Beneficial effects of chronic mexiletine treatment in a human model of SCN5A overlap syndrome

Abstract

Aims

SCN5A mutations are associated with various cardiac phenotypes, including long QT syndrome type 3 (LQT3), Brugada syndrome (BrS), and cardiac conduction disease (CCD). Certain mutations, such as SCN5A-1795insD, lead to an overlap syndrome, with patients exhibiting both features of BrS/CCD [decreased sodium current (I_Na)] and LQT3 (increased late I_Na). The sodium channel blocker mexiletine may acutely decrease LQT3-associated late I_Na and chronically increase peak I_Na associated with SCN5A loss-of-function mutations. However, most studies have so far employed heterologous expression systems and high mexiletine concentrations. We here investigated the effects of a therapeutic dose of mexiletine on the mixed phenotype associated with the SCN5A-1795insD mutation in HEK293A cells and human-induced pluripotent stem cell–derived cardiomyocytes (hiPSC-CMs).

Methods and results

To assess only the chronic effects on trafficking, HEK293A cells transfected with wild-type (WT) SCN5A or SCN5A-1795insD were incubated for 48 h with 10 µm mexiletine followed by wash-out, which resulted in an increased peak I_Na for both SCN5A-WT and SCN5A-1795insD and an increased late I_Na for SCN5A-1795insD. Acute re-exposure of HEK293A cells to 10 µm mexiletine did not impact on peak I_Na but significantly decreased SCN5A-1795insD late I_Na. Chronic incubation of SCN5A-1795insD hiPSC-CMs with mexiletine followed by wash-out increased peak I_Na, action potential (AP) upstroke velocity, and AP duration. Acute re-exposure did not impact on peak I_Na or AP upstroke velocity, but significantly decreased AP duration.

Conclusion

These findings demonstrate for the first time the therapeutic benefit of mexiletine in a human cardiomyocyte model of SCN5A overlap syndrome.

Graphical Abstract

Graphical Abstract.

I _Na, sodium current.

What’s new?

• Mexiletine may acutely inhibit late sodium current and chronically enhance sodium channel trafficking through its ‘chaperone’ effect and hence may be beneficial in SCN5A overlap syndromes.

• We here show that the latter can already be achieved by chronic treatment with a clinically relevant concentration of mexiletine.

• We identify mexiletine as a novel approach to rescue decreased peak sodium current, which may prove beneficial in other inherited arrhythmia disorders.

• We furthermore demonstrate for the first time the therapeutic benefit of mexiletine in a human cardiomyocyte model of SCN5A overlap syndrome.

Introduction

The cardiac sodium channel Na_v1.5 is essential for excitability and electrical conduction in the heart. Na_v1.5 is encoded by the SCN5A gene and is located on the membrane of cardiomyocytes (CMs), where it allows the influx of sodium ions leading to the rapid upstroke of the action potential (AP).¹ Mutations in SCN5A cause biophysical changes in sodium current (I_Na), which can lead to different cardiac disorders with distinct clinical features.^2,3 Gain-of-function SCN5A mutations responsible for the long QT syndrome type 3 (LQT3) typically lead to an increased late I_Na, while loss-of-function SCN5A mutations associated with conduction disease and Brugada syndrome (BrS) cause a decrease in peak I_Na.⁴ In some instances, a single SCN5A mutation can lead to a double phenotype, showing characteristics of both gain- and loss-of-function, as is the case for the SCN5A-1795insD mutation.^5–7 This mutation causes an ‘overlap syndrome’, with affected individuals displaying electrocardiogram (ECG) characteristics of cardiac conduction disorder, BrS, and LQT3, in addition to a high risk for (nocturnal) sudden death. Accordingly, the mutation leads to both a reduced peak I_Na and an increased late I_Na.^5–7

Pharmacological therapeutic options for patients carrying SCN5A mutations are limited. β-Blockers are often prescribed in LQT3 but they do not show benefit in all patients.^8,9 Moreover, the management of individuals with an overlap syndrome may be even more complicated. Pacemaker therapy, aimed at preventing (nocturnal) bradycardia-dependent QT prolongation, was initially considered efficacious in SCN5A-1795insD patients; however, a subset of patients experienced ventricular arrhythmias despite pacemaker therapy,¹⁰ indicating a need for additional therapeutic approaches. Mexiletine, a Class Ib antiarrhythmic drug and sodium channel blocker, may shorten QT interval by inhibiting late I_Na in LQT3 patients and has shown benefits in patients with structural heart disease and at risk of recurrent ventricular arrhythmia,^11,12 although its beneficial effects are mutation-dependent.¹³ In addition, incubation with mexiletine for 24–48 h has been shown to rescue the reduced peak I_Na associated with certain loss-of-function SCN5A mutations, most likely by restoring membrane trafficking of Na_v1.5 channels.^14–18 However, the majority of these studies used very high, non-therapeutic concentrations (>100 µm) of mexiletine. Moreover, the investigations were performed in heterologous expression systems (such as HEK293A cells), which, although informative, do not fully recapitulate the cardiomyocyte environment.

