral Nerve Injury. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jon P. Niemi, Jane A. Lindborg, and Richard E. Zigmond Whole-Tissue Immunolabeling and 3D Fluorescence Imaging to Visualize Axon Degeneration in the Intact, Unsectioned Mouse Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ying Cao and Jing Yang Transmission Electron Microscopy and Morphometry of the CNS White Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Julia M. Edgar, Rebecca Sherrard Smith, and Ian D. Duncan In Vivo Calcium Imaging During Axon Degeneration in Zebrafish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mauricio E. Vargas In Vivo Imaging of Anterograde and Retrograde Axonal Transport in Rodent Peripheral Nerves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James N. Sleigh, Andrew P. Tosolini, and Giampietro Schiavo In Vivo Visualization of Moving Synaptic Cargo Complexes within Drosophila Larval Segmental Axons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rupkatha Banerjee, Joseph A. White II, and Shermali Gunawardena A Chemotherapy-Induced Peripheral Neuropathy Model in Drosophila melanogaster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martha R. C. Bhattacharya Models of Axon Degeneration in Drosophila Larvae . . . . . . . . . . . . . . . . . . . . . . . . . E. J. Brace and Aaron DiAntonio In Vivo Analysis of Glial Immune Responses to Axon Degeneration in Drosophila melanogaster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mary A. Logan and Sean D. Speese

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

207

223

233

263

271

293

301 311

321 339

Contributors ELISABETTA BABETTO • Department of Biochemistry, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA; Department of Pharmacology and Toxicology, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA; Hunter James Kelly Research Institute, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA SELVA BALTAN • Department of Neurosciences, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH, USA; Anesthesia and Perioperative Medicine (APOM), Oregon Health and Science University, Portland, OR, USA RUPKATHA BANERJEE • Department of Biological Sciences, The State University of New York at Buffalo, Buffalo, NY, USA CHINTHASAGAR BASTIAN • Department of Neurosciences, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH, USA BOGDAN BEIROWSKI • Hunter James Kelly Research Institute, New York State Center of Excellence in Bioinformatics and Life Sciences (CBLS) RM B4-314, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA; Department of Biochemistry, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, Buffalo, NY, USA MARTHA R. C. BHATTACHARYA • Department of Neuroscience, University of Arizona, Tucson, AZ, USA E. J. BRACE • Department of Developmental Biology, Washington University School of Medicine, St. Louis, MO, USA JOHN BRAMLEY • Department of Genetics, Washington University in St. Louis, Greater St. Louis, MO, USA ROSALIND BROWN • Centre for Discovery Brain Sciences and the Euan MacDonald Centre for Motor Neurone Disease Research, University of Edinburgh, Edinburgh, UK SYLVAIN BRUNET • Department of Neurosciences, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH, USA WILLIAM BUCHSER • Department of Genetics, Washington University in St. Louis, Greater St. Louis, MO, USA YING CAO • Center for Life Sciences, Peking University, Beijing, China; Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, China ROBERT CHANG-CHIH CHOU • Centre for Discovery Brain Sciences and the Euan MacDonald Centre for Motor Neurone Disease Research, University of Edinburgh, Edinburgh, UK YONGCHEOL CHO • Division of Life Sciences, Korea University, Seoul, Republic of Korea ALEX J. CLARK • Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, UK MICHAEL P. COLEMAN • Department of Clinical Neurosciences, John van Geest Centre for Brain Repair, University of Cambridge, Cambridge, UK FELIPE A. COURT • Center for Integrative Biology, Faculty of Sciences, Universidad Mayor, Santiago, Chile; Fondap Geroscience for Brain in Health and Metabolism, Santiago, Chile CHRISTOPHER DEPPMANN • Department of Biology, University of Virginia, Charlottesville, VA, USA

ix

x

Contributors

AARON DIANTONIO • Department of Developmental Biology, Washington University School of Medicine, St. Louis, MO, USA KOSALA N. DISSANAYAKE • Centre for Discovery Brain Sciences and the Euan MacDonald Centre for Motor Neurone Disease Research, University of Edinburgh, Edinburgh, UK IAN D. DUNCAN • Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin-Madison, Madison, WI, USA CLAIRE S. DURRANT • Centre for Discovery Brain Sciences, The University of Edinburgh, Edinburgh, Scotland, UK JULIA M. EDGAR • Institute of Infection, Immunity and Inflammation, College of Medical Veterinary and Life Sciences, University of Glasgow, Glasgow, UK; Department of Neurogenetics, Max Planck Institute for Experimental Medicine, Goettingen, Germany JOHANNA G. FLYER-ADAMS • Department of Biology, Brandeis University, Waltham, MA, USA JONATHAN GILLEY • Department of Clinical Neurosciences, John van Geest Centre for Brain Repair, University of Cambridge, Cambridge, UK SHERMALI GUNAWARDENA • Department of Biological Sciences, The State University of New York at Buffalo, Buffalo, NY, USA CHRISTOPHER HUGHES • Department of Physics and Astronomy, James Madison University, Harrisonburg, VA, USA JANE A. LINDBORG • Department of Neurosciences, Case Western Reserve University, Cleveland, OH, USA; Department of Neurology, Yale University, New Haven, CT, USA MARY A. LOGAN • Department of Neurology, Jungers Center for Neurosciences Research, Oregon Health and Science University, Portland, OR, USA RODRIGO LO´PEZ-LEAL • Faculty of Sciences, Center for Integrative Biology, Universidad Mayor, Santiago, Chile; Fondap Geroscience for Brain in Health and Metabolism, Santiago, Chile ANDREA LORETO • Department of Clinical Neurosciences, John van Geest Centre for Brain Repair, University of Cambridge, Cambridge, UK LYNDAH LOVELL • Drug Discovery Program, Johns Hopkins University, Baltimore, MD, USA JON P. NIEMI • Department of Neurosciences, Case Western Reserve University, Cleveland, OH, USA RICHARD R. RIBCHESTER • Centre for Discovery Brain Sciences and the Euan MacDonald Centre for Motor Neurone Disease Research, University of Edinburgh, Edinburgh, UK YO SASAKI • Department of Genetics, Washington University in St. Louis, Couch Biomedical Research Building, St. Louis, MO, USA GIAMPIETRO SCHIAVO • Department of Neuromuscular Diseases, UCL Queen Square Institute of Neurology, University College London, London, UK; UK Dementia Research Institute, University College London, London, UK; Discoveries Centre for Regenerative and Precision Medicine, University College London, London, UK JUNG EUN SHIN • Division of Life Sciences, Korea University, Seoul, Republic of Korea JAMES N. SLEIGH • Department of Neuromuscular Diseases, UCL Queen Square Institute of Neurology, University College London, London, UK; UK Dementia Research Institute, University College London, London, UK REBECCA SHERRARD SMITH • Institute of Infection, Immunity and Inflammation, College of Medical Veterinary and Life Sciences, University of Glasgow, Glasgow, UK SEAN D. SPEESE • Jungers Center for Neurosciences Research, Department of Neurology, Oregon Health and Science University, Portland, OR, USA

Contributors

xi

ANDREW P. TOSOLINI • Department of Neuromuscular Diseases, UCL Queen Square Institute of Neurology, University College London, London, UK VERA VALAKH • Department of Biology, Brandeis University, Waltham, MA, USA MAURICIO E. VARGAS • Casey Eye Institute and the Department of Ophthalmology, Oregon Health and Sciences University, Portland, OR, USA JOSEPH A. WHITE II • Department of Biological Sciences, The State University of New York at Buffalo, Buffalo, NY, USA JING YANG • Center for Life Sciences, Peking University, Beijing, China; IDG/McGovern Institute for Brain Research, Peking University, Beijing, China; School of Life Sciences, Peking University, Beijing, China YU YONG • Department of Biology, University of Virginia, Charlottesville, VA, USA RICHARD E. ZIGMOND • Department of Neurosciences, Case Western Reserve University, Cleveland, OH, USA

Part I Introduction

Chapter 1 Axon Degeneration: Which Method to Choose? Michael P. Coleman Abstract Axons are diverse. They have different lengths, different branching patterns, and different biological roles. Methods to study axon degeneration are also diverse. The result is a bewildering range of experimental systems in which to study mechanisms of axon degeneration, and it is difficult to extrapolate from one neuron type and one method to another. The purpose of this chapter is to help readers to do this and to choose the methods most appropriate for answering their particular research question. Key words Primary neuronal culture, Nerve lesion, Explant culture, Drosophila, Axonal transport

1

Introduction The importance of understanding mechanisms of axon degeneration is beyond doubt. Axons die first in most neurodegenerative disorders, neurons cannot function without them, and often the soma will also die without retrograde trophic support. Preserving axons is also an achievable goal. Many recent advances have been made in blocking Wallerian degeneration [1] in boosting axonal transport [2] and in preventing axonal disorders, for example, using antisense oligonucleotides [3]. But how do we advance these and related fields further? Choosing the right experimental system for each research question and understanding both its uses and limitations are a vital part of the answer. While some degeneration mechanisms appear to operate in all axons, many neurodegenerative diseases involve a specific subset of axons and neurons. Unfortunately, some of these neuron types are extraordinarily difficult to isolate and maintain in primary neuronal cultures, and even if this could be overcome, neuron cultures lack the long axons found in vivo. For example, it is not feasible to isolate a pure population of nigrostriatal dopaminergic neurons for primary culture, and even midbrain cultures enriched in these neurons do not come close to replicating their huge terminal branches in the striatum [4]. Does this mean we should only

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_1, © Springer Science+Business Media, LLC, part of Springer Nature 2020

3

4

Michael P. Coleman

work in vivo? Can we learn anything about them from other neuron types or from iPS-derived neurons in culture? How about other organisms such as Drosophila, where the nervous system is very different from ours but experimental methods are so much more powerful? In brief, when can we afford to take a reductionist view and use any axonal system to answer a question, and when does it have to be the right neuron type in the right context? The chapter references below illustrate where the respective methods can usefully contribute to key areas of axon degeneration research.

2

How Reductionist Can We Afford to Be? Some properties are shared by axons in all, or almost all, circumstances. All axons, as far as we know, carry out bidirectional axonal transport (Chapters 20 and 21) driven by kinesins and dynein along dynamic microtubules. There may be some circumstances where transport is less essential, for example, axons several hundred microns in second instar Drosophila larvae can survive without mitochondria. Diffusion of the relevant low molecular weight metabolites (ATP, ADP) to and from the cell body may be sufficient to sustain these short axons without the capacity to metabolize them directly. However, above a threshold length, axonal transport appears to be essential for axons to survive. If axonal transport is prevented by injury, genetic mutation, toxins, ischemia, or inflammatory mechanisms, this results in activation of the Wallerian degeneration mechanism [1] (Chapters 4, 20–23). Failure to deliver NMNAT2 appears to be the initiating event since substituting for NMNAT2 with more stable NMNATs, including the slow Wallerian degeneration protein WLDS, greatly delays axon degeneration in injury and many noninjury models [5]. The mechanism of Wallerian degeneration also appears to be conserved between different neuron types and widely divergent species. WLDS, other NMNATs, or deletion of the downstream prodegenerative protein SARM1 or its orthologues slows Wallerian degeneration in every mouse neuron type studied, including sensory (Chapters 4–8), motor (Chapter 11), sympathetic (Chapters 2 and 3), and hippocampal neurons (Chapter 10) and cerebellar granule cells in vivo and/or in vivo. It also protects axons in neurons from all species studied with the exception of Caenorhabditis elegans [6]. These include mouse, rat, Drosophila melanogaster (Chapters 12, 21–24), zebrafish (Chapter 19), and human neurons (Chapter 9) [7–11]. Thus, Wallerian degeneration is a particularly ancient mechanism of axon degeneration that can be studied in any of these experimental systems. However, not all noninjury degenerative mechanisms proceed in this way (see Subheading 5). Other common features include the presence of, and presumably long-term reliance on, mitochondria (Chapter 15), apart from

Axon Degeneration: Which Method to Choose?

5

the short axons of the milton larvae [12]; the association in vivo with support cells (Chapters 8 and 24), which can be partly bypassed in vitro by supplying the growth factors and nutrients they are normally needed for (Chapters 2, 4, 5); and probably the presence of endoplasmic reticulum (ER) with a crucial role in calcium buffering and perhaps local protein synthesis. Each of these has a key role in axon survival and degeneration, so we can learn something about the underlying mechanisms from widely different experimental systems. However, along with the advantages of doing this, there are unavoidable caveats that should always be remembered. Axon length is one vital consideration. While it appears that local synthesis of proteins is not required for axon survival in primary culture, where axons are a few mm long [5], this does not tell us about its role in a human foot 1 m from the cell body, or even in a mouse brainstem at a few cm, where the axonal transport challenge is much greater (Chapter 20). Extrapolation from in vitro data without considering this could be highly misleading. Another example would be dopaminergic neurons in midbrain primary cultures, or iPS-derived dopaminergic neuron cultures (Chapter 9), from which we can learn a great deal but whose already impressive axonal arbors are still orders of magnitude less complex than those in vivo. It is important not to underestimate the greater demand placed on the soma and axonal transport to support those arbors, which becomes overwhelmed for example in Parkinson’s disease. While it is essential to use these methods for many questions, this difference always needs to be in mind. Developmental stage is important too. Neuronal cultures survive and grow best from early postnatal (Chapters 2, 3, 10) or even embryonic tissue (Chapter 4), but many of the axon degeneration disorders we want to model are late-onset adult conditions. During ageing, axonal transport declines [13] and some axonal arbors expand even more to reinnervate targets vacated by adjacent neurons that have died [14], so the challenges of maintaining an old axon in vivo are far greater than those for a young axon in vitro. It should also be remembered that establishing a primary neuronal culture (Chapters 2–8) necessarily involves axotomy as only the soma is explanted, and this typically reverts the neuronal transcriptome to a more juvenile state. Similarly, Schwann cells in culture (Chapter 8) without axons resemble the repair Schwann cell phenotype after an axon lesion more than that of mature, differentiated Schwann cells. To overcome these limitations, it is important to consider in vivo methods for imaging axon structure (Chapter 17), ultrastructure (Chapter 18), interaction with other cell types (Chapter 16), and axonal transport (Chapters 20 and 21), as well as ex vivo systems that maintain much of the cellular architecture (Chapters 10, 11, 13, 14).

6

Michael P. Coleman

Sometimes the scientific question being asked limits the choice. For example, biochemical analysis of axon-specific material cannot easily be achieved using in vivo samples because it is extraordinarily difficult to avoid glial contaminants. However, in vitro methods yield only small amounts of tissue, and culturing, especially in the absence of glia, may itself influence axonal biochemistry. Similarly, to study specifically anterograde or retrograde axonal transport (Chapters 20 and 21), we must use a system where axonal orientation is known, such as tibial nerve, sparsely labeled primary culture neurons (Chapter 3), or microfluidic or spot cultures (Chapters 4 and 7), rather than corpus callosum or labeling an entire dispersed neuron culture with axons heading in every direction. Thus, many important conclusions can be made about general aspects of axon biology from any neuron type in culture or in vivo, and in a range of different species. This is important because the use of reductionist systems in primary neuronal culture or in simpler organisms can reveal mechanisms it would be difficult or impossible to discover in mammals in vivo. However, the differences between neuron types, especially between CNS and PNS (Chapters 10 and 14), and the differences in axon length and complexity are vital considerations for what the results mean for other neurons in other circumstances. An obvious solution is to use the advantages offered by a reductionist system to make initial discoveries and then to confirm them in a more complex system closer to humans. A good example is the discovery of the prodegenerative action of the gene SARM1 by screening in Drosophila before confirming a similar action of the orthologue in mice [15].

3

In Vivo, In Vitro, or Ex Vivo? Ultimately, evidence for a specific mechanism of axon degeneration, or of intraaxonal functions such as local protein synthesis or axonal transport, needs to be in vivo. Axon length, electrical activity (Chapters 11 and 13), physiological association with glia (Chapters 8 and 24), ageing, and the physical and metabolic stresses of everyday life all influence axon survival in vivo but cannot be fully modeled in vitro. However, in vivo studies, especially in mammals, have significant limitations. First, live imaging is more difficult than in vitro and sometimes involves invasive methods that themselves can influence axons, although there are ways to overcome this (Chapter 20). Second, genetic manipulation typically requires several generations of mice to develop a new strain or relies on viral delivery with variations in transduction efficiency between individuals and limited control of expression level. Third, the use of drugs or inhibitors is limited by blood brain barrier, difficulties in establishing the in vivo half-life, and off-target effects on nonneural tissues. Finally, the cost, regulatory burden, and animal welfare

Axon Degeneration: Which Method to Choose?

7

concerns for in vivo studies, especially in mammals, can be prohibitive. There are several solutions to this conundrum. One is to carry out a detailed study of mechanism in vitro followed by a limited investigation in vivo to confirm the physiological relevance of the most important findings. For example, we used in vitro studies to show that NAD precursor NMN can promote axon degeneration under conditions where it cannot be converted to NAD because use of the NAMPT inhibitor FK866 in vitro to switch off NMN synthesis, and readdition of exogenous NMN, was far more straightforward than in vivo [16]. However, it was the subsequent generation of NMN deamidase transgenic mice and lesion of their sciatic nerves that clearly showed a protective effect of sequestering NMN that matches that of WLDS and which provided sufficient tissue in which to analyze levels of all the relevant nucleotides to confirm the deamidase was acting as expected [17]. On other occasions, such as the ability of NMNAT1 overexpression to preserve injured axons, the results in primary culture and transgenic mice differ illustrating why such in vivo confirmation is important. Another solution is to carry out in vivo studies in simpler organisms. Zebrafish embryos are particularly useful as they can be rendered transparent for live imaging (Chapter 19) but introduce the limitations of an early developmental stage. Drosophila larvae (Chapter 21) have similar advantages and limitations. The use of adult flies or C. elegans offers other solutions, including some axon types that are amenable for live imaging such as wing sensory axons to study degeneration conveniently and longitudinally [18]. A third solution is to use ex vivo cultures from mammals, for example, organotypic hippocampal slice cultures (OHSCs) (Chapter 10) [19]. These retain neuronal circuits formed in vivo along with all associated glial cell types in close to normal cellular architecture. Addition of drugs to the culture medium is straightforward without complications of blood brain barrier or systemic toxicity and can be repeated as often as necessary, as can sampling of the culture medium to measure extracellular proteins and metabolites. Although OHSCs have to be established from juvenile tissue to survive, they do show signs of maturing ex vivo for example in their tau splicing pattern and the initial inflammatory reaction triggered by the slicing injury subsides over the first week in culture. Thus, considerable progress can be made in this way. Ex vivo tissue can also be used to study neuromuscular junction (NMJ) denervation (Chapters 11 and 12), axonal transport [13], Wallerian degeneration [20], mitochondrial function (Chapter 15), and central nervous system (CNS) electrophysiology (Chapter 13), in each case with the same advantages of ease of delivery of exogenous reagents and live imaging. When comparing data from one experimental system with that for another, it is important to remember that response times and

8

Michael P. Coleman

magnitude can change even if mechanistic principles remain similar. For example, the typical lag before an axotomized wild-type neurite in primary culture degenerates is around 6–8 h, compared to around 36 h for transected sciatic nerves in vivo and a similar time ex vivo [20], but the ability of WLDS and deletion of SARM1 to delay degeneration by around tenfold in each case clearly indicates a similar mechanism. Differences in glial support (Chapters 8 and 24) and/or axon length are two possible explanations why the absolute timing differs even if the relative timing is retained. Similarly, the loss of synaptophysin in OHSCs from the TgCRND8 mutant APP mouse is faster and more pronounced in ex vivo OHSCs than in the same strain in vivo (Chapter 10). This could reflect the additional strain placed on the ex vivo tissue by the slicing injury, which forces a remodeling of synaptic connections that would not be required in vivo to this degree, and by atmospheric oxygen, which exceeds the partial pressure of oxygen in the brain by around fourfold. While these remind us that mechanisms in vitro and ex vivo can differ from those in vivo, substantial progress can be made understanding mechanisms in this way before testing whether the findings hold in vivo.

4

Rodents, Flies, Humans, or Other Species? Much of the recent progress in understanding axon biology and mechanisms of degeneration has come from mouse or Drosophila studies, but there have been substantial recent and earlier contributions from studies in squid giant axon, zebrafish, and larger mammals, and confirmation in human tissue often remains the ultimate goal. For a given scientific question, what governs the choice of where to start? For mammalian studies, mouse is a good starting point because of the wealth of genetic mutant resources, well-understood physiology, well-developed disease models, and the low maintenance cost relative to larger mammals. However, despite an almost 1:1 equivalence of mouse and human genes, there are important but more subtle differences. For example, healthy adult humans express similar levels of 3R and 4R tau, whereas mice express predominantly 4R, a crucial cryptic termination site in the transcript of human stathmin-2 implicated in TDP-43-mediated ALS is missing in the mouse gene, and mice have only one gene for SMN, the protein mutated in spinal muscular atrophy, whereas humans have two. Thus, even for mice extrapolation to human conditions requires some confirmatory evidence such as the use of iPS-derived neurons (Chapter 9), human genetics, biopsy tissue, brain imaging, or postmortem studies. For example, the ability of WLDS to preserve axons has been confirmed in cultured human neurons [11] and the identification of human axonal disorders

Axon Degeneration: Which Method to Choose?

9

associated with NMNAT2 loss-of-function mutation [21, 22] further indicates the importance of the Wallerian mechanism in humans. Other differences between mice and humans include differences in axon length, longevity, the use of inbred strains, and pathogen-free conditions. Many human axonal disorders when modeled in mice produce only a subtle phenotype. While this is often thought to reflect longevity, these other differences are also very important. For example, some mouse strains lack α-synuclein or nicotinamide nucleotide transhydrogenase with important consequences for axon degeneration and protein aggregation disorders [23], and WLDS also arose as a “silent” mutation in one substrain until its effect was discovered in nerve lesion experiments [24]. The use of pathogen-free conditions, away from environmental pollutants using a standard diet almost certainly reduces the impact of many genetic models in mice too, especially considering the importance of viral infection and metabolic disorders to human neuropathy. For some mechanism questions, however, it is essential to use a simpler organism or neuronal cultures. An example would be for unbiased mutagenesis or RNAi screens (Chapters 6 and 22) and subsequent gene identification prior to replication in mammals in vivo and for avoiding homozygous lethality using MARCM to understand the impact of mutations and the cellular level [15]. However, it is essential to remember that Drosophila lack myelinated axons and an adaptive immune system before basing mammalian mechanism directly on such data. Zebrafish embryos support in vivo live imaging superior to other systems (Chapter 19), together with convenient in vivo drug screens, but the use of a juvenile system is a limitation.

5

To Cut or Not to Cut? Finally, as one of the simplest models of axon degeneration is axotomy, it is important to consider to what extent this does and does not model axonal disorders where there is no physical injury. Axotomy is the ultimate interruption of axonal transport. Although axonal transport continues for several hours in the distal stump [13], no molecules or organelles can subsequently be delivered from the soma. The failure to replenish one key axonal protein in this way, NMNAT2, together with ongoing turnover of this labile protein, activates Wallerian degeneration [5]. Consequently, depriving axons of NMNAT2 in other ways such as genetic knockdown (Chapter 3) or knockout in mice [25] or missense or frameshift mutation in humans [21, 22] also leads to axon death or failure of axon growth. Supplying a more stable NMNAT temporarily rescues a transected axon, but in the axotomy model, or in a

10

Michael P. Coleman

severe block of axonal transport [26], the rescue can never be permanent because something else will always become limiting eventually. For specific NMNAT2 mutation however, rescue by WLDS can last for many months [25], and with SARM1 deletion, it can be lifelong [27]. Thus, the capacity for axon preservation is significantly underestimated by pure axotomy models. Chemotherapy-induced peripheral neuropathy models (Chapter 22), for example, disrupting microtubule biology with vincristine or paclitaxel, offer one of the simplest noninjury methods to activate the Wallerian pathway. Many studies have demonstrated the ability of WLDS and SARM1 deletion to prevent CIPN degeneration in vitro and pain phenotype in vivo [28–30], and lowering NMNAT2 levels potentiates vincristine-induced degeneration [31]. Importantly, in the absence of a physical injury, there is no calcium influx through a cut end in this model, and the chemotherapy agent can be applied for a limited time, so longer term survival is possible. We made a similar finding when protein synthesis was impaired using the reversible inhibitor cycloheximide: if the drug is washed out before a threshold time, there can be permanent rescue by WLDS [5]. There are many other ways in which axons can be deprived of NMNAT2 activity: specific disruption of the palmitoylation mechanisms that underlie its axonal transport [32], alterations in protein turnover [5], splicing defects, toxins that directly inhibit this enzyme [33], and a deficiency of ATP needed for its synthesis and axonal transport and as an enzyme substrate, so Wallerian-like degeneration could contribute to other disease models. However, there are clearly other disorders where axons degenerate by other mechanisms, for example, by massive calcium influx, widespread disruption of RNA metabolism, failure of glial support, etc. Thus, before using axotomy as a model of any disorder, it is necessary first to show using the closest possible model to the human disease that blocking the Wallerian pathway delays axon degeneration.

6

Summary This chapter describes which axon degeneration mechanisms can be studied in many circumstances, and which differ between different neuron types, different species, and different experimental conditions. It helps to explain why degeneration timing can be different even when mechanisms are similar. It explains not only what makes some experimental systems ideally suited for some questions but also what limitations require subsequent confirmation in another experimental system. This is an important background for choosing the optimal experimental method in the chapters that follow.

Axon Degeneration: Which Method to Choose?

11

References 1. Conforti L, Gilley J, Coleman MP (2014) Wallerian degeneration: an emerging axon death pathway linking injury and disease. Nat Rev Neurosci 15:394–409. https://doi.org/10. 1038/nrn3680 2. Van Helleputte L, Kater M, Cook DP et al (2018) Inhibition of histone deacetylase 6 (HDAC6) protects against vincristineinduced peripheral neuropathies and inhibits tumor growth. Neurobiol Dis 111:59–69. https://doi.org/10.1016/j.nbd.2017.11.011 3. Corey DR (2017) Nusinersen, an antisense oligonucleotide drug for spinal muscular atrophy. Nat Neurosci 20:497–499. https://doi.org/ 10.1038/nn.4508 4. Matsuda W, Furuta T, Nakamura KC et al (2009) Single nigrostriatal dopaminergic neurons form widely spread and highly dense axonal Arborizations in the Neostriatum. J Neurosci 29:444–453. https://doi.org/10. 1523/JNEUROSCI.4029-08.2009 5. Gilley J, Coleman MP (2010) Endogenous Nmnat2 is an essential survival factor for maintenance of healthy axons. PLoS Biol 8: e1000300. https://doi.org/10.1371/journal. pbio.1000300 6. Nichols ALA, Meelkop E, Linton C et al (2016) The apoptotic engulfment machinery regulates axonal degeneration in C. elegans neurons. Cell Rep 14:1673–1683. https:// doi.org/10.1016/j.celrep.2016.01.050 7. Mack TGA, Reiner M, Beirowski B et al (2001) Wallerian degeneration of injured axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat Neurosci 4:1199–1206. https:// doi.org/10.1038/nn770 8. Adalbert R, Gillingwater TH, Haley JE et al (2005) A rat model of slow Wallerian degeneration ( WldS ) with improved preservation of neuromuscular synapses. Eur J Neurosci 21:271–277. https://doi.org/10.1111/j. 1460-9568.2004.03833.x 9. MacDonald JM, Beach MG, Porpiglia E et al (2006) The drosophila cell corpse engulfment receptor draper mediates glial clearance of severed axons. Neuron 50:869–881. https://doi. org/10.1016/j.neuron.2006.04.028 10. Martin SM, O’Brien GS, Portera-Cailliau C, Sagasti A (2010) Wallerian degeneration of zebrafish trigeminal axons in the skin is required for regeneration and developmental pruning. Development 137:3985–3994. https://doi.org/10.1242/dev.053611 11. Kitay BM, McCormack R, Wang Y et al (2013) Mislocalization of neuronal mitochondria

reveals regulation of Wallerian degeneration and NMNAT/WLDS-mediated axon protection independent of axonal mitochondria. Hum Mol Genet 22:1601–1614. https://doi. org/10.1093/hmg/ddt009 12. Stowers RS, Megeath LJ, Go´rska-Andrzejak J et al (2002) Axonal transport of mitochondria to synapses depends on Milton, a novel drosophila protein. Neuron 36:1063–1077. https://doi.org/10.1016/S0896-6273(02) 01094-2 13. Milde S, Adalbert R, Elaman MH, Coleman MP (2015) Axonal transport declines with age in two distinct phases separated by a period of relative stability. Neurobiol Aging 36:971–981. https://doi.org/10.1016/j. neurobiolaging.2014.09.018 14. Valdez G, Tapia JC, Lichtman JW et al (2012) Shared resistance to aging and ALS in neuromuscular junctions of specific muscles. PLoS One 7:e34640. https://doi.org/10.1371/ journal.pone.0034640 15. Osterloh JM, Yang J, Rooney TM et al (2012) dSarm/Sarm1 is required for activation of an injury-induced axon death pathway. Science 337:481–484. https://doi.org/10.1126/sci ence.1223899 16. Di Stefano M, Nascimento-Ferreira I, Orsomando G et al (2015) A rise in NAD precursor nicotinamide mononucleotide (NMN) after injury promotes axon degeneration. Cell Death Differ 22:731–742. https://doi.org/ 10.1038/cdd.2014.164 17. Di Stefano M, Loreto A, Orsomando G et al (2017) NMN deamidase delays Wallerian degeneration and rescues axonal defects caused by NMNAT2 deficiency in vivo. Curr Biol 27:784–794. https://doi.org/10.1016/j.cub. 2017.01.070 18. Neukomm LJ, Burdett TC, Gonzalez MA et al (2014) Rapid in vivo forward genetic approach for identifying axon death genes in Drosophila. Proc Natl Acad Sci U S A 111:9965–9970. https://doi.org/10.1073/pnas.1406230111 19. Harwell CS, Coleman MP (2016) Synaptophysin depletion and intraneuronal Aβ in organotypic hippocampal slice cultures from huAPP transgenic mice. Mol Neurodegener 11:44. https://doi.org/10.1186/s13024-016-01107 20. Beirowski B, Berek L, Adalbert R et al (2004) Quantitative and qualitative analysis of Wallerian degeneration using restricted axonal labelling in YFP-H mice. J Neurosci Methods

12

Michael P. Coleman

134:23–35. https://doi.org/10.1016/j. jneumeth.2003.10.016 21. Huppke P, Wegener E, Gilley J et al (2019) Homozygous NMNAT2 mutation in sisters with polyneuropathy and erythromelalgia. Exp Neurol 320:112958. https://doi.org/10. 1016/j.expneurol.2019.112958 22. Lukacs M, Gilley J, Zhu Y et al (2019) Severe biallelic loss-of-function mutations in nicotinamide mononucleotide adenylyltransferase 2 (NMNAT2) in two fetuses with fetal akinesia deformation sequence. Exp Neurol 320:112961. https://doi.org/10.1016/j. expneurol.2019.112961 23. Navarro SJ, Trinh T, Lucas CA et al (2012) The C57BL/6J mouse strain background modifies the effect of a mutation in Bcl2l2. G3 (Bethesda) 2:99–102. https://doi.org/10. 1534/g3.111.000778 24. Lunn ER, Perry VH, Brown MC et al (1989) Absence of Wallerian degeneration does not hinder regeneration in peripheral nerve. Eur J Neurosci 1:27–33. https://doi.org/10.1111/ j.1460-9568.1989.tb00771.x 25. Gilley J, Adalbert R, Yu G, Coleman MP (2013) Rescue of Peripheral and CNS axon defects in mice lacking NMNAT2. J Neurosci 33:13410–13424. https://doi.org/10.1523/ JNEUROSCI.1534-13.2013 26. Ferri A, Sanes JR, Coleman MP et al (2003) Inhibiting axon degeneration and synapse loss attenuates apoptosis and disease progression in a mouse model of Motoneuron disease. Curr Biol 13:669–673. https://doi.org/10.1016/ S0960-9822(03)00206-9 27. Gilley J, Ribchester RR, Coleman MP (2017) Sarm1 deletion, but not Wld S, confers lifelong

Rescue in a Mouse Model of severe Axonopathy. Cell Rep 21:10–16. https://doi.org/10. 1016/j.celrep.2017.09.027 28. Geisler S, Doan RA, Strickland A et al (2016) Prevention of vincristine-induced peripheral neuropathy by genetic deletion of SARM1 in mice. Brain 139:3092–3108. https://doi.org/ 10.1093/brain/aww251 29. Geisler S, Doan RA, Cheng GC et al (2019) Vincristine and bortezomib use distinct upstream mechanisms to activate a common SARM1-dependent axon degeneration program. JCI Insight 4:e129920. https://doi. org/10.1172/jci.insight.129920 30. Turkiew E, Falconer D, Reed N, Ho¨ke A (2017) Deletion of Sarm1 gene is neuroprotective in two models of peripheral neuropathy: deletion of Sarm1 gene is neuroprotective in two models of peripheral neuropathy. J Peripher Nerv Syst 22:162–171. https://doi. org/10.1111/jns.12219 31. Gilley J, Mayer PR, Yu G, Coleman MP (2019) Low levels of NMNAT2 compromise axon development and survival. Hum Mol Genet 28:448–458. https://doi.org/10.1093/ hmg/ddy356 32. Milde S, Gilley J, Coleman MP (2013) Subcellular localization determines the stability and axon protective capacity of axon survival factor Nmnat2. PLoS Biol 11:e1001539. https:// doi.org/10.1371/journal.pbio.1001539 33. Buonvicino D, Mazzola F, Zamporlini F et al (2018) Identification of the Nicotinamide salvage pathway as a new Toxification route for antimetabolites. Cell Chem Biol 25:471–482. e7. https://doi.org/10.1016/j.chembiol. 2018.01.012

Part II In Vitro Assays

Chapter 2 Axon Degeneration Assays in Superior Cervical Ganglion Explant Cultures Andrea Loreto and Jonathan Gilley Abstract The ability of peripheral nervous system neurons to extend long, axon-like neurites in vitro makes them ideally suited for studies on mechanisms of axon survival and degeneration. In this chapter, we describe how to prepare explant cultures of sympathetic neurons of the superior cervical ganglion (SCG). We also describe how to induce and assess axon degeneration with an injury or a chemical insult. Key words Axon degeneration, Superior cervical ganglion, SCG explant culture, Wallerian degeneration, Injury, Axon transection

1

Introduction Axon loss is an early feature of several neurodegenerative disorders and leads to compromised neuronal function. Understanding how axons die is important to identify therapeutic targets to delay or halt the progression of these pathologies [1]. Much of the current knowledge on mechanisms of axon degeneration comes from studies in primary neuronal cultures of sympathetic superior cervical ganglion (SCG) and sensory dorsal root ganglion (DRG) neurons. This chapter focuses on how to culture and assess neurite (axon) degeneration in SCG explant cultures from postnatal mice. Explant cultures offer a number of advantages when assessing axon degeneration. These are relatively easy to dissect and culture since plating of the whole ganglia avoids complications and variability due to dissociation protocols. Neuronal cell bodies are located within the ganglion, with long axon-like neurites (5–6 mm long at 7 days in vitro, DIV) radially extending away from it; this hugely simplifies experiments of axon degeneration following an injury since all neurites distal to the site of injury are transected (Fig. 1). Furthermore, directly comparable areas of neurites (e.g., the distal ends) can be imaged over time across different dishes and

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_2, © The Author(s) 2020

15

16

Andrea Loreto and Jonathan Gilley

Fig. 1 Representative images of uninjured and injured SCG explant cultures. (a) Representative phase-contrast image of a SCG explant at 7 DIV. Long neurites have extended radially from the ganglion. Different types of transection can be performed: (b) A short transection permits a side-by-side comparison of cut and uncut neurites. (c) Transection of all the neurites on one side of the dish avoids overlapping of uncut and cut neurites and increases the area of degenerating neurites that can be imaged. (d) Transection of all neurites to maximize the area of degenerating neurites to image (see Note 11). (e) Representative phase-contrast images of healthy and degenerating neurites

Studying Axon Degeneration in SCG Explants

17

experimental conditions. Finally, several neurites are imaged at the same time, minimizing differences in the time of degeneration of individual neurites and facilitating quantification of the extent of degeneration. In this chapter, we describe two methods to induce neurite degeneration: (1) the injury model, where degeneration is initiated by a physical transection of the neurites, a process known as Wallerian degeneration, and (2) chemically induced degeneration, where neurite death is caused by the administration of a toxic compound to uninjured neurons. The most widely used toxic compounds to induce neurite degeneration are chemotherapy agents such as vincristine, the mitochondrial toxins Carbonyl cyanide m-chlorophenyl hydrazone (CCCP), and rotenone and inhibitors of protein synthesis such as cycloheximide and emetine [2–5]. Interestingly, all these drugs initiate an axon death program that is mechanistically related to Wallerian degeneration. Quantification of neurite degeneration can be easily performed through analysis of phasecontrast images with an ImageJ plugin to give a degeneration index score [6]. This method allows images to be acquired from the same culture at different timepoints, so that morphological changes occurring to the same group of neurites can be followed over time. Alternatively, neurite morphology can be visualized by immunostaining with antibodies against neuronal and axonal markers, although this requires cultures to be fixed, so a separate dish is needed for each timepoint. Not only does this preclude following changes in the same group of neurites, but it also means that a greater number of cultures are needed for every experiment. Importantly, although not detailed here, the transection experiments described in this chapter can also be used to provide material from the cell body and neurites separately for biochemical and protein expression studies [2, 7, 8]. Finally, although this chapter is specific for mouse SCG explant cultures, similar methods can also be applied to rat SCG and mouse/rat DRG explant cultures.

