Author + information
- Received May 31, 2009
- Revision received August 14, 2009
- Accepted August 24, 2009
- Published online November 24, 2009.
- Wataru Shimizu, MD, PhD⁎,⁎ (, )
- Arthur J. Moss, MD‡,
- Arthur A.M. Wilde, MD, PhD∥,
- Jeffrey A. Towbin, MD#,
- Michael J. Ackerman, MD, PhD⁎⁎,
- Craig T. January, MD, PhD††,
- David J. Tester, BS⁎⁎,
- Wojciech Zareba, MD, PhD‡,
- Jennifer L. Robinson, MS‡,
- Ming Qi, PhD§,
- G. Michael Vincent, MD‡‡,
- Elizabeth S. Kaufman, MD§§,
- Nynke Hofman, MSc¶,
- Takashi Noda, MD, PhD⁎,
- Shiro Kamakura, MD, PhD⁎,
- Yoshihiro Miyamoto, MD, PhD†,
- Samit Shah, BA‡,
- Vinit Amin, MA‡,
- Ilan Goldenberg, MD‡,
- Mark L. Andrews, BBA‡ and
- Scott McNitt, MS‡
- ↵⁎Reprint requests and correspondence:
Dr. Wataru Shimizu, Division of Cardiology, Department of Internal Medicine, National Cardiovascular Center, 5-7-1 Fujishiro-dai, Suita, Osaka 565-8565, Japan
Objectives The purpose of this study was to investigate the effect of location, coding type, and topology of KCNH2(hERG)mutations on clinical phenotype in type 2 long QT syndrome (LQTS).
Background Previous studies were limited by population size in their ability to examine phenotypic effect of location, type, and topology.
Methods Study subjects included 858 type 2 LQTS patients with 162 different KCNH2mutations in 213 proband-identified families. The Cox proportional-hazards survivorship model was used to evaluate independent contributions of clinical and genetic factors to the first cardiac events.
Results For patients with missense mutations, the transmembrane pore (S5-loop-S6) and N-terminus regions were a significantly greater risk than the C-terminus region (hazard ratio [HR]: 2.87 and 1.86, respectively), but the transmembrane nonpore (S1–S4) region was not (HR: 1.19). Additionally, the transmembrane pore region was significantly riskier than the N-terminus or transmembrane nonpore regions (HR: 1.54 and 2.42, respectively). However, for nonmissense mutations, these other regions were no longer riskier than the C-terminus (HR: 1.13, 0.77, and 0.46, respectively). Likewise, subjects with nonmissense mutations were at significantly higher risk than were subjects with missense mutations in the C-terminus region (HR: 2.00), but that was not the case in other regions. This mutation location–type interaction was significant (p = 0.008). A significantly higher risk was found in subjects with mutations located in α-helical domains than in subjects with mutations in β-sheet domains or other locations (HR: 1.74 and 1.33, respectively). Time-dependent β-blocker use was associated with a significant 63% reduction in the risk of first cardiac events (p < 0.001).
Conclusions The KCNH2missense mutations located in the transmembrane S5-loop-S6 region are associated with the greatest risk.
Long QT syndrome (LQTS) is a congenital disorder caused by mutations of several cardiac ion channel genes and is diagnosed clinically by a prolonged QT interval on the electrocardiogram (ECG) and variable clinical outcomes including arrhythmia-related syncope and sudden death (1,2). Mutations involving the KCNH2gene (hERG[human ether-a-go-go-related gene]), which codes for the pore-forming α-subunit of a cardiac K+channel, have been linked to the type 2 LQTS, the second most common variant of LQTS (3). The KCNH2mutations lead to a reduction in the rapid component of the delayed rectifier repolarizing current (IKr), which contributes to lengthening of the QT interval (4). The KCNH2subunits oligomerize to form a tetramer that inserts into the cell membrane to form the functional K+channel. Each subunit comprises 6 α-helical transmembrane segments (S1 to S6), where the K+-selective pore is found between S5 and S6. The transmembrane segments are flanked by amino (N)- and carboxyl (C)-terminus regions (5–8). In a previous study of patients with type 2 LQTS, mutations in the pore region were associated with an increased risk for arrhythmia-related cardiac events when compared with patients with nonpore mutations (9). However, this study was limited by population size in its ability to examine the phenotypic effect of mutations within distinct domains of the nonpore region.
