High conductance ␥-aminobutyric acid type A (GABA A ) channels (>40 picosiemens (pS)) have been reported in some studies on GABA A channels in situ but not in others, whereas recombinant GABA A channels do not appear to display conductances above 40 pS. Furthermore, the conductance of some native GABA A channels can be increased by diazepam or pentobarbital, which are effects not reported for expressed GABA A channels. GABARAP, a protein associated with native GABA A channels, has been reported to cause clustering of GABA A receptors and changes in channel kinetics. We have recorded single channel currents activated by GABA in L929 cells expressing ␣ 1 , ␤ 1 , and ␥ 2S subunits of human GABA A receptors. Channel conductance was never higher than 40 pS and was not significantly increased by diazepam or pentobarbital, although open probability was increased. In contrast, in cells expressing the same three subunits together with GABARAP, channel conductance could be significantly higher than 40 pS, and channel conductance was increased by diazepam and pentobarbital. GABARAP caused clustering of receptors in L929 cells, and we suggest that there may be interactions between subunits of clustered GABA A receptors that make them open co-operatively to give high conductance "channels." Recombinant channels may require the influence of GABARAP and perhaps other intracellular proteins to adopt a fuller repertoire of properties of native channels.

Single channel currents were activated by GABA (0.5 to 5 microM) in cell-attached and inside-out patches from cells in the dentate gyrus of rat hippocampal slices. The currents reversed at the chloride equilibrium potential and were blocked by bicuculline (100 microM). Several different kinds of channel were seen: high conductance and low conductance, rectifying and "nonrectifying." Channels had multiple conductance states. The open probability (Po) of channels was greater at depolarized than at hyperpolarized potentials and the relationship between Po and potential could be fitted with a Boltzmann equation with equivalent valency (z) of 1. The combination of outward rectification and potential-dependent open probability gave very little chloride current at hyperpolarized potentials but steeply increasing current with depolarization, useful properties for a tonic inhibitory mechanism.

Spontaneous, single channel, chloride currents were recorded in 48% of cell-attached patches on neurones in the CA1 region of rat hippocampal slices. In some patches, there was more than 1 channel active. They showed outward rectification: both channel conductance and open probability were greater at depolarized than at hyperpolarized potentials. Channels activated by gamma-aminobutyric acid (GABA) in silent patches on the same neurones had similar conductance and outward rectification. The spontaneous currents were inhibited by bicuculline and potentiated by diazepam. It was concluded that the spontaneously opening channels were constitutively active, nonsynaptic GABA(A) channels. Such spontaneously opening GABA(A) channels may provide a tonic inhibitory mechanism in these cells and perhaps in other cells that have GABA(A) receptors although not having a GABA(A) synaptic input. They may also be a target for clinically useful drugs such as the benzodiazepines.

Native GABA(A) channels display a single-channel conductance ranging between approximately 10 and 90 pS. Diazepam increases the conductance of some of these native channels but never those of recombinant receptors, unless they are coexpressed with GABARAP. This trafficking protein clusters recombinant receptors in the membrane, suggesting that high-conductance channels arise from receptors that are at locally high concentrations. The amphipathic (MA) helix that is present in the large cytoplasmic loop of every subunit of all ligand-gated ion channels mediates protein-protein interactions. Here we report that when applied to inside-out patches, a peptide mimicking the MA helix of the gamma2 subunit (gamma(381-403)) of the GABA(A) receptor abrogates the potentiating effect of diazepam on both endogenous receptors and recombinant GABA(A) receptors coexpressed with GABARAP, by substantially reducing their conductance. The protein interaction disrupted by the peptide did not involve GABARAP, because a shorter peptide (gamma(386-403)) known to compete with the gamma2-GABARAP interaction did not affect the conductance of recombinant alphabetagamma receptors coexpressed with GABARAP. The requirement for receptor clustering and the fact that the gamma2 MA helix is able to self-associate support a mechanism whereby adjacent GABA(A) receptors interact via their gamma2-subunit MA helices, altering ion permeation through each channel. Alteration of ion-channel function arising from dynamic interactions between ion channels of the same family has not been reported previously and highlights a novel way in which inhibitory neurotransmission in the brain may be differentially modulated.

Hippocampal pyramidal neurons potentially express multiple subtypes of GABA(A) receptors at extrasynaptic locations that could therefore respond to different drugs. We activated extrasynaptic GABA(A) receptors in cultured rat hippocampal pyramidal neurons and measured single-channel currents in order to compare the actions of two drugs that potentially target different GABA(A) receptor subtypes. Despite the possible difference in receptor targets of etomidate and diazepam, the two drugs were similar in their actions on native extrasynaptic GABA(A) receptors. Each drug produced three distinct responses that differed significantly in current magnitude, implying heterogeneous GABA(A) receptor populations. In the majority of patches, drug application increased both the single-channel conductance (>40 pS) and the open probability of the channels. By contrast, in the minority of patches, drug application caused an increase in open probability only. In the third group high-conductance channels were observed upon GABA activation and drug application increased their open probability only. The currents potentiated by etomidate or diazepam were substantially larger in patches displaying high-conductance GABA channels compared to those displaying only low-conductance channels. Factors contributing to the large magnitude of these currents were the long mean open time of high-conductance channels and the presence of multiple channels in these patches. In conclusion, we suggest that the local density of extrasynaptic GABA(A) receptors may influence their single-channel properties and may be an additional regulating factor for tonic inhibition and, importantly, differential drug modulation.