Figure 2.

Luminance-driven hyperpolarization is mediated by large increases in synaptic inhibition. A, Effect of varying membrane potential on amplitude of luminance-evoked hyperpolarization. As the membrane potential was depolarized by continuous injection of positive current through the intracellular electrode, the magnitude of the luminance-driven hyperpolarization progressively increased, becoming five times greater at +10 mV (+600 pA; right) than at rest (0 pA injection; left). The stimulus was a transition from a gray to white uniform field, sustained for 1 s. B, Calculation of reversal potential for luminance-evoked hyperpolarization. By varying the amount of current injection, hyperpolarizations were recorded at multiple membrane potentials to assess the approximate potential at which the driving force on the luminance-evoked response was nullified. Because the magnitudes of hyperpolarization differed between cells, the set of responses for each cell was normalized to the response evoked at zero millivolts. (In many cases, this value was extrapolated by fitting the set of points to a linear function.) This graph contains two or three representative points (postnormalization) from each cell, the set of which were fit by a straight line (n = 11 layer 2/3 neurons). The y-intercept, −85 ± 6 mV, indicates the population average reversal potential of the luminance-evoked hyperpolarization. C, Effect of changes in luminance on synaptic conductances for three neurons (i–iii). The top row shows intracellular responses obtained at the resting membrane potential after a transition in mean luminance from black to white (100% luminance step). Additional responses (data not shown) were recorded at multiple levels of current injection to calculate changes in synaptic conductances. Luminance-evoked hyperpolarizations (top row) were correlated with large increases in total conductance (G_total; second row) that were almost entirely attributable to changes in inhibitory conductance (ΔG_inh; bottom row), as excitatory conductance decreased only slightly (ΔG_exc; third row). The quality of conductance values were verified by calculating theoretical responses based on these parameters (gray line overlaying membrane potential; top row). In most cases, the calculated trace was virtually indistinguishable from the true response, indicating that the conductance parameters accurately represent the luminance-evoked changes in synaptic activity.