- Email: [email protected]
Developmental plasticity of selected spinocerebellar axons. Studies using the North American opossum, Didelphis virginiana
Developmental Brain Research 102 Ž1997. 309–314
Short communication
Developmental plasticity of selected spinocerebellar axons. Studies using the North American opossum, Didelphis Õirginiana Jonathan R. Terman, Xian M. Wang, George F. Martin
)
Department of Cell Biology, Neurobiology and Anatomy and Neuroscience Program, The Ohio State UniÕersity, College of Medicine, 333 West Tenth AÕe., Columbus, OH 43210, USA Accepted 17 June 1997
Abstract When the thoracic spinal cord of the opossum is hemisected at postnatal day 5 or 8, but not at day 12 or later ages, spinocerebellar axons which originate from spinal border cells, the sacralrcoccygeal ventrolateral nucleus, and Stilling’s nucleus grow through the lesion and reach the cerebellum. The critical period for such growth is comparable to that reported previously for spinocerebellar axons originating within Clarke’s nucleus and for axons of the fasciculus gracilis, but shorter than that for most descending spinal axons. It appears, therefore, that differences exist in the ability of ascending and descending axons to traverse a lesion of their spinal pathway during development. q 1997 Elsevier Science B.V. Keywords: Clarke’s nucleus; Marsupial; Regeneration; Sacralrcoccygeal ventrolateral nucleus; Spinal border cell; Stilling’s nucleus
Axons fail to grow beyond a lesion of the spinal cord in adult mammals, but at least some do when the lesion is made early in development Žsee reviews w30,33x.. In the North American opossum, Didelphis Õirginiana, supraspinal axons grow around w28,43,49x or through w29,44,45x a lesion of their spinal pathway when it is made prior to postnatal day ŽPD. 30, but not when it is made at older ages. Interestingly, the critical period for such growth is not the same for all axons. For example, the critical period for lateral vestibulospinal and certain pontine reticulospinal axons ends before PD12, yet that for rubrospinal and raphespinal axons continues until PD30 w43,44,49x. Although growth through or around a lesion of the spinal cord has been documented for axons of most, if not all, descending spinal pathways in developing mammals, it has only been described for a limited number of ascending axons. Using the North American opossum, we have shown that axons which originate in Clarke’s nucleus ŽCN., i.e., axons of the dorsal spinocerebellar tract ŽDSCT., grow through a lesion of their spinal pathway when it is made at PD5-8, but not when it is made at PD12 or at later ages w40x. Spinocerebellar axons differ, however, in origin
)
Corresponding author. Fax: q1 Ž614. 292-7659; E-mail: [email protected] 0165-3806r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 1 1 2 - 0
w31,32x, developmental history w1,27,50x, laterality w31,32x, funicular position w23,48x, peduncular course w21,51x, and terminal distribution w5,47x. It is possible, therefore, that differences also exist in their ability to grow through or around a lesion of their spinal pathway. In the present study, we have asked if spinocerebellar axons originating from spinal border cells ŽSBC. w13,37x, neurons within the sacralrcoccygeal ventrolateral nucleus ŽVLN. w31,37x, or neurons within Stilling’s sacral nucleus ŽSN. w9,35,39x are capable of such growth and, if so, whether the critical periods for it are the same as that for axons of the DSCT. The animals used in this study had been employed previously and a detailed description of the methods can be found elsewhere w40x. Opossums with appropriately aged young ŽPD5 w n s 7x, PD8 w n s 4x, PD12 w n s 8x, PD20 w n s 2x, and PD26 w n s 2x. were anesthetized with ketamine and Metafane and placed in a supine position to expose their pouch. During anesthesia, the pouch sphincter relaxes making it possible to anesthetize the pups by hypothermia. The spinal cord of the anesthetized pups was exposed surgically and hemisected at mid-thoracic levels using a microblade. The lesion was made by inserting the microblade slightly to the left of the midline and pulling it to the right until the right half of the cord was severed. To ensure that the hemisection was complete, a microdissection needle with a 908 angle was drawn through the lesion
310
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
cavity. In addition, 3 of the PD5 pups were sacrificed shortly after lesioning and examined for the presence of a complete hemisection. In these cases the spinal cord was removed after perfusion and the lesioned segment was embedded in paraffin, sectioned at 20 m m in the frontal plane, and stained for Nissl substance. In each of these animals the hemisection was obvious grossly and histologically. The remaining animals were maintained 4–9 months before being anesthetized using sodium pentobarbital for bilateral injections of 5% Fluoro-Gold ŽFG. or 2% Fast Blue ŽFB. into the anterior lobe of the cerebellum, the primary target of spinocerebellar axons w23x. No differences have been observed in our laboratory between the ability of FB or FG to retrogradely label neurons Žunpublished results.. FG is preferable to FB since it holds up better to ultraviolet light. Unfortunately, we have found FG to be more toxic to younger animals, so we have employed FB as a retrograde tracer in such cases. After a 7–10 day survival, each animal was reanesthetized and sacrificed by perfusion so its brain and spinal cord could be removed and sectioned using a freezing microtome. Every section of the spinal cord caudal to the lesion site was examined for labeling of SBC Žlower thoracic and upper lumbar levels., neurons in the VLN Žsacral and coccygeal levels., and neurons in SN Žsacral and coccygeal levels.. Since the cerebellar projections of these nuclei are crossed in the opossum Žunpublished results., as in other species w31,32x, labeled neurons contralateral to the lesion were considered to have supported axons which grew through or around the lesion to the cerebellum. Labeled neurons ipsilateral to the lesion were used as internal controls. In addition, the cerebellum and the spinal cord rostral to the lesion were sectioned and examined for symmetry of the injections and labeling. In four of the cases lesioned at PD5, labeled neuronal profiles in each of the above mentioned nuclei were counted in every sixth section on the experimental and control sides. Results were expressed as the mean " the standard error of the mean and analyzed using the paired t-test for significant differences in the number of labeled neurons on the two sides. When the thoracic cord was hemisected on PD5 or PD8 and bilateral injections of FB or FG were made into the cerebellum 4–9 months later, labeling was present bilaterally within SBC, within neurons of the VLN, and within neurons of SN. The results from one case are shown in Fig. 1. The site of the lesion was not evident grossly ŽFig. 1A, arrow. and after sectioning, recognizable spinal cord was present on the lesioned side ŽFig. 1C; right side. although it was abnormal in appearance. Observed abnormalities included fusion of the dorsal horns, decrease in the size of the dorsal columns, an uncharacteristic appearance of the grey matter, and a relative thinness of the white matter on the lesioned side. Examination of the injection sites and the presence of bilaterally symmetrical labeling of CN rostral to the lesion Žopen arrows, Fig. 1B. suggested that the injections were similar bilaterally. Caudal
