THE JOURNAL OF COMPARATIVE NEUROLOGY 304555-568 (1991)

Plasticity of Spinal Systems After Unilateral Lumbosacral Dorsal Rhizotomy in the Adult Rat SHWUN-DE WANG, MICHAEL E. GOLDBERGER, AND MARION MURRAY Department of Anatomy, Medical College of Pennsylvania, Philadelphia, Pennsylvania 19129

ABSTRACT Plasticity of spinal systems in response to lumbosacral deafferentation has previously been described for the cat, by using immunocytochemistry to demonstrate plasticity of tachykinin systems and degeneration methods to demonstrate plasticity of descending systems. In this study, we describe the response to lumbosacral deafferentation in the adult rat. Application of immunocytochemical methods to visualize tachykinins (predominantly substance P [SP]), serotonin (5-HT),and dopamine B-hydroxylase (DBH),the synthesizing enzyme for norepinephrine, permits us to compare the response of SP systems in rat and cat spinal cord and to examine the response of two descending systems, serotoninergic and noradrenergic, to deafferentation. We used image analysis of light microscopic preparations to quantify the immunoreaction product in the spinal cord in order to estimate the magnitude, time course and localization of changes induced by the lesion. The distribution of SP, serotoninergic (5-HT),and noradrenergic staining in the spinal cord of rat is very similar to that of the cat. Unilateral lumbosacral rhizotomy elicits a partial depletion, followed by a partial replacement of tachykinin immunoreactivity in laminae I and 11. This response was similar to that described for the cat, although characterized by a longer time course, and, as in the cat, is likely due to plasticity of tachykinin containing interneurons. The same lesion elicits no depletion but a marked and permanent increase in 5-HT immunoreactivity in laminae I and 11, which develops more rapidly than the response by the SP system. These results indicate sprouting or increased production of SP and 5-HT in response to deafferentation. No change was seen in DBH immunoreactivity, indicating that the noradrenergic system does not show plasticity in response to defierentation. Our results demonstrate that dorsal rhizotomy evokes different effects in different systems in the adult spinal cord of the rat and thus suggests that the response of undamaged pathways to partial denervation of their target is regulated rather than random. Key words: sprouting,deafferentation,SP, serotonin, norepinephrine

Plasticity in the spinal cord of adult mammals in response to partial denervation has been reported by some investigators (Liu and Chambers, '58; Bernstein and Bernstein, '71; Goldberger and Murray, '74, '78, '82, '85; Hulsebosch and Coggeshall, '81; Murray and Goldberger, '74, '86; McNeill and Hulsebosch, '87; Tessler et al., '80, '81,'84; Polistina et al., '87, '89; Wang et al., '87; Bullitt et al., '88. In other studies, evidence for sprouting was not found (Kerr, '72, '75; Rodin et al., '83; Rodin and Kruger, '84; Micevych et al., '86a)b; Pubols and Bowen, '88). Rodin and colleagues ('83, '84) concluded from their studies that sprouting does not occur in rat spinal cord after lesions comparable to those made in cat, leading them to propose that the reports of sprouting in cat spinal cord were erroneous. It is important to resolve the question of lesion induced plasticity in the spinal cord since it is essential to our understanding of mechanisms that may contribute to O

1991 WILEY-LISS, INC.

recovery of function after spinal cord injury and to an understanding of how neuroanatomically demonstrable reorganization may differ within different regions of the CNS. It is also possible that unexpected species differences could account for the differences in the plasticity described for cat and rat. The use of experimental paradigms in the rat comparable to those used in the cat can help determine whether there are marked species differences in spinal plasticity. At the same time these studies may indicate whether plasticity in the spinal cord is markedly different from plasticity in other regions of the adult CNS, e.g., the limbic system and hypothalamus (Raisman, '69; Lynch et al., '76; Cotman and Nadler, '78; McWillams and Lynch, '79; Murray et al., '79;

Accepted November 5,1990.

556

S.-D. WANG ET AL.

Gage et al., '83; Zhou and Azmitia, '84, '86; Artymyshyn tions supplemented by quantitative image analysis measureand Murray, '85; Hoff, '86; Gasser and Dravid, '87; Battisti ments to describe the time course, magnitude, and location et al., '87, '891, the somatosensory system (Wellsand Tripp, of changes in the density of the immunocytochemical '87a,b), the cerebellum (Chen and Hillman, '821, as well as reaction for SP, 5-HT, and dopamine beta-hydroxylase spinal cord (see also Bjorklund and Wiklund, '80; Decima et (DBH), the synthesizing enzyme for noradrenaline, in the spinal cord of the adult rat after unilateral lumbosacral al., '86; Leedyet al., '87). Plasticity of intrinsic and descending pathways has been deafferentation. described after dorsal rhizotomy in the cat spinal cord using immunocytochemistry for substance P (and other tachykiMATERIALS AND METHODS nins, see Too et al., '89) containing systems (Tessler et al., Surgical procedures '80, '81) and degeneration methods to identify descending pathways (Goldbergerand Murray, '74, '78, '821.' TachykiThirty-two adult female Sprague-Dawley albino rats, nin projections to the dorsal horn originate mainly from weighing 200-250 g, were used in these experiments. Eight dorsal root ganglion cells and this projection is eliminated rats served as unoperated or sham-operated controls. by dorsal rhizotomy. In the cat, there is partial recovery of Twenty-four rats were subjected to complete unilateral SP, mediated by SP containing spinal interneurons (Tessler (right side) lumbosacral rhizotomy (Ll-S2),and sacrificed by et al., '81). This has been considered to be plasticity of a vascular perfusion 4 (5 rats), 10 (5 rats), 15 (3 rats), 30 (5 homotypic nature since the system which responds by rats), or more than 60 days postoperatively (6 rats). These increasing its projections shares some features, in this case times were chosen to be comparable with the time course peptide content, with the system which has been elimi- used to show SP recovery in the cat (Tessler et al., '80, '81, nated. The long-term response of the rat SP system to the '84). same lesion has not been described. Animals were deeply anesthetized by intraperitoneal Two important descending pathways, the serotoninergic injection of chloral hydrate (35 mg/100 g body weight). A and the noradrenergic systems, can be studied separately skin incision was made and the deep fascia dissected to using immunocytochemical methods. Their responses to expose the laminae. The intervertebral foraminae were the same lesion may be similar or different, and this enlarged and the ganglia and dorsal roots were visualized information would be useful in extending our understand- with the aid of a Zeiss surgical microscope. Dorsal root ing of the rules that govern plasticity in the adult nervous section and/or ganglionectomies from L, to L, were persystem. The serotoninergic projection to the spinal cord formed using an extradural approach. The S,, S,,and L, originates mainly from brainstem raphe nuclei and this roots were sectioned inside the vertebral canal. The muscle projection therefore remains intact after dorsal rhizotomy. was sutured and the skin was closed with wound clips. In In the cat, there is an increase in density of 5-HT immuno- most cases recovery was uneventful. reactivity in the spinal cord in response to deafferentation At sacrifice, rats were deeply anesthetized and perfused (Goldberger and Paige, '86). Such an increase suggests intracardially with saline followed by 4% paraformaldehyde heterotypic plasticity, since the system which responds in 0.1 M phosphate buffer, pH 7.4. The spinal cord was (5-HT) is dissimilar to the system which has been removed post-fixed in situ for 4 hours and then removed, placed in (dorsal roots, which do not contain 5-HT, but see Kai-Kai 30% sucrose buffer solution overnight, and divided into and Keen, '85; Kai-Kai, '89), at least with respect to segments. The segments from L, to L, were frozen and transmitter content. Biochemical analyses of spinal gray sectioned on a freezing sliding microtome at 30 ym in the matter after unilateral cervical rhizotomy in rats revealed transverse plan. Every tenth section was collected into test no changes in amount of 5-HT (Hadjiconstantinou et al., tubes containing phosphate buffered saline (PBS). '84), but the response of the rat spinal 5-HT system to Completeness of the rhizotomy was determined 1) by dorsal rhizotomy has not been examined using the anatom- microscopicinspection at surgery and after perfusion; 2) by ical resolution afforded by immunocytochemistry. the absence of behavioral responses to intense noxious The noradrenergic projection to the spinal cord origi- stimulation (pinch) of the dederented hindlimb; and 3) nates primarily from the nuclei of the locus coeruleus group confirmed histologically by loss of staining for thiamine and this projection therefore also remains intact after monophosphatase (see Results). dorsal rhizotomy. There is one report in the rat that Thiamine monophosphatase (TMPase) noradrenaline content, measured with HPLC, decreased on both experimental and control sides after unilateral cervicoSections cut at 30 ym were rinsed in 0.2 M Tris-maleate thoracic rhizotomy (Colado et al., '88). Immunocytochemi- buffer at pH 5.0, incubated in 10 ml of 1.25% thiamine cal studies of the effects of deafferentation on noradrenaline monophosphate chloride (0.125 gl10 ml) (Sigma); 10 ml of containing pathways in the spinal cord have not yet been 0.2 M Tris-maleate buffer at pH 5.0 (4.72 d l 0 0 ml); 20 ml, carried out in either rat or cat. 0.2%lead nitrite and 10 ml of distilled water for 2-3 hours Immunocytochemical staining methods can be used to at 37°C (Knyihar-Csillik et al., '86), rinsed in Tris-maleate compare plasticity in these systems in rat spinal cord after buffer for 10 minutes, and developed for 5 minutes in 1% complete unilateral lumbosacral defierentation. Although aqueous ammonium sulfide (Fisher Co.). Sections were these three systems-the tachykinin, serotoninergic, and then rinsed, dehydrated, and coverslipped. noradrenergic systems-have been shown to be capable of plasticity, the response of each to the same lesion has not Immunocytochemistry been compared. In this report we used qualitative observaSP (tachykinins). The immunocytochemical reaction for tachykinins recognizes both SP and substance K, but since SP is the major tachykinin present in the DRG (Too et 'The spinal tachykinin system will, for convenience, be referred to as SP al., '89) we will refer to tachykinins as SP for convenience. (substance P f although other tachykinins, principally substance K, are also present. This reaction was carried out on free-floating sections using

