- Email: [email protected]
An intron polymorphism in the CXCL16 gene is associated with increased risk of coronary artery disease in Chinese Han population: A large angiography-based study
Atherosclerosis 210 (2010) 160–165
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
An intron polymorphism in the CXCL16 gene is associated with increased risk of coronary artery disease in Chinese Han population: A large angiography-based study Mingfang Huang a,b , Yaling Han b,∗ , Xiaolin Zhang b , Fang Pei b , Jie Deng b , Jian Kang b , Chenghui Yan b a b
Department of Cardiology, Xijing Hospotal, Fourth Military Medical University, Xi’an, China Department of Cardiology, Shenyang Northern Hospital, 83 Wenhua Road, Shenyang 110016, China
a r t i c l e
i n f o
Article history: Received 16 June 2009 Received in revised form 3 November 2009 Accepted 3 November 2009 Available online 10 November 2009 Keywords: CXCL16 Chemokine Coronary angiography Coronary artery disease Genetics Polymorphism
a b s t r a c t Objective: Experimental and clinical observations suggest that CXCL16, a recently discovered transmembrane chemokine combining functions of a chemokine and a scavenger receptor, could be an important player in atherosclerosis, but the relationship of its common variants with coronary artery disease (CAD) has not been extensively studied. Methods: We designed an angiography-based case-controlled study consisting of 1176 CAD patients and 850 control subjects to investigate the association between five common single nucleotide polymorphisms (SNPs) of CXCL16 gene and CAD risk in Chinese Han population. The plasma concentration of CXCL16 was measured by enzyme-linked immunosorbent assay. Results: No significant differences were observed for the distributions of rs2250333, rs2304973, rs2277680, and rs1051009 between CAD patients and control subjects. However, both the genotype and allele frequencies of rs3744700 showed significant differences between cases and controls (P = 0.001 and P < 0.001, respectively). The GG homozygotes had significantly higher CAD risk compared with the T allele carriers (GT + TT) (OR, 1.77; 95% CI, 1.28–2.43; adjusted P < 0.001) in a logistic regression model after adjustment for the conventional risk factors for CAD. The GG genotypes also had increased plasma CXCL16 levels compared with T allele carriers in both cases and control subjects. Conclusions: SNP rs3744700 of CXCL16 gene is independently associated with the development of CAD in Chinese Han population, and GG homozygote which is associated with increased expression of CXCL16 may have a promoting effect on CAD. © 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Coronary artery disease (CAD) and its main complication, myocardial infarction (MI), are leading causes of death and disability worldwide. The pathogenesis of coronary atherosclerosis involves inherited, behavioural and environmental factors. Inflammatory processes are thought to promote atherogenesis from early lesion formation to unstable plaque rupture [1]. Infiltrates of activated macrophages and T-cells are prominent in human and experimental atherosclerotic lesions. Recruitment of these cells to lesions is coordinated by the production of chemokines, small proteins which interact with specific chemoattractant receptors belonging to the superfamily of G protein-coupled receptors [2]. Once present in the subendothelial space, macrophages engulf oxidatively modified low-density lipoproteins (ox-LDL) through their scavenger receptors and develop into cholesterol-filled foam
∗ Corresponding author. Tel.: +86 24 23922184; fax: +86 24 23911006. E-mail address: [email protected] (Y. Han). 0021-9150/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2009.11.004
cells, a hallmark of atherosclerotic lesion formation. Therefore, scavenger receptors also play a crucial role in the pathogenesis of atherosclerosis [3]. CXCL16 (CXC ligand 16)/SR-PSOX (scavenger receptor that binds phosphatidylserine and oxidized lipoprotein), a recently discovered transmembrane chemokine [4], is a member of the CXC chemokine subfamily. Unlike other members of this subgroup, it is structurally similar to CX3CL1, containing four distinct domains; that is, a chemokine domain tethered to the cell surface via a mucin-like stack, which in turn is attached to transmembrane and cytoplasmic domains. The chemokine domain acts not only as an attractant for cells expressing the CXCR6 receptor [4], but also as a scavenger receptor facilitating uptake of ox-LDL, phosphatidylserine, bacteria and apoptotic cells [5,6]. The extracellular part of CXCL16 undergoes cleavage by the metalloproteinase ADAM 10 creating a soluble chemokine (sol-CXCL16) [7,8], which activates CXCR6-expressing T-cells. Functioning as a chemokine and a scavenger receptor, CXCL16/SR-PSOX may be an interesting player in the formation of atherosclerotic lesions. CXCL16 is expressed on macrophages,
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
dendritic cells, and aortic smooth muscle cells. Enhanced CXCL16 expression has been detected in atherosclerotic lesions from humans and murine models of atherosclerosis (apoE and LDL receptor knockouts) [9–12]. Furthermore, several pro-atherogenic inflammatory stimuli increase CXCL16 expression [4,8,11,13,14], enhancing uptake of ox-LDL and facilitating foam cell formation, as well as promoting recruitment of T-cell to the atherosclerotic plaque [10]. In apoE−/− mice, genetic deletion of CXCR6 attenuated atherosclerosis, accompanied by a decreased percentage of CXCR6+ T-cells within the aortas, indicating a pro-atherogenic role of CXCL16/CXCR6 interaction [15]. However, a recent rodent study reported that CXCL16-deficient LDL receptor −/− mice were associated with accelerated atherosclerosis [12]. Human epidemiological studies evaluating CXCL16 as a potential biomarker of CAD have produced controversial results. While Sheikine et al. [16] found decreased CXCL16 levels in both stable and unstable angina, Lehrke et al. [17] reported increased CXCL16 levels in a large population of CAD patients, particularly in those with unstable disease. And a recent study further support the concept that CAD patients are characterized by increased plasma levels of CXCL16 [14]. The role of admission CXCL16 levels in patients with acute MI, however, has been less extensively studied. Genetic variation of CXCL16 could modulate CXCL16 production and function. A preliminary study by Lundberg et al. found that a single nucleotide polymorphism (SNP) in the fourth exon of the CXCL16 gene was significantly associated with the severity of coronary artery stenosis in CAD patients [18]. However, the association study is limited to a SNP, which fails to include all the potential risk-conferring variations in the CXCL16 gene and their potential effects on CAD risk. To address these issues, we first conducted an analysis of HapMap-tagged SNPs in the CXCL16 gene, including 181Ala >Val (rs2277680), and its relation to the risk of CAD in a Chinese Han population. Next, we investigated the association of CXCL16 genetic variation with its plasma levels in a subgroup of patients.
