The GABA_A receptor is the main inhibitory neurotransmitter receptor in the brain and is a member of the cysteine loop ligand-gated ion channel family. It has featured as a target for the action of many drugs and is thought to play a pivotal role in particular neurological diseases, including epilepsy and anxiety. In addition, these receptors are also subject to modulation by numerous endogenous agents in the brain, including Zn²⁺, H⁺ and neurosteroids as well as by intracellular regulatory processes such as protein phosphorylation. GABA_A receptors are now accepted to be hetero-pentamers, composed of core subunit members from the α(1-6) and β(1-3) families, which are usually co-expressed with λ(1-3) subunit family members to a stoichiometry of 2α:2β:1λ. There are also minor receptor populations where the λ subunit is believed to be replaced in the receptor by either δ or ε subunits, and a ψ subunit has also been recently cloned.

To date, trying to elucidate the underlying mechanisms by which these receptors operate has relied largely upon the use of site-directed mutagenesis which has outlined those domains on the receptor that are mostly responsible for GABA activation and modulation by benzodiazepines. In the absence of any crystalline structures for the GABA_A receptor, it has proved difficult to obtain precise information on the location of ligand binding sites and on the identity of residues involved in signal transduction. This is particularly true when the agent under investigation exhibits distinct subtype selectivity such as Zn²⁺, which is a potent inhibitor on αβ GABA_A receptors but quite weak on αβλ subunit receptors.

We have utilised two approaches to probing the molecular structure of GABA_A receptors. A molecular modelling comparison was enabled between the GABA_A receptor and the acetylcholine binding protein (AChBP) coupled with a rationale site-directed mutagenesis programme to completely resolve the molecular determinants involved in Zn²⁺ regulation of the receptor. Such an approach, coupled with patch-clamp electrophysiology, identified two main but discrete binding domains for Zn²⁺. These are coupled with residues proximal to the GABA binding site that act as signal moderators. When the binding domains are mutated, they totally accounted for the Zn²⁺ inhibitory effect. The presumed binding domains are located in the N-terminal domain, at an interfacial site between α and β subunits, whilst the other, functionally dominant site, is located in the mouth of the anion-selective ion channel. We can now demonstrate that the inclusion of the λ subunit, which markedly reduces the sensitivity to Zn²⁺, arises largely from disruption of the ion channel and N-terminal domain sites.

This work was supported by the MRC and Wellcome Trust. E.L.D. was a Maplethorpe Fellow.