Divergent Biophysical Defects Caused by Mutant Sodium Channels in Dilated Cardiomyopathy With Arrhythmia
Mutations in SCN5A encoding the principal Na+ channel α-subunit expressed in human heart (NaV1.5) have recently been linked to an inherited form of dilated cardiomyopathy with atrial and ventricular arrhythmia. We compared the biophysical properties of 2 novel NaV1.5 mutations associated with this syndrome (D2/S4 – R814W; D4/S3 – D1595H) with the wild-type (WT) channel using heterologous expression in cultured tsA201 cells and whole-cell patch-clamp recording. Expression levels were similar among WT and mutant channels, and neither mutation affected persistent sodium current. R814W channels exhibited prominent and novel defects in the kinetics and voltage dependence of activation characterized by slower rise times and a hyperpolarized conductance-voltage relationship resulting in an increased “window current.” This mutant also displayed enhanced slow inactivation and greater use-dependent reduction in peak current at fast pulsing frequencies. By contrast, D1595H channels exhibited impaired fast inactivation characterized by slower entry into the inactivated state and a hyperpolarized steady-state inactivation curve. Our findings illustrate the divergent biophysical defects caused by 2 different SCN5A mutations associated with familial dilated cardiomyopathy. Retrospective review of the published clinical data suggested that cardiomyopathy was not common in the family with D1595H, but rather sinus bradycardia was the predominant clinical finding. However, for R814W, we speculate that an increased window current coupled with enhanced slow inactivation and rate-dependent loss of channel availability provided a unique substrate predisposing myocytes to disordered Na+ and Ca2+ homeostasis leading to myocardial dysfunction.
Mutations in SCN5A, the gene encoding the principle α-subunit of the human cardiac voltage-gated sodium channel NaV1.5, cause inherited susceptibility to ventricular tachyarrhythmia (congenital long-QT syndrome [LQTS], Brugada syndrome [BrS]),1–3 familial heart block with or without sinus node dysfunction,4–6 and other more complex phenotypes with features of both arrhythmia susceptibility and impaired atrioventricular conduction.7,8 Subclinical expression of SCN5A mutations may also manifest as drug-induced arrhythmias,9,10 sudden infant death syndrome,11,12 and increased arrhythmia risk in specific populations.13
Conventional wisdom in the era of molecular cardiology predicts that mutations affecting plasma membrane ion channels will primarily predispose to rhythm disorders, but will not directly affect contractile function. A possible exception to this paradigm was provided by the discovery of SCN5A mutations associated with familial dilated cardiomyopathy associated with atrial and ventricular arrhythmias (DCM-arrhythmia syndrome). In 2004, McNair et al14 reported that a missense mutation, SCN5A-D1275N, segregated with a phenotype including DCM of variable expression, abnormal atrioventricular conduction, sinus node dysfunction, atrial and ventricular tachyarrhythmias in 4 generations of a kindred originally described by Greenlee and colleagues.15 The mutation was also discovered by Olson and colleagues, who independently investigated this family.16
Motivated by the association of familial DCM-arrhythmia with SCN5A-D1275N, Olson et al surveyed an additional 156 unrelated DCM probands for SCN5A variants; this survey led to the discovery of other mutations including 1 frameshift (insertion of TG at c.2550) and 3 missense alleles (T220I, R814W, D1595H).16 With the exception of R814W, which was de novo, the other mutations segregated in families with a complex and heterogenous phenotype characterized by variable expression of DCM, sick sinus syndrome, atrial and ventricular arrhythmias. Transmission was consistent with a pattern of autosomal dominant inheritance with incomplete penetrance. In those families, 43% carriers exhibited early-onset atrial fibrillation (AF) or atrial flutter, whereas 65% displayed impaired left ventricular contractile function.16 Cardiomyopathy, however, did not appear to be secondary to AF-associated tachycardia.
