UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Axonal regeneration and functional recovery in the chick following embryonic spinal cord injury Hasan, Sohail Jamil 1992

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

Item Metadata

Download

Media
831-ubc_1992_fall_hasan_sohail_jamil.pdf [ 16.03MB ]
Metadata
JSON: 831-1.0086545.json
JSON-LD: 831-1.0086545-ld.json
RDF/XML (Pretty): 831-1.0086545-rdf.xml
RDF/JSON: 831-1.0086545-rdf.json
Turtle: 831-1.0086545-turtle.txt
N-Triples: 831-1.0086545-rdf-ntriples.txt
Original Record: 831-1.0086545-source.json
Full Text
831-1.0086545-fulltext.txt
Citation
831-1.0086545.ris

Full Text

AXONAL REGENERATION AND FUNCTIONAL RECOVERY IN THE CHICKFOLLOWING EMBRYONIC SPINAL CORD INJURYBySOHAIL JAMIL HASANB.Sc., Daihousie University, 1986A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThE FACULTY OF GRADUATE STUDIES(NEUROSCIENCE)We accept this thesis as conformingto the required standTHE UNIVERSITY OF BRITISH COLUMBIAJuly 1992© Sohail Jamil Hasan, 1992Signature(s) removed to protect privacyIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature___________________________Department of___________________The University of British ColumbiaVancouver, CanadaDate CCT S1Z.DE-6 (2/88)Signature(s) removed to protect privacy11ABSTRACTThe embryonic central nervous system (CNS) is more plastic than adult CNS, andthe early embryonic brain and spinal cord has been suggested to recover more readily from asevere injury. The following studies were designed to determine the stages of chicken(Gallus domesticus) embryonic development during which descending brainstem-spinaltracts maintain their capacity for anatomical and functional repair after complete thoracicspinal cord transection. Brainstem-spinal neurons begin to project their axons through thespinal cord at approximately embryonic day (E)4, and these projections are essentiallycomplete to the lumbar level by E12. Disruption of these brainstem-spinal pathways wasproduced by complete thoracic spinal cord transections at different developmental ages fromE3 through E14. Control sham operations were conducted in parallel. The post-operativeembryonic recovery period varied from 5 to 8 days. Following recovery, the extent ofanatomical repair was assessed by injecting a fluorescent dye into the lumbar spinal cordcaudal to the transection site. Brainstem tissue sections were subsequently examined for thepresence of retrogradely labelled brainstem-spinal neurons. Anatomical results indicatedsimilar distributions of retrogradely labelled neurons within the brainstem of both shamtransected controls and embryos transected prior to E13 (Hasan et al., 1991, 1992).The neuroanatomical recovery shown by chick embryos can be attributed either toaxonal regeneration of previously severed axons or to the subsequent development of newaxonal projections from later developing neurons. In order to address this issue, theembryonic lumbar spinal cord was injected before and after thoracic transection with twodifferent retrograde tract tracing fluorescent dyes. Co-localization of both labels within thesame brainstem-spinal neuron would be indicative of regeneration of previously axotomizedprojections rather than the subsequent development of new axonal projections. Findingsindicated that there were double-labelled brainstem—spinal neurons after a transection prior toE13 and the number of double-labelled brainstem-spinal neurons decreased after an E13-E15transection. In addition, at each subsequent stage of development from E1O-E12, a higher111ratio of double-labelled brainstem-spinal neurons (indicating regeneration of previouslysevered axons) to the number of cell bodies labelled with the second fluorescent tracer alone(indicating possible subsequent development) was observed. This would suggest that duringsuccessive stages of development, regeneration of previously axotomized fibers increasinglycontributes to the observed anatomical and functional recovery after thoracic cordtransections prior to E13 (Hasan et al., 1992).Functional recovery was assessed by focal electrical stimulation of identifiedbrainstem locomotor regions in transected or sham-transected E18-E20 embryos. Legmuscle electromyographic (EMG) recordings were used to monitor brainstem stimulatedlocomotor activity (Valenzuela et al., 1990; Hasan et al., 1991, 1992). Functional repair wasevident among E18-E20 embryos that had had their spinal cords transected prior to E13 andthis brainstem evoked locomotion was indistinguishable from brainstem-evoked locomotionin control (sham-transected or untransected) embryos. In addition, voluntary open-fieldlocomotion and brainstem evoked locomotion in hatchling chicks transected prior to E13 wasindistinguishable from that observed in control hatchlings, indicating that completefunctional recovery had occurred. Embryos and hatchling chicks transected on or after E13showed reduced functional repair abilities (Hasan et al., 1991, 1992).ivTABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS ivLIST OF TABLES vLIST OF FIGURES viLIST OF ABBREVIATIONS viiGeneral Abbreviations viiNeuroanatomical Abbreviations viiiACKNOWLEDGEMENTS ixCHAPTER 1 - General introduction 1CHAPTER 2 - Anatomical repair of transected spinal cord in embryonic chick(single-labelling studies) 9INTRODUCTION 10MATERIALS AND METHODS 12RESULTS 16DISCUSSION 35CHAPTER 3 - Anatomical repair of transected spinal cord in embryonic chick(double-labelling studies) 41INTRODUCTION 42MATERIALS AND METHODS 44RESULTS 51DISCUSSION 71CHAPTER 4- Functional repair of transected spinal cord in embryonic chick(physiological studies) 77INTRODUCTION 78MATERIALS AND METHODS 81RESULTS 90DISCUSSION 111CHAPTER 5 - General discussion 113REFERENCES 121VLIST OF TABLESTable 2-1. Single-label summary (permissive transects). 32Table 2-2. Single-label summary (restrictive transects). 33Table 3-1. Single/double-label summary (permissive/restrictive transects). 52Table 4-1. Locomotor activity summary (embryos). 91Table 4-2. Locomotor activity summary (hatchlings). 102viLIST OF FIGURESFigure 2-1. Spinal cord transection and injection sites (controls). 17Figure 2-2. Lack of dye diffusion (control). 19Figure 2-3. RPgc single-labelling. 22Figure 2-4. VeL single-labelling. 24Figure 2-5. Schematic summary drawings (single-labelling). 26Figure 3-1. Schematic representatation of experimental procedure (double-labelling). 45Figure 3-2. RPgc double-labelling. 54Figure 3-3. RPgc double-labelling II. 57Figure 3-4. VeL double-labelling. 59Figure 3-5. Schematic summary drawings (double-labelling). 61Figure 3-6. RPgc double-labelling control. 69Figure 4-1. Schematic representation of embryonic physiology procedure. 82Figure 4-2. Schematic representation of hatchling physiology procedure. 87Figure 4-3. Stimulation and electrolytic lesion sites. 92Figure 4-4. Evoked locomotor activity in embryonic chicks. 95Figure 4-5. Dependence of in ovo evoked stepping frequency on stimulation currentintensity. 97Figure 4-6. Evoked locomotor activity in hatchling chicks. 103Figure 4-7. EMG5 in hatchling chicks. 105Figure 4-8. Summary of EMG activity patterns. 107Figure 4-9. Burst duration versus cycle duration. 109viiLIST OF ABBREVIATIONSGeneral AbbreviationsBDNF Brain derived neurotrophic factorCAM Cell adhesion moleculeCb CerebellumCBDA Cascade Blue-labelled dextran-amineC CelsiusCC Canalis centralisCF Caudioflexorius muscleCNS Central nervous systemcm CentimeterE Embryonic dayEMG ElectromyographicENP Early neural proteinFDA Fluorescein-labelled dextran-amineFT Femorotibialis muscleGL Gastrocnemius lateralis musclehr HoursHz HertzI Injection site of retrograde tract-tracing chemicalIF Iliofibularis musclekDa kilodaltonL LitersLNP Late neural proteinL. PECT Left pectoralis muscleL. SA Left sartorius muscleM MolarmA milliampereMBP Myelin basic proteinmm minutemL millilitermm millimetermsec millisecondMW Molecular weightNCAM Neuronal cell adhesion moleculeNgCAMNeuron-glia cell adhesion moleculeNGF Nerve growth factorOT Tractus opticusP Posthatching dayPECT Pectoralis musclePNS Peripheral nervous systempsi Pounds per square inchRDA Rhodamine-labelled dextran-amineR. PECT Right pectoralis muscleR. SA Right sartorius muscleSA Sartorius musclesec secondSEM Standard error of the meanT Plane of transection siteTsx Experimental age of transectionWGA-HRP Wheat germ agglutinin- horseradish peroxidasew/v Weight per volumemicroamperep.L microlitermicronviiiNeuroanatomical AbbreviationsMedullaIX Nucleus nervi glossopharyngeiX Nucleus motorius dorsalis nervi vagiID Nucleus olivaris inferiorMLF Fasciculus longitudinalis medialisCnv Nucleus centralis medullae oblongatae, pars ventralisRgc Nucleus reticularis gigantocellularisRpc Nucleus reticularis parvocellularis (medulla)R Nucleus raphes (medulla)lTD Nucleus et tractus descendens nervi trigeminiPonsVI Nucleus nervi abducentisLa Nucleus laminarisMLF Fasciculus longitudinalis medialisN V Nervus trigeminusN VI Nervus abducensN VIII Nervus octavusVeM Nucleus vestibularis medialisVeL Nucleus vestibularis lateralis (Deiters’)R Nucleus raphes (pons)RP Nucleus reticularis pontis caudalisRPgc Nucleus reticularis pontis caudalis, parsgigantocellularisLoC Locus ceruleusScd Nucleus subceruleus dorsalisScv Nucleus subceruleus ventralisMesencephalonIII Nucleus oculomotoriusV VentriculusBC Brachium conjunctivumEW Nucleus of Edinger-WestphalICo Nucleus intercollicularisIs Nucleus interstitialisMLd Nucleus mesencephalicus lateralis, pars dorsalisMRF Formatio reticularis mesencephaliN III Nervus oculomotoriusRu Nucleus ruberTPc Nucleus tegmenti pedunculo-pontinus, pars compactaixACKNOWLEDGEMENTSThere are a number of very special people that have helped make this thesispossible. First and foremost, I would like to dedicate this thesis to my big brother Khalid(baya). He has always been my role model and his example has given me the confidence tosucceed. I’d also like to thank mom and dad, and Sabir and Sarvat for being so supportive.Together, these people have been and always will be my emotional anchor.Several people have been tremendously helpful to me and have also made the lab avery pleasant place within which to work. It was great having Sarah Shull (“Patitchka”)around my first year of graduate studies. We did just about everything together. IgnacioValenzuela (“Iggy”) is a brilliant engineer and his contributions greatly facilitated thecompletion of the physiology experiments in this thesis. Iggy was also instrumental inhelping me to worship the Data Gods with the creation of the Data God symbol, the DataGod cry, and the Data God dance. Thanks Iggy, our efforts seem to have payed-off onceagain. Cranking up the ghetto blaster and doing surgery with Hans Stegmann Keirstead wasalways a riot. Gillian Dawn Muir D.V.M. has been my surrogate big sister. She and herthree legged dog Isadora (“Izzy”, not to be confused with “Iggy”) were always willing to lendan ear when times were tough. Gill also helped my finances immeasurably by introducingme to the Vancouver used clothing scene. Diane Henshel was incredibly helpful to methroughout graduate studies. She very carefully critiqued just about everything I wroteduring my tenure at U.B.C. (from journal papers to this thesis). Thanks Diane, things wouldhave been much more difficult without you. Deirdre Webster and Greg Funk (former labmembers) were also wonderful to consult every once in a while. Also, Bill Milsom’s “openlab door” policy was both a comfort when I needed advice and a big help when it came toposter making time. And then of course there’s our fearless leader Dr. John Steeves. Onething about John is that he genuinely cares about his graduate students and, as a result, Ixnever feared for my graduate career. Thanks for the faith, encouragement, and directionJohn.And then there’s Rhonda Jean Garland (a.k.a. “Bonda Bean”, “Beanus Bondai”,“Pimphales promelas”, or just plain “Beaner”) who was with me (always with homemadecookies to share) through the ups and the downs. Thanks Beaner for being a ray of sunshineand for always saying “that which does not kill me shall make me stronger”. I can’t evenbegin to express how much your friendship has meant to me and I hope it never ends. I’dalso like to thank Tim West both for his friendship and also for all those late night mindcleansing games of hallway frisbee. Special thanks also go to Tim and Rhonda for helpingme to perfect the soon to be famous “Amazmo Burger” on the barbeque.Also, there’s all the folks in the department who have made the social side ofgraduate studies such a great experience. It became especially evident at my thesis defenseparty that I am lucky to have made quite a few friends at U.B.C. These friends includeJoanne Lessard, Richard Kinkead, Claudia Kasserra, Julie Hunter, Steve Land, PaoloDomenici, Nicholas Bernier, Sumi Aota, and Colin Brauner to name just a few. Thank youall for your friendship. I hope that I have the fortune of crossing paths with you again in thefuture.In addition, there are a number of others who, although not directly involved in mystudies, have significantly improved (and/or corrupted) my graduate student life. DaveMacLennan and Ron Aoyama are especially to be thanked for their kind friendship, all thosedinner/movie nights at there place and also for introducing me to Pangalactic Gargoblasters(a big hit at my thesis defense party). Thanks also to Byron Berry and his parents. TheBerrys have been my adopted family in Vancouver and I greatly appreciate that they invitedme into their home for dinner on many a lonely Christmas or Thanksgiving Day.Finally, I would like to thank the Rick Hansen Man in Motion Legacy Fund and theNatural Sciences and Engineering Research Council of Canada whose scholarship supportgreatly expedited the completion of this work.1CHAPTER 1GENERAL INTRODUCTION2Invertebrates and lower vertebrates are able to regenerate axonal processes ofneurons in their central nervous system (CNS). If the brain or spinal cord of a fish or axolotiis damaged, the cut axons will begin to regenerate within a few days. The axons willsubsequently regrow back to their original targets and will form functional connections there.Not only will axons of invertebrates and some lower vertebrates regenerate after beingdamaged, but CNS neurons also show great accuracy in their ability to find and remakeappropriate connections with their original target (Attardi and Sperry, 1963; Gaze andJacobson, 1963; Sperry, 1963; Bullock et al., 1977; Davis et al., 1990). This ability has beengradually lost during evolution, and in adult higher vertebrates (i.e. birds and mammals) isvirtually absent. Frogs are at an intermediate stage, being able to regenerate axons in theiroptic nerves, but not in most of the rest of their CNS (Lyon and Stelzner, 1987).Like the CNS of invertebrates and lower vertebrates, the peripheral nervous system(PNS) of adult higher vertebrates has long been recognized as having a remarkable capacityfor regeneration (Wailer, 1852; Ranvier, 1871, 1873; Cajal, 1928). Subsequent to axotomy(severence of axons) both the proximal and distal stumps of the axons undergo immediatetraumatic degeneration. Distal axonal stumps continue with Wallerian (or secondary)degeneration (Waller, 1852; Ranvier, 1871, 1873), due to their lack of “trophic” support(Cajal, 1928) and ultimately are resorbed by the glia. However, proximal axonal stumpswhich are still connected to the neuronal soma develop new growth cones or “clubs” withinone day of injury (Cajal, 1928). Regeneration proceeds at a rate of approximately 1 mm perday (Cajal, 1928). Although regeneration is greatly facilitated by increased proximity toremnants of the distal nerve portion (i.e. Schwann cell tube), modest regeneration can occurin the absence of any nerve remnants (Cajal, 1928). Near normal functional recovery canoccur within weeks or months following injury to the adult higher vertebrate PNS (Cajal,1928).Unlike the PNS, however, the CNS of adult higher vertebrates has a very limitedcapacity for regeneration. Damage to higher vertebrate adult CNS axons (e.g. due to spinal3cord injuries) results in aborted attempts at regeneration, followed by degeneration (Cajal,1928; Bernard and Carpenter, 1950; Bjorklund et aL, 1971). Again, both proximal and distalstumps undergo immediate traumatic degeneration, and distal axon stumps undergoWallerian degeneration (Cajal, 1928; Bjorkland et al., 1971). Although proximal axonstumps develop new growth cones, they will only elongate 1-2 mm before degenerating(Cajal, 1928; Schnell and Schwab, 1990). Thus, CNS axons usually do not regrow in theenvironment of the higher vertebrate adult spinal cord. However, Aguayo and colleagueshave demonstrated that CNS axons will anatomically regenerate if they are allowed toproject through a peripheral nervous system environment (David and Aguayo, 1981); therebydispelling previous suggestions that the lack of CNS axonal regeneration in higher adultvertebrates was due to the irreversible suppression of intrinsic neuronal growth programs. Itis now generally accepted that, if CNS environmental conditions are favorable, adult CNSneurons should be capable of regeneration (e.g. Crutcher, 1989; Reynolds and Weiss, 1992).Certain factors which are present in the environment of the adult higher vertebrateCNS may have detrimental effects on neurite (growing axons and dendrites) outgrowth(Schwab and Thoenen, 1985; Caroni and Schwab, 1988a; Schwab and Caroni, 1988). Theinhibitory effects of myelin are of particular interest with the recent discovery of two proteins(35 kDa and 250 kDa) expressed by oligodendrocytes (Caroni and Schwab, 1988a; Schwaband Caroni, 1988). These two proteins have been found to inhibit neurite outgrowth in vitro(Caroni and Schwab, l988a; Schwab and Caroni, 1988). In addition, in vivo studies haveshown that one can successfully block the neurite inhibiting properties of these two proteinsand thereby enhance axonal regeneration after CNS injury in the rat (Schnell and Schwab,1990). Using ventricular injections of hybridoma cells which produce antibodies to theseproteins, Schnell and Schwab (1990) have demonstrated enhanced anatomical regenerationof corticospinal axons after pyramidal tract lesions in adult rats.In contrast to higher vertebrate adult CNS, the embryonic CNS (e.g. spinal cordtissue) of higher vertebrates has generally been found to have a greater capacity for repair4(Hasan et at, 1991, 1992; Shimizu et al., 1990). These findings are not surprising since anenvironment capable of supporting the extensive migration of axonal growth cones duringspinal cord development, might be expected to be supportive of regenerating axons. Thus,the embryonic spinal cord may provide insight into the mechanisms which facilitate orobstruct axonal repair/regeneration. An essential first step in understanding this process is todetermine when during development the embryonic spinal cord loses the ability to repair. Ifone could identify a certain stage of embryonic development when the repair and recoveryfollowing spinal injury changes from successful to unsuccessful, then it may be possible inthe future to identify the cellular and molecular changes that underlie the altered response toinjury.Avian embryos have been a popular model for the study of vertebrate development.The developing chicken (Gallus domesticus) is a classic model of vertebrate embryologyand, as such, its stages of development are thoroughly described and well known (Hamburgerand Hamilton, 1951). Avian embryos will survive in vivo experimental manipulations (e.g.surgery) that mammalian embryos (in vivo) will not readily endure. In addition, it is wellestablished that the overall structure and function of the vertebrate brainstem and spinal cordhave remained relatively unchanged throughout evolutionary history (Griliner and Dubuc,1988; Sarnat and Netsky, 1981). For example, the overall development and basicorganization of brainstem and spinal locomotor mechanisms indicates that birds are verysimilar to all other vertebrates, including mammals (ten Cate, 1960, 1962; Bekoff, 1976;Jacobson and Hollyday, 1982b; Okado and Oppenheim, 1985; ODonovan and Landmesser,1987; Sholomenko and Steeves, 1987; Steeves et al., 1987; Webster and Steeves, 1988,1991; Sholomenko et al., 1991a,b,c; Hasan et al., 1991, 1992). Previous studies haveestablished that the descending brainstem-spinal control of avian locomotion is analogous tothat in quadrupedal mammals such as the cat and rat, (Steeves and Jordan, 1980, 1984;Armstrong, 1986; Garcia-Rill et at, 1986; Sholomenko and Steeves, 1987; Steeves et al.,1987; Noga et al., 1988; Sholomenko et al., 1991a,b,c). Birds utilize the same brainstem5spinal pathways for the initiation of locomotion, and after an adult spinal cord injury, sufferthe same motor deficits as a mammal (Eidelberg et aL, 1981; Zemlan et al., 1983;Sholomenko and Steeves, 1987; Webster and Steeves 1991).Specifically, it has long been known that after transection of all spinal dorsal roots(i.e. removal of all peripheral afferent feedback), every vertebrate animal so far examinedcan still generate locomotor movements (Grillner and Wallen, 1985; McClellan, 1986). Thisimplies a basic reliance on CNS networks (often referred to as central pattern generators orrhythmic oscillators) for locomotion. At present, the rhythmic oscillators for vertebratelocomotion are not completely characterized, but they are known to be intrinsic to the spinalcord. These spinal locomotor networks are normally activated by descending neurons fromsupraspinal centers. Several regions of the brain have been identified as locomotor initiationcontrol centers, including: medullary and pontine reticular formation regions, mesencephaliclocomotor regions, subthalamic nuclei, red nuclei, basal ganglia, limbic forebrain, and (inmammals) the motor cortex (Lawrence and Kuypers, 1968 a,b; Steeves and Jordan, 1980,1984; Armstrong, 1986; Garcia-Rill, 1986; Jordan, 1986; McClellan, 1986; Garcia-Rill andSkinner 1987; Steeves et al., 1987; Griliner and Dubuc, 1988). It has also been suggestedthat some of these supraspinal brainstem regions, especially reticulospinal regions, areessential to the generation of any functional locomotion (for review, see Armstrong, 1986).Physiological data suggest that spinal cord neurons participating in the generation oflocomotor behaviors are predominantly activated by direct input from the phylogeneticallyolder brainstem-spinal pathways (i.e. the extrapyramidal motor system). A good proportionof the integrated brainstem motor output is directed to the spinal cord via reticulospinalpathways originating in the pons and medulla (Barnes, 1984; Steeves and Jordan, 1984;Armstrong, 1986; Garcia-Rill, 1986; McClellan, 1986, 1988, 1990; Steeves et al., 1987).The studies in this thesis examine both the development and repair of brainstemspinal (including reticulospinal) pathways in the embryonic chick. I have focussed myattention on determining the duration of development during which an embryonic chick is6capable of axonal repair after complete transection of the thoracic spinal cord.Approximately 5-8 days after transection, anatomical recovery (Chapter 2) can