The SCN5A gene encodes the alpha subunit of the main cardiac sodium channel Na_v1.5, which is known to be responsible for maintaining the normal function of inward sodium current (INa). INa current is the main component in fast depolarization phase after which the excitation–contraction coupling cascade and proper conduction of the electrical impulse is subsequently initiated within the heart (Aronsen et al., 2013).

Genetic variants of SCN5A are involved in a number of inherited cardiac channelopathies including Brugada syndrome (BrS), long QT syndrome (LQT3), and cardiac conduction system dysfunction. Meanwhile, SCN5A variants are also found to be correlated with myocardial contractile dysfunction, dilated cardiomyopathy, and heart failure. According to one of the most complete open database of on-line disease-related genetic variations “ClinVar” (ClinVar, 2018), more than 700 SCN5A variation locations are shown to be associated with cardiac disorders (Table 1) and about 90% variants account for non-synonymous variants, while the rest are due to deletion and duplication. With the help of patch clamp technique and transgenic animal technology, studies have clarified the major pathogenic mechanisms underlying sodium channelopathies due to SCN5A variants.

Recent research results indicated that the function and regulation of Na_v1.5 was more complicated than traditionally assumed. For example, Na_v1.5 has been found to be regulated by more than 20 interacting proteins in distinct membrane compartments (Shy et al., 2013). The dysfunction of Na_v1.5 may not only be a cause but also a consequence in different pathophysiological procedures and various cardiac disorders. Undoubtedly the dysfunction of Na_v1.5 contributes to arrhythmogenesis during pathophysiological conditions. However, sodium channel remodeling manifested with alterations in Na_v1.5 clustering was also discovered within distinct cardiomyocyte microdomains in some pathophysiological procedures such as heart failure (Rivaud et al., 2017). We previously found that the SCN5A variant A1180V could directly lead to cardiac structural impairment without involvement of long-time arrhythmias (Shen et al., 2013). Similarly, the environmental and genetic factors may also play additional roles in pathogenesis of disorders linked to SCN5A variants.

In this review, we will summarize SCN5A variants-related arrhythmia syndromes, structural cardiac disorders and the underlying potential mechanisms. Then a brief description will also be given to introduce the current knowledge regarding the therapy strategies.

The SCN5A gene, located in chromosome 3p21 with 28 exons, is a member of the human voltage-gated sodium channel gene family and encodes alpha subunit of the main cardiac sodium channel Na_v1.5. It was firstly identified by Georgè in 1995 with the help of fluorescence in situ hybridization (George et al., 1995). Na_v1.5 is a large transmembrane protein (227 KDa) with four internally homologous domains (DI-DIV), each containing six transmembrane spanning segments (S1-S6) (Figure 1). S4 segment works as a critical voltage-sensor because of its ample positive charge residues. It is triggered to undergo transmembrane movement when the cell membrane depolarizes, allowing for sodium ion influx to generate sodium current. The P-loops between S5 and S6 segments constitute the central pore forming region, which determine the ion selectivity of the channel. The four domains are interconnected by intracellular peptide chains, with N-terminal and C-end both located in the intracellular side (Catterall, 2014; Chen-Izu et al., 2015). Previous studies suggested the SCN5A gene was mainly expressed in cardiomyocytes, however, recent researches found this gene also expressed and played fundamental roles in other tissues such as brain, gastrointestinal (GI) tract, and cancer tissues (Black and Waxman, 2013; Verstraelen et al., 2015).

