You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Skin-derived precursors are a suitable alternative to peripheral nerve as a source of Schwann cells for… Sparling, Joseph Samuel 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2014_spring_sparling_joseph.pdf [ 6.01MB ]
Metadata
JSON: 24-1.0165873.json
JSON-LD: 24-1.0165873-ld.json
RDF/XML (Pretty): 24-1.0165873-rdf.xml
RDF/JSON: 24-1.0165873-rdf.json
Turtle: 24-1.0165873-turtle.txt
N-Triples: 24-1.0165873-rdf-ntriples.txt
Original Record: 24-1.0165873-source.json
Full Text
24-1.0165873-fulltext.txt
Citation
24-1.0165873.ris

Full Text

SKIN-DERIVED PRECURSORS ARE A SUITABLE ALTERNATIVE TO PERIPHERAL NERVE AS A SOURCE OF SCHWANN CELLS FOR TRANSPLANTATION-BASED REPAIR OF THE INJURED RAT SPINAL CORD    by Joseph Samuel Sparling  M.A., Queen's University, 2005 B.A. (Honours), Brock University, 2002 B.Sc. (Honours), McMaster University, 2000  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Neuroscience)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  February 2014  ? Joseph Samuel Sparling, 2014 ii  Abstract For much of human history the devastating loss of neurological functions that occurs after spinal cord injury (SCI) was thought to be irreversible, so the people afflicted with such injuries were offered no hope of effective medical treatment. Today that has changed, as advances in neurobiology and medicine over the past century have led to the development of treatments aimed specifically at repairing the injured spinal cord. The transplantation of Schwann cells (SCs) has emerged as one promising example of such a treatment, with demonstrated efficacy in multiple animal models of SCI and encouraging preliminary results in clinical trials. Although SCs possess many of the qualities of an ideal cellular therapy, the harvest of autologous SCs from peripheral nerve (N-SCs) causes permanent nerve injury, which could be avoided by generating SCs from an alternative autologous source. One such source is skin-derived precursors (SKPs), which can be isolated from the adult mammalian dermis and differentiated into SCs (SKP-SCs) in vitro. Herein I examined the efficacy of SKP-SCs as a treatment for SCI in rodent injury models and compared those cells to their nerve-derived counterparts. This work provided the first demonstration of efficacy for SKP-SC therapy after thoracic contusion and showed that, much like N-SCs, SKP-SCs myelinate, promote axonal growth, and enhance functional recovery after SCI. In addition, we found evidence that SKP-SCs may have advantages over N-SCs with respect to their ability to interact favourably with spared astrocyte-rich host tissue and promote axonal growth. Subsequently we directly compared neonatal SKP-SCs and N-SCs and found that those cell types were highly similar in terms of their protein/gene expression profiles, migration and integration into astrocyte-rich domains in vitro and in vivo, and many reparative effects following transplantation into the partially crushed cervical spinal cord. Taken together our iii  findings suggest that SKP-SCs and N-SCs have similar therapeutic efficacy, and that where differences between those two cell types exist, they consistently favour the SKP-SCs as the more favourable cell type for SCI repair. Thus, our work to-date supports the notion that SKP-SCs are a suitable alternative to N-SCs for transplantation-based central nervous system repair.  iv  Preface A version of Chapter 2 has been published1. Biernaskie J, Sparling JS, Liu J, Shannon CP, Plemel JR, Xie Y, Miller FD, Tetzlaff W (2007). Skin-derived precursors generate myelinating Schwann cells that promote remyelination and functional recovery after contusion spinal cord injury. J Neurosci 27(36):9545-9559.  Jeff Biernaskie and I shared co-first authorship on that publication as we contributed equally to the work. The original idea for this study came from Freda Miller and Wolfram Tetzlaff, who also supervised the in vitro and in vivo aspects of this collaborative study, respectively. I assisted with experimental design and I was chiefly responsible for planning and organizing many aspects of the experiment, including: all surgeries, animal care, and behavioural testing (including 5 measures), and some of the in vivo histological assessments. I also assisted with contusion surgeries, conducted or supervised all aspects of animal care and behavioural testing, and cut, immunostained, and/or imaged a substantial portion of the tissue used for histological quantifications and figure images. In addition I conducted the statistical analyses for all of the behavioural data and a portion of the histological data, wrote the sections of the methods and results pertaining to those analyses, made about half of the figures in the paper, and provided feedback on drafts of the manuscript prior to publication.  Jeff Biernaskie conducted all of the cell culture work, delivered and resuspended the cells for transplantation, conducted/supervised a significant portion of the histological analysis, and wrote the bulk of the final paper in collaboration with Wolfram Tetzlaff. Jie Liu performed all of the animal surgeries. Jason Plemel and Casey Shannon assisted me with certain behavioural assessments and with perfusions at the end of the study. Robert Xie provided technical assistance on some of the histology. This work was conducted with the approval of the University of British v  Columbia Animal Care Committee. The relevant animal care certificate number is A03-0139 ?Anatomical and functional recovery after spinal cord contusion injury?. A version of Chapter 3 has been prepared for submission to be published2. Sparling JS, Bretzner F, Biernaskie J, Assinck P, Jiang Y, Arisato H, Plunet WT, Borisoff J, Liu J, Miller FD, Tetzlaff W. Neonatal Schwann cells generated from skin-derived precursors or peripheral nerve induce similar levels of functional recovery after transplantation into the partially injured cervical spinal cord of the rat. The in vitro and in vivo portions of that work were again supervised by Freda Miller and Wolfram Tetzlaff, respectively. Frederic Bretzner planned, organized and supervised the surgeries, behavioural analyses and histological assessments conducted in experiment 1, and also conducted the electrophysiological assessment in both experiments, quantified all of the rubrospinal tract axon branching data, and wrote an initial paper based on the results of experiment 1. I planned and organized all aspects of experiment 2. In that experiment I assisted with the culture work and resuspended the cells for transplantation, conducted or supervised all aspects of animal care, behavioural testing, and histological assessments, except the rubrospinal tract branching analysis, which was conducted by Frederic Bretzner on images that I took from tissue that I immunostained. I also resampled all of the histological data from experiment 1 (except the RST axon branching data), ran all final data analyses, created all of the figures and wrote the final draft of the manuscript combining the work from both experiments. Jeff Biernaskie isolated and expanded all of the cells used in this study and resuspended the cells for transplantation in the first study. Peggy Assinck and Yuan Jiang assisted with animal care, behavioural testing and analysis, and/or immunohistochemistry in experiment 2. Hiroki Arisato assisted with animal care and behavioural testing and analysis in experiment 1. Ward vi  Plunet provided conceptual input and technical assistance on forelimb behavioural measures in experiment 1. Jamie Borisoff provided technical assistance on the electrophysiological measures and Jie Liu conducted animal surgeries in both experiments. This work was conducted with the approval of the University of British Columbia Animal Care Committee. The relevant animal care certificate numbers include: A03-0112 ?Regeneration of chronically injured rubrospinal neurons? and A06-1529 ?Anatomical and functional recovery after spinal cord contusion injury?. A version of Chapter 4 is in preparation for publication3. Sparling JS, Plemel JR, Biernaskie J, Miller FD, Tetzlaff W. Schwann cells generated from neonatal rodent skin-derived precursors are functionally indistinguishable from species- and age-matched nerve-derived Schwann cells. I designed all of the experiments included in Chapter 4, conducted all of the Schwann cell culture work and in vitro assessments, conducted or supervised all of the in vivo measures, conducted all of the data analysis, made all of the figures and wrote the manuscript. Jason Plemel kindly provided the astrocytes used for the in vitro analysis and provided useful input regarding cell culture issues in general. Jeff Biernaskie trained me to culture Schwann cells from nerve, provided helpful input regarding Schwann cell culture in general, and cultured the Schwann cells used in the in vivo work, the latter of which was supervised by Freda Miller. Wolfram Tetzlaff supervised all of the in vitro and in vivo experiments. This work was conducted with the approval of the University of British Columbia Animal Care Committee. The relevant animal care certificate numbers include: A08-0200 ?Myelin inhibition of myelination? and A06-1529 ?Anatomical and functional recovery after spinal cord contusion injury?.  vii  Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents ........................................................................................................................ vii List of Figures ............................................................................................................................. xvi List of Abbreviations ............................................................................................................... xviii Acknowledgements ................................................................................................................. xxvii Dedication ................................................................................................................................. xxix Chapter 1: General Introduction .................................................................................................1  Overview of introduction ................................................................................................ 2 1.1 Spinal cord injury ............................................................................................................ 7 1.2 Causes of SCI .............................................................................................................. 7 1.2.1 Relating spinal cord anatomy to functional losses due to SCI.................................... 8 1.2.2 The outcome of SCI in humans ................................................................................ 11 1.2.3 The global cost of SCI .............................................................................................. 15 1.2.4 A brief history of the treatment of SCI in humans.................................................... 17 1.2.5 The current state of treatment for SCI in humans ..................................................... 21 1.2.6 The pathophysiology of SCI ......................................................................................... 24 1.3 Primary injury ........................................................................................................... 24 1.3.1 Secondary injury ....................................................................................................... 27 1.3.21.3.2.1 Secondary injury mechanisms triggered by vascular changes after SCI .......... 27 1.3.2.2 Secondary damage mediated by inflammatory cells following SCI ................. 33 1.3.2.3 Cell death and degeneration following SCI ...................................................... 35 viii   The response of neural cells to SCI .......................................................................... 38 1.3.31.3.3.1 Spinal cord stem/progenitor cells ...................................................................... 38 1.3.3.2 Glial scar formation and astrocyte contributions to secondary injury .............. 39 1.3.3.3 The oligodendrocyte response to SCI: Demyelination vs. Remyelination ....... 42 1.3.3.4 Neuronal and axonal responses to SCI: The failure of CNS regeneration ....... 48  Long-term pathophysiological outcomes of SCI ...................................................... 51 1.3.4 Therapeutic strategies for promoting improved neurological outcome after SCI ........ 54 1.4 Neuroprotective strategies for SCI............................................................................ 54 1.4.1 Neurorepair strategies for SCI .................................................................................. 56 1.4.2 The need for combinatorial treatments ..................................................................... 60 1.4.3 Cellular therapies for SCI ............................................................................................. 62 1.5 Schwann cells................................................................................................................ 71 1.6 The development of SCs in spinal nerves ................................................................. 72 1.6.11.6.1.1 The migratory neural crest: Origin of SCs in spinal nerves.............................. 73 1.6.1.2 The generation of Schwann cell precursors ...................................................... 74 1.6.1.3 The generation of immature SCs ...................................................................... 75 1.6.1.4 The generation of mature SCs........................................................................... 76  SC dedifferentiation .................................................................................................. 80 1.6.2 Sensory versus motor SC phenotypes ....................................................................... 82 1.6.3 Axon-derived signals control SC development, differentiation and dedifferentiation .  1.6.4 ................................................................................................................................... 83  SC functions during development and in adulthood in the intact PNS..................... 86 1.6.5 The role of SCs in spontaneous repair and recovery after PNS injury ..................... 88 1.6.6ix   Therapeutic application of SCs following PNS injury ............................................. 94 1.6.7 Endogenous SCs in the CNS..................................................................................... 97 1.6.81.6.8.1 The source of endogenous SCs in the injured CNS .......................................... 97 1.6.8.2 The role in endogenous SCs in spontaneous CNS repair ................................. 99  Therapeutic applications of SCs in CNS injury and disease................................... 101 1.6.91.6.9.1 Peripheral nerve grafts applied to the CNS ..................................................... 103 1.6.9.2 N-SC transplantation into the intact and demyelinated CNS.......................... 105 1.6.9.3 Bridges containing N-SCs applied to SCI transection models ....................... 107 1.6.9.4 Intraparenchymal transplantation of N-SCs following SCI ............................ 108 1.6.9.5 Summary of the benefits and limitations associated with N-SC transplantation following SCI .................................................................................................................. 111  Overcoming the limitations of N-SCs to enhance CNS repair ........................... 112 1.6.101.6.10.1 Improving the survival of N-SCs in the injured spinal cord ....................... 113 1.6.10.2 Improving N-SC interactions with astrocytes ............................................. 114 1.6.10.3 Co-treatments to enhance the growth of CNS axons in general ................. 117 1.6.10.4 Promoting regeneration across the distal graft-host interface ..................... 118 1.6.10.5 Maximizing the therapeutic efficacy of N-SC therapy for SCI .................. 119  Clinical trials of N-SC transplantation following SCI ........................................ 122 1.6.11 Drawbacks of autologous N-SC application ....................................................... 123 1.6.12 Alternative sources of autologous SCs ....................................................................... 130 1.7 Do tMSCs represent an alternative source of SCs for therapeutic applications? ... 132 1.7.1 SKPs as an alternative source of SCs for therapeutic applications ............................. 134 1.8 The isolation and characterization of SKPs ............................................................ 134 1.8.1x   SKPs generate bona fide SCs under appropriate conditions in vitro and in vivo ... 137 1.8.2 Overview of experiments and hypotheses .................................................................. 140 1.9Chapter 2: Skin-derived Precursors Generate Myelinating Schwann Cells that Promote Remyelination and Functional Recovery after Contusion Spinal Cord Injury ...................143  Introduction ................................................................................................................. 144 2.1 Materials and methods ................................................................................................ 145 2.2 Animals ................................................................................................................... 145 2.2.1 Preparation of SKPs and SKP-SCs ......................................................................... 146 2.2.2 Flow cytometry ....................................................................................................... 147 2.2.3 Spinal cord contusion injury and cell transplantation ............................................. 148 2.2.4 Tissue processing and immunocytochemistry ........................................................ 150 2.2.5 Quantification of cavity and transplant volume, tissue sparing, cell survival, 2.2.6myelination, and axon numbers .......................................................................................... 152  Behavioural assessment .......................................................................................... 154 2.2.72.2.7.1 Open field test ................................................................................................. 154 2.2.7.2 Horizontal ladder test ...................................................................................... 155 2.2.7.3 Mechanical sensitivity test .............................................................................. 156 2.2.7.4 Thermal sensitivity test ................................................................................... 156  Statistical analyses .................................................................................................. 157 2.2.8 Results ......................................................................................................................... 