Projections of digit afferents to the cuneate nucleus in the raccoon before and after partial deafferentation

Functional organization and connectivity of the dorsal column nuclei complex reveals a sensorimotor integration and distribution hub

The dorsal column nuclei complex (DCN-complex) includes the dorsal column nuclei (DCN, referring to the gracile and cuneate nuclei collectively), external cuneate, X, and Z nuclei, and the median accessory nucleus. The DCN are organized by both somatotopy and modality, and have a diverse range of afferent inputs and projection targets. The functional organization and connectivity of the DCN implicate them in a variety of sensorimotor functions, beyond their commonly accepted role in processing and transmitting somatosensory information to the thalamus, yet this is largely underappreciated in the literature. To consolidate insights into their sensorimotor functions, this review examines the morphology, organization, and connectivity of the DCN and their associated nuclei. First, we briefly discuss the receptors, afferent fibers, and pathways involved in conveying tactile and proprioceptive information to the DCN. Next, we review the modality and somatotopic arrangements of the remaining constituents of the DCN-complex. Finally, we examine and discuss the functional implications of the myriad of DCN-complex projection targets throughout the diencephalon, midbrain, and hindbrain, in addition to their modulatory inputs from the cortex. The organization and connectivity of the DCN-complex suggest that these nuclei should be considered a complex integration and distribution hub for sensorimotor information.

Electro-cutaneous stimulation on the palm elicits referred sensations on intact but not on amputated digits

OBJECTIVE

Grasping and manipulation control critically depends on tactile feedback. Without this feedback, the ability for fine control of a prosthesis is limited in upper limb amputees. Early studies have shown that non-invasive electro-cutaneous stimulation (ES) can induce referred sensations that are spread to a wider and/or more distant area, with respect to the electrodes. Building on this, we sought to exploit this effect to provide somatotopically matched sensory feedback to people with partial hand (digital) amputations.

APPROACH

For the first time, this work investigated the possibility of inducing referred sensations in the digits by activating the palmar nerves. Specifically, we electrically stimulated 18 sites on the palm of non-amputees to evaluate the effects of sites and stimulation parameters on modality, magnitude, and location of the evoked sensations. We performed similar tests with partial hand amputees by testing those sites that had most consistently elicited referred sensations in non-amputees.

MAIN RESULTS

We demonstrated referred sensations in non-amputees from all stimulation sites in one form or another. Specifically, the stimulation of 16 of the 18 sites gave rise to reliable referred sensations. Amputees experienced referred sensations to unimpaired digits, just like non-amputees, but we were unable to evoke referred sensations in their missing digits: none of them reported sensations that extended beyond the tip of the stump.

SIGNIFICANCE

The possibility of eliciting referred sensations on the digits may be exploited in haptic systems for providing touch sensations without obstructing the fingertips or their movements. The study also suggests that the phenomenon of referred sensations through ES may not be exploited for partial hand prostheses, and it invites researchers to explore alternative approaches. Finally, the results seem to confirm previous studies suggesting that the stumps in partial hand amputees partially acquire the role of the missing fingertips, physiologically and cognitively.

Somatosensory brainstem, thalamus, and cortex of the California sea lion (Zalophus californianus)

Pinnipeds (sea lions, seals, and walruses) are notable for many reasons, including their ape-sized brains, their adaptation to a coastal niche that combines mastery of the sea with strong ties to land, and the remarkable abilities of their trigeminal whisker system. However, little is known about the central nervous system of pinnipeds. Here we report on the somatosensory areas of the nervous system of the California sea lion (Zalophus californianus). Using stains for Nissl, cytochrome oxidase, and vesicular glutamate transporters, we investigated the primary somatosensory areas in the brainstem, thalamus, and cortex in one sea lion pup and the external anatomy of the brain in a second pup. We find that the sea lion's impressive array of whiskers is matched by a large trigeminal representation in the brainstem with well-defined parcellation that resembles the barrelettes found in rodents but scaled upward in size. The dorsal column nuclei are large and distinct. The ventral posterior nucleus of the thalamus has divisions, with a large area for the presumptive head representation. Primary somatosensory cortex is located in the neocortex just anterior to the main vertical fissure, and precisely locating it as we do here is useful for comparing the highly gyrified pinniped cortex with that of other carnivores. To our knowledge this work is the first comprehensive report on the central nervous system areas for any sensory system in a pinniped. The results may be useful both in the veterinary setting and for comparative studies related to brain evolution.

Functional and structural organization of the forelimb representation in cuneate nucleus in rat

We examined the physiological representation of the forelimb in the cuneate nucleus (CN) of forelimb-intact young adult rats (n=38) as the first part in a series of studies aimed at understanding the possible role that CN plays in delayed cortical reorganization that follows forelimb amputation. Metabolic labeling with cytochrome oxidase (CO) and electrophysiological mapping were used to examine the relationship between the structural and functional organization of CN. CN is a cylinder-shaped structure that lies bilaterally in the brainstem and extends nearly 4mm in the rostrocaudal direction. The forelimb is represented along the rostrocaudal extent. CN contains three zones; the rostral and caudal zones receive input largely from deep muscle and joint receptors and a middle zone, in the vicinity of the obex, receives input primarily from cutaneous receptors in the skin. The middle zone is somatotopically organized with the glabrous digits represented centrally, bordered on the medial side by ulnar wrist, ulnar forearm, and posterior upper arm representations; on the lateral side by radial wrist, radial forearm, and anterior upper arm representations; and on dorsal side by the dorsal digits and dorsal hand. The middle zone also contains well-defined CO-filled glomerular structures, called barrelettes, which are located within a homogenously stained field. The barrelettes are associated with the representation of the glabrous digits, with D5 represented most dorsal followed sequentially in a ventral-to-lateral direction by the representation of D4, D3, D2, and D1. The digit representations are topographically organized with the distal digit surface represented laterally with respect to the more medially lying proximal digit surface. The digit and palmar pads are also represented by barrelettes located on the medial side of CN. In contrast, the dorsal digit surfaces are represented dorsally and the dorsal hand is represented directly beneath the cuneate fasciculus, in a region devoid of barrelettes. The representations of the ulnar and radial wrist, forearm, and upper arm also lie within the homogeneously stained field in CN. The forelimb representation is bordered on the medial side by representation of trunk and hindlimb, and on the lateral side by representation of shoulder, ear, and head. While the present findings support and extend previous electrophysiological and anatomical studies of CN in the rat, they also provide a detailed physiological description of the functional organization of CN that is necessary for subsequent understanding of the functional reorganization of CN that may result following forelimb amputation.

