are associated with the Brugada syndrome

Transcription

1 Cardiovascular Research 44 (1999) locate/ cardiores locate/ cardiores Human SCN5A gene mutations alter cardiac sodium channel kinetics and q are associated with the Brugada syndrome a b,c,g, a Martin B. Rook, Connie Bezzina Alshinawi *, W. Antoinette Groenewegen, f d a Isabelle C. van Gelder, Antoni C.G. van Ginneken, Habo J. Jongsma, b c,e,h Marcel M.A.M. Mannens, Arthur A.M. Wilde a Department of Medical Physiology, Utrecht University, Utrecht, The Netherlands b Department of Clinical Genetics, Academic Medical Center, Amsterdam, The Netherlands c Department of Experimental and Molecular Cardiology, Academic Medical Center, Amsterdam, The Netherlands d Department of Physiology, Academic Medical Center, Amsterdam, The Netherlands e Heart Lung Institute, University Hospital Utrecht, Utrecht, The Netherlands f Department of Cardiology, University Hospital Groningen, Groningen, The Netherlands g Laboratory of Molecular Genetics, Department of Pathology, The Medical School, University of Malta, Msida, Malta h The Interuniversity Cardiology Institute of The Netherlands Received 15 April 1999; accepted 15 September 1999 Abstract Background: Primary dysrhythmias other than those associated with the long QT syndrome, are increasingly recognized. One of these are represented by patients with a history of resuscitation from cardiac arrest but without any structural heart disease. These patients exhibit a distinct electrocardiographic (ECG) pattern consisting of a persistent ST-segment elevation in the right precordial leads often but not always accompanied by a right bundle branch block (Brugada syndrome). This syndrome is associated with a high mortality rate and has been shown to display familial occurrence. Methods and results: Pharmacological sodium channel blockade elicits or worsens the electrocardiographic features associated with this syndrome. Hence, a candidate gene approach directed towards SCN5A, the gene encoding the a-subunit of the cardiac sodium channel, was followed in six affected individuals. In two patients missense mutations were identified in the coding region of the gene: R1512W in the DIII DIV cytoplasmic linker and A1924T in the C-terminal cytoplasmic domain. In two other patients mutations were detected near intron/ exon junctions. To assess the functional consequences of the R1512W and A1924T mutations, wild-type and mutant sodium channel proteins were expressed in Xenopus oocytes. Both missense mutations affected channel function, most notably a 4 5 mv negative voltage shift of the steady-state activation and inactivation curves in R1512W and a 9 mv negative voltage shift of the steady-state activation curve in A1924T, measured at 228C. Recovery from inactivation was slightly prolonged for R1512W channels. The time dependent kinetics of activation and inactivation at 220 mv were not significantly affected by either mutation. Conclusions: Two SCN5A mutations associated with the Brugada syndrome, significantly affect cardiac sodium channel characteristics. The alterations seem to be associated with an increase in inward sodium current during the action potential upstroke Elsevier Science B.V. All rights reserved. Keywords: Ventricular arrhythmias; Sudden death; Ion channels; Genetic code 1. Introduction q M.B.R. and C.B.A. contributed equally to this work. *Corresponding author. Academic Medical Center, University of Amsterdam, Department of Clinical and Experimental Cardiology, P.O. Box 22700, 1100 DE Amsterdam, The Netherlands. Tel.: ; fax: address: c.alshinawi@amc.uva.nl (C. Bezzina Alshinawi) Primary dysrhythmias in disease entities other than the long QT syndrome (LQTS), the paradigm for primary familial cardiac arrhythmias [1,2], although still very Time for primary review 34 days / 99/ $ see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S (99)

2 508 M.B. Rook et al. / Cardiovascular Research 44 (1999) uncommon, are increasingly diagnosed. Patients with a 2.2. SSCP analysis and direct sequencing of PCR history of aborted sudden death and a distinct electro- products cardiographic (ECG) pattern, consisting of persistent STsegment elevation in the right precordial leads in combina- Genomic DNA was extracted from peripheral blood tion with a right bundle branch block (RBBB), in the lymphocytes using standard protocols. All 28 exons of absence of any structural heart disease, may represent SCN5A were amplified by PCR using intronic primers another cardiac ion channel disorder [3]. The syndrome designed by Wang et al. [14], and analyzed by single may