Search

The aim of this investigation was to understand the functions of long horizontal connections projecting from area 17 to area 18 in cats. The animals were anesthetized and prepared for recording single-cell responses to sine-wave gratings in area 18. Neuronal activity was analyzed under three conditions: prior to, during, and after inactivation of a circumscribed region of area 17. The latter was depressed with micro-injections of GABA. Cells in both areas were in close retinotopic correspondence. Cells were classified as simple and complex types. Globally, simple cells were less affected than complex units, and those which were affected shifted their optimal spatial frequency to higher values. Complex neurons were more often influenced by the interruption of area 17 input. Namely, the peaks of the tuning curves were displaced on the x-axis to a new optimal spatial frequency. This effect was obtained by a dual change: a decline in the discharge strength to the optimal spatial frequency and an enhancement to nonoptimal spatial frequency. Contrast sensitivity function disclosed similar shifts of optimal spatial frequencies. Likewise bandwith, spatial resolution, high cutoff, and low cutoff were modified to a greater extent in complex cells. It appears that there is no relationship between areas 17/18 orientation difference and the modifications observed in tuning curves to spatial frequencies. The results suggest that neurons of area 18 may carry multiple-frequency channels and that area 17 facilitates the emergence of one particular spatial frequency.

To understand the influence of the ascending path linking area 17 to area 18 of visual cortices, experiments were carried out in which a small neuronal population of area 17 was inactivated with GABA, while unitary responses were recorded in area 18. In the latter, cells are identified as belonging to the simple or complex family according to their firing pattern evoked in response to sine-wave gratings scrolling through the receptive fields. Anesthetized cats were prepared for single-cell recordings. In area 17, a GABA-containing pipette was placed in superficial layers in order to inactivate reversibly a small neuronal population. Prior to blockade, the orientation tuning curves were obtained in both areas and the difference in optimal orientation between areas 17 and 18 was recorded. In area 18, cells were classified as simple or complex. The strategy was to study the reaction of neurons in area 18 prior to, during and after area 17 depression. In most simple cells, whenever the difference in orientation was in the iso-range, that is when the difference in optimal orientations of the injected site (in area 17) and of the neuron in area 18 was less than 30 degrees, the GABA application produced a decline of the evoked discharges, whereas GABA injection augmented the evoked firing rate when the difference was in the cross-range (60 degrees). In contrast to simple cells, GABA depression enhanced the responses in the majority of complex cells with like orientations in both areas. When the difference between recording sites was in the cross-range, then area 17 depression produced weaker evoked firing. A tangential penetration of the injecting pipette, allowing injection of different orientation sites while testing the same unit in area 18, revealed that the latter could react with an enhancement or a decline of the responses as the injecting pipette shifted from iso (or cross) to cross (or iso) disparity in optimal orientations between areas 17 and 18. These results suggest that the path connecting area 17 to area 18 may be functionally discriminated on the basis of the orientation domain and cell types. In addition, our data suggest that the ascending visual streams are required to generate orientation specificity in area 18.

The functional relationships between areas 17 and 18 are still incompletely understood; many experiments indicate that cells sharing like properties are linked together, whereas other studies suggest a deviation from this principle. Here we report the effect of using GABA to block circumscribed area 17 neuronal populations with known orientation preferences. We studied the interaction of this treatment on the responses of simple and complex neurons in area 18. The data suggest that these interactions follow different rules for simple and complex cells: specifically, an iso-orientation linkage appears to hold for simple cells while a cross-orientation linkage appears to hold for complex cells.