There are two major SCN5A isoforms: “adult” isoform (SCN5A-003, NM_000335) and “neonatal” (or fetal) isoform (SCN5A-001, NM_001099404). The “adult” SCN5A isoform is abundantly expressed in adult human heart while the “neonatal” isoform is abundantly expressed in neonatal heart. The “adult” isoform differs from the “neonatal” isoform in exon 6 within a voltage-sensor domain (D1/S3-S4) (Murphy et al., 2012). In a mutually exclusive splicing manner, the “neonatal” isoform exhibited slower kinetic of activation and inactivation and a lower depolarized threshold of activation compared to the “adult” isoform (Onkal et al., 2008). The latest study found that both fetal (exon 6a) and adult (exon 6) isoforms of SCN5A were expressed after short-term culture of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) with one of SCN5A variants I230T. Interestingly, prolonged culture time increased the expression of adult sodium channel isoform, which was paralleled by a decrease in INa, representing a severe clinical phenotype in patients with recessive cardiac conduction disease (Veerman et al., 2017). This result partly explained the electrophysiological immaturity phenomenon of hiPSC-CMs and refined our understanding on pathophysiological mechanism of a recessive form of SCN5A-related conduction disorder. Apart from these two isoforms, SCN5A-014 (NM_198056) is also expressed in human heart and should be mentioned in some detail. SCN5A-014 includes an additional glutamine at position 1077 (1077Q) at the exon 17–18 splice boundary, resulting in smaller INa current and exhibiting functional differences with the previous standard isoforms (Makielski et al., 2003).

Sodium current participates in phase 0 and 2 of action potential (AP), which is an important determinant on the production of heart rhythms, normal conduction, and maintenance of excitement. Aberrant sodium channels resulting from SCN5A variants would potentially cause disorganization of the cardiac electrophysiological system, produce various arrhythmias and result in structural heart diseases. Inward and outward transmembrane currents are in dynamic balance physiologically, while variants in SCN5A could induce two states in the sodium channel: “gain-of-function state” or “loss-of-function state.” The former state appears as an increase in sodium influx (sodium current) and a delay in inactivation, while the latter one is manifested by a decrease in sodium influx and an acceleration of inactivation. Different variants-related phenotypes correspond to different state.

Dilated cardiomyopathy (DCM) is a myocardial disease characterized by left ventricular systolic dysfunction with dilatation of left ventricle or double ventricles. It is the most common type of primary cardiomyopathy and the most common reason for cardiac transplantation in adults and children (Sanbe, 2013). It was reported that 20–50% DCM patients had a family history, suggesting that genetic factors played an important role in their pathogenesis. Currently, more than 40 causative genes have been found in DCM, including SCN5A (Hershberger et al., 2013). In 2004, Mc Nair firstly identified a heterozygous G-to-A variant of SCN5A, which was co-segregated with DCM phenotype (McNair et al., 2004). Subsequently, SCN5A missense variants (T220I, R814W, D1595H) and truncation variants (2550-2551insTG) were successively discovered in DCM patients (Olson et al., 2005). Most variants are located in the voltage sensor domain (VSD) especially in S3 and S4 segment. As the electrophysiological properties of these variants are not fully clarified, the pathogenesis of DCM due to SCN5A variants remains inconclusive.

Systolic function of the heart was previously thought to be mainly associated with calcium channels. SCN5A variants were traditionally recognized as the cause of arrythmias in DCM patients, but did not directly affect the cardiac systolic function. A primary disruption of an ion channel through gene variants leading to DCM was explained that in one way, the ion channel disturbance caused dysfunction of cytoskeletal protein binding partners and resulted in a DCM phenotype. In another way, the electrical dysfunction caused by variants defects led to mechanical instability, and ultimately led to myocardial dilation (Towbin and Lorts, 2011). Furthermore, gain-of-function SCN5A variants were partly responsible for hyperexcitability of the fascicular-Purkinje system. The incomplete repolarization in Purkinje cells brought premature ventricular action potentials and developed DCM finally (Laurent et al., 2012).

