- Email: [email protected]
Association of genetic variants in SEMA3F, CLEC16A, LAMA3, and PCSK2 with myocardial infarction in Japanese individuals
Atherosclerosis 210 (2010) 468–473
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Association of genetic variants in SEMA3F, CLEC16A, LAMA3, and PCSK2 with myocardial infarction in Japanese individuals Tetsuo Fujimaki a , Kimihiko Kato a , Kiyoshi Yokoi a , Mitsutoshi Oguri b , Tetsuro Yoshida c , Sachiro Watanabe d , Norifumi Metoki e , Hidemi Yoshida f , Kei Satoh f , Yukitoshi Aoyagi g , Yoshinori Nozawa h , Genjiro Kimura i , Yoshiji Yamada j,∗ a
Department of Cardiovascular Medicine, Gifu Prefectural Tajimi Hospital, Tajimi, Japan Department of Cardiology, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan c Department of Cardiovascular Medicine, Inabe General Hospital, Inabe, Japan d Department of Cardiology, Gifu Prefectural General Medical Center, Gifu, Japan e Department of Internal Medicine, Hirosaki Stroke Center, Hirosaki, Japan f Department of Vascular Biology, Institute of Brain Science, Hirosaki University Graduate School of Medicine, Hirosaki, Japan g Department of Genomics for Longevity and Health, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan h Gifu International Institute of Biotechnology and Tokai Gakuin University, Kakamigahara, Japan i Department of Cardio-Renal Medicine and Hypertension, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan j Department of Human Functional Genomics, Life Science Research Center, Mie University, 1577 Kurima-machiya, Tsu, Mie 514-8507, Japan b
a r t i c l e
i n f o
Article history: Received 28 October 2009 Received in revised form 28 November 2009 Accepted 30 November 2009 Available online 5 December 2009 Keywords: Myocardial infarction Coronary heart disease Atherosclerosis Genetics Polymorphism
a b s t r a c t Objective: The purpose of the present study was to identify genetic variants that confer susceptibility to myocardial infarction (MI) in Japanese individuals. Methods: The study population comprised 5014 Japanese individuals, including 1444 subjects with MI and 3570 controls. The 150 polymorphisms examined in the present study were selected by a genomewide association study for ischemic stroke with the use of the GeneChip Human Mapping 500K Array Set (Affymetrix), and were determined by a method that combines the polymerase chain reaction and sequence-specific oligonucleotide probes with suspension array technology. Results: An initial screen by the chi-square test revealed that the A→G polymorphism of SEMA3F (rs12632110), the C→T polymorphism of CLEC16A (rs9925481), the A→G polymorphism of LAMA3 (rs12373237), and the C→G polymorphism of PCSK2 (rs6080699) were significantly (false discovery rate for allele frequencies of <0.05) associated with MI. Subsequent multivariable logistic regression analysis with adjustment for covariates and a stepwise forward selection procedure revealed that the A→G polymorphism of SEMA3F (dominant model; P = 0.0014; odds ratio, 0.76), the C→T polymorphism of CLEC16A (dominant model; P = 0.0009; odds ratio, 0.75), the A→G polymorphism of LAMA3 (recessive model; P = 0.0099; odds ratio, 0.80), and the C→G polymorphism of PCSK2 (recessive model; P = 0.0155; odds ratio, 1.19) were significantly (P < 0.05) associated with the prevalence of MI. Conclusion: Determination of these genotypes may prove informative for assessment of the genetic risk for MI in Japanese individuals. © 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Myocardial infarction (MI) is an important clinical problem because of its large contribution to mortality. Given that the disease prevention is an important strategy for reducing the overall burden of MI, identification of markers for MI is essential for both risk prediction and intervention for reducing the chance of future events. Epidemiological studies have underscored the contribution
∗ Corresponding author. Tel.: +81 59 231 5387; fax: +81 59 231 5388. E-mail address: [email protected] (Y. Yamada). 0021-9150/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2009.11.050
of conventional risk factors such as smoking, dyslipidemia, hypertension, diabetes mellitus, and obesity in the development of MI [1]. In addition to these risk factors, recent genetic studies have suggested the importance of genetic factors and of interactions between genetic and environmental factors in predisposition to MI [2,3]. Although recent genome-wide association studies (GWAS) have identified various polymorphisms that confer susceptibility to MI [4–9], genetic variants for predisposition to MI in Japanese individuals have not determined definitively. We have performed an association study of MI in a total of 5014 Japanese individuals with 150 single nucleotide polymorphisms (SNPs) selected from a GWAS for ischemic stroke [10]. The purpose
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
469
Table 1 Baseline characteristics of the study subjects with myocardial infarction (MI) and controls. Characteristic
MI (n = 1444)
Controls (n = 3570)
P value
Age (years) Sex (male/female) Body mass index (kg/m2 ) Current or former smoker (%) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum HDL–cholesterol (mmol/l) Serum triglycerides (mmol/l)
65.8 ± 10.1 1137/307 23.8 ± 3.3 26.9 90.6 ± 94.3 71.2 142 ± 27 75 ± 15 48.0 7.64 ± 3.48 6.59 ± 1.80 28.5 5.15 ± 1.05 1.19 ± 0.35 1.73 ± 1.15
65.8 ± 11.2 1600/1970 23.4 ± 3.3 19.0 68.7 ± 39.3 39.4 137 ± 21 79 ± 12 13.8 6.48 ± 2.77 5.60 ± 1.32 24.8 5.13 ± 0.91 1.48 ± 0.39 1.54 ± 1.04
0.9101 <0.0001 0.0002 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0070 0.5853 <0.0001 <0.0001
Quantitative data are means ± SD. Hypertension: systolic blood pressure of ≥140 mmHg, diastolic blood pressure of ≥90 mmHg, or taking of antihypertensive medication. Diabetes mellitus: fasting plasma glucose level of ≥6.93 mmol/l, blood glycosylated hemoglobin (hemoglobin A1c ) content of ≥6.5%, or taking of antidiabetes medication. Hypercholesterolemia: serum total cholesterol concentration of ≥5.72 mmol/l or taking of lipid-lowering medication. HDL, high-density lipoprotein.
