Projections from striate and extrastriate visual cortices of the cat to the reticular thalamic nucleus

We have studied the pattern of connectivity of the visual cortical areas 17, 18, 19, 20a, 21a, posteromedial lateral (PMLS), and the posterolateral lateral (PLLS) suprasylvian areas with the reticular thalamic nucleus (RTN) of the cat ventral thalamus. Three cortical areas per hemisphere were injected iontophoretically with either 4% wheat germ agglutinin-horseradish peroxidase, 4% dextran-fluororuby, or 4% dextran-biotin. The visual field representations of the injection sites were determined by reference to previously published visuotopic maps of the cortex. The locations of labelled fibres, presumed terminals and cell bodies were determined with the aid of a camera lucida attachment and computer aided stereometry. In the ventral thalamus, the primary visual cortices (areas 17 and 18) project in a topographic manner to both the perigeniculate nucleus (PGN) and the RTN. By contrast, the "higher" visual cortical areas (areas 19, 21a, 20a, PMLS, and PLLS) project only to the RTN. Our experiments demonstrate the existence of a single, albeit coarse, visuotopic map within the RTN but do not support the notion of separate subregions within the RTN that can be related specifically to a particular visual cortical area. The putative single visuotopic map in the RTN appears to be organised in such a way that the vertical meridians are represented along the rostrocaudal axis of the RTN, whereas the horizontal meridians are mapped within the dorsoventral axis of the nucleus. The upper visual field is represented within regions of the RTN adjacent to the caudal part of the dorsal lateral geniculate nucleus (LGNd), whereas the lower visual field is represented in the parts of the RTN rostral to the LGNd. The map also shows a ventrodorsal shift along the rostrocaudal axis of the RTN such that in the rostral RTN the representation of vertical meridian is placed more ventrally than that in the caudal part of the nucleus.