to the lesion, bilateral labeling was found in SBC ŽFig. 1D, closed arrows., neurons of the VLN ŽFig. 1E, arrows., and neurons of SN ŽFig. 1F, arrows.. Counts from every sixth section revealed that labeled neuronal profiles were fewest in number contralateral to the lesion and the results were statistically significant Ž P - 0.01; data not shown.. It will be recalled that neurons in each of the above nuclei project contralaterally to the cerebellum. Note that bilateral labeling was also present in CN caudal to the lesion ŽFig. 1D, open arrows. and that labeled neurons were fewest ipsilateral to the lesion Žright side. w40x. It will be recalled that the DSCT projects ipsilaterally to the cerebellum. Following hemisection of the mid-thoracic spinal cord on PD12 or at later stages of development and bilateral injections of FB or FG into the cerebellum 4–6 months later, labeling was not found in SBC, within neurons of the VLN, or within neurons of SN contralateral to the lesion, although it was present in neurons within the same areas on the ipsilateral side. The results from a PD12 case are documented in Fig. 2. The lesion site was identifiable grossly ŽFig. 2A, arrow. and histologically it was seen to include over half the cord ŽFig. 2C.. Examination of the injection sites and labeling in CN rostral to the lesion Žopen arrows, Fig. 2B. suggested that the injections were symmetrical bilaterally. Caudal to the lesion, labeling in SBC ŽFig. 2D, closed arrows., the VLN ŽFig. 2E, arrows., and SN ŽFig. 2F, arrow. was only present ipsilateral to the lesion. As expected w40x, labeling in CN caudal to the lesion ŽFig. 2D; open arrow. was limited to the side contralateral to the lesion. In all of the PD5 and 8 cases, labeled spinocerebellar axons were present in their normal position ipsilateral to the lesion in what appeared to be regenerated spinal cord at the lesion site Žarrows, Fig. 1C., as well as in normal spinal cord rostral Žarrows, Fig. 1B., and caudal Žnot seen at this magnification. to it. Although they were not counted, labeled axons ipsilateral to the lesion appeared to be fewer in number than those on the contralateral side ŽFig. 1B and C.. A few labeled axons also were present rostral and ipsilateral to the lesion in the PD12 cases. Such axons likely originated rostral to the lesion, however, since axonal labeling was not found at the lesion site or caudal to it on the ipsilateral side. Our results indicate that, in the opossum, axons originating from SBC, the VLN, and SN, like those which originate in CN, grow through a lesion of their spinal pathway and reach the cerebellum if the lesion is made early enough in development. It is unlikely that our lesions were incomplete in the PD5–8 cases even though recognizable spinal cord was present at the lesion site. Evidence for a complete hemisection included the abnormal appearance of the spinal cord at the lesion site and the obvious presence of a hemisection in the PD5 pups sacrificed shortly after lesioning. Furthermore, when the lesion was made on PD12 using the same technique, it was clear that it included over half the cord.
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
As described previously w29,40,44x, reconstruction of recognizable spinal cord occurs at the lesion site after hemisection or transection in the PD5 opossum. Compara-
311
ble results have been reported after crush lesions of the spinal cord in the South American opossum, Monodelphis domestica Žsee review w33x.. Regeneration of damaged
Fig. 1. A photograph of a portion of spinal cord including the lesion site Žarrow. from an animal subjected to mid-thoracic hemisection on postnatal ŽPD. 5 and bilateral injections of Fluoro-Gold into the cerebellum 6 months later is shown in ŽA.. A section of thoracic cord ŽT4. rostral to the lesion from the same case is illustrated in ŽB.. Note that labeling of Clarke’s nucleus ŽCN. Žopen arrows. was bilaterally symmetrical suggesting that the injections were comparable on the two sides. Labeled axons were found within the lateral and ventral funiculi bilaterally, although they appeared to be fewest on the side of the lesion Žclosed arrows.. The asterisk delineates the side opposite the lesion in this section and in those which follow. A section through the lesion site is shown in ŽC. and labeled axons were seen on the side of the lesion Žarrows, right side. in what appeared to be reconstructed spinal cord. A photomicrograph from the first lumbar section of the cord is shown in ŽD.. Notice that spinal border cells ŽSBC. were labeled bilaterally Žclosed arrows.. Labeled neurons were also present within CN on both sides Žopen arrows.. A section from the second sacral segment of the cord is seen in ŽE. and it can be seen that labeled cells were present bilaterally in the sacralrcoccygeal ventrolateral nucleus ŽVLN. Žarrows.. A section of coccygeal cord is seen in ŽF. and labeled cells were present bilaterally in Stilling’s nucleus ŽSN. Žarrows.. The scale bar in ŽB. applies to ŽC. – ŽF..
312
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
spinal cord occurs in non-mammalian vertebrates and it is apparently initiated by proliferation of ependymal cells which bridge the gap at the lesion site Žsee review w30x.. The mechanisms which underlie spinal cord reorganization in the developing opossum are not known, nor is it known
why it no longer occurs at PD12 or at later ages. Likewise, it is not known whether growth through the lesion resulted from regeneration of cut axons, growth of late arriving axons that were not damaged by the lesion, or both. Both mechanisms contribute to extension of descending axons
Fig. 2. A portion of the spinal cord including the lesion site Žarrow. from an animal subjected to hemisection of the mid-thoracic cord on PD12 and bilateral injections of Fast Blue into the cerebellum 4 months later is shown in ŽA.. A photomicrograph of a section of thoracic cord rostral to the lesion ŽT4. from the same case is shown in ŽB.. Note that labeled cells were present in CN in relatively equal numbers bilaterally Žopen arrows. suggesting that our injections were symmetrical. The asterisk delineates the side opposite the lesion in this section and in those which follow. A section through the lesion is shown in ŽC. where it can be seen that over half of the cord was involved. The spared lateral funiculus ŽLF. and ventral funiculus ŽVF. are indicated. A section through the first segment of the lumbar cord is shown in ŽD.. Labeled SBC Žclosed arrows. were present only ipsilateral to the lesion. In addition, labeling in CN Žopen arrow. was only present contralateral to the lesion. A section from the second segment of the sacral cord is shown in ŽE.. Labeled cells were only seen in the VLN Žarrows. ipsilateral to the lesion. A section of coccygeal cord is shown in ŽF.. Labeled neurons in SN Žarrow. were present only on the side of the lesion. The scale bar in ŽB. applies to ŽC. – ŽF..