of the present study was to identify genetic variants that confer susceptibility to MI in Japanese individuals, and thereby to contribute to the personalized prevention of this condition. 2. Methods 2.1. Study population The study population comprised 5014 unrelated Japanese individuals, including 1444 subjects with MI and 3570 controls, who either visited outpatient clinics of or were admitted to one of the participating hospitals (Gifu Prefetural General Medical Center and Gifu Prefectural Tajimi Hospital in Gifu Prefecture, Japan; Hirosaki University Hospital, Reimeikyo Rehabilitation Hospital, and Hirosaki Stroke Center in Aomori Prefecture, Japan) between October 2002 and March 2008 with various symptoms or for an annual health check up, or who were recruited to a populationbased prospective cohort study of aging and age-related diseases in Nakanojo, Gunma Prefecture, Japan. The 1444 subjects with MI (1137 men, 307 women) all underwent coronary angiography and left ventriculography. The diagnosis of MI was based on typical electrocardiographic changes and on increases both in the serum activity of creatine kinase (MB isozyme) and in the serum concentration of troponin T. The diagnosis was confirmed by the presence of wall motion abnormality on left ventriculography and by identification of the responsible stenosis in any of the major coronary arteries or in the left main trunk by coronary angiography. The 3570 control subjects (1600 men, 1970 women) had no history of coronary heart disease, aortic aneurysm, or peripheral arterial occlusive disease; of ischemic or hemorrhagic stroke or other cerebral diseases; or of other atherosclerotic, thrombotic, embolic or hemorrhagic disorders. The study protocol complied with the Declaration of Helsinki and was approved by the committees on the Ethics of Human Research of Mie University Graduate School of Medicine, Hirosaki University Graduate School of Medicine, Gifu International Institute of Biotechnology, Tokyo Metropolitan Institute of Gerontology, and participating hospitals. Written informed consent was obtained from each participant. 2.2. Selection of polymorphisms Our aim was to identify genetic variants associated with MI in the Japanese population in a case–control association study. A total of 150 polymorphisms examined in the present study (Supplementary Table 1) were selected from an initial screen by a GWAS of ischemic stroke (atherothrombotic cerebral infarc-
tion) on the basis of P value for allele frequency <1.0 × 10−7 with the use of the GeneChip Human Mapping 500K Array Set (Affymetrix, Santa Clara, CA) [10]. Given that the main cause of MI and atherothrombotic cerebral infarction is atherothrombosis, genetic variants associated with MI might be included in those related to atherothrombotic cerebral infarction. We have not examined the relation of these polymorphisms to MI in our previous studies [11–13]. 2.3. Genotyping polymorphisms Venous blood (7 ml) was collected into tubes containing ethylenediaminetetraacetic acid at a final concentration of 50 mM, and genomic DNA was isolated with a kit (Genomix; Talent, Trieste, Italy). Genotypes of 150 polymorphisms were determined at G&G Science (Fukushima, Japan) by a method that combines the polymerase chain reaction and sequence-specific oligonucleotide probes with suspension array technology (Luminex, Austin, TX, USA). Primers, probes, and other conditions for genotyping of polymorphisms related [false discovery rate (FDR) < 0.05] to MI as determined by the initial chi-square test are shown in Supplementary Table 2. Detailed genotyping methodology was described previously [14]. 2.4. Statistical analysis Quantitative data were compared between subjects with MI and controls by the unpaired Student’s t-test. Categorical data were compared by the chi-square test. Allele frequencies were estimated by the genome counting method, and the chi-square test was used to identify departures from Hardy–Weinberg equilibrium. In the initial screen, the genotype distribution (3 × 2) and allele frequencies (2 × 2) of each polymorphism were compared between subjects with MI and controls by the chi-square test. Given the multiple comparisons of genotypes with MI, the FDR was calculated from the distributions of P values for the allele frequencies of the 150 polymorphisms [15]. Polymorphisms with a FDR for allele frequency of <0.05 were further examined by multivariable logistic regression analysis with adjustment for covariates. Such analysis was thus performed with MI as a dependent variable and independent variables including age, sex (0, woman; 1, man), body mass index (BMI), smoking status (0, nonsmoker; 1, smoker), the serum concentration of creatinine, the history of hypertension, diabetes mellitus, and hypercholesterolemia (0, no history; 1, positive history), and genotype of each polymorphism; and the P value, odds ratio, and 95% confidence interval were calculated. Each geno-
470
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Table 2 Polymorphisms related (P value for allele frequency <0.05) to myocardial infarction (MI) as revealed by the chi-square test. Gene
Polymorphism
dbSNP
SEMA3F AA AG GG
A→G
rs12632110
CLEC16A CC CT TT
C→T
LAMA3 AA AG GG
A→G
PCSK2 CC CG GG
C→G
SIX5 GG GT TT
G→T
COL13A1 AA AC CC USP37 AA AG GG
A→C
COLEC11 CC CT TT
C→T
ITPK1 CC CT TT
C→T
LRP5 CC CT TT
C→T
COL6A3 CC CT TT
C→T
SS18L1 CC CT TT
C→T
TSPAN9 AA AG GG
A→G
a
A→G
MIa
Controlsa
356 (24.8) 706 (49.1) 375 (26.1)
720 (20.4) 1759 (50.0) 1040 (29.6)
1142 (79.4) 277 (19.3) 18 (1.3)
2636 (75.0) 815 (23.2) 63 (1.8)
24 (1.7) 335 (23.3) 1078 (75.0)
49 (1.4) 676 (19.2) 2788 (79.4)
96 (6.7) 553 (38.8) 777 (54.5)
317 (8.9) 1448 (40.8) 1782 (50.3)
1003 (69.8) 393 (27.4) 41 (2.8)
2585 (73.5) 871 (24.7) 62 (1.8)
37 (2.6) 374 (26.0) 1026 (71.4)
70 (2.0) 795 (22.6) 2656 (75.4)
1194 (83.7) 222 (15.6) 10 (0.7)
2846 (80.2) 663 (18.7) 38 (1.1)
52 (3.6) 503 (35.3) 871 (61.1)
199 (5.6) 1269 (35.8) 2079 (58.6)
842 (58.6) 505 (35.1) 90 (6.3)
1935 (55.0) 1343 (38.1) 243 (6.9)
724 (50.8) 571 (40.0) 131 (9.2)
1904 (53.7) 1370 (38.6) 274 (7.7)
0 (0.0) 42 (2.9) 1384 (97.1)
0 (0.0) 69 (1.9) 3478 (98.1)
1290 (90.4) 135 (9.5) 1 (0.1)
3138 (88.5) 402 (11.3) 7 (0.2)
231 (16.1) 689 (47.9) 517 (36.0)
640 (18.2) 1694 (48.1) 1185 (33.7)
rs9925481
rs12373237
rs6080699
rs16980013
rs942576
rs526897
rs6739899
rs2295394
rs3736228
rs11690358
rs2427254
rs2011973
P value (genotype)
P value (allele)
FDR
0.0013
0.0004
0.0469
0.0030
0.0006
0.0469
0.0035
0.0011
0.0469
0.0050
0.0012
0.0469
0.0054
0.0024
0.0646
0.0110
0.0025
0.0646
0.0139
0.0034
0.0744
0.0117
0.0173
0.3252
0.0640
0.0263
0.4185
0.0898
0.0280
0.4185
0.0308
0.0318
0.4185
0.0926
0.0368
0.4185
0.1240
0.0430
0.4185
Values in parentheses are percentages.
type was assessed according to dominant, recessive, and additive genetic models. Additive models included the additive 1 (heterozygotes versus wild-type homozygotes) and the additive 2 (variant homozygotes versus wild-type homozygote) models, which were analyzed simultaneously with a single statistical model. We also performed a stepwise forward selection procedure to examine the effects of genotypes as well as of other covariates on MI. In this analysis, each genotype was examined according to a dominant or
recessive model on the basis of statistical significance in the multivariable logistic regression analysis. The P levels for inclusion in and exclusion from the model were 0.25 and 0.1, respectively. With the exception of the initial screen by the chi-square test (FDR < 0.05), a P value of <0.05 was considered statistically significant. Statistical significance was examined by two-sided tests performed with JMP version 6.0 and JMP Genomics version 3.2 software (SAS Institute, Cary, NC).
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
471
Table 3 Multivariable logistic regression analysis of polymorphisms related (FDR for allele frequency < 0.05) to myocardial infarction by the chi-square test with adjustment for age, sex, body mass index, smoking status, serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia. Gene
SEMA3F CLEC16A LAMA3 PCSK2
Polymorphism
A→G C→T A→G C→G
Dominant
Recessive
P
OR (95%CI)
P
0.0014 0.0009 0.9575 0.1342
0.76 (0.64–0.90) 0.75 (0.63–0.89)
0.0566 0.3845 0.0099 0.0155
Additive 1
Additive 2
OR (95%CI)
P
OR (95%CI)
P
OR (95%CI)
0.78 (0.65–0.93) 0.75 (0.63–0.89)
0.0016 0.2840 0.9115 0.0536
0.72 (0.59–0.88)
0.80 (0.68–0.95) 1.19 (1.03–1.38)
0.0068 0.0014 0.5308 0.4156
OR, odds ratio; CI, confidence interval.
