Physiological evidence that the 2-deoxyglucose method reveals orientation columns in cat visual cortex

Abstract

A method for measuring local rates of glucose uptake in mammalian brain tissue^1,2 has been used to study monkey cortex receiving visual stimulation through one eye³. The resulting pattern of glucose uptake resembled that of the ocular dominance columns, identified by several independent methods in the cortex^4–6. In addition, the pattern of glucose uptake produced by binocular viewing of stripes of a single orientation^7–12 was consistent with the arrangement of orientation columns expected from the sequence of preferred orientations of single cells recorded in the visual cortex^4,13–16. It was therefore assumed that the columns of glucose uptake revealed in the studies of orientation selectivity correspond to columns of cortical cells selective for the orientation viewed. Recent findings about the monkey visual cortex led us to question this assumption.^17–20 We have now investigated this question in the cat by making microelectrode recordings to determine preferred orientation of cortical units at known positions within the visual cortex and then stimulating the same animals with a full-field pattern of stripes to measure the rates of glucose uptake at the positions of each of the cortical units. As in previous studies, preferred orientation shifted gradually and progressively along electrode tracks parallel to the cortical surface^4,13–16 and a pattern of densely labelled columns was observed in the 2-deoxyglucose autoradiographs^7–12. The centres of densely labelled cortical columns contained cells selective for the stimulated orientation. Preferred orientation shifted gradually away from the stimulated orientation at sites progressively more distant from the centres of the columns. Control experiments revealed no columns after stimulation with a pattern of changing orientation. Thus the columns revealed by the deoxyglucose method do correspond to the physiological orientation columns.