MedicineLakex

 

Doi:10.1016/s0008-6363(03)00342-0

Cardiovascular Research 59 (2003) 27–36 www.elsevier.com / locate / cardiores common polymorphism in KCNH2 (HERG) hastens cardiac Connie R. Bezzina *, Arie O. Verkerk , Andreas Busjahn , Andreas Jeron , Jeanette Erdmann , Tamara T. Koopmann , Zahurul A. Bhuiyan , Ronald Wilders , Marcel M.A.M. Mannens , Hanno L. Tan , Friedrich C. Luft , Heribert Schunkert , Arthur A.M. Wilde aExperimental and Molecular Cardiology Group, Room M0-052, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands bDepartment of Clinical Genetics, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands cFranz Volhard Clinic and Max Dellbruck Center for Molecular Medicine, Berlin, Germany dInternal Medicine, University of Regensburg, Regensburg, Germany eDepartment of Physiology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Received 30 December 2002; accepted 26 February 2003 Objective: Genetic variants of cardiac ion channels may influence cardiac repolarization. Thereby such variants may modulate the
penetrance of primary electrical disorders, contribute to differences in susceptibility to drug-induced QT-prolongation betweenindividuals, or contribute to rhythm disturbances in the context of structural heart disease. Since the current encoded by KCNH2 (HERG;I ) is a primary determinant of repolarization, we conducted association studies between the respective alleles of the common amino acid-changing polymorphism at codon 897 (2690A.C; K897T) within HERG and rate-corrected QT interval (QTc). Methods and
Results: Association analysis in Caucasian subjects (n51030) revealed a significant association of this polymorphism with QTc
(P50.0025) with CC homozygotes having a significantly shorter QTc (388.562.9 ms) compared to AA homozygotes (398.560.9) and
heterozygotes (AC, 397.261.2). The latter two genotypes were associated with comparable mean QTc's, suggesting that the 2690C-allele
is recessive. After stratification by sex, the polymorphism was more predictive of QTc in females (P50.0021), a finding that was
replicated in a second population sample (n5352) from the same ethnic background (P50.044). To assess whether this polymorphism
could represent a ‘functional' polymorphism, we compared the biophysical properties of K897- and T897-HERG channels by whole-cell
voltage clamp. Compared to the K897 channel, the T897 channel displayed a shift of 27 mV in voltage dependence of activation and
increased rates of current activation and deactivation. Conclusion: As confirmed in modeling studies, these changes are expected to
shorten action potential duration by an increase in I . This recapitulates the shorter QTc in females homozygous for the 2690C-allele.
 2003 European Society of Cardiology. Published by Elsevier Science B.V. All rights reserved.
Keywords: K-channel; Ion channels; ECG; Epidemiology; Arrhythmias . Introduction
baseline QT interval may predispose to sudden death inpatients with coronary artery disease or myocardial The electrocardiographic (ECG) QT interval is a mea- infarction Large population studies have shown that sure of cardiac ventricular repolarization. A prolonged a prolonged QT interval may also predict cardiovasculardeath in normal, apparently healthy individuals Moreover, in dofetilide-treated patients with moderate to *Corresponding author. Tel.: 131-20-566-3265; fax: 131-20-697-
E-mail address: (C.R. Bezzina).
Time for primary review 38 days.
0008-6363 / 03 / $ – see front matter  2003 European Society of Cardiology. Published by Elsevier Science B.V. All rights reserved.