2

Materials Preparation of solutions should be performed in a laminar flow cabinet or Class II biological safety cabinet.

2.1 Dissection and Plating of SCG Explant Cultures

1. No. 5 forceps and microdissection scissors sterilized in 70% ethanol and air-dried before use. 2. Leibovitz’s L-15 medium (with L-glutamine). Store at 4  C. 3. SCG medium: Dulbecco’s Modified Eagle Media (DMEM) (with 4500 mg/l glucose and 110 mg/l sodium pyruvate) supplemented with 10% fetal bovine serum (FBS) or 2% B27

18

Andrea Loreto and Jonathan Gilley

(see Note 1), 1 penicillin/streptomycin, 2 mM L-glutamine, 4 μM aphidicolin, and 25–50 ng/ml 2.5S nerve growth factor (NGF) or 100 ng/ml 7S NGF (see Note 2). A stock of 50 ml (or more) of SCG medium containing DMEM, FBS or B27, penicillin/streptomycin, and L-glutamine can be stored at 4  C for up to 3 weeks. Aphidicolin and NGF should always be added fresh to the required amount of medium. Concentrated stock solutions of each supplement are as follows: 100 penicillin/streptomycin (10,000 U/ml penicillin and 10,000 μg/ ml streptomycin in 10 mM citrate buffer); 100 L-glutamine (200 mM); 4 mM aphidicolin in DMSO; and 100 μg/ml 2.5S or 7S NGF prepared following the manufacturer’s instructions. Aliquots of each supplement are stored at 20  C. 4. Sterile 35 mm tissue culture dishes. 5. Solutions for coating of tissue culture dishes: 20 μg/ml poly-Llysine (Mw  300,000) hydrobromide in sterile, tissue culturegrade water and 20 μg/ml laminin in DMEM (980 μl DMEM per 20 μl laminin aliquot). Do not use SCG medium to dilute the laminin as the presence of FBS will interfere with coating. Concentrated stock solutions of each supplement are as follows: 2.5 mg/ml poly-L-lysine stock solution (125) in sterile, tissue culture-grade water and 1 mg/ml laminin, stored as 20 μl aliquots in 1.5 ml microfuge tubes. Aliquots of each supplement are stored at 20  C. 6. Sterile, disposable blades for attachment to a nondisposable scalpel handle or sterile disposable surgical scalpels with a polystyrene handle. Blades should be curved with a relatively long cutting edge (e.g., No. 10 or No. 22).

3

Methods All steps are performed at room temperature in a horizontal laminar flow cabinet unless indicated otherwise.

3.1 Dissection and Culture of SCG Explants

1. To coat the dishes, add 1 ml of poly-L-lysine hydrobromide solution to cover the whole surface of a 35-mm sterile tissue culture dish and incubate for at least 1 h at room temperature. Next, remove the poly-L-lysine solution and wash the dish twice with sterile water before leaving to air dry completely. Dried poly-L-lysine-coated dishes can be stored at 4  C wrapped in parafilm for several weeks. Add ~200–250 μl of the laminin solution to coat a circular area, approximately 20 mm in diameter, in the center of each dish (see Note 3). Incubate for 1–2 h at room temperature or in a 5% CO2 incubator at 37  C. Remove the laminin solution and immediately add 600 μl of SCG medium to cover the whole surface of a

Studying Axon Degeneration in SCG Explants

19

Fig. 2 Dissection of SCG explants. (a and b) Excessive fat tissue (green arrow) and trachea (white arrow) are removed to expose the ganglia. (c) The two SCG explants (white arrows) have a distinctive oval shape and appear white/translucent compared to surrounding tissue. (d) Uncleaned SCG soon after dissection. (e) Clean SCG ready to be plated

35-mm tissue culture dish before plating of ganglia. The laminin-coated area should not be allowed to dry. 2. Dissect superior cervical ganglia from 0- to 3-day-old mice (see Note 4) using a stereoscopic microscope, sterilized No. 5 forceps, and microdissection scissors. Decapitate mice using sharp scissors, which must be positioned behind the ears to avoid accidental removal of the ganglia. Pin the head to a Syligard-coated dish, and remove excessive tissue and trachea to expose the ganglia. These are located at the point where the carotid artery bifurcates into the internal and external carotid arteries. The SCG has a distinctive oval shape and appears white/translucent compared to surrounding tissue (Fig. 2). Care should be taken to distinguish the SCG from the nodose ganglion, which is also located close the carotid artery branch point. The nodose ganglion is smaller and attached to a thicker nerve fiber. Transfer the dissected ganglia in a 35-mm sterile tissue culture dish filled with Leibovitz’s L-15 medium (see Note 5).

20

Andrea Loreto and Jonathan Gilley

3. Use forceps to gently remove any associated nerve fibers and extraneous tissue (see Note 6). Transfer cleaned SCG explants to another 35 mm dish filled with Leibovitz’s L-15 medium. 4. Plate two-to-three ganglia in the center of the tissue culture dish filled with 600 μl of SCG medium (see Note 7), and transfer the dishes to a 5% CO2 incubator at 37  C overnight. If litters of mixed genotypes are being processed, two ganglia from an individual mouse can be plated in the same dish to ensure that explants deriving from mice with different genotypes are kept separate. 5. The day after plating, check that ganglia are attached and that neurites have extended radially (they should be several hundred microns long by this stage). Carefully flood the dishes with 1 ml of SCG medium prewarmed to 37  C. Replace medium every 2–3 days with 1–2 ml of fresh SCG medium prewarmed to 37  C (see Note 8). 3.2 Neurite Degeneration Assays in SCG Explant Cultures

1. Injury-induced degeneration assay: Perform transection using an inverted microscope with a 2.5 or 5 objective under transmitted light. Position the sterile scalpel blade, and cut by rolling the cutting edge of the blade along the surface of the dish with a single down-and-up movement. Bring the blade back to its tip before removing it from the dish to avoid accidental detachment of the neurite network caused by slippage of the blade (see Note 9). Keep the DIV at which the cultures are manipulated consistent across dishes and experiments to decrease variability in the timing of degeneration (see Note 10). Different types of transection can be performed (Fig. 1) (see Note 11). 2. Chemically induced degeneration assay: Replace the medium with fresh SCG medium containing the compound of interest (or vehicle) at the desired concentration. The concentration should be determined by performing dose-response experiments (see Note 12). 3. Quantification of neurite degeneration: Acquire phase-contrast images of neurites using 10 or 20 objectives. The degeneration index can be calculated using the ImageJ plugin referenced in the introduction of this chapter (see also Chapter 4). The parameters of the plugin should be adjusted for the objective used and/or the resolution of the image. It is advisable to image the distal ends where the neurite network is less dense (see Note 13). Always capture a 0-h timepoint; subsequent timepoints will depend on the type of insult applied (see Note 14). Ideally, the same field of neurites should be imaged at each timepoint. Intact neurites will retain continuity, whereas the appearance of blebs and fragmentation are indicators of degeneration (Fig. 1e).

Studying Axon Degeneration in SCG Explants

4

21

Notes 1. Other sources of FBS can be used but should first be tested rigorously for their ability to maintain healthy cultures. Different lots of FBS from the same source can also vary in their quality. B27 supplement can also be used. While neurite morphology differs slightly when FBS or B27 are used, they respond similarly in degeneration assays. Culture quality can be assessed by checking morphology of the neurites which should look free from blebs, form a dense network and reach 5–6 mm length at 7 DIV. 2. Aphidicolin is used as an antimitotic to limit proliferation of nonneuronal cells. A combination of 20 μM fluorodeoxyuridine and 20 μM uridine can be used as an alternative to aphidicolin but is less potent. 7S and 2.5S NGF can be obtained from a number of commercial sources. SCG neurons require NGF to survive; too low or too high concentrations in the medium will result in unhealthy cultures. It is therefore critical to determine empirically the optimal concentration to use. A comparison of batches from different sources is advisable. 3. We recommend starting the coating procedure (which takes 2.5–3 h) before the dissection of SCG explants, so that the ganglia can be cultured as soon as possible after dissection. It is important that the dishes are fully dried after the poly-L-lysine coating to avoid that the laminin solution spreads across the whole dish surface (to limit the amount of laminin solution used). 4. The age of the mouse pups can influence the growth characteristics of the neurites, so this should be as closely matched as possible. 5. Although not essential, it is advisable to keep mouse heads on ice and to use cold L-15 medium, if the dissection procedure is protracted. Mouse heads can be kept on ice for up to 4 h and SCG ganglia can be stored in L-15 medium for up to 2 h at 4  C before plating, if required. 6. Some SCG explants might get damaged during the dissection and cleaning procedures, but in most cases, incomplete ganglia will survive and grow normally in culture and can be used. 7. SCG explants are plated in the center of a 35-mm dish containing 600 μl of SCG medium (this should be added immediately after laminin is removed, see step 1) to promote attachment and growth. This volume of SCG medium is enough to keep the surface of the dish from drying out overnight in the incubator but small enough that the ganglia are held in the center of the dish by the surface tension. Care should be taken when

22

Andrea Loreto and Jonathan Gilley

transferring dishes to the incubator to avoid ganglia moving from the center dish. The volume will have to be adapted if dishes of different size are used. Plating two-to-three ganglia per dish will increase internal variability and reliability of the experiments, and plating explants close to each other simplifies neurite transection. Do not prewarm the SCG medium if the ganglia are kept in cold L-15 medium during the dissection procedure (see Note 5). 8. It is important to add SCG medium within 16–24 h after plating to prevent cultures from drying out. Care should be taken when medium is added to SCG explant cultures or dishes are moved out of the incubator since ganglia can detach. To avoid this, gently add a couple of small drops of medium directly over the ganglia before adding the remaining medium to the side of the dish. For subsequent medium changes, always remove old medium and add fresh medium slowly to the side of the dish. 9. Great care should also be taken when moving dishes after introducing the cut as transected/degenerating neurites are more prone to detach from the dish surface. 10. Although injury experiments can be performed at any time from 2 DIV, more consistent results are obtained between 5 and 7 DIV, when neurites reach the optimal length for transection. Neurites cut between 5 and 7 DIV consistently degenerate within 6–8 h after injury. 11. A short transection (Fig. 1b) permits a side-by-side comparison of cut and uncut neurites, but it is important to note that intact neurites at the edges of the cut might overlap with injured, degenerating neurites. A field of neurites distant from the edge of the cut should thus be selected if a quantification of degeneration of the transected neurites is performed. Transection of all the neurites on one side of the dish (Fig. 1c) avoids this problem and increases the area of degenerating neurites that can be imaged. However, this type of transection increases the chances of detachment of the neurite network. Finally, transection of all neurites can be performed (Fig. 1d) to maximize the area of degenerating neurites, particularly, when collecting material for biochemical and/or protein analyses, although an internal uncut control will be lacking. 12. Concentrations of the most commonly used agents that cause degeneration of SCG neurites: 20 nM vincristine, 50 μM CCCP, 10 μM rotenone, 10 μg/ml cycloheximide, and 10 μM emetine. The time course of degeneration following these treatments is generally slower than that of transected neurites (6–8 h after injury at 5–7 DIV), the former normally degenerate between 12 and 24 h after drug addition. The

Studying Axon Degeneration in SCG Explants

23

length of time in culture and culturing conditions might impact on the degeneration time course. Most compounds used to induce neurite degeneration are dissolved in DMSO. To avoid any unwanted toxicity, it is preferable that the final concentration of DMSO in the medium does not exceed 0.5%. 13. When quantifying the degeneration index using the ImageJ plugin, the most reliable results are obtained when exactly the same field of neurites is imaged at each timepoint with the quality of images being as consistent as possible (especially background signal). 14. Once the time of degeneration for a specific experimental condition is determined, reduce the timepoints acquired to minimize movement of the dishes out of the incubator and prevent detachment of degenerating neurites. The highest degeneration index score can be assigned to neurites that fully detach from the dish. However, this should only be done when degeneration could be documented in the previous timepoint as dish movements can make transected neurites detach at early timepoints even without degeneration. If neurites have detached before degeneration has been seen, they should be excluded from the analysis.

Acknowledgments The authors were supported by funding from the UK Medical Research Council (grant number MR/N004582/1), Parkinson’s UK (project grant G-1602), and a Sir Henry Wellcome postdoctoral fellowship from the Wellcome Trust (grant number 210904/ Z/18/Z). References 1. Conforti L, Gilley J, Coleman MP (2014) Wallerian degeneration: an emerging axon death pathway linking injury and disease. Nat Rev Neurosci 15:394–409. https://doi.org/10. 1038/nrn3680 2. Gilley J, Coleman MP (2010) Endogenous Nmnat2 is an essential survival factor for maintenance of healthy axons. PLoS Biol 8: e1000300. https://doi.org/10.1371/journal. pbio.1000300 3. Loreto A, Di Stefano M, Gering M, Conforti L (2015) Wallerian degeneration is executed by an NMN-SARM1-dependent late Ca2+ influx but only modestly influenced by mitochondria. Cell Rep 13:2539–2552. https://doi.org/10.1016/ j.celrep.2015.11.032

4. Di Stefano M, Loreto A, Orsomando G et al (2017) NMN Deamidase delays Wallerian degeneration and rescues axonal defects caused by NMNAT2 deficiency in vivo. Curr Biol 27:784–794. https://doi.org/10.1016/j.cub. 2017.01.070 5. Summers DW, DiAntonio A, Milbrandt J (2014) Mitochondrial dysfunction induces Sarm1-dependent cell death in sensory neurons. J Neurosci 34:9338–9350. https://doi.org/10. 1523/JNEUROSCI.0877-14.2014 6. Sasaki Y, Vohra BPS, Lund FE, Milbrandt J (2009) Nicotinamide mononucleotide adenylyl transferase-mediated axonal protection requires enzymatic activity but not increased levels of neuronal nicotinamide adenine dinucleotide. J

24

Andrea Loreto and Jonathan Gilley

Neurosci 29:5525–5535. https://doi.org/10. 1523/JNEUROSCI.5469-08.2009 7. Gilley J, Orsomando G, Nascimento-Ferreira I, Coleman MP (2015) Absence of SARM1 rescues development and survival of NMNAT2deficient axons. Cell Rep 10:1974–1981. https://doi.org/10.1016/j.celrep.2015.02. 060

8. Di Stefano M, Nascimento-Ferreira I, Orsomando G et al (2015) A rise in NAD precursor nicotinamide mononucleotide (NMN) after injury promotes axon degeneration. Cell Death Differ 22:731–742. https://doi.org/10.1038/ cdd.2014.164

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 3 Microinjection of Superior Cervical Ganglion Neurons for Studying Axon Degeneration Jonathan Gilley and Andrea Loreto Abstract Primary cultures of neurons of the peripheral nervous system have been successfully used for studying many aspects of neuronal development and survival, including investigations into the mechanisms of axon degeneration. In this chapter, we describe how to prepare and microinject dissociated cultures of sympathetic neurons of the superior cervical ganglion (SCG) specifically for use in highly controlled and targeted assays of axon survival and degeneration. Key words Superior cervical ganglion, Dissociated SCG culture, Microinjection, Axon survival, Axon degeneration, Axon transection, Wallerian degeneration

1

Introduction Primary cultures of early postnatal sympathetic neurons of the superior cervical ganglion (SCG), which are dependent on nerve growth factor (NGF) for survival, are an established model system for studying many aspects of neuronal biology. SCG neurons predominantly extend axon-like neurites [1], and these can grow several millimeters in just a few days making them ideally suited to investigations of axonal function. Indeed, studies in SCG cultures, together with studies in primary cultures of mid-to-late embryonic stage sensory neurons of the dorsal root ganglion (DRG), have contributed significantly to our understanding of the key molecular regulators of axon degeneration caused by physical injury (Wallerian degeneration) or related degeneration triggered by chemical insults [2, 3]. Crucially, similar mechanisms appear to be involved during the degeneration of axons in a number of neurodegenerative disorders [2]. Targeted suppression of gene expression by RNA interference and exogenous expression of wild-type or modified proteins are useful tools for probing the roles of specific genes in cellular processes. SCG and DRG neuron cultures are relatively resistant to

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_3, © The Author(s) 2020

25

26

Jonathan Gilley and Andrea Loreto

standard chemical transfection, so viral transduction of constructs has become a favored method for achieving these outcomes, especially in DRG explant or spot cultures (see Chapters 4 and 5). However, microinjection has also proved successful for delivery of expression vectors and other materials into cultured neurons and, in particular, SCG neurons, where larger cell bodies/nuclei and a greater ability to tolerate injection stresses make the process much more efficient than for DRG neurons or other neuron types (although DRG neurons can be injected, if desired). Microinjection is a technically challenging and low-throughput method targeting relatively small numbers of cells. While this makes it unsuited to biochemical analyses, it does provide a unique degree of precision that is particularly useful in studies of neurite (axon) degeneration in primary neuron cultures. For example, it provides a greater degree of control over the concentrations of the materials being introduced, as well as greater flexibility over the type and combinations of material that can be used. The ability to target small numbers of neurons in specific locations can also provide better spatial resolution of individual neurites than nontargeted methods. In addition, after an initial financial outlay on some specialized equipment, subsequent costs are relatively low and the ability to use standard mammalian expression vectors in injections, with no theoretical size restriction, is a particular advantage. In this chapter, we present core methods for preparing and microinjecting primary cultures of dissociated mouse SCG neurons with the specific aim of studying neurite degeneration. A procedure for the dissociation of mouse SCG neurons, which is adapted from an earlier protocol for the dissociation of rat SCG neurons [4], is described first. Dissociated cultures are required, so that individual neurons can be targeted easily. Rat SCG neurons can be used if preferred, but most studies of neurite degeneration are performed in mouse cultures due to the growing number of useful mutant mouse models with altered axon survival [2, 3]. Our method also describes a minor modification for the preparation of independent cultures from individual mice, for example, if litters of mixed genotypes are being processed. We next describe the basic procedure for microinjection of SCG neurons for the specific purpose of tracking the survival of their neurites. Crucially, in addition to any test materials being injected, a means for fluorescent labeling must be included in order that the injected neurons and their neurites can be visualized separately from the mass culture (Fig. 1). While we do not provide specific details, injected test materials have, in the past, included mammalian expression vectors and small interfering RNAs (siRNAs) [5–11], although purified proteins and other compounds can also be injected. We additionally describe a simple method for transecting the labeled neurites with a scalpel blade to trigger Wallerian degeneration, if required (Fig. 1).

Microinjection for Studying Axon Degeneration

27

Fig. 1 Representative images of a generic microinjection experiment showing degeneration of labeled neurites after transection. (a) Low-magnification fluorescence and phase-contrast images of a region of a dissociated culture of wild-type SCG neurons 48 h after injection of a small cluster of the neurons with a fluorescent protein expression vector (pDsRed2). The series of images shows the same area before, immediately after (0 h), and 8 h after the introduction of a cut with a scalpel blade. Labeled cell bodies of the injected neurons are seen to the left of the fluorescence images with some of their labeled neurites extending to the right, the distal ends of which become disconnected from their cell bodies following transection and degenerate within 8 h. Phase-contrast images reveal that changes in the relatively few labeled neurites are occurring within a dense network of unlabeled SCG neurons and neurites that appear grossly unaffected except for a small amount of retraction and neurite regeneration occurring at the cut site. (b) Higher magnification fluorescence images of the boxed regions in (a) for better visualization of the complete loss of continuity of the labeled, transected neurites by 8 h after cut. Scale bars ¼ 100 μm

Finally, although this chapter is limited to a description of microinjection as a tool for assessing the survival of individual neurites (intact or transected) in response to various injected materials, it is important to note that there are many other potential applications of the technique. For example, microinjection of SCG neurons has already proven useful for live-imaging studies of axonal transport [7, 11, 12] and could easily be applied to other aspects of axon biology, such as regeneration.

28

2

Jonathan Gilley and Andrea Loreto

Materials

2.1 Dissociation and Culture of SCG Neurons

Solution preparation and coating of dishes should be performed in a laminar flow cabinet or Class II biological safety cabinet. Sterile, ultrapure water (deionized, filtered water purified to 18.2 MΩ·cm resistivity at 25  C) should be used for solutions, as applicable. 1. Trypsin solution: Dissolve lyophilized trypsin from bovine pancreas in PBS() (phosphate-buffered saline withoutcalcium chloride and magnesium chloride) at a concentration of 0.025%. Alternatively, the 0.025% working solution can be diluted from a 0.25% concentrated stock solution. Store aliquots at 20  C. 2. Collagenase solution: Dissolve desiccated type 2 collagenase (storage at 4  C) in PBS (withcalcium chloride and magnesium chloride) at a concentration of 0.2%, and filter sterilize through a 0.22 μm filter. Large batches can be prepared and aliquots stored at 20  C; however, best results are obtained with freshly prepared solution. Activity of the desiccated collagenase may vary between batches, so incubation times should be adjusted accordingly. 3. SCG medium: DMEM (with 4500 mg/l glucose and 110 mg/l sodium pyruvate) supplemented with 10% fetal bovine serum (FBS) (see Note 1), 1 penicillin/streptomycin, 2 mM L-glutamine, 2–4 μg/ml aphidicolin (see Note 2), and 25–50 ng/ml 2.5S NGF or 100 ng/ml 7S NGF (see Note 3). Concentrated stock solutions of each supplement are as follows: 100 penicillin/streptomycin (10,000 U/ml penicillin and 10,000 μg/ ml streptomycin in 10 mM citrate buffer), 100 L-glutamine (200 mM), 4 mg/ml aphidicolin in DMSO, and 100 μg/ml 2.5S or 7S NGF prepared as per supplier instructions. Aliquots of each supplement are stored at 20  C. SCG medium without aphidicolin and NGF can be stored for several weeks at 4  C with the two omitted items being added fresh as needed. 4. Low-sided (uncoated) 35 mm μ-Dishes (ibidi) coated with poly-L-lysine and laminin (see Note 4): First, cover just the uncoated coverslip insert bases of the dishes with 1 ml of a solution of 20 μg/ml poly-L-lysine hydrobromide (Mw  300,000) dissolved in water. The working poly-L-lysine solution can be diluted from aliquots of a 2.5 mg/ml stock solution in water (store at 20  C). After at least 1 h at room temperature, remove the poly-L-lysine solution and wash the coverslip insert base once with sterile, ultrapure water before leaving to air dry completely. Dried poly-L-lysine-coated dishes can be stored for several weeks at 4  C wrapped in parafilm. For subsequent laminin coating, 1 mg/ml laminin (stored as 20 μl aliquots in 1.5 ml microfuge tubes at 20  C) is diluted to

Microinjection for Studying Axon Degeneration

29

20 μg/ml in DMEM (980 μl per 20 μl aliquot) and ~50–100 μl of the solution used to coat a small circular area, typically 7–10 mm in diameter, in the center of each coverslip insert base. It is important not to use SCG medium to dilute the laminin as the presence of FBS will interfere with coating. After 1–2 h at room temperature, the laminin solution is aspirated immediately prior to addition of the dissociated neurons. The laminin-coated area should not be allowed to dry. 2.2

Microinjection

1. A complete microinjection system setup for injecting on an inverted microscope with a 32/40 objective and optional 4/5 objective and a holder for a 35 mm dish (see Note 5). The microscope should sit on an anti-vibration table. The procedures for injection described in this chapter are based on an Eppendorf system (Fig. 2). In it, an Eppendorf FemtoJet® microinjector pressurizes an injection capillary that is attached to a high resolution x,y,z motor module and an Eppendorf TransferMan® micromanipulator. The micromanipulator allows proportional control over capillary movements. The

Fig. 2 Components of a typical microinjection system for injecting SCG neurons in dissociated cultures. The system shown does not include a CO2 supply

30

Jonathan Gilley and Andrea Loreto

motor module can be attached to the microscope via a manufacturer-specific attachment or can be positioned using a universal stand. Ideally, the microscope should have a heated stage. An environment chamber with regulated CO2 supply is optional. 2. FemtoTip® injection capillaries (Eppendorf) with screw-in thread for quick assembly into the capillary holder of the FemtoJet® microinjector (see Note 6). 3. 0.22 μm centrifuge tube filter columns such as Corning® Costar® Spin-X® centrifuge tube filters. 4. Eppendorf Microloader™ pipette tips for loading injection mix into injection capillaries and for removing any excess. 5. Sterile disposable blades for attachment to a nondisposable scalpel handle or sterile disposable surgical scalpels with a polystyrene handle. Blades should be curved with a relatively long cutting edge (e.g. No. 10 or No. 22).

3

Methods All steps are performed at room temperature in a horizontal laminar flow cabinet, unless indicated otherwise.

3.1 Dissociation and Culture of SCG Neurons

1. Dissect superior cervical ganglia from mousepups at any age up to 2–3 days after birth as described in Chapter 2. Typically, 1–2 ganglia will provide sufficient dissociated neurons for a single dish with a laminin-coated area of 7–10 mm in diameter. Nerve fibers, capillaries, or other encapsulating tissue should be cleaned from the ganglia before proceeding. 2. Transfer ganglia to 0.025% trypsin solution, prewarmed to 37  C, using a pair of sterile No. 5 forceps (sterilized in 70% ethanol and air dried before use). Use 5 ml of trypsin solution in a 15-ml tube for ganglia from multiple mice of the same genotype (up to 20–30) or, if processing individual mice separately, use 1 ml of trypsin solution in separate 1.5 ml tubes for each pair of ganglia. Incubate at 37  C in a waterbath for no more than 20–30 min. 3. Transfer the ganglia to 0.2% collagenase solution, prewarmed to 37  C, using a sterile 1 ml tip attached to an appropriate pipetman/pipettor. Transfer in as small a volume of trypsin solution as possible. As in step 2, use 5 ml of collagenase solution in a 15-ml tube for multiple ganglia or 1 ml of collagenase solution in separate 1.5 ml tubes for individual pairs of ganglia. Incubate at 37  C in a waterbath for no more than 20–30 min.

Microinjection for Studying Axon Degeneration

31

4. Transfer the ganglia to a fresh 15 ml tube or separate 1.5 ml tubes, as appropriate, using a sterile 1 ml tip attached to an appropriate pipetman/pipettor. Transfer in as small a volume of the collagenase solution as possible (ideally 10000) { print(p,",",Fname,",",n,",",taa,",",faa,",",faa/taa); } run("Revert"); run("Open Next"); Fname = getTitle(); selectImage(Fname); } //

3. The axon degeneration assay results can be presented as a scatter plot with connecting lines, with degeneration indices plotted against time after axotomy (Fig. 3b) (see Note 19).

4

Notes 1. A pregnant CD1 mouse usually has more than 10 embryos. Inbred strains are not necessary for this assay. Spot cultures have been made using E12.5–E14.5 embryos; however, the

Axon Degeneration Assay in Cultured DRG Neurons

51

dissection method works the most efficiently with E13 embryos. 2. PDL can be prepped as 10–100
stock solution and stored at 20  C in aliquots. Avoid repeated freeze-thaw cycles. 3. To make a mixed solution of 5-fluoro-20 -deoxyuridine and uridine as a 200
stock solution, dissolve 25 mg of 5-fluoro20 -deoxyuridine and 25 mg of uridine in 50 mL of ultrapure water, sterilize with a syringe filter, and store at 20  C in aliquots. 4. Depending on their intended purpose, DRG cultures are grown on different plate types. For example, a 4-well chamber slide with removable wells can be used for immunofluorescence imaging after axotomy, and the glass-bottomed FluoroDish is optimized for live imaging. The volume of coating solutions or culture medium should be adjusted proportionally to the surface area of each dish type. 5. Each 24-well plate requires 1–2 embryos depending on the exact yield of DRGs from the dissection. A skilled person can dissect out more than 45 DRGs from an embryo in about 15 min. 6. To get rid of extra tissue contaminants (e.g., blood cells), dissected spinal cords can be transferred to a new Petri dish filled with dissection medium. 7. Repetitive pipetting of small tissue chunks, such as tail, around 10–20 times will coat the tip. Avoid tissue remains inside the tip. 8. If the dissected embryos are more than five, divide the collected DRGs into two Eppendorf tubes for effective trypsinization. 9. Adjust the trypsin incubation time if necessary. Undertrypsinization results in insufficient dissociation of the tissue clumps, while over-trypsinization will lead to cell death. 10. If the dishes are not completely dried at this step, spots will spread. To avoid this problem, leave the dishes uncovered until spotted. 11. There is often a sticky clump of supposed nerve fibers that does not dissociate; this can be gently discarded using a P20 pipette tip. 12. Adjust the volume of the cell suspension for each spot depending on the density of cells and axons in culture. Guide for troubleshooting: count the number of collected DRGs and make cell suspension at a density of 1 DRG/μL. Using a hemocytometer, determine the cell density. 1 DRG/μL is equivalent to ~5000 cells/μL, including both neuronal and nonneuronal cells. Spotting 2.5–3 μL of the cell suspension

52

Jung Eun Shin and Yongcheol Cho

yields approximately 6000 neurons survived until DIV 7, which should make a suitable density of axons for axon imaging. Cell numbers in a spot can be adjusted by changing the density or volume of cell suspension spotted. If spots dry too fast during the subsequent incubation time, increase the spotting volume by diluting the cell suspension. 13. During the spotting procedure, avoiding the drying of cell spots is important. Complete drying of spots kills the cells while partial dryness may result in meandering axonal growth. To achieve sufficient humidification of the incubator, placing two water pans in the incubator may be helpful. 14. As an alternative to the spot culture described in this chapter, DRG explant cultures can be made by simply placing each dissected DRG tissue in each well containing culture medium by forceps or by pipetting. Use aphidicolin instead of 5-fluoro20 -deoxyuridine and uridine as a mitotic blocker for the explants. Spot cultures are optimal when individual cell bodies are to be treated homogeneously with, for example, lentivirus. 15. Examples of treatment are application of a chemical compound and knockdown or overexpression of a gene using lentivirus. It is important to design experiments with an appropriate control, such as solvent used for chemical treatment or control lentivirus. Because axons may spontaneously degenerate by treatment, consider allotting wells for an uncut control in each experimental condition. 16. Axotomy can be carried out earlier than DIV 7 once axons have grown sufficiently long to be cut by a blade (e.g., at DIV 5); however, cutting at earlier DIVs tends to lead to faster degeneration time courses. 17. Alternative insults to mechanical injury include growth factor deprivation and chemotherapeutic reagents. NGF-deprivationinduced axon degeneration is a model of developmental cell death. Briefly, DRG cultures (DIV 5) are washed twice with warmed neurobasal medium and replenished with culture medium without NGF, containing NGF-neutralizing antibodies [23]. As a model of chemotherapy-induced peripheral neuropathy, 40–100 nM of vincristine can be added to the culture (DIV 7). Axon degeneration is usually saturated by 24 h after both treatments. 18. Axotomy-induced axon fragmentation occurs mostly synchronously and therefore taking images of the exact same axons over time is not necessary, as long as the number of pictures is sufficiently high and the images shown are representative. However, taking images of regions of interest over time using a motorized stage will improve the efficiency of the imaging procedure.

