The genetic landscape underlying cardiac arrhythmia disorders has seen remarkable progress in recent years, with a growing understanding of the intricate interplay between ion channels and their role in maintaining normal cardiac rhythm [1]. SCN5A is a critical gene in electrical signaling, encoding the alpha subunit of the voltage-gated sodium channel (Na_V1.5), which facilitates sodium ion influx into cardiomyocytes during depolarization, a key process in generating cardiac action potentials and conduction [2]. Mutations in SCN5A can lead to functional impairments of the Na_V1.5, potentially resulting in aberrant cardiac electrical activity and an elevated risk of arrhythmogenic events [3]. Such genetic alterations are associated with a spectrum of cardiac arrhythmias, including congenital long QT syndrome (LQTS), Brugada syndrome (BrS), cardiac conduction disorder, dilated cardiomyopathy, and multifocal ectopic Purkinje-related premature contractions [3-5]. Notably, SCN5A mutations associated with BrS leads to decreased sodium channel activity due to reduced Na_V1.5 protein levels, malfunctioning channel expression, or alterations in gating characteristics such as delayed activation and premature inactivation [2].

Initially, different SCN5A-related disorders were regarded as distinct clinical conditions, characterized by unique and isolated symptoms resulting from specific biophysical alterations caused by SCN5A mutations on the Na_V1.5. However, numerous studies have revealed cases of individuals carrying SCN5A variants that do not align with these conventional disorders [6,7]. Instead, these individuals exhibited overlap syndromes, which combine aspects of different canonical SCN5A-related arrhythmia syndromes, or they displayed a variable arrhythmic phenotype among individuals with the same mutation [7]. The concept of SCN5A overlap syndromes has evolved since its first description in 1999 when an SCN5A-1795insD mutation was identified [8]. Various SCN5A mutations have since been associated with a wide range of clinical phenotypes, including LQT3, BrS, conduction disorders, sinus nodal dysfunction, atrial standstill, and more [9]. The p.1795insD, p.L1786Q, and p.E1784K mutations in the SCN5A gene are linked to a combined phenotype of LQT3 and BrS, characterized by reduced peak Na⁺ currents and induction of late Na⁺ currents [10].

These electrophysiological alterations contribute to QT interval prolongation and ST elevation on electrocardiogram (ECG), indicative of the overlap in LQT3 and BrS phenotypes. Nevertheless, the underlying mechanisms linking ion channel function and this clinical phenotype are not yet fully understood due to the complexity of overlap syndrome. Electrophysiological studies on mutant SCN5A are required to understand the biophysical alterations of ion channels that affect clinical phenotypes.

Here we report a functional study of a Na_V1.5 channel involving the novel double missense mutations of p.A385T/R504T in SCN5A in a patient showing an abnormal ECG consistent with QRS complex widening, a coved type elevated ST-segment in V2, and QTc prolongation. The p.A385T and p.R504T are situated in the loop connecting transmembrane segments 5 to 6 in domain 1 (S5-S6 in DI) and segments 6 to 1 between domain 1 and 2 (DI-DII linker), respectively (NM_000335.5). In a previous report, Chae et al. [11] classified the p.A385T/R504T mutation as a cause of LQT3, but without electrophysiological investigation. To elucidate the pathophysiological association of the mutation with overlap phenotypes, we investigated electrophysiological analysis of the double mutations, including the individual contribution of p.A385T and p.R504T.

Clinical data

Clinical data from a patient exhibiting an overlap syndrome of BrS and intraventricular conduction delay were used in this study. The patient underwent clinical evaluation, including an ECG, at Seoul St. Mary's Hospital in Seoul, Korea. Informed consent was obtained from the patient.

DNA constructs, mutagenesis, HEK cell culture and transfection

A human embryonic kidney cell line (HEK293 cell) was cultured in Dulbecco's Modification of Eagle's Medium (Gibco Life Technologies) containing 10% fetal bovine serum and 1% antibiotic-antimycotic solution at 37°C with 5% CO₂. A pCMV6-XL4 vector subcloned SCN5A cDNA, double mutants p.A385T(c.1153G>A)/R504T(c.1511G>C), p.A385T, and p.R504T (NM_000335.5) were generated by site-directed mutagenesis. SCN1B plasmid originates from Origene (RC209565) and co-transfected with the same amounts of SCN5A. Lipofectamine 3000 (Thermo-Fisher, L30000001) transfection reagent was used for electrophysiology analysis to transiently introduce 1 μg of constructs into HEK293 cells according to the manufacturer's instructions. Samples were visualized using a confocal microscope (LSM-700; Carl Zeiss) at the Cardiovascular and Metabolic Research Core Support Center at Inje University, South Korea. Co-transfection of a green fluorescent protein (GFP) ensured successful transfection, and only enhanced green fluorescent protein (EGFP)-positive cells were used for patch clamp analysis at least 24 h post-transfection.