doi:10.1016 / S0008-6363(03)00342-0 C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 severe heart failure and reduced left ventricular systolic on ECG parameters in the general population. Thus we function, a short baseline QT interval was associated with also studied the biophysical consequences of this poly- reduced mortality morphism by whole-cell voltage clamp and computer A prolonged QT interval may reflect inter-individual variability but may also occur as an adverse reaction tosome commonly used antiarrhythmic agents as well asvarious medications including antibiotics, antihistaminesand antipsychotics, which block I (described below) . Methods
Some of these have also been implicated in the develop-ment of life-threatening torsade de pointes ventricular .1. Study population tachycardia, particularly in females and in individualswith a prolonged QT interval at baseline .1.1. Sample I The QT interval reflects ventricular repolarization during We identified 1030 healthy Caucasians through a which a net outward current restores the membrane population-based survey, namely the Augsburg MONICA potential of the depolarized myocyte back to the resting (Monitoring of Trends and Determinants in Cardiovascular membrane potential. The onset of this repolarization phase Disease, study 3) study, with normal ECG without signs of is primarily mediated by the delayed rectifier K acute myocardial infarction or unstable angina pectoris at which is comprised of the rapidly and slowly activating the time of examination and recording of the ECG components I and I , respectively. Mutations in genes Each participant underwent a physical examination and a encoding for ion channel subunits of these currents medical history was obtained before blood was taken for (KCNH2, KCNE2 for I ; KCNQ1, KCNE1 for I ) are genetic studies. The investigation conforms with the associated with specific subtypes (2, 6, 1, and 5, respec- principles outlined in the Declaration of Helsinki and tively) of the long QT syndrome (LQTS) wherein a written informed consent was obtained from all particip- decrease in outward K current results in prolonga- tion of the action potential (AP). This prolongation, which A standard 12-lead resting ECG was recorded and stored is translated into a prolonged QT interval on the surface in a digitized fashion using a Hewlett Packard Pagewriter ECG, renders the heart vulnerable to torsade de pointes XLi (Hewlett Packard, Palo Alto, CA, USA). QT interval and sudden cardiac death.
in individual leads was computed by the appropriate ECG parameters, including QT interval, are biometrical software (Interpretive Cardiograph A.0.1.00; Hewlett Pac- traits known to be influenced by genetic variance and kard) and double-checked by an investigator blinded to the loci for cardiac ion channel genes have been demonstrated clinical data. Mean corrected QT interval (QTc) was to constitute quantitative trait loci for these parameters in calculated according to Bazett's formula. Individuals with normal, healthy individuals These observations imply an ECG showing a bundle branch block pattern with a that, aside from the monogenic disorders of the LQTS, QRS duration greater than 120 ms were excluded from cardiac ion channel genes contain allelic variants that have further analysis.
a functional bearing on ECG parameters including QTc inthe general population. Thereby, such variants may under-lie (at least in part) inter-individual variation in ECG .1.2. Sample II parameters, modulate the penetrance or expressivity of We recruited 176 pairs of twins (monozygotic, 116; primary electrical disorders such as LQTS and dizygotic, 60; ) by advertisement to participate in Brugada syndrome modify the propensity to drug- studies involving blood pressure regulation and car- induced LQTS or contribute to rhythm disturbances in diovascular phenotypes. The subjects were all German the context of structural heart disease, a hypothesis that has Caucasians from various parts of Germany. The protocol gathered considerable support in recent years was approved by the University's (Berlin) committee on is a primary determinant of repolarization, we the protection of human subjects and written informed assessed the impact of the common 2690A.C nucleotide consent was obtained from all participants. Blood was change corresponding to the K(lysine)897T(threonine) obtained for the determination of zygosity and other amino acid polymorphism within the KCNH2 gene (also molecular genetic studies. Each participant underwent a called HERG for human ether-a-go-go-related gene) that medical history and physical examination. None had a encodes the pore-forming subunit of the I family history of chronic medical illness. Subjects under- ECG by association studies between the respective alleles went echocardiography and planar ECG. A standard 12- of this polymorphism and QTc in two independent samples lead electrocardiogram was performed (CARDIOVITS of the general population. The occurrence of such an CS-100, Schiller, Baar, Switzerland). The QT- and RR- amino acid-changing polymorphism raises the possibility intervals were measured in lead II. QTc was determined that the two channel variants (K897 and T897) could differ according to Bazett's formula. The ECG parameters were in their electrophysiological properties and thereby impact scored by a computer and stored for subsequent retrieval.

C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 .2. Genotype analysis carried out on shrimp alkaline phosphatase-treated PCRproducts of primers P3 and P4, using the following primer: Genotyping for the KCNH2 K897T polymorphism was 59TCCGCAGGCGCACGGACA39 in a 10-ml reaction done by restriction analysis (sample I: n5730; sample II: consisting of dCTP, ddATP, ddGTP (100 mmol / l each), 20 all) or by DNA sequencing (sample I: n5300). For pmol extension primer and 1 U Thermo Sequenase DNA restriction analysis, a fragment encompassing the poly- polymerase (Amersham) in Thermo Sequenase buffer.