be evaluatedwith a restricted injection into the lumbar cord (caudal to the transection site) of a fluorescentretrograde tracing dye (Hasan et aL, 1991, 1992). If the spinal transection is made prior toembryonic day (E) 13 (i.e. on E3-E12), the disthbution and number of retrogradely labelledbrainstem-spinal neurons is similar to that observed in sham-transected or untransectedcontrol chicks. Transections conducted on E13 (or later) result in diminished or noretrograde labelling of brainstem nuclei (Hasan et al., 1991, 1992).There are several neuroanatomical responses that could underlie this embryonicrepair process, including: 1) neurogenesis of new descending brainstem or spinal neurons, 2)subsequent projections from late developing brainstem-spinal neurons, and 3) true axonalregeneration of previously axotomized brainstem-spinal projections. It is known thatbrainstem-spinal neurons become post-mitotic very early in development, usually prior to E5(McConnell and Sechrist, 1980; Sechrist and Bronner-Fraser, 1991), and well before E12.Therefore, it is unlikely that my findings of complete functional repair of E7-E12 transectedcords are dependent on the neurogenesis of additional brainstem-spinal neurons (Hasan et al.,1991, 1992).By Eli of normal embryonic development, the distribution and number ofretrogradely labelled brainstem-spinal neurons is equivalent to those labelled in a chick afterhatching (Okado and Oppenheim, 1985; Hasan et al., 1991, 1992). There is apparently noover-production of brainstem-spinal neurons and subsequently no naturally-occurring celldeath of these neurons during development (Oppenheim, personal communication).Therefore, a mid-thoracic spinal transection at E7-E12 would disrupt the many descendingprojections already at lumbar levels. However, it is still possible that the subsequentdevelopment of axonal projections that had not already descended to thoracic levels at thetime of transection may contribute to the anatomical recovery, even after a transection as lateas Ell-E12.7Considering how complete the development of brainstem-spinal pathways is byEli-El 2, it is more probable that the repair process involves axonal projections sproutingfrom the proximal ends of previously axotomized fibers (i.e. regeneration). If repair was notin part due to regeneration, but only due to subsequent axonal projections from latedeveloping neurons, a reduced number of retrogradely labelled brainstem-spinal neuronswould be expected after an El l-E12 transection. The similar numbers and distribution ofretrogradely labelled brainstem-spinal neurons in transected and control Eli embryos arguesin favor of regeneration contributing to the repair process (Hasan et al., 1991, 1992).Nevertheless, these potential repair mechanisms are not mutually exclusive.Two different retrograde fluorescent tracing labels were used to establish that spinalcord repair in the latter stages of the permissive period (El O-E 13) is due, in part, to trueaxonal regeneration (Hasan et al., 1992). The first retrograde tracing label was injected intothe lumbar cord prior to thoracic transection and the second was injected approximately oneweek after transection. Comparing the degree of retrograde single- and double-labelling ofbrainstem-spinal neurons indicated the extent of regeneration (Chapter 3).We then determined whether these repaired brainstem-spinal projections undergosynaptogenesis with appropriate spinal neurons and subsequently form functionalconnections. The functional repair of descending supraspinal pathways (Chapter 4) wasconfinned by behavioral observations of the motor functions of post hatching (P) chicks andalso directly tested by focal electrical stimulation of brainstem locomotor regions, known tohave direct projections to the lumbar cord (Valenzuela et al., 1990; Hasan et al., 1991, 1992).Leg and wing electromyographic recordings were used to monitor brainstem evoked motoractivity. Brainstem stimulation experiments were undertaken on both transected and controlanimals, either in ovo on E18-E20 or after hatching on P1-P2.In comparison to sham-transected and untransected control animals, completefunctional recovery of motor function was evident in chicks that had their spinal cordtransected prior to E13 (Hasan et al., 1991, 1992; Valenzuela et al., 1990). These findings8were consistent, regardless of which method of assessment (behavioral observation orbrainstem stimulation) was used, or whether they were assessed before or after hatching.Therefore, damaged embryonic spinal cord is capable of complete anatomical and functionalrecovery after an El 2 transection, but anatomical and functional repair rapidly diminisheswith transections on El 3-E 14 and is completely non-existent after an El 5 transection.Thus, thoracic transections of the spinal cord in El 3 or later embryos, or in adultbirds, results in incomplete spinal cord repair and often complete paralysis (Hasan et al.,1991, 1992; Shimizu et al., 1990; Sholomenko and Steeves, 1987, Webster and Steeves,1991). Based on current available evidence, it would appear that after El3, embryonicbrainstem-spinal neurons have a similar response to axotomy as brainstem-spinal neurons inadult birds (Sholomenko and Steeves, 1987; Hasan et al., 1991, 1992). Specifically, theneurons will endeavor to regenerate new axons, but since the spinal cord environment is notsupportive for re-growth the neurons will ultimately fail. The transition from a permissive toa restrictive period for complete anatomical repair and functional recovery of descendingsupraspinal projections following thoracic spinal cord transection in chicks appears to bebetween E12 and E13.9CHAPTER 2ANATOMICAL REPAIR OF TRANSECTED SPINAL CORDIN EMBRYONIC CHICK(SINGLE-LABELLING STUDIES)10INTRODUCTIONThe anatomical organization of avian descending brainstem-spinal pathways issimilar to that of other vertebrates, including mammals (Steeves et al., 1987; Webster andSteeves, 1988; Webster et aL, 1990). Additional studies have demonstrated strikingsimilarities among all vertebrates regarding the functional organization of brainstem-spinalprojections involved in the initiation and maintenance of locomotor activity (Jacobson andHollyday, 1982b; Griliner and Wallen, 1985; Armstrong, 1986; McClellan, 1986; Steeves etal., 1987; Griiner and Dubuc, 1988). As in other vertebrates, focal electrical or chemicalstimulation of brainstem regions containing a high density of reticulospinal projections willevoke locomotion in birds (Steeves et al., 1987; Sholomenko et al., 1991a,b,c). Furthermore,in all vertebrates examined, disruption or inactivation of these pontine and medullaryreticulospinal pathways will effectively inhibit or alter any spontaneous or brainstem-evokedlocomotion (Eidelberg, 1981; Shefchyk et al., 1984; Sholomenko and Steeves, 1987; Grilinerand Dubuc, 1988).Okado and Oppenheim (1985) and Glover and Petursdottir (1991) examined theappearance of retrogradely labelled neurons within the developing brainstem after discreteinjections of wheat-germ-agglutinin-congugated horseradish peroxidase (WGA-HRP) or afluorescent dye into embryonic chicken cervical or lumbar spinal cord. The first retrogradelylabelled neurons appeared within the pontine reticular formation. These reticulospinalprojections descended to the cervical spinal cord by embryonic day 4 (E4) and to the lumbarcord by E5. A comparison of the distribution and density of retrogradely labelled brainstemspinal neurons in embryonic versus hatchling chicks led Oppenheim and colleagues toconclude that the descent of all brainstem-spinal projections is essentially complete to thecervical cord by E7-E8 and to lumbar levels by El l-El2 (Okado and Oppenheim, 1985;Shimizu et al., 1990). Preliminary studies have confirmed that focal electrical stimulation ofreticulospinal projections will evoke fictive locomotor activity from an E7 in vitro brainstemspinal cord preparation (Sholomenko and O’Donovan, personal communication). In general,11the ontogeny of descending supraspinal projections in the embryonic chicken is similar tothat observed in a variety of vertebrates from fish to mammals (Martin et al., 1979; Forehandand Farel, 1982; ten Donkelaar, 1982; Okado and Oppenheim, 1985; Hasan et al., 1991,1992).Recent studies on the effects of spinal cord transection of ElO and E15 chickenembryos indicated anatomical repair of brainstem-spinal axons after an E 10 thoracic spinalcord transection; but no anatomical repair of brainstem-spinal axons if the transection wasperformed on embryonic day 15 (Shimizu et al., 1990). Since complete recovery occurredonly after an ElO transection, when some of the brainstem-spinal projections have yet toreach the lumbar cord (Okado and Oppenheim, 1985; Hasan et al., 1991, 1992), it was notclear whether the repair was due to subsequent projections from late developing neurons (i.e.undamaged axons) or to regeneration of previously axotomized projections.I have also examined the response of embryonic chick spinal cord to injury usingretrograde anatomical tracing techniques. However, unlike Oppenheim and colleagues(Shimizu et al., 1990), I have carried out thoracic transections on E3-E14 of chickdevelopment in order to more precisely establish when in embryonic developmentanatomical recovery of descending supraspinal projections disappeared. I found completeanatomical recovery of descending supraspinal projections in chicks transected as late asE12. Diminished numbers of brainstem-spinal projections were observed after thoracic cordtransections on E13-E14. Therefore, the transition from a permissive to a restrictive periodfor complete anatomical repair of descending supraspinal projections following thoracicspinal cord transection in chicks appears to be between E12 and E13.12MATERIALS AND METHODSFertilized white Leghorn chicken eggs were incubated at 37°C in a humid incubator(relative humidity = 60%). Prior to any embryonic manipulation, a small window was madein the egg shell and the embryos were staged according to the criteria of Hamburger andHamilton (1951). Early embryos were held in place by clasping the amniotic membrane withblunt forceps. Damage to the extraembryonic membranes, although unavoidable, was kept toa minimum. E9 and older embryos were manipulated with the use of a rounded glass probepassed under the neck.All animals are referred to throughout this thesis by the embryonic day (E) on whichthey were transected. Chick embryos incubated at 37°C hatch on E21.Spinal Cord TransectionsComplete spinal cord transections for E3-E10 embryos were made in the upperthoracic region (Ti ± 1 segment) using finely sharpened forceps. In embryos older than ElO,the spinal transection was made at a mid-thoracic (T4 ± 1 segment) level in order tominimize bleeding from dorsal spinal blood vessels. Tissue surrounding the spinal cord wasslightly damaged by the transection procedure. To assure that each transection procedurewas complete a #00 pin was passed across the entire spinal column and cord at the site oftransection. The pin was previously marked for the depth of the spinal cord at that particularage of embryonic development. After surgery, the eggs were sealed with a sterile coverslipusing paraffin wax and returned to the incubator for at least 5 days. In 12 randomly selectedembryos (two E5, two E6, two E7, two E9, two El 1, and two E13) the completeness of thetransection was assessed by removing the spinal column for histological examinationimmediately following the transection procedure. The above surgical procedures were alsoused for sham-transected embryos with the exception that the spinal cord and surroundingtissues were not damaged.13Anatomical AssessmentsComplete anatomical protocols (see below) were completed on one E3 (injected onEl 1), one E5, one E7, two E9, one ElO, seven Eli, three E12, six El3, and two E14 spinalcord transected chicks. In order to maintain comparable survival periods following spinalcord transection, embryos operated on E5-E9 were injected on Ei2-E15, while embryosoperated on ElO through E14 were injected on E17-E19. All injected embryos weresacrificed 1-2 days later and the brainstem and spinal tissue were processed for the optimalvisualization of fluorescent retrogradely labelled brainstem-spinal neurons. In addition,lumbar injections of retrograde tracing chemicals were completed on 7 sham-transectedcontrol embryos (one E5, three Eli, two E12, one Ei3; injected on Ei2 through E18) and 6untransected control embryos (two Eli, two Ei2, one Ei3, and one Ei5, injected on El7-E20). Finally, two untransected hatchling control chicks (which were injected on P3), werealso processed and analyzed for retrograde labelling of brainstem-spinal projections.Embryonic InjectionsThe lumbar (L1-L2) spinal cord of each transected, sham-transected, or untransectedcontrol embryo was injected with Rhodamine-labelled dextran-amine (RDA). To prevent theinjected retrograde tracer from diffusing rostrally above the lesion and falsely labellingbrainstem-spinal neurons, injections were made several spinal segments caudal to thetransection site. For the sham-transected and untransected controls, injections were made ata comparable spinal cord level (L1-L2). A glass micropipette (tip diameter =40 tim)containing an aqueous solution of RDA (10,000 MW, lysine fixable, 25% w/v, MolecularProbes Inc., Eugene, Oregon, catalogue # D-1817) was then inserted into the upper lumbarspinal column and advanced into the spinal cord tissue. The micropipette tip was previouslymarked for the depth of the spinal cord at that particular age of development. Approximately0.1-0.6 IlL of RDA was pressure injected slowly over a 10 to 15 minute interval (smaller14volumes were used for E15 and younger embryos). After the injection, the eggs were sealedwith a sterile coverslip and paraffin and returned to the incubator.After a 24-48 hr survival period, the embryos were sacrificed. First, embryos wereanesthetized in ovo with Somnotol (sodium pentobarbitol, 75 mgfkg IP). Next, the embryoswere removed from their shells, and the omphalomesenteric (umbilical) veins and arterieswere cut to allow exsanguination. The embryos were then perfused via the left cardiacventricle over a 30-minute period with 50 mL of 0.9% NaC1 (containing 1000 IU of sodiumheparin) at 37°C, followed by 50 mL of 4% paraformaldehyde in 0. 1M phosphate buffer(pH 8.5, 20°C). The brain and spinal cord were post-fixed for 1-3 hr and then placed in 0.1M phosphate buffer saline containing 10% sucrose (pH 7.4, 4°C). After 24 hr. the tissue wastransferred into 30% sucrose in 0. 1M phosphate buffer saline and stored at 4°C forsubsequent histology.HistologyThe brain and spinal cord were cut at 30 .tm on a freezing microtome and thesections placed in 0. 1M phosphate buffer (pH 7.4 at40C). The brains were cut in coronalsections while the spinal cords were cut in sagittal sections. Tissue sections wereimmediately mounted on slides for subsequent examination under a fluorescent microscope.The positions of retrogradely labelled neurons within the brain tissue sections weremapped with camera lucida drawings. Every third tissue section was counted so as to reducethe possibility of counting the same neuron twice. Brain nuclei were identified using severalavian atlases (Karten and Hodos, 1967; Youngren and Phillips, 1978; Cabot et al., 1982;Gross and Oppenheim, 1985; Okado and Oppenheirn, 1985; Kuenzel and Masson, 1988;Webster and Steeves, 1988). To make comparisons between transected and control animals(sham-transected and untransected), averaged cell counts ± the standard error (S.E.M.), andnumerical range, were measured for each of several brainstem-spinal nuclei on the right sideof the brain. Cell counts were obtained for the nucleus solitarius, nucleus centralis, nucleus15reticularis gigantocellularis, and nucleus raphes in the medulla; the nucleus vestibularislateralis, locus ceruleus, nucleus subceruleus, nucleus reticularis pontis caudalis/parsgigantocellularis, and nucleus raphes in the pons; the nucleus interstitialis and nucleus ruberin the mesencephalon; and the nucleus paraventricularis magnocellularis and stratumcellulare internum/extemum in the diencephalon.Regardless of which currently available anatomical tracing technique is used and nomatter how rigorously standardized the methodology, it is well known that there can bevariability in the number of retrogradely labelled neurons between animals (Heimer andRoBards, 1981). Counting retrogradely labelled cells can lead to mistaken estimates (usuallyunder-estimates) about the number of neurons projecting to the injection site. By calculatingthe mean and standard error of labelled cell numbers in control embryos, I was able toestimate the degree of variability inherent in the labelling technique and thus makecomparisons between control and transected animals.The spinal cord tissue sections were examined for the extent of diffusion of theinjected retrograde tracer chemical. If there was any evidence of the retrograde tracerdiffusing rostrally through the cord to the transection site, the animal was excluded fromfurther anatomical analysis. The one animal in which this occurred was excluded fromanatomical analysis.16RESULTSSpinal TransectionThe completeness of the spinal cord transection was histologically confirmed in 12randomly selected embryos, operated on different days of embryonic development (two E5,two E6, two E7, two E9, two Eli, and two El3). In all cases, regardless of developmentalage, the spinal cord was completely severed at the thoracic level and there were no apparentdifferences in the morphology of the lesion (Fig. 2-1A,2-2A). Since the same surgicalprocedure was undertaken on all experimental animals and histologically confirmed in adozen embryos, I am confident that these techniques resulted in a complete transection ofeach embryo’s spinal cord. When the thoracic transection was performed prior to Ei3 andthe embryo was allowed to recover for a period of approximately 5-8 embryonic days, therewas little, if any, evidence of scar tissue within the transection site (Fig. 2-1B). Since eachspinal cord was processed within the vertebral column, the location of a lesion performedprior to E13 could be subsequently identified by aberrant morphology of the overlyingtissues. If the transection was performed on or after E13, there was histological evidence ofincomplete spinal cord repair (Fig. 2-ic).Injection SitesIt was also critical to confirm that the injection of retrograde tracing chemical wasconfined to the caudal spinal cord (lumbar and lower thoracic levels). If the tracer (RDA)were to directly diffuse to the transection site (or perhaps even to more rostral spinal levels),then the tracer could falsely label brainstem-spinal projections due to axonal transport viaprojections that had not descended beyond the level of the transection site. Histologicalexamination of all spinal injection sites confirmed that, in all but one instance, the injectedretrograde tracer was confined to within 1-2 segments of the rostral lumbar cord and did not17Figure 2-1. Longitudinal sections through the thoracic (A,B and C) and rostral lumbar (D)spinal cord of the embryonic chicken. A. Photomicrograph of a toluidine blue stainedparasagittal section through the spinal cord of an El 1 embryo illustrating the completesevering of the cord immediately after transection at the mid-thoracic level. B.Photomicrograph of a toluidine blue stained parasagittal section through the thoracic spinalcord of an E20 embryo (transection performed on ElO). Embryos with a complete thoraciccord transection before El 3 showed complete anatomical and functional recovery. Little orno sign of any morphological aberration (eg. glial scarring) was noted 10 days after thetransection. The location of a lesion performed during the permissive repair period (ie. <E13) could be subsequently identified by aberrations in overlying tissues. C.Photomicrograph of a toluidine blue stained parasagittal section through the thoracic spinalcord of a 3 day old hatchling chick (transection performed on E13). Embryos with acomplete thoracic cord transection on or after this age showed incomplete anatomical andfunctional recovery. D. Parasagittal section through the rostral lumbar cord illustrating thesite of injection of the retrograde tract-tracing chemical (in this case 0.2 IlL Rhodaminelabelled dextran-amine). Caudal direction is towards left on photomicrograph. Note thatthere is minimal rostral diffusion of the retrograde tracer up the central canal (CC), little ifany subdural diffusion rostral to the injection site. In all animals the injection sites wereconfined to the rostral lumbar cord (several segments caudal to the mid-thoracic transectionsite). Some tissue shrinkage has occurred due to the dehydrating tissue embeddingprocedure. Abbreviations: CC- canalis centralis, I- injection site of retrograde tract-tracingchemical, T- plane of transection site. Calibration Bar = 400 m for A, B, and C; 25 pm forD.19Figure 2-2. Photomontage constructed from a parasagittal section through the mid-thoracic(A) and rostra! lumbar (B) spinal cord of an E12 embryo. The distance (not shown) betweenthe right edge of panel A and the left edge of panel B is 4.4 mm. This particular animalserved as a control experiment to ensure that the spinal cord was completely severed and thatthe injected retrograde tract-tracing chemical (0.2 iL Rhodamine-labelled dextran-amine)did not directly diffuse across the site of transection. In short, the mid-thoracic transection(T) and the rostra! lumbar injection of retrograde tract-tracing chemical were performedduring the same surgical procedure and the animal was sacrificed immediately thereafter.Note that the transection site (T) and the site of injection of retrograde tract-tracing chemical(I) is separated by an unlabelled area of 5.7 mm. At the low magnification of thesephotomicrographs the meninges autofluoresce. Some tissue shrinkage has occurred due tothe dehydrating tissue embedding procedure. Abbreviations: I- injection site of retrogradetract-tracing chemical, T- plane of transection site. Calibration Bar 100 urn21diffuse rostrally to the transection site (Fig. 2-1D,2-2). The one animal, in which this didoccur, was excluded from anatomical analysis.Brainstem-spinal Projections in Control AnimalsRetrograde tracer was injected into the lumbar spinal cord of 7 sham-transected and8 untransected control animals at various developmental ages (see Materials and Methods,Anatomical Assessments for number of control animals at each developmental age). Thiscontrol study was undertaken to determine the number and distribution of brainstem neuronswith lumbar projections. The numbers and distribution of retrogradely labelled neuronal cellbodies found in identified brainstem nuclei were equivalent among control animals,regardless of when the retrograde tracer was injected between Eli through to hatching (Fig.2-3A,2-4A,2-5; Tables 2-i, 2-2). On rare occasions (one Eli sham-transected and one Eliuntransected animal), a brainstem nucleus was observed in which the number of retrogradelylabelled neurons was significantly reduced