The chemo- and somatotopic architecture of the Galago cuneate and gracile nuclei

The pattern of peripheral nerve inputs into the dorsal column nuclei, cuneate and gracile, was investigated in the prosimian Galago garnetti. The major findings were, that there is a greater segregation of the inputs from the fingers/hand within the cuneate compared with input form the toes/foot within the gracile. In both nuclei, cell clusters can be identified as cytochrome oxidase dense blotches, reactive also for the activity-dependent enzyme nitric oxide synthase. In the cuneate, cell clusters were apparent as six main cytochrome oxidase/nitric oxide synthase-reactive ovals arranged in a medial to lateral sequence. In contrast in the gracile, a higher degree of parcellation was noted and several cytochrome oxidase/nitric oxide synthase blotches were distributed along the rostrocaudal axis of the nucleus. This different architecture parallels differences in the organization of the inputs from the hand and from the foot. In the cuneate, cholera toxin B subunit conjugated to horseradish peroxydase labeled terminals from the glabrous and hairy skin of digits d1 to d5 segregated in each of the five most lateral cytochrome oxidase/nitric oxide synthase blotches. Afferents from the thenar, palmar pads and hypothenar overlapped with those from digit 1, digit 2 to digit 4 and digit 5, respectively. Inputs from wrist arm and shoulder were segregated in the most medial blotch. In the gracile, multiple foci of cholera toxin B subunit conjugated to horseradish peroxydase labeled terminals were observed upon injections of single sites in the toes or plantar pads. Although in multiple foci, inputs from different toes segregated from one another as well. Terminals from the plantar pads appeared to converge on the same cytochrome oxidase/nitric oxide synthase blotches targeted by inputs from the toes. In both the cuneate and the gracile, cytochrome oxidase/nitric oxide synthase blotches also presented intense immunoreactivity for GABA, calbindin, parvalbumin, and brain derived neurotrophic factor. Finally, in the cuneate the cell cluster region presented similarities in prosimian galagos and four species of New World monkeys, whereas it appeared more differentiated and complex in the Old Word macaque monkeys. In conclusion, the different pattern of segregation of the inputs from the hand and from the foot can be related to the different metabolic organization of the cuneate and of the gracile, respectively.

Adaptive responses of monkey somatosensory cortex to peripheral and central deafferentation

This study deals with two kinds of activity-dependent phenomena in the somatosensory cortex of adult monkeys, both of which may be related: (1) mutability of representational maps, as defined electrophysiologically; (2) alterations in expression of genes important in the inhibitory and excitatory neurotransmitter systems. Area 3b of the cerebral cortex was mapped physiologically and mRNA levels or numbers of immunocytochemically stained neurons quantified after disrupting afferent input peripherally by section of peripheral nerves, or centrally by making lesions of increasing size in the somatosensory thalamus. Survival times ranged from a few weeks to many months. Mapping studies after peripheral nerve lesions replicated results of previous studies in showing the contraction of representations deprived of sensory input and expansion of adjacent representations. However, these changes in representational maps were in most cases unaccompanied by significant alterations in gene expression for calcium calmodulin-dependent protein kinase isoforms, for glutamic acid decarboxylase, GABA(A) receptor subunits, GABA(B) receptors, alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) or N-methyl-D-aspartate (NMDA) receptor subunits. Mapping studies after lesions in the ventral posterior lateral nucleus (VPL) of the thalamus revealed no changes in cortical representations of the hand or fingers until >15% of the thalamic representation was destroyed, and only slight changes until approximately 45% of the representation was destroyed, at which point the cortical representation of the finger at the center of a lesion began to shrink. Lesions destroying >60% of VPL resulted in silencing of the hand representation. Although all lesions were associated with a loss of parvalbumin-immunoreactive thalamocortical fiber terminations, and of cytochrome oxidase staining in a focal zone of area 3b, no changes in gene expression could be detected in the affected zone until >40-50% of VPL was destroyed, and even after that changes in mRNA levels or in numbers of GABA-immunoreactive neurons in the affected zone were remarkably small. The results of these studies differ markedly from the robust changes in gene expression detectable in the visual cortex of monkeys deprived of vision in one eye. The results confirm the view that divergence of the afferent somatosensory pathways from periphery to cerebral cortex is sufficiently great that many fibers can be lost before neuronal activity is totally silenced in area 3b. This divergence is capable of maintaining a high degree of cortical function in the face of diminishing inputs from the periphery and is probably an important element in promoting representational plasticity in response to altered patterns of afferent input.

Functional Connections Formed by Saphenous Nerve Terminal Sprouts in the Dorsal Horn Following Neonatal Sciatic Nerve Section

The rostrocaudal distribution of saphenous nerve inputs into the lumbar dorsal horn from L2 to L6 has been investigated in urethane anaesthetized rats whose left sciatic nerve was cut and ligated at birth. In normal cord, electrical stimulation of the saphenous nerve evoked dorsal horn spikes in L2 to caudal L4. Few or no spikes were evoked in L5. After neonatal sciatic nerve section, saphenous nerve stimulation evoked spikes throughout segments L2 to L6. Dorsal horn cell receptive fields were also altered following neonatal sciatic nerve section. A somatotopic map of the lumbar enlargement in normal rats was constructed from the receptive fields (RFs) of adjacent dorsal horn cells. Cells with RFs in the saphenous skin region were concentrated in L3 and rostral L4 and very few were found in L5. After neonatal sciatic nerve section, however, a substantial number of cells with low threshold saphenous skin RFs were also found in caudal L4 and throughout L5. These results show that the central saphenous nerve terminal sprouts that grow into the sciatic terminal region following neonatal sciatic nerve section (Fitzgerald, 1985, J. Comp. Neurol., 240, 414-422; Fitzgerald et al., 1990, J. Comp. Neurol., 300, 370-385) form functional connections. This results in dorsal horn cells that are not normally influenced by saphenous nerve inputs developing substantial low threshold RFs in saphenous nerve skin regions.