Nevertheless, given the potential beneficial effects of mexiletine on both gain (increased late I_Na) and loss (decreased peak I_Na) of function mutations, we here hypothesized that it could ameliorate the overlap syndrome phenotype associated with the SCN5A-1795insD mutation. Using HEK293A cells, we show that chronic administration of a therapeutic concentration of 10 µm (clinical range 2.8–11 µm)^19–21 mexiletine increases both wild-type (WT) and SCN5A-1795insD peak I_Na, while it acutely decreased the mutation-associated late I_Na. Importantly, chronic mexiletine therapy increased peak I_Na and AP upstroke velocity in human-induced pluripotent stem cell–derived cardiomyocytes (hiPSC-CMs) carrying the SCN5A-1795insD mutation, while its acute administration decreased AP duration (APD). These findings demonstrate for the first time the therapeutic benefit of mexiletine in a human cardiomyocyte model of SCN5A overlap syndrome.

Methods

Site-directed mutagenesis

Site-directed mutagenesis was performed on SCN5A-WT cDNA cloned in pSP64T, as previously described.⁵SCN5A-1795insD cDNA was then subcloned into the HindIII–XbaI sites of the expression vector pCGI for bicistronic expression of the channel protein and GFP reporter in HEK293A cells.

Transfection of HEK293A cells

HEK293A cells were cultured with DMEM medium (Dulbecco’s modified Eagle’s medium 21969-035 from Thermo Fisher Scientific) supplemented with 10% foetal bovine serum, 1% l-glutamine (Thermo Fisher Scientific), and 1% penicillin–streptomycin (Thermo Fisher Scientific) in a 5% CO₂ incubator at 37°C. At ∼70% confluency, HEK293A cells were transfected with 0.2 µg of cardiac sodium channel α-subunit cDNA (only SCN5A-WT or only SCN5A-1795insD) and 0.2 µg human β₁-subunit cDNA using Lipofectamine 3000 (Thermo Fisher Scientific).

Generation of human-induced pluripotent stem cell–derived cardiomyocytes

SCN5A-1795insD hiPSC were differentiated into CMs, cryopreserved and thawed as previously described.²² In summary, glass coverslips were coated with 1:100 matrigel in DMEM-F12 (Thermo Fisher Scientific) for 2 h at 37°C before thawed hiPSC-CMs were seeded at a density of 2.1–4 × 10³/cm². The day after thawing and then every 2–3 days following, cells were refreshed with RPMI medium (Thermo Fisher Scientific, 21875034) supplemented with B27. Three days prior to patch clamp experiments, the medium was supplemented with 5% foetal calf serum. Patch clamp experiments were performed 7–10 days after initial cell seeding.

Drug incubation

For chronic experiments, transfected HEK293A cells and SCN5A-1795insD hiPSC-CMs were incubated for 48 h in medium containing either vehicle (H₂0) or mexiletine (Tocris Bioscience) at 37°C. HEK293A cells were then dissociated into single cells using trypsin (Thermo Fisher Scientific; final concentration 0.25%) for 2–3 min and re-suspended in medium without the drug (see scheme in Figures 1–6). Cells exhibiting green fluorescence were selected for electrophysiological experiments. Similarly, hiPSC-CMs incubated with mexiletine or vehicle were washed three times with RPMI/B27 medium supplemented with 5% serum prior to patch clamp experiments. To assess the acute effect of the drug, transfected HEK293A cells and SCN5A-1795insD hiPSC-CMs incubated for 48 h with mexiletine or vehicle were superfused for 5 min with 10 µm mexiletine (see scheme in Figures 2 and 4–6).

Figure 1.

Forty-eight hour incubation with 10 µm mexiletine increases peak I_Na in HEK293A cells transfected with SCN5A-WT or SCN5A-1795insD. (A) Schematic representation of the performed experiments. (B, C) Representative I_Na traces recorded in HEK293A cells transfected with either SCN5A-WT (B) or SCN5A-1795insD (C) following 48 h incubation with vehicle or mexiletine (10 µm). (D–G) Average peak I_Na current–voltage (I–V) relationships (D, E) and voltage dependence of activation and inactivation (F, G) in HEK293A cells transfected with SCN5A-WT or SCN5A-1795insD following 48 h incubation with vehicle or mexiletine (10 µm). Insets: voltage protocols; *P < 0.05, two-way RM ANOVA followed by Holm–Sidak test for post hoc analysis. WT, wild-type.

