- Email: [email protected]
Functionally Aberrant Mutant KCNQ1 With Intermediate Heterozygous and Homozygous Phenotypes
Canadian Journal of Cardiology 34 (2018) 1174e1184
Basic Research
Functionally Aberrant Mutant KCNQ1 With Intermediate Heterozygous and Homozygous Phenotypes Zhenning Liu, MD, PhD,a Renjian Zheng, PhD,b Michael J. Grushko, MD,c Vladimir N. Uversky, PhD, DSc,d and Thomas V. McDonald, MDb,e a b
Department of Emergency Medicine, Shengjing Hospital of China Medical University, Shenyang, China
Departments of Medicine and Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, New York, USA c
d
Division of Cardiology, Jacobi Medical Center, Bronx, New York, USA
Department of Molecular Medicine, and USF Health Byrd Alzheimer’s Research Institute, Morsani College of Medicine, University of South Florida, Tampa, Florida, USA e
Department of Cardiovascular Sciences, Morsani College of Medicine, University of South Florida, Tampa, Florida, USA
ABSTRACT
RESUM E
Background: Deleterious mutations in KCNQ1 may lead to an autosomal dominant form of long QT syndrome (LQTS) (Romano-Ward) or autosomal recessive form (Jervell and Lange-Nielsen). Both are associated with severe ventricular tachyarrhythmias due to the reduction of the slowly activating delayed rectifier Kþ current (IKs). Our objective was to investigate the functional consequences of KCNQ1-R562S mutation in an atypical form of KCNQ1-linked LQTS. Methods: Mutant KCNQ1-R562S was analyzed via confocal imaging, surface biotinylation assays, co-immunoprecipitation, phosphatidylinositol-4,5-bisphosphate pulldown test, whole-cell patch clamp, and computational intrinsic disorder analyses. Results: Protein expression, assembly with KCNE1, and trafficking to the surface membrane of KCNQ1-R562S were comparable with wildtype channels. The most significant functional effect of the R562S
le tères du gène KCNQ1 peuvent aboutir à Contexte : Les mutations de une forme de syndrome du QT long autosomique dominante (Romanocessive (Jervell et Lange-Nielsen). Ces deux Ward) ou autosomique re es à des tachyarythmies ventriculaires se vères formes sont associe es par la re duction du courant potassique (Kþ) à rectification cause e et à activation lente (IKs). Notre objectif e tait d’analyser les retarde quences fonctionnelles de la mutation KCNQ1-R562S dans une conse au gène KCNQ1. forme atypique de syndrome du QT long associe thodes : Le gène mutant KCNQ1-R562S a e te analyse par les Me preuve de biotinylation de techniques suivantes : imagerie confocale, e cipitation, test de « pulldown » de phosphatisurface, co-immunopre lectrophysiologie mole culaire dylinositol-4,5-bisphosphate, par e « technique du patch clamp » sur des cellules entières et analyse computationnelle des troubles intrinsèques.
Long QT syndrome (LQTS) is a potentially life-threatening heart rhythm disorder generally due to malfunction of cardiac ion channels.1 Hereditary LQTS exhibits 2 distinct patterns of inheritance: the autosomal dominant RomanoWard syndrome (RWS) and the autosomal recessive Jervell and Lange-Nielsen syndrome (JLNS).2 The cardiac potassium ion channel that carries the slowly activating delayed rectifier Kþ current (IKs) comprises a and b subunits encoded by KCNQ1 and KCNE1 genes,
respectively, and plays a critical role in cardiac repolarization. Deleterious mutations of KCNQ1 lead to the LQTS type 1 variant (LQT1). The dysfunction of mutant IKs may lead to RWS if there are heterozygous loss-of-function mutations in either gene. When both alleles of KCNQ1 or KCNE1 harbor loss-of-function mutations (either homozygous or compound heterozygous) ventricular arrhythmias present at an earlier age, they are more severe, and there is associated congenital sensorineural hearing loss that characterize the JLNS. Patients with mutations that compromise IKs function frequently have arrhythmia triggered by elevated heart rate and adrenergic tone as occur during exercise. Here, we describe the clinical, biochemical, and biophysical characterization of a rare KNCQ1 mutation (c.1686 G>C, amino acids p.R562S) that resided in helix-C within the C-terminal region of the channel protein in a family with both heterozygous and homozygous individuals. Heterozygous members exhibited a mild RWS pattern. The member
Received for publication May 5, 2018. Accepted June 26, 2018. Corresponding author: Dr Thomas V. McDonald, Department of Cardiovascular Sciences, Morsani College of Medicine, University of South Florida, 12901 Bruce B. Downs Blvd, Tampa, Florida 33612, USA. Tel.: þ1813-974-5413. E-mail: [email protected] See page 1183 for disclosure information.
https://doi.org/10.1016/j.cjca.2018.06.015 0828-282X/Ó 2018 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
Liu et al. Functional KCNQ1 Mutation in LQTS
1175
mutation was a depolarizing shift in the voltage dependence of activation that was dependent on association with KCNE1. The biophysical abnormality was only partially dominant over coexpressed wild-type channels. R562S mutation impaired C-terminal association with membrane phosphatidylinositol-4,5-bisphosphate. These changes led to compromised rate-related accumulation of repolarizing current that is an important property of normal IKs. Conclusions: KCNQ1-R562S mutation reduces effective IKs due to channel gating alteration with a mild clinical expression in the heterozygous state due to minimal dominant phenotype. In the homozygous state, it is exhibited with a moderately severe LQTS phenotype due to the incomplete absence of IKs.
sultats : L’expression des prote ines, l’association avec KCNE1 et le Re taient trafic vers la membrane de surface des canaux KCNQ1-R562S e comparables à ceux des canaux de type sauvage. L’effet fonctionnel le tait un changement plus important de la mutation R562S e polarisant de la de pendance de l’activation à l’e gard de la tension, de tait tributaire de l’association avec KCNE1. L’anomalie biophysique n’e que partiellement dominante par rapport aux canaux de type sauvage s. La mutation R562S inhibait l’association de l’extre mite Cco-exprime terminale avec le phosphatidylinositol-4,5-bisphosphate membranaire. Ces changements menaient à une accumulation compromise, en ristique fonction du taux, du courant repolarisant qui est une caracte importante de l’IKs normal. duit l’IKs efficace en raison Conclusions : La mutation KCNQ1-R562S re ration des modalite s d’ouverture des canaux avec une d’une alte tat he te rozygote en raison du expression clinique peu importante à l’e notype dominant minimal. À l’e tat homozygote, un phe notype de phe re ment grave se manifestait en raison de syndrome du QT long mode l’absence presque complète d’IKs.
with homozygous KCNQ1-R562S exhibited severe arrhythmia, however, less so than classically described in JLNS. Moreover, the deafness in the homozygous member was late onset. Although KCNQ1-R562S had been identified from LQT cohorts,3 its functional characterization has not been reported. Our findings show that KCNQ1-R562S channels carried a potassium current with shifted voltage dependence of activation (VDA) compared with IKs. This depolarizing shift in VDA impaired the rate-related accumulation of Kþ conductance in an incompletely dominant fashion. By correlating the genotype-phenotype relationship of this KCNQ1 mutation, we provide mechanistic insight into the spectrum of clinical scenarios in LQT1.
KCNQ1 C-terminus was subcloned into the HindIII and BamHI sites of p3Flag-CMV-10 vector plasmid (SigmaAldrich, MO). The c.1686G>C mutation was introduced into KCNQ1 by site-directed mutagenesis with the following primers: KCNQ1_forward: 50 -CAG AGG AGC CTG GAC CAG TCC ATT-30 ; KCNQ1_reverse: 50 -AAT GGA CTG GTC CAG GCT CCT CTG-30 (IDT, Coralville, IA). Construction and validation of pCMV-3X-FLAG-KCNE1 plasmids have been previously described.6 Mutated cDNA vectors were verified by automated bidirectional DNA sequencing (Genewiz, NJ). Human Embryonic Kidney 293 cells (American Type Culture Collection, Manassas, VA) were transfected with Fugene 6 (Roche, Basel, Switzerland), and the DNA ratio of KCNQ1 and KCNE1 was 1:2 (by mass). For electrophysiology, green fluorescent protein (GFP) was used at a DNA KCNQ1:GFP ratio of 4:1. The biochemical experiments and electrophysiology recordings were performed 48 hours after transfection.
Materials and Methods Clinical and genetic analysis Clinical/research-based DNA sequencing for hereditary cardiac syndromes has been approved by the Albert Einstein College of Medicine Institutional Review Board. All clinical information was deidentified before the assembly of the manuscript in accordance with Institutional Review Board guidelines. The major LQTS gene sequence analysis for LQTS mutations was performed by FAMILION (PGx Health, a division of Clinical Data, Transgenomic, New Haven, CT). Genomic DNA was amplified by polymerase chain reaction to generate templates for direct sequencing of the targeted exons, splice junctions, and flanking regions of the genes KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2, KCNJ2, CACNA1C, CAV3, SCN4B, AKAP9, SNTA1, and KCNJ5. Sequencing results were compared with in a control population of more than 1300 unrelated, ethnically diverse, presumably healthy individuals.4 The R562S variant did not appear in more than 60,000 individuals comprising the ExAC database of exomes.5 Plasmids, cell culture, and transfection Human KCNQ1 was cloned into the HindIII and BamHI sites of pcDNA3.0 vector plasmid (GenScript, NJ). Human
Protein expression and electrophysiological function analysis SDS-PAGE and immunoblots were performed to analyze the protein expression of KCNQ1-WT, KCNQ1-R562S, and KCNE1 as descried previously.6 Co-immunoprecipitation, immunofluorescence, and surface biotinylation assays were performed as previously described.7 Phosphatidylinositol-4,5bisphosphate (PIP2) binding was assayed by pulldown with PIP2-agarose beads (Echelon Biosciences, UT) using the manufacturer’s protocol followed by immunoblot for KCNQ1 C-terminus. The whole-cell configuration of the patch clamp was used to evaluate the potassium channel function of KCNQ1-(WT, R562S, and 50/50 Mix) with/ without KCNE1. Detailed patch clamp methods are described in the Supplemental Methods. Analysis of intrinsic protein disorder Intrinsic disorder propensities of the KCNQ1 protein and its R526S mutant were evaluated using 3 algorithms from the PONDR family: PONDR VLXT, which is known to have high sensitivity to local sequence peculiarities and can be used
1176
for identifying disorder-based interaction sites;8 PONDR VL3, which is characterized by high accuracy for predicting long intrinsically disordered regions;9 and a meta-predictor PONDR FIT.10 After obtaining an average disorder score by each predictor, all predictor-specific average scores were averaged again to generate an average per-protein intrinsic disorder score. Use of consensus for the evaluation of intrinsic disorder is motivated by empirical observations that this approach usually increases the predictive performance compared with the use of a single predictor.11-13 Potential intrinsic disorderebased binding sites were evaluated by the ANCHOR algorithm,14,15 which uses the pairwise energy estimation and is based on the hypothesis that long regions of disorder contain localized potential binding sites that cannot form enough favourable intrachain interactions to fold on their own, but are likely to gain stabilizing energy by interacting with a globular protein partner. Statistical analysis Values presented were mean standard error of the mean. Analysis of variance statistical analysis was used to determine differences between groups, and post hoc analysis was done with the Student-Newman-Keuls test. P < 0.05 was considered significant.