display a familial occurrence [3], seems to be stranded conformation polymorphism (SSCP) analysis on associated with a high mortality rate [4], and is increasing- 12.5% non-denaturing polyacrylamide gels (Pharmacia ly recognized [4 7]. Very recently, mutations in the Biotech) run at 58C and 158C as described in GENEGEL SCN5A gene which encodes the a-subunit of the fast Excel protocols (Pharmacia Biotech). PCR products precardiac sodium channel, were shown to co-segregate with senting an aberrant conformer were re-amplified from electrocardiographically affected individuals within small genomic DNA, purified by micro-spin chromatography families in which a high incidence of sudden cardiac death (Pharmacia Biotech) and sequenced in both directions by was present [8]. the fluorescent dideoxy chain termination method on an Since pharmacological sodium channel blockade elicits ABI 377 (Applied Biosystems) DNA sequencer. or worsens the electrocardiographic features associated with this syndrome [6,9,10], a candidate gene approach 2.3. Functional expression directed towards SCN5A, was followed in six individuals with the typical electrocardiographic features outlined Mutant sodium channel cdnas were prepared by above. In two patients, intronic mutations which could mutagenesis on the double-stranded psp64t-hh1(sp) plasaffect splicing were identified. In two other patients a mid [15,16], a kind gift from A.L. George Jr., (Vanderbilt missense mutation was identified in the coding region of University, Nashville, TN, USA) using the QuikChangeE the gene. One was a substitution of arginine 1512 by (Stratagene) site-directed mutagenesis kit and the followtryptophan (R1512W) in the highly conserved DIII DIV ing oligonucleotides: 59GAAGCCCATCCCATGGCCCCcytoplasmic linker, a domain in which mutations associ- TGAACAAG39 (sense) and 59CTTGTTCAGGGGCCATated with LQTS have been identified [11]. The other was a GGGATGGGCTTC39 (antisense) for the R1512W mutant, substitution of alanine 1924 by threonine (A1924T) in the and, 59CGCTCTTTGAAGCATACCTCCTTCCTCTT- C-terminal cytoplasmic domain, in which very recently CC39 (sense) and 59GGAAGAGGAAGGAGGTATGCtwo mutations have been reported that are involved in the TTCAAAGAGCG39 (antisense) for the A1924T mutant. LQT-3 syndrome [12,13]. Both missense mutations affect- Mutant clones were identified by restriction endonuclease ed channel function. In particular, they caused a negative analyses (Nco I site created by R1512W; Sph I site voltage shift of the steady-state activation curve when abolished by A1924T). A Kpn I-Bst EII fragment (nt expressed in Xenopus oocytes ) and a Kpn I-Sfi I fragment (nt ), in the case of the R1512W and A1924T mutant constructs, respectively, was subcloned into the wild-type psp64t- 2. Methods hh1(sp) plasmid, and mutant inserts and ligation regions were analyzed fully by nucleotide sequence analysis on an 2.1. Phenotypic analysis ABI 377 DNA sequencer to ensure that the clones selected were free of polymerase errors. Wild-type and mutant Informed consent for genetic studies was obtained from constructs were linearized with Xba I and crnas were all individuals in accordance with local institutional ethical synthesized using the mmessage mmachine kit (Ambion). committee guidelines. Brugada syndrome was diagnosed when the typical electrocardiographic features at least Electrophysiology mm ST-segment elevation in leads V 1 V2 or V 1 V 3, were present in the absence of structural heart disease [3]. An Stage V VI Xenopus oocytes were isolated and injected apparent right ventricular conduction delay in the ECG with ng crna according to standard methods [15]. was not considered essential for the syndrome. The oocytes were kept in a modified Barths s solution [17] Family members were not screened because it was containing in mm: NaCl 88, NaHCO 2.4, KCl 1.0, CaCl 3 2 considered ethically unjustified to identify individuals with 0.7, MgSO4 0.8 and HEPES NaOH 15, ph 7.6. The same a potentially malignant genotype without the possibility to solution was used in the electrophysiological experiments. offer them appropriate treatment. In our country ICD Voltage clamp experiments were performed 2 3 days after resource limitations are such that a prophylactic implanta- injection, using a Geneclamp 500 two-electrode voltage tion is currently not accepted. Hence, eventually identified clamp amplifier (Axon Instruments). Steady-state activaasymptomatic gene carriers cannot be offered appropriate tion and inactivation parameters were determined using treatment. protocols similar to those published previously [11].

