Michael E. Goldberger

The recovery of postural reflexes and locomotion following low thoracic hemisection in adult cats involves compensation by undamaged primary afferent pathways.

Spinal hemisection in the adult cat results in motor impairments followed by substantial recovery of function (16, 20, 39, 53). The present study was undertaken to assess the contribution of undamaged ipsilateral segmental and contralateral descending systems to recovery of motor function. Quantitative behavioral methods were used to examine monopedal reflex and bipedal locomotor functions after thoracic hemisection. Different facets of motor behavior recover at different times. The recovery of monopedal postural reflexes precedes the recovery of more complex motor behavior. Since the reflexes tested are initiated by segmental afferent input and show recovery and normal motor patterns during locomotion, as defined by kinematic analysis show recovery, it is likely that dorsal root input compensates for the loss of descending input to one side of the spinal cord. Quantitative immunocytochemical methods for visualizing the central projections of dorsal root fibers (monoclonal antibody RAT-102; 49) and the descending serotoninergic pathway were used to examine the response of these pathways to hemisection. Hemisection results in a permanent decrease in the density of serotoninergic projections and a permanent increase in dorsal root projections in the spinal cord. The increased density of RAT-102 may represent an increase in the projection of dorsal root fibers and provide the increased input necessary to mediate enhanced reflex control. A transient increase in GAP-43 in the dorsal horn ipsilateral to the hemisection suggests that the increased density of RAT-102 immunoreactivity is associated with growth. Taken together, our results suggest that sprouting of primary afferents within the spinal cord is one mechanism underlying the recovery of function after hemisection.

Recovery of substance P in the cat spinal cord after unilateral lumbosacral deafferentation

Changes in substance P in the cat spinal cord after deafferentation of the hindlimb were investigated using the peroxidase-antiperoxidase technique. Unilateral lumbosacral dorsal root section (L1-S3) is followed by a decrease in dorsal horn (laminae I, II and V) substance P reaction product which is most marked at 10-11 days. There is no observable change in the ventral horn. The 13 or 15 day survivors demonstrate an increase over that seen at 10-11 days, and still greater amounts appear in the dorsal horn of 1 month survivors. After 1 month there is little further observable increase. The location of the returned reaction product resembles that of normals, but differs in its staining characteristics. Substance P containing cell bodies are observed in normal animals and on intact and deafferented sides of experimentals, suggesting that interneurons and propriospinal fibers may be a source of the returning substance P reaction product. The decrease in dorsal horn substance P at short times after rhizotomy followed by an increase at longer times is consistent with axonal sprouting. If so, then the time course of sprouting parallels that of locomotor recovery and supports the hypothesis that the two phenomena are related.

Sciatic nerve transection produces death of dorsal root ganglion cells and reversible loss of substance P in spinal cord

Sciatic nerve section has been shown to reduce substance P (SP) in the dorsal horn of the spinal cord, but the mechanism which underlies the reduction is not understood. Whether SP levels subsequently recover as they do after dorsal rhizotomy has also been unknown. To test the hypothesis that transganglionic degeneration of primary afferents contributes to the reduction of SP, we have studied the changes in SP which result from section of the cat sciatic nerve and determined the extent of dorsal root ganglion (DRG) cell death. Sciatic nerve section resulted in DRG cell death, but the amount was variable and not seen in all animals. Reduction in dorsal horn and DRG SP was seen in all animals, and in the spinal cord it was followed by recovery. These sequelae resemble the changes which follow dorsal rhizotomy. After sciatic nerve section, the reduction in dorsal horn SP is smaller than after rhizotomy, the recovery more complete, and both the reduction and the recovery proceed more slowly. Evidence is presented that similar mechanisms may contribute to depletion of intraspinal SP after sciatic nerve section and after dorsal rhizotomy. The mechanisms contributing to recovery of spinal cord SP after sciatic nerve section may resemble known mechanisms of recovery that occur when the lesion is central.

Spinal neurons mediate return of substance P following deafferentation of cat spinal cord

Deafferentation of the cat dorsal horn by complete unilateral lumbosacral dorsal rhizotomy produces a loss and subsequent partial recovery of substance P (SP) immunoreactivity as visualized by the peroxidase-antiperoxidase technique. The present experiments aimed to determine whether this return of SP represents a generalized response of all fiber systems afferent to the denervated segments or a more selective response of a specific spinal system. Although a contribution from other sources cannot be excluded by this qualitative immunocytochemical technique, several observations indicate that the return of SP staining depends on interneurons which contain SP immunoreactivity: (1) the amount of SP staining in the chronically deafferented dorsal horn deprived of extrinsic fiber systems is comparable to that seen after deafferentation alone; (2) SP-containing neurons are present within the lumbar segments; and (3) destruction of lumbar neurons by the intraspinal injection of kainic acid abolishes SP staining from the chronically deafferented dorsal horn. From these observations it would appear that the anatomical plasticity of SP-containing fibers in the deafferented dorsal horn is due to the response of a particular system rather than to a generalized response of all systems which terminate there.

