|Home | About | Journals | Submit | Contact Us | Français|
Individual mutations in the SCN5A-encoding cardiac sodium channel α-subunit cause single cardiac arrhythmia disorders, but a few cause multiple distinct disorders. Here we report a family harboring an SCN5A mutation (L1821fs/10) causing a truncation of the C-terminus with a marked and complex biophysical phenotype and a corresponding variable and complex clinical phenotype with variable penetrance.
A 12-year-old male with congenital sick sinus syndrome (SSS), cardiac conduction disorder (CCD), and recurrent monomorphic ventricular tachycardia (VT) had mutational analysis that identified a 4 base pair deletion (TCTG) at position 5464-5467 in exon 28 of SCN5A. The mutation was also present in six asymptomatic family members only two of which showed mild ECG phenotypes. The deletion caused a frame shift mutation (L1821fs/10) with truncation of the C-terminus after 10 missense amino acid substitutions. When expressed in HEK-293 cells for patch-clamp study, the current density of L1821fs/10 was reduced by 90% compared with WT. In addition, gating kinetic analysis showed a 5 mV positive shift in activation, a 12 mV negative shift of inactivation and enhanced intermediate inactivation, all of which would tend to reduce peak and early sodium current. Late sodium current, however, was increased in the mutated channels.
The L1821fs/10 mutation causes the most severe disruption of SCN5A structure for a naturally occurring mutation that still produces current. It has a marked loss-of-function and unique phenotype of SSS, CCD and VT with incomplete penetrance.
SCN5A encodes the voltage-dependent sodium channel α-subunit protein hNav1.5 , found predominantly in human heart muscle. This channel is responsible for large peak inward sodium current (INa) that underlies excitability and conduction in working myocardium (atrial and ventricular cells) and special conduction tissue (Purkinje cells and others), and also for late INa that influences repolarization and refractoriness. Mutations in SCN5A that increase late INa (“gain of function”) cause type 3 long QT syndrome (LQT3) , and mutations that decrease peak INa (“loss of function”) cause several arrhythmogenic syndromes including type 1 Brugada syndrome (BrS1)  and isolated cardiac conduction disease (ICCD) [4,5]. More rarely, congenital sick sinus syndrome (SSS) has also been linked to mutations in SCN5A [6,7].
To date, at least 4 mutations in the cardiac sodium channel have been identified whereby the same mutation precipitated multiple different disease phenotypes [7–10]. An insertion of an aspartic acid residue in the C-terminus of SCN5A (1795insD) can result in either BrS1 or LQT3 . The mutation of glycine to arginine (G1406R in the original publication, G1408R by present numbering) in the DIII/S5 to DIII/S6 region resulted in either BrS1 or ICCD in several families . The deletion of a lysine in the III–IV linker of SCN5A (ΔK1500) is associated with BrS1, LQT3 and ICCD . A replacement of glutamic acid by the basic residue lysine (E161K) in the DI/S2 transmembrane region contributes to SSS, CCD and BrS1 . The mechanisms by which the same or similar biophysical phenotype such as loss of function can cause multiple distinct clinical phenotypes are unknown. Discovering and characterizing additional novel mutations in families showing multiple phenotypes may provide further insight.
The sodium channel C-terminal domain (residues 1773–2016) represents a crucial structure implicated in channel inactivation modulation . Intramolecular interactions between Nav1.5 III–IV linker and C-terminal domain have been implicated in stabilization of the inactivated state. Estimated three-dimensional structure predicts that the proximal half contain six helical structures, while the distal half consists of an unstructured peptide. Functional experiments revealed no role of the distal unstructured C-terminal domain. Truncation of the predicted sixth helix of the structured C-terminal region uncouples the C-terminal from the III–IV linker . To date, no deletions in this C-terminal domain have been associated with any heritable arrhythmia syndromes. We present a unique phenotype and functional characterization of a case involving a 4 base deletion (TCTG) in SCN5A at position 5464–5467 in exon 28. This deletion produces a frameshift mutation (L1821fs/10) that deletes the terminal 195 amino acids. This severe, premature truncation includes amino acids comprising the predicted sixth helix.