3 M.B. Rook et al. / Cardiovascular Research 44 (1999) Fig. 1. Representative recordings (eight leads) of the ECG of patient 1, several weeks after a successful resuscitation. The right panels show the effect of oral flecainide treatment (tdd 100 mg). Note the increase in ST-segment elevation in particularly lead V1. Calibration is standard (lower right corner).

4 510 M.B. Rook et al. / Cardiovascular Research 44 (1999) Briefly, the voltage dependence of the relative Na-conduct- Recovery from inactivation was measured at 228C using ance activation was measured by applying series of voltage a two pulse protocol in which a conditioning pulse to 230 clamp steps to different membrane potentials (V test) be- mv, allowing inactivation, was followed by a test pulse tween 2120 and 140 mv, starting from a holding (also to 230 mv) after a variable recovery interval ranging potential (V H)of2100 mv (Fig. 3A). Next, peak INa at between 3 and 83 ms. The fraction of channels that had each V was divided by the driving force for Na ions recovered following the various time intervals, was calcutest (Vtest 2 E Na; ENa5reversal potential of sodium current) and lated by dividing the peak Na currents evoked by the test normalized to the maximum Na-conductance. Steady-state pulse by the peak current measured during the conditioning inactivation relationships were obtained by measuring the pulse. Holding potential between conditioning and test peak INa during a 50 ms test step to 230 mv which pulses was 2100 mv. Averaged data were fitted to a single followed a 1 second prepulse to membrane potentials exponential to obtain the time constant of recovery from between 2120 and 220 mv (Fig. 3B). The peak currents inactivation. measured during the test potentials were normalized to the maximum current measured following the first prepulse to 2120 mv. 3. Results Activation and inactivation curves were fitted with a Boltzmann equation to establish the potential for half- Structural heart disease was excluded by normal echomaximal activation or inactivation (V 1/2) and slope factors cardiograms in six out of six patients (6/ 6), coronary (k) for these parameters. The experiments were performed angiograms (performed in 5/ 6), left and right ventriculoat 108C and at room temperature (228C) to ensure reliable voltage clamping of these fast currents. grams (5/ 6) and stress tests (6/ 6). Electrolytes were normal in all patients. The typical electrocardiographic Fig. 2. SCN5A mutations in patients 1 and 2. (Top) Sequence analysis of exon 26 from patient 1 and one control individual. (Middle) Sequence analysis of exon 28 from patient 2 and one control individual. (Bottom) Schematic representation of the SCN5A sodium channel protein and location of the R1512W and A1924T mutations.

