Skip to main content
Download PDF
Download ePub

Genome-wide association study for metabolic syndrome reveals APOA5 single nucleotide polymorphisms with multilayered effects in Koreans

Lipids in Health and Disease volume 23, Article number: 272 (2024) Cite this article

Abstract

Background and purpose

Genome-wide association studies (GWAS) of metabolic syndrome (MetS) have predominantly focused on non-Asian populations, with limited representation from East Asian cohorts. Moreover, previous GWAS analyses have primarily emphasized the significance of top single nucleotide polymorphisms (SNPs), poorly explaining other SNP signals in linkage disequilibrium. This study aimed to reveal the interaction between rs651821 and rs2266788, the principal variants of apolipoprotein A5 (APOA5), within the most significant loci identified through GWAS on MetS.

Methods

GWAS on MetS and its components was conducted using the data from the Korean Genome and Epidemiology Study (KoGES) city cohort comprising 58,600 individuals with available biochemical, demographic, lifestyle factors, and the most significant APOA5 locus was analyzed further in depth.

Results

According to GWAS of MetS and its diagnostic components, a significant association between the APOA5 SNPs rs651821/rs2266788 and MetS/triglycerides/high-density lipoprotein phenotypes was revealed. However, a conditional analysis employing rs651821 unveiled a reversal in the odds ratio for rs2266788. Therefore, rs651821 and rs2266788 emerged as independent and opposing signals in the extended GWAS analysis, i.e., the multilayered effects. Further gene-environment interaction analyses regarding lifestyle factors such as smoking, alcohol consumption, and physical activity underscored these multilayered effects.

Conclusion

This study unveils the intricate interplay between rs651821 and rs2266788 derived from MetS GWAS. Removing the influence of lead SNP reveals an independent protective signal associated with rs2266788, suggesting a multilayered effect between these SNPs. These findings underline the need for novel perspectives in future MetS GWAS.

Introduction

The National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) guideline defines metabolic syndrome (MetS) as an amalgam of metabolic abnormalities, including abdominal obesity, increased blood sugar levels, elevated blood pressure, and dyslipidemia [1]. Based on this definition, the prevalence of MetS was 34.2% in the United States and has been reported to increase throughout the 21st century [2]. In a nationwide Korean survey, approximately 30% of the population was diagnosed with MetS [3]. MetS is clinically important because it is associated with a 2-fold increased risk for cardiovascular diseases such as coronary artery disease and stroke as well as a 5-fold increased risk of type 2 diabetes [4]. As the prevalence of obesity and MetS has increased over the years, the healthcare costs for targeting these diseases have increased. Numerous potential treatments for MetS have been proposed; however, their effectiveness has been relatively moderate, thereby emphasizing the need for advanced precision medicine for clinical applications [5].

Family-based studies have estimated the heritability (h2) for each component of MetS to be 30–50% [6], with the remaining risk attributed to lifestyle and environmental factors such as diet, physical activity, and other non-genetic influences. The largest genome-wide association study (GWAS) on MetS has identified lipid-related (ZPR1, CETP, and LPL), obesity-related (FTO, MC4R, and VEGFA), diabetes-related (SLC18A1, ARAP2, and EXOC6), and hypertension-related (SIPA1) genes, and genes with pleiotropic effects simultaneously affecting two or more MetS components (GALNT2, FADS2, NEU2, and CMIP) [7].

Although previous GWAS have provided valuable insights into the genetic background of MetS, the unexplained heritability remains to be explored. One hypothesis that could explain this is that the contribution of individual variants to the phenotype has not been fully elucidated, possibly owing to limitations in statistical analyses considering linkage disequilibrium (LD). During the interpretation of GWAS results, only a small number of significant single nucleotide polymorphisms (SNPs) that are not within an LD relationship with each other have been selected for each locus, whereas the remaining SNPs have been ignored. However, certain variants that are located nearby within a locus, with a delicate LD relationship with each other, could exhibit independent effects on the phenotype [8]. Moreover, massive parallel reporter assays have revealed that genetic association signals can present through multiple tightly-linked causal variants [9].

Most of the previous GWASs on MetS have been conducted in non-Asian populations, and relatively few involved Korean populations [7, 10, 11]. The heterogeneous characteristics of MetS across populations make it challenging to identify the SNPs that contribute to the phenotype, let alone discover novel SNPs [12]. Certain ethnic groups, such as Hispanics or African Americans, may have a relatively high prevalence of MetS [13]. Furthermore, the prevalence and severity of specific MetS components may differ among populations. For instance, insulin resistance may be more prevalent in South Asians [14], whereas high blood pressure may be more pronounced in African Americans [15].

