The GlyR is pentameric with each subunit
The GlyR is pentameric, with each subunit consisting of an extracellular ligand binding domain (LBD), four transmembrane (TM) segments, and a large intracellular loop between TM3 and TM4. Binding of neurotransmitter produces conformational changes that are rapidly transmitted to the membrane-spanning pore, ultimately opening the ion channel. Previously we showed that mutation of the aspartate-97 residue to arginine (D97R) in loop A in the LBD of the α1 subunit favors spontaneous transitions of channels to the open state (Beckstead et al., 2002). Subsequently we found that this spontaneous activity was due to the disruption of a putative intersubunit electrostatic bond involving D97 and at least R119 of an adjacent subunit (Todorovic et al., 2010). Single channel recordings showed that clusters of spontaneous D97R GlyR openings had a Popen of 0.91, very similar to that observed when wildtype α1 GlyR are exposed to saturating concentrations of glycine. A high Popen of 0.96 is also observed in α1β GlyR exposed to saturating glycine concentrations; however these receptors exhibit a Popen of 0.54 in response to a maximally-effective concentration of taurine (Lape et al., 2008). We thus hypothesized that the efficacy difference between glycine and taurine might be partially due to differences in their abilities to break this intersubunit electrostatic bond in wildtype GlyR, and that in mutated receptors, in which this bond is broken, the efficacy of taurine would increase. Lape et al. (2008) suggested that the efficacies of agonists and partial agonists in the nicotinic C34 receptor superfamily are defined by transitions from a closed state to an intermediate closed state (flipped) that follows ligand binding but precedes channel opening. According to this model, efficacy is determined by the rates governing the transitions between closed and flipped states, as well as transitions between flipped and open. One conclusion reached by the authors was that the determination of whether a compound acts as a partial or full agonist occurs early in the process of receptor activation. In the present report we focus on the D97 residue, which is positioned in the ligand binding domain to play a role in the initial steps of receptor activation.
Discussion Historically, efficacy at ion channels has been defined by the ratio of transitions to and from the open state of the channel (E=β/α), where β describes the transition rate from the shut to open state of the channel and α describes the closing rate. There is thus no upper bound on efficacy, and so compounds described as ‘full agonists’ are simply those with the highest efficacy found to date, with partial agonists possessing lower efficacy. The remarkable complexity of ligand-gated ion channel structures suggests that ligand binding initiates a multitude of protein conformations, ultimately leading to channel opening. For the acetylcholine receptor, this process was mapped and the gating of the receptor was described as a Brownian conformational wave that begins in the extracellular portion of the receptor at the LBD and proceeds to the pore (Grossman et al., 2000, Auerbach, 2005). In the present report we focused on the D97 residue, which is well positioned in the LBD to play a role in receptor activation. This amino acid is conserved across the entire cys-loop subunit superfamily, residing at a subunit interface near residues previously implicated in ligand binding (Brejc et al., 2001). Our previous work strongly suggested that D97 forms an intersubunit electrostatic bond with R119 (Todorovic et al., 2010), a residue earlier shown to play a role in determining glycine affinity (Grudzinska et al., 2005). In our previous study we engineered cysteine mutations at the D97 and R119 positions, and used those to show that receptor function could be markedly affected by addition of oxidizing, cross-linking or reducing agents (Todorovic et al., 2010). This strongly suggests that these two residues are close enough together to be capable of forming intersubunit disulfide bonds. However, not all evidence fully supports this assertion. D97 and R119 are 6.87Å apart in the zebrafish (3JAD) strychnine-bound α1 GlyR crystal structure (Du et al., 2015), and the equivalent residues in the human (5CFB) strychnine-bound GlyR α3 structure are 6.67Å apart (Huang et al., 2015). Charged residues 4Å apart or closer are interpreted as interacting electrostatically (Kumar and Nussinov, 2002). Unfortunately, to date no unliganded GlyR structures have been determined, and so it is not known how far apart these two amino acids are when no ligand has bound. Interestingly, in the zebrafish (3JAE) glycine-bound α1 GlyR structure (Du et al., 2015), the D97 and R119 residues are found 9.45Å apart, consistent with our hypothesis that agonist binding helps break this putative salt bridge.