Through the electrophysiological analysis of A1180V, a novel SCN5A variant that was discovered by our team, we indicated that variant channel expressed a rate-dependent Na⁺ current reduction and a moderate increase in INaL, implying that A1180V caused DCM by disturbing cellular Na⁺ homeostasis. During the follow-up period of the affected pedigree, we found A1180V carriers exhibited a deterioration of cardiac function and progressed to DCM or atrioventricular block (AVB). These results suggested that A1180V was a risk factor for familiar DCM with preceding atrioventricular block (Ge et al., 2008; Shen et al., 2013). The variant might directly damage myocardium by affecting the intracellular sodium homeostasis, which in turn lead to cell membrane Na⁺/Ca²⁺- and Na⁺/H⁺-exchanger disorders and imbalance of intracellular calcium homeostasis. Apart from A1180V, we further explored three novel non-synonymous SCN5A variants in idiopathic DCM patients, including c.674G>A, c.677C>T, and c.4340T>A (Shen et al., 2017). Consistent with previous results, these newly defined idiopathic DCM related variants were mainly located in the S4 segment of domain I (DI-S4). By applying patch clamp technology, we found that R225Q (c.674G>A) predisposed electrical disorders by reducing peak sodium current density. Some Na_v1.5 variants such as R222Q; R225W; R225P; R814W; and R219H were associated with an atypical phenotype combining several cardiac arrhythmias and DCM (McNair et al., 2011; Mann et al., 2012). Chahine's group investigated mutant channels and found the gating pore current generated by these variants with a cation leak through the typically non-conductive VSD. The presence of a proton-based leak current linked Na_v1.5 VSD variants with human cardiac arrhythmias and dilatation of cardiac chambers (Gosselin-Badaroudine et al., 2012; Moreau et al., 2015a,b). As many of the molecular mechanisms between the dilated phenotype and DCM-related SCN5A variants have yet to be elucidated, future novel insights will be available depending on transgenic mice studies and large-scale clinical long-term follow-up studies.

Sudden infant death syndrome (SIDS) denominates cases of infant death, which mostly occurs during sleep without any preceding symptoms. It is the leading cause of post neonatal mortality (Matthews and Macdorman, 2013). Although many pathophysiological studies have been proposed, the specific pathogenic mechanisms remain unclear. SIDS is presumed to be co-regulated by multiple factors including neurotransmission, energy metabolism and genetic disorders. Ion channel genes such as KCNQ1, KCNH2, SCN5A, RYR2, and their molecular modifiers such as GPD1L and β-subunits account for ~10–15% of all SIDS cases, in which SCN5A variants account for approximately half of the channelopathic SIDS cases (Tan et al., 2010; Tester et al., 2018). Variants in SCN5A led to severe trafficking defects and a severely dysfunctional Na_v1.5 channel in a structurally normal heart. The decreased current density was likely to be a major contributing factor to the lethal arrhythmia in SIDS victim (Gando et al., 2017). Given the potential risk of inherited cardiac conditions caused by ultra-rare gene variants, current experts has made consensus statement on post-mortem genetic testing (molecular autopsy) and clinical investigation of surviving blood relatives. However, under pressure from public moral principle and different physical environments, roll-out of this approach seems a long way away.

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited cardiomyopathy characterized by myocardial atrophy with fibro-fatty replacement, eventually resulting in wall thinning and cardiac dilatation. The inflow tract, outflow tract and apex of the right ventricle constitute the so-called “triangle of dysplasia” (Oomen et al., 2018). ARVC manifests with myocardial inflammation, right ventricular aneurysms, and ventricular tachyarrhythmias, which represents one of the leading causes of sudden death in the young and in athletes (Basso et al., 2018). The estimated prevalence of ARVC is 1 per 5000 in the general population. Approximately >60% of patients have a pathogenic variant (Bhonsale et al., 2015; Calkins et al., 2017). Desmosomes variants was traditionally considered as the genetic basis of ARVC (Gandjbakhch et al., 2013). It was known that desmosomal protein desmoglein-2 and Na_v1.5 interacted with each other (Rizzo et al., 2012). Decreased Na⁺ current density could slow down conduction and increase necrosis and fibrosis. An artificial loss of desmoplakin (DSP) expression induced an abnormal distribution of connexin43 (Cx43) and Na_v1.5. As a result, the sodium current decreased and conduction velocity slowed down, which impaired the mechanical and electrical coupling, these factors also contributed to the pathogenesis of ARVC (Zhang et al., 2013). Beside these, Te Riele performed whole-exome sequencing in six ARVD/C patients without desmosomal variants and found a missense variant (R1898H) in SCN5A. Later, they expanded the study population and found almost 2% of ARVD/C patients had rare SCN5A variants (Te et al., 2017). These findings demonstrated that changes in sodium channel might interact with other molecules or form a tighter network of interactions and served as potential causal gene variants of ARVC.

