Elucidating Schwann Cell Reprogramming

ELUCIDATING SCHWANN CELL REPROGRAMMING

Dr Elizabeth Byrne Cell Interactions and Cancer Group Institute of Clinical Science MRC Clinical Sciences Centre Imperial College London

Thesis submitted for the degree of MPhil

Abstract

The peripheral nervous system, unlike the central nervous system, has an exceptional capacity for regeneration following injury. This is due to the remarkable plasticity of the Schwann cells (SC), which are able to reprogramme, following injury, to a progenitor like cell which facilitates peripheral nerve repair. Current knowledge on the molecular basis of this reprogramming is incomplete and we are lacking a global overview of the transcriptional events that occur in SC following nerve injury and how these change over time.

We aimed to characterise transcriptional changes in the SC, over time, following nerve injury using RNAseq. We also aimed to develop an in vitro dedifferentiation assay to use as a screening tool to asses potential key genes found using RNAseq.

We developed a method of reliably extracting good quality, SC specific, RNA from the sciatic nerve of mice using fluorescence activated cell sorting. We performed RNAseq on SC from intact nerves and from the distal stump of nerves 6 days post transection. We validated this method by confirming differential expression of genes known to be up and downregulated following nerve injury, using RNAseq data. In analysing the RNAseq data we identified several potentially exciting, novel key molecular players in SC reprogramming, namely Myc and Runt-related transcription factor 1.

We also developed an in vitro dedifferentiation assay to use as an initial screen for the genes identified using RNAseq. This involved the addition of neuregulin and serum to previously cyclic adenosine monophosphate differentiated SC. This assay was shown to recapitulate the changes seen in vivo, using RNAseq.

Table of contents

ABSTRACT ...... 2 COPYRIGHT DECLARATION...... 5 DECLARATION OF ORIGINALITY ...... 6 ACKNOWLEDGEMENTS ...... 7 LIST OF FIGURES ...... 8 1. INTRODUCTION ...... 11

1.1 PERIPHERAL NERVOUS SYSTEM ...... 11 1.2 SCHWANN CELL DEVELOPMENT ...... 13 1.2.1 Early SC development ...... 13 1.2.2 Schwann Cell Precursors ...... 14 1.2.3 Immature Schwann Cells ...... 15 1.2.4 Radial Sorting ...... 15 1.2.5 Schwann Cell Myelination ...... 16 1.3 MYELIN...... 17 1.4 WALLERIAN DEGENERATION ...... 18 1.5 REGENERATION ...... 19 1.5.1 Schwann Cell Dedifferentiation ...... 19 1.5.2 The Schwann Cell Repair Programme ...... 20 1.5.3 Myelin Clearance ...... 21 1.5.4 Axonal Guidance ...... 22 1.5.5 Proliferation ...... 23 1.6 REINNERVATION ...... 24 1.7 MOLECULAR BASIS OF SC REPROGRAMMING FOLLOWING INJURY ...... 25 1.7.1 MAPK cascades ...... 25 1.7.2 Transcription factors ...... 27 1.7.3 Growth Factors ...... 28 1.7.4 Receptors ...... 29 1.8 CLINICAL RELEVANCE ...... 30 1.8.1 Peripheral nerve injury ...... 30 1.8.2 Peripheral Neuropathies ...... 33 1.8.3 Peripheral Nerve Sheath Tumours ...... 35 1.9 AIMS ...... 37 2. MATERIALS AND METHODS ...... 38

2.1 IN VIVO METHODS ...... 38 2.1.1 Animal housing conditions ...... 38 2.1.2 Mouse line ...... 38 2.1.3 Genotyping ...... 39 2.1.4 Sciatic nerve transaction ...... 40 2.1.5 Harvesting the sciatic nerve ...... 40 2.1.6 Immunoflurescence ...... 41 2.1.7 In vivo ethynyl-2’-deoxyuridine (EDU) proliferation assay ...... 41 2.1.8 Digestion of the Sciatic nerve and SC extraction ...... 42 2.1.9 SC specific RNA extraction ...... 43 2.1.10 Bioanalyzer ...... 43

2.1.11 Plating and Staining SC following Sorting ...... 44 2.1.12 RNA seq library preparation ...... 44 2.2 IN VITRO METHODS ...... 46 2.2.1 Isolation and culture of rat Schwann Cells ...... 46 2.2.2 Dedifferentiation assay ...... 47 2.2.3 RNA extraction ...... 48 2.2.4 DNase Treatment ...... 49 2.2.5 Reverse Transcription ...... 49 2.2.6 qPCR ...... 50 2.2.7 Protein extraction ...... 50 2.2.8 Western Blot ...... 51 2.2.9 Immunofluorescence ...... 51 2.2.10 EDU proliferation assay ...... 52 2.1.11 RNAseq library preparation ...... 53 2.3 MEDIA AND SOLUTIONS ...... 56 2.3.1 Nerve dissociation media ...... 56 2.3.2 Cell culture media ...... 57 2.3.3 Western Blot Solutions ...... 58 2.4 PRIMERS FOR END POINT PCR (GENOTYPING) ...... 59 2.5 MOUSE PRIMERS FOR QPCR ...... 59 2.6 RAT PRIMERS FOR QPCR ...... 60 2.7 ANTIBODIES...... 61 3. RESULTS ...... 62

3.1 IN VIVO RESULTS ...... 62 3.1.1 Validation of the tdTomato mouse line ...... 62 3.1.2 Methods of SC specific RNA extraction ...... 64 3.1.3 Optimisation of FACS method of SC specific RNA isolation ...... 66 3.1.4 Validating the FACS method of extracting SC specific RNA ...... 72 3.1.5 Characterising the timing of the SC response to injury ...... 77 3.1.6 RNAseq of intact vs distal nerves ...... 82 3.2 IN VITRO RESULTS ...... 88 3.2.1 Validating the differentiation assay ...... 88 3.2.2 Optimising the dedifferentiation SC assay ...... 89 3.2.3 Characterising the dedifferentiation SC assay ...... 95 3.2.4 Interrogating the in vitro dedifferentiation response ...... 99 3.2.5 RNA seq of in vitro samples ...... 102 3.3 COMPARING IN VITRO DEDIFFERENTIATION MODEL TO IN VIVO DATA ...... 104 4. DISCUSSION ...... 109

4.1 SC REPROGRAMMING IN VIVO ...... 109 4.2 IN VITRO REPAIR ASSAY ...... 117 5. FUTURE WORK ...... 120 REFERENCES ...... 122 APPENDIX A...... 135 APPENDIX B...... 136 APPENDIX C...... 141 APPENDIX D...... 143

The copyright of this thesis rests with the author and is made available under a Creative Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work.

Declaration of originality

The work described within this thesis is my own, unless otherwise explicitly stated. No part of this thesis has been submitted for any other degree or professional qualifications.

Acknowledgements

I would like to thank my supervisor Dr Simona Parrinello for her advice, input and patience. Extra special thanks to Dr Mel Clements, senior researcher in the lab whose professional and personal support and advice have been invaluable. I would also like to thank all the members of my lab, Dr Cristina Ottone, Tim Davis, Dr Ben Krusche, Ariadne Whitby, Holly Simpson Ragdale and Leila Zakka, for their help and friendly chat. Thank you also to Katrin Goetsch for covering me so well during my maternity leave.

I would like to thank my husband Andy and my son’s Archie and Alfie who have put up with having an absent wife and mother for certain parts of this journey. I would like to thank them for sharing my happiness and excitement when things were going well and cheering me up when things weren’t going so well.

I would finally like to thank the MRC and Imperial BRC for funding my fellowship. I have learnt new skills and new ways of thinking that I would not have been exposed to otherwise and it has definitely helped me work out what I want out of my future career.

List of figures and tables Figure 1 Structure of a peripheral nerve.(Adapted from Marieb and Hoehn, 12 2005.)

Figure 2 Schwann Cell development (Mirsky et al., 2008) 13

Figure 3 Changes seen in SC following nerve injury. (Adapted from Lee et al., 19 2014.)

Figure 4 Overview of the key molecular players involved in SC 25 dedifferentiation

Figure 5 The tdTomato/P0 cre mice have a high recombination efficiency. 63

Figure 6 LCM is a potential method of extracting SC specific RNA from sciatic 64 nerves of the Tdtomato mice.

Figure 7 FACS does not reduce the quality of RNA obtained from sciatic 70 nerves.

Figure 8 A clear group of tdTomato positive cells is demonstrated by FACS in 72 sciatic nerves from tdTomato mice

Figure 9 Viable SC are seen following sciatic nerve digestion and FACS. 74

Figure 10 Mature SC markers are enriched in the FACS SC when compared to 75 the whole nerve

Figure 11 Myelin proteins are downregulated and dedifferentiation markers 76 are upregulated in FACS SC following nerve transection when compared to FACS SC from intact nerves.

Figure 12 Downregulation of myelin proteins and upregulation of 78 dedifferentiation markers at the mRNA level can be seen by 24 hours post nerve transection in FACS SC.

Figure 13 SC proliferation and dedifferentiation, and axonal degeneration are 80 seen by 24-36 hours following nerve transection. SC dedifferation and axonal degeneration is seen throughout the nerve by 72 hours.

Figure 14 Following nerve transection there is downregulation of myelin 83 proteins and, upregulation of some dedifferentiation markers and a proliferation marker demonstrated by RNAseq.

Figure 15 Down and upregulated transcription factors found by GSEA of 85 RNAseq data when comparing SC from the distal nerve to SC from the intact nerve.

Figure 16 Down regulation and upregulation in gene sets identified by GSEA, 85 grouped according to function, reflects SC biology

Figure 17 SC from the distal nerve show a similar gene expression profile to 87 that of nerves from a mouse with a hypomorphic krox 20 allele

Figure 18 cAMP treatment of SC for 3 days is sufficient to induce 88 differentiation.

Figure 19 Treatment of SC with FCS following cAMP induced differentiation 90 does not enhance the dedifferentiation response when compared to cAMP removal.

Figure 20 Treatment of SC with NRG, TGF and HGF, alone or in combination, 91 following cAMP induced differentiation does not enhance the dedifferentiation response when compared to cAMP removal

Figure 21 Treatment of cAMP differentiated SC with NRG and 10% FCS show 93 promising results which is not seen with TGF and FCS alone or augmented by the addition of TGF.

Figure 22 Treatment of cAMP differentiated SC with NRG and 10% FCS 95 enhances the dedifferentiation response at the mRNA level.

Figure 23 Treatment of cAMP differentiated SC with NRG and 10% FCS causes 96 reduced protein expression of the myelin transcription factor krox20, and the myelin protein P0, and increased expression of the dedifferentiation transcription factor Sox2

Figure 24 Treatment of cAMP differentiated SC with NRG and 10% FCS induces 97 increased protein expression of dedifferentiation markers GFAP and N-cadherin.

Figure 25 Treatment of cAMP differentiated SC with NRG and 10% FCS causes 98 the SC to regain proliferative abilities..

Figure 26 Inhibiting the MEK and JNK pathways in the NRG and FCS treated SC 100 did not completely inhibit the dedifferentiation response

Figure 27 Treatment of cAMP differentiated SC with NRG and 10% FCS causes 103 downregulation of myelin protein genes and upregulation of dedifferentiation genes as seen by RNAseq

Figure 28 The in vitro SC dedifferentiation assay shares some common features 104 with our in vivo data..

Figure 29 There is enrichment of genes downregulated in the hypomorphic 105 krox 20 nerves in cAMP treated, differentiated SC in vitro

Figure 30 Some genes up and downregulated in the hypomorphic krox20 106 nerves are also differentially expressed in SC from the distal nerve in vivo and dedifferentiated SC in vitro.

Figure 31 Some genes differentially expressed after crush injury in rat sciatic 107 nerves are also differentially expressed in SC from the distal nerve in vivo and in dedifferentiated SC in vitro.

Table 1 The Sunderland and Seddon Classifications of peripheral nerve 30 injury.

Table 2 Sequences of primers used for genotyping 59

Table 3 Sequences of mouse primers used for qPCR. 59

Table 4 Sequences of rat primers used for qPCR. 60

Table 5 Details of the antibodies used for immunofluorescence and western 61 blot.

Table 6 Digestion using trypsin, DNase, collagenase and hyaluronidase is 67 optimal.

Table 7 Digestion using 0.02% hyaluronidase, 0.3% trypsin and 0.6% 68 collagenase is optimal.

Table 8 Digestion for 20 minutes is optimal.. 68

Table 9 Tissue homogenisation by chopping the nerve and harsh trituration is 69 optimal.

Table 10 Pooling the RNA from SC from 3 sciatic nerves improves the yield of 71 RNA obtained without compromising the quality.

1. Introduction

The peripheral nervous system (PNS), unlike the central nervous system (CNS), has an exceptional capacity for regeneration following injury. This is due to the remarkable plasticity of the Schwann cells (SC), the principle glial cell of the PNS, which is able to reprogramme to a progenitor like cell following injury, and functions to facilitate nerve repair. Although our knowledge of the molecular events that underlie this process has increased greatly in recent years we are still lacking a global picture of the transcriptional changes that occur in SC following nerve injury and how these change over time.

1.1 Peripheral nervous system

The PNS comprises all of the nerves and ganglia outside of the CNS (the brain and spinal cord). It serves to connect the CNS to the limbs and organs (motor neurons) and sends signals from these back to the CNS (sensory neurons). It can be divided into the somatic nervous system, which controls voluntary movements, and the autonomic nervous system which controls unconscious functions

(Crossman and Neary, 2014).

Nerves in the PNS are either sensory; containing axons and dendrites of afferent, sensory neurons, motor; containing axons and dendrites from efferent, motor neurons or mixed; containing both sensory and motor neurons (Crossman and Neary, 2014). Each axonal segment in a nerve is ensheathed by a SC and surrounded by connective tissue known as the endoneurium, which is longitudinally orientated (Sunderland, 1990). Individual nerve fibres are bundled into fascicles by the perinuerial connective tissue layer. The epineurium is a connective tissue layer that surrounds multiple fascicles to make the nerve trunk proper (Campbell, 2008). The perineurium and epineurium are orientated circumferentially (Sunderland, 1990) (Figure 1).

Figure 1. Structure of a peripheral nerve. (Adapted from Marieb and Hoehn, 2005.) Image

reproduced with minor modifications with permission of the rights holder Pearson Education,

Inc.

In terms of blood supply, microvessels form plexuses running longitudinally in the epineurium, these send transverse processes to the perineurium and the vascular network consists mainly of capillaries at the endoneurium (Campbell, 2008). The PNS is offered some protection by the blood nerve barrier (Kanda, 2012). This is found at the microvasculature of the epineurium and at the innermost area of vessels of the perineurium (Bell & Waddell, 1984). In these areas the endothelium of the vessels are non-fenestrated and endothelial cells are connected by continuous tight junctions. This isolates the epineurium from the contents of the blood vessels (Kanda, 2012). Here the endothelial cells also express receptors and transporters which aid in the delivery of vital compounds to the PNS and removal of toxins from the PNS (Sano et al., 2007).

The PNS, unlike the CNS, is not protected by the bony coverings of the skull or spinal canal leaving it exposed to mechanical injury.

1.2 Schwann Cell development

SC are the principal glial cell of the PNS. They originate from neural crest cells in the embryo which first differentiate to SC precursors, then immature SC and finally either mature myelinating or nonmyelinating SC (Jessen and Mirsky, 2005). SC are necessary for axonal survival and support. The myelin sheath, produced by the SC, insulates the axon allowing fast conduction of action potentials along the axon.

Figure 2. Schwann Cell development (Mirsky et al., 2008) The dashed lines indicate

reversibility. Image reproduced with permission of the rights holder 2005 Nature Publishing

Group.

1.2.1 Early SC development Le Douarin and Smith’s studies on a bird model showed that most SC are derived from the neural crest (Le Douarin & Smith, 1988). Neural crest cells are a temporary group of cells that form from the embryonic ectoderm cell layer in vertebrates. The transition from ectoderm to neural crest is controlled by bone morphogenic protein (BMP), fibroblast growth factor (FGF) and Wnt signalling pathways (Stuhlmiller & Garcia-Castro, 2012). Early in development the neural crest cells split away from the dorsal aspect of the neural tube and migrate widely throughout the embryo and proliferate to give rise to a wide variety of cell types, including SC. The crest cells originating in the trunk region of the neural tube give rise to SC. The fate of the neural crest cells is dependent on both receptor signalling and interaction with extracellular matrix (Monk et al., 2015).

1.2.2 Schwann Cell Precursors Neural crest cells differentiate to SC precursors which are migratory and proliferative, like neural crest cells, but differ in the molecular profile and response to signalling. SC precursors express the markers Cadherin 19, desert hedgehog (Dhh), Brain fatty acid binding protein (BFABP) and Growth associated protein 43 (GAP43) along with low levels of the myelin protein markers myelin protein zero (P0), peripheral myelin protein 22 (PMP 22) and myelin proteolipid protein (PLP), these are not expressed in neural crest cells (Woodhoo and Sommer, 2008). They associate with axons and rely on neuregulin (NRG) 1, via the ErbB2/3 receptor, for survival, unlike neural crest cells whose survival is dependent on extracellular matrix (Woodhoo et al. 2004). NRG1 is also essential for proliferation and directed migration of SC precursors (Newbern and Birchmeier, 2010). Shah et al found that

NRG1 was needed for glial specification of neural crest cells and suppressed neuronal differentiation.

(Shah et al., 1994). It is postulated that the neurotropin 3 (NT3) augments the role played by NRG1; as acting via the tyrosine kinase receptor C it stimulates migration of SC and inhibits myelination

(Chan et al., 2001). Expression of NT3 correlates with the initiation of proliferation in SC and decreases when myelination begins (Cosgaya et al., 2002). The Sry-related HMG box (Sox) 10 transcription factor is expressed in neural crest cells and SC precursors but is downregulated in other cell types derived from neural crest cells (Monk et al., 2015). Mutant animals lacking Sox 10 do not have peripheral glia (Britsch et al., 2001; Kelsh and Eisen, 2000). Sox 10 expression is maintained by the transcription factor paired box (Pax) 3, which is activated by the histone deacetylases 1 and 2

(HDAC 1/2) (Doddrell et al., 2012). Pax3 is also a regulator of the expression of key SC genes P0 and

Fatty acid binding protein 7 (Fabp7) (Jacob et al., 2014).

1.2.3 Immature Schwann Cells The differentiation of SC precursors to immature SC is marked by the cessation of migration (Monk et al., 2015) and an upregulation of markers such as glial fibrillary acidic protein (GFAP), Sox2 and

S100β and a downregulation of markers such as cadherin 19 (Jessen and Mirsky, 2005). Woodhoo et al. have shown that Notch signalling is key in the SC precursor to immature SC transition (Woodhoo et al., 2009). Immature SC are able to survive in culture, by releasing a cocktail of growth factors into the culture media. SC precursors do not have this ability (Jessen and Mirsky, 1997). Immature

SC are also able to induce mesenchymal cells to form supporting structures such as vessels, by secretion of vascular endothelial growth factor (VEGF) (Mukouyama et al., 2005) and perineurial/epineurial sheaths, by secretion of desert hedgehog (Dhh) (Parmantier et al., 1999). Thus it is at this point that SC begin to be able to support their own growth and development.

1.2.4 Radial Sorting Radial sorting, the process of separating the larger diameter axons from the smaller axons, occurs when extracellular matrix proteins have been deposited and organised into basal lamina by immature SC. The immature SC’s then make contact with the axons by extending cytoplasmic processes, the SC proliferate and separate individual axons from the fascicle (Webster et al., 1973).

Several small diameter axons are then ensheathed by one non-myelinating SC and are known as

Remak bundles. Each segment of larger diameter axons form a one to one relationship with a SC, at this point the SC stops proliferating and is referred to as a pro-myelinating SC. Once SC have acquired the correct relationships with axons they can differentiate into either myelinating or non myelinating SC (Monk et al., 2015).

1.2.5 Schwann Cell Myelination The transition from a promyelinating SC to a myelinating SC is regulated by NRG1; Taveggia et al. showed that axonal NRG1 type III is the principal growth factor involved in SC myelination, via the

ErbB receptor on SC’s (Taveggia et al., 2005). NRG1 acts via the phosphoinositide 3 (PI-3) kinase/AKT/ mechanistic target of rapamycin (mTOR) (Hnia et al., 2012), mitogen activated protein kinase (MAPK) (Ishiiet al., 2013) and the calcineurin/nuclear factor of activated T-cells (NFAT) (Kao et al., 2009) pathways to induce myelination. SC can also control their own myelination via the G

Protein-Coupled Receptor GPR126 (Mogha et al., 2013; Monk et al., 2009) which increases intracellular cyclic adenosine monophosphate (cAMP). Activation of these pathways results in an increase in expression of the transcription factors Sox 10 (Bremer et al., 2011)and Oct-6

(Bermingham et al., 1996) these act together to induce Krox-20 expression (Jagalur et al., 2011), which is considered the ‘master transcriptional regulator of SC myelination’ (Parkinson et al., 2004).

One of the actions of Krox 20 is to down regulate Notch 1, which is expressed and promotes maturation until the onset of myelination (Woodhoo et al., 2009).

The neurotrophin brain derived neurotrophic factor (BDNF) has been found to promote myelination in the early stages via binding to the p75 neurotrophin receptor (p75 NTR), however p75 NTR is not expressed by mature SC, and, at later stages, BDNF inhibits myelination (Cosgaya et al., 2002). SC also release the neurotrophin nerve growth factor (NGF) which has been found to stimulate myelination in vitro (Chan et al., 2004).

1.3 Myelin

Myelin is the multilayered spiral of specialised cell membrane of the SC (or oligodendrocyte in the

CNS). It is made up of 70-85% lipids, (cholesterol, sphingolipids and some more specialised lipids) and 15-30% proteins (P0, myelin basic protein (MBP) and PMP22) (Morell and Quarles, 1999). P0 is the major protein in PNS myelin (Quarles et al., 2006). The myelin ensheaths the axon with regular interruptions known as the nodes of Ranvier, the areas of myelin between the nodes of Ranvier are called the internodes. Myelin at the internodes is compact, layers of myelin are fused together, whereas myelin adjacent to the nodes is non compact, SC cytoplasm can be found between layers of myelin (Scherer, 1996). At the nodes of Ranvier the axon membrane is in contact with the extracellular space. Myelin facilitates speed of conduction along the axon as it allows the axonal action potential to jump from one node to the next, so called salutatory conduction (Quarles et al.,

2006).

1.4 Wallerian Degeneration

First described by Augustus Volney Waller in 1850 following his observations of frog hypoglossal and glossopharyngeal nerves following axotomy, Wallerian degeneration is the process a nerve undergoes following a cut or crush injury (Waller, 1850). Wallerian degeneration is seen in both the

CNS and the PNS. Following injury the proximal and distal ends of the injured nerves retract creating a gap between the two ends of the nerve. As the axons distal to the site of injury have lost contact with the cell bodies they begin to degenerate. This process is complete within around a day in the

PNS but takes longer in the CNS (Stoll et al., 2002). This is followed by a period of regeneration.

Degeneration of the axon is necessary for the regeneration events which follow; in Wlds mice, a line with a spontaneous mutation which causes delayed Wallerian degeneration as the distal axons can survive up to 2 weeks following the injury, events such as myelin breakdown, macrophage recruitment and axonal regeneration are also delayed (Bisby & Chen, 1990).

1.5 Regeneration

The capability of CNS axons to regenerate following damage is limited but the PNS shows a remarkable capacity to regenerate. This is due to the plasticity of the SC, which are able to reprogramme following injury. When the axon degenerates, as described above, the SC in the distal segment of the nerve loses its axonal contacts. This loss of contact triggers reprogramming of the SC to a ‘repair cell’, the features of which I will outline below (figure 3).

Figure 3. Changes seen in SC following nerve injury. (Adapted from Lee et al., 2014.) Image reproduced with minor modifications with permission of the rights holder Experimental Biology 2014.

1.5.1 Schwann Cell Dedifferentiation

Dedifferentiation is the process of a fully differentiated, mature cell reverting to a progenitor cell type from its own lineage (Jopling et al., 2011). Classically the SC response to injury has been described as a dedifferentiation response (Chen et al., 2007), although it is now known that this is

19 only part of the vast array of changes seen in the SC following nerve injury. The terminally differentiated myelinating SC provides support to the axon and insulates the axon, propagating the speed of the signal down the axon. To provide this function the key genes in mature SC are those coding for the myelin protein machinery such as the transcription factor Krox 20, the master regulator of myelin transcription, myelin associated proteins such as P0, myelin basic protein (MBP),

PMP and PLP, enzymes involved in cholesterol synthesis and membrane associated proteins such as periaxin. These genes are all quickly downregulated when SC become denervated (Chen et al., 2007;

Jessen & Mirsky, 2008). Accompanying this is an upregulation in genes associated with an immature

SC phenotype; sox 2, p75NTR, L1, neural cell adhesion molecule (N-cam) and glial fibrilary acidic protein (GFAP) (Chen et al., 2007; Jessen and Mirsky, 2008). Thus following injury the SC changes gene expression profile to resemble an immature SC, therefore undergoes dedifferentiation, although SC also acquire other properties not shared by immature SC as described below. These changes in gene expression begin within 48 hours of nerve injury (Trapp et al., 1988; White et al.,

1989).

1.5.2 The Schwann Cell Repair Programme

Following injury there is also upregulation of genes not active during SC development. Most striking is the transcription factor cJun; in cJun knockout mice peripheral nerve regeneration is severely compromised whereas development of the peripheral nerve proceeds normally (Arthur Faraj et al.,

2012).

There is upregulation of genes that code for proteins which assist in repairing the nerve and recruiting inflammatory cells to the wound site. N-cadherin, BDNF, glial cell line derived neurotrophic factor (GDNF), artemin, NT3, erythropoietin, VEGF and pleiotrophin are all factors which promote axonal growth and elongation (Boyd and Gordon, 2003). These factors are all

20 upregulated in SC in the distal stump following nerve transection; artemin is not expressed in SC during development (Fantana et al., 2012) and GDNF and N-cadherin, are expressed in the SC precursor but not in the immature SC during development (Jessen and Mirsky, 2005).

The SCs in the distal segment of the nerve following injury also release a cocktail of cytokines that attract and activate inflammatory cells, particularly macrophages, to the wound site, these include tumour necrosis factor (TNF) alpha, leukaemia inhibitory factor (LIF), and interleukin (IL) 6. Some of these cytokines also promote axonal regrowth (Bauer et al., 2007).