3. Results The characteristics of the study subjects are shown in Table 1. The frequency of male subjects, BMI, the percentage of smokers, systolic blood pressure, serum concentrations of creatinine and triglycerides, fasting plasma glucose level, blood glycosylated hemoglobin content, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia were greater, whereas diastolic blood pressure and serum concentration of high-density lipoprotein (HDL)–cholesterol were lower in subjects with MI than in controls. Comparisons of allele frequencies of polymorphisms between subjects with MI and controls by the chi-square test revealed that 13 polymorphisms were related (P < 0.05) to the prevalence of MI (Table 2). Among these polymorphisms, the A→G polymorphism of SEMA3F (rs12632110), the C→T polymorphism of CLEC16A (rs9925481), the A→G polymorphism of LAMA3 (rs12373237), and the C→G polymorphism of PCSK2 (rs6080699) were significantly (FDR for allele frequency of <0.05) associated with MI. The genotype distributions for the 13 polymorphisms related to MI are also shown in Table 2. Genotype distributions of four polymorphisms significantly associated with MI were in Hardy–Weinberg equilibrium for the subjects with MI and controls (Supplementary Table 3). Multivariable logistic regression analysis with adjustment for age, sex, BMI, smoking status, serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia revealed that the A→G polymorphism of SEMA3F (dominant and additive 1 and 2 models), the C→T polymorphism of CLEC16A (dominant and additive 1 models), the A→G polymorphism of LAMA3 (recessive model), and the C→G polymorphism of PCSK2 (recessive model) were significantly (P < 0.05) associated with MI (Table 3). The G allele of SEMA3F, the T allele of CLEC16A, and the G allele of LAMA3 were protective against MI, whereas the G allele of PCSK2 was a risk factor for this condition. A stepwise forward selection procedure was performed to examine the effects of genotypes for the four polymorphisms associated with MI by the chi-square test as well as of age, sex, BMI, smoking status, the serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholes-
terolemia on MI. Diabetes mellitus, male sex, hypertension, serum concentration of creatinine, hypercholesterolemia, SEMA3F (dominant model), CLEC16A (dominant model), LAMA3 (recessive model), and PCSK2 (recessive model), in descending order of statistical significance, were significant (P < 0.05) and independent determinants of MI (Table 4). Finally, we examined the relations of identified polymorphisms to intermediate phenotypes, including BMI, systolic and diastolic blood pressure, fasting plasma glucose level, blood glycosylated hemoglobin content, and serum concentrations of creatinine, total cholesterol, HDL–cholesterol, and triglycerides (Table 5). The A→G polymorphism of SEMA3F was related to the serum concentration of total cholesterol in a dominant model. The C→T polymorphism of CLEC16A was related to BMI and the blood glycosylated hemoglobin content in a recessive model and to the serum concentration of triglycerides in a dominant model. The A→G polymorphism of LAMA3 was related to BMI in a dominant model and to the prevalence of diabetes mellitus and the serum concentration of total cholesterol in a recessive model. The C→G polymorphism of PCSK2 was related to the fasting plasma glucose level in dominant and recessive models. 4. Discussion We have examined the possible relations of 150 polymorphisms selected from the initial screen by a GWAS of ischemic stroke [10] to the prevalence of MI in 5014 Japanese individuals. Our association study with three steps of analysis (chi-square test, multivariable logistic regression analysis with adjustment for covariates, and stepwise forward selection procedure) revealed that four genetic variants, [the A→G polymorphism of SEMA3F (rs12632110) at 3p21.3, the C→T polymorphism of CLEC16A (rs9925481) at 16p13.13, the A→G polymorphism of LAMA3 (rs12373237) at 18q11.2, and the C→G polymorphism of PCSK2 (rs6080699) at 20p11.2] were significantly associated with MI in Japanese individuals. These polymorphisms differed from two SNPs (rs6007897 and rs4044210) associated with ischemic stroke in our previous GWAS [10]. Sema domain, immunoglobulin domain, short basic domain, secreted, 3F (SEMA3F) plays an important role in endothelial
Table 4 Effects of genotypes and other characteristics on the prevalence of myocardial infarction as determined by a stepwise forward selection procedure (P < 0.05). Characteristics
P
R2
Diabetes mellitus Male sex Hypertension Serum creatinine concentration Hypercholesterolemia SEMA3F (A→G, dominant) CLEC16A (C→T, dominant) LAMA3 (A→G, recessive) PCSK2 (C→G, recessive)
<0.0001 <0.0001 <0.0001 <0.0001 0.0002 0.0012 0.0013 0.0159 0.0260
0.1043 0.0657 0.0402 0.0052 0.0024 0.0018 0.0017 0.0010 0.0008
R2 , contribution rate.
472
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Table 5 Relations of four polymorphisms to intermediate phenotypes. SEMA3F
AA
AG
GG
P (dominant)
P (recessive)
Body mass index (kg/m ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.4 ± 3.2 73.5 ± 38.2 50.7 139 ± 23 77 ± 13 23.4 6.78 ± 3.21 5.94 ± 1.67 24.5 5.08 ± 0.94 1.59 ± 1.06 1.38 ± 0.40
23.5 ± 3.2 76.1 ± 66.9 48.9 139 ± 23 77 ± 13 24.4 6.86 ± 2.99 5.88 ± 1.47 26.5 5.16 ± 0.95 1.63 ± 1.12 1.40 ± 0.41
23.6 ± 3.1 74.2 ± 55.9 46.7 139 ± 23 77 ± 13 23.0 6.87 ± 3.09 5.86 ± 1.56 26.2 5.16 ± 0.99 1.59 ± 1.04 1.40 ± 0.40
0.3215 0.3507 0.1275 0.9493 0.9911 0.7475 0.4877 0.4072 0.2235 0.0341 0.6746 0.1399
0.2849 0.5329 0.0786 0.7161 0.7487 0.3896 0.7669 0.5381 0.8552 0.4662 0.4056 0.3762
CLEC16A
CC
CT
TT
P (dominant)
P (recessive)
Body mass index (kg/m2 ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.5 ± 3.1 75.5 ± 62.3 48.9 139 ± 24 77 ± 14 24.0 6.87 ± 3.14 5.92 ± 1.58 25.7 5.14 ± 0.96 1.63 ± 1.14 1.39 ± 0.40
23.5 ± 3.1 73.6 ± 45.8 48.6 139 ± 23 77 ± 13 22.8 6.77 ± 2.85 5.83 ± 1.45 26.1 5.14 ± 0.95 1.54 ± 0.90 1.41 ± 0.41
24.3 ± 4.0 68.9 ± 19.7 44.4 139 ± 23 79 ± 10 22.2 6.93 ± 2.77 5.39 ± 0.79 32.1 5.19 ± 0.87 1.40 ± 0.62 1.44 ± 0.39
0.5536 0.2567 0.7417 0.8912 0.4988 0.3795 0.4614 0.0969 0.5800 0.9895 0.0089 0.1461
0.0366 0.3446 0.4330 0.9212 0.3111 0.7488 0.8133 0.0242 0.2090 0.6374 0.0871 0.3232
LAMA3
AA
AG
GG
P (dominant)
P (recessive)
Body mass index (kg/m ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
24.4 ± 3.8 71.3 ± 24.2 54.8 138 ± 20 76 ± 11 27.4 7.26 ± 3.77 6.25 ± 1.35 27.4 5.22 ± 0.86 1.75 ± 1.20 1.38 ± 0.39
23.5 ± 3.1 76.0 ± 69.5 50.8 139 ± 23 77 ± 14 26.2 6.95 ± 3.30 5.96 ± 1.61 27.1 5.20 ± 0.97 1.60 ± 1.01 1.40 ± 0.41
23.5 ± 3.1 74.8 ± 55.9 48.1 139 ± 24 78 ± 13 23.0 6.81 ± 2.99 5.86 ± 1.53 25.6 5.12 ± 0.96 1.61 ± 1.10 1.39 ± 0.40
0.0132 0.5896 0.2976 0.7734 0.3859 0.4638 0.2874 0.1664 0.7698 0.4847 0.2691 0.8458
0.3902 0.6715 0.0787 0.7857 0.0693 0.0254 0.1717 0.1110 0.3001 0.0230 0.8543 0.5628
PCSK2
CC
CG
GG
P (dominant)
P (recessive)
Body mass index (kg/m2 ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.7 ± 3.3 72.4 ± 54.2 46.0 140 ± 24 78 ± 12 20.3 6.52 ± 2.78 5.80 ± 1.63 26.4 5.17 ± 0.92 1.64 ± 1.07 1.40 ± 0.37
23.5 ± 3.1 75.3 ± 57.9 47.5 138 ± 22 77 ± 13 22.8 6.72 ± 2.99 5.83 ± 1.43 25.2 5.15 ± 0.95 1.61 ± 1.04 1.40 ± 0.40
23.5 ± 3.2 75.0 ± 59.6 49.9 140 ± 24 77 ± 14 24.5 6.97 ± 3.13 5.94 ± 1.61 26.2 5.13 ± 0.97 1.60 ± 1.12 1.39 ± 0.41
0.1603 0.3553 0.2660 0.6433 0.5695 0.1075 0.0448 0.4070 0.7816 0.5847 0.5003 0.7998
0.8476 0.9015 0.0630 0.2692 0.7349 0.0765 0.0023 0.0535 0.5473 0.5706 0.6144 0.7002
2
2
P values of < 0.05 are shown in bold.
cell functions in vasculature [16]. Reduced expression of SEMA3F inhibited by small interfering RNA affected the intracellular localization and function of gap junction protein, alpha 1, 43 kDa (GJA1) [17]. GJA1 is generally expressed in medial smooth muscle cells of normal vascular wall [18], being increased at early stages of atherosclerosis and reduced in advanced lesion [19]. We have now shown that the A→G polymorphism in intron 18 of SEMA3F (rs12632110) was significantly associated with MI in Japanese individuals, with the G allele protecting against this condition. Although the A→G polymorphism of SEMA3F was related to the serum concentration of total cholesterol, functional relevance of this SNP with the pathogenesis of MI remains elucidated. C-type lectin domain family 16, member A (CLEC16A) has been detected mainly in immune cells including B lymphocytes and nat-
ural killer cells [20]. Two independent GWAS in Northern European [21] and American [22] populations showed that several variants of CLEC16A were significantly associated with type 1 diabetes mellitus. It was hypothesized that the altered protein structure could elicit autoimmune response, resulting in the destruction of the islet cells of the pancreas as seen in type 1 diabetes mellitus. We have now shown that the C→T polymorphism in intron 11 of CLEC16A (rs9925481) was significantly associated with the prevalence of MI in Japanese individuals, with the T allele protecting against this condition. Given that the protective (T) allele of this SNP was related to the reduced blood glycosylated hemoglobin content and serum concentration of triglycerides, the association of the C→T polymorphism of CLEC16A with MI might be attributable, at least in part, to the effect of this SNP on glucose and triglycerides metabolism.