2. Methods 2.1. Study populations A total of 1176 patients with CAD and 850 control individuals, aged between 29 and 75 years, were enrolled in the study from June 2006 to April 2009. Patients and controls were unrelated individuals of Chinese Han from the northeast region of China. All the subjects had undergone coronary angiography for the evaluation of suspected or established CAD at Northern Hospital, Shenyang, China. The study complied with the declaration of Helsinki. Approval was obtained from the local Ethics Committees and all patients gave their informed consent. Inclusion criteria for CAD were ≥50% narrowing of the lumina of at least 1 of the major coronary arteries by coronary angiography. Additionally, angiographic severity of disease was defined as 0-, 1-, 2- or 3-vessel disease based on the number of luminal narrowing ≥50% in the three major coronary arteries [19]. The diagnosis of MI was based on typical electrocardiographic changes as well as on increases in the serum activities of enzymes such as creatinine kinase, aspartate aminotransferase, and lactate dehydrogenase and in the serum concentration of troponin T. The diagnosis was confirmed by the presence of a wall-motion abnormality on left ventriculography as well as by identification of the responsible stenosis in any of the major coronary arteries or in the left main trunk by coronary angiography. The control subjects were selected from the subjects admitted to the hospital for the evaluation of chest pain, whose major coronary artery had no more than 20% stenosis, and did not have any vascular disease. Control subjects
161
were recruited during the same period as the patients. A complete clinical history was obtained from all subjects, and the vascular risk factors are summarized in Table 1. We did not include subjects with a history of hematologic, neoplastic, renal, liver, or thyroid disease. Furthermore, patients who had inflammatory, infectious or autoimmune diseases, type-1 diabetes mellitus or familial hyperlipidemia were also excluded. To further investigate the impact of gene polymorphisms on pathological phenotypes, the levels of plasma CXCL16 were determined in a subset population selected from our original cohort between March 2008 and April 2009. They consisted of 160 subjects with first episode of acute MI (AMI) and 160 subjects with completely normal coronary arteries (CAD-free). The subset cases and controls were individually matched in terms of age (≤5 years), gender, smoking status, presence of hypertension, and diabetes mellitus. Because statin therapy can significantly reduces plasma levels of CXCL16 [14], all subset participants were not on lipidlowering medication. 2.2. Characterizing linkage disequilibrium and tagSNPs selection The CXCL16 gene contains five exons and spans 6283 bp on chromosome 17 (NCBI accession NM 022059). SNPs were retrieved from Hapmap database for Han population sample (release no. 21a/phase II, Jan 07, population: CHB; http://www.hapmap.org). Pair-wise LD between SNPs was assessed using Lewontin’s D statistic and the squared correlation statistic r2 . The Haploview program was used to calculate the LD coefficients and define haplotype blocks [20]. TagSNPs were selected using the r2 -based Tagger program [21] with pair-wise r2 ≥ 0.80 and minor allele frequency (MAF) ≥5%. Nine SNPs covering 10.28 kb including CXCL16 (spanning from 3 kb upstream to 1 kb downstream of the gene) were identified. SNP rs8071612 is located in intron 3, and rs1876444, rs2277680 and rs1050998 in exon 4. Rs2277680 captures the other three SNPs, and these four SNPs are in complete association with one another (D = 1, r2 = 1). rs2304973 is located in intron 1, rs2250333 in intron 2, rs3744700 in intron 4, rs1051009 in 3 untranslated region, and rs8123 in the downstream of the CXCL16 gene. Rs2250333 captures rs8123, however, the other three SNPs (rs2304973, rs3744700, and rs1051009) do not capture the other SNPs. So five SNPs were selected as tagSNPs for the following association analysis and all these SNPs are in low pair-wise LD based on r2 values. 2.3. Genotyping Peripheral blood was collected in vacuum tubes containing EDTA. Genomic DNA was extracted from blood leukocytes using a commercial kit (Tiangen Biotech Co., Ltd., Beijing, China) [22]. Five polymorphisms were genotyped in our case–control population. The SNPs detailed sequence information is publicly available (http://www.ncbi.nlm.nih.gov/SNP/). The polymorphism rs2277680 was genotyped by polymerase chain reaction (PCR)based RFLP analysis as described previously [18]. The polymorphisms rs1051009, rs3744700, rs2304973, and rs2250333 were genotyped by PCR direct sequencing. The primers 5 -ATC ATG GAG AAG ATG GCT AG-3 and 5 -TGT CCA AGT TAT TAT CAC CC-3 were used for amplifying and sequencing the target region containing both the rs1051009 and rs3744700 sites and primers 5 -GGG ATT ATG GAT GTT GTC GG-3 and 5 -GAA GTT GCC ACC AAG AAA GAG-3 for the rs2304973 and rs2250333, respectively. PCR cycling conditions were as follows: an initial denaturation at 95 ◦ C for 5 min, followed by 35 cycles at 95 ◦ C for 30 s, 55 ◦ C (rs1051009 and rs3744700) or 59 ◦ C (rs2304973 and rs2250333) for 30 s, and 72 ◦ C for 1 min. Genotyping was performed in a blind manner so that the performers did not know the case/control status of the subjects. For
162
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
Table 1 Baseline characteristics of the patients. Variables
Age (years) Gender (M/F) BMI (kg/m2 ) DM (%) Hypertension (%) Smoking (%) TC (mmol/l) TG (mmol/l) LDL-C (mmol/l) HDL-C (mmol/l) Glucose (mmol/l)
Entire population (n = 2026)
Subgroup population (n = 320)
CAD (n = 1176)
Controls (n = 850)
P
AMI (n = 160)
CAD-free (n = 160)
P
56.5 ± 9.8 824/352 26.1 ± 3.2 21.8 55.8 50.2 4.74 ± 1.15 2.19 ± 1.57 2.58 ± 0.60 1.45 ± 0.32 6.72 ± 2.90
55.4 ± 10.6 451/399 24.6 ± 3.6 11.3 46.4 29.1 4.52 ± 0.99 2.00 ± 1.59 2.45 ± 0.54 1.51 ± 0.36 5.55 ± 1.64
0.021 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.008 <0.001 <0.001 <0.001
56.2 ± 10.2 100/60 25.4 ± 3.2 16.3 47.5 38.8 4.62 ± 1.12 2.15 ± 1.56 2.52 ± 0.56 1.42 ± 0.33 5.66 ± 1.74
55.8 ± 10.5 100/60 24.8 ± 3.6 16.3 47.5 38.8 4.58 ± 1.06 2.10 ± 1.59 2.47 ± 0.53 1.46 ± 0.37 5.54 ± 1.48
0.652 Match 0.247 Match Match Match 0.528 0.355 0.128 0.433 0.368
AMI indicates acute myocardial infarction; BMI, body mass index; CAD, coronary artery disease; DM, diabetes mellitus; F, female; HDL-C, high density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; M, male; TC, total cholesterol; and TG, triglyceride.
quality control, 10% samples of cases and controls were randomly selected to be tested again by different technicians under masked condition and all results were 100% concordance. Genotypes identified by PCR-RFLP were confirmed by DNA sequencing.
of CHB population, and assuming a dominant model, the study had more than 80% statistical power to detect an association (at P < 0.05) with an OR of 1.35. 3. Results
2.4. Determination of plasma CXCL16 levels 3.1. Baseline characteristics of the study population AMI patients had blood withdrawal at admission to the hospital before any invasive procedure and therapy. The median symptom onset-entry time was 5.3 h (range: 1–12 h). Fasting blood samples were collected from CAD-free subjects after angiography. After being centrifuged at 2000 × g for 20 min, plasma samples were frozen and stored at −70 ◦ C until assay. All samples were thawed only once. CXCL16 concentrations in plasma were measured with the human CXCL16 Quantikine ELISA Kit (R&D Systems, Minneapolis, MN) [16] according to the manufacturer’s instructions. Intra-assay and inter-assay coefficients of variation (CV) were approximately 5% and 10%, respectively, and assay sensitivity was <7 pg/ml.
Table 1 shows the distribution of the clinical and biological characteristics of the subjects. Age and the percentage of men were greater among subjects with CAD than among controls. Compared with the control group, the case group had more smokers and higher rates of hypertension and diabetes. The case group also had significantly higher BMI, LDL-C, serum triglyceride levels, and fasting glucose levels, but lower HDL-C levels than the controls. The proportions of subjects reported taking cholesterol-lowering medications such as statins in the cases and controls were 47.5% and 9.3%, respectively. The baseline characteristics of the subgroup subjects are also shown in Table 1. We found no significant differences in the non-matched variables between the cases and controls.