A microarray study comparing the sciatic nerve distal to a crush injury to intact developing and adult sciatic nerves, identified over a hundred genes that were differentially regulated between intact and crushed nerves that were not developmentally regulated (Bosse et al., 2006), this suggests that the

SC response to injury is not purely a dedifferentiation response.

1.5.3 Myelin Clearance

During the first 5-7 days following nerve transection the SC are principally responsible for myelin clearance and about half of the myelin is cleared in this time period (Perry et al., 1995). Although the mechanism of clearance that the SC use has previously been described as phagocytosis (Stoll &

Muller 1999; Hirata & Kawabuti, 2002.) this is not strictly correct as the term phagocytosis applies only when something external to the phagocytic cell is internalised and digested, whereas following nerve transection there is no evidence that the myelin ever breaks away from the SC before being digested. It has been shown that myelin breaks up within the SC into oval shaped fragments which are then processed into smaller myelin segments, this is dependent upon actin polymerisation (Jung et al., 2011). These myelin segments are then digested in lysosomes in a process known as autophagy (Gomez-Sanchez et al., 2015). Following this initial SC phase of myelin clearance, a second

21 phase, which is much more dependent on macrophages proceeds. The macrophages phagocytose myelin debris (Mosley and Cuzner, 1996).

1.5.4 Axonal Guidance

Another key function of the reprogrammed SC is its ability to guide axons to their distal targets, a feature not seen in the developmental setting where axons are able to migrate to their innervation target in the absence of glial cells (Grim et al., 1992) in fact during development the SC follow the axons to the target (Heerman & Schwab, 2013) until the late stages when it is suggested the SC may influence the growth of axons (Gilmour et al., 2002; Morris et al., 1999; Nguyen et al., 2002).

Following injury the SC change morphologically; they elongate and form channels, known as bands of Bungner, which guide the axons from the proximal nerve towards the distal stump, these are first seen around 4 days following nerve injury (Svennigsen and Dahlin, 2013).

Fibroblasts accumulate at many injury sites where they secrete extracellular matrix and contribute to scar formation as well as secreting cytokines which mediate inflammation and angiogenesis, but in respect to peripheral nerve injury Parrinello et al. have shown they also play a key role in guiding

SC migration. Ephrin B on fibroblasts binds to the EphB2 receptor on SC to trigger a switch in the SC from repulsion to attraction thus allowing the SC to form bands of SC which bridge the gap between the nerve ends. The transcription factor Sox 2 has been found to be a downstream target of the

EphB2 receptor and activation causes relocalisation of the transmembrane protein N-cadherin, which mediates cell-cell adhesion, to the SC contacts. These bands of SC are necessary to guide the axons back to the distal target; disruption of these SC bands result in disruption of axonal re-growth

(Parrinello et al., 2010). Recently Cattin et al. have shown that newly formed blood vessels, formed in the nerve bridge in response to VEGF released from hypoxic macrophages, act as a scaffold for the

SC to migrate across with axons following (Cattin et al, 2015).

1.5.5 Proliferation

In contrast to the mature myelinating SC which is quiescent and forms a 1:1 relationship with an axon segment, following injury SC proliferate in order to replace lost or damaged tissue following nerve injury. In development, SC stop proliferating at the promyelinating SC stage. Proliferation of the SC is facilitated by the release of mitogens by macrophages at the injury site (Davies & Stoobant,

1990) and the peak rate of SC proliferation is seen 2-3 days after nerve injury (Clemence et al.,

1989). The work of Monje et al.’s in vitro has suggested that the initiation of proliferation is not regulated by the same genes that cause the other changes that occur in the SC following nerve injury

(Monje et al., 2010.) suggesting that proliferation is uncoupled from dedifferentiation in SC. It has been shown that SC proliferation is not necessary for peripheral nerve regeneration as mice with SC that don’t have the ability to proliferate still have normal nerve regeneration following crush injury

(Yang et al., 2008).

1.6 Reinnervation If the regenerating axons make contact with the distal targets the SC then differentiate and remyelinate the axon (Sulaiman and Gordon, 2009). In these remyelinated nerves the myelin sheath is thinner than prior to injury and the length of the internodes is shortened which leads to a slower speed of conduction of the axonal signal in the remyelinated area (Sherman and Brophy, 2005).

If, however the regenerating axons do not make contact with the intended end target, the SC forming Bands of Bungner at the distal stump will disappear and eventually be replaced by scar tissue (Hall, 1999). If repair is delayed there is a significant increase in SC apoptosis (Saito et al.,

2009). Over time SC lose the ability to promote axonal regeneration, although it seems that chronically denervated SC still retain the ability to re-differentiate to myelinating SC upon axonal contact (Sulaiman and Gordon, 2000). It has been found that the neuron is able to attempt regeneration for at least 12 months after injury (Campbell, 2008). SC in culture can undergo multiple cycles of differentiation/dedifferentiation (Monje et al., 2010)

1.7 Molecular basis of SC reprogramming following injury Although we are lacking a global picture of the transcriptional changes that occur in SC following nerve injury, there have been several key molecular players identified which regulate the changes that occur in the SC following nerve injury (figure 4).

Dedifferentiation

Dedifferentiation Dedifferentiation

Figure 4. Overview of the key molecular players involved in SC dedifferentiation. Receptors are shown in red, signalling pathways in blue, transcription factors in green and the results of activation of the transcription factors or receptors are shown in boxes. Arrows indicate activation.

1.7.1 MAPK cascades a) Ras/Raf/ extracellular signal-regulated kinases (ERK)

The Ras/Raf/ERK intracellular signalling pathway has been found to be of importance; mice which were forced to express Raf kinase showed features consistent with a demyelinating phenotype and activation of Raf alone was found to be sufficient to cause SC to dedifferentiate in the absence of nerve injury, it also caused macrophage recruitment and SC proliferation (Napoli et al. 2012). This has also been shown to occur in vitro; SC in culture respond to cAMP by differentiation; increasing expression of several myelination genes including P0, Krox 20 and MBP (Morgan et al., 1991) but SC

25 over-expressing Raf kinase were resistant to this cAMP induced differentiation (Harrisingh et al.

2004). cJun and notch signalling are thought to be downstream of Raf (Napoli et al., 2012). In SC induced to express Ras, in vitro, differentiation was inhibited and dedifferentiation was induced

(Harrisingh et al 2004). Signalling via both Raf and Ras was found to be ERK dependent (Harrisingh et al 2004, Napoli et al. 2011). High levels of ERK activation have been seen at both the proximal and distal nerve ends within hours of nerve injury (Sheu et al. 2000). Inhibition of ERK prevented SC proliferation, cytokine release and demyelination in vivo but even when ERK is inhibited p75NTR, a

SC dedifferentiation marker, is still expressed and expression of myelin protein genes is reduced

(Napoli et al., 2012; Shin et al., 2013) suggesting other factors are in play.

b) Rac/JNK c-Jun N-terminal kinases (JNK), which are activated by MAP kinase kinase 7 (MKK7), cause expression of cJun; this has been found to play a role in the dedifferentiation of SC in culture when cAMP is removed (Monje et al., 2010). JNK was also found to activate cJun in ex-vivo nerves (Shin et al.,

2013).

Rac is upstream of JNK and has been shown to be activated in SC following nerve injury (Jung et al.,

2011) and play a role in demyelination (Park & Feltri, 2011).

c) p38MAP Kinase p38MAPK activation in vitro was found to induce dedifferentiation, block cAMP activated differentiation and mediate myelin break down. It was also found to induce expression of cJun. P38

MAPK activation occurs within minutes of nerve injury in the distal part of the injured nerve (Zrouri et al., 2004). Inhibition in vivo was found to block dedifferentiation of SC following nerve injury (Yang et al. 2012).

Despite the above, around 60% of gene expression changes in SC following nerve injury occur independently of MAPK’s (Shin et al., 2013).

1.7.2 Transcription factors a) C-Jun

Mice with a conditional deletion of the transcription factor cJun were shown to have delayed loss of myelin proteins, abnormal morphology of the Bands of Bungner, reduced survival of neurones, and delayed functional recovery following nerve injury when compared to wild type mice. (Arthur-Farraj et al. 2012; Parkinson et al., 2008). These mice had normal nerves suggesting cJun is not a key player in developmental regulation (Arthur-Farraj et al., 2012).

It has been suggested that one of the key roles of cJun is to suppress krox 20 expression (Parkinson et al., 2004) as cJun is known to have a cross inhibitory relationship with the myelin transcription factor krox 20 (Parkinson et al., 2008), in that expression of one inhibits expression of the other. cJun is also fundamental in regulating the expression of neurotrophic factors such as GDNF, artemin and

BDNF, as SC lacking cJun expression fail to release these factors which are essential for axonal survival (Fontana et al., 2012).

b) Sox 2

Inducing expression of the transcription factor Sox2 in vitro was found to inhibit the SC differentiation which is usually seen with the addition of forskolin, a cAMP analogue, to the culture, the SC remained in a dedifferentiated state and showed increase in proliferation rates in response to mitogens (Le et al., 2005). As described above, Sox2 is also fundamental in formation of bands of

Bungner which are key in axonal guidance (Parrinello et al., 2010).

1.7.3 Growth Factors a) NRG1

Treatment of SC / dorsal root ganglion (DRG) co-cultures with NRG1 induced a reduction in myelin protein genes and an increase in genes associated with immature SC (Zanazzi et al., 2001). Inhibiting the ERK pathway in the NRG1 treated cells resulted in reduced cJun expression and prevented demyelination (Syed et al., 2010). This suggests that a function of NRG1 is to act via ERK to activate cJun, although reports from in vivo and ex vivo experiments suggest that cJun expression is not controlled by ERK (Shin et al., 2013; Monje et al., 2010).

In other in vivo studies NRG1 has been found to activate three of the MAPK’s (p38 MAPK, JNK and

ERK) and hence regulate cJun expression, if any of the three MAPKs were inhibited the demyelinating effects of NRG1 were not seen (Harrisingh et al., 2004; Yang et al., 2012). Adding

NRG1 to injured nerves facilitated the repair of the nerves (Chen et al., 1998).

b) Transforming growth factor (TGF) beta

TGF beta has been shown to cause proliferation of SC in vitro when used in combination with NRG1, however if NRG1 is not present an increase in apoptosis of SC is seen, a reduction in the number of

SC undergoing apoptosis following nerve injury was also seen in vivo in mice with a deletion of the type II TGF beta receptor in SC (D’Antonio et al., 2006). TGF beta has also been shown to inhibit myelination of SC in vitro (Einheber et al., 1995).

c) Fibroblast Growth Factor (FGF) 2

Zanazzi et al have shown that addition of FGF-2 to SC/neuron co-cultures inhibits myelin formation

(Zanazzi et al., 2001). FGF-2 also inhibits cAMP induced differentiation in vivo (Morgan et al., 1994).

1.7.4 Receptors a) Notch

The transmembrane receptor protein Notch has also been implicated as a key player as loss of Notch in vivo was found to impede SC dedifferentiation after nerve injury and expression of Notch was found to promote SC dedifferentiation in intact nerves (Woodhoo et al. 2009).

b) Netrin

Work by Webber et al., published in 2011 showed that the balance of expression of the netrin 1 receptors deleted in colorectal cancer (DCC) and Uncoordinated (Unc)5H2 also play roles in peripheral nerve repair. In mature myelinating SC the expression of the inhibitory receptor Unc5H2 predominates. Following injury there is upregulation of the axonal growth promoting receptor DCC and inhibition of expression of the Unc5H2 receptor. When expression of DCC was inhibited in the injured nerve in vivo axonal regeneration was impaired and expression of Unc5H2 was increased.

When the Unc5H2 was inhibited axonal growth was facilitated. These changes were independent of levels of netrin 1 (Webber et al., 2011).

1.8 Clinical Relevance

SC are key in several important pathological conditions;

1.8.1 Peripheral nerve injury

Peripheral nerve injury is a relatively common problem affecting 20 million people (Grinsell &

Keating, 2014) and costing around $150 billion annually in the US (Taylor et al., 2008). The most common cause of peripheral nerve injury is trauma, 3% of trauma cases incur a peripheral nerve injury, and most injuries occur in the upper limb (Kouyoumdjian, 2006; Svennigsen and Dahmin,

2013). Peripheral nerve injury can result in significant morbidity for the patient with loss of motor and/ or sensory function of the organ/s the affected nerve innovates and neuropathic pain can also be a significant problem. The psychological impact of these issues can be catastrophic for the patient

(Rosberg et al., 2005).

a) Classification

Peripheral nerve injury can be classified, according to the amount of disruption to the nerve, by either the Seddon or the Sunderland classification systems (see table 1) (Seddon, 1943; Sunderland,

1951.) Of these the Sunderland classification is more robust.

Sunderland Seddon Description

I Neurapraxia Conduction block. Demyelination.

II Axonotmesis Axonal loss.

III Disruption of the endoneurium.

IV Disruption of the perineurium.

V Neurontomesis Disruption of the epineurium.

Table 1. The Sunderland and Seddon Classifications of peripheral nerve injury.

The mildest of these nerve injuries is neuropraxia (Sunderland grade I) where the axons remain intact but there is loss of function of the nerve, usually due to compression. Wallarien degeneration does not occur in this setting and assuming the cause of the injury has resolved normal function of the nerve can be expected from within a few hours to within a few months, with the average neuropraxia injury taking around 12 weeks to recover from (Dumitri et al., 2001).

Disruption of the axon occurs when the injury is more severe such as lacerations, contusions, stretching or severe compression of the nerve. In these cases Wallerian degeneration does occur. In axonotomesis injuries where there is preservation of the surrounding stroma (Sunderland Grade II), the recovery time and extent depends on the amount of axonal disruption and the distance of the axon from its end target, but recovery is usually complete (Campbell, 2008). In neurontomesis

(Sunderland Grade V) there is complete loss of continuity of the nerve trunk and disruption to the surrounding tissues, reinnervation is not possible and without surgery the prognosis is very poor

(Campbell, 2008). In any injury classified as Sunderland Grade III or above, the recovery time is likely to be long and the recovery is, more often than not, incomplete (Sunderland, 1978).

b) Treatment

The first nerve repair was documented in the 7th Century by a military surgeon, Paulus Aegeneta

(Griffin et al., 2014). Since then much of the advancement in peripheral nerve surgery has been on the battle fields. Current options for surgical repair of peripheral nerves include either reattaching the two ends of the cut nerve together, neurorrhopy, or attaching the two ends of the nerve together with a graft, this is favourable if neurorrhopy would place the nerve under too much tension.

The timing of the surgery can be described as either primary, where surgery takes place within a week of the injury to the nerve, or secondary, where surgery to repair the nerve takes place at any time after a week after the damage (Payne, 2001). Results of surgery are much improved in primary rather than secondary repair, but a surgeon may advocate waiting for up to 10 weeks to see if the nerve spontaneously recovers function; it is known that axons grow at a rate of 1-3mm per day

(Deumens et al., 2010) which can aid the decision of whether surgery is necessary or not. Primary repair may also not be possible in some cases where compromise to other systems, such as fracture of the bone or vascular compromise, may be more urgent. Surgery should result in well aligned, healthy nerve fascicles, with a good vascular supply and under minimal tension, if this can not be achieved in the acute setting then secondary repair may be preferable (Payne, 2001).

In only around half of cases does the surgical repair result in a satisfactory outcome (Thomson, 2005) and more favourable outcomes have been found when the patient is younger and when the injury is more distal (Grinsell & Keating, 2014).

There has been significant advances in the knowledge of the molecular biology underpinning peripheral nerve repair and several studies have shown improvement in peripheral nerve repair by targeting these molecular mechanisms in small mammals, but there is still no therapy of this nature available to humans suffering from traumatic nerve injury (Hoke, 2006). Promising results have also been seen in animal studies when SC are transplanted into nerve conduits, these effects can be enhanced when the SC are modified to increase expression of growth factors (Li et al., 2006; May et al., 2008). Despite this the gold standard treatment of epineurial repair surgery for peripheral nerve injury has changed little in the last 25 years.

1.8.2 Peripheral Neuropathies

Peripheral neuropathy describes a wide range of conditions that affect peripheral nerves and result in impaired sensation, movement and/or organ function. The causes are wide ranging and include systemic disease (such as diabetes and HIV), iatrogenic causes (such as chemotherapy), hereditary conditions and trauma. They can be classified as axonal or demyelinating depending on the nature of the pathology seen; axonal neuropathies involve degeneration of axons and demyelinating neuropathies are caused by dysfunctional SC, although these two pathologies overlap in most cases usually one predominates (Feldman et al., 2008). A small pilot study looking at cJun expression in neuropathies found that expression was markedly increased in SC in patients found to have axonal neuropathies and was also found to be upregulated in SC in demyelinating neuropathies (Hutton et al., 2011), suggesting a similar process to that occurring in SC following injury. Below I will highlight some of the areas research into this condition has focused on.

a) Human immunodeficiency virus associated distal sensory polyneuropathy (HIV-SN)

This peripheral neuropathy affects 10-35% of people with HIV (Cornblath & Hoke, 2006; McArthur et al., 2005; Sacktor, 2002). Degeneration of small distal nerve fibres and recruitment of HIV infected macrophages to the dorsal root ganglia are seen microscopically. It is thought to be caused by mediators released by HIV infected macrophages rather direct infection of SC by HIV and many cases are thought to be due to toxicity of antiretroviral medication. A trial in 1996 tested the effects of recombinant NGF on patients with HIV-SN, although promising results were seen in terms of pain reduction many side effects were reported (McArthur et al., 2000; Schifitto et al., 2001).

33 b) Charcot Marie Tooth Disease (CMT)

CMT is a class of hereditary neuropathy which commonly presents in the first and second decade of life with balance difficulties, clumsiness and distal muscle weakness. It has a worldwide prevalence of around 1 in 2500 and is usually inherited in an autosomal dominant fashion (Irobi-Devolder,

2008). There are several subtypes, CMT1 is demyelinating and CMT2 is axonal. The most common subtype is CMT1A where sufferers have three copies of the PMP22 gene and therefore over-express

PMP22 leading to dysfunctional myelin production. Other subtypes of CMT1 are caused by loss of a copy of the PMP22 gene, mutations in the genes coding for PMP22, Krox 20 and P0, which lead to SC dysfunction (Suter and Scherer, 2003).

Observations in the mouse model of CMT1A have found that there is increased autophagy in the SC

(Gomez-Sanchez et al., 2015). When human SC carrying a mutation in PMP22 gene were grafted into nude mice the injured nerve failed to regenerate in the area of the graft indicating that the abnormal

SC also affected the axon (Sahenk et al., 1999) when the mice were treated with the NT3 axonal regeneration was improved (Sahenk et al., 2005). Given these findings Sahenk et al. went on to test

NT3 treatment in a small number of CMT patients; the patients were treated with either NT3 or placebo for 6 months and patients who received NT3 were found to have a higher density of myelinated nerve fibres on nerve biopsy and improved sensory functions when compared to those receiving placebo, this was not seen when it came to motor functions (Sahenk et al., 2005). SC in a mouse model of CMT1a were found to have increased expression of cJun, which is thought to protect from loss of axons, as when cJun was inactivated there was an increase in axonal loss

(Hantke et al., 2014).

34 c) Diabetic Neuropathy

Diabetic neuropathy is the most common peripheral neuropathy in developed countries (Said, 2007) and affects 13-40% of diabetic patients (Zochodnene, 2007). It is characterised by axonal loss, demyelination, remyelination and vessel changes (Behse et al., 1977). It is thought to be caused in part by ischaemic damage to the nerves, and by increased glucose metabolism via the polyol pathway. In vitro studies have shown that activation of this polyol pathway in SC leads to a reduction in NGF production. Replacement of NGF in diabetic neuropathy showed promising results in phase II clinical trials (Apfel et al., 1998) but the phase III clinical trials did not replicate these findings (Apfel et al., 2002).

1.8.3 Peripheral Nerve Sheath Tumours

Peripheral nerve sheath tumours (PNST) are a group of tumours that arise from the cells in a peripheral nerve. They range from benign neurofibromas, schwannomas and perineuromas to malignant peripheral nerve sheath tumours (MPNST), and can occur spontaneously or be associated with the hereditary, autosomal dominantly inherited disease, neurofibromatosis (NF). NF1 is caused by a mutation of the NF 1 gene on chromosome 17 and patients have multiple neurofibromas and less commonly, MPNSTs. NF2 is caused by a mutation of the NF2 gene on chromosome 22 and this is associated with an increased incidence of schwannomas (Mertens and Lothe, 2001). MPNST are relatively rare, account for 5-10% of all malignant soft tissue tumours, 50% are in patients with NF1

(Geller and Gebhardt, 2006).

Neurofibromas, schwannomas and some MPNST arise from SC. In neurofibromas of NF1 patients the timing of the mutation of the NF1 gene during development of the SC is critical; in a mouse model, NF1 loss at the late SC precursor/early immature SC stage resulted in neurofibroma

35 development whereas neurofibromas did not occur in peripheral nerves if loss was later in the immature SC stage (Joseph et al., 2008; Wu et al., 2008; Zaheng et al., 2008). It is suggested that the loss of SC/axon interaction is key for the development of this tumour. In axon/SC co-cultures loss of

NF1 in the SC induced dissociation from axons due to a down regulation in the SC surface protein semaphorin 4F via the Ras-Raf-ERK pathway (Parrinello and Lloyd, 2009). This loss of axonal contact allows the SC to become more responsive to mitogens and, therefore, become more proliferative

(Parrinello and Lloyd, 2009). The stroma of neurofibromas tend to contain numerous inflammatory cells, specifically mast cells, it is thought that the factors realised by these inflammatory cells provide a favourable environment for tumour progression (Le and Parada, 2007) these mast cells have been found to contain the same NF1 mutation as the SC (Yang et al., 2008). Parallels can be drawn between the loss of axonal/SC contact, SC proliferation and inflammatory environment seen pathologically in neurofibromas and physiologically following peripheral nerve injury. Similar parallels can be made in the case of schwannomas; vestibular schwannommas with NF2 gene mutations were found to have reduced expression of myelin proteins and increased expression of N-

Cam, NGFR and L1, resembling a denervated SC following nerve injury (Hung et al., 2002).

Activation of NRG/erbB signalling has been found to be increased in neurofibromas and MPNST

(Stonecypher et al., 2005) as well as schwannomas (Stonecypher et al., 2006).

1.9 Aims

While the work outlined above has identified some of the key players involved in SC reprogramming following nerve injury our knowledge of the mechanisms that initiate and maintain these changes in the SC are still incomplete and fragmented. A systems level, global approach is now needed to fully understand the mechanisms behind the changes seen in SC following nerve injury.

Unpinning these changes will lead to improved understanding of the complex process of repair following nerve injury. As well as furthering our knowledge on the complex process of peripheral nerve regeneration this may also provide important targets for improving nerve regeneration therapy and may also provide potential therapeutic targets for treatment of peripheral neuropathies and peripheral nerve sheath tumours, subtypes of which are characterised by aberrant SC dedifferentiation.

The aims of this project are

 To characterise the global transcriptional changes that occur in SC following

nerve injury, over time, using RNAseq.

 To develop an in vitro dedifferentiation assay to provide an initial screen for

analysing potential key genes found by RNAseq.

2. Materials and Methods

2.1 In Vivo Methods

2.1.1 Animal housing conditions

All animal work was carried out in accordance with the guidelines and regulations set out by the

Home Office. Animals were housed in social groups where possible and environmental enrichment was available in all cages. Animals were kept in rooms at a constant temperature and maintained on a twelve hour light/dark cycle. They were fed on standard rodent chow and water ad libitum. The animals were identified by ear notch.

2.1.2 Mouse line We had available mice from the B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Hze line from The Jackson

Laboratory, these mice have a loxP flanked STOP cassette at the Ai14 conditional allele which prevents transcription of the downstream red fluorescent protein Td Tomato. These mice were crossed with mice from the line B6N.FVB-Tg(Mpz-Cre)26Mes/J, also from The Jackson Laboratory, which expressed Cre recombinase under the control of P0. This resulted in the STOP cassette being deleted in P0 expressing tissues, i.e. SC, and expression of the tdTomato in the SC of the offspring.

The Cre expression in the mouse line is also seen in a small number of periventricular cells giving rise to cortical cells and in a subpopulation of neural crest progenitors found in the heart, as our work looks solely at the peripheral nerves the expression in these cells will not confound our results.

2.1.3 Genotyping The ear notches taken from the mice for identification were used to extract DNA for genotyping the mice. To do this the ear notches were placed in 600µl of sodium hydroxide and boiled at 95°C for 10 minutes. 100µl of Tris-HCl was added to each sample.

The polymerase chain reaction (PRC) mix for each sample was made up of the following:

5µl 5x Taq buffer (Promega)

2µl magnesium chloride (Sigma-Aldrich)

0.5µl Deoxynucleotide (dNTP) solution (New England BioLabs)

1 µl Primer (see table 2)

0.2µl Go Taq DNA polymerase (Promega)

15.3µl H2O

End Point PCR was commenced by incubating the samples at 94°C for 3 minutes 30 seconds, 61°C for

45 seconds then 72°C for 5 minutes 45 seconds for samples containing either of the tdTomato primers or 95°C for 3 minutes 30 seconds, 60°C for 30 seconds and 72°C for 5 minutes and 45 seconds if samples contained the Cre primer.

The PCR products were then separated on a 2% agarose (Invitrogen) gel containing 0.0004%

PeqGreen (Peqlab) and the DNA products were visualised under a UV light to determine the genotype.

2.1.4 Sciatic nerve transaction Sciatic nerve transection operations were carried out on mice aged 6-8 weeks. The mice were anaesthetised using isoflurane. A small longitudinal incision was made inferior to the femur, 1cm from the sciatic notch. Blunt dissection was used to expose the sciatic nerve. The sciatic nerve was then transacted at this point. The incision was closed using clips and vetergesic analgesia was given.

The mice were monitored for any adverse effects.

2.1.5 Harvesting the sciatic nerve

The mice were sacrificed using an appropriate Schedule 1 method of humane killing as outlined in the Animals (Scientific Procedures) Act 1986. The mouse was placed in a prone position and the skin was disinfected with 70% ethanol. The skin was removed and the dorsal aspect of the mouse exposed. The muscles of the posterior thigh were dissected to expose the sciatic nerve. The nerve was cut at the sciatic notch proximally and just before the bifurcation of the nerve distally in order to obtain a similar length of nerve in all cases.