Figure 6.

Forty-eight hour incubation with 10 µm mexiletine increases peak I_Na in SCN5A-1795insD hiPSC-CMs while acute mexiletine has no effect. (A, B) Schematic representations of the performed experiments (see panels) and average peak I_NaI–V relationships for SCN5A-1795insD hiPSC-CMs after 48 h incubation with vehicle or mexiletine (10 µm) followed by removal of the drug (A) and after 5 min of re-administration (wash-in) of mexiletine (10 µm) (B). (C, D) Voltage dependence of (in)activation measured in SCN5A-1795insD hiPSC-CMs after 48 h incubation with vehicle or mexiletine (10 µm) followed by removal of the drug (C) and after 5 min re-administration (wash-in) of mexiletine (D). Insets: voltage protocols; *P < 0.05, two-way RM ANOVA followed by Holm–Sidak test for post hoc analysis.

Figure 2.

Acute mexiletine treatment does not affect peak I_Na in HEK293A cells transfected with SCN5A-WT or SCN5A-1795insD previously incubated with either vehicle or mexiletine. (A) Schematic representation of the performed experiments. (B–E) Average peak I_NaI–V relationships before (baseline) and after 5 min wash-in of 10 µm mexiletine in HEK293A cells transfected with SCN5A-WT (B, C) or SCN5A-1795insD (D, E), following 48 h incubation with either vehicle or mexiletine. WT, wild-type.

Figure 4.

Acute mexiletine treatment blocks late I_Na in HEK293A cells transfected with SCN5A-1795insD. (A) Schematic representation of the performed experiments (top panel). Typical late I_Na measurements obtained using a descending ramp protocol (see inset) in HEK293A cells transfected with SCN5A-1795insD previously incubated for 48 h with vehicle (middle left panel) and 10 µm mexiletine (middle right panel) prior to drug application (baseline), after 5 min wash-in of mexiletine (10 µm) and TTX (30 µm) in the continued presence of mexiletine. The mexiletine-sensitive currents and TTX-sensitive currents in the presence of mexiletine are shown in the bottom panels. (B) Average I–V relationships for mexiletine-sensitive current and TTX-sensitive current in the presence of mexiletine in HEK293A cells transfected with SCN5A-1795insD previously incubated for 48 h with vehicle (left panel) and mexiletine (right panel). The TTX-sensitive current in the presence of mexiletine is around the 0 pA/pF level, indicating that mexiletine already blocked completely late I_Na. TTX, tetrodotoxin.

Patch clamp analysis

Data acquisition

Sodium current and APs were measured with the ruptured and perforated patch clamp technique, respectively, using an Axopatch 200B amplifier (Molecular Devices). Voltage control, data acquisition, and analysis were performed with pClamp10.6/Clampfit (Molecular Devices) and custom-made software for I_Na and APs, respectively. Borosilicate glass patch pipettes (Harvard Apparatus) with a tip resistance of 2–2.5 MΩ were used. Cell membrane capacitance was determined dividing the decay time constant of the capacitive transient in response to 5 mV hyperpolarizing steps from −40 mV, by the series resistance. Series resistance and cell membrane capacitance were compensated for ≥80%. Peak I_Na and APs were filtered at 5 kHz and digitized at 40 kHz, while late I_Na was filtered and digitized at 2 and 1 kHz, respectively. Peak I_Na was measured at room temperature, while late I_Na and APs were recorded at 37°C.

