Cardiac sodium channelopathy associated with SCN5A mutations: electrophysiological, molecular and genetic aspects

Sodium channel structure and function

The voltage-dependent cardiac sodium channel consists of a transmembrane pore-forming α-subunit protein of about 220 kDa associated with a small (30–40 kDa) ancillary modulatory β-subunit and several regulatory proteins (the stoichiometry of α- and β-subunits in heart is as yet incompletely known). The α-subunit protein Na_V1.5 (encoded by the SCN5A gene located on chromosome 3p21) is made up of a cytoplasmic N terminus, four internally homologous domains (DI–DIV, each consisting of six transmembrane α-helical segments S1–S6) interconnected by cytoplasmic linkers, and a cytoplasmic C terminal domain (for details, see legend to Fig. 1). Channel activation (opening of the pore) is triggered by the outward movement of the positively charged S4 segments allowing for sodium ion influx to commence. Simultaneously, the processes of fast and slow inactivation are initiated leading to closure of the channel pore. Fast inactivation of the sodium current involves an inactivation gate formed by a cluster of three hydrophobic amino acids (isoleucine, phenylalanine and methionine, or IFM-motif; see Fig. 1) in the intracellular DIII–DIV linker, together with two docking sites located on the intracellular linkers between S4 and S5 of DIII and DIV (West et al. 1992; Kass, 2006). In addition, the carboxy (COOH-) terminal domain is also involved in the inactivation process and acts through stabilization of the closed gate, thus minimizing channel re-opening (Motoike et al. 2004). During prolonged periods of depolarization, the sodium channel enters the state of slow inactivation, which requires much longer recovery times than fast inactivation. Slow inactivation is an important determinant of sodium channel availability, and most likely involves conformational changes in the S5–S6 P-loops, although the DIII S4–S5 linker may also be of importance (Kass, 2006; Casini et al. 2007). Activation and inactivation properties of sodium channels are tightly regulated during physiological conditions but may be altered in the setting of genetic and acquired sodium channel dysfunction. As a consequence, sodium channel availability and peak sodium current may be decreased, or the channel is not properly inactivated, resulting in a persistent (sustained), non-inactivating sodium current during the action potential plateau (George, 2005). These alterations in sodium channel function may have profound consequences for cardiac electrophysiology and arrhythmogenesis.

Expression and distribution of sodium channels in the heart

Cardiac sodium channels show inhomogeneous expression within the cardiac conduction system and across the ventricular wall. Na_V1.5 protein expression is low to absent in the sinoatrial and atrioventricular nodes, but abundant in the His bundle, bundle branches and Purkinje fibres (Yoo et al. 2006; Remme et al. 2009b). Furthermore, a transmural gradient is observed in left and right ventricle, with lower Na_V1.5 labelling intensity and decreased functional sodium channel availability in the subepicardium as compared to the subendocardium (Ashamalla et al. 2001; Cordeiro et al. 2008; Remme et al. 2009b). In addition, differences in sodium channel (in)activation properties and pharmacological responsiveness have been described between atrial and ventricular myocytes (Li et al. 2002; Burashnikov et al. 2007).

Distinct pools of sodium channels within the cardiomyocyte: intercalated disc versus lateral membrane

More recently, the existence of two distinct functional pools of sodium channels within the cardiomyocyte has been demonstrated. Na_V1.5 localization is observed in both the intercalated discs and the lateral membrane of the myocyte, and different interacting proteins (discussed in more detail below) appear to associate with Na_V1.5 at these distinct areas (Fig. 2A). Sodium channels located at the lateral membrane are associated with the syntrophin–dystrophin complex, and dystrophin-deficient mdx mice display reduced lateral membrane Na_V1.5 expression levels and concomitant changes in conduction velocity (Petitprez et al. 2011). In contrast, Na_V1.5 channels in the intercalated disc region (which is devoid of syntrophin) interact with synapse-associated protein 97 (SAP97), a member of the membrane-associated guanylate kinase (MAGUK) family (Petitprez et al. 2011). Recently, Milstein and colleagues demonstrated that SAP97 allows Na_V1.5 to interact with the K_ir2.1 potassium channel in the heart, further underlining the relevance of the macromolecular complex (Milstein et al. 2012). The functional relevance of the two separate pools of sodium channels within myocytes has been investigated by macropatch measurements. Differences in peak sodium current amplitude and kinetics between channels located at the intercalated discs and at the cardiomyoyte lateral membrane have been observed (Fig. 2B–E), although species-specific differences probably exist (Verkerk et al. 2007; Lin et al. 2011). Nevertheless, these findings suggest that sodium channels are differentially regulated at various subcellular regions within the myocyte (Mohler & Hund, 2011). As yet, little is known about the specific roles for these separate pools of channels or their functional relevance during (patho)physiological conditions.