Results Cases presentation A 64-year-old woman from the Caribbean with a history of hypertension, sensorineuronal hearing loss, seizures, and ventricular tachycardia was referred for cardiogenetic evaluation. She was diagnosed with bilateral hearing loss at age 3 years after being able to hear earlier. On examination she was determined to have postlingual bilateral hearing loss. She could hear only loud noises. She had multiple syncope episodes that were initially diagnosed as seizures, but she had not experienced cardiac arrest. Antiepileptic medications failed to prevent her syncope/seizure episodes. At age 58, she was noted to have ventricular tachycardia during a seizure episode and subsequent QT interval prolongation. She was given betaadrenergic blockers, and an internal cardiac defibrillator was implanted. There have been episodes of ventricular tachycardia storm with appropriate internal cardiac defibrillator shocks coincident with medical noncompliance. She has been symptom free while compliant with daily beta-adrenergic blocker therapy. A family history was taken, and pedigree constructed (Fig. 1A). The identity of the proband’s father was not known. There was no history of cardiovascular events, sudden cardiac death, or hearing impairment in other family members. Her resting electrocardiogram with a markedly prolonged QT and QTc interval is shown in Figure 1B. An echocardiogram demonstrated normal left ventricle size and systolic function. Two of her children and 2 grandchildren demonstrated normal resting QT and QTc intervals along with heart rates (Fig. 1, C-F).16 It was worth noting that the proband (1C) was on beta-adrenergic blocker therapy with Nadolol, but the other family members were not on medication.
Canadian Journal of Cardiology Volume 34 2018
Genetic analysis The proband was homozygous for the KCNQ1 mutation (c.1686G>C, p.R562S) and heterozygous for KCNE1 (c.112G>A, p.G38S). She was also heterozygous for AKAP9 (c.4841G>A, p.R1614Q) and KCNJ5 (c.844G>C, p.E282Q). Sequencing of KCNQ1 in her offspring showed that all her children and 1 grandchild had inherited KCNQ1R562S in a heterozygous state. The point mutation c.1686G>C is in exon 14, which encodes the C-terminal region of KCNQ1 (Fig. 2A). KCNQ1-R562S was reported as class II, variant of uncertain significance by the clinical laboratory. The other 3 heterozygous mutations including KCNE1-G38S, AKAP9-R1614Q, and KCNJ5-E282Q were reported as class III variants (polymorphisms; not expected to cause disease). The alignment of KCNQ1 C-terminal region protein sequences across different species demonstrated R562 evolutionary conservation (Fig. 2A). Protein expression and cellular localization of KCNQ1R562S mutation Immunoblot analysis of heterologously expressed KCNQ1 showed that the WT and mutant channel protein were of comparable abundance (Fig. 2, B-D). Immunofluorescence analysis indicated that KCNQ1-R562S reached the surface of the cell without intracellular accumulation comparable with KCNQ1-WT (Fig. 3A). The cell-surface marker cadherin colocalized with a Pearson correlation of 0.9376 0.0035 for KCNQ1-WT and 0.9279 0.0035 for KCNQ1-R562S. Quantitative surface labelling with NHS-SS-Biotin of KCNQ1-R562S was not significantly different than KCNQ1WT (Fig. 3, B and C). Biophysical effect of KCNQ1-R562S mutation on KCNQ1 function The maximal current density during a depolarizing step to 120 mV was not significantly different among the 3 groups (KCNQ1-WT homotetramers, KCNQ1-R562S homotetramers, and KCNQ1-WT/R562S 50:50 mix) (Supplemental Figure S1). The voltage that was required for half-maximum current activation (Vh) for KCNQ1-R562S was shifted in a depolarizing direction relative to KCNQ1-WT, whereas the KCNQ1-Mix showed an intermediate depolarizing Vh shift (Table 1). The cardiac IKs channel is a macromolecular complex composed of a pore-forming a subunit (KCNQ1) and modulatory b subunit (KCNE1). As shown in Figure 4, there was a significant difference in the current (at test potential from 0 mV to 45 mV)/maximal current ratio among the 3 groups. The Vh values for KCNQ1-R562S/KCNE1 and KCNQ1-Mix/KCNE1 were shifted in a depolarizing direction by 45 and 15 mV, respectively, when compared with KCNQ1-WT/KCNE1 (Table 1). As for the deactivation rates, there was a trend towards slower deactivation for WT KCNQ1/KCNE1 (1392.0 901.6 milliseconds at 40 mV, n ¼ 8) compared with KCNQ1-Mix/KCNE1 (1157.7 600.3 milliseconds at 40 mV, n ¼ 9) and KCNQ1-R562S/ KCNE1 (1158.8 375.7 milliseconds at 40 mV, n ¼ 13), but the differences were not significant.
Liu et al. Functional KCNQ1 Mutation in LQTS
1177
Figure 1. Clinical data of a family. (A) Family pedigree. Proband is indicated by the arrow. Shaded symbols indicate the presence of KCNQ1-R562S mutation (full shade: homozygote, half-shade: heterozygote). (B-G) Representative resting electrocardiogram complexes from proband (B), her 2 children (C, D), and 3 grandchildren (E-G).
Figure 2. Heterologous expression of wild-type and R562S-mutant KCNQ1 and KCNE1. (A) Reference and proband DNA sequence of KCNQ1 in the upper panel. Regional amino acid alignment of the KCNQ1 C-terminus across vertebrate species shows that R562 is highly conserved (lower panel). (B) Immunoblot detection of KCNQ1 and KCNE1 protein expressed in HEK-293 cells. (C) Densitometry quantification of the protein levels of KCNQ1 monomer and oligomers form immunoblot analysis. Expressions level not significantly different (n ¼ 5, P > 0.05). (D) Densitometry quantification of the protein levels of KCNE1. Expressions level not significantly different (n ¼ 5, P > 0.05).
1178
Canadian Journal of Cardiology Volume 34 2018
Figure 3. Comparable surface expression of KCNQ1 WT and R562S mutant channel proteins. (A) Confocal immunofluorescence micrographs of HEK293 cells transfected with either KCNQ1-WT or KCNQ1-R562S and counter-stained with anti-cadherin antibody to indicate the cell membrane. Scale bar ¼ 10 mm. Expressions level not significantly different (n ¼ 10, P > 0.05). (B) Representative immunoblot of KCNQ1 protein subunits for surfacebiotinylation study. Surface proteins were labeled with N-Hydroxysulfosuccinimide (NHS)-biotin in intact cells and precipitated by agarose-streptavidin. Calnexin was served as negative control for surface labeling and cadherin as positive control for surface labeling in the lower panel. (C) Densitometry analysis of the efficiency of surface presentation of KCNQ1 proteins expressed as the amount of surface KCNQ1 divided by the normalized for streptavidin pulldown of biotinylated protein. Expressions level not significantly different (n ¼ 4, P > 0.05). DIC, differential interference contrast.
We also examined if either KCNE1 (S/G)38 polymorphism affected the functional phenotype of the KCNQ1R562S mutation. Patch-clamp results (Supplemental Figure S2) revealed that the current density and the Vh for KCNQ1-R562S/KCNE1-38G did not significantly differ from the KCNQ1-R562S/KCNE1-38S or KCNQ1-R562S/ KCNE1-Mix (Table 1). Biochemical analysis of KCNQ1-R562S interactions with KCNE1 and PIP2 To investigate the interaction of KCNQ1 and KCNE1 C-terminal regions further, co-immunoprecipitation was
performed. Full-length KCNQ1-WT and KCNQ1-R562S channel proteins physically interacted with KCNE1 (38S or 38G) to a comparable degree. Furthermore, a purified C-terminal fragment of KCNQ1-WT and KCNQ1R562S also interacted with KCNE1 WT (38S or 38G) comparably, albeit to a lesser degree than the full-length channel (Fig. 5, A-D). To investigate the interaction of KCNQ1-CT and PIP2, a pulldown assay was performed using PIP2-coated agarose beads. The amount of KCNQ1-R562S C-terminal fragment exhibited significantly lower binding to PIP2 (by 69.92% 4.14%) compared with the KCNQ1-WT C-terminal fragment (Fig. 5, E and F).