There are several coding types of mutations in genes that form the functional K+channel: missense, nonsense, splice site, in-frame deletion, and frameshift mutations (10). Missense mutations are point mutations that result in a single amino acid change within the protein; nonsense mutations generate a stop codon and can truncate the protein. Insertion and deletion mutations cause in-frame or frameshift mutations, the latter of which change the grouping of nucleotide bases into codons. Splice site mutations may alter splicing of messenger ribonucleic acid. In our recent cohort of type 1 LQTS (11), a missense mutation accounted for 81% of all the mutations, and the type of mutation (missense vs. nonmissense) was not an independent risk factor. On the other hand, nonmissense mutations such as frameshift and nonsense mutations have been reported to be more frequently identified in the type 2 LQTS patients (11,12).
Moreover, topology of mutations (α-helical domain, β-sheet domain, and other uncategorized location) has been recently reported to relate to the function of mutated channel in the type 2 LQTS patients (8).
We hypothesized that the distinct location, coding type, and topology of the channel mutation would have important influence on the phenotypic manifestations and clinical course of patients with type 2 LQTS. To test this hypothesis, we investigated the clinical aspects of 858 subjects having a spectrum of KCNH2mutations categorized by the distinct location, coding type, and topology of the channel mutations.
The study population of 858 subjects was derived from 213 proband-identified families with genetically confirmed KCNH2mutations. The proband in each family had corrected QT (QTc) prolongation not due to a known cause. The subjects were drawn from the U.S. portion of the International LQTS (Rochester) Registry (n = 456), the Netherlands' (Amsterdam) LQTS Registry (n = 214), the Japanese (National Cardiovascular Center) LQTS Registry (n = 95), and the Mayo Clinic LQTS Registry (n = 93). All subjects or their guardians provided informed consent for the genetic and clinical studies. Not included in the study population were 58 subjects with evidence of 2 or more LQTS mutations and an additional 18 who had polymorphisms (p.R176W or p.R1047L) that the authors felt might reduce IKrcurrent. A total of 201 of the 456 patients enrolled from the U.S. portion of the International LQTS Registry and 61 of the 95 patients from the Japanese LQTS Registry were reported in our prior reports (9,12).
Routine clinical and electrocardiographic parameters were acquired at the time of enrollment in each of the registries. Follow-up was censored at age 41 years to minimize the influence of coronary disease on cardiac events. Measured parameters on the first recorded ECG included QT and R-R intervals in milliseconds, with QT corrected for heart rate by Bazett's formula. The QTc interval was expressed in its continuous form and categorized into 4 levels: <460, 460 to 499, 500 to 530, and >530 ms. The QTc interval was categorized into 3 levels: <500, 500 to 530, and >530 ms for the end point of lethal cardiac events (aborted cardiac arrest or LQTS-related sudden cardiac death), because there were few lethal cardiac events in the lowest QTc group (<460 ms). Clinical data were collected on prospectively designed forms with information on demographic characteristics, personal and family medical history, electrocardiographic findings, therapy, and end points during long-term follow-up. Data common to all 4 LQTS registries involving genetically identified patients with type 2 LQTS genotype were electronically merged into a common database for this study.
The KCNH2mutations were identified using standard genetic tests performed in molecular-genetic laboratories in the participating academic centers. From the Rochester registry, 60 subjects died of sudden cardiac death at a young age and were not genotyped. These 60 subjects were assumed to have the same KCNH2mutation as other affected close members of their respective family.