Sodium current

Peak I_Na and late I_Na recordings in HEK293A cells were performed using a pipette solution containing the following (in mm): 110 CsF, 10 CsCl, 10 NaF, 1 MgCl₂, 1 CaCl₂, 2 Na₂ATP, 11 EGTA, and 10 HEPES, pH 7.2 (CsOH). The bath solution contained the following (in mm): 140 NaCl, 10 CsCl, 2 CaCl₂, 1 MgCl₂, 5 glucose, and 10 HEPES, pH 7.4 (NaOH). In hiPSC-CMs, peak I_Na and late I_Na measurements were performed using the following external solution (in mm): 130 NaCl, 20 CsCl, 1.8 CaCl₂, 1.2 MgCl₂, 11 glucose, 0.005 nifedipine, and 5.0 HEPES, pH 7.4 (CsOH), and pipette solution containing (in mm) 3.0 NaCl, 133 CsCl, 2.0 MgCl₂, 2.0 Na₂ATP, 2.0 TEACl, 10.0 EGTA, and 5.0 HEPES, pH 7.3 (CsOH). Late I_Na was measured as 30 µm tetrodotoxin (TTX)-sensitive current, using a descending ramp protocol from a holding potential of −120 mV (cycle length of 5 s; see insets in Figures 3 and 4). In hiPSC-CMs, peak I_Na was elicited from a holding potential of −120 mV with a cycle length of 5 s, while in HEK293A cells, the holding potential was −140 mV (see insets in Figures 1, 2, and 6). Current–voltage (I–V) relationships and voltage dependency of activation were characterized with 500 ms depolarizing pulses between −100 and +60 mV (HEK293A cells) or +50 mV (hiPSC-CMs). Voltage dependency of inactivation was analysed using a double-pulse protocol, with a 500 ms pre-pulse and a step to −20 mV. Sodium current was defined as the difference between peak and steady-state current at the end of each step. Current densities were calculated dividing the current amplitude by the cell membrane capacitance. Voltage dependence of activation and inactivation curves was fitted with Boltzmann function (y = [1 + exp{(V − V_1/2)/k}]⁻¹), where V_1/2 is the half-maximal voltage of (in)activation and k is the slope factor.

Figure 3.

Forty-eight hour incubation with mexiletine increases late I_Na in HEK293A cells transfected with SCN5A-1795insD. (A) Schematic representation of the performed experiments and typical late I_Na recordings obtained using a descending ramp protocol in HEK293A cells transfected with SCN5A-1795insD previously incubated for 48 h with vehicle (left panel) or mexiletine (right panel) before (baseline) and after application of 30 µm TTX. TTX-sensitive current was obtained by subtraction of the current before and after application of TTX. (B) Average I–V relationships for TTX-sensitive current in HEK293A cells transfected with SCN5A-1795insD previously incubated for 48 h with vehicle or mexiletine (left panel); *P < 0.05, two-way RM ANOVA followed by Holm–Sidak test for post hoc analysis. Boxplots (right panel) depicting late I_Na densities (median: boxes represent interquartile range; whiskers represent 95% interval) determined at −10 mV; *P < 0.05, Mann–Whitney test. TTX, tetrodotoxin.

Action potentials

Action potentials were measured using a modified Tyrode’s solution containing the following (in mm): 140 NaCl, 5.4 KCl, 1.8 CaCl₂, 1.0 MgCl₂, 5.5 d-glucose, and 5 HEPES, pH 7.4 (NaOH). Pipettes were filled with the following (in mm): 125 K-gluconate, 20 KCl, 5 NaCl, 0.44 amphotericin B, and 10 HEPES, pH 7.2 (KOH). Action potentials were elicited at 1 Hz by 3 ms, ≈1.2× threshold current pulses through the patch pipette. To overcome the lack of the inward rectifying potassium current (I_K1) in hiPSC-CMs, which limits the functional availability of I_Na,²³ we injected a 2 pA/pF in silico I_K1 with kinetics of Kir2.1 channels through dynamic clamp, as previously described and validated.²⁴ Resting membrane potential (RMP), AP amplitude (APA), maximal AP upstroke velocity (V_max), and APD at 20, 50, and 90% repolarization (APD₂₀, APD₅₀, and APD₉₀, respectively) were analysed. Data from 10 consecutive APs were averaged and potentials were corrected for the calculated liquid junction potential.²⁵

Statistical analysis

Data were analysed using Sigma Stat 3.5 (Systat Software Inc.) and GraphPad Prism version 8.4.3(686) for Windows (GraphPad Software). Values are shown as mean ± SEM unless stated otherwise. Normality was tested by Kolmogorov–Smirnov. Unpaired Student’s t-test, paired Student’s t-test, Mann–Whitney test, and two-way RM ANOVA followed by Holm–Sidak test for post hoc analysis were used where appropriate and indicated in the relevant figures and tables. The level of statistical significance was set to P < 0.05.

Results

Chronic mexiletine incubation increases peak sodium current in SCN5A-WT and SCN5A-1795insD transfected HEK293A cells

