CORELEASE OF ACETYLCHOLINE AND GABA BY THE STARBURST AMACRINE CELLS

Rabbit retinas were isolated from the eye and maintained in vitro. When they were incubated for 60 min in the presence of H-3-GABA, subsequent autoradiography showed radioactivity to be present primarily in amacrine cells. Under these conditions, most of the radioactivity contained in the retinas remained in the chemical form of GABA. Autoradiography and immunohistochemistry of alternate sections showed the amacrine cells that accumulate H-3-GABA to be the same cells that contain endogenous GABA immunoreactivity. These include the starburst cells, the indoleamine-accumulating cells, and other, as yet unidentified amacrine cells. The localization confirms previous immunohistochemical findings. When retinas containing H-3-GABA were exposed to elevated concentrations of K+, their content of H-3-GABA decreased. Autoradiography showed a reduced H-3-GABA content in all of the cells that contained H-3-GABA. Since those include the starburst cells, previously shown to be cholinergic, the finding demonstrates that the starburst cells release both ACh and GABA. Retinas simultaneously labeled with C-14-GABA and H-3-ACh were superfused, and the release of radioactive compounds from the retina was studied. Depolarization by elevated K+ caused an increased recovery of both ACh and GABA in the superfusate, but the predominant mechanisms of their release appeared to be different. The stimulated release of ACh was entirely Ca2+ dependent, while the release of radioactivity originating from GABA was much less so. A concentration-dependent counterflux (homoexchange) of intracellular GABA was demonstrated by raising the extracellular concentration of GABA (or nipecotic acid). These results suggest that a large outward flux of GABA occurs via the GABA transporter, probably by the potential-sensitive mechanism studied by Schwartz (1982, 1987). Stimulation of double-labeled retinas by flashing light or moving bars always increased the release of ACh, and the release was entirely dependent on the presence of extracellular Ca2+. Stimulation with light never caused a detectable release of GABA. This was unexpected, since the two neurotransmitters are present in the same amacrine cells: stimulation adequate to release one neurotransmitter should release both. Control experiments showed the following: (1) GABA synthesized endogenously from radiolabeled glutamate was released in the same overall way as GABA accumulated from the medium; (2) inhibition of GABA-transaminase suppressed the degradation of GABA within the retina but did not unmask a light-stimulated GABA release; (3) saturating concentrations of agents that affect GABA reuptake (nipecotic acid, SKF 89976A, unlabeled GABA) increased the recovery of radioactive GABA in the perfusate but did not unmask a light-stimulated release of GABA; and (4) treatment with APB did not reveal a light-evoked response, indicating that its absence was not due to counter-balancing of the retina's ON and OFF responses. An interpretation of these results is that the carrier-mediated release of GABA is greater, overall, than GABA's secretion in synaptic vesicles, so that the carrier-mediated component overwhelms vesicular release when the whole tissue is studied. If this is so, however, the carrier-mediated release must be relatively insensitive to the membrane potential, so that release is detectably increased by the large, long-lasting depolarizations induced by rises in extracellular K+, and not to the small (and brief) depolarizations caused by stimulation of the retina with light.