Skip to content

Advertisement

  • Research
  • Open Access

The rs5888 single nucleotide polymorphism in scavenger receptor class B type 1 (SCARB1) gene and the risk of premature coronary artery disease: a case-control study

Lipids in Health and Disease201615:7

https://doi.org/10.1186/s12944-016-0176-9

Received: 23 November 2015

Accepted: 4 January 2016

Published: 12 January 2016

Abstract

Background

Several single nucleotide polymorphisms (SNPs) in lipid transport genes have been shown to be associated with premature coronary artery disease (PCAD). The scavenger receptor BI (SCARB1) is a key component of the reverse cholesterol transport and lipid metabolism. We aimed to examine the relationship between the rs5888 SNP within SCARB1and the risk of angiographically determined PCAD.

Methods

We used an age cut-off of 55 years for women and 45 years for men to define PCAD. Five-hundred and five patients with newly diagnosed angiographically documented PCAD (≥50 % luminal stenosis of any coronary vessel) as case group compared with 546 controls (subjects with no luminal stenosis at coronary arteries). The severity of CAD was determined by vessel score as well as Gensini score. A real-time polymerase chain reaction (PCR) and High Resolution Melting (HRM) analysis was used to distinguish between genotypes.

Results

T allele as compared to C allele was associated with increased odds of PCAD in total population (adjusted OR = 1.3, 95 % CI = 1.0 to 1.5; p = 0.020), and in women (adjusted OR = 1.3, 95 % CI = 1.0 to 1.8; p = 0.037), but not in men (adjusted OR = 1.2, 95 % CI = 0.9 to 1.5; p = 0.311). There was also no significant association between the examined polymorphism and the severity of CAD in whole or in men or women subgroups.

Conclusions

Our findings suggest that the SNP (rs5888) within SCARB1 is independently associated with PCAD in a sex-dependent manner.

Keywords

  • Coronary artery disease
  • Premature atherosclerosis
  • Coronary angiography
  • Scavenger receptor class B type I
  • rs5888 polymorphism

Background

The occurrence of coronary artery disease (CAD) in younger subjects, although infrequent, causes devastating effect on the more active lifestyle of this group of patients and may result in premature death [1]. Patients with premature CAD (PCAD), defined as the presence of coronary artery atherosclerotic lesion in men ≤45 year and women ≤55 years, have different risk factor profiles, clinical presentations, and prognoses than older patients [2]. Even though the identification of risk factors for PCAD is of essential importance for the understanding of its origins and for establishing preventive strategies, the mechanisms and predictors of PCAD are still not completely understood [3]. Risk factors for CAD are both environmental and genetic. In patients with PCAD, the role of genetic risk factors is expected to be more important than that of environmental factors [4]. Over the past recent years, several single nucleotide polymorphisms (SNPs) in lipid transport genes have been shown to be associated with PCAD [59].

Scavenger receptor class B type1 (SCARB1, also called SR-BI) is a multiligand cell surface receptor expresses both in macrophages and in the liver, indicating a major role for in the clearance of excess cholesterol from the body [10]. This membrane protein facilitates the uptake of cholesteryl esters from high-density lipoprotein cholesterol (HDL-C) and drives cholesterol from tissues to the liver in the various stages of reverse cholesterol transport pathway. This makes SCARB1 gene as an attractive candidate gene for CAD. The SCARB1 gene is located on 12q24.31 encompassing 13 exons and spans approximately 75 kb. Various SCARB1 polymorphisms in humans have been shown to be associated with altered serum lipid profile but their influence on CAD development or severity is still unclear [11]. One exonic SNP (rs5888) within SCARB1, has been linked to lower SR-BI protein expression and function [12]. This SNP is a “C” to “T” substitution at cDNA position 1050 base position on exon 8, and a coding-synonymous polymorphism (A350A) which was initially found by Acton et al. [13]. Several studies on rs5888 polymorphism and the risk of CAD have been previously published with contradictory results [1418]. There is even controversy about which allele confers increased risk. Wu et al. [14] found that the individuals with TT genotype were associated with increased CAD risk, whereas others reported that the T allele or the T allele carrier was associated with decreased risk of CAD [1518]. Moreover, they have investigated this association amongst general CAD populations and there is no study, that we are aware of, studying this association in PCAD patients. Thus, the aim of the present study was to examine the relation of rs5888 SNP with the presence and severity of angiographically determined PCAD in an Iranian coronary population.

Methods

Study participants and coronary angiograms

A prior study has used an age cut-off of 55 years for women and 45 years for men to define “young” patients with CAD [2]. To define PCAD, the same age definition was used in this study. This age- and sex-matched case–control study was conducted on 560 patients with newly diagnosed angiographically documented PCAD and an equal number of controls with normal coronary arteries. Cases were selected consecutively from those who, between August 2009 and May 2011, underwent coronary angiography at the catheterization laboratory of our center because of symptoms related to CAD or the results of non-invasive tests. Unrelated matched controls were selected from a pool of 1547 young patients (419 men and 1128 women) who, between June 2004 and July 2011, consecutively admitted for elective coronary angiography at our hospital. Cases and controls were matched for age within two years. Exclusion criteria included previous history of acute myocardial infarction (MI), stent implantation, cardiopulmonary resuscitation (CPR), and coronary artery bypass graft surgery. Also, patients on lipid-lowering drugs and those with minimal CAD (coronary lesions with less than 50 % luminal stenosis) were excluded. The study protocol was approved by the local ethical committee. Written informed consent was obtained from all patients in agreement with the Declaration of Helsinki for research involving human subjects explicitly provided permission for DNA analyses and gathering the relevant clinical data.