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Laminins are members of glycoprotein family that are the main components of the multifunctional extracellular matrix proteins regulating adhesion, motility, gene expression, and apoptosis [23]. Targeted disruption of LAMA3 in mice caused the formation of lethal epidermal blistering condition similar to human junctional epidermolysis bullosa [23]. It was suggested that LAMA3 has an important role in the regulation of endothelial wound healing [24]. We have now shown that the A→G polymorphism in intron 30 of LAMA3 (rs12373237) was significantly associated with the prevalence of MI in Japanese individuals, with the G allele protecting against this condition. Given that the protective (G) allele of this SNP was related to the decreases in BMI, prevalence of diabetes mellitus, and serum concentration of total cholesterol, the association of the A→G polymorphism of LAMA3 with MI might be attributable, at least in part, to the effect of this SNP on glucose and cholesterol metabolism. Proprotein convertase subtilisin/kexin type 2 (PCSK2) is expressed in the pancreatic islet, pituitary gland, and other neuronal tissues in the brain [25], and has an important role in endocrine function. PCSK2 in the pancreatic islet is involved in processing of proglucagon to glucagon and proinsulin to both insulin and C-peptide [26,27]. The tandem repeat polymorphism in intron 2 of PCSK2 was associated with type 1 diabetes mellitus in Japanese population [28]. The polymorphism of PCSK2 (rs2021785) was also associated with type 2 diabetes mellitus complicated with end stage renal disease in African-American population [29]. We have now shown that the C→G polymorphism in intron 8 of PCSK2 (rs6080699) was significantly associated with the prevalence of MI, with the G allele representing a risk factor for this condition. Given that the risk (G) allele of this SNP was related to the increased fasting plasma glucose level, the relation of the C→G polymorphism of PCSK2 to MI might be attributable, at least in part, to the effect of this SNP on glucose metabolism. Our study has several limitations: (i) It is possible that one or more of the polymorphisms associated with MI in the present study are in linkage disequilibrium with other polymorphisms in the same gene or in other nearby genes that are actually responsible for the development of this condition. (ii) The functional relevance of the identified polymorphisms to gene transcription or to protein structure or function was not determined in the present study. (iii) Given that we adopted the criterion of FDR < 0.05, but not Bonferroni’s correction, for association to compensate for the multiple comparisons of genotypes with MI, it was not possible to exclude completely potential statistical errors such as false positives. (iv) Given that the results of the present study were not replicated, validation of our findings will require their replication with independent subject panels. (v) Given that SNPs at 9p21 were not related to ischemic stroke in the initial screen by a GWAS [10], they were not examined in the present study. In conclusion, our present study suggests the SEMA3F, CLEC16A, LAMA3, and PCSK2 may be susceptibility loci for MI in Japanese individuals. Determination of genotypes for these polymorphisms may prove informative for assessment of the genetic risk for MI in such individuals. Acknowledgments This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (nos. 18209023, 18018021, and 19659149 to Y.Y.). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2009.11.050.
473
References [1] Yusuf S, Hawken S, Ounpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case–control study. Lancet 2004;364:937–52. [2] Kullo IJ, Ding K. Mechanisms of disease: the genetic basis of coronary heart disease. Nat Clin Pract Cardiovasc Med 2007;4:558–69. [3] Arnett DK, Baird AE, Barkley RA, et al. Relevance of genetics and genomics for prevention and treatment of cardiovascular disease: a scientific statement from the American Heart Association Council on Epidemiology and Prevention, the Stroke Council, and the Functional Genomics and Translational Biology Interdisciplinary Working Group. Circulation 2007;115:2878– 901. [4] Erdmann J, Grosshennig A, Braund PS, et al. New susceptibility locus for coronary artery disease on chromosome 3q22.3. Nat Genet 2009;41:280–2. [5] Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 2007;447:661–78. [6] Myocardial Infarction Genetics Consortium. Genome-wide association of earlyonset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 2009;41:334–41. [7] Samani NJ, Erdmann J, Hall AS, et al. Genomewide association analysis of coronary artery disease. N Engl J Med 2007;357:443–53. [8] McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science 2007;316:1488–91. [9] Helgadottir A, Thorleifsson G, Manolescu A, et al. A common variant on chromosome 9p21 affects the risk of myocardial infarction. Science 2007;316:1491– 3. [10] Yamada Y, Fuku N, Tanaka M, et al. Identification of CELSR1 as a susceptibility gene for ischemic stroke in Japanese individuals by a genome-wide association study. Atherosclerosis 2009;207:144–9. [11] Yamada Y, Izawa H, Ichihara S, et al. Prediction of the risk of myocardial infarction from polymorphisms in candidate genes. N Engl J Med 2002;347:1916–23. [12] Yamada Y, Kato K, Oguri M, et al. Genetic risk for myocardial infarction determined by polymorphisms of candidate genes in a Japanese population. J Med Genet 2008;45:216–21. [13] Yamada Y, Matsuo H, Segawa T, et al. Assessment of genetic risk for myocardial infarction. Thromb Haemost 2006;96:220–7. [14] Itoh Y, Mizuki N, Shimada T, et al. High-throughput DNA typing of HLA-A, -B, -C, and -DRB1 loci by a PCR–SSOP–Luminex method in the Japanese population. Immunogenetics 2005;57:717–29. [15] Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Stat Soc Ser B 1995;57: 289–300. [16] Bielenberg DR, Klagsbrun M. Targeting endothelial and tumor cells with semaphorins. Cancer Metastasis Rev 2007;26:421–31. [17] Kawasaki Y, Kubomoto A, Yamasaki H. Control of intracellular localization and function of Cx43 by SEMA3F. J Membr Biol 2007;217:53–61. [18] Severs NJ, Rothery S, Dupont E, et al. Immunocytochemical analysis of connexin expression in the healthy and diseased cardiovascular system. Microsc Res Tech 2001;52:301–22. [19] Blackburn JP, Peters NS, Yeh HI, et al. Upregulation of connexin43 gap junctions during early stages of human coronary atherosclerosis. Arterioscler Thromb Vasc Biol 1995;15:1219–28. [20] Skinningsrud B, Husebye ES, Pearce SH, et al. Polymorphisms in CLEC16A and CIITA at 16p13 are associated with primary adrenal insufficiency. J Clin Endocrinol Metab 2008;93:3310–7. [21] Todd JA, Walker NM, Cooper JD, et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 2007;39:857–64. [22] Hakonarson H, Grant SF, Bradfield JP, et al. A genome-wide association study identifies KIAA0350 as a type 1 diabetes gene. Nature 2007;448:591–4. [23] Ryan MC, Lee K, Miyashita Y, et al. Targeted disruption of the LAMA3 gene in mice reveals abnormalities in survival and late stage differentiation of epithelial cells. J Cell Biol 1999;145:1309–23. [24] Tzu J, Marinkovich MP. Bridging structure with function: structural, regulatory, and developmental role of laminins. Int J Biochem Cell Biol 2008;40: 199–214. [25] Seidah NG, Chretien M, Day R. The family of subtilisin/kexin like pro-protein and pro-hormone convertases: divergent or shared functions. Biochimie 1994;76:197–209. [26] Rouille Y, Westermark G, Martin SK, et al. Proglucagon is processed to glucagon by prohormone convertase PC2 in alpha TC1-6 cells. Proc Natl Acad Sci USA 1994;91:3242–6. [27] Smeekens SP, Montag AG, Thomas G, et al. Proinsulin processing by the subtilisin-related proprotein convertases furin PC2, and PC3. Proc Natl Acad Sci USA 1992;89:8822–6. [28] Yoshida H, Ohagi S, Sanke T, et al. Association of the prohormone convertase 2 gene (PCSK2) on chromosome 20 with NIDDM in Japanese subjects. Diabetes 1995;44:389–93. [29] Leak TS, Keene KL, Langefeld CD, et al. Association of the proprotein convertase subtilisin/kexin-type 2 (PCSK2) gene with type 2 diabetes in an African American population. Mol Genet Metab 2007;92:145–50.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Association of genetic variants in SEMA3F, CLEC16A, LAMA3, and PCSK2 with myocardial infarction in Japanese individuals Tetsuo Fujimaki a , Kimihiko Kato a , Kiyoshi Yokoi a , Mitsutoshi Oguri b , Tetsuro Yoshida c , Sachiro Watanabe d , Norifumi Metoki e , Hidemi Yoshida f , Kei Satoh f , Yukitoshi Aoyagi g , Yoshinori Nozawa h , Genjiro Kimura i , Yoshiji Yamada j,∗ a