In the reports by McNair et al14 and Olson et al,16 no functional studies of the mutant sodium channels were performed, but inferences were drawn from prior work. Specifically, T220I and D1275N have previously been studied in heterologous expression systems because of their associations with congenital sick sinus syndrome17 and familial atrial standstill,18 respectively. Furthermore, even though D1595H affects the same residue previously known to be mutated to asparagine in familial heart block (D1595N),6 it is not clear whether a histidine substitution at this position would manifest similar biophysical abnormalities.
Here we report the functional characterization of 2 novel SCN5A mutations associated with the DCM-arrhythmia syndrome, R814W and D1595H. Our findings reveal divergent biophysical defects for these 2 mutations, suggesting that different electrophysiological derangements can lead to the same complex phenotype or that this syndrome is more heterogenous than originally proposed, an idea supported by retrospective review of the clinical phenotypes. Further, the unique functional properties of R814W suggest a plausible mechanism for myocardial dysfunction based on predicted abnormalities in internal Na+ and Ca2+ homeostasis.
Materials and Methods
Mutagenesis and Heterologous Sodium Channel Expression
The mutations R814W and D1595H were engineered in a recombinant human heart sodium channel α subunit cDNA (hH1)19 and transiently coexpressed in human embryonic kidney tsA201 cells with the human β1 subunit (hβ1). Additional details of these methods may be found in the supplemental materials (available online at http://circres.ahajournals.org).
Sodium currents were recorded using standard whole-cell patch clamp technique.20 Experiments were performed at room temperature (≈22°C), except during deactivation protocols, when the recording bath temperature was decreased to 10.5±0.1°C by water flow through a thermally conductive recording chamber using a CL-100 temperature controller combined with an SC-20 in-line solution cooler (Warner Instruments Inc). Additional details of these methods may be found in the supplemental materials.
Electrophysiological data were analyzed using Clampfit 9.2 (Axon instruments Inc) and Origin 7.0 (Microcal Software Inc). Symbols with error bars represent means±SEM; error bars are occasionally smaller than data symbols. Statistical significance was determined using unpaired Student t test or 1-way ANOVA followed by a Tukey test.
Whole-cell capacitance was assessed by integrating the capacitive transient elicited by a 10-mV voltage step from −120 mV to −110 mV. Apparent gating charge qRT associated with 10% to 90% rise time was derived from first-order exponential decay fit of rise times across a 100-mV range. Steady-state activation (GV), fast inactivation (h∞), and slow inactivation (S∞) curves were obtained by fitting normalized peak conductances or currents with single Boltzmann functions, equation
where V is the membrane potential, V0.5 the half-activation or half-inactivation voltage, q the apparent gating charge, F Faraday’s constant, R the universal gas constant, and T the temperature in Kelvin, with RT/F=25 mV at 22°C. Time constant of deactivation, time constants of fast and slow inactivation (τh and τS), as well as time constant of recovery from fast inactivation (τRFI) were all derived from first-order relaxations, whereas time constants of recovery from slow inactivation (τ1 and τ2) were obtained from second-order exponential decays. The apparent gating charges associated with deactivation and with recovery from fast inactivation (qdeact, qRFI) were obtained from fitting of the respective time constant (τdeact, τRFI) values at different potentials with a single exponential growth function, equation
Persistent current was evaluated by a 100-ms depolarization from −120 mV to either −60 or −10 mV in the absence and presence of 20 μmol/L tetrodotoxin (TTX, Sigma-Aldrich Inc), a potent extracellular sodium channel blocker. Digital subtraction of the time-averaged responses under these 2 conditions yielded a small TTX-sensitive sodium current, from which the amplitude of the last 10 ms of the 100-ms pulse was normalized to the earlier peak amplitude to indicate the percentage of persistent current.
We determined the functional properties of 2 SCN5A mutations associated with the DCM-arrhythmia syndrome with the goal of elucidating the molecular and biophysical defects underlying this complex disorder. The 2 mutations that we characterized are located in distinct regions of the human cardiac sodium channel protein Nav1.5. R814W affects the third conserved arginine residue counting from the extracellular end of the S4 segment in domain 2 (D2/S4), 1 of the 4 sodium channel voltage-sensors. D1595H alters a conserved aspartate residue near the cytoplasmic end of the D4/S3 segment.