Axon Degeneration Assay in Cultured DRG Neurons

53

19. Degeneration indices obtained from images with fully fragmented axons are approximately 0.8. Due to the heterogeneity in size and circularity of axon fragments, a degeneration index hardly reaches 1. If necessary, the indices can be shifted by adjusting the size and circularity setting in the macro.

Acknowledgments This work was supported by the National Research Foundation of Korea (NRF) grant to J.E.S. (NRF-2017R1C1B2008356), Health Technology R&D Project to Y.C. (HI17C1459), and BK21 Plus project of the NRF to J.E.S. and Y.C. References 1. Gerdts J, Summers DW, Milbrandt J, DiAntonio A (2016) Axon self-destruction: new links among SARM1, MAPKs, and NAD+ metabolism. Neuron 89:449–460. https://doi.org/ 10.1016/j.neuron.2015.12.023 2. Wang JT, Medress ZA, Barres BA (2012) Axon degeneration: molecular mechanisms of a selfdestruction pathway. J Cell Biol 196:7–18. https://doi.org/10.1083/jcb.201108111 3. Hill CS, Coleman MP, Menon DK (2016) Traumatic axonal injury: mechanisms and translational opportunities. Trends Neurosci 39:311–324. https://doi.org/10.1016/J. TINS.2016.03.002 4. Pan YA, Misgeld T, Lichtman JW, Sanes JR (2003) Effects of neurotoxic and neuroprotective agents on peripheral nerve regeneration assayed by time-lapse imaging in vivo. J Neurosci 23:11479–11488 5. Grisold W, Cavaletti G, Windebank AJ (2012) Peripheral neuropathies from chemotherapeutics and targeted agents: diagnosis, treatment, and prevention. Neuro-Oncology 14: iv45–iv54. https://doi.org/10.1093/ neuonc/nos203 6. Coleman M (2005) Axon degeneration mechanisms: commonality amid diversity. Nat Rev Neurosci 6:889–898. https://doi.org/10. 1038/nrn1788 7. Cashman CR, Ho¨ke A (2015) Mechanisms of distal axonal degeneration in peripheral neuropathies. Neurosci Lett 596:33–50. https://doi. org/10.1016/J.NEULET.2015.01.048 8. Brennan KM, Bai Y, Shy ME (2015) Demyelinating CMT–what’s known, what’s new and what’s in store? Neurosci Lett 596:14–26. https://doi.org/10.1016/J.NEULET.2015. 01.059

9. Burke RE, O’Malley K (2013) Axon degeneration in Parkinson’s disease. Exp Neurol 246:72–83. https://doi.org/10.1016/j. expneurol.2012.01.011 10. Kneynsberg A, Combs B, Christensen K et al (2017) Axonal degeneration in tauopathies: disease relevance and underlying mechanisms. Front Neurosci 11:572. https://doi.org/10. 3389/fnins.2017.00572 11. Stoll G, Jander S, Myers RR (2002) Degeneration and regeneration of the peripheral nervous system: from Augustus Waller’s observations to neuroinflammation. J Peripher Nerv Syst 7:13–27 12. Simon DJ, Weimer RM, McLaughlin T et al (2012) A caspase cascade regulating developmental axon degeneration. J Neurosci 32:17540–17553. https://doi.org/10.1523/ JNEUROSCI.3012-12.2012 13. Simon DJ, Pitts J, Hertz NT et al (2016) Axon degeneration gated by retrograde activation of somatic pro-apoptotic signaling. Cell 164:1031–1045. https://doi.org/10.1016/J. CELL.2016.01.032 14. Geisler S, Doan RA, Strickland A et al (2016) Prevention of vincristine-induced peripheral neuropathy by genetic deletion of SARM1 in mice. Brain 139(Pt 12):3092–3108. https:// doi.org/10.1093/brain/aww251 15. Chen X, Rzhetskaya M, Kareva T et al (2008) Antiapoptotic and trophic effects of dominantnegative forms of dual leucine zipper kinase in dopamine neurons of the substantia nigra in vivo. J Neurosci 28:672–680. https://doi. org/10.1523/JNEUROSCI.2132-07.2008 16. Le Pichon CE, Meilandt WJ, Dominguez S et al (2017) Loss of dual leucine zipper kinase signaling is protective in animal models of

54

Jung Eun Shin and Yongcheol Cho

neurodegenerative disease. Sci Transl Med 9: eaag0394. https://doi.org/10.1126/ scitranslmed.aag0394 17. Welsbie DS, Yang Z, Ge Y et al (2013) Functional genomic screening identifies dual leucine zipper kinase as a key mediator of retinal ganglion cell death. Proc Natl Acad Sci U S A 110:4045–4050. https://doi.org/10.1073/ pnas.1211284110 18. Ferri A, Sanes JR, Coleman MP et al (2003) Inhibiting axon degeneration and synapse loss attenuates apoptosis and disease progression in a mouse model of Motoneuron disease. Curr Biol 13:669–673. https://doi.org/10.1016/ s0960-9822(03)00206-9 19. Sajadi A, Schneider BL, Aebischer P (2004) Wlds-mediated protection of dopaminergic fibers in an animal model of Parkinson disease. Curr Biol 14:326–330. https://doi.org/10. 1016/j.cub.2004.01.053 20. Riederer BM, Barakat-Walter I (1992) Differential distribution of two microtubuleassociated proteins, MAP2 and MAP5, during chick dorsal root ganglion development in situ and in culture. Dev Brain Res 68:111–123. https://doi.org/10.1016/0165-3806(92) 90253-S 21. Araki T, Sasaki Y, Milbrandt J (2004) Increased nuclear NAD biosynthesis and SIRT1 activation prevent axonal degeneration. Science 305:1010–1013. https://doi.org/10.1126/ science.1098014 22. Shin JE, Miller BR, Babetto E et al (2012) SCG10 is a JNK target in the axonal degeneration pathway. Proc Natl Acad Sci U S A 109: E3696–E3705. https://doi.org/10.1073/ pnas.1216204109 23. Ghosh AS, Wang B, Pozniak CD et al (2011) DLK induces developmental neuronal degeneration via selective regulation of proapoptotic JNK activity. J Cell Biol 194:751–764. https:// doi.org/10.1083/jcb.201103153 24. Nikolaev A, McLaughlin T, O’Leary DDM, Tessier-Lavigne M (2009) APP binds DR6 to trigger axon pruning and neuron death via distinct caspases. Nature 457:981–989. https:// doi.org/10.1038/nature07767

25. Miller BR, Press C, Daniels RW et al (2009) A dual leucine kinase-dependent axon selfdestruction program promotes Wallerian degeneration. Nat Neurosci 12:387–389. https://doi.org/10.1038/nn.2290 26. Sasaki Y, Vohra BPS, Lund FE, Milbrandt J (2009) Nicotinamide mononucleotide adenylyl transferase-mediated axonal protection requires enzymatic activity but not increased levels of neuronal nicotinamide adenine dinucleotide. J Neurosci 29:5525–5535. https:// doi.org/10.1523/JNEUROSCI.5469-08. 2009 27. Sasaki Y, Milbrandt J (2010) Axonal degeneration is blocked by nicotinamide mononucleotide adenylyltransferase (Nmnat) protein transduction into transected axons. J Biol Chem 285:41211–41215. https://doi.org/ 10.1074/jbc.C110.193904 28. Sasaki Y, Nakagawa T, Mao X et al (2016) NMNAT1 inhibits axon degeneration via blockade of SARM1-mediated NAD+ depletion. Elife 5. https://doi.org/10.7554/eLife. 19749 29. Gerdts J, Brace EJ, Sasaki Y et al (2015) Sarm1 activation triggers axon degeneration locally via NAD+ destruction. Science 348:453–457. https://doi.org/10.1126/sci ence.1258366 30. Walker LJ, Summers DW, Sasaki Y et al (2017) MAPK signaling promotes axonal degeneration by speeding the turnover of the axonal maintenance factor NMNAT2. Elife 6. https://doi.org/10.7554/eLife.22540 31. Gerdts J, Summers DW, Sasaki Y et al (2013) Sarm1-mediated axon degeneration requires both SAM and TIR interactions. J Neurosci 33:13569–13580. https://doi.org/10.1523/ JNEUROSCI.1197-13.2013 32. Gerdts J, Sasaki Y, Vohra B et al (2011) Imagebased screening identifies novel roles for IkappaB kinase and glycogen synthase kinase 3 in axonal degeneration. J Biol Chem 286:28011–28018. https://doi.org/10. 1074/jbc.M111.250472

Chapter 5 Viral Transduction of DRG Neurons Yo Sasaki Abstract The manipulation of gene expression is an essential tool to study the function of genes or signaling pathways. Uniform and robust gene manipulation is crucial for successful assays. However, neuronal cells are generally difficult-to-transfect cells with conventional DNA/RNA transfection reagents. Therefore, virus-mediated gene delivery is a primary choice for the studies of gene functions in neurons. In this chapter, we will describe the methods for lentivirus-mediated gene expression or knockdown in DRG neurons. Key words Axon degeneration, Lentivirus, DRG, Wallerian degeneration, Neurodegeneration

1

Introduction The study of the molecular mechanism of axon degeneration is initiated by the finding of the mouse carrying the mutant gene called slow Wallerian degeneration (Wlds) [1]. The expression of Wlds significantly delays axon degeneration both in vivo and in vitro [2]. The function of Wlds or the signaling pathways of axon degeneration has been extensively studied in the past two decades [3, 4, 16]. The expression of wild type, mutant, and ortholog genes, as well as gene knockdown, has been critical methods for these studies [5–9]. Among various methods to alter the gene expression in neuronal cells, virus-mediated gene manipulation is a highly efficient, high-throughput, and cost-effective way compared with conventional DNA/RNA transfection, electroporation, or microinjection. The current lentivirus gene delivery system (lentivirus vectors) is derived from human immunodeficiency virus type-1 (HIV-1) [10] after a series of modifications to obtain minimal essential components, safety features, and a high expression of transgenes [11, 12, 15]. In this chapter, we will describe lentivirus vectors consisting of three plasmids (packaging, envelope, and transfer). This lentivirus retains the ability to integrate its transfer plasmid

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_5, © Springer Science+Business Media, LLC, part of Springer Nature 2020

55

56

Yo Sasaki

into genomes of dividing and nondividing cells; therefore, it is an ideal system for long-term gene manipulation in neuronal cells, which do not undergo mitosis. It is pseudotyped with vesicular stomatitis virus G glycoprotein (VSV-G) envelope protein, and a broad range of cell types are targeted by this virus. Transfer plasmids carry a variety of promotors that regulate the expression of transgenes, shRNAs, and gRNAs, which binds to genomic DNA and recruits CRISPR-associated protein 9 (Cas9), which introduces double strand breaks at a specific site defined by the gRNA. The protocol described in this chapter uses a lentivirus transfer plasmid carrying the ubiquitin promoter for protein coding genes or U6 promoter for shRNA for dorsal root ganglion (DRG) sensory neurons. The expression of gRNA under U6 promoter together with Cas9 is an alternative way to manipulate gene expression in DRG neurons [13, 14]. Other types of neurons including superior cervical ganglion (SCG) and cortical neurons can be infected with the same system. However, the promoter in the transfer plasmid may need to be optimized for a specific type of neuron.

2

Materials

2.1 Plasmids (See Note 1)

1. psPAX2. 2. pMD2.G. 3. FUGW. 4. pLKO.1 puro.

2.2

Transfections

1. FuGENE 6 (see Note 2). 2. Opti-MEM I Reduced Serum Medium.

2.3

Cells Culture

1. 293T cells (see Note 3). 2. Tissue culture incubator (37
C and 5% CO2). 3. 37
C water bath. 4. Cell culture medium: Dulbecco’s Modified Eagle’s Medium (DMEM), 10% heat inactivated fetal bovine serum (FBS) (see Note 4), 100 units/ml penicillin, 100 μg/ml streptomycin. 5. Trypsin 0.05% solution: 5.3 mM KCl, 0.44 mM KH2PO4, 4.2 mM NaHCO3, 138 mM NaCl, 0.33 mM Na2HPO4, 5.6 mM D-Glucose, 0.91 mM EDTA, 0.025 mM phenol red, 0.021 mM Trypsine. 6. Polybrene (hexadimethrine bromide) solution: 4 mg/ml polybrene. Sterilize by filtration through 0.22 μm syringe filter.

Lentivirus Mediated Gene Regulation in Neurons

2.4 Virus Concentration

57

1. Lenti-X™ Concentrator. 2. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, and 1.8 mM KH2PO4. 3. Refrigerated centrifuge for 1.5 ml tubes at 12,000  g.

2.5

Biosafety

1. Tissue culture hood. Biosafety level 2 and institutional regulations should be applied to lentivirus handling. 2. Autoclave. Culture plates, tubes, pipets, etc., that come in contact with lentivirus particles should be autoclaved and disposed according to institutional guidelines.

3

Methods The following protocol is the standard procedure for virus packaging in one well of 6-well culture plate. For different size of wells, the number of cells and the amounts of DNA and transfection reagents should be adjusted by the surface area relative to a 6-well.

3.1 Lentivirus Packaging

1. Frozen stock vial of the 293T cell is thawed quickly in a 37
C water bath and plated in the appropriate size of culture plate (follow the instruction of the frozen vial), containing cell culture medium in the humidified chamber at 37
C and 5% CO2. Immediately after the cells become confluent, remove the culture medium and add the appropriate amount of trypsin (enough volume to cover the surface). Incubate the cell culture plate for 5 min in the cell culture incubator and then add an equal amount of culture medium and remove cells by pipetting. Count and plate cells at a density of 1  106 cells in one 6-well containing 2 ml culture medium. 2. Remove 293T cell culture medium from a 6-well (cell density 70–90% confluent and 3–4 days post passage, see Note 3) and add 0.5 ml trypsin. Incubate the cell culture plate for 5 min in the cell culture incubator. Confirm the detachment of cells by gently tapping the plate, and add 0.5 ml culture medium. Single cell suspension is obtained after three to five cycles of gentle pipetting (see Note 5). Count and plate cells at a density of 1  106 cells in one 6-well containing 1.8 ml culture medium. 3. When the cell density reaches 60–80% of confluency (typically 18–24 h after seeding), prepare the mixture of DNA as shown in Table 1. Each of the DNA stock solution is diluted at the concentration of 0.1 μg/μl in H2O prior to mix (see Note 6). 4. Transfer 6 μl of FuGENE 6 (FuGENE 6: DNA ratio 3:1) into 100 μl Opti-MEM in a 1.5-ml tube without touching the wall

58

Yo Sasaki

Table 1 List of plasmids and respective recommended concentrations needed to produce lentiviral particles Plasmid psPAX2 pMD2.G Transfer plasmid

kb 11 5.8 ~10

μg

Concentration (μg/μl)

Volume (μl)

1.2

0.1

12

0.4

0.1

4

0.4

0.1

4

of the tube, mix by gentle tapping, and incubate for 5 min at room temperature. 5. Add the DNA mixture 20 μl into FuGENE 6/Opti-MEM solution, mix by gentle tapping, and incubate for 20 min at room temperature. 6. Add DNA/FuGENE 6/Opti-MEM mixture dropwise to the culture medium. Gently swell the plate three times in one direction on the flat surface and then additional three times perpendicular to the initial direction. Place the plate in the incubator. 7. At 3 days after transfection (on the day of transfection is 0 day), harvest the culture medium containing virus particles and add fresh 1.8 ml culture medium to the cells for the second virus collection. Confirm high transfection efficiency using fluorescent protein markers if applicable (see Note 7). 8. Centrifuge the culture medium (12,000  g for 5 min at 4
C) to remove contaminated cells and debris, and aliquot 900 μl of cleared supernatant in a 1.5-ml tube (two tubes from one 6-well containing 1.8 ml medium). Do not disturb the debris and cells in the bottom of the tube. This cleared supernatant can be directly used for the lentivirus infection to neurons. However, virus concentration (steps 9–10) is recommended to achieve the higher titer and purity. 9. Add 300 μl Lenti-X Concentrator for 900 μl cleared culture medium containing the lentivirus particle, and incubate for a minimum of 1 h to overnight at 4
C. 10. Centrifuge (1500  g for 45 min or 12,000  g for 5 min at 4
C) Lenti-X/culture medium mixtures. Typically, white particles precipitate. Carefully remove the supernatant without disturbing the precipitate. Resuspend the precipitate with 90 μl sterilized PBS and store virus solutions at 80
C (see Note 8). 11. At 5 days after the transfection, repeat steps 7–10 to obtain additional virus solutions (see Note 9).

Lentivirus Mediated Gene Regulation in Neurons

3.2 Lentivirus Tittering

59

In case the transfer plasmid provides fluorescent proteins or drug resistant genes, virus titer can be measured by counting the number of infected cell colonies as described below (see Note 10). 1. Prepare a series dilution of lentivirus: Add 2 μl of concentrated virus into 18 μl of 293T culture medium (#1; 1 μl virus in 10 μl), dilute 5 μl of #1 in 45 μl of medium (#2; 101 μl virus in 10 μl), and prepare up to sixth dilution (#6; 105 μl virus in 10 μl). 2. Plate 103 293T cells in 96-well and immediately add 10 μl of virus dilutions prepared in step 1 (#1 to #6) in the presence of polybrene (4 μg/ml). 3. Add appropriate chemicals for the selection markers (e.g., puromycin) at 3 days after infection if applicable. 4. At 5 days after infection, identify the colonies that are positive for fluorescent proteins or resistant to selection markers. For example, two colonies are identified in wells infected with #5 dilution in step 1, calculated titer is 2/104 ¼ 2  104 colony forming unit (cfu)/μl. A typical titer of 10 virus solution is between 104 and 105 cfu/μl (see Note 11).

3.3 Lentivirus Infection to DRG Neurons

1. Add lentivirus solution to standard DRG culture medium at the MOI (multiplicity of infection; the ratio of virus particles to the number of cells) about 50–100 (see Note 12). 2. In our experience, the medium containing the virus does not have to be changed. However, consider the adverse effects of 293T medium on DRG neurons, especially if the virus solution is not concentrated.

4

Notes 1. psPAX2 (Didier Trono, Addgene plasmid # 12260) provides the minimal components necessary for lentivirus packaging, and pMD2.G(Didier Trono, Addgene plasmid # 12259) provides the envelop protein, VSV-G, necessary for virus binding and uptake by the host cell. FUGW [5] is a transfer plasmid, expressing ubiquitin promoter-driven EGFP and pLKO.1 puro (David Root, Addgene plasmid # 10878) express short hairpin RNA (shRNA) under the regulation of U6 RNA polymerase III [15]. Each component of the lentivirus vector can be prepared by standard DNA preparation procedures. 2. Other DNA transfection reagents are acceptable. To obtain the best transfection efficiency, optimize transfection parameters (e.g., the ratio of DNA:transfection reagents) according to the instructions for each reagent.

60

Yo Sasaki

3. The quality of 293T cell is critical for DNA transfection and lentivirus production. To maintain 293T cells, 70–90% confluent cells are passaged with 1:3 to 1:5 dilutions (e.g., 2 ml cell suspension from one 6-well is diluted to 6–10 ml and transfer the 2 ml diluted cell suspension in one 6-well). 293T cells should become 70–90% confluent after 3–4 days. Cells that do not proliferate as described above often produce low quantities of the virus. Take new frozen 293T cell aliquot when the proliferation rate slows down or after around 30 passages. 4. We used a variety of FBS from different sources and obtained a similar quality of lentivirus when the 293T cells are passaged as described. If 293T cells do not proliferate as described (see Note 3), then test different FBS bathes or providers. 5. If 293T cells are difficult to detach after trypsinization (requires repeated and hard tapping of culture vessels) or difficult to disperse into single cell suspension (requires more than five times pipetting up and down), watch the resultant virus quality. It is good idea to take a new frozen aliquot of 293T cells at this point. 6. The type of transfer plasmid can be over expression (ex. FUGW) or knockdown (ex. pLKO.1 puro). The size of transfer plasmid ranges from about 10kB (FUGW) to 20kB depending on the size of the promoter and gene. The amount of transfer plasmid can be increased according to the size (ex. 0.8 μg with 20 kb transfer plasmid). The volume of FuGENE 6 should be adjusted to keep FuGENE 6:DNA ratio 3:1. 7. The key to produce a high titer lentivirus is a high co-transfection efficiency of three plasmids, packaging, envelope, and transfer. In our experience, the quality of 293T cell is important. Pay close attention to the proliferation rate and density (see Note 3). It is also important to prepare a large amount of packaging and envelope plasmid and aliquot to obtain stable results. The aliquots are stored at 20
C and, once thawed, the aliquots is stored at 4
C for up to 4 weeks. 8. The centrifuge condition recommended in the manufacture’s protocol is 1500  g for 45 min at 4
C. However, we routinely use the condition 12,000  g for 5 min at 4
C. Aliquot the concentrated virus solution to avoid more than five cycles of freeze and thaw. 9. Lentivirus particles are produced by 293T cells immediately after the co-expression of three components. The first collection of the virus solution at 3 days after transfection contains more virus particles than the second collection (at 5 days), which still contains good amount of virus particles. The virus

Lentivirus Mediated Gene Regulation in Neurons

61

collection can be done between 2 to 5 days post transfection with peak production at around 3 days. 10. If there is no selection markers or fluorescent proteins, virus titer can be determined by using commercially available kits that quantify the amount of p24, a component of virus capsid. 11. The titer of lentivirus will decrease if the transgene is toxic to 293T cells. There are several strategies to overcome this issue. Use pharmacological/chemical treatments to suppress the toxicity or co-express the proteins that block the toxicity of transgenes. Consider applying inducible promoter to regulate transgene expression as well. When the size of transfer plasmids is increased, the virus titer is decreased. If this is an issue, then increase the virus concentration factor up to 100-fold and increase the MOI. 12. The optimal MOI can vary depends on multiple factors including the type of transfer vectors, cells, and expressed genes. Although the current transfer plasmid with the ubiquitin or U6 promoter is highly efficient for DRG neurons, polybrene (4 μg/ml) can be used for infection of difficult-to-express genes or other type of cells (e.g., mouse embryonic fibroblast (MEF)). Polybrene enhances virus absorption to target cell membranes. Polybrene is added to culture medium at the same time of virus addition. Replace medium without polybrene if it is toxic to host cells.

Acknowledgments This work was supported by the National Institutes of Health (Grant RO1AG013730 (Jeffrey Milbrandt: J.M.); RO1NS065053 and RO1NS087632 (J.M. and Aaron DiAntonio: A.D.)). The author thanks J.M., A.D., and members of the J.M. and A.D.’s laboratories for fruitful discussions. References 1. Coleman MP, Conforti L, Buckmaster EA et al (1998) An 85-kb tandem triplication in the slow Wallerian degeneration (Wlds) mouse. Proc Natl Acad Sci U S A 95:9985–9990 2. Coleman MP, Freeman MR (2010) Wallerian degeneration, WldS, and Nmnat. Annu Rev Neurosci 33:245–267 3. Conforti L, Gilley J, Coleman MP (2014) Wallerian degeneration: an emerging axon death pathway linking injury and disease. Nat Rev Neurosci 15:394–409 4. Gerdts J, Summers DW, Milbrandt J et al (2016) Axon self-destruction: new links

among SARM1, MAPKs, and NAD+ metabolism. Neuron 89:449–460 5. Araki T, Sasaki Y, Milbrandt J (2004) Increased nuclear NAD biosynthesis and SIRT1 activation prevent axonal degeneration. Science (New York, NY) 305:1010–1013 6. Sasaki Y, Araki T, Milbrandt J (2006) Stimulation of nicotinamide adenine dinucleotide biosynthetic pathways delays axonal degeneration after axotomy. J Neurosci 26:8484–8491 7. Gilley J, Coleman MP (2010) Endogenous Nmnat2 is an essential survival factor for

62

Yo Sasaki

maintenance of healthy axons. PLoS Biol 8: e1000300 8. Gerdts J, Brace EJ, Sasaki Y et al (2015) SARM1 activation triggers axon degeneration locally via NAD+ destruction. Science (New York, NY) 348:453–457 9. Di Stefano M, Nascimento-Ferreira I, Orsomando G et al (2015) A rise in NAD precursor nicotinamide mononucleotide (NMN) after injury promotes axon degeneration. Cell Death Differ 22:731–742 10. Naldini L, Trono D, Verma IM (2016) Lentiviral vectors, two decades later. Science (New York, NY) 353:1101–1102 11. Miyoshi H, Blo¨mer U, Takahashi M et al (1998) Development of a self-inactivating lentivirus vector. J Virol 72:8150–8157 12. Lois C, Hong EJ, Pease S et al (2002) Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science (New York, NY) 295:868–872 13. Walker LJ, Summers DW, Sasaki Y et al (2017) MAPK signaling promotes axonal

degeneration by speeding the turnover of the axonal maintenance factor NMNAT2. ELife 6:545 14. Summers DW, Milbrandt J, DiAntonio A (2018) Palmitoylation enables MAPKdependent proteostasis of axon survival factors. Proc Natl Acad Sci U S A 115:E8746–E8754 15. Jason Moffat, Dorre A. Grueneberg, Xiaoping Yang, So Young Kim, Angela M. Kloepfer, Gregory Hinkle, Bruno Piqani, Thomas M. Eisenhaure, Biao Luo, Jennifer K. Grenier, Anne E. Carpenter, Shi Yin Foo, Sheila A. Stewart, Brent R. Stockwell, Nir Hacohen, William C. Hahn, Eric S. Lander, David M. Sabatini, David E. Root, (2006) A Lentiviral RNAi Library for Human and Mouse Genes Applied to an Arrayed Viral HighContent Screen. Cell 124 (6):1283–1298 16. Josiah Gerdts, Daniel W. Summers, Jeffrey Milbrandt, Aaron DiAntonio, (2016) Axon SelfDestruction: New Links among SARM1, MAPKs, and NAD+ Metabolism. Neuron 89 (3):449–460

Chapter 6 Planning and Analysis of Axon Degeneration Screening Experiments Lyndah Lovell, John Bramley, and William Buchser Abstract A network of intersecting molecular pathways interacts to initiate and execute axon destruction. Maximum protection against axon degeneration likely requires more than manipulation of a single target. Here, we describe the process of designing a high-throughput arrayed screening assay for the identification of key factors responsible for axon destruction and/or protection. First, we go over some existing screens in the literature, then discuss the planning, tracking, analysis, and statistics around such a screening experiment. Prioritization of perturbations may allow laboratories to cost-effectively explore the process of screening. We also present the pairing of a combinatorial drug screen with a machine learning algorithm, predicting how to best modulate neurodegenerative and neuroprotective components. Key words Axon degeneration, Analysis, Screening, Discovery, High-throughput, Hits, Nonhits, Support vector machine learning, Targets, Antitargets

1

Introduction Axon degeneration is responsible for the progression of neurodegenerative diseases, neuropathy, and impaired function after neuronal injury [1–3]. In order to protect against the functional deficits of these conditions, an important goal is to protect neurons from the degeneration program altogether and prevent major axon loss. High-throughput screening experiments have been undertaken to advance both basic and therapeutic knowledge surrounding the mechanism behind axon degeneration. In this introduction, we provide some background and cite papers that will be helpful for a reader wanting to run their own axon degeneration screening experiment. Many genes and metabolites have already been implicated in the degeneration process such as calcium regulators, Dserine, nicotinamide adenine dinucleotide (NAD)-ases, kinases, the SARM1 gene, and more [4–8]. Recent work has used RNA interference (RNAi) [9] or CRISPR/Cas9 system for high-throughput gene silencing.

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_6, © Springer Science+Business Media, LLC, part of Springer Nature 2020

63

64

Lyndah Lovell et al.

Standard “arrayed” screens use multiwell plates where each well contains a different perturbation (a drug, shRNA, gRNA, or plasmid). Two other variations on screening may be of interest. First, using more sophisticated substrates may improve the physiologic nature of the screening. In addition to 3D substrates like Matrigel and beads, there has now been success in a multiple sclerosis screen utilizing micropillar arrays (in place of axons) which successfully identified eight individual FDA-approved compounds that enhanced oligodendrocyte differentiation [10]. Second, pairing screens with predictive learning algorithms can augment the discovery of new compounds and compound-combinations that enhance desired effects. For example, a phenotypic screen joined with support vector machine (SVM) learning algorithms revealed cotreatment with kinase inhibitors had synergistic effects on promoting neurite outgrowth of hippocampal neurons [11]. Also, RNAi genome-wide screening has identified potential therapies against kinase and polypeptide targets responsible for conditions such as retinal ganglion cell death (a contributor to glaucoma) and Parkinson’s disease (PD) [12, 13]. The first screening experiment to enable automated imaging and use it in a drug screen of axon degeneration is still the basis for most screening experiments [5]. Another related screen examined degeneration specifically at the distal part of the axon (synaptic terminal) [14]. A related series of screens combined degeneration and regeneration to examine regeneration after injury [15, 16]. Neuronal screening is particularly difficult (compared with cell line-based screens), and some recent progress in the field is reported in this journal’s special section [17]. Finally, an in-depth chapter about high-content imaging has also been published [18]. Here, we describe how to design a high-throughput screen for targets related to axon degeneration. Nuances of the design such as the cell type, experimental assay, and type of injury will depend on the researcher’s condition of interest. This chapter also gives special attention to polypharmacology through execution of a combinatorial drug screen.

2

Materials

2.1 Data Storage and Management

1. Image and analysis storage: A suitable file storage option will need to be selected to store images and analysis generated throughout the screen. Network attached storage (NAS) and file servers allow quick access to stored data. Cloud based options such as Amazon AWS (https://aws.amazon.com/), Box (https://www.box.com/), DropBox (https://www. dropbox.com/), Google Drive (https://www.google.com/ drive/), and Microsoft Azure (https://azure.microsoft.com/) will likely become more viable options as the cost and speed of cloud-based storage improves.

Axon Degeneration Screening

2.2 Data Analysis Requirements

65

1. Computational requirements: The computational demands of planning and analyzing the screen can be met by most modern computers. 2. Data Analysis Software: The methods outlined here can be performed using standard tools provided by Microsoft Office (Excel, Access) or alternative open source spreadsheet software (Libre Office, Open Office). 3. Advanced Data Analysis Tools: Business intelligence tools such as Tibco Spotfire (https://spotfire.tibco.com/) and Tableau (https://www.tableau.com/) allow for more sophisticated data analysis and visualization. Scripting languages such as R (CRAN) can also enable more advanced and streamlined data analysis and pipelining.

3

Methods

3.1 Planning the Screen

In designing a drug screen, the utilization of an appropriate in silico simulation can provide a powerful tool in assessing the feasibility of undertaking the screen. We have developed a simulation that will enable the user to vary and assess a series of parameters that can be useful in providing insights into the design and execution of the screen. The simulation presented in this chapter has been optimized for arrayed screens in the setting of axon degeneration. Reference degeneration distributions were based on real screening data related to [19] (Fig. 1). Follow the protocol below to plan out your screen. 1. Start with a background distribution. This is what the majority of the screening results will look like, assuming that most perturbations are negative for your phenotype. We provide a real background distribution for mouse dorsal root ganglia (DRG) neurons 24 h after axotomy in Fig. 1. This will help

Fig. 1 Distribution of degeneration indices of axons following physical injury (axotomy). The degeneration index (DI) ranges from 0 (completely intact) to 1 (completely fragmented). This distribution was generated through the aggregation of experimental datasets containing degeneration indices from physically injured mouse DRG axons, 24 hours after axotomy [19]. An in-depth explanation of axotomy in vitro and DI values can be found in the 4th Chapter

66

Lyndah Lovell et al.

you think about the frequency of false positives or to create your own simulation. 2. Optimize the assay over several days. Specifically, (a) Choose a negative and positive control. Negative controls are typically scrambled shRNA (for RNAi), nontarget gRNA + Cas9 (for CRISPR screens), or vehicle (for drug screens) (see Note 1). We recommend using SARM1 knockout or knockdown as a positive control for screens which try to find genes protecting from degeneration after injury, as previous SARM1 knockout experiments have resulted in axon protection in mice and flies after axonal injury [19–21]. (b) Run an experiment under planned screening conditions. Ensure replicates of both positive and negative controls. Repeat the experiment varying: layout of controls on the plate, day, incubator, animal (if primary cells), and technician. Ideally, use at least two replicates for each of the conditions listed. (c) Analyze the experiment. (d) Calculate Z factor: 

   standard deviation among all the positives absolute difference of the means Z factor ¼ 1  3  þSD among all the negatives l

A larger Z factor is better with a score above 0, meaning screening is feasible. A Z factor below zero means that the noise associated with the experiments causes the positive and negative to overlap too much. A lower Z factor likely requires further optimization.