Department of Cardiovascular Medicine, Gifu Prefectural Tajimi Hospital, Tajimi, Japan Department of Cardiology, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan c Department of Cardiovascular Medicine, Inabe General Hospital, Inabe, Japan d Department of Cardiology, Gifu Prefectural General Medical Center, Gifu, Japan e Department of Internal Medicine, Hirosaki Stroke Center, Hirosaki, Japan f Department of Vascular Biology, Institute of Brain Science, Hirosaki University Graduate School of Medicine, Hirosaki, Japan g Department of Genomics for Longevity and Health, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan h Gifu International Institute of Biotechnology and Tokai Gakuin University, Kakamigahara, Japan i Department of Cardio-Renal Medicine and Hypertension, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan j Department of Human Functional Genomics, Life Science Research Center, Mie University, 1577 Kurima-machiya, Tsu, Mie 514-8507, Japan b
a r t i c l e
i n f o
Article history: Received 28 October 2009 Received in revised form 28 November 2009 Accepted 30 November 2009 Available online 5 December 2009 Keywords: Myocardial infarction Coronary heart disease Atherosclerosis Genetics Polymorphism
a b s t r a c t Objective: The purpose of the present study was to identify genetic variants that confer susceptibility to myocardial infarction (MI) in Japanese individuals. Methods: The study population comprised 5014 Japanese individuals, including 1444 subjects with MI and 3570 controls. The 150 polymorphisms examined in the present study were selected by a genomewide association study for ischemic stroke with the use of the GeneChip Human Mapping 500K Array Set (Affymetrix), and were determined by a method that combines the polymerase chain reaction and sequence-specific oligonucleotide probes with suspension array technology. Results: An initial screen by the chi-square test revealed that the A→G polymorphism of SEMA3F (rs12632110), the C→T polymorphism of CLEC16A (rs9925481), the A→G polymorphism of LAMA3 (rs12373237), and the C→G polymorphism of PCSK2 (rs6080699) were significantly (false discovery rate for allele frequencies of <0.05) associated with MI. Subsequent multivariable logistic regression analysis with adjustment for covariates and a stepwise forward selection procedure revealed that the A→G polymorphism of SEMA3F (dominant model; P = 0.0014; odds ratio, 0.76), the C→T polymorphism of CLEC16A (dominant model; P = 0.0009; odds ratio, 0.75), the A→G polymorphism of LAMA3 (recessive model; P = 0.0099; odds ratio, 0.80), and the C→G polymorphism of PCSK2 (recessive model; P = 0.0155; odds ratio, 1.19) were significantly (P < 0.05) associated with the prevalence of MI. Conclusion: Determination of these genotypes may prove informative for assessment of the genetic risk for MI in Japanese individuals. © 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Myocardial infarction (MI) is an important clinical problem because of its large contribution to mortality. Given that the disease prevention is an important strategy for reducing the overall burden of MI, identification of markers for MI is essential for both risk prediction and intervention for reducing the chance of future events. Epidemiological studies have underscored the contribution
∗ Corresponding author. Tel.: +81 59 231 5387; fax: +81 59 231 5388. E-mail address: [email protected] (Y. Yamada). 0021-9150/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2009.11.050
of conventional risk factors such as smoking, dyslipidemia, hypertension, diabetes mellitus, and obesity in the development of MI [1]. In addition to these risk factors, recent genetic studies have suggested the importance of genetic factors and of interactions between genetic and environmental factors in predisposition to MI [2,3]. Although recent genome-wide association studies (GWAS) have identified various polymorphisms that confer susceptibility to MI [4–9], genetic variants for predisposition to MI in Japanese individuals have not determined definitively. We have performed an association study of MI in a total of 5014 Japanese individuals with 150 single nucleotide polymorphisms (SNPs) selected from a GWAS for ischemic stroke [10]. The purpose
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
469
Table 1 Baseline characteristics of the study subjects with myocardial infarction (MI) and controls. Characteristic
MI (n = 1444)
Controls (n = 3570)
P value
Age (years) Sex (male/female) Body mass index (kg/m2 ) Current or former smoker (%) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum HDL–cholesterol (mmol/l) Serum triglycerides (mmol/l)
65.8 ± 10.1 1137/307 23.8 ± 3.3 26.9 90.6 ± 94.3 71.2 142 ± 27 75 ± 15 48.0 7.64 ± 3.48 6.59 ± 1.80 28.5 5.15 ± 1.05 1.19 ± 0.35 1.73 ± 1.15
65.8 ± 11.2 1600/1970 23.4 ± 3.3 19.0 68.7 ± 39.3 39.4 137 ± 21 79 ± 12 13.8 6.48 ± 2.77 5.60 ± 1.32 24.8 5.13 ± 0.91 1.48 ± 0.39 1.54 ± 1.04
0.9101 <0.0001 0.0002 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0070 0.5853 <0.0001 <0.0001
Quantitative data are means ± SD. Hypertension: systolic blood pressure of ≥140 mmHg, diastolic blood pressure of ≥90 mmHg, or taking of antihypertensive medication. Diabetes mellitus: fasting plasma glucose level of ≥6.93 mmol/l, blood glycosylated hemoglobin (hemoglobin A1c ) content of ≥6.5%, or taking of antidiabetes medication. Hypercholesterolemia: serum total cholesterol concentration of ≥5.72 mmol/l or taking of lipid-lowering medication. HDL, high-density lipoprotein.
