Association between the LIPG polymorphisms and serum lipid levels in the Maonan and Han populations.

Abstract Introduction The Maonan population is a relatively isolated minority in China. Little is known about endothelial lipase gene (LIPG) single nucleotide polymorphisms (SNPs) and serum lipid levels in the Chinese populations. The present study aimed to detect the association of several LIPG SNPs and environmental factors with serum lipid levels in the Chinese Maonan and Han populations. Methods In total, 773 subjects of Maonan ethnicity and 710 participants of Han ethnicity were randomly selected from our previous stratified randomized samples. Genotypes of the LIPG rs2156552, rs4939883 and rs7241918 SNPs were determined by polymerase chain reaction‐restriction fragment length polymorphism, and then confirmed by direct sequencing. Results The allelic (rs2156552, rs4939883 and rs7241918) and genotypic (rs2156552 and rs4939883) frequencies were different between the two ethnic groups (p < 0.05–0.01). The minor allele carriers had lower apolipoprotein (Apo)A1/ApoB ratio (rs2156552 and rs7241918), high‐density lipoprotein cholesterol (HDL‐C) and apolipoprotein (Apo)A1 (rs2156552) levels and higher ApoB levels (rs4939883) in the Han population, and lower HDL‐C (rs2156552, rs4939883 and rs7241918) levels in the Maonan minority than the minor allele non‐carriers (p < 0.0167 after Bonferroni correction). Subgroup analyses according to sex showed that the minor allele carriers had a lower ApoA1/ApoB ratio (rs2156552 and rs7241918) and higher ApoB levels (rs7241918) in Han males, and lower ApoA1 and HDL‐C levels in Maonan females than the minor allele non‐carriers (p < 0.0167–0.001). Conclusions The present study demonstrates the association between the LIPG polymorphsims and serum lipid levels in the two ethnic groups. These associations might have an ethnic‐ and or/sex‐specificity.


| INTRODUCTION
Coronary artery disease (CAD) is the most common cause of mortality, morbidity and a major contributor to the financial burden in developing countries. 1 Dyslipidemia, particularly hypertriglyceridemia and hypercholesterolemia, is a well-described independent predictor for atherosclerosis and CAD. Low-density lipoprotein cholesterol (LDL-C) has been considered to be the major lipid risk factor and main target of lipid-lowering therapy in most national guidelines. Previous studies have shown a highly consistent, inverse correlation between plasma concentrations of high-density lipoprotein cholesterol (HDL-C) and its major protein apolipoprotein (Apo)A1 and atherosclerotic cardiovascular disease risk in humans. 2 In addition, clinical data have indicated that each 1% increase in the serum concentration of HDL-C can decrease cardiovascular risk by 2-3%. 3 Although previous research reported that the risk for dyslipidemia is largely attributed to an unhealthy lifestyle, such as poor nutrition, 4 lack of exercise, 5 excessive drinking and smoking, 6 there is now strong evidence suggesting that predisposition to the development of lipid disorders begins with heredity. 7 The heritability of serum HDL-C levels in the Strong Heart Family study and HERITAGE family study has been estimated at 50% and 52%, respectively. 8,9 By contrast, the associated variants in genome-wide association studies (GWAS) accounted for only 5-8% of the variation in the HDL-C levels. 10 However, the specific genetic variants that contribute to serum HDL-C levels in the diverse ethnic groups are largely unknown.
The endothelial lipase gene (LIPG, as known as EL, EDL, PRO719; Gene ID: 9388; HGNC ID:6623) is the most recent member assigned to the triglycerde (TG) lipase family, which was reported to play a physiological role in the modulation of HDL-C metabolism. 11 Serum HDL-C levels are regulated in part by the lipase enzyme family, and its members include lipase, lipoprotein lipase (LPL) and hepatic lipase (HL). 12 LIPG is highly homologous to LPL and HL, both of which are critical to the metabolism of lipids carried on plasma lipoproteins. 13 A study of the lipolytic activity showed that LIPG has more phospholipase activity and relatively less TG lipase activity and can hydrolyze HDL phospholipids ex vivo. 14 Another study demonstrated that overexpression of LIPG in mice liver by adenovirus-mediated gene transfer results in a remarkable decrease in HDL-C and ApoA1 levels. 15 Antibody inhibition studies in wild-type and LIPG knockout mice demonstrated that inhibition of LIPG causes siginificantly increased HDL-C levels. 16 Vergeer et al. 17 showed that LIPG uses its phospholipase activity to hydrolyze HDL-C (its primary substrate) in a dose-dependent manner. Additionly, a previous study reported that, although the preferred substrate of LIPG is HDL, LIPG is still capable of hydrolyzing apoB-containing lipoproteins [very LDL (VLDL)/ lDL)]. 18 Indeed, Broedl et al. 19 demonstrated that LIPG can reduces the levels of serum VLDL-C and LDL-C in atherosclerosis-prone mouse models. These data suggest that, in addition to its role in HDL metabolism, LIPG may also contribute to VLDL and LDL metabolism. These loss-of function experiments suggest that LIPG could be a physiological regulator of lipid metabolism. Despite the obvious functional evidence for an influence of LIPG on altered serum lipid levels in animal models, it remains to be determined whether this receptor has an equally important function in humans.
The human LIPG is located on chromosome 18q21.1 and is expressed in a variety tissues, including the liver, placenta, lung and testis. 20 Several SNPs in the LIPG have been found to be associated with serum HDL-C concentrations in some studies but not in others. [21][22][23][24][25][26][27][28][29] The main reason for inconsistency in serum lipid levels among these studies may be the different ethnic, genetic, sex, health and environmental factors and their interactions. Therefore, further research will be necessary to characterize the full impact of these SNPs on lipid metabolism in different racial and ethnic groups.
China is a multi-ethnic country with 56 ethnic groups, and the  Figure S1) and all participants were agricultural workers. The age of the participants ranged from 25 to 80 years, with a mean ± SD age of 56.05 ± 11.67 and 57.14 ± 14.99 years in the Han and Maonan populations (p > 0.05), respectively. All study subjects were essentially healthy, with no history of cardiovascular disease such as CAD and stroke, diabetes, hyper-or hypothyroids, and chronic renal disease.
We excluded subjects who had a history of taking medications known to affect serum lipid levels (lipid-lowering drugs such as statins or fibrates, beta-blockers, diuretics, or hormones) before the blood sample was drawn. The present study was approved by the Ethics Committee of the First Affiliated Hospital, Guangxi Medical University