Despite a severe disruption of the channel structure, L1821fs/10-SCN5A produced currents, albeit reduced, with predominantly loss of function but also an increase in late INa relative to peak. The corresponding clinical phenotype was complex with SSS, CCD and recurrent monomorphic ventricular tachycardia (VT).
Comprehensive open reading frame/splice site mutational analysis of SCN5A was performed using denaturing high performance liquid chromatography (DHPLC), and direct DNA sequencing as previously described [12,13]. The study was performed according to the terms required by the Research Ethics Committee of the National Institute of Cardiology “Ignacio Chávez”, México City and the Mayo Foundation Institution Review Board written informed consent was obtained from all participants. The investigation also conforms with the principles outlined in the Declaration of Helsinki” (Cardiovascular Research 1997; 35:2–4).
L1821fs/10 was created by site-direct mutagenesis (mutagenesis kit from Stratagene®) using a PCR technique. The appropriate nucleotide changes for L1821fs/10 was engineered into the most common splice variant of human cardiac voltage-dependent Na channel SCN5A/hNav1.5 [lacking a glutamine at position 1077, we note as Q1077del (Genbank accession No.AY148488)] in the pcDNA3 vector (Invitrogen; Carlsbad, CA). Integrity of the constructs was verified by DNA sequencing. WT and mutant channels were transiently expressed in HEK-293 cells for functional study as described previously [14,15].
Macroscopic INa was measured using a standard whole-cell patch clamp method at a temperature of 22°–24°C. Details have been previously published . The extracellular (bath) solution contained, in mmol/L, NaCl 140, KCl 4, CaCl2 1.8, MgCl2 0.75 and HEPES 5 (pH 7.4 set with NaOH). The pipette solution contained, in mmol/L, CsF 120, CsCl 20, EGTA 2 and HEPES 5 (pH 7.4 set with CsOH). Pipettes had resistances between 1.0 and 2.0 MΩ when filled with recording solution. The data were acquired using pClamp 8.2 (Axon Instruments Inc. Union City, CA) and analyzed using Clampfit (Axon Instruments Inc.). The standard voltage clamp protocols are presented with the data and as described in detail previously .
Forty-eight hours post-transfection, HEK-293 cells transfected with WT and mutant plasmid DNA containing the FLAG epitope, which introduced between S1 and S2 in domain I of hNaV1.5 were used for the immunocytochemistry experiments using confocal microscopy. Transfected and nontransfected HEK-293 cells were fixed with a freshly prepared mixture of methanol: acetone (1:1) for 1 minute. Other details of the procedure have been previously published .
Data are shown as symbols with standard error of the mean (S.E.M.). Determinations of statistical significance were performed using One-way ANOVA for comparisons of two groups. A p value of < 0.05 was considered statistically significant. Curve fits are down using pClamp 8.2 (Axon Instruments). Non-linear curve fitting is performed with Origin 6.0 (Microcal Software).
A 12-year-old male presented with fetal bradycardia and no family history of cardiac disease or sudden death. He was born by C-section without any complication during the prenatal period. At one year of age, right bundle branch block and atrial flutter were documented (Figure 1A, 1B). At 3 years, atrial flutter ablation was performed; sinus bradycardia and pauses up to 3.72 seconds with junctional escapes were observed after the procedure (Figure 1C). At 5 years of age, monomorphic ventricular tachycardia (VT) was documented (Figure 1D) and treated with lidocaine in the emergency room converting to a very slow junctional escape rhythm. A DDDR pacemaker was implanted and amiodarone treatment initiated. Two echocardiograms and a CT scan were performed ruling out structural heart disease. Clinical electrophysiological study showed a prolonged H-V interval (100 ms, normal < 50 ms) and second-degree 4:3 atrioventricular (AV) conduction during atrial pacing (Figure 1E). Radiofrequency ablation of a monomorphic VT with RBBB morphology and a cycle length of 350 ms (Figure 1F) were attempted, without success. A pharmacological test was performed with propafenone (2 mg/kg) to unmask BrS; severe QRS widening was induced without significant changes in repolarization. The patient developed transient asystole with an acute rise in the pacing threshold, reversing spontaneously after a few seconds. The ECG was otherwise negative. The patient was clinically diagnosed as congenital SSS, CCD and recurrent monomorphic VT and LQTS was also considered a possible contributor to the patient’s phenotype based on QT prolongation (QTc 550 ms). However, given that torsades de pointes have never been documented, the QT prolongation might be attributed clinically to the combination of amiodarone therapy and the severity of CCD, both of which can cause QT prolongation.
