Heteromeric wild-type/mutant potassium channel subunit composition as a major determinant of channelopathy phenotype in heterozygous patients

Gly³⁷⁵ is positioned in the S6 pore-lining helix with the side chain facing away from the pore lumen (Fig. 1). The glycine introduces a noticeable bend in S6 and has been proposed to function as a “gating hinge” (Tao and MacKinnon, 2019). In addition, Gly³⁷⁵ is located in the “deep pore” region that has been proposed to undergo structural rearrangements during the voltage-dependent channel gating (Chen et al., 2014). Therefore, substitution of Gly³⁷⁵ with a larger and positively charged arginine would be expected to affect channel gating. Indeed, Geng et al. (2023) found that co-expression of WT and G375R mutant subunits at 1:1 ratio in Xenopus laevis oocytes shifted the half-maximal activation voltage (V_1/2) of BK channels in the hyperpolarized direction by a staggering 120 mV, based on analysis of the macroscopic currents recorded from excised inside-out membrane patches. A similar effect was observed for whole-cell currents. This profound GOF would cause opening of mutant BK channels at much more negative potentials than WT channels, underpinning the molecular mechanism of the severe phenotype observed in patients with a de novo G375R mutation.

Since the macroscopic current analysis indicated that the G375R mutation causes a large left-shift in V_1/2, Geng et al. (2023) used single-channel current analysis to further investigate the functional consequences of the G375R mutation on the single channel open probability (Po) and conductance. Consistent with the results from the macroscopic current analysis, single channel current analysis of homomeric WT and mutant channels indicated that the G375R mutation left-shifted the voltage dependence of Po with an observed shift in V_1/2 of >180 mV.

In addition to the profound shift in voltage dependence of Po, the homomeric G375R channels exhibited an ∼40% decrease in the single-channel conductance compared to WT channels. As proposed by Geng et al. (2023), this DOF could be caused by the decrease in the pore vestibule volume due to the substitution of Gly³⁷⁵ with a larger arginine and/or by an electrostatic repulsion of the permeant K⁺ ions introduced by the substitution. As the macroscopic current analysis indicates, for the averaged cellular response this respectable DOF is thwarted by the GOF effect of the G375R mutation. Importantly, in native cells BK channels are also regulated by a variety of additional factors, including alternative splicing, regulatory β and ɣ subunits, intracellular Mg²⁺, and cholesterol (Cui, 2021; Gonzalez-Perez et al., 2014; Shipston and Tian, 2016; Bukiya and Dopico, 2019). Additional studies are necessary to investigate how the GOF and DOF effects of the G375R mutation integrate with the common pathways of BK channel regulation.

While it is clear that Gly³⁷⁵ is an important residue for BK channel gating, the mechanism of the Po regulation by Gly³⁷⁵ is not clear. Is it the structural flexibility introduced by the glycine residue, and/or size and charge at the 375 position that regulates Po? It has been previously reported that G375D substitution shifted the V_1/2 of BK channel activation to the hyperpolarized potentials by 55 mV, based on the analysis of macroscopic currents recorded in the inside-out patch-clamp configuration (Chen et al., 2014). This shift is less than the one caused by the G375R mutation; however, overall it is in the same direction despite the opposite charges of the substituted residues. Still, it is premature to rule out the possible contribution of a charge at the 375 position without the single-channel current analysis as it is possible the G375D mutation causes opposite effects on Po and single-channel conductance than G375R and this response is masked in the macroscopic currents. Alternatively, the charge substitutions could affect BK channel gating by changing the rotamer preferences of ionized side chains located in the vicinity of the 375 position, as it was shown for the effect of residue substitutions on the ion selectivity of pentameric ligand-gated ion channels (Cymes and Grosman, 2016).

The G375R mutation left-shifted the V_1/2 of BK channel activation in the absence and also presence of intracellular Ca²⁺, although, the shift was seemingly larger in the presence of Ca²⁺, based on the macroscopic currents recorded from excised inside-out membrane patches. Interestingly, for the G375D mutation the left-shift in V_1/2 was smaller in the presence of Ca²⁺ than in the absence (Chen et al., 2014). Single-channel based studies involving substitutions of different size and charge residues at the position 375, investigated over the range of Ca²⁺ concentrations, are needed to investigate if and how Ca²⁺-dependent gating and effects of Gly³⁷⁵ are allosterically coupled.