Table 1. Biophysical characteristics of KCNQ1 mutants with/without KCNE1-WT Current density (pA/pF) KCNQ1-WT KCNQ1-Mix KCNQ1-R562S KCNQ1-WT/KCNE1-38S KCNQ1-Mix/KCNE1-38S KCNQ1-R562S/KCNE1-38S KCNQ1-R562S/KCNE1-Mix KCNQ1-R562S/KCNE1-38G
145.77 109.70 95.34 354.13 436.81 414.88 342.27 387.03
27.85 14.18 22.32 83.23 99.56 77.12 53.05 60.74
Activation time 1/2 max (ms) 9.74 8.22 8.25 492.08 416.29 539.81 719.93 480.00
1.131 0.34 0.80 55.93 58.05 67.62 92.25 62.38
Vh (mV) 21.40 24.33 13.64 10.46 27.62 55.17 54.52 57.26
4.11 0.96 1.32 1.55 1.27 1.29 1.14 0.75
Slope factor
Cell quantity
5 11 12 8 16 9 7 10
19.14 18.81 15.57 17.52 22.13 23.20 24.15 23.72
2.77 0.62 1.06 1.43 1.25 1.12 0.99 0.63
Liu et al. Functional KCNQ1 Mutation in LQTS
1179
Figure 4. Biophysical effect of the KCNQ1-R562S mutation on KCNQ1/KCNE1 channels. (A) Whole-cell current traces of KCNQ1 (WT, R562S, or Mix [50/50])/KCNE1-WT (38S) channels, in response to a series of depolarizing voltage steps (voltage clamp protocol shown above). (B) Itest/Imax ratio (current at 0-45 mV relative to maximal current). n ¼ 8-16, *P < 0.05. (C) Normalized voltage-dependent activation curves. Vh shown in Table 1. n ¼ 8-16, P < 0.05. (D) Rates of activation, measured as half-maximal rise time at 0-45 mV. The rate of activation of KCNQ1-WT/KCNE1 was significantly slower than channels composed of either KCNQ1-R562S/KCNE1 or KCNQ1-Mix (50/50)/KCNE1. n ¼ 8-16, *P < 0.05.
KCNQ1-R562S mutation decreases the rate-related accumulation of IKs IKs plays an important role in dynamic QT duration adaptation to heart rate.17 At high action potential frequencies, the slow activation and deactivation of IKs lead to Kþ conductance accumulation that maintains the QTc. To assess rate-related current accumulation, we simulated trains of action potentials at different rates (60/min, 120/min, and 150/min) and measured the rate of Kþ conductance accumulated at the end of 100 pulses. The KCNQ1-WT/KCNE1 channel exhibited marked accumulation of Kþ conductance at the frequencies between 120 and 150 depolarizations per minute (Fig. 6A). The rate of Kþ conductance accumulation was slower for the KCNQ1-R562S/KCNE1-38S and KCNQ1-Mix/KCNE1-38S channels (Fig. 6, B and C). At 150 pulses/min, the differences were most appreciated in channels with all KCNQ1-R562S subunits compared with a 50:50 mix of WT:mutant KCNQ1. We examined the steadystate current densities (pA/pF) of KCNQ1-WT, R562S, and Mix with KCNE1-38S channels at different pulse rates of 60/ min, 120/min, and 150/min. Interestingly, only at the pulse rate of 120/min was the current density of the KCNQ1-WT/ KCNE1-38S channel (212.25 48.26) significantly higher
than either KCNQ1-R562S/KCNE1-38S (80.31 12.44) or KCNQ1-Mix/KCNE1-38S channels (106.00 24.65) (P < 0.05). However, at the other pulse rates of 60/min and 150/ min, there were no significant differences in current density among these 3 groups. These results were consistent with the clinical manifestation of the heterozygous daughter of the proband (C in pedigree of Fig. 1) during the treadmill exercise stress test. The resting QTc (as calculated by the Hodges formula) value was near normal and rose progressively with exercise and continued to lengthen during exercise and into the recovery phase (peaking over 525 milliseconds at maximal heart rate, Fig. 6D). Moreover, a similar QTc response to exercise was observed for the heterozygous adult male (data not shown). Predicted mutational effects on intrinsic protein disorder The KCNQ C-terminus comprises proximal a-helices A and B that bind calmodulin, whereas the distal coiled-coil helix-C and helix-D tetramerize with an intervening unstructured region that is presumably flexible.18-20 Furthermore, the membrane phospholipid PIP2 interacts with residues in the KCNQ1 C-terminus providing additional regulation of channel gating kinetics.21 R562 resides near the
1180
Canadian Journal of Cardiology Volume 34 2018
Figure 5. Biochemical analysis of KCNQ1-R562S interactions with KCNE1 and PIP2. (A) Immunoblot analysis of whole cell lysis input for KCNQ1 (either while length indicated as KCNQ1-full, or C-terminal domain indicated as KCNQ1-CT) and KCNE1 in various cotransfection combinations. The hash (#) indicated nonspecific band. (B) KCNQ1 was immunoprecipitated followed by immunoblot detection of KCNQ1 protein (top gel) and KCNE1 (lower gel). (C) Control experiment with unrelated normal IgG used for precipitation showed no nonspecific KCNQ1 or KCNE1 detection. (D) Quantification of the KCNE1 protein pulldown to input. KCNE1 protein pulled down with the full length or C-terminus of KCNQ1-WT was not significantly different form KCNQ1-R562S. n ¼ 3, P > 0.05. (E) KCNQ1-R562S mutation in helix C showed impaired binding to PIP2. Representative immunoblot of KCNQ1-CT (WT or R562S mutant) by PIP2-coated agarose beads. (F) Quantification of the KCNQ1-CT pulldown to input shows that KCNQ1-R562S-CT binds to immobilized PIP2 with less avidity than KCNQ1-WT-CT. n ¼ 3, *P < 0.05. PIP2, phosphatidylinositol-4,5-bisphosphate.
juncture of the a-helix-C and the predicted intrinsically disordered region preceding the D-helix.19,20 To investigate the potential consequence of R562S substitution we subjected the amino acid sequence to in silico intrinsic disorder analysis. As shown in Figure 7, R562S is predicted to perturb both intrinsic disorder and local disorder-based binding potential the channel.
Discussion In this study, we identified and characterized a KCNQ1 mutation (R562S) found in the homozygous state in an older female patient with LQTS and atypical JLNS features. Her children with heterozygous inherited KCNQ1-R562S exhibited mild manifestation. The mutant KCNQ1-R562S channels expressed and trafficked normally in the
Liu et al. Functional KCNQ1 Mutation in LQTS
1181
Figure 6. KCNQ1-R562S mutation effect on rate-related accumulation of potassium conductance. A series of 100 depolarizing voltage clamp pulses mimicking action potential trains were delivered at 3 frequencies: 60 pulses/min, 120 pulses/min, and 150 pulses/min. The depolarizing steps were fixed at 325-millisecond duration (illustrated in panel A, inset). Both mutant and mixed mutant-WT IKs channels showed slower accumulation of current at 120 and 150 pulses/min than WT IKs channels. (A-C) Normalized current of KCNQ1-WT, R562S, Mix (50/50), and KCNE1-WT (38S). n ¼ 5, 11, 12. (D) QTc response of heterozygous patient (patient C in Fig. 1) during treadmill exercise stress test showing progressive prolongation to over 500 milliseconds during exercise and recovery of QTc prolongation during the 18-minute recovery period.
heterologous expression system and appropriately associated with coexpressed KCNE1. Currents produced by the mutant channels primarily showed a depolarizing shift in VDA. The VDA shift was most evident when KCNQ1 was coexpressed with KCNE1, as occurs in the heart. The predicted functional effect in the heart is that there will be significantly less IKs activated during normal cardiac myocyte action potentials. To mimic the heterozygous state, we examined currents from cells expressing a 50:50 mix of KCNQ1-WT and R562S subunits with KCNE1. Although there was a depolarizing shift in VDA for the WT/R562S mix compared with WT channels, it was not a strongly dominant current suppression. Patients with LQT1 are characterized by having arrhythmia episodes triggered by physical exertion and higher heart rates. Moreover, they have abnormal QTc prolongation during exercise. To evaluate the effect of KCNQ1-R562S on rate-related Kþ conductance accumulation in vitro we introduced a series of depolarizing steps with varying frequencies. Channels that were composed of mutant subunits or mixed WT/mutant showed slower accumulation of Kþ conductance than WT channels. This was clinically paralleled by the QTc duration response to exercise in the 2 children who were heterozygous for KCNQ1-R562S.
The C-terminus of KCNQ1 is important for channel gating, assembly, and trafficking.18 The direct interaction between the C-termini of KCNQ1 and KCNE1 affects activation and deactivation kinetics of IKs current.18,22 The amino acid sequence of the KCNQ1 region comprising helices C and D is highly conserved throughout evolution suggesting that mutations in the region are not functionally tolerated (Fig. 2). In the present study, we demonstrate that although KCNQ1R562S mutation did not impair the physical interaction with KCNE1-WT, its interaction with KCNE1 subunits amplified the biophysical mutant effects. The distal half of KCNQ1 helix-C encompassing a cluster of basic residues is important for channel modulation by PIP2, and these region mutations can impair interaction with PIP2.23 The potential mechanism of interaction with PIP2 is likely an electrostatic attraction between these side chains and the negatively charged polar heads of the phospholipid. Furthermore, we observe that KCNQ1-R562S mutation exhibited significantly lower binding to PIP2 than KCNQ1-WT, likely reflecting a lower affinity for PIP2. Because R562 resides between helices C and D of the C-terminus, it is reasonable to postulate that flexibility is important in that precise area. Our modeling of intrinsic disorder (Fig. 7), a characteristic of flexible regions,
1182
Canadian Journal of Cardiology Volume 34 2018
Figure 7. Analysis of the effect of R526S mutation on the intrinsic disorder propensity of KCNQ1 protein. (A, B) Evaluating intrinsic disorder propensity of the KCNQ1 protein (A) and its R526S mutant (B) by a series of per-residue disorder predictors. Profiles generated by PONDR VLXT, PONDR VL3, and PONDR FIT are shown by black, red, and pink lines, respectively. Dashed cyan lines show the mean disorder propensity calculated by averaging disorder profiles of individual predictors. Light pink shadow around the PONDR FIT curves shows error distribution. In these analyses, the predicted intrinsic disorder scores above 0.5 are considered to correspond to the disordered regions; scores between 0.2 and 0.5 are considered flexible. (C) Comparison of the mean disorder profiles of KCNQ1 protein and its R526S mutant (bottom panel) and “disorder difference spectrum” calculated as a simple difference between the mean disorder curves for mutant and wild-type protein (middle panel). A negative peak reflects a mutation-induced local decrease in the intrinsic disorder propensity. Top panel shows “ANCHOR difference spectrum” calculated as a simple difference between the ANCHOR curves calculated for the mutant and wild-type protein and showing how the R562S substitution may affect local disorder-based binding potential of the protein.