5 M.B. Rook et al. / Cardiovascular Research 44 (1999) Fig. 2. (continued) features, viz. at least 1 mm ST-segment elevation in leads To determine the functional consequences of the V 1 V2 or V 1 V 3, were present in all six patients. Right R1512W and A1924T mutations, wild-type and mutant ventricular conduction delay was only present in patients 4 sodium channel proteins were expressed in Xenopus and 6. Patients 1, 3 6 were successfully resuscitated, oocytes by injecting the corresponding crnas. All three patient 2 was without complaints. Fig. 1 shows eight crnas resulted in expression of functional channels, since selected leads of an ECG taken from patient 1. The typical in each case typical rapidly activating and inactivating ST-elevation pattern in V 1 V2 was enhanced by sodium Na-currents were recorded under voltage clamp conditions channel blockade in this patient (flecainide, Fig. 1, right (Fig. 3). Maximum current amplitudes ranged between 2 panel), as well as in patients 3 and 5 (not shown) who also and 12 ma. To assess the electrophysiological mechanisms received these drugs during the hospitalization phase. in this syndrome, we compared the voltage dependence of SSCP analysis of the entire coding region of SCN5A in the activation and inactivation kinetics of currents carried six patients identified four different aberrant conformers in by wild-type (I Na2WT) and mutant sodium channels patients 1 4. None of these conformers was present in 100 (INa2R1512W or I Na2A1924T). alleles from unrelated control individuals. DNA sequence At 108C the V1/2 of INa2R1512W activation and inactivaanalysis in patient 1 revealed heterozygosity for a C T tion curves each were shifted about 2.5 mv to more transition in the first nucleotide of codon 1512, resulting in negative voltages as compared to I Na2WT. For INa2A1924T an Arg Trp substitution (R1512W; Fig. 2, top panel) these shifts were larger: 27.5 and 24.4 mv, respectively within the highly conserved DIII DIV cytoplasmic linker (Fig. 4A and Table 1). The slope factors for the voltage (Fig. 2, lower panel). DNA sequence analysis in patient 2 dependence of inactivation for wild-type and mutant revealed heterozygosity for a G A substitution in the first channels were similar. nucleotide of codon 1924, resulting in the substitution of At 228C (Fig. 4B and Table 1), the negative shifts of the Ala by Thr (A1924T; Fig. 2 middle panel) within the V1/2 of activation at 228C were more pronounced than at cytoplasmic C-terminal domain of the protein (Fig. 2, 108C: 25.1 mv for I and 29.0 mv for Na2R1512W lower panel). Patient 3 was heterozygous for a C T I. The V of inactivation at 228C for I Na2A1924T 1/2 Na2R1512W substitution at position 224 within intervening sequence 2 was shifted by 23.8 mv, but for I it was identical Na2A1924T (IVS2 224). This substitution (G[C T]CTAAT) ] to that of I Na2WT. In addition, the slope of the inactivation occurred within a site that bears a strong resemblance to curves was increased in particular. The negative shifts in the lariat branch site consensus sequence (YNYYRAY) V1/2 for activation of the mutant sodium channels found at implicated in splicing, which is typically located bp 228C, suggest an enhancement of excitability of carupstream of the 39 splice site. Finally, patient 4 was heterozygous for a C T substitution at IVS diomyocytes expressing these channel mutants. At 108C the mutations did not noticeably alter time

6 512 M.B. Rook et al. / Cardiovascular Research 44 (1999) Fig. 3. Examples of Na-currents measured at 108C after injection of wild-type or mutant SCN5A crna in Xenopus oocytes. (A) Left panel: selected recordings of INa2WT currents elicited by the voltage clamp protocol used to determine the voltage dependence of activation. Holding membrane potential (V H) was 2100 mv, with increasingly depolarizing test-pulses (V test) more INa2WT was activated. Maximum peak inward current of 212 ma was measured at Vtest between 230 and 220 mv. Right panel: peak I Na2WT /Vtest relationship from which the voltage dependence of activation was calculated (see Methods). Reversal potential for the sodium current (E Na) was estimated from its x-axis intercept as indicated. (B) Left panel: selected recordings of I currents elicited by the voltage clamp protocol used to determine the voltage dependence of inactivation. Only the last 3 ms of the inactivating Na1924T prepulse (V ) are shown. Peak I measured during the 230 mv test-potential was maximal (26.8 ma) when V was more negative than prepulse Na2A1924T prepulse 2120 mv. With increasingly depolarizing V prepulse, INa2A1924T decreased to zero beyond prepulses more positive than 265 mv. Right panel: relationship between Vprepulse and INa A1924T derived from all data points from the experiment shown in the previous panel.

