Vanderbilt Institute of Chemical Biology



Discovery at the VICB







How Potassium Channel Mutations Lead to Heart Arrhythmias

By: Carol A. Rouzer, VICB Communications
Published:  March 8, 2018


Extensive studies of multiple mutations of the KCNQ1 channel reveal that protein instability is a major cause of hereditary type 1 long QT syndrome.


KCNQ1 is a voltage-gated potassium channel, meaning that it opens and closes in response to changes in membrane potential. It plays an important physiological role in many tissues, including the gastrointestinal tract, lungs, ears, and heart. In the heart, KCNQ1 forms a complex with the KCNE1 regulatory protein that modulates its response to voltage changes and its ability to conduct potassium ions (Figure 1). This KCNQ1/KCNE1 complex is responsible for the cardiac myocyte's slow delayed rectifier current, a flow of potassium ions out of the cell that occurs late in the action potential and is critical for membrane repolarization. Mutations that lead to loss of function (LOF) of KCNQ1 are the cause of type 1 long QT syndrome (LQTS), a hereditary heart condition characterized by an abnormally long period between the beginning of the QRS complex and the end of the T wave on an electrocardiogram. This QT segment designates the time between the beginning and the end of the electrical activity associated with contraction of the ventricles of the heart. LQTS predisposes the heart to arrhythmias, and its effects can range from asymptomatic to life-threatening. Over 600 mutations of the gene encoding KCNQ1 are associated with LQTS, and many more "variants of unknown significance" (VUS) have been identified by genome sequencing. However, the mechanisms by which these mutations alter KCNQ1 function in the heart are poorly understood, making it difficult for physicians to predict the best way to protect patients carrying these mutations from potentially dangerous arrhythmias. To address this conundrum, VICB members Chuck Sanders Jens Meiler, and Jarrod Smith along with their collaborators Carlos Vanoye and Alfred George (Northwestern University) have explored the mechanisms by which LQTS-associated mutations lead to KCNQ1 channel dysfunction [H. Huang, et al. (2018) Sci. Adv., 4, eaar2631]. Several Ph.D. and postdoctoral trainees in the Sanders and Meiler labs contributed to this work, including first author Hui Huang, a Sanders Lab postdoc.



FIGURE 1. (Top) Topology of the KCNQ1 and KCNE1 proteins in the membrane. KCNQ1 comprises six transmembrane helices of which S1-S4 form the voltage-sensing domain and S5 and S6 form the potassium channel. (Bottom) Topology of the slow delayed rectifier current channel comprising four subunits each of KCNQ1 and KCNE1. Figure reproduced under the Creative Commons Attribution-Noncommercial 3.0 License from X. Sun, et al. (2012) Front. Pharmacol., 3, 63.



The study was built upon prior work to develop a medium-throughput assay to assess the consequences of mutations in KCNQ1. The assay utilized whole cell electrophysiology to measure the amplitude, voltage-dependence, and activation and deactivation kinetics of wild-type or mutant KCNQ1 proteins expressed along with wild-type KCNE1 in CHO-K1 cells. Evaluation of 78 different KCNQ1 variants enabled the investigators to classify mutations according to their likelihood of causing LQTS, with LOF defined as failure to generate a potassium current of at least 65% of the magnitude of the wild-type current.