(e) Calculate plate/edge effects: l

It is common for there to be temperature and gas exchange differences across multiwell plates. Plot your primary data averaged per column (to look for effects by rows) and per row (effects across the columns). You can also plot the data per well in a grid, to see if the edges of the plate are different than the center.

l

If the effect is consistent across days, then a normalization method can be devised to account for it.

(f) Record the batch effects by day, incubator, animal, or technician. (g) Calculate effectivity of positive control: Can be estimated as the median of the positive controls (if you have enough data than the mode is better). (h) Calculate penetrance of positive control (fraction of positive controls that display the phenotype). A quick

Axon Degeneration Screening

67

approximation is to set a threshold at the halfway point between the mode of the negative control and the mode of the positive control. Percentage of positive controls that cross the threshold is the estimated penetrance. 3. If you are unable to complete the optimization experiments, you can use our estimates below. They are attempting to be on the lower end of normal, so that the screen is conducted a little more stringently. (a) Effectivity (expressed as degeneration index (DI)): 0.25. (b) Penetrance: 40%. 4. Decide on how many experimental replicates are needed. Different versions of a perturbation that impact the same gene can be counted as replicates. For example, in a CRISPR screen, four guide RNAs can count as four replicates. Estimated number of replicates required ¼ 1.6/penetrance. Round up, and do not go below 2. 5. Decide on the best method for aggregating replicates (or shRNAs/gRNAs against the same target). If looking for hits that have penetrance below 0.55, taking the minimum or maximum (Min/Max) is preferred. For higher penetrance, mean is preferred. For more information on the simulation that yielded this recommendation, see Note 2. 3.2

Prioritization

Running a large screening experiment is costly and timeconsuming. Yet, large screens that examine hundreds or thousands of perturbations often find redundant perturbations that are not particularly useful for identifying novel targets. In drug screens, compounds with similar activity profiles can be grouped together and one representative member can be tested, thus maximizing variance of profiles and the amount of information resulting from the screen. The methods of assigning these groups and lead compounds vary depending on the research question of the researcher. We describe below the several compound grouping methods for the researcher to consider. 1. Correlation Coefficient and Similarity Matrix: Iteratively compare activity profiles between pairs of compounds to measure similarities. The activity profile is the percent of inhibition a compound has against its target. To compare these activity profiles, generate a Similarity Matrix from the correlation coefficients between each compound. Correlation coefficients and a subsequent Similarity Matrix can be made in Microsoft Excel using the Data Analysis option (from the Analysis ToolPak Excel Add-in) or using the CORREL() function. The Data Analysis Correlation selection is faster than the CORREL() function in Excel (see Note 3). Here, we used Microsoft Excel; however, any statistical analysis software can be used

68

Lyndah Lovell et al.

according to the researcher’s preference. After the Similarity Matrix is made, select a single compound to represent several compounds that are closely, positively correlated, in other words, whose correlation coefficient meets some threshold close to +1. Repeat until all selected representative compounds represent distinct activity profiles. 2. Euclidean Distance: Measures the distance between pairs of activity profiles (from above) for each compound in multiple dimensions. For compound pairs that have some minimum threshold distance between them, a single compound is selected to serve as the representative. Repeat until all selected compounds represent activity profiles that are most dissimilar. 3. Maximize the Variance: Measure the variance in the activity profiles and select the compounds with the highest variance to use in the phenotypic screen. The compounds with the greatest variance are the compounds that provide the most unique information and thus minimize redundancy and maximize the amount of information tested in the phenotypic screen. Additional tools useful for grouping algorithms are as follows:

3.3 Management and Quality Control

l

Principal component analysis (PCA).

l

Nonnegative matrix factorization (NNMF).

l

Single value decomposition (SVD).

Two factors make management different in screening experiments compared to other “standard” laboratory experiments. These experiments occur over a longer span of time and are often carried out by multiple people. This makes management, tracking, and a system to account for batch effects very important. Here, we will mention a few specific points in this regard. 1. Reagent lot procurement. Try to plan the amount of reagents you will need for the screen, and preorder or reserve enough of the same lot. This is especially important for tissue-culture medias and supplements, probably less important for postfixed reagents, but if possible, it can help maintain stability of the screen. 2. Logging procedures. Since the experiment is often shared by several researchers, standard lab notebook procedures often become unmanageable. Set up a Laboratory Information Management System (LIMS), or at least a shareable document log (Google Docs or Google Sheets for example). 3. On-plate tracking system. In addition to a shared electronic log, printing out labels that have tracking information for plate

Axon Degeneration Screening

69

treatments will help efficiency. In addition to the standard ID, include check boxes for different steps of the process. 4. File management. Raw imaging data and associated metadata should be carefully managed, ideally on a fileserver. Be sure to implement business rules around how the folders are named and how the metadata is connected. The metadata is information about what the source of the neuron were (animal, technicians), which reagents were used (if they are from different lots), which treatments/perturbations were applied, etc. 5. Quality control. After acquiring the data, make sure to use the on-plate positive and negative controls to ascertain whether the screen is still working correctly. Another important metric to track is the cell health and number. Be on the alert for sudden drops in these metrics. A hard threshold can be set (and made known in the eventual methods section), for which plates will be included in the final analysis. 6. Weekly or biweekly checking for batch effects or plate effects. The first few weeks of screening can often be plagued by protocol drift, and this can be avoided by doing an optimization screen as suggested earlier. Recurring batch effects and other nontreatment-related effects are often caused by protocol drift. It may be necessary to discard the data from the first set of screening assays and write them off as training, then continue once the methods are locked-in. 3.4

Screen Analysis

Analysis of primary screening and discovery projects are ultimately simple. The word “screen” has the same meaning as “filter,” with the goal of a screening experiment to filter a large set of perturbations into a smaller subset. Screens allow the researcher to start with a large list of perturbations (all genes in a given genome, for example) and reduce that list to a smaller set, more likely to be involved in the phenotype-of-interest. The smaller the list (called a “hit list”), the fewer false positives that will be found (good) but the fewer true positives will also be found (bad). A larger hit list is more expensive, due to a larger test set, but it is more likely that a true hit will be carried along for validation. We advise approaching a primary screen to maximize sensitivity (retain the most true positives). This is different than a standard experiment which tries to maximize specificity (removes false positives). More on this idea, a discussion of false discovery rate (FDR) and associated statistics is described in Note 4. Below, we use simulations based on actual axon degenerationscreening data (same as Subheading 3.1) to allow the researcher to optimize the cost/benefit ratio. This simulation generates data used to calculate sensitivity and specificity based on 2, 3, or 4 replicates and different methods of hit detection.

70

Lyndah Lovell et al.

At intervals or once the primary screen has finished, follow the steps below to choose your “hit list” to take to validations and follow-up experiments. This method assumes you have followed the steps from the previous chapter to assess the DI. This is known as your analysis pipeline. We will outline a pipeline framework but leave implementation open for optimal customization. Most labs write custom code (R, Python, C#) or utilize commercial programs like Partek Flow for analysis. 1. Apply normalization. (a) Leave the raw analysis files in their managed folders. Avoid copying and moving files once file management rules are implemented. (b) Choose a normalization method. It should eliminate batch effects and scale the data, so that each batch is comparable. The methods listed below can be used in step 3 to find the best normalization method. Once the best method is established, it should be adhered throughout the screen. l

Norm 1. Divide each data point by the average of the negative controls.

l

Norm 2. Divide each data point by the average of all the perturbations on the plate (except the positive controls).

l

Norm 3. Scale the data, so that the negative controls fall at 0 and the positive controls fall at 1 (when looking for a prevention of degeneration, you would make the positive controls fall on 0 and the negative controls fall on 1).

(c) Create a new file, called “adjusted data,” that lists the normalized degeneration indices. 2. Compile the analyzed data. (a) Use a script (or rule set) to read the adjusted analysis files in place and add the results to a new “compile” file. You will end out having several batches of compilation, so keep track of them and add them to your overall pipeline. (b) Controls (negative and positive) should be named differently for different plate sets when compiling. For example, if you have 20 96-well plates for the entire screen, you will have 20 sets of negative controls and 20 sets of positive controls. (c) Keep track of the Batch ID when compiling (i.e., what source plate or source experimental batch this set of perturbations came from).

Axon Degeneration Screening

71

(d) Use a pivot table or other function to lay out the data for easy analysis (see Fig. 2). 3. Quality control the compilation. (a) This should only be necessary early on but could be performed at regular intervals. (b) Check for the success of the normalization. To visualize, plot two of the replicates together and see how close they fall to a perfect diagonal. Whichever method of normalization that puts the replicates closest to the diagonal is the better. (c) Batch effects. This can be done by plotting the data points in order (prenormalization). You will see a whole set of points close together that are lower or higher, if there are large batch effects. If you find that you do have large batch effects, ensure that your normalization scheme removes them (they should no longer appear different in the normalized data). 4. Rank perturbations. Once the data is compiled and normalized, it should look like a single experiment with many perturbations. Aggregate the replicates (or the gRNAs/shRNAs) based on the method chosen in Subheading 3.1. Mean is the safest, but Min/Max may allow for higher sensitivity if suspected penetrance is low. As we mentioned before, while p value is better to maximize specificity at the expense of sensitivity, effect size measurements such as Mean or Max/ Min are usually better in a screening setting. In a spreadsheet, you can apply the aggregation and then sort by the aggregated metric (Fig. 2). The sorted perturbations are now ranked. 5. Draw a waterfall plot. Visualize the ranked list to see the overall spread of the data (Fig. 3). This is also useful to get a feeling of the noise and to start visualizing which of the perturbations will be chosen as hits. 6. Decide on a threshold. We recommend trying several thresholds and testing your screen’s sensitivity to picking up the positive controls. In Fig. 2 for example, using the aggregation method of “mean” and a threshold of 0.4, we will get a sensitivity of 100% (all positive controls would be included). Ultimately, a more inclusive threshold is better, since it will recover more true hits, but this will create a more expensive secondary screening endeavor. 7. Create a “HitList” of perturbations. Based on the threshold above, make your hit list of the top perturbations to proceed with in additional experiments. Sometimes the bottom hits are also of interest (antihits) (see Note 5).

72

Lyndah Lovell et al.

Fig. 2 Example of compiled degeneration indices for four replicates, sorted by mean. The equation to calculate the mean, median, and min is placed just to the right of the replicates. Additional metadata may be added after the Perturbation column (Batch ID is added here for example). Use the sorting function to sort the data by the column chosen in the previous step, so the most interesting hits rise to the top (usually with the positive controls)

Axon Degeneration Screening

73

Fig. 3 Simulated screening results of 100 perturbations, ranked by the mean degeneration index (DI). Each set of points across the plot represents a simulated experiment with four replicates. The black line displayed in the upper panel represents the average DI for that experiment. This chart demonstrates that as the rank increases, the average DI of the perturbation decreases, potentially indicating a hit. Two different styles are shown, the upper panel as box plots and the lower panel as a plot of the raw data (all four replicates)

8. Reproduce the HitList reagents. Recreate the perturbations from the original source or a new source. This means reordering a new batch of the hit drugs or remaking the RNAi/gRNA reagents. The experimenter will usually order larger quantities to enable additional experiments. This step is important since it eliminates the hit arising from a metadata error of mistaken identity. 9. Proceed with a secondary/validation experiment. Now use these hit perturbations in the same screening pipeline, ideally with additional replicates. Make sure to include enough negative controls since the data may be skewed to the positives at this stage. Validated hits that come out of the secondary screen

74

Lyndah Lovell et al.

are sometimes called “leads.” The experimenter should proceed with these leads as he or she sees fit but should always be skeptical and examine alternative explanations for them (e.g., could the hit actually be caused by off-target effects and not by the expected gene target). 3.5 Combinatorial Drug Screening

3.5.1 Phenotypic Screen

Axon degeneration is a highly conserved mechanism where multiple molecular routes can initiation the destructive program. Due to the ubiquitous interactions of molecular targets, simultaneously inhibiting or sparing certain molecular targets can have additive, or possibly synergistic, effects on axon protection [11]; this is best achieved through combinatorial screens. Designing a combinatorial drug screen comprises two major stages: applying prioritized compound treatments in your phenotypic screen to serve as training for your algorithm and selecting/writing an algorithm suited for predictions on your specific dataset. The results of the singlecompound phenotypic screen will guide the best combinedcompound treatments for the combinatorial phenotypic screen. Your algorithm can also make predictions on untested compounds based on the results of the single-compound screen and direct compound combinations in the combinatorial screen. Here, we describe a phenotypic screen of kinase inhibitors on zebra finch embryonic retinal ganglion cells, and a binary SVM learning algorithm. 1. Harvest the primary tissue for neuronal cell culture in your phenotypic screen (see Note 6). Allow for axon outgrowth (see Note 7). 2. Assign control and treatment wells. Add appropriate concentrations of prioritized compounds to wells and replicates. A dose response can also be performed; final concentrations of 32 nM, 160 nM, 800 nM, 4 μM, and 20 μM for each compound can be used [11]. 3. Injure the axons and use microscopy to measure the DI [4]. Establish a threshold for “protected” and “nonprotected” axons compared to controls. 4. Classify compounds of protected wells as “hits,” and compounds of nonprotected wells as “nonhits.” 5. For the combinatorial screen, gather all compounds classified as “hits.” Coinhibition with multiple hit compounds may create a synergistic protective effect not seen upon applying a single compound [22].

3.5.2 The SVM

Machine learning algorithms have been increasingly useful for drug discovery and identifying relevant biological targets. By inputting large datasets already examined by the researcher, the algorithm is

Axon Degeneration Screening

75

trained and learns to identify relationships within the data to predict classifications of new, previously unstudied compounds. One such popular predictive classifier is the SVM; with the capability to make binary or multiclass classifications, the algorithm is attractive for multidimensional large datasets, and SVMs maintain high accuracy and efficiency by building off of previously learned datasets, updating by incrementally training on new “chunks” of data [11, 23, 24]. SVM is usually implemented within R or MATLAB. The initial high-throughput phenotypic screen of kinase inhibitors is a testing paradigm that provides compound classifications (hit or nonhit). A SVM learning algorithm then uses these classifications from the phenotypic screen and the inhibitor profile of the compounds to predict the classifications of new compounds never tested. In other words, while the phenotypic screen provides information on compounds that will protect or damage axons, the algorithm can deduce the targets that are biologically relevant to the axon degeneration program [11]. Inhibition of some kinases may lead to more degeneration compared to controls; in this case, these kinases may be beneficial to axon preservation and should not be inhibited. Inhibited kinases that result in protected axons are “targets,” while inhibited kinases of nonprotected axons are “antitargets.” As each compound may be profiled against hundreds of kinases, the algorithm learns to recognize patterns in kinase inhibition across multiple dimensions to illuminate the kinases and levels of inhibition necessary for axon protection. Below, we describe how to build an SVM using binary classifiers (hit or nonhit) of kinase inhibitors and how this model provides new targets and antitargets relevant to the axon degeneration pathway (see Note 8). MATLAB was used for this SVM model; however, any program language desired by the researcher may be used with these core principles (see Note 9). 1. Hit/Nonhit binary classes: From the phenotypic screen, assign all hit compounds as “1” and all nonhit compounds as “0.” Add these classifications as an additional column in the kinase inhibitor profile matrix. 2. Kinase profiles and training and testing set: Read the entire kinase inhibitor profile matrix with the hit and nonhit classes into the algorithm (see Note 10). The training set consists of the kinase inhibition profile and hit classifications of the associated compounds. The testing set only contains the kinase inhibition profile, blinded to the hit classifications of the compounds. 3. Cross-validation: Generate the training and testing sets. Our model utilizes a tenfold cross-validation where the kinase inhibitor profile matrix is divided into ten equal sections. In the SVM train feature, the program iteratively moves through each section, where 9/10ths of the sections are used to train

76

Lyndah Lovell et al.

the SVM (training set), and 1/10 is used to test the accuracy of the SVM learning (testing set) (see Note 11). During the training phase, the SVM recognizes patterns among kinases, degrees of kinase inhibition, and hit classes. Here, the algorithm develops a hyperplane and support vectors, so that inhibition data points that fall on either side of the hyperplane will fall into the hit or nonhit category. During the testing phase, the model uses the hyperplane developed in training to predict hit classifications on the set of compounds in the testing set based only on their kinase profiles. 4. Accuracy, sensitivity, and specificity: Measure the quality of your final SVM model by calculating accuracy (correctly predicted hits and nonhits/total compounds), sensitivity (correctly predicts hits/total hits), and specificity (correctly predicted nonhits/total nonhits) (see Note 12). 5. Identifying newtargets and antitargets: The SVM now has high performance and can be used to classify new compounds not previously tested in a phenotypic screen. Given their kinase inhibition profile alone, the SVM can predict if a new compound is likely to be either a hit and protect axons after injury or a nonhit and contribute to axon destruction. Consequently, the SVM also reveals that compound substrates are either targets or antitargets. Kinases that are more frequently inhibited by hit compounds are targets, while kinases that are more frequently inhibited by nonhit compounds are antitargets. The degree to which kinase targets and antitargets are inhibited is also important here (see Note 13). In this way, the SVM begins to deconvolve the roles of these kinases in the axon degeneration pathway. For a subsequent combinatorial screen, new compounds not previously tested in the initial phenotypic screen but that were predicted hits can be combined and serve as new treatments, thus manipulating interactions of targets and antitargets more robustly—taking us closer to unraveling the degeneration network. Overall, the SVM is a valuable tool to augment the highthroughput screening process. By recognizing patterns in large screening datasets, the algorithm gains the ability to make predictions on novel compounds’ effects on axon degeneration and the importance of their targets in the degeneration pathway. These imputations efficiently uncover important components where discovery through phenotypic testing alone may not have been feasible. Particularly in a combinatorial view, the SVM not only identifies significant compounds and their targets but also sheds light on how these components may interact given their inhibition profiles. The interactions of these players can be new areas of focus and elicits further study to elucidate the full mechanism of axon degeneration.

Axon Degeneration Screening

4

77

Notes 1. Many types of perturbations are available for the researcher performing a screen. These include increased expression (plasmid overexpression, CRISPR knock-in), decreased expression (RNAi, CRISPR disruption), or pharmacologically interact (by dosing in a chemical compound). Of these methods, only CRISPR produces cells that have a permanent change, the other methods (plasmid overexpression, RNAi, and drugs) only temporarily change the cell. The negative controls establish a baseline for which to compare treatments that may have a protective effect or accelerate degeneration of axons. For axon degeneration, using a scrambled shRNA (for RNAi) or a nontarget gRNA + Cas9 (for CRISPR screens) are the standard choice for baseline of no treatment (negative controls). Using a vehicle for drug screens (usually DMSO) is also standard. 2. The following figures (Figs. 4 and 5) provide further insights into planning and also are generated using the output of the simulation. 3. A generic Similarity Matrix created from Microsoft Excel Data Analysis feature is provided for four compounds (Fig. 6). The Data Analysis ToolPak is an Excel Add-In. If this Add-In is not yet activated in the researcher’s Excel program, please refer to Microsoft Excel Support to learn how to activate this feature. After opening the Data Analysis option, you will be prompted to select your desired analysis tool (Correlation) and input the activity profile data. The results are displayed in Fig. 6. 4. A brief note about FDR, or false discovery rate. In an experiment where one is hypothesis testing with multiple conditions (does 1 of 20 things significantly affect an outcome?), then the standard alpha of 0.05 is inadequate (0.05 ¼ 1/20, so one is expected to be significant). Usually, Bonferroni or Benjamini Hochberg is used to adjust the alpha level of significance. While this can be done, we advise against it. The purpose of a screen is NOT to determine whether something is significantly different but to decide whether to advance it to another round of validation. Selecting significant hits using Benjamini Hochberg will yield very high specificity, but many potentially interesting hits can be missed in the process. If you only use a statistical test to tell you about the best possible hit, it could be that this hit is a false positive, and a lot of energy will be wasted on it. If instead, the researchers select a set of putative hits and validate all of them, a set of true positives is more likely to be discovered with less wasted resources in the end. 5. Most screens use assays that are optimized to see a particular type of result. An axon degeneration screen where all axons are

78

Lyndah Lovell et al.

Fig. 4 Level of specificity across four different methods of threshold selection at a sensitivity of 90%. This figure works to illustrate the advantages and disadvantages of four different thresholding methods. Using a t test (lower right) results in a significant decrease in specificity when compared to other methods shown. The minimum value (lower left) gives the highest specificity but also has most variation across the number of replicates. The mean (upper left) and median (right) display similar results with a slightly lower specificity when compared to the minimum but are much less sensitive to the number of replicates. Examining this output is valuable in selecting a threshold, so that the screening parameters can be optimized with regard to specificity and number of replicates required

cut is good at finding perturbations that can rescue the axons, since most will degenerate. It is possible to design screens that are able to find hits in both directions. In the case of axon degeneration, this could be perturbations that both protect or accelerate axon degeneration. If there is enough dynamic range in the assay, those other hits could also be discovered. 6. Embryonic dissection day: Mouse tissue and zebra finch eggs are quite small and fragile. Harvesting retina tissue at E9 or E10 allows for a larger embryo and easier manipulation of the eye and subsequent retina whole tissue extraction but avoids risk of hatching, which usually occurs at E11 or E12. Zebra finches work well in axon degeneration translational studies,

Axon Degeneration Screening

79

Fig. 5 Specificity of different methods of thresholding across a penetrance of 40% and 66%. As shown in Fig. 4, using the minimum value for thresholding gives the highest specificity. However, the increase in specificity is less pronounced as penetrance increases. The t test remains the least useful method with the lowest specificity under either penetrance. In designing a screen with a high predicted penetrance, the method selected for thresholding will have less of an impact on specificity, while screens with a low predicted penetrance will gain a higher specificity when using the minimum value to establish a threshold. Simulated data such as that shown in Figs. 3 and 4 allow the screen designer to select the most appropriate metric for thresholding to optimize the screens parameters

Fig. 6 General Similarity Matrix created using Microsoft Excel Data Analysis ToolPak. Any compound compared to itself will have a correlation coefficient of 1. The self-comparisons fall along the diagonal. The upper half of the matrix is omitted because these values are the same as the values in the lower half. Compound 1 and Compound 2 have a correlation coefficient of 0.94 and therefore have similar activity profiles, thus either Compound 1 or Compound 2 can be selected to represent the pair in the phenotypic screen. The same principle applies to Compound 3 and Compound 4 with a correlation coefficient of 0.99. One of these compounds can be selected to represent the pair in the phenotypic screen, thus saving time and resources by limiting redundancy

and their retina is functionally more similar to the human retina compared to that of mice or rats [4, 25, 26]. 7. Some notes specific to zebra finch retinal ganglion cell culture: Edge effects cause drying on the outer columns of 96-well plates upon incubation and deter axon outgrowth, so avoid plating in these wells. If there is trouble with the explant not adhering to the plate after addition of remaining cell culture media, consider placing plate in incubator for 45 min to 1 h before adding additional media. Plating all explants must be completed within 2 h to avoid cell death. Placing the explant in

80

Lyndah Lovell et al.

the center of the well allows for maximum space for axongrowth. Do not allow the cell culture incubation to exceed 7 days. 8. Kinase inhibitor compounds were used in our SVM model, but other small molecules are equally capable of being fed into the SVM model when the SVM is trained with their profile dataset. 9. In silico high-throughput techniques that investigate small molecule ligand binding can also provide insight into multiple cell processes by revealing binding behavior of a compound to specific molecular targets. Known 3D structures of small molecules and the intended target serve as input files for a small molecule docking algorithm that predicts the best binding modes and binding affinity; this is especially useful for drug design [27]. The researcher can also use this as a first step to selecting new compounds to classify in the SVM. 10. For simplicity in the algorithm, the researcher can scale the kinase inhibition percentages from 1 to 10, where percentages greater than 0 but less than 10 are scaled to “1,” and percentages greater than 10 but less than 20 are scaled to “2,” and so on. 11. A generic example of a kinase inhibition profile matrix with a tenfold Cross-Validation is included below. When running the profile in the SVM, be sure to randomize the hit class order, so that both hits and nonhits are provided during SVM training (Fig. 7).

Fig. 7 Generalized example of kinase inhibitor profile matrix and tenfold Cross-Validation. Inhibition values are expressed as percentages randomly generated by Microsoft Excel. For simplicity, only ten compounds profiled against five kinases are shown; however, the researcher may use a matrix of any number of compounds and kinases (any dimension size) that best suits their research question

Axon Degeneration Screening

81

12. Omitting inhibition profiles for certain compounds may help to improve accuracy, sensitivity, and specificity. Training the algorithm on compounds that provide the maximum information, or maximum relevance for creating the hyperplane classifier, will enhance the model’s predictive ability. To build this maximum relevance dataset, please refer to paper [11]. 13. The kinase targets and antitargets revealed through SVM may not be the only kinases involved in axon degeneration and therefore should be compared to other pharmacologically related kinases in the respective organism kinome; for steps on how to write this algorithm and further definetargets and antitargets, please refer to the aforementioned paper [11].

Acknowledgments We would like to acknowledge all of the screening work done by the Jeffrey Milbrandt and Aaron DiAntonio laboratories, in particular, by Josiah Gerdts, Daniel Summers, and Chihiro Sato. We would also like to acknowledge the Vance Lemmon and John Bixby laboratories for their screening workflows and experience. References 1. Gooch CL, Pracht E, Borenstein AR (2017) The burden of neurological disease in the United States: a summary report and call to action. Ann Neurol 81(4):479–484 2. Heemels M-T (2016) Neurodegenerative diseases. Nature 539:179–179. https://doi.org/ 10.1038/539179a 3. Wozniak KM, Vornov JJ, Wu Y et al (2018) Peripheral neuropathy induced by microtubule-targeted chemotherapies: insights into acute injury and long-term recovery. Cancer Res 78(3):817–829. https://doi.org/10. 1158/0008-5472.CAN-17-1467 4. Bramley JC, Collins SVA, Clark KB, Buchser WJ (2016) Avian axons undergo Wallerian degeneration after injury and stress. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 202:813–822. https://doi.org/10. 1007/s00359-016-1123-y 5. Gerdts J, Sasaki Y, Vohra B et al (2011) Imagebased screening identifies novel roles for IkappaB kinase and glycogen synthase kinase 3 in axonal degeneration. J Biol Chem 286:28011–28018. https://doi.org/10. 1074/jbc.M111.250472 6. Essuman K, Summers DW, Sasaki Y et al (2017) The SARM1 toll/Interleukin-1 receptor domain possesses intrinsic NAD + cleavage

activity that promotes pathological axonal degeneration. Neuron 93:1334–1343. https://doi.org/10.1016/j.neuron.2017.02. 022 7. Geden MJ, Deshmukh M (2016) Axon degeneration: context defines distinct pathways. Curr Opin Neurobiol 39:108–115 8. Liraz-Zaltsman S, Slusher B, Atrakchi-Baranes D et al (2018) Enhancement of brain D-serine mediates recovery of cognitive function after traumatic brain injury. J Neurotrauma 35 (14):1667–1680. https://doi.org/10.1089/ neu.2017.5561 9. Mohr SE, Smith JA, Shamu CE et al (2014) RNAi screening comes of age: improved techniques and complementary approaches. Nat Rev Mol Cell Biol 15(9):591–600 10. Mei F, Fancy SPJ, Shen YAA et al (2014) Micropillar arrays as a high-throughput screening platform for therapeutics in multiple sclerosis. Nat Med 20(8):954–960. https://doi. org/10.1038/nm.3618 11. Al-Ali H, Lee D-H, Danzi MC et al (2015) Rational Polypharmacology: systematically identifying and engaging multiple drug targets to promote axon growth. ACS Chem Biol 10:1939–1951. https://doi.org/10.1021/ acschembio.5b00289

82

Lyndah Lovell et al.

12. Welsbie DS, Mitchell KL, Jaskula-Ranga V et al (2017) Enhanced functional genomic screening identifies novel mediators of dual Leucine zipper kinase-dependent injury signaling in neurons. Neuron 94(6):1142–1154. https:// doi.org/10.1016/j.neuron.2017.06.008 13. Hamamichi S, Rivas RN, Knight AL et al (2008) Hypothesis-based RNAi screening identifies neuroprotective genes in a Parkinson’s disease model. Proc Natl Acad Sci U S A 105(2):728–733. https://doi.org/10.1073/ pnas.0711018105 14. Bhattacharya MRC, Geisler S, Pittman SK et al (2016) TMEM184b promotes axon degeneration and neuromuscular junction maintenance. J Neurosci 36(17):4681–4689. https://doi. org/10.1523/JNEUROSCI.2893-15.2016 15. Frey E, Karney-Grobe S, Krolak T et al (2018) TRPV1 agonist, capsaicin, induces axon outgrowth after injury via Ca2+/PKA signaling. eNeuro 5. https://doi.org/10.1523/ ENEURO.0095-18.2018 16. Frey E, Valakh V, Karney-Grobe S et al (2015) An in vitro assay to study induction of the regenerative state in sensory neurons. Exp Neurol 263:350–363. https://doi.org/10. 1016/j.expneurol.2014.10.012 17. Lerch JK, Buchser W (2017) Functional genomics and high content screening in the nervous system. Mol Cell Neurosci 80:159–160. https://doi.org/10.1016/j.mcn.2017.03.009 18. Buchser W, Collins M, Garyantes T et al (2012) Assay development guidelines for image-based high content screening, high content analysis and high content imaging. In: Assay guidance manual. Eli Lilly & Company and the National Center for Advancing Translational Sciences, Bethesda (MD) 19. Gerdts J, Summers DW, Sasaki Y et al (2013) Sarm1-mediated axon degeneration requires both SAM and TIR interactions. J Neurosci

33:13569–13580. https://doi.org/10.1523/ JNEUROSCI.1197-13.2013 20. Osterloh JM, Yang J, Rooney TM et al (2012) dSarm/Sarm1 is required for activation of an injury-induced axon death pathway. Science 337:481–484. https://doi.org/10.1126/sci ence.1223899 21. Gerdts J, Brace EJ, Sasaki Y et al (2015) SARM1 activation triggers axon degeneration locally via NAD+ destruction. Science 348:453–457. https://doi.org/10.1126/sci ence.1258366 22. Hopkins AL (2008) Network pharmacology: the next paradigm in drug discovery. Nat Chem Biol 4:682–690. https://doi.org/10. 1038/nchembio.118 23. Mitiche I, Morison G, Nesbitt A et al (2018) Classification of EMI discharge sources using time–frequency features and multi-class support vector machine. Electr Power Syst Res 163:261–269. https://doi.org/10.1016/j. epsr.2018.06.016 24. Gu B, Quan X, Gu Y et al (2018) Chunk incremental learning for cost-sensitive hinge loss support vector machine. Pattern Recogn 83. https://doi.org/10.1016/j.patcog.2018.05. 023 25. Goldman SA, Nottebohm F (1983) Neuronal production, migration, and differentiation in a vocal control nucleus of the adult female canary brain. Proc Natl Acad Sci U S A 80:2390–2394. https://doi.org/10.1073/ pnas.80.8.2390 26. Lind O (2016) Colour vision and background adaptation in a passerine bird, the zebra finch (Taeniopygia guttata). R Soc Open Sci 3. https://doi.org/10.1098/rsos.160383 27. Man SM, Zhu Q, Zhu L et al (2015) Critical role for the DNA sensor AIM2 in stem cell proliferation and cancer. Cell 162:45–58. https://doi.org/10.1016/j.cell.2015.06.001

Chapter 7 A Microfluidic Culture Platform to Assess Axon Degeneration Yu Yong, Christopher Hughes, and Christopher Deppmann Abstract The field of microfluidics allows for the precise spatial manipulation of small amounts of fluids. Within microstructures, laminar flow of fluids can be exploited to control the diffusion of small molecules, creating desired microenvironments for cells. Cellular neuroscience has benefited greatly from devices designed to fluidically isolate cell bodies and axons. Microfluidic devices specialized for neuron compartmentalization are made of polydimethylsiloxane (PDMS) which is gas permeable, is compatible with fluorescence microscopy, and has low cost. These devices are commonly used to study signals initiated exclusively on axons, somatodendritic compartments, or even single synapses. We have also found that microfluidic devices allow for rapid, reproducible interrogation of axon degeneration. Here, we describe the methodology for assessing axonal degeneration in microfluidic devices. We describe several use cases, including enucleation (removal of cell bodies) and trophic deprivation to investigate axon degeneration in pathological and developmental scenarios, respectively. Key words Microfluidic devices, PDMS, Axon degeneration, Injury, NGF deprivation, Wallerian degeneration

1

Introduction Axons are the primary information conduits of the nervous system. Failure to maintain the integrity of axons is a feature of many inherited and acquired neurological disorders and can be attributed to roughly 16.8% of deaths, worldwide (2015) [1]. In response to pathological insult, such as injury (spinal cord injury, traumatic brain injury), exposure to toxic substances (Aβ, α-synuclein), or chemotherapies like vincristine, a process of axonal fragmentation or Wallerian degeneration (WD) occurs, often resulting in permanent loss of neural function [2, 3]. WD of injured axons has been

Electronic supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-07160585-1_7) contains supplementary material, which is available to authorized users. Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_7, © Springer Science+Business Media, LLC, part of Springer Nature 2020

83

84

Yu Yong et al.

observed in various species, and their molecular mechanisms are well conserved from mice to flies [4], but not in worms [5]. Perhaps counterintuitively, regressive/degenerative processes are essential for the formation of the vertebrate nervous system. During nervous system development, neurons, axon branches, and synapses are produced in excess. Components that receive sufficient trophic support stabilize, and those that do not are eliminated via apoptosis, nonapoptotic degenerative mechanisms, or cytoskeletal rearrangement, respectively [6, 7]. Developmental degeneration involves the elimination of significant portions of primary axons and major collaterals. Importantly, this process is evolutionarily conserved and has been observed in cortical layer 5 projection neurons in mouse, elimination of missorted optic axons during retinotectal development in zebrafish, and mushroom body γ neuron pruning during drosophila metamorphosis [8–11]. Despite the differences between developmental and pathological axon degeneration, understanding the cellular basis of both processes holds promise for therapeutic application. Thus, a highly adaptable and reproducible system will expedite our understanding of axon degeneration. Various in vitro and in vivo models allow the study of morphological and biochemical changes associated with pathological and developmental degeneration. In vivo models such as sciatic nerve transection, optic nerve crush, and lateral fluid percussion allow for simulation of degeneration after trauma [12, 13]. Drosophila mushroom body remodeling, terminal arbor pruning at the neuromuscular junction, and retinal ganglion projections during mouse retinotopic mapping provide experimental models for investigating developmental axon degeneration [14, 15]. However, in vitro models such as trophic factor deprivation, mechanical axotomy, and glutamate excitotoxicity allow rapid interrogation of the molecular mechanisms underlying degeneration. These classic in vitro and in vivo approaches are limited in their ability to apply insults locally (i.e., soma vs. axon), which is likely a closer approximation of physiological and pathological degenerative triggers. Moreover, these classic approaches often do not allow for examining biochemical events associated with degeneration in specific subcellular locales. Compartmentalized culture approaches overcome both of these barriers to progress. The compartmentalized “Campenot” chamber is the first device developed to allow local treatment of axons and soma. It consists of a reusable Teflon piece that is affixed to a petri dish with a thin layer of silicone vacuum grease [16] (Fig. 1a). These chambers have revolutionized the way the field considers long-distance signaling (i.e., from the retrograde signaling endosome) but suffered from a few nontrivial limitations as follows. (1) They are only suitable for neurons that project axons robust enough to cross the vacuum grease barrier. As such, most neurons of the central

Microfluidic Chambers to Study Axon Degeneration

85

Fig. 1 Campenot chamber and two chamber microfluidic device. (a) Campenot chamber. A 35-mm petri dish divided into three chambers by a Teflon divider. (b) Two chamber microfluidic device. A PDMS microfluidic device is attached to the 22
22 mm coverslip in a 35-mm petri dish

nervous system (CNS) cannot be grown in these devices. (2) Because these devices are plated on plastic and are opaque, they are incompatible with high-resolution fluorescence microscopy [17]. (3) Finally, preparing Campenot chambers is labor intensive, has a high failure rate (e.g., leaks), and is highly variable with respect to when axons cross to the barrier. Microfluidic compartmentalized neuronal devices have channels with micrometer dimensions, allowing for the manipulation of small volumes and control of the net flow of diffusion [18] (Fig. 1b). By incorporating microfluidic channels and precise control over factors bathing distinct neuronal regions, compartmentalized microfluidic platforms have been developed and used to study neuronal injury and axon degeneration [19–21]. These have several advantages over Campenot chambers: (1) Neurons from a wide range of sources including dorsal root ganglia, cortex, and hippocampus can be grown in these devices. (2) The usage of transparent PDMS offers high efficiency, good biological compatibility, and ability to integrate with other research techniques like electrophysiology and high-resolution microscopy [22]. (3) The process of making microfluidic devices is fast, inexpensive, and highly reproducible. There are three steps in the fabrication of microfluidic devices for neuronal culture: (1) design and print out the mask, (2) fabrication of the master mold using photolithography, and (3) production of microfluidics using polydimethylsiloxane (PDMS) (Subheading 3) [20, 23, 24]. A rudimentary two chamber device has been used to great effect for studies ranging from axonal transport of endosomes to axon regeneration after injury [19, 25]. This design has been modified into platforms with multireserviors for multiple

86

Yu Yong et al.

drug testing on axons [26–28], vertically layered platforms for 3D neuron/glia coculture [29], valve-based chambers for a compressive injury model [30, 31], and neuronal behavior analysis chips for small animals like Caenorhabditis elegans [32, 33]. Based on a protocol to study axon regeneration [20], we have used a similar approach of aspirating cell bodies away from axons to study the molecular mechanisms of Wallerian degeneration [21]. Here, we describe our protocol for examining axon degeneration after injury or trophic deprivation in microfluidic devices using neurons from mouse superior cervical ganglia (SCGs) (Fig. 2).