Coronary angiographies were performed by the percutaneous femoral approach using standard techniques. The presence and severity of CAD was determined by clinical vessel score. All cases showed the evidence of atherosclerosis i.e., ≥ 50 % luminal stenosis in at least one coronary artery or major branch segment in their epicardial coronary tree. Controls had no luminal stenosis at coronary angiograph. The severity of CAD was also determined by a semi-quantitative scoring system (Gensini score) which has been previously described [19]. Two cardiologists, blind to the study results, interpreted all angiograms. Definitions for analyzed risk factors of CAD including, body mass index (BMI), dyslipidemia, hypertension, cigarette smoking, opium addiction, and diabetes have been reported elsewhere [20, 21].

Sample collection and DNA extraction

From June 2004 to July 2011, for the purpose of creating a “DNA-bank of patients with PCAD and controls”, all young patients (age ≤45 years for men, and age ≤55 years for women at the disease onset) admitted for elective coronary angiography at our center were asked to provide a sample of whole blood for DNA extraction Peripheral venous blood samples were collected from an antecubital vein after 12 h overnight fasting of participants. Upon arrival at the laboratory, each participant provided about 15-mL venous blood sample. 5 mL blood was collected in plain tubes and used for biochemical assays immediately; the rest of the sample (10 ml) placed in ethylenediamine tetraacetic acid (EDTA)-containing tubes and stored deep-frozen until later use. Genomic DNA was extracted from leukocytes using the buffy coat of the stored EDTA whole blood samples. DNA extraction carried out using the standard ‘salting out’ method. DNA quantity was evaluated by calculating absorbance at λ = 260 nm, and the quality was assessed by a ratio of λ = 260/280 nm being close to 1.8. The purified DNA was stored in Tris-EDTA buffer (pH 8.0) at –70 °C until further analysis.

PCR amplification and genotype analysis

Real-time polymerase chain reaction (PCR) and High-Resolution Melting (HRM) analysis was applied for the genotype analysis using Corbett Rotor-Gene 6000 real-time rotary analyzer (Corbett Life Science Pty. Ltd., Mortlake, NSW, Australia). One set of primers based on common HRM specifications was designed using Beacon Designer software version 7.0 (Premier Biosoft International, Palo Alto, CA, USA) and synthesized by Bioneer Inc. (Daejeon, Korea). The forward primer was 5′- CTTGTTTCTCTCCCATCCTCA -3′ and the reverse was 5′- GAGTGTGCCTCCTGGTTAG -3′. The final reaction mixture contained Type-it HRM Master Mix (QIAGEN NV, Venlo, Netherlands) with intercalating DNA-binding Evagreen dye (Biotium, Inc. Hayward, CA, USA), DNA polymerase (5U/ml), primer mix (0.4 μM of each primer), genomic DNA (50 ng), and RNase Free Water (QIAGEN NV) in a total volume of 20 μl. The PCR cycling conditions consisted of an initial denaturation at 95 °C for 15 min, followed by 37 cycles of denaturation (96 °C for 20 s), primer annealing (62 °C for 20 s), and extension (72 °C for 15 s). One positive control for each genotype (CC, CT, and TT) and one appropriate negative control were included in each run. The positive controls were verified by using the PCR with restriction fragment length polymorphism (RFLP) technique as described previously with minor modification [22]. The genotypes were separated by electrophoresis on 2 % agarose gel stained with ethidium bromide (Fig. 1) and also confirmed by direct DNA sequence analysis.
Figure 1
Fig. 1

Agarose gel electrophoresis for the rs5888 polymorphism of scavenger receptor class B type 1 (SCARB1) gene. The 218-bp PCR product of SCARB1 rs5888 (C > T) was digested by Hin1I (BsaHI) restriction enzyme. The C allele codes for the presence of the Hin1I site and the T allele codes its absence. The wild-type CC variant produced two fragments with 187 bp and 31 bp while heterozygote CT produced 3 fragments of 218 bp, 187 bp, and 31 bp. The mutated homozygous variant TT produced one fragment of 218 bp. The 31 bp fragment was not visible in the gel due to its fast migration speed. M, 50 bp marker ladder; Lanes 6 and 10, CC genotype; Lanes 2, 5, 7, 8, 9, and 12, CT genotype; Lanes 1, 3, 4, and 11, TT genotype

Following the PCR amplification steps, melt curves for the products were generated by heating in 0.1 °C increments at a rate of 2 s per each step over the temperature range from 65 to 95 °C. The HRM data were analyzed using the Rotor-Gene Q software package supplied with the instrument. Sequence variations were distinguished from wild-type samples by the different shapes of normalized and temperature-shifted melting curves. In samples which failed to give an interpretable HRM pattern (10 %), the PCR-RFLP method was used for genotyping and reconfirmed by direct sequencing in 15 samples.

Statistical analysis

Continuous variables are reported as “mean ± standard deviation” and categorical variables as “numbers and percentage”. Variables were tested for normality by visually inspection of box-plots and by using Kolmogorov–Smirnov normality test. Gensini score, which was skewed, presented as median and inter-quartile range (25th to 75th percentiles). The case and control groups were compared using the independent two-sample Student t-test (or Mann-Whitney U-test if required) for the continuous variables and the chi-square test (or the Fisher exact test, as appropriate) for the categorical variables. Due to the highly skewed distribution of Gensini scores, to compare Gensini score among rs5888 genotypes, the Kruskal-Wallis analysis was chosen and subjects with normal coronary arteries (Gensini score = 0, n = 546) were excluded to ensure any observed effect on graded severity was not being driven by those without any CAD in whom risk allele frequency is expected to be significantly lower.

