Anatomical demonstration of orientation columns in macaque monkey



In the macaque monkey striate (primary visual) cortex, the grouping of cells into ocular dominance and orientation columns leads to the prediction of highly specific spatial patterns of cellular activity in response to stimulation by lines through one or both eyes. In the pesent paper these paterns have been examined by the 2-deoxyuglucose autoradiographic method developed by Sokoloff and his group (Kennedy et al, '76). An anesthetized monkey was given an injection of 14C 2-deoxyglucose and then visually stimulated for 45 minues with a large array of moving vertical stripes, with both eyes open. The 14C autoradiographs of striate cortex showed vertical bands of label extending through the full cortical thickness. Layer I was at most only lightly labelled, and layers IV b and VI wee the most dense. Layer IV c (the site of terminations of most geniculate afferents) was labelled uniformly along its length, as expected from the lack of orientation specificity of units recorded in that leyer. In the other layers the pattern seen in tangential sections was complex, consisting of swirling stripes with many bifurcations and blind endings, but with occasional more regular regions whee the stripes wee roughly parallel. Interstripe distance was rather constant, at 570 μm. Ocular dominance columns were examined in this same monkey, in the same region, by injecting one eye with 3H-proline two weeks before the deoxyglucose experiment, and preparing a second set of autoradiographs of the sections after prolonged washing to remove the 14C-deoxyglucose. As seen in tangential sections through layer IV c, these columns had the usual stripe-like form, with a period of 770 μm, but were simpler in their pattern than the orientation stripes, with fewer bifurcations and less swirling. A comparison of the two sets of columns in the same area showed many intersections, but no strict or consistent relationships: angles of intersection showed a distribution that was not obviously different from that expected for any two randomly superimposed sets of lines.

Another monkey was stimulated with vertical stripes, but with only one eye open. Deoxyglucose autoradiographs of tangential sections showed regular uniform rows of label in layer IV c, with all the characteristic features of eye dominance columns. In the layers above and below IV c the rows in tangential view were broken up into regularly spaced patches of label, presumably representing aggregations of cells responsive to vertically oriented stimuli. The patches showed no consistent alignment across the ocular dominance rows, and indeed no such tendency would be expected, considering the complexity of the orientation columns. This pattern of labelling is again predicted from and confirms the previous physiological studies.