Associations of three lipoprotein lipase gene polymorphisms, lipid profiles and coronary artery disease
MOHAMED S. DAOUD, FARID S. ATAYA, [...], and ABDULAZIZ A. AL-JAFARI
Abstract
Lipoprotein lipase (LPL) plays a central role in lipoprotein metabolism by hydrolyzing the core triglycerides (TGs) of circulating chylomicrons and very-low-density lipoprotein (VLDL). The effects of LPL polymorphisms on lipid levels and coronary artery disease (CAD) have been inconsistent among studies and populations. To assess the lipid profiles and distributions of three LPL gene polymorphisms in Saudi patients with CAD, the HindIII, PvuII and Ser447Ter polymorphisms in the LPL gene were analyzed in 226 patients with CAD and 110 controls. Polymerase chain reaction-restriction fragment length polymorphism was used to detect LPL gene polymorphisms. The plasma lipid profiles of the patients were determined using standard enzymatic methods. Patients in the CAD group had significantly higher triglyceride (TG), total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) levels than controls irrespective of the HindIII, PvuII or Ser447Ter genotype. Compared to the findings in controls, the HindIII TT, PvuII TC and Ser447Ter CC genotypes were associated with significantly reduced high-density lipoprotein cholesterol (HDL-C) levels in patients with CAD (P<0.0001). In summary, there are associations between LPL gene variants and high plasma TG, TC and LDL-C levels as well as low HDL-C levels.
Introduction
Lipoprotein lipase (LPL) is a glycoprotein that is synthesized in the parenchymal cells of different tissues. LPL hydrolyzes circulating core triglycerides (TGs) of exogenous (chylomicron) and endogenous [very-low-density lipoprotein (VLDL)] origins to provide free fatty acids for oxidation and utilization in the heart and other tissues and for storage in adipose tissue (1). LPL affects circulating triglyceride (TG) levels by generating lipoprotein remnants, which are processed by hepatic lipase. Following secretion, LPL attaches to the luminal surface of endothelial cells, in which it has a significant role in the catabolism of lipoproteins in circulation and interacts with lipoproteins locally (2). It was previously reported that LPL increases the retention of low-density lipoprotein (LDL) and VLDL particles by the subendothelial matrix of the arterial wall, which enhances the conversion of these lipoproteins into more atherogenic forms (3). These localized and deleterious effects cannot be assessed by measuring either circulating LPL or TG levels but can be explored by identifying genetic variants of LPL and their relationships with the presence and extent of atherosclerotic lesions. Findings of previous studies revealed that plasma lipoprotein concentrations are significant predictors for the risk of coronary heart disease (CHD) (4). Thus, genes with essential roles in lipoprotein metabolism are excellent candidates for inter-individual variation in the susceptibility to CHD (5), including the LPL gene. The LPL gene spans over 30 kb on chromosome 8p22, and is divided into 10 exons (6). The major products of LPL activity are free fatty acids and glycerol for energy utilization and storage (1) and LDL activity is accompanied by the formation of intermediate-density lipoprotein and chylomicrons remnants (CRs). Independent of its lipolytic activity, LPL binds to and travels with CRs to the liver, where it enhances the clearance of these lipoproteins via LDL receptors (7). Due to the pivotal role of LPL in lipid metabolism, genetic defects in the LPL gene can affect lipoprotein metabolism, resulting in an atherogenic lipid profile. A number of mutation have been identified in this locus. Additionally, several rare mutations in the LPL gene have been associated with markedly reduced enzyme activity and a number of common variants have been associated with moderate changes in LPL catalytic function (8). However, there is controversy with regard to the effects of these variants on LPL activity. For example, Emi et al (9) reported that the G188E mutation leads to the expression of an inactive enzyme, which explains the manifestation of LPL deficiency. However, Hallman et al (10) did not find any association between the HindIII(−) allele and higher LPL activity. The associations of certain polymorphic loci in the LPL promoter, introns or exons with lipid disorders and coronary artery disease (CAD) have been reported by some groups (rs285, rs1800590, rs320, rs268, rs1801177) (6–8,11–14) but contested by other authors (15–17). To the best of our knowledge, no data on the screening of such specific polymorphisms in the Saudi population have been reported. Therefore, the aim of the present investigation was to determine and address the lipid profile and distribution of the HindIII, PvuII and Ser447Ter polymorphisms of LPL in Saudi patients with CAD.