of the present study was to identify genetic variants that confer susceptibility to MI in Japanese individuals, and thereby to contribute to the personalized prevention of this condition. 2. Methods 2.1. Study population The study population comprised 5014 unrelated Japanese individuals, including 1444 subjects with MI and 3570 controls, who either visited outpatient clinics of or were admitted to one of the participating hospitals (Gifu Prefetural General Medical Center and Gifu Prefectural Tajimi Hospital in Gifu Prefecture, Japan; Hirosaki University Hospital, Reimeikyo Rehabilitation Hospital, and Hirosaki Stroke Center in Aomori Prefecture, Japan) between October 2002 and March 2008 with various symptoms or for an annual health check up, or who were recruited to a populationbased prospective cohort study of aging and age-related diseases in Nakanojo, Gunma Prefecture, Japan. The 1444 subjects with MI (1137 men, 307 women) all underwent coronary angiography and left ventriculography. The diagnosis of MI was based on typical electrocardiographic changes and on increases both in the serum activity of creatine kinase (MB isozyme) and in the serum concentration of troponin T. The diagnosis was confirmed by the presence of wall motion abnormality on left ventriculography and by identification of the responsible stenosis in any of the major coronary arteries or in the left main trunk by coronary angiography. The 3570 control subjects (1600 men, 1970 women) had no history of coronary heart disease, aortic aneurysm, or peripheral arterial occlusive disease; of ischemic or hemorrhagic stroke or other cerebral diseases; or of other atherosclerotic, thrombotic, embolic or hemorrhagic disorders. The study protocol complied with the Declaration of Helsinki and was approved by the committees on the Ethics of Human Research of Mie University Graduate School of Medicine, Hirosaki University Graduate School of Medicine, Gifu International Institute of Biotechnology, Tokyo Metropolitan Institute of Gerontology, and participating hospitals. Written informed consent was obtained from each participant. 2.2. Selection of polymorphisms Our aim was to identify genetic variants associated with MI in the Japanese population in a case–control association study. A total of 150 polymorphisms examined in the present study (Supplementary Table 1) were selected from an initial screen by a GWAS of ischemic stroke (atherothrombotic cerebral infarc-
tion) on the basis of P value for allele frequency <1.0 × 10−7 with the use of the GeneChip Human Mapping 500K Array Set (Affymetrix, Santa Clara, CA) [10]. Given that the main cause of MI and atherothrombotic cerebral infarction is atherothrombosis, genetic variants associated with MI might be included in those related to atherothrombotic cerebral infarction. We have not examined the relation of these polymorphisms to MI in our previous studies [11–13]. 2.3. Genotyping polymorphisms Venous blood (7 ml) was collected into tubes containing ethylenediaminetetraacetic acid at a final concentration of 50 mM, and genomic DNA was isolated with a kit (Genomix; Talent, Trieste, Italy). Genotypes of 150 polymorphisms were determined at G&G Science (Fukushima, Japan) by a method that combines the polymerase chain reaction and sequence-specific oligonucleotide probes with suspension array technology (Luminex, Austin, TX, USA). Primers, probes, and other conditions for genotyping of polymorphisms related [false discovery rate (FDR) < 0.05] to MI as determined by the initial chi-square test are shown in Supplementary Table 2. Detailed genotyping methodology was described previously [14]. 2.4. Statistical analysis Quantitative data were compared between subjects with MI and controls by the unpaired Student’s t-test. Categorical data were compared by the chi-square test. Allele frequencies were estimated by the genome counting method, and the chi-square test was used to identify departures from Hardy–Weinberg equilibrium. In the initial screen, the genotype distribution (3 × 2) and allele frequencies (2 × 2) of each polymorphism were compared between subjects with MI and controls by the chi-square test. Given the multiple comparisons of genotypes with MI, the FDR was calculated from the distributions of P values for the allele frequencies of the 150 polymorphisms [15]. Polymorphisms with a FDR for allele frequency of <0.05 were further examined by multivariable logistic regression analysis with adjustment for covariates. Such analysis was thus performed with MI as a dependent variable and independent variables including age, sex (0, woman; 1, man), body mass index (BMI), smoking status (0, nonsmoker; 1, smoker), the serum concentration of creatinine, the history of hypertension, diabetes mellitus, and hypercholesterolemia (0, no history; 1, positive history), and genotype of each polymorphism; and the P value, odds ratio, and 95% confidence interval were calculated. Each geno-
470
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Table 2 Polymorphisms related (P value for allele frequency <0.05) to myocardial infarction (MI) as revealed by the chi-square test. Gene
Polymorphism
dbSNP
SEMA3F AA AG GG
A→G
rs12632110
CLEC16A CC CT TT
C→T
LAMA3 AA AG GG
A→G
PCSK2 CC CG GG
C→G
SIX5 GG GT TT
G→T
COL13A1 AA AC CC USP37 AA AG GG
A→C
COLEC11 CC CT TT
C→T
ITPK1 CC CT TT
C→T
LRP5 CC CT TT
C→T
COL6A3 CC CT TT
C→T
SS18L1 CC CT TT
C→T
TSPAN9 AA AG GG
A→G
a
A→G
MIa
Controlsa
356 (24.8) 706 (49.1) 375 (26.1)
720 (20.4) 1759 (50.0) 1040 (29.6)
1142 (79.4) 277 (19.3) 18 (1.3)
2636 (75.0) 815 (23.2) 63 (1.8)
24 (1.7) 335 (23.3) 1078 (75.0)
49 (1.4) 676 (19.2) 2788 (79.4)
96 (6.7) 553 (38.8) 777 (54.5)
317 (8.9) 1448 (40.8) 1782 (50.3)
1003 (69.8) 393 (27.4) 41 (2.8)
2585 (73.5) 871 (24.7) 62 (1.8)
37 (2.6) 374 (26.0) 1026 (71.4)
70 (2.0) 795 (22.6) 2656 (75.4)
1194 (83.7) 222 (15.6) 10 (0.7)
2846 (80.2) 663 (18.7) 38 (1.1)
52 (3.6) 503 (35.3) 871 (61.1)
199 (5.6) 1269 (35.8) 2079 (58.6)
842 (58.6) 505 (35.1) 90 (6.3)
1935 (55.0) 1343 (38.1) 243 (6.9)
724 (50.8) 571 (40.0) 131 (9.2)
1904 (53.7) 1370 (38.6) 274 (7.7)
0 (0.0) 42 (2.9) 1384 (97.1)
0 (0.0) 69 (1.9) 3478 (98.1)
1290 (90.4) 135 (9.5) 1 (0.1)
3138 (88.5) 402 (11.3) 7 (0.2)
231 (16.1) 689 (47.9) 517 (36.0)
640 (18.2) 1694 (48.1) 1185 (33.7)
rs9925481
rs12373237
rs6080699
rs16980013
rs942576
rs526897
rs6739899
rs2295394
rs3736228
rs11690358
rs2427254
rs2011973
P value (genotype)
P value (allele)
FDR
0.0013
0.0004
0.0469
0.0030
0.0006
0.0469
0.0035
0.0011
0.0469
0.0050
0.0012
0.0469
0.0054
0.0024
0.0646
0.0110
0.0025
0.0646
0.0139
0.0034
0.0744
0.0117
0.0173
0.3252
0.0640
0.0263
0.4185
0.0898
0.0280
0.4185
0.0308
0.0318
0.4185
0.0926
0.0368
0.4185
0.1240
0.0430
0.4185
Values in parentheses are percentages.
type was assessed according to dominant, recessive, and additive genetic models. Additive models included the additive 1 (heterozygotes versus wild-type homozygotes) and the additive 2 (variant homozygotes versus wild-type homozygote) models, which were analyzed simultaneously with a single statistical model. We also performed a stepwise forward selection procedure to examine the effects of genotypes as well as of other covariates on MI. In this analysis, each genotype was examined according to a dominant or
recessive model on the basis of statistical significance in the multivariable logistic regression analysis. The P levels for inclusion in and exclusion from the model were 0.25 and 0.1, respectively. With the exception of the initial screen by the chi-square test (FDR < 0.05), a P value of <0.05 was considered statistically significant. Statistical significance was examined by two-sided tests performed with JMP version 6.0 and JMP Genomics version 3.2 software (SAS Institute, Cary, NC).
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
471
Table 3 Multivariable logistic regression analysis of polymorphisms related (FDR for allele frequency < 0.05) to myocardial infarction by the chi-square test with adjustment for age, sex, body mass index, smoking status, serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia. Gene
SEMA3F CLEC16A LAMA3 PCSK2
Polymorphism
A→G C→T A→G C→G
Dominant
Recessive
P
OR (95%CI)
P
0.0014 0.0009 0.9575 0.1342
0.76 (0.64–0.90) 0.75 (0.63–0.89)
0.0566 0.3845 0.0099 0.0155
Additive 1
Additive 2
OR (95%CI)
P
OR (95%CI)
P
OR (95%CI)
0.78 (0.65–0.93) 0.75 (0.63–0.89)
0.0016 0.2840 0.9115 0.0536
0.72 (0.59–0.88)
0.80 (0.68–0.95) 1.19 (1.03–1.38)
0.0068 0.0014 0.5308 0.4156
OR, odds ratio; CI, confidence interval.