7 M.B. Rook et al. / Cardiovascular Research 44 (1999) Fig. 4. Comparison of the voltage dependence of activation and inactivation of wild-type (WT) and mutant (R1512W and A1924T) cardiac sodium channels. Data were fitted with a Boltzmann equation to estimate the potential for half-maximal activation or inactivation (V 1/2) and slope factor (k). Scale bars represent S.E., accuracy of fits exceed 99%. (A) Voltage dependence of wild-type and mutant cardiac sodium channels measured at 108C. Inactivation of R1512W and A1924T channels was shifted 22.4 and 24.4 mv, respectively, compared to WT channels. Activation of R1512W and A1924T channels was shifted 22.7 and 27.5 mv, respectively. (B) Voltage dependence of wild-type and mutant cardiac sodium channels measured at 228C. Inactivation of R1512W channels was shifted 23.8 mv compared to WT channels. The voltage dependence of inactivation of A1924T channels was virtually similar to that of WT channels. Activation of R1512W and A1924T channels was shifted 25.1 and 29.0 mv, respectively. dependent channel kinetics during a voltage step from double exponential, indicating the presence of a fast and a 2100 mv to 220 mv, as is shown in Fig. 5A. At this slow component. As shown in Fig. 5B, only R1512W temperature, the gating behavior was well described by the channels were marginally different from wild-type chan- 3 classical Hodgkin and Huxley relationship mh. The time nels and showed a slight increase in the tfast value of constants for the activation-parameter m and inactivation- inactivation. parameter h were 1.3 ms and 13 ms, respectively. At The time constant of recovery from inactivation at 228C 228C activation was already so fast that reliable fitting with was 8 ms for Na2WT and A1924T channels, but some- 3 mhwas not possible. Inactivation was best fitted by a what prolonged (9.6 ms) for R1512W channels (Table 1).

8 514 M.B. Rook et al. / Cardiovascular Research 44 (1999) Fig. 5. Time dependence of activation and inactivation of wild-type and mutant channels during a voltage step from holding potential (2100 mv) to 220 mv. Data from experiments shown in Fig. 4 were averaged and normalized. For clarity, only 10% of the digitized data points are plotted. All figures are 3 6S.E. (A) Kinetics at 108C are well fitted by the Hodgkin and Huxley relationship mh, in which the activation (m) and inactivation (h) parameters are represented by mono-exponentials. Kinetics of wild-type and mutant channels are virtually identical. (B) At 208C activation was too fast to be reliably fitted, the time course of inactivation was best fitted by a bi-exponential function, indicating the presence of a fast and a slow component. Only tfast of R1512W channels was marginally slower than that of wild-type channels. 4. Discussion excitable and conducting properties of the heart. Indeed, genes encoding ion channels have been demonstrated to Ventricular fibrillation in the absence of structural heart underlie various forms of the congenital long QT disease is classified as primary electrical disease. Such a syndrome (for reviews see [1,2]). In their initial report on a diagnosis places the arrhythmogenic substrate in the distinct clinical and electrocardiographic entity, consisting

9 M.B. Rook et al. / Cardiovascular Research 44 (1999) Table 1 Summary of the potentials for half-maximal activation or inactivation (V inactivation at 228C 1/2 ) and slope factors (k) at 108C and 228C, and of the time constants of recovery of Temperature (8C) WT R1512W A1924T Activation 10 V 1/2 (mv) k V 1/2 (mv) k Inactivation 10 V 1/2 (mv) k V 1/2 (mv) k Recovery from inactivation 22 t 2100 mv (ms) of the trias aborted sudden cardiac death, right bundle selectively shorten the epicardial action potential as a branch block (RBBB) and right precordial ST-segment result of both this shift and the prominent presence of Ito in elevation, Brugada and Brugada speculated that this epicardial cells [22]. In Brugada patients, flecainide (Fig. syndrome might represent another ion channel disorder [3]. 