Robert W. Dykes

Parallel processing of somatosensory information: a theory.

Abstract We have hypothesized that at each level of the somatosensory system there are groups of neurons that form a number of different spatially segregated populations. These populations are recognizable by their electrophysiological and anatomical characteristics. Members of each population are hypothezized to be intrinsically different because of genetically determined biochemical properties which give rise to different neurotransmitters, different somatal diameters and different axonal and dendritic distributions as well as different functions. The underlying genetic differences are not known so it is not possible to specify their phenotypic expression, however, some examples of the properties thought to arise from these differences have been cited. The organization seen in the adult brain, when described in this way, may be amenable to explanation by known embryological mechanisms that are general enough to require only a limited amount of genetic coding. Spatial segregation according to function exists to a degree much greater than earlier experiments would lead one to expect, and well-documented cellular affinity mechanisms may be all that is required to generate such groupings in the developing embryo. The somatotopic map is thought to result from competitive axonal sorting upon a continuous function or gradient within each of the populations. Afferent input must be sorted out with respect to nearest neighbors and be able to re-arrange so that the nearest neighbor relations existing at lower levels of the neuraxis are reproduced at each higher level. Mech anisms capable of doing this have been proposed by others to account for the data obtained in the retinal-tectal pathway of the visual system. Some of these models seem directly relevant to the somatosensory pathways. The hypo` presented here suggests that somatotopic boundaries observed within a population provide the necessary reference line required to order large portions of the somatotopic map along a gradient expressed throughout those portions of the population that remain in contact at each level of the neuraxis. The major weaknesses of this theory appear to be: (1) an absence of a role for interneurons or local circuit neurons in the theory, and the need to introduce them as an ad hoc mechanism for communications among the submodality-specific populations, (2) the lack of a role for the reciprocal connections found in all parts of the system and (3) an inability to predict what should happen beyond the relay nuclei and primary somatosensory cortex. The lack of a communication link between neurons from different populations is a serious deficiency. It is obvious that ideas similar to these have some heuristic value in the other sensory systems. The auditory and visual pathways consist of multiple nuclear relay regions and multiple cortical representations. In the visual system it is already clear that there are several distinct populations of neurons found at each level of the visual pathway (X, Y and W cells) and it is reasonable to postulate that several populations of neurons exist in the auditory pathway, too. The general organizational rules outlined here should be applicable to comparable submodality-specific populations in those sensory systems thereby yielding further economy of embryological mechanisms yet allowing a great diversity of sensory attributes to be encoded and relayed centrally.

Long-term enhancement of evoked potentials in cat somatosensory cortex produced by co-activation of the basal forebrain and cutaneous receptors

SummaryAveraged evoked potentials from primary somatosensory cortex (SEPs) were recorded before and after pairing the peripheral stimuli with electrical activation of the basal forebrain (BF) in anesthetized cats. Four pulses at 400 Hz were delivered to the BF 120 ms before each cutaneous stimulus and 10 to 660 such pairings were found to produce an enlargement of the SEP in 10 of 11 animals. The average increase in amplitude of the initial peak of the SEP was 69%. The SEP remained enhanced in five of six animals that were tested an hour or more after the pairing, and in one case the SEP was tested 4.5 h after pairing without diminution. The effective BF sites were located in the substantia innominata and at the rostral pole of the globus pallidus, regions known to contain many cholinergic cell bodies. Enhancement occurred consistently only if stimulation of the BF site elicited a positive wave in the cortex at a latency of 11 to 18 ms. Repeated BF stimulation without cutaneous input did not produce a change in subsequent SEPs. The long-term changes described here may be involved in experimentally- and naturally-induced cortical reorganization.

Acetylcholine permits long-term enhancement of neuronal responsiveness in cat primary somatosensory cortex

Acetylcholine (ACh) was administered iontophoretically to single neurons in cat somatosensory cortex. Using extracellular recording techniques, neuronal responsiveness was determined at regular intervals from the number of action potentials produced either by iontophoretically applied glutamate or by tactile stimulation of the cutaneous receptive field. The responses were altered in only 21% (13/61) of the neurons following the application of ACh alone. In contrast, 75% (66/88) of the neurons displayed altered responses during administration of ACh simultaneously with either iontophoretically administered glutamate or with tactile stimulation of the receptive field. Forty-seven percent (29/62) of the responses potentiated in the presence of ACh remained enhanced for periods lasting from 8 min to over 1 h. The responsiveness of cortical neurons to afferent inputs changes during the reorganization of somatotopic maps that occurs after deafferentation, and perhaps during some forms of learning. As ACh has been implicated in some of these processes, it may be that the changes in responsiveness observed here following iontophoretically applied ACh are similar to those which facilitate the acquisition of neuronal responses to altered or novel afferent inputs.