| Epidemiological survey
The survey was carried out using internationally standardized methods, in accordance with a common protocol. 30

| Biochemical measurements
A fasting venous blood sample of 5 ml was drawn from the participants. A part of the sample (2 ml) was collected into glass tubes and used to determine serum lipid levels, and another part (3 ml) was shifted to tubes with anticoagulants (4.80 g/l citric acid, 14.70 g/l glucose and 13.20 g/l tri-sodium citrate) and used to

| DNA amplification and genotyping
Genomic DNA of the samples was extracted from peripheral blood leucocytes in accordance with the phenol-chloroform method. 33 The extracted DNA was stored at 4°C until analysis. Genotyping of the

| DNA sequencing
Eighteen samples (each genotype in two) detected by the PCR-RFLP were also confirmed by direct sequencing. The PCR products were purified by low melting point gel electrophoresis and phenol extraction, and then the DNA sequences were analyzed using an ABI Prism

| Diagnostic criteria
The normal values of serum TC, TG, HDL-C, LDL-C, ApoA1, ApoB Allele frequency was determined via direct counting and the Hardy-Weinberg equilibrium was verified with a standard goodness-of-fit test. Genotype distribution between the two groups was analyzed by the chi-squared test. General characteristics between the two ethnic groups were compared using Student's unpaired t-test. The association between genotypes and serum lipid parameters was tested by analysis of covariance (ANCOVA). Any SNPs associated with the lipid profiles at p < 0.0167 (corresponding to p < 0.05 after adjusting for three independent tests with Bonferroni correction) were considered statistically significant. Sex, age, BMI, blood pressure, alcohol consumption and cigarette smoking were adjusted for the statistical analysis. Multivariable linear regression analyses with stepwise modeling were used to determine the correlation between the genotypes (rs2156552: AT/TT = 0, AA = 1; rs4939883: CT/TT = 0, CC = 1; rs7241918: GT/TT = 0, GG = 1) and several environmental factors with serum lipid levels in a combined population of Maonan and Han, Maonan, Han, males and females; respectively. p < 0.05 (twosided) was considered statistically significant.