Logistic regression model was constructed to test the independent relationship between the rs5888 variants and the presence of CAD with the SNP coded as 0, 1, or 2 based on the number of risk (T) alleles. Analyses were repeated after adjusting for diabetes mellitus, cigarette smoking, opium addiction, hypertension, and family history of CAD as well as serum HDL-C, triglyceride, and LDL-C in an established multivariable model. Odds ratio (OR) and 95 % confidence intervals (CI) were calculated in all patients and in males and females separately. Moreover, a linear multiple regression analysis with the presence of CAD severity (Gensini score as a continuous variable) as the dependent variable, was performed. The following were tested as potential risk factors: diabetes mellitus, cigarette smoking, opium addiction, hypertension, and family history of CAD as well as BMI and serum levels of HDL-C, triglyceride, and LDL-C. Non-normal variables, such as Gensini scores, were natural log-transformed before entering the model. Additionally, a multiple ordinal logistic regression model, with number of diseased vessels as the dependent variable, was established with adjusting for aforementioned potential risk factors. A generalized linear model (GLM) was used to analyze the association between the genotypes and lipid profile adjusted for age, sex, BMI, and cigarette smoking. A p value ≤ 0.05 was considered statistically significant. All Statistical analyses were done by PASW Statistics for Windows, Version 18.0. (Chicago: SPSS Inc.).

Results

Sixty-nine patients (55 cases and 14 controls) were excluded due to missing data or genotyping failure, resulting in 1051 patients entered in the statistical analyses including 505 cases, and 546 controls. The mean age of the study participants was 46 ± 6 years and 45.4 % were men. Baseline characteristics of the study sample stratified by gender and CAD status are presented in Table 1 and 2, respectively. The overall prevalence of cardiovascular risk factors including, hypertension, diabetes mellitus, cigarette smoking, dyslipidemia, and family history of CAD were high. The women had a greater prevalence of hypertension, diabetes mellitus, and dyslipidemia as compared to men but cigarette smoking and opium addiction were much more frequent in men than in women. There was no significant difference in the prevalence of family history of CAD between men and women (Table 1). As seen in Table 2, the prevalence of diabetes, family history of CAD, hyperlipidemia, and cigarette smoking as well as hypertension and opium addiction were significantly higher in CAD group whereas there was no statistically significant difference in age, sex, and BMI between two groups.
Table 1

Baseline characteristics of the study population stratified by sex

 

All (n = 1051)

Men (n = 477)

Women (n = 574)

P-value*

Age (years)

45.6 ± 6.0

41.0 ± 3.5

49.4 ± 4.9

<0.001

Body mass index (kg/m2)

29.7 ± 5.7

28.1 ± 4.4

31.1 ± 5.3

<0.001

Cigarette smoking

   

<0.001

 Current smoker

213 (20.3)

188 (39.4)

25 (4.4)

 

 Ex-smoker

97 (9.2)

82 (17.2)

15 (2.6)

 

 Non-smoker

741 (70.5)

207 (43.4)

534 (93.0)

 

Opium addiction

99 (9.4)

91 (19.1)

8 (1.4)

<0.001

Hypertension

456 (44.3)

126 (26.4)

339 (59.1)

<0.001

Diabetes

240 (22.8)

53 (11.1)

187 (32.6)

<0.001

Dyslipidemia

673 (64.0)

290 (60.8)

383 (66.7)

0.046

Family history of CAD

357 (34.0)

157 (32.9)

200 (34.8)

0.511

Fasting blood glucose (mg/dl)

116.7 ± 50.1

106.8 ± 41.1

124.8 ± 52.2

<0.001

Triglyceride (mg/dl)

182.8 ± 116.4

193.3 ± 123.1

174.1 ± 109.9

0.008

HDL-cholesterol (mg/dl)

42.2 ± 11.7

38.5 ± 10.5

45.2 ± 11.8

<0.001

LDL-cholesterol (mg/dl)

117.0 ± 40.3

113.4 ± 37.1

119.9 ± 42.5

0.011

Total cholesterol (mg/dl)

186.8 ± 48.6

182.2 ± 47.7

190.6 ± 49.5

0.005

Creatinine (mg/dl)

0.9 ± 0.6

1.0 ± 0.8

0.8 ± 0.4

<0.001

LVEF (%)

53.8 ± 9.0

52.7 ± 9.7

54.7 ± 8.3

<0.001

Data are presented as mean ± SD or n (%)

CAD Coronary artery disease, LVEF left ventricular ejection fraction

* P-values for men vs. women

Table 2

Baseline characteristics in the study population based on coronary artery disease presence

Variables

non-CAD (n = 546)

CAD (n = 505)

P value

Age (years)

45.6 ± 6.1

45.6 ± 5.9

0.922

Body mass index (Kg/m2)

29.7 ± 4.9

29.7 ± 5.4

0.996

Male sex

243 (44.5)

234 (46.3)

0.551

Diabetes mellitus

74 (13.6)

166 (32.9)

<0.001

Hypertension

192 (35.2)

273 (54.1)

<0.001

Smoking status

  

0.001

 Current smoker

87 (15.9)

126 (25.0)

 

 Ex-smoker

48 (8.8)

49 (9.7)

 

 Non-smoker

411 (75.3)

330 (65.3)

 

Family history of CAD

148 (27.1)

209 (41.4)

<0.001

Dyslipidemia

315 (57.7)

342 (72.6)

<0.001

Triglyceride (mg/dl)

173.4 ± 114.3

193.0 ± 118.0

0.007

HDL-cholesterol (mg/dl)

43.9 ± 11.5

40.2 ± 11.7

<0.001

LDL-cholesterol (mg/dl)

116.4 ± 36.3

117.6 ± 44.2

0.945

Total cholesterol (mg/dl)

186.9 ± 42.8

186.8 ± 54.3

0.996

Fasting blood glucose (mg/dl)

106.1 ± 36.7

128.3 ± 60.2

<0.001

Creatinine (mg/dl)

0.9 ± 0.6

0.9 ± 0.6

0.981

LVEF (%)

55.7 ± 8.0

51.7 ± 9.6

<0.001

Data are presented as mean ± SD or n (%)

CAD Coronary artery disease, LVEF left ventricular ejection fraction

Allele and genotype frequencies for the SCARB1 (rs5888) polymorphism in the case and control groups, separated by sex, are shown in Table 3. There was no statistically significant difference in the distribution of rs5888 genotypes between two groups and in men and women subgroups after stratification by sex. Table 4 shows that no associations were detected between SCARB1 rs5888 (C > T) genotypes and lipid profile including total cholesterol, HDL-C, LDL-C, and triglyceride levels. Even after analyzing data separately for men or women, the rs5888 SNP in SCARB1 gene was not associated with lipid profile in the subgroup of men or women (data not shown).
Table 3