PLASTICITY IN SPINAL CORD OF ADULT RAT the avidin and biotinylated horseradish peroxidase complex (ABC)method (Hsu et al., '81). The sections were washed in 0.1 M phosphate-buffered saline (PBS), incubated for 30 minutes in PBS containing 3% normal goat serum (GS) and 0.4% Triton X-100 (GS-PBS-T),incubated for 48 hours at 4°C in the SP primary antisera (INCSTAR,Stillwater, MN), diluted 1:5,000, then treated sequentially as follows: two washes in PBS, 5 minutes each; two drops of goat-anti rabbit IgG serum (Vector Lab., CA) in 10 ml GS-PBS-T,60 minutes; two washes in PBS, 10 minutes each; ABC (Vector Lab., CA), two drops of each in 10 ml GS-PBS-T, 90-120 minutes; one wash in PBS, 10 minutes; one wash in 0.05 M Tris buffer (pH 7.6),10 minutes; 0.05% DAB 0.005% H,O,. Following the labelling reaction, the sections were rinsed, mounted, dehydrated, and coverslipped. In order to establish immunostaining specificity,controls consisted of substituting normal rabbit serum or blocked antiserum to SP for SI' antiserum in the staining procedure. Blocked antiserum was produced by adding 25 p,g of SP to 1 ml of diluted SP antiserum. Serotonin immunocytochemistry. The procedure for the immunocytochemical reaction for 5-HT was the same as that for SP except that 5-HT antibody (INCSTAR, Stillwater, MN), diluted 1:10,000-20,000, was used as the primary antiserum. In order to establish immunostaining specificity,controls consisted of substituting normal rabbit serum or blocked antiserum to 5-HT for 5-HT antiserum in the staining procedure. Blocked antiserum was produced by adding 15 pg of 5-HT to 1ml of diluted 5-HT antiserum. DBH immunocytochemistry. DBH, the synthesizing enzyme for noradrenaline, was used to localize noradrenergic projections. The procedure for the immunocytochemical reaction for DBH is the same as that for SP except that the DBH antibody (Eugene Tech., NJ), diluted 1:1,000-2,000, was used as the primary antiserum. In order to establish immunostaining specificity,controls consisted of substituting normal rabbit serum in the staining procedure.

Data analysis In the following measurements, laminae 1-11 on the right and left side were compared on each section and a ratio was generated for that section. Dorsal horn area measurements. In order to determine if significant shrinkage has occurred as a result of dederentation, planimetric measurements were made of cresyl violet stained sections from the L, segment. At least 5 sections, 300 pm apart, were measured using the Bioquant Image Analysis system. A horizontal line perpendicular to the posterior median sulcus was drawn at the level of neck of dorsal horn in all animals. The area of the dorsal horn down to the horizontal line on the right and left sides of the spinal cord was traced, measured, and compared. These area measurements were supplemented by measurements of the area of lamina I and I1 on cresyl violet stained sections in which cytoarchitectural criteria were used to identify the laminar boundaries. Quantitative analysis of immunocytochemistry. The area fraction occupied by SP, 5-HT, and DBH immunocytochemical reaction product were measured using the Bioquant image analysis system. Area fraction is that fraction within a measured area that is occupied by a particular structure. In these studies, the measured area comprises lamina I and 11, and the structures studied within that area are immunolabeled s o n s and varicosities. A section was visualized on the monitor screen of the Bioquant system,

557

lamina I and I1 were outlined on the screen. A detection level representing background staining for that section was set, and the area occupied by the label which exceeded the detection level was computed on both right and left side. A ratio of area occupied by staining above the detection level on experimental (right) to control (left) side was generated for each section. An increase in area fraction therefore reflects an increase in staining within an area but not an increase in the area containing the stained structure. Five to seven sections per segment were chosen for quantitative analysis from L, to L, in unoperated and experimental animals. Means of these ratios were determined for SP, 5HT, and DBH immunoreactivity at each time point. Statistical analysis. The Kruskal-Wallis one way ANOVA (P< 0.01, two-tailed test) was used to analyze area measurements and density of immunoreactivity. Individual post-hoccomparisonswere made using the WilcoxonMann-Whitney test corrected for multiple comparisons (Kirk, '68).The statistical analysis was performed with the aid of the Number Cruncher Statistical System (Dr. Jerry L. Hintze, Kaysville Utah) program on a IBM personal computer.