Electrophysiology

All whole-cell Na⁺ currents were recorded at room temperature using the standard whole cell patch-clamp technique. The pipette solution contained 20 mM NaCl, 115 mM CsCl, 1 mM MgCl₂, 5 mM HEPES, 5 mM EGTA, 0.4 mM NaGTP, and 4 mM MgATP (pH adjusted to 7.2 with CsOH); the bath solution contained 130 mM NaCl, 4 mM CsCl, 10 mM glucose, 10 mM HEPES, 2 mM CaCl₂, 1 mM MgCl₂ (pH 7.4 with NaOH). The currents were amplified and filtered using an Axopatch 200B amplifier and subsequently digitized at 5 kHz using Digidata 1550B (Molecular Devices) after analog filtering with 20 kHz of the amplifier. In most of our experimental conditions, the general access resistance ranged from 2 to 4 MΩ. To minimize voltage-clamp error, we compensated for the series resistance electronically by 80%. The data were acquired using pClamp 11.1 (Molecular Devices) and analyzed using Clampfit 11.1 (Molecular Devices). Voltage-clamp protocols are provided as insets within each figure.

Western blot

For Western blot analysis, the expression vector pEGFP-N1 was used to subclone WT-SCN5A and A385T/R504T-SCN5A upstream of the coding region of a GFP. Western blot was performed using whole cell lysates from HEK293 cells that heterologously expressed wild type (WT), and A385T/R504T with β1-subunit SCN1B. The cells were lysated with a lysis buffer comprising 150 mM NaCl, 50 mM Tris-Cl (pH 7.4), 1 mM EDTA, and 1% Triton-X 100, and the protein concentration was determined using a BCA Protein Assay Kit (Thermo-Fisher, 23227). Next, 5X SDS-PAGE loading buffer (Biosesang, S2002) was added to the lysates, which were then separated on 8% SDS-PAGE gels and transferred to PVDF membranes. The membranes were incubated with primary antibodies, including a rabbit anti-GFP (1:1,000, Invitrogen, A-11122), and mouse anti-Na-K ATPase (1:2,000, Sigma, 05-369) overnight at 4°C, followed by a secondary antibody (1:10,000) for 1 h at room temperature. Chemiluminescence was used to detect the membrane signals, and protein expression levels were normalized to anti Na-K ATPase and quantified using ImageJ (National Institutes of Health).

Statistical analysis

The results are expressed as mean ± standard error of mean, and statistical comparisons were made by Student t-test using GraphPad prism 8, with p < 0.05 indicating significance. Multi-exponential functions were fitted to the data using the nonlinear least-squares method with Origin software.

SCN5A double mutation case in a patient with overlap syndrome of BrS and intraventricular conduction delay

Here, we reported a case showing overlap syndrome with double mutation in SCN5A. After resuscitation from sudden cardiac arrest caused by ventricular fibrillation, a patient showed coved-type elevated ST-segment in leads V2 (2 mm) which meets the ECG criteria for BrS (Fig. 1A, heart rate 132 bpm, QRS duration 100 ms, QT/QTc 300/444 ms), along with the widening of QRS duration to 121 ms (normal value < 100 ms) in ECG examination (Fig. 1B, heart rate 128 bpm, QRS duration 121 ms, QT/QTc 246/359 ms).

Figure 1. Twelve-lead electrocardiogram (ECG) of a patient with overlap syndrome with a double missense mutation in SCN5A. (A) Representative ECG showed a coved type ST-segment elevation in leads V2 (2 mm) and T-wave abnormality. (B) Representative ECG showed intraventricular conduction delay with prolonged QRS widening to 121 ms (normal value < 100 ms).

Indistinguishable I_Na properties of A385T/R504T with SCN5A alone

To find the electrophysiological properties caused by this double mutation, whole cell patch clamp experiment was conducted and compared between WT and double mutant SCN5A (p.A385T/R504T) overexpressed HEK293 cells. Step-like depolarizing pulses were applied to activate I_Na (Fig. 2B, inset). As shown by the average current-voltage (I-V) relations, the peak amplitude of I_Na density was not different between WT and A385T/R504T (WT: maximum I_Na −542.21 ± 130.21 pA/pF, n = 11; A385T/R504T: maximum I_Na −534.46 ± 90.23 pA/pF, n = 12) (Fig. 2C). The I-V curves were converted to the conductance-voltage (G-V) curve in order to analyze the voltage dependence of activation of Na_V1.5 (Fig. 2D), which was not different between WT and p.A385T/R504T (Table 1).