Although the assay provided a new and efficient method to identify potentially troublesome mutations, it offered little insight into the mechanisms by which these mutations cause LOF in KCNQ1 channels. To begin to address this question, the researchers chose 51 variants for further study. Of these, 17 were known to cause LQTS, 21 were VUS mutations, and 13 were mutations predicted to be harmless on the basis that they were found in KCNQ1 orthologs of other species. All mutations were located in the voltage sensor domain of the protein, comprising helices S1-S4 (Figure 2). For each variant KCNQ1 protein, the investigators engineered a Myc epitope into the extracellular loop between helices S1 and S2 of the channel domain. After expressing the engineered wild-type protein in HEK293 cells and verifying that the addition of the epitope had no effect on channel function, they expressed each of the engineered variants in the same way. The presence of the epitope enabled the researchers to label all surface-expressed KCNQ1 proteins by incubating the cells with an anti-Myc antibody conjugated to a phycoerythrin fluorescent tag. They could then label internally expressed KCNQ1 by permeabilizing the cells and incubating them with a second anti-Myc antibody conjugated to the AlexaFluor-647 fluorescent dye. Quantification of each label intensity by fluorescence-activated cell sorting provided a measure of the total amount of expressed protein and the amount of protein that reached the cell surface. The results indicated that many LQTS-associated and VUS mutations led to reduced surface expression and that reduced surface expression also strongly correlated with reduced total expression. The researchers quantified surface tracking efficiency by dividing the proportion of total protein that was surface expressed in the mutant by the proportion of total protein that was surfaced expressed in the wild-type and expressing the result as a percentage. This metric demonstrated that 18 of 23 mutations that exhibited a >35% reduction in total expression relative to that of the wild-type protein also exhibited defects in surface trafficking efficiency, although there were interesting exceptions for which trafficking efficiency was unexpectedly high. Co-expression of the proteins with KCNE1 had little effect on total expression or surface trafficking efficiency. Consistent with predictions, all 13 mutations expected to have no significant effect on channel function were expressed and trafficked in the same quantities as wild-type KCNQ1.


FIGURE 2. Structure of the KCNQ1 tetramer as seen from the extracellular (A) or intracellular (B) side of the membrane. Each subunit is denoted by a different color of its helices. For each subunit, helices S1-S4 form the voltage sensor domain located toward the outside of the complex, and the pore is formed by the interaction of helices S5 and S6 from each of the subunits in the center. Red, cyan, and gray side chains denote the position of LQTS, VUS, and neutral mutations, respectively in the voltage sensor domain. (C) Close-up view of the voltage-sensor domain and the locations of mutations investigated in the study. (D) Location of mutations as seen in an expanded, cutaway view of the interface between the voltage sensor domain and the pore domain.  Figure reproduced under the Creative Commons Attribution-NonCommercial 4.0 International License from H. Huang, et al. (2018) Sci. Adv., 4, eaar2631.


The investigators hypothesized that poor total and surface expression of some KCNQ1 mutants might be due to structural instability leading to improper folding. If so, these proteins would likely be processed by the endoplasmic reticulum-associated degradation (ERAD) pathway, leading ultimately to proteolysis by the proteasome. To test this hypothesis, the researchers expressed the mutant proteins in the presence of the proteasomal inhibitor MG132. They found that MG132 treatment resulted in much higher total levels of expression for many of the poorly expressed proteins; however, surface trafficking efficiency of these proteins was not improved. These results suggest that proteasomal degradation is the fate of many poorly expressed mutant KCNQ1 proteins, but even if they are saved from proteolysis, their structural flaws will most likely prevent them from being trafficked to the plasma membrane. 


Type 1 LQTS is often inherited as a dominant trait, suggesting that expression of a mutant allele of KCNQ1 is deleterious to the expression or function of the wild-type allele. To explore this possibility the investigators expressed various KCNQ1 variants together with the wild-type protein in cells. They discovered that some mutations suppressed the expression of the wild-type protein; however, in some cases, co-expression markedly increased levels of the mutant protein. These variable results are likely due to the formation of tetramers containing mixtures of the wild-type and mutant proteins. In some cases, the wild-type may stabilize the mutant, whereas in others, the presence of the mutant may have an overwhelming destabilizing effect.


To better understand the structural impact of the various mutations, the researchers performed two-dimensional 1H,15N-TROSY nuclear magnetic resonance studies on the voltage sensor domains of the wild-type and mutant proteins expressed in E. coli. This technique provides a distinct peak corresponding to each 1H,15N bond in the peptide linkages of the protein backbone. The wild-type KCNQ1 voltage sensor domain produced a well-resolved fingerprint peak pattern, typical of a stable, well-folded protein. Some of the mutants produced a very similar spectrum, differing only in the position of a few peaks associated with the region of the mutation, or in some cases a more distant region, suggesting a modest change in tertiary structure. For 13 mutants, however, the spectrum exhibited substantial peak broadening indicative of mutation-dependent structural destabilization or the presence of multiple conformations. All of these mutants also exhibited LOF and reduced surface and total expression. Four LOF mutants were so poorly expressed in E. coli that spectra could not be acquired. These results offered further evidence that structural instability plays an important role in LOF of many KCNQ1 mutants.  Of the 23 LOF mutants seen to mistraffick, 17 were clearly destabilized, as detected by NMR.