The neural basis of perceptual learning

Perceptual learning is a lifelong process. We begin by encoding information about the basic structure of the natural world and continue to assimilate information about specific patterns with which we become familiar. The specificity of the learning suggests that all areas of the cerebral cortex are plastic and can represent various aspects of learned information. The neural substrate of perceptual learning relates to the nature of the neural code itself, including changes in cortical maps, in the temporal characteristics of neuronal responses, and in modulation of contextual influences. Top-down control of these representations suggests that learning involves an interaction between multiple cortical areas.

Cortical and subcortical contributions to activity-dependent plasticity in primate somatosensory cortex

After manipulations of the periphery that reduce or enhance input to the somatosensory cortex, affected parts of the body representation will contract or expand, often over many millimeters. Various mechanisms, including divergence of preexisting connections, expression of latent synapses, and sprouting of new synapses, have been proposed to explain such phenomena, which probably underlie altered sensory experiences associated with limb amputation and peripheral nerve injury in humans. Putative cortical mechanisms have received the greatest emphasis but there is increasing evidence for substantial reorganization in subcortical structures, including the brainstem and thalamus, that may be of sufficient extent to account for or play a large part in representational plasticity in somatosensory cortex. Recent studies show that divergence of ascending connections is considerable and sufficient to ensure that small alterations in map topography at brainstem and thalamic levels will be amplified in the projection to the cortex. In the long term, slow, deafferentation-dependent transneuronal atrophy at brainstem, thalamic, and even cortical levels are operational in promoting reorganizational changes, and the extent to which surviving connections can maintain a map is a key to understanding differences between central and peripheral deafferentation.

Evidence for brainstem and supra-brainstem contributions to rapid cortical plasticity in adult monkeys

Cortical maps can undergo amazingly rapid changes after injury of the body. These changes involve functional alterations in normal substrates, but the cortical and/or subcortical location(s) of these alterations, and the relationships of alterations in different substrates, remain controversial. The present study used neurophysiological approaches in adult monkeys to evaluate how brainstem organization of tactile inputs in the cuneate nucleus (CN) changes after acute injury of hand nerves. These data were then compared with analogous data from our earlier cortical area 3b studies, which used the same approaches and acute injury, to assess relationships of cuneate and cortical changes. The results indicate that cuneate tactile responsiveness, receptive field locations, somatotopic organization, and spatial properties of representations (i.e., location, continuity, size) change during the first minutes to hours after injury. The comparisons of cuneate and area 3b organization further show that some cuneate changes are preserved in area 3b, whereas other cuneate changes are transformed before being expressed in area 3b. The findings provide evidence that rapid reorganization in area 3b, in part, reflects mechanisms that operate from a distance in the cuneate nucleus and, in part, reflects supracuneate mechanisms that modify brainstem changes.

Adult cortical dynamics

There are many influences on our perception of local features. What we see is not strictly a reflection of the physical characteristics of a scene but instead is highly dependent on the processes by which our brain attempts to interpret the scene. As a result, our percepts are shaped by the context within which local features are presented, by our previous visual experiences, operating over a wide range of time scales, and by our expectation of what is before us. The substrate for these influences is likely to be found in the lateral interactions operating within individual areas of the cerebral cortex and in the feedback from higher to lower order cortical areas. Even at early stages in the visual pathway, cells are far more flexible in their functional properties than previously thought. It had long been assumed that cells in primary visual cortex had fixed properties, passing along the product of a stereotyped operation to the next stage in the visual pathway. Any plasticity dependent on visual experience was thought to be restricted to a period early in the life of the animal, the critical period. Furthermore, the assembly of contours and surfaces into unified percepts was assumed to take place at high levels in the visual pathway, whereas the receptive fields of cells in primary visual cortex represented very small windows on the visual scene. These concepts of spatial integration and plasticity have been radically modified in the past few years. The emerging view is that even at the earliest stages in the cortical processing of visual information, cells are highly mutable in their functional properties and are capable of integrating information over a much larger part of visual space than originally believed.

Reorganization of the raccoon cuneate nucleus after peripheral denervation

The effects of peripheral nerve transection on the cuneate nucleus were studied in anesthetized raccoons using extracellular, single-unit recordings. The somatotopic organization of the cuneate nucleus first was examined in intact, control animals. The cuneate nucleus in the raccoon is organized with the digits represented in separate cell clusters. The dorsal cap region of the cuneate nucleus contains a representation of the claws and hairy skin of the digits. Within the representation of the glabrous skin, neurons with rapidly adapting properties tended to be segregated from those with slowly adapting properties. The representations of the distal and proximal pads on a digit also were segregated. Electrical stimulation of two adjacent digits provided a detailed description of the responses originating from the digit that contains the tactile receptive field (the on-focus digit) and from the adjacent (off-focus) digit. Stimulation of the on-focus digit produced a short latency excitation in all 99 neurons tested, with a mean of 10.5 ms. These responses had a low threshold (426 microA). Stimulation of an off-focus digit activated 65% of these neurons. These responses had a significantly longer latency (15.3 ms) than on-focus responses and the threshold was more than twice as large. Two to five months after amputation of digit 4, 97 cells were tested with stimulation of digits 3 and 5. A total of 44 were in the intact regions of the cuneate nucleus. They had small receptive fields on intact digits and their responses to electrical stimulation did not differ from the control neurons. The remaining 53 neurons were judged to be deafferented and in the fourth digit region on the basis of their location with respect to intact neurons. All but two of these cells had receptive fields that were much larger than normal, often including more than one digit and part of the palm. When compared with the off-focus control neurons, their responses to electrical stimulation had lower thresholds and an increased response probability and magnitude. The latencies of these cells did not decrease, however, and were the same as the off-focus control values. The enhanced responses of the deafferented neurons to adjacent digit stimulation indicate that there is a strengthening of synapses that were previously ineffective. The increased proportion of neurons that could be activated after amputation suggests that there is also a growth of new connections. This experiment demonstrates that reorganization in the adult somatotopic system does occur at the level of the dorsal column nuclei. As a consequence, many of the changes reported at the cortex and thalamus may be due to the changes occurring at this first synapse in the somatosensory pathway.

