Distal connectivity causes summation and division across mouse visual cortex

Abstract

Neurons in different locations across the cortex are connected through polysynaptic networks involving both excitation and inhibition. To probe the functional effect of such networks, we used optogenetic stimulation to trigger antidromic spikes in a local region of primary visual cortex (V1). This local activity had two effects at distal V1 locations: summation and division. The balance between the two depended on visual contrast, and a normalization model precisely captured these effects.

Fig. 1

Optogenetic antidromic stimulation for probing cortical connectivity in V1. (a) The targets of optogenetic stimulation are L2/3 neurons in the callosal binocular zone (BZ) in the electroporated hemisphere. MZ: monocular zone. (b) Top view of the chamber allowing access to both hemispheres. Rectangles indicate regions in c. (c) Retinotopic maps obtained via intrinsic imaging. V1 and LM were determined based on reversal of retinotopy. Curves: representation of the vertical meridian (dotted) and horizontal meridian (dashed). (d) Fluorescence image showing expression of ChR2-Venus. Rectangle: callosal band, processed separately to enhance signals. Curves are replotted from c. (e,f) Coronal sections of electroporated (e) and contralateral (f) hemispheres (g) Response histograms for two units in putative L2/3 (top) and one in putative L5 (bottom). Inset: spike shapes of the two units. (g) Same, in the presence of iGluR blockers (CNQX, D-AP5).

Fig. 2

Contrast-dependence of effects of distal activation. Electrodes were placed in the far MZ, and neurons there were stimulated with white noise visual stimuli. (a-c) Firing rate responses of three units in the far MZ (RF center = 85°, 55°, and 75°) to antidromic stimulation of the BZ (solid curves), in the absence of visual stimulation (0% contrast). Dashed curves: activity in control condition. In the first two units, distal stimulation caused activation. Inset: an experimental schematic. (d-f) Same, in the presence of visual contrast (25%, 50%, and 25% contrast). In all units the effect of distal stimulation became suppressive. In the second unit, suppression coexisted with activation. (g-i) Same, at higher contrast (50%, 100% and 50%). In all units the effect of distal stimulation was purely suppressive. Blue arrow: laser pulse time (0 ms). (j-l) Summary of the results for the three units, plotted as the difference in firing rate between control and laser conditions. (m) Summary of the results for all sites. Responses were averaged between 50 and 250 ms after laser onset, in the absence (control) and the presence (laser) of distal activation. Red bars are for sites with activation and suppression (n = 56) and blue for sites with suppression alone (n = 47).

Fig. 3

Normalization summarizes the effects of distal activation. (a) Population response as a function of stimulus contrast, for sites showing activation at zero contrast. (b) Same, for sites showing only suppression. Curves: fit of the normalization model (Eq. 1). (c) Effects of changing laser power, for sites showing activation at zero contrast. (d) Same, for the remaining sites. Error bars are median ± median absolute deviation.