Prolonged Release From Gabaergic Amacrine Cells Is Reduced By Ca2+ Buffering

Abstract

Purpose: : We previously showed that GABA release underlying GABAergic light-evoked inhibitory post-synaptic currents (IPSCs) in rod bipolar cells (BC) was prolonged relative to the initial light stimulus (Eggers and Lukasiewicz, 2006). In other systems, prolonged, asynchronous release often results from a large build-up of Ca²⁺ in the presynaptic neuron. However, as the light-evoked release of GABA requires transmission through synapses upstream from the amacrine cells, it is possible that this prolonged release is not due to inherent release properties of the amacrine cells. To test this idea we directly activated amacrine cell inputs to BCs with electrical stimulation and tested the effects of increased Ca²⁺ buffering that decreases asynchronous release.

Methods: : Whole-cell voltage clamp IPSC recordings from BCs were made from mouse retinal slices, holding the BC at the reversal potential for cation-mediated currents to isolate IPSCs. A glass stimulating electrode placed near the BC axon terminals evoked amacrine cell input to BCs. Alexa 488 was included in the pipettes to morphologically identify rod BCs. GABA_C receptor (R) IPSCs were isolated with strychnine (500 nM) and SR95531 (20 µM). Ca²⁺ buffering was increased by the addition of EGTA-AM (50 µM) to the bath, a membrane permeant analogue of slow Ca²⁺ buffer EGTA. The charge transfer (Q) and decay to 37% (D37) of the peak of IPSCs were measured.

Results: : GABA_CR-mediated evoked (e) IPSCs had an average D37 of 489 ± 84 ms, similar to the previously reported value of 472.4 ms for light-evoked IPSCs, and significantly longer than the decay of spontaneous GABA_CR IPSCs (34.1 ms) that reflects GABA_CR kinetics. When Ca²⁺ buffering was increased by the addition of EGTA-AM both the Q (-66 ±7%) and the D37 (-45 ± 6%) were significantly decreased.

Conclusions: : As the timecourse of GABA_CR IPSCs was similar whether the currents were activated by light or by direct electrical stimulation of amacrine cell processes, our results suggest that the slow timecourse of GABA_CR-mediated IPSCs is due to slow release properties of GABAergic amacrine cells and not due to prolonged light-evoked activation of amacrine cells. As this prolonged release was decreased when Ca²⁺ buffering was increased with EGTA-AM, it could be due to a build-up of Ca²⁺ in the amacrine cell terminals. This suggests that differences in Ca²⁺ buffering and Ca²⁺ channels among amacrine cell types could be a mechanism for creating previously observed differences in the release timecourse of neurotransmitter.

Keywords: bipolar cells • amacrine cells • inhibitory receptors