Evolution of morphological and histochemical changes in the adult cat cuneate nucleus following forelimb denervation

Morphological and histochemical changes were studied in the ipsilateral cuneate nucleus between one and 52 weeks after forelimb denervation in adult cats. The deafferented nucleus and neighboring fasciculus were noticeably reduced in size within four weeks and decreased further by 13 weeks. The intensity of acetylcholinesterase staining decreased within one week and was further reduced one month after nerve transections. This reduction in acetylcholinesterase staining was transient, approaching control levels within one year. Parvalbumin immunostaining was also altered by the nerve transections; on the deafferented side, the neuropil staining in the cuneate nucleus and fasciculus decreased, but the number of parvalbumin-positive cells was consistently greater than in the contralateral side. These cell counts returned to normal levels within one year. One month after the injury, cytochrome oxidase activity was reduced. This reduction persisted and was even more apparent after one year. In parallel, the cell clusters of the nucleus became progressively less distinct. These observations in an adult mammal indicate that peripheral nerve injury imposes molecular and morphological changes on second-order sensory neurons which evolve differentially with time. Although some changes developed rapidly after deafferentation, the onset of others was slower; and whereas some seemed irreversible, others eventually regressed. Taken together with the functional studies of others, these findings suggest that early molecular changes observed in cuneate neurons reflect adaptive reactions to lesion-induced alterations in afferent activity. Permanent deprivation of the normal input, however, would eventually lead to chronic, and perhaps irreversible, degenerative changes.

The immediate effects of peripheral deafferentation on neurons of the cuneate nucleus in raccoons

Single-unit recordings were obtained from 42 neurons in the cuneate nucleus of 12 anesthetized raccoons. All neurons had receptive fields on the glabrous skin of a forepaw digit. Temporary removal of the dominant excitatory input to a neuron, by injection of lidocaine into the base of the digit, did not result in any expansion of the excitatory receptive field onto adjacent, "off-focus" digits. Similarly, the responses evoked from the off-focus digits by electrical stimulation, which had a longer latency and a higher threshold, were not improved during the lidocaine block. Inhibition was produced in the majority of neurons by high-intensity mechanical stimulation of the off-focus digits, but this was also unchanged when the dominant excitatory input to the neurons was blocked. Since this from of inhibition is not apparent in the somatosensory thalamus before denervation, the spontaneous activity of thalamic neurons must be controlled by inputs other than the cuneate nucleus. These results also indicate that the long-term reorganization seen in the thalamus and cortex is not attributable to a simple unmasking of connections from the adjacent digits within the cuneate nucleus, but may involve strengthening of the connections responsible for longer-latency responses. The only significant change induced in cuneate neurons by temporary denervation was a decrease in the firing rates of 69% of the neurons that had spontaneous activity. Since it is unlikely that any of the large-diameter afferents from touch receptors can account for this finding, mechanically insensitive afferent fibers from the digit may contribute to the spontaneous activity of cuneate neurons, either directly or via a relay in the spinal cord.

Changes in the organization of the ventroposterior lateral thalamic nucleus after digit removal in adult raccoon

The ventroposterior lateral nucleus of the thalamus was studied in seven raccoons that had undergone amputation of the fourth digit between 2 and 5 months previously. Extracellular recordings were made in a series of closely spaced penetrations through the thalamus in chloralose anesthetized animals. The responses to cutaneous stimulation of the forepaw were used to reconstruct the somatotopic organization of the thalamus and to identify recording sites believed to be located in the digit zone that had lost its peripheral input. Twelve penetrations that passed through both of the adjacent fifth and third digit regions were analyzed in detail to delineate this deafferented region. None of the recording sites in this region were completely silent, indicating that the deafferented thalamus had undergone significant reorganization of its inputs. At most sites, the neurons had receptive fields on the skin surrounding the amputation wound and including one of the adjacent digits. Approximately half of the sites had low thresholds in the range of normal thalamic neurons. These results indicate that the ventroposterior thalamus is capable of substantial reorganization, which may account for much of the reorganization seen in somatosensory cortex.

Reorganisation of primary afferent nerve terminals in the brainstem after peripheral nerve injury. An anatomical study in cats

A pure sensory nerve (the superficial branch of the radial nerve) in adult cats was cut to investigate the changes in the nerve endings (terminals) on the neurons of the nucleus cuneatus of the brainstem. In one group of cats (n = 22) the ends of the cut nerve were approximated immediately by epineural suturing to promote optimum regeneration. In another group (n = 11) the proximals tump of the nerve was enclosed in a capsule to prevent regeneration. Four to 17 months later the same nerve was re-exposed. The sutured nerves were cut and nerve-tracer was exhibited to the proximal end of the cut nerves and to the proximal stump of the nerves which had been encapsulated. The purpose was to investigate the labelling of nerve terminals in the cuneate nucleus, because it receives an input of primary afferents from the front leg. The nerve and the cuneate nucleus of the opposite side served as controls. Labelled terminals were distributed throughout the dorsal part of the entire rostrocaudal extent of the cuneate nucleus. The distribution was patchy and was superimposed on clusters of nerve cells. The quantity of labelled nerve terminals on the experimental and control sides was compared: 60% of the labelling observed on the control side was in the sutured nerves while the encapsulated nerves exhibited only 32%. This difference was apparent 4 months after transection of the nerve. Up to 17 months after the nerve was cut, however, there was some increase in the quantity of labelled nerve terminals and this was most apparent in cats in which the nerves had been sutured.

Effect of vibrissae deprivation on follicle innervation, neuropeptide synthesis in the trigeminal ganglion, and S1 barrel cortex plasticity

Deprivation of vibrissae from an early age causes plasticity in S1 barrel cortex. This method of deprivation is most likely to induce plasticity by altering the balance of primary afferent activity from the deprived and spared vibrissae. To study whether or not induction or expression of this type of plasticity might be affected by follicle nerve injury caused by the deprivation technique, three different methods of detecting nerve injury were used: counting axon numbers in the distal follicle nerve, quantifying morphological changes in axons, and measuring neuropeptide expression in the trigeminal ganglion cells. First, nerves innervating follicles chronically deprived of vibrissae from birth had the same number of myelinated and unmyelinated axons as nerves from normally reared animals. Second, axons innervating deprived follicles showed no morphological changes in myelination or mitochondria characteristic of damaged nerves. Third, the corresponding nerve cell bodies in the trigeminal ganglion did not show upregulation of galanin or neuropeptide Y expression. In contrast, animals receiving mild injury of the follicle nerve endings (by cauterization of the follicle) showed profound changes in axonal myelination and mitochondria and increases in neuropeptide expression. These results imply that vibrissae deprivation does not act by inducing injury of the follicular nerve, suggesting that changes in the balance of follicle nerve activity are the cause of cortical plasticity. Consistent with this notion, a fourth experiment demonstrated that trimming the vibrissae induces cortical plasticity comparable to that induced by complete vibrissae removal.