Thermal cycling conditions were 94 8C for 3 min followed 59GGGAGCTTGGGGCCTGACC39 (forward; P1) and by 30 cycles of 94 8C for 10 s, 57 8C for 1 min, 72 8C for 59TCCCCTCCCCCGCCTCACAC39 (reverse; P2). P2 dif- 30 s. Before mass spectrometric analysis on a MALDI- fered from the KCNH2 sequence such that in combination TOF (Bruker Biflex) system, products of primer extension with the 2690A.C change a restriction site for BsaAI was were purified using the Genopure oligo DNA purification created (). The nucleotide that differed is under- system (Bruker).
For sequencing, a fragment containing the polymor- .3. Site-directed mutagenesis and cell culture 59AAGGGCCCTGATACTGATTTT39 (forward, P3) and The HERG / GFPires construct containing K897-HERG 59TTCCAGCTCCCAGCCTCA39 (reverse, P4). The re- cloned in the GFPires plasmid for bicistronic verse primer was used for sequencing, which was carried expression of the HERG channel protein and GFP reporter out using the ABI BigDye Terminator Sequencing kit (i.e. translation of the protein of interest and marker from (Applied Biosystems; the same mRNA due to the presence of an internal Mass spectrometric analysis of primer extension prod- ribosomal entry site (IRES) between the two cloned ucts, a technique amenable to high-throughput analysis, genes), was kindly provided by David C. Johns (Johns was also set up (data not shown). Primer extension was Hopkins University, Baltimore, MD, USA). T897-HERG Fig. 1. (A) Sequence analysis of a KCNH2 2690 AA homozygote (top), AC heterozygote (middle) and CC homozygotes (bottom). (B) Restriction analysisfor genotyping of the K897T polymorphism. A fragment encompassing the polymorphism was amplified using the primers shown. The reverse primerdiffered from the KCNH2 sequence such that in combination with the polymorphism a restriction site for BsaAI was created. ‘PyAC↓GTPu' is therecognition sequence for the restriction enzyme BsaAI, where Py is a pyrimidine-containing nucleotide (C or T), Pu is a purine-containing nucleotide (A orG) and ↓ is the site of cleavage. (C) Schematic representation of a HERG channel subunit displaying the location of the K897T polymorphism.
C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 was created by PCR overlap-extension mutagenesis. The .5. Computer simulations final PCR product was digested with SacI and XhoI andcloned into HERG / GFPires, creating the construct T897- Functional differences between the K897- and T897- HERG / GFPires. The mutated insert and ligation regions HERG channels were tested by computer simulations using were sequenced fully to ensure that the clone was free of the Priebe and Beuckelmann human ventricular cell polymerase errors.
model. The experimentally observed shift in the steady- Human embryonic kidney (HEK-293) cells were tran- state activation curve of T897 versus K897 was im- siently transfected with the HERG / GFPires or T897- plemented by a 27 mV shift in the I HERG / GFPires construct using lipofectamine and cultured and b . The approximately twofold increase in the rate of at 37 8C. Cells exhibiting green fluorescence were selected activation and deactivation of T897 versus K897 was 36–48 h after transfection for electrophysiological experi- implemented by doubling a and b , respectively.
.6. Statistical analysis .4. Electrophysiological experiments The QTc values of monozygotic twins (sample II) were converted into singleton values by calculating the average.
Cells were superfused with solution (3661 8C) con- Genotype-dependent differences in QTc were determined taining (mmol / l): NaCl 140, KCl 5.4, CaCl 1.8, MgCl with univariate analysis of variance (sample I) or mixed 1.0, glucose 5.5, HEPES 5.0, and pH 7.4 (NaOH).
model analysis (sample II). In the latter model the in- Membrane currents were recorded in the ruptured-patch dividuals were treated as repeated measures within the whole-cell configuration of the patch-clamp technique.
random factor family. The within-family correlation was Patch pipettes (3–5 MV) were pulled from borosilicate 0.45. In both analyses, genotype, gender and genotype– glass, heat-polished, and filled with solution containing gender and age–gender interactions were used as indepen- (mmol / l): K-gluconate 125, KCl 20, MgCl 1.0, EGTA 5, dent between group factors and age as covariate. Further- MgATP 5, HEPES 10, and pH 7.2 (KOH). Voltage control, more, the analyses were repeated for both genders separ- data acquisition, and analysis were accomplished using ately where appropriate. Difference in allele frequency was custom software. Potentials were corrected for the liquid tested with the normal approximation of the binomial junction potential. Membrane currents and potentials were distribution. All numbers are presented as mean6S.E.M.
low-pass filtered on-line (cut-off frequency 2 kHz) and Differences were considered significant for P ,0.05. All digitized at 5–10 kHz. Cell size for cells transfected with statistics were calculated using SPSS software (version 11, HERG / GFPires (8.360.7 pF) did not differ significantly from those with T897-HERG / GFPires (8.961.4 pF).