2.5. Statistical analysis 3.2. Association of CXCL16 SNPs with CAD Distributions of continuous variables were expressed as mean ± SD. Logarithmic transformation was performed on skewed variables, including CXCL16. Statistical significance for differences in quantitative variables was tested by the Student’s independent-samples t test and analysis of covariance (ANCOVA). Categorical data are presented as frequencies (percentage). Differences between categorical variables, genotype/allele frequencies, and Hardy–Weinberg equilibrium were tested by 2 analysis or Fisher’s exact test. Simple correlations between plasma levels of CXCL16 and the other variables were determined by means of the Pearson coefficient; the independence of these associations was evaluated by stepwise multiple linear regression analysis. Independence of associations between SNPs and CAD were assessed after adjustment for other potentially confounding factors (age, gender, body mass index, smoking, cholesterol levels, hypertension and diabetes mellitus) using multiple logistic regression analysis and calculating the adjusted odds ratios and their 95% confidence intervals. All data were analyzed using the statistical software SPSS 11.5 for Windows (SPSS, Inc., Chicago, IL, USA). Two-sided probability values <0.05 were considered significant. In addition, P values were corrected for multiple comparisons for five SNPs using the Bonferroni adjustment method, which changed the required significance level from <0.05 to <0.01 (0.05/5 = 0.01). For power calculations the software package Quanto 1.2 was used (http://hydra.usc.edu/gxe). On the basis of MAF ranging from 0.089 for rs2304973 to 0.444 for rs1051009 reported in HapMap
Genotype and allele frequencies of the CXCL16 gene polymorphisms are shown in Table 2. All five SNPs investigated in both case and control samples followed Hardy–Weinberg equilibrium (HWE) (P > 0.1, Table 2). The minor allele frequencies of the polymorphisms were similar to the reported frequencies in Han Chinese population from the International HapMap Project. The only significant difference between patients with CAD and controls was found for the SNP rs3744700 (P = 0.001, Table 2). As the T allele of rs3744700 is ancestral allele (http://www.ncbi.nlm.nih.gov/SNP/), we used it as referential allele. The G allele frequency was significantly higher in CAD patients than in control subjects (0.960 vs 0.934; P < 0.001). To control Type I error the Bonferroni correction was used, with the adjusted alpha value of 0.01. The association between rs3744700 and CAD was still significant. On the basis of multivariable logistic regression analysis with adjustment for cardiovascular risk factors, subjects bearing the GG homozygotes had significantly increased susceptibility to CAD compared with the T allele carriers (GT + TT) (adjusted OR, 1.77; 95% CI, 1.28–2.43; P < 0.001). Thus, CXCL16 rs3744700 polymorphism was an independent risk factor for CAD. Based on the number of luminal narrowing ≥50% in the three major coronary arteries, we did not observe any significant association between the rs3744700 polymorphism and angiographic severity of CAD (Table 3). There was no significant difference in allele frequencies or genotype distributions of the other four SNPs between CAD patients and controls. Also, there
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
163
Table 2 Genotype distribution of CXCL16 tagSNPs in CAD patients and in controls. TagSNPs
Location
Allelea (1/2)
Subject group
MAF
Pb (HWE)
Pb (genotype)
rs2304973
intron 1
C/T
Case Control
997 (84.8) 171 (14.5) 731 (86.0) 113 (13.3)
8 (0.7) 6 (0.7)
0.08 0.07
0.821 0.479
0.727
0.878
1.02 (0.78–1.34)
rs2250333
intron 2
C/T
Case Control
611 (52.0) 462 (39.3) 426 (50.1) 345 (40.6)
103 (8.7) 79 (9.3)
0.28 0.30
0.243 0.450
0.707
0.335
0.91 (0.75–1.10)
rs2277680
exon 4
A/G
Case Control
500 (42.5) 542 (46.1) 341 (40.1) 391 (46.0)
134 (11.4) 0.34 118 (13.9) 0.37
0.479 0.727
0.208
0.301
0.90 (0.75–1.10)
rs3744700
intron 4
T/G
Case Control
87 (7.4) 1085 (92.3) 0.04 105 (12.3) 741 (87.2) 0.07
0.117 0.892
0.001
<0.001
1.77 (1.28–2.43)
rs1051009
3 UTR
C/T
Case Control
0.190 0.111
0.268
0.148
0.86 (0.71–1.05)
Genotype frequency (%) 1/1
1/2
4 (0.3) 4 (0.5)
414 (35.2) 549 (46.7) 278 (32.7) 396 (46.6)
2/2
Dominant modelc P value
213 (18.1) 0.42 176 (20.7) 0.44
OR (95% CI)
The significance of bold values was 0.01. a Allele 1 represent the ancestral allele of each SNP. b P values were obtained by 2 -test. c For the dominant model, the 1/1 genotype was the referent (for rs3744700, the 1/1 + 1/2 genotype were referent). The associations were performed by multivariate logistic regression analysis adjusted with covariates (age, gender, BMI, smoking, hypertension, diabetes, TC, TG, HDL-C, and LDL-C); SNP, singe nucleotide polymorphism; HWE, Hardy–Weinberg equilibrium; 3 UTR, 3 untranslated region. OR, odds ratio; 95%CI, 95% confidence interval; other abbreviations as in Table 1. Table 3 Relation between the rs3744700 genotypes and the angiographic severity of CAD. No. of significantly diseased vessels
1 (n = 397) 2 (n = 373) 3 (n = 406)
rs3744700 genotypes
Allele frequencies
GG
GT
TT
P value
GT + TT
P value
G
T
P value
366 (92.2) 345 (92.5) 374 (92.1)
30 (7.5) 28 (7.5) 29 (7.2)
1 (0.3) 0 (0) 3 (0.7)
0.506 d.f. = 4
31 (7.8) 28 (7.5) 32 (7.9)
0.979 d.f. = 2
762 (96.0) 718 (96.2) 777 (95.7)
32 (4.0) 28 (3.8) 35 (4.3)
0.856 d.f. = 2
P values were obtained by 2 -test.
was no significant association between the genotype distributions of other four SNPs, including SNP rs2277680, and the number of angiographically documented significantly diseased vessels (data not shown).
CXCL16 level was significantly higher in subjects with GG genotype than those with T allele (GT and TT genotypes) in the control group (P = 0.015). Similarly, AMI patients with GG homozygote had higher plasma CXCL16 concentration than that those with T allele (P = 0.027).
3.3. Identification of regulatory elements in intron 4 3.5. SNP rs3744700 polymorphism and plasma lipid parameters To identify possible alternative splice acceptor sites within intron 4 of human CXCL16, the 439 bp genomic DNA sequence of intron 4 was analyzed using a splice site prediction program from the Berkeley Drosophila Genome Project (http://www.fruitfly.org/seq tools/splice.html). One alternative splice acceptor site demonstrating a confidence score 1.00 was identified within intron 2 of CXCL16, and rs3744700 locates just 61 bp upstream form it. Next, we performed a computer-based analysis to identify putative regulatory elements in intron 4 of the CXCL16 gene. We found that intron 4 harbors several putative regulatory elements, including the consensus sequence of GATA, which is affected by the rs3744700 polymorphism, indicating that the intronic variant may participate in the CXCL16 transcription. 3.4. Plasma CXCL16 levels in subset subjects Plasma CXCL16 levels were significantly higher in AMI patients than CAD-free controls (2.36 ± 0.40 ng/ml vs 1.88 ± 0.36 ng/ml, P < 0.01). However, there was no association between plasma CXCL16 levels and the severity of coronary atherosclerosis presented as the number of ≥50% stenotic vessels: 2.29 ± 0.38 ng/ml in 1-vessel disease, 2.37 ± 0.38 ng/ml in 2-vessel disease, and 2.42 ± 0.43 ng/ml in 3-vessel disease, respectively (Fig. 1). Next, we wanted to know whether plasma CXCL16 levels were related to genotype of rs3744700 (Table 4). Due to only one TT homozygote was seen in the CAD-free group and none in the AMI group, those rare homozygotes were pooled with heterozygotes when performing comparisons of CXCL16 plasma levels. We found that
Based on its potential to act as a scavenger receptor for oxLDL, CXCL16 may be involved in lipid metabolism. Because of the low occurrence of subjects with TT genotype, the T allele carriers (GT + TT) were ascribed to a single group. The lipid profiles of control subjects without lipid-lowering therapy (n = 771) with different rs3744700 genotypes are shown in Table I (available online supplementary material). No significant differences were seen between the rs3744700 genotypes and plasma lipid levels (TG, TC, LDL-C, and HDL-C). 4. Discussion In the present study, we have performed the first comprehensive analysis of five common SNPs in CXCL16 gene in 1176 CAD patients and 850 controls in Chinese Han population. Our results show that after adjustment for various risk factors, one SNP Table 4 Plasma CXCL16 levels in subgroup subjects according to rs3744700 genotype. Group
Plasma CXCL16 levels (ng/ml) GG
GT + TT
CAD-free (n = 160) AMI (n = 160)
1.92 ± 0.35 (n = 138) 2.38 ± 0.40 (n = 148)**
1.64 ± 0.40 (n = 22)* 2.11 ± 0.44 (n = 12)* , **
Results are means ± SD. * P < 0.05 compared with GG. ** P < 0.01 compared with GG CAD-free group.
164
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
Fig. 1. Plasma CXCL16 levels in patients with and without CAD. Compared with CAD-free subjects, AMI patients had significantly higher CXCL16 levels (P < 0.01). There were no differences in CXCL16 levels among AMI patients with 1-, 2- and 3vessel diseases. CAD (-): CAD-free, 1-VD: 1-vessel disease, 2-VD: 2-vessel disease, 3-VD: 3-vessel disease.
