Digenic inheritance novel mutations in SCN5a and SNTA1 increase late I_Na contributing to LQT syndrome

Abstract

SCN5A and SNTA1 are reported susceptible genes for long QT syndrome (LQTS). This study was designed to elucidate a plausible pathogenic arrhythmia mechanism for the combined novel mutations R800L-SCN5A and A261V-SNTA1 on cardiac sodium channels. A Caucasian family with syncope and marginally prolonged QT interval was screened for LQTS-susceptibility genes and found to harbor the R800L mutation in SCN5A and A261V mutation in SNTA1, and those with both mutations had the strongest clinical phenotype. The mutations were engineered into the most common splice variant of human SCN5A and SNTA1 cDNA, respectively, and sodium current (I_Na) was characterized in human embryonic kidney 293 cells cotransfected with neuronal nitric oxide synthase (nNOS) and the cardiac isoform of the plasma membrane Ca-ATPase (PMCA4b). Peak I_Na densities were unchanged for wild-type (WT) and for mutant channels containing R800L-SCN5A, A261V-SNTA1, or R800L-SCN5A plus A261V-SNTA1. However, late I_Na for either single mutant was moderately increased two- to threefold compared with WT. The combined mutations of R800L-SCN5A plus A261V-SNTA1 significantly enhanced the I_Na late/peak ratio by 5.6-fold compared with WT. The time constants of current decay of combined mutant channel were markedly increased. The gain-of-function effect could be blocked by the N^G-monomethyl-l-arginine, a nNOS inhibitor. We conclude that novel mutations in SCN5A and SNTA1 jointly exert a nNOS-dependent gain-of-function on SCN5A channels, which may consequently prolong the action potential duration and lead to LQTS phenotype.

the inherited long qt syndrome (LQTS) is a hereditary cardiac disorder characterized by prolonged QT interval on the surface ECG and increased risk for sudden death due to ventricular tachyarrhythmia (9, 13). It is one of the common causes of unexplained syncope especially in young, seemingly healthy individuals (32). To date more than 600 mutations in 13 genes have been identified to associate with LQTS (18), in which six genes encode cardiac ion channels (KCNQ1, KCNH2, SCN5A, KCNJ2, CACNA1C, and KCNJ5) and seven encode ion channel subunits or channel-interacting proteins (ChIPs) (ANKB, KCNE1, KCNE2, CAV3, SCN4B, AKAP9, and SNTA1) (7).

SCN5A encodes the α-subunit of voltage-gated cardiac sodium channel hNa_v1.5 also denoted SCN5A that is responsible for large peak inward sodium current (I_Na) in the heart (8). The α-subunit has four homologous domains (DI-DIV), and each domain contains six transmembrane segments (S1-S6) (10, 23). LQTS-associated mutations in SCN5A cause LQT3, a gain-of-function leading to prolonged cardiac action potential (AP) duration, lengthened QT interval, and increased risk of arrhythmia (38).

α₁-Syntrophin (SNTA1), a dystrophin-associated protein, is the dominant syntrophin isoform in skeletal and cardiac muscle (4, 21). As a scaffolding adapter, SNTA1 binds to neuronal nitric oxide (NO) synthase (nNOS), which is constitutively expressed in the heart, and the cardiac isoform of the plasma membrane Ca²⁺/calmodulin-dependent ATPase (PMCA4b) to form a complex in which PMCA4b acts as a potent inhibitor of NO synthesis (20, 34). SNTA1 also interacts directly with the PDZ domain-binding motif formed by the last three residues (serine-isoleucine-valine) of Na_v1.5 C terminus (14). We have demonstrated that SNTA1 is associated with α-subunit of hNa_v1.5 and the nNOS-PMCA4b complex in cardiomyocytes (29). We have previously described the mutation A390V-SNTA1 in a LQTS patient that disrupted the association with PMCA4b and antagonized the inhibition of nNOS, resulting in augmentation of both peak and late I_Na, suggesting SNTA1 as a novel LQTS-susceptibility gene (LQT12) (29).