Table 1 . Voltage- and time-dependence of activation and inactivation for WT and mutant channels.

	WT	p.A385T	p.R504T	p.A385T/R504T	WT + β1	p.A385T + β1	p.R504T + β1	p.A385T/R504T + β1
Activation	(n = 12)	(n = 12)	(n = 11)	(n = 12)	(n = 19)	(n = 16)	(n = 12)	(n = 7)
V1/2 (mV)	–37.48	–35.36	–37.33	–38.44	–34.09	–32.77	–34.03	–34.45
k (mV)	7.15	7.14	7.15	7.14	7.76	7.73	7.54	7.68
Inactivation	(n = 16)	(n = 15)	(n = 19)	(n = 13)	(n = 16)	(n = 9)	(n = 9)	(n = 11)
V1/2 (mV)	–98.95	–91.48	–93.57	–93.22	–86.96	–86.22	–86.81	–87.37
k (mV)	6.86	6.68	6.69	6.85	5.7491	5.8	6.08	5.78

WT, wild type.

Figure 2. p.A385T/R504T does not affect Na+ current (INa) density. (A) Schematic drawing of the cardiac voltage-gated Na⁺ channel α-subunit (Na_V1.5) showing position of the A385T/R504T in the loop connecting transmembrane segments 5 and 6 in domain 1 (S5-S6 in DI) and segments 6 and 1 between domain 1 and 2 (DI-DII linker). (B) Representative whole-cell Na⁺ current recordings of WT and p.A385T/R504T. (C) Average Na⁺ current-voltage (I-V) relation for WT and p.A385T/R504T channels. (D) Voltage-dependence of activation for WT and p.A385T/R504T channels by normalizing peak conductance against the membrane voltage. WT, wild type.

To assess the voltage dependency of inactivation, a two-pulse protocol was employed; 500 ms conditioning pre-pulses ranging from –140 to –15 mV to induce steady-state inactivation, followed by a 50 ms test pulse (Fig. 3A, inset). The voltage dependent of inactivation in p.A385T/R504T was shifted to the rightward direction by 5.73 mV (Table 1), implying increased ‘window current’ (Fig. 3B).

Figure 3. Characterization of Na+ channel property in the steady-state activation and inactivation. (A) Representative traces of voltage dependence of inactivation. The protocol is shown in left panel. (B) Voltage dependence of activation and inactivation and window current. The window current refers to the expanded segment of the overlapping region between the activation and inactivation of both the WT and p.A385T/R504T channels. (C) Time constant of recovery from inactivation to A385T/R504T was elicited with a double pulse protocol. (D) The relationship between P2/P1 and the inter-pulse interval was graphed, with curves representing double-exponential functions. The fast and slow time constants were estimated from the plotted double-exponential functions. WT, wild type; N.S, not significant.

To measure the recovery from inactivation of Na_V1.5, another type of two-pulse protocol with incremental increase of the pulse interval was applied (Fig. 3C). This protocol involved the conditioning pre-pulse (P1; –20 mV, 100 ms) to inactivate Na_V1.5, followed by variable hyperpolarized interval (recovery period; –120 mV, 1–1,000 ms) and the second test pulse (P2; –20 mV, 20 ms) (Fig. 3D, inset). The peak amplitudes in response to the P2 were normalized to the peak amplitudes at P1, and the resulting curve was plotted against the inter-pulse recovery period (Fig. 3D). When the curve was fit to a double exponential function, the fast- and slow-time constants of recovery from inactivation showed no difference between WT and p.A385T/R504T (Fig. 3D, right panel bar graphs).

Disclosure of loss-of-function phenotype with SCN1B co-expression

When the β1 subunit of Na_V1.5 (SCN1B) was co-expressed in HEK293 cell, the density of I_Na was markedly reduced in p.A385T/R504T (WT + β1: I_peak −451.31 ± 77.49 pA/pF (n = 12); p.A385T/R504T + β1: I_peak −234.85 ± 28.44 pA/pF (n = 28), Fig. 4A–C), without affecting the voltage dependence of activation (Fig. 4D, Table 1). The immunoblotting analysis showed that the total protein expression of SCN5A was not reduced regardless of the SCN1B expression. However, a significant decrease in surface protein expression was observed upon co-expression with SCN1B (Fig. 4E, p < 0.05). Also, the voltage dependence of inactivation was not different between WT and p.A385T/R504T when co-expressed with SCN1B (Fig. 5A, B, Table 1). Interestingly, the fast component of recovery from inactivation was slower in p.A385T/R504T with SCN1B (Fig. 5C, D).