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
through the lesion site when the spinal cord is injured during early development in the chick w22x and opossum Žunpublished results.. Major differences were not found in the critical periods for developmental plasticity of the spinocerebellar axons studied. In contrast, clear differences exist for descending axons w43,44,49x. For example, lateral vestibulospinal and certain pontine reticulospinal axons do not grow through a lesion when it is made on PD12 or later, but rubrospinal and raphespinal axons do. In fact, some rubral and raphe axons grow through such a lesion when it is made as late as PD30. The critical period for developmental plasticity of spinocerebellar axons in the opossum is shorter than that for many descending spinal axons. Spinocerebellar axons no longer traverse the lesion site after hemisection of the thoracic cord on PD12 Žw40x; present results.; but, as noted above, some supraspinal axons do when the lesion is made as late as PD30 w44x. The critical period for developmental plasticity of axons within the fasciculus gracilis, another ascending spinal pathway, is also relatively short in the opossum w46x. It is possible, therefore, that differences exist in the ability of ascending and descending axons to grow beyond a lesion of their spinal pathway. Such differences also have been reported for reptiles w15,16x, frogs w3,7,11,18x, salamanders w14,38x, fish w4,6x, and lampreys w52x, but the mechanisms which underlie them have not been elucidated. Loss of developmental plasticity may be due to a diminished ability to initiate andror sustain axonal growth with age w10,17x. That hypothesis is supported in the opossum by the finding that supraspinal axons which reach the spinal cord first lose the ability to grow through or around a lesion before those which arrive at a later date w43,44,49x. It is interesting, however, that the critical periods for developmental plasticity of different spinocerebellar axons appear comparable even though they apparently have different developmental histories w1,27,50x. It is possible, therefore, that the development of a non-permissive environment for axonal growth plays a role in the loss of developmental plasticity. In the North American opossum, a temporal correlation exists between the appearance of myelin, which is known to contain proteins inhibitory to neurite extension Že.g., w8x., and the end of the critical period for plasticity of descending spinal axons w19,43,44,49x. In fact, evidence for a cause and effect relationship has been provided for the chick Že.g., w25x.. However, the critical period for developmental plasticity of spinocerebellar axons Žw40x; present study. ends well before the appearance of myelin w19x. If the development of a non-permissive environment is a factor in loss of spinocerebellar plasticity, myelin does not apparently play a role Žsee also w26x.. It appears, therefore, that the factors which result in loss of developmental plasticity are not the same for all axons. Indeed, it is becoming increasingly clear that differences exist in the ability of axons to grow
313
in the same regions of the brain and spinal cord, and on similar substrates in vitro w2,12,20,24,34,36,41,42x.
Acknowledgements The authors wish to thank Ms. Mary Ann Jarrell for surgical assistance and tissue processing, and Drs. James S. King and Phillip G. Popovich for helpful criticism of the manuscript. Our studies were supported by USPHS Grants NS-25095 and NS-10165.
References w1x M.L. Arsenio Nunes, C. Sotelo, Development of the spinocerebellar system in the postnatal rat, J. Comp. Neurol. 237 Ž1985. 291–306. w2x D. Bagnard, F. Mann, S. Henke-Fahle, J. Bolz, Developmental mechanisms underlying the segregation of afferent and efferent cortical projections, Soc. Neurosci. Abstr. 21 Ž1995. 1285. w3x M.J. Beattie, J.C. Bresnahan, G. Lopate, Metamorphosis alters the response to spinal cord transection in Xenopus laeÕis frogs, J. Neurobiol. 21 Ž1990. 1108–1122. w4x T. Becker, M.F. Wullimann, C.G. Becker, R.R. Bernhardt, M. Schachner, Axonal regrowth after spinal cord transection in adult zebrafish, J. Comp. Neurol. 377 Ž1997. 577–595. w5x S. Berretta, V. Perciavalle, R.E. Poppele, Origin of spinal projections to the anterior and posterior lobes of the rat cerebellum, J. Comp. Neurol. 305 Ž1991. 273–281. w6x S.M. Bunt, P. Fill-Moebs, Selection of pathways by regenerating spinal cord fiber tracts, Dev. Brain Res. 16 Ž1984. 307–311. w7x H.L. Campbell, M.S. Beattie, J.C. Bresnahan, The response of dorsal root afferent fibers to dorsal funiculus lesions in developing Rana catesbeiana tadpoles, Soc. Neurosci. Abstr. 10 Ž1984. 1024. w8x P. Caroni, M.E. Schwab, Two membrane protein fractions from rat central myelin with inhibitory properties for neurite growth and fibroblast spreading, J. Cell Biol. 106 Ž1988. 1281–1288. w9x H.T. Chang, Caudal extension of Clarke’s nucleus in the spider monkey, J. Comp. Neurol. 95 Ž1951. 43–51. w10x D.F. Chen, S. Jhaveri, G.E. Schneider, Intrinsic changes in developing retinal neurons result in regenerative failure of their axons, Proc. Natl. Acad. Sci. USA 92 Ž1995. 7287–7291. w11x J.D.W. Clarke, D.A. Tonge, N.H.K. Holder, Stage-dependent restoration of sensory dorsal columns following spinal cord transection in anuran tadpoles, Proc. R. Soc. London B 227 Ž1986. 67–82. w12x S.A. Colamarino, M. Tessier-Lavigne, The axonal chemoattractant netrin-1 is also a chemorepellent for trochlear motor axons, Cell 81 Ž1995. 621–629. w13x S. Cooper, C.S. Sherrington, Gower’s tract and spinal border cells, Brain 63 Ž1940. 123–134. w14x B.M. Davis, J.L. Ayers, L. Koran, J. Carlson, M.C. Anderson, S.B. Simpson Jr., Time course of salamander spinal cord regeneration and recovery of swimming: HRP retrograde pathway tracing and kinematic analysis, Exp. Neurol. 108 Ž1990. 198–213. w15x M.T. Duffy, A. Hawrych, D.R. Liebich, S.B. Simpson Jr., Regeneration of cerebrospinal fluid contacting neurons ŽCSFCN. in the regenerated tail spinal cord of the lizard Anolis carolinensis, Soc. Neurosci. Abstr. 19 Ž1993. 1716. w16x M.T. Duffy, S.B. Simpson Jr., D.R. Liebich, B.M. Davis, Origin of spinal cord axons in the lizard regenerated tail: supernormal projections from local spinal neurons, J. Comp. Neurol. 293 Ž1990. 208–222. w17x J.W. Fawcett, Intrinsic neuronal determinants of regeneration, Trends Neurosci. 15 Ž1992. 5–8.
