Muscarinic receptors: electrifying new insights

Proteins embedded within the cell membrane are exposed to an electric field. The function of some of these proteins is regulated by and dependent upon the transmembrane potential. The degree to which voltage influences a protein's conformation and hence its function varies, with some proteins being strongly voltage dependent while others appear to have little or no voltage dependence at all. G-protein coupled receptors (GPCRs) have been considered to be voltage insensitive because they do not have a transmembrane region rich in charged amino acids as in the case of typical voltage-sensitive proteins. Studies carried out on muscarinic receptors, a subfamily of the rhodopsin-like class of GPCRs, have, however, questioned this view. Ben-Chaim et al. (2003) demonstrated that the sensitivity of muscarinic receptors for the agonist acetylcholine (ACh) was strongly dependent on the membrane potential. Muscarinic-2 receptors (M2Rs) are predominantly expressed in the sino-atrial node, atrial myocytes and some Purkinje fibres, where they play a role in the vagal regulation of heart rate. M2R stimulation by agonists mediates dissociation of G-protein subunits, with the βγ subunit directly activating G protein-coupled inwardly rectifying potassium (GIRK) channels. The resulting outward K⁺ current (GIRK current) hyperpolarizes the cardiomyocyte membrane, in this way elongating periods between contractions, an effect known as negative chronotrophy (Salazar et al. 2007). Voltage-induced conformational changes within M2Rs were suggested to alter the association of the receptor with the trimeric G-proteins, thus regulating the activation of GIRK channels (Ben-Chaim et al. 2006).

In a recent issue of The Journal of Physiology, Navarro-Polanco et al. (2011) shed new light on the relationship between membrane potential and ligand binding to M2Rs. The authors proposed that the membrane potential modifies agonist affinity to M2Rs by inducing conformational changes within the ligand binding pocket rather than altering receptor coupling to G-proteins. This paper therefore offers an alternative mechanism for voltage sensitivity.

The authors noted that the response of M2Rs to two different agonists, ACh and pilocarpine (Pilo), varies with the voltage. The receptor response was assessed by measuring GIRK current in feline atrial myocytes using the whole-cell patch-clamp technique. The magnitude of currents activated by Pilo, which is a partial agonist, was compared with that elicited by ACh, a reference full agonist, over a wide range of concentrations. Concentration–response curves for the agonists showed that membrane depolarization decreased M2R affinity for ACh, while increasing the affinity and efficacy for Pilo. The authors considered various factors that could interfere with their analysis. The use of blockers for a range of K⁺ and Ca²⁺ channels, as well as the highly Ca²⁺-buffered intracellular solution, eliminated the possibility that channels other than GIRK contributed to the agonist-induced responses. Furthermore, the currents activated by Pilo and ACh had similar characteristics, arguing that both agonists activated the same GIRK currents. Finally, the effects of Pilo and ACh on the currents recorded from HEK-293 cells co-transfected with M2Rs and GIRK channels were voltage dependent and virtually identical to those recorded in atrial myocytes. Based on these results, Navarro-Polanco and colleagues (2011) propose that the agonist-specific voltage dependence of M2Rs is due to voltage-induced conformational changes occurring within the agonist binding pocket.

In order to attain a direct indication of the actual conformational changes within the receptor in response to voltage, the authors measured M2R gating currents and assessed how they were modified by agonists. Gating currents correspond to the net charge moved within the receptor upon changes in membrane potential. To record the minuscule gating currents, the authors favoured the cut-open voltage-clamp technique using oocytes expressing heterologous M2Rs. This technique allows clamping of a large surface of the membrane (about half of the oocytes) thereby including many receptors, while ensuring a fast clamping speed. The charge that moved in response to voltage pulses was calculated as the integral of the gating currents. The resulting sigmoidal charge versus voltage distribution (Q–V curve) had a half-point of maximal charge movement (V_1/2) at around –70 mV and the apparent charge moved per receptor, the effective valence, was 0.55 e₀. This value is relatively small compared to the effective valence of voltage-gated K⁺ channels, which has been reported to be ∼12 e₀ (Hille, 2001), indicating that M2Rs undergo rather subtle conformational changes. Importantly, the gating currents occurred at membrane potentials within the physiological range. These gating currents were modulated by agonists: ACh suppressed the gating charge displacement, whereas Pilo increased it in a dose-dependent manner. Interestingly, ACh concentrations that inhibited the M2R gating currents elicited GIRK currents lacking voltage dependence. In contrast, increasing Pilo concentrations enhanced gating charge movement, which was accompanied by pronounced voltage dependence of the ionic currents. This suggests that M2Rs confer voltage sensitivity to GIRK currents.