Here, we report the co-existence of both SCN5A and SNTA1 mutations in the same patient. These two mutations have a combined effect on gain-of-function of late I_Na, which may underlie the mechanism of clinical LQTS phenotype.

MATERIALS AND METHODS

Genetic analysis.

Genetic analysis was performed after written informed consent was obtained from the subjects. The investigation conformed to the principles outlined in the Declaration of Helsinki. The University of Wisconsin institutional review board (IRB) approved this study as exempt from further review. Genomic DNA was extracted from peripheral blood lymphocytes as previously described (31) and screened for the entire open-reading frames of 12 LQTS-susceptibility genes by PCR, denaturing high-performance liquid chromatography (HPLC) and direct DNA sequencing.

Site-directed mutagenesis and heterologous expression.

Mutant human R800L-SCN5A was generated using QuickChange site-directed mutagenesis kits according to the manufacturer's instructions (Stratagene, La Jolla, CA) and was engineered into the most common splice variant of human cardiac voltage-dependent Na channel SCN5A/hNa_v1.5 [lacking a glutamine at position 1077, we note as Q1077del (Genbank accession no. AY148488)] in the pcDNA3 plasmid vector (Invitrogen, Carlsbad, CA) as previously reported (19, 27, 28, 36). Oligonucleotide primers containing the corresponding R800L-SCN5A mutation were synthesized using following sequences: forward 5′-CTGGGCCTGTCCCTCATGAGCAACTTG-3′ and reverse 5′-CAAGTTGCTCATGAGGGACAGGCCCAG-3′ (the mutated site is underlined). The cDNA of SNTA1 (Genbank accession no. NM_003098) was subcloned into pIRES2EGFP plasmid vector (Clontech Laboratories, Palo Alto, CA). A mutation of A261V in SNTA1 was introduced into WT-SNTA1 by site-directed mutagenesis similarly. Oligonucleotide primers containing the corresponding A261V-SNTA1 mutation were synthesized using the following sequences: forward 5′-GCGAGGTCGTGGGTGACTGCCATCCAA-3′ and reverse 5′-TTGGATGGCAGTCACCCACGACCTCGC-3′ (the mutated site is underlined). The cDNAs of nNOS (Genbank accession no. NM_052799) and PMCA4b (Genbank accession no. AY560895) were generous gifts from Solomon H. Snyder (John Hopkins University) and Emanuel E. Strehler (Mayo clinic), respectively. All clones were sequenced to confirm integrity and to ensure the presence of the target mutations without other substitutions. The WT- or R800L-SCN5A cDNA was transiently cotransfected with WT- or A261V-SNTA1 cDNA as well as nNOS and PMCA4b cDNA, at a ratio of 4:1:4:4, respectively, into HEK293 cells with FuGENE6 reagent (Roche Diagnostics, Indianapolis, IN) following manufacturer's instructions.

Chemical reagent.

The NOS inhibitor N^G-monomethyl-l-arginine (l-NMMA) was obtained from Cayman Chemical (Ann Arbor, MI). The l-NMMA was diluted in PBS buffer (pH 7.2) 10 min before use.

Standard electrophysiological measurements for functional characterization.

Macroscopic voltage-gated I_Na was measured 24 h after transfection with the standard whole-cell patch clamp technique at a temperature of 22°–24°C in human embryonic kidney (HEK)293 cells. Cells were continuously perfused with bath (extracellular) solution containing (in mM) 140 NaCl, 4 KCl, 1.8 CaCl₂, 0.75 MgCl₂, and 5 HEPES (pH 7.4 set with NaOH). The pipette (intracellular) solution contained (in mM) 120 CsF, 20 CsCl_2, 5 EGTA, and 5 HEPES and was adjusted to pH 7.4 with CsOH. Microelectrodes were made of borosilicate glass with a puller (P-87; Sutter Instrument Co, Novato, CA) and were heat-polished using a microforge (MF-83; Narishige, Tokyo, Japan). The resistances of microelectrodes ranged from 1.0 to 2.0 MΩ when filled with recording solution. Voltage clamp was generated by Axopatch 200B amplifier (Axon Instruments, Foster City, CA) and controlled using pClamp software 10.2. The series-resistance was compensated. Membrane current data were digitalized at 100 kHz, low-pass filtered at 5 kHz, and then normalized to membrane capacitance.