The activation, deactivation, and inactivation kinetics of the HERG current were determined by the voltage clamp . Results
protocols as diagrammed in and described previous-ly For all protocols, holding potential was 280 .1. Association studies mV and the pacing intervals 15 s. Tail current amplitude,normalized to maximum tail current, was used to construct A total of 1382 Caucasian (German) subjects, recruited steady-state (in)activation curves (). The as two separate samples at two different centers (Regens- (in)activation curves were fitted using the Boltzmann burg and Berlin) were evaluated in this study ( equation I /I 5 A / h1.0 1 exp [(V 2 V ) /k]j to deter- Sample I was initially evaluated. Findings were then mine the membrane potential for half-maximal (in)activa- replicated in a second smaller sample (II). All subjects and the slope factor k. The time course of were genotyped for the K897T polymorphism, with the (in)activation (inset, and ) was fitted by the exception of the monozygotic twins, in whom only one of monoexponential equation I /I 5 A 3 [1 2 exp (2t /t)] the pair was genotyped. As expected of two samples of the The time course of deactivation (inset) same ethnicity, minor allele (2690C) frequency was simi- was fitted by the biexponential equation I /I lar between the two; 23.5% and 24.6% for samples I and [1.0 2 exp (2t /t )] 1 A 3 [1.0 2 exp (2t /t )], where A II, respectively (P 50.725). The allelic distribution was in and A are fractions of fast and slow inactivation com- Hardy–Weinberg equilibrium. We evaluated the influence ponents, and t and t are the time constants of fast and of this polymorphism on QTc, in multifactor analysis that slow inactivating components, respectively The included age, sex, and genotype–sex and age–sex interac- time course of recovery from inactivation () was tions, as factors and covariates.
measured as the fast component of a biexponential equa- Mean QTc values per genotype for the samples as a tion Data are expressed as mean6S.E.M. and whole and after separation by gender are given in compared using a Student's t-test with a significance level In sample I, the K897T polymorphism was significantly P ,0.05.
associated (P 50.0025) with QTc (). CC homo- C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 able 1Study populations Male (n 5509) Female (n 5521) Age, years (median [range]) Sample II
Male (n 576) Female (n 5156) Male (n 542) Female (n 578) Age, years (median [range]) able 2Mean QTc per KCNH2 K897T genotype for total sample and after separation by gender Sample II
Numbers of individuals are given between parentheses. Genotype AA5KK; genotype AC5KT; genotype CC5TT.
a QTc values are mean6S.E.M.
zygotes had the shortest QTc (388.5 ms) compared to AA We next sought to replicate the association of this homozygotes (398.5 ms) and heterozygotes (AC, 397.2 polymorphism with QTc in females in a second indepen- ms). Sex and age were both strongly associated (P ,0.001) dent sample from the same population recruited at a with QTc (). After stratification of the population different center (sample II). Again, the association between by sex, the K897T genotype was more predictive of QTc the K897T genotype and QTc was found to be statistically in females than in males (P 50.002, females; P 50.34, significant in females (P 50.044, ).
males; ). Female CC homozygotes had the shortest Finally, since the mean QTc values of the AA and CC mean QTc (389.0 ms) compared to AA homozygotes homozygotes differed much more than those of the AA (403.0 ms) and AC heterozygotes (402.1 ms) ( homozygotes and AC heterozygotes, a recessive model ). In males, although the association with genotype was tested (sample I: CC n 558, versus AA /AC n 5972).
was not significant, QTc tended to be shorter in CC homozygotes compared to AA homozygotes ( (388.562.9 ms) compared to those with AA /AC genotypes (398.060.7 ms) (P ,0.001). Similarly, female CC homo-zygotes had a significantly shorter QTc (389.063.9 ms;n 527) compared to females with AA /AC genotypes (402.660.9 ms; n 5494) (P ,0.01).