Multiple SCN5A variants caused complex clinical phenotypes, which suggested that different arrhythmias might have the same genetic origin. The clinical and genetic overlap between these arrhythmias are considered overlap syndrome. 1795insD was the first SCN5A variant related to overlap syndrome, carriers not only presented with LQT3, but also with sinus bradycardia, PCCD, and BrS (Postema et al., 2009). Subsequently, similar findings were reported for other SCN5A variants such as DelK1500, E1784K, Q779X (Grant et al., 2002; Sumitomo, 2014; Aoki et al., 2017). Various SCN5A-related arrhythmia syndromes were not separated but closely linked clinical entities with numerous biophysical overlaps. Moreover, there was a strong genotype-phenotype correlation between arrhythmia syndromes and SCN5A variants, although some phenotypes presented during all decades of life, while others developed with increasing age. By observation in various heterologous expression systems, the biophysical characteristics underlying multiple phenotypes were found to delay fast inactivation, reduce peak sodium current density and increase late component of Ina (Remme and Wilde, 2008; Veltmann et al., 2016). The complex phenotype could be explained by the changes in membrane potential and current. Reduced penetrance and variable disease expression might limit the clinical and genetic diagnosis in SCN5A-related overlap syndromes. Even owning the same variant in SCN5A, reasons of the presence of different or even opposite phenotype are still unclear. The fact that the same SCN5A variant might have different phenotypes suggests the phenotype may have individual or family differences, or that certain regulatory genes or environmental factors may also affect the presence of phenotypes. In addition, clinical elements including age, gender, or medications may also modify disease expressivity. Large cohorts studies are thus needed to gain more insight into exact causal relationship between sodium channel function and various phenotypes.

For many years, researchers have been working on reversing of the pathogenic phenotypes caused by SCN5A variants. Although implantation of ICDs can prevent potentially fatal arrhythmias in patients, the main stream of therapy is the use of medications to prevent arrhythmias. Classical sodium channel blockers are classified as Class I anti-arrhythmic agents and furtherly subdivide into classes IA, IB and IC according to their effects on changes of cardiac AP length (Milne et al., 1984). The blockade of sodium channels can reduce the excitability of cardiomyocytes and stop reentrant wavefronts. However, the same mechanism may exert opposing effect at the same time (van Hoeijen et al., 2014). Toxicities include proarrhythmia are common. Whether the sodium channel blocker mexiletine can be used in clinic to shorten the QT interval without side effects among SCN5A variants carriers has long been controversial. As early as 2004, it was found to rescue the membrane expression of the voltage-gated sodium channel and assist in transporting proteins from the sarcoplasmic reticulum to plasma membrane (Valdivia et al., 2004). Later research queried mexiletine might facilitate trafficking of mutant proteins and exacerbating QT prolongation (Ruan et al., 2010). However, recent study demonstrated mexiletine could bring a major reduction of life-threatening arrhythmic events in LQT3 patients in a retrospective cohort study (Mazzanti et al., 2016). Even with so many conflicting studies, selective inhibition of INaL constitutes a promising pharmacological treatment for cardiac channelopathies. Many pharmacists are dedicated to finding a potent, selective inhibitor of INaL. For example, the INaL inhibitor GS967 was discovered to decrease repolarization abnormalities and had anti-arrhythmic effect without damnification on cardiac conduction (Portero et al., 2017). Other traditional drugs such as flecainide, lidocaine, or β-blockers have also been reported to have some clinical benefits in certain population or specific variants (Wilde et al., 2016; Anderson et al., 2017; Chorin et al., 2018). In general, the especially important issue in pharmacological treatment lays in rebranding of old medications and development of novel agents on the basis of ascertainment in the molecular derangements resulting from genetic variants (Varian and Tang, 2017).