Figure 4. Reduced p.A385T/R504T INa co-transfected with β1-subunit. (A) Representative whole-cell Na⁺ current traces of WT and p.A385T/R504T in the presence of β1. (B) I-V relation of WT + β and p.A485T/R504T + β. (C) Comparison of peak I_Na density (p = 0.0027). p-value less than 0.05 was considered to be statistically significant, *p < 0.05; **p < 0.01. (D) Voltage dependence of activation for WT and p.A385T/R504T in the presence of β1-subunit measured form 12-20 cells. (E) Representative Western blotting image of Na_V1.5 in transfected HEK293 cells with and without β1-subunit. WT, wild type; I-V, current-voltage; DB, double mutant; GFP, green fluorescent protein.

Figure 5. Characterization of Na+ channel property in the steady-state activation and inactivation in co-expressed β1-subunit. (A) Representative traces of voltage dependence of inactivation in the presence of β1-subunit. (B) Voltage dependence of activation and inactivation, and window current. (C) Representative traces of time course of recovery from inactivation of WT and p.A385T/R504T with β1-subunit. (D) The P2/P1 ratio was plotted against the inter-pulse interval, and the resulting curves were fitted with double-exponential functions. The time constants for the fast (p = 0.0250) and slow components were estimated from these plotted functions. p-value less than 0.05 was considered to be statistically significant, *p < 0.05. WT, wild type; N.S, not significant.

Different contribution of A385T and R504T in the presence of SCN1B

To elucidate the contribution of individual missense mutation of p.A385T/R504T, site-directed mutants of p.A385T and p.R504T were exclusively induced in SCN5A and co-expressed with SCN1B. The p.R504T + β1, but not p.A385T + β1, resulted in considerably decreased I_Na density when compared WT (WT + β1: I_peak −451.31 ± 77.49 pA/pF, n = 12; A385T + β1: I_peak −430.08 ± 68.76 pA/pF, n = 19; A385T + β1: I_peak −239.73 ± 43.31 pA/pF, n = 13) (Fig. 6A, B). The voltage dependence of activation and inactivation remained unchanged in each mutant (Fig. 6C, Table 1). Interestingly, the recovery from inactivation was significantly slower in p.A385T + β1 while not in p.R504T + β1 (fast time constant; p = 0.0001, slow time constant; p = 0.0046, Fig. 6D).

Figure 6. Whole-cell current recordings of p.A385T and p.R504T. Na⁺ channels were expressed by transfection in HEK293 cells in the presence of β1-subunit. (A) I-V relation of WT + β, p.A385T + β, and p.R504T + β. (B) Comparison of peak I_Na density. Compared to the WT group, p.R504T + β significantly reduced current density (p = 0.0167). *p < 0.05; **p < 0.01; ***p < 0.001. (C) Voltage dependence of activation and inactivation, and window current. (D) Time course of recovery from inactivation of WT, p.A385T, and p.R504T with β1-subunit. WT, wild type; N.S, not significant.

Here we investigated the electrophysiological characteristics of the SCN5A mutant p.A384T/R504T in an overlap syndrome patient presenting QRS widening and spontaneous ST-segment elevation. The density of I_Na was unchanged with p.A385T/R504T alone, but a rightward shift in voltage-dependent of inactivation and faster recovery from inactivation indicated a gain-of-function, contrary to the BrS phenotype. Co-expression with β1 and p.A385T/R504T resulted in reduced I_Na and slowed recovery, consistent with BrS. Total SCN5A protein expression remained constant in p.A385T/R504T, but membrane protein decreased when co-expressed with β1. Both p.A385T and p.R504T showed characteristic β1-dependent biophysical loss of function, with p.R504T reducing I_Na, and p.A385T displaying slowed recovery.