Genetic alterations of the amino acid sequence were characterized by location in the channel protein, by the type of mutation (missense, splice site, in-frame insertions/deletions, nonsense [stop codon], and frameshift), and by the topology of mutation (α-helical domain, β-sheet domain, and other uncategorized location) (Fig. 1).The transmembrane region of the KCNH2encoded channel was defined as the coding sequence involving amino acid residues from 398 through 657 (S5-loop-S6 region: 552 to 657), with the N-terminus region defined before residue 398, and the C-terminus region after residue 657 (Fig. 1) (13,14).
We evaluated the risk associated with 4 main pre-specified regions: 1) N-terminus; 2) transmembrane “nonpore” region (S1–S4); 3) transmembrane “pore” region (S5-loop-S6); and 4) C-terminus. We also evaluated the risk associated with distinct types of mutation and topology of mutation.
Differences in the univariate characteristics by specific groupings were evaluated by standard statistical methods. The primary end point was time to syncope, aborted cardiac arrest, or sudden death, whichever occurred first. The cumulative probability of a first cardiac event was assessed by the Kaplan-Meier method, with significance testing by the log-rank statistic. The Cox proportional-hazards survivorship model was used to evaluate the independent contribution of clinical and genetic factors to the first occurrence of time-dependent cardiac events from birth through age 40 years (15). The Cox regression models, stratified by decade of birth year and allowing for time-dependent covariates, were fit to estimate the adjusted hazard ratio (HR) of each factor as a predictor of first cardiac events. We observed that sex was not proportional as a function of age, with crossover in risk at age 13 on univariate Kaplan-Meier analysis. To fulfill the assumption of proportional hazards for sex over the entire age range, a time-dependent covariate for sex (via an interaction with time) was incorporated, allowing for different hazard ratios by sex before and after age 13 years.
Since almost all subjects were first- and second-degree relatives of probands, the effect of potential lack of independence between subjects was evaluated by refitting the Cox model using the robust sandwich estimator for family membership (16). All significant predictors of risk maintained significance using this robust measure of variance.
Patients who did not have an ECG for QTc measurement were identified in the Cox models as “QTc missing.” Pre-specified covariate interactions between mutation location, type, and α-helical domains were evaluated. Only the mutation location–missense interaction was significant. To test the impact of the interaction between the 4 different mutation locations and missense mutation type, 3 interaction terms were added to the Cox proportional hazards regression model. A 3 degree of freedom likelihood-ratio test was performed to determine their statistical significance. The influence of time-dependent β-blocker therapy (the age at which β-blocker therapy was initiated) on outcome was determined by adding this variable to the final Cox model containing the various covariates.
Total study population
The continuum of KCNH2mutations and their respective number of subjects by location, type, and topology of mutation and contributing registry are presented in the Online Table, and the location, type, and topology of the mutations are diagrammatically presented in Figure 1. A total of 162 different KCNH2mutations were identified in 858 subjects. The mutations were predominantly found in 3 regions: the N-terminus (28.4%, n = 46), the C-terminus (30.9%, n = 50), and the transmembrane domain (40.7%, n = 66). Of the 66 mutations within the transmembrane domain, 78.8% (n = 52) were located within the S5-loop-S6 region. Missense (single amino acid substitutions) accounted for 61.7% (n = 100) of all the mutations, splice site for 1.9% (n = 3), in-frame insertions/deletions for 0.6% (n = 1), nonsense for 10.5% (n = 17), and frameshift for 25.3% (n = 41). Sixty-six mutations (40.7%) were located in the α-helical domain, 17 (10.5%) in the β-sheet domain, and 79 (48.8%) in other uncategorized locations.
The phenotypic characteristics of patients enrolled in each of the 4 registries and by location, type, and topology of mutation are presented in Table 1.The age was younger in the Mayo Clinic registry than in the other 3 registries. The QTc interval was longer and the cardiac events were more frequent in the U.S. and Japanese registries than in the other 2 registries. A pacemaker was more frequently implanted in the U.S. registry, and a defibrillator in the Mayo Clinic registry. LQTS-related death was more frequent in the U.S. registry than in the other 3 registries; that seems mainly because the U.S. registry included the largest proportion of patients missing ECG data and was the longest-standing registry, in which 44 of the 92 deaths occurred before 1980. It is not surprising that the death rate in subjects missing ECG data (i.e., QTc) was very high.