Materials and methods
Study subjects
The study comprised 226 patients (157 males and 69 females, aged 42–82 years) who were admitted to the Department of Cardiology, King Khalid University Hospital, Riyadh, Saudi Arabia and 103 healthy subjects (58 males and 45 females, aged 20–78 years) who had no history of CAD as controls. The subjects included in this study were of unrestricted age and gender. All the subjects provided written informed consent prior to participation as well as to having blood drawn at the time of angiography or time of screening for DNA extraction. The study was reviewed and approved by the Institutional Review Board of the King Khalid University Hospital. This study was conducted in accordance with the guidelines set by the Ethics Committee of the College of Medicine and Research Centre of King Saud University, Riyadh, Saudi Arabia. All subjects enrolled in this study were Saudi residents with similar dietary patterns. The key demographic data of the subjects were recorded including age, gender and lipid profiles. CAD was assessed by review of the patients’ angiograms by their treating cardiologists.
Sample collection and lipid analysis
Blood samples for the glucose and lipid measurements were drawn from the patients and controls subsequent to an overnight fast. The plasma glucose concentration was measured by the glucose oxidase method using a Biotrol kit (Biotrol, Earth City, MO, USA) on a Bayer Opera analyzer [Bayer Diagnostics (Siemens), Munich, Germany]. Serum total cholesterol (TC) was measured using a Biotrol kit, and high-density lipoprotein cholesterol (HDL-C) was measured using a commercial Randox kit (Randox Laboratories Ltd., London, UK). LDL cholesterol [low-density lipoprotein cholesterol (LDL-C)] levels were calculated using the Friedewald formula and TG levels were measured using the lipase/glycerol kinase UV endpoint method of the Opera analyzer.
DNA extraction
Genomic DNA was extracted from peripheral blood specimens, which were drawn into tubes containing ethylenediaminetetraacetic acid, using the QIAamp DNA isolation kit (Qiagen, Hilden, Germany).
Genotyping
The presence of three common polymorphisms of the LPL gene was determined by polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) analysis using genomic DNA. The primer sets were selected on the basis of previously published information as follows (18): HindIII forward (H1): 5′-TGA AGC TCA AAT GGA AGA GT-3′, and reverse (H2): 5′-TAC AAG CAA ATG ACT AAA-3′; PvuII forward (P1): 5′-ATG GCA CCC ATG TGT AAG GTG-3′, and reverse (P2): 5′-GTG AAC TTC TGA TAA CAA TCT C-3′; and Ser447Ter forward: 5′-TAC ACT AGC AAT GTC TAG GTG A-3′, and reverse: 5′-TCA GCT TTA GCC CAG AAT GC-3′. In each reaction, 3 μl (150 ng) of the genomic DNA template was added to the PCR reaction mixture, which consisted of 12.5 μl of 2X Promega master mix (Promega Corporation, Madison, WI, USA), 2 μl of each primer and distilled water to a final volume of 25 μl. The PCR conditions were as follows: an initial denaturation at 94°C for 2 min was followed by 40 cycles of denaturation at 94°C for 15 sec, annealing at 50°C for 30 sec and extension at 72°C for 1 min, with a final extension at 72°C for 2 min. PCR was performed in a MyCycler (Bio-Rad, Hercules, CA, USA).
