Plasticity of ocular dominance columns in monkey striate cortex

Abstract

Ocular dominance columns were examined by a variety of techniques in juvenile macaque monkeys in which one eye had been removed or sutured closed soon after birth. In two monkeys the removal was done at 2 weeks and the cortex studied at 1 1/2 years. Physiological recordings showed continuous responses as an electrode advanced along layer IV C in a direction parallel to the surface. Examination of the cortex with the Fink-Heimer modification of the Nauta method after lesions confined to single lateral-geniculate layers showed a marked increase, in layer IV G, in the widths of columns belonging to the surviving eye, and a corresponding shrinkage of those belonging to the removed eye. Monocular lid closures were made in one monkey at 2 weeks of age, for a period of 18 months, in another at 3 weeks for 7 months, and in a third at 2 days for 7 weeks. Recordings from the lateral geniculate body showed brisk activity from the deprived layers and the usual abrupt eye transitions at the boundaries between layers. Cell shrinkage in the deprived layers was moderate - far less severe than that following eye removal, more marked ipsilaterally than contralaterally, and more marked the earlier the onset of the deprivation. In autoradiographs following eye injection with a mixture of tritiated proline and tritiated fucose the labelling of terminals was confined to geniculate layers corresponding to the injected eye. Animals in which the open eye was injected showed no hint of invasion of terminals into the deprived layers. Similarly in the tectum there was no indication of any change in the distribution of terminals from the two eyes. The autoradiographs of the lateral geniculates provide evidence for several previously undescribed zones of optic nerve terminals, in addition to the six classical subdivisions. In the cortex four independent methods, physiological recording, transneuronal autoradiography, Nauta degeneration, and a reduced-silver stain for normal fibres, all agreed in showing a marked shrinkage of deprived-eye columns and expansion of those of the normal eye, with preservation of the normal repeat distance (left-eye column plus right-eye column). There was a suggestion that changes in the columns were more severe when closure was done at 2 weeks as opposed to 3, and more severe on the side ipsilateral to the closure. The temporal crescent representation in layer IV C of the hemisphere opposite the closure showed no obvious adverse effects. Cell size and packing density in the shrunken IVth layer columns seemed normal. In one normal monkey in which an eye was injected the day after birth, autoradiographs of the cortex at 1 week indicated only a very mild degree of segregation of input from the two eyes; this had the form of parallel bands. Tangential recordings in layer IV C at 8 days likewise showed considerable overlap of inputs, though some segregation was clearly present; at 30 days the segregation was much more advanced. These preliminary experiments thus suggest that the layer IV C columns are not fully developed until some weeks after birth. Two alternate possibilities are considered to account for the changes in the ocular dominance columns in layer IVG following deprivation. If one ignores the above evidence in the newborn and assumes that the columns are fully formed at birth, then after eye closure the afferents from the normal eye must extend their territory, invading the deprived-eye columns perhaps by a process of sprouting of terminals. On the other hand, if at birth the fibres from each eye indeed occupy all of layer IV C, retracting to form the columns only during the first 6 weeks or so, perhaps by a process of competition, then closure of one eye may result in a competitive disadvantage of the terminals from that eye, so that they retract more than they would normally. This second possibility has the advantage that it explains the critical period for deprivation effects in the layer IV columns, this being the time after birth during which retraction is completed. It would also explain the greater severity of the changes in the earlier closures, and would provide an interpretation of both cortical and geniculate effects in terms of competition of terminals in layer IV C for territory on postsynaptic cells.