SYNE1‐QK1 SNPs, G × G and G × E interactions on the risk of hyperlipidaemia

Abstract This study aimed to assess the relationship of 3 spectrin repeat containing nuclear envelope protein 1 (SYNE1) and 4 KH domain containing RNA binding (QK1) single nucleotide polymorphisms (SNPs), their haplotypes, gene‐gene (G × G), gene‐environment (G × E) interactions and hypercholesterolaemia (HCH) and hypertriglyceridaemia (HTG) in the Chinese Maonan minority. The genetic make‐up of the SYNE1‐QK1 SNPs in 1932 unrelated subjects (normal, 641; HCH, 649; and HTG, 642) was obtained by next‐generation sequencing technologies. The genotypic frequencies of following SNPs were suggestively distinctive between the control and HCH groups (rs2623963, rs7745725, rs9459317, rs16897566), or between the control and HTG groups (rs2623963, rs1358317, rs7745725, rs1923608, rs16897566 SNPs; P < .05, respectively). Multiple‐locus linkage disequilibrium analysis indicated that the identified SNPs were not inherited independently. Several haplotypes and gene‐gene interaction haplotypes among the detected SNPs may be related with an increased morbidity of HCH (C‐G‐A, C‐G‐G and C‐G‐G‐T‐C‐A‐T) and HTG (C‐G‐G, G‐T‐G‐C, C‐G‐G‐G‐T‐G‐C and C‐G‐G‐T‐C‐A‐T), whereas others may be related with an decreased risk of HCH (G‐A‐A, G‐C‐A‐T, C‐A‐A‐T‐C‐A‐T and G‐A‐A‐G‐C‐A‐T) and HTG (G‐A‐A, G‐C‐A‐T, C‐A‐A‐T‐C‐A‐T and G‐A‐A‐G‐C‐A‐T). The association evaluation based on haplotypes and gene‐gene interactions could improve the power of detecting the risk of dyslipidaemia than anyone of SNP alone. There was significant three‐locus model involving SNP‐SNP, haplotype‐haplotype/environment and G × G interactions (P < .05‐0.001) that were detected by GMDR in HCH and HTG groups. Different interactions between genetic and environmental factors would produce different redundancy or synergy effects on the morbidity of HCH and/or HTG.


| INTRODUC TI ON
Coronary artery disease (CAD) has become the prominent reason of morbidity, disability, functional deterioration, mortality and costly health care, plus it is responsible for around 30% of all the deaths globally. [1][2][3] Preceding researches have showed that hyperlipidaemia acts as a major risk factor for CAD and its complications, which related to increased serum levels of total cholesterol (TC) and triglyceride (TG) 4 and which is a highly hereditary disease and 40%-60% of variation in blood lipid spectrums was genetically determined. 5,6 Several compelling researches showed that comprehensive lowering TC, 7 TG 7 and low-density lipoprotein cholesterol (LDL-C) 8 levels were more effective in reducing cardiovascular risk than lowering LDL-C levels alone. 9 Meanwhile, PCSK9 inhibitors are recommended as class I drugs to further reduce the risk of cardiovascular events in the acute phase of patients suffering from acute coronary syndrome (ACS) by 2019 ESC/EAS guidelines for the management of dyslipidaemias: Lipid modification to reduce cardiovascular risk. 10 Hence, identification of new lipid-related genes is important for guiding the treatment of hyperlipidaemia to further reducing cardiovascular risk.
Recently, several persuasive genes associated closely with blood lipid levels including the spectrin repeat containing nuclear envelope protein 1 (SYNE1) and KH domain containing RNA binding (QK1) have been reported by genome-wide association studies (GWASes) in the European population. 11 SYNE1 (also known as dJ45H2. Several studies have shown that the SYNE1 was powerfully related to the metabolism of serum TG, LDL-C and high-density lipoprotein cholesterol (HDL-C) levels. [12][13][14] Meanwhile, Sharma et al also documented that SYNE1 was related with inferior levels of apolipoprotein (Apo) A1 as well as HDL-C in severe septic patients. 15  Guangxi, 17 especially the Buyi, 18 is much closer than that between Maonan and Han ethnic group. 19 The marriage culture in Maonan is relatively conservative. Maonan keep the custom of intra-ethnic marriages, and intermarriage with other ethnic groups is rare. Thus, there was less heterogeneity about their genetic background in Maonan population, so that it is particularly suitable as a population to explore the genetic variation related to blood lipid. Therefore, the current research was designed (a) to evaluate the correlation of the SYNE1 (rs2623963, rs7745725, rs1358317) and QK1 (rs9459317, rs1764053, rs1923608, rs16897566) single nucleotide polymorphisms (SNPs) and blood lipid spectrums in participants with hypercholesterolaemia (HCH) and/or hypertriglyceridaemia (HTG); (b) to assess the connection of their haplotypes with the possibility of HCH/HTG; and (c) to recognize the potential G × G as well as G × E interactions among these variants in the Maonan population.