LPL gene polymorphism analysis
Digestion of the PCR products was performed by adding 1 μl of the respective restriction enzyme (HindIII and PvuII, both Promega Corporation; MnlI: New England Biolabs, Ipswich, MA, USA) to 10 μl of the PCR product containing 2 μl of 10X buffer solution (final reaction volume of 20 μl). The mixtures were centrifuged for 2 min at 3,913 x g and incubated in a water bath at 37°C overnight. The resulting fragments were resolved by electrophoresis (80 V, 60 min) on a 2.5% agarose gel and visualized using UV light. The HindIII site (intron 8) produced a 600 bp fragment following digestion. The PvuII restriction site (intron 6) yielded 330 and 110 bp fragments. Genotypes were scored by an experienced reader blinded to the clinical and angiographic results. The polymorphic allele with the restriction site was designated as ‘T’ and the allele without the site as ‘G’ for HindIII. For PvuII, the allele with the restriction site was designated as ‘T’ and that without the site as ‘C’. The 488 bp PCR product contained two MnlI restriction sites, one of which is a polymorphic site indicating the Ser447Ter mutation. Digestion of the PCR product with MnlI resulted in three fragments of 290, 250 and 200 bp (18,19). The identified genotypes were named according to the presence or absence of the enzyme restriction sites; for example, Ser447Ter GG, GC and CC indicated homozygosity for the presence of the site, and heterozygosity and homozygosity for the absence of the site, respectively.
Statistical analysis
Measurement data were presented as the mean ± standard deviation (SD), and compared using the two-sample Student’s t-test. Enumeration count data were summarized as numbers (%) and compared using the χ2 test. Two analyses were used to evaluate the allelic and genotypic frequencies that were calculated from the observed genotypic counts and to assess the Hardy-Weinberg equilibrium expectations. The same methodology was applied to comparisons between allelic and genotypic frequencies. Associations were determined as odds ratios (ORs) and 95% confidence intervals (CIs). The likelihood of carrying a specific allele was defined as the number of subjects who carried the allele divided by the number of subjects in who did not carry the allele. The OR for the LPL genotype distribution was determined using χ2 analysis. The CAD was defined as the odds of allelic carriage in the diseased (CAD) group divided by the odds of allelic carriage in the healthy (control) group. Statistical analysis was performed using the Statistical Package for Social Sciences for Windows, version 20.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.
Results
Patient data
Details regarding the clinical and biochemical characteristics of the study population (226 patients with CAD and 103 controls) are included in Table I. Patients with CAD had significantly lower HDL-C concentrations compared to the control group and significantly higher plasma levels of fasting blood sugar, TG, TC and LDL-C (P<0.0001 for each) compared to the control subjects.
Risk factors for CAD
Diabetes mellitus, dyslipidemia, hyper-tension and smoking were selected as major risk factors. The frequencies of the major CAD risk factors are shown in Table II. Diabetes mellitus, dyslipidemia, hypertension and smoking were more frequently present in the patient group than in the control group and diabetes mellitus, dyslipidemia, hypertension and smoking were identified as risk factors for CAD (OR=22.53, 9.04, 24.79 and 3.59, respectively, P<0.0001 for each).