157 2.3 Isolation, expansion and characterization of SKPs and SKP-derived SCs for 2.3.1transplantation ..................................................................................................................... 157 xi   Transplanted SKPs and SKP-derived SCs survive in the contusion lesion cavity and 2.3.2SKP-derived SCs promote sparing of the host spinal cord tissue ....................................... 158  Transplanted SKP-derived SCs modify the extracellular environment after spinal 2.3.3cord lesion ........................................................................................................................... 164  SKP-derived SCs promote axonal growth and sprouting into the transplant region ....  2.3.4 ................................................................................................................................. 170  Na?ve SKPs and SKP-derived SCs myelinate axons in the injured spinal cord ..... 180 2.3.5 SKPs and SKP-derived SCs promote recruitment of endogenous, myelinating SCs 2.3.6into the injured spinal cord ................................................................................................. 186  SKP-derived SCs promote functional improvement after contusion injury ........... 189 2.3.7 Discussion ................................................................................................................... 195 2.4Chapter 3: Neonatal Schwann Cells Generated from Skin-derived Precursors or Peripheral Nerve Induce Similar Levels of Functional Recovery after Transplantation into the Partially Injured Cervical Spinal Cord of the Rat .................................................................200  Introduction ................................................................................................................. 201 3.1 Materials and methods ................................................................................................ 203 3.2 Animals and experimental design ........................................................................... 204 3.2.1 Cell culture for transplantation ............................................................................... 205 3.2.23.2.2.1 Isolation of SKPs ............................................................................................ 205 3.2.2.2 Differentiation of SKP-SCs ............................................................................ 205 3.2.2.3 Isolation of nerve-derived Schwann cells ....................................................... 206 3.2.2.4 Isolation of dermal fibroblasts ........................................................................ 206  Spinal cord injury and cell transplantation ............................................................. 207 3.2.3xii  3.2.3.1 Dorsolateral funiculus crush ........................................................................... 207 3.2.3.2 Cell transplantation ......................................................................................... 207  Behavioural testing ................................................................................................. 208 3.2.43.2.4.1 Cylinder test .................................................................................................... 208 3.2.4.2 CatWalk .......................................................................................................... 209  Electrophysiology ................................................................................................... 209 3.2.5 Anterograde tracing of rubrospinal axons............................................................... 210 3.2.6 Tissue processing and immunohistochemistry ....................................................... 211 3.2.7 Image analysis and histological quantifications ..................................................... 213 3.2.83.2.8.1 Image acquisition and processing ................................................................... 213 3.2.8.2 Area and volume quantifications .................................................................... 213 3.2.8.3 Immunoreactivity quantifications ................................................................... 214 3.2.8.4 Gray and white matter RST axon quantifications ........................................... 215 3.2.8.5 Spared rim width ............................................................................................. 217  Statistical analysis and data presentation ................................................................ 217 3.2.9 Results ......................................................................................................................... 218 3.3 SC transplantation improves functional recovery after cervical SCI...................... 220 3.3.13.3.1.1 Cylinder test results......................................................................................... 220 3.3.1.2 CatWalk analysis results ................................................................................. 222  SC transplantation improves rubrospinal efficacy after cervical SCI ..................... 223 3.3.23.3.2.1 Motor threshold results ................................................................................... 223 3.3.2.2 EMG latency results ........................................................................................ 226 xiii   Cavitation persists after acute transplantation of SCs, but not fibroblasts, in the 3.3.3partially injured cervical spinal cord................................................................................... 226  Transplanted neonatal SCs survive and primarily occupy the spared tissue rim and 3.3.4mid-lesion bridges in the crushed DLF ............................................................................... 230  Transplanted and endogenous SCs myelinate host axons after cervical SCI ......... 231 3.3.5 SKP-SC transplantation does not increase reactive astrogliosis relative to media 3.3.6treatment after cervical SCI ................................................................................................ 233  SC transplantation is associated with reduced RST atrophy and/or enhanced RST 3.3.7sparing/plasticity ................................................................................................................. 237  SC transplants preserve the spared tissue rim and mid-lesion bridges, providing a 3.3.8substrate for spared and sprouting axons ............................................................................ 245  Discussion ................................................................................................................... 250 3.4Chapter 4: Schwann Cells Generated from Neonatal Rodent Skin-derived Precursors are Functionally Indistinguishable from Species- and Age-Matched Nerve-derived Schwann Cells .............................................................................................................................................257  Introduction ................................................................................................................. 258 4.1 Materials and methods ................................................................................................ 260 4.2 Cell culture for in vitro assays ................................................................................ 261 4.2.14.2.1.1 Isolation of SKPs ............................................................................................ 261 4.2.1.2 Differentiation of SKP-SCs ............................................................................ 261 4.2.1.3 Isolation of nerve-derived Schwann cells ....................................................... 262 4.2.1.4 SC purification and expansion for assays ....................................................... 262 4.2.1.5 Isolation and expansion of astrocytes ............................................................. 263 xiv   Cell culture for in vivo assay ................................................................................... 264 4.2.2 In vitro protein and gene expression analyses ........................................................ 264 4.2.34.2.3.1 Immunocytochemistry cultures ....................................................................... 264 4.2.3.2 Western blot procedures ................................................................................. 265 4.2.3.3 qPCR procedures ............................................................................................ 267  Inverted coverslip migration assay ......................................................................... 269 4.2.4 SC-astrocyte boundary assay .................................................................................. 270 4.2.5 Spinal cord injury, cell transplantation, and tissue processing ............................... 271 4.2.6 Immunocytochemistry and immunohistochemistry ................................................ 272 4.2.7 Image analysis, processing, and quantifications ..................................................... 274 4.2.84.2.8.1 Cell counts for SC purity and maturity analyses ............................................ 274 4.2.8.2 In vitro SC migration analysis ........................................................................ 274 4.2.8.3 In vitro SC-astrocyte integration analysis ....................................................... 275 4.2.8.4 In vitro GFAP-immunoreactivity .................................................................... 276 4.2.8.5 In vivo SC-astrocyte integration and SC migration analyses .......................... 276  Statistical analysis and data presentation ................................................................ 277 4.2.9 Results ......................................................................................................................... 278 4.3 Cultured SKP-SCs and N-SCs have highly similar phenotypes ............................. 278 4.3.1 SKP-SCs and N-SCs both show poor migration over astrocytes in vitro ............... 284 4.3.2 SKP-SCs and N-SCs exhibit similar boundary formation with astrocytes in vitro ......  4.3.3 ................................................................................................................................. 286  SKP-SCs and N-SCs display similar integration and migration potentials following 4.3.4transplantation into the injured rodent spinal cord ............................................................. 289 xv   Discussion ................................................................................................................... 292 4.4Chapter 5: General Discussion .................................................................................................298  Opening statement ...................................................................................................... 299 5.1 Summary of thesis....................................................................................................... 299 5.2 Summary of comparisons between neonatal SKP-SCs and adult/neonatal N-SCs .... 308 5.3 Do neonatal N-SCs have advantages over adult N-SCs?............................................ 310 5.4 Comparing adult SKP-SCs to adult N-SCs ................................................................. 312 5.5 Characterizing the SKP-SC phenotype ....................................................................... 314 5.6 Do the SKP-SCs represent a ?less mature? SC phenotype? .................................... 314 5.6.1 Do the SKP-SCs possess a phenotype that is more comparable to sensory or motor 5.6.2SCs? ................................................................................................................................. 318  Do SKP-SCs have advantages over N-SCs in terms of their suitability for CNS repair?  5.7 ..................................................................................................................................... 320  Are SKP-SCs a suitable alternative to N-SCs for therapeutic application in SCI? .... 322 5.8 Towards the clinical translation of SKP-SCs as a treatment for SCI ......................... 324 5.9 Additional directions for future research .................................................................... 327 5.10 Concluding thoughts ................................................................................................... 329 5.11Bibliography ...............................................................................................................................331  xvi  List of Figures Figure 2.1  Characterization of SKP-derived SCs before transplantation .................................. 160 Figure 2.2  Cavity size and cell survival in the contused rat spinal cord after transplantation of SKPs versus SKP-derived SCs ................................................................................................... 162 Figure 2.3  Na?ve SKPs are multipotent in vivo, and generate inappropriate cell types within the injured spinal cord....................................................................................................................... 165 Figure 2.4  Transplanted SKPs and SKP-derived SCs modify the extracellular environment surrounding the lesion ................................................................................................................. 168 Figure 2.5  SKP-derived SCs support robust axonal growth into the transplant ........................ 171 Figure 2.6  SKP-SCs promote axonal growth and sprouting into the transplant region ............. 172 Figure 2.7  Transplanted SKP-derived SCs enhance growth/sprouting of ascending sensory fibres..................................................................................................................................................... 175 Figure 2.8  SKP-SCs promote robust axonal growth in the injured spinal cord ......................... 178 Figure 2.9  Na?ve SKPs and SKP-derived SCs myelinate axons in the injured spinal cord. ...... 181 Figure 2.10  Transplanted, non-myelinating SKP-derived SCs express high levels of p75NTR ..... ..................................................................................................................................................... 185 Figure 2.11  SKPs and SKP-derived SCs promote recruitment of endogenous SCs into the injured spinal cord....................................................................................................................... 187 Figure 2.12  SKP-derived SCs improve locomotor function after a contusion injury of the spinal cord. ............................................................................................................................................ 190 Figure 2.13  SKP-derived SCs do not reduce sensory thresholds when transplanted into the contused spinal cord .................................................................................................................... 193 Figure 3.1  Experimental Design ................................................................................................ 219 xvii  Figure 3.2  Schwann cell transplants improve functional recovery following incomplete cervical spinal cord injury ........................................................................................................................ 221 Figure 3.3  Schwann cell transplants increase rubrospinal efficacy after spinal cord injury ...... 224 Figure 3.4  Cellular responses and lesion properties 11 weeks after injury/transplantation....... 228 Figure 3.5  SKP-SC transplantation does not increase reactive astrogliosis .............................. 234 Figure 3.6  SKP-SC transplantation increases rubrospinal axon density in the gray matter following incomplete cervical crush injury ................................................................................ 238 Figure 3.7  Compared to fibroblasts, transplanted Schwann cells enhance the amount of substrate available for axons to traverse the lesion from rostral to caudal ................................................ 247 Figure 4.1  SKP-SCs and N-SCs both label with known SC markers and show similar purity and maturity when grown under identical culture conditions ........................................................... 279 Figure 4.2  SKP-SCs and N-SCs show similar levels of expression for a variety of proteins and genes related to SC development in vitro ................................................................................... 282 Figure 4.3  SKP-SCs and N-SCs both show poor migratory potential on astrocytes in vitro .... 285 Figure 4.4  SKP-SCs and N-SCs show similar interactions with astrocytes in vitro in SC-astrocyte boundary assays ........................................................................................................... 287 Figure 4.5  SKP-SCs and N-SCs show similar integration and migration potential in the injured cervical spinal cord ..................................................................................................................... 290  xviii  List of Abbreviations 5-HT  ? 5-hydroxytryptamine  AMCA ? 7-amino-4-methylcoumarin-3-acetic acid AMPA  ? ?-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid  ANOVA ? analysis of variance AP2?  ? activating enhancer binding protein 2 alpha ASC(s) ? adipose stem cell(s) ASIA  ? American Spinal Injury Association ATP  ? adenosine triphosphate  BACE1 ? beta-site APP-cleaving enzyme 1 BBB  ? Basso, Beattie, and Bresnahan locomotor scale BCA  ? bicinchoninic acid BCE  ? before the common/current/Christian era  BDA  ? biotinylated dextran amine BDNF  ? brain-derived neurotrophic factor ?IIItub  ? beta III tubulin bFGF  ? basic fibroblast growth factor (aka: FGF-2) BMP2  ? bone morphogenetic protein 2 BMSC(s) ? bone marrow stromal cell(s) Brn2  ? brain 2 BSA  ? bovine serum albumin BSCB  ? blood-spinal cord barrier  Cad-19 ? cadherin-19 xix  cAMP  ? cyclic adenosine monophosphate  Caspr  ? contactin-associated protein CCL2  ? chemokine (C-C motif) ligand 2 CD34  ? cluster of differentiation 34 Cdc2  ? cell division control protein 2 homologue Cdc42  ? cell division control protein 42 cDNA  ? complementary deoxyribonucleic acid CE  ? common/current/Christian era CGRP  ? calcitonin gene-related peptide ChatABC ? chondroitinase ABC CNPase ? 2?, 3?- cyclic nucleotide 3?