314
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
w18x C. Forehand, P. Farel, Anatomical and behavioral recovery from the effects of spinal cord transection: dependence on metamorphosis in anuran larvae, J. Neurosci. 2 Ž1982. 654–662. w19x G.T. Ghooray, G.F. Martin, The development of myelin in the spinal cord of the North American opossum and its possible role in loss of rubrospinal plasticity. A study using myelin basic protein and galactocerebroside immunohistochemistry, Dev. Brain Res. 72 Ž1993. 67–74. w20x R.W. Guillery, C.A. Mason, J.S.H. Taylor, Developmental determinants at the mammalian optic chiasm, J. Neurosci. 15 Ž1995. 4727– 4737. w21x G. Grant, Q. Xu, Routes of entry into the cerebellum of spinocerebellar axons from the lower part of the spinal cord. An experimental study in the cat, Exp. Brain Res. 72 Ž1988. 543–561. w22x S.J. Hasan, H.S. Keirstead, G.D. Muir, J.D. Steeves, Axonal regeneration contributes to repair of injured brainstem-spinal neurons in embryonic chick, J. Neurosci. 13 Ž1993. 492–507. w23x J.C. Hazlett, G.F. Martin, R. Dom, Spino-cerebellar fibers of the opossum, Didelphis Õirginiana, Brain Res. 33 Ž1971. 257–271. w24x S. Jhaveri, Midline glia of the tectum: a barrier for developing retinal axons, Perspect. Dev. Neurobiol. 1 Ž1993. 237–243. w25x H.S. Keirstead, S.J. Hasan, G.D. Muir, J.D. Steeves, Suppression of the onset of myelination extends the permissive period for the functional repair of embryonic spinal cord, Proc. Natl. Acad. Sci. USA 89 Ž1992. 11664–11668. w26x R.E. MacLaren, Expression of myelin proteins in the opossum optic nerve: late appearance of inhibitors implicates an earlier non-myelin factor in preventing ganglion cell regeneration, J. Comp. Neurol. 372 Ž1996. 27–36. w27x G.F. Martin, J.L. Culbertson, J.C. Hazlett, Observations on the early development of ascending spinal pathways. Studies using the North American opossum, Anat. Embryol. 166 Ž1983. 191–207. w28x G.F. Martin, X.M. Xu, Evidence for developmental plasticity of the rubrospinal tract. Studies using the North American opossum, Dev. Brain. Res. 39 Ž1988. 303–308. w29x G.F. Martin, J.R. Terman, X.M. Wang, Evidence for growth of supraspinal axons through hemisected and transected spinal cord in the developing opossum, Didelphis Õirginiana, Soc. Neurosci. Abstr. 20 Ž1994. 1270. w30x G.F. Martin, G.T. Ghooray, X.M. Wang, X.M. Xu, X.C. Zou, Models of spinal cord regeneration, in: F.J. Seil ŽEd.., Neural Regeneration, Progress in Brain Research, Vol. 103, Elsevier, Amsterdam, 1994, pp. 175–201. w31x M. Matsushita, Y. Hosoya, Cells of origin of the spinocerebellar tract in the rat, studied with the method of retrograde transport of horseradish peroxidase, Brain Res. 173 Ž1979. 185–200. w32x M. Matsushita, Y. Hosoya, M. Ikeda, Anatomical organization of the spinocerebellar system in the cat, as studied by retrograde transport of horseradish peroxidase, J. Comp. Neurol. 184 Ž1979. 81–106. w33x J. Nicholls, N. Saunders, Regeneration of immature mammalian spinal cord after injury, Trends Neurosci. 19 Ž1996. 229–234. w34x A. Nose, M. Takeichi, C.S. Goodman, Ectopic expression of connectin reveals a repulsive function during growth cone guidance and synapse formation, Neuron 13 Ž1994. 525–539. w35x J.M. Petras, Spinocerebellar neurons in the rhesus monkey, Brain Res. 130 Ž1977. 146–151. w36x D.M. Snow, P.C. Letourneau, Neurite outgrowth on a step gradient of chondroitin sulfate proteoglycan ŽCS-PG., J. Neurosci. 23 Ž1992. 322–336.
w37x R.L. Snyder, R.L.M. Faull, W.R. Mehler, A comparative study of the neurons of origin of the spinocerebellar afferents in the rat, cat and squirrel monkey based on the retrograde transport of horseradish peroxidase, J. Comp. Neurol. 181 Ž1978. 833–852. w38x L.J. Stensaas, Regeneration in the spinal cord of the newt Notopthalmus (Triturus) pyrrhogaster, in: C.C. Kao, R.P. Bunge, P.J. Reier ŽEds.., Spinal Cord Reconstruction, Raven Press, New York, 1983, pp. 121–149. w39x B. Stilling, Neue Untersuchungen uber den Bau des Ruckenmarks, ¨ ¨ H. Hotop, Cassel, 1859, 5 Vols., 1192 pp. w40x J.R. Terman, X.M. Wang, G.F. Martin, Growth of dorsal spinocerebellar axons through a lesion of their spinal pathway during early development in the North American opossum, Didelphis Õirginiana, Dev. Brain Res. 93 Ž1996. 33–48. w41x R. Tuttle, B.L. Schlaggar, J.E. Braisted, D.D.M. O’Leary, Maturation-dependent upregulation of growth-promoting molecules in developing cortical plate controls thalamic and cortical neurite growth, J. Neurosci. 15 Ž1995. 3039–3052. w42x J. Walter, S. Henke-Fahle, F. Bonhoeffer, Avoidance of posterior tectal membranes by temporal retinal axons, Development 101 Ž1987. 909–913. w43x X.M. Wang, Y.Q. Qin, X.M. Xu, G.F. Martin, Developmental plasticity of reticulospinal and vestibulospinal axons in the North American opossum, Didelphis Õirginiana, J. Comp. Neurol. 349 Ž1994. 288–302. w44x X.M. Wang, J.R. Terman, G.F. Martin, Evidence for growth of supraspinal axons through the lesion after transection of the thoracic cord in the developing opossum, Didelphis Õirginiana, J. Comp. Neurol. 371 Ž1996. 104–115. w45x X.M. Wang, D.M. Basso, J.C. Bresnahan, J.R. Terman, G.F. Martin, Are supraspinal and propriospinal axons which grow through the lesion after transection of the thoracic cord in developing opossums maintained in the adult animal?, Soc. Neurosci. Abstr. 22 Ž1996. 1489. w46x X.M. Wang, Y.Q. Qin, J.R. Terman, G.F. Martin, Early development and developmental plasticity of the fasciculus gracilis in the North American opossum Ž Didelphis Õirginiana., Dev. Brain Res. 98 Ž1997. 151–163. w47x Q. Xu, G. Grant, Collateral projections of neurons from the lower part of the spinal cord to anterior and posterior cerebellar termination areas. A retrograde fluorescent double labeling study in the cat, Exp. Brain Res. 72 Ž1988. 562–576. w48x Q. Xu, G. Grant, Course of spinocerebellar axons in the ventral and lateral funiculi of the spinal cord with projections to the anterior lobe: an experimental anatomical study in the cat with retrograde tracing techniques, J. Comp. Neurol. 345 Ž1994. 288–302. w49x X.M. Xu, G.F. Martin, Developmental plasticity of the rubrospinal tract. Studies using the North American opossum, J. Comp. Neurol. 279 Ž1989. 368–387. w50x H. Yaginuma, Prenatal development of the spinocerebellar fibers in the rat, Soc. Neurosci. Abstr. 13 Ž1987. 1116. w51x J. Yamada, K. Shirao, T. Kitamura, H. Sato, Trajectory of spinocerebellar fibers passing through the inferior and superior cerebellar peduncles in the rat spinal cord: a study using horseradish peroxidase with pedunculotomy, J. Comp. Neurol. 304 Ž1991. 147– 160. w52x H.S. Yin, M.E. Selzer, Axonal regeneration in lamprey spinal cord, J. Neurosci. 3 Ž1983. 1135–1144.