Decrease and long-term recovery of choline acetyltransferase immunoreactivity in adult cat somatosensory cortex after peripheral nerve transections

The functional reorganization of cerebral cortex following peripheral deafferentation is associated with changes in a number of neurotransmitters and related molecules. Acetylcholine (ACh) enhances neuronal responsiveness and could play a role in activity-dependent cortical plasticity. In this study, choline acetyltransferase (ChAT) immunohistochemistry was used to investigate ACh innervation of the primary somatosensory cortex in cats sustaining complete unilateral forearm and paw denervations. Survival times of 2-52 weeks were examined. The deafferented contralateral cortex was defined electrophysiologically, and quantitative estimates of ChAT-immunoreactive fiber density were obtained from the forelimb and hindlimb sectors of area 3b in both hemispheres. In the 3b forelimb sector contralateral to the deafferentation, a decrease in density of ChAT-positive fibers relative to the ipsilateral hemisphere was apparent at 2 weeks and most pronounced at 13 weeks, involving all cortical layers except layer I. There was no such decrease in the hindlimb sector, but the loss of ChAT immunoreactivity extended to sectors representing proximal forelimb and trunk. Changes in ChAT immunoreactivity were no longer found after 1 year of survival. This long-lasting but reversible lowering of ChAT immunoreactivity could result from a loss of afferent activity in basalis neurons and/or trophic influences retrogradely exerted by cortex on these cells. Reduced ACh transmission might then contribute to the loss of gamma aminobutyric acid (GABA) inhibition in the deafferented cortex by decreasing the activation of inhibitory interneurons. The long-term recovery of a normal ChAT immunoreactivity in cortex could be a consequence of its functional reorganization.

Ultrastructural and immunocytochemical characterization of primary afferent terminals in the rat cuneate nucleus

The cuneate nucleus is a relay center for somatosensory information by receiving tactile and proprioceptive inputs from primary afferent fibers that ascend in the dorsal funiculus. The morphology, synaptic contacts, and neurochemical content of primary afferent terminals in the cuneate nucleus of rats were investigated by combining anterograde transport of horseradish peroxidase conjugated to wheat-germ agglutinin or to cholera toxin (injected in cervical dorsal root ganglia) with postembedding immunogold labeling for glutamate and GABA. Both tracers gave similar results. Two types of terminals were labeled: type I terminals were irregularly shaped, had a mean area of 4.0 microns 2, synapsed on several dendrites, and were contacted by other terminals, some of which were GABA positive. Type II terminals were dome-shaped, had a mean area of 2.18 microns 2, and made synaptic contact on a single dendrite. All the anterogradely labeled terminals (interpreted as endings of primary afferents) were enriched in glutamate but not in GABA. The finding that identified primary afferent terminals are enriched in glutamate with respect to other tissue profiles strongly suggests a neurotransmitter role for glutamate in this afferent pathway to the rat cuneate nucleus.

Effect of unilateral partial cochlear lesions in adult cats on the representation of lesioned and unlesioned cochleas in primary auditory cortex

We examined the effect of unilateral restricted cochlear lesions in adult cats on the topographic representations ("maps") of the lesioned and unlesioned cochleas in the primary auditory cortex (AI) contralateral to the lesioned cochlea. Frequency (tonotopic) maps were derived by conventional multineuron mapping procedures in anesthetized animals. In confirmation of a study in adult guinea pigs (Robertson and Irvine [1989] J. Comp. Neurol. 282:456-471), we found that 2-11 months after the unilateral cochlear lesion the map of the lesioned cochlea in the contralateral AI was altered so that the AI region in which frequencies with lesion-induced elevations in cochlear neural sensitivity would have been represented was occupied by an enlarged representation of lesion-edge frequencies (i.e., frequencies adjacent to those with elevated cochlear neural sensitivity). Along the tonotopic axis of AI the total representation of lesion-edge frequencies could extend up to approximately 2.6 mm rostal to the area of normal representation of these frequencies. There was no topographic order within this enlarged representation. Examination of threshold sensitivity at the characteristic frequency (CF, frequency to which the neurons were most sensitive) in the reorganized regions of the map of the lesioned cochlea established that the changes in the map reflected a plastic reorganization rather than simply reflecting the residue of prelesion input. In contrast to the change in the map of the lesioned contralateral cochlea, the map of the unlesioned ipsilateral cochlea did not differ from those in normal animals. Thus, in contrast to the normal very good congruency between ipsilateral and contralateral AI maps, in the lesioned animals ipsilateral and contralateral maps differed in the region of AI in which there had been a reorganization of the map of the lesioned cochlea. Outside the region of contralateral map reorganization, ipsilateral and contralateral AI maps remained congruent within normal limits. The difference between the two maps in the region of contralateral map reorganization suggested, in light of the physiology of binaural interactions in the auditory pathway, that the cortical reorganization reflected subcortical changes. Finally, response properties of neuronal clusters within the reorganized map of the lesioned cochlea were compared to normative data with respect to threshold sensitivity at CF, the size of frequency "response areas," and response latencies. In the majority of cases, CF thresholds were similar to normative data. The frequency "response areas" were slightly less sharply tuned than normal, but not significantly. Response latencies were significantly shorter than normal in three animals and significantly longer in one animal.

Representation of the body surface in the gracile, cuneate, and spinal trigeminal nuclei of the little red flying fox (Pteropus scapulatus)

The body surface representation in the gracile, cuneate, and spinal trigeminal nuclei of the little red flying fox (Pteropus scapulatus) was examined. As in other species, it was found that any single cross-section through all three nuclei contains a representation of most, or all, of the body surface. In the little red flying fox, however, this representation is arranged as a series of dorsolateral to ventromedially oriented bands, within which there are no apparent topographies. These bands are arranged in such a way that the spatial relationships between body regions in the representation do not reflect those at the periphery.

