Kir2.1 mutations differentially increase the risk of flecainide proarrhythmia in Andersen Tawil Syndrome

Abstract

Background: Flecainide and other class-Ic antiarrhythmic drugs (AADs) are widely used in Andersen-Tawil syndrome type 1 (ATS1) patients. However, class-Ic drugs might be proarrhythmic in some cases. We investigated the molecular mechanisms of class-I AADs proarrhythmia and whether they might increase the risk of death in ATS1 patients with structurally normal hearts.

Methods and results: Of 53 ATS1 patients reviewed from the literature, 54% responded partially to flecainide, with ventricular arrhythmia (VA) reduction in only 23%. Of the latter patients, VA persisted in 20-50%. Flecainide was ineffective in 23%, and surprisingly, 13.5% suffered a non-fatal cardiac arrest. In five cardiac-specific ATS1 mouse models (Kir2.1^Δ314-315, Kir2.1^C122Y, Kir2.1^G215D and Kir2.1^R67W and Kir2.1^S136F), flecainide or propafenone (40 mg/Kg i.p.) differentially prolonged the P wave, and the PR, QRS and QTc intervals compared to Kir2.1^WT; Kir2.1^S136F had milder effects. Flecainide increased VA inducibility in all mutant mice except Kir2.1^S136F, which exhibited significant VA reduction. At baseline, Kir2.1^G215D cardiomyocytes had the lowest inward rectifier K+ channel (I_K1) reduction, followed by Kir2.1^C122Y, Kir2.1^R67W and Kir2.1^S136F. Kir2.1^C122Y cardiomyocytes had a significant decrease in sodium inward current (I_Na). Flecainide (10 μM) slightly increased I_K1 density in Kir2.1^WT and Kir2.1^S136F, while it decreased both I_K1 and I_Na in Kir2.1^C122Y and Kir2.1^R67W, despite normal trafficking of mutant channels. Optical mapping in ATS1 patient-specific iPSC-CM monolayers expressing Kir2.1^C122Y, Kir2.1^G215D and Kir2.1^R67W showed an increase in rotor incidence at baseline and under flecainide, confirming the drugś proarrhythmic effect. Lastly, in-silico molecular docking predicts that the Kir2.1-Cys₃₁₁ pharmacophore-binding site is altered in Kir2.1^C122Y heterotetramers, reducing flecainide accessibility and leading to channel closure and arrhythmias.

Conclusions: Class-Ic AADs are only partially effective and might be proarrhythmic in some ATS1 patients. Kir2.1 mutations impacting the resting membrane potential and cellular excitability create a substrate for life-threatening arrhythmias, raising significant concern about using these drugs in some ATS1 patients.

Keywords: ATS1; Arrhythmias; Flecainide; Ion channel diseases; Kir2.1-NaV1.5 channelosome; Sudden Cardiac Death; class-Ic AADs.

Figure 1.. Arrhythmogenic responses to Ic class AADs in ATS1 patients

A: Recorded polymorphic ventricular tachycardia (VT) from an automated external defibrillator in Kir2.1^C122Y proband (II.2). Pedigree of the family is shown in the upper left square. B: Recorded polymorphic ventricular tachycardia (upper panel) and a high burden of non-sustained ventricular tachycardias (NSVT) in Kir2.1^G215D proband (I.2) under Flecainide + Nadolol (middle panel) or Propafenone + Nadolol (bottom panel). C: ECG from Kir2.1^R67W patient shows polymorphic VT. All Probands are indicated with a black arrow.

Figure 2.. Class-Ic-based antiarrhythmic therapy in AST1.

A-B: Venn diagram (A) and schematic representation (B) of class-Ic therapy response in ATS1 patients. Data show 23% of patients who benefited from a full or greater than 90% suppression of VA; 54% reported substantial but incomplete improvement during flecainide treatment, mostly maintaining 20 to 50% of VA. In 23% of patients, flecainide was completely ineffective; 13.5% experienced an additional non-fatal cardiac arrest with an appropriate ICD shock in the presence of flecainide.

Figure 3.. ATS1 mice recapitulate the pathological ECG phenotype

A-B: Time-course after a single dose of flecainide (A) or propafenone (40 mg/Kg) (B) reveals prolonged P wave, PR, QRS and QTc in Kir2.1 mutant animals compared to controls (black). Every value represents the averaged P waves, PR, QRS and QTc intervals from ten consecutive beatings. Arrows indicate the time of flecainide/propafenone administration. Representative beats of Kir2.1^WT (black), Kir2.1^Δ314−315 (green), Kir2.1^C122Y (red), Kir2.1^G215D (purple), Kir2.1^R67W (blue) and Kir2.1^S136F (orange) mice after 5 min of flecainide administration are also indicated. Statistical analysis by two-tailed ANOVA. * = p<0.05; ** = p<0.01; **** = p<0.0001.

Figure 4.. Flecainide is proarrhythmic and alters cardiac conduction in ATS1 mice

Representative electrocardiograms (ECG) lead-II traces recordings in AAV-transduced Kir2.1^C122Y (A) animals showing frequent premature ventricular complexes (PVCs), Kir2.1^G215D (B) and Kir2.1^Δ314−315 (C) animals showing atrioventricular block, and Kir2.1^R67W (D) animals showing non-sustained ventricular tachycardia (NSVT) after flecainide administration.