2

Materials

2.1 Fabrication of Microfluidic Molds and Devices

1. 100 mm silicon wafers. 2. Sylgard 184 silicone elastomer base and curing agent. 3. ¼00 hollow punch. 4. Razor blade or X-ACTO knife. 5. Vinyl tape. 6. Desiccator. 7. SU-8 5 and SU-8 50 photoresist. 8. Propylene glycol monomethyl ether acetate (PGMEA, photoresist developer). 9. Isopropanol. 10. Hot plates and/or ovens. 11. Kapton tape. 12. Photomask.

2.2 Plating Solutions/Reagents

1. Phosphate-buffered saline (PBS; 10
): 80.6 mM sodium phosphate, 19.4 mM potassium phosphate, 27 mM KCl, 1.37 M NaCl, pH 7.4. Dilute to 1
PBS with distilled water for a solution consisting of 10 mM phosphate, 137 mM NaCl, and 2.7 mM KCl. Sterilize 1
PBS by autoclaving. 2. Coating solution: 50 μg/mL poly-D-lysine and 1 μg/mL laminin in 1
PBS. 3. 22
22–1.5 glass coverslips. 4. 35 mm petri dishes. 5. Cell culture medium (varies depending on the cell type). 6. Live imaging medium: DMEM/F-12, no phenol red. 7. Calcium indicator: Fluo4-AM, cell permeant. 1 mM Fluo4-AM in DMSO. Store at 20  C. 8. Nerve growth factor (NGF).

Microfluidic Chambers to Study Axon Degeneration

87

Fig. 2 Schematic presentation of studying axon degeneration in microfluidics. Two chamber microfluidic devices can be fabricated using photolithography and PDMS. Neurons are plated in the cell body (CB) side channel of microfluidic device. A series of parallel microgrooves guide the growing neurites into the distal axon (DA) side channel. Various degeneration paradigms can be tested in these microfluidic cultures, including enucleation and global and local NGF deprivation. Live imaging or immunofluorescence can be performed in these cultures with different region of interests. ‘hv1’ and ‘hv2’ represent the excitation and emission lights, respectively

88

Yu Yong et al.

9. Antinerve growth-β antibody. 10. Cytosine β-D-arabinofuranoside hydrochloride (Ara-C): 5 mM Ara-C in distilled water. Filter through a 0.20-μM Corning filter. Leave one aliquot at 4  C for current use, and store the remaining aliquot at 20  C. 2.3 Immunostaining Solutions

1. 4% PFA: Add 160 mL water to flask, and warm it up to 50–60  C. Add 16 g paraformaldehyde to 160 mL water in chemical hood, stirring for 15 min. Add 200 μL of 10 N NaOH solution. Add 20 mL 10
PBS. Keep the heat and stir for at least 5 h until the solution is clear. Test and adjust the pH to 7.2–7.4. Make up to final volume of 200 mL with water. Cool to room temperature and aliquot. Make 1:1 dilution with 1
PBS. Store at 20  C (see Note 1). 2. 4% PPS: 4% PFA with 120 mM sucrose, 2 mM MgCl2, 10 mM EGTA, 25 mM HEPES, and 60 mM PIPES. Weigh and add 1.82 g of PIPES, 0.60 g of HEPES, 0.38 g of EGTA, 0.019 g of MgCl2, and 4.11 g of sucrose to 100 mL 4% PFA. Test and adjust the pH to 7.4. Aliquot and store at 20  C. 3. Blocking solution: 5% normal goat serum, 0.05% Triton-X-100 in 1
PBS.

3

Methods

3.1 Fabrication of Master Mold

The steps to make master mold for microfluidics are described here. 1. Design and manufacture two high-resolution photomasks containing a desired microgroove and main chamber patterns with alignment marks, respectively. The autocad files for these masks can be found in the Electronic Supplemental Material. Chamber (red) and groove (magenta) layers with cross alignment marks for microfluidic device are shown in photomask design. ai and photomask design.dwg files. The size of silicon wafer (blue) is 4 in. For chamber layer, the distance between soma and axonal chamber is 350 μm. For groove layer, the width for each groove is 10 μm. 2. Clean the silicon wafer with isopropanol. Dry using pressurized insert gas. 3. Taking care to avoid air bubbles, add about 3 mL of SU-8 5 photoresist on a cleaned 100 mm silicon wafer. Spin at 504
g (3000 rpm) for 1 min (3 μm thick). 4. Bake at 65  C for 1 min, then at 95  C for 1 min. 5. Expose the wafer to a photomask containing a microgroove pattern and alignment marks for 30 s at 200 W. 6. Bake at 65  C for 1 min, then at 95  C for 1 min.

Microfluidic Chambers to Study Axon Degeneration

89

7. Develop with PGMEA for 3–5 min. Rinse with isopropanol (see Note 2). Dry using pressurized insert gas. 8. Bake at 65  C for 1 min, then at 95  C for 10 min. 9. Cover the alignment marks with Kapton tape (PDMS strips or scotch tape are effective alternatives). 10. Taking care to avoid air bubbles, add about 3 mL of SU-8 50 photoresist on the silicon wafer. Spin at 56
g (1000 rpm) for 1 min (100 μm thick). 11. Remove the Kapton tape. Soft bake at 65  C for 10 min, then at 95  C for 30 min. 12. Align the chamber mask to the grooves patterned on the wafer using the alignment marks. Expose for 2 min at 200 W. 13. Bake at 65  C for 1 min, then at 95  C for 10 min. 14. Cool the wafer to room temperature (about 1 min). Develop with PGMEA for 5–10 min as needed (see Note 2). Rinse three times with fresh developer solution. Rinse with isopropanol. Dry using pressurized insert gas. 15. Bake at 65  C for 1 min, then at 95  C for 10 min. Store in clean case. 3.2 Fabrication of Microfluidic Devices

1. Fashion a dish out of aluminum foil 20% larger than the master mold (see Note 3). Place the master mold (silicon wafer) in foil dish, and place it in a 150-mm petri dish (Fig. 3a–c). 2. Measure 24 g of silicone elastomer solution with 2.67 g of curing agent in a weighing boat (see Note 4). Mix well by stirring. 3. Place the plate into the leveled vacuum chamber or desiccator to remove bubbles for 30–60 min (Fig. 3d, see Note 5). 4. Cure the elastomer at 60  C overnight in the oven, which will solidify the PDMS (see Note 6). 5. Take out each PDMS mold and cut out the square block of nine chambers with a razor blade (Fig. 3e, see Note 7). Punch holes directly over the circle reservoirs using sharpened stainless-steel punch (Fig. 3f). Cut each individual chamber such that its dimensions are approximately 100
100 , which can fit on to a square coverslip (22 mm
22 mm). Cut the corners to help with handling (Fig. 3g–h). 6. Use vinyl tape to remove dust sticking to the chamber on the groove-side. 7. Always keep groove-side up, and store the microfluidics in closed plate at room temperature (Fig. 3i, see Note 8).

90

Yu Yong et al.

Fig. 3 Fabrication of microfluidics. (a) Design of two chamber of two chamber microfluidics. The microgroove is 10 μm wide, 350 μm long. The size of microfluidic device is 22
22 mm, which fits the square glass coverslips. (b) Picture of foil cup. (c) Picture of the master mold. (d) Picture of PDMS mold after debubbling. (e) Picture of PDMS mold after cutting put the square block. (f) Picture of punching holes over circle reservoirs. (g) Picture of individual microfluidic device. (h) Picture of individual microfluidic device after cutting corners. (i) Picture of PDMS microfluidics

3.3 Neuron Plating and Culture in Microfluidics

Carry out all procedures in sterile environment (tissue culture hood). 1. Coat the square glass coverslips (22
22–1.5) with coating solution at 37  C 5–10% CO2 incubator overnight (see Note 9). Put the coated coverslips in 35 mm petri dish. Wash them with sterile water three times. Air dry the coverslips. 2. Sterilize the microfluidic devices by dipping them in pure ethanol. Let the devices completely dry (1–2 h). Attach the microfluidic device to the coated coverslip (see Note 10).

Microfluidic Chambers to Study Axon Degeneration

91

3. Prepare P0–P2 mouse superior cervical ganglion (SCG) dissociated cell suspension. Other neuronal cells such as embryonic DRG neurons have also been successfully cultured in microfluidics. Add 3 μL of cell suspension (1000–5000 neurons) to cell body side channel of each microfluidic device. Tap the plate to make sure the fluid connects two cell body side wells. 4. Put the plates into 37  C 5–10% CO2 incubator for 5 min (see Note 11). 5. Add about 125 μL of medium to each cell body side well without flushing out the cells. Incubate at 37  C 5–10% CO2 for 30 min. 6. Push 200 μL of medium into distal axon side channel to make sure that the two distal axon side wells are connected. Add more medium to the distal axon side wells to achieve 125 μL of medium per well. 7. Replace the medium in the wells next day. Add 5 μM Ara-C into the medium to remove glia for 48–72 h at 37  C 5–10% CO2. 8. Change medium in the wells every 2 days for following experiments (see Notes 12 and 13). 3.4 Live Imaging in Microfluidic Cultures

For live imaging, instead of plating and culturing the neurons on the glass coverslips in petri dish, attach the PDMS microfluidic device to glass bottom tissue culture dish. Carry out all procedures in tissue culture hood unless otherwise specified. 1. Wash the microfluidic culture with live imaging medium for 3
5 min. 2. Replace live imaging medium containing 1 μM Fluo4-AM for calcium imaging. Incubate cells at 37  C 5–10% CO2 for 30 min prior to imaging. 3. To remove excess dye, replace live imaging medium. 4. Set up desired microscope system and paradigms to perform live imaging.

3.5 Inducing Axon Degeneration

3.5.1 Axotomy

Always check the plates under microscope to ensure that neurons project their axons to the axonal chamber (3–7 DIV). Few examples of degeneration paradigms are described below. These procedures are also applicable to chemotherapeutics and axon regeneration studies. Carry out all procedures in tissue culture hood unless otherwise specified. 1. Remove medium in the cell body side wells. 2. Use glass Pasteur pipet connected to the vacuum to aspirate 3 mL of 1
PBS through the cell body chamber, leaving the axons intact in their respective chamber (see Note 14).

92

Yu Yong et al.

3. Check the enucleating condition under microscope. Aspirate additional 1–3 mL of 1
PBS through the cell body chamber if necessary. 4. Remove all the remaining medium in four wells. Add prewarmed medium to four wells. Incubate at 37  C 5–10% CO2 for designated times. 3.5.2 Global NGF Deprivation

1. Wash the microfluidic culture with prewarmed medium without NGF for 3
5 min. 2. Replace the medium in all four wells with NGF deficient medium containing 1 μg/mL anti-NGF function neutralizing antibody. Incubate at 37  C 5–10% CO2 for designated time.

3.5.3 Local NGF Deprivation

1. Wash the microfluidic culture with prewarmed medium without NGF for 3
5 min. 2. Add 200 μL of NGF deficient medium containing 1 μg/mL anti-NGF antibody to distal axon chamber. 3. Add about 250 μL of medium with NGF to cell body chamber (see Note 15). Incubate at 37  C 5–10% CO2 for designated time. Check the level of medium in each well every 8 h. Reestablish the volume differential if necessary. Rapid loss of volume differential indicates an improper seal rendering the device inappropriate for differential treatment.

3.6 Immunostaining in Microfluidic Cultures

Carry out all procedures in plates with PDMS microfluidic devices attached to the coverslips. It is not necessary to remove the microfluidic chamber for immunostaining. 1. Remove all the medium from four wells of the microfluidics. 2. To fix axons, add 4% PFA to all four wells. Fix at room temperature for 20 min. To fix delicate structures like growth cones or axon spheroids, add prewarmed 4% PPS to wells and fix at 37  C for 20 min (see Note 16). 3. Wash the plate with 1
PBS for 3
5 min. 4. Replace 1
PBS in the wells with blocking buffer. Incubate at room temperature for 1 h. 5. Replace the blocking buffer in wells with desired primary antibody diluted in blocking buffer (e.g., Tuj1, 1:1000). Incubate at 4  C overnight. 6. Wash the plate with 1
PBS for 3
5 min. 7. Replace 1
PBS in the wells with fluorescent conjugated secondary antibody diluted in blocking buffer (i.e., Goat AntiMouse Alexa 488, 1:800). Incubate at room temperature in dark for 1 h. 8. Wash the plate with 1
PBS for 3
5 min in dark.

Microfluidic Chambers to Study Axon Degeneration

93

9. Carefully remove the microfluidic device with coverslip attached from petri dish, and place it directly onto an inverted microscope stage for square coverslip (see Note 17). Perform imaging. 3.7 Quantification of Axon Degeneration

Axon degeneration in microfluidic culture can be quantified from β-III tubulin immunostained fluorescence images either automatically or manually. Loss of positive staining axons can be quantified by ImageJ software using “threshold” function or Stereo Investigator software using spaceballs or petrimetrics stereological probe in unbiased manner [34]. However, in areas with dense axons, the individual axons may appear to merge after applying the threshold analysis. Parameters and settings vary depending on the staining and experimental groups. Algorithms like AxonQuant provide a fast, microfluidic optimized way to measure “axonal continuity” and morphology in automated manner independent of neuronal or axonal density [35]. Alternatively, manual quantification of axon degeneration has been used in many studies. Below is an example of the manual quantification, which is based on the analysis of morphological characteristics (i.e., beads/blebs) of degenerating axons [36]. 1. Investigator A takes 10 representative images of axons in microfluidics per experimental condition (or biological replicate). Blind the images. 2. Blinded Investigator B randomly assigns ten 50 μm-long rectangular boxes to single axons on each image (see Note 18). 3. The blinded Investigator B counts the number of beads/blebs (n) in each box to determine the degenerating axons. If there are three or more beads, blebs, or breaks, the boxed axon is considered a degenerating axon (Fig. 4). The percentage of degeneration per picture is determined by the number of degenerating boxed axons (i.e., if there are five boxed axons out of ten counted as degenerating axons, the percentage of degeneration is 50%). 4. Investigator A unblinds all the images and performs statistical analysis to determine mean, standard error, and other parameters of axon degeneration in each experimental condition.

4

Notes 1. To avoid repeated thawing, leave one aliquot of 4% PFA at 4  C for current use, maximum 1 month. 2. If the grooves or edges of chamber look cloudy after isopropanol rinse, it requires more time to develop. Change into fresh developer if necessary.

94

Yu Yong et al.

Fig. 4 Quantification of axon degeneration. P0 superior cervical ganglion neurons were cultures in microfluidics. At DIV 5, neurons were treated with global NGF deprivation for 19 h, and then fixed and immunostained for Tuj1. 10 yellow 50 μm-long rectangular boxes were randomly assigned to single axons in the image. Asterisks marked the degenerating axons. The percentage of degeneration in this field is 50%

3. To make casting dishes, you may use the bottom of a 500-mL or 1-L beaker to mold the aluminum foil. 4. The first time pouring requires about two times more silicone elastomer and curing agent to achieve appropriate coverage and thickness. The proportion of silicone elastomer to curing agent should be 9:1. 5. This step takes approximately 1–2 h depending on the vacuum and how many plates are loaded. Take care that all microbubbles are removed curing in the oven. 6. Do not stack the plates. The oven should be leveled. 7. Too much pressure during cutting will crack the mold. 8. Protect the microfluidics from dust. Avoid touching the groove-side as they are very delicate. Check the devices before attaching to the coated coverslips to make sure the grooves and channels are connected. 9. Coverslips in coating solution should be incubated at least 2 h at 37  C and can be used up to 2 weeks when kept in coating solution at 37  C. 10. Press hard on microfluidic devices to ensure proper attachment and prevent leakage, and use light pressure for the middle of the chamber (grooves).

Microfluidic Chambers to Study Axon Degeneration

95

11. You may skip this step if there are too many plates to handle. However, it may affect the seeding of neurons in the microfluidics. 12. When changing medium, point the pipet tip away from the connecting channel of the microfluidics to avoid disturbing soma and axons. 13. To maintain the humidity of long-term in vitro cultures, store the 35 mm plates in a 150 mm plate with one water dish. 14. When enucleating neurons, point the pipet tip close to the connecting channel of the cell body side. 15. For local NGF deprivation on cell body side, switch the volume for the two compartments. To maintain fluidic isolation, reset a volume difference of 30–50 μL between the two compartments if necessary. 16. To fix axonal spheroids, slowly add 4% PPS to microfluidic chambers along the edge of the wells. To maintain the surface tension, do not remove all of the medium in each well. 17. To remove the petri dish, leaving coverslips attached to microfluidic device, squeeze the plastic petri dish and use forceps to take out the microfluidics. 18. The investigator should take care not to box bundles of axons, which may confound analysis.

Acknowledgements We thank Nadine Ly for technical assistance and Kanchana Gamage and Shayla Clark for helpful suggestions and comments on the chapter. We thank Professor Brian Pierchala (University of Michigan) for Campenot chamber image. References 1. Feigin VL, Abajobir AA, Abate KH et al (2017) Global, regional, and national burden of neurological disorders during 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Neurol 16:877–897 2. Vargas ME, Barres BA (2007) Why Is wallerian degeneration in the CNS so slow? Annu Rev Neurosci 30:153–179 3. Waller A (1850) Experiments on the section of the glossopharyngeal and hypoglossal nerves of the frog, and observations of the alterations produced thereby in the structure of their primitive fibres. Philos Trans R Soc London 140:423–429

4. Coleman MP, Freeman MR (2010) Wallerian degeneration, Wld S, and Nmnat. Annu Rev Neurosci 33:245–267 5. Nichols ALA, Meelkop E, Linton C et al (2016) The apoptotic engulfment machinery regulates axonal degeneration in C. elegans neurons. Cell Rep 14:1673–1683 6. Cusack CL, Swahari V, Hampton Henley W et al (2013) Distinct pathways mediate axon degeneration during apoptosis and axonspecific pruning. Nat Commun 4:1876 7. Solomon F, Magendantz M (1981) Cytochalasin separates microtubule disassembly from loss of asymmetric morphology. J Cell Biol 89:157–161

96

Yu Yong et al.

8. Luo L, O’Leary DDM (2005) Axon retraction and degeneration in development and disease. Annu Rev Neurosci 28:127–156 9. Neukomm LJ, Freeman MR (2014) Diverse cellular and molecular modes of axon degeneration. Trends Cell Biol 24:515–523 10. Geden MJ, Deshmukh M (2016) Axon degeneration: context defines distinct pathways. Curr Opin Neurobiol 39:108–115 11. Poulain FE, Chien CB (2013) Proteoglycanmediated axon degeneration corrects pretarget topographic sorting errors. Neuron 78:49–56 12. Cheriyan T, Ryan DJ, Weinreb JH et al (2014) Spinal cord injury models: a review. Spinal Cord 52:588–595 13. Xiong Y, Mahmood A, Chopp M (2013) Animal models of traumatic brain injury. Nat Rev Neurosci 14:128–142 14. Watts RJ, Schuldiner O, Perrino J et al (2004) Glia engulf degenerating axons during developmental axon pruning. Curr Biol 14:678–684 15. Low LK, Cheng HJ (2006) Axon pruning: an essential step underlying the developmental plasticity of neuronal connections. Philos Trans R Soc B Biol Sci 361:1531–1544 16. Campenot RB (1977) Local control of neurite development by nerve growth factor. Proc Natl Acad Sci U S A 74:4516–4519 17. Park JW, Kim HJ, Byun JH et al (2009) Novel microfluidic platform for culturing neurons: culturing and biochemical analysis of neuronal components. Biotechnol J 4:1573–1577 18. Whitesides GM (2006) The origins and the future of microfluidics. Nature 442:368–373 19. Taylor AM, Blurton-Jones M, Rhee SW et al (2005) A microfluidic culture platform for CNS axonal injury, regeneration and transport. Nat Methods 2:599–605 20. Park JW, Vahidi B, Taylor AM et al (2006) Microfluidic culture platform for neuroscience research. Nat Protoc 1:2128–2136 21. Gamage KK, Cheng I, Park RE et al (2017) Death receptor 6 promotes wallerian degeneration in peripheral axons. Curr Biol 27 (6):890–896 22. Millet LJ, Gillette MU (2012) New perspectives on neuronal development via microfluidic environments. Trends Neurosci 35:752–761 23. Beebe DJ, Mensing GA, Walker GM (2002) Physics and applications of microfluidics in biology. Annu Rev Biomed Eng 4:261–286 24. Gross PG, Kartalov EP, Scherer A et al (2007) Applications of microfluidics for neuronal studies. J Neurol Sci 252:135–143

25. Barford K, Keeler A, McMahon L et al (2018) Transcytosis of TrkA leads to diversification of dendritic signaling endosomes. Sci Rep 8:1–14 26. Park J, Koito H, Li J et al (2012) Multicompartment neuron-glia co-culture platform for localized CNS axon-glia interaction study. Lab Chip 12:3296–3304 27. Jocher G, Mannschatz SH, Offterdinger M et al (2018) Microfluidics of small-population neurons allows for a precise quantification of the peripheral axonal growth state. Front Cell Neurosci 12:1–12 28. Kilinc D, Peyrin JM, Soubeyre V et al (2011) Wallerian-like degeneration of central neurons after synchronized and geometrically registered mass axotomy in a three-compartmental microfluidic chip. Neurotox Res 19:149–161 29. Shi M, Majumdar D, Gao Y et al (2013) Glia co-culture with neurons in microfluidic platforms promotes the formation and stabilization of synaptic contacts. Lab Chip 13:3008–3021 30. Park JW, Kim HJ, Kang MW et al (2013) Advances in microfluidics-based experimental methods for neuroscience research. Lab Chip 13:509–521 31. Shrirao AB, Kung FH, Omelchenko A et al (2018) Microfluidic platforms for the study of neuronal injury in vitro. Biotechnol Bioeng 115:815–830 32. Chronis N, Zimmer M, Bargmann CI (2007) Microfluidics for in vivo imaging of neuronal and behavioral activity in Caenorhabditis elegans. Nat Methods 4:727–731 33. Ben-Yakar A, Chronis N, Lu H (2009) Microfluidics for the analysis of behavior, nerve regeneration, and neural cell biology in C. elegans. Curr Opin Neurobiol 19:561–567 34. Kneynsberg A, Collier TJ, Manfredsson FP et al (2016) Quantitative and semi-quantitative measurements of axonal degeneration in tissue and primary neuron cultures. J Neurosci Methods 266:32–41 35. Li Y, Yang M, Huang Z et al (2014) AxonQuant: A microfluidic chamber culturecoupled algorithm that allows highthroughput quantification of axonal damage. Neurosignals 22:14–29 36. Zhai Q, Wang J, Kim A et al (2003) Involvement of the ubiquitin-proteasome system in the early stages of Wallerian degeneration. Neuron 39:217–225

Chapter 8 A Schwann Cell–Neuron Coculture System to Study Neuron–Glia Interaction During Axon Degeneration Elisabetta Babetto Abstract Autonomous mechanisms of axon degeneration are frequently studied in vitro by mechanical axon injury of isolated sensory neurons. This has led to major advances in understanding the molecular pathways governing axon degeneration. However, this approach does not pay attention to potential glial mechanisms for the regulation of axon death. Here, I describe a straightforward protocol to seed purified rat Schwann cells on neuronal cultures in order to study the interaction between axons and these glia during axon degeneration. Key words Schwann cell, Spot culture, Axon, Neuro–glia interaction, Axon degeneration, In vitro injury, Coculture

1

Introduction While the dismantling and clearance of myelin by Schwann cells are well-recognized critical components of Wallerian degeneration [1], new insights into other more instructive roles of glia are starting to be discovered [2–4]. Moreover, glia physiologically maintain the integrity of long axons through metabolic functions, raising the possibility that such functions could also be important for the support of injured axons with especially heightened energy demands [5–7]. Here, I describe a straightforward method to facilitate the observation of glial dynamics, following axon injury in a spatially and temporally controlled manner. Injury-induced experimental axon degeneration is a commonly used technique to elucidate neuron-dependent mechanisms [8– 11]. In vitro sensory neuronal cultures, mostly embryonic dorsal root ganglia (eDRGs) or superior cervical ganglia (SCGs) have been used in many studies to investigate the effects of drugs or neuronal genetic modifications on axon degeneration [12–14]. As described in the other chapters, one of the key benefits of these cultures is the possibility to grow neurons in a manner that clearly

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_8, © Springer Science+Business Media, LLC, part of Springer Nature 2020

97

98

Elisabetta Babetto

spatially separates the cell body compartment from the axonal area, either by plating the entire ganglia or by producing “spot cultures” in which a cell suspension of highly concentrated neurons is plated in a small volume of medium in a defined area of the well. Following cell adhesion to the substrate, the neurons extend neurites away from the cell bodies radially over the course of 4–7 days. This somaaxon spatial segregation is crucial for two reasons. First, it makes possible to study only axons without the attached parent neuronal cell bodies, for example, to perform protein analysis exclusively on axons. Since the axonal fraction is devoid of other cell types, this approach is the only known method to selectively study axonal mechanisms by biochemical techniques. Second, it allows highresolution time-lapse and live-imaging analysis of axons undergoing degeneration. For example, it allowed us to discover that experimental axon degeneration proceeds in a temporally controlled manner [9]. Molecular changes occurring in neurons within the first hours after injury orchestrate a decision-making process that commits the axon to degeneration hours later [11, 13]. This is often referred to as the “commitment phase” of axon degeneration. It is during this phase that pharmacological interventions have the highest potential to curb axon breakdown [13]. Genetic manipulations that result in axon protection have also been hypothesized to act during this crucial time period, for example, by antagonizing the energetic decline invoked by the injury [11, 15]. Once “committed” to axon death, the axons synchronously undergo the structural hallmarks of blebbing and fragmentation. However, the study of such cultures is geared toward the discovery of neuron-specific mechanisms of axon degeneration, neglecting potential regulatory mechanisms in the glia that axons are intimately associated with. This chapter illustrates a method to shed light upon the inextricable relationship between axons and Schwann cells during degeneration, which enables scientists to study novel aspects of glia–neuron interaction during the commitment phase of axon degeneration and beyond. This protocol retains the advantages of the “spot culture” technique, in which axons are separated from the neuronal cell bodies. The method allows for time-lapse imaging but adds the potential to investigate a close glia–neuron interaction that would be very challenging to achieve using in vivo studies. Several previous methods to coculture dorsal root ganglia with the attached endogenous Schwann cells have already been published. In contrast, this method applies exogenous Schwann cells to already developed neurons (after 6–7 days in vitro), allowing the prior genetic manipulation of each cell type individually, thus greatly reducing potential developmental effects that may influence axon stability cell autonomously. Viral transduction or pharmacological manipulations can be carried out before or after the two cells types are put into contact with each other, making this protocol valuable to address novel experimental

Schwann Cell-Neuron Coculture

99

questions. Because of the tight temporal control that the investigator can exert on the experimental conditions, the morphological and functional interaction of Schwann cells and neurons can be studied at different time points after injury, allowing the investigation of early changes in Schwann cells or axons. Still, mouse Schwann cells are notoriously difficult to culture over several days because of their reduced cell division and their intrinsic difficulty to myelinate in vitro. Rat Schwann cells are more often used, but the classical methods to prepare them by immunopanning involves a laborious and time-consuming procedure [16]. Therefore, we have developed a straightforward method that can be carried out within a week to obtain highly purified Schwann cells from postnatal rat pups by selective magnetic pull down purification. Altogether, this mixed culture method is a reproducible and easy-to-carry-out technique that is well suited to investigate axon– glia crosstalk after axon injury.

2 2.1

Materials Reagents

1. Fetal bovine serum (FBS) (see Note 1). When first received, decomplement this serum by heat inactivation. Aliquots can be stored at
80  C until the expiration date. 2. D-glucose suitable for cell culture. 3. Natural mouse laminin (1 mg/mL) (see Note 2). 4. Poly-D-lysin (PDL) (10 mg/mL, diluted in sterile MilliQ water) (see Note 3). 5. Poly-L-lysin (PLL) (0.1 mg/mL, diluted in sterile MilliQ water). 6. B27 supplement serum free (see Note 4). 7. Neurobasal medium. 8. Minimum essential medium Eagle’s medium (MEM). 9. Dulbecco’s modified Eagle’s medium (DMEM). 10. 10% FBS in DMEM: dilute 1 mL of FBS in 9 mL of DMEM and filter for sterility. 11. Hanks Balanced Salt Solution (HBSS) without calcium and magnesium. 12. 40% FBS in HBSS: dilute 4 mL of FBS in 6 mL of HBSS and filter for sterility. 13. eDRG medium: Neurobasal Medium, 1 B27 supplement, 2 mM L-glutamine, penicillin-streptomycin 100 U/mL, 50 ng/mL NGF 2.5S, 1 μM 5-fluoro-20 -deoxyuridine, and 1 μM uridine.