Deafferentation-induced terminal field expansion of myelinated saphenous afferents in the adult rat dorsal horn and the nucleus gracilis following pronase injection of the sciatic nerve

We have previously demonstrated sprouting of small diameter saphenous afferents, labelled with wheat germ agglutinin conjugated with horseradish peroxidase (WGA-HRP) and (HRP), into the sciatic territory of the adult rat superficial dorsal horn following destruction of sciatic afferents by injection of the sciatic nerve with pronase (a combination of proteolytic enzymes). In the present experiments, we examined the response of myelinated saphenous axons, which terminate in lamina I and the deep dorsal horn (laminae III-V) under the same conditions, with the tracer B subunit of cholera toxin conjugated to HRP (B-HRP) which specifically labels myelinated primary afferents when injected into a peripheral somatic nerve. We also examined changes in the nucleus gracilis, another site of sciatic degeneration and a target of saphenous afferents. Four months after injection of the pronase, the area of label determined by measurement of the width of the saphenous territory in lamina III was expanded by 24% on the pronase side. Since there was also expansion throughout the deep dorsal horn, the area measured by tracing the labelled region in transverse sections was actually twice that of the control side, and the intensity of labelling within the traced area increased by 18%. There was no change in grey matter area due to the lesion. The traced area of labelling in the nucleus gracilis increased by 40%, and increased in intensity by 17%. The substantia gelatinosa is not normally supplied by B-HRP-labelled afferents, and there was no expansion of these sprouted saphenous afferents into the gelatinosa. These results indicate that myelinated afferents can sprout as vigorously in lamina I and the deep dorsal horn as the small diameter afferents do in the substantia gelatinosa; that there is no invasion of the substantia gelatinosa by the myelinated afferents at least as long as the small diameter afferents also have the opportunity to sprout; and that primary afferents have the potential to sprout at more than one site of termination, i.e., both the dorsal horn and the dorsal column nuclei.

The immediate effects of peripheral denervation on inhibitory mechanisms in the somatosensory thalamus

Multiunit recordings were made in the ventroposterior lateral nucleus of the thalamus in anesthetized raccoons. During recording from cells responding to cutaneous stimulation of a forepaw digit, the corresponding digit was denervated permanently (by cutting its four digital nerves) or temporarily (by injecting lidocaine into the base of the digit). Both procedures resulted in immediate increases in the inhibition that could be induced by stimulation of the adjacent digits when the original cutaneous receptive field was on the glabrous skin. In each case with temporary denervation, this enhanced off-focus inhibition decreased when the excitatory responses returned to normal. In contrast, temporary denervation of the digit during recording at sites in the hairy skin representation did not reveal this increased inhibition from adjacent digits. When capsaicin was applied to the digital nerves in two animals, the excitatory receptive fields of thalamic neurons increased in area, but were still restricted to the same part of the digit. These data indicate that the immediate unmasking of inhibitory responses, previously reported in primary somatosensory cortex of the raccoon, is also present in the thalamus. The capsaicin-induced expansion of excitatory receptive fields confirms previous experiments in other species, and suggests that C fibers play a role in modulating the size of cutaneous receptive fields. However, the enlargement of excitatory receptive fields by capsaicin is much less than the unmasking of inhibitory fields induced by digit denervation, and indicates that different mechanisms are involved in controlling these various inputs to thalamic neurons.

Current hypotheses of structural pattern formation in the somatosensory system and their potential relevance to humans

Current hypotheses of structural pattern formation in the mammalian somatosensory system are modeled on experimental findings from the trigeminal system of rodents. The present results show that, like rodents, the trigeminal nucleus principalis of humans contains a parcellated pattern of cytochrome oxidase dense patches. These results provide an indication of the potential usefulness of rodent-based hypotheses for understanding pattern formation in human somatosensory connections.

Rapid reorganization of cortical maps in adult cats following restricted deafferentation in retina

The retinotopic map in the visual cortex of adult mammals can reorganize in response to a small injury in a restricted region of retina. Although the mechanisms underlying this neural plasticity in adults are not well understood, it is possible that rapid, adaptive alterations in the effectiveness of existing connections play a key role in the reorganization of cortical topography following peripheral deafferentation. In order to test this hypothesis, a small retinal lesion was made in one eye of adult cats and the visual cortex was mapped before and immediately after enucleating the non-lesioned eye. We found that substantial reorganization takes place within hours of enucleation.

Chronic deafferentation in monkeys differentially affects nociceptive and nonnociceptive pathways distinguished by specific calcium-binding proteins and down-regulates gamma-aminobutyric acid type A receptors at thalamic levels

Chronic deafferentation of skin and peripheral tissues is associated with plasticity of representational maps in cerebral cortex and with perturbations of sensory experience that include severe "central" pain. This study shows that in normal monkeys the nonnociceptive, lemniscal component of the somatosensory pathways at spinal, brainstem, and thalamic levels is distinguished by cells and fibers immunoreactive for the calcium-binding protein parvalbumin, whereas cells of the nociceptive component at these levels are distinguished by immunoreactivity for 28-kDa calbindin. Long-term dorsal rhizotomies in monkeys lead to transneuronal degeneration of parvalbumin cells at brainstem and thalamic sites accompanied in the thalamus by a down-regulation of gamma-aminobutyric acid type A receptors and an apparent increase in activity of calbindin cells preferentially innervated by central pain pathways. Release from inhibition and imbalance in patterns of somatosensory inputs from thalamus to cerebral cortex may constitute subcortical mechanisms for inducing changes in representational maps and perturbations of sensory perception, including central pain.