Genotype and allele frequencies of rs5888 polymorphism in the SCARB1 gene according to CAD status in whole study group and in subgroups separated by gender

 

non- CAD patients

CAD patients

P-value*

 

All (n = 546)

Male (n = 243)

Female (n = 303)

All (n = 505)

Male (n = 234)

Female (n = 271)

All

Male

Female

rs5888 (C > T)

      

0.163

0.781

0.143

 CC

195 (35.7)

88 (36.2)

107 (35.3)

162 (32.1)

81 (34.6)

81 (29.9)

   

 CT

281 (51.5)

121 (49.8)

160 (52.8)

259 (51.3)

115 (49.1)

144 (53.1)

   

 TT

70 (12.8)

34 (14.0)

36 (11.9)

84 (16.6)

38 (16.2)

46 (17.0)

   

 CT + TT

351 (64.3)

155 (63.8)

196 (64.7)

343 (67.9)

153 (65.4)

190 (70.9)

0.214

0.715

0.167

Alleles

      

0.082

0.544

0.070

 C

671 (61.4)

297 (61.1)

374 (61.7)

583 (57.7)

277 (59.2)

306 (56.5)

   

 T

421 (38.6)

189 (38.9)

232 (38.3)

427 (42.3)

191 (40.8)

236 (43.5)

   

* p-values for non-CAD vs. CAD

Data are presented as n (%)

CAD, Coronary artery disease; SCARB1, scavenger receptor class B type 1

Table 4

Lipid profile in total population, and in case and control groups according to rs5888 (C > T) genotypes in SCARB1 gene

 

CC

CT

TT

P value*

All participants

(n = 357)

(n = 540)

(n = 154)

 Triglyceride (mg/dl)

183.6 ± 108.7

179.6 ± 113.7

192.5 ± 140.5

0.572

 HDL cholesterol (mg/dl)

41.8 ± 11.6

42.6 ± 12.1

41.4 ± 10.4

0.569

 LDL cholesterol (mg/dl)

119.5 ± 40.2

115.9 ± 40.3

114.9 ± 40.4

0.351

 Total cholesterol (mg/dl)

188.5 ± 51.1

186.2 ± 47.3

185.1 ± 47.2

0.691

Non-CAD group

(n = 195)

(n = 281)

(n = 70)

 Triglyceride (mg/dl)

175.7 ± 113.2

166.0 ± 99.0

196.4 ± 162.6

0.065

 HDL cholesterol (mg/dl)

43.5 ± 12.6

44.3 ± 11.0

43.8 ± 9.5

0.853

 LDL cholesterol (mg/dl)

120.6 ± 36.2

113.0 ± 34.6

118.2 ± 41.6

0.054

 Total cholesterol (mg/dl)

189.4 ± 44.0

183.8 ± 40.7

191.5 ± 46.6

0.108

CAD group

(n = 162)

(n = 259)

(n = 84)

 Triglyceride (mg/dl)

193.3 ± 102.5

194.2 ± 126.4

189.1 ± 119.4

0.887

 HDL cholesterol (mg/dl)

39.7 ± 10.0

40.8 ± 12. 9

39.3 ± 10.7

0.499

 LDL cholesterol (mg/dl)

118.2 ± 44.5

119.0 ± 45.4

112.0 ± 39.4

0.470

 Total cholesterol (mg/dl)

189.4 ± 58.9

188.8 ± 53.4

179.6 ± 47.5

0.350

*Adjusted for age, sex, cigarette smoking, and body mass index (BMI)

Data are presented as mean ± SD

CAD, Coronary artery disease; SCARB1, scavenger receptor class B type 1

After applying a multiple logistic regression model with adjusting for diabetes, hypertension, cigarette smoking, opium addiction, BMI, and family history of CAD as well as triglyceride, HDL-C, and LDL-C, rs5888 polymorphism was found to be associated with the presence of PCAD in total population and in women but not in men (Table 5). T allele as compared to C allele was associated with increased odds of PCAD in total population (adjusted OR = 1.3, 95 % CI = 1.0 to 1.5; p = 0.020), and in women (adjusted OR = 1.3, 95 % CI = 1.0 to 1.8; p = 0.037), but not in men (adjusted OR = 1.2, 95 % CI = 0.9 to 1.5; p = 0.311). Among total population, those with TT genotype had 1.7-fold increased adjusted odds of PCAD compared to those with CC genotype (p = 0.013). When data were analyzed separately for men or women, SCARB1 (rs5888) TT genotype as compared to CC genotype was associated with increased odds of PCAD in women (adjusted OR = 2.1, 95 % CI = 1.2 to 4.0; p = 0.014), but not in men (adjusted OR = 1.3, 95 % CI = 0.7 to 2.5; p = 0.336).
Table 5

Unadjusted and adjusted odds ratios for the effect of rs5888 genotypes on premature coronary artery disease in whole and in men and women separately

 

Unadjusted

Adjusted*

rs5888 (C > T)

Odds ratio (95 % CI)

p-value

Odds ratio (95 % CI)

p-value

All participants

 

0.164

 

0.046

 CC

1.0 (Ref.)

-

1.0 (Ref.)

 CT

1.1 (0.9 to 1.5)

0.448

1.2 (0.9 to 1.6)

0.204

 TT

1.4 (1.0 to 2.1)

0.058

1.7 (1.1 to 2.6)

0.013

 CT + TT

1.2 (0.9 to 1.5)

0.214

1.3 (1.0 to 1.7)

0.077

 Allele T vs. C

1.2 (1.0 to 1.4)

0.082

1.3 (1.0 to 1.5)

0.020

Male sex

 

0.781

 

0.583

 CC

1.0 (Ref.)