| General characteristics and serum lipid profiles
General characteristics and serum lipid parameters for the Han and Maonan populations are summarized in Table 1. Levels with respect to systolic blood pressure, diastolic blood pressure, pulse pressure, waist circumference and the percentages of subjects who consumed alcohol and smoked cigarettes were lower in the Han than in the Maonan population (p < 0.05-0.001). The levels of serum HDL-C and the ratio of ApoA1 to ApoB were higher in the Han than in the Maonan population, whereas the levels of ApoB were lower in the Han than in the Maonan population (p < 0.05-0.001). There were no significant differences with respect to sex ratio, age structure, height, weight, BMI, glucose, serum TC, TG, LDL-C and ApoA1 levels between the two ethnic groups (p > 0.05 for all).

| Genotypic and allelic frequencies
The PCR products of the samples and the results of genotyping of the LIPG SNPs are shown in Figure 1. The genotypes detected by PCR-RFLP were also confirmed by direct sequencing ( Figure 2). As shown in Figure 3, the genotypic and allelic frequencies of the LIPG rs2156552 and rs4939883 SNPs were different between the two ethnic groups. The rs2156552T and rs4939883T allele frequencies and the rs2156552AT/TT and rs4939883CT/TT genotype frequencies were higher in the Maonan than in the Han population (p < 0.05-0.01). The allelic frequencies of the LIPG rs7241918 SNP were also different between the two ethnic groups. The rs7241918G allele frequency was higher in the Maonan than in the Han population (p < 0.05). There was no significant difference with respect to either genotypic or allelic frequencies between males and females of both ethnic groups ( Figure 4).

| Genotypes and serum lipid levels
As shown in Figure 5, the minor allele carriers had a lower ApoA1/ApoB ratio (rs2156552 and rs7241918) and HDL-C and ApoA1 levels (rs2156552) and higher ApoB levels (rs4939883) in the Han population, and lower HDL-C (rs2156552, rs4939883 and rs7241918) levels in the Maonan minority than the minor allele noncarriers (p < 0.0167). Subgroup analysis according to sex showed that the minor allele carriers had a lower ApoA1/ApoB ratio (rs2156552 and rs7241918) and higher ApoB levels (rs7241918) in Han males but not females, and lower ApoA1 and HDL-C levels in Maonan famales but not males than the minor allele non-carriers (p < 0.0167 for all) ( Figure 6).

| Relative factors for serum lipid parameters
The correlation between genotypes of three SNPs and serum lipid parameters is shown in Table 2. Multivariable linear regression analyses showed that genotypes were associated with HDL-C levels (rs2156552, rs4939883 and rs7241918) in the Maonan minority, and the ApoA1/ApoB ratio (rs2156552 and rs7241918) and ApoB levels (rs4939883) in the Han population. Serum lipid parameters were also correlated with several environmental factors, such as sex, age, alcohol consumption, cigarette smoking, blood pressure, blood glucose, waist circumference and BMI, in both ethnic groups or in males and females (p < 0.05-0.001) (Tables 3 and 4).

| Logistic regression of SNPs and serum lipid levels
As shown in Table 5, the genotype and allelic frequencies of the LIPG rs2156552, rs4939883 and rs7241918 SNPs were significantly different in the normal and dyslipidemia groups (p < 0.05-0.001).