predicts that impairment of movement by the mutant residue may play a role in the functional phenotype. A previously described LQT1 mutation (c.1685G>T, p.R562M) at this site of KCNQ1 has also been identified,24 and functional characterization showed the reduced current amplitude largely due to a depolarizing shift in the voltage dependence of activation with perturbation of PIP2 regulation.23 We observed that KCNQ1-R562S behaved similar to R562M but with 2 notable exceptions. R562S did not appear to perturb interactions with KNCE1 whereas R562M did, and the Vh for R562S was 15 mV more depolarized than R562M.23 Both R562S and R562M result in reduced PIP2 binding. A common polymorphism site in KCNE1 (A/G at nucleotide 112 [G112A] alternately coding glycine or serine at position 38) was noted to be heterozygous in the proband. The relative allele frequency for glycine/serine is approximately 0.66/0.33 in the general population.25 Neither genome wide association studies nor family information supports an association between LQTS and G38S.26,27
Recent studies, however, suggest that homozygous KCNE138S might cause a mild reduction in IKs and might thereby increase an arrhythmogenic potential particularly in the presence of QT prolonging mutations.28,29 In our investigation, the functional electrophysiological characteristics of KCNE1-38S, KCNE1-38G, or KCNE1-Mix coexpressed on KCNQ1-R562S were not significantly different. Accordingly, we conclude that neither KCNE1 allele (38S or 38G) differentially impacted the phenotype of the KCNQ1-R562S mutation. JLNS has been linked to pathogenic gene variants (KCNQ1 and KCNE1) in a homozygous or compound heterozygous state with autosomal recessive inheritance.1,2,30 Individuals with JLNS are typically characterized by complete loss of hearing from birth, markedly prolonged QTc interval, and severe cardiac arrhythmia events at an early age. The clinical course of our proband and family in this present study suggested another genotype-phenotype profile that is characterized by postlingual deafness and later arrhythmia
Liu et al. Functional KCNQ1 Mutation in LQTS
presentation in the homozygous state and benign course for the heterozygous state. The proband did not experience syncope until young adulthood, and the heterozygous children have never had symptomatic arrhythmia despite marked QTc prolongation during exercise. More recent studies have highlighted cases of AR-LQT1 without deafness and conversely AD-LQT1 with deafness, thus making the mechanistic connection between arrhythmia and deafness less straightforward.30 One limitation to our analyses is that there are no clinical characteristics of deafness caused by mutant KCNQ1 that can distinguish from other causes. Furthermore, we do not have DNA sequencing of other deafness-associated genes in this family. Nevertheless, other genetic etiologies for deafness seem unlikely given that no other family member has reported hearing loss at any age. Conclusions LQTS with late-onset hearing-loss and arrhythmia features is presented in a patient homozygous for the KCNQ1-R562S mutation. This mutation resides in a critical area of the channel that produces IKs that is important for functional KCNE1 modulation and membrane lipid interaction. Channels carrying only mutant or a mix of WT and mutant subunits expressed currents with shifted voltage-dependent gating that reduced current at physiological potentials and decreased rate-related accumulation of Kþ conductance. Impaired PIP2 interactions and altered protein flexibility in mutant channel proteins likely underlie the mechanism of channel dysfunction. This study expands the clinical spectrum of KCNQ1 mutations and points to areas needing further research. Future investigation into precise genotype-phenotype relationships will likely shed more light onto this intriguing area. Funding Sources This work was supported by the National Institutes of Health (HL120782 to T.V.M.). Disclosures The authors have no conflicts of interest to disclose. References 1. Schwartz PJ, Stramba-Badiale M, Crotti L, et al. Prevalence of the congenital long-QT syndrome. Circulation 2009;120:1761-7. 2. Jervell A, Lange-Nielsen F. Congenital deaf-mutism, functional heart disease with prolongation of the Q-T interval and sudden death. Am Heart J 1957;54:59-68. 3. Andrsova I, Novotny T, Kadlecova J, et al. Clinical characteristics of 30 Czech families with long QT syndrome and KCNQ1 and KCNH2 gene mutations: importance of exercise testing. J Electrocardiol 2012;45: 746-51. 4. Kapa S, Tester DJ, Salisbury BA, et al. Genetic testing for long-QT syndrome: distinguishing pathogenic mutations from benign variants. Circulation 2009;120:1752-60. 5. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 2016;536:285-91.
1183
6. Melman YF, Um SY, Krumerman A, Kagan A, McDonald TV. KCNE1 binds to the KCNQ1 pore to regulate potassium channel activity. Neuron 2004;42:927-37. 7. Chen J, Weber M, Um SY, et al. A dual mechanism for I(Ks) current reduction by the pathogenic mutation KCNQ1-S277L. Pacing Clin Electrophysiol 2011;34:1652-64. 8. Romero P, Obradovic Z, Li X, et al. Sequence complexity of disordered protein. Proteins 2001;42:38-48. 9. Peng K, Radivojac P, Vucetic S, Dunker AK, Obradovic Z. Lengthdependent prediction of protein intrinsic disorder. BMC Bioinformatics 2006;7:208. 10. Xue B, Dunbrack RL, Williams RW, Dunker AK, Uversky VN. PONDR-FIT: a meta-predictor of intrinsically disordered amino acids. Biochim Biophys Acta 2010;1804:996-1010. 11. Walsh I, Giollo M, Di Domenico T, et al. Comprehensive large-scale assessment of intrinsic protein disorder. Bioinformatics 2015;31:201-8. 12. Fan X, Kurgan L. Accurate prediction of disorder in protein chains with a comprehensive and empirically designed consensus. J Biomol Struct Dyn 2014;32:448-64. 13. Peng Z, Kurgan L. On the complementarity of the consensus-based disorder prediction. Pac Symp Biocomput 2012:176-87. 14. Dosztanyi Z, Meszaros B, Simon I. ANCHOR: web server for predicting protein binding regions in disordered proteins. Bioinformatics 2009;25: 2745-6. 15. Meszaros B, Simon I, Dosztanyi Z. Prediction of protein binding regions in disordered proteins. PLoS Comput Biol 2009;5:e1000376. 16. Luo S, Michler K, Johnton P, Macfarlane PW. A comparison of commonly used QT correction formulae: the effect of heart rate on the QTc of normal ECGs. J Electrocardiol 2004;37(Suppl):81-90. 17. Suto F, Zhu W, Chan A, Gross GJ. IKr and IKs remodeling differentially affects QT interval prolongation and dynamic adaptation to heart rate acceleration in bradycardic rabbits. Am J Physiol Heart Circ Physiol 2007;292:H1782-8. 18. Wiener R, Haitin Y, Shamgar L, et al. The KCNQ1 (Kv7.1) COOH terminus, a multitiered scaffold for subunit assembly and protein interaction. J Biol Chem 2008;283:5815-30. 19. Howard RJ, Clark KA, Holton JM, Minor DJ. Structural insight into KCNQ (Kv7) channel assembly and channelopathy. Neuron 2007;53: 663-75. 20. Sachyani D, Dvir M, Strulovich R, et al. Structural basis of a Kv7.1 potassium channel gating module: studies of the intracellular c-terminal domain in complex with calmodulin. Structure 2014;22:1582-94. 21. Thomas AM, Harmer SC, Khambra T, Tinker A. Characterization of a binding site for anionic phospholipids on KCNQ1. J Biol Chem 2011;286:2088-100. 22. Zheng R, Thompson K, Obeng-Gyimah E, et al. Analysis of the interactions between the C-terminal cytoplasmic domains of KCNQ1 and KCNE1 channel subunits. Biochem J 2010;428:75-84. 23. Dvir M, Strulovich R, Sachyani D, et al. Long QT mutations at the interface between KCNQ1 helix C and KCNE1 disrupt I(KS) regulation by PKA and PIP(2). J Cell Sci 2014;127:3943-55. 24. Van Langen IM, Birnie E, Alders M, et al. The use of genotypephenotype correlations in mutation analysis for the long QT syndrome. J Med Genet 2003;40:141-5.
1184
25. Gouas L, Nicaud V, Berthet M, et al. Association of KCNQ1, KCNE1, KCNH2 and SCN5A polymorphisms with QTc interval length in a healthy population. Eur J Hum Genet 2005;13:1213-22. 26. Newton-Cheh C, Eijgelsheim M, Rice KM, et al. Common variants at ten loci influence QT interval duration in the QTGEN Study. Nat Genet 2009;41:399-406. 27. Kanters JK, Olesen MS, Christiansen M. KCNE1 G38S polymorphism is not the cause of long QT syndrome. J Electrocardiol 2016;49:249-50. 28. Yamaguchi Y, Nishide K, Kato M, et al. Glycine/Serine polymorphism at position 38 influences KCNE1 subunit’s modulatory actions on rapid and slow delayed rectifier Kþ currents. Circ J 2014;78:610-8.
Canadian Journal of Cardiology Volume 34 2018 29. Yamaguchi Y, Mizumaki K, Hata Y, et al. Latent pathogenicity of the G38S polymorphism of KCNE1 Kþ channel modulator. Heart Vessels 2017;32:186-92. 30. Giudicessi JR, Ackerman MJ. Prevalence and potential genetic determinants of sensorineural deafness in KCNQ1 homozygosity and compound heterozygosity. Circ Cardiovasc Genet 2013;6:193-200.
Supplementary Material To access the supplementary material accompanying this article, visit the online version of the Canadian Journal of Cardiology at www.onlinecjc.ca and at https://doi.org/10. 1016/j.cjca.2018.06.015.