Therefore, this study strived to perform a GWAS on MetS and its components in a large Korean cohort to identify numerous significant loci that influence metabolic factors, and further investigate the role of SNPs in the APOA5 region, the most significant locus for MetS and two of its diagnostic constituents, plasma triglyceride (TG) and high-density lipoprotein (HDL) levels. Ultimately, the interaction of relevant SNPs with environmental lifestyle factors was analyzed.

Materials and methods

Study participants

The Korean Genome and Epidemiology Study (KoGES) is a large prospective cohort study designed to identify genetic and environmental factors and their interactions within common complex diseases such as MetS, osteoporosis, and cancer [16]. The KoGES cohort, in which health examinee participants were followed up from 2007 to 2013, mainly consisted of population-based and gene-environment interaction studies. All ethical considerations and requirements were satisfied, and the need for informed consent was waived because the participants were granted complete anonymity. The dataset used for the current analysis was the KoGES Health Examinee (HEXA) cohort, which mainly consisted of patients recruited from hospitals in major cities in South Korea. Based on the results of the pairwise identity-by-descent (IBD) and multidimensional scaling (MDS) analyses, individuals with familial relationships or those likely to be a cause of population stratification were excluded [17]. All participants were ≥ 40 years of age (starting population 58,700), and 100 participants with unavailable blood test results for TG or unavailable demographics were excluded. Consequently, 58,600 study participants with non-missing demographic variables and available blood test results were included. This study was approved by the Institutional Review Board (IRB) of the Center for Disease Control of the Korean government and the Seoul National University Hospital Clinical Research Institute (IRB No. C-2004-080-1117). The genotyping and phenotyping methods were performed according to the Declaration of Helsinki.

Phenotype measurements and definition of MetS

Eligible participants completed a questionnaire on demographic factors, comorbidities, and medications for hypertension, diabetes mellitus, and dyslipidemia. Blood tests, including the measurement of fasting blood sugar (FBS), TG, and HDL levels, were conducted using an automated analyzer after 12 h of fasting. MetS cases and controls were ascertained according to the guidelines of the NCEP-ATP III and the American Heart Association/National Heart Lung and Blood Institute (AHA/NHLBI) [18, 19]. The criterion for abdominal obesity was modified according to East Asian guidelines [20]. Patients who met three or more of the following criteria were selected: (1) waist circumference (WC) ≥ 90 cm for males and ≥ 80 cm for females; (2) TG levels ≥ 150 mg/dL; (3) HDL levels ≤ 40 mg/dL in males or ≤ 50 mg/dL in females; (4) blood pressure (BP) levels ≥ 130/85 mmHg or on hypertensive medication; and (5) elevated FBS levels ≥ 100 mg/dL or on medication for diabetes mellitus. Controls without MetS were defined as those who did not satisfy the MetS criteria, or satisfied two or less of the five MetS criteria.

Ascertainment of covariate variables

Based on their alcohol consumption, the participants were stratified into three groups: non- to low drinkers, moderate drinkers, and heavy drinkers. The dividing criteria for drinking status were assessed based on the frequency of alcohol consumption, volume of alcohol consumed each time, and type of alcohol consumed. Participants who drank ≥ 196 g alcohol per week were defined as heavy drinkers; those who drank 98–196 g of alcohol per week were defined as moderate drinkers; and those who drank < 98 g of alcohol per week were defined as non- to low drinkers. Based on their smoking status, the participants were classified as non- or never-smokers, ex-smokers who quit smoking in the past, or current smokers. Based on frequency and intensity, physical activity was stratified as low (no regular exercise), medium (regular exercise with sweating at a frequency of 1–4 times a week), or high (regular exercise with sweating at a frequency of ≥ 5 times a week).

Genotyping and imputation

The samples were genotyped using the KoreanChip version 1.1, a customized array optimized for the Korean population. Imputation was performed using the Northeast Asian Reference Database panel v.2. Comprehensive details of the genotyping and genetic imputation methods have been previously described [21, 22]. Finally, 7,857,085 variants with minor allele frequency ≥ 0.01, Hardy–Weinberg equilibrium P-value ≥ 1E–06, missing rate < 0.1, and estimated imputation accuracy from Minimac4 (R2) ≥ 0.3 were used for the GWAS analysis. The positions of all variants included in this study were based on hg19 reference genome.