2.1.6 Immunofluorescence Sciatic nerves were harvested as described above. The nerves were placed in 4% PFA for 90 minutes.

Following this the nerves were transferred into 30% sucrose (Sigma-Aldrich) at 4°C overnight. The nerves were then transferred to a 1:1 mix of 30% sucrose and optimal cutting temperature compound (OCT) (Thermo Fisher Scientific) for an hour. The nerves were then embedded in OCT and placed in liquid nitrogen. The embedded nerves were kept at -80C until being sectioned. The nerves were sectioned using a cryostat at a thickness of 10 microns. Slides were thawed at room temperature for 20 minutes. Staining using the antibodies detailed in table 5 commenced as per the protocol detailed in 2.2.9 beginning with fixation using 4% paraformaldehyde.

2.1.7 In vivo ethynyl-2’-deoxyuridine (EDU) proliferation assay EDU was reconstituted in sterile PBS at a concentration of 5mg/ml; 200µl was injected intraperitoneally into adult mice. The mice were left for two hours before being sacrificed. The nerves were harvested, mounted and sectioned as described in 2.1.6. The sectioned nerves were thawed at room temperature for 20 minutes and stained as per the EDU protocol described in the in vitro methods section 2.2.10, commencing at fixation using 4% PFA.

2.1.8 Digestion of the Sciatic nerve and SC extraction The sciatic nerves were harvested as described above and put into hepes (GE Laboratories) buffered

Hanks’ Balanced Salt Solution (HBSS) (Thermo Fisher Scientific) on ice. The nerves were then transferred to a 60mm plate and chopped as finely as possible using a scalpel. Each nerve was covered in a digestion mix containing:

25 µl 0. 3% Trypsin (Glibco) (30mg in 10ml HBSS)

25 µl 0.6% Type 2 collagenase (Worthington) (64.8mg in 10ml HBSS)

2µl 1 % Hyaluronidase (Worthington)

The nerves were incubated in the digestion mix for 10 minutes at 37°C. The mix was then triturated

5 times using a 1ml pipette tip before a further 10 minutes at 37°C. Immediately following this 200µl of dissociation buffer was added to each sample to stop the digestion reaction and the sample was triturated 20 times with a 200µl pipette tip. Each sample was then added to 5ml of dissociation buffer and centrifuged at 1000 revolutions per minute (rpm) for 10 minutes at 4°C. The supernatant was removed and the cell pellet was resuspended in 350µl of FACS buffer containing RNAsin

(Promega) (8.75µl per 350µl of FACs buffer). The samples were then sorted using the Aria III machine into either RLT buffer for RNA extraction or into dissociation media if the cells were to be plated following sorting.

2.1.9 SC specific RNA extraction RNA was extracted using the RNeasy micro kit (Qiagen). The samples were collected into 350µl of buffer RLT containing 3.5µl of beta mercaptoethanol (Sigma-Aldrich). The samples were then spun through a gDNA eliminator column at 10000 rpm for 30 seconds to remove genomic DNA. The flow through was mixed with the same volume (350µl) of 70% ethanol. The sample was then transferred to an RNeasy spin column and centrifuged for 15 seconds at 8000 rpm. The flow through was discarded. Then 70µl of buffer RWI was added to the column and spun at 8000rpm for 15 seconds.

The flow through was discarded. 500µl of buffer RPE was then added to the spin column and then the samples were spun at 8000 rpm for 15 seconds. The flow through was discarded. Next 500µl of

80% ethanol was then added to the column before a spin at 8000rpm for 2 minutes. The flow through was discarded and the RNeasy spin column was transferred to a new collection tube, this was then spun, with the lid open for 5 minutes at full speed, to completely dry the membrane. Next

14µl of RNase free water was added to the spin column and the column spun at 10000 rpm for 1 minute to elute the RNA. The RNA was then either analysed using the bioanalyser or reverse transcribed and used for quantitative PCR (qPCR) in the same way as set out in sections 2.2.5-2.2.6 below.

2.1.10 Bioanalyzer A bioanalyser (Agilent Technologies) was used to assess the quality and quantity of the RNA collected using an RNA pico chip (Agilent Technologies). The bioanalyser electrodes were decontaminated using nuclease free water. The gel dye mix was prepared by spinning 65µl of RNA pico gel mix at 1500 G force (g) for 10 minutes, before adding 1µl of RNA dye concentrate and mixing well. The RNA pico chip was then placed in the chip priming station and 9µl of the gel dye mix was placed in the well marked ‘G’. A plunger at the 1ml position was placed over this well and pressed down to prime the chip. 9µl of gel dye mix was placed in each of the 2 wells marked ‘blank G’ and

5µl of RNA pico marker was pipetted into all other wells. Then 1µl of RNA pico ladder was pipetted

43 into the well marked ladder. A 1.5µl aliquot of the sample RNA was denatured by heating at 70°C for

2 minutes, 1µl of this was then transferred to each sample well. The chip was vortexed for 1 minute then transferred to the bioanalyser for analysis.

2.1.11 Plating and Staining SC following Sorting FAC sorted cells were collected into 350µl of dissociation buffer which was transferred to a PLL coated well of a 4 well plate. The collection tube was rinsed with a further 350µl of dissociation buffer which was added to the same well. The cells, post sort, could then be visualised using bright field and fluorescent microscopy. The cells were maintained in an incubator set at 37°C with 10%

CO2. After being left overnight to adhere to the plate the cells were then stained with Calcein AM, a cell permeable dye which is used to determine cell viability. Calcein AM was added to the media at a concentration of 1:10 000 and incubated for 10 minutes before being visualised under a fluorescent microscope.

2.1.12 RNA seq library preparation The cDNA libraries for the in vivo samples were made by Dr Mel Clements, a senior research fellow. cDNA for RNAseq was prepared from total RNA using the smart seq 2 protocol (Picelli et al., 2013).

For reverse transcription 500pg total RNA was mixed with 1µl of 10µM anchored oligo-dT primer and

1µl of 10mM dNTP mix. The mixture was denatured at 72°C for 3mins and then immediately placed on ice. First strand reaction mix (0.5 µL SuperScript II reverse transcriptase, 0.25µl RNase inhibitor,

2µl Superscript II first strand buffer, 0.25µl DTT, 2µl betaine, 0.9µl MgCl2, 1µl TSO, 0.1µl RNase free water) was made and 7µl of this mix was added to each sample. Reverse transcription was initiated by incubating at 42°C for 90 min, followed by 10 cycles of (50°C for 2 minutes, 42°C for 90min). This was inactivated by incubating at 70°C for 15min. After this PCR preamplification of the first stranded mix commenced; The PCR master mix was made using 10µl of first strand mix, 25µl KAPA HiFi

Hotstart Ready Mix, 1µl ISPCR primers and 14µl nuclease free water. This was subject to 98°C for 3 min then 15 cycles of (98°C 15sec, 67°C 20sec, 72°C 6 min) with a final extension at 72°C for 5 minutes. The PCR mix was purified using 1:1 ratio of AMPure XP beads, with the final elution performed in 15µl of EB solution. Following this the library size was checked using a High sensitivity

DNA chip (Agilent Technologies) following a 1:5 dilution (expected average size should be 1.5-2.0kb and the fragments below 300bp should be negligible).

5ng of cDNA was then used for the tagmentation reaction which was carried out with the Nextera XT

DNA sample preparation kit (Illumina), with the addition of 25µl of 2x Tagment DNA buffer and 5µl of Tagment DNA enzyme, in a final volume of 50µl. The tagmentation reaction was incubated at 55°C for 5 min. The tagmented DNA was then purified using the DNA Clean & Concentrator kit (Zymo

Research), with the final elution in 20µl of resuspension buffer from the Nextera XT kit.

The whole volume was then added to 15µl Nextera PCR primer mix, 5µl of index 1 primers, 5µl of

Index 2 primers and 5µl of PCR primer cocktail for limited cycle PCR enrichment. A second round of amplification was then performed by incubating at 72°C for 3 min then 5 cycles of (98°C 10sec, 63°C

30sec, 72°C 3min). The DNA was then purified using a 1:1 ratio of AMPure XP beads. The quality of the library was then checked using a high sensitivity DNA chip and the DNA was quantified using

Qubit High Sensitivity DNA kit (ThermoFisher Scientific).The libraries were then diluted to a final concentration of 2nM, the samples were pooled and 10pmol were sequenced using the MiSeq sequencer (Illumina).

2.2 In Vitro Methods

2.2.1 Isolation and culture of rat Schwann Cells All cell culture work was carried out in a class II microbiological safety cabinet (Nuaire). Primary

Schwann cells were derived from the sciatic nerves of P7 Sprague-Dawley rats (Mathon et al., 2001).

Briefly sciatic nerves from 10 p7 rat pups were removed and finely chopped before being placed in a digestion mixture of DNase (New England Biolabs), collagenase (Worthington) and trypsin (Gibco,

Life Technologies) for 35minutes at 37°C, with gentle titration at 15 minutes. The reaction was stopped by adding 10% fetal calf serum (Sigma-Aldrich) and the cells were spun at 1300rpm for 6 minutes. The cells were re-suspended in 0.1% bovine serum albumin (BSA) (Sigma-Aldrich) in L15 media (Leibovitz) (Sigma-Aldrich). The cells were filtered through a 20µm gauze and plated on an

Ox42(Harlan sera-lab, MAS 370p) coated plate for 30 minutes at room temperature to remove macrophages. The non-adherent cells were then plated on an Ox7 coated plate and incubated for 30 minutes at room temperature to remove fibroblasts. This was repeated twice and the supernatant then spun in 0.5% BSA. The cells were re-suspendered in SC media and plated onto poly-L-lysin (PLL)

(Sigma-Aldrich) coated dishes. The cells were cultured in a dedicated cell culture incubator in maximum humidity at 37°C in 10% CO2. The media was replaced every other day.

Once confluent cells were used for experiments or split between further plates. The cells were split by aspirating the SC media off and rinsing the cells twice in phosphate buffered saline (PBS) (Sigma-

Aldrich). 1.5ml of pre-warmed trypsin was added until the cells visibly began to curl from the plate.

At this point the trypsin was removed and SC media was added to the plate. The media was used to rinse the plate to collect as many cells as possible. Cells were quantified using a casy-counter and plated onto PLL coated dishes at the appropriate density. The media was replaced the following day and every other day until cells were ready to be split again or used for experiments.

2.2.2 Dedifferentiation assay a) Differentiation assay

Treatment of rat SC in culture with 1mM dibutyryl cAMP (Sigma-Aldrich) for 3 days has been shown to induce differentiation of SC (Morgan et al. 1991). Once SC had grown to confluence, they were rinsed 3 times in serum free defined media (SATO) and transferred to SATO containing 1mM cAMP.

The cells were collected or went on to the dedifferentiation protocol after 3 days, with fresh

SATO/cAMP media at day 2.

b) Dedifferentiation assay

Following 3 days in cAMP the cells were rinsed three times in SATO to remove the cAMP. They were then transferred into, unless otherwise stated, SATO containing 10% charcoal stripped foetal calf serum (FCS) and 200ng/ml Recombinant Human Heregulin beta-1 (NRG) (Peprotech) or control conditions (SATO alone). The media was replaced every other day.

c) Interrogating the dedifferentiation assay

In an attempt to unpick the pathways involved in dedifferentiation of the cultured SC following NRG and FCS treatment the SC were treated with JNK inhibitor and/or MEK inhibitor. Following 3 days in cAMP the cells were rinsed 3 times in SATO to remove the cAMP and transferred to SATO containing

10% charcoal stripped FCS and 200ng/ml NRG with 30µM of the MEK inhibitor U0126 (Cell

Signalling) and/or 30mM of the JNK inhibitor SP600125 (Sigma-Aldrich). Cells were collected for downstream experiments after 24 or 48 hours of treatment.

2.2.3 RNA extraction Before RNA extraction the areas and equipment used were cleaned in RNAZap (Thermo Fisher

Scientific) and all materials used were RNAse free where possible. For RNA extraction cells were directly harvested into 400µl Trizol reagent (ThermoFisher Scientific). The samples were then incubated at room temperature for 5 minutes to allow nucleoprotein complexes to completely dissociate. The samples were then spun at 12 000g for 10 minutes at 4°C after which the supernatant was transferred to a fresh tube. 80µl of chloroform was then added to the supernatant and the samples vortexed well then incubated at room temperature for 15 minutes before being spun at 12000g for 15 minutes at 4°C. This allows the sample to be separated into 3 layers; the bottom organic phase containing protein and lipids, the interphase containing DNA and the aqueous top phase containing RNA. The top aqueous phase was carefully transferred to a fresh tube, ensuring that there was no interphase taken to ensure no DNA contamination of the samples. The aqueous phase was then gently mixed with 200µl of isopropanol and incubated for 10 minutes at room temperature. This was followed by an 8 minute spin at 12000g at 4°C. After this an RNA pellet was visible in the tube. The supernatant was removed without disturbing the pellet and the pellet was washed by adding 400µl of 75% ethanol and spinning at 12000g for 5 minutes. The ethanol was then completely removed and the pellet was air dried for 4 minutes before being re-suspended in 11µl of

RNase free water (Gibco Life Sciences). The amount of RNA present was then quantified using a nanodrop (Thermo Scientific).

2.2.4 DNase Treatment To remove any contaminating DNA from the RNA sample the Turbo DNA-free kit (ThermoFisher

Scientific) was used. A 0.1 volume of 10x DNase buffer and 0.2µl of rDNaseI was added to each RNA sample and gently mixed. The samples were then incubated at 37°C for 20 minutes. 2µl of DNA inactivation reagent was added to each sample and the samples were incubated at room temperature for 2 minutes. To remove the DNA inactivation reagent the samples were spun at

10000g for 1.5minutes and the RNA was transferred to a fresh tube on ice.

2.2.5 Reverse Transcription The RNA samples were then converted to cDNA using the iScript cDNA synthesis kit (BioRad). For each sample 4µl of 5x iScript reaction mix and 1µl of iScript reverse transcriptase were added to the

RNA and each sample was made up to 20µl with RNase free water. The samples were incubated at

25°C for 5 minutes, 42°C for 30 minutes and then 85°C for 5 minutes. The cDNA was then diluted with 40µl of water to a total volume of 60µl.

2.2.6 qPCR All qPCR primers were designed using the online tools GenScript (https://www.genscript.com/ssl- bin/app/primer) and Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). For a complete list of primer sequences used see tables 3 and 4.

For each sample primers were made up as follows;

10uL MESA Blue PCR Master mix (Eurogenetic)

8.5uL Water

0.5uL Primer (10uM) (Sigma-Aldrich)

1uL cDNA

Samples were repeated in duplicate. The plates were analysed using a BioRad CFX96 Real Time system and a cycle over threshold (Ct) value was obtained. The Ct value for each primer was compared to the Ct value for a house keeping gene beta 2 microglobulin (B2M) and the fold change of each primer was found by comparing the differences of each primer to the control group (delta delta Ct-method).

2.2.7 Protein extraction Cells were washed twice in ice cold PBS then 50uL of RIPA buffer (Sigma-Aldrich), with 0.1uL of phosphatase inhibitors (Sigma-Aldrich) and 0.5uL of protease inhibitor (Sigma-Aldrich), was added to the dish. Cells were scrapped into the buffer using a cell scrapper. The cell/buffer mix was transferred to an eppendorf tube and left on ice for 10 minutes. The sample was spun at maximum speed for 1 minute and the supernatant was transferred to a new tube. The protein content was measured against a BSA standard using the Bio-Rad protein assay dye reagent (Bio Rad).

2.2.8 Western Blot The desired amount of protein lysate was added to 4x laemmli buffer (Bio-Rad) and made up to a

50µl volume with water. The samples were then boiled at 95°C for 10 minutes to denature the proteins. The samples were run on polyacrylamide gels and resolved by sodium dodecyl sulphate- polyacrylamide gel electrophoresis (SDS-PAGE) to separate out the proteins. The gel was transferred to a polyvinylidene difluoride (PVDF) membrane. The membrane was blocked in 5% milk (Sigma-

Aldrich) in Tris-Buffered Saline with 0.05% Tween 20 (Prolabo) (TBS-T) for an hour at room temperature. The membranes were then transferred to the primary antibody diluted to the appropriate concentration (see table 2) in 5% milk TBS-T overnight at 4°C. The membranes were then washed three times in TBS-T before being transferred to the appropriate HRP-conjugated secondary antibody (see table 2) diluted in 5% milk TBS-T and incubated at room temperature for an hour. After 2 washes in TBS-T the membranes were transferred to TBS and electrochemiluminescence (ECL) was detected using ECL Luminata Crescendo reagent (Millipore).

2.2.9 Immunofluorescence After treatment with cAMP, NRG or control media the media was removed from the SC and the cells were fixed in 4% PFA at room temperature for 20 minutes. The cells were then washed twice in PBS before being permeabilised in 0.5% Triton X-100 in PBS for 15 minutes at room temperature, SC were then washed twice in PBS and then transferred to 3% BSA in PBS for blocking. The cover slips were then incubated in the stated primary antibodies at the concentrations highlighted in table 2, diluted in 3% BSA, overnight at 4°C. The next day the cover slips were washed 3 times in PBS and transferred to the relevant secondary antibody diluted in 3% BSA (see table 2) for an hour at room temperature. The cover slips were then washed twice in PBS and once in water before being mounted. Fluorescence was visualised using epifluoresence, images were captured using Leica software and images were processed using ImageJ.

2.2.10 EDU proliferation assay The proliferation rate of the cells was measured after treatment with cAMP, NRG or control media using the Click-iT EdU assay (Invitrogen). EDU is a thymidine analogue which is incorporated into the

DNA during cell division. The EDU is detected by a ‘click’ reaction; a covalent reaction between an azide (the Alexa Fluor dye) and an alkyne (the EDU) which is catalysed by copper.

Rat SC were plated on glass cover slips which had been PLL and laminin (Sigma-Aldrich) coated at a density of 80 000 cells per well. The SC were subject to the conditions described in the differentiation and dedifferentiation assays. EDU solution was added to the media of the cells for a final EDU concentration of 10uM, the cells were incubated with the EDU for 4 hours. After this time the media was removed and the SC were fixed at room temperature for 15 minutes using 1ml of 4% paraformaldehyde (PFA) (TAAB Laboratories Equipment) in each well containing SC. The PFA was removed and the cells were washed twice with 1ml 3% BSA in PBS. The cells were then permeabilised in 0.5% Triton X-100 (Sigma-Aldrich) in PBS at room temperature for 20 minutes. The

Click-iT reaction cocktail was prepared; per cover slip 43µl of 1xClickiT reaction buffer, 2µl copper sulphate (CuSO4), 0.12µl Alexa Fluor azide and 5µl reaction buffer additive. The cells were washed twice in 1ml of 3% BSA in PBS. The cells were then incubated in 50µl of the Click iT reaction cocktail for 30 minutes at room temperature. The cells were then washed once in 3% BSA in PBS and once in

PBS alone. The nuclei of the cells were then stained with 4’,6-Diamidino-2-Phenylindole,

Dihydrochloride (DAPI) (Insight Technology) at 1 in 5000 in PBS for 5 minutes at room temperature.

The SC were then washed twice in PBS and mounted. The SC were visualised under a fluorescent microscope.

2.1.11 RNAseq library preparation The cDNA libraries for the in vitro samples were prepared by Miss Katrin Goetsch, a research assistant. cDNA libraries for RNA seq analysis were prepared using the NEBNext Ultra Directional

RNA Library Prep Kit for Illumina (New England Biolabs). RNA was prepared from 4 cAMP treated

(differentiated) samples and 4 NRG and FCS treated (dedifferentiated) samples and treated with

DNase to remove contaminating genomic DNA as detailed in 2.1.3 and 2.1.4. The RNA was quantified using a nanodrop and 500ng of total RNA was used to make libraries. The first strand synthesis reaction buffer and random primer mix was prepared by mixing 8µl NEBNext First Strand Synthesis

Reaction Buffer (5x), 2µl NEBNext random primers and 10µl nuclease free water. The total RNA was diluted with nuclease free water to a volume of 50µl. Then 20µl of NEBNext oligo d(T)25 beads were transferred to a 0.2ml PCR tube. The beads were washed twice by adding 100µl of RNA binding buffer (2x) to the beads and mixing, then transferring the tube to a magnetic rack at room temperature for 2 minutes before removing the supernatant. Following this the beads were re- suspended in 50µl of RNA binding buffer (2x) and the 50µl of total RNA was also added to the tube.

The tube was then placed on the thermal cycler and the sample was heated at 65°C for 5 minutes then kept at 4°C to denature the RNA and help the poly-A mRNA bind to the beads. The samples were then mixed and incubated at room temperature for 5 minutes to allow the mRNA to bind to the beads. The tube was then placed on a magnetic rack at room temperature for 2 minutes to separate the poly-A mRNA bound to the beads from the solution. The supernatant was discarded then the tube was removed from the magnetic rack. The beads were washed twice by adding 200µl of wash buffer and mixing. The tube was placed on the magnetic rack for 2 minutes at room temperature before the supernatant was discarded. 50uL of Tris buffer was added to each tube and mixed. The tube was then placed on the thermal cycler and the samples heated at 80°C for 2 minutes and then held at 2°5C to elute the Poly-A mRNA from the beads. After this 50µl of RNA binding buffer (2x) was added to the sample to allow the mRNA to re-bind to the beads, this was incubated at room temperature for 5 minutes. The tube was then placed on the magnetic rack for 2

53 minutes and the supernatant removed. The beads were then washed with 200µl of wash buffer and placed back onto the magnetic rack for 2 minutes before removing all of the supernatant. The mRNA was eluted from the beads by adding 15.5µl of the First Strand Synthesis Reaction Buffer and

Random Primer Mix (x2) and incubating the samples at 94°C for 15 minutes before immediately placing the tubes on the magnetic rack. The purified mRNA was collected by transferring 13.5µl of the supernatant to a fresh tube on ice. First strand DNA synthesis was performed by adding 0.5µl of murine RNase inhibitor, 5µl actinomycin D (0.1µg/µl) (Sigma-Aldrich) and 1µL ProtoScript II reverse transcriptase to the fragmented and primed mRNA and this was incubated in a thermal cycler with heated lid set at 105°C for 10 minutes at 25°C, 15 minutes at 42°C and 15minutes at 70°C before being held at 4°C. Second strand cDNA synthesis was then performed by the addition of 48µl of nuclease free water, 8µl of Second Strand Synthesis Reaction Buffer (10x) and 4µl of Second Strand

Synthesis Enzyme Mix to the samples. This was gently mixed and incubated on the thermal cycler for

1 hour at 16°C with the heated lid set at 40°C. The double stranded cDNA was then purified by the addition of 144µl of AMPure XP beads to the sample which was vortexed and incubated at room temperature for 5 minutes. The sample was briefly spun and then placed on a magnetic rack for 5 minutes, until the solution was clear. The supernatant was then discarded. The samples were washed twice in 200µl of fresh 80% ethanol. The beads were then air dried for 5 minutes before the

DNA was eluted from the beads into 60µl of 10mM Tris-HCl with a 2 minute incubation at room temperature. The tube was then placed on the magnetic rack and 55.5µl of the supernatant was then removed to a fresh PCR tube. 6.5µl of NEBNext End Repair Reaction Buffer (10x) and 3µl of

NEBNext End Prep Enzyme Mix was added to the purified double stranded cDNA and the samples was incubated on the thermal cycler with the heated lid set at 75°C for 30 minutes at 20°C and 30 minutes at 65°C before being held at 4°C. Immediately following this adaptor ligation was performed by adding 15µl Blunt/TA ligase Master Mix, 1µl of NEBNext Adaptor (diluted 1 in 10 with 10mM

Tris/HCl, pH 7.5 with 10mM NaCl) and 2.5µl of nuclease free water, mixing and incubating at 20°C for

15 minutes on the thermal cycler. Nuclease free water was added to the samples to bring the

54 volume to 100µl and 100µl of AMPure XP beads were added to the sample before mixing and incubating at room temperature for 5 minutes. The samples were then put on a magnetic rack and supernatant was removed before the beads were washed twice with 200µl of 80% ethanol. The beads were then air dried before the DNA target was eluted from the beads by adding 52µl of 10mM

Tris-HCl, incubating at room temperature for 2 minutes and placing on the magnetic then 50µlof the supernatant was then transferred to a fresh tube and the beads were discarded. 50 µl of AMPure XP beads were added to the sample and the process was repeated. After air drying the DNA target was eluted with 19µl of 10mM Tris-HCl and incubated for 2 minutes at room temperature before being placed back on the magnetic rack and 17µl of the supernatant was transferred to a fresh tube and

PCR enrichment was performed by adding 3µl NEBNext USER Enzyme, 25µl NEBNext Q5 Hot Start

HiFi PCR master mix (2x), 2.5µl of universal PCR Primer/i5 Primer and 2.5uL of Index Primer/i7

Primer. The samples were then incubated on the thermal cycler for 15 minutes at 37°C, 30 seconds at 98°C , then 12 cycles of (98°C for 10 seconds and 65°C for 75 seconds) followed by 5 minutes at

65°C before being held at 4°C. The PCR reaction was then purified by adding 45µl of AMPure XP beads, mixing and incubating at room temperature for 5 minutes. The samples were then placed on a magnetic rack and the supernatant was then removed. The samples were washed twice in 200µl of

80% ethanol before being air dried for 5 minutes. The target DNA was then eluted into 23µl of

10mM Tris and incubated for 2 minutes at room temperature before being placed onto a magnetic rack where 20 µl of the supernatant was removed to a fresh PCR tube and an aliquot was used to asses library quality on the bioanalyser (Agilent Technologies) using a DNA high sensitivity chip

(Agilent Technologies). The samples were then sequenced using a Miseq sequencer (Ilumina).

