The association between peroxisome proliferator-activated receptor Δ rs3777744, rs3798343, and rs6922548 and coronary artery disease

Objective: The aim of the present study is to investigate the association between the single nucleotide polymorphism (SNP) sites of peroxisome proliferator-activated receptor Δ (PPARD) and the risk of coronary artery disease (CAD). To this end, a prospective observational single-center study of the clinical data from 880 subjects in a Chinese population was conducted. Methods: A total of 880 subjects, including 609 CAD patients and 271 control subjects, were selected for the present study. All inpatients had 4 ml of venous blood drawn after 12 h of fasting, and then clinical tests were conducted to obtain the biochemical parameters. CAD patients and Controls were distinguished by coronary angiography. Statistical analysis was conducted with SPSS software (ver 16.0). Results: A significant association between the G-alleles of PPARD rs3777744 and rs3798343 and a decreased risk for CAD was found. Moreover, we found an interaction between high fasting high-density lipoprotein cholesterol (HDL-C) serum levels, low serum glucose levels and their genotypes, ultimately decreasing the risk of CAD. Haplotype analysis was conducted on the three SNP sites, rs3777744 and rs3798343 to form a block [r2 = 0.79, D′ = 0.99). The A-C haplotypes were associated with an increased risk of CAD (odds ratio (OR), 95% confidence interval (CI): 1.321 (1.060–1.647), P=0.013], and the G-G haplotypes were associated with a decreased risk [OR, 95% CI: 0.714 (0.567–0.849), P=0.004]. Conclusions: Our study indicates a significant association between the G-alleles of PPARD rs3777744 and rs3798343 and a decreased CAD risk. In addition, genotypes interact with high serum HDL-C levels and low serum glucose levels, resulting in decreased prevalence of CAD.


Introduction
Cardiovascular diseases (CVD), particularly coronary artery disease (CAD), are typically caused by atherosclerosis (AS) and are a major cause of deaths worldwide [1]. CAD is mainly diagnosed by angiography, an invasive examination that can be harmful to patients [2]. Despite great developments in the diagnosis and treatment of CAD, early diagnosis remains a challenge in clinical practice. As a result, specific and sensitive biomarkers for inchoate detection of CAD are necessary.

Primer design and single nucleotide polymorphism selection
The original gene sequence covering the target single nucleotide polymorphism (SNP) sites was obtained from the NCBI website (www.ncbi.nlm.nih.gov) and further processed on the My Agena website. The ensemble genome browser (www.ensembl.org) was used to screen the SNP sites in the 5-near, seed and 3-near regions of LncPPARD. The minor allele frequencies (MAF) value was accessed on the International Genome website (www.internationalgenome. org), and all MAF values for selected SNP sites were confirmed as more than 0.05. The primers, PCR, and single base extension of the primers were designed by AssayDesigner 3.1 software (Sequenom Inc., San Diego, CA, U.S.A.). The primers were compounded by the same professional biotechnology company. All primers were diluted according to the manufacturer's user guide.

Results
The basic characteristics of the participants Altogether, 880 patients (609 CAD and 271 controls) suspected of CAD were recruited for the present study. All participants underwent coronary angiography for the diagnosis of CAD at the Friendship Hospital of Ili Kazakh Autonomous Prefecture from 1 March 2010 to 31 April 2015. As shown in Table 1, 609 subjects were placed into the CAD group, and the control group was composed of 271 subjects. Subjects that were elderly (P<0.001), male (P<0.001), or frequent smokers (P=0.038), and patients with high FBG levels (P<0.001) and low HDL-C (P=0.011) were more susceptible to CAD. Serum levels of TC, TG, LDL-C, Apo-A, Apo-B, and drinking status were comparable between the CAD and control groups.

Chi-square analysis of the association between SNP sites and CAD risk
The distribution of the rs3798343 genotype is significantly different in the control and CAD groups (P=0.006). Conversely, the distribution of the rs3777744 (P=0.064) and rs6922548 (P=0.520) genotypes are comparable between the CAD and control groups ( Table 2).