In addition to pharmacological treatments and ICD implantation, some other new approaches are gradually emerging. Non-sense variant is one type of SCN5A pathogenic variants. Non-sense mutation readthrough is a new kind of gene-specific treatment to reduce or avoid deleterious consequences of non-sense variants which could make sure the ribosomes ignore a premature stop codon and produce a full-length protein (Lee and Dougherty, 2012). Some researchers applied readthrough-enhancing methods such as aminoglycosides, suppressor tRNAs or small interfering RNAs (siRNAs) to suppress variants (Roy et al., 2016; Baradaran-Heravi et al., 2017). The sodium currents were restored, however, the restored channels increased the risk of arrhythmia (Teng et al., 2017). Pre-implantation genetic diagnosis (PGD) is a diagnostic procedure by analyzing genotype of the preimplantation embryo, enabling selective transfer of unaffected embryos to the uterus. This procedure improves embryo selection and avoids pathogenesis-related genetic variants. A meta-analysis including four randomized controlled trials (RCTs) and seven cohort studies showed that comprehensive chromosome screening (CCS)-based PGD has better outcomes for women compared to classical morphological criteria (Chen et al., 2015). Although the prospect is raising difficult ethical considerations, PGD is one of the strategies for prevention of gene variants-related diseases somehow (Resnik, 2014; Vermeesch et al., 2016). It is known that gene works through mRNA level. Some studies attempted to regulate the expression of SCN5A mRNA through exogenous ingestion of the messenger RNA (mRNA) stabilizing protein ELAVL1/HuR and to verify whether it was beneficial to control arrhythmia. Researchers found injection of viral particles carrying HuR increased SCN5A expression and improved AP upstroke and conduction velocity, which reduced reentrant arrhythmia. Increasing mRNA stability to rescue decreased SCN5A expression may represent another new paradigm in antiarrhythmic therapy (Zhou et al., 2018). Abnormal bioelectric phenotype was discovered to be corrected in cystic fibrosis caused by variants of the cystic fibrosis transmembrane conductance regulator gene (CFTR) by administering replication-deficient, recombinant adenovirus vector containing a normal copy of the CFTR cDNA (AdCFTR) in a clinical trial (Hay et al., 1995). But arrhythmia genetic or epigenetic therapy investigations have been limited to short-term demonstrations of efficacy in pre-clinical pilot studies. For many arrhythmia applications, adequate delivery to the target tissue and proarrhythmic toxicity as well as safety risks still remain. With long-term efficacy and safety studies, integrated approaches of normalizing cardiac structure and function may represent the next generation of potential therapeutic strategies for cardiac disorders with gene variants.

Cardiac sodium channel plays a central role in cardiomyocyte excitability and proper conduction of cardiac electrical impulses. The function and mechanisms of sodium channel are complicate and remain a lot of controversies. Even within families, the observed phenotypes carrying the same SCN5A variant are highly diverse. Besides, environmental and epigenetic alterations also determine variable disease severity. Continuous exploration and novel insight of this issue will contribute to clarify detailed mechanisms and ultimately enable improved diagnosis, risk stratification, and development of more effective treatment strategies.

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. In particular, WL, LY, and CS: drafting of the manuscript; KH, JG, and AS: revising the manuscript for important intellectual content critically and for final approval of the manuscript.