Location, type, and topology of mutation on clinical outcome
As to the location of mutation, the QTc interval was longer and cardiac events were more frequent in patients with mutations in the transmembrane pore locations (S5-loop-S6) than in patients with mutations in transmembrane nonpore (S1 to S4), N-terminus, or C-terminus locations. As to the type of mutation, the QTc interval was longer in patients with missense mutations than in patients with either frameshift/nonsense or other mutations. Sudden death was also more frequent among patients with missense mutations. As to the topology of mutation, the QTc interval was longer and cardiac events were more frequent among patients with mutations located in the α-helical domain than among patients with mutations in either the β-sheet domain or other uncategorized location.
The cumulative probabilities of first cardiac event by type, location, and topology of mutation are presented in Figures 2A,2B, and 2C, respectively. No significant difference in event rates was observed among types of mutation (p = 0.68) (Fig. 2A), although missense mutations were more associated with longer QTc interval and increased risk for sudden death compared with other types of mutations. Conversely, significantly higher event rates were found among subjects with transmembrane pore mutations than among subjects with mutations in transmembrane nonpore, N-terminus, or C-terminus regions, with a gradual increase in event rates occurring during ages 5 to 40 years (Fig. 2B). Significantly higher event rates were also observed among subjects with mutations located in the α-helical domains than among subjects with mutations in either the β-sheet domains or other locations (Fig. 2C).
The findings from the Cox regression analysis by location and by topology of KCNH2mutations for first cardiac events and those for aborted cardiac arrest or LQTS-related sudden cardiac death are presented in Table 2.The clinical risk factors associated with first cardiac events involved males before age 13 years (HR: 1.54 vs. females), females after age 13 years (HR: 3.29 vs. males), and longer QTc intervals (HR: 3.33, QTc >530 ms [n = 112] vs. QTc <460 ms [n = 239]; HR: 2.09, QTc 500 to 530 ms [n = 146] vs. QTc <460 ms; HR: 1.56, QTc 460 to 499 ms [n = 251] vs. QTc <460 ms). Mutations located in the transmembrane pore region made significant and independent contributions to the risk model, with C-terminus region as reference (HR: 1.56). Mutations located in the α-helical domains made significant contributions to the risk model with the β-sheet domains as reference (HR: 1.74). A mutation in the α-helical domain located in the transmembrane pore region would have a risk equal to the multiplicative product of the 2 hazard ratios, namely, 1.56 × 1.74 = 2.71. On the other hand, a mutation in the α-helical domain located in the nonpore transmembrane S1 to S4 region would have a risk of 0.61 × 1.74 = 1.06, and this value was very similar to 1. Time-dependent β-blocker use was associated with a significant 63% reduction in the risk of first cardiac events (p < 0.001). The clinical risk factors associated with lethal cardiac events showed similar tendency to those with cardiac events, and involved females after age 13 years (HR: 2.38 vs. males) and longer QTc intervals (HR: 4.97, QTc >530 ms vs. QTc <500 ms; HR: 2.57, QTc 500 to 530 ms vs. QTc <500 ms). History of prior syncope was a significant risk for lethal cardiac events (HR: 3.42). Time-dependent β-blocker use showed a reduction in the risk of lethal cardiac events by 26%, but this did not reach statistical significance.
Combination of location and type of mutation on clinical outcome
The inter-relation between location, type, and topology of mutation is presented in Table 3.Among 52 mutations within the transmembrane pore region, 46 mutations (88.5%) were missense mutations, and only 6 mutations (11.5%) were frameshift/nonsense mutations. Conversely, frameshift/nonsense mutations were more frequently located in the C-terminus region (31 of 50 mutations, 62.0%); 17 mutations (34.0%) were missense mutation, and the remaining 2 mutations (4.0%) were from any other type (splice mutation). Because transmembrane pore mutations are more risky than mutations in the transmembrane nonpore, C-terminus, or N-terminus regions (Fig. 2B), and there is no significant difference in event rates among the types of mutation (Fig. 2A), nonmissense mutations, mainly frameshift/nonsense mutations in the C-terminus region, may be an independent risk. Therefore, we further investigated the risk associated with a combination of location and type of mutation.