1.0 (Ref.)

 CT

1.0 (0.7 to 1.5)

0.874

1.2 (0.8 to 1.8)

0.448

 TT

1.2 (0.7 to 2.1)

0.491

1.3 (0.7 to 2.5)

0.336

 CT + TT

1.1 (0.7 to 1.6)

0.715

1.2 (0.8 to 1.9)

0.348

 Allele T vs. C

1.1 (0.8 to 1.4)

0.544

1.2 (0.9 to 1.5)

0.311

Female sex

 

0.146

 

0.044

 CC

1.0 (Ref.)

1.0 (Ref.)

 CT

1.2 (0.8 to 1.7)

0.354

1.1 (0.7 to 1.8)

0.551

 TT

1.7 (1.0 to 2.9)

0.050

2.1 (1.2 to 4.0)

0.014

 CT + TT

1.3 (0.9 to 1.8)

0.167

1.3 (0.9 to 2.0)

0.202

 Allele T vs. C

1.2 (1.0 to 1.6)

0.070

1.3 (1.0 to 1.8)

0.037

*Adjusted for diabetes, hypertension, cigarette smoking, opium addiction, BMI, and family history of CAD as well as triglyceride, HDL-C, and LDL-C

CAD, coronary artery disease; BMI, body mass index; CI, confidence interval; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; Ref., the reference category

Table 6 shows the association between the studied polymorphism and the severity of CAD with respect to vessel score and Gensini score. The distribution of vessel score was statistically similar among different genotypes (adjusted p = 0.309). Additionally, in total population, the median and inter-quartile range for Gensini score was not significantly different among the CC (42, 20 to 65), CT (42, 23 to 75), and TT (37, 20 to 64) genotypes (adjusted P = 0.540). There was also no significant association between the examined polymorphism and the severity of CAD within men or women subgroups.
Table 6

The genotypes of rs5888 single nucleotide polymorphism in association with CAD severity

 

CC (n = 357)

CT (n = 540)

TT (n = 154)

P-value

P-value*

Vessel score, n (%)

   

0.178

0.247

 0

195 (54.6)

281 (52.0)

70 (45.3)

  

 1VD

72 (20.2)

117 (21.7)

46 (29.9)

  

 2VD

40 (11.2)

74 (13.7)

22 (14.3)

  

 3VD

50 (14.0)

68 (12.6)

16 (10.4)

  

Gensini score, median (25th to 75th percentiles), na.

 All

42 (20 to 65), 162

42 (23 to 75), 259

37 (20 to 64), 84

0.449

0.572

 Males

43 (26 to 77), 81

40 (23 to 67), 115

36 (25 to 61), 38

0.816

0.697

 Females

40 (20 to 64), 81

45 (23 to 78), 144

38 (18 to 66), 46

0.233

0.398

0 no coronary lesion, 1VD one-vessel disease, 2VD Two-vessel disease, 3VD three-vessel disease, CAD coronary artery disease, HDL-C high density lipoprotein cholesterol, LDL-C low density lipoprotein cholesterol

* P-values adjusted for diabetes mellitus, cigarette smoking, opium addiction, hypertension, and family history of CAD as well as body mass index, serum HDL-C, triglyceride, and LDL-C

aA total number of 505 patients after excluding 546 with normal coronary arteries (Gensini score = 0)

Discussion

The importance of SCARB1 in reverse cholesterol transport and atherosclerosis has been previously confirmed in mouse models [11]. Moreover, recent studies have demonstrated rs5888 variant within SCARB1 to be linked with lipid levels and the size of lipoprotein particles, especially in younger women [23, 24]. However, the role of this variant in CAD development among humans is much less investigated [25]. Here we report, for the first time to our knowledge, that the SNP rs5888 in the SCARB1 gene is possibly an independent risk factor for early onset CAD in a cohort of young Iranian patients undergoing elective coronary angiography. In this study we focused on young patients with CAD because we wanted to investigate a cohort of patients who less affected by the age-related confounding factors; this strategy is more likely to detect genetic effects. The logistic regression model demonstrated that T allele in comparison with C allele had 1.3-fold increased risk for PCAD in whole and in women subgroup. Also, in men, the T allele showed a 20 % higher risk for CAD compared to C allele but this did not reach a statistical significance.

Among our study population, rs5888 within SCARB1 gene is a common SNP with minor allele frequency around 40 %. The variant allele of SNP rs5888 (T allele) frequency was 0.403 in total; 0.423 in case group and 0.386 in reference group. In a cross-sectional health survey on Lithuanian population, T allele frequency in the reference and MI groups was 0.362 and 0.400, respectively [18], whereas in Korean subjects the T allelic frequency was 0.330 in control and 0.170 in CAD [16]. In white population, the prevalence of this variant varies between 0.400 and 0.600 [13, 23, 2628]. Among the participant of the Framingham Offspring Study, wherein almost all were Caucasian, the frequency of T allele was 0.486 [28], while in Chinese population T allele frequency varies from 0.217 to 0.290 [22, 29]. These studies suggest that the frequency of rs5888 mutation is different in various surveys with various sample sizes, in various geographical regions, and in various ethnic populations. Our finding that T allele may be a significant, independent risk factor for CAD is consistent with the results of a recent candidate gene association study on Chinese [14], but this is in contrast to the other similar studies [1518]. In a Tunisian population, Rejab et al. [17] reported that subjects carrying T allele were associated with an almost 41 % reduced risk of angiographically documented significant coronary stenosis. Several other studies on Korean, Spanish, and Lithuanian patients also suggested that the T allele was associated with lower CAD risk [15, 16, 18].