(Post-)transcriptional regulation of SCN5A

On the transcriptional level, SCN5A mRNA expression is likely to be regulated by various enhancer and repressor sequences close to or within the promoter region of the gene (Arnolds et al. 2012; van den Boogaard et al. 2012). Furthermore, a circadian expression pattern of cardiac Scn5a mRNA under control of the molecular clock transcription factor Bmal1 was recently demonstrated in mice (Schroder et al. 2013). Post-transcriptionally, alternative splicing of the SCN5A gene produces transcript variants with different functional characteristics (see Schroeter et al. 2010). In human heart, two main alternatively spliced SCN5A variants are expressed, which vary in the presence or absence of a glutamine at position 1077 (SCN5A-Q1077 versus SCN5A-Q1077del; Makielski et al. 2003). These splice variants may be of significant functional relevance for disease expressivity in sodium channelopathy, as evidenced by the differential effect of these variants on reduced sodium channel membrane expression due to defective trafficking of the Brugada syndrome mutation SCN5A-G1406R (Tan et al. 2006). Moreover, SCN5A splicing is also developmentally regulated, and a specific neonatal isoform has been identified which is downregulated after birth (Chioni et al. 2005; Schroeter et al. 2010). Sodium channel dysfunction may be greatly enhanced in the setting of this neonatal SCN5A splice isoform, as demonstrated recently for a mutation identified in a fetus with severe long QT syndrome (LQTS; Murphy et al. 2012). Finally, during pathophysiological conditions such as human heart failure, increased expression of abnormal C-terminal SCN5A splicing variants may result in the formation of truncated, non-functional sodium channels, thereby predisposing to cardiac conduction slowing and arrhythmogenesis (Shang et al. 2007).

Post-translational regulation of sodium channels

Post-translational modifications of Na_V1.5 expression and function include phosphorylation, glycosylation, S-nitrosylation, ubiquitination and methylation (reviewed by Rook et al. 2012). Na_V1.5 harbours conserved amino acid motifs for N-glycosylation in its extracellular domain, and glycosylation may affect sodium channel gating (Zhang et al. 1999; Rook et al. 2012). Furthermore, aberrant expression of glycogens during, for instance, heart failure may modulate voltage-dependent sodium channel-mediated electrical signalling during such pathophysiological conditions (Montpetit et al. 2009). Protein kinase A (PKA)-dependent phosphorylation of Na_V1.5 increases peak sodium current through facilitation of channel transport from intracellular storage compartments to the cell surface (Yarbrough et al. 2002). In contrast, the effect most frequently reported in relation to protein kinase C (PKC) activation concerned a reduction in sodium channel membrane trafficking and current amplitude (Murray et al. 1997; Hallaq et al. 2012). Sodium channel density and kinetics are furthermore also regulated by intracellular calcium levels (Casini et al. 2009), extracellular protons and pH (Jones et al. 2011), reactive oxygen species (Liu et al. 2010), temperature (Amin et al. 2005) and stretch (Beyder et al. 2010).

Sodium channel regulation by interacting proteins within the sodium channel macromolecular complex