The standard voltage clamp protocols for functional characterization were as follows. Activation was measured for steps between −120 mV and +60 mV in 10-mV increments from a holding potential of −140 mV. The midpoint and slope factor of activation was obtained using a Boltzmann function where G_Na = [1 + exp (V_1/2 − V)/K]⁻¹, where V_1/2 and k were the midpoint and slope factor, respectively. G/G_Na = I_Na (norm)/(V − V_rec), where V_rec was the reversal potential and V was the membrane potential. Steady-state inactivation was measured in response to a test depolarization to 0 mV for 24 ms from a holding potential of −140 mV, following 1-s conditioning pulse from −150 to 0 mV in 10-mV increments. The voltage-dependent availability from inactivation relationship was determined by fitting the data to the Boltzmann function: I_Na = I_Na-max [1 + exp (V_c − V_1/2)/K]⁻¹, where V_1/2 and k were the midpoint and the slope factor, respectively, and V_c is the membrane potential. Decay rates and amplitude component were measured from the trace beginning at 90% of peak I_Na to 24 ms and fitted with a sum of exponentials (exp): I_Na (t) = 1 − [A_f exp(−t/τ_f) + A_S exp(−t/τ_S)] + offset, where t was time, and A_f and A_S were fractional amplitudes of fast and slow components, respectively. Late I_Na was measured as the mean current between 600 and 700 ms after the initiation of the depolarization from −140 mV to −20 mV for 700 ms after passive leak subtraction as previously described. The leak current was subtracted as previously described (12).

Statistic analysis.

All data points are shown as the mean value and SE. Determinations of statistical significance were performed using a Student's t-test for comparisons of two means or using one-way ANOVA for comparisons of multiple means. A P value of < 0.05 was considered statistically significant.

RESULTS

Clinical features.

A three-generation Caucasian family with LQTS was identified (Fig. 1A). The proband was a 37-year-old woman who had experienced recurrent bouts of syncope for 30 years. Her resting 12-lead ECG showed a marginal prolongation of QTc interval (480 ms) while the echocardiography revealed no significant abnormality (Fig. 1A). Her two children, two siblings, and both parents were available to be clinically evaluated as the immediate family. The father presented with occasional dizziness, but denied palpitation or syncope. The mother who died at the age of 76 had suffered from schizophrenia and diabetes mellitus type two but no episode of syncope. The brother reportedly had an arrhythmia when administered with anesthesia, and the sister had seizures in her childhood. The proband's 10-year-old son suffered from supraventricular tachycardia, and the ECG showed moderately prolonged QTc (500 ms; Fig. 1A). The 7-year-old daughter was asymptomatic.

Fig. 1.

ECG phenotype and genotype. A: pedigree and ECG tracing of the proband (arrow) and her son. The QTc of the proband exceeds 480 ms, and the QTc of the proband's son is 500 ms. B: sequence chromatograms of SCN5A and SNTA1 mutations and the amino acid conservation of R800 in SCN5A and A261 in SNTA1 across species. Left: arrow indicates the 2399G>T substitution leading to the R800L mutation in SCN5A. Right: arrow indicates the 782C>T substitution leading to the A261V mutation in SNTA1.

Genetic analysis.