Influence of covariates, including K897T genotype, gender and age, in amulti-variate / factor model Sample I
KCNH2 K897T genotype
Sex3K897T genotype Sample II
KCNH2 K897T genotype
Fig. 2. Mean and 95% confidence interval of the QTc intervals in sample I per KCNH2 K897T genotype and gender. Number of individuals per Sex3K897T genotype genotype is given between parentheses. Genotype AA5KK; AC5KT; 59 (2003) 27–36 Fig. 3. Characteristics of K897 and T897 HERG currents. (A) Representative examples of K897 (left) and T897 (right) currents elicited by a two-step voltage clamp protocol (inset). P1 activated steady-state HERG current. Thecurrent magnitude progressively increased and then decreased with voltage according to voltage-dependent inactivation. P2 elicited HERG tail currents; their peak is due to fast recovery from inactivation secondary to repolarization.
The subsequent current decline is due to deactivation. (B) The average amplitudes of steady-state (circles) and peak tail (squares) currents did not differ significantly between K897 and T897. (C) The voltage for half-maximalactivation shifted from 223.761.8 in K897 to 230.861.6 in T897 (P 50.005); the slope factor k (mV) was 7.360.6 and 7.260.5, respectively (P 50.86). The activation time course (fitted to the rising phase of the current during P1,panel A) for T897 was faster than for K897 (inset, asterisks, P ,0.05). (D) Deactivation characteristics determined using a two-step protocol (left inset). P1 activated steady-state HERG current. P2 served to elicit HERG tail currents.
The deactivation current–voltage (I–V) relationship was determined from the peak of the deactivating tail current during P2. This I–V relationship did not differ significantly between K897 and T897. The deactivation time course(analyzed by fitting the decaying phase of the current during P2 to a biexponential function) for T897 was faster than for K897, as shown by the decrease in both time constants (right inset, asterisks, P ,0.05). (E) Voltage dependenceof inactivation of K897 and T897 determined by a three-step protocol (inset). During P1, HERG current activation occurred and was followed by inactivation. Repolarization during P2 allowed recovery from inactivation. The voltagedependence of inactivation was determined from the peak of the tail current measured during P3. This I–V relationship did not differ significantly between K897 and T897. Voltage for half-maximal inactivation: 239.963.7 (K897),240.563.5 (T897) (P 50.90); slope factor k (mV): 20.061.7 (K897), and 1961.2 (T897) (P50.62). (F) Time constants of HERG current inactivation (circles) and recovery from inactivation (squares). The time constant ofinactivation was determined using a three-pulse protocol (right inset). The first two steps served to (in)activate and recover from inactivation, respectively (see E). The time constant of inactivation, derived from fitting the decliningphase of the current during P3, did not differ significantly between K897 and T897. The time constant of recovery from inactivation was determined using a two-pulse protocol (left inset). P1 served to activate and inactivate HERGchannels. Repolarization during P2 elicited a tail current due to fast recovery from inactivation. This recovery process was measured as the fast time constant of a biexponential fit to the tail current (the slower time constant of this fitreflects the concurrent deactivation process). The fast time constant did not differ significantly between K897 and T897.
C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 .2. Electrophysiological data .3. Physiological implications shows typical K897- and T897-HERG currents Our electrophysiological experiments demonstrate that expressed in HEK-293 cells (panel A), and their activation the T897-HERG channel differs from the K897 channel by (panels B–C), deactivation (panel D), and inactivation alterations in both activation (27 mV shift of steady-state (panels E and F) characteristics. The average amplitudes of activation curve and approximately twofold increase in rate steady-state and peak tail currents did not differ sig- of activation) and deactivation parameters (approximately nificantly between K897 and T897 (The voltage twofold increase in rate of deactivation). To assess the dependence of activation of T897 displayed a 7 mV shift physiological implications of these differences, we carried into the negative direction (P 50.005) compared to K897 out computer simulations using the comprehensive mathe- (). Moreover, the activation time course for T897 matical model of a human ventricular cell by Priebe and was faster than for K897 (P ,0.05), as shown by the Beuckelmann shows the effect of altered approximately twofold decrease in the activation time activation parameters. The altered activation of the T897 constant (inset). The deactivation current–voltage channel (solid lines) compared to the K897 channel (I–V) relationship did not differ significantly between (dashed lines) results in a larger I amplitude (bottom). As K897 and T897 (). However, the deactivation time a consequence, repolarization of the AP is accelerated course for T897 was 1.5 times faster (P ,0.05) than for (top), resulting in an 11.5% decrease in AP duration at K897, as shown by the decrease in both time constants 90% repolarization (APD ). shows the effect of (right inset). The voltage dependence of inactiva- altered deactivation parameters. The more rapid deactiva- tion (and the time constants of inactivation ( (bottom) results in a 0.7% increase in APD .
right, and right inset) and recovery from inactivation The net effect of the alterations in both activation and (left, and left inset) did not differ significantly deactivation parameters is a decrease in APD between K897 and T897.
