The transmembrane region of the nicotinic acetylcholine receptor: is it an all-helix bundle?

The nicotinic acetylcholine receptor is the best characterised member of the Ligand-Gated-Ion-Channel family of receptors. In spite of a wealth of data from molecular cloning studies these receptors have so far eluded all attempts at crystallisation; quantitative structural data are few and are at relatively low resolution. The widely accepted current model for the topology of the receptors is that of a pentameric cylindrical bundle that spans the membrane. The disposition of the transmembrane region of the individual subunits is based on hydropathy profiles calculated from sequence data which are interpreted as indicating a common structural motif of four antiparallel alpha-helices, M1 to M4. Until very recently this model has been unquestioned even though there are few direct experimental data to support it. We have constructed models of this key functional region for the nicotinic acetylcholine receptor, building out from the ion-channel. The model of the basic ion-channel comprises a five helical M2 bundle with a left-handed twist. The remainder of the region (M1, M3, M4) was homology modelled, together with M2, as a four helix antiparallel bundle per subunit, using the crystal structure of myohaemerythrin as a template. The models strongly suggest that the four helix bundle model is inappropriate and that recent suggestions of a mixed motif of helix and sheet may better accommodate the existing data.