1) and other sodium channel blockers like procainamide The syndrome may display a familial occurrence [3], and ajmaline augment or elicit the electrocardiographic seems to be associated with a high mortality rate [4], and is abnormalities [6,9,10]. In addition, reduced INa is also increasingly recognized [4,5 7], for review see [18]. It has compatible with the frequently encountered conduction been estimated that this syndrome underlies 40 60% of disturbances (increased HV interval, RBBB) in these patients with idiopathic VF [8]. Recently, mutations in patients [6,9,10], see also [18]. SCN5A, the gene encoding the a-subunit of the cardiac fast The changes in INa carried by the two mutated cardiac sodium channel were shown to co-segregate with affected sodium channels we describe in this paper suggest the individuals within small families in which a high incidence opposite. We found a substantial negative voltage shift of of sudden cardiac death was present [8]. the activation of the mutant channels while, as shown in Electrical heterogeneity in the right ventricular outflow Fig. 5, the time course of activation and inactivation were tract (RVOT) has been postulated as the cellular basis for virtually not affected by the mutations. These observations the ECG pattern and the associated, potentially lethal, suggest that in cardiac myocytes expressing the mutated arrhythmias [19,20]. Ion currents relevant to the early channels the action potential threshold should be lower. phase of the action potential of epicardial cells, differing Since the voltage dependence of inactivation at 228C was from endocardial cells by a characteristic spike and dome not changed for A1924T channels, the net result should be morphology, are the fast inward sodium current (I Na), the an increase in the window current. Such an effect is less transient outward current (I to), and the L-type calcium obvious for R1512W channels since these also exhibit a current (I CaL). Disappearance of the dome due to altered small negative shift in the steady state inactivation curve. characteristics of any of the three channel types will result Moreover, the shift of the activation curve to more in selective shortening of the epicardial action potential. negative potentials, in the absence of changes in time Transmural current flow during the plateau phase of the dependent kinetics, could lead to a larger INa during the action potential will ensue, giving rise to ST-segment action potential upstroke and a larger overshoot. It is of elevation in the leads V1 and V2 facing the right ventricular interest that in the sodium channel mutations associated outflow tract (RVOT) in which the Ito-dependent phase 1 with idiopathic ventricular fibrillation recently described seems most prominently present [21]. In this setting [8], the mutation T1620M resulted in a nearly 110 mv arrhythmia s may result from phase 2 reentry [19,20]. shift in the inactivation curve, which could also result in The mutations described in this study and one of the increased window current and a larger INa during the mutations reported by Chen et al. significantly affect INa action potential upstroke. characteristics, suggesting a causal relationship [8]. It has At 228C, the time constant of recovery from inactivation been suggested that gene mutations that reduce sodium was very similar between wild-type and A1924T channels, channel density or conductance are prime candidates [19]. but for R1512W channels it was slightly prolonged (1.5 A reduction in INa will reduce the action potential am- ms). We do not think that this prolongation will have plitude, which in turn leads to a negative shift of the significant effects on action potential morphology. In voltage at which phase 1 repolarization begins. In ex- theory, a slower recovery from inactivation could lead to a perimental preparations flecainide has been shown to decrease of channel availability at higher rates, but for