100

Elisabetta Babetto

14. C-medium: MEM, 10% FBS, 4 g/L glucose, 2 mM L-glutamine, and 50 ng/mL NGF 2.5S. Filter for sterility. 15. Trypsin-EDTA 0.05%. 16. Leibovitz’s L-15 medium. 17. Nerve growth factor (NGF) 2.5S (100 μg/mL, diluted in sterile MilliQ water) (see Note 5). 18. 70% ethanol for sterilizing surfaces and instruments. 19. Ascorbic acid (diluted in MEM and used at a final concentration of 50 μg/mL) (see Note 6). 20. Dulbecco’s PBS (D-PBS). 21. 0.05% (w/v) Type 1 collagenase in DMEM. 22. 0.125% (w/v) trypsin TRL3 in DMEM. 23. Wash buffer: D-PBS with 2 mM EDTA. Filter and keep cold (on ice or in the refrigerator). 24. Separation buffer: D-PBS, 0.5% (w/v) bovine serum albumin, 2 mM EDTA. Filter, degas, and keep cold (on ice or in the fridge). To degas, place the solution in a vacuum chamber for at least 2 h (or overnight). The solution can be stored in the fridge for up to 3 days. 25. Anti-p75 primary antibody, clone 192–IgG (or analogous antibody against an extracellular epitope of the p75 protein). 26. Rat Anti-Mouse IgG1 Microbeads (Miltenyi Biotech). 27. Gentamycin (final concentration 50 μg/mL). 28. Rat Schwann cell medium: DMEM with high glucose, 10% FBS, 2 mM Glutamine, penicillin–streptomycin 100 U/mL, 2 ng/mL recombinant human neuregulin, and 2 μM forskolin, filtered after mixing. 29. Trypan blue. 30. SOX10 antibody or other Schwann cell marker. 2.2

Equipment

1. Vacuum chamber for degassing. 2. Dissection instruments (laminectomy forceps and small Dumont #5, scissors etc.). 3. Petri dishes. 4. Biosafety cabinet. 5. Laminar flow with dissection microscope. 6. LS-columns with MACS separator (Miltenyi Biotech). 7. 0.22 μm filters. 8. 70 μm nylon cell restrainer. 9. Sterile scalpel for chopping tissue.

Schwann Cell-Neuron Coculture

101

10. Interchangeable flat blades and handle with 1.5 mm cutting edge (e.g., FST blade 10035-05). 11. 5% CO2 cell culture incubator. 12. Hemocytometer for cell counting. 13. 24-well plate. 2.3

Animals

1. Embryonic date 14.5 (E14.5)-timed pregnant CD-1 IGS mouse (see Note 7). 2. Postnatal P2 or P3 rat pups, Sprague Dawley (see Note 8).

3

Method

3.1 Purification of Rat Schwann Cells: Dissection and Plating

1. Coat the petri dishes with PLL (final concentration 0.01 mg/ mL in sterile water) for 30 min at room temperature. 2. Remove the PLL solution and wash three times with D-PBS. Leave at room temperature until when ready to plate. 3. After humanely euthanizing the rat pups by decapitation, work quickly (ideally completing all the dissection within an hour) under a dissection microscope placed in a laminar flow hood. Remove the sciatic nerves of each pup. At this age, the nerves still appear fairly translucent because of the lack of myelination, therefore a good light source and a dissection microscope are crucial to carry out the procedure properly (see Note 9). Place the remaining pups on ice while dissecting. 4. Place the nerves in ice-cold L-15 medium containing gentamycin, while finishing the dissection of all the pups. 5. Transfer 4–6 nerves to a 35-mm petri dish containing L-15 medium without gentamycin, and clean the nerves by removing pieces of tissues and blood that may have remained attached to the nerve during the dissection. 6. Chop the sciatic nerves in pieces as small as possible using a sterile blade. 7. Remove the L-15 with a P1000 pipette, carefully ensuring that the small pieces of tissue remain in the dish, and proceed to the enzymatic digestion. 8. Add 1 mL of 0.05% type I collagenase, dissolved in DMEM, and incubate at 37  C in the cell culture incubator for 30 min. 9. Remove the collagenase solution carefully and follow with 1 mL of 0.125% trypsin TRL3 dissolved in DMEM, incubating at 37  C in for 30 min. 10. Stop the enzyme activity by adding 2.5 mL 40% FBS in HBSS. 11. Transfer the solution containing the small pieces of tissue to a tube and centrifuge at 300  g for 10 min.

102

Elisabetta Babetto

12. Resuspend the pellet in 10% FBS in DMEM. 13. Triturate the tissue by passing it through a P1000 tip about 35 times or until no tissue fragments are visible. In this step, it is crucial not to cause bubbles in the solution as these would strongly decrease cell viability. 14. Remove the remaining membranes and undissolved chunks of tissue by passing the solution though a 70-μm nylon cell restrainer placed on 50 mL falcon tube (see Note 10). 15. Centrifuge the cells at 300  g for 10 min, and resuspend the pellet in rat Schwann cell medium. 16. Plate the cells in petri dishes previously coated with PLL. 17. Rinse the dishes daily with HBSS to remove the debris caused by the dissection and the degenerating axons. 18. Add fresh medium until confluency is reached or fibroblasts start to invade, whichever happens first (usually 3–4 days). 3.2 Purification of RAT SCHWANN CELLS: Magnetic Pull Down Assay

1. Trypsinize the dishes to detach the cells (with 0.05% Trysin with EDTA for 5 min in the cell culture incubator) (see Note 11). 2. After cell detachment, stop the trypsinization by adding 10% FBS in DMEM. 3. Centrifuge the cell suspension at 300  g for 10 min and then resuspend the pellet in 2 mL of wash buffer. 4. Place a 70-μm nylon cell restrainer onto a 50-mL falcon tube and moisten it with 1 mL wash buffer. 5. Pass the cell suspension through a 70-μm nylon cell restrainer to remove cell clumps. 6. The following steps describe the protocol to purify up to 107 cells. Count the cells using a hemocytometer or an automated cell counter; if the number of cells is greater than 107, double the volumes and use two columns from this point onward. 7. Remove the wash buffer by centrifuging at 300  g for 10 min and discarding the supernatant. 8. Resuspend the cells in 100 μL of separation buffer containing 5 μL of anti-p75 primary antibody, clone 192–IgG (see Note 12). This antibody binds to an extracellular epitope of Schwann cells. Incubate at room temperature for 10 min. 9. Wash the excess unbound antibody by adding 10 mL of wash buffer, centrifuging the cells at 300  g for 10 min and discarding the supernatant. 10. Resuspend the cells in 80 μL of separation buffer and 20 μL of Rat Anti-Mouse IgG1 Microbeads and incubate for 15 min in a refrigerator.

Schwann Cell-Neuron Coculture

103

11. Wash the cells with 5 mL of wash buffer and centrifuge 300  g for 10 min. 12. Resuspend the cells with 500 μL of separation buffer and keep on ice. 13. Set up the MACS magnetic separator and rinse the LS column with 3 mL of separation buffer. Discard the flow-through. 14. Apply the cell suspension onto the LS column and collect the flow-through containing unlabeled cells (see Note 13) for celltype characterization. 15. Wash the column with 3 mL of separation buffer and collect the unlabeled cells (repeat for three times). 16. Remove the column from separator and place it onto a 15-mL collection tube. 17. Collect the p75-labeled cells by applying 5 mL of Schwann cell medium to the column. 18. Immediately flush out the positive cells by firmly pushing the plunger into the column. 19. Determine the cell number and vitality by counting the cells in a hemocytometer using trypan blue or similar dyes. 20. Plate the cells in PLL-coated dishes and expand them up to the fifth passage (see Note 14). The purity of Schwann cells can be assessed by immunostaining with a Schwann cell marker, for example, SOX10 (Fig. 1). This protocol usually yields Schwann cells with a purity of about 95%.

Fig. 1 Characterization of the purity of rat Schwann cells prepared by magnetic pull down assay (MACS). Low-magnification epifluorescence images of cells 24 h after plating stained for the glia marker SOX10 and counterstained with DAPI. The protocol yields purified Schwann cells (with a rate of 95% purity). Few glial cells are present in the flow-through fraction that did not bind to the anti-P75 antibody

104

Elisabetta Babetto

21. In order to determine the efficiency of the protocol, the flowthrough collected at point 14 can be also centrifuged at 300  g for 10 min and resuspended in Schwann cell medium. Cells can then be plated and analyzed by immunostaining (Fig. 1). The vast majority of the cells in this fraction are not SOX10-positive; they are more likely to be fibroblasts. 22. Schwann cells can be prepared in advance, expanded, and frozen in liquid nitrogen with standard protocols. About 1 or 2 days after plating eDRGs, a vial of Schwann cells should be defrosted to allow for at least one passage before adding the glia onto the neurons. 3.3 Neuronal Spot Culture Preparation

1. The day before the eDRGs dissection, coat a 24-well plate with PDL, diluted 1:100 in sterile water, overnight in the tissue culture incubator. 2. On the day of the dissection, rinse the plate three times with sterile water, aspirate the last wash, and allow it to dry briefly in a biosafety cabinet for few minutes. 3. Defrost an aliquot of laminin in the refrigerator or on ice, dilute it 1:300 in sterile water, and coat the plate for 2–4 h in the tissue culture incubator (see Note 15). 4. Sterilize all the dissection instrument and the surfaces with 70% ethanol, paying particular attention to ensure they are dry. Do not introduce ethanol in any solution. 5. Humanely euthanize the pregnant mouse, and then extract the embryos by cutting the peritoneum and removing the uterus. 6. Transfer the uterus to a dish containing L-15 medium, and expose the embryos surrounded by vitelline sac. 7. Identify the placenta (a dark red tissue), and cut beneath it to remove the embryo from the vitelline sac. 8. Cut the umbilical cord and transfer the embryo to a clean dish containing cold L-15 medium. 9. The technique to dissect eDRGs is very well described in Chapter 4 of this book. Remove the embryo’s extremities and organs, pull aside the skin of the back of the embryo superficially and carefully, turn the embryo belly up, and make small incisions along the vertebral column to expose the spinal cord. Once the spinal cord has been exposed fully, pull the body tissue laterally to expose the DRGs. 10. With forceps, carefully remove the spinal cord with the DRGs attached (see Note 16) and place it in a new dish with cold L-15 medium (see Note 17).

Schwann Cell-Neuron Coculture

105

11. With fine #5 forceps collect each ganglia and place them one by one in a tube with L-15 medium. Collect 40–60 ganglia for each 24-well plate to be prepared. 12. Gently spin down the tube for 15 s on a table top centrifuge, and carefully remove the L-15 as much as possible. 13. Add 1 mL of 0.05% trypsin with EDTA into each tube, and incubate for 20 min at 37  C in the incubator. 14. During this incubation time, it is advisable to aspirate the laminin solution from the coated 24-well plate and leave the plate open in a biosafety cabinet, so that the plate dries completely. 15. Spin the cell suspension for 15 s, and remove the supernatant (trypsin) without disturbing the pellet of ganglia. 16. Add 1 mL of eDRGs medium, and pipette with a P1000 tip to disrupt the ganglia. 17. Centrifuge at room temperature for 5 min at 800  g, and then resuspend the pellet in 60 μL of eDRG medium. 18. Place a drop of 2.5 μL of cell suspension in the top half of each well, and place the plate in the incubator for about 20–24 min (see Note 18). During this incubation time, the cells adhere to the plastic, but the volume of the medium is enough that the “spot” does not dry up. 19. Slowly add 500 μL of warmed eDRG medium, and place the dishes in the 37  C incubator. This medium contains 5-fluoro20 -deoxyuridine and uridine, which exert an antimitotic effect, limiting the growth of endogenous glia and fibroblasts, while not affecting neuronal survival. 20. Change half of the medium after 4 days. At this time, the cell bodies have extended neurites. There should be no more than a couple of cell bodies outside the spot area, and all the area surrounding the cell body spot should be covered in a thick axonal network. 3.4 Rat Schwann Cell Addition to Mouse Neuronal Cultures

1. Seven days after plating eDRG neurons, the neurites are fully developed. Examine the plate at the microscope, and exclude the wells in which cell bodies are present outside the spot area (see Note 19). 2. Detach the rat Schwann cell from the dish with trypsin as described above in point 1 of Subheading 3.2, and count the cells. 3. Rinse the wells with DMEM to remove the antimitotic reagent. 4. Plate 200,000 rat SCs to each well in C-medium. 5. About 24 h after plating Schwann cells onto neurons, remove the C-medium and replace it with freshly prepared medium

106

Elisabetta Babetto

containing 50 μg/mL ascorbic acid to induce the engulfment of neurites by Schwann cells (ascorbic acid promotes the formation of a basement membrane around Schwann cells and drives glial differentiation). 6. Maintain the cultures in this medium until the desired experimental endpoint. Typically, to test axon injury, the cultures can be maintained until day in vitro (DIV) 12 or longer. 7. Change the C-medium with ascorbic acid every other day, making sure to prepare fresh aliquots of ascorbic acid for each experiment. 3.5

Blade Axotomy

1. Adjust the magnification of a dissection microscope to a setting that allows the operator to clearly visualize the cluster of neuronal cell bodies. 2. Steadily place the flat blade at an angle near the cell body area while looking through the eyepiece, and slowly lower it to cut the neurites. Do not drag the blade as doing so would cause detachment of the neurite network; instead press it on the plastic of the well as to crush the neurites. Injure the neurites by making cuts around the cell body area, but leave one side of the well uninjured to prevent complete detachment of the culture during the following experimental procedure of immunostaining (Fig. 2). Indeed, the neurites form a network that

Fig. 2 Epifluorescent tilescan of a 24-well plate containing axotomixed eDRGs immunostained for Beta-III-tubulin and neurofilament heavy chain. The blade cuts are clearly visible, and demark the edge between the cell body spot and the axonal network

Schwann Cell-Neuron Coculture

107

can lose its attachment to the plastic if degenerated, especially when rinsed multiple times during immunostaining procedures. Having an injured area that does not degenerate helps to anchor the network to the plastic. However, make sure that the uninjured area is confined to one location of the well, clearly distinguishable from the area occupied by the injured neurites, not to confound the experiment. 3. Place the plate back in the incubator for the remainder of the time course. Typically, axon degeneration occurs between 24 and 48 h if axons are injured at DIV 12 after 6 days of contact with Schwann cells. 4. Axon degeneration can be assessed by a variety of methods. The most straightforward is the immunostaining of axonal cytoskeletal proteins, such as Beta-III-tubulin or neurofilament subunits (Fig. 2). Particular attention should be given to the health of the cell bodies as axon degeneration can be secondary to apoptosis.

4

Notes 1. Different batches of serum can influence the experimental outcome, leading to slightly different timecourses. It is therefore recommended to lot-test the serum before buying it. This allows the investigator to buy larger quantities and to reduce experimental variability. Aliquots can be stored for several years until the expiration date. Moreover, it is best to avoid repeated freeze–thawing cycles to ensure experimental reproducibility. 2. Laminin should be aliquoted in small working volumes, and defrosted in the fridge or on ice. It should not be refrozen after thawing to prevent polymerization. 3. Several commercially available PDL formulations can be used for this protocol, but high-molecular weight ranges are preferred. 4. It is advisable to lot-test B27 given the high variability experienced from different lots of this reagent over the last few years [17]. 5. NGF is a labile reagent. It is stable in the freezer for a maximum of 6 months in our experience. If neuronal cultures suddenly stop extending neurites, using fresh NGF often resolves the problem. 6. Ascorbic acid should be prepared freshly (working quickly and in the dark to avoid oxidation), filtered with a 0.22-μm filter, and immediately used or frozen at
80  C.

108

Elisabetta Babetto

7. E14.5 CD1 mice are preferred; however, the protocol can be carried out also with E13.5 and E15.5 mice. Time mating should be set up to obtain the specific embryonic age. The day when the semen plug is noticed is counted as 0.5. Likewise, the cultures can also be prepared from C57BL/6 mice. 8. All the animal work should follow the institutional and federal animal welfare regulations and guidelines. 9. At this age (P3), the bone structure of the newborn pup is still soft and the vertebral column can be easily broken using laminectomy forceps to expose the full length of the sciatic nerve. For this, it is helpful to pin the pup body on a sylgard dish by placing needles through its paws. 10. Scraping the strainer with a cell scraper and rinsing it with additional medium help ensure that all the cells pass through the strainer. 11. Shorter incubations (3 min) are possible as trypsinization can be harmful to Schwann cells. However, proper detachment of most of the cells needs to be ensured, to avoid positive selection of cells with poor adhesion. 12. It is important to avoid the formation of bubbles in the separation buffer after degassing as this would hinder the proper flow of the buffer through LS columns. 13. The flow-through should contain mostly fibroblast, although it is not uncommon to have some Schwann cells as well, because the anti-p75 antibody binding may not be 100% efficient. 14. Repeated passaging causes permanent changes in cell cultures that may affect the biological variant studied. In the case of Schwann cells, their ability to myelinate decreases over time [18]. 15. The coating with laminin can be started right before the dissection as dissection and dissociation should take no longer than 2 h to ensure cell viability. 16. Occasionally, the DRGs remain connected to the body and the spinal cord remains devoid of ganglia when removed from the body. In this case, transfer the body to a clean dish with cold L-15 and pick the ganglia from the body using forceps. To visualize the ganglia, it is helpful to use a swan lamp and angle the light source laterally to create a good contrast. 17. If there are still traces of blood on the forceps or on the tissue itself at this stage, it is advantageous to transfer the spinal cord to another clean dish with cold L-15 to remove the blood further.

Schwann Cell-Neuron Coculture

109

18. This incubation time varies based on the type of plastic of the dish, and temperature and humidity of the incubator. Therefore, this time may need to be empirically adjusted in each laboratory. 19. Only PDL and laminin coating allow for the formation of a proper neuronal spot. Matrigel and collagen, two commonly used coating reagents for myelination studies, do not provide enough substrate for the neurons to remain attached in a spot. On these substrates, neuronal cell bodies detach when the medium is added after incubation, or migrate into the axonal area a few days after plating. References 1. Brosius Lutz A, Chung WS, Sloan SA, Carson GA, Zhou L, Lovelett E, Posada S, Zuchero JB, Barres BA (2017) Schwann cells use TAM receptor-mediated phagocytosis in addition to autophagy to clear myelin in a mouse model of nerve injury. Proc Natl Acad Sci U S A 114: E8072–e8080 2. Stahl, BA, Jaggard, JB, Keene, AC (2019) Sleep regulates the glial engulfment receptor Draper to promote Wallerian degeneration bioRxiv, 716894 3. Vaquie A, Sauvain A, Duman M, Nocera G, Egger B, Meyenhofer F, Falquet L, Bartesaghi L, Chrast R, Lamy CM, Bang S, Lee SR, Jeon NL, Ruff S, Jacob C (2019) Injured axons instruct schwann cells to build constricting actin spheres to accelerate axonal disintegration. Cell Rep 27:3152–3166.e7 4. Wong KM, Babetto E, Beirowski B (2017) Axon degeneration: make the Schwann cell great again. Neural Regen Res 12:518–524 5. Funfschilling U, Supplie LM, Mahad D, Boretius S, Saab AS, Edgar J, Brinkmann BG, Kassmann CM, Tzvetanova ID, Mobius W, Diaz F, Meijer D, Suter U, Hamprecht B, Sereda MW, Moraes CT, Frahm J, Goebbels S, Nave KA (2012) Glycolytic oligodendrocytes maintain myelin and long-term axonal integrity. Nature 485:517–521 6. Lee Y, Morrison BM, Li Y, Lengacher S, Farah MH, Hoffman PN, Liu Y, Tsingalia A, Jin L, Zhang PW, Pellerin L, Magistretti PJ, Rothstein JD (2012) Oligodendroglia metabolically support axons and contribute to neurodegeneration. Nature 487:443–448

7. Beirowski B, Babetto E, Golden JP, Chen YJ, Yang K, Gross RW, Patti GJ, Milbrandt J (2014) Metabolic regulator LKB1 is crucial for Schwann cell-mediated axon maintenance. Nat Neurosci 17:1351–1361 8. Yang J, Wu Z, Renier N, Simon DJ, Uryu K, Park DS, Greer PA, Tournier C, Davis RJ, Tessier-Lavigne M (2015) Pathological axonal death through a MAPK cascade that triggers a local energy deficit. Cell 160:161–176 9. Babetto E, Beirowski B, Janeckova L, Brown R, Gilley J, Thomson D, Ribchester RR, Coleman MP (2010) Targeting NMNAT1 to axons and synapses transforms its neuroprotective potency in vivo. J Neurosci 30:13291–13304 10. Babetto E, Beirowski B, Russler EV, Milbrandt J, DiAntonio A (2013) The Phr1 ubiquitin ligase promotes injury-induced axon self-destruction. Cell Rep 3:1422–1429 11. Gerdts J, Brace EJ, Sasaki Y, DiAntonio A, Milbrandt J (2015) SARM1 activation triggers axon degeneration locally via NAD(+) destruction. Science 348:453–457 12. Gilley J, Coleman MP (2010) Endogenous Nmnat2 is an essential survival factor for maintenance of healthy axons. PLoS Biol 8: e1000300 13. Miller BR, Press C, Daniels RW, Sasaki Y, Milbrandt J, DiAntonio A (2009) A dual leucine kinase-dependent axon self-destruction program promotes Wallerian degeneration. Nat Neurosci 12:387–389 14. Shin JE, Miller BR, Babetto E, Cho Y, Sasaki Y, Qayum S, Russler EV, Cavalli V, Milbrandt J, DiAntonio A (2012) SCG10 is a JNK target in

110

Elisabetta Babetto

the axonal degeneration pathway. Proc Natl Acad Sci U S A 109:E3696–E3705 15. Wang J, Zhai Q, Chen Y, Lin E, Gu W, McBurney MW, He Z (2005) A local mechanism mediates NAD-dependent protection of axon degeneration. J Cell Biol 170:349–355 16. Dong Z, Dean C, Walters JE, Mirsky R, Jessen KR (1997) Response of Schwann cells to mitogens in vitro is determined by pre-exposure to

serum, time in vitro, and developmental age. Glia 20:219–230 17. Cressey D (2009) Neuroscientists claim growing pains. Nature 459:19 18. Rutten MJ, Janes MA, Chang IR, Gregory CR, Gregory KW (2012) Development of a functional schwann cell phenotype from autologous porcine bone marrow mononuclear cells for nerve repair. Stem Cells Int 2012:738484

Chapter 9 Establishing Myelinating Cocultures Using Human iPSC-Derived Sensory Neurons to Investigate Axonal Degeneration and Demyelination Alex J. Clark Abstract Complex signaling between Schwann cells and axons are vital for peripheral neuron development, myelination, and repair. The interaction between these two cell types can be modeled in vitro by coculturing rodent Schwann cells and neurons together. These have in the past been used with great success to help unravel the bidirectional signaling mechanisms that lead to Schwann cell proliferation and myelination. To provide more translatable potential, we have developed myelinating cocultures using human, induced pluripotent stem cell (iPSC)-derived neurons. Under the right conditions, the human neurons are efficiently myelinated by rat Schwann cells, demonstrating successful cross-species signaling. This chapter describes all the necessary steps to generate these myelinating cocultures and methods to investigate and quantify various aspects of myelination. The myelinating cocultures can be maintained in excellent health for over 1 year, facilitating their use to study developmental or chronic disease processes. With this in mind, we have used the cocultures to model a sensory neuropathy which displays clinically with both axonal and demyelinating features. In the cocultures, we found evidence of extensive axonal degeneration and demyelination demonstrated by axonal swelling and fragmentation, and myelin disintegration. The myelinating cocultures can therefore be used to study complex, human disease processes that result in both axonal and myelinassociated degenerative processes. Key words Induced pluripotent stem cells, Schwann cells, Sensory neuron, Coculture, Myelination, Degeneration

1

Introduction The intimate relationship between axons and Schwann cells regulates the health and function of each cell type. In fact, the anatomical and metabolic association is so close that each cell type is ultimately dependent upon each other. Complex axoglial signaling results in the formation of the myelin sheath, a compacted spiral wrap of a lipid rich membrane that protects the underlying axon and facilitates fast saltatory conduction. Axon to glia signaling regulates the ensheathment fate of the axon, through neuregulin-

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_9, © Springer Science+Business Media, LLC, part of Springer Nature 2020

111

112

Alex J. Clark

I type III (TIIINRG1) signaling, and facilitates the proliferation of Schwann cells in development and after nerve injury. Glia to axon signaling aids the long-term survival of the axon and results in the formation of specialized breaks in the myelin sheath known as nodes of Ranvier [1]. Myelin sheaths are formed during early development and are synthesized by oligodendrocytes in the central nervous system (CNS) and Schwann cells in the peripheral nervous system (PNS). In the periphery, damage of the axon can greatly alter the signaling between axon and Schwann cell and results in a cascade that drives the dedifferentiation of Schwann cells [2]. Neuron and Schwann cell cocultures allow the complex axo-glial signaling to take place in vitro and consequently recapitulate many of the sequential features of myelination. These include Schwann cell alignment and ensheathment of the axon, formation of the basal lamina, and finally membrane wrapping and compaction. The interaction between these two cells types in vitro also results in the formation of nodes of Ranvier, with voltage-gated sodium channels clustered at the node and voltage-gated potassium channels clustered at the juxtaparanode, which are separated by the Schwann cell paranodal loops. Induced pluripotent stem cells (iPSCs) have in recent years become a fundamental tool for studying human cells in vitro [3]. This includes modeling neurological disorders, which, through the reprogramming of patient-derived samples, allows the impact of mutations to be studied within a polygenetic background [4, 5]. The utility of generating human neurons in vitro also means they can be used for pharmacological studies, particularly, to study toxicology and immunopathological mechanisms. We have used iPSC-derived sensory neurons cocultured with rat Schwann cells to generate myelinating cocultures. We have previously reported that human neurons can be effectively and reliably myelinated by rat Schwann cells, and these cultures display all the expected features of axo-glial interactions [6]. Furthermore, the longevity of our cocultures is markedly greater than entirely rodent cocultures, with cultures able to be maintained in excellent health for over 1 year. Here, I describe in detail the necessary steps to generate these cocultures, how to use them to study various aspects of myelination and highlight some of the potential pitfalls which may be encountered.

2

Materials

2.1 iPSC Culture and Neuronal Differentiation

1. iPSC medium (StemFlex). 2. Matrigel, with and without phenol. 3. Antibiotic/antimycotic.

iPSC to Study Myelination and Axonal Degeneration

113

4. 6- and 24-well tissue culture plates. 5. PBS. 6. Rho-associated protein kinase (ROCK) inhibitor (Y-27632). 7. Mouse embryonic fibroblast (MEF) conditioned medium . 8. Fibroblast Growth Factor basic (FGF2). 9. Knockout Serum Replacement medium (KSR): Knockout DMEM, 15% knockout-serum replacement, 1% Glutamax, 1% nonessential amino acids, 1% antibiotic/antimycotic, and 100 μM β-mercaptoethanol. 10. N2 medium: Neurobasal medium, 2% B27 supplement, 1% N2 supplement, 1% Glutamax, and 1% antibiotic/antimycotic. 11. Dual SMAD LDN-193189.

inhibitors

(SMADi):

SB431542

and

12. 3 small inhibitors (3i): CHIR99021, SU5402, and DAPT. 13. Human recombinant neurotrophic factors: nerve growth factor (β-NGF), brain-derived neurotrophic factor (BDNF), glialderived neurotrophic factor (GDNF), and neurotrophin-3 (NT3). 14. Coverslips. 15. Hydrochloric acid (37%). 16. Nitric acid (70%). 17. Cytosine arabinoside (araC). 18. 70% (v/v) ethanol diluted in MilliQ water. 2.2 Myelinating Cocultures

1. Schwann cell expansion medium: DMEM/F12, 10% fetal bovine serum, 200 ng/ml NRG1-β1 EGF domain, 10 ng/ml β-NGF (mouse-recombinant), and 4 μg/ml forskolin. 2. Schwann cell cryopreservation medium: 60% DMEM/F12, 20% FBS, and 20% DMSO. 3. Schwann cell basal medium: DMEM/F12, 5 μg/ml insulin, 100 μg/ml transferrin, 25 ng/ml β-NGF (humanrecombinant), 25 ng/ml Selenium, 25 ng/ml thyroxine, 30 ng/ml progesterone, 25 ng/ml triiodothyronine, and 8 μg/ml putrescine. 4. Myelination medium: N2 medium, 1:300 phenol-free Matrigel, 5% charcoal-stripped FBS, 25 ng/ml β-NGF (humanrecombinant), and 50 μg/ml ascorbic acid.

2.3 Immunocytochemistry

1. Electron microscopy grade paraformaldehyde (EM-PFA). 2. PBS. 3. Methanol. 4. Normal goat serum.

114

Alex J. Clark

5. Triton-X 100. 6. Superfrost plus microscopy slides. 7. Vectorshield antifade mounting medium. 8. Fine forceps. 9. Nail varnish. 2.4

Antibodies

1. Neurofilament heavy chain (NF200) (Sigma, raised in mouse) 1:400. 2. Myelin basic protein (MBP) (Abcam, raised in rat) 1:400. 3. N-cadherin (BD Biosciences, raised in rabbit) 1:400. 4. Pan-sodium channel (Sigma, raised in mouse) 1:400. 5. Ezrin-Raxdixin-Moesin (ERM) (Cell Signalling, raised in rabbit) 1:500. 6. Caspr (gift from Prof Baht, raised in guinea pig) 1:400. 7. Pan-neurofascin (Abcam, raised in rabbit) 1:500. 8. Kv1.2 (Neuromab, raised in rabbit) 1:200.

3

Methods Carry out all cell culture procedures in a laminar flow hood. The hood surfaces and equipment within the hood (such as pipettes) should be sterilized before commencing work by wiping with 70% ethanol. While working, any plasticware, bottles, and reagents should also be sprayed and wiped with 70% ethanol before entering the culture hood. Do not spray or wipe any cell culture vessel with ethanol and allow the ethanol to fully evaporate from surfaces before placing cell culture vessels on them. The following sections will describe all the necessary steps required to establish myelinating cocultures using human iPSCderived neurons. This will include the initial thawing of iPSCs and their differentiation to sensory neurons, followed by establishing the myelinating cocultures and methods to investigate and/or quantify myelination.