Bicuculline-induced alterations of response properties in functionally identified ventroposterior thalamic neurones

SummaryExtracellular recordings of 105 neurones in the cats somatosensory thalamus were obtained with carbon fibre-containing multibarrel micropipettes. The responses of cells to natural stimulation of cutaneous or deep structures were characterized and the responses to electrical stimulation of primary somatosensory cortex were determined. Receptive fields were mapped and the functional properties were examined before and during the microiontophoretic administration of glutamate, γ-aminobutyric acid (GABA) and bicuculline methiodide (BMI). Modality and submodality properties of all cells tested apparently remained unchanged qualitatively, despite all pharmacological interventions. BMI lowered the response threshold of a majority of the 48 cells tested for this variable, although almost 25% responded with elevated thresholds. BMI changed the temporal properties of the responses of both thalamocortical relay neurones and of presumed interneurones. Discharges evoked by natural stimuli and by electrical stimulation of the cortex were prolonged and their pattern was altered. Decreases in the frequency of bursts of discharges were often observed with BMI, and these bursts were invariably prolonged and the interspike interval profiles were altered. Receptive field size changes were observed only in 8 of 48 neurones. For two of these, the field size decreased, while for the others there were small increases.

Somatosensory cortical neurons with an identifiable electrophysiological signature

In both cats and rats, neurons with a distinctively narrow action potential were recognized as a small subset of all neurons isolated in the somatosensory cortex. These cells were characterized by generally having a spontaneous activity, some evidence of an afferent input, a sensitivity to glutamate but a relative resistance to depolarization block induced by glutamate and a marked insensitivity to acetylcholine. Two were filled with horseradish peroxidase (HRP) and recovered. Although others have suggested that such neurons are interneurons, following reconstruction it was apparent that the two cells filled with HRP were pyramidal cells. These observations suggest that there may be more than one class of cortical neurons with thin spikes.

Morphological and electrophysiological characteristics of somatosensory thalamocortical axons studied with intra-axonal staining and recording in the cat

SummaryThe intracortical arborizations of thalamocortical fibers arising from the ventroposterolateral (VPL) nucleus in the cat were studied following intra-axonal injections of horseradish peroxidase (HRP). The axons were impaled 1.5 to 3 mm below the surface of the cortex, identified electrophysiologically by stimulating the VPL nucleus and functionally by stimulating the somatic receptive field with natural stimuli. Many of the results obtained in a previous study using similar techniques (Landry and Deschênes 1981) were confirmed by the present experiments. Fibers activated by cutaneous stimulation arborized either in area 3b or 1 but some did send branches to both areas. Also, the intracortical arborization of a rapidly adapting cutaneous afferent fiber in area 2 is described. The size and tangential extent of the fiber in area 2 are similar to those arborizing in other areas of the primary somatosensory cortex and consist of multiple patches separated by uninvaded gaps. One fiber activated by stimulation of deep tissue receptors gave rise to two bushes that arborized along a rostrocaudal axis exclusively in area 3b. Terminal boutons and varicosities were found mostly in layers VI, IV, the bottom third of III and the upper portion of V, but some fibers did send a few collateral branches to layer II and the bottom part of layer I. The results suggest that in the forelimb representation, the same modality and submodality can be recorded in more than one cytoarchitectonic area but that areas 3b, 1 and 2 should not be considered as a single functionally homogeneous area. Counts of terminals suggest that a single fiber arborizing in area 1 makes as many as 3 times the number of synapses made in area 2 or 3b. Since fibers appear to be modality and submodality specific, if convergence of modality, submodality and/or body areas occur in the cortex, then this must be preferentially, but not exclusively, done by thalamic fibers of different functions which arborize in the same cytoarchitectonic area and synapse upon a shared postsynaptic target. In the same experiments intra-axonal recordings revealed the presence of two hyperpolarizing afterpotentials elicited by a preceding action potential. The first afterpotential was associated with a decrease in excitability of the fiber and an increase in membrane resistance. Passage of depolarizing current through the microelectrode was necessary to demonstrate the second afterpotential. These afterpotentials may affect the integrative properties of the axons by modifying impulses originating in the thalamus.

What Makes a Map in Somatosensory Cortex

The quotation from Van Vogt’s novel (1970) has particular relevance for students of somesthesis. Some of the most fundamental concepts about the organization of somatosensory pathways are currently undergoing rapid and radical change. Maps represent some of these concepts pictorially, expressing the relationships between functions and structures. In these efforts to understand and to communicate ideas to others, we often create the “semantic disturbance” noted by Van Vogt. In the interest of clarifying the relationship between a map and the concept it designates, the following pages explore the history of mapmaking in somatosensory cortex and summarize the evolution of the underlying ideas. Following this we examine more carefully how a map is made and what its limitations are.

The effects of strychnine on neurons in cat somatosensory cortex and its interaction with the inhibitory amino acids, glycine, taurine and β-alanine

In area 3b of primary somatosensory cortex, neurons may be classified as either rapidly adapting or slowly adapting to sustained stimuli and may be differentiated further by the presence or absence of a receptive field and by their threshold of activation. It is also possible to use the rate of adaptation of the background activity to a sustained stimulus to divide the cortex into slowly adapting regions or rapidly adapting regions. By blocking GABA-mediated inhibition with iontophoretically administered bicuculline methiodide, others have observed an increase in receptive field size in rapidly adapting regions but not in slowly adapting regions. The present study was designed to look for a different inhibitory transmitter which might control receptive field size in slowly adapting regions. Iontophoretically delivered strychnine was employed as an antagonist because it interferes with glycine-like inhibitory transmitters such as glycine, taurine and beta-alanine. Pharmacological tests were performed on 157 neurons in two series of experiments. In the first series three effects were documented. (i) In rapidly adapting regions, the size of the receptive field increased in 11 out of 25 cases whereas none of the 20 receptive fields tested in slowly adapting regions enlarged. (ii) In 13 of 24 cases a receptive field was revealed for previously unresponsive neurons in rapidly adapting regions whereas only 5 of 22 unresponsive cells tested in slowly adapting regions developed a receptive field. (iii) In 15 of 25 cells with receptive fields tested in rapidly adapting zones, strychnine reduced the threshold for somatic stimuli but only 8 of 20 cells isolated in slowly adapting zones showed this effect. In a second series of experiments, the effect of beta-alanine, glycine and taurine was examined on neurons of the rapidly adapting regions. beta-Alanine and taurine reduced the excitability of all neurons tested. Glycine inhibited most neurons. However, strychnine only antagonized the inhibitory effects of beta-alanine on responses to peripheral stimuli (9 of 11 cases). When neurons could not be driven by peripheral stimuli, the inhibition of spontaneous or glutamate-induced activity could not be blocked by strychnine (0 of 18 cases). We suggest that glycine-like amino acids contribute to the control of receptive field size and the control of neuronal excitability in rapidly adapting regions but not in slowly adapting regions. Our data suggest that strychnine-sensitive synapses are limited only to a subset of cortical neurons driven by somatic inputs.

Simultaneous recordings from pairs of cat somatosensory cortical neurons with overlapping peripheral receptive fields

Fifty-three cell pairs in the somatosensory cerebral cortex were examined in pentobarbital-anesthetized cats for evidence of short latency interactions. Many neuronal pairs separated by distances of 150 to 500 micron were observed to have temporal dependencies. In a subset of 19 pairs where the surrounding multiunit activity could be classified as rapidly adapting or slowly adapting, short latency interactions existed only between cell pairs sharing the same multiunit background activity. If one member of the pair was in a slowly adapting background and the other in a rapidly adapting background, the cells did not influence one another. This observation was taken as evidence for parallel and separate processing of afferent signals from rapidly and slowly adapting cutaneous mechanoreceptors in cat somatosensory cortex.

Neuroanatomical evidence of reinnervation in primate allografted (transplanted) skin during cyclosporine immunosuppression

Histological evidence is presented documenting the reinnervation of sensory mechanoreceptors across major histocompatibility barriers in allografted primate (baboon) skin. Meissner and pacinian corpuscles, as well as hair follicles, showed a spectrum of reinnervation by host axons. Our light and electron microscopic evidence to date has suggested that allografted Merkel cells did not survive transplantation. This, and other instances of tissue and receptor destruction resulting from histoincompatibility, indicated marked differences when skin allografts were compared to autografts with respect to the ability of host axons to locate and reinnervate sensory mechanoreceptors.