Figure 1A through 1C illustrates representative whole-cell currents from cells heterologously expressing WT, R814W, and D1595H channels. Peak current densities from the mutant channels were comparable to those from WT (data not shown) suggesting that neither mutation exerted a major effect on cell surface expression. Furthermore, neither mutation caused an increased level of TTX-sensitive persistent current measured at either −10 or −60 mV (data not shown) as previously observed with most SCN5A mutations causing LQTS.21 However, the mutants exhibited striking differences in the kinetics of either activation (R814W) or inactivation (D1595H) compared with WT channels. These differences are readily appreciated by comparisons of superimposed normalized current traces (Figure 1D and 1E). At −60 mV, R814W channels exhibited a slower activation time course than either WT or D1595H (Figure 1D). By contrast, comparison of superimposed normalized current traces recorded at −20 mV indicated a slower rate of current decay for D1595H than for either WT or R814W (Figure 1E). Normalized current-voltage relationships also demonstrated that R814W, but not D1595H, exhibited a hyperpolarizing shift suggesting a difference in the voltage dependence of activation (Figure 1F).
Disrupted Activation and Deactivation for R814W
Both mutants exhibited rapid activation typical of voltage-gated sodium channels, however R814W displayed notable differences in kinetics and voltage dependence. We observed a hyperpolarizing shift in the R814W steady-state activation (G–V) curve (Figure 2A). A Boltzmann function best fitted to these data revealed reduced voltage dependence with diminished apparent gating charge, consistent with neutralization of a D2/S4 positive charge that is important for voltage-sensing (supplemental Table I).
To further assess the activation time course, we compared the time required for activating currents to increase from 10% to 90% of peak current amplitude during a range of membrane depolarizations (Figure 2B). Whereas D1595H had small yet significant differences in activation kinetics compared with WT, R814W exhibited a substantially slower onset of activation discernible as a rise time that was up to 3-fold slower at the most negative voltages. Further, R814W exhibited abnormal rise times across a wide range of conductance values (see supplemental Figure I), indicating that the slower kinetics of activation are not attributable solely to the shift in steady-state gating properties.
Collectively, these findings indicated that R814W, but not D1595H, has increased channel opening at hyperpolarized membrane voltages but also has a slower rate of entry into the activated state. Abnormal activation kinetics has not been previously reported for other disease-associated SCN5A mutations and thus is a novel feature of R814W associated with the DCM-arrhythmia syndrome.
Because the kinetics and voltage dependence of activation were altered in R814W, we tested whether deactivation was also affected. To examine deactivation, we first activated channels with a brief (≈4 to 6 ms) depolarization to −30 mV and then examined tail current decay elicited by repolarizing pulses to voltages between −140 and −70 mV at 10.5°C (Figure 2C). R814W channels exhibited significantly slower deactivation kinetics between −100 and −70 mV than either WT or D1595H (supplemental Table I), presumably reflecting a more sluggish transition between the open state and a preceding closed state in the activation pathway. Scaled superimposed tail currents at −80 mV illustrated an almost 3-fold slowing of deactivation for R814W compared with WT and D1595H (Figure 2D). Regarding the significant difference in tail current decay at −70 mV between D1595H and WT (Figure 2C), we cannot exclude the possibility that this resulted from differences in fast inactivation rather than a true difference in deactivation. The deactivation of both mutants was more voltage dependent than that of WT channels (see Figure 2 legend).
Defective Fast Inactivation for D1595H
Whereas a prominent functional effect of R814W was to perturb activation, D1595H primarily affected fast inactivation. The kinetics of inactivation were well fit by a single exponential relaxation in all 3 constructs. The voltage dependence of inactivation time constants (τh) was exponential and comparable for WT and R814W channels (Figure 3A). However the voltage dependence of D1595H inactivation time constants required a second-order exponential fit and had almost 3-fold larger time constants as well as less steep voltage dependence (Figure 3A and supplemental Table I). The voltage dependence of steady-state fast inactivation (h∞) further illustrated the divergent effects of the 2 mutants. Whereas the h∞ curve of R814W displayed a small (≈3 mV) but significant depolarizing shift, that of D1595H had a larger (≈−7 mV) hyperpolarizing shift (Figure 3B). No difference in the apparent gating charge was noted for either mutant (supplemental Table I).