Fig. 4. Functional effects of the K897T HERG polymorphism. Dashed lines show the control action potential (top) and the HERG encoded membranecurrent I (bottom) of the Priebe–Beuckelmann human ventricular cell model at steady 1-Hz stimulation. Solid lines show the effects of incorporating the experimentally observed changes in I kinetics of the K897T polymorphism. (A) Effect of altered activation parameters (27 mV shift in steady-state activation curve and twofold increase in rate of activation). (B) Effect of altered deactivation parameters (twofold increase in rate of deactivation). (C)Combined effect.
C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 . Discussion
study and Ref. as well as in the setting of thecongenital LQTS This disparity in QTc between Association studies between the respective alleles of the males and females stems from postpubertal QTc shortening KCNH2 K897T polymorphism and QTc in a large sample in men the mechanism for which is hitherto un- of Caucasian subjects recruited in Germany demonstrated known. The longer QTc in females renders them more that this polymorphism is significantly associated with susceptible to the development of torsade de pointes in QTc. CC homozygotes displayed a significantly shorter QTc compared to AA homozygotes and AC heterozygotes.
congenital LQTS wherein among known mutation carriers, The latter two genotypes were associated with comparable females are more likely to be symptomatic A similar mean QTc's, suggesting that the 2690C-allele is recessive.
female preponderance is found in the setting of drug- After stratification by gender, the effect of the K897T induced QT-prolongation The polymorphism de- genotype on QTc was found to be most predictive in scribed herein could attenuate QTc-prolongation in these females, in whom mean QTc was reduced by 14 ms in settings in female CC homozygotes, and could underlie, at CC homozygotes compared to AA homozygotes and AC least in part, some degree of inter-individual susceptibility heterozygotes. This (female) gender-specific effect of the to the development of QTc-prolongation and life-threaten- K897T genotype on QTc was replicated in a second ing cardiac arrhythmias subsequent to I -blocking drugs (smaller) independent sample of the (German) Caucasian in females. In support of this proposition is the fact that population recruited at a different center.
female CC homozygotes had a mean QTc (389.063.9 ms) We next investigated whether the K897T change itself comparable to the mean (all genotypes) QTc in males could be responsible for these effects, i.e. whether it (392.961.0 ms; ).
represents a ‘functional polymorphism'. Electrophysiologi- In addition, the K897T polymorphism could contribute cal characterization of the T897-HERG channel (reflecting to the variable penetrance in the congenital LQTS. In the situation in CC homozygotes) revealed that this support of such a role, Laitinen et al. noted that channel exhibited biophysical properties similar to those of among 28 families in which the KCNQ1-Fin mutation—a the K897-HERG channel (which reflects the situation in founder mutation for LQTS in Finns—segregated, female AA homozygotes) with the exception of voltage depen- mutation carriers with the CC genotype tended to have dence of steady-state activation (shifted by 27 mV), and shorter QTc values than those with the AA and AC time courses of I activation and deactivation (accelerated approximately twofold). As confirmed by computer simu- Our findings differ from those reported recently by lation, these changes increase I thereby augmenting net Pietila et al. who, in a study on 187 Finnish females, outward current during the repolarization phase of the found hastened repolarization (shorter QTc cardiac AP. This hastens repolarization, which would the longest QTc measured at any of the 12 ECG leads) in translate into a shorter QTc on the surface ECG.
AA homozygotes compared to individuals with AC and The K897 residue in KCNH2 is evolutionary highly CC genotypes. This underscores the importance of con- conserved from mouse and rat to human sug- firming associations in different populations. The disparity gesting that replacing this basic residue with threonine (T) between the two studies could reflect population differ- could affect the function of the channel. Together with the ences such as population-specific differences in the occur- fact that the biophysical properties of the T897 channel rence of additional (functional) polymorphisms (possibly explain the observed phenotype for this allele, this implies in the promotor region) that influence the effect. Besides, that it is highly probable that it is the K897T change itself, our study analysed a much larger number of individuals and not another polymorphism in tight linkage disequilib- (521 females versus 187) and thus we were able to rium with it, that is functionally responsible for the QTc compare all three genotypes separately (AA versus AC phenotype. In further support of this, sequence analysis of versus CC, as opposed to AA versus AC and CC com- the entire coding region of KCNH2 in six CC homo- bined). In particular, we tested association using 27 CC zygotes and 26 AC heterozygotes (all Caucasian) identified homozygotes whereas the number of homozygotes in the no further amino acid-changing polymorphisms in these other study was five. Notwithstanding the disparity, both individuals. Nevertheless, yet-unknown polymorphisms studies point to a role for the KCNH2 gene in inter- within the non-coding regions that could be in tight linkage individual variability of QTc interval.