Modification of astrocytes in the spinal cord following dorsal root or peripheral nerve lesions.

Glial fibrillary acidic protein (GFAP) immunocytochemistry was used to monitor the response of astrocytes in the rat spinal cord to either dorsal root or sciatic nerve lesions. Image analysis methods were used to provide a quantitative correlate of the reactive gliosis. Multiple dorsal root section elicited a rapid increase in GFAP immunoreactivity of astrocytes unilaterally within the spinal cord along the pathway of the degenerating dorsal root axons in the dorsal and ventral horns and this gliosis persisted in the dorsal horn beyond the time at which active phagocytosis of degenerative debris occurred. Labeling of proliferating cells using [3H]thymidine revealed that none of the dividing cells contained detectable GFAP, suggesting that the increased GFAP labeling represents primarily a hypertrophy rather than a proliferation of astrocytes. Comparison of animals that had been deafferented in the early neonatal period with those deafferented as adults indicated that the GFAP immunoreactive response persisted following neonatal lesions but that it was markedly less intense than after adult lesions. Sciatic nerve section in adults does not result in extensive frank degeneration but it does evoke a rapid and marked increase in staining of astrocytes both in the dorsal horn and in the ventral horn. Transganglionic changes in GFAP staining in the dorsal horn occur by 3 days post-operatively, which is much earlier than the time of dorsal root ganglion neuron death caused by the sciatic nerve lesion. These experiments indicate that astrocytes can respond to signals from a variety of changes in neurons, including not only Wallerian degeneration, but also retrograde and transganglionic changes.

Criteria for assessing recovery of function after spinal cord injury: behavioral methods

*Department of Anatomy, The Medical College of Pennsylvania/EPPI Division, 3200 Henry Avenue, Philadelphia, Pennsylvania 19129; TDepartment of Anatomy, Georgetown University, 3900 Reservoir Rd. NW., Washington DC 20007; *Department of Neuroscience, University of Florida College of Medicine, Gainesville, Florida 32610; and BAmerican Paralysis Association, 500 Morris Avenue, Springfield, New Jersey 07081

Locomotor recovery after unilateral hindlimb deafferentation in cats.

After hindlimb deafferentation, recovery of locomotor patterns may be distinguished from recovery of accurate movements. Cats were timed while crossing wide runways requiring that locomotion be present, and narrow runways which require that it also be accurate. Locomotion recovers soon after deafferentation and becomes almost normal in speed, although accuracy is still absent. This accuracy returns 10 days postoperatively. Ventral root afferents with cell bodies in dorsal root ganglia are unnecessary for recovery. Postural reflexes (and other non-topographic feedback) are used as cues for limb guidance after removal of that limbs topographic feedback. Anatomical experiments show that a thoracic root sprouts during the recovery period. This sprouting may increase the potency of postural reflexes. Although lower thoracic roots contribute to recovery, they are not essential. Ipsilateral descending system are essential, and represent the final common path for all recovery after deaffernting a hindlimb. Contralateral descending and afferent systems alone cannot mediate overground locomotion. Thus there is a difference between the systems which mediate overground locomotion and those used by animals in treadmill locomotion or with L-DOPA or clonidine or brain stem stimulation. The proposed mechanism underlying recovery of accurate locomotion is behavioral substitution which may be faciliated by collateral sprouting of relevant systems.

Consequences of damage to the sensorimotor cortex in neonatal and adult cats. II. Maintenance of exuberant projections