We first assessed the chronic effect of mexiletine on WT I_Na and mutant I_Na using HEK293A cells transfected with either SCN5A-WT or SCN5A-1795insD. Cells were incubated with 10 µm mexiletine for 48 h followed by dissociation and resuspension in absence of the drug (see scheme in Figure 1A). SCN5A-1795insD cells showed a markedly reduced I_Na as compared to SCN5A-WT (Figure 1B–E), in line with the loss-of-function phenotype associated with the mutation. Chronic incubation with mexiletine resulted in a significant increase in average I_Na density in both SCN5A-WT (at −30 mV: 62.7%) and SCN5A-1795insD (at −30 mV: 93.6%) as compared to vehicle incubation (Figure 1B–E). Interestingly, the effect of mexiletine on peak I_Na appears larger on SCN5A-1795insD compared to SCN5A-WT, suggesting a possible preferential rescuing by mexiletine of mutant channels; however, further investigations are needed to confirm this. In Figure1F and G and Supplementary material online, Table S1, the effects of 48 h mexiletine incubation on I_Na gating properties are depicted. While mexiletine did not impact on SCN5A-WT gating, it induced a significant shift towards more negative potentials for both voltage dependency of activation and inactivation in HEK293A cells transfected with SCN5A-1795insD. Mexiletine did not affect cell capacitance in either SCN5A-WT or in SCN5A-1795insD (data not shown). Overall, our findings indicate that chronic incubation with mexiletine at a clinically relevant dose of 10 µm is capable of increasing the presence of both SCN5A-WT and SCN5A-1795insD channels on the membrane.

Acute treatment with mexiletine does not affect peak sodium current in HEK293A cells transfected with either SCN5A-WT or SCN5A-1795insD

In the experiments presented in Figure 1, mexiletine was not present in the bath solution during patch clamp measurements in order to investigate specifically the chronic impact of the drug without its direct, acute effect on the channel. We next mimicked the clinical situation of chronic mexiletine therapy with continuous presence of the drug. For this purpose, we employed the same approach as in Figure 1A (48 h of incubation with 10 µm mexiletine or vehicle followed by dissociation and resuspension of cells in absence of the drug), but then acutely re-introduced mexiletine into the bath solution (10 µm for 5 min) (see scheme in Figure 2A). Following chronic incubation with mexiletine or vehicle, acute treatment with mexiletine neither affected peak I_Na in HEK293A cells transfected with SCN5A-WT (Figure2B and C) nor SCN5A-1795insD (Figure2D and E). Hence, the beneficial increase of I_Na following 48 h mexiletine incubation was maintained upon acute (re-)application of mexiletine. Acute exposure to mexiletine did result in changes in voltage dependency of activation and inactivation for both SCN5A-WT and SCN5A-1795insD (data not shown). However, in a subset of cells where wash-out experiments were also performed, we were unable to reverse these effects of mexiletine, and a further negative shift of V_1/2 of (in)activation was observed upon wash-out (see Supplementary material online, Figures S1 and S2). These results suggest a time-dependent effect rather than a direct effect of the drug on voltage dependence of (in)activation.

Chronic effects of mexiletine on late sodium current in HEK293A cells transfected with SCN5A-1795insD

We next assessed the effect of chronic mexiletine incubation on late I_Na in HEK293A cells transfected with SCN5A-1795insD. To measure late I_Na, we used a descending ramp protocol that, in contrast to a single-step protocol, allows measurement of late I_Na across a dynamic voltage range simulating a plateau and repolarization phase of a cardiac AP.²⁶Figure 3A shows typical examples of late I_Na recordings in cells chronically incubated either with vehicle (Figure 3A, left panel) or mexiletine (Figure 3A, right panel). Late I_Na, measured as TTX-sensitive current, was obtained by subtraction of the current recorded in the presence of 30 µm TTX from the current recorded in the absence of the drug (baseline). Forty-eight hour incubation with mexiletine (followed by removal of the drug) significantly increased average late I_Na at −30 to +10 mV as compared to vehicle (Figure 3B, left panel). At −10 mV, late I_Na was −0.33 ± 0.05 pA/pF (n = 16) and −0.58 ± 0.10 pA/pF (n = 18; P < 0.05, Mann–Whitney test) for vehicle and mexiletine, respectively (76.9% increase, Figure 3B, right panel).

Acute effects of mexiletine on late sodium current in HEK293A cells transfected with SCN5A-1795insD

Following the observation that chronic mexiletine incubation increased late I_Na in HEK293A cells transfected with SCN5A-1795insD, we next investigated whether this could be mitigated by the acute effect of mexiletine. Late I_Na was measured in HEK293A cells transfected with SCN5A-1795insD previously incubated for 48 h with vehicle (Figure 4A, left panel) or mexiletine (Figure 4A, right panel), at baseline, after 5 min perfusion with mexiletine (10 µm), and after 5 min of TTX (see scheme in Figure 4A).

Mexiletine-sensitive current was obtained by subtraction of the current recorded in the presence of mexiletine from the baseline current, while TTX-sensitive current was measured by subtraction of the current recorded in the presence of TTX from the current measured after application of mexiletine. As shown in the typical examples in Figure 4A (middle panels), late I_Na is completely blocked by the acute application of mexiletine in cells previously incubated chronically with either vehicle or mexiletine. Consequently, TTX-sensitive currents in acute presence of mexiletine are virtually absent (Figure 4A, bottom panels). This is further substantiated by the average I–V relationships shown in Figure 4B.