(rs3744700) exhibits significant association with CAD. The prevalence of the GG homozygotes is significantly higher among CAD patients than among control subjects. The GG genotypes also had increased plasma CXCL16 levels compared with the T allele carriers in both cases and control subjects. However, we found no evidence for an association of the other gene variations with CAD in Chinese Han population. CAD is a major cause of morbidity and mortality around the would, and inflammation is believed to play an important role in the pathogenesis of the condition [1]. The fundamental importance of chemokines for atherogenesis, progression, and destabilization of atherosclerotic plaques is now widely appreciated [2]. To date, only one study has investigated the association between polymorphism in CXCL16 and CAD risk [18]. They found that, for rs2277680, a nonsynonymous coding SNP on exon 4, there was no difference in the frequency of allele and genotypes between 387 patients who survived the first MI and 387 sex- and age-matched controls. However, carriers of the rare 181Val allele had a higher degree of angiographically determined coronary artery stenosis and smaller luminal diameter before PCI, suggesting this SNP was marginally associated with the severity of coronary artery stenosis in CAD patients. Our results support the finding that SNP rs2277680 is not associated with CAD risk. However, based on the number of luminal narrowing ≥50% in the three major coronary arteries [19], we did not observe any significant association between all five tagSNPs investigated in our study, including rs2277680, and the angiographic severity of CAD. The difference between the Sweden study and ours could be due to the different methods used in evaluating angiographic severity of CAD. They quantitatively analyzed the severity of coronary lesions using the QCA-CMS system [18]. Besides, the different genetic background of the two populations, as well as the relatively small sample size in the Sweden study may also contribute to the different conclusions. The rs3744700 resides in intron 4 of the CXCL16 gene and has not beet investigated before. The observed association with the intronic SNP in this study would be explained by either rs3744700 being directly pathogenic or it being in linkage disequilibrium with another causal variant. SNP rs3744700, though intronic, could itself be causal. It is plausible that disease-associated variants with modest effects will
be distributed proportionately between coding and noncoding sequences of the genome [23]. An increasing amount of evidence indicates that intronic genetic variations can have deleterious implications for gene splicing [24]. Indeed, recent evidence from many laboratories indicates that intronic variants can be functional and as a result are often the cause of human diseases [25–27]. Therefore, the disease-associated rs3744700 in CXCL16 may be functional in terms of splicing regulation or alternatively they may monitor functional elements within the gene associated with CAD risk. Using online programs, we found that rs3744700 locates on 61 bp from the splice acceptor site, and identified the consensus sequence of GATA that is affected by the rs3744700 polymorphism. Furthermore, we found that the plasma CXCL16 levels were significant higher in AMI patients than CAD-free controls, supporting the concept that CAD patients are characterized by higher plasma levels of CXCL16 [14,17]. The GG homozygotes also had higher plasma CXCL16 levels compared with T allele carriers in both cases and control subjects. Results of our computer-based analysis and CXCL16 measurement strongly suggest that rs3744700 may be functional. We speculate that the association of this SNP with CXCL16 plasma levels and CAD risk has arisen from a direct effect of this SNP on CXCL16 expression or mRNA splicing. Additional functional analyses of this SNP are needed to elucidate the underlying mechanism. We cannot exclude the possibility that other common CXCL16 variants may account for the increased risk of CAD we observed. However, the data generated by the HapMap Project for Chinese Han population suggests that this is unlikely. There are nine SNPs with MAF ≥5% in the CXCL16 gene region, including 3 kb upstream to 1 kb downstream of the gene. The rs3744700 is in low pair-wise LD based on r2 values with other eight SNPs. However, the present study did not include SNPs with MAF <5%, and the MAF distribution of SNPs from the International HapMap Project shows that more than 40% of SNPs have MAF <5%. The rare SNPs hypothesis suggests that targeting rare SNPs in large case–control association studies has more power to detect causal SNPs than does targeting common SNPs [28]. Thus, we cannot exclude the possibility that the rs3744700 SNP might be a functionally neutral marker that is in linkage disequilibrium with a rare functional polymorphism located elsewhere at the CXCL16 gene. Recently, Pollin et al. [29] found that associated SNP can be in linkage disequilibrium with causal variant more than 800 kb away. Recent genome-wide association studies (GWAS) of CAD have focused on a few chromosomal regions with strong signals [30–32]. It is also possible that rs3744700 might be linked with a mutation of another gene adjacent to CXCL16 gene, or with the true causative locus far away from the CXCL16 gene, such as one of the variants identified from reported GWAS data. Biologically, atherosclerosis is a continuous process characterized by the activation of a variety of factors including inflammation. So we conducted a stratified analysis by the severity of CAD. Our data show that neither the plasma CXCL16 levels nor the genotype of the five SNPs, including rs3744700, are associated with the severity of CAD. We hypothesize that CXCL16 is not involved in the late stage of multivessel-disease development but instead is involved directly in the early stage of CAD development. The present study has several strengths and limitations that should be addressed briefly. The strengths include the use of a pure Chinese Han population from the northeast region of China to eliminate false positive results due to population stratification. Moreover, the large sample size of the present work, which is essential for the investigation of the mostly modest influence of SNPs on clinical phenotypes, deserves to be mentioned. Finally angiographical profiles of all subjects were evaluated. Regarding limitations, it should be mentioned that different geographical and racial backgrounds of the individuals can affect the consequences
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
of an association study. Therefore our results need to be confirmed in other populations. In summary, our study shows a strong association between CXCL16 and CAD in Chinese Han population. Future research is needed to confirm the present findings, by replicating this association in independent populations of various ethnic origins, especially in large-scale prospective cohort studies. Furthermore, additional research will be required to better define the precise effects of the intronic variant of CXCL16 gene. Conflict of interest There are no conflicts of interest. Acknowledgements The work was supported in part by the Military Medical Science and Technique Foundation during the 11th Five-Year Plan Period (No. 06G021). The authors thank Dr. Lianming Liao for his reading and commenting of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2009.11.004. References [1] Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med 2005;352:1685–95. [2] Zernecke A, Shagdarsuren E, Weber C. Chemokines in atherosclerosis: an update. Arterioscler Thromb Vasc Biol 2008;28:1897–908. [3] Moore KJ, Freeman MW. Scavenger receptors in atherosclerosis: beyond lipid uptake. Arterioscler Thromb Vasc Biol 2006;26:1702–11. [4] Matloubian M, David A, Engel S, Ryan JE, Cyster JG. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat Immunol 2000;1:298–304. [5] Shimaoka T, Kume N, Minami M, et al. Molecular cloning of a novel scavenger receptor for oxidized low density lipoprotein, SR-PSOX, on macrophages. J Biol Chem 2000;275:40663–6. [6] Shimaoka T, Nakayama T, Kume N, et al. Cutting edge: SR-PSOX/CXC chemokine ligand 16 mediates bacterial phagocytosis by APCs through its chemokine domain. J Immunol 2003;171:1647–51. [7] Gough PJ, Garton KJ, Wille PT, et al. A disintegrin and metalloproteinase 10mediated cleavage and shedding regulates the cell surface expression of CXC chemokine ligand 16. J Immunol 2004;172:3678–85. [8] Abel S, Hundhausen C, Mentlein R, et al. The transmembrane CXCchemokine ligand 16 is induced by IFN-gamma and TNF-alpha and shed by the activity of the disintegrin-like metalloproteinase ADAM10. J Immunol 2004;172:6362–72. [9] Minami M, Kume N, Shimaoka T, et al. Expression of SR-PSOX, a novel cellsurface scavenger receptor for phosphatidylserine and oxidized LDL in human atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2001;21:1796–800. [10] Wuttge DM, Zhou X, Sheikine Y, et al. CXCL16/SR-PSOX is an interferon-gammaregulated chemokine and scavenger receptor expressed in atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2004;24:750–5.
165
[11] Wagsater D, Olofsson PS, Norgren L, Stenberg B, Sirsjo A. The chemokine and scavenger receptor CXCL16/SR-PSOX is expressed in human vascular smooth muscle cells and is induced by interferon gamma. Biochem Biophys Res Commun 2004;325:1187–93. [12] Aslanian AM, Charo IF. Targeted disruption of the scavenger receptor and chemokine CXCL16 accelerates atherosclerosis. Circulation 2006;114:583– 90. [13] Tenger C, Sundborger A, Jawien J, Zhou X. IL-18 accelerates atherosclerosis accompanied by elevation of IFN-gamma and CXCL16 expression independently of T cells. Arterioscler Thromb Vasc Biol 2005;25:791–6. [14] Smith C, Halvorsen B, Otterdal K, et al. High levels and inflammatory effects of soluble CXC ligand 16 (CXCL16) in coronary artery disease: down-regulatory effects of statins. Cardiovasc Res 2008;79:195–203. [15] Galkina E, Harry BL, Ludwig A, et al. CXCR6 promotes atherosclerosis by supporting T-cell homing, interferon-gamma production, and macrophage accumulation in the aortic wall. Circulation 2007;116:1801– 11. [16] Sheikine Y, Bang CS, Nilsson L, et al. Decreased plasma CXCL16/SR-PSOX concentration is associated with coronary artery disease. Atherosclerosis 2006;188:462–6. [17] Lehrke M, Millington SC, Lefterova M, et al. CXCL16 is a marker of inflammation, atherosclerosis, and acute coronary syndromes in humans. J Am Coll Cardiol 2007;49:442–9. [18] Lundberg GA, Kellin A, Samnegard A, et al. Severity of coronary artery stenosis is associated with a polymorphism in the CXCL16/SR-PSOX gene. J Intern Med 2005;257:415–22. [19] Chen Q, Reis SE, Kammerer CM, et al. Association between the severity of angiographic coronary artery disease and paraoxonase gene polymorphisms in the National Heart, Lung, and Blood Institute-sponsored Women’s Ischemia Syndrome Evaluation (WISE) study. Am J Hum Genet 2003;72:13–22. [20] Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–5. [21] Ke X, Cardon LR. Efficient selective screening of haplotype tag SNPs. Bioinformatics 2003;19:287–8. [22] Pei F, Han Y, Zhang X, et al. Association of interleukin-18 gene promoter polymorphisms with risk of acute myocardial infarction in northern Chinese Han population. Clin Chem Lab Med 2009;47:523–9. [23] Carlson CS, Eberle MA, Kruglyak L, Nickerson DA. Mapping complex disease loci in whole-genome association studies. Nature 2004;429:446–52. [24] Pagani F, Baralle FE. Genomic variants in exons and introns: identifying the splicing spoilers. Nat Rev Genet 2004;5:389–96. [25] Bouatia-Naji N, Rocheleau G, Van Lommel L, et al. A polymorphism within the G6PC2 gene is associated with fasting plasma glucose levels. Science 2008;320:1085–8. [26] Choi JW, Park CS, Hwang M, et al. A common intronic variant of CXCR3 is functionally associated with gene expression levels and the polymorphic immune cell responses to stimuli. J Allergy Clin Immunol 2008;122(6):1119– 26.e7. [27] Chen Q, Razzaghi H, Demirci FY, Kamboh MI. Functional significance of lipoprotein lipase HindIII polymorphism associated with the risk of coronary artery disease. Atherosclerosis 2008;200:102–8. [28] Gorlov IP, Gorlova OY, Sunyaev SR, Spitz MR, Amos CI. Shifting paradigm of association studies: value of rare single-nucleotide polymorphisms. Am J Hum Genet 2008;82:100–12. [29] Pollin TI, Damcott CM, Shen H, et al. A null mutation in human APOC3 confers a favorable plasma lipid profile and apparent cardioprotection. Science 2008;322:1702–5. [30] McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science 2007;316:1488–91. [31] 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. [32] 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.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
