Abstract
Correlated spiking between cortical neurons affects sensory coding. Previous work has established that neurons across the layers of visual cortex differ systematically in their spontaneous activity as well as in the degree to which they are correlated across trials in their visual response. However, we know little about how spontaneously occurring spike rate correlations are related to those evoked under visual stimulation. Here we investigate correlated variability within and across the layers of macaque primary visual cortex (V1), and compare their spatial structure during fixation on a neutrally grey screen versus stimulation with high contrast Gabor grating patches of varying orientations. We placed linear multielectrode arrays spanning all layers of V1 in two monkeys that were trained to fixate on a screen while stimuli were presented to the receptive field of neurons under study (n=63 penetrations). Electrodes were aligned and laminar compartments delineated using Current Source Density (CSD) analysis. We extracted multiunit activity as a measure of population spiking within the supragranular, granular and infragranular laminar compartments. Correlated trial-by-trial variability was computed within and across compartments for each stimulus type, and then averaged across all stimulus orientations. We found that spontaneously correlated spiking variability across V1 layers is significantly greater than that obtained from a randomized trial-shuffle control. However, the degree of the spontaneously generated correlated spiking variability varied systematically between layers. Visual stimulation altered spike rate correlations in a layer-specific manner even though spiking increased uniformly across the entire cortical column. Taken together, these findings suggest that the laminar structure of visually evoked correlated spiking variability differs significantly from the spatial structure of spontaneously occurring spike rate correlations. We will discuss these findings in the context of sensory coding within V1's laminar microcircuit.
Meeting abstract presented at VSS 2017