The cumulative probabilities of first cardiac event for missense mutations within different locations are presented in Figure 2D, and those for mutations located in the α-helical domains within different locations are presented in Figure 2E. Significantly higher event rates were found in subjects with missense mutations (Fig. 2D) and mutations in the α-helical domains (Fig. 2E) located in transmembrane pore region than in those located in any other regions. Among 261 patients with frameshift/nonsense mutations, the event rates were not different by location of mutations (data not shown).
The Cox regression analysis by a combination of location and type of mutations for first cardiac events and that for aborted cardiac arrest or LQTS-related sudden cardiac death is presented in Table 4.For patients with missense mutations, the transmembrane pore (S5-loop-S6) and N-terminus regions were a significantly greater risk than the C-terminus region (HR: 2.87 and 1.86, respectively), but the transmembrane nonpore (S1 to S4) region was not (HR: 1.19). However, for nonmissense mutations, these other regions were no longer riskier than the C-terminus (HR: 1.13, 0.77, and 0.46, respectively). Likewise, subjects with nonmissense mutations, mainly frameshift/nonsense mutations, were at significantly higher risk than were subjects with missense mutations in the C-terminus region (HR: 2.00), but that was not the case in other regions. This mutation location-type interaction was significant (p = 0.008). However, a mutation topology-type analysis did not reveal a significant interaction (p = 0.11). Also, the mutation location-type interaction was not seen for the aborted cardiac arrest or LQTS-related sudden cardiac death end point reported in Table 4.
Among subjects with missense mutations, the transmembrane pore region was a significantly higher risk than were the transmembrane nonpore, N-terminus, or C-terminus regions (not shown HR: 2.42, 1.54, and 2.87, respectively). For subjects with nonmissense mutations, the transmembrane pore region was not a significantly higher risk than were the transmembrane nonpore, N-terminus, or C-terminus regions (HR: 2.47, 1.48, and 1.13, respectively). It is interesting to note that, while not significant, the effect sizes (HRs) of the pore risk stay relatively constant across mutation type except in the case of the C-terminus, further evidence of the location-type interaction.
The major findings of the present study from 858 type 2 LQTS subjects with genetically confirmed KCNH2mutations derived from 4 LQTS registries are that: 1) there is a significant mutation type-location interaction; specifically, that the relative risk between C-terminus and the regions is different for missense versus nonmissense locations; 2) patients with missense mutations in the transmembrane pore region have significantly higher cardiac event rates than do patients with missense mutations in the N-terminus, transmembrane nonpore, or C-terminus regions; 3) patients with nonmissense mutations were at significantly higher risk than were patients with missense mutations in the C-terminus region; and 4) patients with mutations located in putative α-helical domains have significantly higher cardiac event rates than do patients with mutations in either the β-sheet domains or other uncategorized locations, and these higher event rates are independent of traditional clinical risk factors and of β-blocker therapy. Our data indicate that risk stratification and specific management or treatment by distinct location, coding type, and topology of the channel mutation in addition to classical risk factors such as QTc, sex, or history of prior syncope may be possible in patients with type 2 LQTS, although further studies are definitely required.
A total of 12 forms of congenital LQTS have been reported (2,4,17–20), and clinical studies for genotype-phenotype correlations have been rigorously investigated in the type 1, 2, and 3 LQTS, which constitute >90% of genotyped patients with LQTS (2,21–25). More recently, mutation-location specific differences in the severity of clinical phenotype have been investigated in each genotype (9,11,12,26,27). As to the type 1 LQTS, a large cohort of 600 patients with KCNQ1mutations has demonstrated that location and biophysical function of mutations were independent risk factors influencing the clinical course (11). However, the distribution of mutation location as well as the frequency of mutation type are reported to be different in each of 3 major genotypes (9,11,12,26,27). More recently, putative secondary structures of α-helices or β-sheet are reported to have an important role on the channel function in the type 2 LQTS (8). Therefore, a larger cohort of patients having a spectrum of KCNH2mutations is required to test the hypothesis that the location, coding type, and topology of mutations would influence the clinical course in the type 2 LQTS.
