Abstract

Previous reports have described several associations of PR, QRS, QT and heart rate with genomic variations by genome-wide association studies (GWASs). In the present study, we examined the association of ∼2.5 million SNPs from 2994 Japanese healthy volunteers obtained from the JPDSC database with electrocardiographic parameters. We confirmed associations of PR interval, QRS duration and QT interval in individuals of Japanese ancestry with 11 of the 45 SNPs (6 of 20 for QT, 5 of 19 for PR and 0 of 6 for QRS) observed among individuals of European, African and Asian (Indian and Korean) ancestries. Those results indicate that many of the electrocardiographic associations with genes are shared by different ethnic groups including Japanese. Possible novel associations found in this study were validated by Korean data. As a result, we identified a novel association of SNP rs4952632[G] (maps near SLC8A1, sodium–calcium exchanger) (P = 7.595 × 10−6) with PR interval in Japanese individuals, and replication testing among Koreans confirmed the association of the same SNP with prolonged PR interval. Meta-analysis of the Japanese and Korean datasets demonstrated highly significant associations of SNP rs4952632[G] with a 2.325-ms (95% CI, 1.693–2.957 ms) longer PR interval per minor allele copy (P = 5.598 × 10−13). Cell-type-specific SLC8A1 knockout mice have demonstrated a regulatory role of sodium–calcium exchanger in automaticity and conduction in sinoatrial node, atrium and atrioventricular node. Our findings support a functional role of sodium–calcium exchanger in human atrial and atrioventricular nodal conduction as suggested by genetically modified mouse models.

INTRODUCTION

The electrocardiographic (ECG) parameters, PR interval, QRS duration and QT interval, can be regarded as quantitative individual traits that are affected by both environmental and genetic factors and that vary considerably even among healthy subjects. For example, QT interval is inherently longer in women than in men and is inversely correlated with heart rate. Thus, QT interval is usually corrected against the heart rate to allow interpretation independent of heart rate variability (corrected QT; QTc). Indeed, each ECG parameter reflects a specific cardiac function. In brief, PR interval corresponds to the sum of atrial and atrioventricular nodal conduction, the disturbance of which is related to a higher incidence of atrial fibrillation (1–3) and associated with a potent risk for pacemaker implantation (2), heart failure, stroke and all-cause mortality (4). The width of the QRS complex reflects ventricular depolarization and conduction time, and longer QRS duration is a predictor of mortality, sudden death and incident heart failure (5–7). The QT interval has long been useful as a clinical index of the duration of ventricular repolarization, and prolongation of QT interval confers an increased risk of sudden cardiac death (8,9).

ECG phenotypes have substantial heritable components, and the search for common genetic variants influencing these quantitative ECG traits is clinically important. Recent genome-wide association studies (GWASs) have identified genetic determinants of PR interval (10–12), QRS duration (13,14) and QT interval (15–17) in individuals of various ancestries. However, few studies have examined genetic factors associated with ECG parameters in East Asian populations including Japanese. The present study analyzed genome-wide data obtained from 3041 healthy Japanese volunteers in a Japan Pharmacogenomics Data Science Consortium (JPDSC) project to examine whether we could replicate previously described genetic associations with the three ECG parameters. Any associated SNPs in previously unrecognized loci were also tested for associations in a Korean population.

RESULTS

Clinical and electrocardiographic characteristics of JPDSC data

The phase 1 JPDSC dataset obtained from a total of 1032 Japanese healthy volunteers (619 males and 413 females) and the phase 2 JPDSC dataset obtained from a total of 2009 Japanese healthy volunteers (1002 males and 1007 females) were used for the present analysis. The mean, median, minimum and maximum of several clinical parameters (age, BMI, heart rate, systolic and diastolic blood pressure, serum potassium and serum calcium) and ECG parameters [PR = PR interval (ms), QRS = the width of the QRS complex (ms), QT = QT interval (ms)] are shown in Supplementary Material, Table S1.

Clinical factors accounting for individual variation in ECG parameters

We used multiple linear regression analysis to estimate clinical factors accounting for individual variation in ECG parameters, after removing outliers defined as values either higher or lower than 4 × SD from the mean (Supplementary Material, Fig. S1, Supplementary Material, Table S2). We performed preliminary analysis to test whether gender, age, log-transformed (log) BMI, log HR, systolic pressure, diastolic pressure, serum potassium, serum calcium and geographic region in Japan should be included as covariates. Based on these results, we used gender, age, log HR and log BMI as covariates for log-transformed (log) PR, gender, age, log HR and systolic pressure as covariates for log QRS, and gender, age and log (60/HR) as covariates for log QT. We did not include serum potassium or serum calcium as covariates because contribution rates of these variables to each ECG parameter were low. The results of our GWASs on ECG parameters are summarized in Fig. 1.

Figure 1.

The Manhattan plots and quantile–quantile plots of genome-wide association results from analysis of Japanese datasets are shown.

Figure 1.

The Manhattan plots and quantile–quantile plots of genome-wide association results from analysis of Japanese datasets are shown.

Replication of previous associations

We examined whether previously described SNPs associated with QT interval, PR interval or QRS duration are also associated in individuals of Japanese ancestry (Table 1), as described in the following text.

Table 1.