Neuronal response properties within subregions of raccoon somatosensory cortex 1 week after digit amputation

Multiple penetrations in the somatosensory cortex of three anesthetized raccoons 1 week following amputation of the fourth digit provided detailed information about somatotopy and neuronal responsiveness in the deafferented cortex. Recordings in a total of 601 penetrations (292 in deafferented cortex and 309 in the surrounding cortex) were compared with those from intact control animals described previously (Rasmusson et al., 1991). The level of spontaneous activity increased within the deafferented cortex, with 42% of the sites having high or moderate levels of spontaneous activity, in comparison with 18% in control animals. There was also an increase in the incidence of inhibitory responses to stimulation of adjacent digits (26% of the penetrations vs. 10% in control animals), confirming previous findings. These two variables, increased spontaneous activity and the presence of strong lateral inhibition, were highly correlated in individual penetrations. An unexpected finding was that the cortex representing the intact parts of forepaw was also disrupted with respect to these two measures, suggesting that amputation had an effect outside the deafferented region. In contrast, response properties that are more clearly a reflection of information processing in the dorsal column-medial lemniscal pathway (adaptation and threshold) were altered only within the deafferented region. The deafferented region was not homogeneous immediately after amputation, but consisted of a radically affected core region and a slightly affected fringe adjacent to the intact representations. This inhomogeneity had also been apparent with partial digit deafferentation, reported previously. The fringe, approximately 1 mm in width, may reflect overlapping projections from adjacent digits at one or more levels of the somatosensory pathway. Since the size of the fringe is similar to the maximum extent of reorganization found in other models of reorganization, the mechanisms of plasticity within this region may involve an unmasking of pre-existing synapses with slight modification in synaptic strength. However, the plasticity within the core region of the raccoon seen in these experiments, which may be 5 mm from nondeafferented cortex, requires more extensive changes, perhaps via polysynaptic pathways.

Parcellated organization in the trigeminal and dorsal column nuclei of primates

The subdivisions of the brainstem trigeminal complex in non-primate mammals are characterized by aggregated or parcellated patterns of neural organization. The present studies used cytochrome oxidase histochemistry to test if parcellated organization patterns also occur in the brainstems of primates. The results demonstrate that a parcellated pattern of neural organization exists in the trigeminal nucleus principalis, but not in the spinal trigeminal nuclei, of macaque and squirrel monkeys. The results further suggest that parcellation in the nucleus principalis qualitatively resembles the aggregated organization in dorsal column nuclei. Taken together with previous findings from non-primates, these results indicate that central parcellation is an organizational feature of specific ascending somatosensory projections in many mammals including primates.

Somatotopic organization of the dorsal column nuclei in the rat: transganglionic labelling with B-HRP and WGA-HRP

To analyze the patterns of cutaneous primary afferent fibers projecting to the dorsal column nuclei in the rat, horseradish peroxidase (HRP)-based tracers were injected intracutaneously into a number of discrete regions of the forelimbs and hindlimbs. Three-4 days following the HRP injections, the rats were perfused transcardially; 60 microns transverse sections were cut, and the HRP was reacted using the tetramethyl benzidine method. Comparisons were made of projections following injections with choleragenoid-conjugated horseradish peroxidase (B-HRP) or with wheat-germ agglutinin conjugated to horseradish peroxidase (WGA-HRP). B-HRP and WGA-HRP produced similar patterns of labelling, but B-HRP produced greater intensity of labelling and slightly larger projection areas. In the cuneate nucleus (CN), HRP labelling of primary afferents from small, delimited regions, e.g., from a portion of the skin of a single digit, appeared to be precisely restricted in rostrocaudally oriented columns, with little or no overlap (in the mediolateral and dorsoventral plane) into adjacent regions. With respect to rostrocaudal organization, a region in the CN containing a dense population of cutaneous primary afferent fibers appeared to be similar to the middle, or cluster, region in cats and in raccoons and the pars rotunda in primates. Projection patterns were very consistent from rat to rat, but their somatotopic organization differed from that suggested by electrophysiological studies: cutaneous afferents from forelimb digit 1 projected near the ventral border of the CN; those from digit 5 projected dorsomedially to those from digit 1; the projections from the remaining digits formed a crescent between the projections from digits 1 and 5. In the gracile nucleus, the organization of cutaneous afferent projections from hindlimb digits was more variable and complex than that found in the CN.

Central projections from the skin of the hand in squirrel monkeys

Central termination patterns of afferents from the hands of squirrel monkeys were studied after subdermal injections of wheat germ agglutinin conjugated with horseradish peroxidase (WGA-HRP) or cholera toxin subunit B conjugated to HRP (BHRP). WGA-HRP more effectively labeled axons terminating in the superficial dorsal horn of the spinal cord, while BHRP more effectively labeled axons terminating in the deeper layers. Injections of both tracers, when restricted to parts of glabrous digits, palm, or dorsal hand, revealed somatotopic patterns in the spinal cord and pars rotunda of the cuneate nucleus that were, in some respects, similar and, in other respects, quite different from those previously reported for macaque monkey (Florence et al., J. Comp. Neurol. 286:48-70, '89). As in macaques, injections in digits 1-5 produced a rostrocaudal sequence of foci of terminations in the cervical spinal cord. However, inputs from the palm were located medial to those from the digits, whereas the palm is represented lateral to the digits in macaque monkeys. Since inputs from the palm is also medial in the dorsal horn in cats (Nyberg and Blomqvist, J. Comp. Neurol. 242:28-39, '85), the condition in squirrel monkeys may be similar to the generalized state. In the cuneate nucleus, single injections in the hand produced dense label in the pars rotunda, and sparse label in the rostral and caudal poles. As in macaque monkeys, inputs from specific parts of the hand related to rostrocaudal clusters of cells that are cytochrome oxidase dense. The representation of the digits differed from macaques in that the digits were represented dorsal to the palm, rather that ventral to the palm as in macaques. Again, comparisons with cats suggest that squirrel monkeys have the more generalized pattern. Finally, inputs from the hair, dorsal surfaces of the digits terminated on the same clusters as the inputs from the glabrous, ventral surfaces, apparently overlapping somewhat. The proximity of these terminations from dorsal and ventral surfaces of the digits may be related to observations that cortical representations of the glabrous surfaces of digits become responsive to dorsal surfaces of the same digits when inputs from glabrous skin are chronically deactivated (e.g., Merzenich et al., Neuroscience 3:33-55, '83).