Short communication
Developmental plasticity of selected spinocerebellar axons. Studies using the North American opossum, Didelphis Õirginiana Jonathan R. Terman, Xian M. Wang, George F. Martin
)
Department of Cell Biology, Neurobiology and Anatomy and Neuroscience Program, The Ohio State UniÕersity, College of Medicine, 333 West Tenth AÕe., Columbus, OH 43210, USA Accepted 17 June 1997
Abstract When the thoracic spinal cord of the opossum is hemisected at postnatal day 5 or 8, but not at day 12 or later ages, spinocerebellar axons which originate from spinal border cells, the sacralrcoccygeal ventrolateral nucleus, and Stilling’s nucleus grow through the lesion and reach the cerebellum. The critical period for such growth is comparable to that reported previously for spinocerebellar axons originating within Clarke’s nucleus and for axons of the fasciculus gracilis, but shorter than that for most descending spinal axons. It appears, therefore, that differences exist in the ability of ascending and descending axons to traverse a lesion of their spinal pathway during development. q 1997 Elsevier Science B.V. Keywords: Clarke’s nucleus; Marsupial; Regeneration; Sacralrcoccygeal ventrolateral nucleus; Spinal border cell; Stilling’s nucleus
Axons fail to grow beyond a lesion of the spinal cord in adult mammals, but at least some do when the lesion is made early in development Žsee reviews w30,33x.. In the North American opossum, Didelphis Õirginiana, supraspinal axons grow around w28,43,49x or through w29,44,45x a lesion of their spinal pathway when it is made prior to postnatal day ŽPD. 30, but not when it is made at older ages. Interestingly, the critical period for such growth is not the same for all axons. For example, the critical period for lateral vestibulospinal and certain pontine reticulospinal axons ends before PD12, yet that for rubrospinal and raphespinal axons continues until PD30 w43,44,49x. Although growth through or around a lesion of the spinal cord has been documented for axons of most, if not all, descending spinal pathways in developing mammals, it has only been described for a limited number of ascending axons. Using the North American opossum, we have shown that axons which originate in Clarke’s nucleus ŽCN., i.e., axons of the dorsal spinocerebellar tract ŽDSCT., grow through a lesion of their spinal pathway when it is made at PD5-8, but not when it is made at PD12 or at later ages w40x. Spinocerebellar axons differ, however, in origin
)
Corresponding author. Fax: q1 Ž614. 292-7659; E-mail: [email protected] 0165-3806r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 1 1 2 - 0
w31,32x, developmental history w1,27,50x, laterality w31,32x, funicular position w23,48x, peduncular course w21,51x, and terminal distribution w5,47x. It is possible, therefore, that differences also exist in their ability to grow through or around a lesion of their spinal pathway. In the present study, we have asked if spinocerebellar axons originating from spinal border cells ŽSBC. w13,37x, neurons within the sacralrcoccygeal ventrolateral nucleus ŽVLN. w31,37x, or neurons within Stilling’s sacral nucleus ŽSN. w9,35,39x are capable of such growth and, if so, whether the critical periods for it are the same as that for axons of the DSCT. The animals used in this study had been employed previously and a detailed description of the methods can be found elsewhere w40x. Opossums with appropriately aged young ŽPD5 w n s 7x, PD8 w n s 4x, PD12 w n s 8x, PD20 w n s 2x, and PD26 w n s 2x. were anesthetized with ketamine and Metafane and placed in a supine position to expose their pouch. During anesthesia, the pouch sphincter relaxes making it possible to anesthetize the pups by hypothermia. The spinal cord of the anesthetized pups was exposed surgically and hemisected at mid-thoracic levels using a microblade. The lesion was made by inserting the microblade slightly to the left of the midline and pulling it to the right until the right half of the cord was severed. To ensure that the hemisection was complete, a microdissection needle with a 908 angle was drawn through the lesion
310
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
cavity. In addition, 3 of the PD5 pups were sacrificed shortly after lesioning and examined for the presence of a complete hemisection. In these cases the spinal cord was removed after perfusion and the lesioned segment was embedded in paraffin, sectioned at 20 m m in the frontal plane, and stained for Nissl substance. In each of these animals the hemisection was obvious grossly and histologically. The remaining animals were maintained 4–9 months before being anesthetized using sodium pentobarbital for bilateral injections of 5% Fluoro-Gold ŽFG. or 2% Fast Blue ŽFB. into the anterior lobe of the cerebellum, the primary target of spinocerebellar axons w23x. No differences have been observed in our laboratory between the ability of FB or FG to retrogradely label neurons Žunpublished results.. FG is preferable to FB since it holds up better to ultraviolet light. Unfortunately, we have found FG to be more toxic to younger animals, so we have employed FB as a retrograde tracer in such cases. After a 7–10 day survival, each animal was reanesthetized and sacrificed by perfusion so its brain and spinal cord could be removed and sectioned using a freezing microtome. Every section of the spinal cord caudal to the lesion site was examined for labeling of SBC Žlower thoracic and upper lumbar levels., neurons in the VLN Žsacral and coccygeal levels., and neurons in SN Žsacral and coccygeal levels.. Since the cerebellar projections of these nuclei are crossed in the opossum Žunpublished results., as in other species w31,32x, labeled neurons contralateral to the lesion were considered to have supported axons which grew through or around the lesion to the cerebellum. Labeled neurons ipsilateral to the lesion were used as internal controls. In addition, the cerebellum and the spinal cord rostral to the lesion were sectioned and examined for symmetry of the injections and labeling. In four of the cases lesioned at PD5, labeled neuronal profiles in each of the above mentioned nuclei were counted in every sixth section on the experimental and control sides. Results were expressed as the mean " the standard error of the mean and analyzed using the paired t-test for significant differences in the number of labeled neurons on the two sides. When the thoracic cord was hemisected on PD5 or PD8 and bilateral injections of FB or FG were made into the cerebellum 4–9 months later, labeling was present bilaterally within SBC, within neurons of the VLN, and within neurons of SN. The results from one case are shown in Fig. 1. The site of the lesion was not evident grossly ŽFig. 1A, arrow. and after sectioning, recognizable spinal cord was present on the lesioned side ŽFig. 1C; right side. although it was abnormal in appearance. Observed abnormalities included fusion of the dorsal horns, decrease in the size of the dorsal columns, an uncharacteristic appearance of the grey matter, and a relative thinness of the white matter on the lesioned side. Examination of the injection sites and the presence of bilaterally symmetrical labeling of CN rostral to the lesion Žopen arrows, Fig. 1B. suggested that the injections were similar bilaterally. Caudal