Na_V1.5 is known to interact with a large number of proteins which regulate sodium channel expression, trafficking and/or function (Table 1). For many of these regulatory proteins, their interaction site within various domains of Na_V1.5 has been identified (Fig. 1). The intracellular C-terminal part of Na_V1.5 contains several protein–protein interaction sites, including the calmodulin-binding IQ motif, the PY motif, and the PDZ domain-binding motif. Proteins such as PTPH1, SAP97 and syntrophins have been described to interact with the PDZ domain-binding motif (Jespersen et al. 2006; Gavillet et al. 2006; Petitprez et al. 2011). The ubiquitin-protein ligase Nedd4-2 directly binds to the PY motif and regulates Na_V1.5 function by ubiquitylation of the channel (van Bemmelen et al. 2004). The IQ motif constitutes a binding site for calmodulin (CaM), a ubiquitous calcium-binding protein. Although discrepant effects have been described, CaM may confer sensitivity to intracellular calcium levels on Na_V1.5 (see Shy et al. 2013). Fibroblast growth factor homologous factors (FHF) family members can also interact with the C-terminus of Na_V1.5, and fibroblast growth factor homologous factor 13 (FGF13) has been shown to regulate sodium channels and conduction velocity in murine hearts (Wang et al. 2011). The cytosolic adaptor protein 14-3-3η interacts with the DI–DII linker region, thereby modulating steady-state inactivation of the channel (Allouis et al. 2006). This region furthermore contains multiple PKA and PKC phosphorylation sites in addition to an interaction site for the calcium/calmodulin-dependent protein kinase II (CAMKII), which has multiple regulatory effects on both sodium channel availability and persistent current magnitude (Wagner et al. 2006; Ashpole et al. 2012). Ankyrin-G and MOG1 have been identified to interact with the II–III linker segment of Na_V1.5, and both proteins regulate cell surface expression of sodium channels (Lemaillet et al. 2003; Mohler et al. 2004; Kattygnarath et al. 2011). A member of the superfamily of F-actin cross-linking proteins, α-actinin-2, interacts with the cytoplasmic loop connecting domains III and IV, thereby increasing sodium current density without affecting gating properties (Ziane et al. 2010). Furthermore, β-subunits (β1–β4 encoded by the SCN1B–SCN4B genes) probably bind to the extracellular connecting loops between S5 and S6, allowing them to modulate sodium channel density and kinetics (Malhotra et al. 2001; Ko et al. 2005; Meadows & Isom, 2005; Medeiros-Domingo et al. 2007). Other proteins that may directly or indirectly interact with Na_V1.5 and functionally regulate this channel include caveolin-3 (Vatta et al. 2006), glycerol-3-phosphate dehydrogenase 1-like protein (GPD1L; London et al. 2007), plakophilin-2 (Sato et al. 2009), desmoglein-2 (Rizzo et al. 2012), telethonin (Mazzone et al. 2008), and Z-band-alternatively spliced-PDZ motif protein (ZASP; Li et al. 2010; Table 1).

Functional and clinical relevance of Na_V1.5-interacting proteins

Thus, sodium channels are not isolated units within the myocyte membrane, but are functional components of a macromolecular complex through which they associate with proteins that participate in cell adhesion, signal transduction, and cytoskeleton anchoring. The functional relevance of such interactions is evidenced by the fact that mutations in these modulatory proteins are associated with sodium channel dysfunction and arrhythmia (Table 1). Of the sodium channel interacting proteins, caveolin-3 and α1-syntrophin have been implicated in LQTS; Vatta et al. 2006; Wu et al. 2008) whereas mutations in GPD1L and MOG1 have been associated with Brugada syndrome (London et al. 2007; Abriel, 2010; Kattygnarath et al. 2011). Furthermore, mutations in sodium channel β-subunits have been identified in patients presenting with Brugada syndrome (BrS) and/or conduction disease (β1 and β3), atrial fibrillation (β1, β2 and β3), LQTS (β4), and idiopathic ventricular fibrillation (IVF) in the absence of a BrS phenotype (β3) (Medeiros-Domingo et al. 2007; Watanabe et al. 2008, 2009; Hu et al. 2009; Valdivia et al. 2010; Olesen et al. 2011, 2012).

More recently, mutations in the sarcolemmal membrane-associated protein (SLMAP), a protein of unknown function localizing at T-tubules and sarcoplasmic reticulum, were identified in BrS patients. SLMAP was found to modulate intracellular trafficking of hNa_V1.5 channels, but no direct interaction between SLMAP and hNa_V1.5 was observed (Ishikawa et al. 2012). In addition, the anticancer drug taxol, which changes the properties of the cytoskeletal component tubulin, affects sodium current density and gating in vitro and is associated with increased occurrence of cardiac conduction disorders and arrhythmias in vivo (Rowinsky et al. 1991; Casini et al. 2010).