An intron polymorphism in the CXCL16 gene is associated with increased risk of coronary artery disease in Chinese Han population: A large angiography-based study Mingfang Huang a,b , Yaling Han b,∗ , Xiaolin Zhang b , Fang Pei b , Jie Deng b , Jian Kang b , Chenghui Yan b a b
Department of Cardiology, Xijing Hospotal, Fourth Military Medical University, Xi’an, China Department of Cardiology, Shenyang Northern Hospital, 83 Wenhua Road, Shenyang 110016, China
a r t i c l e
i n f o
Article history: Received 16 June 2009 Received in revised form 3 November 2009 Accepted 3 November 2009 Available online 10 November 2009 Keywords: CXCL16 Chemokine Coronary angiography Coronary artery disease Genetics Polymorphism
a b s t r a c t Objective: Experimental and clinical observations suggest that CXCL16, a recently discovered transmembrane chemokine combining functions of a chemokine and a scavenger receptor, could be an important player in atherosclerosis, but the relationship of its common variants with coronary artery disease (CAD) has not been extensively studied. Methods: We designed an angiography-based case-controlled study consisting of 1176 CAD patients and 850 control subjects to investigate the association between five common single nucleotide polymorphisms (SNPs) of CXCL16 gene and CAD risk in Chinese Han population. The plasma concentration of CXCL16 was measured by enzyme-linked immunosorbent assay. Results: No significant differences were observed for the distributions of rs2250333, rs2304973, rs2277680, and rs1051009 between CAD patients and control subjects. However, both the genotype and allele frequencies of rs3744700 showed significant differences between cases and controls (P = 0.001 and P < 0.001, respectively). The GG homozygotes had significantly higher CAD risk compared with the T allele carriers (GT + TT) (OR, 1.77; 95% CI, 1.28–2.43; adjusted P < 0.001) in a logistic regression model after adjustment for the conventional risk factors for CAD. The GG genotypes also had increased plasma CXCL16 levels compared with T allele carriers in both cases and control subjects. Conclusions: SNP rs3744700 of CXCL16 gene is independently associated with the development of CAD in Chinese Han population, and GG homozygote which is associated with increased expression of CXCL16 may have a promoting effect on CAD. © 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Coronary artery disease (CAD) and its main complication, myocardial infarction (MI), are leading causes of death and disability worldwide. The pathogenesis of coronary atherosclerosis involves inherited, behavioural and environmental factors. Inflammatory processes are thought to promote atherogenesis from early lesion formation to unstable plaque rupture [1]. Infiltrates of activated macrophages and T-cells are prominent in human and experimental atherosclerotic lesions. Recruitment of these cells to lesions is coordinated by the production of chemokines, small proteins which interact with specific chemoattractant receptors belonging to the superfamily of G protein-coupled receptors [2]. Once present in the subendothelial space, macrophages engulf oxidatively modified low-density lipoproteins (ox-LDL) through their scavenger receptors and develop into cholesterol-filled foam
∗ Corresponding author. Tel.: +86 24 23922184; fax: +86 24 23911006. E-mail address: [email protected] (Y. Han). 0021-9150/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2009.11.004
cells, a hallmark of atherosclerotic lesion formation. Therefore, scavenger receptors also play a crucial role in the pathogenesis of atherosclerosis [3]. CXCL16 (CXC ligand 16)/SR-PSOX (scavenger receptor that binds phosphatidylserine and oxidized lipoprotein), a recently discovered transmembrane chemokine [4], is a member of the CXC chemokine subfamily. Unlike other members of this subgroup, it is structurally similar to CX3CL1, containing four distinct domains; that is, a chemokine domain tethered to the cell surface via a mucin-like stack, which in turn is attached to transmembrane and cytoplasmic domains. The chemokine domain acts not only as an attractant for cells expressing the CXCR6 receptor [4], but also as a scavenger receptor facilitating uptake of ox-LDL, phosphatidylserine, bacteria and apoptotic cells [5,6]. The extracellular part of CXCL16 undergoes cleavage by the metalloproteinase ADAM 10 creating a soluble chemokine (sol-CXCL16) [7,8], which activates CXCR6-expressing T-cells. Functioning as a chemokine and a scavenger receptor, CXCL16/SR-PSOX may be an interesting player in the formation of atherosclerotic lesions. CXCL16 is expressed on macrophages,
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
dendritic cells, and aortic smooth muscle cells. Enhanced CXCL16 expression has been detected in atherosclerotic lesions from humans and murine models of atherosclerosis (apoE and LDL receptor knockouts) [9–12]. Furthermore, several pro-atherogenic inflammatory stimuli increase CXCL16 expression [4,8,11,13,14], enhancing uptake of ox-LDL and facilitating foam cell formation, as well as promoting recruitment of T-cell to the atherosclerotic plaque [10]. In apoE−/− mice, genetic deletion of CXCR6 attenuated atherosclerosis, accompanied by a decreased percentage of CXCR6+ T-cells within the aortas, indicating a pro-atherogenic role of CXCL16/CXCR6 interaction [15]. However, a recent rodent study reported that CXCL16-deficient LDL receptor −/− mice were associated with accelerated atherosclerosis [12]. Human epidemiological studies evaluating CXCL16 as a potential biomarker of CAD have produced controversial results. While Sheikine et al. [16] found decreased CXCL16 levels in both stable and unstable angina, Lehrke et al. [17] reported increased CXCL16 levels in a large population of CAD patients, particularly in those with unstable disease. And a recent study further support the concept that CAD patients are characterized by increased plasma levels of CXCL16 [14]. The role of admission CXCL16 levels in patients with acute MI, however, has been less extensively studied. Genetic variation of CXCL16 could modulate CXCL16 production and function. A preliminary study by Lundberg et al. found that a single nucleotide polymorphism (SNP) in the fourth exon of the CXCL16 gene was significantly associated with the severity of coronary artery stenosis in CAD patients [18]. However, the association study is limited to a SNP, which fails to include all the potential risk-conferring variations in the CXCL16 gene and their potential effects on CAD risk. To address these issues, we first conducted an analysis of HapMap-tagged SNPs in the CXCL16 gene, including 181Ala >Val (rs2277680), and its relation to the risk of CAD in a Chinese Han population. Next, we investigated the association of CXCL16 genetic variation with its plasma levels in a subgroup of patients.
2. Methods 2.1. Study populations A total of 1176 patients with CAD and 850 control individuals, aged between 29 and 75 years, were enrolled in the study from June 2006 to April 2009. Patients and controls were unrelated individuals of Chinese Han from the northeast region of China. All the subjects had undergone coronary angiography for the evaluation of suspected or established CAD at Northern Hospital, Shenyang, China. The study complied with the declaration of Helsinki. Approval was obtained from the local Ethics Committees and all patients gave their informed consent. Inclusion criteria for CAD were ≥50% narrowing of the lumina of at least 1 of the major coronary arteries by coronary angiography. Additionally, angiographic severity of disease was defined as 0-, 1-, 2- or 3-vessel disease based on the number of luminal narrowing ≥50% in the three major coronary arteries [19]. The diagnosis of MI was based on typical electrocardiographic changes as well as on increases in the serum activities of enzymes such as creatinine kinase, aspartate aminotransferase, and lactate dehydrogenase and in the serum concentration of troponin T. The diagnosis was confirmed by the presence of a wall-motion abnormality on left ventriculography as well as by identification of the responsible stenosis in any of the major coronary arteries or in the left main trunk by coronary angiography. The control subjects were selected from the subjects admitted to the hospital for the evaluation of chest pain, whose major coronary artery had no more than 20% stenosis, and did not have any vascular disease. Control subjects
161