The comparison to previous GWAS

Previous reports
 
Present study
 
SNP in original paper Alleles (allele corresponding to positive beta)a P-value References Tested SNPb R-squarec Allelesd MAFe P-valuef Betag 
(A) QT 
ATP1B1 
rs1320976 AG (G) 2.E−10 18 sa  AG 0.093 5.16E−01 0.475 
rs10919071 GA (A) 1.E−15 10 rs7540067 0.529 AG 0.129 1.36E−01 0.980 
SLC8A1 
rs13017846 GA (A) 8.E−14 17 rs4952632 0.975 GT 0.278 3.28E−04* −1.772 
GPR133 
rs885389 GA (G) 4.E−08 19 rs7489187 0.421 GA 0.285 4.14E−01 −0.399 
SLC35F1, C6orf204, PLN 
rs11153730 CT (C) 2.E−29 20 rs72967533 0.772 GA 0.215 1.30E−04* 2.036 
NOS1AP 
rs12143842 TC (T) 2.E−78 10,16 sa  TC 0.39 3.92E−18* 3.888 
CNOT1, GINS3, NDRG4, SLC38A7, GOT2 
rs37062 GA (A) 3.E−25 16 rs3743567 0.323 AG 0.166 2.69E−02 −1.302 
c6orf204, SLC35F1, PLN, ASF1A 
rs11756438 AC (A) 5.E−22 16 rs9489456 0.58 AG 0.298 1.69E−02 1.131 
KCNQ1 
rs12296050 TC (T) 8.E−11 16 rs7939542 AG 0.364 5.52E−04* 1.552 
KCNH2 
rs4725982 CT (T) 5.E−16 16 sa  CT 0.321 3.24E−01 −0.466 
rs2968864 CT (T) 8.E−16 16 rs2968863 AG 0.042 1.11E−01 −1.699 
rs2968863 TC (C) 2.E−15 10 sa  TC 0.042 1.11E−01 −1.699 
LITAF, CLEC16A, SNN, ZC3H7A, TNFRSF17 
rs8049607 CT (T) 5.E−15 16 sa  CT 0.413 1.09E−01 −0.705 
SCN5A 
rs11129795 AG (G) 5.E−14 10 sa  AG 0.098 5.67E−01 −0.420 
rs12053903 CT (T) 1.E−14 16 sa  CT 0.486 3.27E−01 −0.425 
LIG3, RFFL 
rs2074518 TC (C) 6.E−12 16 rs3135967 0.968 TC 0.189 4.76E−01 −0.390 
NDRG4 
rs7188697 GA (A) 7.E−25 16 rs9926577 0.969 CT 0.329 2.46E−05* −1.936 
PLN 
rs11970286 TC (T) 2.E−24 16 sa  TC 0.182 1.47E−04* 2.141 
RNF207 
rs846111 CG (G) 4.E−16 16 rs709209 0.707 GA 0.338 1.28E−01 0.715 
KCNJ2 
rs17779747 TG (G) 6.E−12 10 sa  TC 0.107 9.43E−01 −0.050 
(B) PR 
Previous reports Present study 
SCN5A 
rs3922844 TC (C) 5.E−43 10 sa  TC 0.132 7.16E−02 −1.271 
rs11708996 GC (C) 6.E−26 11 sa  GC 0.041 5.67E−01 0.694 
TBX5 
rs1895585 AG (A) 1.E−19 10 rs7312625 GA 0.26 5.15E−01 0.348 
rs3825214 GA (G) 3.E−12 10 sa  GA 0.442 2.44E−01 0.565 
TBX5, TBX3 
rs1896312 CT (C) 3.E−17 11 sa  CT 0.425 7.75E−03 1.295 
CAV1 
rs11773845 CA (C) 4.E−12 10 rs7801180 0.599 CA 0.425 5.70E−04* 1.679 
rs3807989 AG (A) 7.E−13 10 sa  AG 0.337 2.46E−06* 2.423 
MEIS1 
rs3891585 AG (A) 1.E−11 11 rs997153 0.973 AG 0.252 2.80E−02 1.206 
rs11897119 CT (C) 5.E−11 10 rs997153 AG 0.252 2.80E−02 1.206 
ITGA9 
rs267567 GA (A) 4.E−11 11 rs768354 0.436 GA 0.155 7.44E−01 0.214 
ARHGAP24 
rs11732231 CG (C) 3.E−09 11 rs34949750 0.445 GA 0.122 7.87E−01 0.199 
rs7660702 TC (T) 3.E−17 10 sa  TC 0.122 7.67E−01 0.218 
rs7692808 GA (G) 6.E−20 11 rs7660702 AG 0.122 7.67E−01 0.218 
SCN10A 
rs6801957 TC (T) 9.E−09 11 sa  TC 0.218 2.75E−07* 3.012 
rs6800541 CT (C) 2.E−74 11 sa  CT 0.152 6.53E−06* 3.033 
rs6795970 AG (A) 1.E−58 10 sa  AG 0.153 2.51E−05* 2.815 
SOX5, C12orf67 
rs11047543 AG (G) 3.E−13 10 rs4963776 0.902 AC 0.121 4.05E−01 −0.600 
NKX2-5, C5orf41 
rs251253 TC (T) 9.E−13 10 sa  TC 0.152 3.28E−01 0.656 
MYH6 
rs365990 GA (G) 9.E−11 10 sa  GA 0.132 3.76E−01 −0.621 
(C) QRS 
Previous reports Present study 
TBX5 
rs3825214 GA (G) 3.E−13 10 sa  GA 0.442 3.35E−01 0.218 
SCN10A 
rs6795970 AG (A) 5.E−27 14 sa  AG 0.153 2.75E−01 0.339 
CDKN1A 
rs1321313 AG (A) 5.E−25 21 sa  AG 0.13 2.35E−02 0.763 
SCN5A 
rs1805126 CT (C) 3.E−20 14 sa  CT 0.489 7.95E−01 0.058 
NFIA 
rs2207790 TC (C) 6.E−18 21 rs6587924 0.978 AC 0.421 3.00E−02 −0.487 
C6orf204 
rs6906287 CT (C) 6.E−16 14 rs9489456 0.739 AG 0.298 3.77E−01 0.216 
Previous reports
 