Basic Research
Functionally Aberrant Mutant KCNQ1 With Intermediate Heterozygous and Homozygous Phenotypes Zhenning Liu, MD, PhD,a Renjian Zheng, PhD,b Michael J. Grushko, MD,c Vladimir N. Uversky, PhD, DSc,d and Thomas V. McDonald, MDb,e a b
Department of Emergency Medicine, Shengjing Hospital of China Medical University, Shenyang, China
Departments of Medicine and Molecular Pharmacology, Albert Einstein College of Medicine, Bronx, New York, USA c
d
Division of Cardiology, Jacobi Medical Center, Bronx, New York, USA
Department of Molecular Medicine, and USF Health Byrd Alzheimer’s Research Institute, Morsani College of Medicine, University of South Florida, Tampa, Florida, USA e
Department of Cardiovascular Sciences, Morsani College of Medicine, University of South Florida, Tampa, Florida, USA
ABSTRACT
RESUM E
Background: Deleterious mutations in KCNQ1 may lead to an autosomal dominant form of long QT syndrome (LQTS) (Romano-Ward) or autosomal recessive form (Jervell and Lange-Nielsen). Both are associated with severe ventricular tachyarrhythmias due to the reduction of the slowly activating delayed rectifier Kþ current (IKs). Our objective was to investigate the functional consequences of KCNQ1-R562S mutation in an atypical form of KCNQ1-linked LQTS. Methods: Mutant KCNQ1-R562S was analyzed via confocal imaging, surface biotinylation assays, co-immunoprecipitation, phosphatidylinositol-4,5-bisphosphate pulldown test, whole-cell patch clamp, and computational intrinsic disorder analyses. Results: Protein expression, assembly with KCNE1, and trafficking to the surface membrane of KCNQ1-R562S were comparable with wildtype channels. The most significant functional effect of the R562S
le tères du gène KCNQ1 peuvent aboutir à Contexte : Les mutations de une forme de syndrome du QT long autosomique dominante (Romanocessive (Jervell et Lange-Nielsen). Ces deux Ward) ou autosomique re es à des tachyarythmies ventriculaires se vères formes sont associe es par la re duction du courant potassique (Kþ) à rectification cause e et à activation lente (IKs). Notre objectif e tait d’analyser les retarde quences fonctionnelles de la mutation KCNQ1-R562S dans une conse au gène KCNQ1. forme atypique de syndrome du QT long associe thodes : Le gène mutant KCNQ1-R562S a e te analyse par les Me preuve de biotinylation de techniques suivantes : imagerie confocale, e cipitation, test de « pulldown » de phosphatisurface, co-immunopre lectrophysiologie mole culaire dylinositol-4,5-bisphosphate, par e « technique du patch clamp » sur des cellules entières et analyse computationnelle des troubles intrinsèques.
Long QT syndrome (LQTS) is a potentially life-threatening heart rhythm disorder generally due to malfunction of cardiac ion channels.1 Hereditary LQTS exhibits 2 distinct patterns of inheritance: the autosomal dominant RomanoWard syndrome (RWS) and the autosomal recessive Jervell and Lange-Nielsen syndrome (JLNS).2 The cardiac potassium ion channel that carries the slowly activating delayed rectifier Kþ current (IKs) comprises a and b subunits encoded by KCNQ1 and KCNE1 genes,
respectively, and plays a critical role in cardiac repolarization. Deleterious mutations of KCNQ1 lead to the LQTS type 1 variant (LQT1). The dysfunction of mutant IKs may lead to RWS if there are heterozygous loss-of-function mutations in either gene. When both alleles of KCNQ1 or KCNE1 harbor loss-of-function mutations (either homozygous or compound heterozygous) ventricular arrhythmias present at an earlier age, they are more severe, and there is associated congenital sensorineural hearing loss that characterize the JLNS. Patients with mutations that compromise IKs function frequently have arrhythmia triggered by elevated heart rate and adrenergic tone as occur during exercise. Here, we describe the clinical, biochemical, and biophysical characterization of a rare KNCQ1 mutation (c.1686 G>C, amino acids p.R562S) that resided in helix-C within the C-terminal region of the channel protein in a family with both heterozygous and homozygous individuals. Heterozygous members exhibited a mild RWS pattern. The member
Received for publication May 5, 2018. Accepted June 26, 2018. Corresponding author: Dr Thomas V. McDonald, Department of Cardiovascular Sciences, Morsani College of Medicine, University of South Florida, 12901 Bruce B. Downs Blvd, Tampa, Florida 33612, USA. Tel.: þ1813-974-5413. E-mail: [email protected] See page 1183 for disclosure information.
https://doi.org/10.1016/j.cjca.2018.06.015 0828-282X/Ó 2018 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
Liu et al. Functional KCNQ1 Mutation in LQTS
1175
mutation was a depolarizing shift in the voltage dependence of activation that was dependent on association with KCNE1. The biophysical abnormality was only partially dominant over coexpressed wild-type channels. R562S mutation impaired C-terminal association with membrane phosphatidylinositol-4,5-bisphosphate. These changes led to compromised rate-related accumulation of repolarizing current that is an important property of normal IKs. Conclusions: KCNQ1-R562S mutation reduces effective IKs due to channel gating alteration with a mild clinical expression in the heterozygous state due to minimal dominant phenotype. In the homozygous state, it is exhibited with a moderately severe LQTS phenotype due to the incomplete absence of IKs.
sultats : L’expression des prote ines, l’association avec KCNE1 et le Re taient trafic vers la membrane de surface des canaux KCNQ1-R562S e comparables à ceux des canaux de type sauvage. L’effet fonctionnel le tait un changement plus important de la mutation R562S e polarisant de la de pendance de l’activation à l’e gard de la tension, de tait tributaire de l’association avec KCNE1. L’anomalie biophysique n’e que partiellement dominante par rapport aux canaux de type sauvage s. La mutation R562S inhibait l’association de l’extre mite Cco-exprime terminale avec le phosphatidylinositol-4,5-bisphosphate membranaire. Ces changements menaient à une accumulation compromise, en ristique fonction du taux, du courant repolarisant qui est une caracte importante de l’IKs normal. duit l’IKs efficace en raison Conclusions : La mutation KCNQ1-R562S re ration des modalite s d’ouverture des canaux avec une d’une alte tat he te rozygote en raison du expression clinique peu importante à l’e notype dominant minimal. À l’e tat homozygote, un phe notype de phe re ment grave se manifestait en raison de syndrome du QT long mode l’absence presque complète d’IKs.
with homozygous KCNQ1-R562S exhibited severe arrhythmia, however, less so than classically described in JLNS. Moreover, the deafness in the homozygous member was late onset. Although KCNQ1-R562S had been identified from LQT cohorts,3 its functional characterization has not been reported. Our findings show that KCNQ1-R562S channels carried a potassium current with shifted voltage dependence of activation (VDA) compared with IKs. This depolarizing shift in VDA impaired the rate-related accumulation of Kþ conductance in an incompletely dominant fashion. By correlating the genotype-phenotype relationship of this KCNQ1 mutation, we provide mechanistic insight into the spectrum of clinical scenarios in LQT1.
KCNQ1 C-terminus was subcloned into the HindIII and BamHI sites of p3Flag-CMV-10 vector plasmid (SigmaAldrich, MO). The c.1686G>C mutation was introduced into KCNQ1 by site-directed mutagenesis with the following primers: KCNQ1_forward: 50 -CAG AGG AGC CTG GAC CAG TCC ATT-30 ; KCNQ1_reverse: 50 -AAT GGA CTG GTC CAG GCT CCT CTG-30 (IDT, Coralville, IA). Construction and validation of pCMV-3X-FLAG-KCNE1 plasmids have been previously described.6 Mutated cDNA vectors were verified by automated bidirectional DNA sequencing (Genewiz, NJ). Human Embryonic Kidney 293 cells (American Type Culture Collection, Manassas, VA) were transfected with Fugene 6 (Roche, Basel, Switzerland), and the DNA ratio of KCNQ1 and KCNE1 was 1:2 (by mass). For electrophysiology, green fluorescent protein (GFP) was used at a DNA KCNQ1:GFP ratio of 4:1. The biochemical experiments and electrophysiology recordings were performed 48 hours after transfection.
Materials and Methods Clinical and genetic analysis Clinical/research-based DNA sequencing for hereditary cardiac syndromes has been approved by the Albert Einstein College of Medicine Institutional Review Board. All clinical information was deidentified before the assembly of the manuscript in accordance with Institutional Review Board guidelines. The major LQTS gene sequence analysis for LQTS mutations was performed by FAMILION (PGx Health, a division of Clinical Data, Transgenomic, New Haven, CT). Genomic DNA was amplified by polymerase chain reaction to generate templates for direct sequencing of the targeted exons, splice junctions, and flanking regions of the genes KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2, KCNJ2, CACNA1C, CAV3, SCN4B, AKAP9, SNTA1, and KCNJ5. Sequencing results were compared with in a control population of more than 1300 unrelated, ethnically diverse, presumably healthy individuals.4 The R562S variant did not appear in more than 60,000 individuals comprising the ExAC database of exomes.5 Plasmids, cell culture, and transfection Human KCNQ1 was cloned into the HindIII and BamHI sites of pcDNA3.0 vector plasmid (GenScript, NJ). Human
Protein expression and electrophysiological function analysis SDS-PAGE and immunoblots were performed to analyze the protein expression of KCNQ1-WT, KCNQ1-R562S, and KCNE1 as descried previously.6 Co-immunoprecipitation, immunofluorescence, and surface biotinylation assays were performed as previously described.7 Phosphatidylinositol-4,5bisphosphate (PIP2) binding was assayed by pulldown with PIP2-agarose beads (Echelon Biosciences, UT) using the manufacturer’s protocol followed by immunoblot for KCNQ1 C-terminus. The whole-cell configuration of the patch clamp was used to evaluate the potassium channel function of KCNQ1-(WT, R562S, and 50/50 Mix) with/ without KCNE1. Detailed patch clamp methods are described in the Supplemental Methods. Analysis of intrinsic protein disorder Intrinsic disorder propensities of the KCNQ1 protein and its R526S mutant were evaluated using 3 algorithms from the PONDR family: PONDR VLXT, which is known to have high sensitivity to local sequence peculiarities and can be used
1176
for identifying disorder-based interaction sites;8 PONDR VL3, which is characterized by high accuracy for predicting long intrinsically disordered regions;9 and a meta-predictor PONDR FIT.10 After obtaining an average disorder score by each predictor, all predictor-specific average scores were averaged again to generate an average per-protein intrinsic disorder score. Use of consensus for the evaluation of intrinsic disorder is motivated by empirical observations that this approach usually increases the predictive performance compared with the use of a single predictor.11-13 Potential intrinsic disorderebased binding sites were evaluated by the ANCHOR algorithm,14,15 which uses the pairwise energy estimation and is based on the hypothesis that long regions of disorder contain localized potential binding sites that cannot form enough favourable intrachain interactions to fold on their own, but are likely to gain stabilizing energy by interacting with a globular protein partner. Statistical analysis Values presented were mean standard error of the mean. Analysis of variance statistical analysis was used to determine differences between groups, and post hoc analysis was done with the Student-Newman-Keuls test. P < 0.05 was considered significant.