| Subjects
A total of 1932 discrete individuals (22- sex ratio between normal and HCH/HTG groups. Patients suffering from HCH did not have high triglyceride levels, and patients suffering from HTG also did not have high cholesterol levels, all subjects were independent and unrelated individuals. They were basically healthy and none of them had any history of type-2 diabetes mellitus (T2DM), CAD, ischaemic stroke or myocardial infarction. They were not taking any medicines that could alter the serum lipid levels.
Prior to the study, all subjects had signed informed consent. The protocol was authorized by the Ethics Committee of the First Affiliated Hospital, Guangxi Medical University.

| Epidemiological analysis
Universally standardized methods and protocols were used to conduct the epidemiological survey. 20 Using a standard set of questionnaires, detailed lifestyle as well as demographic characteristics were collected. Alcohol consumption was categorized into groups of grams of alcohol per day: 0 (non-drinker), ≤ 25 and > 25. 21 Smoking status was categorized into groups of cigarettes per day: 0 (non-smoker), ≤ 20 and > 20. 22 The inclusion criteria for smoking and drinking have been described in our previous epidemiological studies. 23,24 The alcohol information included questions about the number of grams of rice wine, wine, beer or liquor consumed during the preceding 12 months. Current smoking was defined as more than one cigarette per day. Participants who reported having smoked ≥ 100 cigarettes during their lifetime were classified as current smokers if they currently smoked and former smokers if they did not. 23,24 The blood pressure, body mass index (BMI), height, waist circumference and weight were measured as previous description. 25

| Biochemical assays
A fasting venous blood sample of 5 mL was taken from each subject. Two-fifths of the sample (2 mL) was used to measure serum lipid levels. The remaining three-fifths of sample (3 mL) was utilized to extract deoxyribonucleic acid (DNA). The methods of serum ApoA1, LDL-C, ApoB, TG, HDL-C and TC measurements were referred to a previous study. 26

| SNP selection
Following steps were utilized to select seven SNPs in SYNE1 and

| DNA amplification and genotyping
Genomic DNA was isolated from white blood cells in blood samples by phenol-chloroform method. 28,29 All extracted DNA samples were stored at 4°C until experiment. Genotyping of 7 SNPs was performed by the next-generation sequencing technology (NGS) at the Center for Human Genetics Research, Shanghai Genesky Bio-Tech Co. Ltd. 30 Primer sequences of the 7 SNPs are shown in Table S1.  4 The diagnostic criteria of overweight, obesity and normal weight 31 and hypertension, 32 were referred to our previous study.

