The effects of long-standing limb loss on anatomical reorganization of the somatosensory afferents in the brainstem and spinal cord

We examined the terminations of sensory afferents in the brainstem and spinal cord of squirrel monkeys and prosimian galagos 4-8 years after a therapeutic forelimb or hindlimb amputation within 2 months of birth. In each animal, the distributions of labeled sensory afferent terminations from remaining body parts proximal to the limb stump were much more extensive than in normal animals. These sprouted afferents extended into the portions of the dorsal horn of the spinal cord as well as the cuneate and external cuneate nuclei of the brainstem (forelimb amputees) or spinal Clarke's column (hindlimb amputee) related to the amputated limb. Such reorganization in sensory afferents along with reorganization of the motor efferents to muscles (Wu and Kaas, J Neurosci 19: 7679-7697, 1999, Neuron 28: 967-978, 2000) may provide a basis for mislocated phantom sensations of missing forelimb movements accompanying actual shoulder movements during cortical stimulation or movement imagery in patients with amputations.     

Brachially innervated ectopic hindlimbs in the chick embryo. I. Limb motility and motor system anatomy during the development of embryonic behavior

The functional status of brachially innervated hindlimbs, produced by transplanting hindlimb buds of chick embryos in place of forelimb buds, was quantified by analyzing the number and temporal distribution of spontaneous limb movements. Brachially innervated hindlimbs exhibited normal motility until E10 but thereafter became significantly less active than normal limbs and the limb movements were more randomly distributed. Contrary to the findings with axolotls and frogs, functional interaction between brachial motoneurons and hindlimb muscles cannot be sustained in the chick embryo. Dysfunction is first detectable at E10 and progresses to near total immobility by E20 and is associated with joint ankylosis and muscular atrophy. Although brachially innervated hindlimbs were virtually immobile by the time of hatching (E21), they produced strong movements in response to electrical stimulation of their spinal nerves, suggesting a central rather than peripheral defect in the motor system. The extent of motoneuron death in the brachial spinal cord was not significantly altered by the substitution of the forelimb bud with the hindlimb bud, but the timing of motoneuron loss was appropriate for the lumbar rather than brachial spinal cord, indicating that the rate of motoneuron death was dictated by the limb. Measurements of nuclear area indicated that motoneuron size was normal during the motoneuron death period (E6-E10) but the nuclei of motoneurons innervating grafted hindlimbs subsequently became significantly larger than those of normal brachial motoneurons. Although the muscle mass of the grafted hindlimb at E18 was significantly less than that of the normal hindlimb (and similar to that of a normal forelimb), electronmicroscopic examination of the grafted hindlimbs and brachial spinal cords of E20 embryos revealed normal myofiber and neuromuscular junction ultrastructure and a small increase in the number of axosomatic synapses on cross-sections of motoneurons innervating grafted hindlimbs compared to motoneurons innervating normal forelimbs. The anatomical data indicate that, rather than being associated with degenerative changes, the motor system of the brachial hindlimb of late-stage embryos is intact, but inactive. © 1993 John Wiley & Sons, Inc.     

Nonoverlapping Thalamocortical Connections to Normal and Deprived Primary Somatosensory Cortex for Similar Forelimb Receptive Fields in Chronic Spinal Cats

The fluorescent dye retrograde tracing technique, using fast blue in combination with fluorogold, was used to examine thalamocortical projections from the ventrobasal complex to primary somatosensory cortex in chronic spinal cats that sustained T12 cord transection at 2 weeks of age. Following cord transection at this age, it has been shown that forelimb afferents can excite the deprived hindlimb projection zone, in addition to the region of somatosensory cortex that they normally occupy (McKinley et al, 1987). These two regions of cortex are separated by over 10 mm, thus facilitating the determination of whether the forelimb representation in “hindlimb cortex” is derived from the sector of the ventrobasal complex of the thalamus representing the forelimb, hindlimb, or both. Injections of the two dyes into separate regions of the cortex that were excited by the same peripheral forelimb receptive fields produced single labeling of two nonoverlapping clusters of thalamic neurons. This finding suggests that the projections for these two areas are independent and distinct, and indicates that altered thalamocortical projections do not contribute the critical component underlying reorganizational changes observed at the cortical level after spinal cord transection. It is hypothesized that the degree of reorganization required to achieve the magnitude of change observed in the cortex must occur below the level of the thalamocortical relay.     

