Thalamic connections of three representations of the body surface in somatosensory cortex of gray squirrels

Abstract The anatomical tracer, wheat germ agglutinin, was used to determine the connections of electrophysiologically identified locations in three architectonically distinct representations of the body surface in the somatosensory cortex of gray squirrels. Injections in the first somatosensory area, S‐I, revealed reciprocal connections with the ventroposterior nucleus (VP), a portion of the thalamus just dorsomedial to VP, the posterior medial nucleus, Pom, and sometimes the ventroposterior inferior nucleus (VPI). As expected, injections in the representation of the face in S‐I resulted in label in ventroposterior medial (VPM), the medial subnucleus of VP, whereas injections in the representation of the body labeled ventroposterior lateral (VPL), the lateral subnucleus of VP. Furthermore, there was evidence from connections that the caudal face and head are represented dorsolaterally in VPM, and the forelimb is represented centrally and medially in VPL. The results also support the conclusion that a representation paralleling that in VP exists in Pom, so that the ventrolateral part of Pom represents the face and the dorsomedial part of Pom is devoted to the body. Because connections with VPI were not consistently revealed, the possibility exists that only some parts or functional modules of S‐I are interconnected with VPI. Two separate small representations of the body surface adjoin the caudoventral border of S‐I. Both resemble the second somatosensory area, S‐II, enough to be identified as S‐II in the absence of evidence for the other. We term the more dorsal of the two fields S‐II because it was previously defined as S‐II in squirrels (Nelson et al., '79), and because it more closely resembles the S‐II identified in most other mammals. We refer to the other field as the parietal ventral area, PV (Krubitzer et al., '86). Injections in S‐II revealed reciprocal connections with VP, Pom, and a thalamic region lateral and caudal to Pom and dorsal to VP, the posterior lateral nucleus, Pol. Whereas major interconnections between S‐II and VPI have been reported for cats, raccoons, and monkeys, no such interconnections were found for S‐II in squirrels. The parietal ventral area, PV, was found to have prominent reciprocal interconnections with VP, VPI, and the internal (magnocellular) division of the medial geniculate complex (MGi). The pattern of connections conforms to the established somatotopic organization of VP and suggests a crude parallel somatotopic organization in VPI. Less prominent interconnections were with Pol. Sparse, fine label in part of the ventral (principal) nucleus of the medial geniculate complex (MGv) suggests the existence of some input from PV. The connections demonstrated in the present study help characterize three somatosensory areas in squirrels. Such information is essential for identifying homologous areas of cortex across species, and several possibilities are outlined in the Discussion. In addition to VP, the results suggest the presence of two somatic nuclei, the medial and lateral nucleus of the Po group; and evidence is provided for the existence of VPI in rodents. Furthermore, consistent with many recent reports, we found that each cortical area has interconnections with more than one thalamic nuclei, and each thalamic nucleus has interconnections with more than one cortical field. Finally, in keeping with the responsiveness of neurons in PV to both somatic and auditory stimuli (Krubitzer et al., '86), PV was found to have interconnections with both somatic and auditory thalamic nuclei.