Present study
 
SNP in original paper Alleles (allele corresponding to positive beta)a P-value References Tested SNPb R-squarec Allelesd MAFe P-valuef Betag 
(A) QT 
ATP1B1 
rs1320976 AG (G) 2.E−10 18 sa  AG 0.093 5.16E−01 0.475 
rs10919071 GA (A) 1.E−15 10 rs7540067 0.529 AG 0.129 1.36E−01 0.980 
SLC8A1 
rs13017846 GA (A) 8.E−14 17 rs4952632 0.975 GT 0.278 3.28E−04* −1.772 
GPR133 
rs885389 GA (G) 4.E−08 19 rs7489187 0.421 GA 0.285 4.14E−01 −0.399 
SLC35F1, C6orf204, PLN 
rs11153730 CT (C) 2.E−29 20 rs72967533 0.772 GA 0.215 1.30E−04* 2.036 
NOS1AP 
rs12143842 TC (T) 2.E−78 10,16 sa  TC 0.39 3.92E−18* 3.888 
CNOT1, GINS3, NDRG4, SLC38A7, GOT2 
rs37062 GA (A) 3.E−25 16 rs3743567 0.323 AG 0.166 2.69E−02 −1.302 
c6orf204, SLC35F1, PLN, ASF1A 
rs11756438 AC (A) 5.E−22 16 rs9489456 0.58 AG 0.298 1.69E−02 1.131 
KCNQ1 
rs12296050 TC (T) 8.E−11 16 rs7939542 AG 0.364 5.52E−04* 1.552 
KCNH2 
rs4725982 CT (T) 5.E−16 16 sa  CT 0.321 3.24E−01 −0.466 
rs2968864 CT (T) 8.E−16 16 rs2968863 AG 0.042 1.11E−01 −1.699 
rs2968863 TC (C) 2.E−15 10 sa  TC 0.042 1.11E−01 −1.699 
LITAF, CLEC16A, SNN, ZC3H7A, TNFRSF17 
rs8049607 CT (T) 5.E−15 16 sa  CT 0.413 1.09E−01 −0.705 
SCN5A 
rs11129795 AG (G) 5.E−14 10 sa  AG 0.098 5.67E−01 −0.420 
rs12053903 CT (T) 1.E−14 16 sa  CT 0.486 3.27E−01 −0.425 
LIG3, RFFL 
rs2074518 TC (C) 6.E−12 16 rs3135967 0.968 TC 0.189 4.76E−01 −0.390 
NDRG4 
rs7188697 GA (A) 7.E−25 16 rs9926577 0.969 CT 0.329 2.46E−05* −1.936 
PLN 
rs11970286 TC (T) 2.E−24 16 sa  TC 0.182 1.47E−04* 2.141 
RNF207 
rs846111 CG (G) 4.E−16 16 rs709209 0.707 GA 0.338 1.28E−01 0.715 
KCNJ2 
rs17779747 TG (G) 6.E−12 10 sa  TC 0.107 9.43E−01 −0.050 
(B) PR 
Previous reports Present study 
SCN5A 
rs3922844 TC (C) 5.E−43 10 sa  TC 0.132 7.16E−02 −1.271 
rs11708996 GC (C) 6.E−26 11 sa  GC 0.041 5.67E−01 0.694 
TBX5 
rs1895585 AG (A) 1.E−19 10 rs7312625 GA 0.26 5.15E−01 0.348 
rs3825214 GA (G) 3.E−12 10 sa  GA 0.442 2.44E−01 0.565 
TBX5, TBX3 
rs1896312 CT (C) 3.E−17 11 sa  CT 0.425 7.75E−03 1.295 
CAV1 
rs11773845 CA (C) 4.E−12 10 rs7801180 0.599 CA 0.425 5.70E−04* 1.679 
rs3807989 AG (A) 7.E−13 10 sa  AG 0.337 2.46E−06* 2.423 
MEIS1 
rs3891585 AG (A) 1.E−11 11 rs997153 0.973 AG 0.252 2.80E−02 1.206 
rs11897119 CT (C) 5.E−11 10 rs997153 AG 0.252 2.80E−02 1.206 
ITGA9 
rs267567 GA (A) 4.E−11 11 rs768354 0.436 GA 0.155 7.44E−01 0.214 
ARHGAP24 
rs11732231 CG (C) 3.E−09 11 rs34949750 0.445 GA 0.122 7.87E−01 0.199 
rs7660702 TC (T) 3.E−17 10 sa  TC 0.122 7.67E−01 0.218 
rs7692808 GA (G) 6.E−20 11 rs7660702 AG 0.122 7.67E−01 0.218 
SCN10A 
rs6801957 TC (T) 9.E−09 11 sa  TC 0.218 2.75E−07* 3.012 
rs6800541 CT (C) 2.E−74 11 sa  CT 0.152 6.53E−06* 3.033 
rs6795970 AG (A) 1.E−58 10 sa  AG 0.153 2.51E−05* 2.815 
SOX5, C12orf67 
rs11047543 AG (G) 3.E−13 10 rs4963776 0.902 AC 0.121 4.05E−01 −0.600 
NKX2-5, C5orf41 
rs251253 TC (T) 9.E−13 10 sa  TC 0.152 3.28E−01 0.656 
MYH6 
rs365990 GA (G) 9.E−11 10 sa  GA 0.132 3.76E−01 −0.621 
(C) QRS 
Previous reports Present study 
TBX5 
rs3825214 GA (G) 3.E−13 10 sa  GA 0.442 3.35E−01 0.218 
SCN10A 
rs6795970 AG (A) 5.E−27 14 sa  AG 0.153 2.75E−01 0.339 
CDKN1A 
rs1321313 AG (A) 5.E−25 21 sa  AG 0.13 2.35E−02 0.763 
SCN5A 
rs1805126 CT (C) 3.E−20 14 sa  CT 0.489 7.95E−01 0.058 
NFIA 
rs2207790 TC (C) 6.E−18 21 rs6587924 0.978 AC 0.421 3.00E−02 −0.487 
C6orf204 
rs6906287 CT (C) 6.E−16 14 rs9489456 0.739 AG 0.298 3.77E−01 0.216 

*Significant.

aNucleotide in the parenthesis denotes one that increases the phenotype.

bWhen the reported SNP is present in our platform, the same SNP was genotyped (sa) whereas a different proxy SNP that is in tight LD was genotyped.

cr2-value between the original SNP and the proxy SNP.

dAlleles of the SNP genotyped in our study. The order of the alleles here corresponds to that in the original SNP in Column 2. Thus, the left-side nucleotide in this column is associated with the left-side nucleotide in Column 2.

eMAF: minor allele frequency of the SNP genotyped in our study.

fThreshold P-value was 0.0011.

gStandardized regression coefficient was converted to ms.