Such disparity in study results may be ascribed to methodological heterogeneity (differences in study designs, sample sizes, definition of the phenotype, and, age and gender), different genetic background, and differences in the nature of various populations. For example, the precise definition of phenotype in studies on CAD association is of essential importance because individuals with significant CAD who are clinically silent may be classified as controls leading to a higher likelihood of null results. To avoid this potential bias in the current study, we defined phenotype based on objective angiographic documentation of coronary artery status. Notably, the examined polymorphism (rs5888) is not among the list of 46 loci that previously described to be robustly associated with CAD in large-scale genome-wide association studies (GWAS) [30, 31]. In the largest GWAS meta-analysis of CAD undertaken to date, the Coronary ARtery DIsease Genome wide Replication and Meta-analysis (CARDIoGRAM) plus The Coronary Artery Disease (C4D) consortium (CARDIoGRAMplusC4D), involving European and South Asian descent, 15 new susceptibility loci were identified [31]. Four of the 30 CAD risk loci previously reported in individuals of European ancestry (CDKN2B-AS1, COL4A2, CXCL12 and APOE) were detected as independent additional SNPs not in LD (r2 < 0.5) with the lead SNP, which also reached genome-wide significance. The additional SNP in the APOE locus was rs445925 (P = 9.42 × 10−11; r2 = 0.015 with rs207560 in 1000 Genomes Project data) [31]. There was no trend for higher odds ratios (ORs) in any of the subgroup analyses; however, one new locus reached genome-wide significance in males and in young CAD cases (rs16986953; P = 1.89 × 10−8 and 1.67 × 10−8, respectively), which was located in a gene desert (with nearest transcript AK097927), 1.3 Mb away from the APOB gene [31].

Even though several studies having suggested association between the genotypes of rs5888 gene polymorphism and various clinical parameters, the causative role of this variant is still unknown. Indeed, the exon 8 variant (rs5888) is a common synonymous or silent SNP that does not encode for a change in amino acid sequence of the protein product. Although this silent polymorphism was conventionally believed to be a nonfunctional marker in linkage disequilibrium with an as-yet-unidentified causal variant [13, 15, 32], no linkage has been found between this SNP and a known functional promoter variant [15]. Also, the promoter variant appears to be extremely rare (0.020) in white populations [15]. On the other hand, recent accumulating evidence indicates that silent SNPs could directly influence expression and/or function of their gene products [3335]; using in vitro models, Constantineau et al. [12] demonstrated that synonymous rs5888 variant affected the RNA secondary structure and conformation of its gene product, SR-BI, and was significantly associated with lower SR-BI expression and its functionality. Hence, SCARB1 rs5888 T allele may possibly alter SR-BI expression, and in turn, influence risk for development of CAD. Moreover, another possible mechanism for this association may be related to some splicing regulatory role [36].

We found no associations between SCARB1 rs5888 (C > T) genotypes and lipid profile in our studied patients. In two other studies, the investigators also did not find such associations, when combined with other polymorphisms in the SCARB1 gene, the combined genotypes were related to triglyceride and HDL-C in the women [13, 37]. However, the association between SCARB1 gene and lipid profile is complex and still debated [18]. Another notable issue is that subgroup analysis of men and women demonstrated significant different frequencies of rs5888 variant of SCARB1 gene exclusively in women. Significant sex-dependent associations between SCARB1 genetic variations with lipid profile, BMI, and risk of cardiovascular disease have been previously shown [13, 17, 18, 27, 32]. As noted by Ritsch et al. [32], this phenomenon may be justified by a likely interaction between SCARB1 genotype and estrogen-dependent regulation of SR-BI expression and lipid parameters or by a possible effect of sex on factors affecting the function of SR-BI; factors such as cholesteryl gradient within the cells, HDL-C plasma concentration and composition of its particles, and triglyceride levels that influence HDL-C metabolism via cholesterol ester transfer protein.

There are several limitations in the present study that should be noted. First, because of case-control design, a selection bias may have been introduced. Second, the control group is not representative of the general healthy population, because it is enriched for people with a high clinical suspicion of having significant CAD, and although coronary stenosis is ruled out, cardiovascular risk factors can be assumed to be higher in this group as compared to the general population. Third, although we tested a relatively large number of patients with PCAD, our study was under-powered to detect the 1.2-fold effect size in male subjects with risk allele (T). Fourth, our results from a single-center may not be generalized to other ethnic groups; however, our tertiary referral center receives patients from all over the country with various ethnic backgrounds. Finally, the lack of ethnicity data in the present study is a major limitation that might explain some of the discrepancy with other studies particularly because the population was from various ethnic backgrounds; therefore, caution needs to be exercised in drawing conclusions about the impact of the examined SNP on PCAD.

Conclusion

Using angiographic data and thus refining the phenotype, we found that the SNP (rs5888) within SCARB1 is independently associated with PCAD in whole study population and in women, but no association was observed in men. However, the statistical power was not sufficient to detect an association in male subgroup. Further studies, particularly in those with PCAD, with larger sample size in sex subgroups are required to confirm such association in well-defined ethnic groups. Moreover, because many genes are involved in the development of CAD due to dyslipidemia, further investigations are needed to test the relationship between polymorphisms associated with lipid metabolism and the risk of CAD. In addition, the exact mechanism and biological basis of possible role of rs5888 variant in pathogenesis of CAD needs to be explained.