2.3 Media and Solutions

2.3.1 Nerve dissociation media

Dissociation Media

19 mls low glucose DMEM (lonza)

5% Horse Serum (Invitrogen)

10µl/ml Penicillin/Streptomycin (TermoFisher Scientific)

______

FACs Buffer

Sterile PBS

1mM Ethylenediaminetetraacetic acid (EDTA)(Sigma-Aldrich)

25mM Hepes (GE laboratories)

1% FCS

2.3.2 Cell culture media SC Media

500ml DMEM (lonza)

15mls charcoal stripped foetal calf serum

10mls Glutamine

1.2 mls Kanamycin/Gentamicin

600µl Glial Growth Factor

500µl Forskolin

______

SATO

400ml Dulbecco’s modified Eagle’s medium (DMEM)

100µg/ml BSA

60ng/ml Progesterone (Sigma-Aldrich)

16µg/ml Putrescine (Sigma-Aldrich)

40ng/ml Selenium (Sigma-Aldrich)

50ng/ml Thyroxine (Sigma-Aldrich)

50ng/ml Triiodothyrine (Sigma-Aldrich)

______

2.3.3 Western Blot Solutions Resolving Gel 4% 6% 8% 10% 12%

30% Acrylamide 2ml 3ml 4ml 5ml 6ml

4x Tris-HCl-SDS pH8.8 3.75ml 3.75ml 3.75ml 3.75ml 3.75ml

APS 100µl 100µl 100µl 100µl 100µl

TEMED 20µl 20µl 20µl 20µl 20µl

H2O 9.13 8.13 7.13 6.13 5.13

Stacking Gel 4%

30% Acrylamide 666µl

Tris-HCl-SDS pH 6.8 1.25ml

APS 50µl

TEMED 10µl

H2O 3.024ml

Running Buffer (pH8.3)

1L distilled H2O

25mM Tris pH9.3 (30.28g)

192mM Glycine (144.13g)

0.1% SDS

______

Transfer Buffer (pH8.3)

800ml of distilled H2O

25mM Tris pH9.3 (30.28g)

192mM Glycine (144.13g)

20% Methanol

2.4 Primers for end point PCR (genotyping) Primer Forward sequence Reverse sequence (5’ to 3’) (5’ to 3’) tdTomato wild type AAGGGAGCTGCAGTGGAGT CCGAAAATCTGTGGGAAGTC tdTomato mutant GGCATTAAAGCAGCGTATCC CTGTTCCTGTACGGCATGG

Cre AATGCTTCTGTCCGTTTGCC CTACSCCAGAGACGGAAATC

Table 2: sequences of primers used for genotyping

2.5 Mouse primers for qPCR Primer Forward sequence Reverse sequence

(5’ to 3’) (5’ to 3’)

P0 CTTTGCCCTACCCCAGCTATG ACGGCACCATAGATTTCCCT

PMP ACTGTACCACATCCGCCTTG CGCACAGACCAGCAAGGATT

PLP CACCTGTTTATTGCTGCGTT ATGAAGGTGAGCAGGGAAAC

MBP ATCCAAGTACCTGGCCACAG TGTGTGAGTCCTTGCCAGAG

P75NTR AGCTCCCAGCCTGTAGTGAC GCAGCTGTTCCATCTCTTGA cJun AGCAGACGCTTGAGTTGAGA GGGTCCCTGCTTTTGAGATAA

Krox 24 CCATCACCTATACTGGCCGC AAAGGGGTTCAGGCCACAAA

Sox2 CATGGGCTCTGTGGTCAAGT TACATGGTCCAATTCCCCCG

N-cadherin ATGCTGACCACTCTCACTGC CCCGTTCACAGGGTCTATTT

Cyclin D1 GCACAACGCACTTTCTTTCC TCCAGAAGGGCTTCAATCTG

Cdh5 TGTGGGAAAGATCAAGTCCA TTCCCTGTGTTAGCATCGAC

Nefl CAAGGACGAGGTGTCGGAAA TGATTGTGTCCTGCATGGCG

Table 3: Sequences of mouse primers used for qPCR.

2.6 Rat primers for qPCR Primer Forward sequence Reverse sequence

(5’ to 3’) (5’ to 3’)

P0 ATTTATGTGCGGGGAGG TCTTTAGGAGGAGGCGACCA

PMP GCGGTGCTAGTGTTGCTCTT GATCAGTCCTGTGTCCATTGC

PLP CACCTGTTTATTGCTGCATT ATGAAGGTGAGCAGGGAAA

MBP TGACAGACTCCAAGCACACA GGCCTGTCTTTGAAGGTGTT

Periaxin GAGCGGAGTTGGTGGAGATT GTCCCCTTCCTGCAAACTGA

P75NTR ATTCTCCGATGTGGTGAGCG CCAGTCTCCTCGTCCTGGTA cJun GAAATAGGCGAGCGGCTAC TTTGCAAAAGTTCGCTCCCG

Krox 24 CCATCACCTATACTGGCCGC AAAGGGGTTCAGGCCACAAA

Sox2 TAGGGCTGGGAGAAAGAAGA GAAGTGCAATTGGGATGAAA

N-cadherin ATGCTGACCACTCTCACTGC CCCGTTCACAGGGTCTATTT

Cyclin D1 GCACAACGCACTTTCTTTCC TCCAGAAGGGCTTCAATCTG

Table 4: Sequences of rat primers used for qPCR.

2.7 Antibodies Antibody Company Raised Western IF Secondary Antibody in Concentration Concentration

S100 Dako Rabbit 1 in 1000 Alexa 488 anti Rabbit (Invitrogen) p75NTR Cell Rabbit 1 in 1000 Alexa 488 anti Signalling Rabbit (Invitrogen)

Neurofilament Abcam Chicken 1 in 5000 Alexa 488 Anti chicken (Invitrogen)

N-Cadherin BD Mouse 1 in 400 Alexa 594 Anti Biosciences mouse (Invitrogen)

GFAP Abcam Rabbit 1 in 500 Alexa 488 Anti Rabbit (Invitrogen)

P0 Sigma- Goat 1 in 1000 ECL Anti Goat IgG Aldrich horseradish peroxidase linked (GE Life Sciences)

Krox 20 Covance Rabbit 1 in 1000 ECL Anti Goat IgG horseradish peroxidase linked (GE Life Sciences)

Sox 2 Cell Rabbit 1 in 2000 ECL Anti Goat IgG Signalling horseradish peroxidase linked (GE Life Sciences)

Beta Tubulin Sigma- Mouse 1 in 5000 ECL Anti Goat IgG Aldrich horseradish peroxidase linked (GE Life Sciences)

Table 5. Details of the antibodies used for immunofluorescence and western blot.

3. Results

3.1 In vivo results We aimed to characterise the transcriptional changes seen in SC, in vivo, following nerve injury over time. Current knowledge on the key factors that control SC reprogramming following peripheral nerve injury is patchy and fragmented, and we are lacking a global overview of the changes occurring in the SC specifically. To fulfil our aim we used a mouse sciatic nerve transection model and developed a method of extracting SC specific RNA from the sciatic nerve. Below I describe the development and validation of this technique.

3.1.1 Validation of the tdTomato mouse line We generated a P0aCre;tdTomato mouse line (as described in materials and methods) in which SC are selectively labelled with a red fluorescent protein (as described in materials and methods). We began by determining the recombination efficiency in these animals by staining nerves from the tdTomato mice with the SC marker S100 and assessing the percentage of co-labelling by counting the number of cells (DAPI positive nuclei) per high power field (x40 magnification) which expressed s100 and calculating the percentage of these which also expressed tdTomato. We assessed 3 sciatic nerves from 3 different tdTomato mice, 3 high power fields were assessed in each nerve. The average number of S100 expressing cells per high power field was 279. The percentage of S100 cells also expressing tdTomato was calculated for each high power field and an average over the three high power fields for each nerve was calculated (Figure 5). We found that more than 92% of S100 positive SC also expressed tdTomato, confirming a high recombination efficiency. tdTomato expression was not seen in any other cell type examined. These results confirmed that this mouse line would prove an excellent tool for identifying and isolating SC from the nerve.

TdTomato S100 a b.

c DAPI d Merge

95 94 labelled

- 93 92 91 90 89 88 87

Percentage Percentage cells of co 86 Nerve 1 Nerve 2 Nerve 3 e Figure 5. The tdTomato/P0 cre mice have a high recombination efficiency. (a-d) Representative immunofluorescence images from (a) sciatic nerve from a TdTomato mouse which was also stained with (b) SC marker S100 and (c) the nuclear stain DAPI. (d) is the merge of images a-c. Scale bar represents 100µm. (e) Graph showing the percentage of

S100 cells co-labelled with tdTomato in three nerves from tdTomato mice. The percentage of cells co-labelled (y axis) for each nerve represents the average percentage of S100 expressing cells which also expressed tdTomato in 3 high power fields In nerve 1 868 of 926 s100 expressing cells expressed tdTomato, in nerve 2 668 of 727 s100 expressing cells expressed tdTomato and in nerve 3 792 of 861 s100 expressing cells also expressed

tdTomato. The error bars represent standard error of the mean (SEM).

3.1.2 Methods of SC specific RNA extraction Both single cell laser capture microdissection and fluorescence-activated cell sorting (FACS) are methods which can be used to isolate cells of interest based on their fluorescent properties. We thought that these methods, which exploit the red fluorescent properties of the SC in this tdTomato mouse line, may provide a way of isolating SC from the nerve. Below I describe how we began to asses these techniques.

a) Single Cell Laser Capture Microdissection

This method allows selection and collection of single cells from tissue sections using a UVA laser focused through a microscope to free the cell which is then transferred to a collection cap. This system relies on being able to readily identify the cell type of interest, the SC, under a fluorescent microscope. This can be easily done in nerves from the tdTomato mouse, without the need for a staining step. The tissue is visualised under a microscope and the cells of interest are drawn round free hand on a computer, this informs the machine of where to focus the laser on the slide to cut around the cell of interest. This is a very time and labour intensive process.. During the time it takes to draw around the cells there is bleaching of the fluorescence (as seen in figure 6 (a)), which makes identification of SC more difficult as the process goes on.

a b

Figure 6. LCM is a potential method of extracting SC specific RNA from sciatic nerves of the Tdtomato mice. (a and b) Representative images from LCM isolation of SC in tdTomato sciatic nerves. The nerves were embedded as (a) cross section and (b) longitudinally. Scale bar represents 30µm (a) and 75µm (b). Images captured from LCM software.

We optimised this method in terms of method of OCT removal, embedding the nerves and the lazer enrgy used. The RNA extracted using this method was consistently of poor quality (RNA integrity number (RIN)<2), therefore we decided that it would be best to abandon this method and focus on the FACS method I will go on to descibe.

b) Fluorescence-activated cell sorting (FACS)

FACS is a method by which a mixed population of cells is sorted in to separate groups depending on the fluorescent properties of those cells. As the mouse line we are using selectively expresses the red fluorescent protein tdTomato in the SC, FACS would allow the SC (the red cells) to be separated from the rest of the cells in the nerve. To do this the nerve first needs to be digested to a single cell suspension that can pass through the FACS machine. The cells are then sorted by the FACS machine and used for further downstream analysis. As initial attempts using this method proved promising we decided to concentrate our efforts on optimising it. I will go onto discuss the optimisation and validation of the methods used to obtain good quality SC specific RNA (RIN>7) of sufficient quantity to make libraries for RNAseq.

3.1.3 Optimisation of FACS method of SC specific RNA isolation a) Optimisation of the digestion method

To obtain a single cell suspension needed for FACS the nerve must first be digested. This is a difficult process as the nerve is ensheathed in connective tissue which is not readily digestable. The timing of the digestion is also critical; this needed to be as short as possible so that cell death is avoided and potential changes to RNA expression caused by the digestion were kept to a minimum.

We optimised the digestion in terms of enzymes used, concentration of those enzymes, and the length of time of the digestion.. We assesed the outcome by plating the cells following digestion and visualising the cells 2 hours after plating. We assesed the quality of digestion based on the number of single cells compared to the amount of nerve tissue which remained in clumps and assesed cell death based on morphological criteria as outlined in the legend below. The results of this optimisation process are detailed in tables 6-8.

Legend for tables 6-8:

Quality of Cell death digestion

Poor Mainly clumps Minimal <25% cells are of tissue. Very rounded and few single cells small.

Good Several clumps Moderate <50% of cells of tissue with are rounded easily and small. identifiable single cells.

Excellent One or two Excessive <75% of cells clumps of are rounded tissue with and small. numerous single cells identified.

Enzyme mix Quality of digestion Cell death

Trypsin, DNase and collagenase Poor Minimal

Trypsin, DNase collagenase and Good Minimal hyaluronidase

Trypsin, DNase collagenase and Good Moderate pronase

Trypsin, collagenase, DNase Excellent Excessive hyaluronidase and pronase

Table 6. Digestion using trypsin, DNase, collagenase and hyaluronidase is optimal. The table shows the outcome of different cocktails of enzymes on the quality of digestion and cell viability.

Enzyme mix Quality of digestion Cell death

Hyaluronidase 0.08% Excellent Excessive (Trypsin o.3%, collagenase 0.6%) Hyaluronidase 0.04% Good Excessive (Trypsin o.3%, collagenase 0.6%) Hyaluronidase 0.02% Good Minimal (Trypsin o.3%, collagenase 0.6%) Trypsin 0.6% Good Excessive (Collagenase 0.6% Hyaluronidase 0.02%) Collagenase 1.2% Good Excessive (Trypsin 0.3%, Hyaluronidase 0.02%)

Table 7. Digestion using 0.02% hyaluronidase, 0.3% trypsin and 0.6% collagenase is optimal. Table shows the outcome of different concentrations of enzymes on the quality of digestion and cell viability.

Time Quality of digestion

10 minutes Poor

20 minutes Good

30 minutes Good

40 minutes Excellent

Table 8. Digestion for 20 minutes is optimal. Table shows the outcome of different lengths of time of digestion on the quality of digestion.

We also assesed the outcome of different methods of tissue homogenistation prior to digestion. The

results (table 9) show the RNA integrity number (RIN), and quantity of RNA obtained following using

a TissueRuptor and chopping the nerves using a scalpel to homogenise the tissue prior to digestion.

The RIN and quantity of RNA were calculated using the Agilent bioanalyser. The RIN is calculated by

analysing the entire electrophoretic trace from the RNA sample which includes the presence or

absence of RNA degradation products Traditionally RNA integrtiry was confirmed by obtaining a 28s

to 18s peak ratio of 2 but this has been shown to provide a weak correlation with RNA integrity and

RIN has been shown to be a more reporducible and reliable method of RNA integrity analysis

(Schroeder et al., 2006).

Homogenisation Conditions RIN Quantity (pg/uL)

TissueRuptor in enzymes 3.5 14

TissueRuptor post enzymes 1.7 11

Chopping + harsh trituration (5 times 7.8 134 during and 20 times post digestion)

Chopping + gentle trituration (10 7.6 48 times post digestion)

Table 9. Tissue homogenisation by chopping the nerve and harsh trituration is optimal. The table shows the effects of different methods of tissue homogenisation on RNA quality (RIN) and RNA quantity

Having completed the above optimisation steps we concluded that 20 minutes digestion in 0.3%

trypsin, 0.6% collagenase type II and 0.02% hyaluronidase, chopping the nerves finely before

digestion and harsh tituration once the nerves were in the digestion mix were the optimal conditions

for good quality digestion with minimal cell death (detailed in materials and methods).

b) FACs conditions

To ensure that the FACS sorting was not causing a significant reduction in the quality of RNA we

compared the RIN of RNA from an intact nerve that had been digested and FACs sorted before the

RNA was extracted to that of an intact nerve that was digested in the same way but RNA was

extracted directly after digestion so the FACS sorting step was omitted. The bioanalyser plots with

RIN values are seen in figure 7.

RIN 7.6

RIN 7.3

Figure 7.FACS does not reduce the quality of RNA obtained from sciatic nerves. (a and b)

Representative bioanalyser plots showing RNA quality (RIN) from (a) a nerve digested prior to RNA being extracted and (b) SC from a nerve digested and FACS sorted prior to RNA being extracted.

Figure 7 shows that the FACS step was not having a significant impact on the quality of RNA (RIN 7.6 in unsorted cells vs RIN 7.3 in sorted cells). However we were seeing a decrease in the quantity of

RNA obtained (940 pg/uL in unsorted cells vs 36 pg/uL in sorted cells). This reduction in quantity would be expected as in the unsorted nerves there is RNA from other cell types found in the intact nerve as well SC RNA whereas the sorted cells should contain SC RNA only, there is also some inevitable loss of material associated with FACS. To overcome this we decided to pool nerves as described below.

70 c) Optimising the number of nerves needed

As the yield of RNA achieved after FACS was low we decided to see if this could be improved by pooling more than one nerve for each sample. We looked at the difference in quantities and quality

(RIN) of RNA obtained when pooling 1, 2 and 3 intact nerves, the results are shown in table 10.

Number of Intact Nerves RIN Quantitiy of RNA (pg/uL)

1 7.3 104

2 7.8 325

3 7.7 717

Table 10. Pooling the RNA from SC from 3 sciatic nerves improves the yield of RNA obtained without compromising the quality. The table shows the RNA quality (RIN) and quantity obtained when 1, 2 and 3 intact sciatic nerves were harvested, digested and FACS to obtain SC specific RNA.

Table 10 shows that following the change described above, pooling nerves had no detrimental effect on the RIN and did increase the yield of RNA. We therfore decided to pool 3 nerves for future experiments, as this reliably resulted in good quality RNA of sufficient quantities to make cDNA libraries for RNAseq. We considered any sample with a RIN greater than 7 good quality

(Sigurgeirsson et al., 2014).

3.1.4 Validating the FACS method of extracting SC specific RNA Having optimised the protocol to extract SC specific RNA using FACS as outlined above, and reliably

obtaining RNA of a sufficient quantity and quality using this method, we went on to try to ensure

that the RNA was indeed SC specific. We used several techiniques to do this as I outline below.

a) Flow cytometry

We first collected, digested and FACS sorted intact nerves from wild type (WT) and

tdTomato mice and compared the FACs plots obtained (figure 8). This shows that there was

very little autofluresence in the tdTomato spectrum and in the cells from the tomato mice a

clear population of tdTomato positive cells could be identified. We used a stringent gate

well above where the level of autoflurescence in the tdTomato spectrum in the WT mice

stopped in order to ensure there was no contamination of the tdTomato SC with other cell

types. We also did not collect any cells with high levels of green autoflourescence as these

were likely to represent a population of dead or cells.

Green fluorescence Green

a b TdTomato fluorescence TdTomato fluorescence

Figure 8. A clear group of tdTomato positive cells is demonstrated by FACS in sciatic nerves from tdTomato mice (a and b) show fluorescence of cells from the sciatic nerves of WT and tdTomato mice. The images are representative FACS plots of digested sciatic nerves from (a)

wild type and (b) tdTomato mice highlighting levels of tdTomato (red) and green autofluorescence. The lines represent the gates used to collect the tdTomato positive b)fraction Inspecting cells after FACS sorting

In order to know exactly what cells were obtained after sorting we harvested and digested 3 intact nerves from from tdTomato mice and FACs sorted the tdTomato positive and the tdTomato negative cells into dissociation media. The cells were then plated onto a PLL coated 4 well plate and left for 4 hours to attach. After this time the cells were treated with Calcein AM, a live/dead cell indicator which is able to enter live cells where it is converted to green flourescent calcein by the action of esterases in the living cell. We then visualised the cells under a flourescent microscope (see figure 9) which showed that in the tdTomato positive fraction the majority of red cells were staining green, meaning that most of the SC were still viable following sorting and there were very few cells staining green which were not also co-stained red, meaning that the vast majority of the live cells were tdTomato positive therefore SC, which suggests we have minimal contamination in the FACS sorted SC from intact nerves. No red cells were seen in the tdTomato negative fraction. Around 200 SC were seen from each digested nerve.

a tdTomato b Calcein

c Bright field d Merge Figure 9.Viable SC are seen following sciatic nerve digestion and FACS.(a-d) Representative images at x40 magnification of cells obtained after FACS showing (a) tdTomato expression, (b) calcein AM (live cell indicator) staining, (c) bright field image and (d) merge images of a-c. Scale bars represent 50µm.. c) Using qPCR

We decided to further validate the method by harvesting, digesting and FACs sorting intact nerves and collecting the tdTomato positive fraction (the SC) and the tdTomato negative fraction

(the other cells found in the nerve), RNA was made from these two fractions. We also harvested and digested whole intact nerves and made RNA, omiting the sorting step. The RNA from the three groups (tdTomato positive, tdTomato negative and whole nerve) was made into cDNA and qPCR was performed using markers of mature SC (myelin proteins P0, PLP and PMP22) and

74 markers to help to detect axonal (Nefl) or endothelial (Cdh5) contamination, in order to help confirm the SC specific nature of the samples. When comparing the tdTomato positive fraction to the whole nerve (figure 10a) and to the tdTomato negative fraction (figure 10b) there was an increase in the differentiated SC markers and dowregulation of genes associated with axons and fibroblasts, showing enrichment of SC markers in the FACS sorted sample.

18 Differentiated SC Axon Endothelium 16 14

12 Whole 10 Intact 8 Sorted 6 Intact

Relative mRNA expression mRNA Relative 4 2

0 a P0 PLP PMP22 Nefl Cdh5

7 tdTomato 6 Negative Intact 5

3 tdTomato Positive Relative mRNA expression mRNA Relative 2 Intact 1

0 P0 PLP PMP22 Nefl Cdh5 Figure 10. Mature SC markers are enriched in the FACS SC when compared to the whole nerve and to the tdTomato negative fraction. (a and b) Graphs showing the expression of differentiated SC (P0, PLP and PMP22), axonal (Nefl) and endothelial (Cdh5) markers in FACS sorted SC from intact nerves compared to (a) whole intact nerve (n=2) Error bars represent SEM of biological replicates. and (b) the tdTomato negative fraction of FACS sorted nerve (n=1) Error bars represent SEM of technical replicates. Each gene was normalised to the house keeping gene B2M. 75

We went on to validate this method further by ensuring the changes we would expect to see in the

SC following nerve injury are seen when SC sorted using this method from intact nerves were compared to SC from the distal stump of sciatic nerves 3 days post transection (figure 11). These showed the changes expected; downregulation of differentiated SC markers (myelin proteins P0,

PLP, PMP22 and MBP), and upregulation of the dedifferentiation markers (p75, sox2 and cJun) and a proliferation marker (cyclin D1) in the sorted cells from the distal stump of the cut nerve.

Differentiated SC Dedifferentiated SC Proliferation 100

Sorted Intact

Sorted 10 Cut Relative mRNARelative expression

1 P0 PLP PMP22 MBP p75 sox2 cJun cyclin D1

Figure 11. Myelin proteins are downregulated and dedifferentiation markers are upregulated in FACS SC following nerve transection when compared to FACS SC from intact nerves. The graph 0 shows expression levels of differentiated (P0, PLP, PMP22 and MBP) and dedifferentiated (p75, Sox2 and cJun) SC markers and a proliferation marker (cyclin D1) in digested and FACs sorted SC from intact nerves and the distal end of nerves 3 days post sciatic nerve transaction. n=1. Error

bars represent the SEM of technical replicates. Each gene was normalised to the house keeping gene B2M.

We have used flow cytometry, immunoflurescence and qPCR to validate this method of sorting a population of SC from sciatic nerves in mice and reliably extracting good quality (RIN>7) RNA of sufficient quanitiy to use for RNAseq.

3.1.5 Characterising the timing of the SC response to injury The SC response to injury begins within minutes of the injury and can last several months, dependent on when and if the axons make contact with the distal targets. Therefore, it stands to reason that different sets of genes will be upregulated and downregulated over this period of time and when we look at the SC at a certain time point we are just getting a snapshot of that instant and missing the bigger picture. We aimed to get a temporal picture of the transciptional changes in SC following nerve injury by performining RNAseq on SC from the distal end of the transected nerve at various time points.

In order to pick the best time points at which to analyse the SC transcriptome we began to characterise the timining of the dedifferentiation response. We did this by performing qPCR on cDNA made from RNA extracted from SC sorted using FACs from intact nerves and the distal stump of sciatic nerves 24hours, 48 hours, 72 hours and 7 days following sciatic nerve transection (figure

12). We assesed expression of differeantiated SC markers (myelin proteins P0, PMP, PLP and MBP), dedifferentated SC markers (p75, sox2, cJun and N-cadherin) and a proliferation marker (cyclin D1).

Differentiated SC Dedifferentiated SC Proliferation

64.00 * * 32.00

16.00 * * Intact 8.00 * * * * 24hours 4.00 48hours 2.00 72hours 1.00 7days P0 PMP PLP MBP p75 Sox2 cJun N cyclin D1 0.50 cadherin 0.25 * 0.13 * * 0.06 * 0.03 Relative mRNA expression (log scale) (log expression mRNA Relative 0.02 * * * 0.01 * 0.00 * 0.00 Figure 12. Downregulation of myelin proteins and upregulation of dedifferentiation markers at the mRNA level can be seen by 24 hours post nerve transection in FACS SC from the distal cut nerve. The graph shows the expression of differentiated (P0, PMP, PLP and MBP), dedifferentiated (p75, Sox2, cJun and N-cadherin) SC markers and a proliferation marker

(cyclin D1) in SC from the distal sciatic nerve at the stated time points following nerve transection. (n=3) * signifies p<0.05 using a 2 tailed unpaired students t test, comparing each time point to the intact sample. Error bars represent SEM of biological replicates. Each gene was normalised to the house keeping gene B2M.

Figure 12 shows that there is down regulation of some mature SC markers by 24 hours and by 72 hours all the mature SC markers are downregulated. It also shows there is upregulation of dedifferentiated SC markers by 24 hours. There is also increse in the proliferation marker by 24 hours.

Finally we decided to further characterise the timing of the SC repair response in the first 3 days following nerve injury at the protein level and in terms of proliferation. We did this by harvesting nerves 12 hours, 24 hours, 36 hours, 48 hours, 60 hours and 72 hours following sciatic nerve transection. For each time point 2 nerves were analysed The animals had recieved an intraperitoneal

EDU injection 2 hours before being sacrificed. The contralateral intact nerves were taken as controls.

The nerves were sectioned and stained with EDU (prolifferation marker), p75 (a dedifferentiation marker) and neurofilament (NF) (an axonal marker). For EDU quantification the number of tdTomato positive SC in three fields per high power field for each nerve were counted and the number of these

SC expressing the green fluorescence indicating EDU incorporation and therefore proliferation were counted and a pecentage of SC undergoing proliferation was calculted, the results are shown in figure 13a. The average number of cells counted per high power field was 310 and the average number of SC analysed per nerve was 808.

In the intact nerve p75 expression is minimal (figure 13f). Dedifferentiated SC express p75, this can be seen in figure 13g. p75 expression was analysed by measuring the total length of the nerve in milimeters (mm) and the length of nerve with p75 expression (in mm) and expressing the p75 expression as a percentage of the total length of the nerve ((length of nerve expressing p75(mm)/total length of nerve (mm))x100) this is seen in figure 13b.