QT interval

We attempted to examine 20 SNPs reported to be associated with QT interval (Table 1). Reported genes for these SNPs were ATP1B1 (two SNPs), SLC8A1, GPR133, SLC35F1/C6orf204/PLN, NOS1AP, CNOT1/GINS3/NDRG4/SLC38A7/GOT2, c6orf204/SLC35F1/PLN/ASF1A, KCNQ1, KCNH2 (three SNPs), LITAF/CLEC16A/SNN/ZC3H7A/TNFRSF17, SCN5A (two SNPs), LIG3/RFFL, NDRG4, PLN, RNF207 and KCNJ2 (Table 1). When the SNPs reported in the previous studies were included in our platform, we used those SNPs; however, SNPs that were in tight linkage disequilibrium (LD) with the reported SNPs were used otherwise. Among the 20 SNPs examined, associations with QT interval were replicated in our data for 6 SNPs. Among them, the replications were performed with the same SNPs as in the previous reports in two SNPs, i.e. rs12143842 (NOS1AP) and rs11970286 (PLN) whereas proxy SNPs that were in tight linkage equilibrium were used for the other four SNPs, i.e. rs4952632 (SLC8A1), rs72967533 (SLC35F1, C6orf204, PLN), rs7939542 (KCNQ1) and rs9926577 (NDRG4) (Table 1). Consistent with previous GWAS studies, the strongest main association signal maps to the NOS1AP locus, and there were 12 SNPs in this area with genome-wide significant associations (P < 5 × 10−8) in our data. The most significant SNP in NOS1AP in our study was rs10494365[C] (MAF = 0.385, P = 3.117 × 10−19). The NOS1AP SNP rs12143842[A] (MAF = 0.390, P = 3.918 × 10−18) is in strong LD with our top signal, rs10494365 (r2 = 0.802 in JPDSC), and was the most significant SNP in NOS1AP reported by previous GWAS of QT interval in populations of European descent (15,16).

The rs10919071[A] in ATP1B1, the index SNP in the QTSCD study (15), was not included in our Illumina Chip. We searched for proxy SNPs that were in tight LD and found rs7540067 (Table 1). However, in our data, P-value for this SNP was only 0.136. Another SNP rs1320976 that was reported to be associated with QT interval in a previous study was present in our data, but P-value in our data was only 0.516 (Table 1). Instead, the SNP in ATP1B1 that showed the lowest P-value in our study was rs11809180[G] (MAF = 0.335, P = 9.720 × 10−7), which is in low LD with rs10919071 (r2 = 0.176 in the 1000 Genome JPT). The rs12296050[T] in KCNQ1, the index SNP for the QTSCD study (15), was not included in the Illumina Chip, but a proxy SNP rs7939542 showed a significant association with QT interval in our study (Table 1). The SNP in KCNQ1 that showed the lowest P-value in individuals of Japanese descent was rs7947981[G] (MAF = 0.111, P = 3.003 × 10−7).

PR interval

We attempted to examine 19 SNPs reported to be associated with PR (Table 1). Reported genes for these SNPs were SCN5A (two SNPs), TBX5 (two SNPs), TBX5-TBX3, CAV1 (two SNPs), MEIS1 (two SNPs), ITGA9, ARHGAP24 (three SNPs), SCN10A (three SNPs), SOX5-C12orf67, NKX2-5-C5orf41 and MYH6 (Table 1). Among the 19 SNPs examined, associations with PR were replicated in our data for 5 SNPs. Among them, the replications were performed with the same SNPs as in the previous reports in four SNPs, i.e. rs3807989 (CAV1), rs6801957 (SCN10A), rs6800541 (SCN10A) and rs6795970 (SCN10A), whereas proxy SNPs that were in tight linkage equilibrium were used for one SNP, i.e. rs7801180 (CAV1) (Table 1).

QRS duration

We attempted to examine six SNPs reported to be associated with QRS (Table 1). Four of the six SNPs were in our platform whereas the other two were not, and the associations of the latters were examined by proxy SNPs (Table 1). We identified significant associations for none of the six SNPs with QRS duration (Table 1) although MAFs of the SNPs were not very low in JPDSC data (0.13–0.489).

Possible novel loci associated with QT interval, PR interval and QRS duration

Using a pre-specified P-value threshold of 1 × 10−5, we searched for novel genetic loci harboring significant SNPs associated with QT interval, PR interval and QRS duration in Japanese. We selected seven SNPs at four loci (one each associated with QT interval and QRS duration, and two with PR interval) based on a quality assessment survey of SNP genotyping and clarity of cluster (Tables 2–4). Among seven SNPs surpassing a pre-specified threshold of 1 × 10−5 for Japanese datasets, six SNPs at three loci (one with QRS duration and two with PR interval) were replicated in a Korean dataset (n = 6805 people of Korean ancestry), i.e. SLC8A1 and DCP1A|CACNA1D were associated with PR interval, and KLHL38 was associated with QRS duration (Tables 2–4).

Table 2.

Meta-analysis of Japanese and Korean datasets for possible novel association (A) QT NCRNA00092|LOC100128771

Summary SNP (rs) rs10819861 rs4078421a 
 Chr 
Position 98 867 813 98 863 169 
Minor  
Major  
JPDSC n 2981  
MAF 0.476  
Coefficient (ms) 2.298  
P-value 1.128E−07  
Korea n  6491 
MAF  0.387 
Coefficient (ms)  0.695 
P-valueb  1.421E−02 
Meta-analysis Coefficient (ms)  1.157 
P-valuec  6.629E−07 
Summary SNP (rs) rs10819861 rs4078421a 
 Chr 
Position 98 867 813 98 863 169 
Minor  
Major  
JPDSC n 2981  
MAF 0.476  
Coefficient (ms) 2.298  
P-value 1.128E−07  
Korea n  6491 
MAF  0.387 
Coefficient (ms)  0.695 
P-valueb  1.421E−02 
Meta-analysis Coefficient (ms)  1.157 
P-valuec  6.629E−07 

GWASs, genome-wide association studies; JPDSC, Japan Pharmacogenomics Data Science Consortium; KARE, Korea Association Resource; KoGES, Korean Genome and Epidemiology Study; ECG, electrocardiographic; SNP, single-nucleotide polymorphism; MAF, minor allele frequency; NCX1, sodium–calcium exchanger.

aProxy SNP that was in LD with rs10819861 (r2 = 1).

bThreshold P-value was 0.007.

cThreshold for genome-wide significance was 5E−8.

Table 3.