Long QT syndrome type 3 (LQT3)

Long QT syndrome (LQTS) is characterized by prolonged QT intervals on the ECG and increased risk for sudden death due to ventricular tachyarrhythmias, in particular torsades de pointes. Various subtypes of LQTS exist, each associated with distinct clinical features and underlying genetic defect. Patients with LQTS type 3 (LQT3, associated with SCN5A mutations) display arrhythmias predominantly during rest or sleep (at slow heart rates), and they are often relatively bradycardic (Schwartz et al. 2001; Schwartz, 2006). Compared to other LQT subtypes, LQT3 patients are particularly at risk for sudden death, and cardiac arrest (rather than syncope) is often the first clinical event (Zareba et al. 2001; Schwartz et al. 2001). Most SCN5A mutations associated with LQT3 typically disrupt fast inactivation of the sodium current, allowing for sodium channels to re-open, resulting in a persistent (late) inward current during the action potential plateau phase (Bennett et al. 1995). Consequently, delayed repolarization and action potential prolongation occurs, and early after-depolarizations may subsequently trigger torsades de pointes and sudden death. Alternatively, SCN5A mutations less commonly cause LQT3 through slowed inactivation (resulting in channel openings of longer duration), faster recovery from inactivation (causing increased sodium channel availability), and increased peak sodium current density (Wedekind et al. 2001; Rivolta et al. 2001; Clancy et al. 2003). Therapeutic options for LQT3 include β-blockers, sodium channel blockers mexiletine or flecainide (depending on gene mutation), and implantable cardioverter defibrillator in high risk patients (see Remme & Bezzina, 2010). However, β-blockers may not be ideal in LQT3 (since cardiac events occur mostly during rest or sleep), and the effects of sodium channel blockers are probably mutation specific. More recently, the late sodium current inhibitor ranolazine has been proposed as a potential pharmacological therapy for cardiac sodium channel disease related to gain of sodium current function, including LQT3 (Fredj et al. 2006, Moss et al. 2008; Remme & Wilde, 2013). Indeed, in a small set of five patients carrying the SCN5A-ΔKPQ mutation, short-term ranolazine infusion significantly decreased QTc-intervals in the absence of adverse effects (Moss et al. 2008), indicating a potential beneficial effect of late sodium current inhibition in LQT3 (discussed further below). In general, however, efficacy of pharmacological treatment strategies in LQT3 has not been extensively studied due to small patient population sizes, and variation in results is probably further increased by genotype-specific differences in response to treatment.

Brugada syndrome

Brugada syndrome (BrS) is a familial disorder characterized by ventricular arrhythmias and sudden cardiac death occurring in otherwise healthy individuals at a relatively young age (<40 years). On ECG analysis, a typical pattern is observed comprising ST-segment elevation in the right-precordial leads V1–V3, which is unmasked or increased after administration of Class 1A or 1C anti-arrhythmic sodium channel-blocking drugs (ajmaline, flecainide), or during exercise (see Meregalli et al. 2005). In most cases signs of conduction disease are also present on the ECG, e.g. prolonged PQ- and QRS-duration in the presence of a normal QT-interval (Antzelevitch, 2001). Arrhythmias and sudden death occur mostly during rest or sleep, and arrhythmia incidence is higher in males than females (Antzelevitch, 2006). Furthermore, episodes of fever and other factors may provoke or exacerbate the typical ECG pattern and promote onset of arrhythmias (Dumaine et al. 1999; Chockalingam et al. 2011).

In general, SCN5A mutations associated with BrS are loss-of-function mutations, leading to reduced sodium channel availability, either through decreased trafficking and membrane surface channel expression, or through altered channel gating properties (Tan et al. 2003; Viswanathan & Balser, 2004; Kapplinger et al. 2010). Reduced sodium current decreases action potential upstroke velocity, leading to atrial and ventricular conduction slowing accompanied by prolongation of PR- and QRS-intervals on the ECG. The right-precordial ST segment elevation and its relation to arrhythmogenesis is less well understood, but two major hypotheses have been proposed. The repolarization disorder hypothesis involves increased transmural heterogeneity in action potential duration, which occurs preferentially in the right ventricle due to the more pronounced presence of the transient outward potassium current (I_to) in the right ventricular epicardium (Antzelevitch, 2006). Alternatively, conduction slowing in the right ventricular outflow tract may be exacerbated in the setting of decreased sodium current (depolarization hypothesis; Meregalli et al. 2005). These characteristics may render the right ventricle more susceptible to the deleterious effects secondary to sodium channel dysfunction, thereby promoting right ventricular arrhythmogenesis observed in BrS patients (Tukkie et al. 2004; Antzelevitch et al. 2005). Structural discontinuities in the epicardium of the right ventricle may further contribute to the typical BrS ECG pattern as well as the onset of ventricular arrhythmias (Hoogendijk et al. 2010).