Areal Distributions of Cortical Neurons Projecting to Different Levels of the Caudal Brain Stem and Spinal Cord in Rats

Distributions of corticospinal and corticobulbar neurons were revealed by tetramethylbenzidine (TMB) processing after injections of wheatgerm agglutinin conjugated to horseradish peroxidase (WGA:HRP) into the cervical or lumbar enlargements of the spinal cord, or medullary or pontine levels of the brain stem. Sections reacted for cytochrome oxidase (CO) allowed patterns of labeled neurons to be related to the details of the body surface map in the first somatosensory cortical area (SI). The results indicate that a number of cortical areas project to these subcortical levels: (1) Projection neurons in granular SI formed a clear somatotopic pattern. The hindpaw region projected to the lumbar enlargement, the forepaw region to the cervical enlargement, the whisker pad field to the lower medulla, and the more rostral face region to more rostral brain stem levels. (2) Each zone of labeled neurons in SI extended into adjacent dysgranular somatosensory cortex, forming a second somatotopic pattern of projection neurons. (3) A somatotopic pattern of projection neurons in primary motor cortex (MI) paralleled SI in mediolateral sequence corresponding to the hindlimb, forelimb, and face. (4) A weak somatotopic pattern of projection neurons was suggested in medial agranular cortex (Ag_m), indicating a premotor field with a rostromedial-to-caudolateral representation of hindlimb, forelimb, and face. (5) A somatotopic pattern of projection neurons representing the foot to face in a mediolateral sequence was observed in medial parietal cortex (PM) located between SI and area 17. (6) In the second somatosensory cortical area (SII), neurons projecting to the brain stem were immediately adjacent caudolaterally to the barrel field of SI, whereas neurons projecting to the upper spinal cord were more lateral. No projection neurons in this region were labeled by the injections in the lower spinal cord. (7) Other foci of projection neurons for the face and forelimb were located rostral to SII, providing evidence for a parietal ventral area (PV) in perirhinal cortex (PR) lateral to SI, and in cortex between SII and PM. None of these regions, which may be higher-order somatosensory areas, contained labeled neurons after injections in the lower spinal cord. Thus, more cortical fields directly influence brain stem and spinal cord levels related to sensory and motor functions of the face and forepaw than the hindlimb.

The termination patterns of corticospinal and corticobulbar projections were studied in other rats with injections of WGA:HRP in SI. Injections in lateral SI representing the face produced dense terminal label in the contralateral trigeminal complex. Injections in cortex devoted to the forelimb and forepaw labeled the contralateral cuneate nucleus and parts of the dorsal horn of the spinal cord. The cortical injections also demonstrated interconnections of parts of SI with some of the other regions of cortex with projections to the spinal cord, and provided further evidence for the existence of PV in rats.     

Ultrastructural immunocytochemical localization of excitatory amino acids in the somatosensory system.

We studied the ultrastructure and the synaptic arrangement of glutamate-immunoreactive terminals in rats, in the superficial laminae of the spinal cord, the brainstem cuneate nucleus, and the thalamic ventroposterolateral nucleus, where a role for glutamate as neurotransmitter has been suggested by biochemical, physiological and pharmacological approaches. The antiserum employed was raised against glutaramate conjugated to keyhole limpet hemocyanin with glutaraldehyde, and was used for pre-embedding staining with an avidin-biotin-peroxidase method and for post-embedding staining with an immunogold procedure. Both methods yielded similar results, consisting of labeling of selected terminals in all the areas examined. Double immunogold labeling on the same thin section using antisera against gamma-amino-butyric acid (GABA) or substance P (SP), in combination with the anti-glutamate serum, showed that staining for glutamate and GABA was present in different terminals in all the regions examined; glutamate and SP were co-localized in a few terminals only in the superficial laminae of the spinal cord. By performing immunogold staining in combination with anterograde tracing, glutamate immunoreactivity could be localized in identified primary afferents to the dorsal spinal cord and cuneate nucleus, and in lemniscal afferents to the thalamus.     