RESULTS Qualitative changes in staining Thiamine monophosphatase (TMPase). TMPase reaction product is distributed across inner lamina TI (IIi), where it appears as intensely staining coarse, clumped varicosities. The distribution of reaction product is comparable on the left and right sides in control animals. TMPase reaction product is absent from other areas of the spinal cord (Fig. la). This distribution is similar to that reported by Knyihar-Csilliket al. ('86). Following unilateral lumbosacral dorsal root section and/or ganglionectomies, TMPase activity in the ipsilateral dorsal horn was completely abolished by 4 days (Fig. lb) and did not recover as long as 8 months postoperatively (Fig. lc). The TMPase enzyme reaction on the contralateral side was unaffected by the rhizotomy. These results indicate that all TMPase is supplied to the dorsal horn by the dorsal roots and that there was no lesion induced plasticity of this system at the levels which we examined. The elimination of the TMPase reaction product could thus be used as evidence of completeness of the rhizotomy. Distribution of SP immunoreactivity. In normal adult rats, SP immunocytochemical reaction product is present in all laminae of the spinal grey matter and also in the dorsal lateral funiculus of the white matter. The densest reaction product is found in lamina I and outer lamina I1 (1101, where the reaction product appears as irregular, coarse, clumped granules (Fig. 2a,b). In lamina V, the reaction product forms a reticulated plexus; the staining of this plexus shows considerable variability from section to section within normal animals. Fine, more regular punctate granules appear at the medial border of dorsal horn, extending ventrally to lamina X and around the central canal. In the ventral horn, fine granular reaction product is distributed throughout laminae VII and IX. In the white matter, dense varicosities and plexus-like SP reaction product are present in the dorsal lateral funiculus. The normal distribution of SP in the rat lumbar cord is thus virtually identical to the distribution in the cat lumbar cord (Tessler et al., '80,'81).

S.-D. WANG ET AL.

558

R \

R

b

R

C Fig. 1. Photomicrogal j showing TMPase in dorsal horn c normal adult rats and following right-sided dorsal rhizotomy. a: Normal L, dorsal horn. b: Dorsal horn 4 days postoperative. c: Dorsal horn 247 days postoperative. Note symmetrical staining in superficial dorsal horn of normal spinal cord and complete and permanent loss postoperatively. Calibration bar = 300 pm.

Following unilateral lumbosacral rhizotomy, stainable SP reaction product within the ipsilateral dorsal horn was reduced in comparison to the control side. By 4 days after rhizotomy, the SP-like reaction product density was noticeably decreased in lamina I, lamina 110, and lamina V (Fig. 2c). Ten days postoperatively, the density of SP reaction product was maximally decreased in lamina I, lamina 110, and lamina V, but not completely abolished (Figs. 2d, 3b). The distribution pattern of SP reaction product remained normal, but the irregular, coarse granules were almost entirely absent and only fine punctate staining could be seen. In the other laminae and the dorsal lateral funiculus, SP reaction density did not change. At longer survival times, the SP staining density began to recover. The amount of SP reaction product in the dorsal horn of 15 day survivors appeared to be increased over that seen in 10 day survivors. The SP staining density in the dorsal horn ipsilateral to the rhizotomy gradually increased

for 2 months, but appeared to stabilize after that time (Figs. 2e,f, 3d). The SP reaction product in lamina I and lamina 110in deafferented animals differed in appearance from that of normal animals. The dorsal horn on the intact side of deafferented and in normal animals contained densely staining irregular, coarse, clumped granules as well as finer granules. After deafferentation, only the finer, more regular and less densely staining granules remain (Fig. 3b) and staining of this type increased after 10 days postoperatively (Fig. 3d). The coarse granular reaction product thus appears to be supplied by dorsal roots and to be permanently eliminated by rhizotomy. Fine reaction product survives rhizotomy and increases during recovery. The response to dorsal rhizotomy in the rat lumbar spinal cord therefore appears to be virtually identical to the response seen in the cat lumbar spinal cord (Tessler et al., '80, '81; Micevych et al., '86a,b). The intact dorsal roots (Fig. 20 show virtually no staining despite the fact that SP containing axons course within the roots to their termination in the dorsal horn. The central stumps of the sectioned dorsal roots also show no punctate or granular staining or staining of varicosities. Nevertheless the degenerated central stumps are markedly shrunken and, because of gliosis, have become densely cellular and appear dense in this preparation (Fig. 2F). The density of degenerated central stumps is not accompanied by greater background staining within the spinal gray matter (compare Fig. 3a with 3b and 3c with 3d), perhaps because shrinkage and gliosis are much less marked in the dorsal horn than in the distal stump of the cut dorsal root. Distribution of serotonin (5-HT) immunoreactivity. In normal adult rats, 5-HT immunocytochemical reaction product is distributed to all laminae of the spinal gray matter (Fig. 4). The reaction product is most densely distributed in lamina I and lamina 110 of the dorsal horn, the intermediolateral column, lamina X surrounding the central canal, and the ventral horn. Serotonin immunoreactivity in laminae I and 110 of the dorsal horn is present as isolated, punctate profiles with a few short strands of 5-HT containing axons with varicosities and thus differs from the cat. In some sections, there are a few longer 5-HT immunoreactive axons traveling along the dorsal border of lamina I. Lamina IG contains less immunoreactive staining product than any other lamina of the gray matter. In general, 5-HT immunoreaction product in rat lumbar spinal cord is similar in distribution, although it appears to be less dense than that in cat spinal cord (Ruda et al., '82;Goldberger and Paige, '86). By 4 days after unilateral lumbosacral rhizotomy, the 5-HT reaction product appears to be slightly, but insignificantly (see Table I), increased in laminae I and 110 of the ipsilateral dorsal horn compared to the contralateral side. The density of the reaction product appears t o be further elevated in lamina I and lamina 110 at later survival periods (Fig. 5 ) . At all postoperative times, the pattern of 5-HT reaction product within the superficial laminae remains normal, but the area fraction occupied by the staining in laminae I and I1 increases as a result of the increase in clumped punctate 5-HT staining. In other laminae, density of the 5-HT reaction product does not appear to change either ipsilaterally or contralaterally. The rat thus appears to resemble the cat by responding to dorsal rhizotomy by an increase in 5-HT in the dorsal horn (Goldberger and Paige, '86).

PLASTICITY IN SPINAL CORD OF ADULT RAT

-.

.

Ic.’

-.

559

Fig. 2. Photomicrographs showing tachykinin (SP) immunoreactivity in spinal cord of normal adult rats and following right-sided dorsal rhizotomy. a: Normal L,. b Sham operation. c: 4 days postoperative. d 10 days postoperative. e: 30 days postoperative. f: 64 days postoperative. Note partial loss of staining, first apparent at 4 days, and partial

recovery at 30 and 60 days. In f, compare very light staining in intact dorsal root (single arrow, left side) with dense shrunken, gliotic central stump on the experimental side (double arrow, right side). Calibration bar = 300 pm.

Distribution of dopamine B-hydroxglase immunoreactiuity. DBH immunoreactive product is distributed through all laminae of the spinal gray matter. The stained punctate reaction product is intermingled with labeled fine fibers in laminae I and 110of the dorsal horn (Fig. 6). In the

ventral horn, DBH reactive fibers are coarser than those in the dorsal horn. This distribution is similar to the pattern of DBH distribution previously described for the dorsal horn of rat (Westlund et al., ’83).Following acute or chronic deafferentation, both the pattern and density of DBH

S.-D. WANG ET AL.

