Cyclic nucleotide-gated channels: structural basis of ligand efficacy and allosteric modulation

Abstract

Most working proteins, including metabolic enzymes, transcription regulators, and membrane receptors, transporters, and ion channels, share the property of allosteric coupling. The term ‘allosteric’ means that these proteins mediate indirect interactions between sites that are physically separated on the protein. In the example of ligand-gated ion channels, the binding of a suitable ligand elicits local conformational changes at the binding site, which are coupled to further conformational changes in regions distant from the binding site. The physical motions finally arrive at the site of biological activity: the ion-permeating pore. The conformational changes that lead from the ligand binding to the actual opening of the pore comprise ‘gating’. In 1956, del Castillo and Katz suggested that the competition between different ligands at nicotinic acetylcholine receptors (nAChRs) could be explained by formation of an intermediate, ligand-bound, yet inactive state of the receptor, which separates the active state of the receptor from the initial binding of the ligand (del Castillo & Katz, 1957). This ‘binding-then-gating’, two-step model went beyond the then-prevailing drug-receptor model that assumes a single bimolecular binding reaction, and paralleled Stephenson's conceptual dichotomy of ‘affinity’ and ‘efficacy’ (Stephenson, 1956). In 1965 Monod, Wyman and Changeux presented a simple allosteric model (the MWC model) (Monod et al. 1965) that explained the cooperative binding of oxygen to haemoglobin; it was adopted as an important paradigm for ligand-gated channels soon after its initial formulation (Changeux et al. 1967; Karlin, 1967; Colquhoun, 1973).