Meta-analysis of Japanese and Korean datasets for possible novel association (B) PR SLC8A1|LOC400950

Summary SNP (rs) rs17026114 rs4952632 rs13026826  
 Chr  
Position 40 743 183 40 748 005 40 753 603  
Minor  
Major  
JPDSC n 2974 2976 2976  
MAF 0.278 0.278 0.278  
Coefficient (ms) 2.438 2.452 2.419  
P-value 8.908E−06 7.595E−06 1.080E−05  
Korea n 6710 6700 6689  
MAF 0.389 0.389 0.389  
Coefficient (ms) 2.229 2.259 2.263  
P-value 2.117E−08* 1.392E−08* 1.320E−08*  
Meta-analysis Coefficient (ms) 2.300 2.325 2.316  
P-value 1.013E−12* 5.598E−13* 7.320E−13*  
DCP1A|CACNA1D 
Summary SNP (rs) rs10452033 rs77862767 rs12631120a rs1384970a 
 Chr 
 Position 53 441 376 53 436 024 53 437 002 53 445 491 
 Minor   
 Major   
JPDSC n 2975 2974   
 MAF 0.136 0.124   
 Coefficient (ms) −3.193 −2.985   
 P-value 3.852E−06 3.617E−05   
Korea n   6512 6693 
 MAF   0.169 0.166 
 Coefficient (ms)   −1.413 −1.477 
 P-value   5.522E−03* 4.377E−03* 
Meta-analysis Coefficient (ms)   −2.086 −2.122 
 P-value   6.713E−07 3.976E−07 
Summary SNP (rs) rs17026114 rs4952632 rs13026826  
 Chr  
Position 40 743 183 40 748 005 40 753 603  
Minor  
Major  
JPDSC n 2974 2976 2976  
MAF 0.278 0.278 0.278  
Coefficient (ms) 2.438 2.452 2.419  
P-value 8.908E−06 7.595E−06 1.080E−05  
Korea n 6710 6700 6689  
MAF 0.389 0.389 0.389  
Coefficient (ms) 2.229 2.259 2.263  
P-value 2.117E−08* 1.392E−08* 1.320E−08*  
Meta-analysis Coefficient (ms) 2.300 2.325 2.316  
P-value 1.013E−12* 5.598E−13* 7.320E−13*  
DCP1A|CACNA1D 
Summary SNP (rs) rs10452033 rs77862767 rs12631120a rs1384970a 
 Chr 
 Position 53 441 376 53 436 024 53 437 002 53 445 491 
 Minor   
 Major   
JPDSC n 2975 2974   
 MAF 0.136 0.124   
 Coefficient (ms) −3.193 −2.985   
 P-value 3.852E−06 3.617E−05   
Korea n   6512 6693 
 MAF   0.169 0.166 
 Coefficient (ms)   −1.413 −1.477 
 P-value   5.522E−03* 4.377E−03* 
Meta-analysis Coefficient (ms)   −2.086 −2.122 
 P-value   6.713E−07 3.976E−07 

GWASs, genome-wide association studies; JPDSC, Japan Pharmacogenomics Data Science Consortium; KARE, Korea Association Resource; KoGES, Korean Genome and Epidemiology Study; ECG, electrocardiographic; SNP, single-nucleotide polymorphism; MAF, minor allele frequency; NCX1, sodium–calcium exchanger.

*Significant.

aProxy SNPs tdat were in LD witd rs10452033 (r2 = 1).

Table 4.

Meta-analysis of Japanese and Korean datasets for possible novel association (C) QRS KLHL38

Summary SNP (rs) rs11991744 
 Chr 
Position 124 659 329 
Minor 
Major 
JPDSC n 2975 
MAF 0.230 
Coefficient (ms) −1.255 
P-value 1.935E−06 
Korea n 6796 
MAF 0.267 
Coefficient (ms) −0.562 
P-value 1.796E−03* 
Meta-analysis Coefficient (ms) −0.787 
P-value 1.251E−07 
Summary SNP (rs) rs11991744 
 Chr 
Position 124 659 329 
Minor 
Major 
JPDSC n 2975 
MAF 0.230 
Coefficient (ms) −1.255 
P-value 1.935E−06 
Korea n 6796 
MAF 0.267 
Coefficient (ms) −0.562 
P-value 1.796E−03* 
Meta-analysis Coefficient (ms) −0.787 
P-value 1.251E−07 

GWASs, genome-wide association studies; JPDSC, Japan Pharmacogenomics Data Science Consortium; KARE, Korea Association Resource; KoGES, Korean Genome and Epidemiology Study; ECG, electrocardiographic; SNP, single-nucleotide polymorphism; MAF, minor allele frequency; NCX1, sodium–calcium exchanger.

*Significant.

Meta-analysis of Japanese dataset and Korean dataset for possible novel associations

To obtain P-values based on the results from our Japanese data and those from the Korean data, we performed a meta-analysis by the inverse variance method (Tables 2–4). We observed a highly significant association of SNP rs4952632 in SLC8A1|LOC400950 with a 2.325-ms (95% CI, 1.693–2.957 ms) longer PR interval per minor allele copy (P = 5.598 × 10−13). This SNP was submitted to the eQTL analysis using the data from lymphoblastoid cell lines from HapMap JPT samples, but the results were negative (P = 0.139, rs4952632, probe ID = ILMN_1699520_SLC8A1).

The results of meta-analysis of the other possible loci did not reach pre-specified genome-wide significance (5 × 10−8) (Tables 2–4), and further studies are required to replicate these findings, identify the causal variants at or near these loci, characterize their functional significance and determine whether they predict cardiovascular events.

DISCUSSION

A genome-wide association approach is an established methodology that enables the identification of genetic loci and variants associated with complex traits and diseases unconstrained by prior knowledge. However, it is important with such an approach to robustly identify the candidate genes to guide future deep sequencing studies.

In the present study, we examined the association of ∼2.5 million SNPs from 2994 Japanese healthy volunteers obtained from the JPDSC database with ECG parameters. We identified associations of PR interval, QRS duration and QT interval in individuals of Japanese ancestry with the SNPs in the same genes as observed among individuals of European (10,13–16), African (11) and Asian [Indian (12) and Korean (17)] ancestries. These findings support the concept that genetic architecture for myocardial electrophysiological activity appears to be similar among individuals from different populations around the world. Primary signals reported here were not necessarily the same as the index SNPs reported in other populations.

We identified a novel association of SNP rs4952632[G] in SLC8A1 (P = 7.595 × 10−6) with PR interval in Japanese individuals. The meta-analysis using both our and Korean samples gave a P-value lower than the genome-wide threshold (Tables 2–4). SLC8A1 encodes a sodium–calcium exchanger (NCX1) that serves as a key sarcolemmal protein for the maintenance of calcium homeostasis in the heart. By regulating intracellular Ca2+ levels, Na+–Ca2+ exchange is a determinant of cardiac contractility, and acute alterations in exchange activity have major effects on contractile strength. For example, small changes in intracellular Na+ in response to digitalis produce positive inotropy. In the absence of NCX1, it would be expected that cardiac myocytes would be overloaded with Ca2+. Unexpectedly, the ventricular cardiomyocyte-specific NCX1 knockout mice live to adulthood with only modestly reduced cardiac function (22). In addition to its role in Ca2+ handling, NCX1 plays an important role in the electrical activity of the heart. As the consequence of NCX1 activation generates a net inward current (three Na+ in and one Ca2+ out), NCX1 is electrogenic (23). Mice lacking NCX1 selectively in cells of the cardiac pacemaking and conduction system developed severe bradycardia and rhythm disturbances including SN arrhythmia, SN pauses, atrioventricular (AV) block and ventricular tachycardia (24).