The entire study population was genotyped for three LPL gene polymorphisms: HindIII (alleles designated as T and G), PvuII (alleles designated as T and C) and Ser447Ter (alleles designated G and C). Table III shows the genotypes and frequencies for the observed alleles. For all polymorphisms in the patients and controls, the distribution of genotypes yielded from the Hardy-Weinberg equilibrium were as expected. Allele frequencies did not differ significantly between the CAD and control groups. Within the CAD group (n=226), the TT HindIII genotype was identified in 102 patients (45.1%), whereas 81 (35.8%) and 43 patients (19%) carried the TG and GG genotypes, respectively. Within the control group (n=103), the TT genotype was identified in 42 subjects (40.8%), whereas 35 (34%) and 26 subjects (25.2%) carried the TG and GG genotypes, respectively. Regarding HindIII, patients with CAD were less likely to carry the GG genotype than the control group and patients with CAD had higher frequencies of the TT and TG genotypes. For the PvuII genotype, the TT genotype was identified in 89 patients (39.4%) within the CAD group, whereas 44 (42.7%) and 35 patients (15.5%) carried the TC and CC genotypes, respectively. Within the control group, the TT genotype was identified in 46 patients (44.7%), while 44 (42.7%) carried the TC genotype and 13 patients (12.6%) carried the CC genotype. Concerning PvuII, patients with CAD had higher frequencies of the TC and CC genotypes but a lower frequency of the TT genotype compared to the controls. For the Ser447Ter genotype, the CC genotype was found in 185 patients (81.9%) within the CAD group, whereas 41 patients (18.1%) carried the GC genotype. Within the control group (n=103), the CC genotype was identified in 92 patients (89.3%) and 11 patients (10.7%) carried the GC genotype. We did not find any GG genotypes in either group for this gene. A higher frequency of the GC genotype and a lower frequency of the CC genotype were observed in the CAD group compared to the control group.
Data are presented as the observed number of cases and the expected number of cases in parentheses. The χ2 test was used for comparisons between positive and negative risk factors.
Relationships between the frequencies of major CAD risk factors and LPL gene polymorphisms
The relationships between the frequencies of major CAD risk factors and those of LPL gene polymorphisms in the patients with CAD are presented in Table IV. The results revealed associations between LPL HindIII genotypes and hypertension (χ2=6.68, P=0.03) and smoking (χ2=5.80, P=0.05). Additionally, a positive relationship between PvuII genotypes and smoking (χ2=6.964, P=0.03) was noted as well as a relationship between Ser447Ter genotypes and diabetes (χ2=6.74, P=0.009).
Relationships between the HindIII, PvuII and Ser447Ter genotypes and lipid parameters of CAD and the control groups
The relationships between the HindIII, PvuII and Ser447Ter genotypes and the lipid parameters of the CAD and control groups are shown in Table V. TG, TC and LDL-C levels were significantly higher in the CAD group compared to the control group irrespective of the HindIII, PvuII or Ser447Ter genotype (P<0.0001 for each), whereas compared to the levels in the control group, the HindIII TT (P=0.03), PvuII TC (P=0.006) and Ser447Ter CC genotypes (P=0.003) were associated with decreased HDL-C levels in the CAD group.
Association among the HindIII, PvuII and Ser447Ter genotypes
As shown in Table VI, no association was observed between the HindIII and PvuII genotypes. However, an association between the HindIII and Ser447Ter genotypes was noted (χ2=7.33, P=0.03). A significant relationship was observed between the PvuII and Ser447Ter genotypes (χ2=7.3, P=0.02).
Haplotype frequencies for the three LPL polymorphisms
Haplotype reconstruction for the three LPL polymorphisms under study revealed 18 haplotypes possessing HindIII (intron 8 T481G), PvuII and Ser447Ter polymorphisms. The GGTTCC and TGCCCC haplotypes were significantly more common in the control group compared to the CAD group (P=0.003 and 0.007, respectively, Table VII).
Discussion
CAD is a complex disease with well-documented genetic and environmental components. The findings of most genetic studies of CAD are controversial, likely because the genetic risk of CAD is not based on a single gene but is instead based on interactions among several pathophysiological pathways involving multiple genes and environmental risk factors. LPL is a potential target for the treatment of CAD because LPL gene variants are involved in a number of pathophysiological conditions associated with CAD. Several mutations of the LPL gene have been identified thus far and 20% of these mutations occur in noncoding regions (20–21). The LPL gene is one of the most appealing candidate genes that may be used to explain some of the lipid and lipoprotein abnormalities encountered in numerous cases of CAD. In this study, we presented results on polymorphisms and haplotypes of the LPL gene in a Saudi population and the possible association of these polymorphisms with CAD and lipid profiles in some detail.