Divergent effects on the recovery from fast inactivation were also observed (Figure 3C and supplemental Table I). Whereas R814W exhibited a consistently slower time course of recovery from fast inactivation compared with WT over the entire voltage range studied (−150 to −100 mV), D1595H displayed a crossover effect at −110mV with slower recovery at voltages negative to this crossover point and hastened recovery at voltages positive to this point as judged by significant differences in recovery time constants (supplemental Table I). A slightly reduced apparent gating charge associated with recovery from fast inactivation for D1595H suggested diminished voltage dependence (supplemental Table I). Altogether, these findings indicated that D1595H and to a lesser extent R814W had discernible defects in both the kinetics and voltage dependence of fast inactivation. Compared with WT, D1595H had slower rates of entry and recovery from fast inactivation but less stable steady-state availability, whereas R814W had slower recovery from inactivation and greater availability.
Increased Window Current for R814W
The large hyperpolarizing shift in the conductance-voltage relationship coupled with the small depolarizing shift in steady-state inactivation observed for R814W gives rise to an increased “window current” defined as the area beneath the point of intersection of these 2 curves. Figure 4A illustrates an expanded view of superimposed steady-state inactivation and conductance-voltage relationships for WT, R814W, and D1595H. There is a prominently increased window current for R814W as compared with both WT and D1595H. We examined the potential impact of increased window current on sodium channel function by using a voltage-ramp protocol. As illustrated in Figure 4B, only R814W channels exhibited aberrant behavior during the ramp with peak activating inward current occurring at −70 mV consistent with the effects of increased window current. This finding suggests that there is an increased range of membrane potentials where greater Na+ conductance may occur.
Enhanced Slow Inactivation and Use-Dependence for R814W
We also investigated whether R814W and D1595H affected slower forms of inactivation that may be important for regulating channel availability on a time scale of several seconds to minutes. Compared with WT, R814W exhibited significantly faster onset of and slower recovery from slow inactivation (Figure 5A and 5B; supplemental Table II). The voltage dependence of steady-state slow inactivation was not different for R814W (Figure 5C), but this mutant exhibited a significantly greater extent of slow inactivation compared with WT and D1595H (supplemental Table II). Collectively, these results indicated that R814W confers greater stability to slow-inactivated states of the channel. For D1595H, the extent, voltage dependence and onset of slow inactivation were indistinguishable from WT, but recovery from slow inactivation was intermediate between WT and R814W (Figure 5; supplemental Table II).
These findings indicated that R814W and to a lesser extent D1595H had enhanced slow inactivation, which predicts a greater loss of channel availability with accumulated exposure time to a depolarized membrane, such as occurs during repetitive activity, possibly leading to greater use-dependent reduction in peak current. We tested this hypothesis by measuring peak current in cells stimulated by trains of 30 depolarizing pulses (500 ms) to +10 mV from a holding potential of −100 mV with frequencies that simulated heart rates of 60, 100, and 115 beats per minute (bpm). Figure 6 illustrates similar degrees of use dependence for WT and D1595H, but there was a significantly greater loss of channel availability with rapid repetitive pulsing for R814W channels. For example, with simulated tachycardia at 115 bpm, there was up to 55% use-dependent loss of R814W channel availability with normalized currents recorded after the 30th pulse being 30.3±1.5% for WT (n=15), 30.4±1.9% for D1595H (n=16), and only 13.6±2.0% for R814W (n=13; P<0.001). These results further emphasize the divergent nature of the functional defects exhibited by R814W and D1595H.