560

Fig. 3. Higher power photomicrographs of SP immunoreactivity in control and experimental dorsal horns from rats deafferented 10 days and 64 days prior to sacrifice to show depletion and partial recovery of SP immunoreactivityin the dorsal horn. Panel a shows normal levels of staining on the control side and b shows depletion of staining at the

postoperative time during which depletion is maximal. Panel c shows normal levels of staining on the control side and d shows partial recovery of staining on the experimental side 2 months postoperatively. Calibration bar = 100 wm.

immunoreactive staining appear to be unchanged compared to the control side (Fig. 7).

tively. At 10 days, staining was further decreased by 65% compared to normal ( P < 0.01, two-tailed test). At 30 days postoperatively, the SP immunoreactive staining has substantially and significantly increased over the 10 day level ( P < 0.01, two-tailed test) but was still 50% less (P < 0.01, two-tailed test) than that seen on the unoperated side. At survival times greater than 60 days, the area fraction remained significantly increased over that seen at 10 days but reduced by 24%compared to the unoperated side. The quantitative evidence thus supports a loss followed by a partial recovery of SP in response to deafferentation. 5-HT immunoreactiuity. At 4 days postoperatively, a slight, statistically insignificant increase in area fraction occupied by 5-HT reaction product in lamina I and I1 was found. At 10 days postoperative, the area fraction occupied by 5-HT immunoreaction product was increased by 75% ( P < 0.01, two-tailed test) compared to the area fraction in laminae I and I1 on the control side and the mean ratio of experimental to control increased to 210% after 60 days. The increase in 5-HT immunoreactivity greatly exceeds the decrease in area associated with shrinkage and thus appears to represent a real increase in the amount of immunoreactivity in the dorsal horn rather than an artifactual increase secondary to shrinkage.

Quantitative analysis (Table 1, Figs. 8,9> A computer assisted densitometer system was used to quantify and compare area measurements and density of immunoreactivity in laminae I and I1 of ipsilateral and contralateral sides at 4 postoperative periods. Effectof deafferentationon area of laminae I and 11 and of dorsal horn. Two measurements, the area of the dorsal horn and the area of laminae I and 11, were used to determine extent of shrinkage in the dorsal horn following defierentation. The estimates of shrinkage given by the two measurements were comparable (r = .91). Unilateral lumbosacral rhizotomy produces 11-15% shrinkage of laminae I and 11,which is significant ( P < 0.01) by 10 days and which does not change subsequently. The extent of shrinkage in the rat lamina I and I1 is thus comparable to that seen in cat after complete lumbosacral dorsal rhizotomy, although the shrinkage in the cat did not reach statistical significance (Goldberger and Murray, '82). SP immunoreactivity. The area fraction of SP staining in laminae I and I1 was decreased by 24% (P < 0.01, two-tailed test) on the operated side at 4 days postopera-

PLASTICITY IN SPINAL CORD OF ADULT RAT

561

Fig. 4. Photomicrographs showing 5-HT immunoreactivity in dorsal horn of normal adult rat. a: Normal 5-HT distribution. calibration bar = 200 pm. Higher magnification of left (b)and right ( c ) sides from section shown in a. Note similar pattern and density of staining on the right and left sides. Calibration bar = 100 wm.

TABLE 1. QuantitativeAnalysis of SP, 5-HT, and DBH-ImmunoreactivityAfter Unilateral Lumbosacral (L,-S,) Rhizotomy in Adult Rats' Mean ratio of R (deaferentated) to L (control)side Area of lamina

Area fraction

Group

N

I + I1

SP

5-HT

DBH

Normal adult 4 days postoperative 10 days postoperative 30 days postoperative 64-354 days postoperative

5

1.01 t 0.03 0.96 t 0.06 0.86 2 0.04' 0.89 2 0.03* 0.85 2 0.12'

1.01 i 0.02 0.76 f 0.15*,** 0.34f 0.05* 0.51 f 0.15' 0.76 f 0.19*,**

1.03 t 0.18 1.35 t 0.20 1.75 ? 0.43* 1.91 2 0.56* 2.09 t 0.76*

1.06 ? 0.20 1.12 f 0.05 1.12 ? 0.45 1.14 2 0.19 0.97 t 0.16

5 5 5 6

'Values are the Mean 2 S.E.M.

"Significantlydifferent compared to the normal (P< 0.01, two tailed test); **Significantlydifferent compared to the 10 days postoperative (P< 0.01, two tailed test).

DBH irnmunoreactivitg. The area fraction occupied by DBH immunoreactivity in laminae I and I1 increased by a maximum of 14%,a slight and insignificant difference. The slight increase in DBH immunoreactivity is comparable in amount to the shrinkage measured in the dorsal horn and thus appears to be attributable to shrinkage.

DISCUSSION The results of the present study demonstrate that dorsal rhizotomy elicits different responses from different systems

in the mature spinal cord. Histochemical techniques show that TMPase completely disappears following lumbosacral deafferentation (Knyihar-Csillik et al., '86). This result confirms that 1) TMPase is located in dorsal root terminals which degenerate after rhizotomy (Coimbra et al., "74; Knyihar-Csillik et al., '86; Ribeiro-da-Silva et al., '86); 2) there are no TMPase containing dorsal root projections to superficial layers of the lumbar dorsal horn from the contralateral side (Swett and Woolf, '85; Molander, '86); and 3) the completeness of rhizotomy can be determined using this reaction.

S.-D. WANG ET AL.

562

Fig. 5. Photomicrographs showing 5-HT immunoreactivity in dorsal horn of adult rat following right-sided dorsal rhizotomy 30 days postoperatively. a: Low power micrograph. Calibration bar = 200 pm. Higher magnification of left (b)and right (c) sides from section shown in a. Note increased density of stainingon deaferented side (c). Calibration bar = 100 pm.

Lesion induced plasticity of SP containing neuronal systems has been shown in other regions of brain (Artymyshyn and Murray, '85) as well as in spinal cord (Tessler et al., '80, '81; Micevych et al., '86a,b; Vacca-Galloway and Sternberger, '86). Our qualitative and quantitative results using immunocytochemicaltechniques show that following rhizotomy, the density of SP immunoreactivity is dramatically decreased in laminae I and I1 of the dorsal horn but then recovers partially. Our results in the rat are therefore similar to those reported in the cat (Tessler et al., '80; Micevych et al., '86a); the normal distribution of SP and the time course and extent of depletion in response to deafferentation are almost the same in the two species although the recovery time is considerably longer in the rat. The demonstration of recovery of SP in the dorsal horn after deafferentation in the rat, which has not been reported by others (Jessell et al., '79; Hunt et al., '821, was made possible by examining the reaction at longer survival times and by the use of quantitative methods. Both descending and intrinsic SP systems project to the dorsal horn (Tuchscherer et al., '87; Wessendorf and Elde, '87)and either or both could account for the recovery. Some SP staining remains in laminae I and I1 at 10 days, the time

of maximum depletion, consistent with the suggestion that the SP containing axons which are responsible for recovery may be among those fiber systems which are normally located in the dorsal horn. Therefore, the return of SP reaction product appears to follow the general rule that, in the adult, plasticity may increase an already existing projection rather than create an aberrant projection (Raisman, '69; Goldberger and Murray, '78; Tessler et al., '81; Nakamura et al., '84). At the light microscopic level, the SP reaction product differs morphologically at long post-operative survivals from the normal reaction product in that coarse granular staining is absent and all SP reaction product appears as fine punctate staining. Electron microscopic studies (Coimbra et al., '84; Murray and Goldberger, '86) and an EM immunocytochemical study (Ribeiro-da-Silva et al., '89) reveal a class of glomerular or complex terminals, some of which contain SP (Ribeiro-da-Silvaet al., '89), all of which disappear after dorsal rhizotomy (Murray and Goldberger, '86). It seems likely that the loss of these terminals accounts for the loss of coarse granular staining. The difference in the appearance of the reaction product is thus likely to reflect a change in the population of synaptic