Progressive cardiac conduction defect (PCCD) and sick sinus syndrome

PCCD, also called Lenègre or Lev disease, is characterized by progressive conduction slowing through the His-Purkinje system, with right and/or left bundle branch block and QRS-widening, leading to complete atrioventricular block, syncope and sudden death. In some cases, inherited PCCD is associated with mutations in SCN5A (Schott et al. 1999; Wolf & Berul, 2006). From a mechanistic point of view, SCN5A mutations underlying PCCD lead to reduced sodium channel availability (loss-of-function) similar to BrS, and considerable overlap exists between these two clinical entities (see also below) (Kyndt et al. 2001). Similarly, inherited sick sinus syndrome has also been associated with loss-of-function mutations in SCN5A (Benson et al. 2003; Smits et al. 2005), in addition to atrial standstill (Groenewegen et al. 2003; Takehara et al. 2004; Lopez et al. 2011). Although sodium channels are not considered essential for sinoatrial nodal pacemaking, they do, however, contribute to cardiac automaticity through the depolarizing effects of inward sodium current (albeit relatively small). In addition to reduced automaticity of sinoatrial pacemaking tissue, decreased sodium channel availability may also cause bradycardia by slowing or block of conduction from the central sinoatrial region to the surrounding atrial tissue (see Lei et al. 2008). Furthermore, sinus bradycardia and/or arrest may also occur in LQT3 patients with SCN5A gain-of-function mutations, as a consequence of action potential prolongation caused by increased persistent inward sodium current (Veldkamp et al. 2003).

Atrial fibrillation

Atrial fibrillation is the most prevalent clinical arrhythmia affecting mostly elderly patients with underlying structural cardiac abnormalities, but it may also occur as a hereditary disease in young patients with structurally normal hearts. Both loss-of-function and gain-of-function mutations in SCN5A have been described in this familial form, which are thought to induce atrial fibrillation through decreased atrial conduction velocity and increased atrial action potential duration and excitability, respectively (Darbar et al. 2008; Ellinor et al. 2008; Makiyama et al. 2008; Li et al. 2009). In addition, a role for structural remodelling of the atria secondary to sodium channel dysfunction is suspected. Mutations in the β-subunits causing sodium current reduction have also been associated with atrial fibrillation (Watanabe et al. 2009; Olesen et al. 2011).

Dilated cardiomyopathy

Familial forms of dilated cardiomyopathy (DCM) are mostly associated with mutations in cytoskeletal proteins, but have also been reported in patients with SCN5A mutations (Bezzina et al. 2003; McNair et al. 2004), often in combination with atrial arrhythmias and/or fibrillation (Olson et al. 2005). The question remains whether arrhythmias and structural defects are a direct effect of sodium current alterations, or merely secondary to long-standing cardiac abnormalities. Biophysical properties consistent with both loss and gain of sodium channel function have been observed in SCN5A mutations associated with DCM (McNair et al. 2004; Ge et al. 2008; Nguyen et al. 2008; Bezzina & Remme, 2008). Very few commonalities between various SCN5A mutations related to DCM exist (Bezzina & Remme, 2008), although they appear to be preferentially localized to the region of the channel containing the voltage sensor (McNair et al. 2011). The mechanisms underlying SCN5A mutation-related DCM remain unclear, and may involve a complex interplay of altered sodium channel current, (pre-existent) myocardial structural abnormalities, and occurrence of long-standing arrhythmias (Fig. 3). Recently, it was demonstrated that the SCN5A-R219H mutation causes a proton leak current, suggesting that intracellular acidification may contribute to the DCM (and arrhythmias) observed in carriers of this mutation (Gosselin-Badaroudine et al. 2012).

Overlap syndromes of cardiac sodium channelopathy

In some cases, one single SCN5A mutation can result in multiple rhythm disturbances within one family. The first such mutation, the SCN5A-1795insD mutation, was characterized by our group in a large Dutch family presenting with extensive variability in type and severity of symptoms, including ECG features of sinus node dysfunction, bradycardia, conduction disease, Brugada syndrome (ST-segment elevation), and long QT syndrome (QT-interval prolongation), either in isolation or in combinations thereof (Bezzina et al. 1999; van den Berg et al. 2001). Therefore, an intriguing clinical and genetic overlap exists between these arrhythmia syndromes, previously considered separate entities. Similar findings for other mutations were subsequently reported for other SCN5A mutations by others (Kyndt et al. 2001; Grant et al. 2002; Rossenbacker et al. 2004). Today, extensive clinical and biophysical overlap among the various types of SCN5A mutations is known to exist, now referred to as ‘overlap syndromes’ of cardiac sodium channel disease (see Remme et al. 2008).