| SNP-environmental interactions on serum lipid levels
The interactions of the LIPG SNPs and sex, age, BMI, smoking and drinking on serum lipid levels are shown in Table 6       and allele frequencies of the rs2156552, rs4939883 and rs7241918 SNPs shared a racial/ethnic specificity.
In the last decade, several HDL-C candidate genes have been identified via association studies and one of the candidate genes found to be associated in GWAS is LIPG. Studies using genetically modified mice have suggested that LIPG activity negatively influences plasma HDL-C levels. 16,17 The high-level overexpression of LIPG in the liver significantly decreased the levels of serum HDL-C and ApoA1. 17 A relevant study showed that the LIPG mutations are involved in increased monocyte adhesion and uptake in the vessel wall, which is used to indicate the early inflammation step of atherosclerosis.
Because a reduced plasma HDL-C level is a well-documented and modifiable risk factor for atherosclerotic diseases, including CAD and ischemic stroke, many genetic association studies have investigated the effects of common sequence variants in LIPG on HDL-C levels and diseases related to HDL-C levels. However, previous findings regarding the association of these SNPs with changes in serum lipid levels are inconsistent. The LIPG rs2156552 SNP has been associated   with HDL-C in many nationalities. [23][24][25] A large Diabetes Genetic Initiative GWAS reported that subjects with the minor allele of rs2156552 had lower serum HDL-C concentrations than subjects with the major allele (p = 2 × 10 −7 ). 22 Another GWAS conducted in 17 723 participants of white European descent also confirmed that the T allele carriers of rs2156552 in LIPG were associated with lower levels of HDL-C (p = 1.7 × 10 −12 ). 23 Dumitrescu et al. 24 replicated the association of rs2156552 with plasma HDL-C levels in diverse populations and found that the minor allele carriers had a negative correlation with

HDL-C in European Americans but an opposite effect in African
Americans and American Indians. By contrast, Chung et al. 39 reported that the rs2156552 SNP was associated with the serum HDL-C level in a Korean population (p = 2.61 × 10 −4 ). Likewise, the potential association of LIPG rs4939883 SNP and serum lipid profiles is contradictory. Kathiresan et al. 25 observed that the rs4939883 SNP mutant allele was related to a significant decrease in plasma HDL-C compared to subjects who carried the wild-type allele in European populations (p = 1.6 × 10 −11 ). The association of rs4939883 SNP and HDL-C levels was replicated in another GWAS (p = 7 × 10 −5 ). 26 Nevertheless, a study investigating lipid-related genetic loci in Caucasian women showed that the rs4939883 variant was highly associated with a lower level of ApoA1 (p = 1.9 × 10 −5 ), although there was no association with HDL-C. 27 Teslovich et al. 28   pickle sour meat and animal offals, which contain abundant saturated fatty acids. A long-term diet high in saturated fats has been associated with deleterious effects on serum lipid metabolism, especially because of their influence on serum TG, TC and LDL-C levels. 40 In the present study, we observed that the values for waist circumference and serum TG and ApoB were higher in the Maonan than in the Han population. Most of the local adult Maonan men like to drink and smoke, and consider it impolite not to give their guests wine. We also found that the the percentages of subjects who consumed alcohol and smoked cigarettes were higher in the Maonan than in the Han population. Excessive alcohol consumption and cigarette smoking have been shown to lead to dyslipidemia and related diseases. 41 There is strong evidence for an adverse relationship between smoking and the metabolic and physiologic responses of different heart and blood vessels. 42 Smoking initiates and promotes atherosclerosis by altering cardiac hemodynamics, causing dyslipidemia and the increased production of free oxygen radicals as a result of the oxidative stress of nicotine. 43 Moderate alcohol consumption was shown to be causally related to a low risk of CAD by mainly increasing serum HDL-C and ApoA1 concentrations; however, an excessive intake of alcohol has been shown to lead to hypertriglyceridemia. 44 A previous meta-analysis indicated that 30 g of alcohol daily was associated with a plasma TG increase of 5.69 mg/dl. 45 Another study conducted in older Italian subjects (65-84 years old) found that alcohol intake increases serum LDL-C levels. 46  There are several potential limitations to the present study. First, the sample size was relatively small compared to many GWAS and replication studies and further studies with larger sample sizes are needed to confirm our results. Second, serum lipid levels were examined only once, and so it hard to represent the long-term status of lipid levels. Third, we were unable to alleviate the effect of diet and several environmental factors during the statistical analysis. Thus, LIPG expression in adipose tissue and serum lipid levels needs to be investigated in further studies.

| CONCLUSIONS
Overall, the present study shows that the allelic (rs2156552, rs4939883 and rs7241918) and genotypic (rs2156552 and rs4939883) frequencies were different between the two ethnic groups. The association of LIPG polymorphisms and serum lipid levels was also different between the Maonan and Han populations, or between males and females. These results suggest that there may be a sex and/or racial/ethnic-specific association of the LIPG SNPs and some serum lipid parameters in our study populations. These differences in the association of LIPG polymorphisms and serum lipid profiles between the two ethnic groups might partly result from different LIPG-enviromental interactions.