The somatotopic organization of the ventroposterior thalamus of the squirrel monkey, Saimiri sciureus.

Multiunit microelectrode mapping techniques were used to investigate the organization of the somatosensory. thalamus in squirrel monkeys. Receptive fields and response characteristics were determined for closely spaced recording sites along arrays of electrode penetrations that passed through the ventral thalamus dorsoventrally, rostrocaudally, or lateromedially. The results were related to thalamic architecture and led to the following conclusions: (1) A large, single, systematic representation of the body surface occupied most or all of the ventroposterior nucleus, VP. The nucleus was further defined by a distinct cytoarchitectonic appearance, produced by densely packed, deeply stained neurons. (2) Recording sequences in VP were characterized by (a) abrupt shifts in receptive field locations over short recording distances indicating that the electrode had crossed discontinuities or folds in the representation, (b) long sequences of overlapping receptive fields indicating regions of continuous representation and the maintenance of adjacency in the map, and (c) similar receptive Held locations for sites along the trajectory of a penetration indicating regions of isorepresentation. Major somatotopic discontinuities were associated with crossing narrow cell-poor laminae that partially divided VP into subnuclei related to the hand, foot, trunk, and tail in lateral VP and the face in medial VP. Somatotopic discontinuities occurred for electrode penetrations in all three planes, but discontinuities were greater and more frequent for lateromedial electrode penetrations. Lines of isorepresentation and gradual change were most extensive in the rostrocaudal and dorsoventral planes. We hypothesize that the disruptions, regions of isorepresentation, and regions of gradual change result from the thickening, splitting, and folding of a two-dimensional representation of the skin surface to occupy a three-dimensional volume. (3) The magnifications of various skin surfaces in VP were variable so that some skin surfaces, especially the tips of the digits, occupied relatively large portions of the nucleus, while other skin surfaces such as the trunk activated little tissue. It appeared that regions of isorepresentation varied in extent according to magnification factor and position in the map. (4) Within VP, neurons could be classified as slowly adapting or rapidly adapting to maintained skin indentation. Each type of neuron formed small groups or clusters in the nucleus so that several successive recording sites typically encountered one type before a sequence of the other type was observed. (5) Neurons ventrocaudal to VP formed a second representation of the body in the ventroposterior inferior nucleus (VPI). Recordings and cytoarchitecture suggest that VPI extends further caudal than generally recognized, to include a part of the region caudal to VP that is sometimes described as part of the posterior group (Po) of thalamic nuclei. Body parts represented in VP were represented again in VPI in a roughly parallel sequence so that in both nuclei the caudal body was lateral and the rostral body was medial. Neurons in VPI were activated by inputs from pacinianlike receptors. Such activation was not observed in VP. (6) A 1.0-1.5 mm thick band of tissue along the dorsal and rostrodorsal borders of VP was responsive to manipulation of deep body tissues including muscles and joints. The mapping evidence was compatible with the existence of one or more representations of deep receptors roughly parallel to the cutaneous representation in VP. Patterns of thalamocortical connections suggest subdividing the thalamic zone of activation by deep receptors into a ventroposterior oral nucleus, VPO, relaying muscle spindle information to cortical area 3a, and a ventroposterior superior nucleus, VPS, relaying deep receptor inputs to area 2.