10 516 M.B. Rook et al. / Cardiovascular Research 44 (1999) R1512W channels this would be counteracted by the impact of these changes on action potential morphology negative voltage shift of the steady state activation. and the resulting ECG in the patients, in which the mutant To explain the ECG of the Brugada patients carrying the sodium channels are presumably heterozygously ex- R1512W or the A1924T sodium channel mutations on the pressed, is uncertain. A particular complicating factor is basis of an increase in I Na, we tentatively suggest the the potential temperature dependence of Na-channel gating following mechanism. Altered I will have an impact on parameters. Expression studies at physiological tempera- Na I to (and I CaL) characteristics. A larger action potential ture in mammalian cell lines and direct measurements on overshoot due to increased INa may result in increased I to, epicardial cells from affected patients may be the only way potentially also leading to a negative shift of the epicardial to provide answers to these uncertainties. We have not phase 1 nadir. The resultant epicardial action potential studied the interaction with the putative b-subunit [28] for shortening and related ECG changes have been discussed the sodium channel encoded by SCN5A, which could be above. Interestingly, in the family described by Chen et al. affected in the case of the A1924T mutant channel, as has and in the patients 1 and 2 with mutations in SCN5A exons been shown for the other C-terminal mutation D1790G we describe here, no RV conduction delay is observed [8], [12]. which also argues against a substantial reduction of I Na. In conclusion, functional studies in a heterologous Alternatively, to explain our results still assuming a expression system of Brugada syndrome related mutant reduction in I Na, the shift of the activation curves towards sodium channels, reveal significant shifts of the activation the resting membrane potential at physiological tempera- curves in the negative direction. These shifts could result ture could be such that the fraction of open sodium channels during diastole is increased. This would cause some persistent depolarization and consequently a reduced action potential upstroke. In patients 3 and 4, we detected intronic mutations which may possibly affect splicing. It has been shown that changes in the consensus sequence at splice sites or lariat branch-point regions cause exon skipping. Moreover, intron or exon mutations (including silent mutations) outside these regions can also affect splicing [23 26]. Whenever these mutations lead to alternative splicing, it is likely that non-functional channels will ensue. Chen et al. also described SCN5A gene mutations in small Brugada families leading to premature truncation of the sodium channel protein [8]. It may be speculated that in these particular cases heterozygosity for the mutation will lead to a reduction in expression of functional sodium channels. In these cases the Brugada phenotype would result from a reduction of I Na. The fact that no SCN5A mutations were detected in two patients (5 and 6) suggests genetic heterogeneity for this in increased sodium current during the action potential upstroke. Our results may provide the basis for new insights into the mechanisms of arrhythmogenesis in patients with idiopathic VF. Acknowledgements We thank Drs. R.N. Hauer, E. Wever, D. Duren, H. Crijns, A. Gorgels and Boehmer for their help in collecting the patients and G. Salib, M. Schipper, K. Jansema and A.V. Postma for technical assistance. We are grateful to Drs. O. Destree and L. Tertoolen and Mr. K. Koster (Hubrecht Laboratory, Utrecht) for allowing the use of their facilities and for the generous supply of Xenopus oocytes. This work was supported by grants: D95/ 014, M and from the Dutch Heart Foundation. disorder. However, the SSCP methodology (the most References widely used screening method) for mutation detection can detect between 88 and 90% of mutations [27] and therefore [1] Kass RS, Davies MP. The roles of ion channels in an inherited heart disease: molecular genetics of the long QT syndrome. Cardiovasc one cannot entirely exclude presence of SCN5A mutations Res 1996;32: within this gene in these individuals. [2] Roden DM, Lazzara R, Rosen M, Schwartz PJ, Towbin J, Vincent GM. Multiple mechanisms in the long QT syndrome. Current 4.1. Limitations of the study knowledge, gaps, and future directions. Circulation 1996;94: [3] Brugada P, Brugada J. Right bundle branch block, persistent ST- In this study we have not shown co-segregation of the segment elevation and sudden cardiac death: a distinct clinical and SCN5A mutations with the typical electrocardiographical electrocardiographic syndrome. J Am Coll Cardiol 1992;20:1391 features. However, we believe that the mutations described in this study could be