Results Cases presentation A 64-year-old woman from the Caribbean with a history of hypertension, sensorineuronal hearing loss, seizures, and ventricular tachycardia was referred for cardiogenetic evaluation. She was diagnosed with bilateral hearing loss at age 3 years after being able to hear earlier. On examination she was determined to have postlingual bilateral hearing loss. She could hear only loud noises. She had multiple syncope episodes that were initially diagnosed as seizures, but she had not experienced cardiac arrest. Antiepileptic medications failed to prevent her syncope/seizure episodes. At age 58, she was noted to have ventricular tachycardia during a seizure episode and subsequent QT interval prolongation. She was given betaadrenergic blockers, and an internal cardiac defibrillator was implanted. There have been episodes of ventricular tachycardia storm with appropriate internal cardiac defibrillator shocks coincident with medical noncompliance. She has been symptom free while compliant with daily beta-adrenergic blocker therapy. A family history was taken, and pedigree constructed (Fig. 1A). The identity of the proband’s father was not known. There was no history of cardiovascular events, sudden cardiac death, or hearing impairment in other family members. Her resting electrocardiogram with a markedly prolonged QT and QTc interval is shown in Figure 1B. An echocardiogram demonstrated normal left ventricle size and systolic function. Two of her children and 2 grandchildren demonstrated normal resting QT and QTc intervals along with heart rates (Fig. 1, C-F).16 It was worth noting that the proband (1C) was on beta-adrenergic blocker therapy with Nadolol, but the other family members were not on medication.
Canadian Journal of Cardiology Volume 34 2018
Genetic analysis The proband was homozygous for the KCNQ1 mutation (c.1686G>C, p.R562S) and heterozygous for KCNE1 (c.112G>A, p.G38S). She was also heterozygous for AKAP9 (c.4841G>A, p.R1614Q) and KCNJ5 (c.844G>C, p.E282Q). Sequencing of KCNQ1 in her offspring showed that all her children and 1 grandchild had inherited KCNQ1R562S in a heterozygous state. The point mutation c.1686G>C is in exon 14, which encodes the C-terminal region of KCNQ1 (Fig. 2A). KCNQ1-R562S was reported as class II, variant of uncertain significance by the clinical laboratory. The other 3 heterozygous mutations including KCNE1-G38S, AKAP9-R1614Q, and KCNJ5-E282Q were reported as class III variants (polymorphisms; not expected to cause disease). The alignment of KCNQ1 C-terminal region protein sequences across different species demonstrated R562 evolutionary conservation (Fig. 2A). Protein expression and cellular localization of KCNQ1R562S mutation Immunoblot analysis of heterologously expressed KCNQ1 showed that the WT and mutant channel protein were of comparable abundance (Fig. 2, B-D). Immunofluorescence analysis indicated that KCNQ1-R562S reached the surface of the cell without intracellular accumulation comparable with KCNQ1-WT (Fig. 3A). The cell-surface marker cadherin colocalized with a Pearson correlation of 0.9376 0.0035 for KCNQ1-WT and 0.9279 0.0035 for KCNQ1-R562S. Quantitative surface labelling with NHS-SS-Biotin of KCNQ1-R562S was not significantly different than KCNQ1WT (Fig. 3, B and C). Biophysical effect of KCNQ1-R562S mutation on KCNQ1 function The maximal current density during a depolarizing step to 120 mV was not significantly different among the 3 groups (KCNQ1-WT homotetramers, KCNQ1-R562S homotetramers, and KCNQ1-WT/R562S 50:50 mix) (Supplemental Figure S1). The voltage that was required for half-maximum current activation (Vh) for KCNQ1-R562S was shifted in a depolarizing direction relative to KCNQ1-WT, whereas the KCNQ1-Mix showed an intermediate depolarizing Vh shift (Table 1). The cardiac IKs channel is a macromolecular complex composed of a pore-forming a subunit (KCNQ1) and modulatory b subunit (KCNE1). As shown in Figure 4, there was a significant difference in the current (at test potential from 0 mV to 45 mV)/maximal current ratio among the 3 groups. The Vh values for KCNQ1-R562S/KCNE1 and KCNQ1-Mix/KCNE1 were shifted in a depolarizing direction by 45 and 15 mV, respectively, when compared with KCNQ1-WT/KCNE1 (Table 1). As for the deactivation rates, there was a trend towards slower deactivation for WT KCNQ1/KCNE1 (1392.0 901.6 milliseconds at 40 mV, n ¼ 8) compared with KCNQ1-Mix/KCNE1 (1157.7 600.3 milliseconds at 40 mV, n ¼ 9) and KCNQ1-R562S/ KCNE1 (1158.8 375.7 milliseconds at 40 mV, n ¼ 13), but the differences were not significant.
Liu et al. Functional KCNQ1 Mutation in LQTS
1177
Figure 1. Clinical data of a family. (A) Family pedigree. Proband is indicated by the arrow. Shaded symbols indicate the presence of KCNQ1-R562S mutation (full shade: homozygote, half-shade: heterozygote). (B-G) Representative resting electrocardiogram complexes from proband (B), her 2 children (C, D), and 3 grandchildren (E-G).
Figure 2. Heterologous expression of wild-type and R562S-mutant KCNQ1 and KCNE1. (A) Reference and proband DNA sequence of KCNQ1 in the upper panel. Regional amino acid alignment of the KCNQ1 C-terminus across vertebrate species shows that R562 is highly conserved (lower panel). (B) Immunoblot detection of KCNQ1 and KCNE1 protein expressed in HEK-293 cells. (C) Densitometry quantification of the protein levels of KCNQ1 monomer and oligomers form immunoblot analysis. Expressions level not significantly different (n ¼ 5, P > 0.05). (D) Densitometry quantification of the protein levels of KCNE1. Expressions level not significantly different (n ¼ 5, P > 0.05).
1178
Canadian Journal of Cardiology Volume 34 2018
Figure 3. Comparable surface expression of KCNQ1 WT and R562S mutant channel proteins. (A) Confocal immunofluorescence micrographs of HEK293 cells transfected with either KCNQ1-WT or KCNQ1-R562S and counter-stained with anti-cadherin antibody to indicate the cell membrane. Scale bar ¼ 10 mm. Expressions level not significantly different (n ¼ 10, P > 0.05). (B) Representative immunoblot of KCNQ1 protein subunits for surfacebiotinylation study. Surface proteins were labeled with N-Hydroxysulfosuccinimide (NHS)-biotin in intact cells and precipitated by agarose-streptavidin. Calnexin was served as negative control for surface labeling and cadherin as positive control for surface labeling in the lower panel. (C) Densitometry analysis of the efficiency of surface presentation of KCNQ1 proteins expressed as the amount of surface KCNQ1 divided by the normalized for streptavidin pulldown of biotinylated protein. Expressions level not significantly different (n ¼ 4, P > 0.05). DIC, differential interference contrast.
We also examined if either KCNE1 (S/G)38 polymorphism affected the functional phenotype of the KCNQ1R562S mutation. Patch-clamp results (Supplemental Figure S2) revealed that the current density and the Vh for KCNQ1-R562S/KCNE1-38G did not significantly differ from the KCNQ1-R562S/KCNE1-38S or KCNQ1-R562S/ KCNE1-Mix (Table 1). Biochemical analysis of KCNQ1-R562S interactions with KCNE1 and PIP2 To investigate the interaction of KCNQ1 and KCNE1 C-terminal regions further, co-immunoprecipitation was
performed. Full-length KCNQ1-WT and KCNQ1-R562S channel proteins physically interacted with KCNE1 (38S or 38G) to a comparable degree. Furthermore, a purified C-terminal fragment of KCNQ1-WT and KCNQ1R562S also interacted with KCNE1 WT (38S or 38G) comparably, albeit to a lesser degree than the full-length channel (Fig. 5, A-D). To investigate the interaction of KCNQ1-CT and PIP2, a pulldown assay was performed using PIP2-coated agarose beads. The amount of KCNQ1-R562S C-terminal fragment exhibited significantly lower binding to PIP2 (by 69.92% 4.14%) compared with the KCNQ1-WT C-terminal fragment (Fig. 5, E and F).