A 4 base pair deletion (TCTG) at position 5464–5467 in exon 28 of SCN5A was elucidated (bottom panel of Figure 2B). This deletion produced a frameshift mutation annotated as L1821fs/10 indicating that the final normal amino acid in the protein is the leucine (L) at position 1821 followed by 10 “scrambled” (frame-shifted) amino acids before truncating prematurely. Thus, the gene product from this mutant allele ends at amino acid 1831 resulting in a severe truncation of the C-terminus.
The mutation was confirmed in the proband’s mother as well (Figure 2C) who was asymptomatic. One female sibling and one female cousin were genotype positive but had a mild clinical phenotype of bradycardia with intraventricular conduction delay and abnormal repolarization on ECG with 1st degree AV block respectively. Genotype carriers in an aunt, uncle, and grandmother were apparently asymptomatic and without ECG abnormalities. Thus, despite a severe truncation deletion predicted at the molecular/genetic level, L1821fs/10 was associated with incomplete penetrance and variable expressivity.
Transfected HEK-293 cells transiently expressing the WT and L1821fs/10 mutant channels were voltage clamped after 24h and 48h incubation and showed reduced currents (Figure 3A and 3B) for the mutant channel compared to WT. Summary of INa density in 3C showed that after 24h of incubation, the mutant channel had dramatic reduction of INa density (−4 ± 2 pA/pF, n=14) compared to WT (−302 ± 48 pA/pF, n=21, p value < 0.0001). Although the current density of L1821fs/10 was increased after 48h of incubation (−22 ± 8 pA/pF, n=33), this was still much less than WT levels after 24h of incubation (p< 0.001). Note that among the 47 experiments performed for L1821fs/10, only 5 cells had macroscopic currents after 48h of incubation sufficiently large to do kinetic analysis. Extending the incubation time to 72h did not increase the current for L1821fs/10. Incubation with the antiarrhythmic drug (mexiletine 500μM, quinidine 100μM and cisapride 10μM), and other attempts to restore the current density in L1821fs/10 such as incubation at low temperature and co-expression with the β1 and β3 subunits, known to increase INa density of some expression defect mutants, failed to restore the current densities for L1821fs/10 (data not shown). When L1821fs/10 was co-expressed with WT (0.75 μg DNA each, n=10 cells at 24h) the current amplitude was about 55% of that in WT alone (1.5 μg DNA n=10 cells) consistent with haploinsufficiency in the heterologous system, and no dominant negative effect.
Late INa, for L1821fs/10 mutant channels was increased in response to long (700 ms) depolarizing steps to various test potentials from a holding potential of −140 mV (Figure 3D). After leak subtraction, late INa was measured as the mean between 600 and 700 ms after the initiation of the depolarization and normalized by dividing by the maximum peak amplitude of INa from the current-voltage relationship (Figure 3E, ,2.9%2.9% ± 0.3%, n=5, for L1821fs/10 and 0.8% ± 0.1%, n=5, for WT, p<0.001). Late INa were confirmed using 1 μM STX (data not shown).