Sensory-evoked turning locomotion in red-eared turtles: kinematic analysis and electromyography

We examined the limb kinematics and motor patterns that underlie sensory-evoked turning locomotion in red-eared turtles. Intact animals were held by a band-clamp in a water-filled tank. Turn-swimming was evoked by slowly rotating turtles to the right or left via a motor connected to the shaft of the band-clamp. Animals executed sustained forward turn-swimming against the direction of the imposed rotation. We recorded video of turn-swimming and computer-analyzed the limb and head movements. In a subset of turtles, we also recorded electromyograms from identified limb muscles. Turning exhibited a stereotyped pattern of (1) coordinated forward swimming in the hindlimb and forelimb on the outer side of the turn, (2) back-paddling in the hindlimb on the inner side, (3) a nearly stationary, “braking” forelimb on the inner side, and (4) neck bending toward the direction of the turn. Reversing the rotation caused animals to switch the direction of their turns and the asymmetric pattern of right and left limb activities. Preliminary evidence suggested that vestibular inputs were sufficient to drive the behavior. Sensory-evoked turning may provide a useful experimental platform to examine the brainstem commands and spinal neural networks that underlie the activation and switching of different locomotor forms.     

Convergence of forelimb afferent actions on C7-Th1 propriospinal neurones bilaterally projecting to sacral segments of the cat spinal cord

Propriospinal neurones located in the cervical enlargement and projecting bilaterally to sacral segments of the spinal cord were investigated electrophysiologically in eleven deeply anaesthetized cats. Excitatory or inhibitory postsynaptic potentials from forelimb afferents were recorded following stimulation of deep radial (DR), superficial radial (SR), median (Med) and ulnar (Uln) nerves. 26 cells were recorded from C7, 22 from C8 and 3 from Th1 segments. The majority of the cells were located in the Rexed's laminae VIII and the medial part of the lamina VII. In 10 cases no afferent input from the forelimb afferents was found. In the remaining neurones effects were evoked mostly from DR (88%) and Med (63%), less often from SR (46%) and Uln (46%). Inhibitory actions were more frequent than excitatory. The highest number of IPSPs was evoked from high threshold flexor reflex afferents (FRA)--all connections were polysynaptic. However, inhibitory actions were often evoked from group I or II muscle afferents (polysynaptic or disynaptic) and, less frequently, from cutaneous afferents (mostly polysynaptic). Di- or polysynaptic IPSPs often accompanied monosynaptic EPSPs from group I or II muscle afferents. Disynaptic or polysynaptic EPSPs from muscle and cutaneous afferents were also recorded in many neurones, while polysynaptic EPSPs from FRA were observed only exceptionally. Various patterns of convergence in individual neuronal subpopulations indicate that they integrate different types of the afferent input from various muscle and cutaneous receptors of the distal forelimb. They transmit this information to motor centers controlling hind limb muscles, forming a part of the system contributing to the process of coordination of movements of fore--and hind--limbs.     

