Purchase this article with an account.
Ilya Buldyrev, William Taylor; Feed-forward Inhibition Controls Spike Onset Latency in Rabbit Local Edge Detectors. Invest. Ophthalmol. Vis. Sci. 2013;54(15):3398.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In some retinal ganglion cells (RGCs), the latency-to-first-spike has been shown to vary systematically with the spatial phase of a stimulus, and therefore, populations of such RGCs have been proposed to provide rapidly encoded spatial information at short latencies after shifts in visual attention. The rabbit retina has two distinct populations of small-field RGCs that could encode the visual scene at high spatial resolution; the local edge detector (LED) and the OFF brisk-sustained (BS) RGCs. The goals of this study were to determine whether the latency-to-first-spike varied systematically with the position of a stimulus edge in either of these cell types, and if so, how spike latency was determined by the timing of inhibitory and excitatory synaptic inputs.
Extracellular and whole-cell voltage clamp recordings from LEDs and OFF-BS RGCs were performed in whole-mount retinal preparations. Spike latencies were measured with centered spot stimuli of increasing diameter, and with a 200 by 800 μm dark bar that was presented at a range of positions across the receptive field.
OFF responses in LEDs had a pronounced increase in delay to spike onset with increased stimulus size, while in OFF BS RGCs spike onset latency did not vary with stimulus size. Furthermore, the spike latency was strongly sensitive to the position of an edge within the receptive field for LEDs, but not for OFF BS RGCs. We hypothesized that this difference arises from distinct inhibitory amacrine cell circuitry, and tested whether glycinergic inputs affect the timing of spike onset in LEDs by blocking glycinergic transmission with strychnine, or suppressing it from inside the cell with the glycine receptor channel blocker Ginkgolide B. These drugs significantly decreased spike latency in LEDs, and reduced, but did not eliminate its dependence on stimulus size or position. In voltage clamp recordings, the ratio of inhibition to excitation increased with stimulus size. For stimuli larger than 75 μm, NMDA receptors on amacrine cells contributed to a slow component of the glycinergic IPSCs. While blocking glycinergic transmission, revealed a fast GABAergic IPSC for larger stimuli, suggesting a role for feedforward GABAergic inputs.
We found that LEDs in the rabbit can transmit stimulus dimensions and position using a spike latency code, and that feedforward inhibition from amacrine cells contributes to this property.
This PDF is available to Subscribers Only