Twelve LQTS candidate genes were screened in five family members. As a result, two novel mutations were detected in the proband (Fig. 1A). One was a missense mutation in SCN5A, a single-base transition at nucleotide 2399 (2399G>T) resulting in a replacement of an arginine by a leucine at codon 800 (designated R800L; Fig. 1B). It was located within the extracelluar linker between S3 and S4 segment of hNa_v1.5 DII. The same nucleotide change was confirmed in the proband's son and in the deceased mother on DNA extracted from a postmortem sample (Fig. 1A). The proband was also a carrier of a rare missense mutation in SNTA1 (Fig. 1A). The SNTA1 mutation A261V was derived from a C-to-T substitution at nucleotide 782, causing an amino acid change from alanine to valine at codon 261 (designated A261V; Fig. 1B). The A261V-SNTA1 mutation was identified in the proband's father, son, and daughter as well (Fig. 1A). Both mutations involved residues conserved across a variety of species (Fig. 1B). R800L was absent from 1300 SCN5A alleles (17), and A261V was absent from 400 SNTA1 alleles (Transgenomic; Familion Tests) of an ethnically matched, healthy control cohort.

The combined effect of R800L-SCN5A and A261V-SNTA1 on late I_Na.

Whole cell patch-clamp recordings were used to study I_Na of WT-SCN5A or R800L-SCN5A plus WT-SNTA1 or A261V-SNTA1 heterologously expressed in HEK293 cells. Channels were coexpressed with nNOS and PMCA4b, known to modulate cardiac I_Na through a macromolecular protein complex. Representative I_Na traces are shown in Fig. 2A. When compared with the WT-SCN5A plus WT-SNTA1 channel, R800L-SCN5A plus A261V-SNTA1, WT-SCN5A plus A261V-SNTA1, and R800L-SCN5A plus WT-SNTA1 channels showed small increases in peak I_Na that did not reach statistical significance (Fig. 2B).

Fig. 2.

Electrophysiological properties of SCN5A in human embryonic kidney (HEK)293 cells coexpressing PMCA4b, neuronal nitric oxide synthase (nNOS), wild-type (WT), or mutant SCN5A and WT or mutant SNTA1. N^G-monomethyl-l-arginine (l-NMMA) was introduced into HEK293 cell culture media 6 h before testing. A: whole-cell current traces from representative experiments of each group after 24 h transfection. B: summary data of peak sodium current (I_Na) density of each group showing no significant difference of peak I_Na in R800L-SCN5A plus A261V-SNTA1, WT-SCN5A plus A261V-SNTA1, or R800L-SCN5A plus WT-SNTA1 compared with WT-SCN5A plus WT-SNTA1. Currents were elicited by test depolarization to 24 ms from a holding potential of −140 mV. Numbers of tested cells are indicated above the bar. Symbols represent means, and bars represent SE.

Late I_Na was measured as the leak subtracted inward current remaining at the end of a 700-ms-long depolarization (Fig. 3A) and expressed as a ratio of late/peak I_Na. The WT-SCN5A plus WT-SNTA1 channel had the smallest late I_Na, but late I_Na was increased for R800L-SCN5A plus WT-SNTA1, WT-SCN5A plus A261V-SNTA1, and R800L-SCN5A plus A261V-SNTA1. Summary data (Fig. 3B) showed that R800L-SCN5A plus WT-SNTA1 channel had twofold greater late I_Na compared with WT-SCN5A plus WT-SNTA1 channel, although it did not reach statistical significance. However, WT-SCN5A plus A261V-SNTA1 and R800L-SCN5A plus A261V-SNTA1 channels had 3.4-fold and 5.6-fold increased late I_Na, respectively (P < 0.05, shown in Table 1).

Fig. 3.