Extensive dual innervation and mutual inhibition by forelimb and hindlimb inputs to ventroposterolateral nucleus projection neurons in the rat

The stimulation of brachial plexus and sciatic nerve resulted in a precisely timed, synchronous volley of inputs to ventroposterolateral (VPL) neurons from either forelimb or hindlimb. Such stimulation activated sensory fibers of all modalities and was therefore modality-nonspecific. Extracellular recordings of modality-nonspecific single-unit evoked responses from VPL showed that 13% of VPL projection neurons responded to both forelimb and hindlimb inputs. We also demonstrated mutually inhibitory interactions between inputs from forelimb and hindlimb in 45% of VPL units. Unlike the somatotopic map produced by others using modality-specific inputs, the modality-nonspecific evoked response map of VPL had a broadly overlapping distribution of evoked responses. This was especially true for the more caudal aspects of VPL. When the delivery of stimuli was appropriately timed, forelimb inputs caused the inhibition of responses to forelimb stimulation; similarly, hindlimb inputs inhibited responses to forelimb stimulation. The inhibition had a variable duration that may reflect a combination of processes, including recurrent inhibitory collateral input from the thalamic reticular nucleus (TRN) or an intrinsic hyperpolarizing inhibitory afterpotential of the VPL neuron. The presence of an extensive converging input on VPL neurons and an inhibitory correlate to this overlapping of inputs may explain the shifting of VPL maps following lesions of peripheral nerve, spinal cord, or dorsal column nuclei (DCN).     

Areal distributions of cortical neurons projecting to different levels of the caudal brain stem and spinal cord in rats

Distributions of corticospinal and corticobulbar neurons were revealed by tetramethylbenzidine (TMB) processing after injections of wheatgerm agglutinin conjugated to horseradish peroxidase (WGA:HRP) into the cervical or lumbar enlargements of the spinal cord, or medullary or pontine levels of the brain stem. Sections reacted for cytochrome oxidase (CO) allowed patterns of labeled neurons to be related to the details of the body surface map in the first somatosensory cortical area (SI). The results indicate that a number of cortical areas project to these subcortical levels: (1) Projection neurons in granular SI formed a clear somatotopic pattern. The hindpaw region projected to the lumbar enlargement, the forepaw region to the cervical enlargement, the whisker pad field to the lower medulla, and the more rostral face region to more rostral brain stem levels. (2) Each zone of labeled neurons in SI extended into adjacent dysgranular somatosensory cortex, forming a second somatotopic pattern of projection neurons. (3) A somatotopic pattern of projection neurons in primary motor cortex (MI) paralleled SI in mediolateral sequence corresponding to the hindlimb, forelimb, and face. (4) A weak somatotopic pattern of projection neurons was suggested in medial agranular cortex (Agm), indicating a premotor field with a rostromedial-to-caudolateral representation of hindlimb, forelimb, and face. (5) A somatotopic pattern of projection neurons representing the foot to face in a mediolateral sequence was observed in medial parietal cortex (PM) located between SI and area 17. (6) In the second somatosensory cortical area (SII), neurons projecting to the brain stem were immediately adjacent caudolaterally to the barrel field of SI, whereas neurons projecting to the upper spinal cord were more lateral. No projection neurons in this region were labeled by the injections in the lower spinal cord. (7) Other foci of projection neurons for the face and forelimb were located rostral to SII, providing evidence for a parietal ventral area (PV) in perirhinal cortex (PR) lateral to SI, and in cortex between SII and PM. None of these regions, which may be higher-order somatosensory areas, contained labeled neurons after injections in the lower spinal cord. Thus, more cortical fields directly influence brain stem and spinal cord levels related to sensory and motor functions of the face and forepaw than the hindlimb. The termination patterns of corticospinal and corticobulbar projections were studied in other rats with injections of WGA:HRP in SI.(ABSTRACT TRUNCATED AT 400 WORDS)     

Quadrupedal locomotion in squirrel monkeys (Cebidae: Saimiri sciureus): a cineradiographic study of limb kinematics and related substrate reaction forces

