Abstract
A debate surrounds the question of whether the orientation tuning of neurons in primary visual cortex (V1) is sharpened by intracortical circuitry. Such sharpening would be apparent in dynamical changes in orientation tuning during a response. Extracellular measurements have argued for such changes (Ringach et al., 1997), but intracellular measurements have found otherwise (Gillespie et al., 2001). We sought to resolve this question by measuring the responses of an entire population rather than those of individual neurons. We imaged voltage-sensitive dye fluorescence in layers 2–3 in V1 of anesthetized, paralyzed cats (imaged areas: ∼30 mm2; sampling rate: 110 Hz). We established the orientation preference of each pixel from responses to flickering gratings (8 orientations). We then presented random sequences of static gratings (40 ms) with one of four phases and one of four orientations (Ringach et al., 1997). Reverse correlation analysis yielded the responses to each grating at each instant in time. We summarized these data by plotting the average response of each pixel as a function of its preferred orientation relative to the stimulus. These responses were well fitted by a Gaussian function. Responses rise 40–50 ms after stimulus onset, reach their maximum after 80–100 ms, and return to baseline after ∼150 ms. Tuning width shows no sign of sharpening. It remains largely constant at ∼40 degrees (half-width-half-height), a value somewhat larger than the ∼30 degrees measured intracellularly (Gillespie et al., 2001). An analysis restricted to pixels with particularly sharp orientation tuning yielded similar results. We conclude that the orientation selectivity of neurons in the superficial layers of primary visual areas is constant through time. This result argues against the hypothesis that orientation selectivity is sharpened through intracortical circuitry.