PLASTICITY IN SPINAL CORD OF ADULT RAT

563

Fig. 6. Photomicrographs showing DBH immunoreactivity in dorsal horn of normal adult rat. a: Low power micrograph showing normal DBH distribution. Calibration bar = 200 pm. Higher magnification of left (b)and right ( c )sides of section shown in a. Note symmetrical staining. Calibration bar = 100 km.

terminals remaining in the superficial laminae of the dorsal horn and is therefore consistent with replacement by the small terminals found in the EM study (Murray and Goldberger, '86). Results from 5-HT immunocytochemical studies demonstrate an increased density of the descending serotoninergic projection following partial deaerentation in those areas normally receiving the greatest dorsal root input. The increased density of the 5-HT immunoreactive staining was noticeable by 4 days post-operatively before detectable shrinkage occurred. Although lamina I and I1 on the experimental side was on average 12-13% smaller than the control side by 10-354 days postoperative, the area fraction of lamina I and I1 occupied by the 5-HT projection more than doubled. Shrinkage does not therefore account for the dramatic increase in the 5-HT projection. There were no obvious changes in the density of serotonin immunoreactivity in other laminae. The changes within the superficial laminae, while dramatic, are not likely to be of a magnitude adequate to be demonstrable in biochemical analyses from homogenized whole spinal cord, and this probably accounts for the failure of Hadjiconstantinou et al. ('84) to detect differences in 5-HT content after cervical rhizotomy. Plasticity of the 5-HT system has been shown in the spinal cord (Bjorklund and Wiklund, '80; Bregman, '87a,b),

inferior olivary complex (Pare et al., ,871, hypothalamus (Frankfurt and Beaudet, '87), zona incerta (Frankfurt and Beaudet, '881, and hippocampus (Zhou and Azmitia, '84, '86; Gasser and Dravid, '87). In these studies, plasticity was elicited by partial damage to the 5-HT system and was therefore homotypic in nature. In the present study in the adult rat, the 5-HT projection was not directly damaged, and thus the 5-HT staining was not reduced but rather an increase was seen in the superficial dorsal horn only a few days after rhizotomy. Loss of dorsal root afferents thus resulted in heterotypic plasticity mediated by an intact projection. The normal distribution of 5-HT immunoreactive product (Ruda et al., '82) and the response to deafferentation (Goldberger and Paige, '86) seen in the rat are comparable to those described in the cat. Our light microscopic studies have shown that the increase in density of the 5-HT projection and the decrease and recovery in density of the SP projection were most apparent in laminae I and 11. The recovered tachykinin staining suggests a return of tachykinin innervation, likely to be mediated by intrinsic neurons, to the partially denervated neurons of the superficial laminae. Serotoninergic projections modulate the transmission of noxious stimuli (Fields et al., '77) and the projections of the serotoninergic system are closely associated with the small diameter

S.-D. WANG ET AL.

564

Fig. 7. Photomicrographs showing DBH immunoreactivity in dorsal horn of adult rats following right sided rhizotomy 97 days postoperatively. a: Low power micrograph. Calibration bar = 200 pm. Higher magnification of left (b)and right ( c )sides from section shown in a. Note symmetrical staining. Calibration bar = 100 pm.

SP-containing dorsal root axons (Gobel et al., '81; DiGiulio et al., '87). Therefore, the finding that the greatest sprouting of 5-HT projections is in the superficial dorsal horn suggests that the central pathways conveyingpainful sensation may be specifically changed as a result of the response to rhizotomy. Dorsal rhizotomy eliminates perception of noxious and other sensory events, but one of the reported consequences of such lesions in humans is paresthesia. In addition a common complication of dorsal root lesions in rodents is autophagy (Colado et al., '88), which may be stimulated by paresthetic sensations referred to the deafferented limb. Our observation that the dorsal rhizotomy induces changes in both the descending 5-HT projection and the interneuronal tachykinin projection to laminae I and 11, the normal target for the dorsal root s o n s which mediate painful sensation, suggests that the abnormal circuitry provided by sprouting may mediate transmission of aberrant information. Our results using DBH immunostaining show that this projection is not appreciably changed following rhizotomy. The increased DBH immunostaining in laminae I and I1 was not significant and was comparable to the extent of shrinkage of the dorsal horn and is therefore likely due to the shrinkage of these laminae. Plasticity of the noradrenergic system has been observed under other conditions. Noradrenergic axons have been shown to reinnervate tar-

gets after neurotoxin lesions (6-OHDA or 5,7-DHT) which damage the noradrenergic terminals (Bjorklund and Lindvall, '79; Schmidt and Bhatnagar, '79; Levitt and Moore, '80; Gustafson and Moore, '87). Also, homotypic sprouting has been shown in the same system by partial denervation of the target areas (Pickel et al., '73; Reis et al., '78; Gage et al., '83; Sakaguchi et al., '84; Haring and Davis, '85; Gasser and Dravid, '87) and heterotypic sprouting of noradrenergic projections has also been demonstrated in the septal area (Mooreet al., '71) and the interpeduncular nucleus (Battisti et al., '87). We found, however, no evidence of increased noradrenergic innervation following unilateral lumbosacral rhizotomy. Dorsal rhizotomy in the adult rat does not, therefore, provide an adequate stimulus for heterotypic sprouting by the descending noradrenergic system, which may reflect minimal convergence by noradrenergic and dorsal root axons upon dorsal horn neurons. Careful qualitative analysis of the characteristics of specific staining reactions and the extent of tissue shrinkage must be used in the interpretation of the quantitative changes. Immunocytochemical preparations show some variability in intensity of staining from one section to the next and from one staining run to the next but not between the two sides of the spinal cord. The variability makes quantitative assessment of the changes difficult when different animals and different preparations are being com-

565

PLASTICITY IN SPINAL CORD OF ADULT RAT

Ratio Exp/Cont

4-4

c

0

0 1.5 \ Q X

w

"

0

4

10 30'60

DBH

SP

Area Measurement

0

4

10 30'60

0

4

10 30b60

0

4

10 30,60

Days Post Operative Fig. 8. Bar graph representation of the quantitative analysis for the mean (2S.E.M.) area of laminae I and 11, and density of SP, 5-HT, and DBH immunoreactivities expressed as ratio of experimental to control sides. The values shown in this graph are taken from Table 1. Stars indicate values are significantly different from normal values, P < 0.001,2-taiIed Wilcoxon-Mann Whitneytest.

2.5

2

1.5

1

0.5

0 0

4

10

30

Days Post Operative Fig. 9. Line graph representation of the quantitative analysis for the mean area of laminae I and 11, and SP, 5-HT, and DBH immunoreactivities, expressed as ratio of experimental to control sides, which permits comparison of the time course of changes for each of the parameters measured. The values shown in this graph are taken from Table 1.