After chronic sensorimotor cortex ablations, sparing and greater recovery of function were seen in neonatally operated cats compared with adult operated cats. These results suggested that undamaged cortex in neonatal operates might display projections different from those of adult operates. Injections of horseradish peroxidase-wheat germ agglutinin (HRP-WGA) were made in ipsilateral parietal cortex adjacent to the sensorimotor cortex ablations or in the contralateral sensorimotor cortex. No changes in the projections of the parietal cortex were seen in operated cats or in the projections of the undamaged sensorimotor cortical projections of adult operates. In contrast, the intact sensorimotor cortex of neonatal operates exhibited crossed corticothalamic and corticorubral projections not present in normal or adult operated animals, whereas the corticospinal tract (CST) was unchanged by the ablations. Analysis of neurons within the ventroanterior-ventrolateral nuclear complex of the thalamus ipsilateral to the ablation showed that the surviving cells of neonatal operates were equal in number but were, on average, larger than those of normals and adult operates. Some neurons in neonatal operates were larger than any seen in adult operates and normals. Injections of HRP/WGA were also made into the sensorimotor cortex of normal newborn animals. Dense bilateral corticothalamic and corticorubral projections were present. The CST had extended to lumbar levels by the day of birth but projections to the grey matter were sparse. Thus, bilateral projections seen in neonatal operates probably represent retention of some exuberant projections present in normal neonatal animals. The CST which exhibited no exuberant projection was unchanged by the lesion. The crossed corticothalamic and corticorubral projections are likely to play a role in sparing and recovery of function particularly in sparing of contact placing.

Infant lesions effect: I. Development of motor behavior following neonatal spinal cord damage in cats

This study was undertaken to determine the effect of spinal cord damage on motor development, and to determine whether there is greater survival of motor function in those motor patterns with a later onset of function than in those which are present at birth. The postnatal development of postural reflexes and locomotion was examined during the first 4 months of life in normal kittens and in those which had received a spinal cord lesion (at high cervical or low thoracic levels) at birth. The results suggest that there are some similarities in normal development, recovery of function after adult lesions and recovery and/or development of function after neonatal lesions. After neonatal lesions, just as after lesions in adults, reflex recovery appears to underlie recovery of locomotion. After spinal lesions, the pattern and sequence of motor development was identical to that seen in normal animals. Hindlimb motor development was normal for some time after the spinal lesion, but deficits appeared later. These observations suggest that postural reflexes and locomotion are not dependent upon ipsilateral descending input for their onset, but only for their maturation. Unexpectedly, tactile placing developed after neonatal spinal cord lesions. This represents sparing of function, for tactile placing is abolished and does not recover after the same lesion sustained in adulthood. Tactile placing is the last of the series of postural reflexes to develop. It depends on the last of the spinal pathways to develop, the corticospinal tract. Two aspects of this study support the hypothesis that later developing motor patterns will have a greater chance for survival and subsequent development than those which are present at birth. First, the immediate effects of spinal cord lesions on postural reflexes are more severe on those reflexes that are more mature at birth. Second, the spinal cord lesions produce more severe impairment of the more mature forelimb motor function than of the less mature hindlimb motor function. The hypothesis is not supported, however, when the long-term effect of spinal cord lesions on the maturation of motor behavior is considered. All postural reflexes and locomotion fail to mature fully, i.e. they retain characteristics of the immature responses.

Lack of sprouting and its presence after lesions of the cat spinal cord

Degeneration methods were used to study the dorsal root and descending projections after chronic partial denervation of adult cat spinal cord. Conventional mapping methods were used, supplemented in some cases by densitometric measurements of the amounts of degeneration present. The amount of shrinkage of spinal gray matter in some sections was estimated by planimetric measurement. Two preparations were used: (1) partial unilateral rhizotomy in which all dorsal roots caudal to T4 were cut except L6 (spared root preparation); (2) complete unilateral deafferentation. The projection of L6 dorsal roots was examined in spared root preparations. T13 dorsal root projections were examined in deafferented preparations in which T13 was the lowest remaining root. The projection of descending systems was mapped in spared root and deafferented preparations. The spared root displayed an increased projection in the lateral portion of the dorsal horn, in the zona intermedia, Clarkes nucleus and in the base and reticular zone of nucleus gracilis. The lowest remaining root (T13) increased its projection to laminae VII and VIII and to the base and reticular zone of nucleus gracilis. In all cases, when an increased projection resulted from prior denervation, the increase never exceeded the boundaries of the normal projection. No sprouting was observed in those regions with the strictest topographical organization, the cell nests of nucleus gracilis or lamina IX of the spinal cord, even though these regions were partially denervated by the chronic lesions. Descending projections were increased on the experimental side of deafferented preparations12 but not of spared root preparations, suggesting that the presence of the spared root may prevent sprouting by descending systems. Because measurements of gray matter indicated that maximal sprouting occurred in segments showing least shrinkage (sprouting of L6 spared root into L6 segment), in this case shrinkage cannot account for the increased density of degeneration. These results suggest that certain conditions are important for the regulation of sprouting in the adult CNS. Firstly, sprouting is limited by a requirement for proximity and/or overlap. Secondly, the strictness of topographical localization within a particular region may limit the likelihood of sprouting into that region. Finally, a competitive or hierarchical relationship among the remaining systems may modify the capacity of a particular system to sprout.