3. Results The characteristics of the study subjects are shown in Table 1. The frequency of male subjects, BMI, the percentage of smokers, systolic blood pressure, serum concentrations of creatinine and triglycerides, fasting plasma glucose level, blood glycosylated hemoglobin content, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia were greater, whereas diastolic blood pressure and serum concentration of high-density lipoprotein (HDL)–cholesterol were lower in subjects with MI than in controls. Comparisons of allele frequencies of polymorphisms between subjects with MI and controls by the chi-square test revealed that 13 polymorphisms were related (P < 0.05) to the prevalence of MI (Table 2). Among these polymorphisms, the A→G polymorphism of SEMA3F (rs12632110), the C→T polymorphism of CLEC16A (rs9925481), the A→G polymorphism of LAMA3 (rs12373237), and the C→G polymorphism of PCSK2 (rs6080699) were significantly (FDR for allele frequency of <0.05) associated with MI. The genotype distributions for the 13 polymorphisms related to MI are also shown in Table 2. Genotype distributions of four polymorphisms significantly associated with MI were in Hardy–Weinberg equilibrium for the subjects with MI and controls (Supplementary Table 3). Multivariable logistic regression analysis with adjustment for age, sex, BMI, smoking status, serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholesterolemia revealed that the A→G polymorphism of SEMA3F (dominant and additive 1 and 2 models), the C→T polymorphism of CLEC16A (dominant and additive 1 models), the A→G polymorphism of LAMA3 (recessive model), and the C→G polymorphism of PCSK2 (recessive model) were significantly (P < 0.05) associated with MI (Table 3). The G allele of SEMA3F, the T allele of CLEC16A, and the G allele of LAMA3 were protective against MI, whereas the G allele of PCSK2 was a risk factor for this condition. A stepwise forward selection procedure was performed to examine the effects of genotypes for the four polymorphisms associated with MI by the chi-square test as well as of age, sex, BMI, smoking status, the serum concentration of creatinine, and the prevalence of hypertension, diabetes mellitus, and hypercholes-
terolemia on MI. Diabetes mellitus, male sex, hypertension, serum concentration of creatinine, hypercholesterolemia, SEMA3F (dominant model), CLEC16A (dominant model), LAMA3 (recessive model), and PCSK2 (recessive model), in descending order of statistical significance, were significant (P < 0.05) and independent determinants of MI (Table 4). Finally, we examined the relations of identified polymorphisms to intermediate phenotypes, including BMI, systolic and diastolic blood pressure, fasting plasma glucose level, blood glycosylated hemoglobin content, and serum concentrations of creatinine, total cholesterol, HDL–cholesterol, and triglycerides (Table 5). The A→G polymorphism of SEMA3F was related to the serum concentration of total cholesterol in a dominant model. The C→T polymorphism of CLEC16A was related to BMI and the blood glycosylated hemoglobin content in a recessive model and to the serum concentration of triglycerides in a dominant model. The A→G polymorphism of LAMA3 was related to BMI in a dominant model and to the prevalence of diabetes mellitus and the serum concentration of total cholesterol in a recessive model. The C→G polymorphism of PCSK2 was related to the fasting plasma glucose level in dominant and recessive models. 4. Discussion We have examined the possible relations of 150 polymorphisms selected from the initial screen by a GWAS of ischemic stroke [10] to the prevalence of MI in 5014 Japanese individuals. Our association study with three steps of analysis (chi-square test, multivariable logistic regression analysis with adjustment for covariates, and stepwise forward selection procedure) revealed that four genetic variants, [the A→G polymorphism of SEMA3F (rs12632110) at 3p21.3, the C→T polymorphism of CLEC16A (rs9925481) at 16p13.13, the A→G polymorphism of LAMA3 (rs12373237) at 18q11.2, and the C→G polymorphism of PCSK2 (rs6080699) at 20p11.2] were significantly associated with MI in Japanese individuals. These polymorphisms differed from two SNPs (rs6007897 and rs4044210) associated with ischemic stroke in our previous GWAS [10]. Sema domain, immunoglobulin domain, short basic domain, secreted, 3F (SEMA3F) plays an important role in endothelial
Table 4 Effects of genotypes and other characteristics on the prevalence of myocardial infarction as determined by a stepwise forward selection procedure (P < 0.05). Characteristics
P
R2
Diabetes mellitus Male sex Hypertension Serum creatinine concentration Hypercholesterolemia SEMA3F (A→G, dominant) CLEC16A (C→T, dominant) LAMA3 (A→G, recessive) PCSK2 (C→G, recessive)
<0.0001 <0.0001 <0.0001 <0.0001 0.0002 0.0012 0.0013 0.0159 0.0260
0.1043 0.0657 0.0402 0.0052 0.0024 0.0018 0.0017 0.0010 0.0008
R2 , contribution rate.
472
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Table 5 Relations of four polymorphisms to intermediate phenotypes. SEMA3F
AA
AG
GG
P (dominant)
P (recessive)
Body mass index (kg/m ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.4 ± 3.2 73.5 ± 38.2 50.7 139 ± 23 77 ± 13 23.4 6.78 ± 3.21 5.94 ± 1.67 24.5 5.08 ± 0.94 1.59 ± 1.06 1.38 ± 0.40
23.5 ± 3.2 76.1 ± 66.9 48.9 139 ± 23 77 ± 13 24.4 6.86 ± 2.99 5.88 ± 1.47 26.5 5.16 ± 0.95 1.63 ± 1.12 1.40 ± 0.41
23.6 ± 3.1 74.2 ± 55.9 46.7 139 ± 23 77 ± 13 23.0 6.87 ± 3.09 5.86 ± 1.56 26.2 5.16 ± 0.99 1.59 ± 1.04 1.40 ± 0.40
0.3215 0.3507 0.1275 0.9493 0.9911 0.7475 0.4877 0.4072 0.2235 0.0341 0.6746 0.1399
0.2849 0.5329 0.0786 0.7161 0.7487 0.3896 0.7669 0.5381 0.8552 0.4662 0.4056 0.3762
CLEC16A
CC
CT
TT
P (dominant)
P (recessive)
Body mass index (kg/m2 ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.5 ± 3.1 75.5 ± 62.3 48.9 139 ± 24 77 ± 14 24.0 6.87 ± 3.14 5.92 ± 1.58 25.7 5.14 ± 0.96 1.63 ± 1.14 1.39 ± 0.40
23.5 ± 3.1 73.6 ± 45.8 48.6 139 ± 23 77 ± 13 22.8 6.77 ± 2.85 5.83 ± 1.45 26.1 5.14 ± 0.95 1.54 ± 0.90 1.41 ± 0.41
24.3 ± 4.0 68.9 ± 19.7 44.4 139 ± 23 79 ± 10 22.2 6.93 ± 2.77 5.39 ± 0.79 32.1 5.19 ± 0.87 1.40 ± 0.62 1.44 ± 0.39
0.5536 0.2567 0.7417 0.8912 0.4988 0.3795 0.4614 0.0969 0.5800 0.9895 0.0089 0.1461
0.0366 0.3446 0.4330 0.9212 0.3111 0.7488 0.8133 0.0242 0.2090 0.6374 0.0871 0.3232
LAMA3
AA
AG
GG
P (dominant)
P (recessive)
Body mass index (kg/m ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
24.4 ± 3.8 71.3 ± 24.2 54.8 138 ± 20 76 ± 11 27.4 7.26 ± 3.77 6.25 ± 1.35 27.4 5.22 ± 0.86 1.75 ± 1.20 1.38 ± 0.39
23.5 ± 3.1 76.0 ± 69.5 50.8 139 ± 23 77 ± 14 26.2 6.95 ± 3.30 5.96 ± 1.61 27.1 5.20 ± 0.97 1.60 ± 1.01 1.40 ± 0.41
23.5 ± 3.1 74.8 ± 55.9 48.1 139 ± 24 78 ± 13 23.0 6.81 ± 2.99 5.86 ± 1.53 25.6 5.12 ± 0.96 1.61 ± 1.10 1.39 ± 0.40
0.0132 0.5896 0.2976 0.7734 0.3859 0.4638 0.2874 0.1664 0.7698 0.4847 0.2691 0.8458
0.3902 0.6715 0.0787 0.7857 0.0693 0.0254 0.1717 0.1110 0.3001 0.0230 0.8543 0.5628
PCSK2
CC
CG
GG
P (dominant)
P (recessive)
Body mass index (kg/m2 ) Serum creatinine (mol/l) Hypertension (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes mellitus (%) Fasting plasma glucose (mmol/l) Blood glycosylated hemoglobin (%) Hypercholesterolemia (%) Serum total cholesterol (mmol/l) Serum triglycerides (mmol/l) Serum HDL–cholesterol (mmol/l)
23.7 ± 3.3 72.4 ± 54.2 46.0 140 ± 24 78 ± 12 20.3 6.52 ± 2.78 5.80 ± 1.63 26.4 5.17 ± 0.92 1.64 ± 1.07 1.40 ± 0.37
23.5 ± 3.1 75.3 ± 57.9 47.5 138 ± 22 77 ± 13 22.8 6.72 ± 2.99 5.83 ± 1.43 25.2 5.15 ± 0.95 1.61 ± 1.04 1.40 ± 0.40
23.5 ± 3.2 75.0 ± 59.6 49.9 140 ± 24 77 ± 14 24.5 6.97 ± 3.13 5.94 ± 1.61 26.2 5.13 ± 0.97 1.60 ± 1.12 1.39 ± 0.41
0.1603 0.3553 0.2660 0.6433 0.5695 0.1075 0.0448 0.4070 0.7816 0.5847 0.5003 0.7998
0.8476 0.9015 0.0630 0.2692 0.7349 0.0765 0.0023 0.0535 0.5473 0.5706 0.6144 0.7002
2
2
P values of < 0.05 are shown in bold.
cell functions in vasculature [16]. Reduced expression of SEMA3F inhibited by small interfering RNA affected the intracellular localization and function of gap junction protein, alpha 1, 43 kDa (GJA1) [17]. GJA1 is generally expressed in medial smooth muscle cells of normal vascular wall [18], being increased at early stages of atherosclerosis and reduced in advanced lesion [19]. We have now shown that the A→G polymorphism in intron 18 of SEMA3F (rs12632110) was significantly associated with MI in Japanese individuals, with the G allele protecting against this condition. Although the A→G polymorphism of SEMA3F was related to the serum concentration of total cholesterol, functional relevance of this SNP with the pathogenesis of MI remains elucidated. C-type lectin domain family 16, member A (CLEC16A) has been detected mainly in immune cells including B lymphocytes and nat-
ural killer cells [20]. Two independent GWAS in Northern European [21] and American [22] populations showed that several variants of CLEC16A were significantly associated with type 1 diabetes mellitus. It was hypothesized that the altered protein structure could elicit autoimmune response, resulting in the destruction of the islet cells of the pancreas as seen in type 1 diabetes mellitus. We have now shown that the C→T polymorphism in intron 11 of CLEC16A (rs9925481) was significantly associated with the prevalence of MI in Japanese individuals, with the T allele protecting against this condition. Given that the protective (T) allele of this SNP was related to the reduced blood glycosylated hemoglobin content and serum concentration of triglycerides, the association of the C→T polymorphism of CLEC16A with MI might be attributable, at least in part, to the effect of this SNP on glucose and triglycerides metabolism.