The simultaneous presence of long QT syndrome and Brugada syndrome or conduction disease in one patient was considered improbable prior to our description of the 1795insD family. Expression studies of the 1795insD mutation in Xenopus oocytes and the human embryonic kidney cell line HEK293 revealed opposing and inconclusive effects on sodium current density and kinetics (Bezzina et al. 1999; Veldkamp et al. 2000), necessitating the analysis of the biophysical properties of this mutation in native myocytes. Therefore, we generated transgenic mice carrying the heterozygous Scn5a^1798insD/+ mutation, equivalent to human SCN5A-1795insD (Remme et al. 2006). In Scn5a^1798insD/+ myocytes, a drastic reduction in peak sodium current density was found, underlying the observed conduction slowing in mice with the mutation. A small but relevant persistent inward sodium current was measured in myocytes from Scn5a^1798insD/+ mice, explaining the presence of increased action potential duration and QTc prolongation. Disruption of fast inactivation of the sodium current, as evidenced by a delayed time course of sodium current decay in Scn5a^1798insD/+ myocytes, potentially underlies the observed persistent current. These findings confirmed that the presence of a single SCN5A mutation is indeed sufficient to cause an overlap syndrome of cardiac sodium channel disease.

Disease variability in SCN5A mutation carriers

Similar to most primary electrical disorders and Mendelian disorders in general, patients harbouring mutations in SCN5A may display profound variability in the clinical expression of the disease (Probst et al. 2009). This variability may stem from the different severities of the biophysical defect associated with the various mutations. For instance, in a study conducted among patients with loss-of-function mutations in SCN5A, Meregalli and co-workers demonstrated that truncating mutations were associated with more severe conduction disease (Meregalli et al. 2009). However, variability in disease severity is also observed among family members with the same primary genetic defect. Here, while some family members display electrocardiographic signs of the disease, others may be unaffected in spite of having the same mutation, as was observed in the SCN5A-1795insD family (Bezzina et al. 1999; Fig. 4A). Moreover, among the electrocardiographically affected family members, some individuals present with arrhythmia while others remain symptom-free throughout life. Thus, identification of mutation carriers at risk for serious cardiac events is essential.

Modifiers of disease expression: age and sex

Kyndt and colleagues suggested that the patient's sex is a possible modifier of disease expression in a family carrying the SCN5A-G1406R mutation. In this family females displayed predominantly with conduction defects whereas males presented mostly with clinical features pertaining to the Brugada syndrome (Kyndt et al. 2001). The preponderance of Brugada syndrome in males also suggested sex as a modifier in sodium channelopathy. The contribution of age in modulation of the phenotype in the setting of sodium channelopathy is most elegantly observed in patients with progressive cardiac conduction disease (Probst et al. 2003), where age-related fibrosis is thought to underlie increased cardiac conduction delay with advancing age. This hypothesis is supported by experimental findings in Scn5a^+/− mice that are haploinsufficient for the cardiac sodium channel (Royer et al. 2005). Interestingly, fibrotic changes were most pronounced in aged male Scn5a^+/− mice, in line with the male predominance observed in Brugada syndrome patients (Jeevaratnam et al. 2012).

Genetic modifiers

While evidence points to genetic background effects on the severity of ECG manifestations and arrhythmia risk, genetic modifiers of clinical variability in sodium channelopathy remain largely unknown (Kolder et al. 2012). We have previously investigated the effects of genetic factors on disease severity by comparing the phenotype in two distinct strains of mice carrying the Scn5a^1798insD/+ mutation equivalent to SCN5A-1795insD in humans. We found that phenotype severity was more pronounced in 129P2 mice as compared to FVBN/J mice, providing the first conclusive evidence that genetic modifiers determine disease expressivity in cardiac sodium channelopathy (Remme et al. 2009a). By comparing cardiac gene expression between these two strains of mice we subsequently identified Scn4b, encoding a sodium channel β-subunit, as a modifier of conduction disease severity in these mice (Remme et al. 2009a). More recently, in a systems genetics approach on F2 progeny arising from the 129P2 and FVBN/J mouse strains carrying the Scn5a^1798insD/+ mutation, we have identified Tnni3k encoding troponin 1-interacting kinase as a novel modulator of cardiac conduction (Scicluna et al. 2011, Lodder et al. 2012; Fig. 4B). We are currently correlating findings from these mouse studies to clinical and genetic data from carriers and non-carriers in the large family segregating the SCN5A-1795insD mutation.