CAD is a multifactorial disorder believed to result from an interaction between the genetic background and environmental factors such as diet, smoking and physical activity. CAD is usually associated with conventional risk factors including hypertension, diabetes mellitus and hypercholesterolemia (22). In the present study, diabetes mellitus, dyslipidemia, hypertension and smoking were identified as risk factors for CAD and a positive association was found between LPL HindIII genotypes and hypertension and smoking. Additionally, an association between PvuII genotypes and smoking was observed. The association between Ser447Ter genotypes and diabetes was evident in the present study. In general, individuals with hypertension, diabetes mellitus and hypercholesterolemia were considered at high risk for CAD, while subjects that did not exhibit any of these factors were considered low-risk subjects. Age, smoking, hypertension and diabetes have been established as independent risk factors for ischemic cardiovascular disease (23). Hypertension and dyslipidemia were found to influence CAD (24). Genetic factors were statistically independent of age, smoking, hyperuricemia, hypertension, diabetes mellitus and hypercholesterolemia. Smoking was considered an important environmental factor for CAD in men at low risk, consistent with the hypothesis that the cessation of smoking is important in the prevention of CAD in these individuals (23). It has been documented that dyslipidemia, diabetes mellitus and obesity were more common among patients with a +/− genotype than among controls of the same genotype, while dyslipidemia and hypertension were identified as independent risk factors for CAD (24).
The frequencies of the HindIII (TT, 45.1; TG, 35.8 and GG, 19%) and PvuII genotypes (TT, 39.4; TG, 42.7 and CC, 15.5%) in the present study were similar to those found in different ethnic groups, including Northern Europeans (7), Caucasians (25), Russians (26) and Tunisians (27). The allele frequencies of PvuII (C) and HindIII (G) in this study were similar to those found in other Caucasian populations. Ahn et al (25) reported allele frequencies of 0.45 and 0.26 for PvuII(−) and HindIII(−), respectively, in a control population (n=539). Peacock et al (28) also reported frequencies of 0.435 and 0.228 for PvuII(−) and HindIII(−), respectively, among 92 healthy control subjects. A strong linkage disequilibrium was also identified between the two PvuII and HindIII polymorphisms in several other studies (28–33).
It was of interest to compare the Ser447Ter gene polymorphism frequency of our population (81.9 CC and 18.1% GC) with other published data as there is considerable variation in the reported frequency of this polymorphism in different populations (7.4–20%) (12,34,35). Komurcu-Bayrak et al (36) reported a Ser447Ter allele frequency of 11% among randomly selected Turkish participants and obtained a similar minor allele frequency for Ser447Ter in the group without CAD and a lower minor allele frequency in the group with significant CAD.
The HindIII TT (P=0.03), PvuII TC (P=0.006) and Ser447Ter CC genotypes (P=0.003) were associated with significantly lower HDL-C levels in the CAD group compared to the control group (Table V).
The present study clarified the associations among the HindIII, PvuII and Ser447Ter genotypes of the LPL gene and the lipid parameters of patients with CAD. TG, TC and LDL-C levels were higher in the CAD group compared to the control group, irrespective of the LPL genotype, whereas compared to the control group, HDL-C levels were significantly reduced in patients with CAD who had the HindIII TT, PvuII TC and Ser447Ter CC genotypes. Previous studies have investigated the effect of LPL gene activity on the plasma levels of TG, TC, HDL-C and apolipoproteins; however, the results have been inconsistent (19,37–40). Clee et al (38) demonstrated that decreased plasma LPL activity is associated with high TG and low HDL-C levels in patient samples, and the Ser447Ter mutation is associated with higher plasma LPL activity. Common variants of the LPL gene have differential effects on plasma lipid concentrations and the development of atherosclerosis. The association of PvuII and HindIII polymorphisms with hypertriglyceridemia, low HDL-C levels and CAD were reported in several studies (6–8). These effects of the HindIII polymorphism on TG and HDL-C levels are consistent with the physiological role of LPL, which both hydrolyzes TG-rich lipoproteins and modulates plasma HDL-C levels. The HindIII polymorphism is located in intron 8 of the LPL gene, and therefore, should not be the cause of the observed effects (41).