Acute mexiletine treatment was able to block both the baseline mutation-induced late I_Na in vehicle-treated cells (at −10 mV, chronic vehicle: −0.33 ± 0.05 pA/pF, n = 16, vs. after acute exposure to mexiletine: −0.03 ± 0.03 pA/pF, n = 8) and the increased late I_Na induced by chronic exposure to mexiletine (at −10 mV, chronic mexiletine: −0.58 ± 0.10 pA/pF, n = 18, vs. after acute exposure to mexiletine: −0.06 ± 0.04 pA/pF, n = 9).

Chronic mexiletine increases action potential upstroke velocity and peak sodium current in SCN5A-1795insD human-induced pluripotent stem cell–derived cardiomyocytes

To investigate the effect of mexiletine in a cardiomyocyte model that resembles more closely the heterozygote situation in patients, we performed experiments in hiPSC-CMs obtained from a SCN5A-1795insD patient.⁷ We assessed the functional impact of mexiletine on AP properties in SCN5A-1795insD hiPSC-CMs, employing dynamic clamp for in silico I_K1 injection in cells paced at 1 Hz at physiological temperature. Incubation of hiPSC-CMs with mexiletine for 48 h, followed by removal of the drug, significantly increased V_max (76.6% increase) (Figure5A and B and Supplementary material online, Table S2), thus importantly rescuing the loss-of-function part of this particular mutation. Acute re-administration of mexiletine following chronic exposure did not significantly affect V_max (Figure5C and D and Supplementary material online, Table S3) and hence did not counteract the beneficial effects of chronic mexiletine on V_max. Action potential amplitude and RMP were not altered in response to chronic (Figure 5B and Supplementary material online, Table S2) or acute mexiletine treatment (Figure 5D and Supplementary material online, Table S3). Chronic administration of mexiletine for 48 h significantly increased APD₂₀, APD₅₀, and APD₉₀ in SCN5A-1795insD hiPSC-CMs (Figure5A and B and Supplementary material online, Table S2), while acute re-administration of mexiletine significantly reduced APD₅₀ and APD₉₀ (Figure5C and D and see Supplementary material online, Table S3). In line with the observed increase in V_max, 48 h incubation with mexiletine significantly increased I_Na density (at 15 mV: 119.2% increase; Figure 6A and Supplementary material online, Table S4) without any changes in voltage dependence of (in)activation (Figure 6C and Supplementary material online, Table S4); mexiletine also did not affect cell capacitance (data not shown). Acute re-administration of mexiletine following 48 h exposure to the drug affected neither I_Na density nor voltage dependence of (in)activation (Figure6B and D and see Supplementary material online, Table S5), indicating that continued exposure to a therapeutic dose of mexiletine maintains the beneficial increase in peak I_Na induced by chronic mexiletine administration.

Figure 5.

Chronic and acute effects of mexiletine on action potential properties in SCN5A-1795insD hiPSC-CMs. (A) Schematic representation of the performed experiments and typical examples of APs recorded at 1 Hz in SCN5A-1795insD hiPSC-CMs after 48 h incubation with either vehicle or mexiletine (10 µm). (B) Average data for maximal upstroke velocity (V_max), AP amplitude, resting membrane potential, AP duration at 20, 50, and 90% repolarization (APD₂₀, APD₅₀, and APD₉₀). *P < 0.05, unpaired Student’s t-test. (C) Schematic representation of the performed experiments and typical examples of APs recorded at 1 Hz in SCN5A-1795insD hiPSC-CMs incubated for 48 h with mexiletine followed by removal of the drug (baseline) and after 5 min acute re-administration (wash-in) of mexiletine. (D) Average data for V_max, APA, RMP, APD₂₀, APD₅₀, and APD₉₀. Insets: time derivatives of the AP upstrokes. *P < 0.05, paired Student’s t-test. AP, action potential; APA, AP amplitude; RMP, resting membrane potential.

