Uneven densities of corticopontine neurons in the somatosensory cortex: a quantitative experimental study in the cat

By use of large injections in the pontine nuclei of wheat germ agglutinin-horseradish peroxidase conjugate, the distribution of corticopontine cells in the primary somatosensory cortex (SI) was mapped quantitatively. The borders of the cytoarchitectonic areas 3a, 3b, 1 and 2, together constituting SI, were determined cyto- and myeloarchitectonically. Flat maps showing the distribution of labelled cells in SI were constructed. All labelled cells were confined to lamina V, and present in deep as well as superficial parts of this lamina. The size spectrum of the corticopontine cells seemed to encompass the whole range, from the smallest to the largest seen in lamina V of SI in thionin stained sections. Average densities of labelled corticopontine cells in SI varied from 114 to 248 per mm2 in different cases. No significant differences appear to exist between average densities in areas 3a, 3b, 1 and 2. However, densities vary consistently from medial to lateral within SI. When comparing this pattern with the physiological maps showing the somatotopical organization of SI, it appears that regions representing the trunk and proximal parts of the extremities have higher densities of corticopontine neurons than regions representing distal parts of the extremities and the face. The latter parts of SI have in common a much larger magnification factor than the former, that is, the volume of cortex devoted to a certain area in the periphery is much larger in e.g. the SI face region than in the trunk region. It thus would appear that the over-representation of the face and distal extremities in terms of cortical volume devoted to them, is not upheld in terms of number of corticopontine neurons. Although apparently not paralleled in other connections of SI, the uneven densities of corticopontine projections from SI are very similar to what has been described previously in the corticopontine projections from visual areas 17, 18, and 19, where regions with the largest magnification factors have the lowest densities of corticopontine neurons. On the basis of these findings we suggest that sensory information from proximal body parts and peripheral parts of the visual field is relatively more important for the cerebellum, with its main function in movement control, than it is for parts of the brain engaged in more direct analysis of sensory messages.