-phosphodiesterase  CNS  ? central nervous system CNTF  ? cilliary neurotrophic factor CPG(s) ? central pattern generator(s) CsA  ? cyclosporine A CSPG(s) ? chondroitin sulfate proteoglycan(s) CST  ? corticospinal tract  CT  ? computerized tomography  CTB  ? Cholera toxin B CT dye ? CellTracker dye CXCL10 ? chemokine (C-X-C motif) ligand 10 Cy3  ? cyanine 3 D?H  ? dopamine beta-hydroxylase xx  DCC  ? deleted in colorectal cancer DHH  ? desert hedgehog DLF  ? dorsolateral funiculus  DMEM ? Dulbecco?s modified Eagle?s medium DMSO  ? dimethyl sulfoxide DNA  ? deoxyribonucleic acid  DPBS  ? Dulbecco?s phosphate buffered saline DRG  ? dorsal root ganglion EAE  ? experimental autoimmune encephalomyelitis  ECL  ? enhanced chemiluminescence ECM  ? extracellular matrix EDTA  ? ethylenediaminetetraacetic acid EGF  ? epidermal growth factor Egr2  ? early growth response gene 2 (aka: Krox-20) EM  ? electron microscopy  EMG  ? electromyography EPI-NCSC(s) ? epidermal neural crest stem cell(s) ErbB2/3 ? erythroblastic leukemia viral oncogene homolog-2/3 ES cell(s) ? embryonic stem cell(s) EYFP  ? enhanced yellow fluorescent protein FBS  ? fetal bovine serum FDA  ? Food and Drug Administration  FGF  ? fibroblast growth factor xxi  Fibro(s) ? dermal fibroblast(s) FITC  ? fluorescein isothiocyanate GalC  ? galactocerebroside GAP-43 ? growth-associated protein-4 GAPDH ? glyceraldehyde-3-phosphate dehydrogenase GDNF  ? glial-derived neurotrophic factor GFAP  ? glial fibrillary acidic protein  GFP  ? green fluorescent protein GGF  ? glial growth factor GM  ? gray matter GRP(s) ? glial-restricted precursor(s) GTPases ? guanosine triphosphate hydrolase enzymes HBSS  ? Hank?s balanced salt solution HCl  ? hydrogen chloride HDAC(s) ? histone deacetylase(s) hfPS cell(s) ? hair follicle pluripotent stem cell(s) HNK-1 ? human natural killer-1 Ho  ? Hoechst nuclear stain HSC(s) ? hematopoietic stem cell(s) HSPG(s) ? heparin sulfate proteoglycan(s) Id2/4  ? inhibitor of DNA binding 2/4 IGF  ? insulin-like growth factor IL  ? interleukin (e.g., IL-1: interleukin-1) xxii  iNOS  ? inducible nitric oxide synthase  iNPC(s) ? induced neural precursor cell(s) iPS cell(s) ? induced pluripotent stem cell(s) IR  ? immunoreactivity (e.g., GFAP-IR) LIF  ? leukemia inhibitory factor MAG  ? myelin associated glycoprotein MBP  ? myelin basic protein MCP-1 ? monocyte chemoattractant protein-1 MEK  ? MAPK/ERK kinase  MHC  ? major histocompatibility complex miRNA(s) ? micro ribonucleic acid(s) MMP2  ? matrix metalloprotease 2 mpz  ? myelin protein zero gene MRI  ? magnetic resonance imaging  mRNA  ? messenger RNA MS  ? multiple sclerosis MSC(s) ? mesenchymal stem cell(s) NCad  ? neural cadherin NCAM ? neural cell adhesion molecule NCSC(s) ? neural crest stem cell(s) NCX  ? sodium-calcium exchange protein  NF  ? neurofilament-200 Nf1  ? neurofibromatosis type 1 gene xxiii  NFATc4 ? nuclear factor of activated T cells, cytoplasmic, calcineurin-dependent-4 NF-?B  ? nuclear factor kappa-light-chain enhancer of activated B cells  NFM  ? neurofilament M NG2  ? neuron-glia antigen 2 NGF  ? nerve growth factor NgR  ? Nogo receptor  NMDA ? N-Methyl-D-aspartate  NO  ? nitric oxide  Nogo-A ? neurite outgrowth inhibitor A NRG1  ? neuregulin 1 NRP(s) ? neural-restricted precursor(s) NSCISC ? National Spinal Cord Injury Statistical Center  N-SC(s) ? nerve-derived Schwann cell(s) NSC(s) ? neural stem cell(s) NSPC(s) ? neural stem/progenitor cells  NT  ? neurotrophin (e.g., NT-3: neurotrophin-3) O1  ? oligodendrocyte marker 1 O4  ? oligodendrocyte marker 4 Oct6  ? octamer-binding transcription factor 6 OEC(s) ? olfactory ensheathing cell(s) OEG  ? olfactory ensheathing glia OMgp  ? oligodendrocyte-myelin glycoprotein  OPC(s) ? oligodendrocyte precursor cell(s) xxiv  P0  ? myelin protein zero P2X7  ? Purinergic receptor P2X, ligand-gated ion channel, 7 P2Y12  ? purinergic receptor P2Y, G-protein coupled, 12 p75  ? low-affinity nerve growth factor receptor (aka: p75NTR) PAN/PVC ? polyacrylonitrile / polyvinylchloride Pax3  ? transcription factor paired box 3 PBS  ? phosphate buffered saline PDGF  ? platelet-derived growth factor PDGFR? ? platelet-derived growth factor receptor alpha PDL  ? poly-D-lysine PFA  ? paraformaldehyde PI3K  ? phosphatidylinositide 3-kinase PLC?  ? phospholipase C, gamma PLP  ? proteolipid protein  PMP-22 ? peripheral myelin protein-22 PN graft(s) ? peripheral nerve graft(s) PNS  ? peripheral nervous system  PPAR  ? peroxisome proliferator-activated receptor PrPc  ? protease-resistant protein c PSA-NCAM ? polysialylated neural cell adhesion molecule PTEN  ? phosphatase and tensin homologue  PVDF  ? polyvinylidene difluoride qPCR  ? quantitative real-time polymerase chain reaction xxv  RGM  ? repulsive guidance molecule  RM-ANOVA ? repeated-measures analysis of variance RNA  ? ribonucleic acid ROS  ? reactive oxygen species  RST  ? rubrospinal tract  S100?  ? S100 calcium-binding protein ? SC(s)  ? Schwann cell(s) SCI  ? spinal cord injury SCP(s)  ? Schwann cell precursor(s) SD  ? Sprague Dawley SDS-PAGE ? sodium dodecyl sulphate-polyacrylamide gel electrophoresis SEM  ? standard error of the mean SERT  ? serotonin transporter SKP(s)  ? skin-derived precursor(s) SKP-SC(s) ? SKP-derived Schwann cell(s) Sox2  ? sex determining region Y-related high mobility group-box-2 Sox10  ? sex determining region Y-related high mobility group-box-10 SREBPs  ? sterol regulatory element-binding proteins  Sur1  ? sulfonylurea receptor 1 SVZ  ? subventricular zone TACE  ? tumour necrosis factor alpha-converting enzyme TBST  ? Tris-buffered saline containing 0.1% Tween-20 TGF-?1 ? transforming growth factor beta 1 xxvi  TH  ? tyrosine hydroxylase TLR(s) ? Toll-like receptor(s) tMSC(s) ? transdifferentiated mesenchymal stem cell(s) TNF?  ? tumour necrosis factor alpha  TRPM7 ? transient receptor potential cation channel, subfamily M, member 7 Unc5H2 ? uncoordinated 5H2 US  ? United States VEGF ? vascular endothelial growth factor  WM  ? white matter wpi  ? weeks post-injury WT  ? wildtype YFP  ? yellow fluorescent protein Yy1  ? Yin yang 1 xxvii  Acknowledgements First and foremost I would like to thank my supervisor Wolfram Tetzlaff for giving me the opportunity to do my PhD in his laboratory and for his patience and continued support, both financial and otherwise, throughout my rather lengthy degree. I am also indebted to my supervisory committee members, Drs.  Jane Roskams, Fabio Rossi, and Tim O?Connor for their time, patience, and guidance throughout the course of my degree, and I would like to thank my previous/comprehensive committee members, Drs. Janice Eng, Steven Vincent, and Matt Ramer for their valuable input and feedback earlier in my PhD training. Thank you all for making me a better scientist and for encouraging me to expand my technical skills.  I am indebted to our collaborators in the laboratory of Dr. Freda Miller (Toronto). The work herein would not have been possible without the contributions and support of Drs. Miller and Jeff Biernaskie, and I am truly thankful for the culture training provided by both Dr. Biernaskie and Shaalee Dworski during my PhD.  Thank you to the funding agencies that supported me personally, or the projects I worked on, during my graduate training at UBC, including: the Canadian Institutes of Health Research, the Michael Smith Foundation for Health Research, the Stem Cell Network, ICORD, and the faculties of Neuroscience, Medicine, and Graduate Studies (FoGS) at UBC. A special thank you as well, to Rebecca Trainor at FoGS. Your personal support, guidance, and encouragement were instrumental in the completion of this work Rebecca, and I simply cannot thank you enough for pushing me back on track when I wandered off into the woods.  Throughout my degree I was fortunate to be surrounded by highly skilled, intelligent, creative, and supportive people in the Tetzlaff lab. Thank you to Dr. Jason Plemel and Peggy Assinck, whose friendship and support provided a silver lining to even the toughest days of my xxviii  PhD. Your words of encouragement and editorial assistance helped make the completion of this work possible Jason. Peggy, thank you for reminding me that helping others is its own reward. I am indebted to you both for helping to keep me sane/motivated throughout my PhD, and I will remember our times together fondly in the future ? particularly whenever I smell paraformaldehyde and/or Scotch. I am also very grateful to those who assisted with my training and provided technical assistance on my projects in the Tetzlaff lab, including Clarrie Lam, Jie Liu, Ward Plunet, Loren Oschipok, Casey Shannon, Catherine Hall, Nicole Boeder, Fernando Lucero Villegas, Petra Schreiner, Greg Duncan, Brett Hilton, and Darren Sutherland, and I would like to particularly thank Yuan Jiang, for his amazing technical support and the frequent provision of much needed entertainment/distraction. Thank you as well to the graduate trainees, faculty, and administrative personnel at ICORD who have provided support and encouragement throughout my degree, in particular: Drs. Lowell McPhail, Leanne Ramer, Tom Oxland, Angela Scott, Andrew Gaudet, Brian Kwon, and John Steeves, as well as Tracey Chang, Jeremy Green, Ben Nguyen, Cheryl Niamath, Lisa Anderson, Mark Crawford, Diana Hunter, Jacquelyn Cragg, Jessica Inskip, and Tim Bhatnagar.  To my family I offer my sincere thanks for a lifetime of love, constant encouragement, and unwavering support. Mom, I cannot thank you enough for the years of unconditional love and support that made this accomplishment possible, and I will forever be indebted to you and Sam for giving me a home during the many months it took me to finish this work. Marina, Brianna, Bella, Miguel, Jesus, Anita, Aunt Bonnie, Joan, and Chachi, thank you all for believing in me and reminding me to smile. And finally, thank you to my wife, Monica Rachel Sparling, for her continued patience, love, encouragement, and support throughout my PhD. You shared the weight of this work with me Mon, and for that I will always be immeasurably grateful. xxix  Dedication  To my beautiful wife, Monica Rachel Sparling. I?m sorry this took so long.       1  Chapter 1:                                                                                                      General Introduction 2    Overview of introduction 1.1Spontaneous repair processes generally fail following injury to the adult mammalian central nervous system (CNS), resulting in permanent functional deficits, the variety and severity of which depend on the location and extent of neurologic damage. In the context of spinal cord injury (SCI), this commonly translates into life-long deficits in motor, sensory, and/or autonomic functions below the level of injury, as well as a high frequency of related secondary complications (e.g., urinary tract infections, pressure ulcers, etc.); the combination of which is often devastating, not only to a person?s physical well-being, but also to his/her quality of life and life expectancy (Dijkers, 2005; Dietz and Curt, 2006; Strauss et al., 2006; Grossman et al., 2012; NSCISC, 2012). Although considerable advances have been made in the surgical and medical management of SCI and related secondary complications (Knoeller and Seifried, 2000; Lifshutz and Colohan, 2004; Donovan, 2007; Parent et al., 2011), at present, there exists no therapeutic intervention proven effective in restoring the loss of neurologic functions following SCI (Harrop et al., 2012; Hug and Weidner, 2012; Illis, 2012; Lammertse, 2013). Towards that goal, a wide variety of promising therapeutics are currently being investigated in pre-clinical animal studies (see reviews by: Kwon et al., 2011a; Kwon et al., 2011b; Tetzlaff et al., 2011) as well as human clinical trials (see reviews by: Gensel et al., 2011; Harrop et al., 2012; Wilcox et al., 2012; Lammertse, 2013). One promising potential therapy for SCI treatments involves the transplantation of Schwann cells (SCs) into the injured spinal cord. SC transplantation to treat SCI is particularly attractive because those cells have demonstrated functional efficacy in a wide variety of animal models of SCI and they can be generated from autologous tissue sources in the adult; thereby 3  obviating the need to use potentially harmful immunosuppressive drugs to prevent the rejection of transplanted cells following clinical application (reviewed in: Bunge and Pearse, 2003; Oudega et al., 2005; Oudega and Xu, 2006; Tetzlaff et al., 2011). Indeed, the potential of these cells to improve recovery from SCI has prompted multiple clinical trials to examine the safety and potential benefits of SC transplantation in humans with SCI (Tator, 2006; Saberi et al., 2008; Saberi et al., 2011; Talan, 2012; Xian-Hu et al., 2012). There remains however, one major drawback to the clinical application of autologous SC therapies, as these cells are typically generated via the excision of sural nerve in humans (nerve-derived SCs; N-SCs) and this procedure not only requires an additional surgery, but also results in a permanent peripheral sensory deficit and carries the risk of painful neuroma formation (Wolford and Stevao, 2003; Hood et al., 2009). Although the loss of sensation in the periphery may be well worth any gain in central functions, particularly given that the negative effects of peripheral nervous system (PNS) injury may go unnoticed in patients with existing SCI-related sensory deficits, it is obviously not ideal to injure the PNS in order to treat the CNS, particularly given that one aim of SC transplantation is to ameliorate the loss of sensation caused by SCI. This situation would clearly be improved if SCs could be generated from autologous tissue sources other than peripheral nerve, and indeed this can be done, as alternative sources of SCs have now been found in adult rodent and human tissues. Arguably the most promising alternative source of SCs that has been identified to date are the skin-derived precursors (SKPs); a resident stem cell found in adult mammalian skin (Biernaskie et al., 2009). Under appropriate conditions, isolated SKPs can be differentiated into SCs (SKP-SCs) in culture, and preliminary work with those cells demonstrated that they display 4  SC-like morphology, label with many of the antibodies typically associated with N-SCs, and show characteristic SC myelination of both PNS and CNS axons (Biernaskie et al., 2006; McKenzie et al., 2006). In light of those characteristics, we hypothesized that SKP-SCs would be a suitable alternative to N-SCs for therapeutic transplantation following SCI. To test that hypothesis, we began by assessing the benefits of SKP-SC transplantation following thoracic contusion in rats (Chapter 2); a model that has been used repeatedly to demonstrate the efficacy of N-SC transplantation after SCI (Tetzlaff et al., 2011). The results of that study demonstrated that SKP-SCs provide many of the same benefits typically associated with N-SC transplantation after SCI, including a modest, but significant improvement in hindlimb motor function. In addition, we noted that the SKP-SCs appeared to have advantages over N-SCs transplanted under similar conditions in terms of the degree of reactivity they elicit from nearby astrocytes and their abilities to migrate/integrate into astrocyte-rich host tissue and support the growth of certain axon populations, particularly across the distal graft-host interface. In light of the fact that cultured SKP-SCs are newly generated SCs, and in recognition of evidence that less mature SCs (i.e., Schwann cell precursors harvested from embryonic nerve) display similar advantages over N-SCs harvested from postnatal nerve (Woodhoo et al., 2007), we hypothesized that SKP-SCs may simply represent a population of less mature SCs compared to the N-SCs harvested from peripheral nerve. As such, and in light of our previous findings, we also hypothesized that SKP-SCs have advantages over N-SCs in terms of their interactions with astrocytes and their ability to support axonal growth, and we predicted that those advantages would endow the SKP-SCs with greater efficacy as a treatment for SCI. 5  To test those hypotheses, we conducted two concurrent studies comparing SKP-SCs and N-SCs harvested from neonatal rat tissue side-by-side both in vitro and in vivo. The first of those studies examined the SKP-SCs as a treatment for incomplete cervical SCI and directly compared the behaviour and reparative efficacy of SKP-SCs and N-SCs transplanted into the injured cervical spinal cord in an effort to determine whether either cell type had an advantage in terms of their efficacy as a treatment for SCI. The second study examined the interactions of SCs from either source with astrocytes in vitro and astrocyte-rich host tissue in the injured spinal cord to more precisely delineate the advantages of SKP-SCs over N-SCs, and assessed the expression of a variety of proteins and genes related to SC development in an effort to determine whether the SKP-SCs truly represent a less mature SC phenotype than their nerve-derived counterparts.   Consistent with our hypotheses, the SKP-SCs proved to be efficacious as a treatment for incomplete cervical SCI. However, contrary to our hypotheses, we found few significant differences between SKP-SCs and N-SCs, either in vitro or following transplantation into the injured spinal cord, which indicated that SKP-SCs do not have an advantage over N-SCs with respect to their suitability or efficacy in CNS repair when those cell types are both generated from neonatal tissue sources. Furthermore, SKP-SCs and N-SCs were not found to differ in terms of their expression of any of the characteristic SC marker proteins/genes that we examined, and our examination indicated that cells from either source were relatively immature and shared highly similar phenotypes in terms of their interactions with astrocytes overall.  Thus, our findings suggest that SKP-SCs may possess a phenotype that is more similar to that of N-SCs generated from neonatal, as opposed to adult, peripheral nerve. Our findings also support the notion that SKP-SCs are a suitable alternative to N-SCs for transplantation-based 6  repair of the injured CNS, and add to a growing body of evidence suggesting that SKP-SCs may have therapeutically relevant advantages over their counterparts generated from adult nerve. Although more work comparing SKP-SCs and N-SCs from adult rodent and human tissue sources is required to confirm the present findings, and definitively establish the suitability of SKP-SCs for clinical applications, this work carries important implications for the potential clinical translation of SKP-SCs as a cellular therapy for SCI. To facilitate a detailed discussion of the research described in Chapters 2 through 4, the remainder of this introductory chapter provides a rather thorough review of the major topics pertinent to the experiments presented herein. I begin by discussing SCI with a particular focus on the need for effective treatments to promote neurological repair and functional recovery. Next I describe the pathophysiology of SCI including a detailed account of secondary injury and the response to injury displayed by various CNS resident cell types. Following that I delve into a general discussion regarding therapeutic strategies for CNS repair, with an emphasis on cellular therapies for SCI. Next I examine the topic of SCs including a detailed discussion of their development, their functions in the intact and injured PNS, the benefits and limitations of N-SC transplantation as a treatment for CNS injury/disease, and finally the drawbacks of using peripheral nerve as a source for autologous SCs for therapeutic clinical applications. At that point I critically analyze the alternative (i.e., non-nerve) sources of autologous SCs, other than SKPs, that have been suggested as potentially suitable replacements for N-SCs in the clinic. And finally, I review what was known about the SKPs and SKP-SCs prior to the initiation of the present work; which leads rather directly into the overview of the experiments conducted and hypotheses tested in the present work. 7   Spinal cord injury 1.2 Causes of SCI 1.2.1SCI is extremely heterogeneous in humans, as a variety of both traumatic (e.g., motor vehicle accidents) and non-traumatic (e.g., ongoing infection/disease) events can cause damage to spinal cord tissue, and the functional ramifications of that damage depend on multiple factors including the pattern of pathology associated with a particular mechanism of injury, as well as the precise location of the injury and the extent of tissue damage (Kirshblum et al., 2002; Norenberg et al., 2004). The animals models used to test the potential therapeutic benefit of SKP-SC transplantation in the present work (thoracic contusion and dorsolateral funiculus crush) are traumatic in nature, and thus traumatic SCI is the focus of this discussion and the term SCI is used throughout this work to refer to traumatic SCI unless otherwise stated. In humans, the frequency of different causes of SCI varies according to geographical location, often reflecting cultural and socioeconomic differences among nations and regions (Farry and Baxter, 2010; Cripps et al., 2011), but worldwide SCI appears to disproportionately affect males as compared to females, as the estimated incidence in males across multiple studies is 2-4 times that in females (Dryden et al., 2003; Wyndaele and Wyndaele, 2006; Turner et al., 2009; Couris et al., 2010; NSCISC, 2012). The exception to the latter appears to be in the elderly population where men and women show similar rates of injury, mostly due to falls at home (Spivak et al., 1994). In North America, as in most developed nations, the most common cause of traumatic SCI has historically been motor vehicle accidents, but recent reports indicate that work accidents and falls at home are becoming more prevalent as the median age of the general population rises (Tator, 1995a; Dryden et al., 2003; Pickett et al., 2006; Turner et al., 2009; 8  Couris et al., 2010; Farry and Baxter, 2010; Cripps et al., 2011; NSCISC, 2012). These types of incidents tend to result in vertebral dislocations and fractures of the spinal column, which represent the most common injuries observed in the clinic, where vertebral dislocations (with and without fracture) occur in 29-45% of cases and burst fractures occur in 30-48% of cases (Sekhon and Fehlings, 2001; Pickett et al., 2006). Vertebral dislocations and fractures most often result in contusion and some degree of compression and/or distraction of the spinal cord, and although laceration and even full transection of the spinal cord can occur as a result of severe fracture-dislocation or foreign bodies (e.g., knives and bullets) traversing the vertebral canal, those primary mechanisms of injury are much less common clinically (Bunge et al., 1993; Dumont et al., 2001; Norenberg et al., 2004).   Relating spinal cord anatomy to functional losses due to SCI 1.2.2With the exception of functions subserved by the cranial nerves that arise directly from the brain or brainstem, all communication between the brain and peripheral organs and tissues occurs via the spinal cord. The spinal cord contains numerous ascending (sensory) and descending (motor) tracts of axons that run along its length in the white matter at specific locations in the dorsoventral-mediolateral axes (i.e., transverse plane) and carry information between structures in the brain and each level of the spinal cord (Guertin, 2012). Sensory and motor axons enter/exit the spinal cord via the sensory and motor nerve roots that comprise the spinal nerves, which are continuous with the peripheral nerves that innervate specific organs and tissues at each vertebral level (Guertin, 2012). At the cervical level, the spinal nerves innervate the head, neck, shoulders, chest, upper limbs and respiratory organs, whereas at the thoracic level the spinal nerves primarily innervate the trunk and its organs, and at the lumbar and sacral levels 9  they innervate the lower limbs and pelvic organs (McDonald, 1999). Thus the spinal cord is anatomically organized to enable communication between specific regions of the brain and specific peripheral organs/tissues at each vertebral level, but it is important to note that the spinal cord is not a passive conduit for that communication, but rather an active participant in the integration and modulation of inputs from descending supraspinal pathways and peripheral afferents (Flynn et al., 2011). In addition to the long ascending and descending axon tracts in the white matter, the neuronal content of the spinal cord is largely composed of interneurons that form circuits within the gray matter of each spinal segment and propriospinal interneurons that give rise to intraspinal axons that project via the white matter to the gray matter in other spinal segments, sometimes over substantial distances (Conta and Stelzner, 2009). The intraspinal network of interneurons, including propriospinal neurons, plays a critical role in motor reflexes, voluntary movement, and sensory processing (reviewed by Flynn et al., 2011). Gray matter interneurons form simple spinal reflex pathways that enable involuntary autonomic and somatic motor responses to peripheral stimuli in the absence of supraspinal input, but they also form more complex networks known as central pattern generators (CPGs), that generate patterned or rhythmic responses in multiple muscle groups and provide the basic motor commands for more complex behaviours such as locomotion, scratching, micturition and even ejaculation (Guertin and Steuer, 2009; Flynn et al., 2011; Guertin, 2012; Courtois et al., 2013). Unlike simple reflexes, CPG-driven functions are subject to conscious/voluntary control, as locomotor CPG activity is induced and modulated by supraspinal input, although it is also modulated directly by primary afferent input, which enables the automatic adjustment of CPG-generated motor output in response to sensory feedback 10  (Guertin, 2012). Interneurons in the gray matter of the cervical and lumbar enlargements are believed to form motor CPGs that generate the basic motor patterns underlying rhythmic forelimb/hindlimb or upper/lower limb movements, respectively, in all vertebrates including humans (Dietz et al., 1994; Kiehn and Butt, 2003; van Hedel and Dietz, 2010; Harkema et al., 2011; Guertin, 2012). Although those circuits are normally under supraspinal control, there is evidence from many species, including humans, that propriospinal connections couple the cervical and lumbar CPGs and thus enable the unconscious coordination of forelimb/hindlimb or arm/leg movements during locomotion to be achieved largely at the spinal level (Juvin et al., 2005; Dietz and Michel, 2009; Zehr et al., 2009; Juvin et al., 2012). In adult humans, as with all adult mammals, spontaneous repair of the injured CNS generally fails, so SCI causes permanent damage to both ascending and descending white matter tracts as well as the interneurons and propriospinal axons of the gray matter, leading to permanent neurological impairments that affect motor, sensory, and autonomic functions. As a result of the anatomical organization of the white matter tracts, damage to specific tracts in the spinal cord disrupt the connection between the brain and portions of the periphery innervated by that tract below the level of injury (Blight, 2002). Thus, an injury at the level of the 4th cervical vertebra (C4) disrupts communication between the brain and all peripheral structures innervated by the injured spinal tract(s) from C5 down, including thoracic, lumbar, and sacral levels, but does not disrupt communication between the brain and peripheral organs/tissues innervated by spinal nerves at levels above the injury (e.g., C2). In addition to the effects of damage to axon tracts in the white matter, SCI often results in damage to the gray matter, causing the loss of interneurons involved in local reflex pathways, CPGs, and/or propriospinal connections between 11  spinal segments, which lead to functional deficits in reflexes, patterned motor responses, and/or the coordination of responses, respectively (Flynn et al., 2011). Even if a CPG remains intact, injury in rostral segments can cut off the supraspinal input required to induce or properly modulate its activity, thereby disturbing or eliminating voluntary control of CPG-dependent patterned activities such as locomotion. Although there is evidence to support the existence of CPGs in humans, their role in locomotion remains hotly debated, as it is thought that supraspinal input may play a more extensive part in the regulation of locomotion in man compared with lower vertebrates (Dietz and Michel, 2009; Guertin, 2012).  The outcome of SCI in humans 1.2.3In humans the neurological impairments caused by SCI often include paralysis, stiffness or spasticity of the skeletal muscles, lost or abnormal sensation (anesthesia, hypoesthesia, paresthesia, or dysesthesia) across multiple sensory modalities (e.g., touch, temperature, vibration), persistent neuropathic pain, as well as deficits in cardiovascular, respiratory, bowel, bladder, and/or sexual functions (Levi et al., 1995c; Watanabe et al., 1996; Stiens et al., 1997; Blight, 2002; Winslow and Rozovsky, 2003; Anderson, 2004; Gustin et al., 2010; Popa et al., 2010; Teasell et al., 2010; Zeilig et al., 2012). The majority of traumatic SCI occurs at the cervical level and thus impairs function in the pelvic organs, trunk, and upper and lower limbs (i.e., tetraplegia / quadriplegia), while the remaining injuries occur at the thoracic, lumbar or sacral vertebral levels and result in lower limb and trunk impairments only (i.e., paraplegia) (Farry and Baxter, 2010; NSCISC, 2012). The degree of functional loss varies substantially according to the severity of neurological damage, but a general distinction is often made between ?complete? injuries, which are characterized by a total loss of sensory and motor functions below 12  the level of damage, and ?incomplete? injuries in which some degree of residual neurological function is maintained. The most severe neurologic impairments usually occur in individuals with complete tetraplegia resulting from high cervical (C1-C4) SCI, as in addition to the complete loss of motor and sensory function of all limbs, and a range of severe autonomic disturbances, these individuals often require assistance breathing (e.g., mechanical ventilation or phrenic nerve pacing) due to the loss of control of the muscles required for respiration (Winslow and Rozovsky, 2003). Most individuals with SCI have less severe neurologic impairments, as complete tetraplegia is relatively rare compared to all other types of SCI combined, only occurring in 16% of new injuries in the United States (US) since 2005 for example, and only a fraction of those cases involve injury at high cervical levels (NSCISC, 2012). However, complete neurologic recovery is rare (<1%) after any trauma to the spinal cord, so most people with SCI will suffer some degree of long-term disability, and many require assistive / adaptive devices (e.g., wheel chair or vehicular hand controls), assistance in daily living (e.g., homecare), and continued/frequent medical supervision (Dryden et al., 2004; Turner et al., 2009; Guilcher et al., 2010; DeVivo et al., 2011; NSCISC, 2012). The neurological impairments caused by SCI are often directly associated with a host of secondary health complications (e.g., deep vein thrombosis, urinary tract infections, pressure sores / ulcers and respiratory infections), the variety and severity of which depend largely on the degree of disability, and particularly the severity of autonomic dysfunctions (Levi et al., 1995c; Winslow and Rozovsky, 2003; Dryden et al., 2004; Christie et al., 2011; DeVivo and Farris, 2011). These complications contribute substantially to SCI-related health issues, resulting in increased need for medical attention and rehospitalization long after initial discharge, and they 13  are often the cause of mortality in the SCI population (Levi et al., 1995b; Dryden et al., 2004; Shavelle et al., 2006; Strauss et al., 2006; Christie et al., 2011; DeVivo and Farris, 2011; NSCISC, 2012). In developed nations, survival after SCI, particularly during the first year or two post-injury, has improved dramatically since the 1940s due to advances in emergency medical services and the treatment of SCI and related secondary complications (see below), and yet individuals living with SCI continue to have a reduced life expectancy compared to their age-matched uninjured peers in the general population (Frankel et al., 1998; Imai et al., 2004; Shavelle et al., 2006; Strauss et al., 2006; Farry and Baxter, 2010). Mortality rates are significantly higher during the first year or two after SCI, particularly for people with more severe injuries and those injured at an older age (Shavelle et al., 2006; Strauss et al., 2006; NSCISC, 2012), but those rates remain elevated in the SCI population compared to the general population, even after that critical period. For example, a 20 year old male with paraplegia who survives at least a year post-injury is expected to live to reach 65 years of age, whereas a person in the general population without a SCI has a life expectancy of 79 years (NSCISC, 2012). Had the same 20 year old received an injury causing high tetraplegia with ventilator dependence, he would only be expected to reach the age of 45 (NSCISC, 2012). In addition to the negative impact of SCI on physical health and well-being, the combination of neurological impairment, persistent pain, and secondary health complications can be devastating in terms of psychological health, as individuals with SCI are at increased risk for depression and anxiety and often report lower perceived quality of life than the general population (Davidoff et al., 1992; Evans et al., 1994; Levi et al., 1995b; Elliott and Frank, 1996; 14  Dijkers, 1997; Krause, 1998; Tate et al., 2002; Sakakibara et al., 2009; Shin et al., 2012). This is particularly the case for individuals with more severe injuries and older people with SCI who tend to require more physical assistance (Gerhart et al., 1993; Evans et al., 1994; Shin et al., 2012). For non-disabled individuals, the mental and emotional toll of SCI may be difficult to fully comprehend, but the culmination of those effects is clearly and sadly demonstrated by the fact that a person with SCI is five times more likely to commit suicide than someone in the general populace (Dijkers et al., 1995). Beyond the physical, mental, and emotional burdens, there is also a high financial cost to living with SCI. In addition to expenses related directly to compensating for physical disabilities (e.g., assistive / adaptive devices, home renovations/moving costs to ensure accessibility, etc.), individuals with SCI also tend to require more frequent and longer hospitalizations, greater medical attention in general, more prescription medication, and significantly more homecare service compared to people in the general population (Levi et al., 1995b; Dryden et al., 2004; Munce et al., 2009; Guilcher et al., 2010; DeVivo and Farris, 2011). Depression, anxiety, sleep disturbance and fatigue are commonly cited as reasons for seeking medical attention, so clearly the psychological toll of SCI contributes to increased health care utilization in this population (Levi et al., 1995c; Levi et al., 1995b; Dryden et al., 2004). Recurring or persistent secondary complications and hospital admissions represent disruptions to daily life, which are costly both financially and socially, and may negatively impact quality of life (Menter et al., 1991; Johnson et al., 1996; Krause, 1998). SCI often has a negative effect on employment and income (Krause et al., 2011) and tends to disproportionately affect individuals from low income families (at least in the US) (Turner et al., 2009). As such, many people living with SCI have neither the financial 15  means nor sufficient medical insurance coverage to accommodate SCI-related medical, homecare, and living expenses and a large proportion of those individuals are forced to rely on publicly funded health care and informal homecare by friends and relatives (Turner et al., 2009). The situation is far more dire for individuals in undeveloped nations (e.g., Nigeria, Zimbabwe, and South Africa), where most patients die either before they reach a treatment centre or within the first year post-injury (Cripps et al., 2011).  The global cost of SCI  1.2.4SCI is a global health problem that afflicts millions of people worldwide and represents a substantial burden on society and the global economy. According to data from studies conducted worldwide between 1995 and 2005, the incidence of SCI varies considerably from country to country, and the global incidence of SCI ranges from 10.4-83 cases per million inhabitants per year, depending on the country in question and the methods of estimation used (Wyndaele and Wyndaele, 2006). Excluding those who die prior to hospitalization, a more recent estimate places the annual incidence of SCI at about 40 cases per million people, or approximately 12,000 new cases each year in the US (NSCISC, 2012). Similar values have been found in Canada, where the Rick Hansen Institute reports about 1,400 new cases of traumatic SCI (not resulting in death prior to hospitalization) are estimated to occur each year; which translates to 41 cases per million inhabitants per year (Farry and Baxter, 2010; Noonan et al., 2012). Among developed nations, the average estimated prevalence of SCI is 485 cases per million inhabitants based on SCI registry data (Wyndaele and Wyndaele, 2006). Assuming that a similar rate applies worldwide, and given a current world population of roughly 7 billion (http://www.census.gove/popclock/), that prevalence translates into nearly 3.5 million people 16  living with SCI on our planet. According to estimates from a recent household survey by the Christopher and Dana Reeve Foundation, nearly 1.3 million of those individuals live in the US alone (Turner et al., 2009), although prevalence rates based on SCI registry and hospitalization data are much lower, with an upper range of approximately 327,000 persons (NSCISC, 2012). In Canada the most recent estimate of the prevalence of SCI is ~85,000 persons, and nearly half of those cases occurred as a result of non-traumatic causes, such as diseases, infections, or tumours (Farry and Baxter, 2010; Noonan et al., 2012). In contrast, the data from the Reeve Foundation showed that only 18% of patients with SCI in the US did not directly attribute that injury to physical trauma (Turner et al., 2009). The estimated combined annual cost of health care and living expenses for individuals with SCI varies greatly according to the level and severity of injury, as do the estimated lifetime costs directly attributable to SCI (Cao et al., 2011; DeVivo et al., 2011). According to the National Spinal Cord Injury Statistical Center (NSCISC; Burmingham, Alabama), the average estimated expenses in the first year after injury range from $334,000 for individuals with incomplete motor function at any level to just over $1 million for those with high tetraplegia (C1-C4), and the estimated costs each subsequent year range from $41,000 to $178,000 for those same groups; all values in Feb 2012 US dollars (NSCISC, 2012). The NSCISC also provides the most recent estimates of the lifetime costs for an individual with SCI, which range from $1.5 million (incomplete motor function at any level) to $4.5 million (C1-C4 tetraplegia) for someone injured at 25 years of age, and $1 million to $2.5 million for those same groups if the individual is injured at 50 years of age; all values in Feb 2012 US dollars (NSCISC, 2012). Importantly, those figures do not include any indirect costs (e.g., lost wages, benefits, and/or productivity), 17  which costs individuals with SCI an estimated $69,000 (Feb 2012 US dollars) per person per year, though that value varies considerably depending on the level and severity of injury, education, and pre-injury employment history (NSCISC, 2012). In 2009, SCI was estimated to cost the US health care system $40.5 billion (2009 US dollars) annually (Turner et al., 2009). Based on an estimated prevalence of 3.5 million people with SCI worldwide (Wyndaele and Wyndaele, 2006) and the fact the US is estimated to account for only perhaps 37% of that population (Turner et al., 2009), the extrapolated health care cost for SCI worldwide is $109 billion (2009 US dollars) annually. That estimate is unlikely to accurately represent the worldwide SCI-related health care expenditures, as the availability and cost of SCI-related health care varies substantially from nation to nation, but it does drive home a significant point ? SCI represents a massive economic burden to people and societies around the globe.  A brief history of the treatment of SCI in humans 1.2.5SCI has obviously been occurring in humans since the dawn of the species, but the first known medical documentation of SCI cases occurred in the Edwin Smith papyrus (Hughes, 1988), which is an ancient Egyptian surgical treatise that dates to roughly 1600 BCE (Allen, 2005). In that ancient text, SCI was described as ?an ailment not to be treated? (Hughes, 1988), and sadly, that viewpoint continued to dominate the care of patients with SCI for about 3,500 years thereafter; during which time SCI was generally considered a death sentence (Donovan, 2007). During most of that period, the only intervention that was relatively widely adopted for SCI was non-surgical (i.e., closed) traction to treat spinal fractures and dislocations; a method originally championed by Hippocrates (460-360 BCE), although he did not actually advocate it when fracture was coupled with paralysis (Richards, 1968; Knoeller and Seifried, 2000; Lifshutz 18  and Colohan, 2004). The first recorded surgical intervention for SCI was the removal of bone fragments impinging on the spinal cord, a procedure performed by Paulus of Aegina (625-690 CE) over 2,000 years after the ancient Egyptians first described SCI (Knoeller and Seifried, 2000; Lifshutz and Colohan, 2004). However, that and other surgical techniques generally failed to gain wide acceptance over the next 1,300 years, as spinal column surgery was extremely painful, often failed, and frequently resulted in death due to infection (Knoeller and Seifried, 2000; Donovan, 2007). Between the late 19th and early 20th centuries, the ease and safety of all surgical interventions were drastically improved due to the application (and refinement) of aseptic techniques, general anesthetics, and antibiotics in medicine, and these advances prompted increased efforts to develop surgical interventions to treat spinal trauma (Donovan, 1994; Knoeller and Seifried, 2000; Donovan, 2007). With increased surgical efforts, came improved surgical techniques, and combined with improvements in imaging techniques (i.e., the development of computerized tomography [CT] and magnetic resonance imaging [MRI]) and spine surgery instrumentation (e.g., fusion rods, plates, and pedicle screws) these advances eventually led to the development of the safe and effective procedures for stabilizing the fractured spine and decompressing the injured spinal cord that are used today (Knoeller and Seifried, 2000; Donovan, 2007). In the early 20th century, medical advances had yet to impact the treatment of SCI, so the prospect of surviving a SCI remained extremely poor. For example, in World War I, 80% of American soldiers with traumatic SCI died within the first two weeks, and only 10% survived to 1 year after injury (Guttmann, 1946). However, the 1940s heralded marked improvement in survival after SCI, largely due to the application of antibiotics, but also the introduction of SCI 19  management units dedicated to the prevention and management of secondary complications and comprehensive rehabilitation, and this trend in improved survival has continued since then, with more recent advances in emergency medical services and the development of effective surgical interventions for spinal fractures and dislocations, in the 1960s and 1970s respectively (Whiteneck et al., 1992; Samsa et al., 1993; Knoeller and Seifried, 2000; Donovan, 2007). The improved survival of patients with SCI meant that for the first time in history there was a growing population of individuals with chronic SCI and therefore a growing need for long-term care and effective treatment of this condition, and by the 1980s, recognition of the latter finally started spawning large-scale clinical trials for interventions aimed at treating the injured spinal cord itself, rather than merely the bones of the spinal column. Over the preceding century, the testing of experimental therapeutics for SCI had expanded substantially, fuelled by growing knowledge of neuroanatomy and neurobiology, and new insights into the underlying pathology of SCI from animal research (Lifshutz and Colohan, 2004). Promising results from work in animals with SCI prompted clinical testing of a wide variety of experimental treatments throughout that period (e.g., enzymes to dissolve scars, electrical stimulation, nerve transplants, omental transposition, and spinal cord cooling), but most were only examined in a single patient or a group of patients treated by a single clinician, and despite the fact that many of those experiments reported ?successful? results, none were ever proven efficacious or widely adopted (Tator, 2006). One exception to that was the use of corticosteroids, which became a ?standard of care? for acute SCI by the 1980s, despite a lack of firm clinical evidence that it was beneficial (Lammertse, 2013). 20  The late 20th century saw a gradual shift in medicine towards evidence-based practice, which emphasizes that empirical evidence from properly conducted clinical research should be the cornerstone of clinical decision making, rather than subjective clinical experience (Sackett et al., 1996; Howick, 2011). This shift was facilitated by many factors, not the least of which was the adoption of new premarket approval regulations for novel pharmaceutical drugs by the US Food and Drug Administration (FDA) in 1962 (Kaplan et al., 2011). Those new regulations led to the adoption of randomized controlled clinical trials, which were mandatory to prove the effectiveness of new drugs in order to attain FDA approval to market a drug as a treatment for a specific condition (Kulynych, 1999). Over time the use of randomized controlled trials became the ?gold standard? for establishing efficacy in all medical research, and its application expanded from drug testing to the evaluation of surgical and even diagnostic procedures (Kaplan et al., 2011). In the field of SCI, the clinical trial era began in February of 1979 with the enrolment of the first patient in the multicentre double-blind randomized controlled trial of methylprednisolone (Bracken et al., 1984; Lammertse, 2013). However, it wasn?t until the mid-1980s and beyond that the large-scale, randomized, placebo-controlled trials began producing data regarding the efficacy of interventions aimed at treating the underlying neuropathology of SCI; e.g.: methylprednisolone  (Bracken et al., 1990; Bracken et al., 1997) and GM-1 ganglioside (Geisler et al., 1991; Geisler et al., 2001). This marked an important transition in SCI clinical research, as the focus had now largely shifted from promoting survival and minimizing physical trauma in SCI patients to promoting neuroprotection and regeneration using pharmacological / biological interventions directed at the underlying neuropathology responsible for functional 21  deficits. During these early trials the transplantation of tissue and cells was still in its infancy, and remained largely limited to single case reports or small non-randomized exploratory trials that mainly addressed treatment safety (i.e., Phase I clinical trials) (for examples see: Falci et al., 1997; Wirth et al., 2001; Tadie et al., 2002; Rabinovich et al., 2003; von Wild and Brunelli, 2003; Cheng et al., 2004; Feron et al., 2005; Knoller et al., 2005; Park et al., 2005; Schwartz and Yoles, 2005); with the exception of the transplantation of autologous activated macrophages, which was the first cell-based therapy subjected to randomized controlled trial in studies initiated in 2003 (Lammertse et al., 2012). In the meantime, a vast amount of preclinical research in animal models of SCI continued to increase knowledge regarding the underlying mechanisms of injury and suggest novel treatments for clinical application, prompting ever increasing numbers of clinical trials in the field of SCI (Tator, 2006; Lammertse, 2013).  The current state of treatment for SCI in humans 1.2.6To date, there have been an impressive number of clinical trials for treatments aimed at improving neurological recovery following SCI (e.g.: Bracken et al. 1990; Bracken et al. 1997; Cardenas et al. 2007; Casha et al. 2012; Dobkin et al. 2006; Fehlings et al. 2011; Fehlings et al. 2012a; Geffner et al. 2008; Geisler et al. 2001; Kapadia et al. 2011; Karamouzian et al. 2012; Kumar et al. 2009; Lammertse et al. 2012; Levi et al. 2009; Lima et al. 2010; Lorenz et al. 2012; Pal et al. 2009; Pitts et al. 1995; Pointillart et al. 2000; Popovic et al. 2011; Rabinovich et al. 2003; Saberi et al. 2011b; Shapiro et al. 2005; Tadie et al. 2003; Tator et al. 1987; Triolo et al. 2012; Vaccaro et al. 1997; Vale et al. 1997; Wu et al. 2012; Yang et al. 2012; Yoon et al. 2007; Zariffa et al. 2012). However, the testing of many of those potential treatments remains in the preliminary stages (i.e., Phase I or II clinical trials), where small groups of patients are exposed 22  to a treatment to establish safety and appropriate dosing regimens, as relatively few treatments have been subjected to the kind of testing that is necessary to establish clinical efficacy (e.g., Phase III clinical trials using large randomized placebo-controlled trials; or large non-randomized prospective observational studies). Of those treatments that have undergone adequate testing to definitively determine efficacy, most have failed to demonstrate sufficient neurological benefit to warrant their adoption as a standard of care for the treatment of SCI (Bracken et al., 1990; Bracken et al., 1997; Pointillart et al., 2000; Geisler et al., 2001; Tadie et al., 2003; Lammertse et al., 2012; Yang et al., 2012). The corticosteroid methylprednisolone was thought to be a rare exception to that, as initial reports from randomized placebo-controlled trials indicated a modicum of therapeutic efficacy (Bracken et al., 1990; Bracken et al., 1997), which led to rapid and widespread clinical application, but clinical use of methylprednisolone has subsequently declined significantly due to continued scrutiny of trial evidence and research demonstrating a high risk of serious health complications at the effective dosage used to treat SCI (Hurlbert, 2000; Pointillart et al., 2000; Hugenholtz, 2003; Hurlbert and Hamilton, 2008). Of all of the treatments tested so far, only surgery and rehabilitation approaches have yielded sufficient safety and therapeutic efficacy (i.e., evidence of neurological or functional recovery) in large enough groups of patients to justify their adoption as standards of care following SCI. Early (<24 hours post-SCI) surgical stabilization and decompression of the spinal cord was recently shown to be associated with significant neurological improvement at 6 month follow-up in patients with cervical SCI regardless of the completeness of injury (Fehlings et al., 2012b). In contrast, although physical rehabilitation using manual/automated activity-based rehabilitation (Thomas and Gorassini, 2005; Wirz et al., 2005; Dobkin et al., 2006; Harkema et 23  al., 2012b; Harkema et al., 2012a; Lorenz et al., 2012; Zariffa et al., 2012) with/without functional electrical stimulation (Ladouceur and Barbeau, 2000; Kapadia et al., 2011; Popovic et al., 2011) has been found to improve neurological function in patients over time following injury, those effects are generally only significant for patients with incomplete injuries (i.e., some preservation of function). Thus, although the medical, surgical, and rehabilitative treatments for SCI and related secondary complications advanced considerably during the late 20th century and beyond, standard medical practice following SCI has largely remained focused on the same approaches: the prevention and management of secondary complications to maintain health, surgical decompression and stabilization to minimize damage, and physical rehabilitation to optimize recovery (Mothe and Tator, 2012). And so, despite advances in the treatment of SCI over the last 30-40 years, neuroprotection and neurological repair (i.e., regeneration) remain quite limited, and SCI continues to cause lifelong disability and neurological dysfunction. Although that summary appears quite bleak and it is true that no regenerative therapy has proven to be beneficial in human SCI to date (Fawcett, 2002; Illis, 2012), it is important to keep in mind that there are many therapeutic options that have yet discovered, let alone adequately tested for efficacy following SCI. Thus, although many treatments have been proven ineffective, a variety of potential treatments have demonstrated preliminary signs of efficacy in early (Phase I/II) clinical trials and currently await testing in large randomized controlled trials designed to definitively determine efficacy (e.g.: Casha et al. 2012; Fehlings et al. 2011; Kumar et al. 2009; Lima et al. 2010; Saberi et al. 2011b; Yoon et al. 2007). In addition, there are many clinical trials aimed at neuroprotection and/or neurological repair after SCI that are currently in the planning, recruiting, or analysis phases (Fehlings et al., 2012a; Talan, 2012; Tator et al., 2012; Wilcox et 24  al., 2012), any one of which may also provide a new standard of care for the treatment of SCI in humans.   The pathophysiology of SCI 1.3The long-term pathological consequences of traumatic SCI are the result of two distinct, yet related, processes commonly referred to as primary and secondary injury. Primary injury refers to tissue damage that is directly attributable to the initial trauma to the spinal cord, whereas secondary injury refers to a cascade of pathophysiological events at the tissue, cellular, and molecular levels (e.g., ischemia, inflammation, oxidative stress, etc.) that are triggered by primary injury and contribute to tissue damage over time following injury (Tator and Fehlings, 1991; Tator, 1995b). The only treatment that can be directed at primary injury is prevention, as there is no intervention that can stop trauma from occurring when external forces are applied directly to the spinal cord. In contrast, secondary injury mechanisms involve protracted tissue damage that accumulates over time, and may therefore be ameliorated by interventions designed to counteract the underlying pathological processes. This is not as straightforward as it sounds however, as there are a large number of secondary injury mechanisms at work following SCI and many of those mechanisms are inter-related.  Primary injury 1.3.1Primary injury mechanisms typically include contusion (i.e., impact), compression, distraction (i.e., forcible stretching of the spinal column in the axial plane), and laceration of the spinal cord (Dumont et al., 2001). In humans, SCI often involves multiple primary injury mechanisms. For example, contusions resulting from spinal fractures are always associated with 25  some degree of compression whether transient or persistent, whereas dislocations of the spinal column tend to involve contusion and some degree of compression and distraction, and either type of injury may result in laceration of the spinal cord by sharp bone fragments or shearing, respectively (Dumont et al., 2001). Even lacerations due to stab injuries in humans regularly involve some degree of contusion/compression due to the impact of the spinal cord with the surrounding vertebrae (Lipschitz and Block, 1962). This is often not the case in preclinical SCI research however, as animal models of SCI typically involve contusion (Allen, 1911; Bresnahan et al., 1987; Noyes, 1987b; Behrmann et al., 1992; Gruner, 1992; Stokes, 1992; Scheff et al., 2003), compression (Rivlin and Tator, 1978; Dolan et al., 1980), distraction (Myklebust et al., 1988; Dabney et al., 2004), partial or complete crush (Plemel et al., 2008), or partial or complete transection (Theriault and Tator, 1994; Chadi et al., 2001; Kwon et al., 2002a; Hendriks et al., 2006; Pettersson et al., 2007) of the spinal cord in near-complete isolation from other mechanisms of injury. Given that most cases of SCI in humans include contusion as a primary mechanism of injury (Bunge et al., 1993; Schwab and Bartholdi, 1996; Dumont et al., 2001; Norenberg et al., 2004), contusive injury models tend to mimic the typical human pathology best, and are therefore most commonly used to study secondary mechanisms of injury and to assess the efficacy of potential treatments in animals (Kwon et al., 2002a; Dietz and Curt, 2006). Primary injury typically results in two major and immediate pathological events: 1) physical damage to blood vessels in the spinal cord, particularly the microvasculature of the gray matter, causing hemorrhage, edema and impaired microcirculatory tissue perfusion at the site of initial trauma and 2) disruption of neuronal, glial, and endothelial cellular membranes causing the rapid necrotic death of neurons, astrocytes, oligodendrocytes, and endothelial cells at the 26  lesion epicentre and less severe damage to cells, axons, and myelin in adjacent regions (Hausmann, 2003; Choo et al., 2007; Mothe and Tator, 2012); Mautes et al., 2000). More severe injuries are known to cause more extensive hemorrhage and necrosis (Anderson, 1985; Noyes, 1987a; Boldin et al., 2006), but the initial pattern of neuropathology also varies according to the specific mechanisms of primary injury that occur during SCI. For example, in the case of laceration injuries, the pattern of initial damage is clearly dictated by the trajectory of foreign bodies or bone through the spinal canal, but in all other mechanisms of primary injury it is generally recognized that initial mechanical damage occurs primarily in the central gray matter, where the tissue is less resilient and more vascularized (Wolman, 1965; Kakulas, 1984; Tator and Koyanagi, 1997; Dumont et al., 2001; Norenberg et al., 2004). Thus in most cases, primary injury preferentially affects the capillaries/venules, neurons/axons and glia in the central gray matter. However, even amongst those primary injury mechanisms that preferentially affect the gray matter there can be substantial variability in the pattern of hemorrhage and cellular membrane compromise. For example, just 5 minutes after injury in the rat, Choo et al. (2007) found that contusion and dislocation injuries both produced substantial hemorrhage in the gray matter at the site of injury, whereas distraction injuries caused little to no noticeable hemorrhage. Furthermore, although contusion and dislocation both resulted in damage to cellular membranes in the gray and white matter at the lesion epicentre, as well as near-immediate tissue necrosis in the gray matter, distraction was not associated with obvious necrosis and both dislocation and distraction were associated with more extensive, diffuse disruption of membranes on neuronal somata and axons both rostral and caudal to injury (Choo et al., 2007). Subsequent work by that 27  same group demonstrated that the differences in primary injury resulting from those three mechanisms lead to substantial differences in secondary pathology as well (Choo et al., 2008), clearly demonstrating that the mechanism of primary injury plays a key role in determining the pattern and extent of spreading secondary damage; which may explain why the extent of hemorrhage has proven highly predictive of functional outcomes following both experimental and clinical SCI (Noyes, 1987a; Boldin et al., 2006).  Secondary injury 1.3.2The necrotic death of cells, and the severe axonal and vascular compromise caused by the initial trauma to the spinal cord triggers a cascade of secondary injury mechanisms at the tissue, cellular and molecular levels, which contribute to the degeneration of tissue adjacent to the lesion site that was not damaged by the initial injury itself. Many of these processes are inter-related, arising from and/or contributing to one another in a manner that promotes further tissue loss over time.  1.3.2.1 Secondary injury mechanisms triggered by vascular changes after SCI Primary injury generally causes a relatively small region of hemorrhage to appear almost immediately at the injury epicentre, but over the next 24 hours, very small regions of hemorrhage (i.e., petechiae) often emerge in adjacent tissue that was spared the initial trauma, and these small hemorrhages coalesce over time into one large hemorrhagic lesion during secondary injury (Simard et al., 2007). This progressive hemorrhagic necrosis is thought to substantially increase the loss of cells, as the volume of hemorrhage typically doubles within 12-18 hours after contusion injuries (Simard et al., 2012). Although a number of destructive processes are likely to contribute to progressive hemorrhagic necrosis during secondary injury, one of the prime culprits 28  is the non-selective cation channel sulfonylurea receptor 1 (Sur1). Sur1 is not expressed constitutively, but is upregulated de novo in both neural and endothelial cells following CNS injury, and this channel has been linked to microvascular (e.g., capillary) fragmentation, leading to vasogenic edema and delayed secondary petechial hemorrhage formation after SCI (Simard et al., 2012). As the term hemorrhagic necrosis implies, hemorrhage is well associated with the necrotic death of neural cells. Although there are many potentially neurotoxic substances that may be present in blood, the majority of necrosis due to hemorrhage appears to result from hemoglobin neurotoxicity (Regan et al., 2008). The degradation of hemoglobin generates iron, which accumulates in tissue following hemorrhage and can persist for months thereafter (Hua et al., 2006). As a catalytic metal ion, iron serves as a substrate for free radical chain reactions, particularly during lipid peroxidation (Mautes et al., 2000), and thus causes oxidative damage to cells (see below), leading to necrosis in cells that accumulate too much damage to survive. Neurons are particularly vulnerable to hemoglobin toxicity, primarily due to their limited capacity to sequester and detoxify iron (Regan et al., 2008). In addition to hemorrhage, SCI is also associated with widespread edema and ischemia (Dumont et al., 2001). Vasogenic edema results from a breakdown of the blood-spinal cord barrier (BSCB), which occurs due to the damage/loss of endothelial cells and astrocytes that maintain the tight endothelial junctions of the BSCB (Saadoun and Papadopoulos, 2010; Bartanusz et al., 2011). Ischemia is thought to result from a combination of local vascular damage, local responses to that damage (i.e., vasospasm, thrombosis, and a loss of autoregulatory homeostasis), and systemic hypoperfusion that occurs due to injury-induced 29  neurogenic shock; the latter of which is a common cardiovascular response to severe SCI, characterized by bradycardia, hypotension, and decreased cardiac output and peripheral resistance (Dumont et al., 2001). As secondary injury progresses, vascular damage and BSCB breakdown spreads, as numerous vasoactive and/or endothelial destructive compounds (e.g., endothelin-1, excitatory neurotransmitters, reactive oxygen species, bradykinins, and histamines) are released into the local environment due to the compromise of neural and endothelial cellular membranes, the activation of adjacent endothelial and glial cells (particularly astrocytes), and the arrival of blood-borne inflammatory cells (Noble and Wrathall, 1989; Tator and Fehlings, 1991; McKenzie et al., 1995; Popovich et al., 1996b; Schnell et al., 1999; Bartanusz et al., 2011). Thus, the breakdown of the BSCB spreads to spared segments both rostral and caudal to injury, peaking within days of injury and gradually declining over the subsequent weeks (Noble and Wrathall, 1989; Popovich et al., 1996b; Schnell et al., 1999; Bilgen et al., 2001), but remaining elevated even 8 weeks after SCI in rats (Cohen et al., 2009). As a result of the combined effect of those events, spinal cord tissue (particularly at the lesion epicentre and in adjacent regions) is subjected to ischemia, edema, disturbed ionic homeostasis, and exposure to a variety of cells and proteins normally excluded from the CNS by the BSCB (Griffiths and Miller, 1974; Beggs and Waggener, 1975, 1976; Lemke et al., 1987; Noble and Wrathall, 1988; Tator and Fehlings, 1991; Popovich et al., 1996b). Vasogenic edema caused by BSCB dysfunction causes swelling of the spinal cord that can lead to compression against the surrounding dura, resulting in high intraparenchymal pressure that contributes to ongoing ischemia (Saadoun et al., 2008; Saadoun and Papadopoulos, 2010). Damage to the vasculature also provides a route for the invasion of the spinal cord by 30  blood-borne inflammatory cells (see below) and the breakdown of the BSCB releases serum into the spinal cord parenchyma that likely contributes to the activation of resident microglia (Ransohoff and Perry, 2009; Takigawa et al., 2010), even in regions relatively removed from the lesion site. The necrotic death of cells due to severe membrane disruption, and widespread hemorrhage, edema, and ischemia serves to perpetuate secondary damage, due to the uncontrolled release of ions, excitatory neurotransmitters, and reactive oxygen species into the extracellular space. However, of all the vascular events induced by SCI, ischemia may contribute the most to secondary injury, as it triggers the widest array of secondary cytotoxic events that contribute directly to the progressive loss of neuronal and glial cells over time (Tator and Fehlings, 1991; Pantoni et al., 1996; Mautes et al., 2000; Dumont et al., 2001; Hausmann, 2003; Szydlowska and Tymianski, 2010). Ischemia results in a lack of sufficient oxygen (hypoxia) and glucose to support cellular metabolism, leading to a reduction in the synthesis of adenosine triphosphate (ATP) and the rapid depletion of cellular energy (i.e., ATP stores) that is required to maintain the ionic gradients necessary for cellular function and homeostasis in the CNS (Saikumar et al., 1998; Harris and Attwell, 2012). As a result of this energy failure, ischemia results in cellular edema, chronic membrane depolarization, the release of excitatory neurotransmitters (e.g., glutamate), and reduced neurotransmitter reuptake from the extracellular space (Liang et al., 2007; Szydlowska and Tymianski, 2010). Cellular edema almost inevitably leads to oncotic cell death, which is a form of necrotic cell death due to cytotoxic swelling, but it also contributes to vasogenic edema and may therefore serve to further exacerbate ischemia following SCI (Liang et al., 2007; Saadoun and Papadopoulos, 2010). Although all CNS cell types swell during ischemia, 31  astrocytes appear to be particularly vulnerable to this phenomenon, due to the presence of aquaporin-4 water channels on the astrocytic endfeet that directly contact endothelial cells of the BSCB (Swanson et al., 2004; Saadoun and Papadopoulos, 2010). The release of excitatory amino acids following SCI occurs due to ischemia, but also due to necrotic cell death and SCI-induced astrocyte dysfunction (see below), and in combination those events cause extracellular glutamate to reach excitotoxic levels rapidly following SCI (Panter et al., 1990; Liu et al., 1991; Wrathall et al., 1996). The accumulation of glutamate in the extracellular space activates glutamate receptors (e.g., N-Methyl-D-aspartate [NMDA], ?-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid [AMPA] and kainate receptors), which induces the release of even more glutamate into the extracellular space (Zhou et al., 2013), thus contributing to excessive glutamate receptor activation; the net effect of which is a rise in intracellular calcium (Ca2+) (Choi, 1988; Matute et al., 2007). In addition to glutamate receptors, the intracellular accumulation of Ca2+ may also occur under ischemic conditions, due to the dysfunctional activation of a variety of other channels and transporters (e.g., TRPM7, NCX, P2X7) and/or the release of calcium from internal stores in the endoplasmic reticulum (Szydlowska and Tymianski, 2010; Arbeloa et al., 2012). The excessive intracellular accumulation of Ca2+ is associated with a wide range of harmful downstream events, including: the over-activation of various proteases, lipases, phosphatases, and endonucleases; further disruption of ion transporters; increased production of reactive oxygen species; endoplasmic reticulum dysfunction; and mitochondrial dysfunction (Kroemer et al., 1998; Dumont et al., 2001; Szydlowska and Tymianski, 2010). Those processes culminate in a variety of negative consequences for the cell, including the persistence of Ca2+ elevation, acidosis, cellular edema, 32  reduced ATP synthesis, increased oxidative and electrophilic stress, DNA fragmentation, cytoskeletal breakdown, and damage to cellular and/or mitochondrial membranes (Lewen et al., 2000; Higuchi et al., 2005; Szydlowska and Tymianski, 2010). In regions of severe/prolonged ischemia, such as the lesion epicentre, these processes commonly lead to necrotic cell death, but in regions subjected to sub-lethal ischemia, apoptotic cell death (which requires energy) is more likely, and is often triggered by the reoxygenation when tissue perfusion improves (Saikumar et al., 1998; Dumont et al., 2001; Szydlowska and Tymianski, 2010). Mitochondrial damage caused by elevated intracellular Ca2+ is thought to play a key role in apoptosis due to ischemia/reperfusion, as this leads to the release of cytochrome c, which acts in concert with apoptosis activating factor-1 to activate caspase-9 to induce caspase-3 and caspase-6, leading to programmed cell death (Budd et al., 2000; Eldadah and Faden, 2000; Kuida, 2000). Oxidative stress resulting from the elevated production of reactive oxygen species (ROS) during ischemia, is elevated even further during reperfusion, and is thought to play a particularly important role in secondary damage following SCI (Demopoulos et al., 1982; Hall and Braughler, 1982; Anderson et al., 1985; Braughler et al., 1987; Hall et al., 1992; Hamann et al., 2008; Yune et al., 2008; Hamann and Shi, 2009), as this process leads to the oxidation of proteins, lipids, and nucleic acids causing macromolecular damage to cellular structures, including the cell membrane and mitochondria, the latter of which may result in permanently decreased ATP synthesis and increased ROS production (Schmidley, 1990; Barja, 2004; Jia et al., 2012). The CNS is particularly susceptible to oxidative stress due to the high concentration of polyunsaturated fatty acids that are vulnerable to lipid peroxidation in neuronal and glial cell membranes, the high rate of oxidative metabolic activity, the related production of reactive 33  oxygen and nitrogen metabolites, and the relatively low concentration of antioxidants (Hall et al., 1992; Storey, 1996; Vaziri et al., 2004; Logan et al., 2005; Hamann et al., 2008). Widespread lipid peroxidation is known to occur after SCI, and contributes substantially to membrane damage (Jia et al., 2012), but also produces non-radical oxidants such as the conjugated aldehyde, acrolein, which is thought to be one of the main culprits responsible for membrane damage, mitochondrial dysfunction and myelin disruption (Hamann et al., 2008; Hamann and Shi, 2009; Shi et al., 2011). The half-life of acrolein is much longer than that of conventional ROS, so it can diffuse out from injured tissue into adjacent spared tissue, propagating secondary injury more readily (Shi et al., 2011).  1.3.2.2 Secondary damage mediated by inflammatory cells following SCI Neutrophils are the first blood-borne inflammatory cells to arrive in large numbers after SCI, as within hours of injury these cells begin to accumulate in the vascular endothelium and the parenchyma adjacent to regions of hemorrhagic necrosis, peaking in number at 12-24 hours in the rat and 1-3 days in humans, and disappearing from the injured cord altogether within about 5-10 days in both of those species (Taoka et al., 1997; Carlson et al., 1998; Fleming et al., 2006; Beck et al., 2010). Although neutrophils help to sterilize the injury site via the release of oxidative and proteolytic enzymes, excessive release of those substances due to the robust neutrophil response following SCI is thought to cause damage to uninjured (?bystander?) cells in adjacent spared tissue (Taoka et al., 1997), and the release of the protease elastase may play a particularly important role in facilitating continued secondary damage by injuring endothelial cells to further increase vascular permeability (Harlan, 1987; Zimmerman and Granger, 1990; Carlos and Harlan, 1994; Taoka et al., 1998). 34  Resident microglia in the CNS are known to respond to local tissue damage within seconds of injury (Davalos et al., 2005; Nimmerjahn et al., 2005; Hines et al., 2009), at least in part due to the activation of P2Y12 purinergic receptors by nucleotides (e.g., ATP) released into the extracellular space by damaged cells and reactive astrocytes (Davalos et al., 2005; Haynes et al., 2006), but perhaps also in response to serum components and/or extracellular glutamate released as a result of damage to endothelial and/or neural cells (Ransohoff and Perry, 2009). In response to injury, activated microglia proliferate, upregulate the expression of cytokines (e.g., interleukin-1 [IL-1] and tumour necrosis factor alpha [TNF?]) and chemokines (e.g., leucotrienes) and eicosanoids (e.g., prostaglandins), and these cells often take on a phagocytic phenotype that is indistinguishable from hematogenous macrophages (Hausmann, 2003; David and Kroner, 2011). The number of activated microglia peaks within the lesion site by 1 week post-injury and in the white matter by 3 weeks later in the rat (Popovich et al., 1997). At the same time, hematogenous macrophages begin to accumulate at the lesion site in the injured spinal cord, peaking around 5-7 days post-injury (Popovich et al., 1997; Carlson et al., 1998; Fleming et al., 2006). Unlike neutrophils, activated microglia and phagocytic macrophages (from resident microglia or hematogenous monocytes) persist in the injured spinal cord for weeks to months after SCI in both the rat and human (Popovich et al., 1997; Carlson et al., 1998; Fleming et al., 2006). Although these cells play important roles in the clearance of debris and tissue repair, they are also known to cause bystander damage in tissue adjacent to the lesion site, via the release of a variety of neurotoxic compounds including proteolytic enzymes, free radicals (e.g., nitric oxide [NO]), and pro-inflammatory cytokines (e.g., IL-1 and TNF?) (Chao et al., 1992; Hausmann, 35  2003; Gensel et al., 2009; David and Kroner, 2011). Following SCI, the macrophages found in the spinal cord predominantly display a pro-inflammatory or ?M1? phenotype (Mosser and Edwards, 2008; Kigerl et al., 2009). In tissues outside of the CNS, this pro-inflammatory (i.e., ?M1?) macrophage response eventually declines as M1 macrophages are replaced by anti-inflammatory or ?M2? macrophages (Mosser and Edwards, 2008; Laskin, 2009). However, this transition does not occur following SCI, where pro-inflammatory M1 macrophages continue to dominate the lesion indefinitely, thus contributing substantially to secondary tissue damage following SCI (Kigerl et al., 2009). T-lymphocytes also invade the injury site during the first week after SCI, predominantly within the lesion epicentre, and although their numbers remain elevated for 6-10 weeks after injury, the number of these cells remains relatively low compared to other inflammatory cells recruited after SCI (Popovich et al., 1997; Schnell et al., 1999; Sroga et al., 2003; Fleming et al., 2006). Lymphocytes can lyse oligodendrocytes directly (Antel et al., 1994) and/or induce apoptosis by releasing molecules that activate cellular death receptors (Almad et al., 2011), and although the deliberate addition of CNS-reactive T-cells has been shown to contribute to damage and functional deficits following SCI (Popovich et al., 1996a; Jones et al., 2002), the extent to which those cells contribute to naturally occurring secondary damage after SCI remains a matter of some debate (Popovich and Jones, 2003; Crutcher et al., 2006). 1.3.2.3 Cell death and degeneration following SCI Whereas primary injury results in the rapid necrotic death of many neurons, astrocytes and oligodendrocytes after SCI, secondary injury contributes even further to cell losses starting immediately after primary injury and continuing for weeks thereafter. For example, secondary 36  damage appears as early as 24 hours after contusion injury, as the spared/intact ventromedial white matter near the injury site shows a significant loss of axons, the development of severe myelin pathology, and a 50% reduction in the number of oligodendrocytes and astrocytes (Rosenberg and Wrathall, 1997; Grossman et al., 2001), but the apoptotic death of oligodendrocytes in the spared white matter adjacent to injury continues for at least 3 weeks following contusion in rats, and is thought to contribute substantially to the demyelination of intact axons that survived the initial trauma (Blight, 1983, 1985; Shuman et al., 1997; Totoiu and Keirstead, 2005; Almad et al., 2011). Apoptosis occurs in both gray and white matter, and appears to be most prominent in oligodendrocytes, neurons, and microglia, but can also effect astrocytes after severe SCI (Crowe et al., 1997; Liu et al., 1997; Lou et al., 1998; Yong et al., 1998; Casha et al., 2001). Following SCI apoptosis is thought to occur via multiple pathways, including extrinsic (receptor-mediated) apoptosis induced by activation of death receptors (e.g., Fas and p75) (Casha et al., 2001) and/or the production of inducible nitric oxide synthase (iNOS) by macrophages (Satake et al., 2000), and intrinsic (receptor-independent) apoptosis induced by mitochondrial damage (Eldadah and Faden, 2000) or direct activation of the caspase-3 proenzyme (Citron et al., 2000). The contribution of apoptosis to secondary damage is perhaps best demonstrated by evidence that the inhibition of caspases (mediators of apoptosis) reduces lesion size and improves motor performance after SCI in mice (Li et al., 2000). In addition to widespread necrosis and apoptosis, autophagic cell death has recently been reported in neurons, astrocytes, but especially oligodendrocytes for at least 21 days following SCI in mice (Kanno et al., 2009; Kanno et al., 2011) and has been implicated in the loss of neurons following contusion injury in rats (Chen et al., 2012a). Under normal or mild 37  pathological conditions, autophagy is thought to be a protective cellular response, as it allows for the removal of aggregated proteins and damaged organelles, including mitochondria, thereby promoting the health and longevity of cells (Mao and Reddy, 2010; Xu and Zhang, 2011). However, excessive autophagy can be triggered by severe cellular insults, such as those associated with SCI, and may actively contribute to axonal degeneration, and cellular atrophy and death (Cherra and Chu, 2008; Xu and Zhang, 2011). Besides contributing to the loss of cells due to secondary damage, apoptosis is known to continue for weeks and even months after SCI in adult rodents, monkeys, and even humans in axon tracts undergoing Wallerian degeneration (Katoh et al., 1996; Li et al., 1996; Crowe et al., 1997; Liu et al., 1997; Shuman et al., 1997; Emery et al., 1998; Lou et al., 1998; Yong et al., 1998; Abe et al., 1999; Li et al., 1999; Casha et al., 2001; Warden et al., 2001). Wallerian degeneration is a normal response to injuries that involve axonal transection in both the PNS and the CNS, and refers specifically to the disintegration of distal axon segments that have been cut off from their neuronal somata and the degradation of associated myelin sheaths (Vargas and Barres, 2007; Gaudet et al., 2011). In the PNS, Wallerian degeneration only takes 1-2 weeks to complete, and is thought to play an essential role in remodelling the environment of the distal nerve segment into one that supports the successful regeneration of injured axons (Griffin et al., 1992; George and Griffin, 1994; Gaudet et al., 2011), however, in the adult mammalian CNS, the rate of Wallerian degeneration is much slower, taking months to years to reach completion after SCI (Bignami and Ralston, 1969; Miklossy and Van der Loos, 1987; Perry et al., 1987; Miklossy and Van der Loos, 1991; George and Griffin, 1994; Becerra et al., 1995; Buss et al., 2004). As a result, the degeneration of axons and myelin and the loss of oligodendrocytes continue for a very 38  long time in injured axon tracts rostral and caudal to the injury site after SCI, but this process is distinct from secondary injury as the loss of axons and myelin from tracts that have already been severed by SCI does not appear to result in any additional meaningful loss of function (Vargas and Barres, 2007).  The response of neural cells to SCI 1.3.3In addition to cell death, the responses of local stem/progenitor cells, neurons, oligodendrocytes, and astrocytes also play a role in determining the extent of secondary damage and/or spontaneous recovery following SCI. 1.3.3.1 Spinal cord stem/progenitor cells The existence of multiple populations of neural stem/progenitor cells (NSPCs) or oligodendrocyte progenitor cells (OPCs) in the adult mammalian spinal cord is well established, with proposed niches in the ependymal, sub-pial, and intraparenchymal regions (Johansson et al., 1999; Namiki and Tator, 1999; Horner et al., 2000; Yamamoto et al., 2001a; Martens et al., 2002; Barnabe-Heider et al., 2010; Petit et al., 2011). Many of these cells are known to proliferate extensively and rapidly in response to SCI (Johansson et al., 1999; Namiki and Tator, 1999; Yamamoto et al., 2001a; Mothe and Tator, 2005; Zai and Wrathall, 2005; Barnabe-Heider et al., 2010), beginning as early as 24 hours after injury (Horky et al., 2006), generally peaking within 1 week post-injury (wpi), and subsiding to insignificant levels by 4 wpi (McTigue et al., 2001; Zai and Wrathall, 2005; Zai et al., 2005). NSPCs can be isolated and differentiated into neurons, astrocytes, and oligodendrocytes under appropriate culture conditions (Weiss et al., 1996; Shihabuddin et al., 1997; Yamamoto et al., 2001a; Kulbatski et al., 2007; Mothe et al., 39  2011), and the transplantation of human NSPCs from spinal cord ependyma after SCI in the rat has been shown to produce neurons as well as glia in vivo (Mothe et al., 2011). However, despite all of the evidence that NSPCs in the adult mammalian spinal cord are capable of generating neurons, spontaneous neurogenesis largely fails to occur in the adult mammalian spinal cord under normal conditions (Horner et al., 2000; Horky et al., 2006) or following injury (Yamamoto et al., 2001b; Mothe and Tator, 2005; Zai and Wrathall, 2005; Horky et al., 2006). Instead spinal progenitor cells preferentially generate glia, and thus replace lost astrocytes and/or oligodendrocytes, but not neurons following SCI (Johansson et al., 1999; Horner et al., 2000; Martens et al., 2002; Mothe and Tator, 2005; Zai and Wrathall, 2005; Horky et al., 2006; Yang et al., 2006; Vessal et al., 2007; Meletis et al., 2008). Thus, despite the fact that approximately 50% of the oligodendrocytes and astrocytes in the spared white matter die within 24 hours after contusion injury (Grossman et al., 2001), glial densities generally return to normal levels in the spared tissue rim between 2-6 weeks post-injury (Wrathall et al., 1998; Frei et al., 2000; Rosenberg et al., 2005; Tripathi and McTigue, 2007), and this occurs concurrent with ongoing apoptosis in those regions (Crowe et al., 1997; Liu et al., 1997; Yong et al., 1998; Li et al., 1999; Warden et al., 2001). 1.3.3.2 Glial scar formation and astrocyte contributions to secondary injury Although many astrocytes die during primary and secondary injury (Jaeger and Blight, 1997; Yong et al., 1998; Grossman et al., 2001), surviving astrocytes in the vicinity of the lesion site respond to SCI by undergoing reactive astrogliosis. That process involves proliferation followed by hypertrophy and increased production of glial fibrillary acidic protein (GFAP) and chondroitin sulfate proteoglycans (CSPGs) by astrocytes at the site of injury (Eng and Ghirnikar, 40  1994; Fitch and Silver, 1997; Popovich et al., 1997; Stichel and Muller, 1998; Fawcett and Asher, 1999; Schnell et al., 1999). The number of reactive astrocytes in the lesion site increases over time and these cells surround and enclose the injury site, forming a physical barrier commonly referred to as the glial scar, which walls off areas of necrosis from the surrounding spared parenchyma (Faulkner et al., 2004; Okada et al., 2006; Wanner et al., 2013) and acts, in combination with increased CSPG production, to inhibit axonal regeneration through the site of injury (Reier et al., 1983; Davies et al., 1997; Popovich et al., 1997; Davies and Silver, 1998; Davies et al., 1999; Fawcett and Asher, 1999; McKeon et al., 1999; Hermanns et al., 2001; Bradbury et al., 2002; Silver and Miller, 2004). In addition, newly generated astrocytes that arise from spinal cord stem/progenitor cells also contribute to glial scar formation (Johansson et al., 1999; Mothe and Tator, 2005; Meletis et al., 2008; Barnabe-Heider et al., 2010), and some of those cells appear to generate a transient population of phagocytic astrocytes early after hemisection SCI; at least in mice (Sellers et al., 2009). The formation of the glial scar is thought to limit the spread of hematogenous inflammatory cells and help restore the BSCB, and therefore astrocytes are generally thought to play a protective role following SCI as they limit the spread of damage into adjacent uninjured tissue (Fitch and Silver, 1997; Stichel and Muller, 1998; McGraw et al., 2001; Okada et al., 2006; Sofroniew, 2009). However, the astrocyte response to SCI appears to come at the expense of axonal regeneration, as the end result of reactive astrogliosis is a lesion site that is much less accessible to regenerating CNS axons (Silver and Miller, 2004; Busch and Silver, 2007). In addition to the formation of the glial scar, and in spite of numerous potentially protective functions served by these cells (e.g., uptake of excitatory glutamate), astrocytes have 41  also been implicated in the exacerbation of secondary injury, particularly due to their role in modulating injury-induced inflammation (Swanson et al., 2004; Farina et al., 2007). Following injury astrocytes release ATP to trigger microglial activation (Davalos et al., 2005) and produce a variety of chemokines (e.g., CXCL10 and CCL2) and cytokines (e.g., TNF?, IL-1 and IL-6) that further propagate the inflammatory response by promoting inflammatory cell infiltration and proliferation (Swanson et al., 2004; Brambilla et al., 2005; Brambilla et al., 2009). Indeed, transgenic inhibition of the transcription factor nuclear factor kappa B (NF-?B; a key regulator of inflammation and secondary injury processes after SCI) specifically in astrocytes is associated with reduced inflammation, and enhanced neuroprotection and functional recovery after contusive SCI in mice (Brambilla et al., 2005). Astrocytes are also implicated in the development of both vasogenic and cytotoxic edema, because they secrete the angiogenic factor vascular endothelial growth factor A (VEGF-A) and matrix metalloproteases (e.g., MMP2), which are both known to contribute directly to BSCB disruption (Swanson et al., 2004; Argaw et al., 2009), and they express aquaporin-4 which is elevated after SCI, making astrocytes highly susceptible to cytotoxic edema, which can spread to adjacent cells via gap junctions and may lead to increased vasogenic edema (Swanson et al., 2004; Liang et al., 2007; Saadoun and Papadopoulos, 2010). In addition, astrocytes show enhanced NO production in the presence of elevated transforming growth factor ?1 (TGF-?1) (Hamby et al., 2006), and they release glutamate rather than taking it up in the presence of elevated TNF? (Bezzi et al., 2001). Given that both astrocytes and microglia produce TGF-?1 and TNF? following injury (John et al., 2003), the release of NO and glutamate from astrocytes likely contributes to ongoing excitotoxicity and free radical and lipid peroxidation damage after SCI. 42  1.3.3.3 The oligodendrocyte response to SCI: Demyelination vs. Remyelination In contrast to the resilient and multifaceted (although somewhat detrimental) response to injury displayed by astrocytes, the primary responses of oligodendrocytes following injury appear to be necrosis, apoptosis and quiescence (Ludwin, 1990; Crowe et al., 1997; Liu et al., 1997; Li et al., 1999; Casha et al., 2001; Almad et al., 2011). Oligodendrocytes are particularly susceptible to oxidative stress and glutamate excitotoxicity during ischemia (Merrill et al., 1993; Oka et al., 1993; Noble et al., 1994; Husain and Juurlink, 1995; Pantoni et al., 1996; McDonald et al., 1998b; McDonald et al., 1998a; Li and Stys, 2000; Matute et al., 2007), and also appear to be vulnerable to elevated extracellular ATP (Wang et al., 2004; Matute et al., 2007; Huang et al., 2012) as well as elevated levels of proinflammatory cytokines (e.g., TNF? and IL-1?) (Vartanian et al., 1995; Curatolo et al., 1997; Sherwin and Fern, 2005; Li et al., 2008; Steelman and Li, 2011). In addition, oligodendrocytes are known to express cell surface ?death receptors? including Fas and p75, the activation of which appear to be involved in delayed apoptosis in white matter tracts following SCI (Casha et al., 2001; Beattie et al., 2002), and although those receptors are also expressed on astrocytes, the latter are largely resistant to death receptor-induced apoptosis (Song et al., 2006). As a result of these vulnerabilities, large numbers of oligodendrocytes are lost following SCI, with 93% of those cells dying within 7 days post-injury at the lesion epicentre (McTigue et al., 2001), and ongoing apoptosis in the adjacent degenerating white matter tracts for 3-6 weeks after contusion injuries in rats and monkeys (Crowe et al., 1997; Shuman et al., 1997) and at least 60 days after SCI in humans (Emery et al., 1998). Indeed, the majority of apoptosis of neural cells following SCI appears to involve oligodendrocytes rather than neurons, particularly during Wallerian degeneration of injured axon tracts in the white matter rostral and caudal to injury (Crowe et al., 1997; Liu et al., 1997; Shuman et al., 1997; Almad et al., 2011). 43  Although primary and secondary injury both cause direct damage to myelin sheaths, the loss of oligodendrocytes due to prolonged apoptosis in spared white matter tracts is thought to largely contribute to the demyelination of intact axons adjacent to the lesion site following SCI (Crowe et al., 1997; Shuman et al., 1997; Casha et al., 2001); particularly considering that each oligodendrocyte myelinates segments on 30-80 different axons (Chong et al., 2012). Demyelination is a prominent feature after SCI in both animals (Gledhill et al., 1973a; Bresnahan et al., 1976; Blight, 1983, 1985; Totoiu and Keirstead, 2005; Smith and Jeffery, 2006; Siegenthaler et al., 2007; Lasiene et al., 2008; James et al., 2011; Powers et al., 2012), and humans (Bunge et al., 1993; Kakulas, 1999; Norenberg et al., 2004; Guest et al., 2005). However, spontaneous remyelination is also a consistent finding after SCI, typically beginning around 2-3 weeks post-injury (Bunge et al., 1961; McDonald and Ohlrich, 1971; Gledhill et al., 1973a; McDonald, 1975; Harrison and McDonald, 1977; Kakulas, 1999; Totoiu and Keirstead, 2005; Smith and Jeffery, 2006; Lasiene et al., 2008; Powers et al., 2012), and is associated with the restoration of saltatory conduction with normal or near-normal action potential conduction properties (Smith et al., 1979; James et al., 2011). Remyelinated axons have traditionally been identified by abnormally thin or short myelin sheaths (i.e., larger g-ratios [axon diameter / total fibre thickness (including myelin)] and shorter internodal lengths), largely based on electron microscopy (EM) work from the 1970s and 1980s, describing presumptive remyelinated CNS fibres a few weeks after demyelination (Bunge et al., 1961; Gledhill and McDonald, 1977). However, recent work using an inducible membrane-bound reporter to label glial progenitor cells in the injured spinal cord of transgenic mice allowed for more definitive identification of regenerated myelin after SCI, and revealed that newly generated oligodendrocyte myelin sheaths thicken over time, approaching uninjured values (based on g-ratio) within 6 months post-injury, 44  but remaining significantly thinner than newly generated SC myelin sheaths (Powers et al., 2013). Thus, newly formed oligodendrocyte myelin sheaths do not remain thinner than the myelin sheaths formed during natural development (Powers et al., 2013). Despite ongoing remyelination, the presence of denuded axons in the spinal cords of chronically injured animals, and to lesser extent in human tissue, has long been viewed as evidence that endogenous remyelination is insufficient following SCI, and thus represents a therapeutic target even in chronic SCI (Gledhill et al., 1973b; Blakemore, 1974; Gledhill and McDonald, 1977; Blight, 1983; Bunge et al., 1993; Guest et al., 2005; Totoiu and Keirstead, 2005). However, more recent evidence indicates that the techniques used to estimate the number of demyelinated fibres in earlier work likely included demyelinated segments of degenerating/severed axons, thereby overestimating the extent of demyelination that is functionally relevant. Indeed, all intact/functional axons are remyelinated by 3 months after SCI in mice and rats, which suggests that only injured / degenerating / dysfunctional axons remain demyelinated in the chronic injury setting (Lasiene et al., 2008; Powers et al., 2012). Thus, although demyelination may persist in the chronic SCI setting, it appears to be a target of limited utility, as the only axons that need remyelinating are those that are damaged or dysfunctional (Powers et al., 2012). Given the injured axons are likely to degenerate over time, this may explain why only limited demyelination has been observed in most human tissue from patients with chronic SCI examined years/decades after injury, except in conditions involving some degree of ongoing cord compression (Bunge et al., 1993; Gensert and Goldman, 1997; Kakulas, 1999; Norenberg et al., 2004; Guest et al., 2005). 45  Mature oligodendrocytes are post-mitotic, and thus unable to divide to replace the cells that are lost following SCI (Keirstead and Blakemore, 1997), but new myelinating oligodendrocytes are generated by one of the resident spinal cord progenitor populations, known as oligodendrocyte progenitor cells (OPCs) or polydendrocytes. OPCs are typically identified by the expression of nerve/glial antigen 2 (NG2) or platelet-derived growth factor receptor alpha (PDGFR?) (Nishiyama et al., 2009; Trotter et al., 2010). These cells are present throughout the gray and white matter in the adult CNS, and respond to demyelination by proliferating, migrating to areas of demyelination, and remyelinating denuded axons (Wolswijk and Noble, 1989; Carroll et al., 1998; Blakemore and Keirstead, 1999; Keirstead and Blakemore, 1999; Watanabe et al., 2002; Lytle et al., 2009). In addition to newly generated oligodendrocytes, endogenous SCs are also well known to play a role in remyelination in the injured CNS, particularly following contusion or in demyelinating injury/disease states (Bunge et al., 1993; McTigue et al., 2001; Guest et al., 2005; Totoiu and Keirstead, 2005). While traditionally those cells were assumed to have migrated into the injured spinal cord from the periphery, recent evidence suggests that the majority of the endogenous myelinating SCs found in the CNS after toxin-induced focal demyelination actually arise from local CNS precursors, rather than peripheral origins (Zawadzka et al., 2010). Although spontaneous remyelination by OPCs and SCs ensures that the vast majority of intact axons do not remain permanently denuded after SCI, remyelination by endogenous cells is delayed following SCI, so demyelination is quite prevalent during the first couple of weeks following SCI (Gledhill et al., 1973a; Harrison and McDonald, 1977; Blight, 1985), and appears to affect a large number of axons in spared white matter tracts, given that 53% of spared 46  rubrospinal tract axons showed signs of remyelination following contusion injury in the rat (Powers et al., 2012). Demyelination has long been known to result in the impairment of rapid axonal conduction (Bostock and Sears, 1976; Bostock et al., 1978), but more recent evidence indicates that myelinating glia are also required for the maintenance of normal axon transport and long-term survival in long projecting axons (Nave and Trapp, 2008; Nave, 2010a, b). Furthermore, recent work has demonstrated that demyelination and/or the depletion of oligodendrocytes are both associated with increased axonal degeneration, and that efficient remyelination is protective against demyelination-associated axonal damage and degeneration (Irvine and Blakemore, 2008; Pohl et al., 2011). Even less severe alterations of oligodendrocyte function, such as the deletion of certain myelin proteins (e.g., proteolipid protein [PLP]; 2?, 3?- cyclic nucleotide 3?-phosphodiesterase [CNPase]; myelin associated glycoprotein [MAG]) (Griffiths et al., 1998; Sheikh et al., 1999; Lappe-Siefke et al., 2003; Pan et al., 2005; Petzold, 2005; Nguyen et al., 2009) or the disruption of oligodendrocyte peroxisomes (Kassmann et al., 2007) have been found to lead to axonal degeneration, indicating the critical nature of oligodendrocyte support for axons under normal conditions. In the pathological environment of the injured spinal cord the dependence of axons on oligodendrocyte-derived support is probably even greater, as bare axons are likely to be more vulnerable to toxic insults and to require increased metabolic support (Redford et al., 1997; Dutta and Trapp, 2007; Nave and Trapp, 2008; Nave, 2010a, b). Thus, although demyelination appears to be a transient phenomenon after SCI, even temporary demyelination increases the vulnerability of axons to secondary damage, and may therefore contribute to the early loss of intact axons in the spared white matter following injury (Lasiene et al., 2008; Almad et al., 2011; Powers et al., 2012). As such, facilitating remyelination in the acute and sub-acute setting after SCI remains a viable 47  therapeutic target, as earlier remyelination may prevent the loss of some axons due to secondary injury and thereby preserve some neurological function. This may explain why increased remyelination following NSPC transplantation in sub-acute SCI is often correlated with improved functional recovery in animal models (Cao et al., 2005; Cummings et al., 2005; Hofstetter et al., 2005; Keirstead et al., 2005; Lee et al., 2005; Mitsui et al., 2005; Karimi-Abdolrezaee et al., 2006). In addition to facilitating axonal degeneration, demyelination results in the accumulation of myelin debris for many months following SCI, and in light of the protracted nature of Wallerian degeneration in the CNS (Vargas and Barres, 2007), that debris tends to linger in degenerating axon tracts for a very long time, persisting for months and even years after SCI (Perry et al., 1987; George and Griffin, 1994; Becerra et al., 1995; Buss et al., 2004). Myelin debris is known to inhibit the differentiation of OPCs, which may partly explain the lag in spontaneous remyelination following SCI (Kotter et al., 2006; Baer et al., 2009; Plemel et al., 2013), as the rate of remyelination in the CNS correlates with the rate of myelin clearance (Shields et al., 1999; Zhao et al., 2006; Ruckh et al., 2012). In addition, myelin contains a variety of proteins that are known to inhibit axonal regeneration, including: Nogo-A (Chen et al., 2000; GrandPre et al., 2000; Prinjha et al., 2000), MAG (McKerracher et al., 1994; Mukhopadhyay et al., 1994), oligodendrocyte-myelin glycoprotein (OMgp) (Wang et al., 2002), repulsive guidance molecule (RGMa) (Hata et al., 2006), ephrinB3 (Benson et al., 2005), semaphorin4D (Moreau-Fauvarque et al., 2003), and netrin-1 (L?w et al., 2008). As such, myelin debris makes the environment of the injured CNS less favourable to axonal growth and thus contributes to the 48  failure of axonal regeneration following CNS injury (Filbin, 2003; Schwab, 2004; Liu et al., 2006; Yiu and He, 2006; Cafferty et al., 2010). 1.3.3.4 Neuronal and axonal responses to SCI: The failure of CNS regeneration The neuronal response to SCI is largely similar to that of oligodendrocytes, as primary injury generally results in widespread necrosis of neurons, and both apoptosis and autophagy contribute to more extensive cell losses during secondary injury (Crowe et al., 1997; Liu et al., 1997; Lou et al., 1998; Yong et al., 1998; Kanno et al., 2009; Kanno et al., 2011; Chen et al., 2012a). Unlike lost oligodendrocytes and astrocytes, NSPCs fail to spontaneously replace lost neurons following SCI (see above). In addition to the death of neurons, SCI is well associated with widespread axonal injury and degeneration. Primary injury causes the physical disruption of axons that pass through the lesion site, whereas secondary injury contributes to damage to axons adjacent to the site of the initial trauma, leading to the transection of axons in the ascending and descending axon tracts that pass through the level of injury in the white matter adjacent to injury  (Schwab and Bartholdi, 1996). Although the response of CNS axons to transection can vary substantially (Hawthorne et al., 2011), most axons show a similar pattern of behavior following SCI, wherein the distal segments of severed axons undergo protracted Wallerian degeneration and the proximal segments persist near the lesion site, displaying dystrophic, swollen endbulbs that lack filopodia and likely provide no meaningful function (Coleman and Perry, 2002; Gu?zar-Sahag?n et al., 2004; Tom et al., 2004). While some populations of axons (e.g., serotonergic fibres) are known to persist at the lesion edge and show considerable spontaneous sprouting, the proximal segments of transected cortical axons typically retract from the lesion site (Houle and Jin, 2001; Hawthorne et al., 2011).  49  Thus, in contrast to the outcome of PNS injury, spontaneous axonal regeneration fails following injury to the adult mammalian CNS. The failure of CNS regeneration is thought to arise from a number of factors, including: 1) the lack of an adequate intrinsic regenerative response in mature CNS neurons (Cai et al., 2001; Bulsara et al., 2002; Harel and Strittmatter, 2006; Szpara et al., 2007); 2) the lack of growth permissive substrates at the lesion site, including the presence of gaps and/or large cystic cavities (McDonald, 1999); 3) the formation of the inhibitory glial scar (Silver and Miller, 2004; Yiu and He, 2006; Busch and Silver, 2007); 4) the persistence of inhibitory myelin proteins, including Nogo, MAG, and OMgp (McGee and Strittmatter, 2003; Liu et al., 2006; Cafferty et al., 2010); 5) the lack of adequate trophic support, which is thought to contribute to retrograde atrophy of neurons whose axons are damaged (Kobayashi et al., 1997; Jones et al., 2001; Hains et al., 2003; Zhou and Snider, 2006); and 6) the lack of appropriate spatial and temporal gradients of growth factors and guidance cues necessary to stimulate and guide growing axons to their appropriate targets (Alto et al., 2009; Blesch and Tuszynski, 2009). Although the relative contribution of the various growth inhibitory elements present in the CNS remains unclear, it was recently demonstrated that CSPGs produced by reactive astrocytes can act via Nogo receptors (NgR1 and NgR3), indicating that CSPGs and myelin proteins may at least partly use shared mechanisms to exert their inhibitory effects (Dickendesher et al., 2012). In addition, to myelin proteins and CSPGs, myelin and the glial scar are also known to contain other inhibitory guidance molecules, such as RGM (Schwab et al., 2005a; Schwab et al., 2005b; Hata et al., 2006), ephrins/Ephs (Goldshmit et al., 2004; Benson et al., 2005), semaphorins (Moreau-Fauvarque et al., 2003; Kantor et al., 2004); netrins (L?w et al., 2008) and slits (Wehrle et al., 2005), all of which are known/thought to contribute to the inhibition of CNS axon regeneration (Schwab et al., 2006; Cafferty et al., 2010). 50  The lack of regeneration following SCI eventually leads to atrophy of the cell bodies of many long projecting axons that reside in the brain or in the spinal cord below the level of injury, despite the fact that those neuronal somata are well removed from the site of injury (Kobayashi et al., 1997; Kwon et al., 2002b). Indeed, in humans, SCI is often associated with notable atrophy of the primary motor and sensory cortices (Freund et al., 2011). Neuronal atrophy leads to decreased responsiveness to growth factors and other treatments delivered to the site of injury, which makes promoting regeneration increasingly difficult in the chronic setting (Kwon et al., 2004). However, not all neurons atrophy following injury, and in contrast to the situation at the site of injury, spontaneous growth and sprouting of both injured and uninjured CNS axons has been found to occur after incomplete SCI. For example, following hemisection or discrete transection of the corticospinal tract (CST), spontaneous plasticity in the CST or rubrospinal tract (RST) has been found to occur, resulting in the formation of new neural circuits that bypass the lesion site in mice (Steward et al., 2008), rats (Raineteau et al., 2002; Bareyre et al., 2004), and monkeys (Rosenzweig et al., 2010). This plasticity is thought to play a key role in the spontaneous, and often substantial, recovery of locomotor function that can occur following incomplete SCI in both animals and humans (Little and Halar, 1985; Roth et al., 1991; Rossignol et al., 1999; Courtine et al., 2005), and at least in rats, the transection of newly formed circuits has been shown to result in the loss of recovered electrophysiological and behavioral functions (Bareyre et al., 2004). In parallel with neuronal plasticity in the spinal cord, there is also a growing body of evidence that indicates substantial cortical plasticity occurs following SCI, particularly involving the functional reorganization of the sensorimotor cortex, in both rats and humans (Bareyre et al., 2004; Freund et al., 2011; Henderson et al., 2011). 51   Long-term pathophysiological outcomes of SCI 1.3.4Although the initial spread of secondary damage is rapid, resulting in substantial and irreversible damage to the gray matter within 1 hour of injury and the white matter within 72 hours of injury following contusive SCI, the expansion of the lesion site slows over time and eventually stops (Blight and Young, 1989; Dumont et al., 2001). Neurogenic shock subsides and the vasculature stabilizes, leading to relatively normal tissue perfusion and adequate oxygen and glucose to once again support aerobic respiration (Dumont et al., 2001). Although the reperfusion of ischemic tissue is known to initially exacerbate the loss of cells due to reperfusion injury, more extensive cellular damage would likely occur if ischemia was left unchecked. The permeability of the BSCB returns to normal in intact regions by 1-2 months post-injury (Bartanusz et al., 2011), which helps to prevent further edema. Edema itself eventually subsides after SCI, although precisely how that occurs remains a mystery, as little is known about the mechanisms of excess fluid elimination from the spinal cord (Saadoun and Papadopoulos, 2010). While the initially massive cellular inflammatory response eventually subsides, macrophages and reactive microglia are often found in the spinal cord months and even years after injury ? particularly in regions of ongoing Wallerian degeneration (Schmitt et al., 2000; Popovich and Jones, 2003; Buss et al., 2004; Norenberg et al., 2004). As the production of free radicals decreases, it eventually reaches a point where the level of antioxidants is sufficient to maintain balance, thus reducing/eliminating oxidative stress. Excess levels of excitatory amino acids are eventually cleared from the extracellular space, as is myelin and other debris; although some portion of debris is commonly found in regions of ongoing Wallerian degeneration, where degenerating and severed axons with denuded segments persist after SCI, even decades later in 52  some human cases (Gledhill et al., 1973a; Norenberg et al., 2004; Guest et al., 2005; Totoiu and Keirstead, 2005). In most, if not all types of SCI, secondary injury actually results in more tissue damage than primary injury (Hausmann, 2003). The combined result of those two processes is often the loss of a substantial portion of tissue, as indicated by a 30% reduction in cord area according to MRI in human SCI patients; indicative of gross atrophy of the spinal cord following injury (Freund et al., 2011). In injuries involving contusion/compression (including dislocation), primary and secondary damage typically culminate in the formation of an oval-shaped fluid-filled cyst/cavity that extends 2-3 spinal segments surrounded by an inhibitory glial scar and a rim of spared/intact white matter that is far from normal, as it contains a reduced density of axons, particularly those with large (? 5?m) diameter, abnormal myelin sheaths (due to remyelination), and often contains macrophages, and myelin debris (Blight, 1985, 1991, 1992; Beattie et al., 1997; Shuman et al., 1997; Bunge, 2001; Silver and Miller, 2004; Rosenberg et al., 2005; Dietz and Curt, 2006) The extent of damage following contusion typically includes a degree of peripheral degeneration, which is thought to contribute to the resultant functional deficits, but commonly receives little attention in experimental treatment paradigms (Dietz and Curt, 2006; Van De Meent et al., 2010). In contrast to contusion injuries, transections/hemisections of the spinal cord are characterized by limited rostrocaudal spread of secondary injury, and thus tend to result in a much smaller lesion overall (Hausmann, 2003) and much less extensive demyelination (Siegenthaler et al., 2007). In addition, transections, hemisections, and crush injuries limited to specific tracts (e.g., dorsolateral funiculus crush used here) all involve damage to the dura and 53  meninges, and thus allow for more rapid and increased invasion by peripheral monocytes/macrophages, as well as extensive infiltration of the spinal cord by cells that are normally restricted to the meninges, such as fibroblasts (Hausmann, 2003; Norenberg et al., 2004); the latter of which create a mesenchymal/fibroblastic scar composed mainly of collagen that impedes axonal growth even more than the astroglial scar (Berry et al., 1983).  In the face of the extensive loss of tissue caused by SCI, the delay in cell replacement (despite the presence of endogenous spinal cord stem/progenitor cells), and the failure of CNS axonal regeneration, most of the damage caused by SCI is permanent, and thus results in persistent functional deficits, the range and extent of which are determined by the particular neural circuits left disrupted or dysfunctional after injury (Dietz and Curt, 2006). However, some degree of spontaneous functional recovery is common in the weeks and months following SCI in both animals and humans (Onifer et al., 2011), and is primarily thought to occur via three mechanisms: 1) the resolution of spinal shock (Holaday and Faden, 1983; Hiersemenzel et al., 2000); 2) remyelination of denuded intact/spared axons (Blight and Young, 1989; Gensert and Goldman, 1997); 3) the creation of new functional neural circuits via the sprouting of injured/uninjured axons both in the spinal cord, brainstem, and cortex (Bareyre et al., 2004; Bareyre, 2008). Although spontaneous recovery can be substantial following incomplete SCI, particularly for less severe injuries, even with the addition of modern surgical and rehabilitative interventions, recovery eventually plateaus in all patients, typically within the first year (Lorenz et al., 2012), so the vast majority of people with SCI are left with significant functional impairments and lifelong disability.  54   Therapeutic strategies for promoting improved neurological outcome after SCI  1.4There is nothing that can be done to stop the immediate damage produced by the initial trauma that causes SCI, so research is generally focused on developing treatments that act to limit the spread of secondary damage (i.e., neuroprotective strategies) and/or promote tissue repair (i.e., neurorepair strategies) during or subsequent to secondary injury. The most promising treatments for SCI are generally those that provide multiple neuroprotective and/or neuroreparative benefits, and cellular transplantation is one such multidimensional therapeutic strategy that has garnered considerable attention in the field of SCI (for reviews see: Bunge and Pearse, 2003; Tator, 2006; Fehlings and Vawda, 2011; Tetzlaff et al., 2011; Mothe and Tator, 2012; Ruff et al., 2012).   Neuroprotective strategies for SCI   1.4.1Examples of neuroprotective strategies tested in animals models or humans with SCI include: 1) surgical decompression and spinal stabilization to minimize physical tissue damage (Dolan et al., 1980; Guha et al., 1987; La Rosa et al., 2004; Carreon and Dimar, 2011); 2) management of blood flow to minimize ischemia (Guha et al., 1989; Tator, 1992); 3) preservation of vascularity (e.g., promotion of endothelial survival) to minimize hemorrhage and edema (Han et al., 2010; Lutton et al., 2012; Simard et al., 2012); 4) prevention or reversal of membrane and cytoskeletal damage to limit axonal degeneration, cell death and/or atrophy (Borgens