Abstract
Purpose: :
The selective activation of the direct response of Retinal Ganglion Cells (RGC), while avoiding stimulation of the retinal network, is a powerful strategy for epiretinal prosthetic applications. However, existing stimulation techniques offer little control over which cell types or neuronal structures are activated. Here, we determine how stimulus waveform, amplitude, and pulse width affect the threshold of direct and network-mediated responses of RGCs in normal and degenerate (RCS) retina using epiretinal stimulation.
Methods: :
Whole-cell recordings were obtained from RGCs in normal and RCS rats (P45-60) using flat-mount retinal preparations and standard patch clamp technique. Timed pulses (0.1-4ms) of monophasic (anodal, cathodal) and biphasic (anodal-first or cathodal-first) currents were applied at 1 Hz frequency via stimulating electrode (10 μm tip ID) with a mean distance from target cell of 25 μm. Action potentials were recorded in current-clamp configuration.
Results: :
Stimulation thresholds for direct responses of RGCs were lowest for cathodal and highest (3-9x) for anodal pulses across all durations tested, in both wild type and RCS retina. With biphasic waveforms, activation thresholds were comparable to those evoked via cathodal currents for longer pulse widths of 1-4 ms, but increased with shorter pulses, in particular with anodal-first. Selectivity for direct RGC stimulation, defined as the ratio between thresholds of network and direct activation, was lower in wild type compared to RCS retina. With monophasic waveforms, selectivity increased for cathodal stimulation from 3 to 6 and from 2.5 to 12 in wild type and RCS, respectively, when pulse durations decreased from 1 to 0.1 ms. Short pulses of anodal currents could rarely evoke network responses in normal and degenerate retina. For biphasic stimulation, there was no selectivity for pulses of ≥ 1 ms. However, shorter pulses with cathodal-first polarity had selectivity of 2.7 for 0.5 ms and 5.8 for 0.1 ms pulses. High selectivity for both anodal and anodal-first polarity was due to lack of network responses with short pulse durations.
Conclusions: :
These results suggest that with epiretinal stimulation (1) thresholds for direct RGC responses are comparable for normal and RCS retina, while thresholds for network-mediated responses are higher in RCS, (2) cathodal currents have the lowest activation threshold for both wild type and RCS, (3) anodal as well asanodal-first waveforms demonstrate high activation thresholds and best selectivity for durations ≤ 1 ms in both normal and RCS retina, (4) cathodal-first at pulse widths of ≤ 0.5 ms provides low activation threshold and high selectivity, and thus is most suitable for selective epiretinal stimulation strategy.
Keywords: retinal connections, networks, circuitry • signal transduction: pharmacology/physiology