Whole-cell voltage-clamp study of the fading of GABA-activated currents in acutely dissociated hippocampal neurons.

The lability of the responses of mammalian central neurons to gamma-aminobutyric acid (GABA) was studied using neurons acutely dissociated from the CA1 region of the adult guinea pig hippocampus as a model system. GABA was applied to the neuronal somata by pressure ejection and the resulting current (IGABA) recorded under whole-cell voltage clamp. In initial experiments we examined several basic properties of cells in this preparation. Our data confirm that passive and active membrane properties are similar to those which characterize cells in other preparations. In addition, GABA-dependent conductance (gGABA), reversal potential (EGABA), and the interaction of GABA with pentobarbital and bicuculline all appeared to be normal. Dendritic GABA application could cause depolarizing GABA responses, and somatic GABA application caused hyperpolarizations due to chloride (Cl-) movements. Repetitive brief applications (5-15 ms) of GABA (10(-5) to 10(-3) M) at a frequency of 0.5 Hz led to fading of successive peaks of IGABA until, at a given holding potential, a steady state was reached in which IGABA no longer changed. Imposing voltage steps lasting seconds during a train of steady-state GABA responses led initially to increased IGABA that then diminished with maintenance of the step voltage. The rate of decrease of IGABA at each new holding potential was independent of the polarity of the step in holding potential but was highly dependent on the rate of GABA application. Application rates as low as 0.05 Hz led to fading of IGABA, even with activation of relatively small conductances (5-15 nS). Since IGABA evoked by somatic GABA application in these cells is carried by Cl-, the Cl- equilibrium potential (ECl) is equal to the reversal potential for IGABA, i.e., to EGABA. The fading of IGABA with changes in holding potential can be almost entirely accounted for by a shift in ECl resulting from transmembrane flux of Cl- through the GABA-activated conductance. Maneuvers that prevent changes in the intracellular concentration of Cl-ions, [Cl-]i, including holding the membrane potential at EGABA during repetitive GABA application or buffering [Cl-]i with high pipette [Cl-], prevent changes in EGABA. Desensitization of the GABA response (an actual decrease in gGABA) occurs in these neurons during prolonged application of GABA (greater than 1 s) but with a slower time course than changes in EGABA. Whole-cell voltage-clamp techniques applied to tissue-cultured spinal cord neurons indicated that rapid shifts in EGABA result from repetitive GABA application in these cells as well.(ABSTRACT TRUNCATED AT 250 WORDS)