The local domain for divergence of subcortical afferents to the striate and extrastriate visual cortex in the common...

In the common marmoset (Callithrix jacchus), the cortical projection from the pulvinar and other diencephalic structures into the striate and prestriate cortex was investigated with various fluorescent retrograde tracers. Single cortical injections as well as multiple injections at distances of 1-2 mm with one tracer into an extended but coherent cortical region were applied. Fields with multiple injections were placed so that they touched each other (minimal distances 2 to 3 mm). Retrogradely labelled cells in the LGN and/or the pulvinar were arranged in coherent columns, volumes or slabs, but cell volumes resulting from neighbouring cortical injections overlapped at their border (for details of the thalamo-cortical topography see the companion paper Dick et al. (1991]. Double labelled cells (dl) were only found in the zones of overlap of the cell volumes labelled by the respective tracers. The relative number of dl-cells in these overlap zones was 6.2 +/- 3.1%. The dl-frequency was the same in the various nuclei of the pulvinar and the LGN. In the main layers of LGN, dl-cells were found only in the overlap zone of two injection fields into area 17, but a few dl-cells were found in interlaminar cells after injections into area 17 and 18. Maximal cortical distances between injection fields which produced dl in the pulvinar, were 3 to exceptionally 4 mm but dl was highest at injection distances less than or equal to 2.5 mm and decreased sharply at wider distances. Such overlap zones were concerned with identical or overlapping regions of visual field representation in the cortex and probably also in the pulvinar. Although in individual experiments up to four different tracers were injected into different striate/prestriate regions, often embracing the same visual field representation, individual cells in the pulvinar showed dl from maximally only two tracers injected into neighbouring cortical regions. We conclude that dl in the posterior thalamic projection nuclei is determined essentially by cortical distance and thus reflects the local domain of branching of thalamo-cortical afferents. Pruning of such branches during development may further restrict bifurcating axons to identical visual field representations, but representation of identical visual field regions in different visual areas is not, per se, a sufficient condition for dl. It is not found if such regions are further apart from each other than the typical local domain of 2-3 mm, exceptionally up to 4 mm in one experiment after injections into area 17 and MT.(ABSTRACT TRUNCATED AT 400 WORDS)