We also identified possible associations of PR interval with the DCP1A|CACNA1D locus, and QRS duration with the KLHL38 locus in our Japanese datasets, and confirmed them in the Korean datasets, although none of them reached the genome-wide significance level.

CACNA1D, encoding the voltage-dependent L-type calcium channel subunit alpha-1D, also known as Ca(v)1.3, is a strong regional candidate associated with variations in PR interval. Although the main L-type calcium channel in the heart is Ca(v)1.2 (alpha-1C) Ca2+ channel, Ca(v)1.3 Ca2+ channel is predominantly expressed in the sinoatrial node, atria, AN node and proximal ventricular conduction system along with Cav1.2. In contrast, ventricles express only Cav1.2 (25–27). Ca(v)1.3 Ca2+ channel-null mice demonstrated sinus bradycardia with prolonged PR interval and were prone to develop atrial fibrillation (28).

There was a suggestive association of QRS duration with KLHL38, 1 of 43 Kelch-like (KLHL) family members encoded in the human genome (29). KLHL proteins contain one BTB/POZ domain, one BACK domain and five to six Kelch motifs and are implicated in the ubiquitination process, although the specific roles for KLHL38 have not yet been elucidated.

We identified some loci associated with PR interval, QRS duration and QT interval in the population of Japanese ancestry. However, we were unable to confirm previously reported associations of these loci with the above-mentioned phenotypes at the significance level required for GWASs. These failures may imply that our study was underpowered to detect these loci.

In summary, our results demonstrated that the JPDSC dataset is useful to map previously unrecognized loci to ECG parameters in Japanese and that many of the associations between genes and QT, PR or QRS are shared worldwide.

MATERIAL AND METHODS

JPDSC

The JPDSC comprising six pharmaceutical companies, Astellas, Daiichi Sankyo, Mitsubishi Tanabe, Otsuka, Taisho and Takeda, has assembled a database for conducting pharmacogenomics (PGx) studies in Japanese subjects. The database contains the genotypes of 2.5 million single-nucleotide polymorphisms (SNPs) and 5 human leukocyte antigen loci per person from 2994 Japanese healthy volunteers, as well as 121 kinds of clinical information, including self-reports, physiological data, hematological data and biochemical data. DNA and clinical information from healthy volunteers were obtained in 2 phases; the first set of samples was collected from 2000 to 2003 and the rest of the samples were collected from 10 geographic regions in Japan (second phase). All the samples were collected from the Japanese population as validated by the principal component analysis. Written informed consent was obtained from all the subjects, and this study was approved by the ethical committee of JPDSC.

GWAS

The final number of subjects included in the GWAS was 2994. Genomic DNAs were genotyped on an Illumina Human Omn 2.5.8 BeadChip, and SNPs used in the GWAS were selected according to the following criteria: SNP call rates (≥95%), the Hardy–Weinberg equilibrium test (P ≥ 0.01) and minor allele frequency (≥1%). ECG measures were adjusted for independent variables within each phase by the linear regression model. The standardized residuals in each phase were combined and used as quantitative phenotypes for the association analysis, performed with PLINK (version 1.1.3). The significance of the association between a trait and genotype was evaluated using Wald test. Genome-wide significance was inferred at P < 5 × 10−8. A minimal set of summary results of all SNPs will be sent to researchers on requests.

Replications of previously reported SNPs

Among the SNPs reported to be associated with QT, PR and QRS, the associations with P-values of <5 × 10−8 in previous reports were examined. As GWAS platforms were different between ours and those used in previous reports, some SNPs previously reported to be associated with ECG traits were not present in our data. In those cases, we searched for proxy SNPs with high r2 values with the reported SNPs. Using both previously reported SNPs and proxy SNPs with high r2 values, we attempted to replicate the previously reported associations. As we attempted to replicate 45 different associations, the P-value threshold of 0.05/45 = 0.0011 was used according to Bonferroni's correction. The direction of each association was carefully examined to see whether the direction of the association in the previous report is the same as in the present study. When a proxy SNP rather than the same SNP was used in the replication, it was carefully examined whether the directions of the effects are concordant between the previous study and our study.

Replications of novel SNPs

SNPs with P-values under the pre-specified threshold of 1 × 10−5 were tested for replication using datasets obtained from Korea Association Resource [KARE (n = 6805)] dataset, which was a subset of Korean Genome and Epidemiology Study (KoGES) cohort (17). The covariates used for the correction in the replication study included age, sex, systolic and diastolic blood pressure, heart rate, height, weight and resident area of participants (i.e. Ansan or Ansung city).

As we attempted to replicate seven different SNPs, the P-value threshold of 0.05/7 = 0.007 was used according to Bonferroni's correction. The direction of each association was carefully examined to see whether the direction of the association in the previous report is the same as in the present study. Genomic DNAs were genotyped on Affymetrix 5.0 SNP array. Meta-analysis was performed using inverse variance method.

Statistical analysis

All statistical analyses were performed using R environment (version 2.15.0).

eQTL analysis

eQTL analysis was performed according to Yang et al. (30) using the data from lymphoblastoid cell lines (31) from HapMap JPT samples. Software used was Genevar (GENe Expression VARiation) by Sanger Institute (30).

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was supported by notable awards for research achievements in science and technology (Takeda Science Foundation). Korean Study Funding source: The Korean genotype and epidemiological data were provided by the Korean Genome Analysis Project (4845-301) and by the Korean Genome and Epidemiology Study (4851-302), funded by the Ministry for Health and Welfare, Republic of Korea.