T. Fujimaki et al. / Atherosclerosis 210 (2010) 468–473
Laminins are members of glycoprotein family that are the main components of the multifunctional extracellular matrix proteins regulating adhesion, motility, gene expression, and apoptosis [23]. Targeted disruption of LAMA3 in mice caused the formation of lethal epidermal blistering condition similar to human junctional epidermolysis bullosa [23]. It was suggested that LAMA3 has an important role in the regulation of endothelial wound healing [24]. We have now shown that the A→G polymorphism in intron 30 of LAMA3 (rs12373237) was significantly associated with the prevalence of MI in Japanese individuals, with the G allele protecting against this condition. Given that the protective (G) allele of this SNP was related to the decreases in BMI, prevalence of diabetes mellitus, and serum concentration of total cholesterol, the association of the A→G polymorphism of LAMA3 with MI might be attributable, at least in part, to the effect of this SNP on glucose and cholesterol metabolism. Proprotein convertase subtilisin/kexin type 2 (PCSK2) is expressed in the pancreatic islet, pituitary gland, and other neuronal tissues in the brain [25], and has an important role in endocrine function. PCSK2 in the pancreatic islet is involved in processing of proglucagon to glucagon and proinsulin to both insulin and C-peptide [26,27]. The tandem repeat polymorphism in intron 2 of PCSK2 was associated with type 1 diabetes mellitus in Japanese population [28]. The polymorphism of PCSK2 (rs2021785) was also associated with type 2 diabetes mellitus complicated with end stage renal disease in African-American population [29]. We have now shown that the C→G polymorphism in intron 8 of PCSK2 (rs6080699) was significantly associated with the prevalence of MI, with the G allele representing a risk factor for this condition. Given that the risk (G) allele of this SNP was related to the increased fasting plasma glucose level, the relation of the C→G polymorphism of PCSK2 to MI might be attributable, at least in part, to the effect of this SNP on glucose metabolism. Our study has several limitations: (i) It is possible that one or more of the polymorphisms associated with MI in the present study are in linkage disequilibrium with other polymorphisms in the same gene or in other nearby genes that are actually responsible for the development of this condition. (ii) The functional relevance of the identified polymorphisms to gene transcription or to protein structure or function was not determined in the present study. (iii) Given that we adopted the criterion of FDR < 0.05, but not Bonferroni’s correction, for association to compensate for the multiple comparisons of genotypes with MI, it was not possible to exclude completely potential statistical errors such as false positives. (iv) Given that the results of the present study were not replicated, validation of our findings will require their replication with independent subject panels. (v) Given that SNPs at 9p21 were not related to ischemic stroke in the initial screen by a GWAS [10], they were not examined in the present study. In conclusion, our present study suggests the SEMA3F, CLEC16A, LAMA3, and PCSK2 may be susceptibility loci for MI in Japanese individuals. Determination of genotypes for these polymorphisms may prove informative for assessment of the genetic risk for MI in such individuals. Acknowledgments This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (nos. 18209023, 18018021, and 19659149 to Y.Y.). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2009.11.050.
473
References [1] Yusuf S, Hawken S, Ounpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case–control study. Lancet 2004;364:937–52. [2] Kullo IJ, Ding K. Mechanisms of disease: the genetic basis of coronary heart disease. Nat Clin Pract Cardiovasc Med 2007;4:558–69. [3] Arnett DK, Baird AE, Barkley RA, et al. Relevance of genetics and genomics for prevention and treatment of cardiovascular disease: a scientific statement from the American Heart Association Council on Epidemiology and Prevention, the Stroke Council, and the Functional Genomics and Translational Biology Interdisciplinary Working Group. Circulation 2007;115:2878– 901. [4] Erdmann J, Grosshennig A, Braund PS, et al. New susceptibility locus for coronary artery disease on chromosome 3q22.3. Nat Genet 2009;41:280–2. [5] Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 2007;447:661–78. [6] Myocardial Infarction Genetics Consortium. Genome-wide association of earlyonset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 2009;41:334–41. [7] Samani NJ, Erdmann J, Hall AS, et al. Genomewide association analysis of coronary artery disease. N Engl J Med 2007;357:443–53. [8] McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science 2007;316:1488–91. [9] Helgadottir A, Thorleifsson G, Manolescu A, et al. A common variant on chromosome 9p21 affects the risk of myocardial infarction. Science 2007;316:1491– 3. [10] Yamada Y, Fuku N, Tanaka M, et al. Identification of CELSR1 as a susceptibility gene for ischemic stroke in Japanese individuals by a genome-wide association study. Atherosclerosis 2009;207:144–9. [11] Yamada Y, Izawa H, Ichihara S, et al. Prediction of the risk of myocardial infarction from polymorphisms in candidate genes. N Engl J Med 2002;347:1916–23. [12] Yamada Y, Kato K, Oguri M, et al. Genetic risk for myocardial infarction determined by polymorphisms of candidate genes in a Japanese population. J Med Genet 2008;45:216–21. [13] Yamada Y, Matsuo H, Segawa T, et al. Assessment of genetic risk for myocardial infarction. Thromb Haemost 2006;96:220–7. [14] Itoh Y, Mizuki N, Shimada T, et al. High-throughput DNA typing of HLA-A, -B, -C, and -DRB1 loci by a PCR–SSOP–Luminex method in the Japanese population. Immunogenetics 2005;57:717–29. [15] Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Stat Soc Ser B 1995;57: 289–300. [16] Bielenberg DR, Klagsbrun M. Targeting endothelial and tumor cells with semaphorins. Cancer Metastasis Rev 2007;26:421–31. [17] Kawasaki Y, Kubomoto A, Yamasaki H. Control of intracellular localization and function of Cx43 by SEMA3F. J Membr Biol 2007;217:53–61. [18] Severs NJ, Rothery S, Dupont E, et al. Immunocytochemical analysis of connexin expression in the healthy and diseased cardiovascular system. Microsc Res Tech 2001;52:301–22. [19] Blackburn JP, Peters NS, Yeh HI, et al. Upregulation of connexin43 gap junctions during early stages of human coronary atherosclerosis. Arterioscler Thromb Vasc Biol 1995;15:1219–28. [20] Skinningsrud B, Husebye ES, Pearce SH, et al. Polymorphisms in CLEC16A and CIITA at 16p13 are associated with primary adrenal insufficiency. J Clin Endocrinol Metab 2008;93:3310–7. [21] Todd JA, Walker NM, Cooper JD, et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 2007;39:857–64. [22] Hakonarson H, Grant SF, Bradfield JP, et al. A genome-wide association study identifies KIAA0350 as a type 1 diabetes gene. Nature 2007;448:591–4. [23] Ryan MC, Lee K, Miyashita Y, et al. Targeted disruption of the LAMA3 gene in mice reveals abnormalities in survival and late stage differentiation of epithelial cells. J Cell Biol 1999;145:1309–23. [24] Tzu J, Marinkovich MP. Bridging structure with function: structural, regulatory, and developmental role of laminins. Int J Biochem Cell Biol 2008;40: 199–214. [25] Seidah NG, Chretien M, Day R. The family of subtilisin/kexin like pro-protein and pro-hormone convertases: divergent or shared functions. Biochimie 1994;76:197–209. [26] Rouille Y, Westermark G, Martin SK, et al. Proglucagon is processed to glucagon by prohormone convertase PC2 in alpha TC1-6 cells. Proc Natl Acad Sci USA 1994;91:3242–6. [27] Smeekens SP, Montag AG, Thomas G, et al. Proinsulin processing by the subtilisin-related proprotein convertases furin PC2, and PC3. Proc Natl Acad Sci USA 1992;89:8822–6. [28] Yoshida H, Ohagi S, Sanke T, et al. Association of the prohormone convertase 2 gene (PCSK2) on chromosome 20 with NIDDM in Japanese subjects. Diabetes 1995;44:389–93. [29] Leak TS, Keene KL, Langefeld CD, et al. Association of the proprotein convertase subtilisin/kexin-type 2 (PCSK2) gene with type 2 diabetes in an African American population. Mol Genet Metab 2007;92:145–50.