Late I_Na induced by cardiac sodium channel in HEK cells coexpressing PMCA4b, nNOS, WT or mutant SCN5A, and WT or mutant SNTA1. l-NMMA was introduced into HEK293 cell culture media 6 h before testing. A: representative traces showing increased late I_Na associated with R800L-SCN5A plus A261V-SNTA1 and WT-SCN5A plus A261V-SNTA1 compared with WT-SCN5A plus WT-SNTA1. B: summary data for late I_Na normalized to peak I_Na after leak subtraction. Currents were elicited by a test depolarization pulse from −140 mV to −20 mV for 700 ms after the initiation of the depolarization. Numbers of tested cells are indicated above the bar. Symbols represent means, and bars represent SE. *P < 0.05 vs. WT-SCN5A plus WT-SNTA1. See table 1 for n numbers.

Table 1.

Electrophysiological properties of hNa_v1.5 channels in human embryonic kidney 293 cells coexpressing neuronal nitric oxide synthase, cardiac isoform of plasma membrane Ca²⁺/calmodulin-dependent ATPase, WT- or R800L-SCN5A, and WT- or A261V-SNTA1

Samples	Peak INa		Activation			Inactivation			Late INa
	pA/pF	n	V1/2, mV	k	n	V1/2, mV	k	n	%	n
WT-SCN5A+WT-SNTA1	−181 ± 16	10	−40 ± 1.6	4.4	9	−83 ± 1.0	6.4	10	0.13 ± 0.03	5
R800L-SCN5A+WT-SNTA1	−195 ± 16	13	−41 ± 1.2	4	9	−81 ± 1.0	7.3	10	0.31 ± 0.1	8
WT-SCN5A+A261V-SNTA1	−223 ± 17	15	−42 ± 1.2	4	11	−80 ± 0.7	6.6	13	0.44 ± 0.1*	7
R800L-SCN5A+A261V-SNTA1	−212 ± 12	17	−43 ± 0.6	5	14	−81 ± 0.8	7.0	14	0.73 ± 0.1*	11
R800L-SCN5A+A261V-SNTA1+l-NMMA	−191 ± 11	9	−37 ± 1.0	5	6	−83 ± 1.2	6.5	6	0.21 ± 0.1	5

Values are means ± SE. I_Na, sodium current; V_1/2 and k, midpoint and slope factor, respectively.

^*P < 0.05 vs. wild-type (WT)-SCN5A+WT-α₁-syntrophin (SNTA1).

R800L-SCN5A plus A261V-SNTA1 increased the hNa_v 1.5 window current.

To investigate the gating properties of mutant Na_v1.5 channels, we analyzed the kinetic parameters concerning activation and inactivation of R800L-SCN5A plus WT-SNTA1, WT-SCN5A plus A261V-SNTA1, and R800L-SCN5A plus A261V-SNTA1 channels and compared the data with that of WT-SCN5A plus WT-SNTA1 channel. Peak I_Na at each voltage were normalized to the whole cell capacitance and plotted against test voltages. The steady-state activation of R800L-SCN5A plus A261V-SNTA1 was shifted toward hyperpolarized potential by −3 mV, and the steady-state inactivation was slightly shifted toward depolarized potential by 1 mV compared with WT-SCN5A plus WT-SNTA1 (Fig. 4, A and B). Although no significant difference was shown in activation or inactivation parameters, the incomplete inactivation of R800L-SCN5A plus A261V-SNTA1 channel enlarged the overlap of the activation and inactivation curves (Fig. 4, C and D), which resulted in increased window current. Parameters for the recovery from inactivation for WT and mutant channels were not significantly different (data not shown).

Fig. 4.

Voltage-dependence of steady-state activation and inactivation was measured from HEK293 cells transiently expressing PMCA4b, nNOS, WT or mutant SCN5A, and WT or mutant SNTA1. l-NMMA was introduced into HEK293 cell culture media 6 h before testing. A: R800L-SCN5A plus A261V-SNTA1 did not alter steady-state activation parameters significantly. B: R800L-SCN5A plus A261V-SNTA1 did not affect steady-state inactivation significantly. C: peak current activation data are replotted as a conductance (G) curve with steady-stated inactivation relationships, showing that R800L-SCN5A plus A261V-SNTA1 increases the overlap of these relationships (window current) compared with WT-SCN5A plus WT-SNTA1. After incubation with l-NMMA, the increased window current of combined mutant channel was diminished. D: enlarged scale from C to better show the window current. See table 1 for n numbers.