Declarations

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
Department of Cardiac Research, Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran
(2)
Department of Molecular Pathology, Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran
(3)
Department of Genetics, School of Biological Sciences, Tarbiat Modares University, Tehran, Iran
(4)
Department of Laboratory Medicine and Pathology, Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran

References

  1. Celik T, Iyisoy A. Premature coronary artery disease in young patients: an uncommon but growing entity. Int J Cardiol. 2010;144(1):131–2. doi:10.1016/j.ijcard.2008.12.150. Epub 2009 Jan 26.PubMedView ArticleGoogle Scholar
  2. De Sutter J, De Bacquer D, Kotseva K, Sans S, Pyörälä K, Wood D, et al. Screening of family members of patients with premature coronary heart disease Results from the EUROASPIRE II family survey. Eur Heart J. 2003;24(3):249–57.PubMedView ArticleGoogle Scholar
  3. Vaisi-Raygani A, Ghaneialvar H, Rahimi Z, Nomani H, Saidi M, Bahrehmand F, et al. The angiotensin converting enzyme D allele is an independent risk factor for early onset coronary artery disease. Clin Biochem. 2010;43(15):1189–94. doi:10.1016/j.clinbiochem.2010.07.010.PubMedView ArticleGoogle Scholar
  4. Cengel A, Tanindi A. Myocardial infarction in the young. J Postgrad Med. 2009;55(4):305–13. doi:10.4103/0022-3859.58944.PubMedView ArticleGoogle Scholar
  5. Kay A, Marz W, Hoffmann MM, Zhang Q, Masana L, Cavanna J, et al. Coronary artery disease and dyslipidemia within Europe: genetic variants in lipid transport gene loci in German subjects with premature coronary artery disease. Atheroscler Suppl. 2002;3(1):27–33.PubMedView ArticleGoogle Scholar
  6. Kumar P, Luthra K, Dwivedi M, Behl VK, Pandey RM, Misra A. Apolipoprotein E gene polymorphisms in patients with premature myocardial infarction: a case-controlled study in Asian Indians in North India. Ann Clin Biochem. 2003;40(Pt 4):382–7. doi:10.1258/000456303766477020.PubMedView ArticleGoogle Scholar
  7. Almeida KA, Strunz CM, Maranhao RC, Mansur AP. The S447X polymorphism of lipoprotein lipase: effect on the incidence of premature coronary disease and on plasma lipids. Arq Bras Cardiol. 2007;88(3):297–303.PubMedView ArticleGoogle Scholar
  8. Ji J, Herbison CE, Mamotte CD, Burke V, Taylor RR, van Bockxmeer FM. Hepatic lipase gene -514 C/T polymorphism and premature coronary heart disease. J Cardiovasc Risk. 2002;9(2):105–13.PubMedView ArticleGoogle Scholar
  9. Abd El-Aziz TA, Mohamed RH, Hagrass HA. Increased risk of premature coronary artery disease in Egyptians with ABCA1 (R219K), CETP (TaqIB), and LCAT (4886C/T) genes polymorphism. J Clin Lipidol. 2014;8(4):381–9. doi:10.1016/j.jacl.2014.06.001.PubMedView ArticleGoogle Scholar
  10. Van Eck M, Pennings M, Hoekstra M, Out R, Van Berkel TJ. Scavenger receptor BI and ATP-binding cassette transporter A1 in reverse cholesterol transport and atherosclerosis. Curr Opin Lipidol. 2005;16(3):307–15.PubMedView ArticleGoogle Scholar
  11. Mineo C, Shaul PW. Functions of scavenger receptor class B, type I in atherosclerosis. Curr Opin Lipidol. 2012;23(5):487–93. doi:10.1097/MOL.0b013e328357ba61.PubMedView ArticleGoogle Scholar
  12. Constantineau J, Greason E, West M, Filbin M, Kieft JS, Carletti MZ, et al. A synonymous variant in scavenger receptor, class B, type I gene is associated with lower SR-BI protein expression and function. Atherosclerosis. 2010;210(1):177–82. doi:10.1016/j.atherosclerosis.2009.11.029.PubMedPubMed CentralView ArticleGoogle Scholar
  13. Acton S, Osgood D, Donoghue M, Corella D, Pocovi M, Cenarro A, et al. Association of polymorphisms at the SR-BI gene locus with plasma lipid levels and body mass index in a white population. Arterioscler Thromb Vasc Biol. 1999;19(7):1734–43.PubMedView ArticleGoogle Scholar
  14. Wu DF, Yin RX, Cao XL, Chen WX, Aung LH, Wang W, et al. Scavenger receptor class B type 1 gene rs5888 single nucleotide polymorphism and the risk of coronary artery disease and ischemic stroke: a case-control study. Int J Med Sci. 2013;10(12):1771–7. doi:10.7150/ijms.7044.PubMedPubMed CentralView ArticleGoogle Scholar
  15. Rodriguez-Esparragon F, Rodriguez-Perez JC, Hernandez-Trujillo Y, Macias-Reyes A, Medina A, Caballero A, et al. Allelic variants of the human scavenger receptor class B type 1 and paraoxonase 1 on coronary heart disease: genotype-phenotype correlations. Arterioscler Thromb Vasc Biol. 2005;25(4):854–60. doi:10.1161/01.atv.0000157581.88838.03.PubMedView ArticleGoogle Scholar
  16. Hong SH, Kim YR, Yoon YM, Min WK, Chun SI, Kim JQ. Association between HaeIII polymorphism of scavenger receptor class B type I gene and plasma HDL-cholesterol concentration. Ann Clin Biochem. 2002;39(Pt 5):478–81. doi:10.1258/000456302320314485.PubMedView ArticleGoogle Scholar
  17. Rejeb J, Omezzine A, Boumaiza I, Rebhi L, Kacem S, Rejeb NB, et al. Association of three polymorphisms of scavenger receptor class BI gene (exon8, exon1, intron5) with coronary stenosis in a coronary Tunisian population. Gene. 2012;511(2):383–8. doi:10.1016/j.gene.2012.09.070.PubMedView ArticleGoogle Scholar