Table 1. Biophysical characteristics of KCNQ1 mutants with/without KCNE1-WT Current density (pA/pF) KCNQ1-WT KCNQ1-Mix KCNQ1-R562S KCNQ1-WT/KCNE1-38S KCNQ1-Mix/KCNE1-38S KCNQ1-R562S/KCNE1-38S KCNQ1-R562S/KCNE1-Mix KCNQ1-R562S/KCNE1-38G
145.77 109.70 95.34 354.13 436.81 414.88 342.27 387.03
27.85 14.18 22.32 83.23 99.56 77.12 53.05 60.74
Activation time 1/2 max (ms) 9.74 8.22 8.25 492.08 416.29 539.81 719.93 480.00
1.131 0.34 0.80 55.93 58.05 67.62 92.25 62.38
Vh (mV) 21.40 24.33 13.64 10.46 27.62 55.17 54.52 57.26
4.11 0.96 1.32 1.55 1.27 1.29 1.14 0.75
Slope factor
Cell quantity
5 11 12 8 16 9 7 10
19.14 18.81 15.57 17.52 22.13 23.20 24.15 23.72
2.77 0.62 1.06 1.43 1.25 1.12 0.99 0.63
Liu et al. Functional KCNQ1 Mutation in LQTS
1179
Figure 4. Biophysical effect of the KCNQ1-R562S mutation on KCNQ1/KCNE1 channels. (A) Whole-cell current traces of KCNQ1 (WT, R562S, or Mix [50/50])/KCNE1-WT (38S) channels, in response to a series of depolarizing voltage steps (voltage clamp protocol shown above). (B) Itest/Imax ratio (current at 0-45 mV relative to maximal current). n ¼ 8-16, *P < 0.05. (C) Normalized voltage-dependent activation curves. Vh shown in Table 1. n ¼ 8-16, P < 0.05. (D) Rates of activation, measured as half-maximal rise time at 0-45 mV. The rate of activation of KCNQ1-WT/KCNE1 was significantly slower than channels composed of either KCNQ1-R562S/KCNE1 or KCNQ1-Mix (50/50)/KCNE1. n ¼ 8-16, *P < 0.05.
KCNQ1-R562S mutation decreases the rate-related accumulation of IKs IKs plays an important role in dynamic QT duration adaptation to heart rate.17 At high action potential frequencies, the slow activation and deactivation of IKs lead to Kþ conductance accumulation that maintains the QTc. To assess rate-related current accumulation, we simulated trains of action potentials at different rates (60/min, 120/min, and 150/min) and measured the rate of Kþ conductance accumulated at the end of 100 pulses. The KCNQ1-WT/KCNE1 channel exhibited marked accumulation of Kþ conductance at the frequencies between 120 and 150 depolarizations per minute (Fig. 6A). The rate of Kþ conductance accumulation was slower for the KCNQ1-R562S/KCNE1-38S and KCNQ1-Mix/KCNE1-38S channels (Fig. 6, B and C). At 150 pulses/min, the differences were most appreciated in channels with all KCNQ1-R562S subunits compared with a 50:50 mix of WT:mutant KCNQ1. We examined the steadystate current densities (pA/pF) of KCNQ1-WT, R562S, and Mix with KCNE1-38S channels at different pulse rates of 60/ min, 120/min, and 150/min. Interestingly, only at the pulse rate of 120/min was the current density of the KCNQ1-WT/ KCNE1-38S channel (212.25 48.26) significantly higher
than either KCNQ1-R562S/KCNE1-38S (80.31 12.44) or KCNQ1-Mix/KCNE1-38S channels (106.00 24.65) (P < 0.05). However, at the other pulse rates of 60/min and 150/ min, there were no significant differences in current density among these 3 groups. These results were consistent with the clinical manifestation of the heterozygous daughter of the proband (C in pedigree of Fig. 1) during the treadmill exercise stress test. The resting QTc (as calculated by the Hodges formula) value was near normal and rose progressively with exercise and continued to lengthen during exercise and into the recovery phase (peaking over 525 milliseconds at maximal heart rate, Fig. 6D). Moreover, a similar QTc response to exercise was observed for the heterozygous adult male (data not shown). Predicted mutational effects on intrinsic protein disorder The KCNQ C-terminus comprises proximal a-helices A and B that bind calmodulin, whereas the distal coiled-coil helix-C and helix-D tetramerize with an intervening unstructured region that is presumably flexible.18-20 Furthermore, the membrane phospholipid PIP2 interacts with residues in the KCNQ1 C-terminus providing additional regulation of channel gating kinetics.21 R562 resides near the
1180
Canadian Journal of Cardiology Volume 34 2018
Figure 5. Biochemical analysis of KCNQ1-R562S interactions with KCNE1 and PIP2. (A) Immunoblot analysis of whole cell lysis input for KCNQ1 (either while length indicated as KCNQ1-full, or C-terminal domain indicated as KCNQ1-CT) and KCNE1 in various cotransfection combinations. The hash (#) indicated nonspecific band. (B) KCNQ1 was immunoprecipitated followed by immunoblot detection of KCNQ1 protein (top gel) and KCNE1 (lower gel). (C) Control experiment with unrelated normal IgG used for precipitation showed no nonspecific KCNQ1 or KCNE1 detection. (D) Quantification of the KCNE1 protein pulldown to input. KCNE1 protein pulled down with the full length or C-terminus of KCNQ1-WT was not significantly different form KCNQ1-R562S. n ¼ 3, P > 0.05. (E) KCNQ1-R562S mutation in helix C showed impaired binding to PIP2. Representative immunoblot of KCNQ1-CT (WT or R562S mutant) by PIP2-coated agarose beads. (F) Quantification of the KCNQ1-CT pulldown to input shows that KCNQ1-R562S-CT binds to immobilized PIP2 with less avidity than KCNQ1-WT-CT. n ¼ 3, *P < 0.05. PIP2, phosphatidylinositol-4,5-bisphosphate.
juncture of the a-helix-C and the predicted intrinsically disordered region preceding the D-helix.19,20 To investigate the potential consequence of R562S substitution we subjected the amino acid sequence to in silico intrinsic disorder analysis. As shown in Figure 7, R562S is predicted to perturb both intrinsic disorder and local disorder-based binding potential the channel.
Discussion In this study, we identified and characterized a KCNQ1 mutation (R562S) found in the homozygous state in an older female patient with LQTS and atypical JLNS features. Her children with heterozygous inherited KCNQ1-R562S exhibited mild manifestation. The mutant KCNQ1-R562S channels expressed and trafficked normally in the
Liu et al. Functional KCNQ1 Mutation in LQTS
1181
Figure 6. KCNQ1-R562S mutation effect on rate-related accumulation of potassium conductance. A series of 100 depolarizing voltage clamp pulses mimicking action potential trains were delivered at 3 frequencies: 60 pulses/min, 120 pulses/min, and 150 pulses/min. The depolarizing steps were fixed at 325-millisecond duration (illustrated in panel A, inset). Both mutant and mixed mutant-WT IKs channels showed slower accumulation of current at 120 and 150 pulses/min than WT IKs channels. (A-C) Normalized current of KCNQ1-WT, R562S, Mix (50/50), and KCNE1-WT (38S). n ¼ 5, 11, 12. (D) QTc response of heterozygous patient (patient C in Fig. 1) during treadmill exercise stress test showing progressive prolongation to over 500 milliseconds during exercise and recovery of QTc prolongation during the 18-minute recovery period.
heterologous expression system and appropriately associated with coexpressed KCNE1. Currents produced by the mutant channels primarily showed a depolarizing shift in VDA. The VDA shift was most evident when KCNQ1 was coexpressed with KCNE1, as occurs in the heart. The predicted functional effect in the heart is that there will be significantly less IKs activated during normal cardiac myocyte action potentials. To mimic the heterozygous state, we examined currents from cells expressing a 50:50 mix of KCNQ1-WT and R562S subunits with KCNE1. Although there was a depolarizing shift in VDA for the WT/R562S mix compared with WT channels, it was not a strongly dominant current suppression. Patients with LQT1 are characterized by having arrhythmia episodes triggered by physical exertion and higher heart rates. Moreover, they have abnormal QTc prolongation during exercise. To evaluate the effect of KCNQ1-R562S on rate-related Kþ conductance accumulation in vitro we introduced a series of depolarizing steps with varying frequencies. Channels that were composed of mutant subunits or mixed WT/mutant showed slower accumulation of Kþ conductance than WT channels. This was clinically paralleled by the QTc duration response to exercise in the 2 children who were heterozygous for KCNQ1-R562S.
The C-terminus of KCNQ1 is important for channel gating, assembly, and trafficking.18 The direct interaction between the C-termini of KCNQ1 and KCNE1 affects activation and deactivation kinetics of IKs current.18,22 The amino acid sequence of the KCNQ1 region comprising helices C and D is highly conserved throughout evolution suggesting that mutations in the region are not functionally tolerated (Fig. 2). In the present study, we demonstrate that although KCNQ1R562S mutation did not impair the physical interaction with KCNE1-WT, its interaction with KCNE1 subunits amplified the biophysical mutant effects. The distal half of KCNQ1 helix-C encompassing a cluster of basic residues is important for channel modulation by PIP2, and these region mutations can impair interaction with PIP2.23 The potential mechanism of interaction with PIP2 is likely an electrostatic attraction between these side chains and the negatively charged polar heads of the phospholipid. Furthermore, we observe that KCNQ1-R562S mutation exhibited significantly lower binding to PIP2 than KCNQ1-WT, likely reflecting a lower affinity for PIP2. Because R562 resides between helices C and D of the C-terminus, it is reasonable to postulate that flexibility is important in that precise area. Our modeling of intrinsic disorder (Fig. 7), a characteristic of flexible regions,
1182
Canadian Journal of Cardiology Volume 34 2018
Figure 7. Analysis of the effect of R526S mutation on the intrinsic disorder propensity of KCNQ1 protein. (A, B) Evaluating intrinsic disorder propensity of the KCNQ1 protein (A) and its R526S mutant (B) by a series of per-residue disorder predictors. Profiles generated by PONDR VLXT, PONDR VL3, and PONDR FIT are shown by black, red, and pink lines, respectively. Dashed cyan lines show the mean disorder propensity calculated by averaging disorder profiles of individual predictors. Light pink shadow around the PONDR FIT curves shows error distribution. In these analyses, the predicted intrinsic disorder scores above 0.5 are considered to correspond to the disordered regions; scores between 0.2 and 0.5 are considered flexible. (C) Comparison of the mean disorder profiles of KCNQ1 protein and its R526S mutant (bottom panel) and “disorder difference spectrum” calculated as a simple difference between the mean disorder curves for mutant and wild-type protein (middle panel). A negative peak reflects a mutation-induced local decrease in the intrinsic disorder propensity. Top panel shows “ANCHOR difference spectrum” calculated as a simple difference between the ANCHOR curves calculated for the mutant and wild-type protein and showing how the R562S substitution may affect local disorder-based binding potential of the protein.
