Abstract
Abstract: :
Purpose: The inhibitory surrounds of bipolar and ganglion cells get contributions from horizontal cells and possibly from amacrine cells. Neither the relative strengths of these contributions nor the circuitry that mediates surround inhibition are well understood. The classical hypothesis is that surround inhibition is mediated primarily by horizontal cell feedback onto cones. In teleost retina, the gap junction blocker carbenoxolone eliminates feedback from horizontal cells to cones. Our purpose was to determine if carbenoxolone has a similar effect on feedback in macaque monkey retina and if so whether blocking it reduces the strength of ganglion cell surrounds. Methods: H1 cells and ganglion cells were microscopically targeted for recording and light stimulation in an in vitro preparation of the intact macaque monkey retina. H1 cell responses to flashing spots were measured before and after superfusion with 100 µM carbenoxolone. Ganglion cell receptive fields were measured before and after application of the drug using alternating spots and annuli as well as drifting sine wave gratings. All stimuli were presented on a background of the same space-averaged luminance. Results: As in teleost retina, carbenoxolone changed the H1 response to a flashing spot. Before drug application, light onset caused a rapid hyperpolarization, followed by a small relatively slow depolarization back to the new resting potential. This depolarization, which has been attributed to feedback, was abolished by the drug. Carbenoxolone also eliminated ganglion cell receptive field surrounds. Ganglion cells gave large sustained responses of opposite sign to flashing spots and annuli. Application of the drug abolished the response to the annulus. Additionally, the spatial tuning curve measured with a drifting grating changed shape. Before drug application, response amplitude increased from low to intermediate spatial frequencies, a feature attributed to surround inhibition, and then decreased rapidly with further increases in spatial frequency. Application of the drug eliminated the initial amplitude increase. The drug effects washed out after 30 minutes. Conclusion: These data are strong evidence that in macaque the inhibitory surrounds of inner retinal neurons get their major contribution from the H1 horizontal cell network and that H1 horizontal cells contribute to the surround by inhibitory feedback onto cones.
Keywords: 559 retinal connections, networks, circuitry • 540 receptive fields • 415 ganglion cells