EXPERIMENTAL NEUROLOGY ARTICLE NO. 143, 45–60 (1997) EN966318 Fetal Spinal Cord Transplants Rescue Some Axotomized Rubrospinal Neurons from Retrograde Cell Death in Adult Rats FUTOSHI MORI,*,† B. TIMOTHY HIMES,*,‡ MASAYOSHI KOWADA,† MARION MURRAY,* AND ALAN TESSLER*,‡ *Department of Neurobiology and Anatomy, The Medical College of Pennsylvania/Hahnemann University, Philadelphia, Pennsylvania 19129; †Department of Neurosurgery, Akita University School of Medicine, 1-1-1 Hondo, Akita City, Akita 010, Japan; and ‡The Philadelphia Veterans Administration Medical Center, Philadelphia, Pennsylvania 19104 the site of a complete spinal cord transection (45) or, after a delay, into an incomplete contusion injury (67). The mechanisms by which transplants enhance locomotor performance are unknown. Rescue of neurons from axotomy-induced retrograde cell death represents one possible mechanism. Fetal spinal cord transplants keep alive all the 35–40% of the ipsilateral Clarke’s nucleus neurons at the L1 segment that would otherwise die after hemisection of the midthoracic spinal cord of either newborn or adult rats (27). They also prevent death of all the 56% of the contralateral red nucleus (RN) neurons that die after midthoracic spinal cord hemisection in newborn rats (12). Rubrospinal neurons atrophy (20, 68, 72) and die (20, 22, 50, 60, 72) after axotomy in adults, but it has not yet been determined whether transplants rescue RN neurons in these animals. One aim of the present investigation was to provide this information. In both newborns and adults, transplants are thought to rescue axotomized neurons at least in part by serving as a surrogate source of target-derived neurotrophic factors (12, 18, 24, 27). Additional mechanisms may apply, however, and these are likely to differ in newborns and adults. After spinal cord injury in newborns, for example, interrupted and late-developing supraspinal axons are able to grow through transplants and into caudal host spinal cord, where they appear to establish connections with their normal targets (4, 8). This has provided part of the evidence suggesting that growth and synapse formation favor permanent survival of axotomized neurons (9, 12, 38). Growth into intraspinal fetal spinal cord transplants has also been observed in adult recipients, but this is largely restricted to axons whose perikarya are located within 0.5 mm of the transplants; only rare supraspinal axons grow into these transplants in adults (32). A second goal of the present study was to examine whether the spinal projections of RN neurons that survive axotomy can extend into fetal spinal cord transplants and into caudal host spinal cord. Because the RN plays an important role in the control of limb movement (42; Intraspinal transplants of fetal spinal cord may contribute to recovery after spinal cord injury by keeping axotomized neurons alive. In this study we examined whether transplants rescued axotomized red nucleus (RN) neurons from retrograde cell death in adult rats. RN neurons were labeled by retrograde transport of Fluorogold (FG); 1 week later right-sided RN neurons were axotomized by left-sided hemisection at C3–4 vertebral level, and Embryonic Day 14 spinal cord or gelfoam was introduced into the cavity. Additional rats received hemisection and a transplant of fetal spinal cord or gelfoam without FG injection. At 2 and 4 months, the number of neurons in the magnocellular portion of the RN contralateral to the hemisection decreased 35–40% in rats that received gelfoam; mean soma area of surviving neurons decreased 40%. RN cell loss was reduced to 20% in rats that received fetal spinal cord transplants, but the decrease in mean soma area was unchanged. Transplants therefore rescued about half of the axotomized RN neurons that otherwise would have died but did not prevent perikaryal atrophy. Anterograde transport of WGA–HRP injected into RN 2 months after transplantation showed that rubrospinal axons reached the site of injury but rarely entered transplants; FG injections caudal to transplants showed that axons of transplant neurons extended at least two segments into host spinal cord. Fetal spinal cord transplants may therefore contribute to locomotor recovery in adults with spinal cord injuries both by preventing retrograde cell death and by establishing novel circuits across the site of injury. r 1997 Academic Press INTRODUCTION Intraspinal transplants of fetal spinal cord enhance the development of locomotor performance after the spinal cord of newborn animals has been either completely transected (29, 31, 45) or overhemisected (11, 35). More limited recovery has been observed in adult recipients when transplants were placed acutely into 45 0014-4886/97 $25.00 Copyright r 1997 by Academic Press All rights of reproduction in any form reserved. 46 MORI ET AL. reviewed in 28), we also examined the projections formed between fetal spinal cord transplants and caudal host spinal cord. Information about the anatomical connections between RN neurons and transplants and between transplants and host spinal cord would provide additional insight into the mechanisms by which transplants contribute to the recovery of locomotor performance after spinal cord injury in adults. MATERIALS AND METHODS Animals Sixty adult female Sprague–Dawley rats weighing 225–300 g (Zivic-Miller Laboratories, Allison Park, PA) were used in these studies. Fourteen rats underwent spinal cord hemisection alone, and 30 rats received fetal spinal cord transplants into the hemisection site. Eight age-matched animals served as intact, unoperated controls and 4 received fluorescent tracer Fluorogold (FG; Fluorochrome, Inc., Englewood, CO) injection only as control. The surgical procedures were carried out in accordance with the Laboratory Animal Welfare Act Guide for the Care and Use of Laboratory Animals (NIH, DHEW Publication No. 78-23, Revised 1978) after review and approval by the Animal Care and Use Committee of the Medical College of Pennsylvania/ Hahnemann University. Spinal Cord Hemisection and Transplantation After rats were deeply anesthetized with an intraperitoneal injection of xylazine (5 mg/kg), ketamine (50 mg/kg), and acepromazine (0.4 mg/kg), the spinal cord was exposed by a partial laminectomy of the C3 and C4 vertebrae, and a hemisection cavity 3–4 mm in length was formed in the left side of the spinal cord with microscissors and gentle aspiration. The lesion completely severed the axons of the rubrospinal tract in the ipsilateral lateral funiculus as well as the ipsilateral dorsal and ventral funiculi and gray matter. In the hemisection only group, the cavity was filled with gelfoam soaked with saline. In transplant recipients, the cavity was filled with one or two whole pieces of Embryonic Day 14 (E14) spinal cord removed from embryos of timed pregnant Sprague–Dawley rats. The day of insemination was considered E0. The dura matter was closed with interrupted 10-O sutures and covered with a piece of hydrocephalus shunt film (Durafilm; Codman Surlef, Inc.), the muscle and skin were closed in layers. The techniques for transplantation have been more fully described in previous publications (12, 53). Prelabeling of RN Neurons by Retrograde