to the lesion, bilateral labeling was found in SBC ŽFig. 1D, closed arrows., neurons of the VLN ŽFig. 1E, arrows., and neurons of SN ŽFig. 1F, arrows.. Counts from every sixth section revealed that labeled neuronal profiles were fewest in number contralateral to the lesion and the results were statistically significant Ž P - 0.01; data not shown.. It will be recalled that neurons in each of the above nuclei project contralaterally to the cerebellum. Note that bilateral labeling was also present in CN caudal to the lesion ŽFig. 1D, open arrows. and that labeled neurons were fewest ipsilateral to the lesion Žright side. w40x. It will be recalled that the DSCT projects ipsilaterally to the cerebellum. Following hemisection of the mid-thoracic spinal cord on PD12 or at later stages of development and bilateral injections of FB or FG into the cerebellum 4–6 months later, labeling was not found in SBC, within neurons of the VLN, or within neurons of SN contralateral to the lesion, although it was present in neurons within the same areas on the ipsilateral side. The results from a PD12 case are documented in Fig. 2. The lesion site was identifiable grossly ŽFig. 2A, arrow. and histologically it was seen to include over half the cord ŽFig. 2C.. Examination of the injection sites and labeling in CN rostral to the lesion Žopen arrows, Fig. 2B. suggested that the injections were symmetrical bilaterally. Caudal to the lesion, labeling in SBC ŽFig. 2D, closed arrows., the VLN ŽFig. 2E, arrows., and SN ŽFig. 2F, arrow. was only present ipsilateral to the lesion. As expected w40x, labeling in CN caudal to the lesion ŽFig. 2D; open arrow. was limited to the side contralateral to the lesion. In all of the PD5 and 8 cases, labeled spinocerebellar axons were present in their normal position ipsilateral to the lesion in what appeared to be regenerated spinal cord at the lesion site Žarrows, Fig. 1C., as well as in normal spinal cord rostral Žarrows, Fig. 1B., and caudal Žnot seen at this magnification. to it. Although they were not counted, labeled axons ipsilateral to the lesion appeared to be fewer in number than those on the contralateral side ŽFig. 1B and C.. A few labeled axons also were present rostral and ipsilateral to the lesion in the PD12 cases. Such axons likely originated rostral to the lesion, however, since axonal labeling was not found at the lesion site or caudal to it on the ipsilateral side. Our results indicate that, in the opossum, axons originating from SBC, the VLN, and SN, like those which originate in CN, grow through a lesion of their spinal pathway and reach the cerebellum if the lesion is made early enough in development. It is unlikely that our lesions were incomplete in the PD5–8 cases even though recognizable spinal cord was present at the lesion site. Evidence for a complete hemisection included the abnormal appearance of the spinal cord at the lesion site and the obvious presence of a hemisection in the PD5 pups sacrificed shortly after lesioning. Furthermore, when the lesion was made on PD12 using the same technique, it was clear that it included over half the cord.
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
As described previously w29,40,44x, reconstruction of recognizable spinal cord occurs at the lesion site after hemisection or transection in the PD5 opossum. Compara-
311
ble results have been reported after crush lesions of the spinal cord in the South American opossum, Monodelphis domestica Žsee review w33x.. Regeneration of damaged
Fig. 1. A photograph of a portion of spinal cord including the lesion site Žarrow. from an animal subjected to mid-thoracic hemisection on postnatal ŽPD. 5 and bilateral injections of Fluoro-Gold into the cerebellum 6 months later is shown in ŽA.. A section of thoracic cord ŽT4. rostral to the lesion from the same case is illustrated in ŽB.. Note that labeling of Clarke’s nucleus ŽCN. Žopen arrows. was bilaterally symmetrical suggesting that the injections were comparable on the two sides. Labeled axons were found within the lateral and ventral funiculi bilaterally, although they appeared to be fewest on the side of the lesion Žclosed arrows.. The asterisk delineates the side opposite the lesion in this section and in those which follow. A section through the lesion site is shown in ŽC. and labeled axons were seen on the side of the lesion Žarrows, right side. in what appeared to be reconstructed spinal cord. A photomicrograph from the first lumbar section of the cord is shown in ŽD.. Notice that spinal border cells ŽSBC. were labeled bilaterally Žclosed arrows.. Labeled neurons were also present within CN on both sides Žopen arrows.. A section from the second sacral segment of the cord is seen in ŽE. and it can be seen that labeled cells were present bilaterally in the sacralrcoccygeal ventrolateral nucleus ŽVLN. Žarrows.. A section of coccygeal cord is seen in ŽF. and labeled cells were present bilaterally in Stilling’s nucleus ŽSN. Žarrows.. The scale bar in ŽB. applies to ŽC. – ŽF..
312
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
spinal cord occurs in non-mammalian vertebrates and it is apparently initiated by proliferation of ependymal cells which bridge the gap at the lesion site Žsee review w30x.. The mechanisms which underlie spinal cord reorganization in the developing opossum are not known, nor is it known
why it no longer occurs at PD12 or at later ages. Likewise, it is not known whether growth through the lesion resulted from regeneration of cut axons, growth of late arriving axons that were not damaged by the lesion, or both. Both mechanisms contribute to extension of descending axons
Fig. 2. A portion of the spinal cord including the lesion site Žarrow. from an animal subjected to hemisection of the mid-thoracic cord on PD12 and bilateral injections of Fast Blue into the cerebellum 4 months later is shown in ŽA.. A photomicrograph of a section of thoracic cord rostral to the lesion ŽT4. from the same case is shown in ŽB.. Note that labeled cells were present in CN in relatively equal numbers bilaterally Žopen arrows. suggesting that our injections were symmetrical. The asterisk delineates the side opposite the lesion in this section and in those which follow. A section through the lesion is shown in ŽC. where it can be seen that over half of the cord was involved. The spared lateral funiculus ŽLF. and ventral funiculus ŽVF. are indicated. A section through the first segment of the lumbar cord is shown in ŽD.. Labeled SBC Žclosed arrows. were present only ipsilateral to the lesion. In addition, labeling in CN Žopen arrow. was only present contralateral to the lesion. A section from the second segment of the sacral cord is shown in ŽE.. Labeled cells were only seen in the VLN Žarrows. ipsilateral to the lesion. A section of coccygeal cord is shown in ŽF.. Labeled neurons in SN Žarrow. were present only on the side of the lesion. The scale bar in ŽB. applies to ŽC. – ŽF..