NF expression is uniform and linear in intact nerves (figure 13d), when the axon undergoes degeneration following nerve injury the NF staining is disrupted (figure 13e). To quantify this the total length of the nerve was measured in mm and the length of nerve where NF staining was disrupted was measure in mm and the percentage of the nerve with disrupted NF was calculated

((legnth of nerve with disrupted NF staining (mm)/total length of nerve (mm))x100), this is seen in figure 13 c.

100 10 80 8 60 6 40 4 p75 with 20 2 % of EDU positive cells positive EDU % of 0 stained nerve of length % of 0 a Intact 12 h 24 h 36 h 48 h 60 h 72 h b Intact 12 h 24 h 36 h 48 h 60 h 72 h

Intact Distal 72 hours

100 DAPI , 80 tdTomato , ,

60 NF d e 40 NF NF staining DAPI

20 ,

% of nerve with disrupted disrupted with nerve % of 0 tdTomato Intact 12 h 24 h 36 h 48 h 60 h 72 h , p75 f g c Figure 13:SC proliferation and dedifferentiation, and axonal degeneration are seen by 24-36 hours following nerve transection. SC dedifferentiation and axonal degeneration is seen

throughout the nerve by 72 hours. The graphs show (a) percentage of EDU positive SC (b) the percentage of the nerve staining with p75NTR and (c) the percentage of the nerve that showed axonal degradation as seen by disrupted NF staining in intact sciatic nerves and the distal end of sciatic nerves at the stated time points following transection. (n=2). (d-g) are representative images at x40 magnification of NF staining in green (d and e) and p75 staining in green (f and g) in intact (d and f) and the distal nerve stumps72 hours post transection (e and g) .tdTomato is in red and nuclear DAPI staining is in blue. Scale bar represents 100µm.

Figure 13 shows that SC prolifereation is seen initially at 36hours following transection and the number of proliferating SC continue to rise with time (a). The protein expression of p75NTR, a key dedifferentiation marker, begins at 36 hours following nerve transection, by 60 hours p75NTR expression is seen throughout the examined distal nerve (b). Axonal degeneration, as evidenced by disruped NF staining, begins at 24 hours after nerve transection and by 48 hours post sciatic nerve transection the entire length of the examined distal nerve showed evidence of axonal degeneration

(c). When analysing these sections we noted that EDU expression, p75 expression and NF disruption started at the proximal tip of the distal nerve end and progessed distally along the nerve as time went on, which suggests that dedifferentiation begins at the cut end of the distal nerve and progresses in a wave like fashion distally along the nerve, by 60-72hours the entire nerve appeared to be dedifferentiated.

Unfortunately, due to time constraints it was not possible to perform RNAseq at multiple time points but this data shows that to collect the most infomative data several early time points should be analysed up to 3 days; we decided that collection 18, 36, 54 and 72 hour time points should catch key changes in the SC transcriptome. We also decided a further later time point at 6 days following nerve transection would provide useful information to compare the early time points to a later time point where the wound would be more established.

3.1.6 RNAseq of intact vs distal nerves In order to assess the global transcriptional changes that occur following nerve injury in SC we performed RNAseq on cDNA libraries made from the RNA extracted from 3 groups of SC FACS sorted from intact nerves and 3 groups of SC FACS sorted from the distal end of a nerve 6 days post transection. See appendix A for RIN values of samples used for RNAseq. With the help of our colleagues at the MRC bioinformatics facility a list of normalised RNA-seq expression data was generated which included the fold change of 10005 known genes (24107 in total) between the intact and distal samples, along with the significance of the change (p value). 2347 were found to be significantly changed between the intact and distal group (p<0.05), of these 1138 (48.5%) were up regulated and 1209 (51.5%) were down regulated in the SC from the distal nerve compared to SC from intact nerve.

We began to analyse the RNAseq data by looking at expression changes between the distal and intact nerve of the key genes we used for qPCR in validating the technique (figure 14). These represented mature, differentiated SC markers (P0, PMP22, PLP and MBP) and markers of dedifferentiation (p75NTR, Sox2, cJun, N-cadherin) and the proliferation marker cyclin D1.

Differentiated SC Dedifferentiated SC Proliferation

-2

-4 Fold Change in the Distal Nerve the Distal in Fold Change -6

-8

Figure 14.Following nerve transection there is downregulation of myelin proteins and, upregulation of some dedifferentiation markers and a proliferation marker demonstrated by RNAseq. The graph shows the fold change in expression of myelin proteins(P0, PMP22, PLP and MBP), dedifferentiated SC markers (p75, Sox2, cJun and N-cadherin) and the proliferation

marker cyclin D1 in SC from the distal stump 6 days post nerve transection when compared to SC from intact nerves using RNA seq.

Figure 14 shows there is downregulation of myelin associated genes and the dedifferentiation markers cJun and N-cadherin, and upregulation of the dedifferentiation markers p75NTR and Sox2, and the proliferation marker cyclin D1 when comparing the SC from the distal end of nerves 6 days post transection to SC from intact nerves. These results recapitulate the results seen by qPCR on SC sorted from the distal nerve at the 7 day time point in figure 12.

Gene set enrichment analysis (GSEA) was performed on the normalised RNAseq data of differentially expressed genes between samples from the SC of the distal end of cut nerves 6 days following transection compared to from SC from intact nerves. GSEA compares the data to known groups of genes which share common biological functions, location or regulation (Subramanian et al., 2005). It involves generating an enrichment score which reflects how well the genes known to be up or

83 downregulated in the given functional groups match the levels of expression of genes in the RNAseq data. The GSEA calculates 4 key statistics for each of the groups of genes the data is compared to

(Subramanian et al., 2005) . These are;

 Enrichment score (ES): this reflects the degree to which genes are overlap at the top or

bottom of a ranked list of differentially expressed genes. A positive ES indicates enrichment

of genes that are upregulated and a negative ES indicates enrichment of genes that are

downregulated. The enrichment plot is a graphical representation of the ES for a gene set

(Subramanian et al., 2005).

 Normalised enrichment score (NES): ES are normalised to account for differences in gene set

size therefore can be used to compare results from different gene sets (Subramanian et al.,

2005).

 False discovery rate (FDR): is the estimated probability that the NES for a gene set is a false

positive. In most cases where the results are used to generate protential future research

targets an FDR cut off of 25% is appropriate (q<0.25) (Subramanian et al., 2005).

 Nominal P value: is the estimated statistical significance of the ES for each gene set. The p

value is not adjusted for gene set size and multiple hypotheses testing (unlike FDR)

(Subramanian et al., 2005).

Using GSEA to compare our RNAseq data to groups of genes regulated by different transcription factors showed, when using an false discovery rate (FDR) cut off of q<0.25, that no transcription factors were significantly upregulated. However, using a nominal P value cut off of p<0.01, 3 significantly upregulated transcription factors were identified. Also 2 significantly down regulated transcription factors were identified using a FDR cut off of q<0.25. (Figure 15).

Oct1

Androgen Receptor

ETS2

Runx1

Myc

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Normalised Enrichment Score

Figure 15.Down and upregulated transcription factors found by GSEA of RNAseq data when comparing SC from the distal nerve to SC from the intact nerve. The bar chart shows the normalised enrichment scores of significantly upregulated (p<0.01) and down regulated (FDR q<0.25) gene sets of transcription factors.

Figure 16 summarises the results of GSEA, when looking at genes enriched in the SC from the distal nerve compared to those from the intact nerve.

extracellular matrix interation (2) metabolism (protein) (2) lipid synthesis (13) cell adhesion (3) stem cell (2) transcription and translation (14) neural development (1) metabolism (protein) (4) lipid degradation (2) immune response/inflammation (16) glycolysis (5) cell cycle (18) extracellular matrix interation (9)

-3 -2 -1 0 1 2 3 Average normalised enrichment score Figure 16.Downregulation and upregulation in gene sets identified by GSEA, grouped according to function, reflects SC biology. The bar chart shows the average normalised enrichment scores of significantly upregulated (FDR<0.25 and p<0.01) and down regulated (FDR<0.25) functional

groups of genes. The number of individual gene sets involved in each group is indicated in brackets. Results are from the GSEA analysis of the normalised RNAseq results when comparing the SC from the distal nerve to the intact nerve. 85

Figure 16 shows that in SC from the distal nerve there is downregulation of genes involved in lipid synthesis and cell adhesion. This is explained by the fact that the SC turn off the myelination programme and aquire migratory properties following nerve injury. There is also downregulation of genes involved in extracellular matrix interactions and protein metabolism in SC from the distal nerve as compared to SC from the intact nerve. There is upregulation of genes involved in transcription and translation, protein metabolism, glycolysis and cell cycle as would be expected in these transcriptionally active proliferating SC. Upregulation in the extracellular matrix interaction genes is in keeping with the migratory abilities the SC acquire and upregulation in genes associated with lipid degradation is in keeping with the myelin clearance function the SC have following nerve injury. There is also upregulation of genes involved in the inflammatory resonse in the SC of the dsital nerve in keeping with the recruitment of inflammatory cells seen following nerve injury. Also upregulated are groups of genes associated with stem cell renewal and neural development.

In 2005 Le et al profiled sciatic nerves from mice with a hypomorphic Krox20 allele, As krox20 is the master transcriptional regulator of myelination, the hypomorphic krox20 allele is essentially driving

SC dedifferentiaion without nerve injury. Le et al. used microarray technology and found the genes differentially expressed between the hypomorphic krox20 nerve and an age matched nerve from a

WT mouse.. When comparing these differentially expressed genes to our in vivo data there was significant enrichment of genes differentially expressed in the hypomorphic krox20 nerve when compared to a WT mouse with those expressed in the SC from the distal nerve 6 days following transection when compared to SC from intact nerves. Figure 17 shows the enrichment plot of genes upregulated (a) and downregulated (b) in the hypomorphic krox 20 nerves when compared to the genes differentially expressed in the SC from the distal nerves as compared to the intact nerve.

86 a b Figure 17. SC from the distal nerve show a similar gene expression profile to that of nerves from a mouse with a hypomorphic krox 20 allele. Enrichment plots from our GSEA showing the enrichment of genes (a) upregulated and (b)downregulated in nerves of mice with hypomorphic krox 20 gene when compared to a WT mouse(Le et al., 2005) and those genes differentially expressed in SC from distal nerves when compared to intact nerves.

As krox 20 is known to be down regulated in SC following nerve injury (Warner et al., 1998) these striking similarities between our in vivo data and the expression profiles of nerves with SC with reduced krox20 function help to confirm that our samples are SC specific and we are reflecting SC biology when analysing our results.

3.2 In vitro Results Our aim was to develop a dedifferentiation assay that would recapitulate the changes seen in vivo in

SC following nerve injury, in a cell culture system. This would provide an initial screening method to analyse target genes implicated in the dedifferentiation process.

Treatment of rat SC in culture with 1mM cyclic adenosine monophosphate (cAMP) for 3 days has been shown to induce differentiation of SC (Morgan et al., 1991), We modified this protocol by introducing a step, following this cAMP induced differentiation, to induce the SC to dedifferentiate.

Below I describe the development and validation of this assay.

3.2.1 Validating the differentiation assay To ensure that SC were differentiating optimally when treated with cAMP we carried out RNA extraction and performed qPCR on SC treated with cAMP in serum free defined media (SATO) for 3,

5 and 7 days. We looked at expression of markers of differentiated SC (PMP, MBP and periaxin) and of dedifferentiated SC (p75 and cJun), see figure 18.

Differentiated Dedifferentiated 64.00 32.00 16.00 No cAMP 8.00 4.00 cAMP 3 2.00 days 1.00 0.50 PMP MBP Periaxin p75 cJun cAMP 5 0.25 days 0.13 cAMP 7 0.06 days 0.03 0.02 Relative mRNA expression (logscale) expression mRNA Relative 0.01 0.00 0.00

Figure 18. cAMP treatment of SC for 3 days is sufficient to induce differentiation. The graph shows the fold change of expression differentiated and dedifferentiated SC genes after treatment with cAMP for set time periods. n=1. Error bars represent SEM of technical repeats. Each gene was normalised to the house keeping gene B2M. 88

We found that treatment with cAMP for 3 days was sufficient to induce differentiation, with no apparent improvement in differentiation seen in cells treated for 5 and 7 days. This is consistent with the literature (Morgan et al., 1991).

3.2.2 Optimising the dedifferentiation SC assay It is known that in these cAMP differentiated SC the removal of cAMP alone induces SC to down regulate the mature myelin gene expression and increase expression of some dedifferentiated SC markers (Monje et al., 2010). My goal was to enhance this dedifferentiation aspect of the assay. To begin we treated SC with cAMP in SATO for 3 days. Following removal of cAMP we then attempted to induce the SC to dedifferentiate by treatment with varying levels of FCS as following nerve injury the blood nerve barrier is breached and the SC come into contact with serum. This was done by washing off cAMP then treated the differentiated SC with SATO alone and SATO supplemented with

3%, 5% and 10% FCS for 3 days. We collected the samples, extracted RNA, reverse transcribed this to cDNA and looked at expression of differentiated SC markers (P0, PLP and PMP) and the dedifferentiated SC markers (cJun, Krox24 and p75NTR) by qPCR (figure 19).

Differentiated Dedifferentiated 4.00

2.00 cAMP 1.00 P0 PLP PMP c-Jun Krox 24 p75NTR cAMP 0.50 removal

0.25 3% Serum

0.13 5% Serum 0.06 10% 0.03 Serum

0.02 Relative mRNA expression (ddct) scale log (ddct) expression mRNA Relative 0.01

0.00

Figure 19. Treatment of SC with FCS following cAMP induced differentiation does not enhance the dedifferentiation response when compared to cAMP removal. The graph shows the fold change of expression of mature SC markers (P0, PLP and PMP) and dedifferentiated SC markers (cJun, krox24 and p75) when cAMP is removed from SC and when SC are treated with 3%, 5% or 10% FCS following cAMP induced differentiation (n=2). Error bars represent SEM of biological replicates. Each gene was normalised to the house keeping gene B2M.

Figure 19 shows that FCS did not seem to reliably improve the downregulation of differentiated SC markers or the increase the expression of dedifferentiated SC markers. However the cells treated with FCS looked healthier and there was less cell death than in the sample without FCS.

We then decided to investigate the effect of treating cAMP differentiated SC with cocktails of different growth factors for 3 days. The growth factors used were chosen from factors shown in the literature to prevent myelination (NRG, TGF beta), cause demyelination (NRG)( Zanazzi et al. 2001) and cause SC proliferation (NRG, TGF beta and hepatocyte growth factor (HGF)) (Einheber et al.

1995; Krasnoselsky et al. 1994), the results are seen in figure 20.

Differentiated Dedifferentiated Differentiated Dedifferentiated 4.00 4.00 2.00 2.00 cAMP 1.00 cAMP 1.00 0.50 0.50 0.25 0.25 cAMP 0.13 cAMP 0.13 Removal ( log scale) 0.06 removal

(log (log scale) 0.06 0.03 TGF 0.02 0.03

Relative mRNA expression NRG 0.01 * 0.02 0.00 Relative mRNA expression 0.01 a b 0.00 0.00

4.00 2.00 cAMP 4.00 1.00 2.00 cAMP 0.50 1.00 0.25 0.50 0.13 cAMP removal 0.25 0.06 0.13 cAMP Removal (log (log scale) 0.03

( log scale) 0.06 0.02 HGF 0.03 0.01 0.02 NRG TGF Relative mRNA expression

0.00 Relative mRNA expression 0.00 0.01 c d 0.00 0.00

4.00 2.00 4.00 cAMP 1.00 2.00 cAMP 0.50 1.00 0.25 0.50 P0 PLP PMP MBP Sox2c-jun 0.13 0.25 0.06 cAMP 0.13 cAMP 0.03 Removal 0.06 Removal (log (log scale)

0.02 (log scale) 0.03 0.01 0.02 0.00 NRG HGF 0.01 NRG TGF Relative mRNA expression 0.00 Relative mRNA expression 0.00 HGF 0.00 0.00 e 0.00 f 0.00 Figure 20. Treatment of SC with NRG, TGF and HGF, alone or in combination, following cAMP induced differentiation does not enhance the dedifferentiation response when compared to cAMP removal The graphs show the fold change of expression of differentiated (P0, PLP, PMP and MBP) and dedifferentiated(sox2, krox 24 and cJun) SC genes after treatment with (a) NRG n=4, (b) TGF n=4, (c) HGF n=2, (d) NRG and TGF n=4, (e) NRG and HGF n=2 and(f) NRG, TGF and HGF n=2. * signifies p<0.05 using a 2 tailed unpaired students t test when compared to cAMP removal. Error bars represent SEM of biological replicates. Each gene was normalised to the house keeping gene B2M.

Figure 20 shows that when cultured with growth factors alone, other than PMP being significantly down regulated in the NRG treated SC when compared to the SC where cAMP was removed, there is

91 no significant improvement in the dedifferentiation response seen with the addition of growth factors alone. The cells also looked unhealthy and a lot of cell death was observed. We postulated that the addition of 10% FCS may have a synergistic effect when used in combination with the added growth factors, improving the dedifferentiation response observed and suspected it may also improve cell survival. As significant cell death was seen in the samples treated with HGF we decided to omit this from further analysis.

Differentiated Dedifferentiated

16.00 8.00 4.00 2.00 cAMP 1.00 0.50 P0 PMP MBP Periaxin Krox24 c-Jun

0.25 cAMP Removal 0.13 0.06 0.03 NRG FCS 0.02

Relative mRNA expression (log scale) (log expression mRNA Relative 0.01 0.00 0.00 a

Differentiated Dedifferentiated Differentiated Dedifferentiated 4.00 4.00 cAMP 2.00 cAMP 2.00 1.00 0.50 1.00 0.25 0.50 0.13 cAMP cAMP removal 0.25 removal 0.06 0.13 (log (log scale) (log (log scale) 0.03 0.02 0.06 TGF FCS NRG

0.01 Relative mRNA expression Relative mRNA expression 0.03 TGF FCS 0.00 0.00 0.02 b 0.00 c 0.01 * Figure 21. Treatment of cAMP differentiated SC with NRG and 10% FCS show promising results which is not seen with TGF and FCS alone or augmented by the addition of TGF. The graphs show the fold change in expression of differentiated (P0, PLP, PMP22 and MBP) and dedifferentiated (Sox2, krox24 and cJun) SC markers when treated with (a) NRG and 10% FCS n=1, (b) TGF and 10% FCS n=4 and (c) NRG, TGF and 10% FCS n=2. * signifies p <0.05 using a 2 tailed unpaired students t test comparing to cAMP removal Error bars represent SEM of technical replicates (a) and biological replicates (b and c). Each gene was normalised to the house keeping gene B2M.

When 10% FCS was used in combination with TGF beta (figure 21b), other than a significant reduction in MBP, there was no significant improvement in the dedifferentiation response. However, these preliminary studies did show promising results when NRG was used in combination with 10%

93 serum (figure 21 a); there was down regulation of the differentiated SC genes tested and up regulation of the dedifferentiated SC marker genes to a greater extent than that seen with cAMP removal alone. These effects on the repair genes seemed to be lost when TGF was added to the NRG and 10% serum (figure 21 c). We therefore decided to further characterise the changes seen when cAMP differentiated cells were treated with NRG with 10% FCS as outlined below.

3.2.3 Characterising the dedifferentiation SC assay We went on to characterise this dedifferentiation assay by looking at expression of key differentiated and dedifferentiated SC markers by qPCR. Treating SC with a combination of NRG with

10% FCS after cAMP induced differentiation of the cells caused a reproducible and reliable downregulation of differentiated SC genes (P0, PMP, MBP and periaxin) and up regulation of genes associated with a dedifferentiation (cJun, p75, sox2 and N-cadherin). This was seen at the mRNA level (figure 22).

Differentiated Dedifferentiated 16.00

8.00 * * 4.00 cAMP 2.00 *

1.00 cAMP 0.50 removal

0.25 * * NRG 0.13 FCS

0.06

0.03 *

Relative mRNA expression (ddct) scale log (ddct) expression mRNA Relative 0.02 * 0.01 * 0.00 Figure 22. Treatment of cAMP differentiated SC with NRG and 10% FCS enhances the

dedifferentiation response at the mRNA level. The graph shows relative expression levels of differentiated (P0, PMP, MBP and periaxin) and dedifferentiated (krox24, cJun, p75, Sox2 and N- cadherin) SC markers in cells treated with NRG and 10% FCS. n=4. * signifies p<0.05 using a 2 tailed unpaired students t test comparing each treatment group to cAMP. Error bars represent SEM of biological replicates. Each gene was normalised to the house keeping gene B2M.

This was also seen at the protein level, with a reduction in the protein concentrations of krox 20 and

P0 (compared to cAMP removal) and an increase in the protein levels of sox 2 (as compared to cAMP) seen by western blot (figure 23), and by an increase in expression of the dedifferentiation markers GFAP and N Cadherin seen by immunofluorescence (figure 24) in the cells treated with NRG and 10% FCS as compared to those treated with cAMP.

cAMP NS Media Removal cAMP NRG + FCS Krox 20 Differentiated P0

Beta Tubulin

Sox 2 Dedifferentiated

Figure 23. Treatment of cAMP differentiated SC with NRG and 10% FCS causes reduced protein

expression of the myelin transcription factor krox20, and the myelin protein P0, and increased expression of the dedifferentiation transcription factor Sox2. This is a representative western blot showing protein expression of mature SC (krox 20 and P0) and dedifferentiated SC (Sox2) markers. A loading control (beta tubulin) was used.

a cAMP b cAMP Removal

GFAP – Green

N cadherin – Red

DAPI - Blue

c NRG 10% FCS

Figure 24.Treatment of cAMP differentiated SC with NRG and 10% FCS induces increased protein expression of dedifferentiation markers GFAP and N-cadherin. (a-c) Representative images at x40 magnification of cells treated with (a) cAMP (b) cAMP then SATO and (c) cAMP then NRG and 10% FCS and stained with GFAP (green), N cadherin (red) and DAPI (blue). Scale bar represents 50µm.

Following nerve injury SC regain proliferative properties. To determine if SC treated with NRG and

FCS also had this characteristic feature we performed EDU proliferation assays on cells treated with cAMP, cells where the cAMP had been removed and replaced with SATO, cells treated with NRG and

FCS after cAMP removal and SC growing in the normal SC media. The experiment involved analysing

2 cover slips per condition, each with 80,000 SC seeded on them. 100 SC (DAPI positive nuclei) from

10 different fields over the 2 cover slips for each condition were counted and the number of these that were EDU positive was also recorded and expressed as a percentage (total of 1000 SC analysed per condition). The experiment was repeated 3 times The results show that when cells were treated

97 with cAMP the proliferation rate was reduced, treatment with NRG and FCS caused an increase in the proliferation rate which was not a feature of the cAMP removal group (figure 25).

EDU DAPI EDU DAPI

a SC media b cAMP

EDU DAPI 18 * 16 * 14 12 10 8 6 4

% of EDU positive cells positive EDU % of 2 * 0 NS cAMP cAMP NRG FCS c NRG 10% FCS d Removal

Figure 25. Treatment of cAMP differentiated SC with NRG and 10% FCS causes the SC to regain proliferative abilities. (a-c) Representative images at x20 magnification of cells treated with EDU after being cultured in (a) standard SC media, (b) cAMP and (c) cAMP for 3 days then NRG with 10% FCS for 3 days EDU is red and nuclear stain in DAPI blue. Scale bar represents 30µm (d) quantifies the percentage of EDU positive cells seen in each condition In each experiment 1000 SC were analysed for each condition. n=3. *indicates p<0.05 using a 2 tailed unpaired students t

test when compared to cAMP. Error bars represent SEM of biological replicates

These results show that treatment of SC that have undergone cAMP induced differentiation with

NRG and 10% FCS induces the SC to down regulate differentiated SC genes, up regulate dedifferentiation genes (as seen at mRNA and protein levels) and proliferate, as is seen in SC following nerve injury in vivo.

3.2.4 Interrogating the in vitro dedifferentiation response In order to get a better understanding of the pathways involved in the dedifferentiation response seen when the cells are treated with NRG and 10% serum, and to begin to understand how well these overlap with the pathways involved with dedifferentiation in vivo we tested the effect of inhibitors of MEK and JNK pathways on the dedifferentiation response.

MEK is upstream of ERK, which is thought to be a key signalling pathway involved in SC reprogramming after injury (Harrisingh et al., 2004; Napoli et al., 2012) and NRG1 is known to inhibit myelination in a MEK dependent manor (Syed et al., 2010). Although we found that the addition of the MEK inhibitor to the NRG 10% FCS media in the dedifferentiation SC assay did not prevent the down regulation of mature SC genes (P0, PMP, MBP and periaxin) or up regulation genes associated with SC dedifferentiation (p75, sox2 and cJun) as seen in figure 26. This indicates that NRG does not act via the MEK pathway in our in vitro system.

It is known that the increase in expression of cJun in vivo (Blom et al., 2014) and in vitro when cAMP is removed from SC in culture (Monje et al., 2010) is via JNK activation. Monje et al., also showed that inhibiting JNK did not prevent krox 20 from being lost when cAMP was removed from SC (Monje et al., 2010). Our data supports this as adding a JNK inhibitor to the NRG FCS media for 48 hours prevented the increased expression of dedifferentiated SC markers (p75, sox2 and cJun) but did not stop the differentiated SC genes (P0, PMP, MBP and periaxin) from being downregulated, as seen in figure 26.

Differentiation Dedifferentiation 16.00 8.00 4.00 2.00 1.00 cAMP P0 PMP22 MBP periaxin p75 Sox 2 cjun 0.50 NRG FCS 24

0.25 * MEK 24 0.13 JNK 24 0.06 MEK/JNK 24 0.03

Relative mRNA expression (log scale) (log expression mRNA Relative 0.02 0.01 a 0.00

16.00 8.00 4.00 2.00 * 1.00 0.50 P0 PMP22 MBP periaxin p75 Sox 2 cjun cAMP 0.25 0.13 NRG FCS 48 0.06 MEK 48 0.03 JNK 48 0.02 0.01 MEK/JNK 48 0.00 0.00 Relative mRNA expression (log scale) (log expression mRNA Relative 0.00 0.00 b 0.00

Figure 26. Inhibiting the MEK and JNK pathways in the NRG and FCS treated SC did not completely inhibit the dedifferentiation response The graphs show the relative expression of differentiated (P0, PMP22, MBP and periaxin) and dedifferentiated (p75, Sox2 and cJun) SC markers following treatment with cAMP, NRG and 10% FCS with MEK inhibitor (n=2), JNK inhibitor (n=4) and both MEK and JNK inhibitors (n=2) for (a)24 hours and (b) 48 hours. * signifies p<0.05 using a 2 tailed unpaired students t test compared to the NRG FCS treated SC without inhibitors. Error bars represent SEM of biological replicates. Each gene was normalised to the house keeping gene B2M.