or the label was particularly faint in appearance.Due to the rarity of this event, I attribute these differences to “vagaries” that sometimes occurwith retrograde labelling techniques (Heimer and RoBards, 1981); in any case, they did notsignificantly alter these results. The combined results from these 15 control animals arepresented in Tables 2-1 and 2-2. In general, the distribution and temporal development ofbrainstem-spinal neurons observed here agrees with that previously reported by Okado andOppenheim (1985).Brainstem-spinal Projections in Spinal Transected AnimalsComplete anatomical protocols were completed on one E3 (injected on El 1), oneE5, one E7, two E9, one ElO, seven Eli, three E12, six E13, and two El4 spinal cordtransected embryos. If the thoracic transection was performed on E3-El2 the subsequentlylabelled brainstem nuclei contained an equivalent mean number, range, and distribution ofbrainstem-spinal neurons to the values observed in control animals (Table 2-1, Fig. 2-3, 2-422Figure 2-3. Photomicrographs of retrogradely labelled gigantocellular reticulospinal neuronswithin the ventromedial reticular formation of the caudal pons (RPgc) in an E18 control embryo,injected with RDA on E16 (A), an E18 embryo previously transected on Eli (permissivetransect) and injected with RDA on E16 (B), and an E20 embryo previously transected on Ei3(restrictive transect) and injected with RDA on E18 (C). Plate D indicates the location ofphotomicrographs in A,B, and C. Embryos were injected with 0.1-0.6 jiL of RDA into therostral lumbar cord. Note that the distribution and number of retrogradely labelled brainstemspinal neurons in control embryos (A) and embryos transected prior to Ei3 (B) were equivalent(see Table 2-i). Abbreviations: Cb- cerebellum, NV- Nervus trigeminus, RDA- Rhodaminelabelled dextran-amine, RPgc- N. reticularis pontis caudalis, pars gigantocellularis. CalibrationBar= 100 rim.24Figure 2-4. Photomicrographs of retrogradely labelled vestibulospinal neurons within the lateralvestibular nucleus of the dorsolateral pons (VeL) in an E18 control embryo, injected with RDAon E16 (A), an E18 embryo previously transected on El 1 (permissive transect) and injected withRDA on E16 (B), and an E20 embryo previously transected on E13 (restrictive transect) andinjected with RDA on El8 (C). Plate D indicates the location of photomicrographs in A,B, andC. Embryos were injected with 0.1-0.6 iL of RDA into the rostral lumbar cord. Note that thedistribution and number of retrogradely labelled brainstem-spinal neurons in control embryos(A) and embryos transected prior to E13 (B) were equivalent (see Table 2-1). Abbreviations:Cb- cerebellum, NV- Nervus trigeminus, RDA- Rhodamine-labelled dextran-amine, VeLnucleus vestibularis lateralis. Calibration Bar = 100 pPm.Cbn—flN2‘‘1VeL*jjNVD26Figure 2-5. Camera lucida drawings summarizing the representative distribution and number ofretrogradely labelled brainstem-spinal neurons at four different levels of the brainstem: A: rostralmedulla, B: caudal pons (at the level of the vestibular nuclei), C: rostral pons (at the level of thelocus coeruleus), and D: mesencephalon (at the level of the red nucleus). For each level (A-D):Part 1 (upper drawing) is an E18 control embryo; Part 2 (middle drawing) is an El7-18 embryotransected on El 1 (permissive transect); Part 3 (lower drawing) is an El9-20 embryo transectedon El 3 (restrictive transect). Note in all cases that the distribution and number of retrogradelylabelled brainstem-spinal neurons are similar in a control and Eli transected embryo, but thenumber of retrogradely labelled neurons diminishes after a thoracic transection on El 3 (see alsoTables 2-1 and 2-2). Small circles represent a single retrogradely labelled brainstem-spinalneuron/tissue section; large circles represent five retrogradely labelled brainstem-spinalneurons/tissue section. Abbreviations: see list of abbreviations.270 ATELENCEPHALONCEREBELLUMRuOTvi-oRPgc1mmIIC BAl Ix—x 28RRR29Bi VeM VeLMLFNRP/RPB2 VeLMLFNRP/RPB3 VeLMLFNRP/RP30Cl00NRPgcNVIC3N00 RPgcNVI031Dl020332Table 2-1. Number of labelled brainstem neurons after injection of a retrograde tracer into thelumbar cord. The spinal cord of the experimental animals was transected at the thoracic level.Control(n = 15) E3-E12 Transect(n = 16)Brainstem-sDinal nucleusMEDULLAMean÷S.E.M. Ranae Mean+S .E .M. RanaeDIENCEPHALONN. paraventricularismagnocellularisStratum cellulare internum/externumN. solitarius 11.8÷ 1.6 3—26 15.0± 0.9 0—30N. centralis medullae 70.2+ 7.2 0—119 84.4±10,4 51—300oblongataeN. reticularis giganto— 27.4± 4.1 10-62 25.8± 1.7 15-50cellularisN. raphes (medulla) 41.8± 2.5 26-60 44.4k 5.0 20-80PONSN. vestibularis lateralis 139.8±20.0 13—272 146.3± 7.5 70—200Locus ceruleus 98.2± 9.4 27-144 116.3± 4.2 70-160N. subceruleus 44.8± 5.9 17-72 35.6+ 5.6 10-100N. reticularis pontis 146.6±24.8 14—320 230.8±13.6 80—290caudalis & pars gigantocellularisN. raphes (pons) 29.7k 3.5 16—57 31.3± 3.4 20—70MESENCEPHALONN. interstitialis 80.8±16.0 40—150 73.6±5.6 60—100N. ruber 178.3±24.0 100—250 185.7±14.3 100—20060.8±17.3 0—100 117.8±26.6 0—20055- 190 60-170110. 0.22 .9 117.1±13.133Table 2-2. Number of labelled brainstem neurons after injection ofa retrograde tracer into thelumbar cord. The spinal cord of the experimental animals was transected at the thoracic level.Control(n = 15) E13-E14 Transect(n = 8)Brainstem-sDinal nucleusMEDULLAMean+S . E .M. Ranae Mean+S.E.M. RanceDIENCEPHALONN. paraventricularismagnocellularisStratum cellulare internum/externumN. solitariusN. centralis medullaeoblongataeN. reticularis gigantocellularisN. raphes (medulla)PONSN. vestibularis lateralisLocus ceruleusN. subceruleusN. reticularis pontiscaudalis & pars gigantocellularisN, raphes (pons)MESENCEPHALONN. interstitialisN. ruber11.8± 1.670.2± 7.227.4± 4.141.8± 2.5139. 8±20. 098.2± 9.444.8± 5.9146. 6±24 . 829.7k 3.580.8±16.0178.3±24.03-260—11910—6226—6013—27227— 14 417—7 214—32 016—574 0-150100—2500.0± 0.00.0± 0.03.0± 0.015. 0±15. 087.5±77 .522.5± 2.50.0± 0.022. 5±13 . 50.0± 0.00.0± 0.027.5± 2.50—00—00—60—3010—16520—2 50—09—360—00-025—3060.8±17.3 0-100 0.0± 0.0 0—055—19 0 0—0110 . 0±22 . 9 0.0± 0.034and 2-5). There were no significant or consistent differences in the anatomical data obtainedfrom animals transected on E3-E12, thus the results from these animals were groupedtogether. The only minor difference between the spinal transected and control groups was aslight increase in the average number of retrogradely labelled neurons within the locusceruleus, nucleus reticularis pontis caudalis and nucleus paraventricularis magnocellularis ofthe E3-E12 spinal transected group. The range of labelled neurons, however, was similar forboth spinal transected and control groups and the differences were not statisticallysignificant. In addition, one Eli transected embryo exhibited reduced retrograde labelling ofthe nucleus centralis medullae oblongatae, however the labelling of descending projectionsfrom other nuclei in the pons and mesencephalon was similar to the other Eli transectedembryos and control animals. These discrepancies are most likely due to the inherentvariability of retrograde labelling techniques (Heimer and RoBards, 1981).Among embryos transected on E13-E14, very few brainstem nuclei containedlabelled neurons. The medulla was virtually unlabelled, except for the raphe nuclei and afew cells within the gigantocellular reticular formation. Interestingly, in the pons, therewere a substantial number of labelled cells within the nucleus vestibularis lateralis(approximately 63% of the number of neurons labelled in control embryos). A smallernumber of retrogradely labelled neurons were also observed within the locus ceruleus andnucleus reticularis pontis caudalis (Table 2-2 and Fig. 2-3, 2-4, and 2-5). In themesencephalon, only the nucleus ruber was labelled at either age, and its cell count was sixtimes less than the control average (Table 2-2 and Fig. 2-SD). No labelled cells were foundin the diencephalon of embryos transected on E13-E14.35DISCUSSIONThe experiments reported here examined the capacity of brainstem-spinalprojections to anatomically recover from complete spinal transection at various stages ofembryonic development. The control (untransected and sham-transected) embryos, injectedwith retrograde tracers between El 1 and E18, displayed an equivalent distribution andnumber of retrogradely labelled neurons in each descending brainstem nucleus whencompared with results obtained from untransected hatchling chicks. These results confirmthe results previously reported by Oppenheim and colleagues that brainstem-spinalprojections complete development to the lumbar cord on El l-E12 (Okado and Oppenheim,1985; Shimizu et aL, 1990). More importantly, when compared to control animals, theresults demonstrate complete anatomical repair of descending brainstem-spinal projections inembryos that underwent thoracic cord transection as late in embryonic development as E12.Since the axonal projections of developing brainstem-spinal pathways to the lumbar cord arecomplete around El2, these findings also provide further evidence for the suggestion thatregeneration of previously axotomized projections can contribute to the observed anatomicalrepair after spinal cord transection in the embryonic chick (Hasan et al., 1992; Shimizu et al.,1990; Chapter 3). Examination of brainstem-spinal projections in El3-El4 transectedembryos indicated only minimal anatomical repair. This is not an unexpected finding since arecent parallel study (Shimizu et al., 1990) had noted complete functional recovery inhatchling chicks transected on ElO and no functional recovery if the transection wasperformed on E15. Shimizu and coworkers also showed (partial) anatomical recoveryfollowing spinal cord transections on ElO and no anatomical brainstem-spinal recoveryfollowing spinal cord transections on El5. These present results more precisely pinpoint thetransition from the permissive to the restrictive period of embryonic spinal cord repair toaround El3.The possibility of either an incomplete transection of the thoracic cord or directdiffusion of the retrograde tracer to levels at or above the transection site, yielding false36positive results, has been carefully examined in several animals and discounted (see Resultsand Fig. 2-1, 2-2). I also considered the possibility that the retrograde labelling of brainstemneurons following spinal transection was due to transynaptic labelling, via propriospinalneurons that had descended below the transection site and retrogradely transported the tracer.Retrograde transynaptic labelling of neurons can be an undesired outcome with someanatomical tracing chemicals (eg. WGA-HRP; Heimer and RoBards, 1981), especially if thedistance between the injection site and the labelled soma is short and/or the post-injectionsurvival period is long. Since my post-injection survival periods were relatively short (24-48hr), it is unlikely that transynaptic labelling contributed to the observed results. Furthermore,the position and density of retrogradely labelled brainstem-spinal neurons in E3-E12transected embryos was similar to that seen in intact embryos and adult birds (Okado andOppenheim, 1985; Webster and Steeves, 1988). This would be an unlikely coincidence iftransynaptic labelling was responsible.I conclude that direct retrograde labelling of brainstem-spinal neurons followingembryonic spinal transection occurred as a result of some, as yet unidentified, centralnervous system (CNS) repair process. The diminished repair of descending supraspinalpathways in the E13-E14 embryos cannot be attributed to a shorter post-transection recoveryperiod, since both Eli-El 2 transects and El 3-E14 transects had equivalent recovery periods(5-8 days).Another finding of the present study is that different descending spinal tracts appearto lose their capacity for repair at different ages of development. For example, in the El 3-E14 transected embryos, neurons descending from the nucleus vestibularis lateralis werelabelled in numbers that were within the range of control animals, averaging 63% of controlvalues.There are a number of possible explanations for the diminished projections fromdifferent brainstem-spinal nuclei in El 3-El 4 transected embryos. For example, glial scarringmight have been restricted to localized regions of the cord. This might prevent the37subsequent growth or regrowth of those tracts that normally run through these spinal funiculi(Barrett et al., 1984). One argument against this suggestion is that many of the observeddescending pontine pathways project through the same funiculi as the absent descendingmedullary- and mesencephalic-spinal projections (Glover and Pettersdottir, 1988; Websterand Steeves, 1988). Further studies are required to determine any possible glial scardifferences after transections at different stages of embryonic development.Another explanation is that the cellular mechanisms mediating axonal outgrowth,and/or repair of previously axotomized projections, may stop functioning at different stagesof development for distinct brainstem-spinal nuclei. The stage at which each chickbrainstem-spinal nucleus becomes post-mitotic (ie. the neuronal birthdate) has beendetermined (McConnell and Sechrist, 1980; Sechrist and Bronner-Fraser, 1991), as well asthe embryonic day when each supraspinal nucleus begins to send axonal projections to thecord (Okado and Oppenheim, 1985). There appears to be no correlation between birthdate,initiation of axonal development and late repair capacity. Different brainstem-spinal nuclei,having either similar birthdates or axonal development periods demonstrated entirelydifferent repair capacities. For example, the cells of the nucleus vestibularis lateralis (VeL)and the nucleus reticularis pontis cauadalis, pars gigantocellularis (RPgc) both send outaxons to the cervical and lumbar spinal cord between E5 and E5.5 but have dramaticallydifferent quantities of retrogradely labelled cells after an El 3 transection and subsequentinjection of dye (see Table 2-2). Subsequent investigations are planned to determine theexact nature of the differential repair capacities of the various brainstem-spinal nuclei. Thesefindings will also be correlated with the extent of functional recovery in each animal.It could also be argued that the blood supply to the spinal cord below the thoracictransection site was compromised by the surgical procedure at E13. This experimentalartifact could conceivably inhibit repair of descending pathways and/or axonal transport ofretrograde tracer. Physiological evidence (see Chapter 4), however, suggests that spinal38neurons below the transection site remain viable after surgery (Okado and Oppenheim, 1984)and thus the spinal vascular system was probably spared any permanent disruption.Regardless of which factors within the extra-neuronal environment are critical to therepair of a transected spinal cord, it is also important to understand which intrinsic neuronalmechanisms may be contributing to the observed repair of descending supraspinal pathways.Neuronal repair after spinal cord transection may be due to: 1) the mitotic production of newneurons in the brainstem or spinal cord, with the subsequent projection of new fibers throughthe lesion site, 2) subsequent projections from later developing (undamaged) neurons, 3)regrowth of previously axotomized projections (regeneration), or 4) any combination of theseprocesses.The first possibility, that neuronal repair involves the production and differentiationof new neurons in response to transection (ie. continued or renewed neurogenesis) is not alikely mechanism for spinal regeneration in higher vertebrates. Neurogenesis in matureanimals is found in fish and amphibian retina, spinal cord and tectum (fish and premetamorphic Rana pipiens), rodent olfactory epithelium and hippocampus, and songbirdtelencephalic vocal control centers (Anderson and Waxman, 1985; Holder and Clarke, 1988).Other findings in birds, however, have indicated that brainstem-spinal pathways becomepost-mitotic very early in development (usually prior to E5; McConnell and Sechrist, 1980;Sechrist and Bronner-Fraser, 1991). Even in amphibians (e.g. tadpole and salamander),several studies have failed to find evidence for neurogenesis in the brainstem after repair of atransected thoracic spinal cord. This led to the conclusion that anatomical and functionalrecovery in the amphibians resulted from axonal regeneration (Forehand and Farel, 1982;Clarke et al., 1988; Davis et aL, 1989). Therefore it is unlikely that neurogenesis contributessignificantly to the anatomical repair of transected chick embryo spinal cord.The first brainstem-spinal pathways descend to lumbar levels of the cord by E5(Okado and Oppenheim, 1985). Previous and present results suggest that by El 1-E12brainstem-spinal neurons have completed their projections to the lumbar cord and are39equivalent in number and distribution to those observed in a hatchling chick (Okado andOppenheim, 1985; Hasan et a!., 1991, 1992). For the E3-E5 chick embryos examined in thisand previous studies (Clearwaters, 1954; Shimizu et al., 1990; Hasan et al., 1991; Shiga eta!., 1991), few if any brainstem-spinal fibers are present in the thoracic cord at the time oftransection (Okado and Oppenheim, 1985; Hasan et aL, 1991). Therefore, as long as thespinal neurons and glial cells at the lesion site maintain a permissive environment for axonelongation prior to the arrival of supraspinal projections, the correct descending pathwayscan be established. This type of repair cannot be called axonal regeneration, as the fibers arenot directly damaged during the transection.For the E5-E12 chick embryos examined in this study, significant descendingsupraspinal fibers are present within the thoracic cord at the time of transection (Hasan et al,1991; Okado and Oppenheim, 1985). Thus it is possible that some, if not all, of the observedrepair of brainstem-spinal pathways is due to regeneration of previously axotomizedprojections. If the observed anatomical repair were primarily due to subsequent projectionsfrom later developing neurons, it would be unlikely that the number and distribution ofretrogradely labelled brainstem-spinal neurons would be similar in E9-E12 transected andcontrol chicks (see Table 2-1). Terminal sprouting, or regrowth from the proximal cut end ofthe axon, or collateral sprouting has been implicated in the regeneration and functionalrecovery of supraspinal projections after spinal transection in many lower vertebrates,including lamprey (McClellan, 1988, 1990; Lurie and Selzer, 1991), salamander (Davis etal., 1989), axoloti (Clarke et al., 1988), tadpole (Forehand and Farel, 1982), and teleost fish(Bernstein and Gelderd, 1970). Shorter collateral sprouting, or limited lateral regrowth fromthe proximal axon stump, has been reported after spinal cord transection of descendingsupraspinal projections in mammals (Bernstein and Bernstein, 1971; Goldberger, 1973;Goldberger and Murray, 1974; Puchala and Windle, 1977; Martin et al., 1979; Bregman andGoldberger, 1982; Bregman and Bernstein-Goral, 1991). In order to determine if40regeneration (i.e. terminal sprouting) contributes to the observed anatomical repair, doubleretrograde tract-tracing experiments were conducted.41CHAPTER 3ANATOMICAL REPAIR OF TRANSECTED SPiNAL CORDIN EMBRYONIC CHICK(DOUBLE-LABELLING STUDIES)42INTRODUCTIONChapter 2 showed that the embryonic chick is capable of axonal repair aftercomplete transection of the thoracic spinal cord (Hasan et al., 1991, 1992). If the spinaltransection was made prior to embryonic day (E)13 (i.e. on E3-E12), the distribution andnumber of retrogradely labelled brainstem-spinal neurons was similar to sham-transected anduntransected control chicks. Transections conducted on E13 or later resulted in dramaticallyreduced or no retrograde labeffing of brainstem nuclei.There are several neuroanatomical responses that could underlie this embryonicrepair process, including: 1) neurogenesis of new descending brainstem-spinal neurons, 2)subsequent projections from late developing brainstem-spinal neurons, or 3) true axonalregeneration of previously axotomized brainstem-spinal projections. Brainstem-spinalprojections to the lumbar cord develop over a period extending from E4 to El 1-E12 (Okadoand Oppenheim, 1985; Glover and Petursdottir, 1991; Hasan et al., 1991, 1992). It is knownthat brainstem-spinal neurons become post-mitotic very early in development, usually priorto E5 (McConnell and Sechrist, 1980; Sechrist and Bronner-Fraser, 1991), and well beforeE12. Therefore, it is unlikely that the complete anatomical repair of an E7-E12 transectedcord is dependent on the neurogenesis of additional brainstem-spinal neurons (Shimizu et al.,1990; Hasan et al., 1991, 1992). There is apparently no over-production of locus ceruleus orreticulospinal neurons and subsequently no naturally-occurring cell death of these neuronsduring development (Oppenheim, personal communication). Furthermore, by Eli of normalembryonic development, the distribution and number of retrogradely labelled brainstemspinal neurons is equivalent to those labelled in a chick after hatching (Okado andOppenheim, 1985; Hasan et al., 1991, 1992). Therefore, a mid-thoracic spinal transection atE7-E12 would disrupt the many descending projections already at lumbar levels.Nevertheless, it is possible that the subsequent development of axonal projections,that had not descended to thoracic levels at the time of transection, might contribute to theanatomical repair even after a transection as late as Eli-El 2. However, since the43development of brainstem-spinal pathways is nearly complete by El l-E12, it is moreprobable that the repair process involves axonal growth sprouting from axon collaterals orthe proximal ends of previously axotomized fibers (i.e. regeneration). Moreover, these twopotential repair mechanisms are not mutually exclusive.Here, I have investigated the contributions of these two mechanisms to embryonicspinal cord repair. By using two different retrograde fluorescent tracing labels, the firstinjected into the lumbar cord prior to thoracic transection and the second injected severaldays after transection, I established that spinal cord repair in the latter stages of thepermissive period (ElO-El2) is increasingly the result of true axonal regeneration (Hasan etaL, 1992).44MATERIALS AND METHODSEmbryos were incubated, staged, and handled using the same methods detailed inChapter 2 (Materials and Methods).Experimental AnimalsThe entire anatomical protocol (see below) was successfully completed on 31animals which were either ElO (5), El 1 (5), El2 (6), E13 (6), E14 (5), or El5 (4) at the timeof spinal cord transection. Each animal underwent embryonic surgery in ovo on threeseparate occasions during development. The entire procedure is schematically represented inFigure 3-1.On E8-E13, the upper lumbar spinal column (Ll-L2) was pierced with a glassmicropipette (tip diameter = 30 .tm) containing either an aqueous solution of Rhodaminelabelled dextran-amine (RDA; 10,000 MW, lysine fixable, 25% w/v, Molecular Probes Inc.,Eugene, Oregon, catalogue # D-1817), Fluorescein-labelled dextran-amine (FDA; 10,000MW, lysine fixable, 25% w/v, Molecular Probes Inc., Eugene, Oregon, catalogue # D-1820)or Cascade Blue-labelled dextran-amine (CBDA; 10,000 MW, lysine fixable, 25% w/v,Molecular Probes Inc., Eugene, Oregon, catalogue # D-1976). The glass micropipette wasthen inserted into the spinal cord tissue. The micropipette tip was previously marked for thedepth of the spinal cord at that particular age of development. Each dye was diluted to a25% concentration in 0. 