The researchers noted that five LQTS-associated mutations that appear to be unstable are in the S0 segment of KCNQ1's voltage sensor domain (Figure 3). These and other mutations in S0 interact with residues in the S1 or S2 helices that are also sites of mutations associated with poor expression and low trafficking efficiency. To better understand the role of the S0 segment, they performed molecular dynamics simulations of its interactions with the remainder of the voltage sensor domain (Figure 4). The results revealed a dense network of interactions between S0 and the remainder of the domain. Many of these interactions provide key insights into the impact of some LOF mutations (Figure 5). For example, Ala-178 and Cys-180 in helix S2 and Tyr-184 and Phe-193 in the S2-S3 linker form a hydrophobic pocket into which the side chain of Tyr-111 from the S0 segment inserts. Tyr-111 also forms polar interactions with Arg-174, Arg-181, and Lys-196. The Tyr-111-Cys mutation disrupts not only the hydrophobic interactions, but the polar ones as well, explaining why this mutation is unstable as detected by NMR.


FIGURE 3. Three views of the structure of the voltage sensor domain showing the positions of the 5 LQTS-associated mutations in the S0 helix (red) and the positions of 4 residues (magenta) that interact with them and are also subject to destabilizing mutations. Figure reproduced under the Creative Commons Attribution-NonCommercial 4.0 International License from H. Huang, et al. (2018) Sci. Adv., 4, eaar2631.



FIGURE 4. Model of the voltage sensor domain of wild-type KCNQ1 in a phospholipid bilayer following 500 ns of molecular dynamics simulations. The S1-S4 helices are shown in green, and the S0 segment is in cyan. Atoms of the phospholipid molecules are indicated as spheres. Figure reproduced under the Creative Commons Attribution-NonCommercial 4.0 International License from H. Huang, et al. (2018) Sci. Adv., 4, eaar2631.




FIGURE 5. Interactions involving sites in S0 and those interacting with S0  identified by molecular dynamics simulations. Sites of LQTS and VUS mutations are shown in red and blue, respectively. Residues for which mutations are benign or uncharacterized are shown in gray and green, respectively. Black dotted lines denote hydrogen bonds. Figure reproduced under the Creative Commons Attribution-NonCommercial 4.0 International License from H. Huang, et al. (2018) Sci. Adv., 4, eaar2631.



In summary, the results demonstrated that 23 of 32 LOF KCNQ1 mutants exhibit reduced cell surface expression with or without abnormal electrophysiological properties. Nine of 32 LOF mutants exhibited normal surface expression, but reduced conductance, whereas 4 of 32 exhibited normal surface expression but abnormal voltage-dependent activation and/or deactivation kinetics. Fifteen of the 51 proteins examined functioned with expression levels and electrophysiological parameters comparable to those of the wild-type protein. Eleven of these bore mutations that were originally predicted to be benign. The finding that 2 out of 13 predicted benign mutations were not, in fact, benign indicates that one cannot always assume that interspecies variability in protein sequence has no functional consequences.

The researchers concluded that the most common cause of inherited type 1 LQTS is mutation-induced protein destabilization, leading to ERAD-dependent proteasomal degradation. Thus, KCNQ1 joins PMP22 as the only two membrane proteins linked to human disease for which instability has now been documented as the major cause of dysfunction for more than a few mutations.  It is possible the mutation-induced destabilization of structure will prove to be a very common cause of LOF in membrane proteins for which amino acid mutations cause or predispose a patient to disease. In addition, of the 17 KCNQ1 mutations shown to result in significant misfolding, five are in the S0 segment, and 7 are in sites in S1 or S2 that interact with the S0 segment sites. This suggests that the S0 segment plays a critical, and until now under-appreciated, role in the stability of KCNQ1. The findings provide important new insights into personalized medical approaches that might be used to classify and treat patients suffering from type 1 LQTS. Finding ways to correct protein structural defects that lead to instability is a currently unmet challenge, but progress is being made through a new class of compounds, the "pharmacological chaperones" that is being explored as a therapeutic approach for other protein misfolding diseases.



View Science Advances article: Mechanisms of KCNQ1 channel dysfunction in long QT syndrome involving voltage sensor domain mutations





The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center

  • Areas of Interest