100

Figure 26 shows that the dedifferentiation response induced in SC by NRG with 10% FCS cannot be completely inhibited by either JNK inhibition, MEK inhibition or a combination of the two. Although the genes associated with a progenitor state are significantly inhibited with JNK inhibition the myelin genes are suppressed to similar levels in all conditions. This suggests that multiple signalling pathways regulate the repair response seen with the addition of NRG and 10% FCS to SC in culture and further work is necessary to unpick the pathways involved.

101

3.2.5 RNA seq of in vitro samples In order to assess the global transcriptional changes that occur in the differentiated cAMP treated cells and the dedifferentiated NRG and FCS treated cells we performed RNAseq on cDNA libraries made from the RNA extracted from 4 different 10cm plates of SC treated with cAMP (differentiated) and 4 different 10cm plates of SC treated with cAMP then NRG and 10% FCS (dedifferentiated). Each

10cm plate contained approximately 800,000 SC. With the help of our colleagues at the MRC bioinformatics facility a list of normalised RNAseq expression data was generated which included the fold change of 10124 known genes between the differentiated and dedifferentiated samples, along with the significance of the change (p value). 6121 were found to be significantly changed between the differentiated and dedifferentiated group (p<0.05), of these 3565 (59.7%) were up regulated and

2556 (40.3%) were downregulated in the dedifferentiated SC compared to the differentiated SC.

We began to analyse the RNAseq data by looking at expression changes, between the dedifferentiated and differentiated in vitro SC, of the key genes we used for qPCR in validating the assay (figure 27). We looked at expression of the mature, differentiated SC markers (P0, PMP22, periaxin and MBP) and of differentiated SC markers (p75, Sox 2, cJun and N-cadherin).

102

Differentiated Dedifferentiated 4

0 P0 PMP 22 periaxin MBP p75 Sox2 cJun N cadherin

SC -2

-4

-6 Fold Change in the Dedifferentiated the Dedifferentiated in Fold Change

-8

Figure 27.Treatment of cAMP differentiated SC with NRG and 10% FCS causes downregulation of myelin protein genes and upregulation of dedifferentiation genes as seen by RNAseq The graph shows the fold change of differentiated (P0, PMP22, periaxin and MBP) and dedifferentiated (p75, Sox2, cJun and N-cadherin) in the dedifferentiated SC group (NRG/FCS treated) when compared to the differentiated (cAMP treated) SC group using RNAseq.

Figure 27 shows that there is down regulation of the differentiated SC genes and up regulation of the dedifferentiated SC genes seen when in the in vitro dedifferentiated SC (NRG and FCS treated).

This recapitulates the changes seen when qPCR was used to validate the assay (figure22).

103

3.3 Comparing in vitro dedifferentiation model to in vivo data To begin to understand how the in vitro assay recapitulates SC reprogramming that occurs following peripheral nerve injury in vivo, we have compared the RNAseq GSEA data from the in vitro and in vivo models.

We began by looking at the expression of functional groups of genes expressed in the SC from the distal nerve (in vivo) and how these overlapped with those expressed in the dedifferentiated SC in culture (figure 28a) and the same with the intact nerve (in vivo) and compared to the differentiated

SC in culture (figure 28b).

(11) (10) (3) a

(3) (7) (8)

Figure 28 .The in vitro SC dedifferentiation assay shares some common features with our in vivo data. This shows an intersection of functional groups of genes upregulated in (a)SC from the distal nerve in vivo (green set) and dedifferentiated SC in vitro (yellow set) and (b) SC from the intact nerve in vivo (green set) and differentiated SC in vitro (yellow set). The numbers in brackets are the numbers of functional groups in each set. 104

Figure 28 shows that there were 21 different functional groups of genes with increased expression in the SC of the distal nerve when compared to SC from the intact nerve and that there was upregulation of 13 different functional groups of genes in the dedifferentiated SC when compared to differentiated SC. Of these 10 groups were upregulated in both the SC from the distal nerve and the deddifferentiated SC. There was upregulation of 10 different functional groups of genes in the SC of the intact nerve when compared to SC from the distal nerve and 11 groups upregulated in the differentiated SC. Of these 3 groups were upregulated in both the SC from the intact nerve and differentiated SC.

The GSEA showed that there was enrichment of genes downregulated in nerves from mice with a hypomorphic krox20 nerve when compared to nerve from a WT mouse (Le et al., 2005) seen in the differentiated SC in vitro (figure 29).

Figure 29.There is enrichment of genes downregulated in the hypomorphic krox 20 nerves in cAMP treated, differentiated SC in vitro The figure shows an enrichment plot from our GSEA showing the enrichment of genes down regulated in nerves from mice with hypomorphic krox20 allele when compared to nerves from WT mice (Le et al., 2005) with those in the differentiated SC in culture when compared to dedifferentiated SC in vitro.

105

This significant enrichment of genes downregulated in the hypomorphic krox 20 nerves with those expressed in differentiated SC in vitro (figure 29) is also seen in vivo when comparing the distal nerve to the intact (figure 17b). However there was not an enrichment of the genes upregulated in the nerve from the hypomorphic krox20 nerve with those upregulated in the dedifferentiated SC in vitro, as was seen in vivo (figure 17a). Figure 30 compares the number of individual genes upregulated (a) and downregulated (b) in the nerves from the hypomorphic krox20 mouse to those up and downregulated in our in vivo and in vitro models.

UP DOWN

a b

Figure 30. Some genes up and downregulated in the hypomorphic krox20 nerves are also differentially expressed in SC from the distal nerve in vivo and dedifferentiated SC in vitro. Intersection of genes (a) upregulated and (b) down regulated in the nerves of hypomorphic krox20 mice (red set) with those (a) up and (b) down regulated in the SC from the distal nerve (green set) and the dedifferentiated SC (yellow set) when compared to the intact nerve and differentiated SC.

Figure 30 shows that of the 87 genes that were upregulated in nerves from the hypomorphic krox20 mouse when compared to nerves from WT mice, 44 (50.6%) were also upregualted in the SC from the distal nerve when compared to SC from the intact nerve and 18 (20.7%) were also upregulated in the differentiated SC when compared to the dedifferentiated SC. Of these 7 genes were upregulated in all 3 conditions. Of the 87 genes downregulated in nerves from the hypomorphic krox20 mouse when compared to nerves from WT mice, 57 (65.5%) were also upregualted in the SC from the distal nerve when compared to SC from the intact nerve and 48 (55.2%) were also upregulated in the

106 differentiated SC when compared to the dedifferentiated SC. Of these 33 genes were upregulated in all 3 conditions.

We also compared our RNAseq data to genes found to be differentially expressed in crush injured rat nerves when compared to intect nerves using microarray technology (Bosse et al., 2006) (figure 31).

Figure 31. Some genes differentially expressed after crush injury in rat sciatic nerves are also differentially expressed in SC from the distal nerve in vivo and in dedifferentiated SC in vitro.

Intersection of genes differentially expressed in crush injured nerves when compared to intact nerves (Bosse et al.) (red set) to those in the SC of distal nerve compared to SC of intact nerve in

vivo (green set) and dedifferentiated SC compared to differentiated SC in vitro (yellow set).

Figure 31 shows that of the 84 genes differentially regulated in the sciatic nerves of rats following crush injury (Bosse et al., 2006) 37 (44%) were also differentially regulated in the SC from the distal nerve and 36 (42.9%) were also up regulated in the dedifferentiated SC. Of these 21 genes were differentially expressed in all 3 conditions.

Appendix B is a list of the genes differentially regulated in hypomorphic krox20 nerves (Le et al.,

2005) which are also differentially expressed in our in vivo SC from distal nerves post transection when compared to SC from intact nerves and our in vitro NRG/FCS, dedifferentiated SC when compared to cAMP treated differentiated SC. Appendix C is a list of the genes differentially regulated in crushed rat nerves (Bosse et al., 2006) which are also differentially expressed in our in vivo SC

107 from distal nerves post transection when compared to SC from intact nerves and our in vitro

NRG/FCS, dedifferentiated SC when compared to cAMP treated differentiated SC. Appendix D is a list of the genes differentially regulated in both the hypomorphic krox20 nerves (Le et al., 2005) and the crush injured rat nerves (Bosse et al., 2006) which are also differentially expressed in our in vivo SC from distal nerves post transection when compared to SC from intact nerves and our in vitro

NRG/FCS, dedifferentiated SC when compared to cAMP treated differentiated SC.

These results show that the in vitro dedifferentiation assay shares some genetic features with its in vivo counterpart, this was seen when comparing in vitro data to pulished literature (Le et al., 2005;

Bosse et al., 2006) and to our own in vivo data. This provides evidence that this dedifferentiation assay would provide a useful screening tool for validating potential key genes implicated in the dedifferentitation response in the future.

108

4. Discussion

4.1 SC reprogramming in vivo

Although there have been significant advances in our understanding of the molecular mechanisms that underlie peripheral nerve repair over recent years, our knowledge is still patchy and we are lacking a global understanding of the changes that occur and the factors that control these.

Of the nine papers which use global profiling techniques to compare injured to intact nerves seven use microarray technology (Arthur Farraj et al., 2012; Barrette et al., 2010; Bosse et al., 2006;

D’Antonio et al., 2006; Kubo et al., 2002; Nagarajan et al., 2002; Bosse et al., 2001) which has the disadvantage of needing transcript specific probes so cannot detect novel transcripts and is also less sensitive and specific at detecting differential expression of genes than RNAseq. Only two recent papers (Han et al., 2016, Yi et al., 2015) used RNAseq technology to profile transcriptional changes between intact and injured nerves. In six of these papers (Han et al., 2016; Yi et al., 2015; Barrette et al., 2010; Bosse et al., 2006;Nagarajan et al., 2002; Bosse et al., 2001) a crush method of injuring the sciatic nerve (as opposed to transection) was used; after crush injury the basal membrane of the nerve is still intact therefore peripheral nerve regeneration is facilitated. In all nine of these papers

(Han et al., 2016; Yi et al., 2015; Arthur Farraj et al., 2012; Barrette et al., 2010; Bosse et al., 2006;

D’Antonio et al., 2006; Kubo et al., 2002; Nagarajan et al., 2002; Bosse et al., 2001) the whole nerve was used meaning the transcriptional changes seen could be due to changes in SC, axons, fibroblasts or inflammatory cells. Our work is novel in the fact that it isolates SC from the nerve to look for SC specific transcriptional changes following nerve injury. It also uses a transection model of sciatic nerve injury, as opposed to crush. This makes our work more clinically relevant as medical intervention is more likely to be needed following a transection injury than a crush injury.

109

Our aim was to get a global picture of the transcriptional changes that occur in SC following nerve injury. To do this we developed a method of isolating SC specific RNA from sciatic nerves using FACS, the RNA we collected was of sufficient quantity and quality to perform RNAseq. One drawback of the method was the apparent loss of material caused by sorting (36pg/µL of RNA was obtained from one

FAC sorted intact sciatic nerve compared to 940pg/µL from an intact nerve when only the FACS step was omitted). We overcame this issue by pooling nerves so that we had sufficient quantities of RNA to make cDNA libraries from. We also used a method of library preparation that was designed for very small quantities of input RNA (Picelli et al., 2013).

RNAseq was performed on SC specific RNA extracted from nerves distal to the injury site at 6 days post sciatic nerve transaction and intact nerves. The normalised RNAseq data gave results for a total of 10005 known genes and significant changes between the distal and intact groups were seen in

2347 known genes (p<0.05).

When comparing the results of the RNAseq data (figure 14) to the qPCR data from FACS sorted SC from the distal nerve at set time points following nerve transection (figure 12), the RNAseq data recapitulated the changes seen at 7 days by qPCR. It may seem surprising that at this time point cJun, a transcription factor known to be a key player in SC repair response (Arthur-Farraj et al.,

2012), appeared to be down regulated. This is also the case for N-cadherin, which was also shown to be downregulated in SC from the distal nerve in our RNAseq data (figure 14) and also by qPCR at 7 days (figure 12). However when GSEA was performed on the RNAseq data cJun was identified as an up regulated transcription factor (p<0.05), meaning that genes known to be targets of c-Jun overlapped with genes up regulated in the SC of the distal nerve 6 days following nerve injury. This suggests it would be beneficial to perform RNAseq at earlier time points to get a temporal picture of

110 the transcriptional changes in SC following nerve injury. Collection of samples at the 6 day time point may be too late to catch changes in transcription factors which may happen initially and then be switched off at later stages. The fact that our RNAseq data could identify no significantly up regulated transcription factors using FDR <0.25 cut off, supports this.

The genes which were found to be up-regulated in the SC from the distal nerve 6 days following transaction in vivo (figure 16) were genes known to play roles in functions that SC are known to acquire following nerve injury; an enrichment of genes associated with transcription and translation, protein metabolism, glycolysis and the cell cycle was seen in keeping with the proliferative properties the SC regain, also genes associated with inflammation and extracellular matrix were found to be up regulated which is in keeping with the role the SC have in recruiting inflammatory cells to the wound site and the migratory abilities the SC acquire following nerve injury. Genes associated with lipid degradation were also shown to be up regulated in the SC of the distal nerve, this would fit with the myelin breakdown function the SC serve following nerve injury. There was also a down regulation in genes involved in lipid synthesis in the SC from the distal nerve when compared to SC from the intact nerve in vivo, this fits with the switching off of the myelination programme which occurs in SC following nerve injury. The enrichment of genes which play roles in these key functions in SC of the distal nerve following injury reflect known SC biology and reassure us that the samples we have analysed are SC specific.

The GSEA analysis of the RNAseq data showed that genes found to be up in the nerves from a mouse with reduced function of Krox 20 when compared to nerves from WT mice (Le et al., 2005) were enriched in the SC from the distal nerve in vivo (figure 17a) and genes down regulated in the hypomorphic krox 20 mouse were enriched in SC from the intact nerve in vivo (see figure 17b). Krox

20 has been reported to be the ‘master regulator of myelin transcription’; it is known to be essential for the myelination of SC during development (Nagarajan et al., 2001; Topilko et al., 1994), and is

111 downregulated following nerve injury (Warner et al., 1998). Krox 20 has been reported to have a cross-inhibitory relationship with cJun, which is thought to be mediated via JNK signalling (Shin et al.,

2013). This striking overlap of genes from our RNAseq data with those from the nerves from the hypomorphic krox 20 mouse helps to confirm the SC specific nature of our samples and further reassures us that the changes we are seeing in the in vivo samples do reflect SC biology.

In analysing the RNAseq GSEA results we came across several areas which might prove interesting for future analysis;

 Myc Transcription Factor

Genes regulated by the expression of the transcription factor Myc were found to significantly

overlap with genes differentially expressed in the distal nerve compared to the intact nerve

(p<0.01). Myc target genes and the genes involved in the Myc activation pathway were also

found to highly significantly correlate with the genes differentially expressed in the distal nerve

compared to the intact (FDR<0.01). Myc is a transcription factor which plays a role in cell

proliferation, apoptosis and cell transformation. In humans, Myc is thought to control the

expression of around 15% of all genes (Patel et al., 2004). N-myc downstream regulated gene 1

(NDRG1), which is known to be repressed by myc during mouse development (Shimono et al.,

1999) is known to be highly expressed in mature SC and lost in SC following nerve injury at the

late stages of myelin breakdown, with protein expression of the gene seen again at the

remyelinating stage (Hirata et al., 2004). In 2004 Berger et al. showed that loss of this NDRG1

gene led to a demyelinating neuropathy (Berger et al., 2004). The highly significant overlap of

these genes which change in the SC following injury with genes regulated by Myc suggest that

further work into the role of Myc and its downstream targets in peripheral nerve repair may be

an interesting area of future research.

112

 Runt-related transcription factor 1 (RUNX1)

Runx1 was identified as a significantly up regulated transcription factor in the GSEA of our in vivo

RNAseq data (p<0.01). Runx1 or acute myeloid leukaemia 1 protein (AML1) is a transcription factor that regulates the maturation of hematopoietic stem cells. It is also known to play a key role in the development of unmyelinated sensory neurons (Liu and Ma, 2011; Lallemend and

Ernfors, 2012). Runx1 has recently been found to contribute to neurofibroma formation seen with neurofibromatosis type 1; Runx 1 over expression was seen at the mRNA and protein levels in mouse neurofibromas and inhibition of Runx1 in vitro decreased neurofibroma sphere formation (Li et al., 2015). This suggests that characterising Runx1 expression following nerve injury may provide interesting results.

 Oct 1 transcription factor

Of the transcription factors targets which GSEA revealed to be significantly down regulated in the SC of the distal nerve, Oct1 was found to be significant (FDR<0.25). Oct 1 is a ubiquitously expressed transcription factor with many downstream targets. It has been found to serve as a sensor of cell stress and via cAMP is shuttled from the nucleus to the cytoplasm in response to cell stress (Zwilling et al., 1994). Oct1 is a member of the POU family of transcription factors, another member of this family, Oct6, is known to be an important regulator of myelination in SC

(Kawasaki et al., 2003), this family of transcription factors have a common DNA binding domain

(POU domain) therefore it may be the case that the reported down regulation of Oct1 may actually be reflecting changes in Oct6 expression which would be expected to be downregulated following nerve transection. However Oct1 can bind to the TATA box binding protein in the transcription factor IID (TFIID), the TATA binding protein was also found to be significantly down regulated in the SC from the distal nerve (p<0.01). It is reported that c-Jun binds to TATA-binding

113 protein-associated factor-1 (TAF1) which causes a reduction in TFIID driven transcription (Lively et al., 2001). This may suggest that some of the actions of c-Jun activation following nerve injury may be mediated via down regulation of Oct1 and TATA binding protein therefore loss of interaction with TFIID. This may be a potential area of future research.

 Integrins

Integrins are transmembrane receptors which facilitate interactions between cells and extracellular matrix and mediate signals from the extracellular matrix to modify cell behaviour.

The beta 4 integrin subunit has long been reported to be expressed by both mature myelinating

SC and SC exposed to cAMP in culture and to fall rapidly after nerve injury (Feltri et al., 1994). It comes as no surprise therefore that GSEA showed the integrin 4 and A6B1_A6B4 integrin pathways to be significantly downregulated (FDR<0.25) in SC from the distal nerve compared to the intact. Interestingly Feltri et al reported that the beta 1 integrin subunit levels did not fall after nerve injury and incidentally noted that they may rise (Feltri et al., 1994), our work has supported this observation as the Integrin1 pathway and the integrin A9B1 pathway were significantly up regulated in SC from the distal nerve (FDR<0.25). The integrin beta 1 subunit is known to be essential for radial sorting during development (Feltri et al., 2002). This indicates that nerve injury may cause a switch in integrin expression from B4 subunit to B1 subunit which may be an interesting area to examine.

 Epithelial mesenchymal transition (EMT) genes

GSEA analysis of our data showed enrichment of genes associated with the EMT (FDR<0.01). .

EMT is a process in which epithelial cells lose their relationship with their surrounding cells and acquire the migratory and invasive properties of mesenchymal stem cells. Loss of E cadherin is

114

thought to be fundamental in EMT and in our samples E cadherin was downregulated more than

fivefold in the distal nerve. Several signalling pathways thought to induce EMT, such as TGF-beta,

FGF, HGF and notch, (Wang et al., 2010) are also thought to be key player in the reprogramming

of mature SC to a repair SC following nerve injury . EMT plays a key role in development, wound

healing and cancer metastasis. The migratory ability of SC following nerve injury plays a key role

in axonal guidance, suggesting that this EMT switch of SC is also fundamental in peripheral nerve

repair. GSEA also showed that there was enrichment of genes associated with extracellular

matrix interaction (n=8, FDR<0.02) and downregulation of groups of genes associated with cell

contact and adhesion (n=3, FDR<0.25) in the SC of the distal sample as compared to the intact

which fits with the migratory properties the SC develop following nerve injury. In 2010 Arima et

al., suggested that EMT played a key role in the development of neurofibromas caused by the

loss of neurofibromin (Arima et al., 2010). Recently SC have been shown to have the ability to

induce EMT in salivary cystic adenoid carcinoma cells (Shan et al., 2016) which may suggest that

factors released by SC act in an autocrine fashion to promote EMT. This suggests that EMT genes

could provide key therapeutic targets for enhancing peripheral nerve repair and treating SC

associated tumours in the future.

In summary we developed a method of extracting SC specific RNA from sciatic nerves and used this method to look at the transcriptional changes between SC in intact nerves and SC in the distal end of nerves 6 days after transection, using RNAseq. These results have highlighted several potentially interesting areas to study to further our knowledge of peripheral nerve regeneration following injury.

Unfortunately, due to time constraints, it was not possible to sequence RNA from the SC of distal nerves at different time points following nerve injury. However, our preliminary data has suggested

115 the concentrating on time pints within the first 72 hours following nerve injury with a later time point at 6 days would potentially give the most informative results.

116

4.2 In vitro repair assay We aimed to develop an in vitro dedifferentiation assay to use as a high throughput means of analysing potential key genes found to be differentially regulated in SC in vivo following nerve injury.

We have shown that, following cAMP induced differentiation of SC in culture, it is possible to enhance the dedifferentiation response known to occur upon cAMP removal, with the addition of

NRG and 10% FCS to the media.

We have demonstrated that the mRNA levels of the differentiation SC markers P0 and MBP decrease and PMP and periaxin significantly decrease, and that the dedifferentiated SC genes N-cadherin increase and cJun, p75 NTR and sox 2 increase significantly (p<0.05) when cAMP is removed from the media and NRG and 10% FCS is added (see figure 22). Only the downregulation of PMP and periaxin was also significant in SC treated with cAMP removal. Additionally the protein levels of krox

20 and P0 are decreased to a greater extent than seen with removal of cAMP alone (see figure 23) and the protein levels of GFAP and N-cadherin are increased in the dedifferentiated SC group compared to the cAMP removal group (see figure 24). We have also shown that these NRG and FCS treated cells also regain proliferative properties, this was not seen in SC where cAMP was removed

(see figure 25).

NRG appears to have opposing actions; during development not only is SC interaction with axonal- bound NRG essential for SC survival, proliferation and directed migration at the SC precursor stage it also is the principle factor involved in SC myelination later in development, and axons lacking NRG1 type III leads to hypomyelination of the nerve (Taveggia et al., 2005). The addition of NRG to

SC/DRG co-cultures has demonstrated that soluble NRG1 type II is sufficient to induce demyelination and promote dedifferentiation of SC (Zanazzi et al., 2005) and addition of NRG1 to injured nerves facilitates nerve repair (Chen et al., 1998). Syed et al., have shown that these opposing actions may

117 be down to the concentration of NRG1; at low concentrations it has a myelin promoting effect while at higher concentrations it inhibits myelination (Syed et al., 2010).

To assess how closely our in vitro dedifferentiation assay recapitulated the transcriptional changes seen in SC in vivo following nerve injury, we went on to send samples of cAMP treated SC

(differentiated) and SC treated with cAMP then NRG and 10% FCS (dedifferentiated) for RNAseq. The normalised RNAseq data gave results for a total of 10124 known genes (26405 genes in total) and significant changes (p<0.05) between the differentiated and dedifferentiated groups were seen in

6121 known genes.

When comparing the GSEA results of functional groups of genes enriched in the dedifferentiated SC,

10/13 (76.9%) groups that were significantly enriched in the dedifferentiated cells (FDR <0.25) overlapped with functional groups found to be enriched in SC from the distal nerve (figure 28a). 3/12

(25%) of functional groups of genes enriched in the differentiated SC overlapped with the functional groups found to be enriched in SC from the intact nerve (figure 28b).

GSEA showed that, as seen with the in vivo RNAseq samples, there was significant overlap of genes down regulated the dedifferentiated SC (compared to differentiated SC), with the genes with reduced expression in the nerve of the hypomorphic krox 20 mouse (Le et al., 2005) (as compared to

WT nerve) (figure 27 and 28(b)). When looking at genes up regulated in the nerve of the hypomorphic krox20 mouse the overlap of genes was not as great; 18/87 (20%) genes were up regulated in the dedifferentiated SC and 7 (8%) of these also overlapped with genes up regulated in the distal nerve in our in vivo samples (figure 28 a).

118

When comparing the genes differentially expressed in the dedifferentiated SC compared to the differentiated SC 36/84 (43%) were also found to be differentially expressed in rat nerves following crush injury when compared to intact nerve (Bosse et al., 2006), of these 21 (25%) were also differentially expressed in the distal nerve 6 days following transection compared to intact nerve

(figure 29). However, Bosse et al. use the whole nerve rather than isolating SC, therefore the changes in gene expression reported in this paper may be due to changes in fibroblasts or inflammatory cells, therefore we would not expect those gene changes to overlap with the expression profile seen in our in vitro and in vivo data as these are SC specific.

The above suggests that the SC dedifferentiation assay we have developed shares some common features with the SC reprogramming following nerve injury in vivo. Due to the complex nature of the environment of the wounded distal nerve, involving SC interactions with fibroblasts, regenerating axons, vessels, inflammatory cells etc, it would be virtually impossible to completely recreate the SC reprogramming seen in vivo, in vitro. However the similarities seen between the differentiated SC in vitro and the intact nerve in vivo and those seen between the dedifferentiated SC in vitro and the SC from the distal nerve in vivo show that this dedifferentiation assay would be a useful first screening tool for assessing novel targets found in the analysis of the in vitro work.