The genes SCN1B-SCN4B encode the five voltage-gated sodium channel (VGSC) β subunits, with SCN1B encoding the β1-subunit and its splice variant, β1B, and SCN2B-SCN4B encoding the β2, β3, and β4 subunits, respectively [12]. The expression of VGSC β-subunits varies among tissues and cell types, including the heart, where SCN1B is abundantly expressing β1-subunit [13,14]. While the precise function of the interaction between SCN5A and SCN1B is unclear, it is established that modulating SCN5A function through SCN1B is important [12]. Therefore, we identified the p.A385T/R504T mutant disturb interaction with SCN1B. Surprisingly, co-expression with β1-subunit reduced peak I_Na density in mutant to 48%, compared to WT (Fig. 4C). VGSC β-subunits collaborate with α-subunits to facilitate the transportation of channels to the cell membrane and modulate α-subunit activity [15,16]. Our study revealed consistent overall protein expression of SCN5A regardless of mutation without SCN1B expression. Co-expression with SCN1B, however, led to a significant decrease in surface protein expression, indicating likely interference of the β subunit with mutant protein trafficking. In light of diminished surface expression levels regulated by SCN1B, potential influences on other ion channels involved in channelosome formation were suggested [17]. For instance, Clatot et al. [18] recently demonstrated SCN5A and KCND3 inter-regulation, revealing that SCN5A variants in BrS reduce I_Na and increase transient outward K⁺ current (I_to). KCND3 variants associated with spinocerebellar ataxia exhibit variable effects on I_Na based on I_to function. Interaction studies highlight the intricate modulation of SCN5A and KCND3 channels, emphasizing the complex interplay in cardiac and neuronal syndromes [18]. Further investigations will be required to ascertain whether SCN5A mutations exert influences on other ion channels.

VGSC α and β subunits interact through two mechanisms: β1 and β3 noncovalently interact with α subunits via N and C termini, while β2 and β4 form covalent interactions with α subunits through a single N-terminal cysteine [12]. In a previous study conducted by Wan et al. [19], co-transfection of SCN5A double missense mutation (p.R1232W/T1620M) with the β1-subunit resulted in a significant reduction of I_Na density, compared to the transfection of p.R1232W/T1620M alone. These results suggest that the p.T1620M in the extracellular S3-S4 linker in domain IV may affect its interaction with the β1-subunit, resulting in reduced stability of the α-β-subunit complex. Moreover, our findings emphasized a notable reduction in current in the presence of the p.R504T mutation when co-transfected with β1, underscoring the significance of the β1 interaction, especially with p.R504T compared to p.A385T. However, the lack of identification of specific binding sites in the α-β complex was acknowledged as a limitation in our experiments.

Slower recovery from inactivation is one of the mechanisms underlying BrS [20]. Mutations in SCN5A genes can result in delayed recovery from inactivation, which can lead to reduced availability of sodium channels during the cardiac action potential, causing a reduction in I_Na and ultimately contributing to arrhythmias such as BrS and cardiac conduction disorder [21]. In this study, similar to the β1-subunit-dependent regulation seen in the I_Na density reduction results (Fig. 4), the recovery of inactivation of p.A385T/R504T was slower when expressed with the β1-subunit (Fig. 5D). This suggests that the presence of β1 interaction is necessary for the reduced I_Na observed in a BrS patient with the novel double missense mutation (p.A385T/R504T) of SCN5A. When comparing recovery from inactivation for p.A385T and p.R504T, p.A385T + β1 exhibited significantly slower kinetics than WT + β1 (Fig. 6D). This result contrasts with the notable decrease in I_Na density observed in p.R504T + β1 (Fig. 6A). p.A385T exhibits loss-of-function characteristics in the rate of recovery from inactivation, while p.R504T displays loss-of-function properties in regulating I_Na density. The study's findings point to the importance of the β-subunit in modulating the effects of the p.A385T/R504T mutation on sodium channel function. These insights into the impact of specific SCN5A mutations and their interactions with auxiliary subunits could provide a deeper understanding of the pathophysiology of BrS. Furthermore, to understand the impact of the biophysical properties of SCN5A mutations on overlapping phenotypes, it would be beneficial to construct a computational cardiomyocyte-based 3D heart model for in silico simulation in further studies.

In conclusion, this study investigated the electrophysiological properties of novel SCN5A missense mutations (p.A385T/R504T). The interaction between the α-subunit SCN5A and β1-subunit SCN1B has been identified as causing loss-of-function in mutations, underscoring the potential for overlapping clinical phenotypes including BrS and CCD with LQTs. This finding also highlights the modulatory role of SCN1B on SCN5A mutations. Investigating the electrophysiological properties of these mutations could provide valuable insights into the molecular basis of and contribute to the development of more effective diagnostic and therapeutic strategies.

None.

This work was supported by the National Research Foundation of Korea (NRF) funded by the Korea government (MSIT) [NRF-2018R1A5A2025964 and NRF-2021R1A2C2007 to S.J.K, NRF-2022R1A2C2009067 to H.-J.P, RS-2023-00213304 to S.W.C] and by the Dongguk University Research Program of 2021 to S.W.C.

The authors declare no conflicts of interest.