Abstract
Purpose :
The retinal ganglion cell (RGC) receptive field represents the cell's space-time processing. While the connectivity of cells in the retina is well characterized, the functional implications of this circuitry on the RGC receptive field are poorly understood. In this study, we use white noise stimulation and multi-electrode recording to map RGC receptive fields in rod and cone knockout (Trα -/- and GNAT2 -/-) mice to characterize the effect of removing rods on the spatiotemporal processing of RGCs.
Methods :
Retinas from 12-16 week old dark-adapted male mice (3 C57/B6 and 1 Trα -/-) were whole-mounted on a multi-electrode array for extracellular recording of ganglion cell action potentials. Retinas were stimulated with binary white noise checkerboards at 15 Hz at both scotopic and photopic light levels. Single units (70 C57/B6 and 12 Trα -/-) were isolated from recordings and spike-triggered averages (STAs) were computed for each unit to map ganglion cell space-time receptive fields. STAs were fit with a spatiotemporal model composed of a difference of Gaussians spatial model and the impulse response function of a bandpass filter. Parameters of these models were used to compare RGC properties.
Results :
Our results show differences in spatiotemporal processing between wild type and Trα -/- mice. First, RGCs in the Trα -/- mouse had responsive STAs under photopic but lacked them under scotopic conditions, while those in wild-type mice had responsive STAs in both conditions. Comparing the photopic receptive fields, the Trα -/- RGCs had faster temporal processing in their center than wild type cells. However, we see that knockout of Trα -/- has no effect on the antagonistic surround, as it is present in both groups of RGCs. Lastly, we saw no change in RGC receptive field center size in the Trα -/- mouse.
Conclusions :
By combining white noise analysis with genetic knockout of rods and cones, we can single out their contribution to the receptive field of retinal ganglion cells. Unsurprisingly, rods are necessary for RGC spatiotemporal processing in dim light. Further, consistent with their longer integration time, we find that RGC temporal processing is faster in the absence of rods. Lastly, we see that rods have no effect on gross spatial processing in RGCs in photopic conditions.Investigation of spatiotemporal properties of RGCs in mice without cones is underway.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.