Steady-state catecholamine distribution in chromaffin granule preparations: a test of the pump-leak hypothesis of general anesthesia

The molecular mechanism of general anesthesia is not understood. Possible modes of action include binding at a protein site, such as a receptor or channel, or physical effects on membrane lipid properties. The pump-leak hypothesis suggests that anesthetics perturb the bilayer of synaptic vesicles, thereby increasing ionic permeability. This results in decay of proton gradients required for transport and accumulation of neurotransmitters. The subsequent loss of neurotransmitters from synaptic vesicles reduces the efficiency of synaptic transmission and results in the anesthetized state. We have determined the effects of general anesthetics on certain parameters of enzyme activity and membrane permeability relevant to the pump-leak hypothesis. We used chromaffin granules as a convenient model system and focused on clinically relevant anesthetic concentrations (ED50), quantitative measurements of permeability changes, and the kinetics of gradient decay. General anesthetics at ED50 have little or no effect on the proton-transport ATPase activity, but do cause modest increments in proton permeability that change the catecholamine distribution in actively pumping chromaffin granule preparations. We found that pH gradients do not collapse entirely under these conditions and that only a fraction of total catecholamine is lost from the chromaffin granules. When total collapse is induced by other means, efflux of catecholamines occurs with a half-time near 30 min. These results suggest that if the pump-leak hypothesis is valid, then very small losses of catecholamines must be sufficient to induce anesthesia. We conclude that the weight of evidence favors other mechanisms, notably direct binding of anesthetics to sensitive proteins.