Logistic regression analysis of the association between SNP sites and CAD
To further investigate the relationship between the selected SNP sites and CAD, logistic regression analysis was conducted. The genotype distribution of PPARD polymorphism rs3777744 (A>G) is consistent with the Hardy-Weinberg Equilibrium in both the CAD (P=0.309) and control groups (P=0.766). The frequencies of allele A are 73.4 and 67.9% in the CAD and control groups, respectively. All genotypes (AA, P=0.065; AG, P=0.056,   GG, P=0.069) had different distributions in the CAD and control groups, and the GG genotype distribution was significantly different between the CAD patients and the control subjects (adjusted odds ratio (AOR), 95% confidence interval (CI): 0.724, 0.532-0.985; P*=0.04) when the P-value is adjusted for FBG. The genotype distribution of rs3777744 conforms to the dominant model (P*=0.023), and the GG+GA genotype serves as a protective factor in the prevalence of CAD (AOR, 95% CI: 0.714, 0.534-0.954) ( Table 3). The genotype distribution of SNP site rs3798343 (C>G) is consistent with the Hardy-Weinberg Equilibrium in the control group (P=0.894) but not in the CAD group (P=0.033). The frequencies of allele C are 77.83 and 71.22% in CAD and control groups, respectively. The CC and GG genotypes (CC, P=0.005; GG, P=0.001) were distributed differently in the CAD and control groups when the P-value was adjusted for FBG. The genotype distribution of rs3798343 conforms to the dominant model (P*=0.001), and the GG+GC genotype also serves as a protective factor for CAD (AOR, 95% CI: 0.617, 0.460-0.826) ( Table 4). The genotype distribution of rs6922548 was comparable in the two groups. However, small mutations are contained in the subjects, which leads to uncertainty in results (Table 5).

Gene-environment interaction analysis for predicting CAD prevalence
As is shown in

Haplotype analysis of the SNP sites and CAD
Strong LD can be found between the rs3777744 and rs3798343 genotypes in normal control subjects and CAD patients (r 2 = 0.79, D = 0.99) (Figures 1 and 2). Furthermore, haplotype analyses, including those of the two SNPs    and the associations between the different haplotypes and the risk of CAD, have also been carried out. The primary haplotypes are presented in Table 9. The A-C haplotypes were linked with an increased risk of CAD (OR, 95% CI: 1.321 (1.060-1.647), P=0.013), and a decreased risk was associated with the G-G haplotypes (OR, 95% CI: 0.714 (0.567-0.849), P=0.004).