Quadrupedal locomotion of squirrel monkeys on small-diameter support was analyzed kinematically and kinetically to specify the timing between limb movements and substrate reaction forces. Limb kinematics was studied cineradiographically, and substrate reaction forces were synchronously recorded. Squirrel monkeys resemble most other quadrupedal primates in that they utilize a diagonal sequence/diagonal couplets gait when walking on small branches. This gait pattern and the ratio between limb length and trunk length influence limb kinematics. Proximal pivots of forelimbs and hindlimbs are on the same horizontal plane, thus giving both limbs the same functional length. However, the hindlimbs are anatomically longer than the forelimbs. Therefore, hindlimb joints must be more strongly flexed during the step cycle compared to the forelimb joints. Because the timing of ipsilateral limb movements prevents an increasing amount of forelimb retraction, the forelimb must be more protracted during the initial stance phase. At this posture, gravity acts with long moment arms at proximal forelimb joints. Squirrel monkeys support most of their weight on their hindlimbs. The timing of limb movements and substrate reaction forces shows that the shift of support to the hindlimbs is mainly done to reduce the compressive load on the forelimb. The hypothesis of the posterior weight shift as a dynamic strategy to reduce load on forelimbs, proposed by Reynolds ([1985]) Am. J. Phys. Anthropol. 67:335-349; [1985] Am. J. Phys. Anthropol. 67:351-362), is supported by the high correlation of ratios between forelimb and hindlimb peak vertical forces and the range of motion of shoulder joint and scapula in primates.     

Decoding Hindlimb Movement for a Brain Machine Interface after a Complete Spinal Transection

Stereotypical locomotor movements can be made without input from the brain after a complete spinal transection. However, the restoration of functional gait requires descending modulation of spinal circuits to independently control the movement of each limb. To evaluate whether a brain-machine interface (BMI) could be used to regain conscious control over the hindlimb, rats were trained to press a pedal and the encoding of hindlimb movement was assessed using a BMI paradigm. Off-line, information encoded by neurons in the hindlimb sensorimotor cortex was assessed. Next neural population functions, or weighted representations of the neuronal activity, were used to replace the hindlimb movement as a trigger for reward in real-time (on-line decoding) in three conditions: while the animal could still press the pedal, after the pedal was removed and after a complete spinal transection. A novel representation of the motor program was learned when the animals used neural control to achieve water reward (e.g. more information was conveyed faster). After complete spinal transection, the ability of these neurons to convey information was reduced by more than 40%. However, this BMI representation was relearned over time despite a persistent reduction in the neuronal firing rate during the task. Therefore, neural control is a general feature of the motor cortex, not restricted to forelimb movements, and can be regained after spinal injury.     

Blastema cell proliferation in vitro: effects of limb amputation on the mitogenic activity of spinal cord extracts

Primary cultures of mesenchymal cells of axolotl limb blastemas provide a very sensitive in vitro bioassay for studying nerve dependence of newt regeneration. These cells can be stimulated by crude spinal cord extracts of non-amputated animals in a dose-dependent manner up to 60 micrograms protein/ml of culture medium; at this concentration the mitotic index is increased 4-fold. Spinal cord extracts of axolotls 14 days after forelimb amputation (i.e., late bud stage) are more efficient in stimulating blastema cell proliferation (+50%) than extracts of axolotls 7 days after forelimb amputation (i.e., early bud stage) or of axolotls without amputation. In a similar manner, spinal cord extracts of young axolotls 14 days after forelimb amputation, are more stimulatory than older axolotls 14 d after forelimb amputation which regenerate only a very small blastema during the same time. It appears that spinal cord mitogenic activity is enhanced after limb amputation, probably in correlation with blastema cell requirements for limb regeneration.     