causally linked to the disease for two [4] Brugada J, Brugada R, Brugada P. Right bundle-branch block and reasons. First, Chen et al. recently described mutations in ST-segment elevation in leads V1 through V3. A marker for sudden the SCN5A gene that co-segregated with the clinical death in patients without demonstrable structural heart disease. Circulation 1998;97: syndrome [8]. Second, the estimated effect of the func- [5] Atarashi H, Ogawa S, Harumi K et al. Characteristics of patients tionally studied mutation [8], and of our mutations is a with right bundle branch block and ST-segment elevation in right consistent increase in window current. However, the precordial leads. Am J Cardiol 1996;78:

11 M.B. Rook et al. / Cardiovascular Research 44 (1999) [6] Brugada J, Brugada P. Further characterization of the syndrome of [17] Stuhmer W. Electrophysiological recording from Xenopus oocytes. right bundle branch block, persistent ST-segment elevation, and Methods Enzymol 1992;207: sudden cardiac death. J Cardiovasc Electrophysiol 1997;8: [18] Alings AMW, Wilde AAM. The Brugada syndrome. Clinical data [7] Matsuo K, Shimizu W, Kurita T, Inagaki M, Aihara N, Kamakura S. and suggested pathophysiological mechanism. Circulation Dynamic changes of 12-lead electrocardiograms in a patient with 1999;99: Brugada syndrome. J Cardiovasc Electrophysiol 1998;9: [19] Antzelevitch C. The Brugada syndrome. J Cardiovasc Electrophysiol [8] Chen Q, Kirsch GE, Zhang D et al. Genetic basis and molecular 1998;9: mechanism for idiopathic ventricular fibrillation. Nature [20] Yan G, Antzelevitch C. Cellular basis for the electrocardiographic 1998;392: J-wave. Circulation 1996;93: [9] Miyazaki T, Mitamura H, Miyoshi S et al. Autonomic and antiar- [21] Di Diego JM, Sun ZQ, Antzelevitch C. Ito and action potential notch rhythmic drug modulation of ST segment elevation in patients with are smaller in left vs. right canine ventricular epicardium. Am J Brugada syndrome. J Am Coll Cardiol 1996;27: Physiol 1996;271:H548 H561. [10] Nademanee K, Veerakul G, Nimmannit S et al. Arrhythmogenic [22] Krishnan SC, Antzelevitch C. Flecainide-induced arrhythmia in marker for the sudden unexplained death syndrome in Thai men. canine epicardium. Phase 2 reentry? Circulation 1993;87: Circulation 1997;96: [23] Liu W, Qian C, Francke U. Silent mutation induces exon skipping of [11] Wang DW, Yazawa K, George AL, Bennet PB. Characterization of fibrillin-1 gene in Marfan syndrome. Nature Genet 1997;16:328 1 human cardiac Na channel mutations in the congenital long QT 329. syndrome. Proc Natl Acad Sci 1996;93: [24] Dyhr-Mikkelsen H, Kjems J. Inefficient spliceosome assembly and [12] An RH, Wang XL, Kerem B et al. Novel LQT-3 mutation affects abnormal branch site selection in splicing of an HIV-1 transcript in 1 Na channel activity through interactions between alpha- and beta1- vitro. J Biol Chem 1995;270: subunits. Circ Res 1998;83: [25] Hornig H, Aebi M, Weismann C. Effect of mutations at the lariat [13] Wei J, Wang DW, Alings M et al. Congenital long-qt syndrome branch acceptor site on b-globin pre-mrna splicing in vitro. Nature caused by a novel mutation in a conserved acidic domain of the 1986;324: cardiac Na channel. Circulation 1999;99: [26] Nakai K, Sakamoto H. Construction of a novel database containing [14] Wang Q, Li Z, Shen J, Keating MT. Genomic organization of the aberrant splicing mutations of mammalian genes. Gene human SCN5A gene encoding the cardiac sodium channel. Gen- 1994;141: omics 1996;34:9 16. [27] Condie A, Eeles R, Borresen A-L, Coles C, Cooper C, Prosser J. [15] Gellens ME, George AL, Chen L et al. Primary structure and Detection of point mutations in the p53 gene: comparison of singlefunctional expression of the human cardiac tetrodotoxin-insensitive strand conformation polymorphism, constant denaturant gradient voltage-dependent sodium channel. Proc Natl Acad Sci electrophoresis, and hydroxylamine and osmium tetroxide technique. 1992;89: Hum Mutat 1993;2: [16] Makita N, Bennet PB, George Jr. AL. Multiple domains contribute 1 [28] Kupershmidt S, Yang T, Roden DM. Modulation of cardiac Na to the distinct inactivation properties of human heart and skeletal current phenotype by beta1-subunit expression. Circ Res 1 muscle Na channels. Circ Res 1996;78: ;83: