Monocular deprivation produces an imbalance in visual drive from the two eyes, which in adult macaque V1 leads to marked changes in the neurochemistry of GABA interneurons. Such changes were further examined by studying immunoreactivity for calbindin, calretinin, and parvalbumin, three calcium-binding proteins that mark distinct subpopulations of GABA neurons, in macaques that had been monocularly deprived by intravitreal injection of tetrodotoxin. Deprivation for 5 d or longer produced a reversal in the normal pattern of calbindin immunostaining in layer III, from one in which intense neuronal immunostaining surrounded the cytochrome oxidase-rich puffs to one in which it occupied the puffs. Over the same period, calbindin immunostaining in other layers was reduced across the entire width of deprived-eye columns or extended into flanking regions of normal-eye columns. In contrast, reduction in parvalbumin immunostaining occurred only in deprived-eye columns and included only terminals with short periods of deprivation (up to 17 d) but both terminals and somata with longer periods. No change in calretinin immunoreactivity was observed. These findings demonstrate that GABA neurons of macaque V1 vary in their response to monocular deprivation according to their neurochemistry and position, suggesting that the weight of inputs from the two eyes and the intrinsic characteristics of each GABA population determine how a neuron responds to a change in visual input.

The cytoarchitecture was studied in a segment of the ferret suprasylvian gyrus containing at least two and possibly four somesthetic representations of the face that were observed in the primary somatosensory cortex. These representations were restricted to the crown of the gyrus and were surrounded by somesthetically unresponsive cortex that extended down both sides to the base of adjacent sulci. Numerous cytoarchitectonic subdivisions were found on a qualitative basis, and were confirmed quantitatively by cluster analyses and relevant statistical tests of 10 prominent features from layers III, IV, and V. Four distinct cytoarchitectonic subdivisions, each with a well-developed and homogeneous granular layer IV, were found distributed from anterior to posterior along the crown of the gyrus at sites corresponding to the locations of the four facial representations. The surrounding unresponsive cortex had a fragmented cytoarchitecture, especially along the medial bank and base of the coronal sulcus. This unresponsive cortex separated the facial representations from the body representations, which were located on the adjacent posterior cruciate gyrus. Most of the unresponsive subdivisions had a heterogeneous or agranular layer IV and fairly well-developed sublamination in layer III, which may be indicative of extensive corticocortical connections. One set of unresponsive subdivisions had comparable cytoarchitectures that directly bordered the facial representations. Another set of unresponsive subdivisions with comparable architectures occupied most of the lateral bank of the gyrus. The implications of multiple representations and cytoarchitectonic fragmentation of the ferret primary somatosensory cortex are discussed in relation to the organization of the primary somatosensory cortex in other species.

Using high-resolution microelectrode mapping methods, we explored the organization of the face representation within the primary somatosensory cortex of ferrets, finding evidence for at least two and probably four representations of the face distributed consecutively from anterior to posterior along the long axis of the suprasylvian gyrus. Examination of the cytoarchitecture (Rice et al., this issue) revealed that these four areas corresponded to four different cytoarchitectonic fields within the crown of the suprasylvian gyrus. The two central, most completely defined representations were oriented so that the dorsal cutaneous surfaces of the face were represented on the lateral side of the gyrus, while the perioral and ventral surfaces were represented on the medial side. The rostral-to-caudal organization within these two representations was reversed; the glabrous rhinaria were represented at the opposite ends of the maps, and penetrations progressively further away from the cortex serving the rhinaria encountered neurons activated by sites progressively more caudal on the face. Receptive fields obtained more rostrally on the gyrus suggested another reversal, implying a third representation. A small area with large receptive fields near the caudal and medial border of the two central maps suggested the presence of a fourth representation. Since the projections of adjacent skin surfaces overlapped considerably, cortical sites serving a particular cutaneous surface were illustrated as enclosed areas that overlapped the territories of other, adjacent representations. The results of this study and of others suggest a need for a re-evaluation of the hypothesis establishing a homology between the representation found in area 3b of primates and that of the primary somatosensory area in nonprimates.

Multiunit recordings along mediolateral rows in the primary somatosensory cortex of the animals described by C. Avendaño, D. Umbriaco, R.W. Dykes, and L. Descarries (1995, J. Comp. Neurol. 354:321-332) provided information about the functional status of the regions in and near the deafferented cortex. Responses changed along this axis from normally organized receptive fields in the hindlimb representation through a transition zone of unusually small receptive fields into the clearly deafferented forelimb representation, where receptive fields were uncommon and often had unusual characteristics. The most abrupt change along this axis was the appearance of a repetitive, bursting discharge pattern in the multiunit activity near the border of the deprived cortex. The appearance of this pattern was used as a reference to describe differences between normal and deprived cortices. The nature of these differences evolved with time. Much of the deprived cortex lacked identifiable receptive fields for months after the nerve transections and, 1 year later, still only about half of the recording sites within the deprived region displayed organized receptive fields. Some sites within the deprived region lacking definable receptive fields could be excited at long latencies by somatic stimuli anywhere on the body. With time, regions of normal cortex near the border with the deprived zone became more involved in these processes. Spontaneous activity and thresholds also changed with time in both normal and deprived cortices.(ABSTRACT TRUNCATED AT 250 WORDS)

Most anatomical and electrophysiological studies of the cat primary somatosensory cortex rely on Hassler and Muhs-Clement's (J. Hirnforsch. 6:377-420, 1964) cyto- and myeloarchitectonic description distinguishing area 3a from area 3b; however, discrepancies in the delineation of these areas in published studies suggest that many workers have found it difficult to apply those criteria systematically. We examined the cytoarchitecture of area 3b in Nissl stained sagittal sections from which electrophysiological data had been obtained prior to sacrifice. Rostrocaudal rows of electrode penetrations placed at different mediolateral positions in the gyrus located regions responsive to stimulation of either cutaneous or deep structures. Small electrolytic lesions allowed these data to be related to the cytoarchitecture. A systematic study throughout the trunk and limb representations found cutaneous responses in cortical regions characterized by a thick and cell-dense granular layer IV, however these same regions had a variable population of medium-sized and/or large pyramidal cells in layer V. Pyramidal cells were practically absent from the forelimb representation, but were present to varying degrees in the trunk and hindlimb representations. Moreover, the relative thickness and cell-density in layer IV were greater in the forelimb than in the hindlimb representations. Deep responses were found in cortex characterized by a thinner layer IV. Since the characteristics of layer V in area 3a were variable, it was less useful for identification of the border between areas 3a and 3b. Clear changes in the intensity and laminar distribution of acetylcholinesterase staining occurred between areas 3a and 3b, making this a useful adjunct to the Nissl stain.