Discussion

Given the complex mixed phenotype of both gain and loss of sodium channel function involved, SCN5A overlap syndromes may be difficult to treat.^27–30 We here assessed the beneficial effects of a clinically relevant concentration of mexiletine on the overlap mutation SCN5A-1795insD (associated with features of LQT3, BrS, and CCD, consequent to decreased peak I_Na and increased late I_Na) and investigated for the first time its potential benefit in a human cardiomyocyte model. We first explored the effects of mexiletine in HEK293A cells, allowing us to assess its impact on SCN5A-WT vs. SCN5A-1795insD channels. By incubating HEK293A cells with mexiletine for 48 h followed by removal of the drug, we investigated specifically the chronic impact of mexiletine without its acute effect. Similar to what has previously been observed for other SCN5A mutations,^14–18 chronic mexiletine incubation of HEK293A cells transfected with SCN5A-1795insD increased peak I_Na and hence restored the loss-of-function phenotype. However, most previous studies employed high concentrations of mexiletine (>100 µm),^19–21 which are far higher than the therapeutic concentration of 2.8–11 µm, and are expected to decrease peak I_Na acutely. We observed chronically an increase in peak I_Na using a clinically relevant mexiletine concentration of 10 µm; a similar effect was previously reported by Ruan et al., who demonstrated that 10 µm mexiletine rescued the trafficking defect of the SCN5A-F1473S and restored peak I_Na.¹⁶ Crucially, in our experiments, the acute re-administration of mexiletine at this concentration did not affect peak I_Na and thus did not counteract its chronic effect. In line with the apparent increase in membrane trafficking of SCN5A-1795insD channels, chronic incubation with mexiletine also resulted in a larger late I_Na. However, this detrimental effect was abrogated by the subsequent re-administration of mexiletine which blocked late I_Na, indicating its efficacious impact on the gain-of-function consequences of this mutation. Interestingly, chronic mexiletine incubation also significantly increased SCN5A-WT peak I_Na, which was again not abrogated by acute re-administration of the compound. This indicates that mexiletine at a concentration of 10 µm is also capable of enhancing membrane trafficking of WT Na_v1.5.

While the HEK293A cell data provided information on the distinct effects of mexiletine on SCN5A-WT vs. SCN5A-1795insD channels, this does not reflect the actual situation in patients who carry the mutation in heterozygous form. Moreover, HEK293A cells do not fully recapitulate the cardiomyocyte environment and lack many Na_v1.5 interacting proteins.³¹ We therefore also investigated the effect of mexiletine in hiPSC-CMs derived from a patient with the SCN5A-1795insD mutation, allowing us to assess its impact in a more clinically relevant model. Chronic incubation of SCN5A-1795insD hiPSC-CMs resulted in a robust increase in peak I_Na without impact on the gating properties. This indicates that the effects of mexiletine acted, as expected, by enhancing trafficking. In line with this increase in peak I_Na, chronic mexiletine also increased V_max and hence rescued the loss-of-function phenotype of the mutation. Acute re-administration of mexiletine following chronic incubation did not impact significantly on either peak I_Na or V_max, indicating that the concentration used was sufficiently high to promote trafficking of channels to the membrane, while not too high to cause an acute inhibition of peak I_Na. Chronic incubation of SCN5A-1795insD hiPSC-CMs with mexiletine also resulted in prolongation of APD₂₀, APD₅₀, and APD₉₀. While this may reflect an increase in late I_Na as a result of enhanced trafficking of mutant channels (as observed in our HEK293A cell experiments), it may also be the consequence of mexiletine-mediated effects on other ion channels. In fact, acute mexiletine is known to also affect, at various concentrations, potassium channels such as hERG^32,33 and calcium channels.^34–36 Nevertheless, the fact that acute mexiletine reverted the prolongation of APD₅₀ and APD₉₀ induced by its chronic treatment is in line with a direct inhibiting effect of the drug on late I_Na. Since previous studies have shown a virtually undetectable late I_Na in HEK293 cells transfected with SCN5A-WT³⁷ and in control hiPSC-CMs,^38,39 we did not expect a measurable effect of mexiletine on late I_Na on WT channels or in control hiPSC-CMs and therefore did not investigate this.