Chronic effects of total or partial digit denervation on raccoon somatosensory cortex

Electrophysiological recordings were made in the primary somatosensory cortex of anesthetized raccoons 14 to 169 days following digit amputation or 60 to 129 days after transection of the two nerves innervating the ventral surface of the fourth digit. The incidence of inhibitory responses decreased from 50% of the penetrations immediately after amputation to 35% over the first 3 weeks and to almost zero after 2 months. The number of sites with low-threshold excitatory responses increased from 4% to 14% to 50% during these same intervals. Initially, the excitatory fields were small and located over the nerve stumps, and were therefore probably due to direct stimulation of the damaged nerves. At 2 months after amputation, the excitatory receptive fields were large and diffuse. Although the size of receptive fields decreased during the later period (when the thresholds were also decreasing), there was no recovery of any precise somatotopic organization in the deafferented cortex. The reorganization process in the raccoon thus consists of at least two stages: The early stage is dominated by inhibitory connections, whereas the second involves a recovery and restructuring of excitatory inputs. From 2 to 4 months after partial digit denervation, there were only minor changes in response properties or somatotopic organization in the deafferented cortex as compared to immediately after nerve transection. Thus, few of the characteristics of reorganization induced by digit amputation were elicited by this treatment, which leaves some of the digit innervation intact. There was, however, an unexpected increase in the portion of the ventral digit that was able to activate the cortex, suggesting complexities in the peripheral innervation of the digit that need to be resolved.

Somatotopic organization of the cuneate nucleus in the rat: transganglionic labelling with WGA-HRP

A novel somatotopic map of primary cutaneous afferents projecting to the cuneate nucleus in the rat was determined by transganglionic transport of wheat-germ agglutinin conjugated to horseradish peroxidase and free horseradish peroxidase. Intracutaneous injections of tracer into different limited regions of the forelimb resulted in discrete areas of label for each injection site, with little or no overlap into other projection areas. The map of cutaneous projections onto the cuneate nucleus revealed by our anatomical tracing provided much more detail than any previous study in the rat, and demonstrated some significant differences from earlier maps based on electrophysiological recordings.

Acute effects of total or partial digit denervation on raccoon somatosensory cortex

The immediate effects of total or partial denervation of single digits (0-16 hr after nerve transection) on primary somatosensory cortex were studied electrophysiologically. Comparisons of response properties and cortical somatotopy were made between intact raccoons and four groups of raccoons with transection of some or all of the nerves innervating the fourth or fifth digit. Animals with all four digital nerves cut (amputation of the digit) were most different from normal. Approximately half of the penetrations in the affected cortical region showed inhibitory responses to stimulation of adjacent skin regions. These consisted of a strong response to stimulus offset and/or a suppression of spontaneous activity during indentation. Since these responses were substantially different from those recorded several months after digit amputation, additional changes in connectivity and synaptic strength must occur with chronic denervation. These inhibitory responses were not seen in animals with one, two, or three nerves cut per digit. In the animals with partial denervation of a digit, the greatest disruption occurred when both ventral nerves to the glabrous skin were transected. This yielded cell clusters with abnormally large receptive fields, disruptions in somatotopic organization, and a decreased occurrence of low-threshold responses. If only one nerve to glabrous skin was transected, there was less change, even if it was combined with transection of both nerves to hairy skin. These results suggest that the release of inhibitory responses in a cortical digital region by amputation is prevented by the retention of even one ventral nerve. None of the denervation conditions produced large nonresponsive areas of cortex, which would have indicated a loss of all inputs.

Somatotopic organization of inputs from the hand to the spinal gray and cuneate nucleus of monkeys with observations on the cuneate nucleus of humans

Central termination patterns of primary afferents from the hand and forelimb were studied following subdermal injections of HRP conjugates in macaque monkeys. In the middle layers of the dorsal horn of the spinal cord, afferents from digits 1-5 terminated in a rostrocaudal sequence in separate, elongated columns at cervical levels 5-7. Afferents from the glabrous digits extended to the medial margin of the dorsal gray, while afferents from the dorsal skin of the digits terminated more laterally. Afferents from the dorsal hand and palm terminated lateral to those from the digits, while inputs from the forearm occupied tissue rostral and caudal to the representation of the hand. In the cuneate nucleus, terminations from each digit formed an elongated column that was densely labelled in the central pars rotunda and sparsely labelled in both the rostral and caudal reticular poles. Within the pars rotunda, digits 1-5 were represented in order from lateral to medial. Inputs from the digit tips terminated ventral to inputs from the proximal digits. Afferents from the dorsal skin of the digits terminated in an even more dorsal position, while the most dorsal portion of the pars rotunda related to the glabrous and dorsal hand. Within the pars rotunda, terminations from specific parts of the hand overlapped parcellated clusters of neurons. These clusters were densely reactive for cytochrome oxidase (CO) and were surrounded by myelinated fibers. Much sparser label in the reticular poles was found consistently only after injections in the glabrous digits. Inputs to the poles appeared diffuse and overlapping while preserving some somatotopic order. When treated for CO or stained for Nissl substance or myelin, the pars rotunda of humans showed parcellation patterns that closely resembled the patterns seen in monkeys. From the relationship of inputs to the CO dense cell clusters in monkeys, it was possible to postulate in detail the somatotopic organization of inputs to pars rotunda of humans. The present results provide a comprehensive description of the somatotopic patterns of termination of afferents from the skin of the hand and forearm in the spinal cord and cuneate nucleus of macaque monkeys. A direct relationship of afferent somatotopy and identifiable cell clusters in the pars rotunda of the cuneate nucleus is further demonstrated. Finally, the patterns of cell clusters in the pars rotunda of macaque monkeys and humans suggest that the somatotopic organization of the cuneate nucleus may be very similar in human and nonhuman primates.

The projection pattern of forepaw nerves to the cuneate nucleus of the raccoon

The transganglionic transport of horseradish peroxidase (HRP) was used to determine the projection pattern within the cuneate nucleus of the 4 major nerves innervating the forepaw of the raccoon, a carnivore noted for its tactile and manipulative abilities. The two nerves innervating the dorsal, hairy skin and claws (the radial and dorsal ulnar nerves) projected to the marginal rim of the cuneate nucleus, but not to the middle cluster region or to the caudal region of the nucleus. The two nerves innervating glabrous skin (median and ulnar) projected heavily to the cluster region as well as to rostral and caudal levels of the nucleus. This organization, with dorsal nerves ending above the ventral nerves, is similar in the raccoon, rat and tree squirrel, but reversed in the cat. However, the medio-lateral topography is similar in all species with the ulnar and dorsal ulnar nerves projecting medially within the nucleus compared to the median and radial nerves.