,60

566

S.-D. WANG ET AL.

pared. The lumbar spinal cord is symmetrical with comparable numbers of dorsal root ganglion cells and axons in the same root on the left and right side (Ygge et al., '81; Arvidsson et al., '86; Molander, '86; Himes and Tessler, '89). Comparisons of staining intensity between left and right sides in control animals yield ratios close to 1.0. The variability of immunocytochemical staining reaction between two sides is thus acceptably low and side to side comparisons therefore permit a more sensitive analysis than interanimal comparisons. Two other factors complicate the evaluation of the quantitative results in lesioned tissue. Shrinkage accompanies lesions and can lead to an incorrect interpretation of increased density of staining if account is not taken of the amount of shrinkage; the assessment of the amount of shrinkage is critically dependent on the ease with which boundaries can be identified. The cytoarchitectonic boundaries of lapinae I and I1 are readily identified (Ralston and Ralston, '79; Snyder, '82; Murray and Goldberger, '86) and in the present study we found a shrinkage in laminae I and I1 of about 15%which stabilized by 10 days postoperatively. The changes in area fraction occupied by SP and 5HT showed greater changes and different time courses and magnitude and therefore the changes in area fraction occupied by these markers in laminae I and I1 cannot be attributed solely to shrinkage. The change in DBH staining is, however, comparable to the change in area of lamina I and 11, and we can therefore conclude that this marker does not show differences resulting from the deafferentation. Another problem in assessing quantitative evaluation of immunocytochemicalstaining is the possibility that background staining is different in lesioned and control material and that background staining rather than specific staining accounts for the differences in staining intensity (Murray et al., '90). In this material, increased density was evident in the shrunken and degenerated dorsal roots but not obvious within the neuropil of laminae I and I1 even at the later survival times. Our results indicate that three different systems exhibit different responses to the same lesion. SP decreases and shows partial recovery, probably as a result of homotypic sprouting or increased production of SP by interneurons. The serotoninergic pathway increases its projection, a response consistent with heterotypic sprouting or increased production of 5-HT by this descending system. The noradrenergic system shows no change in its projections to the spinal cord. Thus our results suggest that the response of intact systems to partial denervation is subject to some regulation, an observation consistent with behavioral recovery following lumbosacral deafferentation (Goldberger, '88a,b). The methods used cannot distinguish between an anatomically increased projection and an increase in the transmitters synthesized and transported within the pathways examined. Morphologicalstudies indicate a rapid and virtually complete replacement of terminals after dorsal rhizotomy (Murray and Goldberger, '861, and this study was undertaken to determine which systems might contribute to the reinnervation. The solution to this problem awaits the application of EM-immunocytochemicalmethods which have the resolution required to identify changes in innervation patterns at the level of synaptic terminals.

ACKNOWLEDGMENTS Shwun-De Wang was supported by a Chung Shan Institute of Science and Technology (CSIST, Taiwan, R.O.C.)

fellowship, from Department of Biology and Anatomy, National Defense Medical Center, Taipei, Taiwan, Republic of China. The research was supported by NS-24707. The advice and support of Drs. George Martin, Pat Levitt, Hazel Murphy, Alan Tessler, Wendy Battisti and Theresa Eckenrode is greatly appreciated.

LITERATURE CITED Artymyshyn, R., and M. Murray (1985) Substance P in the interpeduncular nucleus of the rat: Normal distribution and the effects of dederentation. J. Comp. Neurol. 231:78-90. Arvidsson, J., J. Ygge, and G. Grant (1986) Cell loss in the lumbar dorsal root ganglion and transganglionic degeneration after sciatic nerve resection in the rat. Brain Res. 373:15-21. Battisti, W.P., B.E. Levin, and M. Murray (1987) Norepinephrine in the interpeduncular nucleus of the rat: Normal distribution and the effects of the deafferentation. Brain Res. 418.287-300. Battisti, W.P., R. Artymyshyn, and M. Murray (1989) B1 and B2 adrenergic lz5Ipindolol binding sites in the interpeduncular nucleus. J. Neurosci. 9.2509-2518. Bernstein, J.J., and M.E. Bernstein (1971) Axonal regeneration and forniation of synapses proximal to the site of lesion following hemisection of the rat spinal cord. Exp. Neurol. 303336-351. Bjorklund, A,, and 0. Lindvall (1979) Regeneration of normal terminal innervation patterns by central noradrenergic neurons after 5,7dihydro+ryptamine-induced axotomy in the adult rat. Brain Res. 171:271-293. Bjorklund, A,, and L. Wiklund (1980) Mechanisms of regrowth of the bulbospinal serotonin system following 5,6-dihydroxytryptamine induced axotomy. I. Biochemical correlates. Brain Res. 1913109-127. Bregman, B.S. (1987a) 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. (198713) Spinal cord transplants permit the growth of serotoninergic axons across the site of neonatal spinal cord transection. Dev. Brain Res. 34265-279. Bullitt, E., W.D. Stofer, C.J. Vierck, and E.R. Per1 (1988) Reorganization of primary afferent nerve terminals in the spinal dorsal horn of the primate caudal to anterolateral chordotomy. J. Comp. Neurol. 270:549-558. Chen, S., and D.E. Hillman (1982) Plasticity of the parallel fiber-purkinje cell synapse by spine takeover and new synapse formation in adult rat. Brain Res. 240:205-220. Coimbra, A,, B.P. Sodre-Borges, and M.M. Magalhaes (1974) The suhstantia gelatinosa Rolandi of the rat. Fine structure, cytochemistry (acid phosphatase) and changes after dorsal root section. J. Neurocytol. 3:199-217. Coimbra, A,, A. Ribeiro-da-Silva, and D. Pignatelli (1984) Effects of dorsal rhizotomy on the several types of primary afferent terminals in laminae 1-111 of the rat spinal cord. Anat. Embryol. (Berl.) 170:279-287. Colado, M.I., A. Amedo, E. Peralta, and J. DelRio (1988) Unilateral dorsal rhizotomy decreases monoamine levels in the rat spinal cord. Neurosci. Lett. 87:302-306. Cotman, C.W., and J.V. Nadler (1978) Reactive synaptogenesis in the hippocampus. In C. Cotman (ed): Neuronal Plasticity. New York: Raven Press, pp. 227-273. Decima, E.E., D. Lorenzo, and L. Malluch (1986) Effects of partial axotomy upon synaptic function. Exp. Brain Res. 643464468, DiGiulio, A.M., B. Tenconi, A. Mannavola, P. Mantegazza, A. Schiavinato, and A. Gorio (1987) Spinal cord interneuron degenerative atrophy caused by peripheral nerve lesions is prevented by serotonin depletion. J. Neurosci. Res. 183443448, Fields, H.I., A.I. Basbaum, C.H. Clanton, and S.D. Anderson (1977) Nucleus raphe magnus inhibition of spinal cord dorsal horn neurons. Brain Res. 126:441453. Frankfurt, M., and A. Beaudet (1987) The ultrastmctural organization of regenerated serotonin axons in the dorsomedial hypothalamus of the adult rat. J. Neurocytol. 16:799-809. Frankfurt, M., and A. Beaudet (1988) Reinnervation of dopamine neurons by regenerating serotonin axons in the rat medial zona incerta. Exp. Brain Res. 72473480. Gage, F.H., A. Bjorklund, U. Stenevi, and S. Dunnett (1983) Functional correlates of compensatory collateral sprouting by aminergic and cholinergic afferents in the hippocampal formation. Brain Res. 26833947.

