Characterization of GABA- and glycine-induced currents of solitary rodent retinal ganglion cells in culture

Abstract

Ganglion cells were fluorescently labeled, dissociated from 7- to 11-day-old rodent retinas, and placed in tissue culture. Whole-cell recordings with patch electrodes were obtained from solitary cells lacking processes, which permitted a high-quality space clamp. Both GABA (1–200μM) and glycine (10–300μM) produced large increases in membrane conductance in virtually every ganglion cell tested, including ganglion cells from different size classes in both rats and mice. Taurine evoked responses similar to those of glycine, but considerably greater concentrations of taurine (150–300 μM) were necessary to observe any effect. Since 20 μM GABA produced approximately the same response as 100 μM glycine, the effects of these two concentrations were compared under various conditions. When recording with chloride distributed equally across the membrane, the reversal potential of the agonist-induced currents was approximately 0 mV. When the internal chloride was reduced by substitution with aspartate, the reversal potential shifted in a negative direction by about 42 mV, indicating that the current was carried mainly by chloride ions. Strychnine (1.5 μM) completely and reversibly blocked the actions of glycine (100 μM) but not those of GABA (20 μM); however, higher concentrations of strychnine (20 μM) nearly totally inhibited the current elicited by GABA (20 μM). The responses to glycine (100 μM) were not affected by bicucullinc methiodide (20 μM) or picrotoxinin (20 μM). In contrast, bicucullinc methiodide (10 μM) and picrotoxinin (10 μM) reversibly blocked the current evoked by GABA (20 μM); d-tubocurarinc (100 μM) only slightly decreased the response to GABA (20 μM). The antagonists were effective over a wide range of holding potentials (−90mV to + 30mV). The responses to a steady application of both GABA and glycine decayed in a few seconds when recorded under conditions of both symmetric and asymmetric chloride across the membrane. During this decay the current and conductance decreased simultaneously, reflecting receptor dcsensitization rather than a change in the driving force lor chloride caused by agonist-induced ionic fluxes. The time-course of descnsitization was usually described by a single exponential with time constants for GABA (20 μM) and glycine (100 μM) of 4.0 ± 1.6s and 4.4 ± 1.9s ( mean ± S.D. ), respectively. These results directly demonstrate that mammalian retinal ganglion cells are sensitive to the presumptive inhibitory neurotransmitters GABA and glycine; thus, synaptic inhibition at the level of the ganglion cell could contribute to receptive field properties. High doses of strychnine at least partially blocked GABA as well as glycinc-induced currents. Thus, previous studies on the intact retina using either high or indeterminate doses of strychnine to selectively antagonize the action of glycine may have inadvertently confused GABA and glycine effects. In addition, the desensitization of retinal ganglion cell responses to GABA and glycine demonstrated here complicates the interpretation of prior studies that used prolonged applications of the agonists: effects attributed to the action of exogenous GABA or glycine may have actually resulted from the desensitized state and thus the absence of a response to that agent. Work in the intact retina using these putative inhibitory transmitters must therefore be re-interpreted in light of the more precise pharmacologic data now available.