The relationship between hypertriglyceridemia and the PvuII RFLP polymorphism of the LPL gene was investigated in different populations. An association between the LPL PvuII polymorphism and lipid disorders was recorded in a Japanese population and this association occurred due to significantly higher triglyceride concentrations in the PvuII +/+ compared to the −/− genotype group (30). Similar results have been reported for Australian (42), French (43) and Japanese school children (44) and the Welsh (32) population. Authors of those studies also reported a significant decrease in HDL-C concentrations in the PvuII −/− genotype group of CAD patients not receiving lipid-lowering drugs. In contrast to these findings, the PvuII −/− genotype was observed to be associated with higher TG concentrations than the +/+ genotype in a Chinese population in the Beijing area (45). Although Jemaa et al (46) failed to demonstrate any significant relationship between lipid concentrations and PvuII polymorphisms in a French population, their observations suggested an association between PvuII polymorphisms and the severity of coronary lesions. Similar results were reported by Wang et al (42) in a study of 500 Australian cardiology patients. In addition to a significant relationship between PvuII RFLP polymorphisms and hyper-triglyceridemia, this study also identified correlations between this polymorphism and CAD and diabetes.
In several studies, the Ser447Ter allele was associated with a reduced risk of CAD (12,47–50). The beneficial effects of the Ser447Ter (S447X) polymorphism may be related to favorable effects on lipid levels. Previous studies in the Turkish population demonstrated that carriers of the X447 allele, compared with noncarriers, had lower plasma TG levels and higher levels of HDL-C and they were protected against metabolic syndrome (18). Metabolic syndrome and hyperlipidemia increase the severity of CAD (51). Thus, we hypothesized that there is a relationship between the Ser447Ter polymorphism of the LPL gene and the severity of CAD.
In our study, the GGTTCC and TGCCCC haplotypes were significantly more common among the controls. However, the TGCCGC, TTCCGC and GGCCGC haplotypes were observed only in the CAD group. Anderson et al (7) reported that in 483 patients with CAD, the estimated frequency of the HindIII(+)/PvuII(−) (TC) haplotype was nominally greater than that in the control subjects, whereas for the other haplotypes (TT, GT), the estimated frequencies were lower than those of the control subjects. Additionally, the frequency of the GC genotype was similar between the CAD and control groups. After adjustment for potential confounders, the OR for CAD associated with the GGTTCC haplotype compared to the wild type was 0.263 (95% CI, 0.11–0.63, P=0.003) and that for TGCCCC was 0.159 (95% CI, 0.041–0.62, P=0.007) (Table VII). Consequently, the GGTTCC and TGCCCC haplotypes may have been protective against CAD in our studied population.
Goodarzi et al (52) found several differences in the allele and haplotype frequencies of the three LPL markers. Such differences may affect the results of association studies conducted in different populations. Rebhi et al (27) reported that the CTGTAA haplotype was significantly more common among patients with significant coronary stenosis than among controls. However, the CGGGAA haplotype occurred significantly more frequently in the control group compared to the coronary stenosis group. The TGGTAG, TTAGAA and CGGGAG haplotypes were observed only in the coronary stenosis group.
In conclusion, differences in relative genotype frequencies were noted between the patient and control groups. However, these differences did not reach statistical significance (Table III). There were significant differences in the plasma levels of TC, LDL-C, HDL-C and TG in association with the LPL genotypes (Table V), suggesting an association between these polymorphisms and the lipid profiles of patients with CAD. CAD may therefore be a complex disorder caused by a combination of genetic and environmental factors that may influence the onset of disease (Tables II and andIVIV).
Acknowledgments
The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this study through the research group project no. RGP-VPP-173.