Many diseases are manifested in heterozygous patients that carry both a WT and a mutant allele. For channelopathies associated with potassium channels this would lead to the assembly of heteromeric as well as homomeric channels from WT and mutation-carrying subunits. Geng et al. (2023) used the binomial equation to predict BK channel subunit assembly from WT and mutant subunits, assuming a random subunit assembly and an even production of WT and mutant subunits. They point out that the most populous channel assembly would be a heteromer with two mutant and two WT subunits, accounting for 38% of the assembled channels, followed by a heteromer with either three mutant and one WT subunits, or vice-versa, each accounting for 25% of the assembled channels (Fig. 2). Homomeric WT and mutant channels would then each account for a meager 6% of the assembled channels. To test this random assembly scenario, Geng et al. (2023) mimicked BK channel assembly in heterozygous patients by injecting equal amounts of cRNA encoding the WT and G375R mutant subunits into Xenopus oocytes. The analysis of histograms of the V_1/2 values for single-channel recordings from 33 assembled channels indicated that a heteromer with 2 mutant and 2 WT subunits was indeed the most populous channel, accounting for 34% of the assembled channels, followed by a heteromer with 3 WT and 1 mutant subunits, and a heteromer with 1 WT and 3 mutant subunits, accounting for 30 and 21% of the assembled channels, respectively (Fig. 2). Homomeric WT and mutant channels accounted for only 3 and 12% of the assembled channels, respectively. Overall the experimentally suggested channel stoichiometry distribution was consistent with the theoretical predictions. Simulations of 1 million groups of 33 assembled channels suggested that the differences between the experimentally suggested and theoretically predicted populations are most likely insignificant, although, some tendency towards preferential assembly and/or uneven subunit expression can’t be excluded. Nevertheless, the main experimental finding is loud and clear: in heterozygous patients with equal expression of WT and mutant subunits, 85% of the assembled channels are expected to be heteromers.

The functional changes due to the GOF and DOF effects of the G375R mutation showed a linear dependence on the number of mutant subunits present in a tetramer. Substitution of each WT subunit with a mutant one resulted in a ∼60 mV leftward shift in V_1/2 and ∼10% decrease in the single-channel conductance (Fig. 2). Such a linear dependence suggests that there is little subunit cooperativity in the GOF and DOF effects. Interestingly, the GOF and DOF effects for channels with two WT and mutant subunits assembled in adjacent and diagonal arrangement were not clearly distinct. It is possible that analysis of a larger number of channels could reveal some differences for the two arrangements.

The experimental results suggest that in heterozygous individuals over 95% of the assembled channels will have at least one mutant subunit incorporated and, therefore, will display an assembly-mediated dominant GOF phenotype, even though only 50% of all expressed channel subunits carry the mutation. Moreover, due to the huge leftward shift introduced by the G375R mutation and population advantage, the heteromeric and homomeric mutant channels would completely functionally dominate homomeric WT channels under physiological conditions. For instance, assuming a normal neuronal resting potential of approximately −70 mV, the first to activate would be the homomeric mutant channels with the averaged V_1/2 of approximately −80 mV and some heteromeric channels assembled from one WT and three mutant subunits with the averaged V_1/2 of approximately −20 mV. Activation of the channels with these two assemblies would hyperpolarize the membrane potential below the normal resting potential and up to the K⁺ reversal potential, making it harder to activate BK channels with other subunit assemblies. Essentially, this would make homomeric WT channels and even channels with only one G375R subunit functionally irrelevant. Not only at rest but also during a strong excitation, e.g., large amplitude post-synaptic potentials, activation of BK channels assembled from predominantly mutant subunits at the onset of depolarization would prevent membrane potential from reaching values necessary for the activation of channels assembled from predominantly WT subunits.

Because of the activation at resting membrane potential, the homomeric mutant channels would have substantial impact on cellular excitability, despite accounting for only ∼10% of the assembled BK channels. The hyperpolarization caused by the opening of homomeric mutant channels could prevent action potential initiation. The physiological effect of the homomeric channels in native cells would also have to be assessed within the context of the regulatory subunits, intracellular Ca²⁺ levels and other BK channel regulatory pathways, that could shape the physiological impact of G375R mutation in a cell type–specific manner.

The study by Geng et al. (2023) drives home the important message that in heterozygous patients, assuming equal expression of both WT and mutant subunits, over 95% of disease-associated potassium channels in the membrane will contain at least one mutant subunit. This suggests a potential therapeutic strategy: a compound that inhibits (for GOF mutations) or activates (for DOF mutations) mutant subunit-containing channels with higher affinity than WT channels would increase the relative contribution of channels composed exclusively of WT subunits, potentially rescuing the WT phenotype. The benefit of such a strategy would have to be weighed against whether enough predominantly WT channels might be available to sustain normal physiological function.

Taken together, the work of Geng et al. (2023) highlights that the subunit composition and phenotype of a mutation have major implications for preclinical studies of potassium channel channelopathies and must be investigated in detail to better assess potential physiological impact of the genetic mutation.

Christopher J. Lingle served as editor.

This work was supported by the National Institute of General Medicine grant R01GM124020 and National Cancer Institute grant R01CA252969.