ACKNOWLEDGEMENTS

The authors thank the Japan Pharmacogenomics Data Science Consortium (JPDSC), which is composed of Astellas Pharma, Inc., Otsuka Pharmaceutical Co., Ltd, Daiichi-Sankyo Co., Ltd, Taisho Pharmaceutical Co., Ltd, Takeda Pharmaceutical Co, Ltd and Mitsubishi Tanabe Pharma Corporation and chaired by Dr Ichiro Nakaoka (Takeda Pharmaceutical Co, Ltd) for kindly providing the data.

Conflict of Interest statement. None declared.

REFERENCES

1
Soliman
E.Z.
Prineas
R.J.
Case
L.D.
Zhang
Z.M.
Goff
D.C.
Jr.
Ethnic distribution of ECG predictors of atrial fibrillation and its impact on understanding the ethnic distribution of ischemic stroke in the Atherosclerosis Risk in Communities (ARIC) study
Stroke.
 , 
2009
, vol. 
40
 (pg. 
1204
-
1211
)
2
Cheng
S.
Keyes
M.J.
Larson
M.G.
McCabe
E.L.
Newton-Cheh
C.
Levy
D.
Benjamin
E.J.
Vasan
R.S.
Wang
T.J.
Long-term outcomes in individuals with prolonged PR interval or first-degree atrioventricular block
JAMA.
 , 
2009
, vol. 
301
 (pg. 
2571
-
2577
)
3
Schnabel
R.B.
Sullivan
L.M.
Levy
D.
Pencina
M.J.
Massaro
J.M.
D'Agostino
R.B.
Sr.
Newton-Cheh
C.
Yamamoto
J.F.
Magnani
J.W.
Tadros
T.M.
, et al.  . 
Development of a risk score for atrial fibrillation (Framingham Heart Study): a community-based cohort study
Lancet.
 , 
2009
, vol. 
373
 (pg. 
739
-
745
)
4
Roy
D.
Talajic
M.
Dubuc
M.
Thibault
B.
Guerra
P.
Macle
L.
Khairy
P.
Atrial fibrillation and congestive heart failure
Curr. Opin. Cardiol.
 , 
2009
, vol. 
24
 (pg. 
29
-
34
)
5
Desai
A.D.
Yaw
T.S.
Yamazaki
T.
Kaykha
A.
Chun
S.
Froelicher
V.F.
Prognostic Significance of Quantitative QRS duration
Am. J. Med.
 , 
2006
, vol. 
119
 (pg. 
600
-
606
)
6
Elhendy
A.
Hammill
S.C.
Mahoney
D.W.
Pellikka
P.A.
Relation of QRS duration on the surface 12-lead electrocardiogram with mortality in patients with known or suspected coronary artery disease
Am. J. Cardiol.
 , 
2005
, vol. 
96
 (pg. 
1082
-
1088
)
7
Oikarinen
L.
Nieminen
M.S.
Viitasalo
M.
Toivonen
L.
Jern
S.
Dahlof
B.
Devereux
R.B.
Okin
P.M.
QRS duration and QT interval predict mortality in hypertensive patients with left ventricular hypertrophy: the Losartan Intervention for Endpoint Reduction in Hypertension Study
Hypertension.
 , 
2004
, vol. 
43
 (pg. 
1029
-
1034
)
8
Tomaselli
G.F.
Beuckelmann
D.J.
Calkins
H.G.
Berger
R.D.
Kessler
P.D.
Lawrence
J.H.
Kass
D.
Feldman
A.M.
Marban
E.
Sudden cardiac death in heart failure. The role of abnormal repolarization
Circulation.
 , 
1994
, vol. 
90
 (pg. 
2534
-
2539
)
9
Schouten
E.G.
Dekker
J.M.
Meppelink
P.
Kok
F.J.
Vandenbroucke
J.P.
Pool
J.
QT interval prolongation predicts cardiovascular mortality in an apparently healthy population
Circulation.
 , 
1991
, vol. 
84
 (pg. 
1516
-
1523
)
10
Pfeufer
A.
van Noord
C.
Marciante
K.D.
Arking
D.E.
Larson
M.G.
Smith
A.V.
Tarasov
K.V.
Muller
M.
Sotoodehnia
N.
Sinner
M.F.
, et al.  . 
Genome-wide association study of PR interval
Nat. Genet.
 , 
2010
, vol. 
42
 (pg. 
153
-
159
)
11
Butler
A.M.
Yin
X.
Evans
D.S.
Nalls
M.A.
Smith
E.N.
Tanaka
T.
Li
G.
Buxbaum
S.G.
Whitsel
E.A.
Alonso
A.
, et al.  . 
Novel loci associated with PR interval in a genome-wide association study of 10 African American cohorts
Circ. Cardiovasc. Genet.
 , 
2012
, vol. 
5
 (pg. 
639
-
646
)
12
Chambers
J.C.
Zhao
J.
Terracciano
C.M.
Bezzina
C.R.
Zhang
W.
Kaba
R.
Navaratnarajah
M.
Lotlikar
A.
Sehmi
J.S.
Kooner
M.K.
, et al.  . 
Genetic variation in SCN10A influences cardiac conduction
Nat. Genet.
 , 
2010
, vol. 
42
 (pg. 
149
-
152
)
13
Sotoodehnia
N.
Isaacs
A.
de Bakker
P.I.
Dorr
M.
Newton-Cheh
C.
Nolte
I.M.
van der Harst
P.
Muller
M.
Eijgelsheim
M.
Alonso
A.
, et al.  . 
Common variants in 22 loci are associated with QRS duration and cardiac ventricular conduction
Nat. Genet.
 , 
2010
, vol. 
42
 (pg. 
1068
-
1076
)
14
Holm
H.
Gudbjartsson
D.F.
Arnar
D.O.
Thorleifsson
G.
Thorgeirsson
G.
Stefansdottir
H.
Gudjonsson
S.A.
Jonasdottir
A.
Mathiesen
E.B.
Njolstad
I.
, et al.  . 
Several common variants modulate heart rate, PR interval and QRS duration
Nat. Genet.
 , 
2010
, vol. 
42
 (pg. 
117
-
122
)
15
Pfeufer
A.
Sanna
S.
Arking
D.E.
Muller
M.
Gateva
V.
Fuchsberger
C.
Ehret
G.B.
Orru
M.
Pattaro
C.
Kottgen
A.
, et al.  . 
Common variants at ten loci modulate the QT interval duration in the QTSCD Study
Nat. Genet.
 , 
2009
, vol. 
41
 (pg. 
407
-
414
)
16
Newton-Cheh
C.
Eijgelsheim
M.
Rice
K.M.
de Bakker
P.I.
Yin
X.
Estrada
K.
Bis
J.C.
Marciante
K.
Rivadeneira
F.
Noseworthy
P.A.
, et al.  . 
Common variants at ten loci influence QT interval duration in the QTGEN Study
Nat. Genet.
 , 
2009
, vol. 
41
 (pg. 
399
-
406
)
17
Kim
J.W.
Hong
K.W.
Go
M.J.
Kim
S.S.
Tabara
Y.
Kita
Y.
Tanigawa
T.
Cho
Y.S.
Han
B.G.
Oh
B.
A common variant in SLC8A1 is associated with the duration of the electrocardiographic QT interval
Am. J. Hum. Genet.
 , 
2012
, vol. 
91
 (pg. 
180
-
184
)
18
Smith
J.G.
Avery
C.L.
Evans
D.S.
Nalls
M.A.
Meng
Y.A.
Smith
E.N.
Palmer
C.
Tanaka
T.
Mehra
R.
Butler
A.M.
, et al.  . 
Impact of ancestry and common genetic variants on QT interval in African Americans
Circ. Cardiovasc. Genet.
 , 
2012
, vol. 
5
 (pg. 
647
-
655
)
19
Marroni
F.
Pfeufer
A.
Aulchenko
Y.S.
Franklin
C.S.
Isaacs
A.
Pichler
I.
Wild
S.H.
Oostra
B.A.
Wright
A.F.
Campbell
H.
, et al.  . 
A genome-wide association scan of RR and QT interval duration in 3 European genetically isolated populations: the EUROSPAN project
Circ. Cardiovasc. Genet.
 , 
2009
, vol. 
2
 (pg. 
322
-
328
)
20
Nolte
I.M.
Wallace
C.
Newhouse
S.J.
Waggott
D.
Fu
J.
Soranzo
N.
Gwilliam
R.
Deloukas
P.
Savelieva
I.
Zheng
D.
, et al.  . 
Common genetic variation near the phospholamban gene is associated with cardiac repolarisation: meta-analysis of three genome-wide association studies
PLoS One.
 , 
2009
, vol. 
4
 pg. 
e6138
 