In contrast to our cohort of 600 type 1 LQTS patients in which the majority of mutations were found in the transmembrane region (66.2%) (11), in the present study, mutations in KCNH2were more evenly distributed in the N-terminus, the transmembrane domain, and the C-terminus. As to the type of mutation, missense mutations dominated (80.5%), and only 13% of the mutations were frameshift/nonsense mutations in our type 1 LQTS cohort (11). In contrast, missense mutations accounted for 61.7%, and frameshift/nonsense mutations were more frequently observed (35.8%) in this type 2 LQTS cohort. Interestingly, most of the mutations located in the transmembrane pore region were missense mutations (46 of 52, 88.5%) in the present study, a finding concordant with the previous type 2 LQTS cohort by Moss et al. (9) (13 of 14, 92.9%). This indicated that the severe phenotype in patients with mutations located in the transmembrane pore region was probably because missense mutations that are expected to cause dominant negative effects were predominant in this region. However, our type 2 LQTS patients with missense mutations located in the N-terminus, transmembrane nonpore, and C-terminus regions were at significantly less risk than were patients with missense mutations in the transmembrane pore region. These data suggest that location of mutation, in other words, the transmembrane pore region, itself was an independent risk in type 2 LQTS patients with KCNH2missense mutation.
Conversely, patients with nonmissense mutations, mainly frameshift/nonsense mutations, were at significantly higher risk than were patients with missense mutations in the C-terminus region, and the event rates in patients with frameshift/nonsense mutations were not different among the transmembrane pore, transmembrane nonpore, N-terminus, and C-terminus regions. Gong et al. (28) recently suggested that most frameshift/nonsense mutations would cause nonsense-mediated decay (NMD), thereby producing less messenger ribonucleic acid from the mutant alleles (28). This potentially would allow for the wild type allele to express more normal channels. Therefore, it is expected that the type 2 LQTS patients with frameshift/nonsense mutation causing NMD would have a mild phenotype. In contrast, the type 2 LQTS patients with frameshift/nonsense mutation without NMD would be expected to have a more severe phenotype because a truncated protein would be produced. Thus, the fact that some frameshift/nonsense mutations show NMD, whereas the other mutations do not, makes the clinical phenotype in the type 2 LQTS patients with frameshift/nonsense mutations more complicated, although this scenario is only a speculation. The present study confirmed the higher risk in patients with nonmissense mutations than in patients with missense mutations in the C-terminus region, suggesting that more careful follow-up is required for type 2 LQTS patients with nonmissense mutations in the C-terminus region.
With regard to the topology of mutation, Anderson et al. (8) recently reported that missense mutations located in a highly ordered structure as α-helices or β-sheet correlated with a class 2 trafficking-deficient phenotype in the type 2 LQTS patients. In the present cohort, mutations located in the α-helical domains were associated with a significantly higher risk compared with mutations in either the β-sheet domains or other uncategorized locations. It is possible that missense mutations in α-helices, where secondary protein structure is thought to be highly ordered, lead to altered secondary and tertiary channel protein structure and abnormal trafficking. This new analysis considering putative secondary structures of mutated channel would be a useful approach in stratifying the risk of cardiac events in patients with LQTS.