119

5. Future Work As a next step it would be interesting to perform RNAseq at earlier time points after nerve transection. This would give us a temporal view of the reprogramming the SC undergoes following transection. Looking at one time point is merely taking a snapshot and may miss key factors which regulate the process initially or have are switched on later in the process. Figure 11 shows that by 24 hours there is already down regulation in the mRNA levels of some myelin proteins and up regulation of dedifferentiation markers, it therefore would be important to have a time point earlier than 24 hours to catch these early transcriptional changes. When we looked at the protein expression of p75NTR, we saw that expression began at the proximal end of the distal nerve at 36 hours and by 60 hours post sciatic nerve transection, the entire length of the nerve examined showed expression of this repair SC marker (figure 12). This suggests that the key events which regulate the transition of a mature myelinating SC to a dedifferentiated SC following nerve injury occur early in the process meaning that the majority of the time points which would provide useful information for a temporal analysis of SC reprogramming should focus on the first 72 hours following nerve injury. As events are likely to change as nerve repair progresses and the wound site becomes more mature it would be interesting and informative to also have a later time point, around 6 days to compare the early changes to.

Studies using microarray (Bosse et al., 2006) and deep sequencing (Yi et al., 2015) technology have looked at various early and late time points following nerve injury, but SC were not specifically analysed in either of these cases. Bosse et al reported that most genetic up or down regulation was transient and changed over time (Bosse et al., 2006). This reiterates the need for a SC specific, temporal overview of the process, with early and late time points.

120

Shue et al have reported that reprogramming of SC occurs in a wave like fashion down the distal nerve starting at the tip (Shue at al., 2000), our data (figure 12) supports this. Although by 6 days

(the time point we used in our study) it is likely all the SC in the distal nerve would have underdone reprogramming, as evidenced by protein expression of p75NTR seen down the entire length of the distal nerve. This may be something to consider in the future if earlier time points are analysed as sequencing SC from the entire length of the distal nerve may mask key genetic changes if only the SC at the tip of the nerve are undergoing reprogramming. A single cell sequencing approach at different areas of the distal nerve might be a way to overcome this.

Once SC specific RNA from several early and a late time point had been collected (e.g. RNA from the

SC of the distal end of nerves 18 hours, 36 hours, 54 hours, 72 hours and 6 days following nerve transection) and sequenced, we would use genome wide co-expression analysis to identify key hub genes. We would focus on novel transcription factors identified and screen a number of these using the in vitro dedifferentiation assay we have developed; inducing the SC in culture to over or under express the target genes identified and analyse this effected the differentiation response seen when

NRG and FCS is added to the cells . Over a longer time frame it may be possible to breed mouse lines with a conditional knock out of target genes specifically in the SC, and to asses nerve repair in these animals.

It may also be possible to look at expression of any genes identified in peripheral nerve sheath tumour samples from humans, following a tissue bank application. This translational aspect of the work would assess if the genes found to be key players in peripheral nerve repair may also provide useful therapeutic targets or prognostic indicators in tumours derived from SC.

121

References

Akassoglou, K., Akpinar, P., Murray, S., and Strickland, S. (2003) Fibrin is a regulator of Schwann cell migration after sciatic nerve injury in mice. Neurosci. Lett. 338 (3), 185-188. Apfel, S. C. (2002) Nerve growth factor for the treatment of diabetic neuropathy: what went wrong, what went right, and what does the future hold? Int Rev Neurobiol. 50, 393–413. Apfel, S. C., Kessler, J. A., Adornato, B. T., Litchy, W. J., Sanders, C., and Rask C. A. (1998) Recombinant human nerve growth factor in the treatment of diabetic polyneuropathy. NGF Study Group. Neurology. 51, 695–702. Arima, Y., Hayashi, H., Kamata, K., Goto, T. M., Sasaki, M., Kuramochi, A., and Saya, H. (2010) Decreased expression of neurofibromin contributes to epithelial-mesenchymal transition in neurofibromatosis type 1. Experimental Dermatology. 19 (8), 136-41. Arthur-Farraj, P.J. Latouche, M., Wilton, D.K., Quintes, S., Chabrol, E., Banerjee, A., Woodhoo, A., Jenkins, B., Rahman, M., Turmaine, M., Wicher, G.K., Mitter, R., Greensmith, L., Behrens, A., Raivich, G., Mirsky, R., and Jessen, K.R. (2012). c-Jun reprograms Schwann cells of injured nerves to generate a repair cell essential for regeneration. Neuron. 75, 633-47. Barrette, B., Calvo, E., Vallieres, N., and Lacroix, S. (2010). Transcriptional profiling of the injured sciatic nerve of mice carrying the Wld(S) mutant gene: identification of genes involved in neuroprotection, neuroinflammation, and nerve regeneration. Brain Behav Immun. 24 (8), 1254-67. Bauer, S., Kerr, B.J., and Patterson, P. H. (2007). The neuropoietic cytokine family in development, plasticity, disease and injury. Nat Rev Neurosci. 8, 221–32. Behse, F., Buchthal, F., and Carlsen, F. (1977). Nerve biopsy and conduction studies in diabetic neuropathy. J Neurol Neurosurg Psychiatry. 40, 1072–82. Bell, M. A., and Waddell, A. G. (1984). A morphometric study of intrafascicular vessels of mammalian sciatic nerve. Muscle Nerve. 7, 524-34. Berger, P., Sirkowski, E. E., Scherer, S. S., and Suter, U. (2004). Expression analysis of the N-Myc downstream-regulated gene 1 indicates that myelinating Schwann cells are the primary disease target in hereditary motor and sensory neuropathy-Lom. Neurobiol Dis. 17 (2), 290-9. Bermingham, J. R., Scherer, S. S., Oconnell, S., Arroyo, E., Kalla, K. A., Powell, F. L., and Rosenfeld, M. G. (1996). Tst-1/Oct-6/SCIP regulates a unique step in peripheral myelination and is required for normal respiration. Genes Dev. 10, 1751-62. Bisby, M. A., and Chen, S. (1990). Delayed Wallerian degeneration in sciatic nerves of C57Bl/Ola mice is associated with impaired regeneration of sensory axons. Brain Research. 530, 117–20. Blom, C. L., Martensson, L. B., and Dahlin, L. B. (2014). Nerve injury-induced c-Jun activation in Schwann cells is JNK independent. Biomed Res Int. 2014, 1-11. Bosse, F., Hasenpusch-Theil, K., Küry, P., and Müller, H. W. (2006). Gene expression profiling reveals that peripheral nerve regeneration is a consequence of both novel injury-dependent and reactivated developmental processes. J Neurochem. 96, 1441–57.

122

Bosse, F., Küry, P., and Müller, H. W. (2001). Gene expression profiling and molecular aspects in peripheral nerve regeneration .Restor. Neurol. Neurosci. 19, 5 – 18. Boyd JG & Gordon T (2003). Neurotrophic factors and their receptors in axonal regeneration and functional recovery after peripheral nerve injury. Mol Neurobiol 27, 277–324. Bremer, M., Frob, F., Kichko, T., Reeh, P., Tamm, E. R., Suter, U., and Wegner, M. (2011) Sox 10 is required for Schwann-cell homeostasis and myelin maintenance in the adult peripheral nerve. Glia. 59, 1022-32. Britsch, S., Goerich, D. E., Riethmacher, D., Peirano, R. I., Rossner, M., Nave, K., Birchmeier, C., and Wegner, M. (2001). The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15, 66-78. Campana, W.M., Li, X., Shubayev, V.I., Angert, M., Cai, K., and Myers, R. R. (2006). Erythropoietin reduces Schwann cell TNF-alpha, Wallerian degeneration and pain related behaviours after peripheral nerve injury. Eur. J. Neurosci. 23, 617-626. Campbell, W. W. (2008). Evaluation and management of peripheral nerve injury. Clin Neurophysiol. 199 (9), 1951-65. Cattin, A. L., Burden, J. J., Van Emmenis, L., Mackenzie, F. E., Hoving, J. J., Garcia Calavia, N., Guo, Y., McLaughlin, M., Rosenberg, L. H., Quereda, V., Jamecna, D., Napoli, I., Parrinello, S., Enver, T., Ruhrberg, C., and Lloyd, A. C. (2015). Macrophage-induced blood vessels guide schwann cell- mediated regeneration of peripheral nerves. Cell. 162, 1127–39. Chan, J. R., Cosgaya, J. M., Wu, Y. J., and Shooter, E. M. (2001). Neurotrophins are key mediators of the myelination program in the peripheral nervous system. 98 (5). 14661-8. Chan JR, Watkins TA, Cosgaya JM, Zhang C, Chen L, Reichardt LF, Shooter EM, Barres BA. (2004). NGF controls axonal receptivity to myelination by Schwann cells or oligodendrocytes. Neuron 43:183–91. Chen, Z.L., Yu, W.M., and Strickland, S. (2007). Peripheral Regeneration. Annu. Rev. Neurosci. 30, 209-33. Cheng, L., Esch, F. S., Marchionni, M. A., and Mudge, A. W. (1998) Control of Schwann cell survival and proliferation; autocrine factors and neuregulins. Mol. Cell. Neurosci. 12, 141–56. Clemence, A., Mirsky, R., and Jessen, K. R. (1989). Non-myelin-forming Schwann cells proliferate rapidly during Wallerian degeneration in the rat sciatic nerve. J Neurocytol. 18(2), 185–92. Cornblath, D. R., and Hoke, A. (2006) Recent advances in HIV neuropathy. Curr Opin Neurol. 19, 446– 50. Cosgaya, J. M., Chan, J. R., and Shooter E. M. (2002). The neurotrophin receptor p75NTR as a positive modulator of myelination. Science. 298, 833-41. Crossman, A.R., and Neary D. (2014) Neuroanatomy An Illustrated Colour Text. 5th Ed. Churchill Livingstone, London. D’Antonio, M., Michalovich, D., Paterson, M., Droggiti, A., Woodhoo, A., Mirsky, R., and Jessen, K. R. (2006) Gene profiling and bioinformatic analysis of Schwann cell embryonic development and myelination. Glia. 53, 501–15.

123

Davis, J. B., and Stroobant, P. (1990). Platelet-derived growth factors and fibroblast growth factors are mitogens for rat Schwann cells. J Cell Biol. 110(4), 1353–60. Deumens, R., Bozkurt, A., Meek, M. F., Marcus, M. A., Joosten, E. A., Weis, J., and Brook, G. A. (2010). Repairing injured peripheral nerves: Bridging the gap. Prog Neurobiol. 92, 245–76. Doddrell, R. D., Dun, X. P., Moate, R. M., Jessen, K. R., Mirsky, R., and Parkinson, D. B. (2012). Regulation of Schwann cell differentiation and proliferation by the Pax-3 transcription factor. Glia. 60, 1269-78. Dumitru, D., Zwarts, M. J., and Amato, A. A. (2001) Peripheral nervous system’s reaction to injury. In: Dumitru, D., Amato, A. A., Zwarts, M., editors. Electrodiagnostic medicine. 2nd ed. Philadelphia: Hanley and Belfus, 115–56. Einheber, S., Hannocks, M. J., Metz, C. N., Rifkin, D. B., and Salzer, J. L. (1995). TGF-β1 regulates axon-Schwann cell interactions. J. Cell Biol. 129, 443–58. Feldman, E. L. (2008). Diabetic neuropathy. Curr Drug Targets. 9, 1–2. Feltri, M.L., Graus Porta, D., Previtali, S. C., Nodari, A., Migliavacca, B., Cassetti, A., Littlewood-Evans, A., Reichardt, L. F., Messing, A., Quattrini, A., Mueller, U., and Wrabetz, L. (2002). Conditional disruption of β1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156, 199–209. Feltri, M. L., Scherer, S. S., Nemni, R., Kamholz, J., Vogelbacker, H., Scott, M. O., Canal, N., Quaranta, V., and Wrabetz, L. (1994) Beta 4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development. 120, 1287–301. Fontana, X., Hristova, M., Da Costa, C., Patodia, S., Thei, L., Makwana, M., Spencer-Dene, B., Latouche, M., Mirsky, R., Jessen, K. R., Klein, R., Raivich, G., and Behrens, A. (2012) c-Jun in Schwann cells promotes axonal regeneration and motoneuron survival via paracrine signaling. J Cell Biol. 198, 127–41. Geller, D. S., and Gebhardt, M. (2006). Malignant peripheral nerve sheath tumors (MPNSTs) Electronic Sarcoma Update Newsletter. 3 (3). Gilmour, D. T., Maischein, H.-M., and Nüsslein-Volhard, C. (2002). Migration and function of a glial subtype in the vertebrate peripheral nervous system. Neuron. 34, 577-88. Gomez-Sanchez, J. A., Carty, L., Iruarrizaga-Lejarreta, M., Palomo-Irigoyen, M., Varela-Rey, M., Griffith, M., Hantke, J., Macias-Camara, N., Azkargorta, M., Aurrekoetxea, I., De Juan, V. G., Jefferies, H. B., Aspichueta, P., Elortza, F., Aransay, A. M., Martínez-Chantar, M. L., Baas, F., Mato, J. M., Mirsky, R., Woodhoo, A., and Jessen, K. R. (2015). Schwann cell autophagy, myelinophagy, initiates myelin clearance from injured nerves. J Cell Biol. 210, 153–68. Griffin, M. F., Malahias, M., Hindocha, S., and Khan, W. S. (2014). Peripheral nerve injury: principles for repair and regeneration. Open Orthop J. 8. 199-203. Grim, M., Halata, Z., and Franz, T. (1992). Schwann cells are not required for guidance of motor nerves in the hindlimb in Splotch mutant mouse embryos. Anat Embryol (Berl). 186, 311–8. Grinsell, D., and Keating, C. P. (2014). Peripheral nerve reconstruction after injury: a review of clinical and experimental therapies. BioMed Research International. 2014. 1-13.

124

Hall, S. M. (1999). The biology of chronically denervated Schwann cells. In: Ann. N. Y. Acad. Sci. 883 , 215-33. Han, D., Chen, Y., Kou, Y., Weng, J., Chen, B., Yu, Y., Zhang, P., and Jiang, B. (2016). Profiling of the dynamically altered gene expression in peripheral nerve injury using NGS RNA sequencing technique. Am J Transl Res. 8 (2), 871-84. Hantke, J., Carty, L., Wagstaff, L. J., Turmaine, M., Wilton, D. K., Quintes, S., Koltzenburg, M., Baas, F., Mirsky, R., and Jessen, K. R. (2014). c-Jun activation in Schwann cells protects against loss of sensory axons in inherited neuropathy. Brain. 137:2922–37. Harrisingh, M.C., Perez—Nadales, E., Parkinson, D.B., Malcolm, D.S., Mudge, A.W., and Lloyd, A.C. (2004)The Ras/Raf/ERK signalling pathway drives Schwann cell dedifferentiation. EMBO J. 23, 3061- 71. Heermann, S., and Schwab, M. H. (2013). Molecular control of Schwann cell migration along peripheral axons: keep moving! Cell Adh Migr. 7, 18–22. Hirata, K., and Kawabuchi, M. (2002). Myelin phagocytosis by macrophages and nonmacrophages during Wallerian degeneration. Microsc Res Tech. 57, 541–7. Hirata, K., Masuda, K., Morikawa, W., He, J., Kuraoka, A., Kuwano, M., and Kawabuchi, M. (2004). N- myc downstream-regulated gene1 expression in injured sciatic nerves. Glia. 4, 325-34. Hnia, K., Vaccari, I., Bolino, A., and Laporte, J. (2012). Myotubularin phosphoinositide phosphatises: Cellular functions and disease pathophysiology. Trends Mol Med. 18, 317-27. Hoke, A. (2006). Mechanisms of Disease: what factors limit the success of peripheral nerve regeneration in humans? Nat. Clin. Pract. Neurol. 2, 448–54. Hung, G., Li, X., Faudoa, R., Xeu, Z., Kluwe, L., Rhim, J. S., Slattery, W., and Lim, D. (2002). Establishment and characterization of a schwannoma cell line from a patient with neurofibromatosis 2. Int. J. Oncol. 20, 475–82. Hutton, E. J., Carty, L., Laurá, M., Houlden, H., Lunn, M. P., Brandner, S., Mirsky, R., Jessen, K., and Reilly, M. M. (2011) c-Jun expression in human neuropathies: a pilot study. J Peripher Nerv Syst.16, 295–303. Irobi-Devolder, J. (2008). A molecular genetic update of inherited distal motor neuropathies. Verh K Acad Geneeskd Belg. 70 (1), 25-46. Ishii, A., Furusho, M., and Bansai, R. (2013). Sustained activation of ERK1/2 MAPK in oligodendrocytes and Schwann cells enhances myelin growth and stimulates oligodendrocyte progenitor expansion. J Neurosci. 33, 175-86. Jacob, C., Lotscher, P., Engler, S., Baggiolini, A., Tavares, S. V., Brugger, V., John, N., Buchmann- Moller, S., Snider, P. L., Conway, S. J., Yamaguchi, T., Matthias, P., Sommer, L., Mantei, N., and Suter, U. (2014). HDAC1 and HDAC2 control the specification of neural crest cells into peripheral glia. J Neurosci. 34 (17), 6112-22. Jagalur, N. B., Ghazvini, M., Mandemakers, W., Driegen, S., Maas, A., Jones, E. A., Jaegle, M., Grosveld, F., Svaren, J., and Meijer, D. (2011). Functional dissection of the Oct6 Schwann cell enhancer reveals an essential role for dimeric Sox10 binding. J Neurosci. 31, 8585–94.

125

Jessen, K. R., and Mirsky, R. (1997). Embryonic Schwann cell development: the biology of Schwann cell precursors and early Schwann cells. J Anat. 191, 501-5. Jessen, K.R., and Mirsky, R. (2005). The origin and development of glial cells in peripheral nerves. Nat. Rev. Neurosci. 6, 671-82. Jessen, K.R., and Mirsky, R. (2008). Negative regulation of myelination: relevance for development, injury and demyelinating disease. Glia. 56, 1552-65. Jopling, C., Boue, S., and Ipisua Belmonte, J.C. (2011). Dedifferentiation, transdifferentiation and reprogramming: three routes to regeneration. Nat. Rev. Mol. Cell Biol. 12, 79-89. Joseph, N. M., Mosher, J. T., Buchstaller, J., Snider, P., Mckeever, P. E., Lim, M., Conway, S. J., Parada, L. F., Zhu, Y., and Morrison, S. J. (2008). The loss of Nf1 transiently promotes self-renewal but not tumorigenesis by neural crest stem cells. Cancer Cell. 13, 129–40.

Jung, J., Cai, W., Lee, H. K., Pellegatta, M., Shin, Y. K., Jang, S. Y., Suh, D. J., Wrabetz, L., Feltri, M. L., and Park, H. T. (2011). Actin polymerization is essential for myelin sheath fragmentation during Wallerian degeneration. J Neurosci. 31, 2009–15. Jungnickel, J., Claus, P., Gransalke, K., Timmer, M., and Grothe, C. (2004) Targeted disruption of the FGF-2 gene affects the response to peripheral nerve injury. Mol. Cell. Neurosci. 25, 444-52. Jungnickel, J., Hasse, K., Konitzer, J., Timmer, M., and Grothe, C. (2006) Faster nerve regeneration after sciatic injury in mice over-expressing basic fibroblast growth factor. J. Neurobiol. 66, 940-8. Kanda, T. (2013) Biology of the blood-nerve barrier and its alteration in immune mediated neuropathies. J Neurol Neurosurg Psychiatry. 84 (2), 208-12. Kao, S. C., Wu, H., Xie, J., Chang, C. P., Ranish, J. A., Graef, I. A., and Crabtree, G. R. (2009). Calcineurin/NFAT signalling is required for neuregulin-regulated Schwann cell differentiation. Science. 323, 651-54. Kawasaki, T., Oka, N., Tachibana, H., Akiguchi, I., and Shibasaki H. (2003). Oct6, a transcription factor controlling myelination, is a marker for active nerve regeneration in peripheral neuropathies. Acta Neuropathol. 105, 203–8. Kelsh, R. N., and Eisen, J. S. (2000). The zebrafish colourless gene regulates development of non- ectomesenchymal neural crest derivatives. Development. 127, 515-25. Kouyoumdjian, J. A. (2006). Peripheral nerve injuries: a retrospective survey of 456 cases. Muscle Nerve. 34 (6), 785-8. Krasnoselsky, A., Massay, M. J., DeFances, M. C., Michalopoulos, G., Zarnegar, R., and Ratner, N. (1994). Hepatocyte growth factor is a mitogen for Schwann cells and is present in neurofibromas. 14 (12), 7284-90. Kubo, T., Yamashita, T., Yamaguchi, A., Hosokawa, K., and Tohyama, M. (2002). Analysis of genes induced in peripheral nerve after axotomy using cDNA microarrays. J Neurochem. 82 (5), 1129-36. Lallemend, F., and Ernfors, P. (2012). Molecular interactions underlying the specification of sensory neurons. Trends Neurosci. 35, 373–81.

126

Le, N., Nagarajan, R., Wang, J.Y.T., Araki, T., Schmidt, R.E., and Milbrandt, J. (2005). Analysis of congenital hypomyelinating EgrLo/Lo nerves identifies Sox2 as an inhibitor of Schwann cell differentiation and myelination. Proc Natl Acad. Sci. 102(7), 2596-601. Le, L. Q., and Parada, L. F. (2007). Tumor microenvironment and neurofibromatosis type I: connecting the GAPs. Oncogene. 26, 4609–16.

Le Douarin, N. M., and Smith, J. (1988). Development of the peripheral nervous system from the neural crest. Ann Rev Cell Biol. 4, 375-404. Lee, H. J., Shin, Y. K., and Park, H. T. (2014) Mitogen activated protein kinase family proteins and c- jun signaling in injury-induced Schwann cell plasticity. Exp Neurobiol. 23, 130–61. Lehmann, H.C., and Hoke, A. (2010). Schwann cells as a therapeutic target for peripheral neuropathies. CNS Neurol. Disord. Drug Targets. 9, 801-6. Lengner, C. J. (2010). iPS cell technology in regenerative medicine. Annals of the New York Academy of Sciences. 1192, 38-44. Li, Q., Ping, P., Jiang, H., and Liu, K. (2006). Nerve conduit filled with GDNF gene-modified Schwann cells enhances regeneration of the peripheral nerve. Microsurgery. 26, 116-21. Li, X., Gonias, S. L., and Campana, W.M. (2005). Schwann cells express erythropoietin receptor and represent a major target for Epo in peripheral nerve injury. Glia. 51, 254-65. Li, H., Zhao, X., Yan, X., Jessen, W. J., Kim, M. O., Dombi, E., Liu, P. P., Huang, G., and Wu, J. (2015). Runx1 contributes to neurofibromatosis type 1 neurofibroma formation. Oncogene. 35 (11), 1468- 74. Liu, Y., and Ma, Q. (2011). Generation of somatic sensory neuron diversity and implications on sensory coding.Curr Opin Neurobiol. 21, 52–60. Lively, T. N., Ferguson, H. A., Galasinski, S. K., Seto, A. G., and Goodrich, J. A. (2001). c-Jun binds the N terminus of human TAF(II)250 to derepress RNA polymerase II transcription in vitro. J Biol Chem. 276, 25582–8. Mathon, N. F., Malcolm, D. S., Harrisingh, M. C., Cheng, L., and Lloyd, A. C. (2001). Lack of replicative senescence in normal rodent glia. Science. 291, 872–5. May, F., Matiasek, K., Vroemen, M., Caspers, C., Mrva, T., Arndt, C., Schlenker, B., Gais, P., Brill, T., Buchner, A., Blesch, A., Hartung, R., Stief, C., Gansbacher, B., and Weidner, N. (2008). GDNF- transduced Schwann cell grafts enhance regeneration of erectile nerves. Eur Urol. 54 (5), 1179-87. McArthur, J. C., Brew, B. J., and Nath, A. (2005). Neurological complications of HIV infection. Lancet Neurol. 4, 543–55. McArthur, J. C., Yiannoutsos, C., Simpson, M. D., Adornato, B. T., Singer, E. J., Hollander, H., Marra, C., Rubin, M., Cohen, B. A., Tucker, T., Navia, B. A., Schifitto, G., Katzenstein, D., Rask, C., Zaborski, L., Smith, M. E., Shriver, S., Millar, L., Clifford, D. B., and the AIDS Clinical Trials Group Team 291. (2000). A phase II trial of nerve growth factor for sensory neuropathy associated with HIV infection. Neurology. 54, 1080-8.

127

Mertens, F., and Lothe, R. A. (2001). Nervous system: peripheral nerve sheath tumours. Atlas of Genetics and Cytogenetics in Oncology and Haematology. 5. 5094. Mogha, A., Benesh, A. E., Patra, C., Engel, F. B., Schoneberg, T., Liebscher, I., and Monk, K. R. (2013). GPR126 functions in Schwann cells to control differentiation and myelination via G-protein activation. J Neurosci. 33, 17976-85. Monje, P. V., Soto, J., Bacallao, K., and Wood, P. M. (2010). Schwann cell dedifferentiation is independent of mitogenic signaling and uncoupled to proliferation: role of cAMP and JNK in the maintenance of the differentiated state. J Biol Chem. 285, 31024–36. Monk, K. R., Feltri, M. L., and Taveggia, C. (2015). New insights on Schwann cell development. Glia. 63 (8), 1376-93. Monk, K. R., NaylorS. G., Glenn, T. D., Mercurio, S., Perlin, J. R., Dominguez, C., Moens, C. B., and Talbot, W.S. (2009). A G protein-coupled receptor is essential for Schwann cells to initiate myelination. Science. 325, 1402-5. Morgan, L., Jessen, K.R., and Mirsky, R. (1991). The effects of cAMP on differentiation of cultured Schwann cells: progression from an early phenotype (04+) to a myelin phenotype (P0+, GFAP-, N- CAM-, NGF-receptor-) depends on growth inhibition. The Journal of Cell Biology. 112, 457-67. Morrell, P., and Quarles, R. H. (1999). Myelin formation, structure, and biochemistry. In: Siegel, G. J., Agranoff, B. W., Albers, R. W., Fisher, S. K., Uhler, M. D., editors. Basic neurochemistry, 6th edition. Philadelphia: Lippincott-Raven. Chapter 4. Morris, J. K., Lin, W., Hauser, C., Marchuk, Y., Getman, D., and Lee, K-F. (1999). Genetic rescue of cardiac defect in erbB2 null mutant mice reveals essential roles of erbB2 in development of the peripheral nervous system. Neuron. 23, 273–83. Mosley, K., and Cuzner, M. L. (1996). Receptor-mediated phagocytosis of myelin by macrophages and microglia: effect of opsonization and receptor blocking agents. Neurochem Res. 21, 481–7. Mukouyama, Y. S., Gerber, H. P., Ferrara, N., Gu, C., and Anderson, D. J. (2005). Peripheral nerve- derived VEGF promotes arterial differentiation via neuropilin 1-mediated positive feedback. Development. 132, 941-52. Nagarajan, R., Le, N., Mahoney, H., Araki, T., and Milbrandt, J. (2002). Deciphering peripheral nerve myelination by using Schwann cell expression profiling. ProcNatlAcadSci. 99, 8998–9003. Nagarajan, R., Svaren, J., Le, N., Araki, T., Watson, M., and Milbrandt, J. (2001). EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron. 30 (2), 355-68. Napoli, I., Noon, L.A., Ribeiro, S., Kerai, A.P., Parrinello, S., Rosenberg, L.H., Collins, M.J., Harrisingh, M.C., White, A.J., Woodhoo, A., and Lloyd, A.C. (2012). A central role for the ERK-signalling pathway in controlling Schwann cell plasticity and peripheral nerve regeneration in vivo. Neuron. 73, 729-42. Newbern, J., and Birchmeier, C. (2010). Nrg1/ErbB signalling networks in Schwann cell development and myelination. Semin Cell Dev Biol. 21 (9), 922-8. Nguyen, Q. T., Sanes, J. R., and Lichtman, J. W. (2002). Pre-existing pathways promote precise projection patterns. Nat Neurosci. 5, 861–7.