were recruited during the same period as the patients. A complete clinical history was obtained from all subjects, and the vascular risk factors are summarized in Table 1. We did not include subjects with a history of hematologic, neoplastic, renal, liver, or thyroid disease. Furthermore, patients who had inflammatory, infectious or autoimmune diseases, type-1 diabetes mellitus or familial hyperlipidemia were also excluded. To further investigate the impact of gene polymorphisms on pathological phenotypes, the levels of plasma CXCL16 were determined in a subset population selected from our original cohort between March 2008 and April 2009. They consisted of 160 subjects with first episode of acute MI (AMI) and 160 subjects with completely normal coronary arteries (CAD-free). The subset cases and controls were individually matched in terms of age (≤5 years), gender, smoking status, presence of hypertension, and diabetes mellitus. Because statin therapy can significantly reduces plasma levels of CXCL16 [14], all subset participants were not on lipidlowering medication. 2.2. Characterizing linkage disequilibrium and tagSNPs selection The CXCL16 gene contains five exons and spans 6283 bp on chromosome 17 (NCBI accession NM 022059). SNPs were retrieved from Hapmap database for Han population sample (release no. 21a/phase II, Jan 07, population: CHB; http://www.hapmap.org). Pair-wise LD between SNPs was assessed using Lewontin’s D statistic and the squared correlation statistic r2 . The Haploview program was used to calculate the LD coefficients and define haplotype blocks [20]. TagSNPs were selected using the r2 -based Tagger program [21] with pair-wise r2 ≥ 0.80 and minor allele frequency (MAF) ≥5%. Nine SNPs covering 10.28 kb including CXCL16 (spanning from 3 kb upstream to 1 kb downstream of the gene) were identified. SNP rs8071612 is located in intron 3, and rs1876444, rs2277680 and rs1050998 in exon 4. Rs2277680 captures the other three SNPs, and these four SNPs are in complete association with one another (D = 1, r2 = 1). rs2304973 is located in intron 1, rs2250333 in intron 2, rs3744700 in intron 4, rs1051009 in 3 untranslated region, and rs8123 in the downstream of the CXCL16 gene. Rs2250333 captures rs8123, however, the other three SNPs (rs2304973, rs3744700, and rs1051009) do not capture the other SNPs. So five SNPs were selected as tagSNPs for the following association analysis and all these SNPs are in low pair-wise LD based on r2 values. 2.3. Genotyping Peripheral blood was collected in vacuum tubes containing EDTA. Genomic DNA was extracted from blood leukocytes using a commercial kit (Tiangen Biotech Co., Ltd., Beijing, China) [22]. Five polymorphisms were genotyped in our case–control population. The SNPs detailed sequence information is publicly available (http://www.ncbi.nlm.nih.gov/SNP/). The polymorphism rs2277680 was genotyped by polymerase chain reaction (PCR)based RFLP analysis as described previously [18]. The polymorphisms rs1051009, rs3744700, rs2304973, and rs2250333 were genotyped by PCR direct sequencing. The primers 5 -ATC ATG GAG AAG ATG GCT AG-3 and 5 -TGT CCA AGT TAT TAT CAC CC-3 were used for amplifying and sequencing the target region containing both the rs1051009 and rs3744700 sites and primers 5 -GGG ATT ATG GAT GTT GTC GG-3 and 5 -GAA GTT GCC ACC AAG AAA GAG-3 for the rs2304973 and rs2250333, respectively. PCR cycling conditions were as follows: an initial denaturation at 95 ◦ C for 5 min, followed by 35 cycles at 95 ◦ C for 30 s, 55 ◦ C (rs1051009 and rs3744700) or 59 ◦ C (rs2304973 and rs2250333) for 30 s, and 72 ◦ C for 1 min. Genotyping was performed in a blind manner so that the performers did not know the case/control status of the subjects. For
162
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
Table 1 Baseline characteristics of the patients. Variables
Age (years) Gender (M/F) BMI (kg/m2 ) DM (%) Hypertension (%) Smoking (%) TC (mmol/l) TG (mmol/l) LDL-C (mmol/l) HDL-C (mmol/l) Glucose (mmol/l)
Entire population (n = 2026)
Subgroup population (n = 320)
CAD (n = 1176)
Controls (n = 850)
P
AMI (n = 160)
CAD-free (n = 160)
P
56.5 ± 9.8 824/352 26.1 ± 3.2 21.8 55.8 50.2 4.74 ± 1.15 2.19 ± 1.57 2.58 ± 0.60 1.45 ± 0.32 6.72 ± 2.90
55.4 ± 10.6 451/399 24.6 ± 3.6 11.3 46.4 29.1 4.52 ± 0.99 2.00 ± 1.59 2.45 ± 0.54 1.51 ± 0.36 5.55 ± 1.64
0.021 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.008 <0.001 <0.001 <0.001
56.2 ± 10.2 100/60 25.4 ± 3.2 16.3 47.5 38.8 4.62 ± 1.12 2.15 ± 1.56 2.52 ± 0.56 1.42 ± 0.33 5.66 ± 1.74
55.8 ± 10.5 100/60 24.8 ± 3.6 16.3 47.5 38.8 4.58 ± 1.06 2.10 ± 1.59 2.47 ± 0.53 1.46 ± 0.37 5.54 ± 1.48
0.652 Match 0.247 Match Match Match 0.528 0.355 0.128 0.433 0.368
AMI indicates acute myocardial infarction; BMI, body mass index; CAD, coronary artery disease; DM, diabetes mellitus; F, female; HDL-C, high density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; M, male; TC, total cholesterol; and TG, triglyceride.
quality control, 10% samples of cases and controls were randomly selected to be tested again by different technicians under masked condition and all results were 100% concordance. Genotypes identified by PCR-RFLP were confirmed by DNA sequencing.
of CHB population, and assuming a dominant model, the study had more than 80% statistical power to detect an association (at P < 0.05) with an OR of 1.35. 3. Results
2.4. Determination of plasma CXCL16 levels 3.1. Baseline characteristics of the study population AMI patients had blood withdrawal at admission to the hospital before any invasive procedure and therapy. The median symptom onset-entry time was 5.3 h (range: 1–12 h). Fasting blood samples were collected from CAD-free subjects after angiography. After being centrifuged at 2000 × g for 20 min, plasma samples were frozen and stored at −70 ◦ C until assay. All samples were thawed only once. CXCL16 concentrations in plasma were measured with the human CXCL16 Quantikine ELISA Kit (R&D Systems, Minneapolis, MN) [16] according to the manufacturer’s instructions. Intra-assay and inter-assay coefficients of variation (CV) were approximately 5% and 10%, respectively, and assay sensitivity was <7 pg/ml.
Table 1 shows the distribution of the clinical and biological characteristics of the subjects. Age and the percentage of men were greater among subjects with CAD than among controls. Compared with the control group, the case group had more smokers and higher rates of hypertension and diabetes. The case group also had significantly higher BMI, LDL-C, serum triglyceride levels, and fasting glucose levels, but lower HDL-C levels than the controls. The proportions of subjects reported taking cholesterol-lowering medications such as statins in the cases and controls were 47.5% and 9.3%, respectively. The baseline characteristics of the subgroup subjects are also shown in Table 1. We found no significant differences in the non-matched variables between the cases and controls.
2.5. Statistical analysis 3.2. Association of CXCL16 SNPs with CAD Distributions of continuous variables were expressed as mean ± SD. Logarithmic transformation was performed on skewed variables, including CXCL16. Statistical significance for differences in quantitative variables was tested by the Student’s independent-samples t test and analysis of covariance (ANCOVA). Categorical data are presented as frequencies (percentage). Differences between categorical variables, genotype/allele frequencies, and Hardy–Weinberg equilibrium were tested by 2 analysis or Fisher’s exact test. Simple correlations between plasma levels of CXCL16 and the other variables were determined by means of the Pearson coefficient; the independence of these associations was evaluated by stepwise multiple linear regression analysis. Independence of associations between SNPs and CAD were assessed after adjustment for other potentially confounding factors (age, gender, body mass index, smoking, cholesterol levels, hypertension and diabetes mellitus) using multiple logistic regression analysis and calculating the adjusted odds ratios and their 95% confidence intervals. All data were analyzed using the statistical software SPSS 11.5 for Windows (SPSS, Inc., Chicago, IL, USA). Two-sided probability values <0.05 were considered significant. In addition, P values were corrected for multiple comparisons for five SNPs using the Bonferroni adjustment method, which changed the required significance level from <0.05 to <0.01 (0.05/5 = 0.01). For power calculations the software package Quanto 1.2 was used (http://hydra.usc.edu/gxe). On the basis of MAF ranging from 0.089 for rs2304973 to 0.444 for rs1051009 reported in HapMap
Genotype and allele frequencies of the CXCL16 gene polymorphisms are shown in Table 2. All five SNPs investigated in both case and control samples followed Hardy–Weinberg equilibrium (HWE) (P > 0.1, Table 2). The minor allele frequencies of the polymorphisms were similar to the reported frequencies in Han Chinese population from the International HapMap Project. The only significant difference between patients with CAD and controls was found for the SNP rs3744700 (P = 0.001, Table 2). As the T allele of rs3744700 is ancestral allele (http://www.ncbi.nlm.nih.gov/SNP/), we used it as referential allele. The G allele frequency was significantly higher in CAD patients than in control subjects (0.960 vs 0.934; P < 0.001). To control Type I error the Bonferroni correction was used, with the adjusted alpha value of 0.01. The association between rs3744700 and CAD was still significant. On the basis of multivariable logistic regression analysis with adjustment for cardiovascular risk factors, subjects bearing the GG homozygotes had significantly increased susceptibility to CAD compared with the T allele carriers (GT + TT) (adjusted OR, 1.77; 95% CI, 1.28–2.43; P < 0.001). Thus, CXCL16 rs3744700 polymorphism was an independent risk factor for CAD. Based on the number of luminal narrowing ≥50% in the three major coronary arteries, we did not observe any significant association between the rs3744700 polymorphism and angiographic severity of CAD (Table 3). There was no significant difference in allele frequencies or genotype distributions of the other four SNPs between CAD patients and controls. Also, there
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
163
Table 2 Genotype distribution of CXCL16 tagSNPs in CAD patients and in controls. TagSNPs
Location
Allelea (1/2)
Subject group
MAF
Pb (HWE)
Pb (genotype)
rs2304973
intron 1
C/T
Case Control
997 (84.8) 171 (14.5) 731 (86.0) 113 (13.3)
8 (0.7) 6 (0.7)
0.08 0.07
0.821 0.479
0.727
0.878
1.02 (0.78–1.34)
rs2250333
intron 2
C/T
Case Control
611 (52.0) 462 (39.3) 426 (50.1) 345 (40.6)
103 (8.7) 79 (9.3)
0.28 0.30
0.243 0.450
0.707
0.335
0.91 (0.75–1.10)
rs2277680
exon 4
A/G
Case Control
500 (42.5) 542 (46.1) 341 (40.1) 391 (46.0)
134 (11.4) 0.34 118 (13.9) 0.37
0.479 0.727
0.208
0.301
0.90 (0.75–1.10)
rs3744700
intron 4
T/G
Case Control
87 (7.4) 1085 (92.3) 0.04 105 (12.3) 741 (87.2) 0.07
0.117 0.892
0.001
<0.001
1.77 (1.28–2.43)
rs1051009
3 UTR
C/T
Case Control
0.190 0.111
0.268
0.148
0.86 (0.71–1.05)
Genotype frequency (%) 1/1
1/2
4 (0.3) 4 (0.5)
414 (35.2) 549 (46.7) 278 (32.7) 396 (46.6)
2/2
Dominant modelc P value
213 (18.1) 0.42 176 (20.7) 0.44
OR (95% CI)
The significance of bold values was 0.01. a Allele 1 represent the ancestral allele of each SNP. b P values were obtained by 2 -test. c For the dominant model, the 1/1 genotype was the referent (for rs3744700, the 1/1 + 1/2 genotype were referent). The associations were performed by multivariate logistic regression analysis adjusted with covariates (age, gender, BMI, smoking, hypertension, diabetes, TC, TG, HDL-C, and LDL-C); SNP, singe nucleotide polymorphism; HWE, Hardy–Weinberg equilibrium; 3 UTR, 3 untranslated region. OR, odds ratio; 95%CI, 95% confidence interval; other abbreviations as in Table 1. Table 3 Relation between the rs3744700 genotypes and the angiographic severity of CAD. No. of significantly diseased vessels
1 (n = 397) 2 (n = 373) 3 (n = 406)
rs3744700 genotypes
Allele frequencies
GG
GT
TT
P value
GT + TT
P value
G
T
P value
366 (92.2) 345 (92.5) 374 (92.1)
30 (7.5) 28 (7.5) 29 (7.2)
1 (0.3) 0 (0) 3 (0.7)
0.506 d.f. = 4
31 (7.8) 28 (7.5) 32 (7.9)
0.979 d.f. = 2
762 (96.0) 718 (96.2) 777 (95.7)
32 (4.0) 28 (3.8) 35 (4.3)
0.856 d.f. = 2
P values were obtained by 2 -test.