predicts that impairment of movement by the mutant residue may play a role in the functional phenotype. A previously described LQT1 mutation (c.1685G>T, p.R562M) at this site of KCNQ1 has also been identified,24 and functional characterization showed the reduced current amplitude largely due to a depolarizing shift in the voltage dependence of activation with perturbation of PIP2 regulation.23 We observed that KCNQ1-R562S behaved similar to R562M but with 2 notable exceptions. R562S did not appear to perturb interactions with KNCE1 whereas R562M did, and the Vh for R562S was 15 mV more depolarized than R562M.23 Both R562S and R562M result in reduced PIP2 binding. A common polymorphism site in KCNE1 (A/G at nucleotide 112 [G112A] alternately coding glycine or serine at position 38) was noted to be heterozygous in the proband. The relative allele frequency for glycine/serine is approximately 0.66/0.33 in the general population.25 Neither genome wide association studies nor family information supports an association between LQTS and G38S.26,27
Recent studies, however, suggest that homozygous KCNE138S might cause a mild reduction in IKs and might thereby increase an arrhythmogenic potential particularly in the presence of QT prolonging mutations.28,29 In our investigation, the functional electrophysiological characteristics of KCNE1-38S, KCNE1-38G, or KCNE1-Mix coexpressed on KCNQ1-R562S were not significantly different. Accordingly, we conclude that neither KCNE1 allele (38S or 38G) differentially impacted the phenotype of the KCNQ1-R562S mutation. JLNS has been linked to pathogenic gene variants (KCNQ1 and KCNE1) in a homozygous or compound heterozygous state with autosomal recessive inheritance.1,2,30 Individuals with JLNS are typically characterized by complete loss of hearing from birth, markedly prolonged QTc interval, and severe cardiac arrhythmia events at an early age. The clinical course of our proband and family in this present study suggested another genotype-phenotype profile that is characterized by postlingual deafness and later arrhythmia
Liu et al. Functional KCNQ1 Mutation in LQTS
presentation in the homozygous state and benign course for the heterozygous state. The proband did not experience syncope until young adulthood, and the heterozygous children have never had symptomatic arrhythmia despite marked QTc prolongation during exercise. More recent studies have highlighted cases of AR-LQT1 without deafness and conversely AD-LQT1 with deafness, thus making the mechanistic connection between arrhythmia and deafness less straightforward.30 One limitation to our analyses is that there are no clinical characteristics of deafness caused by mutant KCNQ1 that can distinguish from other causes. Furthermore, we do not have DNA sequencing of other deafness-associated genes in this family. Nevertheless, other genetic etiologies for deafness seem unlikely given that no other family member has reported hearing loss at any age. Conclusions LQTS with late-onset hearing-loss and arrhythmia features is presented in a patient homozygous for the KCNQ1-R562S mutation. This mutation resides in a critical area of the channel that produces IKs that is important for functional KCNE1 modulation and membrane lipid interaction. Channels carrying only mutant or a mix of WT and mutant subunits expressed currents with shifted voltage-dependent gating that reduced current at physiological potentials and decreased rate-related accumulation of Kþ conductance. Impaired PIP2 interactions and altered protein flexibility in mutant channel proteins likely underlie the mechanism of channel dysfunction. This study expands the clinical spectrum of KCNQ1 mutations and points to areas needing further research. Future investigation into precise genotype-phenotype relationships will likely shed more light onto this intriguing area. Funding Sources This work was supported by the National Institutes of Health (HL120782 to T.V.M.). Disclosures The authors have no conflicts of interest to disclose. References 1. Schwartz PJ, Stramba-Badiale M, Crotti L, et al. Prevalence of the congenital long-QT syndrome. Circulation 2009;120:1761-7. 2. Jervell A, Lange-Nielsen F. Congenital deaf-mutism, functional heart disease with prolongation of the Q-T interval and sudden death. Am Heart J 1957;54:59-68. 3. Andrsova I, Novotny T, Kadlecova J, et al. Clinical characteristics of 30 Czech families with long QT syndrome and KCNQ1 and KCNH2 gene mutations: importance of exercise testing. J Electrocardiol 2012;45: 746-51. 4. Kapa S, Tester DJ, Salisbury BA, et al. Genetic testing for long-QT syndrome: distinguishing pathogenic mutations from benign variants. Circulation 2009;120:1752-60. 5. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 2016;536:285-91.
1183
6. Melman YF, Um SY, Krumerman A, Kagan A, McDonald TV. KCNE1 binds to the KCNQ1 pore to regulate potassium channel activity. Neuron 2004;42:927-37. 7. Chen J, Weber M, Um SY, et al. A dual mechanism for I(Ks) current reduction by the pathogenic mutation KCNQ1-S277L. Pacing Clin Electrophysiol 2011;34:1652-64. 8. Romero P, Obradovic Z, Li X, et al. Sequence complexity of disordered protein. Proteins 2001;42:38-48. 9. Peng K, Radivojac P, Vucetic S, Dunker AK, Obradovic Z. Lengthdependent prediction of protein intrinsic disorder. BMC Bioinformatics 2006;7:208. 10. Xue B, Dunbrack RL, Williams RW, Dunker AK, Uversky VN. PONDR-FIT: a meta-predictor of intrinsically disordered amino acids. Biochim Biophys Acta 2010;1804:996-1010. 11. Walsh I, Giollo M, Di Domenico T, et al. Comprehensive large-scale assessment of intrinsic protein disorder. Bioinformatics 2015;31:201-8. 12. Fan X, Kurgan L. Accurate prediction of disorder in protein chains with a comprehensive and empirically designed consensus. J Biomol Struct Dyn 2014;32:448-64. 13. Peng Z, Kurgan L. On the complementarity of the consensus-based disorder prediction. Pac Symp Biocomput 2012:176-87. 14. Dosztanyi Z, Meszaros B, Simon I. ANCHOR: web server for predicting protein binding regions in disordered proteins. Bioinformatics 2009;25: 2745-6. 15. Meszaros B, Simon I, Dosztanyi Z. Prediction of protein binding regions in disordered proteins. PLoS Comput Biol 2009;5:e1000376. 16. Luo S, Michler K, Johnton P, Macfarlane PW. A comparison of commonly used QT correction formulae: the effect of heart rate on the QTc of normal ECGs. J Electrocardiol 2004;37(Suppl):81-90. 17. Suto F, Zhu W, Chan A, Gross GJ. IKr and IKs remodeling differentially affects QT interval prolongation and dynamic adaptation to heart rate acceleration in bradycardic rabbits. Am J Physiol Heart Circ Physiol 2007;292:H1782-8. 18. Wiener R, Haitin Y, Shamgar L, et al. The KCNQ1 (Kv7.1) COOH terminus, a multitiered scaffold for subunit assembly and protein interaction. J Biol Chem 2008;283:5815-30. 19. Howard RJ, Clark KA, Holton JM, Minor DJ. Structural insight into KCNQ (Kv7) channel assembly and channelopathy. Neuron 2007;53: 663-75. 20. Sachyani D, Dvir M, Strulovich R, et al. Structural basis of a Kv7.1 potassium channel gating module: studies of the intracellular c-terminal domain in complex with calmodulin. Structure 2014;22:1582-94. 21. Thomas AM, Harmer SC, Khambra T, Tinker A. Characterization of a binding site for anionic phospholipids on KCNQ1. J Biol Chem 2011;286:2088-100. 22. Zheng R, Thompson K, Obeng-Gyimah E, et al. Analysis of the interactions between the C-terminal cytoplasmic domains of KCNQ1 and KCNE1 channel subunits. Biochem J 2010;428:75-84. 23. Dvir M, Strulovich R, Sachyani D, et al. Long QT mutations at the interface between KCNQ1 helix C and KCNE1 disrupt I(KS) regulation by PKA and PIP(2). J Cell Sci 2014;127:3943-55. 24. Van Langen IM, Birnie E, Alders M, et al. The use of genotypephenotype correlations in mutation analysis for the long QT syndrome. J Med Genet 2003;40:141-5.
1184
25. Gouas L, Nicaud V, Berthet M, et al. Association of KCNQ1, KCNE1, KCNH2 and SCN5A polymorphisms with QTc interval length in a healthy population. Eur J Hum Genet 2005;13:1213-22. 26. Newton-Cheh C, Eijgelsheim M, Rice KM, et al. Common variants at ten loci influence QT interval duration in the QTGEN Study. Nat Genet 2009;41:399-406. 27. Kanters JK, Olesen MS, Christiansen M. KCNE1 G38S polymorphism is not the cause of long QT syndrome. J Electrocardiol 2016;49:249-50. 28. Yamaguchi Y, Nishide K, Kato M, et al. Glycine/Serine polymorphism at position 38 influences KCNE1 subunit’s modulatory actions on rapid and slow delayed rectifier Kþ currents. Circ J 2014;78:610-8.
Canadian Journal of Cardiology Volume 34 2018 29. Yamaguchi Y, Mizumaki K, Hata Y, et al. Latent pathogenicity of the G38S polymorphism of KCNE1 Kþ channel modulator. Heart Vessels 2017;32:186-92. 30. Giudicessi JR, Ackerman MJ. Prevalence and potential genetic determinants of sensorineural deafness in KCNQ1 homozygosity and compound heterozygosity. Circ Cardiovasc Genet 2013;6:193-200.
Supplementary Material To access the supplementary material accompanying this article, visit the online version of the Canadian Journal of Cardiology at www.onlinecjc.ca and at https://doi.org/10. 1016/j.cjca.2018.06.015.