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
through the lesion site when the spinal cord is injured during early development in the chick w22x and opossum Žunpublished results.. Major differences were not found in the critical periods for developmental plasticity of the spinocerebellar axons studied. In contrast, clear differences exist for descending axons w43,44,49x. For example, lateral vestibulospinal and certain pontine reticulospinal axons do not grow through a lesion when it is made on PD12 or later, but rubrospinal and raphespinal axons do. In fact, some rubral and raphe axons grow through such a lesion when it is made as late as PD30. The critical period for developmental plasticity of spinocerebellar axons in the opossum is shorter than that for many descending spinal axons. Spinocerebellar axons no longer traverse the lesion site after hemisection of the thoracic cord on PD12 Žw40x; present results.; but, as noted above, some supraspinal axons do when the lesion is made as late as PD30 w44x. The critical period for developmental plasticity of axons within the fasciculus gracilis, another ascending spinal pathway, is also relatively short in the opossum w46x. It is possible, therefore, that differences exist in the ability of ascending and descending axons to grow beyond a lesion of their spinal pathway. Such differences also have been reported for reptiles w15,16x, frogs w3,7,11,18x, salamanders w14,38x, fish w4,6x, and lampreys w52x, but the mechanisms which underlie them have not been elucidated. Loss of developmental plasticity may be due to a diminished ability to initiate andror sustain axonal growth with age w10,17x. That hypothesis is supported in the opossum by the finding that supraspinal axons which reach the spinal cord first lose the ability to grow through or around a lesion before those which arrive at a later date w43,44,49x. It is interesting, however, that the critical periods for developmental plasticity of different spinocerebellar axons appear comparable even though they apparently have different developmental histories w1,27,50x. It is possible, therefore, that the development of a non-permissive environment for axonal growth plays a role in the loss of developmental plasticity. In the North American opossum, a temporal correlation exists between the appearance of myelin, which is known to contain proteins inhibitory to neurite extension Že.g., w8x., and the end of the critical period for plasticity of descending spinal axons w19,43,44,49x. In fact, evidence for a cause and effect relationship has been provided for the chick Že.g., w25x.. However, the critical period for developmental plasticity of spinocerebellar axons Žw40x; present study. ends well before the appearance of myelin w19x. If the development of a non-permissive environment is a factor in loss of spinocerebellar plasticity, myelin does not apparently play a role Žsee also w26x.. It appears, therefore, that the factors which result in loss of developmental plasticity are not the same for all axons. Indeed, it is becoming increasingly clear that differences exist in the ability of axons to grow
313
in the same regions of the brain and spinal cord, and on similar substrates in vitro w2,12,20,24,34,36,41,42x.
Acknowledgements The authors wish to thank Ms. Mary Ann Jarrell for surgical assistance and tissue processing, and Drs. James S. King and Phillip G. Popovich for helpful criticism of the manuscript. Our studies were supported by USPHS Grants NS-25095 and NS-10165.
References w1x M.L. Arsenio Nunes, C. Sotelo, Development of the spinocerebellar system in the postnatal rat, J. Comp. Neurol. 237 Ž1985. 291–306. w2x D. Bagnard, F. Mann, S. Henke-Fahle, J. Bolz, Developmental mechanisms underlying the segregation of afferent and efferent cortical projections, Soc. Neurosci. Abstr. 21 Ž1995. 1285. w3x M.J. Beattie, J.C. Bresnahan, G. Lopate, Metamorphosis alters the response to spinal cord transection in Xenopus laeÕis frogs, J. Neurobiol. 21 Ž1990. 1108–1122. w4x T. Becker, M.F. Wullimann, C.G. Becker, R.R. Bernhardt, M. Schachner, Axonal regrowth after spinal cord transection in adult zebrafish, J. Comp. Neurol. 377 Ž1997. 577–595. w5x S. Berretta, V. Perciavalle, R.E. Poppele, Origin of spinal projections to the anterior and posterior lobes of the rat cerebellum, J. Comp. Neurol. 305 Ž1991. 273–281. w6x S.M. Bunt, P. Fill-Moebs, Selection of pathways by regenerating spinal cord fiber tracts, Dev. Brain Res. 16 Ž1984. 307–311. w7x H.L. Campbell, M.S. Beattie, J.C. Bresnahan, The response of dorsal root afferent fibers to dorsal funiculus lesions in developing Rana catesbeiana tadpoles, Soc. Neurosci. Abstr. 10 Ž1984. 1024. w8x P. Caroni, M.E. Schwab, Two membrane protein fractions from rat central myelin with inhibitory properties for neurite growth and fibroblast spreading, J. Cell Biol. 106 Ž1988. 1281–1288. w9x H.T. Chang, Caudal extension of Clarke’s nucleus in the spider monkey, J. Comp. Neurol. 95 Ž1951. 43–51. w10x D.F. Chen, S. Jhaveri, G.E. Schneider, Intrinsic changes in developing retinal neurons result in regenerative failure of their axons, Proc. Natl. Acad. Sci. USA 92 Ž1995. 7287–7291. w11x J.D.W. Clarke, D.A. Tonge, N.H.K. Holder, Stage-dependent restoration of sensory dorsal columns following spinal cord transection in anuran tadpoles, Proc. R. Soc. London B 227 Ž1986. 67–82. w12x S.A. Colamarino, M. Tessier-Lavigne, The axonal chemoattractant netrin-1 is also a chemorepellent for trochlear motor axons, Cell 81 Ž1995. 621–629. w13x S. Cooper, C.S. Sherrington, Gower’s tract and spinal border cells, Brain 63 Ž1940. 123–134. w14x B.M. Davis, J.L. Ayers, L. Koran, J. Carlson, M.C. Anderson, S.B. Simpson Jr., Time course of salamander spinal cord regeneration and recovery of swimming: HRP retrograde pathway tracing and kinematic analysis, Exp. Neurol. 108 Ž1990. 198–213. w15x M.T. Duffy, A. Hawrych, D.R. Liebich, S.B. Simpson Jr., Regeneration of cerebrospinal fluid contacting neurons ŽCSFCN. in the regenerated tail spinal cord of the lizard Anolis carolinensis, Soc. Neurosci. Abstr. 19 Ž1993. 1716. w16x M.T. Duffy, S.B. Simpson Jr., D.R. Liebich, B.M. Davis, Origin of spinal cord axons in the lizard regenerated tail: supernormal projections from local spinal neurons, J. Comp. Neurol. 293 Ž1990. 208–222. w17x J.W. Fawcett, Intrinsic neuronal determinants of regeneration, Trends Neurosci. 15 Ž1992. 5–8.