The mechanism by which mexiletine acutely inhibits late I_Na is relatively well-established. Na_v1.5 is formed by four domains (DI, DII, DIII, and DIV), each of which consists of six transmembrane segments (S1–S6), where S1–S4 helices include the voltage-sensing domains (VSD), S5 and S6 helices form the central pore domain, and DIII–DIV is the inactivation loop of the channel.⁴⁰ Mexiletine can bind two sites of Na_v1.5 depending on the conformation of the channel: mexiletine-binding site 1, formed by the S6 segments from all four domains of the channel, and mexiletine-binding site 2 located within the fenestration between DIII and DIV.⁴¹ The blocking effect of mexiletine on Na_v1.5 is state-dependent⁴² and concentration-dependent.^43–46 Mexiletine binds preferentially to the inactivated state of Na_v1.5, and consequently, SCN5A mutations that favour the inactivated state can facilitate mexiletine binding to Na_v1.5, thus resulting in a better clinical response to mexiletine treatment.^13,16,47 Previous studies have shown that mexiletine, at low concentration such as 10 μm, inhibits preferentially late I_Na rather than peak I_Na,^43–46 which is important to avoid pro-arrhythmic effects. How mexiletine increases the trafficking of Na_v1.5 channels is not completely clear. It has been suggested that the drug acts as a ‘chemical chaperone’ by promoting correct protein folding and facilitating the exit of mutant ion channels trapped within the ER due to misfolding, allowing them to reach the membrane.^15–17 This same mechanism also has been suggested for chronic lidocaine treatment, another sodium channel blocker that is able to rescue sodium channel expression on the cell membrane.⁴⁸ Moreover, this concept of chemical rescue of impaired trafficking by drugs acting as chaperones has been demonstrated for other channels.^49–51 The molecular chaperone effect of mexiletine on Na_v1.5 has been shown by immunofluorescence experiments^15–17 and more recently by surface biotinylation assays in HEK293 cells.⁵² Interestingly, in the latter study, the authors demonstrated that this trafficking rescuing effect is not universal but mutation-dependent. Moreover, when residues involved in mexiletine binding were mutated, mexiletine no longer impacted Na_v1.5 trafficking, demonstrating that the ‘chaperone’ effect of mexiletine requires its binding to the channel.⁵²

Initially, SCN5A-1795insD patients appeared to be successfully managed by pacemaker therapy, aimed at preventing (nocturnal) bradycardia-dependent QT prolongation. However, ventricular arrhythmias and/or sudden cardiac death have recently been observed in a subset of patients despite pacemaker therapy.¹⁰ Most of these patients were older than 40 years and were found to have hypertension as well as signs of cardiac hypertrophy, indicating potential differential age-dependent arrhythmia mechanisms. One may speculate that at later ages, the loss-of-function consequences of the mutation become more relevant for arrhythmogenesis. Indeed, development of cardiac fibrosis has been demonstrated in aged Scn5a haploinsufficient mice,⁵³ and a role for Na_v1.5 in modulating cardiac structure and function is increasingly recognized.⁵⁴ Hence, restoring the reduced I_Na is important to prevent conduction abnormalities, arrhythmias, and secondary cardiac remodelling. Our present observations that chronic mexiletine at a clinically relevant concentration rescues I_Na in SCN5A-1795insD hiPSC-CMs, without acutely affecting I_Na, demonstrate the first successful pharmacological approach to restore I_Na. However, it should be noted that mexiletine may also promote trafficking and plasma membrane localization of mutant channels and hence inadvertently increase late I_Na and repolarization disturbances, as occasionally observed clinically.⁵⁵ Given the fact that this effect on trafficking occurs only when used chronically, it will be essential to clinically monitor patients with (suspected) overlap syndrome mutations for such potential side effects, e.g. by performing ECG analysis 48 h after initiation of mexiletine treatment. Furthermore, our findings indicate that chronic administration of mexiletine can also increase SCN5A-WT peak I_Na, raising the intriguing possibility that it may also prove efficacious in (inherited) conditions associated with reduced I_Na aside from SCN5A channelopathy, including BrS and arrhythmogenic cardiomyopathy.

In conclusion, here we demonstrate for the first time the therapeutic benefit of a clinically relevant dose of mexiletine in a human cardiomyocyte model of SCN5A overlap syndrome. Furthermore, we identified mexiletine as a novel approach to rescue decreased peak I_Na, which may prove beneficial in other inherited arrhythmia disorders.

Supplementary material

Supplementary material is available at Europace online.

Funding

This study was supported by the Netherlands CardioVascular Research Initiative (CVON-PREDICT2 2018-30 and CVON-eDETECT 2015-12), Fondation Leducq (17CVD02), a ZonMw Priority Medicines (PMRare) grant (113303006), a ZonMw PSIDER grant (10250042110010), and a Novo Nordisk Foundation grant (NNF21CC0073729).

Data availability

All relevant data are within the manuscript and its online supplementary material.

Translational perspective

The sodium channel blocker mexiletine may reduce the enhanced late sodium current secondary to gain-of-function SCN5A mutations and consequently have beneficial effects in long QT syndrome type 3. In addition, high concentrations of mexiletine have been shown to rescue peak sodium current for certain loss-of-function mutations, most likely by restoring trafficking of mutant channels. Here, we demonstrate that a clinically relevant concentration of mexiletine has a dual beneficial effect in a human cardiomyocyte model of SCN5A overlap syndrome. Our findings furthermore identify mexiletine as a therapeutic approach to rescue decreased peak sodium current, with potential benefit for other inherited arrhythmia disorders.