  18. Stanislovaitiene D, Lesauskaite V, Zaliuniene D, Smalinskiene A, Gustiene O, Zaliaduonyte-Peksiene D, et al. SCARB1 single nucleotide polymorphism (rs5888) is associated with serum lipid profile and myocardial infarction in an age- and gender-dependent manner. Lipids Health Dis. 2013;12:24. doi:10.1186/1476-511x-12-24.PubMedPubMed CentralView ArticleGoogle Scholar
  19. Gensini GG. A more meaningful scoring system for determining the severity of coronary heart disease. Am J Cardiol. 1983;51(3):606.PubMedView ArticleGoogle Scholar
  20. Boroumand M, Ghaedi M, Mohammadtaghvaei N, Pourgholi L, Anvari MS, Davoodi G, et al. Association of estrogen receptor alpha gene polymorphism with the presence of coronary artery disease documented by coronary angiography. Clin Biochem. 2009;42(9):835–9. doi:10.1016/j.clinbiochem.2009.01.005.PubMedView ArticleGoogle Scholar
  21. Boroumand M, Pourgholi L, Ziaee S, Anvari MS, Jalali A, Goodarzynejad H. The association between Factor V Leiden with the presence and severity of coronary artery disease. Clin Biochem. 2014;47(6):356–60.PubMedView ArticleGoogle Scholar
  22. Wu DF, Yin RX, Yan TT, Aung LH, Cao XL, Miao L, et al. The SCARB1 rs5888 SNP and serum lipid levels in the Guangxi Mulao and Han populations. Int J Med Sci. 2012;9(8):715–24. doi:10.7150/ijms.4815.PubMedPubMed CentralView ArticleGoogle Scholar
  23. Roberts CG, Shen H, Mitchell BD, Damcott CM, Shuldiner AR, Rodriguez A. Variants in scavenger receptor class B type I gene are associated with HDL cholesterol levels in younger women. Hum Hered. 2007;64(2):107–13. doi:10.1159/000101962.PubMedPubMed CentralView ArticleGoogle Scholar
  24. Richard E, von Muhlen D, Barrett-Connor E, Alcaraz J, Davis R, McCarthy JJ. Modification of the effects of estrogen therapy on HDL cholesterol levels by polymorphisms of the HDL-C receptor, SR-BI: the Rancho Bernardo Study. Atherosclerosis. 2005;180(2):255–62. doi:10.1016/j.atherosclerosis.2004.12.013.PubMedView ArticleGoogle Scholar
  25. Shen WJ, Hu J, Hu Z, Kraemer FB, Azhar S. Scavenger receptor class B type I (SR-BI): a versatile receptor with multiple functions and actions. Metabolism. 2014;63(7):875–86. doi:10.1016/j.metabol.2014.03.011.PubMedView ArticleGoogle Scholar
  26. Morabia A, Ross BM, Costanza MC, Cayanis E, Flaherty MS, Alvin GB, et al. Population-based study of SR-BI genetic variation and lipid profile. Atherosclerosis. 2004;175(1):159–68. doi:10.1016/j.atherosclerosis.2004.03.014.PubMedView ArticleGoogle Scholar
  27. McCarthy JJ, Lewitzky S, Reeves C, Permutt A, Glaser B, Groop LC, et al. Polymorphisms of the HDL receptor gene associated with HDL cholesterol levels in diabetic kindred from three populations. Hum Hered. 2003;55(4):163–70.PubMedView ArticleGoogle Scholar
  28. Osgood D, Corella D, Demissie S, Cupples LA, Wilson PW, Meigs JB, et al. Genetic variation at the scavenger receptor class B type I gene locus determines plasma lipoprotein concentrations and particle size and interacts with type 2 diabetes: the framingham study. J Clin Endocrinol Metab. 2003;88(6):2869–79.PubMedView ArticleGoogle Scholar
  29. Wu DF, Yin RX, Hu XJ, Aung LH, Cao XL, Miao L, et al. Association of rs5888 SNP in the scavenger receptor class B type 1 gene and serum lipid levels. Lipids Health Dis. 2012;11:50. doi:10.1186/1476-511x-11-50.PubMedPubMed CentralView ArticleGoogle Scholar
  30. den Hoed M, Strawbridge RJ, Almgren P, Gustafsson S, Axelsson T, Engstrom G, et al. GWAS-identified loci for coronary heart disease are associated with intima-media thickness and plaque presence at the carotid artery bulb. Atherosclerosis. 2015;239(2):304–10. doi:10.1016/j.atherosclerosis.2015.01.032.View ArticleGoogle Scholar
  31. Deloukas P, Kanoni S, Willenborg C, Farrall M, Assimes TL, Thompson JR, et al. Large-scale association analysis identifies new risk loci for coronary artery disease. Nat Genet. 2013;45(1):25–33. doi:10.1038/ng.2480.PubMedView ArticleGoogle Scholar
  32. Ritsch A, Sonderegger G, Sandhofer A, Stanzl U, Tancevski I, Eller P, et al. Scavenger receptor class B type I polymorphisms and peripheral arterial disease. Metabolism. 2007;56(8):1135–41. doi:10.1016/j.metabol.2007.04.009.PubMedView ArticleGoogle Scholar
  33. Komar AA. Genetics. SNPs, silent but not invisible. Science. 2007;315(5811):466–7. doi:10.1126/science.1138239.PubMedView ArticleGoogle Scholar
  34. Kimchi-Sarfaty C, Oh JM, Kim IW, Sauna ZE, Calcagno AM, Ambudkar SV, et al. A “silent” polymorphism in the MDR1 gene changes substrate specificity. Science. 2007;315(5811):525–8. doi:10.1126/science.1135308.PubMedView ArticleGoogle Scholar
  35. Komar AA. Silent SNPs: impact on gene function and phenotype. Pharmacogenomics. 2007;8(8):1075–80. doi:10.2217/14622416.8.8.1075.PubMedView ArticleGoogle Scholar
  36. Cartegni L, Chew SL, Krainer AR. Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet. 2002;3(4):285–98. doi:10.1038/nrg775.PubMedView ArticleGoogle Scholar
  37. McCarthy JJ, Lehner T, Reeves C, Moliterno DJ, Newby LK, Rogers WJ, et al. Association of genetic variants in the HDL receptor, SR-B1, with abnormal lipids in women with coronary artery disease. J Med Genet. 2003;40(6):453–8.PubMedPubMed CentralView ArticleGoogle Scholar

Copyright

© Goodarzynejad et al. 2016

Advertisement