disequilibrium with the polymorphism described herein The kinetic differences between the T897 and K897 and that could impact on gene function cannot be com- channels are modest compared with those observed for pletely ruled out.
mutations found in patients with manifest LQTS Although the mechanism whereby the K897T poly- consistent with the proposition that polymorphisms act to morphism is preferentially associated with a shorter modify the phenotype of stresses such as pathogenic baseline QTc in females remains to be elucidated, this mutations or pharmacological provocation Further- finding is important for several reasons. Compared with more, this finding is consistent with the increasingly men, women exhibit a longer QTc both at baseline (this recognized view that inter-individual susceptibility to QT- C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 HERG K -channel dysfunction in an inherited cardiac arrhythmia.
prolongation lies in the co-inheritance of multiple subsets Proc Natl Acad Sci USA 1996;93:2208–2212.
of polymorphisms that act in concert in determining one's anson B, Tuna N, Bouchard T et al. Genetic factors in the risk In such cases, determination of their separate electrocardiogram and heart rate of twins reared apart and together.
contribution requires association studies in large popula- Am J Cardiol 1989;63:606–609.
tion samples and their effect also being substantiated usjahn A, Knoblauch H, Faulhaber HD et al. QT interval is linkedto 2 long-QT syndrome loci in normal subjects. Circulation through functional studies.
riori SG, Napolitano C, Schwartz PJ. Low penetrance in thelong-QT syndrome: clinical impact. Circulation 1999;99:529–533.
riori SG, Napolitano C, Gasparini M et al. Clinical and geneticheterogeneity of right bundle branch block and ST-segment eleva-tion syndrome. A prospective evaluation of 52 families. Circulation Dr. J.M. Ruijter and Prof. A.H. Zwinderman, biostatisti- cians, were consulted for statistical analysis. Ms. M.E.
yndt F, Probst V, Potet F et al. Novel SCN5A mutation leading Jakobs is thanked for assistance in mass spectrometric either to isolated cardiac conduction defect or Brugada syndrome in analysis. This study was financially supported by the a large French family. Circulation 2001;104:3081–3086.
Netherlands Heart Foundation (NHS grants 2000.059 to rugada R, Roberts R. Brugada syndrome: why are there multipleanswers to a simple question? Circulation 2001;104:3017–3019.
C.R.B. and A.A.M.W., and 2002B191 to H.L.T), the ang P, Kanki H, Drolet B et al. Allelic variants in long-QT disease Netherlands Organization for Scientific Research (grant genes in patients with drug-associated torsades de pointes. Circula- 805-06.155 to A.O.V.), and a fellowship to H.L.T. by the Royal Netherlands Academy of Arts and Sciences. More- uscholl MW, Hense HW, Brockel U et al. Changes in left over, support by the Wilhelm-Vaillant-Stiftung, the Ernst- ventricular structure and function in patients with white coathypertension: cross sectional survey. Br Med J 1998;317:565–570.
und-Berta-Grimmke-Stiftung, the Deutsche Herzstiftung, Marban E, Johns DC. Distinct gene-specific mechanisms and the Deutsche Forschungsgemeinschaft (A.J. and H.S.) of arrhythmia revealed by cardiac gene transfer of two long QT is acknowledged. F.C.L. and A.B. received support from disease genes, HERG and KCNE1. Proc Natl Acad Sci USA the European Community, EurHypGen.
[21] J ohns DC, Nuss HB, Marban E. Suppression of neuronal and cardiac transient outward currents by viral gene transfer of dominant-negative Kv4.2 constructs. J Biol Chem 1997;272:31598–31603.
eferences
anguinetti MC, Jiang C, Curran ME, Keating MT. A mechanisticlink between an inherited and an acquired cardiac arrhythmia: uddu PE, Bourassa MG. Prediction of sudden death from QTc HERG encodes the IKr potassium channel. Cell 1995;81:299–307.
interval prolongation in patients with chronic ischemic heart disease.
mith PL, Baukrowitz T, Yellen G. The inward rectification J Electrocardiol 1986;19:203–212.
mechanism of the HERG cardiac potassium channel. Nature chwartz PJ, Wolf S. QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation nyders DJ, Chaudhary A. High affinity open channel block by dofetilide of HERG expressed in a human cell line. Mol Pharmacol hnve S. QT interval prolongation in acute myocardial infarction.