Transport of Fluorogold To determine that cells surviving hemisection in the presence of transplants include RN neurons that had been axotomized, we prelabeled neurons using the FG. This procedure produces labeling of neurons that lasts at least 2 months (27, 39) but does not cause degeneration of retrogradely labeled cells at concentrations equal to or less than 2.5% (62). After the rats had been anesthetized as described above and the spinal cord exposed by C6 laminectomy, a single pressure injection of 2% FG (2 µl) was made into both sides of the spinal cord through a glass pipette (tip diameter 50–60 µm) attached to a 10-µl Hamilton syringe. The wound was closed as described above. One week later, the rats underwent left-sided hemisection of the C3–4 spinal cord with insertion of a spinal cord transplant or gelfoam into the lesion cavity. Tissue Processing and Staining Animals survived 2 or 4 months after surgery. There were no consistent differences in the results between these survival times. For perfusion the rats received an overdose of sodium pentobarbital (75 mg/kg, ip) and were perfused through the left ventricle with phosphatebuffered saline (PBS) followed by fixative containing 4% paraformaldehyde and 0.02% picric acid in 0.1 M phosphate buffer, pH 7.3. The perfused rats were then positioned on a stereotaxic frame, the brain was divided into two parts in the plane of 7.0 mm from bregma (48) and the brain stem and spinal cord were removed. The tissue was cryoprotected by immersion in an ascending series of 10–30% sucrose dissolved in 0.1 M phosphate buffer. In FG-injected rats, tissue blocks containing the RN were serially sectioned at 16 µm on a cryostat and every third section was mounted onto gelatin-coated slides for analysis of FG-labeled neurons. The sections were air dried and briefly immersed in xylene, coverslipped using DPX mounting media (Fluka Chemical Co., Ronkonkoma, NY) and observed with a fluorescence microscope. Additional 16-µm sections from both FG-injected and non-FG-injected rats were directly mounted onto gelatin-coated slides for Nissl–Myelin stain (15) or immunocytochemistry (MAP1B, ED-1, OX-42, GFAP). The C3–4 spinal cord was serially sectioned at 20 µm and stained with Nissl–Myelin stain to confirm the completeness of the hemisection and the presence of a transplant. Animals were eliminated from the study if the hemisection did not sever the ipsilateral lateral funiculus and gray matter (N 5 2) or if no transplant remained in rats that had received a transplant or if prominent cysts were present at the rostral interface between host spinal cord and transplant (N 5 4). Cell Counting and Cell Size Analysis of RN Neurons were counted in the magnocellular portion of the RN at the level of the interpeduncular nucleus, which comprises the caudalmost 480 µm of the RN (48, TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS plate 24) and is the source of most rubrospinal axons (13, 30, 52, reviewed in 58). In rats that did not receive FG injections, the quantitative analysis was performed on every third section stained for Nissl–Myelin; in rats that received FG, the quantitative analysis was performed on every third section prepared for FG visualization as described above. RN neurons with a visible nucleus were counted and the cell soma area was measured with the Bioquant image analysis system attached to a Leitz Dialux microscope at a final magnification of 3676 under bright-field illumination for Nissl–Myelin stain and 3961 under fluorescent illumination for FG. Neurons were classified on the basis of soma area as small (,300 µm2 ), medium (300–500 µm2 ), large (500–1000 µm2 ), or giant (.1000 µm2 ). Corrected cell counts were obtained with the following procedures. Nuclear diameters were measured at a magnification of 31303 for 105 nuclear profiles from both lesioned and unlesioned RN. These data were then processed by means of the Hendry (26) analysis as modified by Smolen et al. (66) to obtain a factor for corrected neuron counts. Total neuron number was determined by the formula (correction factor 3 raw cell counts 3 distance between sections)/section thickness. Because this correction factor assumes that the neuron nucleus is a sphere, the present analysis provided absolute numbers of neurons biased to the extent that the shape of the neuronal nucleus differs from this assumption (47). Stereological counting methods might therefore provide different absolute numbers of neurons. Our primary interest was in side-to-side comparisons between right and left RN neuron numbers, however, and these will be reliably determined by our methods unless axotomy or transplantation changes the extent to which the neuron nucleus can be considered to be spherical. Nuclear shape did not differ on right and left sides. A paired t test was used to analyze the numerical data for side-to-side differences within groups for two variables: mean cell size and corrected total cell numbers. The groups compared were normal animals, animals with hemisection alone, and hemisected animals with a transplant of fetal spinal cord. In all cases, analysis of the numerical data with a one-way ANOVA showed that the RN of unoperated controls did not differ statistically from the ipsilateral RN of lesioned animals in any group for any of the variables measured. Therefore, we used the lesion/control ratio of each animal as the basis for comparison. Overall significance of the ratio data was determined by the KruskalWallis one-way ANOVA (P , 0.05). If significance differences were present, individual post hoc comparisons were made with the Mann–Whitney U test. All analyses were performed using the StatView statistical computer program (Abacus Concepts, Berkeley, CA). 47 Immunocytochemistry Antibodies raised against several cell-specific markers were used to ensure that cells counted in the RN were in fact neurons, particularly because of the need to distinguish neurons labeled with FG from macrophages and microglia that can phagocytose neurons (56). These antibodies were (1) antimicrotubule-associated protein 1B (MAP1B), one of the components of crossbridges between microtubules (61), which recognizes neurons; (2) ED-1, which recognizes an uncharacterized cytoplasmic antigen found in all cells of the rat macrophage/monocyte line (17); (3) OX-42, which recognizes the complement C3bi receptor expressed by neutrophils, monocytes, macrophages, and microglia (59, 57); and (4) antiglial fibrillary acidic protein (GFAP), which recognizes the major protein found in glial intermediate filaments (5, 19), present in astrocytes (36). The immunocytochemical methods were modifications of the procedures of Milligan et al. (43, 44). The antibodies and dilutions used were (1) MAP1B (a generous gift from Dr. Itzhak Fischer), a polyclonal antibody diluted 1:250 (27, 59); (2) ED-1 (Harlan Bioproducts for Science, Indianapolis, IN), a monoclonal antibody diluted 1:150 (27, 43, 44); (3) OX-42 (Harlan Bioproducts for Science), a monoclonal antibody diluted 1:500 (28, 44, 45); and (4) GFAP (Biomedical Technologies, Inc., Stoughton, MA), a polyclonal antibody diluted 1:1000 (27). MAP1B was diluted in 0.1 M PBS, pH 7.4, containing 1% nonfat milk; the others in 0.1 M PBS, pH 7.4, containing 4% nonfat milk. Sections were incubated at room temperature overnight in a humid chamber. After several rinses in PBS, the appropriate peroxidase-conjugated secondary antibody was applied, also diluted in PBS containing nonfat milk: for MAP1B and GFAP, a goat anti-rabbit IgG (Jackson Immunoresearch Labs, Inc., West Grove, PA.) diluted 1:75; for ED-1 and OX-42, a goat anti-mouse IgG (Jackson Immunoresearch Labs, Inc.) diluted 1:50. Specific staining was visualized by incubating the tissue with 3,38-diaminobenzidine tetrahydrochloride (DAB; Sigma Chemical Co., St. Louis, MO) in distilled water. The sections were then air dried, dehydrated, and coverslipped with DPX mounting media and observed using bright-field microscopy. Axon Tracing Studies Two or 4 months after transplantation, retrograde transport of FG was used to study the projections from transplant neurons into caudal host spinal cord, and anterograde transport of wheat germ agglutininconjugated horseradish peroxidase (WGA–HRP) was used to identify projections between RN neurons and transplants. FG injection. Rats used for this study (N 5 6) were not prelabeled by FG injections prior to transplanta- 48 MORI ET AL. FIG. 1. Nissl–Myelin-stained cross sections show (A) that the lesion completely severed the lateral funiculus, where the axons of the rubrospinal tract course, and (B) that the transplants (TP) were well integrated with host spinal cord (H). Bar, 500 µm. tion. They were anesthetized as described above and their spinal cord exposed by laminectomy two segments caudal to the transplants. A 2% FG solution (1 µl) was pressure injected into the host spinal cord bilaterally as described above. The dura was closed with interrupted 10-O sutures and covered with Durafilm, and muscle and skin were closed in layers. Three days later the rats were perfused as described. A spinal cord block containing the transplants and two segments of the adjacent rostral and two segments of the adjacent caudal spinal 49 TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS cord including the injection site were removed and immersed in an ascending series of 10–30% sucrose dissolved in 0.1 M phosphate buffer. Frozen tissues were cut at 20 µm on a cryostat. Every third section was collected in consecutive order; additional sections were stained for Nissl–Myelin. Sections for FG study were processed as described above. WGA–HRP injection. Two or 3 months after transplantation, the rats (N 5 8) were positioned on a stereotaxic apparatus and a burr hole made by a dental drill according to the coordinates from a rat brain atlas (48). The coordinates for the injection were 5.9 mm caudal from the bregma, 0.7 mm from the midline, and 7.0 mm from the dura matter of the brain. A 1% WGA–HRP solution (Sigma Chemical Co.; 0.03–0.1 µl) was pressure injected into the right magnocellular RN neurons through a micropipette (tip diameter 30–40 µm) attached to a 1-µl Hamilton syringe. Seventy-two hours later the rats were anesthetized with an overdose of sodium pentobarbital and perfused with 150 ml heparinized saline followed by 500 ml of fixative (1.25% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4). Brain stem and spinal cord segments were removed and immersed in an ascending series of 10–30% sucrose in 0.1 M phosphate buffer. The brain stem, including the RN, was cut coronally at 30 µm; the spinal cord was cut sagittally at 30 µm. Midbrain and spinal cord sections were reacted immediately according to the tetramethylbenzidine procedure described by (41) and counterstained with 1% neutral red. RESULTS Spinal Cord Hemisection The spinal cord lesion severed the left lateral funiculus at the C3–4 segments (Fig. 1A). Because rubrospinal tract axons are almost entirely crossed at this level (reviewed in 58), the lesion axotomized neurons of the RN on the right side of the midbrain and spared the left RN. In all animals included in the quantitative study (N 5 12), the lesion was limited to the left side of the spinal cord and completely severed the dorsal, lateral, and ventral funiculi and destroyed the gray matter on the left side. Transplant Morphology Twenty-six of 30 recipients contained surviving transplants, and prominent cysts had not formed at the rostral interface between host spinal cord and transplant. Transplants contained many mature neurons and a neuropil that included both myelinated and unmyelinated axons and resembled other intraspinal embryonic spinal cord transplants that have been described (12, 27, 53). The extent of gliosis at the transplant–spinal cord interface varied among animals, but we included all animals in our quantitative analyses and axon tracing studies except those with cysts (Fig. 1B). Effects of Axotomy and Transplants on RN Neuron Survival and Size (Table 1) In unoperated control rats the magnocellular portion of the right and left RN contained similar numbers of neurons, which were almost equally distributed into small, medium and large classes; about 2% of cells were giants (Figs. 2A, 3A). Neuron numbers in the left RN of normal did not differ from those in the left RN of operated rats 2 or 4 months after C3–4 left-sided hemisection alone or hemisection plus transplant. Neuron cell size distribution and the mean soma area of left RN in normals were also similar to those of experimental animals (Figs. 3B, 3C). Two and 4 months after hemisection alone (N 5 8), the number of neurons in the magnocellular portion of the axotomized RN had decreased by 38% (Figs. 2B, 4B). The cell size distribution was markedly skewed TABLE 1 Mean Cell Size and Cell Number in RN (Nissl–Myelin) Mean cell size (µm2) Group Left Right Total cell number Ratio (R/L) Left Right Ratio (R/L) Control, mean 6 SD (N 5 8) 463.830 6 39.920 465.070 6 47.278 1.003 6 0.059 1296.847 6 187.519 1287.358 6 183.881 0.997 6 0.089 Hemisection, mean 6 SD (N 5 8) 462.213 6 22.096 291.236 6 16.195* 0.630 6 0.018 1340.331 6 152.103 833.650* 6 108.425 0.621 6 0.025 Transplant, mean 6 SD (N 5 8) 433.201 6 34.609 283.992 6 25.576* 0.657 6 0.045 1327.518 6 185.824 1042.386 6 150.884* 0.785 6 0.032 Significant differences among groups** Control . Hemisection, Transplant Control . Transplant . Hemisection * Right differs from left (P , 0.05) using paired comparison t test. ** Overall significant difference among group ratios was determined by Kruskal–Wallis one-way ANOVA (P , 0.05). Individual post hoc comparisons made with the Mann–Whitney test. 50 MORI ET AL. FIG. 2. Transverse section of RN about 400 µm rostral to the caudal pole. (A) Normal. (B) Left-sided hemisection. (C) Left-sided hemisection with transplant. Note cell loss after hemisection (B) and partial preservation of cells in the presence of a transplant. Nissl–Myelin stain. Bar, 200 µm. TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS 51 toward the small cell size, and giant cells were completely absent (Fig. 3B). The mean soma area of surviving neurons was 37% smaller than that of neurons in the right RN (Fig. 4A). In animals with hemisection and transplant (N 5 8), about 80% of the normal number of neurons remained in the magnocellular portion of the axotomized RN (Figs. 2C, 4B). Transplants therefore rescued about half of the RN neurons that would have died without a transplant. As in the axotomized RN after hemisection FIG. 4. Bar graphs representing mean 6 SD of (A) cell size and (B) cell survival in RN as represented by the ratio of the neurons on the lesion side (R)/neurons on the control side (L). (A) Mean cell size decreased similarly in RN contralateral to the hemisection and to the transplant. (B) 38% of cells died contralateral to the hemisection, but only 20% died in the presence of the transplant. *, ,Nor. (normal); **, ,Tp. ,Nor. Data obtained from Nissl–Myelin-stained material. alone, however, the neuron size distribution was shifted toward the small cell size; the number of small neurons increased to twice that of normal, and giant cells were absent. The mean cell soma area was reduced to an extent (35%) similar to that in the axotomized RN following hemisection alone (Fig. 4A). Survival of Rubrospinal Neurons (Table 2) FIG. 3. Soma size distribution in RN in (A) normal, (B) after left-sided hemisection, and (C) after left-sided hemisection plus transplant. Cell size distribution is symmetrical in normal animals (A). The distribution was markedly skewed to the small cell size after hemisection (B) or after hemisection plus transplant (C). The change in size distribution was similar in the two experimental groups. Nissl–Myelin stain. We first examined the number and sizes of magnocellular RN neurons projecting to the spinal cord by injecting FG into the C6 spinal cord of normal rats (N 5 4). Two months after injection, 66% of RN neurons were FG labeled. There were no differences in cell numbers or mean cell size between the right and left RN. The mean size of FG-labeled neurons was 491 µm2. 52 MORI ET AL. TABLE 2 Mean Cell Size and Cell Number in RN (Fluorogold) Mean cell size (µm2) Group Left Right Total cell number Ratio (R/L) Left Right Ratio (R/L) Control, mean 6 SD (N 5 4) 493.562 6 20.434 487.687 6 11.585 0.989 6 0.040 804.562 6 99.690 825.562 6 94.191 1.032 6 0.103 Hemisection, mean 6 SD (N 5 4) 504.930 6 33.786 344.592 6 34.484* 0.683 6 0.060 769.735 6 75.736 500.800 6 108.425* 0.654 6 0.043 Transplant, mean 6 SD (N 5 4) 483.967 6 25.681 327.333 6 28.297* 0.676 6 0.041 795.973 6 26.859 669.223 6 47.219* 0.836 6 0.033 Significant differences among groups** Control . Hemisection, Transplant Control . Transplant . Hemisection * Right differs from left (P , 0.05) using paired comparison t test. ** Overall significant difference among group ratios was determined by Kruskal–Wallis one-way ANOVA (P , 0.05). Individual post hoc comparisons made with the Mann–Whitney test. This was 10% larger than the mean cell size determined from adjacent sections stained for Nissl–Myelin. Small neurons represent about 29% of neurons in the magnocellular portion of RN; 35% of these were FG labeled. Medium-sized neurons constitute 37% of the neurons in the normal magnocellular RN; 84% were labeled by FG. The large cells are 31% of the normal neuron population; 94% were FG labeled. Giant cells make up ,3% of the normal neurons; 65% were FG labeled (Figs. 5A, 6A). Magnocellular RN neurons of all sizes therefore send axons to the spinal cord, but even some of the largest neurons do not. Labeling magnocellular RN neurons with FG prior to cervical hemisection and transplant enabled us to compare the number of neurons projecting to the spinal cord that survived axotomy with and without a transplant (28). Control RN of experimental animals had neuron numbers and neuron size distributions similar to those of unoperated animals. Two months after hemisection alone (N 5 4), the axotomized RN contained 65% of the number of FG-labeled neurons present in the control RN (Fig. 5B). The mean size of surviving axotomized FG-labeled neurons was 345 µm2; this was 32% smaller than in the control RN (Fig. 6B). Following transplantation of fetal spinal cord into the hemisection cavity (N 5 4), the magnocellular portion of the control RN had neuron numbers and neuron size distributions similar to those of unoperated control animals. The magnocellular portion of the axotomized RN contained about 84% of the neurons present in the control RN (Fig. 5C). The mean size of labeled neurons in the axotomized RN was 327 µm2; this was 32% smaller than that in the control RN (Fig. 6C). These results indicate that transplants rescue about half of the magnocellular RN neurons projecting to the spinal cord that were otherwise destined to die after axotomy, but that transplants do not prevent atrophy of these cells. Immunocytochemistry To confirm that the cells which we counted were neurons (27), we stained sections of the RN with four immunocytochemical markers that are specific for different cell types. Rats that survived for 2 months after cervical hemisection (N 5 4) or hemisection plus transplant (N 5 4) were studied. MAP1B (Fig. 7B). MAP1B was seen in the cell bodies of larger RN neurons of normal animals. The axotomized RN had fewer immunopositive cells in both the hemisection alone group and the hemisection plus transplant group. Cells in RN which were FG positive were also MAP1B positive, identifying them as neurons. ED-1. ED-1-positive cells were present next to blood vessels (43, 44). The FG-positive cells were ED-1 negative, indicating that they were not macrophages or monocytes. OX-42. OX-42-positive cells were present in the axotomized and control RN in both hemisection alone and hemisection plus transplant groups. The FGpositive cells were, however, OX-42 negative, indicating that they were not microglia. GFAP. GFAP was localized to the cell bodies and processes of astrocytes in both axotomized and control RN of all animals. FG-positive cells within the RN were GFAP negative, indicating they were not astrocytes. The results of these FG and immunocytochemical studies therefore support the identification of the cells we counted in the RN as neurons. Axon Tracing Studies Projections from transplant neurons into caudal host spinal cord. In preliminary experiments FG injected into segments adjacent to the transplant labeled large numbers of host and transplant cells. To rule out TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS 53 FIG. 5. Fluorescence photographs of FG labeling (A) in normal RN, (B) after hemisection, and (C) after hemisection plus transplant about 400 µm rostral to the caudal pole. Note decreased number of labeled cells contralateral to the hemisection but sparing of cells contralateral to the transplant. Bar, 200 µm. spurious labeling through diffusion, further studies utilized FG injections into host spinal cord two segments caudal to the transplant. These injections labeled by retrograde transport the perikarya of many host neurons adjacent to the injection site. Fewer labeled neurons were present in the transplants, and the interface between host gray matter and transplants was apparent because of a sharp decrease in the number of labeled neurons between host spinal cord and transplants (Figs. 8A, 8B). All of the transplants that we studied (N 5 6) contained small numbers of retrogradely labeled cells, which in some transplants 54 MORI ET AL. spread into surrounding white matter were not further studied even if the RN was well labeled. In all of the rats that received WGA–HRP and in which reaction product was adequately seen (N 5 7/8), labeled axons and collaterals with varicosities descended in the lateral funiculus of the host spinal cord contralateral to the site of WGA–HRP application. Some of these axons penetrated into the glial scar formed between the host spinal cord and the transplant but did not extend into the transplant or caudal host spinal cord (Fig. 10). DISCUSSION The results of these experiments show that cervical hemisection of the adult rat spinal cord causes about 40% of the neurons present in the magnocellular portion of the contralateral RN to die and many of the surviving neurons to atrophy; transplants of fetal spinal cord acutely placed into the hemisection site rescue about half of these neurons from retrograde cell death but do not prevent atrophy of the survivors; in the presence of a transplant, axotomized rubrospinal axons reach the level of the transplant, and, if rescued RN neurons establish connections with neurons in the caudal host spinal cord, these must occur via relays across the transplants. Axotomy Induces RN Neuron Death FIG. 6. Soma size distribution of FG-labeled neurons in RN. (A) In normal animals, all size classes of RN neurons project to cervical cord. In (B) hemisected alone and (C) hemisected plus transplants RN, the distribution was markedly shifted to the smaller sizes. The results from FG labeling show the same pattern of atrophy as the results from Nissl–Myelin preparations (Fig. 3). formed clusters of labeled neurons. These were distributed throughout the transplants and were not preferentially located in the caudal regions closest to the site of labeling. In no case were labeled cells observed rostral to a transplant. Projections from RN neurons to transplant. Injection of WGA–HRP into the rostral half of the right side of the midbrain formed an injection site 500–700 µm in length, which was largely centered on the magnocellular portion of the RN (Fig. 9). Cases in which the tracer The present study has demonstrated a ,40% loss of neurons in the magnocellular portion of the contralateral RN following hemisection of the cervical spinal cord of adult rats. These counts, as well as those showing transplant-mediated rescue, were obtained from Nissl-stained material and confirmed by counts of RN neurons labeled by retrograde transport of FG prior to spinal cord injury. Counts of FG-labeled neurons showed ,35% cell loss after cervical hemisection. The cell loss determined by these two methods cannot be directly compared since FG was applied at the C6 vertebral level, between two and three segments caudal to the site of the spinal cord hemisection. Some axotomized neurons therefore were unlabeled by FG because they terminated rostral to the site at which label was applied (14, 65). We also cannot be certain about the proportion of axotomized RN neurons that die; neurons in the magnocellular portion of the RN that send axons to spinal cord segments rostral to C3–4 escaped injury, as did RN neurons that project elsewhere (reviewed in 58). Our immunocytochemical studies established that FG-containing cells surviving in the RN were neurons by showing them to be immunoreactive for MAP1B, a neuron-specific marker (61), but not for ED-1, OX-42, or GFAP, which identify macrophages, monocytes, microglia, and astrocytes. Our conclusion that a portion of axotomized RN TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS 55 FIG. 7. Double staining of sections labeled with (A) FG and (B) an antibody to MAP1B to identify cells in RN contralateral to a hemisection with transplant. (A) FG-labeled cells (arrows) are (B) MAP1B positive (arrows). A is a fluorescence photomicrograph. B is a brightfield photomicrograph. The DAB reaction product obscures cytoplasmic FG labeling of MAP1B-positive neurons, but FG labeling remains visible within the nuclei. Bar, 200 µm. neurons dies agrees with several previous reports of axotomy-induced retrograde cell loss in the RN after spinal cord injury both in neonates (12, 50) and in adults (20, 22, 60, 72). We cannot, however, exclude the possibility that our cell counts overestimate the extent of neuron death if some surviving neurons have atrophied below our level of detectability (68, 69) and do not produce detectable levels of MAP1B. The cell loss that we observed after hemisection of the cervical cord exceeds that observed in adult rats after hemisection of the thoracic cord (22, 50). This would be consistent with greater vulnerability to axotomy close to the perikaryon (2, 21, 51), but is also due to axotomy of greater numbers of RN neurons at the cervical than at the thoracic level (69). Since RN neurons are also lost when the spinal cord is compressed severely rather than cut (71), these studies emphasize that neuron death is one factor likely to limit recovery after both penetrating and nonpenetrating spinal cord injury. Our results do not allow us to distinguish whether cells from all size categories of RN neurons were lost or only a subset. We observed a 37% decrease in the mean 56 MORI ET AL. FIG. 8. (A) Fluorescence micrograph showing retrogradely labeled neurons within the host spinal cord (H) and transplant (TP). Arrowheads indicate host–transplant interface. FG was injected into host spinal cord two segments caudal to the transplant. Bar, 200 µm. (B) Four of the FG-labeled neurons in the transplant are shown at higher magnification. Bar, 300 µm. (A) and (B) are from sagittal sections. cell size of RN neurons that survived axotomy for 2 to 4 months, but this shift could reflect a selective loss of large and giant neurons, atrophy of these neurons or both. Previous reports have demonstrated atrophy of RN neurons after axotomy in both neonates and adults (12, 50, 69, 72). Transplants Rescue Some Axotomized Neurons Transplants of embryonic spinal cord kept alive about half of the axotomized RN neurons that would have died in the absence of a transplant. Survival appears to have been permanent; neurons were maintained for at least 2–4 months, the longest survival periods studied, and their numbers did not decrease further between 2 and 4 months. The mechanisms that account for permanent rescue from axotomy-induced retrograde cell loss, however, remain unknown. One possibility is that transplants serve as a surrogate source of neurotrophic factors that normally derive from the targets of the rescued neurons and can be taken up by the cut axons. Neurotrophic factors administered exogenously have prevented the death of immature axotomized motoneurons (63, 64, 73) and of several different types of axotomized neurons in adult brain (1, 23, 25, 33, 34, 46, 69). Consistent with this idea is the finding that transplants of embryonic CNS tissues that are normal targets of RN neurons ensured long-term survival of virtually all rubrospinal neurons injured in newborn rats, but CNS tissues that are not normal targets or sciatic nerve grafts produced only temporary rescue (9, 12). The available evidence indicates that brain-derived neurotrophic factor (BDNF) and/or neurotrophin-3 (NT-3) are among the neurotrophic factors that may be responsible for the transplant-mediated rescue of axotomized RN neurons. These are members of a family of neurotrophins whose structure closely resembles that of nerve growth factor (NGF) and which act on neurons primarily through specific protein tyrosine kinase (trk) receptors located on the cell surface (reviewed in 3). NGF binds with greatest affinity to trkA, BDNF to trkB, and NT-3 to trkC (reviewed in 3). RN neurons express mRNA for trkB (40, 69, 70, reviewed in 3) and trkC (69, 70), but not trkA (40, 69, 70; reviewed in 3), suggesting responsiveness to BDNF and NT-3, but not NGF. E14 spinal cord, the age of our transplants, expresses high levels of mRNA for NT-3 and considerably lower levels of mRNA for NGF and BDNF (37). Neither the presence nor the time course of expression of the mRNA for neurotrophic factors nor the factors themselves have been examined in transplants of embryonic spinal cord. The notion that transplantmediated rescue of axotomized RN neurons depends at least in part on BDNF, however, receives support from the finding that administration by gelfoam pledget at the site of spinal cord hemisection in newborn rats produces partial, but permanent rescue of axotomized RN neurons from retrograde cell death (16). The superiority of transplants in rescuing axotomized RN neurons in newborns suggests that transplants provide a single neurotrophic factor in larger amounts or for a longer time than does a gelfoam pledget or that survival depends on multiple factors or additional conditions which are satisfied by a transplant. The contrast between transplant-mediated survival of rubrospinal neurons that is virtually complete in newborns and only partial in adults, despite the generally greater vulnerability of newborn neurons to axotomy, is part of the evidence suggesting that transplants in some cases also contribute to permanent TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS 57 FIG. 9. Injection site of WGA–HRP into the rostral half of the right side of the midbrain which was largely centered on the magnocellular portion of the RN (arrowheads). No labeled neurons are seen in the contralateral RN. Bar, 200 µm. survival of injured neurons by providing a matrix that allows continued axon growth and synapse formation with target neurons (9, 12, 38). In the newborn rat, regenerating or late-developing axons of corticospinal and bulbospinal neurons, including rubrospinal, serotonergic, and other brain-stem neurons, grow through fetal CNS transplants and into host spinal cord many segments caudal to the transplant (6, 8, 10). In the adult host spinal cord, however, most of the cell bodies of neurons that send axons into transplants are located within a few millimeters of the transplants, and even these axons have not been shown to grow through the transplants and into caudal spinal cord (32). In the present investigation anterograde transport of WGA– HRP showed many axons of surviving RN neurons reaching as far as the glial scar at the transplant interface, rare axons penetrating the transplant, and none extending into caudal host spinal cord. This contrasts with the results found with peripheral nerve grafts placed into the cervical cord, which elicited or supported regeneration by a few rubrospinal neurons (55, 69). The explanation for this difference between embryonic spinal cord transplants and peripheral nerve grafts is unknown but is likely to be due to different molecular environments. Failure to regenerate and establish synaptic connections with target neurons in caudal spinal cord may nevertheless contribute to the inability of transplants to produce long-term survival of all axotomized RN neurons in adults. Transplants of embryonic spinal cord, cerebellum, and neocortex secured the permanent survival of virtually all of the 35–40% of Clarke’s nucleus neurons at the L1 segment that normally die after hemisection of the adult or newborn spinal cord at T8 (27). The requirements for transplant-mediated cell rescue appear to differ not only for newborn and adult neurons in RN, but also among different populations of neurons in RN and Clarke’s nucleus. The features of the neurons that account for these differences remain to be identified. The present investigation found that transplants did not prevent perikaryal atrophy of adult RN neurons that survived axotomy. Transplants also did not prevent atrophy of Clarke’s nucleus neurons that survived ipsilateral hemisection of the spinal cord in adult or newborn rats (27). In contrast, BDNF applied adjacent to the RN has been reported to prevent perikaryal atrophy of RN neurons after axotomy in adult rats (69, 70), and transplants of embryonic spinal cord into the site of spinal cord hemisection prevent shrinkage of surviving RN neurons in newborns (12). Neuron survival may therefore require smaller quantities of neurotrophic factors than are required to maintain perikaryal size. In the absence of continued growth and contact with target neurons, as observed with severed rubrospinal axons and transplants in newborns, transplants may fail to prevent or reverse perikaryal atrophy because they provide insufficient quantities of neurotrophic factors to replace those supplied from the normal targets. Transplant Neurons Send Projections into Host Spinal Cord The results of the present investigation showed that injection of FG two segments caudal to the transplants labeled donor neurons by retrograde transport. Labeled 58 MORI ET AL. efferent axons presumably entered host spinal cord by penetrating regions of the host–graft interface where they were not impeded by a glial scar. Outgrowth by donor axons into host spinal cord has been reported previously (32, 53). Our tracing studies of surviving rubrospinal axons with anterograde labeling of RN neurons revealed only an occasional labeled axon within the transplants and none caudal to the transplants, but many labeled axons terminated in host spinal cord just rostral to the transplants. It is therefore possible that neuronal relays were formed across regions of damaged spinal cord via local axonal connections between host spinal cord and transplants. Further studies will be necessary to test whether such novel circuits were established and their functional significance. Intraspinal embryonic spinal cord transplants have been shown to promote limited locomotor recovery after several types of spinal cord injury in adults (45; reviewed in 7, 54). The present study has demonstrated that partial rescue of axotomized supraspinal neurons is one mechanism that may contribute to this recovery; the establishment of functional relays across transplants represents a second possible mechanism. ACKNOWLEDGMENTS We are grateful to Brian Tryon for his contributions and to Theresa Connors, Kathleen Bozek, Michael Spector, and Jean-Manuel Nothias for their technical assistance. The study was supported by NIH Grant NS 24707 and the VA Medical Research Service. REFERENCES 1. 2. 3. 4. 5. 6. FIG. 10. Three semiserial sections in the sagittal plane showing WGA–HRP-labeled rubrospinal (RST) axons reaching the region of the transplant (TP). Some (arrows) enter the glial scar (*), but few if any penetrate the transplant. There is little evidence of retraction of axons severed by the original lesion. Bar, 200 µm. neurons were distributed throughout the transplants, indicating growth or regeneration of their axons into adjacent segments of the host spinal cord. Because we did not inject FG into more caudal segments, the full length of axon regeneration was not determined. These 7. 8. 9. ARENAS, E., AND H. PERSSON. 1994. Neurotrophin-3 prevents the death of adult central noradrenergic neurons in vivo. Nature 367: 368–371. ARMSTRONG, D. M., R. D. TERRY, R. M. DETERESA, G. BRUCE, L. B. HERSH, AND F. H. GAGE. 1987. Response of septal cholinergic neurons to axotomy. J. Comp. Neurol. 264: 421–436. BARBACID, M. 1994. The trk family of neurotrophin receptors. J. Neurobiol. 25(11): 1386–1403. BERNSTEIN-GORAL, H., AND B. S. BREGMAN. 1993. Spinal cord transplants support the regeneration of axotomized neurons after spinal cord lesions at birth: A quantitative double-labeling study. Exp. Neurol. 123: 118–132. BIGNAMI, A., AND D. DAHL. 1976. The astroglial response to stabbing. Neuropathol. Appl. Neurobiol. 2: 99–111. BREGMAN, B. S. 1987. Development of serotonin immunoreactivity in the rat spinal cord and its plasticity after neonatal spinal cord lesions. Dev. Brain Res. 34: 245–263. BREGMAN, B. S. 1994. Recovery of function after spinal cord injury: Transplantation strategies. in Pages 489–592 (S. B. Dunnett and A. Björklund, Eds.), Functional Neural Transplantation. Raven Press, New York. BREGMAN, B. S., AND H. BERNSTEIN-GORAL. 1991. Both regenerating and late-developing pathways contribute to transplantinduced anatomical plasticity after spinal cord lesions at birth. Exp. Neurol. 112: 49–63. BREGMAN, B. S., AND E. KUNKEL-BAGDEN. 1988. Effect of target and non-target transplants on neuronal survival and axonal TRANSPLANTS SAVE AXOTOMIZED ADULT RUBROSPINAL NEURONS 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. elongation after injury to the developing spinal cord. in Pages 205–211 (D. M. Gash and J. R. Sladek, Eds.), Progress in Brain Research Elsevier, New York. BREGMAN, B. S., E. KUNKEL-BAGDEN, M. MCATEE, AND A. O’NEILL. 1989. Extension of the critical period for developmental plasticity of the corticospinal pathway. J. Comp. Neurol. 282: 355–370. BREGMAN, B. S., E. KUNKEL-BAGDEN, P. J. REIER, H. N. DAI, M. MCATEE, AND D. GAO. 1993. Recovery of function after spinal cord injury: Mechanisms underlying transplant-mediated recovery of function differ after spinal cord injury in newborn and adult rats. Exp. Neurol. 123: 3–16. BREGMAN, B. S., AND P. J. REIER. 1986. Neural tissue transplants rescue axotomized rubrospinal cells from retrograde death. J. Comp. Neurol. 244: 86–95. BROWN, L. T. 1974. Rubrospinal projections in the rat. J. Comp. Neurol. 154: 169–188. CABANA, T., AND G. F. MARTIN. 1986. The adult organization and development of the rubrospinal tract. An experimental study using the orthograde transport of WGA–HRP in the North American opossum. Dev. Brain. Res. 30: 1–11. CLARK, G. 1981. Animal Histotechnic. Pages 109–110 and 142– 144 in (G. Clark, Ed.), Staining Procedures, 4th ed. Williams & Wilkins, Baltimore. DIENER, P. S., AND B. S. BREGMAN. 1994. Neurotrophic factors prevent the death of CNS neurons after spinal cord lesions in newborn rats. Neuroreport 5: 1913–1917. DIJKSTRA, C. D., E. A. DOPP, P. JOLING, AND G. KRAAL. 1985. The heterogeneity of mononuclear phagocytes in lymphoid organs: Distinct macrophage subpopulations in the rat recognized by monoclonal antibodies ED1, ED2 and ED3. Immunology 54: 589–599. EAGLESON, K. L., T. J. CUNNINGHAM, AND F. HAUN. 1992. Rescue of both rapidly and slowly degenerating neurons in the dorsal lateral geniculate nucleus of adult rats by a cortically derived neuron survival factor. Exp. Neurol. 116: 156–162. ENG, L. F., J. J. VANDERHEAGEN, A. BIGNAMI, AND B. GERSTL. 1971. An acidic protein isolated from fibrous astrocytes. Brain Res. 28: 351–354. FERINGA, E. R., R. L. MCBRIDE, AND J. N. PRUITT II. 1988. Loss of neurons in the red nucleus after spinal cord transection. Exp. Neurol. 100: 112–120. GAGE, F. H., K. WICTORIN, W. FISCHER, L. R. WILLIAMS, S. VARON, AND A. BJÖRKLUND. 1986. Retrograde cell changes in medial septum and diagonal band following fimbria–fornix transection: Quantitative temporal analysis. Neuroscience 19(1): 241–255. GOSHGARIAN, H. G., J. M. KOISTINEN, AND E. R. SCHMIDT. 1983. Cell death and changes in the retrograde transport of horseradish peroxidase in rubrospinal neurons following spinal cord hemisection in the adult rat. J. Comp. Neurol. 214: 251–257. HAGG, T., D. QUON, J. HIGAKI, AND S. VARON. 1992. Ciliary neurotrophic factor prevents neuronal degeneration and promotes low affinity NGF receptor expression in the adult rat CNS. Neuron 8: 145–158. HAUN, F., AND T. J. CUNNINGHAM. 1984. Cortical transplants reveal CNS trophic interactions in situ. Dev. Brain Res. 15: 290–294. HEFTI, F. 1986. Nerve growth factor promotes survival of septal cholinergic neurons after fimbrial transections. J. Neurosci. 6(8): 2155–2162. HENDRY, I. A. 1976. A method to correct adequately for the change in neuronal size when estimating neuronal numbers after nerve growth factor treatment. J. Neurocytol. 5: 337–349. HIMES, B. T., M. E. GOLDBERGER, AND A. TESSLER. 1994. Grafts of fetal central nervous system tissue rescue axotomized Clarke’s nucleus neurons in adult and neonatal operates. J. Comp. Neurol. 339: 117–131. 59 28. HOUK, J. C. 1991. Red nucleus: Role in motor control. Curr. Opin. Neurobiol. 1: 610–615. 29. HOWLAND, D. R., B. S. BREGMAN, A. TESSLER, AND M. E. GOLDBERGER. 1995. Transplants enhance locomotion in neonatal kittens whose spinal cords are transected: A behavioral and anatomical study. Exp. Neurol. 135: 123–145. 30. HUISMAN, A. M., H. G. J. M. KUYPERS, AND C. A. VERBURGH. 1981. Quantitative differences in collateralization of the descending spinal pathways from red nucleus and other brain stem cell groups in rat as demonstrated with the multiple fluorescent retrograde tracer technique. Brain Res. 209: 271–286. 31. IWASHITA, Y., S. KAWAGUCHI, AND M. MURATA. 1994. Restoration of function by replacement of spinal cord segments in the rat. Nature 367: 167–170. 32. JAKEMAN, L. B., AND P. J. REIER. 1991. Axonal projections between fetal spinal cord transplants and the adult rat spinal cord: A neuroanatomical tracing study of local interactions. J. Comp. Neurol. 307: 311–334. 33. KOLIATSOS, V. E., H. J. W. NAUTA, R. E. CLATTERBUCK, D. M. HOLTZMAN, W. C. MOBLEY, AND D. L. PRICE. 1990. Mouse nerve growth factor prevents degeneration of axotomized basal forebrain cholinergic neurons in the monkey. J. Neurosci. 10(12): 3801–3813. 34. KROMER, L. F. 1987. Nerve growth factor treatment after brain injury prevents neuronal death. Science 235: 214–216. 35. KUNKEL-BAGDEN, E., AND B. S. BREGMAN. 1990. Spinal cord transplants enhance the recovery of locomotor function after spinal cord injury at birth. Exp. Brain Res. 81: 25–34. 36. LAZARIDES, E. 1980. Intermediate filaments as mechanical integrators of cellular space. Nature 283: 249–256. 37. MAISONPIERRE, P. C., L. BELLUSCIO, B. FRIEDMAN, R. F. ALDERSON, S. J. WIEGAND, M. E. FURTH, R. M. LINDSAY, AND G. D. YANCOPOULOS. 1990. NT-3, BDNF, and NGF in the developing rat nervous system: Parallel as well as reciprocal patterns of expression. Neuron 5: 501–509. 38. MANSOUR-ROBAEY, S., D. B. CLARKE, Y.-C. WANG, G. M. BRAY, AND A. J. AGUAYO. 1994. Effects of ocular injury and administration of brain-derived neurotrophic factor on survival and regrowth of the axotomized retinal ganglion cells. Proc. Natl. Acad. Sci. USA 91: 1632–1636. 39. MCBRIDE, R. L., E. R. FERINGA, AND B. E. SMITH. 1988. The fate of prelabeled Clarke’s column neurons after axotomy. Exp. Neurol. 102: 236–243. 40. MERLIO, J.-P., P. ERNFORS, M. JABER, AND H. PERSSON. 1992. Molecular cloning of rat trkC and distribution of cells expressing messenger RNAs for members of the trk family in the rat central nervous system. Neuroscience 51(3): 513–532. 41. MESULAM, M.-M. 1982. Principles of horseradish peroxidase neurohistochemistry and their applications for tracing neural pathways—Axonal transport, enzyme histochemistry and light microscopic analysis. Pages 1–151 in M.-M. Mesulam, Ed., Tracing Neural Connections with Horseradish Peroxidase. Wiley, New York. 42. MILLER, L. E., P. L. E. VAN KAN, T. SINKJÆR, T. ANDERSEN, G. D. HARRIS, AND J. C. HOUK. 1993. Correlation of primate red nucleus discharge with muscle activity during free-form arm movements. J. Physiol. 469: 213–243. 43. MILLIGAN, C. E., T. J. CUNNINGHAM, AND P. LEVITT. 1991. Differential immunochemical markers reveal the normal distribution of brain macrophages and microglia in the developing rat brain. J. Comp. Neurol. 314: 125–135. 44. MILLIGAN, C. E., P. LEVITT, AND T. J. CUNNINGHAM. 1991. Brain macrophages and microglia respond differently to lesions of the developing and adult visual system. J. Comp. Neurol. 314: 136–146. 45. MIYA, D. Y., B. CLARK, A. TESSLER, AND M. MURRAY. 1994. 60 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. MORI ET AL. Transplant mediated behavioral recovery of weight support and interlimb coordination in normal, spinal and spinal 1 transplant rats. Soc. Neurosci. Abstr. 20: 1706. MORSE, J. K., S. J. WIEGAND, K. ANDERSON, Y. YOU, N. CAI, J. CARNAHAN, J. MILLER, P. S. DISTEFANO, C. A. ALTAR, R. M. LINDSAY, AND R. F. ALDERSON. 1993. Brain-derived neurotrophic factor (BDNF) prevents the degeneration of medial septal cholinergic neurons following fimbria transection. J. Neurosci. 13(10): 4146–4156. OORSCHOT, D. E. 1994. Are you using neuronal densities, synaptic densities or neurochemical densities as your definitive data? There is a better way to go. Prog. Neurobiol. 44: 233–247. PAXINOS, G., AND C. WATSON. 1986. The Rat Brain in Stereotaxic Coordinates, 2nd. ed. Academic Press, San Diego. PERRY, V. H., D. A. HUME, AND S. GORDON. 1985. Immunohistochemical localization of macrophages and microglia in the adult and developing mouse brain. Neuroscience 15(2): 313–326. PRENDERGAST, J., AND D. J. STELZNER. 1976. Changes in the magnocellular portion of the red nucleus following thoracic hemisection in the neonatal and adult rat. J. Comp. Neurol. 166: 163–172. PRICE, D. L., V. E. KOLIATSOS, T. O. CRAWFORD, W. C. MOBLEY, AND W. L. PRICE. 1990. Axotomy or treatment with nerve growth factor (NGF) induces the expression of NGF receptor in adult spinal motor neurons: Implications for the pursuit of a motor neuron trophic factor. Soc. Neurosci. Abstr. 16: 478. REID, J. M., D. G. GWYM, AND B. A. FLUMERFELT. 1975. A cytoarchitectonic and Golgi study of the red nucleus in the rat. J. Comp. Neurol. 162: 337–362. REIER, P. J., B. S. BREGMAN, AND J. R. WUJEK. 1986. Intraspinal transplantation of embryonic spinal cord tissue in neonatal and adult rats. J. Comp. Neurol. 247: 275–296. REIER, P. J., B. T. STOKES, F. J. THOMPSON, AND D. A. ANDERSON. 1992. Fetal cell grafts into resection and contusion/compression injuries of the rat and cat spinal cord. Exp. Neurol. 115: 177–188. RICHARDSON, P. M., V. M. K. ISSA, AND A. J. AGUAYO. 1984. Regeneration of long spinal axons in the rat. J. Neurocytol. 13: 165–182. RINAMAN, L., C. E. MILLIGAN, AND P. LEVITT. 1991. Persistence of fluoro-gold following degeneration of labeled motoneurons is due to phagocytosis by microglia and macrophages. Neuroscience 44(3): 765–776. ROBINSON, A. P., T. M. WHITE, AND D. W. MASON. 1986. Macrophage heterogeneity in the rat as determined by two monoclonal antibodies MRC OX-41 and MRC OX-42, the latter recognizing complement receptor type 3. Immunology 57: 239–247. RUIGROK, T. J. H., AND F. CELLA. 1995. Precerebellar Nuclei and Red Nucleus. Pages 277–308 in G. Paxinos, Ed., The Rat Nervous System, 2nd ed. Academic Press, San Diego. SAFAEI, R., AND I. FISCHER. 1989. Cloning of a cDNA encoding 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. MAP1B in rat brain: Regulation of mRNA levels during development. J. Neurochem. 52: 1871–1879. SANNER, C. A., T. J. CUNNINGHAM, AND M. E. GOLDBERGER. 1994. NMDA receptor blockade rescues Clarke’s and red nucleus neurons after spinal hemisection. J. Neurosci. 14(11): 6472– 6480. SATO-YOSHITAKE, R., Y. SHIOMURA, H. MIYASAKA, AND N. HIROKAWA. 1989. Microtubule-associated protein 1B: Molecular structure, localization, and phosphorylation-dependent expression in developing neurons. Neuron 3: 229–238. SCHMUED, L. C., C. BELTRAMINO, AND W. SLIKKER, JR. 1993. Intracranial injection of fluoro-gold results in the degeneration of local but not retrogradely labeled neurons. Brain Res. 626: 71–77. SENDTNER, M., B. HOLTMANN, R. KOLBECK, H. THOENEN, AND Y.-A. BARDE. 1992. Brain-derived neurotrophic factor prevents the death of motoneurons in newborn rats after nerve section. Nature 360: 757–759. SENDTNER, M., G. W. KREUTZBERG, AND H. THOENEN. 1990. Ciliary neurotrophic factor prevents the degeneration of motor neurons after axotomy. Nature 345: 440–441. SHIEH, J. Y., S. K. LEONG, AND W. C. WONG. 1985. Effect of spinal cord hemisection on rubrospinal neurons in the albino rat. J. Anat. 143: 129–141. SMOLEN, A. J., L. L. WRIGHT, AND T. J. CUNNINGHAM. 1983. Neuron numbers in the superior cervical sympathetic ganglion of the rat: A critical comparison of methods for cell counting. J. Neurocytol. 12: 739–750. STOKES, B. T., AND P. J. REIER. 1992. Fetal grafts alter chronic behavioral outcome after contusion damage to the adult rat spinal cord. Exp. Neurol. 116: 1–12. TETZLAFF, W., N. R. KOBAYASHI, AND A. M. BEDARD. 1993. BDNF prevents atrophy of rat rubrospinal neurons after axotomy. Soc. Neurosci. Abstr. 19: 1104. TETZLAFF, W., N. R. KOBAYASHI, K. M. G. GIEHL, B. J. TSUI, S. L. CASSAR, AND A. M. BEDARD. 1994. Response of rubrospinal and corticospinal neurons to injury and neurotrophins. Pages 271– 286 in F. J. Seil, Ed., Progress in Brain Research Elsevier, Amsterdam. TETZLAFF, W., C. A. LEONARD, AND K. C. HARRINGTON. 1992. Expression of neurotrophin receptor mRNAs in axotomized facial and rubrospinal neurons. Soc. Neurosci. Abstr. 18: 1294. THERIAULT, E., AND C. H. TATOR. 1994. Persistence of rubrospinal projections following spinal cord injury in the rat. J. Comp. Neurol. 342: 249–258. XU, X. M., AND G. F. MARTIN. 1990. The response of rubrospinal neurons to axotomy in the adult opossum, Didelphis virginiana. Exp. Neurol. 108: 46–54. YAN, Q., J. ELLIOTT, AND W. D. SNIDER. 1992. Brain-derived neurotrophic factor rescues spinal motor neurons from axotomyinduced cell death. Nature 360: 753–755.