β-blockers have long been the first choice of therapy for patients with congenital LQTS (2,29). However, it has been shown in previous studies that the protection that β-blockers provide against cardiac events for type 2 and 3 LQTS patients is somewhat less effective than for type 1 LQTS patients (23,30). A variety of experimental data also support the genotype-specific efficacy of β-blockers for type 1 LQTS (31). In the present study, time-dependent β-blocker use significantly reduced the risk of first cardiac events by 63% (p < 0.001), confirming the efficacy of β-blockers as a first line of therapy in patients with type 2 LQTS as well as suggesting more prophylactic use of β-blockers, especially for high-risk patients with type 2 LQTS. However, β-blocker use was associated with less protection (29%) in the prevention of lethal cardiac events compared to first cardiac events (mostly syncope), indicating that additional treatment such as potassium supplement or an implantable cardioverter-defibrillator implantation may be considered in high-risk patients with type 2 LQTS. The patients who have aborted cardiac arrest/sudden death may have a more malignant pathophysiology that is more resistant to β-blockers than are syncopal episodes. We purposely included “ECG missing” in the Cox model so that the β-blocker effect is actually adjusted for subjects with “ECG missing” who probably did not receive β-blockers.
We did not evaluate the risk associated with distinct type of biophysical ion-channel dysfunction (dominant-negative or haplotype insufficient), because only a small percentage of the mutations present within our patient population have been studied extensively in identical cellular expression experiments. There were 60 patients who were not genotyped, and they had an increased risk for events mainly because their fatal events occurred at a young age before they were genotyped. When these patients were excluded from the analysis, the pattern of risk in the missense and nonmissense subgroups remains similar to that of the total population, but the significance of the effect is attenuated because of the reduced number of events.
For a table on the KCNH2mutations by location, coding type, and topology of mutation, and contributing registry, please see the online version of this article.
Genotype-Phenotype Aspects of Type-2 Long-QT Syndrome
Dr. Ackerman has a consulting relationship and license agreement/royalty arrangement with PGxHealth (FAMILION). This work was supported in part by a Health Sciences Research Grant (H18-Research on Human Genome-002) and a Research Grant for the Cardiovascular Diseases (21C-8) from the Ministry of Health, Labour and Welfare, Japan (to Dr. Shimizu); research grants HL-33843 and HL-51618 (to Dr. Moss) and HL-60723 (to Dr. January) from the National Institutes of Health, Bethesda, Maryland; and grant 2000.059 from the Nederlandse Hartstichting, Amsterdam, the Netherlands (to Dr. Wilde). Dr. Ackerman has received support from Medtronic, PGxHealth, and Pfizer. Dr. January has received support from Cellular Dynamics International. Dr. Tester receives modest royalties from PGxHealth. Dr. Kaufman receives research support from CardioDx and St. Jude Medical. Drs. Shimizu, Moss, Wilde, Towbin, Ackerman, and January contributed equally to the original concept of this investigation.
- Abbreviations and Acronyms
- rapid component of the delayed rectifier repolarizing current
- long QT syndrome
- nonsense-mediated decay
- corrected QT
- Received May 31, 2009.
- Revision received August 14, 2009.
- Accepted August 24, 2009.
- American College of Cardiology Foundation
- Shimizu W.
- Anderson C.L.,
- Delisle B.P.,
- Anson B.D.,
- et al.
- Moss A.J.,
- Zareba W.,
- Kaufman E.S.,
- et al.
- McConkey E.H.
- Moss A.J.,
- Shimizu W.,
- Wilde A.A.,
- et al.
- Splawski I.,
- Shen J.,
- Timothy K.W.,
- et al.
- Swiss Institute of Bioinformatics
- Cox D.R.
- Ueda K.,
- Valdivia C.,
- Medeiros-Domingo A.,
- et al.
- Wilde A.A.,
- Jongbloed R.J.,
- Doevendans P.A.,
- et al.
- Schwartz P.J.,
- Priori S.G.,
- Spazzolini C.,
- et al.
- Shimizu W.,
- Horie M.,
- Ohno S.,
- et al.
- Gong Q.,
- Zhang L.,
- Vincent G.M.,
- Horne B.D.,
- Zhou Z.
- Moss A.J.,
- Zareba W.,
- Hall W.J.,
- et al.
- Shimizu W.,
- Antzelevitch C.