During the last decade there has been growing awareness that genetic variation of SCN5A predisposes to a range of heart rhythm disorders. More recently, SCN5A mutations have been unexpectedly linked also to familial cardiomyopathy, generating an expanded view of the clinical spectrum of cardiac sodium channelopathies. In this study, we set out to determine the functional consequences of 2 SCN5A mutations (R814W, D1595H) associated with the DCM-arrhythmia syndrome to further understand the molecular defect responsible for this complex disorder.16
Divergent Functional Effects of R814W and D1595H
Our data demonstrated that these two mutations have divergent biophysical defects. Among the prominent features of R814W were the abnormal kinetics of activation and deactivation, the hyperpolarized shift of the conductance-voltage relationship, and the altered voltage dependence of activation and deactivation. These perturbations are consistent with neutralization of a conserved D2/S4 positive charge that contributes to voltage-sensing. The net physiological effects of these abnormalities can be predicted to be complex. The hyperpolarized voltage dependence of activation coupled with the depolarized steady-state inactivation suggests that R814W channels will open at more negative potentials and exhibit greater window current. The physiological significance of slowed deactivation observed for R814W channels is less certain, however it is known that abnormal deactivation of skeletal muscle sodium channels contributes to the hyperexcitability that occurs in certain myotonic disorders.22,23
In addition to the effects on activation and deactivation, R814W also exhibits enhanced slow inactivation, a gating process that is functionally distinct from fast inactivation. Slow inactivation is evoked by very long sustained membrane depolarization or by the accumulation of time in a depolarized state such as during repetitive stimulation. This process is thought to contribute to controlling channel availability over a time course of several seconds to minutes. The physiological significance of enhanced slow inactivation can be demonstrated dynamically by observing the reduced channel availability during repetitive stimulations (Figure 6). Reduced channel availability, especially at faster rates combined with the effects on activation, suggests that R814W may be associated with a rate-dependent reduction in sodium current.
By contrast, the primary defects exhibited by D1595H lie in the kinetics and voltage dependence of fast inactivation, with a less prominent effect on recovery from inactivation (accelerated rate at −100 mV). Slowing of fast inactivation coupled with faster recovery from inactivation may promote greater sodium conductance per unit time, but this may be offset by the hyperpolarized shift in steady-state channel availability. This pattern of biophysical disturbances is somewhat similar to that observed for D1595N, a mutation associated with familial heart block.6 D1595N also impaired fast inactivation but did not cause a shift in the steady-state channel availability curve. These subtle variations in functional consequences of D1595H and D1595N may account for the differences in the associated clinical phenotypes (see below).
Whereas it is relatively easy to understand why R814W interferes with activation through replacement of a conserved positive charge in a voltage-sensing segment with a more hydrophobic and bulky tryptophan residue, much less is known about the role of the D4/S3 segment in gating. In the skeletal muscle sodium channel isoform (NaV1.4), residues in the extracellular end of D4/S3 exhibit depolarization-dependent accessibility indicating voltage-dependent movements, but residues near the cytoplasmic side do not display this behavior.24 Cysteine replacement of D1420 in NaV1.4, which is equivalent to D1595 in NaV1.5, slows the time course of fast inactivation.24 Similarly, neutralization of the this conserved negatively charged residue in the cardiac channel with asparagine (D1595N) impairs fast inactivation as has been reported in association with a form of familial heart block.6 These previous observations combined with our present results suggest that the cytoplasmic end of D4/S3 is important for inactivation, and this effect may depend on side chain charge. The equivalent S3 residues in voltage-gated potassium channels participate in electrostatic interactions with positively charged amino acids in S4, and these interactions help stabilize certain channel conformations.25
Functional Divergence and Clinical Heterogeneity