21
Ritchie
M.D.
Denny
J.C.
Zuvich
R.L.
Crawford
D.C.
Schildcrout
J.S.
Bastarache
L.
Ramirez
A.H.
Mosley
J.D.
Pulley
J.M.
Basford
M.A.
, et al.  . 
Genome- and phenome-wide analyses of cardiac conduction identifies markers of arrhythmia risk
Circulation.
 , 
2013
, vol. 
127
 (pg. 
1377
-
1385
)
22
Henderson
S.A.
Goldhaber
J.I.
So
J.M.
Han
T.
Motter
C.
Ngo
A.
Chantawansri
C.
Ritter
M.R.
Friedlander
M.
Nicoll
D.A.
, et al.  . 
Functional adult myocardium in the absence of Na+-Ca2+ exchange: cardiac-specific knockout of NCX1
Circ. Res.
 , 
2004
, vol. 
95
 (pg. 
604
-
611
)
23
Janvier
N.C.
Boyett
M.R.
The role of Na-Ca exchange current in the cardiac action potential
Cardiovasc. Res.
 , 
1996
, vol. 
32
 (pg. 
69
-
84
)
24
Herrmann
S.
Lipp
P.
Wiesen
K.
Stieber
J.
Nguyen
H.
Kaiser
E.
Ludwig
A.
The cardiac sodium-calcium exchanger NCX1 is a key player in the initiation and maintenance of a stable heart rhythm
Cardiovasc. Res.
 , 
2013
, vol. 
99
 (pg. 
780
-
788
)
25
Mangoni
M.E.
Couette
B.
Bourinet
E.
Platzer
J.
Reimer
D.
Striessnig
J.
Nargeot
J.
Functional role of L-type Cav1.3 Ca2+ channels in cardiac pacemaker activity
Proc. Natl Acad. Sci. USA.
 , 
2003
, vol. 
100
 (pg. 
5543
-
5548
)
26
Platzer
J.
Engel
J.
Schrott-Fischer
A.
Stephan
K.
Bova
S.
Chen
H.
Zheng
H.
Striessnig
J.
Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2+ channels
Cell.
 , 
2000
, vol. 
102
 (pg. 
89
-
97
)
27
Zhang
Z.
Xu
Y.
Song
H.
Rodriguez
J.
Tuteja
D.
Namkung
Y.
Shin
H.S.
Chiamvimonvat
N.
Functional Roles of Ca(v)1.3 (alpha(1D)) calcium channel in sinoatrial nodes: insight gained using gene-targeted null mutant mice
Circ. Res.
 , 
2002
, vol. 
90
 (pg. 
981
-
987
)
28
Zhang
Z.
He
Y.
Tuteja
D.
Xu
D.
Timofeyev
V.
Zhang
Q.
Glatter
K.A.
Xu
Y.
Shin
H.S.
Low
R.
, et al.  . 
Functional roles of Cav1.3(alpha1D) calcium channels in atria: insights gained from gene-targeted null mutant mice
Circulation.
 , 
2005
, vol. 
112
 (pg. 
1936
-
1944
)
29
Dhanoa
B.S.
Cogliati
T.
Satish
A.G.
Bruford
E.A.
Friedman
J.S.
Update on the Kelch-like (KLHL) gene family
Hum. Genomics.
 , 
2013
, vol. 
7
 pg. 
13
 
30
Yang
T.P.
Beazley
C.
Montgomery
S.B.
Dimas
A.S.
Gutierrez-Arcelus
M.
Stranger
B.E.
Deloukas
P.
Dermitzakis
E.T.
Genevar: a database and Java application for the analysis and visualization of SNP-gene associations in eQTL studies
Bioinformatics.
 , 
2010
, vol. 
26
 (pg. 
2474
-
2476
)
31
Stranger
B.E.
Montgomery
S.B.
Dimas
A.S.
Parts
L.
Stegle
O.
Ingle
C.E.
Sekowska
M.
Smith
G.D.
Evans
D.
Gutierrez-Arcelus
M.
, et al.  . 
Patterns of cis regulatory variation in diverse human populations
PLoS Genet.
 , 
2012
, vol. 
8
 pg. 
e1002639
 

Supplementary data