Table 1 and Figure 4 summarize voltage-dependent gating for L1821fs/10 and WT channels. The voltage-dependent activation mid-point was significantly positive shifted by 5 mV, and the slope factor was larger for L1821fs/10 compared to WT (Table 1 and Figure 4A). The steady state inactivation mid-point was significantly negative shifted by 12 mV, and slope factor was larger for L1821fs/10 compared to WT (Table 1 and Figure 4B). Slower time constants of recovery were observed for L1821fs/10 compared to WT (Table 1 and Figure 4C). The mutant channel showed enhanced intermediate inactivation compared with WT for prepulse durations longer than 50 ms (Figure 4D). The voltage dependence of decay of the current traces (Figure 5) showed that the mutation lost the voltage dependence of decay, with the decay being slower than WT over the range of potentials >−30 mV where inactivation dominates the decay phase relative to deactivation at more negative potentials.
The voltage dependence of time to peak amplitude of INa, time constants of INa decay, and fractional amplitude of INa decay were measured at the test potentials of −50, −40, −30, −20, −10, 0, 10, 20, and 30 mV (summary data in Figure 5). The INa for L1821fs/10 tended to peak earlier than the WT at −50 mV, but tended to peak later at more depolarized potentials (Figure 5A). Like time to peak, the fast and slow decay time constants were faster at −50 mV for L1821fs/10 but were significantly slower for the depolarization > −20 mV compared with WT channels (Figure 5B, 5C). This result of slower inactivation may account for the gain the late INa for mutated channel.
Transfected or non-transfected HEK-293 cells were labeled with anti-FLAG antibody (Figure 6A: in nonpermeabilized cells, and left panel of Figure 6B: in permeabilized cells) to mark the location of channels. Calnexin is a chaperon protein present in the endoplasmic reticulum and was used as an endoplasmic reticulum marker with anti-calnexin antibody labeling (middle panel of Figure 6B) to show co-localization of both markers. Forty-eight hours post-transfection, both WT and mutant had a similar peripheral SCN5A localization. Thus L1821fs/10 reached the plasma membrane ruling out a loss-of-function mechanism involving defective trafficking as the sole defect.
The mutant channel L1821fs/10 exhibited a loss of function biophysical phenotype resulting in a mixed clinical syndrome characterized by SSS, CCD and recurrent VT. This discovery provides the fifth SCN5A mutation associated with multiple clinical phenotypes [7–10]. Generally, mutations in SCN5A that cause a decrease in peak INa density appear to be associated with BrS1/CCD [3,5,9]. This can occur by two general mechanisms: a decrease in channel expression or a change in channel kinetics that tends to decrease peak INa.
How does a loss of function mutation in some patients cause BrS1, SSS, or CCD, or in some patients, mixtures of these syndromes? It has been proposed  that a rightward shift in the voltage dependence of the mutant sodium channel activation curve, as found with L1821fs/10 (Figure 4A) is a common feature of CCD [5,9,10,19]. The slower time to peak INa (Figure 5A) in our study indicates that the mutant channels require a more positive membrane potential to fully activate and more time to reach the membrane potential at which the maximum current amplitude occurs than WT channels, which might manifest on the ECG in a widening of the QRS in CCD [5,10,18,19]. However, the mechanisms by which the proband showed SSS and monomorphic VT but not BrS1, are unknown.
The mutation, L1821fs/10, also showed decreased INa with slow decay and a noninactivating component (late INa) in those cells that had measurable macroscopic currents (Figure 3 and and5),5), similar to the gain-of-functional molecular phenotype of LQT3. This introduces seemingly paradoxical terminology, how can a mutation cause both “loss of function” and “gain of function”. The key distinction is timing, “loss of function” generally refers to a loss of peak or early INa, and “gain of function” generally refers to late INa. For L1821fs/10 mutant channels, peak INa is decreased but late INa is increased relative to peak INa.