1M Tris-Buffer pH 8.5 containing 2.5% Triton X- 100.Approximately 0.1-0.6 IlL of RDA, FDA, or CBDA was pneumophoreticafly injected over aone minute interval using a Picospritzer II (General Valve Corp. Fairfield, New Jersey) thatwas pressurized by nitrogen (3-30 psi). This procedure resulted in the retrograde labelling ofbrainstem-spinal neurons that had projected to the lumbar cord by that age of development(c.f. Fig. 3-lA and 3-2A). Note that potentially late developing neurons (Neuron #2 inFigure 3-lA) remained unlabelled since they had yet to project to the level of the lumbar45Figure 3-1. Schematic representation of experimental procedure for retrograde double-labelling of axotomized brainstem-spinal neurons. A. On E8-E13, the lumbar spinal cord isinjected with the first fluorescent tracing dye. Note that neuron #1 has reached the level ofthe lumbar cord but the late developing neuron (#2) has not yet reached the level of thelumbar cord where the first tracing dye is injected. Neuron #1 therefore becomesretrogradely labelled with the first tracing dye while the late developing neuron (#1) remainsunlabelled. B. One to two days later the mid-thoracic spinal cord is completely transected.Note that the axon of neuron #1 is transected. The late developing neuron (#2), however, hasnot yet reached the level of the transection site and remains intact. C. After an additional 7-8days, the second fluorescent tracing dye is injected into the lumbar cord, caudal to the site oftransection. Note that both neuron #1 and the late developing neuron (#2) have now reachedthe level of the lumbar cord where the second tracing dye is injected. Neuron #1 thereforebecomes retrogradely double-labelled while neuron #2 is retrogradely labelled by only thesecond tracing dye.BrainstemaCervicala IThoracicII .LumbarA1 201 2(’I‘TNIIIII..IIIIIIIII..lstTracerIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII..TransectionIII—-I-—--—IA B CNItA IIHIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII1rr.cS&5S5SIIIIII‘I’ I2ndTracer.r. 0,47cord where the first tracing dye was injected. The egg was then sealed with a sterilecoverslip and paraffin and returned to the incubator.One to two days later (on El0-E15), a complete spinal cord transection was made inthe upper thoracic region Cr4 ± 1 segment) using finely sharpened forceps (Fig. 3-1B). Notethat any late developing neurons (Neuron #2 in Figure 3-1B) would not have beenaxotomized if they had not descended to the level of the transection site. Vertebral tissuesurrounding the spinal cord transection site was slightly damaged by this procedure. Toassure that each transection procedure was complete, a #00 pin was passed laterally throughthe entire spinal cord at the site of transection. The tip of the pin was previously marked forthe depth of the spinal cord at that particular age of embryonic development. The egg wasagain sealed with a sterile coversllp and returned to the incubator.After an additional 7-8 days, a different second fluorescent tracing dye (RDA, FDA,or CBDA) was injected into the lumbar cord in a manner identical to the first injection (Fig.3-iC). Note in the schematic diagram that axonal processes of any late developing neurons(Neuron #2 in Figure 3-iC) would have now reached the level of the lumbar cord where thesecond tracing dye was injected. Therefore, late developing neurons would be retrogradelylabelled by only the second tracing dye (c.f. Fig. 3-2B). To prevent the second injectedretrograde tracing dye from diffusing rostrally to and above the transection site therebyfalsely labelling brainstem-spinal neurons, the second tracing dye was injected several spinalsegments (at least 5 mm) caudal to the lesion at the upper lumbar level. In order to maintaincomparable survival periods following spinal cord transection, embryos transected on ElOthrough E12 received a second injection on E17-E20, while embryos transected on E13-E15received a second injection on E20 to post hatching day (P)2. Twenty four to forty eighthours after the second injection, embryos were anesthetized in ovo with Somnotol (sodiumpentobarbitol, 75 mg/kg IP), removed from their shells, and the omphalomesenteric(umbilical) veins and arteries were cut to allow exsanguination. Hatchlings were alsoanesthetized with Somnotol (sodium pentobarbitol, 75 mg/kg IP). The embryos and48hatchlings were immediately perfused via the left cardiac ventricle over a 30-minute periodwith 50 mL of 0.9% NaC1 (containing 1000 IU of sodium heparin) at 37°C, followed by50 mL of 4% paraformaldehyde in 0. 1M phosphate buffer (pH 8.5, 20°C). The brain andspinal cord were post-fixed for 24 hr and then placed in 0.1 M phosphate buffer salinecontaining 10% sucrose (pH 8.5 , 4°C). After 24 hr, the tissue was transferred to 30%sucrose in 0.1M phosphate buffer saline and stored at 4°C for subsequent sectioning.Equivalent results were obtained with all three fluorescent tracing dyes (RDA, FDA,CBDA) regardless of the order in which they were injected. When two tracing dyes weremixed and injected together in order to detect the relative uptake of each dye, approximatelyequivalent numbers of brainstem-spinal neurons were retrogradely labelled with each dye,and a maximum of 30% of the labelled cells contained both tracing dyes. RDA consistentlylabelled cells the brightest and FDA consistently labelled cells least brightly.Control AnimalsTwo lumbar injections of different retrograde tracing chemicals were completed on26 sham-transected control embryos (four ElO, four El 1, five E12, five E13, four E14, fourE15) and 12 untransected control embryos (two of each embryonic ages ElO to E15). Thesurgical procedures described above were used for all controls. Untransected controlembryos did not have their egg shell opened at the time when they should have beentransected (i.e. they only had two surgical procedures for injection of the two differentretrograde tracers). All experimental and control embryos were treated the same with respectto incubation conditions.To assess the completeness of the transection procedure, 14 experimental embryosthat were transected at different developmental ages (two El 0, two Eli, three El 2, threeEi3, two E14, two E15), were randomly selected from each group of eggs which hadundergone transection. The spinal column was immediately removed following thetransection procedure and embedded in Epon or paraffin. Parasagittal 10 p.m sections were49mounted on gelatin-coated slides, stained with toluidine blue, and examined under a lightmicroscope in order to determine whether the transection severed all spinal cord pathways.HistologyEach animal’s brain and spinal cord were cut at 30 im on a freezing microtome andthe sections placed in 0. 1M phosphate buffer (pH 8.5 at 4°C). The brains were cut in coronalsections while the spinal cords were cut in sagittal sections. Tissue sections wereimmediately mounted on slides for subsequent examination using a Zeiss Axiophotepifluorescent microscope equipped with standard filter blocks. Sections were photographedusing Fujichrome 1600 daylight color reversal film push processed to 3200 ASA (Fuji).Exposure times ranged from 10 to 30 seconds.The positions of retrogradely double-labelled neurons containing both the first andsecond injected dyes within each tissue section were mapped by eye onto drawings ofbrainstem tissue sections. Every third tissue section was counted so as to reduce thepossibility of counting the same neuron twice. The positions of retrogradely single-labelledneurons containing only the first or second injected dye (i.e., RDA, FDA, or CBDA) werealso mapped. Brainstem nuclei were identified using several avian atlases (Karten andHodos, 1967; Youngren and Phillips, 1978; Cabot et al., 1982; Gross and Oppenheim, 1985;Okado and Oppenheim, 1985; Kuenzel and Masson, 1988; Webster and Steeves, 1988). Toallow comparisons between experimental and control animals, cell counts ± the standarderror (S.E.M.), and numerical range, were obtained for each of several brainstem-spinalnuclei on the nght side of each brain. Cell counts were obtained for the nucleus solitarius,nucleus centralis, nucleus reticularis gigantocellularis, and nucleus raphes in the medulla; thenucleus vestibularis lateralis, locus ceruleus, nucleus subceruleus, nucleus reticularis pontiscaudalis/pars gigantocellularis, and nucleus raphes in the pons; and the nucleus interstitialisand nucleus ruber in the mesencephalon. The nucleus paraventricularis magnocellularis andstratum cellulare intemum/externum were also examined for retrograde labelling.50The spinal cord tissue sections were examined for the extent of diffusion of theinjected retrograde tracer chemical as well as the persistance of the dye after injection. Therewas never any evidence of the retrograde tracer diffusing rostrally to the level of thetransection site. As for how long a RDA, FDA, or CBDA dye remains in the spinal cord forretrograde transport by brainstem-spinal neurons, see results (below).51RESULTSResults obtained from experimental animals will be discussed first followed byresults from control animals.Experimental AnimalsTable 3-i is a summary of the number of single- and double-labelled brainstemspinal neurons for several nuclei in ElO, E12, and E14 transected animals. Both control andtransected data are shown.The mean number, range and distribution of brainstem-spinal neurons labelled withthe first dye (injected on E8-E13) increased for each subsequent age of development up toEli. These results were seen in every nucleus examined except for the medullary nucleusreticularis gigantocellularis. This result is expected since the number and distribution ofbrainstem neurons with lumbar projections increases up to El 1-E12 (Okado and Oppenheim,1985; Hasan et al., 1991, 1992).The mean number, range and distribution of brainstem-spinal neurons labelled withthe second dye only (injected on E17-P2) were not significantly different between animalstransected on ElO, Eli or E12. This labelling was also similar to the labelling in controlanimals injected on E17-P2. The mean number, range and distribution of brainstem-spinalneurons retrogradely double-labelled with both the first and second dye, however, increasedfor each subsequent age of development from E1O-Ei2. These results were also seen inevery nucleus examined. These results indicate that brainstem-spinal projections to thelumbar cord were essentially complete by E12. For reasons unknown, double-labelledneurons in the nucleus ruber were observed in only three of 31 experimental animals and twoof 38 control animals. The ratio of double-labelled brainstem-spinal neurons (indicatingregeneration of previously severed axons) to the number of neurons labelled with only thesecond fluorescent tracer (potentially indicating subsequent axonal development of lateTable3-1.Number (mean±S.E.M.)of single-anddouble-labelledbrainstem-spinalneuronsafter temporallyseparatedinjections(shown inbrackets)oftwodifferentfluorescentretrogradetracingdyesintothelumbarcord.Dataareshownforthreerepresentativeandspatiallyseparatednuclei.Thespinalcordoftheexperimentalanimalswastransectedat thethoraciclevel,n=5(E1O),6(E12),and5(E14)forexperimentalsandn=4(E1O),S(E12),and4(E14)forcontrols.Abbreviations:E-embryonicday,P-posthatchingday,Tsx-experimental ageof transection. NucleusVestibularisNucleusLateralisInterstitialis(Pon’J(MsricnhaJOn’)NucleusReticularisGigantocellularis(Mc]ii11’)TsxInjectionExperim.ControlExperim.ControlExperim.Control1st(E8)27.2±5.218.3±3.468.67.852.2±7.221.3±4.428.1±3.6ElO2nd(E18)33.43.337.1±6.878.9±8.181.2±12.596.4±8.870.0±15.0Double3.6±0.84.21.24.8±0.37.8±2.15.3±1.68.0±2.8%Double10.811.36.19.65.511.41st(ElO)17.73.714.1±1.1111.8±22.8102.0±4.288.2±17.879.1j9.7E122nd(E20)27.7±1.022.4±3.296.1±27.272.8±6.580.4±7.787.6±2.9Double8.2±1.16.5±1.620.5±4.523.9±7.417.8±0.922.9±3.1%Double29.629.021.332.822.126.11st(E12)31.76.428.9±1.885.6±18.189.89.979.5j7.168.213.4E142nd(P1)0.OjO.036.3±2.40.0±0.097.6±6.80.0±0.072.3±5.6Double0.OjO.09.6j1.10.0±0.021.5±3.90.0±0.019.5±3.1%Double0.026.40.022.00.026.9L’353brainstem-spinal projections) is expressed as a percentage and is an index of axonalregeneration by severed brainstem-spinal axons at that particular age of embryonicdevelopment. I use the ratio of double-labelled brainstem-spinal neurons to the number ofneurons labelled with only the second fluorescent tracer instead of the first fluorescent tracerbecause the number of neurons labelled with the first tracer increased in controls for eachsubsequent age of development up to Eli. Since the second dye was injected late indevelopment (El7-P2), the number of neurons labelled with only the second fluorescenttracer did not significantly differ within the group of experimental animals and betweenexperimental and control animals. The mean percentage of double-labelled brainstem-spinalneurons increased from 5.1% for animals transected on El 0 to 21.7% for animals transectedon El 2. The index of regeneration was recalculated using the sum of neurons labelled withthe first dye only plus neurons labelled with the second dye only in the denominator. Theresultant change in the ratios from ElO to E12 were not significantly different from thosereported above (see Table 3-1). In addition, very few single- or double-labelled cells werefound in the diencephalon of embryos transected on ElO-El2.The mean number, range and distribution of brainstem-spinal neurons labelled withonly the second dye (injected on E17-P2) were much lower in animals transected on El3-El5 when compared to animals transected on ElO-E12. The mean number, range anddistribution of brainstem-spinal neurons labelled with only the second dye was clearlydecreased in animals transected on E14, compared to animals transected on E13, since therewere no double-labelled cells in animals transected on or after E14. The mean number ofbrainstem-spinal neurons double-labelled with both the first and second dye also decreased inanimals transected on E13-E15. The mean percentage of double-labelled neurons decreasedfrom 2.9% for animals transected on E13 to 0.0% for animals transected on E15.Figure 3-2 shows photomicrographs of retrogradely labelled gigantocellularreticulospinal neurons within the ventromedial reticular formation of the caudal pons in anE20 embryo previously transected on Eli. The first photomicrograph taken with a54Figure 3-2. Photomicrographs of retrogradely labelled gigantocellular reticulospinalneurons within the ventromedial reticular formation of the caudal pons in E20 embryospreviously transected on Eli (permissive transect). A. Brainstem-spinal projections, presentat the time of transection, were previously retrogradely labelled with a lumbar cord injectionof the first label (0.1 iL of FDA) on E9. B. After the thoracic cord transection (El 1), thesecond retrograde tracer (0.3 p.L of RDA) was injected into the lumbar cord on El8. Thepresence of double-labelled brainstem-spinal neurons (n = 4) indicates regeneration ofpreviously axotomized fibers, whereas, the presence of single-labelled RDA neurons (n = 2)suggests subsequent projections from late developing brainstem-spinal neurons. Singlelabelled neurons are indicated by open arrows and double-labelled neurons are indicated bysolid arrows. Plate C indicates the location of photomicrographs in A and B.Abbreviations: Cb- cerebellum, FDA- Fluorescein-labelled dextran-amine, NV- Nervustrigeminus, RDA- Rhodamine-labelled dextran-amine, RPgc- N. reticularis pontis caudalis,pars gigantocellularis. Calibration Bar =50 urn.56fluorescein filter in place (A) shows reticulospinal neurons labelled after the first fluorescenttracing dye (FDA) was injected on E9, prior to the transection. The second photomicrographtaken with a rhodamine filter in place (B) shows reticulospinal neurons labelled after a post-transection injection of the second fluorescent tracing dye (RDA) on El 8. The presence ofdouble-labelled brainstem-spinal neurons indicates regeneration of previously axotomizedfibers, whereas, the presence of single-labelled RDA neurons suggests subsequentprojections from late developing brainstem-spinal neurons.Figures 3-3 and 3-4 show double-exposed photomicrographs of brainstem sectionsfrom one day old (P1) hatchlings retrogradely labelled after a lumbar cord injection of thefirst fluorescent tracing dye (CBDA) on ElO, followed by a lumbar cord injection of thesecond fluorescent tracing dye (RDA) on E20 in an untransected control (A), and in anembryo transected on E12 (B). Figure 3-3 is of reticulospinal neurons within theventromedial reticular formation of the caudal pons, while Figure 3-4 is of vestibulospinalneurons within the lateral vestibular nucleus. Note the similar distribution and number ofretrogradely double-labelled brainstem-spinal neurons in A and B. Panel C (in both Figures3-3 and 3-4) shows double-exposed photomicrographs of three day old (P3) hatchlingsretrogradely labelled after a lumbar cord injection of the first fluorescent tracing dye(CBDA) on E12 followed by a lumbar cord injection of the second fluorescent tracing dye(RDA) on P1 in an animal transected on E14. The absence of double-labelled brainstemspinal neurons indicates a lack of regeneration of previously axotomized fibers. Thepresence of only the first tracer and the absence of brainstem-spinal neurons containing thesecond tracer alone indicates that late developing brainstem-spinal projections did notcontribute to spinal cord repair.Figure 3-5 shows drawings of brainstem tissue sections summarizing the numberand distribution of single- and double-labelled brainstem-spinal neurons at four differentlevels of the brainstem in spinal transected and control animals.57Figure 3-3. Photomicrographs of retrogradely labelled gigantocellular reticulospinalneurons within the ventromedial reticular formation of the caudal pons in hatchling chicks.A&B. Double-exposed photomicrographs of one day old (P1) hatchling brainstem-spinalprojections retrogradely labelled with a lumbar cord injection of the first label (0.1 $.LL ofRDA, red) on ElO followed by a lumbar cord injection of the second label (0.3 tL of CBDA,blue) on E20. A is from an untransected control and B is from an embryo transected on E12(permissive transect). The presence of double-labelled brainstem-spinal neurons (pink)indicates regeneration of previously axotomized fibers, whereas, the presence of single-labelled Cascade Blue neurons suggests subsequent projections from late developingbrainstem-spinal neurons. Note the similar distribution and number of retrogradely double-labelled brainstem-spinal neurons in A and B (see also Table 3-1). A few representativesingle-labelled neurons are indicated by open arrows and a few representative (but not all)double-labelled neurons are indicated by solid arrows. C. Double-exposed photomicrographof three day old (P3) hatchling brainstem-spinal projections retrogradely labelled with alumbar cord injection of the first label (RDA) on E12 followed by a lumbar cord injection ofthe second label (CBDA) on P1 in an embryo transected on E14 (restrictive transect). Theabsence of single- or double-labelled brainstem-spinal neurons containing the second tracer(CBDA) indicates a lack of regeneration of previously axotomized fibers. In order todouble-expose this photograph, the film was first exposed with a RDA filter in place, andthen with a CBDA filter in place. Plate D indicates the location of photomicrographs in A,B, and C. Abbreviations: Cb- cerebellum, CBDA- Cascade Blue-labelled dextran-amine,NV- Nervus trigeminus, RDA- Rhodamine-labelled dextran-amine, RPgc- N. reticularispontis caudalis, pars gigantocellularis. Calibration Bar =50 rim.YRPgc:(SI59Figure 3-4. Photomicrographs of retrogradely labelled vestibulospinal neurons within thelateral vestibular nucleus of the caudal pons in hatchling chicks. A&B. Double-exposedphotomicrographs of one day old hatchling brainstem-spinal projections retrogradely labelledwith a lumbar cord injection of the first label (0.1 iL of RDA, red) on ElO followed by alumbar cord injection of the second label (0.3 tL of CBDA, blue) on E20. A is from anuntransected control and B is from an embryo transected on E12 (permissive transect). Thepresence of double-labelled brainstem-spinal neurons (pink) indicates regeneration ofpreviously axotomized fibers, whereas, the presence of single-labelled Cascade Blue neuronssuggests subsequent projections from late developing brainstem-spinal neurons. Note thesimilar distribution and number of retrogradely double-labelled brainstem-spinal neurons inA and B (see also Table 3-1). A few representative single-labelled neurons are indicated byopen arrows and a few representative (but not all) double-labelled neurons are indicated bysolid arrows. C. Double-exposed photomicrograph of three day old hatchling brainstemspinal projections retrogradely labelled with a lumbar cord injection of the first label (RDA)on E12 followed by a lumbar cord injection of the second label (CBDA) on P1 in an embryotransected on E14 (restrictive transect). The absence of single- or double-labelled brainstemspinal neurons containing the second tracer (CBDA) indicates a lack of regeneration ofpreviously axotomized fibers. In order to double-expose this photograph, the film was firstexposed with a RDA filter in place, and then with a CBDA filter in place. Plate D indicatesthe location of photomicrographs in A, B, and C. Abbreviations: Cb- cerebellum, CBDACascade Blue-labelled dextran-amine, NV- Nervus trigeminus, RDA- Rhodamine-labelleddextran-amine, RPgc- N. reticularis pontis caudalis, pars gigantocellularis. Calibration Bar =50 jim.61Figure 3-5. Drawings summarizing the representative distribution and number ofretrogradely double-labelled brainstem-spinal neurons at four different levels of thebrainstem: A: rostral medulla, B: caudal pons (at the level of the vestibular nuclei), C: rostralpons (at the level of the locus ceruleus), D: mesencephalon (at the level of the red nucleus).For each level (A-D): Part 1 (upper drawing) is an E20 control embryo; Part 2 (middledrawing) is an E20 embryo transected on Eli (permissive transect); Part 3 (lower drawing) isa P2 hatchling transected on E13 (restrictive transect). Note that the distribution and numberof retrogradely double-labelled brainstem-spinal neurons are similar in a control and Elitransected embryo, but the number of retrogradely double-labelled neurons is diminishedafter a thoracic transection on El 3 (see also Table 3-1). Small circles represent oneretrogradely double-labelled brainstem-spinal neuron/tissue section; large circles in thenucleus interstitialis