Cl- accumulation does not account for the depolarizing phase of the synaptic GABA response in hippocampal pyramidal cells

It has been proposed that the depolarizing phase of the biphasic synaptic GABA response could be mediated by HCO3- passing through GABA(A) channels after dissipation of the transmembrane Cl- gradient due to intracellular Cl- accumulation. To test this hypothesis, giant GABA-mediated postsynaptic currents (GPSCs) were recorded from pyramidal cells in slices of adult guinea pig hippocampus in the presence of 4-aminopyridine. GPSCs consisted of an early outward current (GABA(A) component) followed by a late inward current (GABA(D) component). Spontaneous outward inhibitory postsynaptic currents (IPSCs) occurred during the GABA(D) component of the GPSC. GPSCs that were evoked 1-12 s after the preceding GPSC (short interval, siGPSCs) showed no GABA(A) component even though in many cells the amplitude of the siGPSC was greater than the amplitude of the GABA(D) component of the preceding spontaneous GPSC. In addition, the siGPSC evoked during the GABA(D) component of a spontaneous GPSC was an outward current. To test whether the siGPSC lacked a GABA(D) component because it was generated predominantly at the soma, where less of an increase in [Cl-](i) would occur, picrotoxin was applied to the soma of the pyramidal cell. To the contrary, this focal application of picrotoxin caused less of a reduction in the amplitude of the siGPSC than in the amplitude of the GABA(A) component of the GPSC. Furthermore when a GPSC and siGPSC were evoked 10 s apart using identical stimuli, the area under the outward current curve was sometimes greater for the siGPSC than for the GPSC and yet the siGPSC had no inward component. This result indicates that even when the location of Cl- entry was the same, more Cl- could enter the cell during the siGPSC than during the outward component of the GPSC and yet not lead to an inward current. In addition, when the second of two identical stimuli was applied during the inward GABA(D) component of the first evoked GPSC, the GABA(A) response it generated was always outward, demonstrating that the equilibrium potential for GABA(A) responses did not become more positive than the holding potential during a GPSC. Finally, evoking GPSCs at a hyperpolarized potential revealed that the siGPSC actually lacked a GABA(D) conductance. These results disprove the Cl- accumulation hypothesis of the synaptic depolarizing GABA response and suggest the possibility that a separate channel type may mediate the GABA(D) component of the GPSC.