This combination of molecular phenotypes has been reported previously for ΔK1500, a deletion of a lysine in the III–IV linker of SCN5A, which is associated with BrS1, LQT3 and CCD  and for 1795insD, an insertion of aspartic acid in the C-terminal domain of SCN5A, which causing both LQT3 and BrS1 . In heterologous expression system, L1821/fs10 also has very similar molecular phenotype of both loss-of-function and gain-of-functional sodium channel features with a previously reported mutation of a methionine to a leucine (M1766L) from a patient with LQT3 . When expressed, M1766L showed a profound trafficking defect, but had increased late INa and the expression defect was “rescued” by mexiletine, or, as in the patient, the mutation was rescued when expressed in the context of a sodium channel containing the common polymorphism H558R .
In this case, the L1821fs/10 proband did have a prolonged QTc, but it was in the presence of right bundle branch block, and his VT was monomorphic, not torsades. Although this mutation has a biophysical phenotype that could potentially cause LQT3, there is no evidence for LQT3 in the patient. The reduced current expression was not restored with mexiletine, quinidine, or cisapride, drugs previously reported to rescue expression defective SCN5A mutations [17,20–22]. Co-expression with β subunits and low temperature of incubation, which increased some expression levels in M1766L , did not increase expression level for L1821fs/10. Further, the cell surface staining with anti-Flag antibody (Figure 6A: in nonpermeabilized cells) and co-labeling with anti-Flag antibody and anti-calnexin antibody (Figure 6B: in permeabilized cells) indicated that the L1821fs/10 channel reached the plasma membrane. From this we conclude that the markedly decreased peak INa of L1821fs/10 was mainly caused by biophysical abnormalities.
Recently, the Nav1.5 C-terminal domain has attracted considerable attention for having a direct structural role in the control of channel inactivation in the studies by Kass and colleagues [11, 23–25]. They postulate that the proximal part of the C-terminal domain is a critical structure for stabilizing the inactivated state of the channel during prolonged depolarization. The naturally occurring C-terminal truncation mutation, L1821/fs10 (Q1832 stop) in our study not only showed results similar to the experimental mutation (S1885 stop) designed to test biophysical function of the C-terminus, namely increased late INa resulting from channels failing to inactivate during prolonged depolarization and slows the channel’s recovery from inactivation, as well as shifts the steady state inactivation curve in the hyperpolarizing direction [23–25].
It is interesting that selected family members preferentially presented the phenotype with different degrees of severity, the marked clinical phenotype suffered by the proband contrasts sharply with the lack of symptoms in four carriers in the preceding generations and the very mild clinical phenotypes observed in a sister and a cousin. In general, but with one exception, the asymptomatic carriers were females, which agrees with a male predominance for loss-of-function SCN5A mutations causing BrS1. The mechanisms for this variability in this instance and in general are incompletely understood. We previously reported that the two common polymorphisms (H558R and S524Y) caused a profound expression defect in the Q1077 background, but not the Q1077del background [16, 26], and that H558R can modify expression of an arrhythmia causing mutation M1766L . These and other polymorphisms were absent in these patients. Loss of function for the BrS1 mutation G1406R was more severe in the Q1077 background . Although mRNA for Q1077 and Q1077del are present in equal ratios in different individuals , should the protein levels vary this could account for clinical variability. Other unknown genetic, developmental, or acquired abnormalities may also account for clinical variability. For example structural abnormalities such as increased fibrosis has been shown to interact with decreased INa  to affect excitability and conduction. Understanding the mechanisms for variability of the clinical phenotype may lead to improved understanding of pathogenesis and of possible therapy.
It is important to emphasize that these studies in heterologous systems, like previous work in the field, may not reflect what occurs in the myocytes, which has additional subunits and interacting proteins. Also these clones do not contain the introns or promoters that might affect expression in the myocytes. These experiments in heterologous systems only suggest a possible biophysical phenotype that may lead to the clinical syndromes, but further studies in more integrated systems is necessary to describe the full pathogenetic pathway.
This work was supported by an American Heart Association (AHA), Greater Midwest Affiliate Postdoctoral Fellowship to BH.T., AHA Established Investigator Award to M.J.A., and NIH grant HD42569 to M.J.A and HL71092 to J.C.M.. We are particularly indebted to family members for their participation in this study.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.