R800L-SCN5A plus A261V-SNTA1 caused slower decay of I_Na.

Time constants (τ_f, τ_s) were obtained from 2-exponential fits of decay phase of macroscopic I_Na measured at various potentials. When compared with WT-SCN5A plus WT-SNTA1 channel, the R800L-SCN5A plus A261V-SNTA1 channel significantly increased τ_f and τ_s values through a wide range of test potentials (P < 0.05; Fig. 5, A and B). This suggested that inactivation of I_Na in R800L-SCN5A plus A261V-SNTA1 was impaired.

Fig. 5.

Decay of macroscopic current. A: voltage dependence of inactivation fast time constants. When compared with WT-SCN5A plus WT-SNTA1, R800L-SCN5A plus A261V-SNTA1 showed larger fast component (τ_f) values through a wide range of test potentials from −40 mV to 10 mV. B: voltage dependence of inactivation slow time constants. When compared with WT-SCN5A plus WT-SNTA1, R800L-SCN5A plus A261V-SNTA1 showed larger slow component (τ_s) values through a wide range of test potentials from −40 mV to 10 mV. Both effects can be abolished by l-NMMA. *P < 0.05 vs. WT-SCN5A plus WT-SNTA1. See table 1 for n numbers.

R800L-SCN5A plus A261V-SNTA1 altered sodium channel gating properties through a nNOS-dependent mechanism.

To further observe the effect of NOS inhibition on late I_Na, we introduced l-NMMA (100 μmol/l) into the HEK293 cell culture medium 6 h before testing. The remarkable augmentation in late I_Na produced by R800L-SCN5A plus A261V-SNTA1 channel was blocked by l-NMMA (Fig. 3, A and B), indicating that NO is crucial to maintaining proper function of SCN5A-SNTA1-nNOS-PMCA4b complex. Notably, time constants (τ_f, τ_s) of R800L-SCN5A plus A261V-SNTA1 also returned to normal levels after the application of l-NMMA (Fig. 5, A and B), suggesting the alteration of sodium channel gating properties caused by the SCN5A and SNTA1 mutations was mediated through a nNOS-dependent mechanism.

To clarify whether nNOS and PMCA4b (i.e., the complete SCN5A-SNTA1-nNOS-PMCA4b complex) are required for the observed effects, we performed functional characterization in HEK293 cells only coexpressing WT- or R800L-SCN5A and WT- or A261V-SNTA1 without nNOS and PMCA4b expression. Peak I_Na, late I_Na, and channel kinetics showed no significant differences compared with WT channel alone (data not shown). These data imply that the combined effects of SCN5A and SNTA1 mutations on sodium channel are mediated by the entire SCN5A-SNTA1-nNOS-PMCA4b complex.

DISCUSSION

Major findings.

In the present study, we identified two novel mutations, R800L-SCN5A and A261V-SNTA1, that coexisted in the proband and her son. They both had symptoms of arrhythmia and prolonged QT interval, whereas other family members with only one of the mutations had weaker cardiac phenotypes, suggesting the impact of a double-hit. Functional assays using a heterologous expression system revealed that late I_Na of R800L-SCN5A plus A261V-SNTA1 channel was significantly increased and was reversed by nNOS inhibitor. In addition, incomplete inactivation of combined mutant channel enlarged the overlap of activation and inactivation, thus increased the window current as shown in Fig. 4D, which has been implicated in increased cardiac vulnerability.

Electrophysiological mechanisms for altered hNa_v1.5 function.

We suggest two possible mechanisms that may contribute to the gain-of-function effect on hNa_v1.5 function. First, the incomplete inactivation resulted in an enlarged overlapping area between the steady-state activation and inactivation curves, thereby increased the window current of combined mutant channel. The window current is primarily determined by the overlap of sodium channel activation and inactivation (16, 26). It identifies a range of voltages where channels have a small probability of being partially but not fully inactivated (6). A small percentage of non-inactivated channels are available to be activated within this voltage range, resulting in a sustained sodium inward current. When the membrane reenters this voltage range during the repolarization phase, the window current will flow in during the cardiac AP (38). The window current tends to depolarize the resting membrane potential and may alter the excitability of cardiomyocytes (16). Due to the low probability of hNa_v1.5 channel opening within the window, the non-inactivating component of the WT channel was relatively small and did not substantially alter the window amplitude. However, R800L-SCN5A plus A261V-SNTA1 channel enlarged the overlap of activation and inactivation due to incomplete inactivation. Such alterations would increase both the critical voltage range and the magnitude of the resulting window current (38). As previously discussed by other authors (2, 22, 33), such a maintained inward current might prolong AP duration. Furthermore, the inward window current would lead to small depolarization near the resting membrane potential as well as reduced threshold for initiating cardiac APs (16). Myocardial hyperexcitability coupled with prolongation of APD might lay the basis for LQTS phenotype in the patient. Second, the remarkably slowed decay of currents observed in combined mutant channel indicated that both fast inactivation and slow inactivation were impaired. Decelerated decay of macroscopic currents, stemming from delayed onset of inactivation, often occurs in combination with a persistent current (25). A slower current decay alone should not directly lengthen APD, but might affect voltage-dependent activity of other outward or inward currents that are crucial for APD, since additional sodium inward current was generated during the early phase of the action potential plateau, which resulted in augmentation of the net ionic current (24).

Disruption in SCN5A-SNTA1-nNOS-PMCA4b complex.

Na_v1.5 has been reported to be part of the dystrophin multi-protein complex, in which syntrophin regulates Na_v1.5 gating by a PDZ domain-mediated interaction (15). Like other isoforms, α₁-syntrophin consists of four conserved domains. Two pleckstrin homology domains (PH1 and PH2) are responsible for recruiting proteins to the sarcolemma (37). The PDZ domain interacts with the COOH-terminal PDZ domain-binding motif of Na_v1.5, which is composed of the last three residues of the protein, serine-isoleucine-valine. The syntrophin unique COOH-terminal domain (SU) can bind SNTA1 to dystrophin (3).

Gavillet et al. identified that the protein level of Na_v1.5 was decreased in the heart of dystrophin-deficient mice (mdx), and this reduced expression could not be explained by a decrease in the mRNA level of SCN5A (14). It was previously confirmed in cardiomyocytes that the activity of nNOS is negatively regulated by PMCA4b through PDZ domain-mediated interaction (20). Williams et al. reported that when SNTA1 was applied to the nNOS-PMCA4b complex to form the larger nNOS-PMCA4b-SNTA1 complex, the inhibition of nNOS precipitated by PMCA4b was elevated to maximum, compared with the nNOS-PMCA4b complex, suggesting that the interaction of PMCA4b and SNTA1 synergistically inhibit NOS-mediated NO production (34). Thereafter, two missense mutations in SNTA1 were found in patients with inherited LQTS. Wu et al. showed that one SNTA1 mutation, A257G, when coexpressed with Na_v1.5 in HEK293 cells or in neonatal cardiomyocytes, the midpoint of steady-state activation was negatively shifted and the peak current density was elevated (35). Our group demonstrated the existence of the macromolecular complex SCN5A-SNTA1-nNOS-PMCA4b in cadiomyocytes and reported another LQTS-related SNTA1 mutation, A390V, which disrupted association with PMCA4b, released the inhibition of nNOS, and resulted in the peak and late I_Na through S-nitrosylation of cardiac sodium channel mediated by local accentuated NO concentration (29). We also identified three mutations in SNTA1 (11) (S287R, T372M, G460S) associated with sudden infant death syndrome that caused similar effects as the LQTS-linked A390V mutation (i.e., increased peak and late I_Na as well as positive shifts of steady-state inactivation). In this study, the WT-SCN5A plus A261V-SNTA1 and R800L-SCN5A plus A261V-SNTA1 channels showed similar gain-of-function effects on cardiac sodium channel via the SCN5A-SNTA1-nNOS-PMCA4b macromolecular complex. Particularly interesting was the finding that combined mutant channel exhibited the most prominent gain-of-function effect (i.e., increased late I_Na and slowed current decay). Moreover, the effects on I_Na were reversed by the nNOS inhibitor l-NMMA, which further supported the fact that combined mutant channel disturbed the normal function of SCN5A-SNTA1-nNOS-PMCA4b complex through NO regulation.