PLASTICITY IN SPINAL CORD OF ADULT RAT Gasser, U.E., and A.R. Dravid (1987) Noradrenergic, serotoninergic and cholinergic sprouting in the hippocampus that follows partial or complete transection of the septohippocampal pathway: Contributions of spared inputs. Exp. Neurol. 96:352-364. Gobel, S., W.M. Falls, and E. Humprey (1981) Morphology and synaptic connections of ultrafine primary axons in laminae I of the spinal dorsal horn: Candidates for the terminal axonal arbors of primary neurons with unmyelinated (C) axons. J. Neurosci. 1t1163-1179. Goldberger, M.E. (1988a) Spared root defierentation of a cat's hindlimb: Hierarchical regulation of pathways mediating recovery of motor behavior. Exp. Brain Res. 73t329-342. Goldberger, M.E. (1988b) Partial and complete defierentation of cat hindlimb: The contribution of behavioral substitution to recovery of motor function. Exp. Brain Res. 73:343-353. Goldberger, M.E., and M. Murray (1974) Restitution of function and collateral sprouting in cat spinal cord: The defierented animal. J. Comp. Neurol. 1583-57. Goldberger, M.E., and M. Murray (1978) Axonal sprouting and recovery of function may obey of the same laws. In C. Cotman (ed): Neuronal Plasticity. New York Raven Press, pp. 73-96. Goldberger, M.E., and M. Murray (1982) Lack of sprouting and its presence after lesion of the cat spinal cord. Brain Res. 241:227-239. Goldberger, M.E., and M. Murray (1985) Recovery of function and anatomical plasticity after damage to the adult and neonatal spinal cord. In c. Cotman (ed): Synaptic Plasticity. Raven Press, pp. 77-110. Goldberger, M.E., and E. Paige (1986) Anatomical plasticity of serotonin projections after defierentation of cat spinal cord. SOC.Neurosci. Abstr. 12:512. Gustafson, E., and R.Y. Moore (1987) Noradrenaline neuron plasticity in developing rat brain: Effects of neonatal 6-hydroxydopamine demonstrated by dopamine B-hydroxylase immunocytochemistry. Dev. Brain Res. 37:143-155. Hadjiconstantinou, M., P. Panula, Z. Lackovic, and N.H. Neff (1984) Spinal cord serotonin: A biochemical and immunohistochemical study following transection. Brain Res. 322:245-254. Haring, J.H., and J.N. Davis (1985) Retrograde labeling of locus coeruleus neurons after lesion-induced sprouting of the coeruleohippocampal projections. Brain Res. 360:384-388. Himes, B.T., and A. Tessler (1989) Death of some dorsal root ganglion neurons and plasticity of others following sciatic nerve section in adult and neonatal rats. J. Comp. Neurol. 284215-230. Hoff, S.F. (1986) Lesion induced transneuronal plasticity in the adult rat hippocampus. Neuroscience 19: 1227-1 233. Hsu, S.M., L. Raine, and H. Fanger (1981) The use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem. 29:579-580. Hulsebosch, C.E., and R.E. Coggeshall(1981) Sproutingof dorsal root axons. Brain Res. 224:170-174. Hunt, S.P., M.N. Rossor, P.C. Emson, and V. Clement-Jones (1982) Substance P and enkephalins in spinal cord after limb amputation. Lancet 1t1023. Jessell, T., A. Tsunoo, I. Kanazawa, and M. Otsuka (1979) Substance P: Depletion in the dorsal horn of rat spinal cord after section of the peripheral processes of primary sensory neurons. Brain Res. 168.947259. Kai-Kai, M.A. (1989) Cytochemistry ofthe trigeminal and dorsal root ganglia and spinal cord of the rat. Comp. Biochem. Physiol. [A1 93:183-193. Kai-Kai, M.A., and P. Keen (1985) Localization of 5-hydroxytryptamine t o neurons and endoneurial mast cells in rat sensory ganglia. J. Neurocytol. 14t63-78. Kerr, F.W. (1972) The potential of cervical primary fierents to sprout in the spinal nucleus of V following long term trigeminal denervation. Brain Res. 43t547-560. Kerr, F.W. (1975) Neuroplasticity of primary fierents in the neonatal cat and some results of early deafferentation of the trigeminal spinal nucleus. J. Comp. Neurol. 163t305-328. Kirk, R.E. (1968) Experimental Design: Procedures for the Behavioral Sciences.Belmont, CA: Brooks/Cole. Knyihar-Csillik, E., A. Bezegh, S. Boti, and B. Csillik (1986) Thiamine monophosphatase: A genuine marker for transganglionic regulation of primary sensory neurons. J. Histochem. Cytochem. 34t363-371. Leedy, M.G., M.S. Beattie, and J.C. Brenahan (1987) Testosterone induced plasticity of synaptic inputs to adult mammalian motoneurons. Brain Res. 424:386-390.