Discussion
Hyperlipidemia is a major risk factor for AS and CAD [15,16]. Abnormal fatty acid metabolism was found to be related to the severity of ischemic heart disease (IHD) [17]. PPARD is a protein-coding RNA located in chromosome 6 with a length of 85633 bp, and studies show that PPARD plays a significant role in lipid metabolism and oxidation regulation in endothelial cells, vascular smooth muscle cells, and cardiomyocytes [9,18]. In animal experiments with goats, PPARD was found to work as a key transcription factor, together with PPARG, that participates in multiple physiological processes to facilitate lipid secretion and catabolism of fatty acids in mammary epithelial cells [19,20]. Activated PPARD affecting the plasma lipid profile has been reported in humans and animals. GW501516, the PPARD agonist, has been shown to up-regulate the expression of ABCA1 in human monocytic cell lines and increase high-density lipoprotein cholesterol (HDL-C) in monkeys [21]. In another previous study, subjects with low plasma HDL-C concentrations (<1.6 mmol/l) were given different concentrations of GW501516 resulting in the improvement of lipid metabolism [22]. In our study, in the gene-environment interaction analysis, HDL-C played a potential role in the prevalence of CAD as it interacted with different genotypes.
The association between diabetes and CAD has been well established by numerous studies. It is well known that hyperglycemia is a risk factor for coronary heart disease [23]. The pathogenesis of insulin resistance is caused by many factors, including ectopic fat accumulation, increased glucose output, and decreased glucose utilization. Metabolism of glucose is regulated by the liver and skeletal muscle, where PPARD is highly expressed [24,25]. The activation of PPARD strengthens muscular fatty acid oxidation and oxidative phosphorylation, and the oxidative capacity of muscle is positively correlated with systemic insulin sensitivity, meaning that PPARD has a beneficial effect on insulin sensitivity in vivo [23]. In our own research, the interaction between elevated FBG and genotype increases the prevalence of CAD, which confirms that hyperglycemia interaction with genotype is a risk factor for CAD.
Recently, SNPs analysis was conducted to explore the molecular mechanisms by which genetic predispositions impact diseases.
Several SNPs in PPARD, including rs2016520 and rs9794, were found to be associated with CVDs [1,7]. In addition, the PPARD SNP rs2016520 (+294C) seemed to be significantly associated with cholesterol metabolism and the risk of CAD [11,26].
To further investigate the relationship between PPARD and CAD, several SNPs including rs3777744, rs3798343, and rs6922548, were selected for the present study.
In a cohort of 880 consecutively sampled patients (609 cases and 271 controls), we found that the distribution of CAD differed between the rs3777744 and rs3798343 polymorphisms. The GG genotype (wild-type) of PPARD rs3777744 decreased the risk of CAD by 27.6% (AOR, 95% CI: 0.724, 0.532-0.985; P*=0.04), and in a dominant model, the GG+GA genotype decreased the risk of CAD by 28.6% (AOR, 95% CI: 0.714, 0.534-0.954; P=0.023). Additionally, the GG genotype of PPARD rs3798343 decreased the risk of CAD by 39.4% (AOR, 95% CI: 0.606, 0.445-0.825; P*=0.001), and the GG+GA genotype decreased the risk of CAD by 38.3% (AOR, 95% CI: 0.617, 0.460-0.826; P*=0.001). We found that high serum glucose levels and low high-density lipoprotein cholesterol levels were shown to increase the risk of CAD, and the effect was more pronounced when the serum indicators interacted with genetic factors. A sequential haplotype analysis was conducted and linkage disequilibrium was found between rs3777744 and rs3798343 (r 2 = 0.79, D = 0.99) (Figures 1 and 2). This is the first study on the relationship amongst rs3777744, rs3798343, and CAD.
The intron variants rs3777744 and rs3798343 occur within an intron located in the gene for PPARD. Transcription variants within an intron have numerous possibilities for regulating genes, such as affecting alternative splicing of the mRNA and enhancing the expression of multiple genes. Although a significant association was found between CAD AP2A, Transcription factor AP-2α. and the PPARD polymorphisms in the present study, more evidence is needed to conclude that SNPs are functional. SNPs can be associated in some populations due to correlations with other SNPs that actually impact the regulation of the gene. The linkage disequilibrium was found between rs3777744 and rs3798343 (r 2 = 0.79, D = 0.99) in haplotype analysis. Subsequently, a methylated CpG islands prediction was conducted using the USUC database. Areas upstream and downstream (1000 bp) of the studied SNPs were included in the analyzed sequences. However, none of the methylation islands were found in the target sequences. In addition, a transcription factor prediction using the PROMO database was conducted, and the analyzed sequences included the areas upstream and downstream (30 bp) of the three SNPs. The possible related transcription factors that bind to the functional variants are shown in . The data implied that CETS1, ELK1, PAX5, and P53 bind to rs3777744, P53 binds to rs3798343, and AP2A and PAX5 specifically bind to rs6922548. Additional transcription factors were predicted to bind to the analyzed sequences (Figures 3-5). P53 (also called TP53, tumor protein P53) is a widely studied protein-coding gene that is significantly related to neoplasms. In addition, P53, along with its functional polymorphism, has been linked to CAD in recent years [27][28][29].
A subsequent KEGG pathway analysis within the transcription factors was conducted. TNF (P=0.170) and ErbB (P=0.136) signaling pathways, which have been found to play key roles in CAD occurrence and development [30][31][32], were suggested as critical regulatory factors.
In the present study, rs3777744 and rs3798343 influenced the prevalence of CAD. We hypothesized that these two polymorphisms participated in the occurrence and development of CAD by targetting P53, and the TNF and ErbB signaling pathways acted as metabolic regulatory factors in CAD under the influence of the variants.
However, this pathway prediction is unreliable due to the small number of predicted transcription factors. Further studies are needed to explore the pathological mechanism.

Limitations
First, the allele distribution of rs6922548 is consistent with the Hardy-Weinberg Equilibrium, but the genotype distribution of rs6922548 shows no difference in between the CAD and control groups, so a larger sample size is needed to study the relationship between CAD and rs6922548. Second, the results of the analysis in this work reveal only the statistical correlation between PPARD polymorphisms and CAD, so a professional analysis of the data should be conducted for more convincing results. Third, a molecular mechanism study is needed to explore the role that PPARD plays in CAD. NFY(A/B/C), Official Symbol nuclear transcription factor Y subunit α/β/γ; STAT4: signal transducer and activator of transcription 4; GR-BETA: Official Symbol NR3C1 nuclear receptor subfamily 3 group C member 1-β; XBP1: X-box binding protein 1; GR-ALPHA, Official Symbol NR3C1 nuclear receptor subfamily 3 group C member 1-α; NR3C1 nuclear receptor subfamily 3 group C member 1; THRA, thyroid hormone receptor α.

Conclusion
In conclusion, there is a significant association between the PPARD rs3777744 G-allele, rs3798343 G-allele, and a decreased risk for CAD. Furthermore, we found that high HDL-C serum levels, low glucose levels, and genotypes interact, ultimately decreasing the risk for CAD.