Although a few studies have investigated the effects of certain candidate genetic variants in the form of single nucleotide polymorphisms (SNPs) on disease severity in small families with cardiac sodium channelopathy (Groenewegen et al. 2003), systematic genome-wide searches for such genetic modifiers are as yet unavailable. SNPs recently identified in large genome-wide association studies conducted in the general population as modulators of heart rate and ECG indices of conduction and repolarization (reviewed in Kolder et al. 2012), constitute ideal candidates for modulatory effects.

Sodium channels and cardiac development/structure

Through its macromolecular complex, the sodium channel interacts with many cytoskeletal proteins and components of the extracellular matrix, and defects in some of these interacting proteins have been shown to negatively affect sodium channel function (see Abriel, 2010). It is currently unknown whether sodium channel dysfunction in turn can also destabilize cytoskeletal integrity. However, studies in embryos of mice carrying homozygous Scn5a mutations (which are embryonically lethal) indicate that sodium channels are essential for maintaining cardiac structure during development (Nuyens et al. 2001; Papadatos et al. 2002). Similarly, knockdown of Scn5a in zebrafish resulted in abnormal heart development, which appeared to be independent of electrophysiological effects (Chopra et al. 2010).

Potential role for late sodium current and/or CAMKII

Murine models of LQT3 have demonstrated that the increased late or persistent sodium current associated with gain-of-function Scn5a mutations can disrupt intracellular calcium homeostasis, thereby providing a pro-arrhythmic substrate and a potential mechanism for development of fibrosis and/or hypertrophy (Lindegger et al. 2009; Remme et al. 2010). Interestingly, the late sodium current blocker ranolazine was effective in reducing or preventing intracellular calcium overload in vitro (Lindegger et al. 2009). Thus, pharmacological late sodium current inhibition by compounds such as ranolazine may prove a useful and beneficial therapeutic approach in LQT3 patients, not only through stabilizing repolarization abnormalities, but also by preventing more long-term detrimental effects of intracellular calcium overload (see Remme & Wilde, 2013). Moreover, increased sodium influx with subsequent activation of CAMKII was observed in cardiomyocytes from mice overexpressing the human SCN5A-N1325S gain-of-function mutation (Yao et al. 2011), providing another potential therapeutic target.

Involvement of the TGFβ pathway

Electrical activity-dependent stimulation of pro-fibrotic factors of the transforming growth factor β (TGFβ) pathway may play a role in sodium channel dysfunction (Leask, 2007). In skeletal muscle, inhibition of sodium current was shown to increase levels of TGFβ1 receptor levels, indicating a potential regulatory effect of electrical activity on TGFβ signalling (Ugarte & Brandan, 2006). Indeed, TGFβ1 transcripts were upregulated in aged heterozygous Scn5a knockout mice, suggesting development of TGFβ1-mediated cardiac fibrosis secondary to sodium channel dysfunction (Hao et al. 2011).

Novel role for sodium channels in regulating myocardial structure and function?

Taken together, these observations are intriguing since they imply that sodium channels not only determine electrophysiological characteristics of the myocardium, but also exert as yet unknown regulatory effects on myocardial structure and function, implying a functional contribution of cardiac sodium channels to signal transduction. Different types of SCN5A mutations (gain- versus loss-of-function) may lead to structural abnormalities through various mechanisms (Fig. 3). Regardless of the underlying mechanism, cardiac structural abnormalities secondary to sodium channel dysfunction may further predispose to arrhythmogenesis, and will determine overall cardiac function, thereby contributing significantly to disease severity, progression and prognosis. Thus, novel treatment strategies aimed at reduction and/or prevention of structural abnormalities secondary to sodium channel dysfunction may prove more effective than existing therapy.

Competing interests

None.

Funding

This research was supported by the Netherlands Heart Institute (grant 061.02), and the Division for Earth and Life Sciences (ALW; project 836.09.003) with financial aid from the Netherlands Organization for Scientific Research (NWO).