314
J.R. Terman et al.r DeÕelopmental Brain Research 102 (1997) 309–314
w18x C. Forehand, P. Farel, Anatomical and behavioral recovery from the effects of spinal cord transection: dependence on metamorphosis in anuran larvae, J. Neurosci. 2 Ž1982. 654–662. w19x G.T. Ghooray, G.F. Martin, The development of myelin in the spinal cord of the North American opossum and its possible role in loss of rubrospinal plasticity. A study using myelin basic protein and galactocerebroside immunohistochemistry, Dev. Brain Res. 72 Ž1993. 67–74. w20x R.W. Guillery, C.A. Mason, J.S.H. Taylor, Developmental determinants at the mammalian optic chiasm, J. Neurosci. 15 Ž1995. 4727– 4737. w21x G. Grant, Q. Xu, Routes of entry into the cerebellum of spinocerebellar axons from the lower part of the spinal cord. An experimental study in the cat, Exp. Brain Res. 72 Ž1988. 543–561. w22x S.J. Hasan, H.S. Keirstead, G.D. Muir, J.D. Steeves, Axonal regeneration contributes to repair of injured brainstem-spinal neurons in embryonic chick, J. Neurosci. 13 Ž1993. 492–507. w23x J.C. Hazlett, G.F. Martin, R. Dom, Spino-cerebellar fibers of the opossum, Didelphis Õirginiana, Brain Res. 33 Ž1971. 257–271. w24x S. Jhaveri, Midline glia of the tectum: a barrier for developing retinal axons, Perspect. Dev. Neurobiol. 1 Ž1993. 237–243. w25x H.S. Keirstead, S.J. Hasan, G.D. Muir, J.D. Steeves, Suppression of the onset of myelination extends the permissive period for the functional repair of embryonic spinal cord, Proc. Natl. Acad. Sci. USA 89 Ž1992. 11664–11668. w26x R.E. MacLaren, Expression of myelin proteins in the opossum optic nerve: late appearance of inhibitors implicates an earlier non-myelin factor in preventing ganglion cell regeneration, J. Comp. Neurol. 372 Ž1996. 27–36. w27x G.F. Martin, J.L. Culbertson, J.C. Hazlett, Observations on the early development of ascending spinal pathways. Studies using the North American opossum, Anat. Embryol. 166 Ž1983. 191–207. w28x G.F. Martin, X.M. Xu, Evidence for developmental plasticity of the rubrospinal tract. Studies using the North American opossum, Dev. Brain. Res. 39 Ž1988. 303–308. w29x G.F. Martin, J.R. Terman, X.M. Wang, Evidence for growth of supraspinal axons through hemisected and transected spinal cord in the developing opossum, Didelphis Õirginiana, Soc. Neurosci. Abstr. 20 Ž1994. 1270. w30x G.F. Martin, G.T. Ghooray, X.M. Wang, X.M. Xu, X.C. Zou, Models of spinal cord regeneration, in: F.J. Seil ŽEd.., Neural Regeneration, Progress in Brain Research, Vol. 103, Elsevier, Amsterdam, 1994, pp. 175–201. w31x M. Matsushita, Y. Hosoya, Cells of origin of the spinocerebellar tract in the rat, studied with the method of retrograde transport of horseradish peroxidase, Brain Res. 173 Ž1979. 185–200. w32x M. Matsushita, Y. Hosoya, M. Ikeda, Anatomical organization of the spinocerebellar system in the cat, as studied by retrograde transport of horseradish peroxidase, J. Comp. Neurol. 184 Ž1979. 81–106. w33x J. Nicholls, N. Saunders, Regeneration of immature mammalian spinal cord after injury, Trends Neurosci. 19 Ž1996. 229–234. w34x A. Nose, M. Takeichi, C.S. Goodman, Ectopic expression of connectin reveals a repulsive function during growth cone guidance and synapse formation, Neuron 13 Ž1994. 525–539. w35x J.M. Petras, Spinocerebellar neurons in the rhesus monkey, Brain Res. 130 Ž1977. 146–151. w36x D.M. Snow, P.C. Letourneau, Neurite outgrowth on a step gradient of chondroitin sulfate proteoglycan ŽCS-PG., J. Neurosci. 23 Ž1992. 322–336.
w37x R.L. Snyder, R.L.M. Faull, W.R. Mehler, A comparative study of the neurons of origin of the spinocerebellar afferents in the rat, cat and squirrel monkey based on the retrograde transport of horseradish peroxidase, J. Comp. Neurol. 181 Ž1978. 833–852. w38x L.J. Stensaas, Regeneration in the spinal cord of the newt Notopthalmus (Triturus) pyrrhogaster, in: C.C. Kao, R.P. Bunge, P.J. Reier ŽEds.., Spinal Cord Reconstruction, Raven Press, New York, 1983, pp. 121–149. w39x B. Stilling, Neue Untersuchungen uber den Bau des Ruckenmarks, ¨ ¨ H. Hotop, Cassel, 1859, 5 Vols., 1192 pp. w40x J.R. Terman, X.M. Wang, G.F. Martin, Growth of dorsal spinocerebellar axons through a lesion of their spinal pathway during early development in the North American opossum, Didelphis Õirginiana, Dev. Brain Res. 93 Ž1996. 33–48. w41x R. Tuttle, B.L. Schlaggar, J.E. Braisted, D.D.M. O’Leary, Maturation-dependent upregulation of growth-promoting molecules in developing cortical plate controls thalamic and cortical neurite growth, J. Neurosci. 15 Ž1995. 3039–3052. w42x J. Walter, S. Henke-Fahle, F. Bonhoeffer, Avoidance of posterior tectal membranes by temporal retinal axons, Development 101 Ž1987. 909–913. w43x X.M. Wang, Y.Q. Qin, X.M. Xu, G.F. Martin, Developmental plasticity of reticulospinal and vestibulospinal axons in the North American opossum, Didelphis Õirginiana, J. Comp. Neurol. 349 Ž1994. 288–302. w44x X.M. Wang, J.R. Terman, G.F. Martin, Evidence for growth of supraspinal axons through the lesion after transection of the thoracic cord in the developing opossum, Didelphis Õirginiana, J. Comp. Neurol. 371 Ž1996. 104–115. w45x X.M. Wang, D.M. Basso, J.C. Bresnahan, J.R. Terman, G.F. Martin, Are supraspinal and propriospinal axons which grow through the lesion after transection of the thoracic cord in developing opossums maintained in the adult animal?, Soc. Neurosci. Abstr. 22 Ž1996. 1489. w46x X.M. Wang, Y.Q. Qin, J.R. Terman, G.F. Martin, Early development and developmental plasticity of the fasciculus gracilis in the North American opossum Ž Didelphis Õirginiana., Dev. Brain Res. 98 Ž1997. 151–163. w47x Q. Xu, G. Grant, Collateral projections of neurons from the lower part of the spinal cord to anterior and posterior cerebellar termination areas. A retrograde fluorescent double labeling study in the cat, Exp. Brain Res. 72 Ž1988. 562–576. w48x Q. Xu, G. Grant, Course of spinocerebellar axons in the ventral and lateral funiculi of the spinal cord with projections to the anterior lobe: an experimental anatomical study in the cat with retrograde tracing techniques, J. Comp. Neurol. 345 Ž1994. 288–302. w49x X.M. Xu, G.F. Martin, Developmental plasticity of the rubrospinal tract. Studies using the North American opossum, J. Comp. Neurol. 279 Ž1989. 368–387. w50x H. Yaginuma, Prenatal development of the spinocerebellar fibers in the rat, Soc. Neurosci. Abstr. 13 Ž1987. 1116. w51x J. Yamada, K. Shirao, T. Kitamura, H. Sato, Trajectory of spinocerebellar fibers passing through the inferior and superior cerebellar peduncles in the rat spinal cord: a study using horseradish peroxidase with pedunculotomy, J. Comp. Neurol. 304 Ž1991. 147– 160. w52x H.S. Yin, M.E. Selzer, Axonal regeneration in lamprey spinal cord, J. Neurosci. 3 Ž1983. 1135–1144.