Fig. 4: Suppression following luminance steps.
From: Suppression without inhibition: how retinal computation contributes to saccadic suppression
a Population modulation index (mean ± s.e.m.) of ON (light gray, N = 259) and OFF (dark gray, N = 107) RGCs for probe flashes following luminance steps (blue line). Modulation index for each RGC was based on its average response to 56 or 156 luminance-step sequences (Supplementary Fig. 1b) spanning a contrast range of −0.5 to +0.5 Michelson contrast (“Methods”). Probe flashes were presented at 17 ms, 33, 50, 100, 250, 500, 1000, and 2000 (baseline) after luminance steps. Probe flash responses were suppressed in both ON and OFF RGCs, with similar time course and recovery as in the saccade paradigm with textures (Fig. 1e). Error bars are not visible due to small s.e.m. b Same as in (a), except that the modulation index for each RGC was separately based on average responses to probe flashes after positive-contrast luminance steps (left panel; +0.03 to +0.5 Michelson contrast), and after negative-contrast luminance steps (right panel; −0.03 to −0.5 Michelson contrast). Underlying population data are shown in Supplementary Fig. 8. c Same as in (b), for a subset of ON RGCs (N = 115) in control conditions (light-gray lines) and with GABAA,C and glycine receptors blocked (green lines; cocktail of 5 μM SR-95531, 100 μM Picrotoxin and 1 μM Strychnine). Hash symbol: significant difference between modulation of ON and OFF RGCs in a or between ON RGCs without and with pharmacological blockers in (c) (P < 10−4, two-tailed Wilcoxon rank-sum test).