Figure 5.. Flecainide increases susceptibility to arrhythmias in ATS1 mutant mice.

A: Representative ECG lead-II traces after a train of intracardiac ventricular pulses in AAV-transduced Kir2.1^WT (black; N=10), Kir2.1^Δ314−315 (green; N=8), Kir2.1^C122Y (red; N=8), Kir2.1^G215D (purple; N=6), Kir2.1^R67W (blue; N=6) and Kir2.1^S136F (orange; N=8) animals with periods of polymorphic ventricular tachycardia (PVT) at baseline. A short duration of non-sustained ventricular tachycardias (NSTV) is indicated. B: Contingency plots showing the number of animals with the arrhythmogenic response after intracardiac stimulation at baseline (top) and under flecainide (20mg/Kg) (bottom). C: Representative ECG lead-II trace showing longer NSVT runs after flecainide administration. D: Graph shows ventricular tachycardias (VT) duration at baseline and under flecainide administration. Statistical analysis using the Fisher’s exact test. * = p<0.05; ** = p<0.01; *** = p<0.001; **** = p<0.0001

Figure 6.. Cardiac expression of ATS1 mutations alter mouse ventricle electrophysiology in isolated cardiomyocytes.

A: Current-voltage (I/V) relationships of inward rectifying potassium current I_K1 and B: I_Na density in Kir2.1^WT (black), Kir2.1^Δ314−315 (green), Kir2.1^C122Y (red), Kir2.1^G215D (purple), Kir2.1^R67W (blue) and Kir2.1^S136F (orange) cardiomyocytes at baseline and under flecainide superfusion. C: I_K1 slope from −140 to −60 mV at baseline and under flecainide administration. D: Sodium density I_Na peak (pA/pF) at −35mV. E: Percentage of remaining I_Na density after flecainide inhibition. Statistical analyses were conducted using two-tailed ANOVA. * = p<0.05; ** = p<0.01; **** = p<0.0001.

Figure 7.. Flecainide leads to conduction defects and re-entrant arrhythmias in patient-specific iPSC-CMs.

A: Representative single camera pixel recording from an optical mapping experiment show velocity maps with 1 ms activation isochrones in patient-specific iPSC-CMs^C122Y (red) and CRISPR-mediated isogenic control iPSC-CMs^C122Y (blue) hearts paced at a basic cycle length (BCL) of 400 ms. The color bar indicates conduction velocity (CV; cm s-1). The white asterisks indicate the pacing point and the propagation direction is indicated by the white arrow. B: Re-entrant arrhythmias from C122Y, G215D and R67W patient-specific iPSC-CMs monolayers under flecainide treatment. Below each map is a single pixel recording revealing different patterns of monomorphic or polymorphic re-entrant tachycardia maintained by one self-sustaining rotor. C: The CV restitution curve displayed slower velocities in iPSC-CMs^C122Y monolayers at all frequencies tested. All groups presented slightly slower velocities at higher frequencies. Each value is the mean ± SEM (N=5 differentiations; n indicates number of monolayers per condition; two-way ANOVA corrected by Tukeýs multiple comparisons test, * p<0.05; ** p<0.01; a=IC C122Y vs C122Y; b= IC C122Y Fleca vs C122Y Fleca; c= IC C122Y vs 122Y Fleca; and d= IC 122Y vs IC C122Y Fleca). D: Contingency plots of number of monolayers show arrhythmia inducibility for each group. Data show a high rate of arrhythmia susceptibility in mutant iPSC-CMs. Each value is the mean ± SEM (Fisher’s exact test for contingency data).

Figure 8.. Kir2.1 mutated channels display conformational alteration in the pharmacophore binding site at Cys₃₁₁.

A: Tridimensional (3D) representation of Kir2.1^WT showing cavities C (magenta) and D (yellow) in the absence (top) and presence (bottom) of PIP₂ molecules. Note that PIP₂ is required for proper flecainide binding in c and d cavities. PIP₂ (red arrow) and flecainide (blue arrow) are also indicated. B: 3D representation of flecainide binding in cavity a (green) and b (blue) in the absence (top) and presence (bottom) of PIP₂ molecules in Kir2.1^WT. Gibbs free energy values are shown, being more stable in the presence of PIP₂. C: 3D representation of flecainide binding in Kir2.1^C122Y heterotetramer. Flecainide is fully present in Kir2.1^C122Y heterotetramer. D: Solvent Accessible Surface Area (SASA) of each Cys₃₁₁-pharmacophore in Kir2.1^WT (blue) and Kir2.1^C122Y (red) in sequential incorporation flecainide molecules. E: Comparative table of Gibbs free energy values of flecainide binding to each chain in Kir2.1^WT, Kir2.1^C122Y and Kir2.1^S136F channels. F: Schematic representation of Kir2.1^WT (top), Kir2.1^C122Y (middle) and Kir2.1^S136F (bottom) channel showing a similar pore length (13.71 nm; 13.88 nm; and 16.86 nm) and maxRadius (5.69 nm; 3.27 nm; and 5.43 nm), respectively.