Long-QT founder variant T309I-Kv7.1 with dominant negative pattern may predispose delayed afterdepolarizations under β-adrenergic stimulation

The variant c.926C > T (p.T309I) in KCNQ1 gene was identified in 10 putatively unrelated Czech families with long QT syndrome (LQTS). Mutation carriers (24 heterozygous individuals) were more symptomatic compared to their non-affected relatives (17 individuals). The carriers showed a mild LQTS phenotype including a longer QTc interval at rest (466 ± 24 ms vs. 418 ± 20 ms) and after exercise (508 ± 32 ms vs. 417 ± 24 ms), 4 syncopes and 2 aborted cardiac arrests. The same haplotype associated with the c.926C > T variant was identified in all probands. Using the whole cell patch clamp technique and confocal microscopy, a complete loss of channel function was revealed in the homozygous setting, caused by an impaired channel trafficking. Dominant negativity with preserved reactivity to β-adrenergic stimulation was apparent in the heterozygous setting. In simulations on a human ventricular cell model, the dysfunction resulted in delayed afterdepolarizations (DADs) and premature action potentials under β-adrenergic stimulation that could be prevented by a slight inhibition of calcium current. We conclude that the KCNQ1 variant c.926C > T is the first identified LQTS-related founder mutation in Central Europe. The dominant negative channel dysfunction may lead to DADs under β-adrenergic stimulation. Inhibition of calcium current could be possible therapeutic strategy in LQTS1 patients refractory to β-blocker therapy.


Clinical diagnostics
Patients with suspected LQTS are regularly investigated at the Department of Internal Medicine and Cardiology, and at the Department of Paediatrics (both at the University Hospital Brno and Faculty of Medicine, Masaryk University, Brno, Czech Republic). The diagnosis is established according to ESC Guidelines. 1 The study conformed to the principles outlined in the Declaration of Helsinki. All participants signed a written consent form prior to their inclusion in the study. In the case of participants under the age of 18 years, the written informed consent was obtained from a parent and/or legal guardian. The study was approved by the Multicenter Ethical Committee, University Hospital Brno (Brno, Czech Republic).
All individuals included in this study underwent clinical examination and bicycle ergometry to obtain ECG traces at different adrenergic states. A 12-lead ECG with Mason-Likar modification was used. The initial stress was set to 0.5 W/kg, and increased by 0.5 W/kg every three minutes to achieve a heart rate higher than the submaximal value with respect to age and sex.
All ECGs were recorded as paper printings at the speed of 50 mm/s and voltage of 20 mm/mV, and QT and RR intervals were measured manually for the periods of rest and in the fourth minute of the recovery period of the exercise test. In the majority of cases, the QT interval was measured in the lead V5; the other leads were used only when the end of the T wave could not be discriminated in this lead. In cases when the end of T wave was not clearly visible, the threshold method was used. The QT intervals were corrected for the respective heart rate using Bazett's formula: QTc = QT/√RR (both intervals were measured in seconds).

Genetic testing
Between 2000 and 2018, 132 unrelated index cases with susceptibility to LQTS were examined at the Department of Medical Genetics (University Hospital Brno and Faculty of Medicine, Masaryk University, Brno, Czech Republic). Informed consents and peripheral blood samples were collected from patients. DNA was extracted by standard molecular techniques. Molecular analysis of LQTS-associated genes including the KCNQ1 gene was performed according to current practises for molecular genetics diagnostics. The classical method (multiplex PCR/SSCP analysis of 3 LQTS major genes) was followed by Sanger sequencing on ABI 3100 Genetic Analyser (Applied Biosystems™, Foster City, CA, USA).
Primers for the screening of entire coding regions of genes KCNQ1, KCNH2 and SCN5A were designed with Primer-BLAST tool. 2  and 2) Target enriched library prepared with hybridization capture-based method using KAPA HyperPlus Kit with SeqCap EZ Choice Library Kit (Roche, Basel, Switzerland). The procedure has been performed as recommended by the manufacturer. Genetic counselling and testing of first-degree relatives have been offered to patients at risk.

In silico analysis
Various in silico tools were used to predict the possible clinical impact of the identified sequence variant (Suppl. Tab. S1). The functional impact of the amino acid substitution was predicted with SIFT, Provean, 3 MutationTaster, FATHMM, 4 and PMUT. 5 The conservation of the impacted amino acid position was measured with LRT and MutationAssessor. The visualization of protein conservation across species was performed in MEGA7 software. 6 Protein sequences of included species were obtained from Ensembl Genome Browser (www.ensembl.org/index.html). Sequences were aligned with ClustalW with preset parameters. Allele frequency of the substitution was determined from online databases ExAC 7 and GnomAD.

Haplotype analysis
For the haplotype analysis, 9 STR (short tandem repeats) markers spanning the ~11.9-Mb region of chromosome 11 (including the KCNQ1 gene) were chosen from UCSC Genome  The voltage dependence of steady-state activation was fitted using the Boltzmann cell. This implicates that the fraction of L-type Ca 2+ channels in the t-tubular membrane fCaL,t and related fKL,t should be around 0.8.
The time constants related to the rate of ion diffusion from the dyadic spaces to the subsarcolemmal spaces (τdsss = 0.819 ms, τdtst = 0.214 ms) were set to be consistent with the rate of ion diffusion from dyads in our previous model. 8 The time constants of ion diffusion  12 and that inhibition of IK1 and IKs causes only a minor change of the human AP (prolongation of APD90 by 3 to 5%) 13 , a partial modification of the parameters describing membrane transport system in ref. 8   The t-tubular fractions of ion transporters (f x,t ) were adopted from 8     Suppl. Fig. S4: Comparison of the effect of total suppression of IKs on AP in the presented model and in the validated model of human epicardial myocyte published by O'Hara et al. 17 .
The red traces represent the first AP elicited from 1 Hz steady state after IKs suppression.

Statistical analysis
The data are mostly presented by the arithmetic mean (± SD from n patients, or ± SEM from n cells; Origin, version 8.5.1; OriginLab Corporation). To determine statistical significance of differences, paired/unpaired t-tests, and one-way/repeated measures ANOVA with the Bonferroni post-test were used; P < 0.05 was considered statistically significant. If the difference between the arithmetic and geometric means was >10%, the geometric mean x (3)

Clinical diagnostics
Representative examples of electrocardiograms (EGCs) of two T309I carriers, recorded at rest and in the 4 th minute of recovery after the exercise, are shown in Suppl. Fig. S7, the first one recorded in a female who suffered cardiac arrest and was successfully resuscitated (parts A and B), the second one from an asymptomatic female (parts C and D). For results characterizing differences in QTc lengths in T309I carriers and their unaffected relatives, see No significant changes were apparent in PQ interval and QRS complex (Suppl. Fig. S8).
In the population over 16 years old, a non-significant tendency to a prolonged PQ interval (P = 0.07) and a shortened QRS complex (P = 0.11) was apparent in T309I carriers. The proarrhythmic action of WT/T309I channels was prevented by 5%-inhibition of cardiac calcium current (ICa); contemporary change in the peak value of Ca 2+ transient was negligible (Suppl. Fig. 11). Suppl.