| Statistical analyses
All experimental data were statistically assessed by means of SPSS (version 22.0). Values were manifested as mean ± SD, but the levels of TG were manifested as medians and interquartile ranges. Direct counting was used to determine allele frequency. Independent sample t test was implemented to assess the common characteristics between the two groups. Chi-square test was utilized to determine the genotype distribution among different groups. Hardy-Weinberg equilibrium (HWE), pairwise LD (measured by D′), haplotype frequencies and gene-gene interactions were analysed using Haploview.
Analysis of covariance (ANCOVA) was utilized to examine the correlation among serum lipid parameters and genotypes; and P < .0071 (equivalent to P < .05 after adjusting for 7 independent assessments by the Bonferroni correction) reflected statistical significance.
Unconditional logistic regression analysis was utilized to detect the associations among the haplotypes, genotypes and G × G interactions and the probability of HCH/HTG. Related parameters including smoking, gender, blood pressure, BMI, alcohol consumption, blood glucose and age were adjusted for the statistical evaluation. The best interaction combination among SNP-SNP, haplotype-haplotype/environment, gene-gene exposures were screened by generalized multifactor dimensionality reduction (GMDR). 33 The degree of cross-validation consistency was an effective method to identify the finest model among all considered probabilities. The score between 0.50 (representing that the model projects no better than chance) and 1.00 (representing impeccable projection) of cross-validation consistency was an indicator that precisely calculates the extent of case-control status. P < .05 represented statistical significance. Table 1 shows that the levels of ApoB, TG, LDL-C, TC, waist circumference, systolic blood pressure and the proportion of smokers, blood glucose, BMI, diastolic blood pressure, weight and pulse pressure were greater in HCH and HTG than in control groups (P < .05-0.001). The levels of serum HDL-C, ApoA1 and the ApoA1/ApoB ratio were less in HCH and HTG than in control groups (P < .05-0.001). There was no any obvious difference in the factors including age distribution, height, sex, alcohol consumption between the HCH/HTG and normal groups.

| Genotypic and allelic occurrences and the association with serum lipid levels
Seven SNPs of SYNE1-QK1 cluster were detected in a close genomic area of chromosome 6 ( Figure 1). The genotypic as well as allelic occurrences of the SYNE1 (rs2623963, rs7745725 and rs1358317) and QK1 (rs9459317, rs1764053, rs1923608 and rs16897566) SNPs are represented in Table 2. The genotype distribution of 7 SNPs met the Hardy-Weinberg equilibrium in HCH, HTG and control groups. The genotypic frequencies of following SNPs were suggestively distinctive between the control and HCH groups (rs2623963, rs7745725, rs9459317, rs16897566), or between the control and HTG groups (rs2623963, rs1358317, rs7745725, rs1923608, rs16897566 SNPs; P < .05). The dominant models of rs2623963, rs7745725 and rs9459317 SNPs showed an increased morbidity of HCH, whereas the dominant model of rs16897566 revealed a protective effect (P < .05). At the same time, the dominant models of rs2623963 and rs7745725 SNPs showed an increased morbidity of HTG, while the dominant models of rs1923608 and rs16897566 indicated a protective effect.

| Haplotype-based association with HCH/HTG
As shown in Figure 4, there was a strong pairwise LD among the detected loci in control, HCH as well as HTG groups. As shown in Table 3,

| Gene-gene interaction-based association with HCH as well as HTG
The commonest gene-gene interaction haplotype was the SYNE1-

| Gene-gene/environment interaction on hyperlipidaemia
GMDR was utilized to evaluate the association between the G × G/ G × E factor (comprising blood pressure (BP), age, drinking, BMI, glucose, smoking and sex) interactions and the risk of hyperlipidaemia, after adjusting for covariates. A significant three-locus model (a cross-validation constancy of 9 of 10, the assessment accurateness of 66.69%, # P < .001) comprising rs7745725, rs9459317 and rs16897566 SNPs was noticed in HCH group and another significant three-locus model (a cross-validation constancy of 9 of 10, the assessment accurateness of 58.53%, # P = .006) comprising rs2623963, rs1923608 and rs16897566 SNPs was noticed in HTG group (Table 5, representing a possible SNP-SNP interaction among the above SNPs).
In addition, other significant models including the haplotype-haplo-    Table S2.