Coordinated network functioning in the spinal cord: An evolutionary perspective

The successful achievement of harmonious locomotor movement results from the integrated operation of all body segments. Here, we will review current knowledge on the functional organization of spinal networks involved in mammalian locomotion. Attention will not simply be restricted to hindlimb muscle control, but by also considering the necessarily coordinated activation of trunk and forelimb muscles, we will try to demonstrate that while there has been a progressive increase in locomotor system complexity during evolution, many basic organizational features have been preserved across the spectrum from lower vertebrates through to humans. Concerning the organization of axial neuronal networks that control trunk muscles, it has been found across the vertebrate range that during locomotor movement a motor wave travels longitudinally in the spinal cord via the coupling of rhythmic segmental networks. For hindlimb activation it has been found in all species studied that the rostral lumbar segments contain the key elements for pattern generation. We also showed that rhythmic arm movements are under the control of cervical forelimb generators in quadrupeds as well as in human. Finally, it is highlighted that the coordination of quadrupedal movements during locomotion derives principally from an asymmetrical coordinating influence occurring in the caudo-rostral direction from the lumbar hindlimb networks.     

Lack of evidence for cell death among avian spinal cord interneurons during normal development and following removal of targets and afferents.

Chick embryos and posthatched chicks were examined at several ages for the presence of pyknotic interneurons in the lumbar spinal cord. Because no pyknotic interneurons were found, direct cell counts of healthy interneurons were carried out and a comparison made between early- and late-stage embryos and hatchlings. There was no decrease in the number of interneurons in the ventral intermediate gray matter of the spinal cord between embryonic day (E) 8 and 2 weeks posthatching (PH) or in the dorsal horn between E10 and 2 weeks PH. To study whether interneuron survival is regulated by targets or afferents, a situation known to exist in other developing neural populations, early embryos were subjected to (1) removal of one limb, resulting in the loss of lateral motor column motoneurons and dorsal root ganglion sensory afferents; (2) transection of the thoracic spinal cord, thereby removing both descending afferents and rostral targets of spinal interneurons, or (3) a combination of the two operations. No reductions in interneuron numbers were found as a result of these operations. Furthermore, morphometric analysis also revealed no change in neuronal size following these experimental manipulations. By contrast, there was a slight decrease in the total area of spinal gray matter that was most prominent in the dorsal region following limb bud removal. Our results indicate (1) that spinal interneurons fail to exhibit the massive naturally occurring death of postmitotic neurons that has been observed for several other populations of spinal neurons, and (2) spinal interneurons appear to be relatively resistant to induced cell death following the removal of substantial numbers of afferent inputs and targets.     

Peripheral specification of Ia synaptic input to motoneurons innervating foreign target muscles

Muscle sensory neurons, called Ia afferents, make monosynaptic connections with functionally related sets of motoneurons in the spinal cord. Previous work has suggested that peripheral target muscles play a major role in determining the central connections of Ia afferents with motoneurons. Here, we ask whether motoneurons can also be influenced by their target muscles in terms of the monosynaptic input they receive from Ia afferents, by transplanting thoracic motoneurons into the lumbosacral spinal cord so that they innervate foreign muscles. Three or four segments of thoracic neural tube from stage 14-15 chicken embryos were transplanted to the lumbosacral region of stage 16-17 embryos, and electrophysiological recordings were made from transplanted motoneurons after the embryos had reached stage 38-40. Transplanted thoracic motoneurons innervated limb muscles and received monosynaptic inputs from Ia afferents. These connections were not random: Most of the connections were formed between Ia afferents and motoneurons projecting to the same muscle (homonymous connections). Few aberrant connections were found although the anatomical distribution of afferents in the transplant indicated that they had ample opportunity to contact inappropriate motoneurons. We conclude that although peripheral target cues are not sufficient to respecify an already committed motoneuron (turn a thoracic motoneuron into a lumbosacral motoneuron), they do provide sufficient information for Ia afferent input to be functionally correct.     

