Background-induced flicker enhancement in cat retinal horizontal cells. I. Temporal and spectral properties.

1. Dim backgrounds can enhance small-spot flicker responses of cat retinal horizontal cells by a factor of 2 or more. 2. Intracellular marking with horseradish peroxidase (HRP) reveals that this enhancement effect occurs in--but is not necessarily limited to--the cone-connected, A-type horizontal cell. 3. Flicker amplitudes decrease over a frequency range from 3 to 36 Hz of square-wave photic stimulation. There is little evidence of flicker-response enhancement at 3 Hz. Flicker-response enhancement is typically 2-6 times larger at 35 than at 6 Hz. 4. Inspection of flicker waveforms indicates both a scaling-up of response signals with backgrounds and a distortion composed of 2- to 5-ms-latency decrease, expressed primarily within a quick component of OFF-repolarization. 5. Flicker enhancement first increases as a function of background irradiance and then decreases. The increasing limb has the dynamic range and spectral sensitivity of cat rods (507-nm peak). Enhancement is maintained during rod after-effects. The decreasing limb of the background-versus-intensity function results from light adaptation of cat, long-wavelength (red) cones. 6. The flicker responses themselves peak spectrally at approximately 555 nm and reflect only the activity of cat long-wavelength (red) cones, without evidence of intermixing of other photoreceptor mechanisms. 7. Thus within the first synaptic layer of the cat visual system, rod signals interact with the flicker responses of red cones, both increasing cone-signal amplitudes and modifying cone-signal waveforms. 8. The results are closely analogous to "suppressive rod-cone interaction" (SRCI) as described in human psychophysics. 9. An outer-plexiform-layer circuit involving rods, horizontal cells and cones may mediate rod-induced enhancement of cone flicker. This being the case, notions of horizontal-cell feedback interactions with cones may have to be modified and extended. A specific feedback model is elaborated in the companion paper.