Abstract
Cells in the P-pathway show L/M-cone opponency. Whether or not cone opponency derives from cone-specific connectivity to the receptive field surround is controversial. To answer this question, spatial frequency (SF) tuning curves were obtained with achromatic, equiluminant red-green, M-cone, and L-cone isolating gratings. A band-pass index (BPI = ratio of low frequency/peak response) was calculated. BPI = 1 means the spatial frequency response was low-pass with no trace of a surround. BPI = 0 means the center and surround were exactly balanced and opposite in sign. For achromatic gratings, BPI was in the range 〈0,1〉 with mean 0.53 (retina) and 0.44 (LGN). However, in about 80% of P and Parvo cells, BPI = 1 for the center cone (mean BPI = 0.94). Therefore, in most P-pathway cells, the center cone did not contribute to the antagonistic surround. Another issue concerns the possible existence of P cells with mixed cone input to the center (Crook et al., J. Neurosci., 31, 2011) but with strong cone opponency with large fields. We observed a few cells showing this behavior at eccentricities >=10 deg; this behavior could not be modeled by a random wiring approach. These data suggest considerable cone-specificity in the wiring of the midget, P-pathway.