| Relative factors for serum lipid parameters
As presented in Figure 7, the integrative variants and haplotypes linked with the SYNE1 rs2623963, rs1358317 and rs7745725 and QK1 rs9459317, rs1764053, rs1923608 and rs16897566 SNPs to lipid parameters. A series of environmental parameters, for example cigarette smoking, age, alcohol consumption, sex as well as common cardiovascular risk factors just as blood glucose, BP and BMI, were also related with blood lipid levels in control, HCH and HTG groups.

TA B L E 6 (Continued)
Maonan ethnic group is famous in China for its unique eating habits. Rice is the staple food for Maonan people. In addition, they also eat corn, potatoes, wheat, sorghum, etc. Maonan people especially prefer the foods of spicy, acid and rich in salt as well as oil.
This type of diet rich in long-term high-saturated fat might lead to high blood glucose levels, obesity, hyperlipidaemia, hypertension and atherosclerosis. 40 The main saturated long-chain fatty acid in the diet could produce harmful effects on serum lipid metabolism, especially impact the levels of serum TC and TG. 41 Unhealthy lifestyle including cigarette smoking and excessive drinking was directly related to the occurrence and development of hyperlipidaemia. 42 In the current research, we found that the percentages of participants who smoked were greater in HCH and HTG than in control groups. In recent years, the effect of smoking on hyperlipidaemia has gained more and more attention worldwide. Several recent reports have shown that there were lower serum HDL-C and greater serum TC, LDL-C and TG levels in smokers compared to non-smokers. 43,44 Moderate drinking reduced the incidence of cardiovascular events 45 ; and the potential mechanism may be related to increased HDL-C 46

F I G U R E 7
Correlation among environmental exposures, the genotypes and haplotypes of SYNE1-QK1 cluster and serum lipid variables in the control, HCH and HTG groups. TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; ApoA1/B, the ratio of apolipoprotein A1 to apolipoprotein B; BMI, body mass index, S1-S4 are combined with SYNE1 rs2623963-rs7745725-rs1358317, Q1-Q5 are combined with QK1 rs9459317-rs1764053-rs1923608-rs16897566 and H1-H7 are combined with SYNE1 rs2623963-rs7745725-rs1358317 and QK1 rs9459317-rs1764053-rs1923608-rs16897566 not recruited in this research; the genetic information of such patients may be different from those patients suffering from pure HCH or HTG. Consequently, although we have examined the effects of 7 SNPs in the SYNE1-QK1 cluster on lipid levels, there are numerous potential lipid-related SNPs have been overlooked in the current study. In addition, in order to further explore the molecular mechanism of the identified SNPs associated with the development of hyperlipidaemia, several further in-depth studies including incorporating the genetic information of single nucleotide mutations in SYNE1-QK1 cluster, their haplotypes, G × G and G × E interactions in vitro and in vivo functional researches are needed to confirm the impact of a variant on a molecular level.

| CON CLUS IONS
There were potential correlations between the SYNE1-QK1, environment exposures and serum lipid spectrums in the Maonan population. Moreover, the association evaluation based on haplotypes and gene-gene interactions could improve the power of detecting the risk of dyslipidaemia than anyone of SNP alone, and probably illuminated more changes of serum lipids especially for HDL-C.
When the GMDR was used to analyse the interactions, different interactions between gene and environment factors would produce different redundancy or synergy effects on the morbidity of HCH or HTG. In addition to genetic factors, the influence of environmental exposures on lipid levels would be an important factor that cannot be ignored.

ACK N OWLED G EM ENTS
This study was supported by the National Natural Science Foundation of China (No. 81460169).

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets generated during the present study are not publicly available, because detailed genetic information of each participant was included in these materials.