One possible interpretation of our findings is that multiple electrophysiological disorders caused by mutant sodium channels can lead to the same clinical end point. Another equally plausible conclusion based on our study is that the clinical phenotypes associated with R814W and D1595H represent distinct syndromes. Closer inspection of individual mutation carriers reported by Olson et al16 illustrate this point. In the family with D1595H, only 1 of 6 proven mutation carriers had dilated cardiomyopathy whereas another presumed mutation carrier was diagnosed with heart failure at age 70 years, raising the possibility that this individual developed ventricular dysfunction for unrelated reasons (ie, phenocopy). A more prominent feature affecting 4 of 6 mutation carriers in this family was sinus bradycardia, a common finding with other SCN5A mutations.26 Only one presumed mutation carrier exhibited atrial fibrillation whereas another was asymptomatic in the sixth decade. This clinical picture is distinct from that of the single proband with the R814W mutation who presented with early onset (age 23 years) dilated cardiomyopathy, atrial fibrillation, and nonsustained ventricular tachycardia. Because this was a de novo mutation, it was not possible to construct a genotype-phenotype correlation or assess penetrance. We suggest that the phenotype associated with D1595H more closely resembles a conduction system disorder or sick sinus syndrome, and the pattern of biophysical disturbances we observed for this allele are more similar to SCN5A mutations found in subjects with these disorders6,17 than to R814W. However, it is conceivable that bradycardia may predispose to cardiomyopathy through a mechanism involving volume overload-induced hypertrophy similar to that observed in the chronic AV block canine model.27
Relationship of Sodium Channel Dysfunction to Cardiomyopathy
How are sodium channel mutations linked pathophysiologically to cardiomyopathy? In general, sodium channel dysfunction may evoke cardiac disease by 3 possible mechanisms: (1) altered membrane excitability, (2) disordered intracellular sodium homeostasis, and (3) disrupted interactions between sodium channels and other proteins. Although, much is known about the relationships of sodium channel mutations with altered membrane excitability as it pertains to ventricular arrhythmia susceptibility, it is not clear how altered membrane excitability might predispose to cardiomyopathy directly. However, indirect mechanisms involving disordered intracellular sodium homeostasis are conceivable. Passage of sodium ions through sodium channels is one of the major pathways for sodium influx in cardiomyocytes.28 Alterations in intracellular sodium concentration will affect transport pathways for other ions including Ca2+ by altered Na+/Ca2+ exchange, and intracellular pH through effects on Na+/H+ exchange.
In the case of R814W, we speculate that an increased window current (Figure 4) will cause greater Na+ influx at least transiently during diastole and contribute to a higher steady-state intracellular Na+ concentration ([Na+]i) in myocytes that carry this mutant allele.29 An elevated [Na+]i will consequently promote an increase in intracellular Ca2+ concentration by blunting the driving force for Ca2+ efflux from myocytes through Na+/Ca2+ exchange leading to increased Ca2+ loading of the sarcoplasmic reticulum, greater diastolic tension, and possibly abnormal diastolic relaxation.30 In normal myocardium under physiological conditions, [Na+]i rises at faster heart rates promoting increased Ca2+ influx and a positive force-frequency relationship (FFR).31 However, the rate-dependent reduction in peak current carried by R814W (Figure 6) implies that there may be a blunted or possibly negative FFR in myocytes carrying this mutant allele. A negative FFR is a hallmark of failing myocardium in several animal models of heart failure and in human cardiomyopathic myocytes.32 This hypothesis is also consistent with the dysfunctional internal Na+ and Ca2+ homeostasis observed in heart failure.30 Cytoplasmic Na+ overload may also predispose to osmotic swelling and dysfunction of mitochondria, a phenomenon recognized in models of myocardial ischemia-reperfusion injury.33
In summary, we have demonstrated divergent biophysical defects associated with 2 cardiomyopathy-associated mutations in the cardiac sodium channel. These observations suggest that either multiple molecular mechanisms underlie this association or that this clinical phenotype is more heterogenous than originally thought. Our results also contribute to understanding the expanding spectrum of clinical disease associated with sodium channel dysfunction and stimulate new hypotheses regarding the pathophysiological mechanisms that underlie these genetic events.
The authors thank Drs Richard Horn and Dan Roden for critical review of the manuscript.
Sources of Funding
This work was supported in part by a grant from the National Institutes of Health (NS32387 to A.L.G.).
Original received May 4, 2007; resubmission received September 20, 2007; revised resubmission received November 13, 2007; accepted November 15, 2007.
Wang Q, Shen J, Li Z, Timothy K, Vincent GM, Priori SG, Schwartz PJ, Keating MT. Cardiac sodium channel mutations in patients with long QT syndrome, an inherited cardiac arrhythmia. Hum Mol Genet. 1995; 4: 1603–1607.