Mesodermal and neuronal retinoids regulate the induction and maintenance of limb innervating spinal motor neurons

During embryonic development, the generation, diversification and maintenance of spinal motor neurons depend upon extrinsic signals that are tightly regulated. Retinoic acid (RA) is necessary for specifying the fates of forelimb-innervating motor neurons of the Lateral Motor Column (LMC), and the specification of LMC neurons into medial and lateral subtypes. Previous studies implicate motor neurons as the relevant source of RA for specifying lateral LMC fates at forelimb levels. However, at the time of LMC diversification, a significant amount of retinoids in the spinal cord originates from the adjacent paraxial mesoderm. Here we employ mouse genetics to show that RA derived from the paraxial mesoderm is required for lateral LMC induction at forelimb and hindlimb levels, demonstrating that mesodermally synthesized RA functions as a second source of signals to specify lateral LMC identity. Furthermore, reduced RA levels in postmitotic motor neurons result in a decrease of medial and lateral LMC neurons, and abnormal axonal projections in the limb; invoking additional roles for neuronally synthesized RA in motor neuron maintenance and survival. These findings suggest that during embryogenesis, mesodermal and neuronal retinoids act coordinately to establish and maintain appropriate cohorts of spinal motor neurons that innervate target muscles in the limb.     

Anatomical specificity of regenerated muscle sensory afferents in the spinal cord of the bullfrog

The specificity of central projections made by regenerated muscle sensory fibers in the brachial spinal cord was studied with anatomical tracing methods. Sensory fibers were interrupted by freezing dorsal roots in postmetamorphic bullfrogs. After several months, regenerated sensory fibers were labeled with horseradish peroxidase applied to the triceps brachii muscle nerve, and their arborizations within the spinal cord were reconstructed from serial cross sections. Most of the regenerated projections from triceps muscle sensory afferents ended in or near their normal terminal field. A few branched and appeared to terminate more dorsally than normal, however, sometimes within the region where cutaneous afferents normally terminate. In contrast to the normal pathway followed by muscle afferents within the spinal cord, many regenerated afferents grew along the circumference of the spinal cord, just under the pial surface, and then turned abruptly toward the midline and into their appropriate terminal region. This suggests that regenerating afferents may actively seek out their appropriate targets and are not simply passively guided to them.     

Lack of evidence for cell death among avian spinal cord interneurons during normal development and following removal of targets and afferents

Chick embryos and posthatched chicks were examined at several ages for the presence of pyknotic interneurons in the lumbar spinal cord. Because no pyknotic interneurons were found, direct cell counts of healthy interneurons were carried out and a comparison made between early-and late-stage embryos and hatchlings. There was no decrease in the number of interneurons in the ventral intermediate gray matter of the spinal cord between embryonic day (E) 8 and 2 weeks posthatching (PH) or in the dorsal horn between E10 and 2 weeks PH. To study whether interneuron survival is regulated by targets or afferents, a situation known to exist in other developing neural populations, early embryos were subjected to (1) removal of one limb, resulting in the loss of lateral motor column motoneurons and dorsal root ganglion sensory afferents; (2) transection of the thoracic spinal cord, thereby removing both descending afferents and rostral targets of spinal interneurons, or (3) a combination of the two operations. No reductions in interneuron numbers were found as a result of these operations. Furthermore, morphometric analysis also revaled no change in neuronal size following these experimental manipulations. By contrast, there was a slight decrease in the total area of spinal gray matter that was most prominent in the dorsal region following limb bud removal. Our results indicate (1) that spinal interneurons fail to exhibit the massive naturally occurring death of postmitotic neurons that has been observed for several other populations of spinal neurons, and (2) spinal interneurons appear to be relatively resistant to induced cell death following the removal of substantial numbers of afferent inputs and targets.     

An unsupervised neural network model for the development of reflex co-ordination

In this paper, we present a model for the development of connections between muscle afferents and motoneurones in the human spinal cord. The model consists of a limb with six muscles, one motoneurone pool, one pooled (Ia-like) afferent for each muscle and a central programme generator. The weights of the connections between the afferents and the motoneurone pools are adapted during centrally induced movements of the limb. The connections between the afferents and the motoneurone pools adapt in a hebbian way, using only local information present at the synapses. This neural network is tested in two examples of a limb with two degrees of freedom and six muscles. Despite the simplifications, the model predicts the pattern of autogenic and heterogenic monosynaptic reflexes quite realistically.