Clinical implications.

The pathogenicity of inherited arrhythmias was once thought to stem solely from primary abnormalities of ion channel proteins leading to the term ion channelopathies (5). Recently, it has become apparent that ChIPs including auxiliary β-subunits and anchoring/adaptor protein regulate ion channels (1). These regulatory proteins directly modify ion channel functions and harbor mutations found in patients with inherited cardiac arrhythmia including LQTS (30).

SCN5A, encoding the α-subunit of hNa_v1.5, has been known to be related to LQT3 (38). SNTA1, encoding α₁-syntrophin, was implicated as a novel LQTS-susceptibility gene (LQT12) (29). Here we show a double hit had a combined effect on late I_Na and caused symptoms in those individuals bearing both mutations. In this study, we exhibited molecular and functional evidence demonstrating that mutations in SCN5A and SNTA1 jointly increased the late I_Na and contributed to LQTS phenotype.

Study limitations.

There are limitations to the present investigation that need to be addressed. First, we generated the eletrophysiological data from in vitro experiments by coexpressing the macrocomplex SCN5A-SNTA1-nNOS-PMCA4b in HEK293 cells. In a more native environment, e.g., human cardiomyocytes, whether these two mutations could jointly exert similar effect on late I_Na represents a possible future direction for this work. Second, because α₁-syntrophin is a scaffolding protein with several protein interaction motifs, A261V-SNTA1 may not be specific to SCN5A but also may interact with other ion channel complexes, producing additional effects. Third, although we have shown that the effect involves direct nitrosylation of the SCN5A protein (29), the sites of nitrosylation, detailed biophysical mechanisms, and possible effects of nitrosylation on other channel related proteins remain to be studied.

Conclusion

In conclusion, two novel mutations R800L-SCN5A and A261V-SNTA1 in the same patient have an additive effect on gain of function of late I_Na, which may consequently prolong the action potential and lead to LQTS phenotype. Our finding provides the first example that ChIPs and SCN5A may jointly contribute to the pathogenesis of LQTS.

GRANTS

This work was supported by University of Wisconsin Cellular and Molecular Arrhythmia Research Program (to J. C. Makielski), the National Heart, Lung, and Blood Institute (HL-71092 to J. C. Makielski), the American Heart Association National Center (11SDG7470009 to B.-H. Tan), and the National Science Foundation of China (81070150 to J. Pu).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: R.-M.H. performed experiments; R.-M.H. analyzed data; R.-M.H., B.-H.T., K.M.O., and C.R.V. interpreted results of experiments; R.-M.H., K.M.O., and A.P. prepared figures; R.-M.H. drafted manuscript; R.-M.H., B.-H.T., J.P., and J.C.M. edited and revised manuscript; R.-M.H., B.-H.T., K.M.O., C.R.V., A.P., J.P., and J.C.M. approved final version of manuscript; B.-H.T., J.P., and J.C.M. conception and design of research.