Inspired by the diverse effects of agonists on gating currents, the authors questioned whether mutations within the ligand binding pocket could also affect gating charge displacement. Five conserved amino acids within the ACh binding site were systematically replaced with alanine (W99A, D103A, Y104A, S107A and Y403A). Three types of mutation were identified: those that shifted the Q–V distribution towards hyperpolarized potentials (D103A, Y104A and Y403A), depolarized potentials (W99A) or did not perturb gating charge movement (S107A). None of these mutants showed a decrease in the effective valence indicating that these amino acids do not form part of the primary voltage sensor. The authors also tested how two of the mutants, S107A and W99A, determine the voltage dependence of ACh-activated GIRK currents. As might be expected, both mutants showed significantly decreased affinity for ACh; however, the S107A mutant, which did not affect gating charge displacement, mediated GIRK currents with similar voltage dependence to that of wild-type. Interestingly, one of the mutations which altered the gating charge movement (W99A) was associated with GIRK currents lacking voltage dependence. This suggests that the affinity of W99A-M2R for ACh was the same at positive and negative potentials implying that the W99A mutation might have disrupted a link between voltage sensor movement and conformational changes within the ligand binding pocket. It would be interesting to test the voltage dependence of GIRK currents using Pilo and other agonists, as the ligand binding pockets may not be identical for all agonist due to their structural differences. In addition, structural analogues of Pilo and ACh could be employed to determine the structural features of the ligands responsible for their opposing effects. It would also be useful to know whether mutants that shifted the Q–V relationship towards hyperpolarized potentials could maintain the voltage sensitivity of agonist binding. This might help us to understand how the primary voltage sensor is linked to the ligand binding site.

A key question still remains: what is the origin of the voltage sensor? Previously proposed candidates for the primary voltage sensor, D120 and R121 (Ben-Chaim et al. 2006), were excluded in this study as the D120N–R121N mutant generated gating currents with the same effective valence as the wild-type M2R. Mutation of the negatively charged amino acid within the transmembrane electric field (D69A) resulted in a mutant with impaired protein expression. Therefore, the contribution of D69 to voltage sensing has not been determined.

This research provides new important information on how muscarinic receptors operate at the molecular level. The authors introduce a new idea in the field, suggesting that voltage controls M2R function by changing its affinity for specific agonists, independent of their coupling to downstream proteins. The observed correlation between M2R gating charge movement and voltage sensitivity of agonist-induced ionic currents suggests that when gating charge movement is suppressed, the conformation of the ligand binding pocket becomes voltage independent, such that agonist binds to the receptor with the same affinity at positive and negative potentials. In voltage-gated ion channels there is a distinct voltage sensor region, composed of charged amino acids. In contrast, M2Rs have only two charged amino acids in the transmembrane electric field and at least one of them (D103) has been shown not to be a part of the voltage sensor. This implies that dipoles rather than charged residues move in the electric field in response to changes in membrane potential. Voltage-sensing residues are unlikely to be localized within one relatively well-defined domain of the receptor, as in the case of voltage-gated K⁺ channels, but rather distributed within protein regions that are confined within the membrane electric field. The exact structure of the voltage sensor may therefore be difficult to identify.

In heart, ACh at concentrations occurring under physiological conditions activates GIRK channels, which carry currents with a conductance that decreases when their membranes are depolarized. GIRK current activation results in elongated periods between contractions without affecting the contractile force (Salazar et al. 2007). The higher M2R affinity for ACh at negative potentials may contribute to this phenomenon. The discovery that Pilo has the opposite voltage sensitivity to ACh might form the basis for the discovery of novel compounds that interact with ACh receptors in specific voltage dependent ways to modulate cardiac bioelectricity and contractile function more selectively.