was no significant association between the genotype distributions of other four SNPs, including SNP rs2277680, and the number of angiographically documented significantly diseased vessels (data not shown).
CXCL16 level was significantly higher in subjects with GG genotype than those with T allele (GT and TT genotypes) in the control group (P = 0.015). Similarly, AMI patients with GG homozygote had higher plasma CXCL16 concentration than that those with T allele (P = 0.027).
3.3. Identification of regulatory elements in intron 4 3.5. SNP rs3744700 polymorphism and plasma lipid parameters To identify possible alternative splice acceptor sites within intron 4 of human CXCL16, the 439 bp genomic DNA sequence of intron 4 was analyzed using a splice site prediction program from the Berkeley Drosophila Genome Project (http://www.fruitfly.org/seq tools/splice.html). One alternative splice acceptor site demonstrating a confidence score 1.00 was identified within intron 2 of CXCL16, and rs3744700 locates just 61 bp upstream form it. Next, we performed a computer-based analysis to identify putative regulatory elements in intron 4 of the CXCL16 gene. We found that intron 4 harbors several putative regulatory elements, including the consensus sequence of GATA, which is affected by the rs3744700 polymorphism, indicating that the intronic variant may participate in the CXCL16 transcription. 3.4. Plasma CXCL16 levels in subset subjects Plasma CXCL16 levels were significantly higher in AMI patients than CAD-free controls (2.36 ± 0.40 ng/ml vs 1.88 ± 0.36 ng/ml, P < 0.01). However, there was no association between plasma CXCL16 levels and the severity of coronary atherosclerosis presented as the number of ≥50% stenotic vessels: 2.29 ± 0.38 ng/ml in 1-vessel disease, 2.37 ± 0.38 ng/ml in 2-vessel disease, and 2.42 ± 0.43 ng/ml in 3-vessel disease, respectively (Fig. 1). Next, we wanted to know whether plasma CXCL16 levels were related to genotype of rs3744700 (Table 4). Due to only one TT homozygote was seen in the CAD-free group and none in the AMI group, those rare homozygotes were pooled with heterozygotes when performing comparisons of CXCL16 plasma levels. We found that
Based on its potential to act as a scavenger receptor for oxLDL, CXCL16 may be involved in lipid metabolism. Because of the low occurrence of subjects with TT genotype, the T allele carriers (GT + TT) were ascribed to a single group. The lipid profiles of control subjects without lipid-lowering therapy (n = 771) with different rs3744700 genotypes are shown in Table I (available online supplementary material). No significant differences were seen between the rs3744700 genotypes and plasma lipid levels (TG, TC, LDL-C, and HDL-C). 4. Discussion In the present study, we have performed the first comprehensive analysis of five common SNPs in CXCL16 gene in 1176 CAD patients and 850 controls in Chinese Han population. Our results show that after adjustment for various risk factors, one SNP Table 4 Plasma CXCL16 levels in subgroup subjects according to rs3744700 genotype. Group
Plasma CXCL16 levels (ng/ml) GG
GT + TT
CAD-free (n = 160) AMI (n = 160)
1.92 ± 0.35 (n = 138) 2.38 ± 0.40 (n = 148)**
1.64 ± 0.40 (n = 22)* 2.11 ± 0.44 (n = 12)* , **
Results are means ± SD. * P < 0.05 compared with GG. ** P < 0.01 compared with GG CAD-free group.
164
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
Fig. 1. Plasma CXCL16 levels in patients with and without CAD. Compared with CAD-free subjects, AMI patients had significantly higher CXCL16 levels (P < 0.01). There were no differences in CXCL16 levels among AMI patients with 1-, 2- and 3vessel diseases. CAD (-): CAD-free, 1-VD: 1-vessel disease, 2-VD: 2-vessel disease, 3-VD: 3-vessel disease.
(rs3744700) exhibits significant association with CAD. The prevalence of the GG homozygotes is significantly higher among CAD patients than among control subjects. The GG genotypes also had increased plasma CXCL16 levels compared with the T allele carriers in both cases and control subjects. However, we found no evidence for an association of the other gene variations with CAD in Chinese Han population. CAD is a major cause of morbidity and mortality around the would, and inflammation is believed to play an important role in the pathogenesis of the condition [1]. The fundamental importance of chemokines for atherogenesis, progression, and destabilization of atherosclerotic plaques is now widely appreciated [2]. To date, only one study has investigated the association between polymorphism in CXCL16 and CAD risk [18]. They found that, for rs2277680, a nonsynonymous coding SNP on exon 4, there was no difference in the frequency of allele and genotypes between 387 patients who survived the first MI and 387 sex- and age-matched controls. However, carriers of the rare 181Val allele had a higher degree of angiographically determined coronary artery stenosis and smaller luminal diameter before PCI, suggesting this SNP was marginally associated with the severity of coronary artery stenosis in CAD patients. Our results support the finding that SNP rs2277680 is not associated with CAD risk. However, based on the number of luminal narrowing ≥50% in the three major coronary arteries [19], we did not observe any significant association between all five tagSNPs investigated in our study, including rs2277680, and the angiographic severity of CAD. The difference between the Sweden study and ours could be due to the different methods used in evaluating angiographic severity of CAD. They quantitatively analyzed the severity of coronary lesions using the QCA-CMS system [18]. Besides, the different genetic background of the two populations, as well as the relatively small sample size in the Sweden study may also contribute to the different conclusions. The rs3744700 resides in intron 4 of the CXCL16 gene and has not beet investigated before. The observed association with the intronic SNP in this study would be explained by either rs3744700 being directly pathogenic or it being in linkage disequilibrium with another causal variant. SNP rs3744700, though intronic, could itself be causal. It is plausible that disease-associated variants with modest effects will
be distributed proportionately between coding and noncoding sequences of the genome [23]. An increasing amount of evidence indicates that intronic genetic variations can have deleterious implications for gene splicing [24]. Indeed, recent evidence from many laboratories indicates that intronic variants can be functional and as a result are often the cause of human diseases [25–27]. Therefore, the disease-associated rs3744700 in CXCL16 may be functional in terms of splicing regulation or alternatively they may monitor functional elements within the gene associated with CAD risk. Using online programs, we found that rs3744700 locates on 61 bp from the splice acceptor site, and identified the consensus sequence of GATA that is affected by the rs3744700 polymorphism. Furthermore, we found that the plasma CXCL16 levels were significant higher in AMI patients than CAD-free controls, supporting the concept that CAD patients are characterized by higher plasma levels of CXCL16 [14,17]. The GG homozygotes also had higher plasma CXCL16 levels compared with T allele carriers in both cases and control subjects. Results of our computer-based analysis and CXCL16 measurement strongly suggest that rs3744700 may be functional. We speculate that the association of this SNP with CXCL16 plasma levels and CAD risk has arisen from a direct effect of this SNP on CXCL16 expression or mRNA splicing. Additional functional analyses of this SNP are needed to elucidate the underlying mechanism. We cannot exclude the possibility that other common CXCL16 variants may account for the increased risk of CAD we observed. However, the data generated by the HapMap Project for Chinese Han population suggests that this is unlikely. There are nine SNPs with MAF ≥5% in the CXCL16 gene region, including 3 kb upstream to 1 kb downstream of the gene. The rs3744700 is in low pair-wise LD based on r2 values with other eight SNPs. However, the present study did not include SNPs with MAF <5%, and the MAF distribution of SNPs from the International HapMap Project shows that more than 40% of SNPs have MAF <5%. The rare SNPs hypothesis suggests that targeting rare SNPs in large case–control association studies has more power to detect causal SNPs than does targeting common SNPs [28]. Thus, we cannot exclude the possibility that the rs3744700 SNP might be a functionally neutral marker that is in linkage disequilibrium with a rare functional polymorphism located elsewhere at the CXCL16 gene. Recently, Pollin et al. [29] found that associated SNP can be in linkage disequilibrium with causal variant more than 800 kb away. Recent genome-wide association studies (GWAS) of CAD have focused on a few chromosomal regions with strong signals [30–32]. It is also possible that rs3744700 might be linked with a mutation of another gene adjacent to CXCL16 gene, or with the true causative locus far away from the CXCL16 gene, such as one of the variants identified from reported GWAS data. Biologically, atherosclerosis is a continuous process characterized by the activation of a variety of factors including inflammation. So we conducted a stratified analysis by the severity of CAD. Our data show that neither the plasma CXCL16 levels nor the genotype of the five SNPs, including rs3744700, are associated with the severity of CAD. We hypothesize that CXCL16 is not involved in the late stage of multivessel-disease development but instead is involved directly in the early stage of CAD development. The present study has several strengths and limitations that should be addressed briefly. The strengths include the use of a pure Chinese Han population from the northeast region of China to eliminate false positive results due to population stratification. Moreover, the large sample size of the present work, which is essential for the investigation of the mostly modest influence of SNPs on clinical phenotypes, deserves to be mentioned. Finally angiographical profiles of all subjects were evaluated. Regarding limitations, it should be mentioned that different geographical and racial backgrounds of the individuals can affect the consequences
M. Huang et al. / Atherosclerosis 210 (2010) 160–165
of an association study. Therefore our results need to be confirmed in other populations. In summary, our study shows a strong association between CXCL16 and CAD in Chinese Han population. Future research is needed to confirm the present findings, by replicating this association in independent populations of various ethnic origins, especially in large-scale prospective cohort studies. Furthermore, additional research will be required to better define the precise effects of the intronic variant of CXCL16 gene. Conflict of interest There are no conflicts of interest. Acknowledgements The work was supported in part by the Military Medical Science and Technique Foundation during the 11th Five-Year Plan Period (No. 06G021). The authors thank Dr. Lianming Liao for his reading and commenting of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2009.11.004. References [1] Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med 2005;352:1685–95. [2] Zernecke A, Shagdarsuren E, Weber C. Chemokines in atherosclerosis: an update. Arterioscler Thromb Vasc Biol 2008;28:1897–908. [3] Moore KJ, Freeman MW. Scavenger receptors in atherosclerosis: beyond lipid uptake. Arterioscler Thromb Vasc Biol 2006;26:1702–11. [4] Matloubian M, David A, Engel S, Ryan JE, Cyster JG. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat Immunol 2000;1:298–304. [5] Shimaoka T, Kume N, Minami M, et al. Molecular cloning of a novel scavenger receptor for oxidized low density lipoprotein, SR-PSOX, on macrophages. J Biol Chem 2000;275:40663–6. [6] Shimaoka T, Nakayama T, Kume N, et al. Cutting edge: SR-PSOX/CXC chemokine ligand 16 mediates bacterial phagocytosis by APCs through its chemokine domain. J Immunol 2003;171:1647–51. [7] Gough PJ, Garton KJ, Wille PT, et al. A disintegrin and metalloproteinase 10mediated cleavage and shedding regulates the cell surface expression of CXC chemokine ligand 16. J Immunol 2004;172:3678–85. [8] Abel S, Hundhausen C, Mentlein R, et al. The transmembrane CXCchemokine ligand 16 is induced by IFN-gamma and TNF-alpha and shed by the activity of the disintegrin-like metalloproteinase ADAM10. J Immunol 2004;172:6362–72. [9] Minami M, Kume N, Shimaoka T, et al. Expression of SR-PSOX, a novel cellsurface scavenger receptor for phosphatidylserine and oxidized LDL in human atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2001;21:1796–800. [10] Wuttge DM, Zhou X, Sheikine Y, et al. CXCL16/SR-PSOX is an interferon-gammaregulated chemokine and scavenger receptor expressed in atherosclerotic lesions. Arterioscler Thromb Vasc Biol 2004;24:750–5.
165
[11] Wagsater D, Olofsson PS, Norgren L, Stenberg B, Sirsjo A. The chemokine and scavenger receptor CXCL16/SR-PSOX is expressed in human vascular smooth muscle cells and is induced by interferon gamma. Biochem Biophys Res Commun 2004;325:1187–93. [12] Aslanian AM, Charo IF. Targeted disruption of the scavenger receptor and chemokine CXCL16 accelerates atherosclerosis. Circulation 2006;114:583– 90. [13] Tenger C, Sundborger A, Jawien J, Zhou X. IL-18 accelerates atherosclerosis accompanied by elevation of IFN-gamma and CXCL16 expression independently of T cells. Arterioscler Thromb Vasc Biol 2005;25:791–6. [14] Smith C, Halvorsen B, Otterdal K, et al. High levels and inflammatory effects of soluble CXC ligand 16 (CXCL16) in coronary artery disease: down-regulatory effects of statins. Cardiovasc Res 2008;79:195–203. [15] Galkina E, Harry BL, Ludwig A, et al. CXCR6 promotes atherosclerosis by supporting T-cell homing, interferon-gamma production, and macrophage accumulation in the aortic wall. Circulation 2007;116:1801– 11. [16] Sheikine Y, Bang CS, Nilsson L, et al. Decreased plasma CXCL16/SR-PSOX concentration is associated with coronary artery disease. Atherosclerosis 2006;188:462–6. [17] Lehrke M, Millington SC, Lefterova M, et al. CXCL16 is a marker of inflammation, atherosclerosis, and acute coronary syndromes in humans. J Am Coll Cardiol 2007;49:442–9. [18] Lundberg GA, Kellin A, Samnegard A, et al. Severity of coronary artery stenosis is associated with a polymorphism in the CXCL16/SR-PSOX gene. J Intern Med 2005;257:415–22. [19] Chen Q, Reis SE, Kammerer CM, et al. Association between the severity of angiographic coronary artery disease and paraoxonase gene polymorphisms in the National Heart, Lung, and Blood Institute-sponsored Women’s Ischemia Syndrome Evaluation (WISE) study. Am J Hum Genet 2003;72:13–22. [20] Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–5. [21] Ke X, Cardon LR. Efficient selective screening of haplotype tag SNPs. Bioinformatics 2003;19:287–8. [22] Pei F, Han Y, Zhang X, et al. Association of interleukin-18 gene promoter polymorphisms with risk of acute myocardial infarction in northern Chinese Han population. Clin Chem Lab Med 2009;47:523–9. [23] Carlson CS, Eberle MA, Kruglyak L, Nickerson DA. Mapping complex disease loci in whole-genome association studies. Nature 2004;429:446–52. [24] Pagani F, Baralle FE. Genomic variants in exons and introns: identifying the splicing spoilers. Nat Rev Genet 2004;5:389–96. [25] Bouatia-Naji N, Rocheleau G, Van Lommel L, et al. A polymorphism within the G6PC2 gene is associated with fasting plasma glucose levels. Science 2008;320:1085–8. [26] Choi JW, Park CS, Hwang M, et al. A common intronic variant of CXCR3 is functionally associated with gene expression levels and the polymorphic immune cell responses to stimuli. J Allergy Clin Immunol 2008;122(6):1119– 26.e7. [27] Chen Q, Razzaghi H, Demirci FY, Kamboh MI. Functional significance of lipoprotein lipase HindIII polymorphism associated with the risk of coronary artery disease. Atherosclerosis 2008;200:102–8. [28] Gorlov IP, Gorlova OY, Sunyaev SR, Spitz MR, Amos CI. Shifting paradigm of association studies: value of rare single-nucleotide polymorphisms. Am J Hum Genet 2008;82:100–12. [29] Pollin TI, Damcott CM, Shen H, et al. A null mutation in human APOC3 confers a favorable plasma lipid profile and apparent cardioprotection. Science 2008;322:1702–5. [30] McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science 2007;316:1488–91. [31] 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. [32] 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.