represent five retrogradely double-labelled brainstem-spinalneurons/tissue section. Abbreviations: see list of abbreviations.62Telencephalon.IIIIIIIIIIIIIIIII’IIIIIIIIIIIIRuCerebelluI• I• III •p •IIIIRPgc1mmCnvII• I• II I• II • II • II • II • II II • II • IDO63‘x-xIxexlIDTTDAlRA2TTDR64RP/RPgcRP/RPRP/RPgcBi0000o00o0LaVeMVeLMLFN V0000RB20o000OoVeLMLFNV00 0000RB30 VeLMLF NV0R65Cl00 Scd N0 000o RPgcNVIC200 Scd N0000 oRPgcNVIC3NRPgcNVI66DlD2D367Control AnimalsThe completeness of the spinal cord transection was histologically confirmed in 14randomly selected embryos operated on different days of embryonic development. In allcases, regardless of developmental age, the spinal cord was completely severed at thethoracic level (c.f. Fig. 2-lA and 2-2A and 3-6). Since the same surgical procedure wasundertaken on all experimental animals, I am confident that the surgical technique resulted ina complete transection of each embryo’s spinal cord.It was also necessary to confirm that the injection of the first and second retrogradetracing chemicals was confined to the caudal lumbar spinal cord. If the second fluorescenttracing dye (RDA, FDA, or CBDA) were to directly diffuse to the transection site (or perhapseven to more rostral spinal levels), then it could falsely label brainstem-spinal projectionsthat had not descended caudal to the level of the transection site (i.e. had not regenerated, andtherefore a false positive result). Histological examination of all spinal injection sitesconfirmed that the injected retrograde tracer was confined to within 1-2 segments of therostral lumbar cord and did not diffuse rostrally to the transection site which was at least 5mm away (c.f. Fig. 2-iD and 2-2).It was also critical to discount the possibility that the first injected tracer remained inthe spinal cord long enough for late developing brainstem-spinal neurons to simultaneouslytake up and retrogradely transport both the first and second fluorescent tracing dyes. A totalof twelve embryos were used for this control experiment. Three ElO embryos received alumbar injection of 0.2 pL of CBDA followed by an immediate spinal cord transectionrostral to the injection site. The transection served to prevent the retrograde transport of thefirst tracing dye to the cell bodies of origin in the brainstem. The same animals then receivedan injection of 0.3 ilL of RDA caudal to the transection site on E18. The presence of RDAlabelled neurons indicated repair of brainstem-spinal axons subsequent to the ElOtransection. The lack of CBDA labelled brainstem-spinal neurons indicated that the firstinjected dye did not diffuse rostral to the site of transection, nor did it remain in the spinal68cord for co-transport with RDA by axons projecting caudal to the site of transection (Fig. 3-6). An identical experiment (with three ElO embryos) was conducted in which the order ofthe dyes was reversed. That is, RDA was injected on ElO and CBDA was injected on E18.In this case, as expected, CBDA labelled neurons were present while RDA labelled neuronswere not present.Three E14 embryos received a lumbar injection of 0.1 jiL of CBDA followed by animmediate spinal cord transection rostral to the injection site. The same animals received aninjection of 0.2 tL of RDA caudal to the transection site on P1. The total lack of RDAlabelled neurons indicated little or no repair of brainstem-spinal axons subsequent to the E 14transection. The absence of CBDA labelled brainstem-spinal neurons indicated that this dyedid not diffuse across the site of transection. An identical experiment (with three E14embryos) was conducted in which the order of the dyes was reversed. In this case, asexpected, the same results (i.e. lack of RDA and CBDA labelled neurons) were obtained.Note that these experiments also verified that the spinal cord was completely severed at thetime of transection and that the first injected tracer did not directly diffuse rostrally to the siteof transection.Two retrograde tracers were injected at appropriate developmental ages into thelumbar spinal cord of 26 sham-transected (four ElO, four Eli, five E12, five El3, four E14,four El5) and 12 untransected (two animals at each embryonic age from El0-E15) controlanimals. These control animals were incubated and each retrograde tracer was injected at thesame age as in corresponding experimental transected animals. This was undertaken todetermine the maximum number and distribution of double-labelled brainstem-spinalneurons with lumbar projections for each developmental age. The numbers and distributionof retrogradely single- and double-labelled neuronal cell bodies found in identified brainstemnuclei were equivalent to the labelled neuronal somata found in the experimental animalswhich received a thoracic spinal cord transection on E10-E12.69Figure 3-6. Double-exposed photomicrograph of retrogradely labelled gigantocellularreticulospinal neurons within the ventromedial reticular formation of the caudal pons in anE20 control embryo following an injection of 0.2 iL CBDA into the lumbar cord at ElO andan immediate spinal cord transection rostral to the injection site. The same animalsubsequently received an injection of 0.3 p1 RDA caudal to the transection site at E18. A.The lack of Cascade Blue labelled brainstem-spinal neurons indicates that this dye did notdiffuse across the site of transection nor did it remain in the spinal cord (extracellularly) longenough to be co-transported with RDA by axons crossing the site of transection. Plate Bindicates the location of photomicrograph in A. Abbreviations: Cb- cerebellum, CBDACascade Blue-labelled dextran-amine, NV- Nervus trigeminus, RDA- Rhodamine-labelleddextran-amine, RPgc- N. reticularis pontis caudalis, pars gigantocellularis. Calibration Bar =50 pm.71DISCUSSIONThe anatomical experiments reported here examined whether (and to what extent)axonal regeneration contributes to the anatomical repair of brainstem-spinal projections aftercomplete spinal cord transection at various ages of embryonic development. Whencompared to control animals, these results demonstrate anatomical regeneration ofdescending brainstem-spinal projections in embryos that underwent thoracic cord transectionas late in embryonic development as E12. These data also suggest that regeneration ofpreviously axotomized fibers increasingly contributes to the observed anatomical repair afteran embryonic spinal cord transection prior to E13. Embryos transected on E13-E15 showedminimal to no anatomical regeneration.Anatomical AssessmentsThere are a number of potentially confounding factors which must be consideredbefore accepting the axonal regeneration results. These factors include: 1) the completenessof the transection, 2) direct diffusion of the second tracing dye across the transection site, 3)the possibility of the first injected tracing dye remaining in the spinal cord long enough forlate developing brainstem-spinal neurons to simultaneously take up and retrogradelytransport both the first and second dye, 4) the variability in the quality or number of neuronslabelled with each tracing dye, and finally, 5) transynaptic labelling of brainstem-spinalneurons via intrinsic spinal neurons. The first three issues have been addressed in the resultssection (see Results, Control Animals) and can thus be dismissed as significant confoundingfactors.Regardless of which currently available anatomical tracing technique is used and nomatter how rigorously standardized the methodology, it is well known that there can bevariability between animals in the quality or number of neurons labelled by a retrogradetracing dye (Heimer and RoBards, 1981). Counting retrogradely labelled cells can lead tomistaken estimates (usually under-estimates) about the number of neurons projecting to the72injection site. By calculating the mean and standard error of labelled cell numbers in controlembryos and hatchlings, I was able to estimate the degree of variability inherent in thelabelling technique and thus make comparisons between control and transected animals. Themean number of brainstem-spinal neurons labelled with the first injected dye (injected onE8-E13) increased for each subsequent age of development up to Eli. These results confirmthe findings of Chapter 2 (Hasan et al., 1991, 1992) and those reported by Oppenheim andcolleagues that brainstem-spinal projections complete development to the lumbar cordbetween El 1-E12 (Okado and Oppenheim, 1985; Shimizu et a!., 1990). A comparison to mydouble-label data confirms that no additional neurons project to the lumbar cord after E12.By calculating the ratio of double-labelled brainstem-spinal neurons (indicating regenerationof previously severed axons) to the number of neurons labelled with the second fluorescenttracer alone (potentially indicating the subsequent development of late brainstem-spinalprojections), I determined an index for the extent of regeneration of severed brainstem-spinalaxons at that particular age of embryonic development. This ratio was expressed as apercentage. The mean percentage of double-labelled neurons increased from 5.1% foranimals transected on ElO to 21.7% for animals transected on E12. This increase in thepercentage of double-labelled cells is probably due to the increase in first labelled cells seenbetween ElO and E12. The extent of apparent axonal regeneration decreased, however, foranimals transected on or after E13. Note that these percentages are at best a conservativemeasure of the degree of axonal regeneration since the transport of both retrograde tracers ina brainstem-spinal axon required that the axon be damaged (see below) at the time of eachinjection. The extent of damage could not be guaranteed. The maximum amount of doublelabelling seen in control animals was 32.8% (in the nucleus vestibularis lateralis). Thisindicates that it is virtually impossible to double-label 100% of the neurons in any givenbrainstem nucleus even if all of the neurons in the nucleus project to the injection level of thespinal cord.73Cell death is also a potential variable which may have contributed to the low amountof double-labelling in all of the nuclei examined in all animals. Cell death is unlikely to haveplayed a major role, however, because the percentage of double-labelled cells in theexperimental transected animals was not apparantly significantly lower than the percentageof double-labelled cells in the untransected control animals (see Table 3-1).I also considered the possibility that the labelling of brainstem-spinal neurons withthe second retrograde tracer following spinal cord transection was due to transynapticlabelling via intrinsic spinal neurons that had descended below the transection site andcontacted the second retrograde tracer. This can sometimes be a problem if the post-injection survival time is long as the dextran-amine dyes are transported transynaptically,although poorly. Since the post-injection survival time (following injection of the secondretrograde tracer) was relatively short (48 hr), it is unlikely that transynaptic labellingcontributed to the observed results. Furthermore, it has been shown that uptake of thefluorescent dextran-amines is very dependent on axon damage. Thus, they are not taken upby intact axons en passant, and uptake by terminals is much less efficient than by damagedaxons (Glover et al., 1986). Finally, the distribution of brainstem-spinal neurons retrogradelylabelled with the second tracer in E1O-E12 transected embryos was the same as that seen inintact embryos, hatchlings, and adult birds (Okado and Oppenheim, 1985; Webster andSteeves, 1988; Hasan et al., 1991, 1992).Transynaptic labelling within the brainstem was also contraindicated. Only thosebrainstem nuclei with known direct projections to the lumbar spinal cord were labelled withthe first, second, or both retrograde tracing dyes. No other brainstem nuclei (e.g. the nucleusvestibularis medialis which projects to the cervical cord) were retrogradely labelled. Sincethe first label remained in the cells for up to ten days, it is theoretically possible that someform of transynaptic labelling occurred. Again, the observation that only brainstem nucleithat project to the lumbar cord were labelled discounts this possibility. In addition, controlexperiments in which the embryo was sacrificed and the brainstem was inspected two days74after the first label was injected (data not shown), revealed that the pattern and density oflabelling was the same as when the animal was sacrificed ten days later. These results alsosuggest that transynaptic labelling was not a complicating factor. Very few labelled cellswere found in the diencephalon of embryos transected on E10-E12. Experiments by otherresearchers (Webster and Steeves, 1988) confirm the relatively poor retrograde labelling ofthe diencephalon.Factors Contributing to the Permissive Period of Spinal Cord RepairA number of intrinsic mechanisms may be contributing to the observed anatomicalrepair/regeneration of descending brainstem-spinal neurons after spinal cord transection priorto E13. The possibilities include 1) neurogenesis of new descending brainstem or spinalneurons, 2) subsequent projections from late developing brainstem-spinal neurons, or 3) trueaxonal regeneration of previously axotomized brainstem-spinal projections.Neurogenesis is not a likely mechanism for the repair observed following a thoracicspinal cord transection on E10-E12 because the brainstem-spinal pathways of birds becomepost-mitotic on or before E5 (McConnell and Sechrist, 1980; Sechrist and Bronner-Fraser,1991). There is always the possibility that neurons are generated (from undifferentiatedprecursors) in response to the transection injury. However, this is unlikely, as has beenpreviously discussed (see Chapter 2, Discussion).The results in this and previous studies (Hasan et al., 1991, 1992) together with thework of Okado and Oppenheim (1985) suggest that brainstem-spinal projections havecompleted their projections to the lumbar cord by El l-E12 and are equivalent in number anddistribution to those observed in the hatchling chick. In other words, most of the descendingbrainstem-spinal fibers are present within the thoracic cord at the time of an El 0-El 2thoracic transection. It is highly unlikely, therefore, that the anatomical repair observedfollowing a thoracic spinal cord transection on E10-E12 is exclusively due to subsequentprojections from late developing neurons. Also, the number and distribution of brainstem75spinal neurons retrogradely labelled with the second dye was similar in both El0-E12transected and control chicks. This would not be the case if projections from late developingneurons were substantially contributing to the repair process in E10-E12 transected chicks.The presence of double-labelled brainstem—spinal neurons following a transection aslate as E12 suggests that regeneration of previously axotomized fibers contributes to theobserved anatomical and functional recovery (Hasan et al., 1992). Terminal or axoncollateral sprouting, has been implicated in the regeneration and restitution of function foraxotomized brainstem-spinal projections in many lower vertebrates (Bernstein and Geiderd,1970; Forehand and Farel, 1982; Clarke et al., 1988; McClellan, 1988, 1990; Davis et al.,1989; Lurie and Seizer, 1991). Anatomical evidence for regeneration has also been reportedfor brainstem-spinal projections in mammalian spinal cord (Bernstein and Bernstein, 1971;Goidberger, 1973; Goldberger and Murray, 1974; Puchala and Windle, 1977; Martin et al.,1979; Bregman and Goldberger, 1982; Bregman and Bernstein-Goral, 1991).Factors Contributing to the Restrictive Period ofSpinal Cord RepairThe diminished anatomical repair/regeneration of descending brainstem-spinalneurons in embryos transected on E13-E15 is unlikely to be attributed to a shorter post-transection recovery period. Care was taken to ensure that both E10-E12 transects and E13-E15 transects were allowed equivalent post-transection survival times (7-8 days). There is apossibility that animals transected late in development (e.g. on E14 or E15) could have aslower axonal regrowth rate compared to animals transected earlier (e.g. on ElO or E12).However, when the injection of the second dye was delayed as long as three weeks followingan E14 transection, there were still no double-labelled neurons found in the brainstem (datanot shown). Therefore, it is unlikely that a decrease in axonal regrowth rate is responsiblefor the lack of regeneration seen after an E14 or E15 transection.Giial scarring may have acted as a physical barrier to prevent the subsequent growthor regrowth of brainstem-spinal neurons following a spinal cord transection on E13-E15.76Like others, I was unable to detect any notable glial scarring upon histological examinationof spinal cord sections following thoracic transections at different ages of embryonicdevelopment (Shimizu et al., 1990; Hasan et al., 1991, 1992).A spinal cord transection on or after E13 may have restricted blood supply to thespinal cord below the transection site. This is highly unlikely since physiological evidence(i.e. normal reflex activity) suggests that spinal neurons below the transection site remainviable after surgery (Okado and Oppenheim, 1984). This would suggest that the spinalvascular system must have been spared any permanent disruption.One additional possibility for the initiation of the restrictive period might be thedisappearance of a permissive factor(s) which might facilitate repair and/or the appearance ofa restrictive factor(s) which could inhibit repair. For example, results by other researchers aswell as by my coinvestigators in this laboratory suggest that myelin associated factors mayinhibit axonal regeneration in the spinal cord (Keirstead et al., 1992; Schnell and Schwab,1990; see Chapter 5).77CHAPTER 4FUNCTIONAL REPAIR OF TRANSECTED SPINAL CORDIN EMBRYONIC CHICK(PHYSIOLOGICAL STUDIES)78INTRODUCTIONIn the preceding chapters, I have documented the apparently complete anatomicalrecovery of descending supraspinal projections in embryonic chickens that underwentcomplete thoracic spinal cord transection as late as embryonic day 12 (E12). It has also beenestablished that the anatomical repair in the latter stages (El O-E 12) of the permissive periodfor spinal cord repair is increasingly due to true axonal regeneration of brainstem-spinalaxons. Diminished numbers of retrogradely single- and double-labelled brainstem-spinalneurons were observed when thoracic cord transections were performed on El 3-E 15.Although these retrograde tracing techniques clearly demonstrate that descending supraspinalaxons are able to bridge a thoracic spinal cord transection, the anatomical data cannotdetermine whether these repaired brainstem-spinal projections undergo synaptogenesis withappropriate spinal neurons to form functional circuits and stimulate behavioral recovery.One way to assess functional recovery following spinal cord transection is to allowthe embryos to complete development, hatch, and then observe their locomotor behavior. Ihave observed embryos hatch after spinal transection and have also assisted the hatching oflate transected embryos (those transected on E14 [see Results, Behavioral Observations ofHatchling Motor Function]). The motor and sensory behaviors of El0-E12 transectedanimals were equivalent to age-matched control animals. Hatchling chicks transected onE14, however, had difficulty maintaining an upright standing posture. Thus, the behavioralfindings also parallel the observed anatomical recovery presented in Chapters 2 and 3. Oneshortcoming of this approach, however, is that it is difficult to determine if any apparentbehavioral recovery is due to the establishment of functional connections betweendescending brainstem and spinal neurons or if it is simply due to spontaneously generatedspinal locomotor activity (Shimizu et al., 1990).Therefore, I determined if descending brainstem-spinal projections originating fromdefined avian brainstem locomotor regions (Steeves et al., 1987; Webster and Steeves, 1988;Valenzuela et al., 1990; Sholomenico et al., 199 la,b,c) became functionally active following79repair after complete thoracic spinal cord transection. To accomplish this, establishedtechniques for the stimulation of discrete brainstem locomotor regions in adult birds (Steeveset al., 1987) were adapted for use in embryos and hatchling chicks (Valenzuela et al., 1990;Hasan et al., 1991, 1992). Tn adult ducks and geese, focal electrical or chemical stimulationof brainstem-spinal neurons in the medullary and pontine reticular formation reliably evokesavian locomotor activity (Steeves et a!., 1987; Sholomenko et al., 1991a,b,c). We previouslydemonstrated that stimulation of the same brainstem locomotor regions in intact embryonicchicks can also evoke in ovo locomotion (Valenzuela et a!., 1990; Hasan et a!., 1991). I usedfocal electrical stimulation of these brainstem regions (Rgc- nucleus reticularisgigantocellularis [medullaj, RPgc- nucleus reticularis pontis caudalis pars gigantocellularis[pons]) to assess the locomotor abilities of E18-E20 embryos and one to two day old (Pl-P2)hatchlings that underwent complete high thoracic spinal cord transections at various ages ofembryonic development (E9-E14). Functional locomotor recovery, based on thesephysiological assessments paralleled the observed anatomical repair (Chapters 2 and 3) andwas evident among embryos transected on E9-E12 and tested on E18-E20 or P1-P2.Conversely, limited functional recovery was observed in embryos transected on E13-E14 andtested on E18-E20 or P1-P2.In order for the brainstem stimulation to cause motor activity in the legs, thebrainstem-spinal neurons must have axons which ultimately connect (via the thoracic spinalcord) to lumbar spinal cord interneurons and/or to motoneurons involved in the generation oflocomotion. I found that the muscle activity patterns and the relationships between muscleburst duration and step cycle duration were similar for corresponding leg muscles inhatchling chicks transected on or before E12 and control chicks. In contrast, functionallocomotor recovery was greatly reduced in hatchlings transected on E14. Regardless ofwhether they were assessed before or after hatching, these results indicate that completefunctional repair was evident for chicks that had their spinal cords transected on or beforeE12. These behavioral and physiological findings corroborate the anatomical evidence of the80preceeding chapters which showed complete anatomical repair of brainstem-spinal pathwaysafter spinal transection on or before E12.81MATERIALS AND METHODSBrainstem Stimulation ofEmbryosFocal electhcal stimulation of identified locomotor regions within the reticulospinalregions of the pontine and medullary reticular formation were undertaken on 59 spinaltransected embryos (three E9, ii ElO, 21 Eli, 15 E12, and 9 E13) and 10 control embryos (4untransected embryos, and 2 sham-transected embryos on each of Eli, E12, and E13). Allembryonic brainstem stimulation experiments were undertaken on E18-E20 embryos.Brainstem evoked motor responses were categorized on the basis