128

Park, H. T., and Feltri L. M. (2011). Rac1 GTPase controls myelination and demyelination. Bioarchitecture. 1 (3), 110-3. Parkinson, D. B., Bhaskaran, A., Arthur-Farraj, P., Noon, L. A., Woodhoo, A., Lloyd, A. C., Feltri, M. L., Wrabetz, L., Behrens, A., Mirsky, R., and Jessen, K. R. (2008). c-Jun is a negative regulator of myelination. J Cell Biol. 181, 625–37. Parkinson, D. B., Bhaskaran, A., Droggiti, A., Dickinson, S., D'Antonio, M., Mirsky, R., and Jessen, K. R. (2004). Krox-20 inhibits Jun-NH2-terminal kinase/c-Jun to control Schwann cell proliferation and death. J Cell Biol. 164, 385–94. Parmantier, E., Lynn, B., Lawson, D., Turmaine, M., Namini, S. S., Chakrabarti, L., McMahon, A. P., Jessen, K. R., and Mirsky, R. (1999). Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths. Neuron. 23 (4), 713-24. Parrinello, S., and Lloyd A.C. (2009). Neurofibroma development in NF1 – insights into tumour initiation. Trends in Cell Biology. 19, 395-403. Parrinello, S., Napoli, I., Ribeiro, S., Wingfield Digby, P., Fedorova, M., Parkinson, D. B., Doddrell, R. D., Nakayama, M., Adams, R. H., and Lloyd, A. C. (2010). EphB signaling directs peripheral nerve regeneration through Sox2-dependent Schwann cell sorting. Cell. 143, 145–55. Patel, J. H., Loboda, A. P., Showe, M. K., Showe, L. C., and McMahon, S. B. (2004). Analysis of genomic targets reveals complex functions of MYC. Nat Rev Cancer. 4 (7), 562-8. Payne, S. H. (2001). Nerve repair and grafting in the upper extremity. J South Orthop. Assoc. 10 (2), 1-7. Perry, V. H., Tsao, J. W., Fearn, S., and Brown, M. C. (1995). Radiation-induced reductions in macrophage recruitment have only slight effects on myelin degeneration in sectioned peripheral nerves of mice. Eur J Neurosci. 7,271–80. Picelli, S., Faridani, O. R., Bjorklund, A. K., Winberg, G., Sagasser, S., and Sandberg, R. (2014). Full- length RNA-seq from single cells using Smart-seq2. Nature Protocols. 9(1), 171–81. Quarles, R. H., Macklin, W. B., and Morell, P. (2006). Myelin formation, structure, and biochemistry, in Basic Neurochemistry: Molecular, Cellular and Medical Aspects, 7th edn (Siegel, G. J., Albers, R. W., Brady, S. T., and Price, D., eds.) Academic Press Elsevier, New York. 51–71. Rogister, B., Delree, P., Leprince, P., Martin, D., Sadzot, C., Malgrange, B., Munaut, C., Rigo, J.M., Lefebvre, P.P., and Octave, J.N. (1993). Transforming growth factor beta as a neurological signal during peripheral nervous system response to injury. J. Neurosci. Res. 34(1), 32-43. Rosberg, H. E., Carlsson, K. S., Hojgard, S., Lindgren, B., Lundborg, G., and Dahlin, L. B. (2005). Injury to the human median and ulnar nerves in the forearm–analysis of costs for treatment and rehabilitation of 69 patients in southern Sweden. J Hand Surg Br. 30, 35–9. Sacktor, N. (2002). The epidemiology of human immunodeficiency virus-associated neurological disease in the era of highly active antiretroviral therapy. J. Neurovirol. 8 (2), 115-21. Sahenk, Z., Chen, L., and Mendell, J. R. (1999). Effects of PMP22 duplication and deletions on the axonal cytoskeleton. Ann Neurol. 45, 16–24.

129

Sahenk, Z., Nagaraja, H. N., McCracken, B. S., King, W. M., Freimer, M. L., Cedarbaum, J. M., and Mendell, J. R. (2005). NT-3 promotes nerve regeneration and sensory improvement in CMT1A mouse models and in patients.Neurology. 65, 681–9. Saito, H., Nakao, Y., Takayama, S., Toyama, Y., and Asou, H. (2005). Specific expression of an HNK-1 carbohydrate epitope and NCAM on femoral nerve Schwann cells in mice. Neurosci. Res. 53, 314– 22. Sano, Y., Shimizu, F., Nakayama, H., Abe, M., Maeda, T., Ohtsuki, S., Terasaki, T., Obinata, M., Ueda, M., Takahashi, R., and Kanda, T. (2007). Endothelial cells constituting blood-nerve barrier have highly specialized characteristics as barrier-forming cells. 32 (2), 139-47. Sanz, E., Yang, L. Su, T., Morris, D.R., Stanley McKnight, G., and Amieux, P.S. (2009). Cell-type-specific isolation of ribosome-associated mRNA from complex tissues. PNAS. 106 (33), 13939-44. Scherer, S. S. (1996). Molecular specializations at nodes and paranodes in peripheral nerve. Microsc Res Tech. 34, 452–61. Schifitto, G., Yiannoutsos, C., Simpson, D. M., Adornato, B. T., Singer, E. J., Hollander, H., Marra, C. M., Rubin, M., Cohen, B. A., Tucker, T., Koralnik, I. J., Katzenstein, D., Haidich, B., Smith, M. E., Shriver, S., Millar, L., Clifford, D. B., and McArthur, J. C. (2001). Long-term treatment with recombinant nerve growth factor for HIV-associated sensory neuropathy. Neurology. 57, 1313–6. Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, Gassmann M, Lightfoot S, Menzel W, Granzow M and Ragg T. (2006). The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Molecular Biology. 7 (3),

Seddon, H. J., Medawar, P. B., and Smith, H. (1943). Rate of regeneration of peripheral nerves in man. J Physiol. 102, 191–215. Shah, N. M., Marchionni, M. A., Isaacs, I., Stroobant, P., and Anderson, D. J. (1994). Glial growth factor restricts mammalian neural crest stem cells to a glial fate. 77 (3), 349-60. Shan, C., Wei, J., Hou, R., Wu, B., Yang, Z., Wang, L., Lei, D., and Yang, X. (2016). Schwann cells promote EMT and the Schwann-like differentiation of salivary adenoid cystic carcinoma cells via the BDNF/TrkB axis. Oncol Rep. 35 (1), 427-35. Sherman, D. L., and Brophy, P. J. (2005). Mechanisms of axon ensheathment and myelin growth. Nat Rev Neurosci. 6, 683–90.

Sheu, J. Y., Kulhanek, D. J. and Eckenstein, F. P. (2000). Differential patterns of ERK and STAT3 phosphorylation after sciatic nerve transection in the rat. Exp Neurol. 166, 392–402. Shimono, A., Okuda, T., and Kondoh, H. (1999). N-myc-dependent repression of ndr1, a gene identified by direct subtraction of whole mouse embryo cDNAs between wild type and N-myc mutant. Mech Dev. 83, 39-52. Shin, Y. K., Jang, S. Y., Park, J. Y., Park, S. Y., Lee, H. J., Suh, D. J., and Park, H. T. (2013). The Neuregulin-Rac-MKK7 pathway regulates antagonistic c-jun/Krox20 expression in Schwann cell dedifferentiation. Glia. 61, 892–904. Sigurgeirsson, B., Emanuelsson, O., and Lundeberg, J. (2014). Sequencing degraded RNA addressed by 3’ tag counting. PLoS ONE. 9 (3), 1-11.

130

Stoll, G., Jander, S., and Myers, R. R. (2002). Degeneration and regeneration of the peripheral nervous system: from Augustus Waller's observations to neuroinflammation. J Peripher Nerv Syst. 2002, 7: 13-27. Stoll, G., and Müller, H. W. (1999). Nerve injury, axonal degeneration and neural regeneration: basic insights. Brain Pathol. 9, 313–25. Stuhlmiller, T. J., and Garcia-Castro, M. I. (2012). Current perspectives of the signalling pathways directing neural crest induction. Cell Mol Life Sci. 69 (22), 3715-37. Sulaiman, O. A., and Gordon, T. (2009). Role of chronic Schwann cell denervation in poor functional recovery after nerve injuries and experimental strategies to combat it. Neurosurgery. 65(4), A105– A114. Sunderland, S. (1951). A classification of peripheral nerve injuries producing loss of function. Brain. 74, 491–516. Sunderland, S. (1978). Nerves and nerve injuries. 2nd ed. Baltimore: Williams and Wilkins. Sunderland, S. (1990). The anatomy and physiology of nerve injury. Muscle Nerve. 13 (9), 771-84. Suter, U., and Scherer, S. S. (2003). Disease mechanisms in inherited neuropathies. Nat Rev Neurosci. 4, 714-26. Svenningsen, A., and Dahlin, L. B. (2013). Repair of the peripheral nerve—remyelination that works. Brain Sci. 3, 1182–97. Syed, N., Reddy, K., Yang, D. P., Taveggia, C., Salzer, J. L., Maurel, P., and Kim, H. A. (2010). Soluble neuregulin-1 has bifunctional, concentration-dependent effects on Schwann cell myelination. J. Neurosci. 30, 6122–31. Taveggia, C., Zanazzi, G., Petrylak, A., Yano, H., Rosenbluth, J., Einheber, S., Xu, X., Esper, R. M., Loeb, J. A., Shrager, P., Chao, M. V., Falls, D. L., Role, L., and Salzer, J. L. (2005). Neuregulin-1 type III determines the ensheathment fate of axons. Neuron. 47 (5), 681-94. Taylor, C. A., Braza, D., Rice, J. B., and Dillingham, T. (2008). The incidence of peripheral nerve injury in extremity trauma. Am J Phys Med Rehabil. 87, 381–5. Thomson, J. G. (2005). Median and ulnar nerve injuries: A meta-analysis of predictors of motor and sensory recovery after modern microsurgical nerve repair: Discussion. Plastic and Reconstructive Surgery. 116 (2), 495–6. Topilko, P., Schneider-Maunoury, S., Levi, G., Baron-Van Evercooren, A., Chennoufi, A. B., Seitanidou, T., Babinet, C., Charnay, P. (1994). Krox-20 controls myelination in the peripheral nervous system. Nature. 371, 796 –9. Trapp, B. D., Hauer, P., and Lemke, G. (1988). Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J Neurosci. 8, 3515-21. 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. Phil Transact Royal Soc London.140, 423–9.

131

Wang, Z., Li, Y., Kong, D., and Sarkar, F. H. (2010). The role of notch signaling pathway in epithelial- mesechymal transition (EMT) during development and tumor aggressiveness. Curr Drug Targets. 11 (6), 745-51. Warner, L. E., Mancias, P., Butler, I. J., McDonald, C. M., Keppen, L., Koob, K. G., Lupski, J. R. (1998). Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat Genet. 18, 382–4. Webber, C. A., Christie, K. J., Cheng, C., Martinez, J. A., Singh, B., Singh, V., Thomas, D., and Zochodne, D. W. (2011). Schwann cells direct peripheral nerve regeneration through the Netrin-1 receptors, DCC and Unc5H2. 59 (1), 1503-17. Webster, H. D., Martin, R., and O’Connell. (1973). The relationships between interphase Schwann cells and axons before myelination: a quantitative electron microscopic study. Dev. Biol. 32, 401-16. White, F. V., Toews, A. D., Goodrum, J. F., Novicki, D. L., Bouldin, T. W., and Morel, P. (1989). Lipid metabolism during early stages of Wallerian degeneration in the rat sciatic nerve. J Neurochem. 52, 1085-92. Woodhoo, A., Alonso, M. B., Droggiti, A., Turmaine, M., D’Antonio, M., Parkinson, D. B., Wilton D. K., Al-Shawi, R., Simons, P., and Shen, J. (2009). Notch controls embryonic Schwann cell differentiation, postnatal myelination and adult plasticity. Nat Neurosci. 12, 839-47. Woodhoo, A., Dean, C. H., Droggiti, A., Mirsky, R., and Jessen K. R. (2004). The trunk neural crest and its early glial derivatives: a study of survival responses, developmental schedules and autocrine mechanisms. 25 (1), 30-41. Woodhoo, A., and Sommer, L. (2008). Development of the Schwann cell lineage: from the neural crest to the myelinated nerve. Glia. 56 (14), 1481-90. Wu, J., Williams, J. P., Rizvi, T. A., Kordich, J. J., Witte, D., Meijer, D., Stemmer-Rachamimov, A. O., Cancelas, J. A., and Ratner, N. (2008). Plexiform and dermal neurofibromas and pigmentation are caused by Nf1 loss in desert hedgehog-expressing cells. Cancer Cell. 13, 105–16.

Yang, D. P., Kim, J., Syed, N., Tung, Y. J., Bhaskaran, A., Mindos, T., Mirsky, R., Jessen, K. R., Maurel, P., Parkinson, D. B., and Kim, H. A. (2012). p38 MAPK activation promotes denervated Schwann cell phenotype and functions as a negative regulator of Schwann cell differentiation and myelination. J Neurosci. 32, 7158–68. Yang, D.P., Zhang, D.P., Mak, K.S., Bonder, D.E., Pomeroy, S.L., and Kim, H.A. (2008). Schwann cell proliferation during Wallerian degeneration is not necessary for regeneration and remyelination of the peripheral nerves: axon-dependent removal of newly generated Schwann cells by apoptosis. Mol. Cell. Neurosci. 38, 80–8. Yi, S., Zhang, H., Gong, L., Wu, J., Zha, G., Zhou, S., Gu, X., and Yu, B. (2015). Deep sequencing and bioinformatic analysis of lesioned sciatic nerves after crush injury. PLoS ONE. 10 (12), 1-18. Zanazzi, G., Einheber, S., Westreich, R., Hannocks, M. J., Bedell-Hogan, D., Marchionni, M. A., and Salzer, J. L. Glial growth factor/neuregulin inhibits Schwann cell myelination and induces demyelination. J Cell Biol. 152, 1289–99.

132

Zheng, H., Chang, L., Patel, N., Yang, J., Lowe, L., Burns, D. K., and Zhu, Y. (2008). Induction of abnormal proliferation by nonmyelinating schwann cells triggers neurofibroma formation Cancer Cell. 13, 117–28.

Zochodne, D. W. (2007). Diabetes mellitus and the peripheral nervous system: manifestations and mechanisms. Muscle Nerve. 36 (2), 144-66. Zrouri, H., Le Goascogne, C., Li, W. W., Pierre, M., and Courtin, F. (2004). The role of MAP kinases in rapid gene induction after lesioning of the rat sciatic nerve. Eur J Neurosci. 20, 1811–8. Zwilling, S., Annweiler, A., and Wirth, T. (1994). The POU domains of Oct1 and Oct2 transcription factors mediate specific interaction with TBP. Nucleic Acids Res. 22, 1655–62.

133

Appendix A

RIN values for the groups of nerves used for RNAseq. Each group included 3 pooled nerves.

Nerve RIN

Intact 1 7.7

Intact 2 7.1

Intact 3 7.5

Distal 1 7.5

Distal 2 7.5

Distal 3 8.1

134

Appendix B

Genes which are differentially regulated in the hypomorphic krox 20 nerve when compared to nerves from WT mice (Le et al., 2005) which are also differentiailly regulated in our in vivo (SC from the distal nerve 6 days following transection when compared to SC from intact nerves) and in vitro

(NRG/FCS dedifferentiated SC when compared to cAMP treated differentiated SC) RNAseq data.

Differentially regulated in Differentially regulated in Differentially regulated in Hypomorphic Krox 20 and in Hypomorphic Krox 20 and in Hypomorphic Krox 20, in vivo vivo data vitro data data and in vitro data

Tenascin C UHRF1 UHRF1

Cyclin B2 RNA binding motif 3 RNA binding motif 3

Cyclin D1 High mobility group AT hook 2 High mobility group AT hook 2

Neuron navigator2 Ribonucleotide reductase M2 Ribonucleotide reductase M2

CDC 28 protein kinase regulatory subunit 2 Phosphodiesterase 8A Phosphodiesterase 8A

Collagen type 18, alpha 1 Chemokine ligand 2 Chemokine ligand 2

Myristoylated alanine rich protein Pleckstrin homology domain Pleckstrin homology domain kinase C substrate containing family O member 1 containing family O member 1

UHRF1 CD44 CD44

RNA binding motif protein 3 Kinesin family member 22 Kinesin family member 22

Coactosin-like F actin binding protein 1 SOX2 CD9

Protein phosphatiase regulatory 5’nucleotidase domain containing 2 subunit 14 Carbohydrate sulphotransferase 2

Cell devision cycle 20 Histone deacytalase 2 ELOVL6

Haematological and neurological TCF19 expressed 1 EPH receptor B6

High mobility group AT hook 2 Coronin 1C Fatty acid desaturase 1

Brain abundant membrane attached Farnesyl-diphosphate Protein regulator of cytokines 1 signal protein 1 farnesyltransferase 1

Topoisomerase 2 alpha Activating transcription factor 3 FGF1

Myristoylated alanine rich protein Kinesin family member 11 kinase C substrate GJB1

CKS1B Enhancer of zeste homolog 2 3-Hydroxyl-3-methylglutaryl CoA

135

Isopentenyl diphosphate delta MAD2L1 ETS1 isomerise 1

Ribosomal Protein L13A Cell adhesion molecule L1 IL16

Ribonucleotide reductase M2 CD55 LIM and calponin homology domains 1

Ki-67 CD9 Lysyl oxidase

Polo like kinase 1 Carbohydrate sulfotransferase 2 MBP

Microsomal glutathione S transferase Phosphodiesterase 8A Clathrin light chain B 3

Tumour necrosis factor receptor superfamily member 21 CMTM6 MPZ

Sperm antigen with coiled domains 1 EF hand domain family member D1 Methylsterol methoxygenase 1

Cyclin A2 ELOVL6 Mevolonate decarboxylase

Nuclear receptor subfamily 4 group A Retinoblastoma like 1 EMI domain containing 1 member 2

CXC chemokine receptor 4 EPH receptor B6 Nuclear receptor binding protein 2

Structural maintenance of chromosome 22 Fatty acid desaturase 1 PMP22

Farnesyl diphosphate EPH receptor A5 farnesyltransferase 1 Periaxin

Sodium voltage gated channel alpha Basic leucine zipper and W2 domains 2 FGF1 subunit 7

INCENP Ferritin heavy chain 1 Semaphorin 3B

Chemokine ligand 2 Gap junction B1 Sirtuin 2

Solute carrier organic anion H2A histone family member Z HMGCR transporter family member 3A1

3-hydroxy-3-methylglutaryl CoA Cyclin B1 reductase Synuclein alpha

Isopentenyl diphosphate delta TCDD inducible poly (ADP-ribose) Endothelin receptor type B isomerise 1 polymerase

Aurora kinase A IL16 Transmembrane protein 30A

High mobility group protein B2 Inositol triphosphate 3 kinase B Transmembrane protein 40

Tripartite motif containing 59 LIM and calponin homology domains 1 UDP glycosyltransferase 8

Nuclear and spindle associated protein 1 Lysyl oxidase

Non-SMC condensin complex subunit MBP

136

Pleckstrin homology domain containing family O member 1 Malic enzyme 1

Microsomal glutathione S transferase CD44 3

Karyopherin alpha 2 MPZ

Kinesin family member 22 Methylsterol methoxygenase 1

Baculoviral IAP repeat containing 5 Mevolonate decarboxylase

ADAM10 N acytyl transferase 2

Nuclear receptor subfamily 4 group A Adenylate kinase 3 member 2

CD9 Nuclear receptor binding protein 2

Phosphoinositide 3 kinase regulatory Cyclin dependent kinase inhibitor 1A subunit 1

Ceramide synthase 2 PMP22

Carbohydrate sulfotransferase 2 Periaxin

Claudin 5 Prostoglandin D2 synthase

24 dehydrocholesterol reductase Ring finger protein 13

Sodium voltage gated channel alpha ELOVL6 subunit 7

EPH receptor B6 Semaphorin 3B

Fatty acid amide hydrolase Sirtuin 2

Solute carrier organic anion Fatty acid desaturase 1 transporter family member 3A1

Farnesyl diphosphate farnesyltransferase 1 Synuclein alpha

Farnesyl pyrophosphate synthetase TGF alpha

TCDD inducible poly (ADP-ribose) FGF1 polymerase

Fibrinogen like 2 Transketolase

Fucosyltransferase 8 Transmembrane protein 30A

Gap junction protein B1 Transmembrane protein 40

137

Glyoxalase domain containing 4 UDP glycosyltransferase 8

Guanine nucleotide binding protein alpha inhibiting activity polypeptide 1

3-hydroxy-3-methylglutaryl CoA reductase

Isopentenyl diphosphate delta isomerase 1

Interleukin 16

LIM and calponin homology domains 1

Lysyl oxidase

Lanosterol synthase

MBP

Microsomal glutathione S transferase 3

Multiple PDZ domain protein

MPZ

Methylsterol monooxygenase 1

Metallothionein 3

Mevalonate diphosphate

Myosin heavy chain 11

N-myc downstream regulated 1

Nuclear receptor subfamily 4 group A member 2

Nuclear receptor binding protein 2

NAD (P) dependent steroid dehydrogenase like

Osteoglycin

Oxysterol binding protein like 5

Pericentriolar material 1

Phosphate cytidylytransferase 2

138

PMP22

Periaxin

Sodium voltage gated channel alpha subunit 7

Semaphorin 3B

Serpin family D member 1

Sirtuin 2

Solute carrier family 25 member 1

SLC25A1

SLC6A15

SLCO3A1

Synuclein alpha

Synuclein gamma

Thyroid hormone responsive

TCDD inducible poly (ADP-ribose) polymerase

Transmembrane protein 30A

Transmembrane protein 40

UDP glycosyltransferase 8

139

Appendix C

Genes which are differentially regulated in crushed rat nerves when compared to intact nerves

(Bosse et al., 2006) which are also differentiailly regulated in our in vivo (SC from the distal nerve 6 days following transection when compared to SC from intact nerves) and in vitro (NRG/FCS dedifferentiated SC when compared to cAMP treated differentiated SC) RNAseq data.

Differentially regulated in Differentially regulated in Differentially regulated in crushed rat nerves and in vivo crushed rat nerves and in vitro crushed rat nerves, in vivo data data data and in vitro data

CD63 CD63 CD63

CD44 CD44 CD44

CD9 CD9 CD9

Integrin beta4 Integrin beta4 Integrin beta4

CXC chemokine receptor 4 Platelet derived growth factor FGF 5 receptor alpha

Cyclin dependent kinase 4 FGF 5 GADD45

FGF 5 GADD45 P75NTR

Growth arrest and DNA damage Cyclin D3 Epithelial membrane protein 1 inducible cyclin D1 P75NTR Synaptotagmin IV

P75NTR Epithelial membrane protein 1 ATP1A2

Epithelial membrane protein 1 Microtubule associated protein IB ATP1B2

ILGFBP3 Interferon regulatory factor 1 Cytochrome p450 51

Prothymosin alpha TGF beta 1 Potassium voltage gated channel shaker 1

Cytochrome c oxidase subunit IV YWHAZ MBP

Synaptotagmin IV Karyopherin beta P0

ApoD Cytochrome c oxidase polypeptide PLP Vb

140

ATPase Na+/K+ transporting Synaptotagmin IV Nestin subunit alpha 2

ATPase Na+/K+ transporting Arachidonate 12-lipoxygenase Periaxin subunit beta 2

Cytochrome p450 51 ATPase Na+/K+ transporting PMP 22 subunit alpha 1

Fatty acid binding protein 5 ATPase Na+/K+ transporting plasmolipin subunit alpha 2

Potassium voltage gated ATPase Na+/K+ transporting Long chain acyl-CoA synthatase channel shaker 1 subunit beta 2 2

MBP Cytochrome p450 51

P0 Potassium voltage gated channel shaker 1

PLP sortilin

Nestin ADP ribosylation factor 6

Periaxin MBP

PMP 22 P0 plasmolipin PLP

Carbonic anhydrase III Nestin

Long chain acyl-CoA synthatase Periaxin 2

Carboxypeptidase Z PMP 22

Ribosomal protein L5 plasmolipin

Ribosomal protein L10a Plasminogen activator 1

Ribosomal protein L18 Lipoprotein lipase

Ribosomal protein S24

141

Appendix D

Genes which are differentially regulated in both the hypomorphic krox 20 nerves when compared to

WT nerves (Le et al., 2005) and crushed rat nerves when compared to intact nerves (Bosse et al.,

2006) which are also differentiailly regulated in our in vivo (SC from the distal nerve 6 days following transection when compared to SC from intact nerves) and in vitro (NRG/FCS dedifferentiated SC when compared to cAMP treated differentiated SC) RNAseq data.

Differentially regulated in Differentially regulated in Differentially regulated in Hypomorphic Krox 20, crushed Hypomorphic Krox 20, crushed Hypomorphic Krox 20, crushed rat nerves and in vivo data rat nerves and in vitro data rat nerves in vivo data and in vitro data

CD44 CD44 CD44

CD9 CD9 CD9

CXC chemokine receptor 4 MBP MBP

Cyclin D1 P0 P0

MBP Periaxin Periaxin

P0 PMP 22 PMP 22

Periaxin

PMP 22

142