Eur Heart J 1985;6(Suppl. D):85–95.
hou Z, Gong Q, Ye B et al. Properties of HERG channels stably chouten EG, Dekker JM, Meppelink P et al. QT interval prolonga- expressed in HEK 293 cells studied at physiological temperature.
tion predicts cardiovascular mortality in an apparently healthy Biophys J 1998;74:230–241.
population. Circulation 1991;84:1516–1523.
riebe L, Beuckelmann DJ. Simulation study of cellular electric kin PM, Devereux RB, Howard BV et al. Assessment of QT properties in heart failure. Circ Res 1998;82:1206–1223.
interval and QT dispersion for prediction of all-cause and car- armke JW, Ganetzky B. A family of potassium channel genes diovascular mortality in American Indians: The Strong Heart Study.
related to eag in Drosophila and mammals. Proc Natl Acad Sci USA IAMOND Study Group, Brendorp B, Elming H, Jun L et al. QTc ees-Miller JP, Kondo C, Wang L, Duff HJ. Electrophysiological interval as a guide to select those patients with congestive heart characterization of an alternatively processed ERG K failure and reduced left ventricular systolic function who will benefit mouse and human hearts. Circ Res 1997;81:719–726.
auer CK, Engeland B, Wulfsen I et al. RERG is a molecular correlate of the inward-rectifying K current in clonal rat pituitary oden DM. Pharmacogenetics and drug-induced arrhythmias. Car- cells. Receptors Channels 1998;6:19–29.
diovasc Res 2001;50:224–231.
erri M, Benhorin J, Alberti M et al. Electrocardiographic quantita- akkar RR, Fromm BS, Steinman RT et al. Female gender as a risk tion of ventricular repolarization. Circulation 1989;80:1301–1308.
factor for torsades de pointes associated with cardiovascular drugs. J ehmann MH, Timothy KW, Frankovich D et al. Age–gender Am Med Assoc 1993;270:2590–2597.
influence on the rate-corrected QT interval and the QT-heart rate ohnloser SH, Singh BN. Proarrhythmia with class III antiarrhyth- relation in families with genotypically characterized long QT mic drugs: definition, electrophysiologic mechanisms, incidence, syndrome. J Am Coll Cardiol 1997;29:93–99.
predisposing factors, and clinical implications. J Cardiovasc Elec- autaharju PM, Zhou SH, Wong S et al. Sex differences in the evolution of the electrocardiographic QT interval with age. Can J riori SG, Barhanin J, Hauer RN et al. Genetic and molecular basis of cardiac arrhythmias: impact on clinical management. Parts I and areba W, Moss AJ, le Cessie S et al. Risk of cardiac events in II. Circulation 1999;99:518–528.
family members of patients with long QT syndrome. J Am Coll anguinetti MC, Curran ME, Spector PS, Keating MT. Spectrum of C.R. Bezzina et al. / Cardiovascular Research 59 (2003) 27–36 ehmann MH, Hardy S, Archibald D et al. Sex difference in risk of oden DM, Balser JR. A plethora of mechanisms in the HERG- torsade de pointes with d,l-sotalol. Circulation 1996;94:2535–2541.
related long QT syndrome. Genetics meets electrophysiology.
aitinen P, Fodstad H, Piippo K et al. Survey of the coding region of Cardiovasc Res 1999;44:242–246.
the HERG gene in long QT syndrome reveals six novel mutations plawski I, Timothy KW, Tateyama M et al. Variant of SCN5A and an amino acid polymorphism with possible phenotypic effects.
sodium channel implicated in risk of cardiac arrhythmia. Science Mutation in brief [334. Online Hum Mutat 2000;15:580–581.
ietila E, Fodstad H, Niskasaari E et al. Association between HERGK897T polymorphism and QT interval in middle-aged Finnishwomen. J Am Coll Cardiol 2002;40:511–514.

Source: http://www.healthtwist.de/sites/default/files/fachartikel/herg.pdf