of visual observations andelectromyographic (EMG) recordings from leg muscles (see below).Prior to the assessment of brainstem-evoked leg activity, electrical stimulation (15-100 pA) of the ventromedial reticular formation had to first evoke synchronous wingflapping in each experimental and control embryo. For experimental animals, the precedingthoracic spinal transection would not have disrupted any descending locomotor pathways tothe lower cervical cord which controls wing movements (Sholomenko and Steeves, 1987;Steeves et al., 1987). Thus, brainstem-evoked wing flapping assured that the lightlyanesthetized embryo was a healthy, viable animal and that all stimulation procedures wereoperating effectively (see below). Consequently, any failure to evoke leg activity could notbe a result of the poor condition of an animal or due to faulty stimulation techniques. All 59spinal transected embryos and all 10 control embryos successfully demonstrated wingflapping in response to focal brainstem stimulation and were subsequently analyzed forevoked leg activity.El8-E20 embryos were maintained at 37°C in a circulating warm water jacket thatsurrounded the eggshell and were anesthetized throughout all surgical and subsequentexperimental procedures by passing 1% Halothane /99% 02 (2L/min) over the eggshell(Fig. 4-i). After 5 minutes of anesthesia, a hole was made in the shell above the air sac andthe chorioallantoic and amnionic membranes were incised and reflected, taking care to avoid82Figure 4-1. Schematic representation of experimental procedures used for in ovo electricalstimulation of brainstem locomotor regions in E18-E20 embryos. Embryos either had theirthoracic spinal cord transected or underwent a sham-operation between E7 and E13. The eggwas placed in a circulating warm water jacket maintained at 37°C. Anesthesia was inducedand maintained by passing 1% halothane / 99% 02 (2L/min) over the surface of the eggshell.After a craniotomy, brainstem locomotor regions were activated by monopolar square wavepulses (current strength = 15-100 tA). Electromyographic (EMG) recordings from variousleg and wing muscles illustrated the timing and pattern of limb movements in response tobrainstem stimulation.8337°CH20STIMULATIONSPINAL TRANSECTIONEMG99% 021% HALOTHANE• •- GAS- EXCHANGE84damage to the major blood vessels. To assure that all animals were maintained at adequatelevels of anesthesia throughout all subsequent surgical and experimental procedures, Irepeatedly tested for a lack of reflex withdrawal to a noxious stimulus (eg. pinching the skinof the scalp or wing).With the embryo thus exposed, the head was placed in a small stereotaxic headholder of our own design. The scalp was incised and reflected, but the cranium was leftintact, since it was soft enough to be easily pierced by the stimulating electrode.Thin flexible stainless steel bipolar hook electrodes were then implanted in thesartorius (SA) muscle of each leg, and the pectoralis (PECT) muscle of each wing. The SAmuscle acts as a strong hip flexor and weak knee extensor and the PECT muscle is the majorwing depressor (Jacobson and Hollyday, 1982a; Dyce et al., 1987). EMG activity wasamplified (l000x), bandpass filtered (100-1000 Hz) and recorded on both FM tape andelectrostatic chart recorders.Brainstem regions within the ventromedial pontine and medullary reticularformation were serially probed with a monopolar stimulating electrode made of glass coatedstainless steel. Stimulation was in the form of square wave pulses (pulse duration 0.5msec) at 60Hz. Electrical current strengths varied from 15-100 tA. When a stimulation trialevoked locomotion (wing and/or leg), the lowest effective stimulation strength (threshold)was subsequently determined and the evoked locomotor EMG patterns were then recorded.Following the stimulation trials, effective brainstem stimulation sites were marked bymaking a small electrolytic lesion (imA DC for 5 sec). The embryo or hatchling wassacrificed by decapitation after which the brain and spinal cord were removed and immersedin 4% paraformaldehyde (24-48 hr) for subsequent histological analysis. The brain was cutcoronally at 30 Lm on a freezing microtome and the tissue sections were immediatelymounted on slides and stained with cresyl violet. The positions of electrolytic lesions used tomark effective locomotor stimulation sites were mapped by eye onto drawings of brainstemtissue sections.85Hatching of Operated Chicks and Behavioral ObservationsA second set of 20 transected (four ElO, eleven E12, five E14), five sham-transected(one ElO, two E12, and two E14), and eight untransected chicks successfully hatched andwere used for both behavioral observations and hatchling brainstem stimulation experiments.On E20-E21, eggs were placed in individual containers in the same incubator.When pipping occurred, the coverslip and paraffin were removed from the operated eggs andthe chicks were allowed to hatch. All of the embryos transected or sham-transected on ElOE12 hatched unassisted, as did all of the control (untransected) animals. All of the embryostransected or sham-transected later in development (E14) only reached the stage where theypipped the shell with their beaks, but failed to hatch on their own. This implies that thespinal cord transection itself was probably not responsible for the hatching failure. I couldsuccessfully assist hatching by cracking open the egg shell and, if necessary, partiallyremoving the chick from the shell and shell membranes when the following conditions wereachieved: 1) the chick had pipped and was breathing on its own, 2) the chorioallantoicmembrane had very little blood flow, and 3) the yolk-sac had become enclosed within thebody cavity. Locomotor activity (volitional walking or running towards a food reward) andconscious sensation (vocalization in response to toe pinch) were assessed in all transectedand sham-transected chicks. Proprioception (the ability to right body when placed on back)was also examined. These data were compared to data obtained from normal (untransected)hatchlings. After behavioral observation, all of the transected and control hatchling chickswere examined for brainstem-evoked locomotion (see below).Brainstem Stimulation ofHatchlingsFocal electrical stimulation of identified locomotor regions within the reticulospinalregions of the pontine and medullary reticular formation was undertaken on 20 (four ElO,eleven E12, five E14) spinal transected and 13 control animals (one ElO, two E12, and two86E14 sham-transects and eight untransected animals) at the ages of P1 or P2. Brainstemstimulation was used to assess whether the observed anatomical regeneration also correlatedwith the functional repair of brainstem-spinal pathways. Brainstem-evoked motor responseswere categorized on the basis of visual observations and electromyographic (EMG)recordings from leg muscles (see below and Valenzuela et al., 1990, and Hasan et aL, 1991,1992).For implantation of EMG recording leads, hatchling chicks were deeplyanesthetized with 3% Halothane / 97% 02 (2lJmin). Thin flexible stainless steel bipolarhook electrodes were implanted in the sartorius, femorotibialis, iliofibularis, caudioflexorius,gastrocnemius lateralis, and tibialis anterior muscle of the right leg. In addition, an electrodewas also implanted in either the sartorius, femorotibialis, or the gastrocnemius lateralismuscle of the left leg in order to record any alternating leg activity (Fig. 4-2). The sartoriusmuscle is a hip flexor, knee extensor; the femorotibialis muscle is a knee extensor; theiliofibularis muscle is a hip extensor, knee flexor; the caudioflexorius muscle is a hipextensor, knee flexor; the gastrocnemius lateralis muscle is an ankle extensor; and the tibialisanterior muscle is an ankle flexor (Jacobson and Hollyday 1982a). Electrodes were alsoimplanted in the pectoralis (PECT) muscle of each wing. The PECT muscle is the majorwing depressor (Dyce et al., 1987). Chicks were allowed to recover (18-24 hr) beforebrainstem stimulation experiments.Prior to the brainstem stimulation experiments, hatchling chicks were again deeplyanesthetized with 3% Halothane / 97% 0 (2L/min). The head was then placed in astereotaxic head holder and the hatchling was allowed to stand on a treadmill situated justunder the head holder. The calvarium and dura mater were resected and the cerebralhemispheres removed by suction first along a plane extending from the habenular nucleus(dorsally) to the optic chiasm (ventrafly), and then forward to the olfactory bulbs. Thisregion of brain includes parts of the hypothalamus and thalamus. Upon completion of thedecerebration, anesthesia was discontinued since decerebrate animals are devoid of any87Figure 4-2. Schematic representation of experimental procedures used for in vivo electricalstimulation (current strength = 15-50 j.LA) of brainstem locomotor regions in transected,sham-transected, and untransected 1-2 day old hatchling chicks. Electromyographic (EMG)recordings were utilized to monitor brainstem-evoked locomotor activity in six muscles ofthe right leg including the sartorius (hip flexor, knee extensor), femorotibialis (kneeextensor), iliofibularis (hip extensor, knee flexor), caudioflexorius (hip extensor, kneeflexor), gastrocnemius lateralis (ankle extensor), and tibialis anterior (ankle flexor) muscle.In addition, either the sartorius, femorotibialis, or the gastrocnemius lateralis muscle of theleft leg was implanted in order to record alternating leg activity. EMG recordings were alsomonitored in the pectoralis wing (wing depressor) muscles (not shown). Lower right panelshows diagram summarizing the positions of effective brainstem locomotor stimulation siteswithin the ventromedial reticular formation of the rostral medulla of sham-transectedhatchlings (filled circles) and hatchlings that had their thoracic spinal cord transected on ElOor E12 (open circles). Abbreviations: IX- N. nervi glossopharyngei, EMGelectromyographic, R- N. raphes, Rgc- N. reticularis gigantocellularis, Rpc- N. reticularisparvocellularis (medulla), TTD- N. et tractus descendens nervi trigemini.II,89conscious perception of pain (Wall, 1975; Wall and Sternbach, 1976; Steeves et aL, 1987;Valenzuela et al., 1990).Brainstem regions within the ventromedial pontine and medullary reticularformation were sequentially probed with a glass coated stainless steel monopolar stimulatingelectrode as the treadmill moved beneath the hatchling’s feet (treadmill speed = 0.1 meters/s).Stimulation was in the form of square wave pulses (pulse duration = 0.5 msec) at 60Hz.Electrical current strengths varied from 15-50 jiA.Prior to the assessment of brainstem-evoked leg activity, electrical stimulation (15-50 pA) of the ventromedial reticular formation had to first evoke synchronous wing flappingin each experimental and control hatchling (see above). As in the embryonic brainstemstimulation experiments, brainstem-evoked wing flapping assured that the decerebratehatchling was a healthy, viable animal and that all stimulation procedures were effective (seebelow). All 20 spinal transected animals and all 13 control chicks successfully demonstratedwing flapping in response to focal brainstem stimulation and were subsequently analyzed forevoked leg activity.When a stimulation trial evoked locomotion (wing and/or leg) on the movingtreadmill, the lowest effective stimulation strength (threshold) was subsequently determinedand the evoked locomotor EMG patterns were then recorded. EMG’s were amplified(l000x), bandpass filtered (100-1000 Hz), digitally converted (R.C. Electronics Inc., SantaBarbara, California) and stored on computer disk for subsequent analysis. Following thestimulation trials, brainstem stimulation sites effective at eliciting leg movement weremarked by making a small electrolytic lesion (1 mA DC for 5 sec).At the end of the experiment, the hatchling was sacrificed by decapitation and thebrain and spinal cord were then removed and immersion fixed in 4% paraformaldehyde forsubsequent histological confirmation of the stimulation site. Standard histologicaltechniques were used for preparation of brainstem tissue sections as discussed above (seeMaterials and Methods, Brainstem Stimulation of Embryos).90RESULTSBrainstem Stimulation ofEmbryosThere were no significant differences in the motor responses evoked from thedifferent brainstem stimulation sites. Nor were there any significant or consistent differencesin the evoked motor responses of embryos transected between E9 and E12. Table 4-1summarizes the number of embryos that were observed to have: 1) brainstem-evoked legactivity of any type (often uncoupled, arrhythmical leg movements as judged by visualobservations and EMG recordings of muscle activity patterns); and 2) brainstem-evokedstepping (based on visual observations and EMG recordings of coordinated, rhythmicalalternating activity from leg muscles). To differentiate between spontaneous spinalgenerated activity and brainstem-evoked leg locomotor activity, I defined the “evoked” legmovements as commencing with the onset of electrical stimulation and terminating with theoffset of stimulation. In addition, the frequency of brainstem-evoked stepping had to varywith changes in the electrical stimulation strength.Brainstem-evoked leg activity ranged from a low of 33% (3/9) in the E13 transectedembryos to a high of 82% (9/11) of the ElO transected embryos. A sharp contrast inbrainstem-evoked leg activity was noted between embryos transected on E12 and thosetransected on E13. The percentage of E9-E12 transected animals that responded with evokedleg activity was comparable to the 80% average for controls (sham-transected=5/6 and intact=3/4) and contrasted with the 33% average for E13 transected animals.Of the E9-E12 transected embryos that exhibited brainstem-evoked leg activity,approximately 64% (19/30) showed coordinated stepping. Of the E13 transected embryosdemonstrating brainstem-evoked leg activity, only 33% showed coordinated stepping (1/3).The identified brainstem locomotor stimulation sites (Fig. 4-3, Steeves et al., 1987;Valenzuela et al., 1990; Hasan et al., 1991, 1992) within the medial reticular formationcorrespond to the gigantocellular divisions of the pons (RPgc) and medulla (Rgc), both ofTable4-1.Number ofembryosshowinglocomotoractivityinresponsetofocalelectricalstimulationof abrainstemlocomotorregion.ExperimentalControl(transectedon)LocomotorActivityE9.EUEUEUEUShamIntactBrainstem-Evoked2/39/1110/219/153/95/63/4LegMovements(uncoordinated)Brainstem-Evoked1/36/116/216/151/93/63/4Stepping(coordinated)I-J92Figure 4-3. A. Photomicrograph of an E18 chick brainstem section stained with cresylviolet. This photomicrograph displays the electrolytic lesion used to mark an effectivelocomotor stimulation site in the nucleus reticularis gigantocellularis, adjacent to the nucleusraphes (R) on the midline. B. Diagrams of representative brainstem levels in a chick,summarizing the positions of effective brainstem locomotor stimulation sites in shamtransected (control) embryos and hatchlings (filled circles) and in embryos and hatchlingsthat had their thoracic spinal cord transected on or prior to E12 (open circles).Abbreviations: III- N. oculomotorius, IX- N. nervi glossopharyngei, BC- brachiumconjunctivum, EW- N. of Edinger-Westphal, ICo- N. intercollicularis, MLd- N.mesencephalicus lateralis, pars dorsalis, MRF- Formatio reticularis mesencephali, R- N.raphes, Rgc- N. reticularis gigantocellularis, Rpc- N. reticularis parvocellularis (medulla),TPc- N. tegmenti pedunculo-pontinus, pars compacta, TT’D- N. et tractus descendens nervitrigemini.AgEWICocD U)94which have direct projections to the lumbar spinal cord. These sites are identical to regionsshowing retrogradely labelled brainstem-spinal neurons after a thoracic spinal cordtransection performed prior to E13 (see results above and Fig. 2-3, 2-5, 3-2, 3-3, 3-5; Okadoand Oppenheim, 1985; Steeves et al., 1987; Hasan et aL, 1991, 1992).The thresholds for evoking uncoupled arrhythmical leg activity or coordinatedlocomotor activity were comparable in all animals (experimental and control) and averaged50 iiA (range = 10-100 j.tA). When a brainstem locomotor region was effective in evokinglocomotion on El8-E20, the frequency of the evoked stepping was similar in bothexperimental embryos Iransected on or before E12 and sham-transected and untransectedcontrol chicks (Fig. 4-4A and 4-4B). The average step cycle duration for transected embryoswas 0.186±0.021 sec (step frequency = 5.4Hz), which was equivalent to the average stepcycle duration in control embryos of 0.191±0.025 sec (step frequency = 5.2Hz). Thesebrainstem-evoked step cycle durations are faster than the average cycle duration recorded forspontaneous stepping activity in the same embryos (cycle duration = 0.415±0.0 15 sec; stepfrequency = 2.4Hz). Simultaneous brainstem-evoked stepping and wing flapping could alsobe evoked in E9-E 12 transected chicks. The EMG patterns of muscle activity were similar tothose evoked in control animals. Finally, I found that in both the E9-E12 transected andcontrol animals an increase in the brainstem stimulation current strength evoked a fasterstepping frequency (Fig. 4-5). This further substantiates the equivalence of evoked steppingin the E9-E12 spinal iransected and control groups.Behavioral Observations ofHatchling Motor FunctionMi 15 of the embryos transected on Ei0-E12 hatched unassisted (four ElO, elevenEl2). All animals were healthy, ate and drank well, and showed no signs of discomfort. TheEl0-E12 transected chicks were able to walk, run, hop and roll over in a manner that wasindistinguishable from El0-E12 sham-transected control hatchling chicks. All three of theembryos sham-transected on E10-El2 hatched unassisted (one ElO, two E12). Vocalization95Figure 4-4. Brainstem evoked locomotor activity in a sham-transected control (A) and aspinal transected (B) embryonic chick. A. Alternating electromyographic (EMG) activityfrom the left and right sartorius (L. SA, R.SA) leg muscles during evoked stepping in ovo byan embryonic day 19 (E19) sham-transected control embryo. The in ova stepping wasevoked by focal electrical stimulation of the ventromedial pontine reticular formation(cuffent strength = 50 pA). B. Alternating EMO activity from the left and right leg sartoriusmuscles evoked in ova from an E19 embryo whose thoracic spinal cord had been transectedon E12 (permissive transect). The pattern of locomotor activity was similar to evokedstepping in sham-transected embryos and was initiated in response to stimulation of theventromedial pontine reticular formation (50 jiA). The evoked embryonic leg activity wasalso visually observed. Abbreviations: L. SA- left sartorius muscle. R. SA- right sartoriusmuscle.ALSAR.SABsecLIIrALiLLLJi’lJ’i-jiiL.LIL-jLiu1‘.0 ON94which have direct projections to the lumbar spinal cord. These sites are identical to regionsshowing retrogradely labelled brainstem-spinal neurons after a thoracic spinal cordtransection performed prior to E13 (see results above and Fig. 2-3, 2-5, 3-2, 3-3, 3-5; Okadoand Oppenheim, 1985; Steeves et a!., 1987; Hasan et al., 1991, 1992).The thresholds for evoking uncoupled arrhythmical leg activity or coordinatedlocomotor activity were comparable in all animals (experimental and control) and averaged50 p.A (range = 10-100 ptA). When a brainstem locomotor region was effective in evokinglocomotion on E18-E20, the frequency of the evoked stepping was similar in bothexperimental embryos transected on or before E12 and sham-transected and untransectedcontrol chicks (Fig. 4-4A and 4-4B). The average step cycle duration for transected embryoswas 0.186±0.021 sec (step frequency = 5.4Hz), which was equivalent to the average stepcycle duration in control embryos of 0.191±0.025 sec (step frequency = 5.2Hz). Thesebrainstem-evoked step cycle durations are faster than the average cycle duration recorded forspontaneous stepping activity in the same embryos (cycle duration = 0.415±0.0 15 sec; stepfrequency = 2.4Hz). Simultaneous brainstem-evoked stepping and wing flapping could alsobe evoked in E9-E12 transected chicks. The EMG patterns of muscle activity were similar tothose evoked in control animals. Finally, I found that in both the E9-E12 transected andcontrol animals an increase in the brainstem stimulation current strength evoked a fasterstepping frequency (Fig. 4-5). This further substantiates the equivalence of evoked steppingin the E9-E12 spinal transected and control groups.Behavioral Observations ofHatchling Motor FunctionAll 15 of the embryos transected on E10-E12 hatched unassisted (four ElO, elevenE12). All animals were healthy, ate and drank well, and showed no signs of discomfort. TheE10-E12 transected chicks were able to walk, run, hop and roll over in a manner that wasindistinguishable from E10-E12 sham-transected control hatchling chicks. All three of theembryos sham-transected on E10-E12 hatched unassisted (one ElO, two E12). Vocalization95Figure 4-4. Brainstem evoked locomotor activity in a sham-transected control (A) and aspinal transected (B) embryonic chick. A. Alternating electromyographic (EMG) activityfrom the left and right sartorius (L. SA, R.SA) leg muscles during evoked stepping in ovo byan embryonic day 19 (E19) sham-transected control embryo. The in ovo stepping wasevoked by focal electrical stimulation of the ventromedial pontine reticular formation(current strength = 50 tA). B. Alternating EMG activity from the left and right leg sartoriusmuscles evoked in ovo from an E19 embryo whose thoracic spinal cord had been transectedon El 2 (permissive transect). The pattern of locomotor activity was similar to evokedstepping in sham-transected embryos and was initiated in response to stimulation of theventromedial pontine reticular formation (50 tA). The evoked embryonic leg activity wasalso visually observed. Abbreviations: L. SA- left sartorius muscle, R. SA- right sartoriusmuscle.97Figure 4-5. Dependence of in ovo evoked stepping frequency on brainstem stimulationcurrent intensity. Focal electhcal stimulation of the ventromedial pontomedullary reticularformation evoked alternating EMG activity from the left and right sartorius (L. SA, R. SA)leg muscles of an E18 embryo whose thoracic spinal cord had been transected on Eli(permissive transect). A. Brainstem simulation of 30 tA evoked in ovo stepping at a stepfrequency of 1.25 Hz (cycle duration = 0.8 sec). B. Increasing the current strength of thebrainstem stimulation to 60 p.A increased the step frequency to 3.0 Hz (cycle duration = 0.33sec). Similar current intensity locomotor frequency relationships have been noted in all otherbrainstem stimulated locomoting preparations. Abbreviations: L. SA- left sartorius muscle,R. SA- right sartorius muscle.98[rALSAR.SABLSA1 secR. SA99and limb withdrawal in response to toe pinch was the same for both groups. AM E10-E12transected, sham-transected and untransected chicks exhibited similar goal-directedmovement (volitional walking or running) towards a food reward. In short, the motor,sensory and hatching behaviors of the El0-E12 transected animals were equivalent to age-matched sham-transected animals and untransected hatchling chicks.All five of the embryos transected on E14 survived an assisted hatching. Bothembryos sham-transected on E14 also survived an assisted hatching. This suggests thatdisruption of the shell and chorioallantoic membranes and experimental movement of theembryo during any embryonic surgery after E12 requires that the embryo be assisted duringhatching.Chicks transected on E14 had difficulty maintaining an upright standing posture.These chicks also had difficulty maintaining lateral stability and were only capable of takinga few stumbling steps. Conversely, chicks sham-transected on E14 showed normal posture,volitional walking and running. After behavioral observation, all of the transected andcontrol hatchling chicks were examined for brainstem-evoked locomotion (see below).Brainstem Stimulation ofHatchlingsAll hatchling brainstem stimulation experiments were performed on Pl-P2. Focalelectrical stimulation of discrete brainstem locomotor regions was undertaken on 20experimental animals and 13 control hatchlings. The identified brainstem locomotorstimulation sites (c.f. Fig. 4-2, see also Steeves et al., 1987; Valenzuela et al., 1990; Hasan etal., 1991, 1992) within the medial reticular formation correspond to the gigantocellulardivisions of the pons (RPgc) and medulla (Rgc), both of which have direct axonal projectionsto the lumbar spinal cord (c.f. Fig. 2-3, 2-5, 3-2, 3-3, and 3-5, see also Okado andOppenheim, 1985; Steeves et a!., 1987; Hasan et al., 1991, 1992) and are identical to thosebrainstem regions showing retrogradely double-labelled brainstem-spinal neurons after athoracic transection performed prior to E13 (see results above and Fig. 2-3, 2-5, 3-2, 3-3, and1003-5). There were no significant differences in the motor responses evoked from the differentbrainstem stimulation sites. There were no significant or consistent differences in the evokedmotor responses of embryos transected on ElO as compared to those embryos transected onE12.Table 4-2 summarizes the number of chicks that showed brainstem-evoked stepping(based on visual observations and EMG recordings of coordinated, rhythmical alternatingactivity from leg muscles). To differentiate between any spontaneous spinal generatedactivity and true brainstem-evoked leg locomotor activity, the leg movements had tocommence with the onset of electhcal stimulation and terminate with the offset ofstimulation. Brainstem-evoked coordinated stepping ranged from a low of 0% (0/5) of theE14 transected hatchlings to a high of 87% (13/15) of the E10-E12 transected hatchlings.The percentage of E10-E12 transected animals that responded with coordinated stepping wascomparable to the 85% average (11/13) for control hatchlings.The thresholds for evoking coordinated leg locomotor activity in hatchlings werecomparable in all animals (experimental and control) and averaged 25 tA (range = 15-50pA). When a brainstem locomotor region was effective in evoking locomotion on P1-P2, themuscle activity patterns were not significantly different between experimental hatchlingstransected on or before E12, sham-transected and untransected control chicks (Fig. 4-6, 4-7,4-8). In addition, the relationships between muscle burst duration and step cycle durationwere similar for corresponding muscles in hatchling chicks transected on E12 and controlchicks (Fig. 4-9). The step cycle duration for transected animals ranged from 0.1 - 1.6seconds, comparable to step cycle durations in control animals of 0.1 - 1.8 seconds.Simultaneous brainstem-evoked stepping and wing flapping could also be evoked in ElOE 12 transected chicks and the EMG patterns of activity were similar to those evoked incontrol hatchlings. Finally, I found in both the El0-E12 transected and control animals thatan increase in the brainstem stimulation current strength evoked a faster stepping frequency.Figure 4-5 is an example of this relationship in an embryo (E18). No data is shown for the101hatchlings. This relationship between stimulation current strength and stepping frequencyfurther substantiates that the evoked stepping in the spinal transected and control groups wasdue to functional brainstem-spinal projections.Table4-2.Number ofanimalsshowingevokedsteppingactivityinresponsetofocalelectricalstimulationofabrainstemlocomotorregiononposthatchingday(P)1orP2.ExperimentalControl(transectedon)LocomotorActivityElOE12E14ShamIntactBrainstem-Evoked2/411/110/53/58/8Stepping(coordinated)C103Figure 4-6. Brainstem evoked locomotor activity in a sham-transected control (A) and aspinal transected (B) hatchling chick. A. Co-activation of leg and wing locomotor activity inresponse to focal electrical stimulation of the ventromedial mesencephalic reticularformation (75mA) in a two day old (P2) hatchling chick (untransected control). Synchronous(in phase) bilateral wing flapping activity from the left and right pectoralis (L. PECT, R.PECT) muscles and alternating EMG activity from the left and right sartorius (L. SA, R. SA)leg muscles were simultaneously evoked. B. Co-activation of leg and wing locomotoractivity in response to focal electrical stimulation of the ventromedial pontine reticularformation (75mA) of a two day old (P2) hatchling, previously transected on E12 (permissivetransect). The evoked locomotor activity consisted of in-phase wing flapping movements,corresponding to bilateral EMG activity recorded from the pectoralis muscles (top twotraces), and simultaneous alternating activity from the left and right leg sartorius muscles.Abbreviations: L. PECT- left pectoralis muscle, R. PECT- right pectoralis muscle, L. SAleft sartorius muscle, R. SA- right sartorius muscle.ill A1U1II1B II.UJLLILI.UShIWI £1LUIIAJUliiiILIL r.AL.PECTR.PECTLSAR.SAT.;TrI,ry.T’TTT’I‘r”rj ‘F‘rTTU’rfl?TTTT}!TTT1111lITIT77pJnTTi Till!1Nfl141N4)F-1 0 .r.1sec105Figure 4-7. Simultaneous electromyographic recordings from four muscles of the right legduring brainstem-evoked locomotion in a two day old (P2) hatchling chick previously shamtransected (A) or transected (B) on E12 (permissive transect). The EMG recordings fromhatchling chicks transected on El 2 show the same muscle activity patterns as recordingsfrom hatchling chicks sham-transected on E12. A more detailed description of muscleactivity patterns is presented in Figure 4-8. Abbreviations: Ff- femorotibialis muscle, GLgastrocnemius lateralis muscle, IF- iliofibularis muscle, SA- sartorius muscle.106Ankle GIHip SAA CONTROLun 4 $Ø[ iii4I44J$tahIflv4Ln4PflL:ie IuIwr- .111*.TWW1 jriIj-,wlF’-’ wyr- fr_WTLjIII [ i. Li! ti 4$..44uiw..Ør4.a*rkij1 secB E12 TRANSECT* *t *4•‘I’*n1$*.4 +‘-4•inØ ‘iq in i’ in‘- áia.aiiei’--.-1 ljdII-””a1uAlhiiFT Ai&KneeIFAnkle GIHip SAFTKneeIF“IF rrrlr •p” •-—hI!7-1 jr .—j ‘1M -—107Figure 4-8. Summary of temporal relationships for leg muscle activity patterns duringoverground walking. Horizontal axis is time normalized for the duration of the step cycle.Two complete step cycles are presented in each diagram. Thick bars represent the proportionof the step cycle that each muscle is active. Filled bars represent stance phase activity andopen bars represent swing phase activity. Lines at the ends of each bar represent the standarderror for the onset/offset time for each muscle. A. Mean muscle activity pattern in controlhatchlings (n = 6). The onset of activity of the sartorius muscle (SA- hip flexor, kneeextensor) has been shown to occur just prior to the onset of the swing phase of the step cycle(Jacobson and Hollyday, 1982a) and was considered here to be the beginning of the stepcycle. The sartorius muscle continues to be active for most of the swing phase. Thefemorotibialis (Fr- knee extensor) and iliofibularis (IF- hip extensor, knee flexor) musclesare also active during the swing phase. The iiofibularis muscle is active twice during eachstep cycle, once during the swing phase and once during the stance phase. The bursting ofthe caudioflexorius (CF- hip extensor, knee flexor) and gastrocnemius lateralis (GL- ankleextensor) muscles alternate with the bursting of the sartorius muscle and are active during thestance phase. B. Mean muscle activity patterns in hatchlings which underwent spinal cordtransection at E12 (n = 6). There were no significant differences from muscle activitypatterns recorded in control hatchlings. The similarity of the muscle activity patterns of E 12transected hatchlings to those of control hatchlings supports the observation that E12transected chicks show complete functional recovery. Abbreviations: CF- caudioflexoriusmuscle, FT- femorotibialis muscle, GL- gastrocnemius lateralis muscle, IF- iliofibularismuscle, SA- sartorius muscle.108A CONTROLSB E12 TRANSECTSKneeSAl I—HipCF—cKneeIF —1ZEJ---•Ankle GL1-I- I-00.0I I I I I0.5 1.0 1.5 2.0HipSAl I—CFrrI IIF .—lJJ——Ankle GL0.0 0.5 1.0 1.5 2.0CONTROLSE12TRANSECTSI___________________________________________________________________________________0.80.80.7•slope=O.450.7a0.6-—10.6•slope=0.4950.5eN•0.40.3I-0.20.1•-•—0.1•——.••__•&u0:2040:508i1.214161.8V02040508I121:416cycleduration(sends)cyclecitation(sen)4.1.0.8•0.8Vo.islope=0.040.7Io.0.6&ope=O.08Sos505_•_—0.4••0.4•—•0.2.:.-02-0.1——••0.1•00:20:40:6081121:416180:20:4•0:608I:sücycledsation(sends)cyclechxation(sends)109Figure 4-9. Linear regression of leg muscle activity (burst) duration versus step cycleduration for gastrocnemius lateralis (ankle extensor, stance phase) muscle (A and B) andsartorius (hip flexor, knee extensor, swing phase) muscle (C and D) during brainstem-evokedlocomotion of hatchling chicks. Typical of birds (Jacobson and Hollyday, 1982a), as cycleduration increases, the burst duration of the lateral gastrocnemius muscle increases (A and B)while the burst duration of the sartorius muscle remains constant (C and D). Controlhatchlings (A and C) display the same relationships between burst duration and cycleduration as do El 2 (permissive) transected hatchlings (B and D), as indicated by the lack ofsignificant difference between the slopes of the regression lines in A and B and between theslopes of the lines in C and D. The similarity of the burst duration vs. cycle durationrelationships between control hatchlings and E12 (permissive) transected hatchlings supportsthe observation that E12 transected chicks show complete functional recovery. R2=0.49,0.06, 0.65, 0.13 for A, B, C, D, respectively, n =6 for controls and n =6 for E12 transects.111DISCUSSIONThese physiological results suggest complete functional repair of descendingbrainstem-spinal projections in embryos and hatchlings that underwent thoracic cordtransection as late in embryonic development as E12. The extent of the functional repair wasdiminished or altogether absent in those animals transected after El 2. In order to ensure thatany failure to evoke leg activity was not due to the poor condition of the animal or to faultyexperimental techniques, brainstem-evoked wing flapping was assured prior to examiningevoked leg activity as an internal control for each animal.After an embryonic spinal cord transection, brainstem-evoked leg activity inembryos or hatchlings can be accepted as a reliable indicator for the functional repair ofbrainstem-spinal projections if the following criteria are satisfied: 1) the leg activity mustcommence with the onset of brainstem stimulation and terminate with the offset ofstimulation, 2) the leg activity must be evoked in response to current strengths that would notdirectly activate the spinal cord, 3) an increase in the brainstem stimulation current strengthshould evoke leg movements at a faster frequency, 4) the pattern and frequency of legactivity evoked in a transected animal should be similar to the pattern and frequency ofevoked leg activity in an untransected or sham-transected control animal, and 5) therelationships between muscle burst duration and step cycle duration should be similar forcorresponding muscles in transected and control animals (only done in hatchlings).The present results satisfy all of the above criteria. Brainstem-evoked leg activity(uncoordinated or coordinated stepping movements) always began within 0.1-0.5 sec of theonset of brainstem stimulation and terminated within 0.1 sec when stimulation ceased(Valenzuela et al., 1990; Hasan et al., 1991, 1992). The brainstem stimulation currentstrengths necessary to evoke stepping were 15-100 tA in embryos and 15-50 jiA inhatchlings. These differences could be due to: a) the fact that the embryos were anesthetizedand the hatchlings were not; or b) differences in nervous system maturity between theembryos and the hatchlings. The effective current spread from a monopolar electrode using112a 100 tA stimulus has been calculated to be at most 0.5 mm from the electrode tip (Stoney etal., 1968; Wise, 1972). This is well short of the 0.5 cm distance between the stimulation siteand the cervical spinal cord and the 2.0 cm distance from brainstem locomotor regions to thethoracic transection site. The evoked stepping frequency could be increased as the result ofan increase in the stimulation current strength (c.f. Fig. 4-5), which is a characteristic ofbrainstem-evoked locomotion in all vertebrates, including birds (Steeves et al., 1987). Thepattern of evoked stepping in transected embryos and hatchlings was similar to the evokedalternating pattern of stepping in control embryos and hatchlings (c.f. Fig. 4-4, 4-6, 4-7, 4-8).In addition, the relationships between muscle burst duration and step cycle duration weresimilar for corresponding muscles in hatchling chicks transected on E12 and control chicks(c.f. Fig. 4-9).How late in development can effective functional repair of a transected spinal cordoccur? The previous study of Oppenheim and colleagues narrowed the “window” to betweenEl 0 and El 5 (Shimizu et al., 1990). The combined results of Chapters 2 and 3 haveestablished that there is virtually complete anatomical repair of brainstem-spinal pathwaysfollowing an E12 transection and greatly diminished repair following an E13 transection.Furthermore, axonal regeneration contributes to the repair of injured spinal cord. The resultsof this chapter extend these observations and demonstrate that the anatomicalrepair/regeneration is correlated with functional recovery. This suggests that there is apermissive and a restrictive embryonic period for both anatomical repair and functionalrecovery after spinal cord injury. Thus, during the transition to the restrictive period therepair process appears to be transforming into the limited repair abilities of the adult spinalcord (Eidelberg, 1981; Sholomenko and Steeves, 1987).113CHAPTER 5GENERAL DISCUSSION114I have observed apparently complete anatomical repair of descending brainstemspinal projections in chicken embryos that underwent thoracic cord transection prior toembryonic day (E)13 (Hasan et al., 1991, 1992). Subsequent experiments using a doubleretrograde tract-tracing protocol suggested that regeneration of previously axotomized fibersincreasingly contributes to the observed anatomical recovery after thoracic cord transectionson E1O-E12 (Hasan et al., 1991, 1992). In animals transected prior to E13, voluntary open-field locomotion (hatchling chicks) and brainstem evoked locomotion (embryonic orhatchling chicks) were indistinguishable from the locomotion observed in sham-transected oruntransected control chicks. This indicated that, after injury, complete functional recoveryhad also occurred (Hasan et al., 1991, 1992). When considered along with the anatomicalresults, the implication is that the functional recovery is due, at least in part, to true axonalregeneration. Conversely, anatomical and functional recovery is drastically diminishedfollowing thoracic cord transection on E13 and E14. After an E15 transection there is noanatomical repair or functional recovery. Thus, there appears to be a limited developmentalperiod for damaged brainstem-spinal axons to regenerate after severe injury. The limitedrecovery of embryos transected on El 3 or later strongly resembles the limited recovery in anadult bird or mammal (including human) following damage to the spinal cord. Therefore, Ihave documented that there is both a permissive period for spinal cord repair (andregeneration) followed by a restrictive period when spinal cord repair is drasticallydiminished or non-existent. What is different about these two periods of embryonic spinalcord repair? In other words, what are the mechanisms that facilitate or inhibit regeneration inthe damaged spinal cord?Many growth cues are controlled by biochemical, often protein based, factors.Thus, it is of great interest to understand how protein expression differs between early andlate embryonic development and whether there are specific proteins (and correspondingparent genes) that are critical for the differential repair of embryonic spinal cord after injury.If one can identify the important proteins, then one can employ molecular techniques to115identify mammalian (eg. human) homologues as well as potentially manipulate theexpression of these proteins after an adult spinal cord injury.Changes in protein expression have been examined using high resolution twodimensional gel electrophoresis (Ethell et aL, in preparation). Certain changes in proteinexpression have been shown to correlate with the transition from a permissive to a restrictiveperiod of embryonic chick spinal cord repair. Specifically, two sets of proteins have beenidentified; early neural proteins (ENPs) which are continuously present during spinal corddevelopment between E7-E12 and then decrease to relatively low levels, and late neuralproteins (LNPs) which are only expressed at high levels after E13. Some of these ENPs andLNPs may play direct or indirect roles in establishing permissive or non-permissiveenvironments for the outgrowth of axotomized fibers. Although this expressional associationis close, correlation does not imply causality. The function of these proteins cannot bedetermined from two dimensional gel electrophoresis alone. Even though it is possible thatnone of the identified proteins have any relationship to regenerative ability, it is also possiblethat these proteins correspond to specific proteins required for initial axonal growth (ENPs)or to proteins which specifically impair axonal outgrowth (LNPs). The ENPs are transientlyexpressed during the earlier ages of spinal cord development. The virtual absence of theseproteins at later ages of embryonic development may partially underlie the inability of lateembryonic and adult spinal cord to repair themselves after damage. It is also possible thatthe lack of spinal cord repair is due to interfering proteins (Caroni and Schwab, 1988a;Schwab and Caroni, 1988; Crutcher, 1989) being expressed in the cord after E13. In otherwords, these proteins may be involved in the inhibition of spinal cord repair. Further studiesto characterize these proteins and their relationship to spinal cord repair are currently beingcarried out.The lack of spinal cord repair in older avian embryos may be due to several factors,including: 1) the absence of essential growth promoting proteins necessary for axon growthand/or growth cone attachment (e.g. a growth or trophic factor, Thoenen and Edgar, 1985; or116extracellular matrix protein, Rathjen and Rutishauser, 1984; Thanos et al., 1984), or 2) thepresence of growth inhibiting factors (eg. glial scarring or proteins interfering with axonoutgrowth; Caroni and Schwab, 1988a; Schwab and Caroni, 1988). Obviously, one or moreof these influences may play a role in the ineffective regenerative capacity of olderembryonic or adult spinal cord tissue. A problem of this complexity, however, is bestapproached by dividing it into two specific topics: growth promotion and growth inhibition.1. Growth PromotionThere are three major environmental factors which regenerating growth conespotentially require for proper growth. First, growth cones may use the extracellular mathxfor anchorage. Second, growth cones may adhere to other cells (e.g. neural cells, glial cells,etc.). Third, growth cones may require a trophic (or maintenance) factor (or several trophicfactors).Laminin, fibronectin, and collagen are all extracellular matrix components whichprovide a scaffolding commonly used for cell migration and anchorage (Carbonetto, 1984;Bernfield, 1989). Membrane-bound receptors such as integrins have been found inabundance on growth cones (Carbonetto, 1984). Specific membrane proteins such asintegrins are thought to mediate interactions between cells and the extracellular matrix.Cell adhesion molecules (CAMs) mediate adhesion to other cells (Rutishauser et al.,1988; Kemler et a!., 1989). There are two major classes of cell adhesion molecules. Theclassification is based on their calcium binding requirements. Cadherin and NgCam(Neuron-glia Cell adhesion molecule) are highly calcium dependent proteins. NCAM(Neuronal CAM), however, is an example of a calcium independent protein (Rutishauser,1988).Nerve Growth Factor (NGF) and Brain Derived Neurotrophic Factor (BDNF) areexamples of trophic factors responsible for altering gene expression and sustaining themetabolism of neurons so that they can support the cellular machinery required to maintain117the survial of neurons and build axons (Levi-Montalcini and Angeletti, 1968; Greene andTischler, 1976; Thoenen and Barde, 1980; Barde et aL, 1982). A wide variety of othertrophic factors have been, and are currently being identified. Systemic trophic factors mayfacilitate early embryonic spinal cord repair since development of the blood brain barrier alsobegins around E13 of chick development (Wakai and Hirokawa, 1978; Risau and Wolburg,1990).2. Growth InhibitionFactors which inhibit axonal outgrowth may do so by creating a physical barrier(e.g. a glial scar) or by altering growth cone motility and elongation. The effects on thegrowth cone might be mediated via biochemical factors such as proteins which interfere withaxon outgrowth. Schwab and colleagues have demonstrated that two proteins associatedwith myelin in the rat are associated with inhibition of axonal outgrowth (Caroni andSchwab, 1988a; Schwab and Caroni, 1988).Cells dying as a result of injury are endocytosed by mobile macroglia and digested.Cavities resulting from necrosis are filled by reactive astrocytes which can then form glialscars. It has been suggested that this physical barrier is a major impediment to axonsregenerating within the cental nervous system (Reier and Houle, 1988). Glial scarring iscommonly seen following central nervous system injury in the adult (Reier and Houle,1988). It is possible that glial scarring may have acted as a physical barrier to prevent thesubsequent growth or regeneration of brainstem-spinal neurons following a spinal cordtransection on E13-E15. Like others, however, I was unable to detect any notable glialscarring upon histological examination of spinal cord sections following thoracic transectionsat different ages of embryonic development (Shimizu et al., 1990; Hasan et al., 1991, 1992).Central nervous system myelin-specific proteins interfering with axon outgrowth area relatively recent discovery (Caroni and Schwab, 1988a; Schwab and Caroni, 1988).Myelin is produced by CNS glial cells, specifically oligodendrocytes (Bensted et aL, 1957).118Myelin basic protein (MBP) is only associated with myelin and is one of the first proteinsexpressed at the onset of myelination of CNS axonal fibres (Hartman et al., 1979). Schnelland Schwab (1990) have recently shown axonal regeneration in the adult rat spinal cordproduced by an antibody (iN-i) against certain myelin associated proteins that inhibit thegrowth of neurons. CNS myelin and cultured oligodendrocytes are strong inhibitors ofneuron growth in culture (Schwab and Caroni, 1988). Caroni and Schwab (1988a) haveshown that this inhibitory property of myelin appears to be associated with two definedmyelin membrane proteins: (NI-35 and NI-250). They have introduced the monoclonalantibody IN-i to young rats by implanting antibody-producing tumours in the brain. Thisantibody when released in the brain neutralizes the inhibitory effect of both of the myelinmembrane proteins in the spinal cord (Caroni and Schwab, 1988b) and facilitates improvedaxonal regeneration of transected corticospinal fibers. Schnell and Schwab (1990) havecompletely severed the corticospinal tract at the thoracic level in 2-6 week old rats. AmongSchwab’s IN-i treated rats, a significant amount of axonal growth ocurred at the site of spinalcord damage, and fine axons could be observed up to 7-il mm below the injury site within2-3 weeks. Among control animals (not treated with IN-i), the maximal distance of axonalgrowth rarely exceeded 1 mm beyond the site of damage (Schnell and Schwab, 1990).Notably, previous studies had shown a complete absence of corticospinal tract regenerationafter the first postnatal week in rats (Bernstein et al., 1983), and in adult hamsters and cats(Kalil and Reh, 1982; Tolbert and Der, 1987).These results demonstrate the regenerative capacity of CNS axons after theneutralization of myelin-associated neuron growth inhibitors. Owing to the difficulty ofdefining a clear and unique function for the corticospinal tract in rats, one limitation of theSchnell and Schwab (1990) study was the inability to assess whether functional recovery ofcorticospinal projections was improved by IN-i treatment. The chick model is presently theonly higher vertebrate to exhibit both anatomical regeneration and functional recovery afterspinal cord injury.119In view of Schwab et al.?S work, different ways to inhibit myelin growth are nowbeing investigated in our laboratory in an effort to enhance regeneration of damaged spinalcord neurons in the embryonic chick (Keirstead et al., 1992). It is noteworthy that the onsetof myelination in the developing embryonic chick spinal cord starts around E13 (Bensted etal., 1957; Hartman et al., 1979; Macklin and Weill, 1985; Keirstead et aL, 1992). In otherwords, the onset of myelination is coincident with the transition from the permissive to therestrictive period for embryonic chick spinal cord repair. To assess a potential inhibitory rolefor myelin in the regeneration of brainstem-spinal projections after spinal transection, theonset of myelination is immunologically delayed (dysmyelination) until well into therestrictive period for embryonic spinal cord repair (Keirstead et al., 1992). The subsequentanatomical and functional assessment of spinal cord repair after a restrictive periodtransection (eg. E15) in such an unmyelinated environment then serves as a direct test ofwhether myelin inhibits anatomical and functional repair.A thoracic spinal cord transection as late as E15 (i.e. during the normal restrictiveperiod for repair) in a previously dysmyelinated embryo resulted in complete anatomical andfunctional recovery. This is in sharp contrast to embryos which were not previouslydysmyelinated, but were transected on E15. Upon hatching, these embryos were completelyparalyzed and unable to stand. In brief, the number and distribution of retrogradely labelledbrainstem-spinal neuronal cell bodies in dysmyelinated El 5-transected animals wasindistinguishable from untransected control embryos or embryos transected during thepermissive period (eg. E12, when myelin has yet to appear). After hatching,electromyographic recordings from the leg muscles of dysmyelinated E15-transected chicksduring overground walking, indicate that locomotor functions were also similar tountransected control animals or embryos transected during the permissive period. It isnotable that this dysmyelination procedure also suppresses the expression of several proteins(LNPs, see above) which normally appear at the developmental onset of myelination (Ethellet al., in preparation).120In conclusion, these findings are the first demonstration that the suppression ofmyelination extends the permissive period for both anatomical repair and functional CNSrecovery after an injury. It remains to be determined: 1) how the immunological proceduresuppresses the onset of myelination (i.e. is there lysis of the oligodendrocyte cell bodies orjust oligodendrocyte processes); 2) whether a modified neuroimmunological approach willimprove the repair and recovery of CNS function after injury to the post-hatching CNS; and3) what the relationship is between the appearance of the LNPs (see above) and thesuppression of regeneration, and whether the inhibition of LNP expression or the reexpression of ENPs would enhance regeneration during the resthctive period of spinal cordrepair. Nevertheless, the data clearly confirm and extend the proposition that the presence ofCNS myelin conthbutes to the inhibition of neuronal repair after a CNS injury (Schnell andSchwab, 1990).121REFERENCES122Anderson MJ, Waxman SG (1985) Neurogenesis in adult vertebrate spinal cord in situ and invitro: a new model system. Ann NY Acad Sci 457:213-233.Armstrong DM (1986) Supraspinal contributions to the initiation and control of locomotionin the cat. Prog Neurobiol 26:273-36 1.Attardi DG, and Sperry RW (1963) Preferential selection of central pathways byregenerating optic fibers. Exp Neurol 7:46-64.Barde YA, Edgar D, Thoenen H (1982) Purification of a new neurotrophic factor frommammalian brain. EMBO J 1:549-553.Barnes C (1984) Brainstem control of spinal cord function. Orlando, FL: Academic Press.Barrett CP, Donati EJ, Guth L (1984) Differences between adult and neonatal rats in theirastroglial response to spinal injury. Exp Neurol 84:374-385.Bekoff A (1976) Ontogeny of leg motor output in the chick embryo: a neural analysis. BrainRes 106:271-291.Bensted JPM, Dobbing J, Morgan RS, Reid RTW, Payling Wright G (1957) Neuroglialdevelopment and myelination in the spinal cord of the chick embryo. J Embryol Exp Morph5:428-437.Bernard JW, Carpenter W (1950) Lack of regeneration in spinal cord of rat. J Neurophysiol13:223-228.Bernfield M (1989) Extracellular matrix. Current Opinion in Cell Biology 1:953-955.Bernstein JJ, Bernstein ME (1971) Axonal regeneration and formation of synapses proximalto the site of lesion following hemisection of the rat spinal cord. Exp Neurol 30:336-35 1.Bernstein J, Gelderd J (1970) Regeneration of the long spinal tracts in the goldfish. BrainRes 20:33-38.Bernstein DR, Stelzner DJ (1983) Plasticity of the corticospinal tract following midthoracicspinal injury in the postnatal rat. J Comp Neurol 22 1:382-400.Bjorklund A, Katzman R, Stenevi U, West KA (1971) Development and growth of axonalsprouts from noradrenaline and 5-hydroxytryptamine neurones in the rat spinal cord. BrainRes 3 1:21-33.Bregman BS, Bernstein-Goral H (1991) Both regenerating and late-developing pathwayscontribute to transplant-induced anatomical plasticity after spinal cord lesions at birth. ExpNeurol 112:49-63.Bregman BS, Goldberger M (1982) Anatomical plasticity and sparing of function after spinalcord damage in neonatal cats. Science 2 17:553-4.Bullock TH, Orkand R, Grinnell A (1977) Introduction to nervous systems. San Francisco,CA: Freeman.123Cabot TB, Reiner A, Bogan N (1982) Avian bulbospinal pathways: anterograde andretrograde studies of cells of origin, funicular trajectories and laminar terminations. In:Descending pathways to the spinal cord (Kuypers HGJM, Martin GF, eds), pp 79-108.Amsterdam: Elsevier.Cajal RS (1928) Cajal’s degeneration and regeneration of the nervous system, englishtranslation (DeFelipe J, Jones EG, eds). New York, NY: Oxford UP.Carbonetto S (1984) The extracellular matrix of the nervous system. TINS 7:382-387.Caroni P. Schwab M (1988a) Two membrane protein fractions from rat central myelin withinhibitory properties for neurite growth and fibroblast spreading. J Cell Biol 106:1281-1288.Caroni P. Schwab M (1988b) Antibody against myelin-associated inhibitor of neurite growthneutralizes nonpermissive substrate properties of CNS white matter. Neuron 1:85-96.Clarke JDW, Alexander R, Holder N (1988) Regeneration of descending axons in the spinalcord of the axoloti. Neurosci Lett 89:1-6.Clearwaters KP (1954) Regeneration of the spinal cord of the chick. J Comp Neurol 101:3 17-329.Crutcher KA (1989) Tissue sections from the mature rat brain and spinal cord as substratesfor neurite outgrowth in vitro: extensive growth on gray matter but little growth on whitematter. Exp Neurol 104:39-54.David S, Aguayo Al (1981) Axonal elongation into peripheral nervous system “bridges”after central nervous system injury in adult rats. Science 214:931-933.Davis BM, Duffy MT, Simpson SB (1989) Bulbospinal and intraspinal connections innormal and regenerated salamander spinal cord. Exp Neurol 103:41-5 1.Davis BM, Ayers IL, Koran L, Carison J, Anderson MC, Simpson SB Jr (1990) Time courseof salamander spinal cord regeneration and recovery of swimming: HRP retrograde pathwaytracing and kinematic analysis. Exp Neurol 108:198-213.Dyce KM. Sack WO, Wensing CJG (1987) Textbook of veterinary anatomy. Philadelphia,PA: Saunders.Eidelberg E (1981) Consequences of spinal cord lesions upon motor function with specialreference to locomotor activity. Prog Neurobiol 17:185-202.Eidelberg E, Story JL, Walden 3G. Meyer BL (1981) Anatomical correlates of return oflocomotor function after partial spinal cord lesions in cats. Brain Res 42:81-88.Forehand C, Farel PB (1982) Anatomical and behavioural recovery from the effects of spinalcord transection: dependence on metamorphosis in anuran larvae. 3 Neurosci 2:654-662.Garcia-Rill E (1986) The basal ganglia and the locomotor regions. Brain Res Rev 11:47-63.Garcia-Rill E, Skinner RD (1987) The mesencephalic locomotor region. II. Projections toreticulospinal neurons. Brain Res 411:13-20.124Garcia-Rill E, Skinner RD. Conrad C, Mosley D, Campbell C (1986) Projections of themesencephalic locomotor region in the rat. Brain Res Bull 17:33-40.Gaze RM, and Jacobson M (1963) A study of the retinotectal projection during regenerationof the optic nerve in the frog. Proc R Soc Lond (Biol) 157:420-448.Glover JC, Pettersdottir G (1988) Pathway specificity of reticulospinal and vestibulospinalprojections in the 11-day chicken embryo. I Comp Neurol 270:25-38.Glover JC, Pettersdottir G (1991) Regional specificity of developing reticulospinal,vestibulospinal, and vestibulo-ocular projections in the chicken embryo. J Neurobiol 22:353-376.Glover JC, Petursdottir 0, Jansen JKS (1986) Fluorescent dextran-amines used as axonaltracers in the nervous system of the chicken embryo. J Neurosci Methods 18:243-254.Goldberger M (1973) Restitution of function and collateral sprouting in the cat spinal cord:the deafferented animal. Anat Rec 175-329.Goldberger M, Murray M (1974) Restitution of function and collateral sprouting in the catspinal cord: the deafferented animal. J Comp Neurol 158:37-54.Greene LA, Tischler AS (1976) Establishment of a noradrenergic clonal cell line of ratpheocromocytoma cells which respond to nerve growth factor. Proc Nat! Acad Sci USA73:2424-2428.Grillner S, Dubuc R (1988) Control of locomotion in vertebrates: spinal and supraspinalmechanisms. In: Advances in neurology, Vol 47: functional recovery in neurological disease(Waxman SG, ed), pp 425-453. New York: Raven Press.Grillner S. Wallen P (1985) Central pattern generators for locomotion, with special referenceto vertebrates. Annu Rev Neurosci 8:233-261.Gross GH, Oppenheim RW (1985) Novel sources of descending input to the spinal cord ofthe hatchling chick. J Comp Neurol 232:162-179.Hamburger V, Hamilton HL (1951) A series of normal stages in the development of thechick embryo. J Morph 88:49-92.Hartman BK, Agrawal HC, Kaimbach S, Shearer WT (1979) A comparative study of theimmunohistochemical localization of basic protein to myelin and o!igodendrocytes in rat andchicken brain. J Comp Neurol 188:273-290.Hasan SJ, Nelson BH, Valenzuela II, Keirstead HS, Shull SE, Ethel! DW, Steeves ID (1991)Functional repair of transected spinal cord in embryonic chick. Restor Neurol Neurosci2:137-154.Hasan SJ, Keirstead HS, Muir GD, Steeves JD (1992) Axonal regeneration contributes torepair of injured brainstem-spinal neurons in embryonic chick. J Neurosci (In Press).Heimer L, RoBards MJ (1981) Neuroanatomical tract-tracing methods. New York, NY:Plenum Press.125Holder N, Clarke JDW (1988) Is there a correlation between continuous neurogenesis anddirected axon regeneration in the vertebrate nervous system? TINS 11:94-99.Jacobson R, Hollyday M (1982a) A behavioral and electromyographic study of walking inthe chick. J Neurophysiol 48:238-255.Jacobson R, Hollyday M (1982b) Electrically evoked walking and fictive locomotion in thechick. J Neurophysiol 48:257-270.Jordan LM (1986) Initiation of locomotion from the mammalian brainstem. In: Neurobiologyof vertebrate locomotion (Griliner S, Stein PSG, Stuart DG, Forssberg H, Herman RM, eds),pp 2 1-37. London: MacMillan Press.Kalil K, Reh T (1982) A light and electron microscopic study of regrowing pyramidal tractfibers. J Comp Neurol 211:265-275.Karten HJ, Hodos W (1967) A stereotaxic atlas of the brain of the pigeon (Columbia livia).Baltimore, MD: Johns Hopkins UP.Keirstead HS, Hasan SJ, Muir GD, Steeves JD (1992) Suppression of the onset ofmyelination extends the permissive period for the functional repair of embryonic spinal cord.Proc Natl Acad Sci USA (Accepted).Kemler R, Ozawa M, Ringwald M (1989) Calcium-dependent cell adhesion molecules.Current Opinion in Cell Biology 1:892-897.Kuenzel WJ, Masson M (1988) A Stereotaxic atlas of the brain of the chick. Baltimore, MD:Johns Hopkins UP.Lawrence DG, Kuypers HGJM (1968a) The functional organization of the motor system inthe monkey. I. The effects of bilateral pyramidal lesions. Brain 91:1-14.Lawrence DG, Kuypers HGJM (1968b) The functional organization of the motor system inthe monkey. 11. The effects of lesions of the descending brain-stem pathways. Brain 91:15-36.Levi-Montalcini R, Angletti PU (1968) Nerve Growth Factor. Physiol Rev 8:534-569.Lurie DI, Seizer ME (1991) Axonal regeneration in the adult lamprey spinal cord. J CompNeurol 306:409-416.Lyon MJ, Stelzner DJ (1987) Tests of the regenerative capacity of tectal efferent axons in thefrog, Ranapipiens. J Comp Neurol 255:511-525.Macklin WB, Weill CL (1985) Appearance of myelin proteins during development in thechick central nervous system. Dev Neurosci 7:170-178.Martin GF, Humbertson AO, Laxon LC, Panneton WM, Tschismadia 1(1979) Spinalprojections from the mesencephalic and pontine reticular formation of the North Americanopossum: a study using axonal transport techniques. J Comp Neurol 187:373-400.McClellan AD (1986) Command systems for intiating locomotion in fish and amphibians:parallels to initiation in mammals. In: Neurobiology of vertebrate locomotion (Grillner 5,Stein PSG, Stuart DG, Forssberg H, Herman RM, eds), pp 3-20. London: MacMillan Press.126McClellan AD (1988) Functional regeneration of descending brainstem command pathwaysfor locomotion demonstrated in the in vitro lamprey CNS. Brain Res 448:339-345.McClellan AD (1990) Locomotor recovery in spinal-transected lamprey: role of functionalregeneration of descending axons from brainstem locomotor command neurons. Neurosci37:781-798.McConnell JA, Sechrist J (1980) Identification of early neurons in the brainstem and spinalcord: an autoradiographic study in the chick. J Comp Neurol 192:769-783.Noga BR, Kettler J, Jordan LM (1988) Locomotion produced in mesencephalic cats byinjection of putative transmitter substances and antagonists into the medial reticularformation and the pontomedullary locomotor strip. J Neurosci 8:2074-2086.O’Donovan MJ, Landmesser L (1987) The development of hindlimb motor activity studied inthe isolated spinal cord of the chick embryo. J Neurosci 7:3256-3264.Okado N, Oppenheim RW (1984) Cell death of motoneurons in the chick embryo spinalcord. IX. The loss of motoneurons following removal of afferent inputs. J Neurosci 4:1639-1652.Okado N, Oppenheim RW (1985) The onset and development of descending pathways to thespinal cord of the chick embryo. J Comp Neurol 232: 143-161.Puchala E, Windle W (1977) The possibility of structural and functional restitution afterspinal cord injury. A review. Exp Neurol 55:1-42.Ranvier LA (1871) De la degenerescence des nerfs apres leur section. Compt rend.Ranvier LA (1873) De la regeneration des nerfs sectionnes. Compt rend., t.76.Rathjen FG, Rutishauser U (1984) Comparison of two cell surface molecules involved inneural cell adhesion. EMBO J 3:461-465.Reier PJ, Houle JD (1988) The glial scar: its bearing on axonal elongation andtransplantation approaches to CNS repair. Functional recovery in neurological disease. In:Advances in neurology, Vol 47: functional recovery in neurological disease (Waxman SG,ed), pp 87-138. New York: Raven Press.Reynolds AB, Weiss S (1992) Generation of neurons and astrocytes from isolated cells of theadult mammalian nervous system. Science 255: 1707-1710.Risau W, Wolburg H (1990) Development of the blood-brain barrier. TINS 13(5): 174-178.Rutishauser U, Acheson A, Hall AK, Mann DM, Sunshine J (1988) The neural cell adhesionmolecule (NCAM) as a regulator of cell-cell interactions. Science 240:53-57.Sarnat H, Netsky M (1981) Evolution of the nervous system. New York, NY: Oxford UP.Schnell L, Schwab M (1990) Axonal regeneration in the rat spinal cord produced by anantibody against myelin associated with neurite-growth inhibitors. Nature 343:269-272.127Schwab M, Carom P (1988) Oligodendrocytes and CNS myelin are nonpermissive substratesfor neurite growth and fibroblast spreading in vitro. J Neurosci 8:2381-2393.Schwab M, Thoenen H (1985) Dissociated neurons regenerate into sciatic but not optic nerveexplants in culture irrespective of neurotrophic factors. J Neurosci 5:2415-2423.Sechrist 3, Bronner-Fraser M (1991) Birth and differentiation of reticular neurons in thechick hindbrain: ontogeny of the first neuronal population. Neuron 7:947-963.Shefchyk SJ, Jell RM, Jordan LM (1984) Reversible cooling of the brainstem reveals areasrequired for mesencephalic locomotor region evoked treadmill locomotion. Exp Brain Res56:257-262.Shiga T, Kunzi R, Oppenheim RW (1991) Axonal projections and synaptogenesis bysupraspinal descending neurons in the spinal cord of the chick embryo. J Comp Neurol305:83-95.Shimizu I, Oppenheim RW, O’Brien M, Shneiderman A (1990) Anatomical and functionalrecovery following spinal cord transection in chick embryo. J Neurobiol 21:918-937.Sholomenko GN, Funk GD, Steeves JD (1991a) Avian locomotion activated by brainsteminfusion of neurotransmitter agonists and antagonists. I. Acetyicholine, excitatory aminoacids and substance P. Exp Brain Res 85:659-673.Sholomenko GN, Funk GD, Steeves JD (1991b) Avian locomotion activated by brainsteminfusion of neurotransmitter agonists and antagonists. II. Gamma-aminobutyric acid. ExpBrain Res 85:674-68 1.Sholomenko GN, Funk GD, Steeves JD (1991c) Locomotor activities in the decerebrate birdwithout phasic afferent input. Neuroscience 40:257-266.Sholomenko GN, Steeves 3D (1987) Effects of selective spinal cord lesions on hind limblocomotion in birds. Exp Neurol 95:403-418.Sperry RW (1963) Chemoaffinity in the orderly growth of nerve fiber patterns andconnections. Proc Nail Acad Sci USA 50:703-709.Steeves JD, Jordan LM (1980) Localization of a descending pathway in the spinal cordwhich is necessary for controlled treadmill locomotion. Neurosci Lett 20:283-288.Steeves JD, Jordan LM (1984) Autoradiographic demonstration of the projections from themesencephalic locomotor region. Brain Res 307:263-276.Steeves JD, Sholomenko GN, Webster DMS (1987) Stimulation of the pontomedullaryreticular formation initiates locomotion in decerebrate birds. Brain Res 401:205-212.Stoney SD, Thompson WD, Asanuma H (1968) Excitation of pyramidal tract cells byintracortical microstimulation: effective extent of stimulating current. 3 Neurophysiol31:659-669.ten Cate J (1960) Locomotor movements in the spinal pigeon. 3 Exp Biol 37:609-613.ten Cate J (1962) Innervation of locomotor movements by the lumbosacral cord in birds andmammals. 3 Exp Biol 39:239-242.128ten Donkelaar HJ (1982) Organization of descending pathways to the spinal cord inamphibians and reptiles. In: Progress in brain research, Vol. 57: descending pathways to thespinal cord (Kuypers HGJM, Martin GF, eds), pp 25-67. Amsterdam: Elsevier.Thanos 5, Bonhoeffer F, Rutishauser U (1984) Fiber-fiber interaction and tectal cuesinfluence the development of the chicken retinotectal projection. Proc Natl Acad Sci USA81:1906-1910.Thoenen H, Barde YA (1980) Physiology of nerve growth factor. Physiol Rev 60:1284-1335.Thoenen H, Edgar D (1985) Neurotrohic factors. Science 229:238-242.Tolbert DL, Der T (1987) Redirected growth of pyramidal tract axons following neonatalpyramidotomy in cats. J Comp Neurol 260:299-311.Valenzuela JI, Hasan SJ, Steeves JD (1990) Stimulation of the brainstem reticular formationevokes locomotor activity in embryonic chicken (in ovo). Dev Brain Res 56:13-18.Wakai S, Hirokawa N (1978) Development of the blood-brain barrier to horseradishperoxidase in the chick embryo. Cell Tissue Res 195:195-203.Wall PD (1975) Editorial. Pain 1:1-2.Wall PD, Sternbach RA (1976) Editorial. Pain 2:1-4.Wailer AV (1852) Sur la reproduction des nerfs at sur la structure et les fonctions desganglions spinaux. Arch Mat Physiol, pp 392-40 1.Webster DMS, Rogers U, Pettigrew JD, Steeves JD (1990) Origins of descending spinalpathways in prehensile birds: do parrots have a homologue to the mammalian corticospinaltract? Brain Behav Evol 36:216-226.Webster DMS, Steeves JD (1988) Origins of brainstem-spinal projections in the duck andgoose. I Comp Neurol 273:573-583.Webster DMS, Steeves JD (1991) Funicular organization of avian brainstem-spinalprojections. J Comp Neurol 3 12:467-476.Weinstein GN, Anderson C, Steeves JD (1984) Functional characterization of limb musclesinvolved in locomotion in the Canada goose, Branta canadensis. Can 3 Zool 62:1596-1604.Wise RA (1972) Spread of current from monopolar stimulation of the lateral hypothalamus.Amer 3 Physiol 210:1181-1186.Youngren OM, Phillips RE (1978) A stereotaxic atlas of the brain of the three-day olddomestic chick. 3 Comp Neurol 18 1:567-600.Zemlan FP, Kow LM, Pfaff DW (1983) Effect of interuption of bulbospinal pathways onlordosis, posture and locomotion. Exp Neurol 81:177-194.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0086545/manifest

Comment

Related Items