3.1 Establishing Myelinating Cocultures 3.1.1 iPSC Thawing and Differentiation

1. Beforehand, coat plasticware with substrate of choice according to manufacturer’s recommendations. For the following work, we have used Matrigel as the substrate and StemFlex as the iPSC medium. Matrigel should be diluted in Knockout DMEM. Matrigel is a solubilized protein extract from Engelbreth-Holm-Swarm mouse sarcoma [7]; because it is secreted by living cells, the protein content varies from batch to batch. Therefore, to dilute the Matrigel, follow the dilution factor given for each batch of Matrigel. Concentrated Matrigel will gel at temperatures over 10
C, therefore care must be

iPSC to Study Myelination and Axonal Degeneration

115

taken when thawing the Matrigel to aliquot into working sizes. Matrigel must be thawed overnight, on ice in a refrigerator, and tips and tubes should be chilled before pipetting. We aliquot concentrated Matrigel into the volume required to formulate 12.5 ml diluted Matrigel solution (approximately 120 μl, but as discussed, this will vary from batch to batch). 2. Plasticware is then coated with the diluted Matrigel solution for a minimum of 1 h at 37
C. 3. iPSCs should be removed from the cryostorage facility and thawed quickly in 37
C water until a small piece of ice remains. Remove from the water and quickly dilute the contents of the vial 1:10 with sterile PBS (see Note 1). 4. Centrifuge at 400  g for 5 min at 5
C. 5. Resuspend the cell pellet in StemFlex medium with Rock Inhibitor (Y-27632, 10 μM), by gently pipetting up and down with a 1-ml pipette tip. The pellet should easily dissociate leaving no visible cell clumps. 6. Aspirate Matrigel from the culture plates/dishes and gently pipette the cell suspension directly on to the coated plasticware without washing. Do not allow the Matrigel to dry before adding the cell suspension. Approximately, one million iPSCs should be plated onto 9 cm2 (see Note 2). 7. Change the medium the following day and exclude ROCK inhibitor if cell to cell contact is observed, and if not, include ROCK inhibitor for an additional 24 h. 8. It is recommended to perform a passage before initiating the differentiation. Standardly, a passage using EDTA (see Note 3) would be performed 2 days after thawing (see Note 4). 9. Sensory neurons are differentiated according to the protocol published by Chambers et al. (2012) (Fig. 1 and Table 1). The

Fig. 1 Timeline showing the crucial steps in establishing a myelination coculture. In order to start the sensory neuron differentiation, iPSC medium is changed for MEF-conditioned medium, which remains for 1–2 days or until the cells are approximately 50% confluent. An 11-day differentiation period then follows which uses dual SMAD inhibitors (SMADi) and 3 small inhibitors (3i) to generate sensory neurons. During the differentiation period the medium transitions from KSR to N2. Neurons can then be matured for weeks or months, depending on the experiment. Once Schwann cells are seeded into the neuronal culture, a 1-week period of Schwann cell basal medium follows, to allow proliferation and alignment. Myelination is initiated by changing to myelination medium. The myelinating cocultures are extremely stable and have been maintained for over 1 year with no loss of myelin or neuronal integrity

116

Alex J. Clark

Table 1 Sensory neuron differentiation protocol with medium composition and procedure Day

Medium composition and procedure

Day-1

Passage iPSCs using EDTA and change the medium to MEF-conditioned + 10 ng/ml FGF-2. Plate out at low density to achieve approximately 50% confluency 24 h later, if less than 50% maintain in MEF-conditioned medium until this is achieved

Day 0

Differentiation should be started when cells are approximately 50% confluent Medium: KSR medium Inhibitors: 2 SMAD inhibitors (10 μM SB431542 and 100 nM LDN-193189)

Day 1 Day 2 Day 3

Medium: KSR medium Inhibitors: 2 SMAD inhibitors (10 μM SB431542 and 100 nM LDN-193189) +3i (3 μM CHIR99021, 10 μM SU5402, 10 μM DAPT)

Day 5

Medium: 75% KSR medium, 25 % N2 medium Inhibitors: 2 SMAD inhibitors (10 μM SB431542 and 100 nM LDN-193189) +3i (3 μM CHIR99021, 10 μM SU5402, 10 μM DAPT)

Day 6 Day 7

Medium: 50% KSR medium, 50% N2 medium Inhibitors: 3i (3 μM CHIR99021,10 μM SU5402,10 μM DAPT)

Day 8 Day 9

Medium: 25% KSR medium, 75% N2 medium Inhibitors: 3i (3 μM CHIR99021, 10 μM SU5402, 10 μM DAPT)

Day 10

Medium: N2 medium Inhibitors: 3i (3 μM CHIR99021,10 μM SU5402, 10 μM DAPT)

Day 4

Day 11+ Replate the immature neurons onto coverslips or plasticware using TrypLE Medium: N2 Medium Inhibitors: ROCK inhibitor (10 μM Y-27632) for 24 h and 3 μM CHIR99021—until day 14 Growth Factors: NGF, BDNF, GDNF, NT3 (all at 25 ng/ml)

confluency of cells when initiating the differentiation is very important and has a large impact on the efficiency of sensory neurogenesis (see Note 5). Aim for approximately 50% confluency when beginning the inhibitor treatment on day 0 (Table 1). The differentiation protocol involves the use of dual SMAD inhibition and 3 small inhibitors (3i) to generate bona fide sensory neurons over an 11-day period while transitioning the medium from KnockOut Serum Replacement (KSR) medium to N2 medium (Fig. 1 and Table 1). 10. Immature neurons are plated out on day 11 or 12. At this stage, neurons can be plated onto coverslips (ideal for immunocytochemistry, electrophysiology, or live cell imaging) or larger plastic dishes (optimal if intending to extract protein, mRNA, DNA, etc.). Myelinating cocultures have successfully been cultured on both forms of tissue culture surfaces.

iPSC to Study Myelination and Axonal Degeneration

117

11. Glass coverslips should be acid treated (1/3 hydrochloric acid (37%), 2/3 nitric acid (70%)) for a minimum of 1 h (3–4 h is recommended) in a fume hood. Occasionally swirl the bottle, so all coverslips are treated. Wash extensively with distilled water, tip out excess water, and top up with 70% ethanol. Coverslips can be stored in ethanol for several months. We have found acid treatment to greatly improve the longevity of iPSC-derived neuronal cultures. 12. Remove coverslips from ethanol and allow to dry in coverslip racks. Place coverslips in a culture plate (13 mm coverslips are ideal in 24-well plates). Coat the coverslips with bubbles of Matrigel (see Note 6). Coat for at least 1 h within the culture hood to minimize movement of the culture plates and therefore reducing the risk of the “bubbles” flowing off the coverslip. 13. The efficiency of neuronal differentiation can vary between iPSC lines. Commonly, we achieve >90% neuronal conversion using the protocol described above. However, if neuronal numbers are particularly low, a purification step using MACS is recommended at this stage (see Note 7). 14. Determine cell number and plate out 30,000 cells per 13 mm coverslip. Remove most of the Matrigel before plating, and plate out the cells in a volume of 75 μl. Use N2 medium (see Note 7) with CHIR99021 and 25 ng/ml human recombinant NGF, BDNF, GDNF, and NT3. Do not let Matrigel dry before plating. Carefully, move the plate into an incubator and allow the cells to adhere for 90 min. 15. Gently flood the cells with 275 μl of medium per well (see Note 8). 3.1.2 Maturation of Neurons

1. Change medium twice weekly after replating using N2 medium with 25 ng/ml human recombinant NGF, BDNF, GDNF, and NT3. CHIR99021 should be included in the medium for 3 days after replating. 2. Nonneuronal cells can often be observed soon after neuronal plating on day 11. It is important these are eradicated before they extensively proliferate. Include 1 μM araC (cytosine arabinoside) the day after replating. Remove araC at the next medium change, then include it in the following change. Two cycles of 1 μM is usually sufficient to eliminate all nonneuronal proliferating cells. If necessary, include an additional cycle if nonneuronal cells are still visible (see Note 9). 3. Neurons can now be maintained in culture with twice weekly medium changes for over 1 year. Include phenol-free Matrigel (1:300 dilution directly into the neuronal medium) once per week.

118

Alex J. Clark

3.1.3 Coculture with Rodent Schwann Cell and Initiation of Myelination

1. Schwann cells are derived from early postnatal rat pups (P3–P5) by dissecting the sciatic nerve and brachial plexus. 2. These are enzymatically digested in collagenase/dispase solution and mechanically dissociated using fire-polished glass Pasteur pipettes. The cell suspension (usually still containing visible “clumps”) is plated onto poly-D-lysine/laminin coated culture flasks. 3. Schwann cells are expanded in expansion medium and cryopreserved in 60% DMEM/F12, 20% FBS, and 20% DMSO. When cryopreserving the Schwann cells, aliquot one million cells per vial. 4. When the neurons are ready for Schwann cell coculture (see Note 10), change the neuronal medium to Schwann cell basal medium (350 μl/well is sufficient for a 24-well plate). 5. Thaw a cryotube of frozen Schwann cells in warm water until a small pellet of ice remains in the tube (see Note 1). Quickly dilute the contents of the cryotube 1:10 in PBS. 6. Centrifuge at 400  g for 5 min. 7. Resuspend the cell pellet in Schwann cell basal medium, by gently pipetting up and down with a P200 pipette (see Note 11). 8. 25 μl of the Schwann cell suspension now needs to be added to each well of iPSC-derived neurons. Gently mix the Schwann cell suspension before each well. Place the pipette tip containing 25 μl of Schwann cell suspension close to the surface of the medium covering the neurons. Slowly pipette out small droplets (approximately 5 μl at a time) and touch the droplet to the surface of the medium. Move the pipette tip around the coverslip, aiming for a droplet in each corner and one in the middle to evenly distribute the Schwann cells around the neuronal culture. Take care not to touch the underlying neurons with the pipette tip. When viewed under the microscope, the Schwann cells can be seen floating in the medium above the adhered neurons. 9. Gently replace the culture plate into the incubator. 10. Schwann cells will sink and adhere to the coverslip within 2 h. Do not move the culture plate until the next day (see Note 12). 11. For the next week, the cocultures remain in Schwann cell basal medium to stimulate proliferation and alignment of Schwann cells (Fig. 1). Replace the medium 1 day (to remove any dead cells) and 4 days following Schwann cell seeding. 12. Seven days after Schwann cell seeding, replace the medium with myelination medium. 13. Following the transition into myelination medium, Schwann cells will rapidly upregulate the well characterized

iPSC to Study Myelination and Axonal Degeneration

119

Fig. 2 Representative DIC images. (a) A representative example of a pure neuronal culture 4 weeks after replating onto coverslips. The neuronal density and axonal coverage in this image would be optimal for coculturing with Schwann cells. Representative DIC images of myelinating cocultures 2 (b), 6 (c), and 20 (d) weeks after initiating myelination. Myelinated axons appear darker with DIC microscopy

promyelination transcription factor—KROX20 [6]. Furthermore, as Schwann cells ensheath the axon, typical molecular markers are expressed. These include N-cadherin, demonstrating a functional interaction between human neurons and rodent Schwann cells [8] and collagen IV, confirming the onset of basal lamina formation. Internodes first become visible approximately 7–10 days after initiation of myelination. After 2–3 weeks, internodes should be abundant throughout the cultures, and myelination continues to increase with age (see Fig. 2) [6]. 3.2 Methods to Investigate Myelination 3.2.1 Immunocytochemistry

Immunocytochemistry is an excellent tool to qualitatively investigate the localization of proteins throughout the internode and nodal complex, and quantitatively analyze the extent of myelination. Myelinating cocultures are highly amenable to immunocytochemistry and below are general protocols to immunostain the myelin internodes and proteins of the nodal complex. Up until recently, techniques to quantify myelination in vitro have predominantly relied upon counting individual myelin segments. The following will also describe a far quicker, automated image processing technique for accurately determining levels of myelination that is normalized to axonal outgrowth. In order to quantify myelination, markers for the neurons and myelin internodes are required. Therefore, a protocol will be described for the immunostaining of myelinating cocultures using antibodies directed against myelin basic protein (MBP) and neurofilament-heavy chain (NF200) (see Note 13). MBP and NF200 immunocytochemistry protocol (see Note 14): 1. Remove culture medium. 2. Wash cell in PBS. 3. Fix cells in 1% paraformaldehyde for 20 min at room temperature.

120

Alex J. Clark

Fig. 3 Representative image of an immunostained myelinating coculture. Using the immunocytochemistry protocol described above the neurons (NF200) and myelin segments (MBP) can be stained with very high contrast and no background noise

4. Wash cells two times in PBS. 5. Permeabilize cells in ice cold methanol (100%) for 20 min on ice. 6. Wash three times in PBS. 7. Incubate with primary antibodies in PBS with 1 mg/ml serum (see Note 15) overnight at 4
C. (a) Myelin basic protein (raised in rat) 1:400. (b) NF200 (raised in mouse) 1:500. 8. Wash three times in PBS. 9. Incubate with secondary antibodies in PBS with 1 mg/ml serum at room temperature for 2 h in the dark. Secondary antibodies below have been used in Fig. 3, these can be changed according to the individual user’s preferences. (a) Alexa Fluor Goat anti-Rat 488. (b) Alexa Fluor Goat anti-Mouse 546. 10. Wash three times in PBS. 11. Place a drop of Vectorshield antifade mounting medium (with or without DAPI) on a Superfrost Plus microscopy slide. 12. Carefully remove coverslips from the culture plate using fine forceps and briefly blot the edge of the coverslip on tissue to remove excess PBS. 13. Place the coverslip cell-side down on the droplet of Vectorshield, ensuring no bubbles are trapped underneath the coverslip. 14. Place tissue on top of the coverslip to remove excess Vectorshield. 15. Seal the edge of the coverslip with nail varnish.

iPSC to Study Myelination and Axonal Degeneration

121

Fig. 4 Nodal structures can be detected in the myelinating cocultures by immunohistochemistry. (a) Demarkaction of the node (p-Nav), paranode (Caspr), and internode (MBP) is shown. (b) The node (ERM), paranode (Caspr), and internode (MBP) is shown. (c) The node and paranode (pan-neurofascin) and internode (MBP) is shown

The same immunocytochemistry protocol described above can also be used to visualize nodal structures (Fig. 4). We have successfully used antibodies against the following proteins to delineate nodal structures:

3.2.2 Antidisialosyl Antibodies Induce Axonal Degeneration and Demyelination

l

Pan-sodium channel (Node).

l

Ezrin-Raxdixin-Moesin (ERM) (Node).

l

Caspr (Paranode).

l

Pan-neurofascin-binds to NF-155 and NF-186 (Paranode and Node, respectively).

l

Kv1.2 (Juxtaparanode).

The PNS is highly enriched in many different myelin- and neuronal-associated gangliosides. Gangliosides which contain disialosyl moieties are expressed by peripheral sensory neurons, and autoantibodies which bind to these gangliosides are associated with axonal degeneration and demyelination. We have therefore exploited the myelinating cocultures to investigate the topographical targets and pathological effects of a human IgM antidisialosyl antibody (hADA), associated with CANOMAD (chronic ataxic neuropathy with ophthalmoplegia M-protein, cold agglutinins, and antidisialosyl antibodies), a condition with both axonal and demyelinating features [9]. After overnight incubation, hADA binding was observed at areas of exposed axolemma including the node of Ranvier and unmyelinated axons (Fig. 5a, b). After overnight antibody incubation with a source of complement (normal human serum, NHS), extensive axonal degeneration was observed, evidenced by blebbing and fragmentation of the axons. In addition, characteristics of demyelination were seen, with internodes swelling and breaking apart (Fig. 5e). Axons and myelin internodes

122

Alex J. Clark

Fig. 5 Human antidisialosyl antibodies cause axonal degeneration and demyelination. (a) hADA (green) binds to nodal axolemma (flanked by Caspr, red) and to (b) unmyelinated axons (NF200, red), without any reactivity to Schwann cells (S100, blue). (c and d) NHS or hADA alone had no acute effect on the myelinating coculture. (e) Overnight incubation with hADA followed by 20% NHS for 2 h leads to acute axonal degeneration. (g) Chronic exposure of hADA on established myelinated cocultures led to demyelination without any loss of the underlying axon, and (f) no change in myelin integrity was observed in control conditions. (Adapted from Clark et al. [6])

remained intact when cocultures were acutely incubated with NHS or antibody alone (Fig. 5c, d); however, prolonged incubation with the antibody without any source of complement resulted in the demyelination of established myelinated cocultures, observed by clusters of disintegrating myelin surrounding intact axons (Fig. 5g). 3.2.3 Live FluoroMyelin Stain

In order to track myelination over time in live cells, we have utilized the myelin stain—FluoroMyelin. While traditional immunocytochemistry is unparalleled when requiring high resolution images, it is a relatively time-consuming technique requiring multiple incubation and wash steps. FluoroMyelin, however, is a 20-min, one-step procedure that can be performed on live cells. FluoroMyelin takes advantage of the high lipid content of myelin to provide a readout of myelination that can be imaged with standard fluorescent microscopes (Fig. 6a–c). One of the main advantages of FluoroMyelin is that it is nontoxic to the cells and therefore allows repeated incubations. In order to track myelination over time, we have reincubated myelinating cocultures with FluoroMyelin and imaged the same area at 4-, 8-, and 12-weeks post induction of myelination (Fig. 6d–f). The appearance of new myelin internodes can be observed as the cultures age. To apply this technique to our in vitro model of CANOMAD, we have used FluoroMyelin to track the demyelination induced by

iPSC to Study Myelination and Axonal Degeneration

123

Fig. 6 FluoroMyelin can be used to track myelination over time. (a) DIC image of myelinating cocultures with myelin visible as thicker and brighter segments. (b) 20-min incubation with FluoroMyelin clearly reveals the myelin segments. (c) Merged DIC with fluorescent images showing all myelin segments are labeled with FluoroMyelin. Reincubation of the same culture with FluoroMyelin at 4 (d), 8 (e), and 12 (f) weeks allows the appearance of new myelin segments to be observed

long-term exposure of hADA to the myelinating cocultures. We have imaged myelination at baseline using FluoroMyelin, prior to any hADA exposure. Four weeks later, after continuous exposure to hADA, we then fixed and immunostained the cultures (using the protocol above). By subsequently aligning the same region, we were able to identify myelin internodes that had degenerated over the 4-week period [6]. This provides a valuable technique to track myelination over time and quantify demyelination in response to a treatment. 3.2.4 Axo-glial Signaling

Axo-glial signaling is crucial for the initiation and maintenance of myelination and can be disturbed in pathological states [10]. We have demonstrated successful cross-species signaling to induce myelination of human neurons with rat Schwann cells. These myelinating cocultures are highly amenable to methods to investigate these complex signaling pathways. We have successfully used two approaches to interfere with the Neuregulin-erbB pathway. Firstly, viral transduction of an AAV containing a NRG1 type III construct was used to upregulate this signaling pathway, and an erbB receptor inhibitor was used to pharmacologically prevent signaling. If using a viral method to exclusively transduce the neurons, it is necessary to perform this step before Schwann cells are added to the culture. To overexpress NRG1 type III, we transduced when the neurons were 3 weeks post differentiation. The medium with the AAV was incubated with the neurons for 3 days, two medium changes were then performed before seeding the Schwann cells into the neuronal cultures 7 days later. The transduction of NRG1 type

124

Alex J. Clark

Fig. 7 Overexpression of TIIINRG1 significantly increases myelination in the cocultures. (a–d) Human iPSCderived neurons were transduced with an AAV containing a TIIINRG1 construct at a range of MOIs. Cultures were fixed for immunocytochemistry or lysed for protein extraction at 4 weeks post induction of myelination. Viral transduction had no effect on axonal outgrowth or morphology, with no visible signs of neurite blebbing (insets in a–d). Scale bar ¼ 50 μm. (f) Quantification of myelination was similar in (a) untransduced and (b) 1  104 MOI conditions, whereas a significant and dose-dependent increase in myelination was observed in (c) 1  105 (∗p < 0.05) and (d) 1  106 (∗∗∗p < 0.001) conditions. (f) In the transduced neurons, western blot shows the increase in both the full length (FL) (135 kDa) and cleaved-terminal fragment (CTF) (60 kDa) of TIIINRG1 compared to untransduced control. The calnexin control demonstrates an even protein loading across conditions. (Adapted from Clark et al. [6])

iPSC to Study Myelination and Axonal Degeneration

125

III led to a significant and dose-dependent increase in myelination (Fig. 7a–e), which corresponded with an increase in expression of both the full length and cleaved-terminal fragment of NRG1 type III compared to untransduced control (Fig. 7f). Conversely, blockade of NRG signaling using a selective and irreversible ErbB inhibitor (PD168393) dose dependently reduced myelination [6]. These experiments demonstrate that crucial signaling pathways can be investigated in a variety of ways using myelinating cocultures. 3.2.5 Semiautomated Quantification of Myelination Using ImageJ

Quickly quantifying the levels of myelination achieved in vitro is invaluable when assessing the contribution of treatments or genetic alteration on the rate and extent of myelination. The below will discuss an automated method we developed to quickly quantifying myelination while normalizing to neuronal coverage. This method circumvents the more traditional and time-consuming method of counting individual myelin segments and is performed using systematic random sampling to ensure objective sampling across the coverslip. The protocol below describes the quantification using images acquired from NF200/MBP immunostained coverslips (protocol above). Systematic random sampling was performed using Zen Black Software (Zeiss) on a Zeiss LSM 510 inverted confocal microscope. A digital diamond shape of 13 points (2 mm horizontal/vertical distance between each point) was centered on each coverslip. A tile scan (each 1792 μm  1792 μm in size, using either a 10 or 20 objective) was taken at each point, resulting in 13 nonoverlapping images that equally sampled all quadrants of the coverslip (31.4% of the total coverslip area) (Fig. 8). 1. Open the images in ImageJ and split the channels. Image>Color>Split Channels. 2. Create binary image using threshold ge>Adjust>Threshold (see Note 16).

tool.

Ima-

3. Select to measure the area of the thresholded images. Analyze>Set Measurements. Select “Area” and “Limit to Threshold.” 4. Measure area for Analyze>Measure.

both

NF200

and

MBP

images.

5. The proportion of axonal area covered by myelin can then be calculated by dividing NF200 area by myelin area, and multiplying by 100. For the representative images shown in Fig. 8, the calculation is shown below: 347,884  100 ¼ 3.2% (347,884/11,143) x 100 ¼ 3.2% 11,143 6. Therefore, 3.2% of axonal outgrowth is myelinated. Measuring MBP area was a novel approach to quantify myelination in vitro, therefore we cross-validated our results, by

126

Alex J. Clark

Fig. 8 Schematic demonstrating the method developed to quantify myelination in vitro. Tile scan images from systematically sampled regions across the coverslip are split into separate channels. The neuronal (NF200) and myelin (MBP) images are then thresholded to create a binary image using the clustering-based thresholding method—Otsu. Area measurements are then taken of the thresholded pixels from each channel. The percentage of axonal area that is myelinated can then be calculated

comparing MBP area measurements to the most commonly used alternative method (myelin segment count). Across 18 separate coverslips of myelinated cocultures, we counted the number of myelin segments and performed the area measurements described above. We found both set of results were highly correlated (Pearson’s r; r ¼ 0.962, n ¼ 18, p < 0.001) [6], and therefore, we now routinely use the quicker area measurement method to quantify myelination.

4

Notes 1. It is important not to warm the cells up at this stage, hence thawing until a small piece of ice remains. Speed is essential here as prolonged exposure to DMSO is toxic to the thawed cells. Therefore, quickly dilute the thawed cell suspension 1:10 in PBS. 2. At this stage, it is important not to swirl the culture plates or dishes as the cells will be pushed to the edges, resulting in an uneven density of cells across the dish. Once cells have been seeded, move the culture dish in an “up down, left right” movement (do not swirl), then leave the dish in the culture hood for 10 min. This allows the cells to adhere before moving the dish to the incubator. 3. There are several enzymatic methods for passaging iPSCs including TrypLE and Accutase; however, these individualize

iPSC to Study Myelination and Axonal Degeneration

127

the iPSCs, and as a result, ROCK inhibitor must be used to improve cell survival [11]. To improve cell survival during passaging, an enzyme-free method has been developed which uses EDTA to partially dissociate iPSCs into small aggregates [12]. As cell-to-cell contact is maintained in the aggregates, ROCK inhibitor is not required during the procedure. This protocol has been shown to increase cell survival, and in addition to not requiring ROCK inhibitor, it can be performed without pelleting the cells with centrifugation. 4. iPSC recovery after thawing varies according to cell line. It is normal to observe many dead and floating cells in the culture medium, the day after thawing. Therefore, if differentiating multiple iPSC lines, a passage before differentiation is a good opportunity to standardize the confluency across lines. 5. We have found that passaging early during the differentiation at day 2 or 3 can improve the efficiency of sensory neuron generation. Use EDTA and do not include ROCK Inhibitor. Do not passage any later than day 4. 6. Matrigel coating is only applied to the coverslip itself in a “bubble,” without coating the underlying plastic. After acid treatment, 75 μl of Matrigel should flow to the edge of a 13-mm coverslip without flowing off onto the plastic underneath. This coating strategy prevents cells from flowing off the coverslip when initially plating them out and also stops cells migrating off the coverslip onto the underlying plastic. If both coverslip and plastic are coated, as the neurons mature, the extensive neurite network that develops will extend over the edge of the coverslip linking with neurons adhered to the plastic. If the coverslip is then removed from the plate (for experimental reasons), it is common that the whole mesh of neurons becomes detached, making subsequent experiments impossible. Treating the coverslip with a Matrigel “bubble” entirely prevents this and allows easy removal of the coverslip with all neurons attached. 7. In order to MACS to purify neurons, we have successfully used Neural Crest Stem Cell Microbeads in combination with MS Columns and MiniMACS Separator magnets (all from Miltenyi Biotec), according to manufacturer’s instructions. 8. When flooding, the coverslips can sometimes float on top of the medium if an air bubble is trapped underneath. Coverslips can easily be pushed down with a pipette tip. Once any air bubbles are displaced, coverslips will not float up again. It is important to check coverslip have not floated up as the neurons will quickly dry out.

128

Alex J. Clark

9. It is important to have a break between araC treatments, so that the health of the neurons is not affected. Never include it for consecutive medium changes. 10. Neurons can be cocultured with Schwann cells once all nonneuronal cells have been depleted. Do not start the coculture if still treating the neurons with araC. In our hands, successful alignment and myelination has been achieved with neurons as young as 2 weeks (post differentiation) all the way up to 8 months old. However, we have observed that if neurons are left to mature before coculturing, Schwann cells align and myelinate quicker, and the total levels of myelination are greater. Therefore, our recommendation is to begin the coculture once the neurons are 4–6 weeks old (post differentiation). We find this optimal in terms of time/myelination, cost-benefit ratio. 11. In all, 25,000 Schwann cells should be plated onto a single 13 mm coverslip (scale up/down if necessary) in a volume of 25 μl. Therefore, resuspend the cell pellet in a volume to achieve 1000 cells per μl. For example, if frozen in one million aliquots, resuspend in 1 ml. This is then suitable for 40 coverslips (25 μl of 25,000 cells per coverslip). 12. When observing the cocultures the next day, it is not always possible to see the Schwann cells. If the neurons are relatively dense, Schwann cells will quickly migrate into neuronal cell body clusters and axonal bundles. Over the next 2–3 days, Schwann cells will proliferate extensively and become visible. 13. While NF200 is commonly used as a marker of large diameter DRG neurons in the mouse, recent literature has confirmed all human DRG neurons express NF200 [13, 14]. We have found that all neurons derived from the Chambers’ differentiation protocol are positive for NF200. Furthermore, in our hands, antibodies against NF200 consistently give excellent signal and as such are now standardly used in our lab to detect iPSCderived sensory neurons. 14. We recommend avoiding the use of Triton-X 100 to permeabilize when immunostaining myelinating cocultures. We have found that the myelin sheath will fragment when Triton-X 100 is used at concentrations greater than 0.1%. If Triton-X 100 is required, we recommend using it in a low concentration (0.05%) and to only include it in the wash steps after the methanol treatment. 15. Inclusion of serum stabilizes the antibodies and reduces background signal. Include serum from the same species that the secondary antibodies are raised in (in this case, normal goat serum).

iPSC to Study Myelination and Axonal Degeneration

129

16. It is important to standardize the thresholding across all images, so that myelination is not under- or overrepresented. We found the clustering-based thresholding method (Otsu) optimal for creating a binary image that detects all neurites and myelin segments. If a preloaded thresholding method is not suitable, it will be necessary to manually set the threshold level. As such, it is essential to be blind to any treatment or genotype. Otsu has been used to threshold both the NF200 and MBP images in Fig. 8.

Acknowledgments I would like to thank Alexander Davies and Malte Kaller (both at Nuffield Department of Clinical Neurosciences, University of Oxford) for their helpful assistance in generating figures. References 1. Taveggia C (2016) Schwann cells-axon interaction in myelination. Curr Opin Neurobiol 39:24–29 2. Jessen KR, Mirsky R (2008) Negative regulation of myelination: relevance for development, injury, and demyelinating disease. Glia 56:1552–1565 3. Karagiannis P, Takahashi K, Saito M et al (2018) Induced pluripotent stem cells and their use in human models of disease and development. Physiol Rev 99:79–114 4. Sandoe J, Eggan K (2013) Opportunities and challenges of pluripotent stem cell neurodegenerative disease models. Nat Neurosci 16:780–789 5. Corti S, Faravelli I, Cardano M et al (2015) Human pluripotent stem cells as tools for neurodegenerative and neurodevelopmental disease modeling and drug discovery. Expert Opin Drug Discov 10:615–629 6. Clark AJ, Kaller MS, Galino J et al (2017) Co-cultures with stem cell-derived human sensory neurons reveal regulators of peripheral myelination. Brain 140:898–913 7. Kleinman HK, Martin GR (2005) Matrigel: basement membrane matrix with biological activity. Semin Cancer Biol 15(5):378–386 8. Wanner IB, Wood PM (2018) N-cadherin mediates axon-aligned process growth and

cell–cell interaction in rat Schwann cells. J Neurosci 22:4066–4079 9. Willison HJ, O’Leary CP, Veitch J et al (2001) The clinical and laboratory features of chronic sensory ataxic neuropathy with anti-disialosyl IgM antibodies. Brain 124:1968–1977 10. Taveggia C, Feltri ML, Wrabetz L (2010) Signals to promote myelin formation and repair. Nat Rev Neurol 6:276–287 11. Watanabe K, Ueno M, Kamiya D et al (2007) A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat Biotechnol 25:681–686 12. Beers J, Gulbranson DR, George N et al (2012) Passaging and colony expansion of human pluripotent stem cells by enzyme-free dissociation in chemically defined culture conditions. Nat Protoc 7:2029–2040 13. Chang W, Berta T, Kim YH et al (2018) Expression and role of voltage-gated sodium channels in human dorsal root ganglion neurons with special focus on Nav1.7, species differences, and regulation by paclitaxel. Neurosci Bull 34:4–12 14. Rostock C, Schrenk-Siemens K, Pohle J et al (2018) Human vs. mouse nociceptors–similarities and differences. Neuroscience 387:13–27

Part III Ex Vivo Assays

Chapter 10 Preparation of Organotypic Hippocampal Slice Cultures for the Study of CNS Disease and Damage Claire S. Durrant Abstract Organotypic hippocampal slice cultures (OHSCs) retain in vivo-like neuronal architecture, synaptic connections, and resident cell populations but gain in vitro advantages of accessibility to experimental manipulation and observation. This chapter describes how to prepare OHSCs from neonatal mice to study mechanisms of neuronal damage, including synapse loss and quantifying Aβ-containing axonal swellings from Alzheimer’s disease transgenic mice. Key words Organotypic hippocampal slice culture, OHSC, Brain slice culture, Synapse, Hippocampus, Drug screening, Axonal swellings, Alzheimer’s disease, Neurodegeneration

1

Introduction Axonal damage, synapse loss, and eventual death of neurons in the brain are key features of many neuropathological processes, including neurodegenerative diseases, neuroinflammation, and trauma [1–5]. Understanding the processes that lead to neuronal damage is essential for the development of effective therapeutics, and choosing an appropriate experimental system is crucial for the success of this work. The central nervous system (CNS) is a complex environment consisting of multiple cell types, specialized synaptic circuits, and cytoarchitecture. While in vivo models retain this complexity, experimental work can be limited by the relative inaccessibility of the brain to live imaging, repeated measurements, and pharmacological manipulation (particularly deep regions such as the hippocampus). Studies in primary neuronal cultures mitigate many of these complications but risk oversimplification, particularly relating to non-cell-autonomous processes. Organotypic hippocampal slice cultures (OHSCs) represent an excellent compromise between in vivo and primary culture studies and are an ideal system to study mechanisms of CNS damage [6– 8]. OHSCs retain in vivo-like resident cell populations (including

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_10, © Springer Science+Business Media, LLC, part of Springer Nature 2020

133

134

Claire S. Durrant

neurons, astrocytes, microglia, and oligodendrocytes), synaptic connections (both excitatory and inhibitory), and neuronal architecture (preservation of the hippocampal subfields) for several months in vitro [6, 7, 9]. The in vitro nature of OHSCs permits tight control of the extracellular environment and allows for repeated non-invasive sampling of the culture medium [7, 10– 12]. Drugs and experimental treatments can be easily and rapidly assessed without issues of blood–brain barrier penetration or systemic toxicity that may mask relevant biological phenotypes. OHSCs represent an excellent refinement method for the use of animals in research as although tissue is still required from mice, experimental treatments or the effects of disease-related mutations can be explored in ex vivo tissue, without the animal experiencing the effects. In addition, many OHSCs can be made from a single pup, permitting control and test treatments to be compared in tissue from the same animal. This reduces the number of mice required and prevents inter-animal variability from confounding experimental effects [9]. This chapter will describe how to prepare OHSCs from neonatal mice for the study of neuropathological processes. OHSCs can be prepared from wild-type or genetically modified mice, and the methods for exploring mechanisms of neuronal damage can be easily adapted for study of genetic or pharmacological insultinduced pathology. In addition to a detailed description of OHSC preparation and maintenance, experimental outcomes, such as ELISA analysis of culture medium, synaptic protein and puncta quantification, immunofluorescence, and live imaging are discussed.

2

Materials Preparation of all reagents should be conducted under sterile conditions in a laminar flow cabinet or class II biological safety cabinet. If constitutive ingredients are weighed out under non-sterile conditions, solutions should be filtered through a 0.22 μm filter prior to use.

2.1

Brain Removal

1. Thick forceps with the end taped together (used to perform cervical dislocation (CD)). 2. Large, sharp scissors (for decapitation/exsanguination after CD). 3. Two pairs of No. 5 forceps. 4. Small dissection scissors (for cutting open skull). 5. Small stainless steel spatula (100 mm). 6. 50 ml falcon tubes for storage of brains (1 per pup). 7. 70% ethanol (for sterilization of tools and area).

Preparation of OHSCs

135

8. Dissection medium: Earle’s Balanced Salt Solution (EBSS) (1.8 mM CaCl2, 0.81 mM MgSO4·7H2O, 5.3 mM KCl, 26.2 mM NaHCO3, 117.2 mM NaCl, 1.01 mM NaH2PO4·H2O, 5.5 mM D-glucose, 0.025 mM phenol Red) + 25 mM HEPES + 100 units/ml penicillin/streptomycin. Store at 4  C (for a maximum of 2 weeks), use when it is ice cold. 2.2 Dissection of OHSCs

1. 70% ethanol (for sterilization of tools and area). 2. Dissection medium (as in Subheading 2.1, item 8). 3. Leica VT1000 S Vibratome (or similar). 4. Stainless steel vibratome blade (Campden model 752-1-SS or similar). 5. 2 dissection needles: 1 ml syringe mounted with needle (25G  0.5 mm). 6. Sterile disposable scalpel (No. 10 blade). 7. Superglue (“Loctite Instant Power Universal Superglue” is strongly recommended). 8. Large (145/20 mm) petri dish. 9. Sterile filter paper (autoclaved prior to use). 10. Modified sterile 3 ml plastic Pasteur pipette (end 10–20 mm cut off with a sterile scalpel to widen mouth). 11. 50 ml falcon tubes containing 5 ml of dissection medium for the collection of slices (kept on ice).

2.3 Plating and Maintenance of OHSCs

1. Incubator (37  C, 5% CO2, 100% humidity). 2. 145/20 mm petri dishes (with vents) to house up to eight individual culture dishes (see Note 1). 3. Sterile 35 mm tissue culture dishes. 4. Sterile, blunt forceps (stored in 70% ethanol, air dried before use). 5. Cell culture membrane inserts: 30 mm, hydrophilic, PTFE, 0.4 μm pore size. 6. Modified sterile 3 ml plastic Pasteur pipette (as in Subheading 2.2, item 10). 7. Maintenance medium: 50% Minimum Essential Medium (MEM) (with 1 GlutaMAX™-1, 25 mM HEPES, 5.5 mM D-glucose, and 0.027 mM phenol Red), 23% EBSS (same composition as used in Subheading 2.1, item 8), 0.65% Dglucose, 100 units/ml penicillin–streptomycin, 6 units/ml nystatin, and 25% horse serum, heat inactivated (see Note 2). Filter sterilize through a 0.22-μm filter, and store at 4  C for a maximum of 2 weeks.

136

3

Claire S. Durrant

Methods

3.1 Dissection and Culture of OHSCs

1. Preparation of culture dishes. This stage must be performed under sterile conditions and should be completed at least 2 h prior to plating slices to allow for equilibration of maintenance medium in the incubator (see Note 3). For each pup to be dissected, prepare 2–4 culture dishes (see Note 4). For each culture, add 1 ml of maintenance medium to a 35-mm culture dish. Remove the membrane from its sterile packaging, and transfer it to the dish ensuring there are no trapped bubbles and no liquid flows over the edges of the membrane (see Note 5). The membrane will turn transparent upon contact with the liquid (see Note 6). Up to eight individual culture dishes can be stored in a single large petri dish (see Note 7). Place the dishes in the incubator until use. 2. Preparation for dissection. For each pup to be collected, place a 50-ml falcon containing 10 ml of dissection medium on ice for at least 1 h before dissection. Pack ice around the vibratome buffer tray, and fill with ice cold dissection medium. Attach the blade to the blade holder, ensuring the base of the blade holder is at a 90 angle with the back of the holder. 3. Collection of brains. Mouse pups (of your preferred strain) between P6 and P9 days old (see Note 8) should be culled by CD (using the blunt end of thick forceps), then immediately decapitated. Alternatively, local guidelines of the Institutional Animal Care Committee for proper euthanasia should be followed. The tail tip can be taken (after death) to genotype the pup if required. Drench the head in 70% ethanol to prevent contamination from the animal during the brain removal. Insert one pair of forceps into the eye sockets to hold the head in place. Using the small scissors, cut away the skin around the skull. Make a lateral cut on each side of the skull under the ears, then cut the skull open at the midline (see Note 9). Using another pair of forceps, peel open the skull from the midline (this should come away easily like an eggshell at this age). Use the forceps to push the brain onto the spatula, severing the optic nerves at the front of the brain that will be holding it in place. Immediately transfer the brain into a falcon tube containing ice cold dissection buffer. Each pup used should be culled, and the brain immediately dissected out before moving to the next animal (see Note 10). 4. Preparing the brain for vibratome sectioning. Gently tip the brain and dissection medium out from the collection falcon into a large petri dish and flip the brain such that the top of the brain is facing upward (Fig. 1a). Cut the brain down the midline and gently separate the two hemispheres by pushing

Preparation of OHSCs

137

them apart with a piece of sterile filter paper (Fig. 1b). Slide a small spatula under one hemisphere such that the cut side faces downward and the cerebellum is closest to you (Fig. 1c). Dab the base of the spatula on tissue paper to dry and remove excess liquid from around the brain by gently dabbing with filter paper (see Note 11). On the vibratome specimen holder, make two oval shaped deposits of thinly spread glue slightly larger than the size of the brain hemisphere (see Note 12). Transfer the brain hemisphere from the spatula to the glue by gently pushing it off using sterile filter paper, ensuring that the cut side lands onto the glue (see Note 13). Keep the orientation such that the cerebellum is nearest you, at the bottom of the specimen holder (Fig. 1d). Repeat for the second brain hemisphere. 5. Vibratome sectioning of the brain. Wait for 15–30 s for the glue to bond to the brain before gently transferring the specimen holder into the buffer tray. Orientate the brains such that the olfactory bulb is closest to the blade, and the cerebellum is facing you (Fig. 1e). Set the vibratome to take 350 μm sections and move the blade so it is positioned just before the olfactory bulb and just above the highest point of the brain. Set the start and end position of the cuts, then set the continuous cutting program. Set vibration to the maximum frequency and keep speed at around the setting: 6–8 (out of 10) (see Note 14). As the vibratome cuts sections of brain, gently flick away excess tissue with the dissection needles. After several cuts, the hippocampus will become visible. When a section with the hippocampus becomes visible, pause the cutting program with the brain slice resting on the blade (Fig. 1f). Gently dissect out the hippocampus and adjacent entorhinal cortex using the dissection needle (Fig. 1g). Usually, only two cuts are required, one upward from the hippocampus through the cortex and the other severing the fimbria. The section of the hippocampus can then be moved away from the rest of the brain, collected using with the modified pipette, and transferred to the collection tube on ice (see Note 15). Repeat until all useable sections have been made. Usually 3–4 per brain hemisphere is realistic, depending on the age of the pup. It is important to work quickly, but gently, as time post dissection will affect slice viability. Ice surrounding the buffer tray should be regularly topped up to ensure the tissue remains ice cold. 6. Assessing slice quality. Examine the slices under a light microscope and discard any that are damaged. The dentate gyrus and CA1 should be clearly visible and undisturbed (Fig. 1h). 7. Plating OHSCs. Moving to a sterile, laminar flow or class II tissue culture hood, gently tip out slices from a single pup into a sterile petri dish. Transfer the collected slices onto the

138

Claire S. Durrant

Fig. 1 Preparation of organotypic hippocampal slice cultures from neonatal mouse brain. (a) Neonatal mouse brain orientated with top of the brain facing upward. (b) Brain cut at the midline. (c) Brain hemisphere transferred to spatula. (d) Brain hemispheres glued onto specimen holder. (e) Specimen holder with brains

Preparation of OHSCs

139

membrane inserts, using a fresh, sterile modified Pasteur pipette, plating up to three slices per membrane. Add each slice to the membrane in a drop of dissection medium, and gently position (see Note 16) (Fig. 1j). Remove the dissection medium entirely using a 200 μl pipette, being careful not to touch the slices (Fig. 1k). Place the completed dishes back in the incubator. 8. Feeding and culture maintenance. Culture medium should be changed completely between 4 and 24 h after plating to ensure removal of any inflammatory cytokines released after the cutting procedure. A further 100% medium change should be carried out on day 4 after plating. From day 7 onward, a weekly 50% medium exchange is sufficient (see Note 17). The incubator should be opened as little as possible to retain a high humidity environment throughout the culture period. 9. Assessing slice health. Healthy slice cultures will start off opaque white and will gradually flatten out and become translucent during the first 2 weeks in vitro (see Note 18). Slices should have a plump, moist appearance. Any cultures that remain opaque, become chalky, develop holes, show intense yellowing of the culture medium, or show signs of contamination should be immediately discarded. Slices that initially appear healthy but then turn white again should also be discarded (see Note 19). 10. Experimental treatments. Cultures should be left for 2 weeks in vitro before beginning experimental treatments to ensure complete recovery from the slicing procedure (see Note 20). Treated cultures can be compared to untreated controls from the same animal, reducing potential confounds of interanimal variability. When adding drug treatments, apply a small drop of treated medium onto the surface of each slice to ensure complete penetration of the slice tissue (see Note 21). For comparing the effects of two different treatments either alone or in combination, a “2:2:2:2” experiment can be conducted, where eight slices from the same animal are plated on 4 dishes (each containing 2 slices): no treatment, treatment A, treatment B, and treatment A + B. All treatment effects can then be normalized to the “no treatment” control for each animal.

ä Fig. 1 (continued) positioned in vibratome buffer tray. (f) Section of brain resting on the vibratome blade with hippocampus visible (blue dashed outline on left slice). (g) Hippocampus and surrounding cortex dissected out. (h) Healthy slice under a light microscope, dentate gyrus (blue arrow) and CA1 (green arrow) indicated. (i) Slices positioned on membrane in dissection medium droplets. (j) Completed slice culture with medium droplets removed

140

Claire S. Durrant

3.2 Experimental Outcomes

OHSCs are an incredibly versatile experimental system allowing multiple different readouts, depending on the hypothesis being tested. Some examples of the assessment of effects of Alzheimer’s disease-related mutations on neuronal pathology are outlined below. Such protocols can be adapted to suit pharmacological, injury, or genetic models as desired.

3.2.1 Medium Assays for Protein Production/Drug Screening

Drug treatments, such as those affecting the production of Aβ [7, 12] or targeting specific cell types [10], can be assessed for the impact on specific proteins detected in the culture medium. Collect 50–100 μl of culture medium at desired intervals and flash freeze on dry ice. When all samples have been collected, thaw and analyze by ELISA (according to kit instructions) for your chosen protein.

3.2.2 Western Blot for Synaptic Proteins

The loss of synaptic proteins is a key hall mark of many neurodegenerative and neuropathological processes [2]. The effect of experimental treatments (such as induction of neuroinflammation via the addition of lipopolysaccharide) [10] or disease-related mutations (in slices from transgenic animals) [7] on synaptic proteins can be assessed by western blot. At a desired timepoint, scrape slices off the membrane using a scalpel, then lyse in 33 μl ice cold RIPA buffer per slice and sonicate using a probe sonicator (for 2  5 s, avoiding bubbles). 8–15 μl of boiled lysate per lane (consisting of 50% RIPA protein lysate + 50% laemmli buffer with 10% 2-mercaptethanol) is usually sufficient for the detection of synaptic proteins (see Note 22). Western blots can then be run as per standard procedures, the details of which are beyond to scope of this chapter. Synaptic proteins of interest should be normalized to a neuronal protein, such as β-iii tubulin (Tuj1), to control for any loss of neurons. Axonal markers such as RT97 can be normalized to total neuronal markers to assess axonal-specific changes [7].

3.2.3 Immunofluorescence of Synapses and Axonal Swellings

End-stage immunostaining of OHSCs can provide an excellent readout for a number of cellular changes. Transfer the membrane to a 6-well plate containing 1 ml of 4% paraformaldehyde (in PBS), then add a further 1 ml gently on top of the membrane and fix for 20 min. Cut the membrane with a scalpel blade to permit transfer of fixed slices to a 24-well plate to minimize the use of antibodies (see Note 23). After block (1 h), primary (overnight), and secondary (2 h in the dark) antibody stages (where antibodies are diluted in block solution), transfer the stained slices to a microscope slide, with the membrane touching the glass slide, and the slices facing upward (see Note 24). Add a few drops of mounting medium (see Note 25) and place a cover slip over the slices, sealing with nail varnish [7]. Slices can then be imaged using a confocal microscope. For synapse counts, z-stacks are created from 63 images and colocalization of pre- and postsynaptic markers quantified

Preparation of OHSCs

141

using ImageJ plugins [7, 13]. Axonal swellings often form in the alveus above CA1 and, in the case of OHSCs taken from huAPP mutant mice, can be quantified by the presence of Aβ accumulation using ImageJ plugins [7]. 3.2.4 Live imaging

4

Slice cultures can take advantage of a number of transgenic mouse models, such as mito-CFP, which have labeled mitochondria that permit live imaging of axonal transport [7] (see Note 26). Live imaging is preferably performed using a water-immersion upright microscope objective to avoid problems imaging through the culture membrane; however low-resolution imaging using a low-magnification (5) inverted objective can be sufficient for some readouts, such as fluorescence intensity of large regions. Viral transfection of fluorescent proteins, such as GCaMP6, permits imaging of functional calcium activity [14] and application of live stains, such as isolectin-B4 conjugated to Alexa fluorophores allows live observation of microglia [7].

Notes 1. Cultures can be stored in 6-well plates, if desired; however, OHSCs can be more prone to contamination in this format as individual cultures cannot be easily isolated from others. 2. Different lot numbers of horse serum have been found to have different effects on OHSC survival. Before beginning experiments, it is strongly recommended to lot-test horse serum batches, selecting (and bulk purchasing) batches that provide optimal slice survival. 3. This can be performed up to 24 h before plating if convenient. 4. Up to eight OHSCs can per prepared per pup, with more slices being obtainable from a P9 pup compared to a P6. Between 2 and 3 slices should be plated per dish. 5. When placing membranes in the dish, grip the membrane on the plastic rim using sterile forceps, do not touch the membrane itself as this could puncture the membrane. Place the membrane in the dish such that one edge touches the dish first, slowly lowering the raised side so that bubbles are avoided. If bubbles remain, gently repeat this motion, lifting one side of the membrane up and lowering slowly until the bubbles have been forced to the edge. 6. Any membranes that fail to turn completely transparent when placed in culture medium are defective and should be discarded. 7. Slice cultures survive best under conditions of high humidity. As such, keeping more dishes within a large holding dish is

142

Claire S. Durrant

preferable. If less than 6 dishes per large dish are required, placement of a 35-mm dish containing sterile water in the center of the holding dish will help ensure the cultures do not dry out. 8. OHSCs from P6 to P9 mice show the best survival in culture [11]. Cultures from older mice tend to fail in this protocol (although protocols have been described for the culture of OHSCs from adult mice [15]). OHSCs can be made from younger mice (P0–P6), although the brain is much smaller at this age which complicates dissection of the hippocampus. 9. Be careful to keep the scissors close to the surface of the skull to prevent damage to the underlying brain tissue. 10. The brain collection process should be done as quickly as possible to minimize time from death to culture. Dissection speed should be assessed and the numbers of pups used in one dissection session adjusted, so that time from cull to final placement of finished OHSCs in the incubator never exceeds 2 h. More pups can be added with practice and increased speed. For beginners, it is advised to keep batches below six pups at a time. 11. Removal of liquid from around the brain and spatula is crucial to ensure effective bonding to the glue and vibratome specimen holder. Excess liquid can cause the glue to flow over the entire surface of the brain or cause the brain to detach during sectioning. 12. Getting the right amount of glue in the right place is crucial for the success of the slicing procedure. Too little and the brain hemisphere will detach, too much and the glue may flow over the brain or cut into the brain during sectioning (the glue dries very hard and can be forced through the brain tissue if not controlled). During slicing, if sections of hard glue can be seen, push these down with your dissection needle to ensure they do not damage the brain tissue. 13. Transferring the brain hemisphere from the spatula to the glue is often the trickiest part of the whole procedure and may take practice. Move slowly and carefully to avoid damaging the tissue. If the brain falls in the wrong orientation, you can gently try to reposition using filter paper, but great care should be taken not to stretch the brain tissue. If your filter paper sticks to the brain/glue, do not pull it away, as this will stretch or tear the brain. If possible, gently cut the piece away with sprung scissors to avoid further damage. Any damaged brains in this process must be discarded as stretching will cause axon damage that may confound your results.

Preparation of OHSCs

143

14. Speed can be adjusted during the procedure to suit requirements. Start slower and build up speed as you become proficient. 15. When moving the brain slices with the modified pipette, great care must be taken. Do not allow bubbles to flow over the slice or the slice to become stuck to the side. Slow, gentle movements must be made to prevent damage. 16. Slices should be positioned toward the center of the membrane, but kept at least 5 mm apart from one another. This is to prevent slices merging into one another as they spread and thin out over time in culture. 17. When performing a 50% medium change, to account for any evaporation that has taken place, around 400 μl is removed for every 500 μl of fresh medium added (see Note 7 for comment on minimizing evaporation). 100% medium changes can be performed if desired; however, for studying neurodegenerative disease processes where extracellular protein accumulation may be important (e.g., in Alzheimer’s disease models), the 50% exchange permits accumulation of relevant pathological proteins over time. 18. This process happens gradually, and often the dentate gyrus and CA1 are the last to turn transparent, giving a “swirled” appearance by eye in early stages of successful cultures. When placing a gloved hand under the culture dish, you should be able to see your glove through the slice in a healthy culture. 19. Many seemingly trivial factors can affect slice culture health, so caution should be taken when starting OHSCs for the first time or moving the technique to a new lab/room or altering protocols. For example, incubators should not be placed on surfaces with excessive vibration (such as next to a high speed centrifuge), as we have found this damaged cultures. 20. Microglia have been found to return to their resting state after this time in vitro [10]. 21. A small droplet should be rapidly absorbed into the slice, but check this has occurred prior to replacing the dish in the incubator. Never allow the surface of the membrane to become flooded as this can kill the slices by hampering oxygen exchange. 22. Some synaptic proteins aggregate when the protein lysate is boiled, so check the specific requirements of your protein of interest when running western blots. 23. To easily cut the membrane without damaging the slices, move the membrane insert onto a glass microscope slide, moistened by PBS. The membrane will stick to the slide and is easily cut using a sharp scalpel. Cut a safe distance from the slices to

144

Claire S. Durrant

prevent damage, but close enough to minimize excess membrane. The plastic insert can then be thrown away, and the remaining loose membrane, with slices adhered, easily moved by gripping the edge with a pair of forceps. 24. Keep the membrane wet at all points during the staining procedure to prevent it becoming opaque. 25. Do not use hard-set mounting medium as this will ultimately dry out the membrane, causing it to become opaque and making repeated imaging difficult. 26. Ensure when selecting a mouse line that expresses fluorescent proteins that the transgene is expressed early in development. As OHSCs are made from neonatal mice, fluorescence may not be detectable in lines where expression is not complete until several weeks in vivo. References 1. Terry RD, Masliah E, Salmon DP et al (1991) Physical basis of cognitive alterations in Alzheimer’s disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol 30:572–580 2. Wishart TM, Parson SH, Gillingwater TH (2006) Synaptic vulnerability in neurodegenerative disease. J Neuropathol Exp Neurol 65:733–739 3. Ransohoff RM (2016) How neuroinflammation contributes to neurodegeneration. Science 353:777–783 4. Hill CS, Coleman MP, Menon DK (2016) Traumatic Axonal Injury: Mechanisms and Translational Opportunities. Trends Neurosci 39:311–324 5. Gao X, Deng P, Xu ZC et al (2011) Moderate traumatic brain injury causes acute dendritic and synaptic degeneration in the hippocampal dentate gyrus. PLoS One 6:e24566 6. De Simoni A, Yu LMY (2006) Preparation of organotypic hippocampal slice cultures: interface method. Nat Protoc 1:1439–1445 7. Harwell CS, Coleman MP (2016) Synaptophysin depletion and intraneuronal Aβ in organotypic hippocampal slice cultures from huAPP transgenic mice. Mol Neurodegener 11:44 8. Humpel C (2015) Organotypic brain slice cultures: a review. Neuroscience 305:86–98 9. Croft CL, Noble W (2018) Preparation of organotypic brain slice cultures for the study of Alzheimer’s disease. F1000Res 7:592

10. Sheppard O, Coleman MP, Durrant CS (2019) Lipopolysaccharide-induced neuroinflammation induces presynaptic disruption through a direct action on brain tissue involving microglia-derived interleukin 1 beta. J Neuroinflammation 16:106 11. Croft CL, Wade MA, Kurbatskaya K et al (2017) Membrane association and release of wild-type and pathological tau from organotypic brain slice cultures. Cell Death Dis 8:e2671 12. Durrant CS, Ruscher K, Sheppard O, et al (2020) Beta secretase 1-dependent amyloid precursor protein processing promotes excessive vascular sprouting through NOTCH3 signalling. Cell Death Dis 11:1–15 13. Ippolito DM, Eroglu C (2010) Quantifying synapses: an immunocytochemistry-based assay to quantify synapse number. J Vis Exp 45. https://doi.org/10.3791/2270 14. Jacob T, Lillis K, Wang Z et al (2018) A proposed mechanism for spontaneous transitions between interictal and ictal activity. J Neurosci 39(3):557–575 15. Humpel C (2015) Organotypic vibrosections from whole brain adult Alzheimer mice (overexpressing amyloid-precursor-protein with the Swedish-Dutch-Iowa mutations) as a model to study clearance of beta-amyloid plaques. Front Aging Neurosci 7:47

Chapter 11 Organotypic Culture Assay for Neuromuscular Synaptic Degeneration and Function Kosala N. Dissanayake, Robert Chang-Chih Chou, Rosalind Brown, and Richard R. Ribchester Abstract We describe here an organotypic culture system we have used to investigate mechanisms that maintain structure and function of axon terminals at the neuromuscular junction (NMJ). We developed this by taking advantage of the slow Wallerian degeneration phenotype in mutant Wlds mice, using these to compare preservation of NMJs with degeneration in nerve-muscle preparations from wild-type mice. We take hind limb tibial nerve/flexor digitorum brevis and lumbrical muscles and incubate them in mammalian physiological saline at 32  C for 24–48 h. Integrity of NMJs can then be compared using a combination of electrophysiological and morphological techniques. We illustrate our method with data showing synaptic preservation ex vivo in nerve-muscle explants from Sarm-1 null-mutant mice. The ex vivo assays of NMJ integrity we describe here may therefore be useful for detailed investigation of synaptic maintenance and degeneration. Key words Wallerian degeneration, Neuromuscular junction, Electrophysiology, Immunostaining, Ex vivo assay

1

Introduction Culture systems have been utilized for decades as models for advancing fundamental biology and for understanding mechanisms of disease. Dissociated primary cells, differentiated stem cells, or immortalized cell lines are commonly used for these purposes in contemporary neurobiology [1]. Organotypic culture of brain slices are also used to study neuronal and synaptic degeneration and function [2], but, by comparison, there are few studies utilizing the organotypic culture approach that have taken advantage of the accessibility of neuromuscular junctions (NMJs) to morphological and physiological investigation. This is principally because the process of isolation of intact nerve-muscle preparations almost inevitably requires section of peripheral nerve axons, which normally triggers anterograde Wallerian-like degeneration of motor axon

Elisabetta Babetto (ed.), Axon Degeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2143, https://doi.org/10.1007/978-1-0716-0585-1_11, © Springer Science+Business Media, LLC, part of Springer Nature 2020

145

146

Kosala N. Dissanayake et al.

terminals, within 15–24 h [3–5]. To mitigate this, we designed a method to exploit the slow-synaptic degeneration phenotype of the WldS strain of mice [6–8]. We describe here the protocol we developed for short-term (1–2 days) culture of whole nerve-muscle preparations utilizing this mouse line. Our approach has several advantages over coculture of dissociated neurones with muscle cells and, by contrast, in vivo investigations. First, “ex vivo” cultures permit study of the maintenance of mature NMJs rather than generation or regeneration of new synapses. Second, they enable investigators to screen and study in a controlled fashion, the effects of exogenous compounds or other conditions (e.g., neuromuscular activity) that may influence the preservation or degeneration of synapses. These procedures bypass pharmacokinetic and pharmacodynamic issues that may arise following drug administration in vivo, as well as avoiding potentially toxic or other systemic side effects. Third, our ex vivo approach also addresses some of the ethical issues associated with nerve lesions or administration of substances in living animals, thereby embracing goals of reduction, refinement, and replacement of laboratory animals in testing for toxicity, as well as avoiding the potentially painful and overtly paralyzing effects associated with major peripheral (e.g., sciatic) nerve lesions. Crucially, our ex vivo assay takes advantage of a slow Wallerian degeneration phenotype, discovered originally in the WldS mouse strain and which has been replicated in genetically modified Sarm-1 null-mutant mice [6, 8–11]. In WldS mice, preservation of neuromuscular synapses after axotomy in vivo can be as long as 10 days [7, 12, 13]. Axon protection in WldS mice is now understood to be due to a spontaneous mutation that causes overexpression of a stable protein, comprising a short N-terminal sequence of the ubiquitination cofactor Ube4b fused to the complete amino acid sequence for nicotinamide mononucleotide adenylyl transferase-1 (Nmnat-1) [9, 14, 15]. This chimeric protein substitutes for a more labile axoplasmic Nmnat isoform, Nmnat-2. This isoform has a short half-life in cut distal axons, and the reduction in its levels after a few hours is sufficient to trigger synaptic and axonal degeneration [8, 16]. Subsequently, it was shown that the WldS phenotype is emulated both in transgenic mice or rats that overexpress the WldS chimeric protein [12, 15, 17] and in Sarm-1 null-mutant flies and mice [11], suggesting that pathways linking expression of the two proteins to axonal and synaptic protection are mechanistically linked [8, 18].  We found that when cultured ex vivo at 32 C, Wlds synaptic terminals are almost completely preserved 24 h later. For reasons we still do not understand this preservation is not as prolonged as it is in vivo [7, 13, 19]. By contrast, synaptic degeneration in preparations from wild-type mice is virtually complete within 15–24 h [7]. Nevertheless, a window of opportunity opens up 24–48 h after

Ex-vivo Assay of NMJ Degeneration

147

isolation and preparation of the explants, which allows the extent of synaptic protection or degeneration in wild-type versus WldS mice to be compared. We use a combination of electrophysiological and morphological assays, enabling quantitative evaluation of a range of treatments that would otherwise be more difficult to control or evaluate in vivo. Our method and assays therefore provide opportunities to understand better the pathways that are involved in the synaptic maintenance or degeneration. We illustrate our protocol here with previously unpublished data obtained after 24–48 h culture of nerve-muscle explants from Sarm-1 null mutant mice (Figs. 1 and 2).

Fig. 1 Illustration of the method. (a) Schematic diagram of the ex vivo nerve-muscle explant preparation. Combined flexor digitorum brevis/deep lumbrical (FDB/DL) muscle preparatiions (“fans”), with the medial plantar nerve (MPN) and lateral plantar nerve (LPN) attached, are secured with fine minutien pins to a strip of dental wax immersed in a screw-top universal tube containing bicarbonate-buffered mammalian physiological saline (MPS). The lid of the tube is modified to accommodate two hollow hypodermic needles, one connected to a gas supply and the other left open for venting gas outflow. After loosely affixing the lid, the tube is immersed in a 32  C water bath (or a temperature-controlled heating block) to the same level of fluid contained in the immersed tube(s) and incubated for 24–48 h. The gas flow should be adjusted to give a gentle, slow stream of bubbles without causing frothing of the medium. (b) Example of a universal tube with hypodermic cannulae fitted. (c) Setup with water bath containing two incubation tubes connected to a gas supply. (d) Freshly dissected FDB/DL fan nerve-muscle preparation, ready for transfer to dental wax and the incubation tube. (e) Separated FDB (left) and DL fan comprising the four deep lumbrical muscles (right, arrows), ready for electrophysiological and imaging assays, respectively

148

Kosala N. Dissanayake et al.

Fig. 2 Degeneration ex vivo requires expression of Sarm-1 protein. (a–c) Neurofilament/SV2 immunostained preparations from Sarm-1 (a and b) and wild-type explants (c) of DL muscles, counterstained with TRITC-α-BTX and imaged in a confocal microscope. Note good preservation of motor nerve terminals (green/yellow overlay of ACh receptors) in the Sarm-1 preparations and complete degeneration in the wildtype preparation. (d) Intracellular recording of nerve-evoked EPP and spontaneous MEPP (arrows) recorded from a Sarm-1 FDB muscle fiber 24 h after incubation. (e) Electrophysiological assay data (mean  SEM) from Sarm-1 and wild-type preparations showing the percentage of responsive fibers with evidence of innervation, based on presence of spontaneous MEPPs and/or nerve-evoked EPPs. Samples of 30 fibers in 1–3 muscles in each group. (f) Box plots (median, interquartile range, 5–95% range) of morphological assay data showing percentage innervation of motor endplates based on sampling of 30–85 NMJs in 10–20 images from 1 to 3 muscles in each group. Differences between groups are all significant p < 0.001, ANOVA with Tukey’s post hoc test, suggesting Sarm-1 dependent and independent neurodegenerative mechanisms could be explored using this methodology

2

Materials 1. Conduct all the experimental preparation at room temperature (18–25  C) unless otherwise stated. 2. Use distilled water for preparing all the solutions.

2.1

Animals

Homozygous thy1.2-YFP16:Wlds mice (see Note 1); Sarm-1 nullmutant (knockout) mice; or wild-type C57Bl6. Mice should be age matched (ideally ~5–10 weeks old). Male or female mice may be used.

Ex-vivo Assay of NMJ Degeneration

149

2.2 Mammalian Physiological Saline (MPS)

120 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 0.4 mM NaH2PO4, 24 mM NaHCO3, and 5 mM D-glucose (see Note 2) MPS should then be equilibrated by bubbling through a sintered glass tube with 95% oxygen and 5% carbon dioxide for ~15 min in order to bring the pH to 7.2–7.4, prior to use.

2.3 Dissection Materials

Bone scissors, iris scissors, petri dish lined with Sylgard, minutien pins, and dissecting microscope.

2.4 Culturing Materials

Dental wax, water bath, 30 ml universal disposal plastic tubes, hypodermic needles (25G and 21G), epoxy resin, thermal resistant clear medical tubing, antibiotics, medical gas (95% oxygen and 5% carbon dioxide), and heat resistant test tube rack.

2.5 Electrophysiological Analysis Materials

Sylgard-lined recording chamber, microelectrode puller, 1 mm glass capillary with in-built filament, fine cannula, 3 M KCl or 4 M potassium acetate, high impedance head stage of suitable amplifier, micromanipulator, silver–silver chloride wire or pellet, Faraday cage, dissecting stereomicroscope, oscilloscope, constant voltage stimulator, personal computer, digital computer interface, and recording software.

2.6 Morphological Analysis Materials

Tetramethylrhodamine-isothiocyanate (TRITC) conjugated with α-bungarotoxin, rocking platform, 4% paraformaldehyde (PFA), antibodies against neurofilament and synaptic vesicle proteins, mounting medium (Vectashield or Mowiol), conventional fluorescence microscope, or a laser-scanning confocal microscope.

3

Methods All the animals used in our studies are housed and maintained in accordance with requirements of the United Kingdom Animals (Scientific Procedures) Act 1986. Animals are normally sacrificed by cervical dislocation and exsanguination in accordance with UK Home Office regulations, Schedule 1.

3.1

Dissections

1. Use a large scissors/bone scissors to remove the both hind legs near hip joint. 2. Using iris scissors, make an incision through the anterior surface of the skin (dorsal midline) and continue to along the dorsal surface of the foot. 3. Carefully peel the skin from the leg (with practice this can be done in one continuous stripping action) until most of the skin has been removed from the leg and the dorsal and plantar surfaces of the foot, including the glabrous and hairy skin around the toes.

150

Kosala N. Dissanayake et al.

4. Secure the leg by minutien pins in a petri dish lined with Sylgard. The plantar surface of the foot and the medial side of the leg should be facing up (see Note 3). 5. Carefully expose the tibial nerve by dissecting away superficial connective tissue and muscle on the medial side of the tibia. Extend the dissection of the tibial nerve distally to the metacarpal junction (ankle) and proximally to its branch point from the sciatic nerve. The tibial nerve branches at the heel into the medial plantar nerve and the lateral plantar nerve, which supply nerve branches respectively to the flexor digitorum brevis (FDB) muscle and the four deep lumbrical (DL) muscles. Dissect both branches of the nerve along with the attached FDB and lumbrical muscles to their distal insertions on the toes. 6. Remove the nerve-muscle preparations and dispose of the remainder of the leg. 7. Pin the dissected muscles on the Sylgard-lined petri dish using minutien pins (see Note 4). 8. Pin the proximal end of the tibial nerve to the dish, but secure it with minimal tension to avoid stretching and damaging the axons (see Note 5). 9. Carefully trim off the excess connective tissue and remnants of blood vessels from the nerve (up to the nerve entry point into the muscle). Remove excess connective tissue from the muscle as well. 10. Pin the nerve-muscle preparation onto a strip of Sylgard or dental wax (1 cm  5 cm) (see Note 6). 3.2 Setting up the Water Bath

1. Set the water bath to the required temperature (normally 32  C) prior to start of the experiment (see Note 7).

3.3

1. Culture tubes are made from 30 ml universal disposal plastic tubes.

Culture Tubes

2. Two hollow hypodermic needles (25G and 21G) are fitted through the lid of the tube (Fig. 1), the larger gauge needle in the middle and the other toward the edge. After adjusting for height (see below), the needles should be sealed in place using epoxy resin. Assembled tubes can be reused after sterilization. 3. Thermal resistant clear medical tubing (about 5 cm) is fitted to the middle hypodermic needle (25G) to facilitate delivery of gases to the bottom of the tube. 4. The hypodermic needle fitted to the periphery (21G) is trimmed half way through its length to avoid contact of the saline inside the tube. This needle serves as a vent for gases delivered by the first hypodermic needle.

Ex-vivo Assay of NMJ Degeneration

3.4 Culturing the Samples for 24 h

151

1. Two-thirds fill the tubes with MPS, adding any test compound (see Note 8). Antibiotics (e.g., 50 μg ml 1 gentamycin and 100 μg ml 1 kanamycin) may be added but have not normally been required in our hands. 2. Insert the nerve-muscle explant mounted on the dental wax, into the tube (one nerve-muscle explant per tube). 3. Close the tube with the lid (see Note 9). 4. Connect the central hypodermic to the continuous gas supply (normally 95% oxygen and 5% carbon dioxide). Adjust the gas flow to produce gentle bubbling of the bathing solution (see Note 10). 5. Place the tube in the water bath and incubate for 24–48 h.

3.5 Harvesting the Samples

1. After 24/48 h of incubation with continuous gas flow, remove the nerve-muscle explants from the tube and transfer them to a Sylgard-lined petri dish containing fresh MPS (see Note 11). 2. Carefully dissect away the FDB muscle along with its tibial nerve supply. 3. Transfer the tibial nerve-FDB muscle preparation to a Sylgardlined recording chamber containing MPS in order to perform electrophysiological analysis. Retain the lumbrical muscles for immunostaining of axons and nerve terminals and labeling of postsynaptic acetylcholine receptors, for subsequent morphological analysis.

3.6 Electrophysiological Analysis

1. Pin the FDB muscle in the recording chamber through the proximal tendon and three distal tendons, stretching them apart to optimize visualization of the pennate muscle fibers (see Note 12). The muscle should be submerged in the MPS and continuously perfused at a rate of 1 ml/min with solution bubbled with the 95% oxygen and 5% carbon dioxide, to maintain pH. Alternatively, the preparation can be bathed in a HEPES-buffered physiological solution (see, for example, [7]), previously equilibrated by bubbling with oxygen or air, in which case bath perfusion is not normally required for recordings made at room temperature (18–25  C). 2. Connect a suction electrode to the nerve (see Note 13). 3. Pull recording electrodes from 1 mm glass capillary with in-built filament using appropriate microelectrode puller (see Note 14). 4. Back fill the microelectrode from a syringe fitted with a fine cannula (