567 Levitt, P., and R.Y. Moore (1980) Organization of brainstem noradrenaline hyperinnervation following neonatal 6-hydroxydopamine treatment in rat. Anat. Embryol. (Berl.) 158:133-150. Liu, C.N., and W.W. Chambers (1958) Intraspinal sprouting of dorsal root axons. Arch. Neurol. 79:46-61. Lynch, G., C. Gall, G. Rose, and C. Cotman (1976) Changes in the distribution of the dentate gyrus associational system following unilateral or bilateral lesions of the entorhinal cortex. Brain Res. 11Ot57-71. McNeill, D.L., and C.E. Hulsebosch (1987) Intraspinal sprouting of rat primary fierents after deafhentation. Neurosci. Lett. 81:57-62. McWilliams, R., and G. Lynch (1979) Terminal proliferation in the partially defierented dentate gyrus: Time course for the appearance and removal of degeneration and the replacement of lost terminals. J. Comp. Neurol. 187:191-198. Micevych, P.E., A. Stroink, T. Yaksh, andV.L.W. Go (1986a) Immunochemical studies of substance P and cholecystokinin octapeptide recovery in dorsal horn following unilateral lumbosacral ganglionectomy. Somatosens. Res. 3t239-260. Micevych, P.E., B.E. Rodin, and L. Kruger (1986b) The controversial nature of the evidence for neuroplasticity of afferent axons in the spinal cord. In T.L. Yaksh (ed): Spinal Afferent Processing. New York: Plenum Press, pp. 417443. Molander, C. (1986) On the Organization of Spinal Cord Projections From Hindlimb Sensory Nerves. GOTAB, Stockholm: ISBN. Moore, R.Y., A. Bjorklund, and U. Stenevi (1971) Plastic changes in the adrenergic innervation of the rat septal area in response to denervation. Brain Res. 33:13-35. Murray, M., and M.E. Goldberger (1974) Restitution of function and collateral sprouting in the cat spinal cord: The partially hemisected animals. J. Comp. Neurol. 155t19-36. Murray, M., and M.E. Goldberger (1986) Replacement of synaptic terminals in lamina I1 and Clarke's nucleus after unilateral lumbosacral dorsal rhizotomy in adult cat. J. Neurosci. 6t3205-3217. Murray, M., J. Zimmer, and G. Raisman (1979) Quantitative electron microscopic evidence for reinnervation in the adult rat interpeduncular nucleus after lesions of the fasciculus retroflexus. J. Comp. Neurol. 187t447468. Murray, M., S.-D.Wang, M.E. Goldberger, and P. Levitt (1990) Modification of astrocytes in the spinal cord following dorsal root or peripheral nerve lesions. Exp. Neurol. 11Ot248-257. Nakamura, S., T. Shirokawa, and T. Sakaguchi (1984) Increased projection from the locus. Coeruleus to the lateral geniculate nucleus and visual cortex in young adult rats following unilateral enucleation. Neurosci. Lett. 49t77-80. Pare, M., L. Descarries, and L. Wiklund (1987) Innervation and reinnervation of rat inferior olive by neurons containing serotonin and substance P: An immunohistochemical study after 6,6-dihydroxytryptaminelesioning. J. Neurocytol. 16t155-167. Pickel, V.M., H. Krebs, and F.E. Bloom (1973) Proliferation of norepinephrine-containing axons in rat cerebellar cortex after peduncle lesions. Brain Res. 59:169-179. Polistina, D., M. Murray, and M.E. Goldberger (1987) Myelinated fibers contributed to dorsal root sprouting after partial defierentation in rat. SOC.Neurosci. Abstr. 17Purt 1):166. Polistina, D., M. Murray, and M.E. Goldberger (1989) Plasticity of dorsal root and descending serotoninergic projections after partial defierentation of adult rat spinal cord. J. Comp. Neuro. 299:349-363. Pubols, L., and D. Bowen (1988) Lack of central sprouting of primary fierent fibers after ricin dederentation. J. Comp. Neurol. 275282-287. Raisman, G. (1969) Neuronal plasticity in the septal nuclei of the adult rat. Brain Res. 14:2548. Ralston, H.J., and D.D. Ralston (1979) The distribution of dorsal root axons in laminae I and I1 ofthe macaque monkey spinal cord. J. Comp. Neurol. 184:643-684. Reis, D.J., R.A. Ross, G. Gilad, and T.H. Joh (1978) Reaction of central catecholaminergic neurons to injury: Model systems for studying the neurobiology of central regeneration and sprouting. In C.W. Cotman (ed):Neuronal Plasticity. New York: Raven Press, pp. 197-226. Ribeiro-da-Silva,A,, J.M. Castro-lopes, and A. Coimbra (1986) Distribution of glomeruli with fluoride-resistance acid phosphatase (FRAP)-containing terminals in the substantia gelatinosa of the rat. Brain Res. 377:323-329. Ribeiro-da-Silva, A,, P. Tagari, and A.C. Cuello (1989) Morphological characterization of substance P-like immunoreactive glomeruli in the superficial dorsal horn of the rat spinal cord and trigeminal subnucleus caudalis: A quantitative study. J. Comp. Neurol. 281:497-515.

568 Rodin, B.E., and L. Kruger (1984) Absence of evidence for intraspinal sprouting in dorsal root axons caudal to a partial spinal hemisection: A horseradish peroxidase transport study. Somatosens. Res. 2:171-192. Rodin, B.E., S.L. Sampogna, and L. Kruger (1983) An examination of intraspinal sprouting in dorsal root axons with the tracer horseradish peroxidase. J. Comp. Neurol. 215:187-198. Ruda, M.A., J. Coffield, and H.W.M. Steinbusch (1982) Immunocytochemical analysis of serotoninergic axons in laminae I and I1 of the lumbar spinal cord of thecat. J. Neurosci. 2:1660-1671. Sakaguchi, T., T. Shirokawa, and S. Nakamura (1984) Changes in projection from locus coeruleus to lateral geniculate nucleus following ablation of visual cortex in adult rats. Brain Res. 321:319,320. Schmidt, R.H., and R.K. Bhatnagar (1979) Regional development of norepinephrine, dopamine-B-hydroxylaseand tyrosine hydroxylase in the rat brain subsequent to neonatal treatment with subcutaneous 6-hydroxydopamine. Brain Res. 166:293-308. Snyder, R.L. (1982) Light and electron microscopic autoradiographic study of the dorsal root projections in the cat dorsal horn. Neuroscience 7: 14 1 7 4 3 7. Swett, J.E., and C.J. Woolf (1985) The somatotopic organization of primary afferent terminals in the superficial dorsal horn of the rat spinal cord of the rat. J. Comp. Neurol. 231:6677. Tessler, A,, E. Glazer, R. Artymyshyn, M. Murray, and M. Goldberger (1980) Recovery of substance P in the cat spinal cord after unilateral lumbosacral deafferentation. Brain Res. 191:459-470. Tessler, A., B.T. Himes, R. Artymyshyn, M. Murray, and M. Goldberger (1981) Spinal neurons mediate return of substance P followingdeafferentation of cat spinal cord. Brain Res. 230:263-281. Tessler, A,, B.T. Himes, K. Soper, M. Murray, M.E. Goldberger, and S. Reichlin (1984) Recovery of substance P but not somatostatin in the cat spinal cord after unilateral lumbosacral dorsal rhizotomy: A quantitative study. Brain Res. 30595-102.

S.-D. WANG ET AL. Too, H.P., J.L. Cordova, and J.E. Maggio (1989) Heterogeneityoftachykininlike immunoreactive peptides in the rat spinal cord and dorsal root ganglia. Peptides (Fayetteville) 10:25-30. Tuchscherer, M.H., C. Knox, and V.S. Seybold (1987) Substance P and cholecystokinin-like immunoreactive varicosities in somatosensory and autonomic regions of the rat spinal cord: A quantitative study of coexistence.J. Neurosci. 733984-3995. Vacca-Galloway, L.L., and C.C. Steinberger (1986) Substance P neurons sprout in the cervical spinal cord of the wobbler mouse: A model for motor neuron disease. J. Neurosci. Res. 16:657-670. Wang, S.-D., T.C. Eckenrode, M.E. Goldberger, and M. Murray (1987) Plasticity of intrinsic and extrinsic systems after rhizotomy of rat spinal cord. SOC.Neurosci. Abstr. 13(Purt 1):166. Wells, J., and L.N. Tripp (1987a) Time course of reactive synaptogenesis in the subcortical somatosensory system. J. Comp. Neurol. 255466475. Wells, J., and L.N. Tripp (1987b) Time course of the reaction of glial fibers in the somatosensory thalamus after lesions in the dorsal column nuclei. J. Comp. Neurol. 255:476-482. Wessendorf, M.W., and R. Elde (1987) The coexistence of serotonin- and substance P-like immunoreactivity in the spinal cord of the rat as shown by immunofluorescent double labelling. J. Neurosci. 7t2352-2363. Westlund, K.N., R.M. Bowker, M.G. Ziegler, and J.D. Coulter (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res. 263: 15-3 1. Ygge, J., H. Aldskogius and G. Grant (1981)Asymmetries and symmetries in the number of thoracic dorsal root ganglion cells. J. Comp. Neurol. 202:365-372. Zhou, F.C., and E. Azmitia (1984) Induced homotypic collateral sprouting of serotoninergic fibers in the hippocampus ofrat. Brain Res. 308:53-62. Zhou, F.C., and E. Azmitia (1986) Induced homotypic sprouting serotoninergic fibers in hippocampus. 11. An immunocytochemicalstudy. Brain Res. 373:337-348.