Chen Q, Kirsch GE, Zhang D, Brugada R, Brugada J, Brugada P, Potenza D, Moya A, Borggrefe M, Breithardt G, Ortiz-Lopez R, Wang Z, Antzelevitch C, O’Brien RE, Schulze-Bahr E, Keating MT, Towbin JA, Wang Q. Genetic basis and molecular mechanism for idiopathic ventricular fibrillation. Nature. 1998; 392: 293–296.
Wang DW, Viswanathan PC, Balser JR, George AL Jr, Benson DW. Clinical, genetic, and biophysical characterization of SCN5A mutations associated with atrioventricular conduction block. Circulation. 2002; 105: 341–346.
Bezzina C, Veldkamp MW, van Den Berg MP, Postma AV, Rook MB, Viersma JW, van Langen IM, Tan-Sindhunata G, Bink-Boelkens MT, Der Hout AH, Mannens MM, Wilde AA. A single Na+ channel mutation causing both long-QT and Brugada syndromes. Circ Res. 1999; 85: 1206–1213.
Makita N, Horie M, Nakamura T, Ai T, Sasaki K, Yokoi H, Sakurai M, Sakuma I, Otani H, Sawa H, Kitabatake A. Drug-induced long-QT syndrome associated with a subclinical SCN5A mutation. Circulation. 2002; 106: 1269–1274.
Liu K, Yang T, Viswanathan PC, Roden DM. New mechanism contributing to drug-induced arrhythmia: rescue of a misprocessed LQT3 mutant. Circulation. 2005; 112: 3239–3246.
Splawski I, Timothy KW, Tateyama M, Clancy CE, Malhotra A, Beggs AH, Cappuccio FP, Sagnella GA, Kass RS, Keating MT. Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science. 2002; 297: 1333–1336.
McNair WP, Ku L, Taylor MR, Fain PR, Dao D, Wolfel E, Mestroni L. SCN5A mutation associated with dilated cardiomyopathy, conduction disorder, and arrhythmia. Circulation. 2004; 110: 2163–2167.
Groenewegen WA, Firouzi M, Bezzina CR, Vliex S, van Langen IM, Sandkuijl L, Smits JP, Hulsbeek M, Rook MB, Jongsma HJ, Wilde AA. A cardiac sodium channel mutation cosegregates with a rare connexin40 genotype in familial atrial standstill. Circ Res. 2003; 92: 14–22.
Gellens ME, George AL, Chen L, Chahine M, Horn R, Barchi RL, Kallen RG. Primary structure and functional expression of the human cardiac tetrodotoxin-insensitive voltage-dependent sodium channel. Proc Natl Acad Sci U S A. 1992; 89: 554–558.
Nguyen TP, Horn R. Movement and crevices around a sodium channel S3 segment. J Gen Physiol. 2002; 120: 419–436.
Veldkamp MW, Wilders R, Baartscheer A, Zegers JG, Bezzina CR, Wilde AA. Contribution of sodium channel mutations to bradycardia and sinus node dysfunction in LQT3 families. Circ Res. 2003; 92: 976–983.
de Groot SH, Schoenmakers M, Molenschot MM, Leunissen JD, Wellens HJ, Vos MA. Contractile adaptations preserving cardiac output predispose the hypertrophied canine heart to delayed afterdepolarization-dependent ventricular arrhythmias. Circulation. 2000; 102: 2145–2151.
Bers DM, Barry WH, Despa S. Intracellular Na+ regulation in cardiac myocytes. Cardiovasc Res. 2003; 57: 897–912.
Despa S, Islam MA, Weber CR, Pogwizd SM, Bers DM. Intracellular Na+ concentration is elevated in heart failure but Na/K pump function is unchanged. Circulation. 2002; 105: 2543–2548.
Pieske B, Houser SR. [Na+]i handling in the failing human heart. Cardiovasc Res. 2003; 57: 874–886.
Pieske B, Maier LS, Piacentino V III, Weisser J, Hasenfuss G, Houser S. Rate dependence of [Na+]i and contractility in nonfailing and failing human myocardium. Circulation. 2002; 106: 447–453.