Statistical analysis and visualization of results

Student’s t-test was performed for continuous variables and the chi-square test was performed for categorical variables to compare the groups with and without MetS. For smoking, alcohol consumption, and exercise intensity, the Cochran-Armitage trend test was used to evaluate the P-values corresponding to deviations from linear trends. GWAS assessed MetS and its five diagnostic criteria as binary variables. The genetic association between MetS and its five diagnostic criteria, adjusted for age and sex, using logistic regression analysis was implemented via PLINK (version 1.9; Free Software Foundation Inc., Boston, MA, USA). All genotyping data were analyzed using the computing server at the Genomic Medicine Institute Research Service Center. Power analyses and sample size calculations were also performed, in which TG values were calculated for the rs651821 and rs2266788 genotypes. With alpha set to 0.05 and power set to 0.8, the required sample size was 1,158 people, which was well below the population number of the current study or that of any genotype subgroup (minimum subgroup population was 2,857 for rs2266788 GG genotype).

The Genome-wide Complex Trait analysis package, Conditional and Joint association analysis using GWAS summary statistics (GCTA-COJO) [23], was used to extract additional statistically significant and independent variants beyond the most significant variants at each locus contributing to each specific phenotype of BP, WC, FBS, TG, HDL, and MetS. ANNOVAR was used to annotate each variant with information regarding function, neighboring genes, and population frequency [24]. Fine-mapping was performed using POLYgenic FUNctionally informed fine-mapping (PolyFUN), implementing the Sum of Single Effects(SuSiE) statistical model [25]. From these methods, the credible set was restricted to a total of 10 sets for each analysis, and posterior inclusion probability (PIP) values were calculated for each section of ± 100 kilobases for each independent SNP derived from GCTA-COJO. This revealed the PIP values for each SNP. A PIP > 0.9 was defined as a high probability of causality, and a PIP 0.1 cutoff was applied to achieve minimum causality. Fine-mapping was performed for each of the UK Biobank and KoGES city cohorts, with the TG binary variable (cutoff of 150 mg/dL) as the phenotype.

Baseline demographics were analyzed using STATA v.17.0 (StataCorp LLC, College Station, TX, USA). The R package ‘qqman’ was used for the GWAS Manhattan plot and qqplot visualization [26]. The genome-wide significance level for the P-value was set at 5E-08. LocusZoom plot was used for magnifying the view of major SNPs of APOA5 on chromosome 11 [27].

Results

Genetic associations from MetS GWAS and its components

Among the 58,600 KoGES study population, approximately 24% of the total study population satisfied 3 or more out of the total 5 MetS diagnostic criteria. Age, sex, body mass index, and lifestyle factors such as smoking, drinking, and physical activity differed significantly between the patients with and without MetS (Table 1). GWAS and conditional and joint analysis (COJO) for MetS revealed 11 independent signals: APOA5 on chromosome 11 was the most significant signal (rs651821, odds ratio [OR] = 1.37, P-value = 1.25E-98), which was followed by CETP on chromosome 16 (rs17231506, OR = 0.83, P-value = 2.57E-23), and LPL on chromosome 8 (rs59147390, OR = 0.83, P-value = 1.00E-18) (Fig. 1; Table 2). GWAS and COJO performed on each of the five MetS diagnostic criteria (Figure S1, Table S1) showed that the number of independent signals that reached a genome-wide significance level was 16 for BP, four for WC, 25 for FBS, 18 for TG, 34 for HDL (a total of 97 SNPs for MetS diagnostic components) (Table S1). APOA5, BUD13, LPL, and APOE loci were all associated with MetS/TG/HDL (Figure S2). These genes can be functionally explained or have been identified in GWAS related to lipid metabolism [28,29,30,31]. Linkage disequilibrium score (LDSC) was also calculated for MetS GWAS. The genomic inflation factor (λGC) was 1.083 (Fig. 1), LDSC intercept was 1.010 (standard error (SE) = 0.004), LDSC ratio was 0.099 (SE = 0.042); implying that polygenicity, but not confounding factors, such as population stratification or cryptic relatedness, was the principal cause of genomic inflation [32]. In addition to age/sex adjustment, further adjustment of lifestyle factors(smoking, alcohol drinking, physical activity) and 5 genetic principal components in GWAS analyses revealed minute differences in the results (Table S2).

Table 1 Baseline characteristics of the study population (N = 58,600)
Full size table
Fig. 1
figure 1

Manhattan and Quantile-Quantile plot of the genome-wide association study of metabolic syndrome(MetS) in Koreans. The x-axis represents the SNP markers on each chromosome. y-axis shows the minus log10 scale. The red horizontal line represents the genome-wide significant threshold P = 5E-08 and the blue horizontal line represents the genome-wide suggestive threshold P = 1E-05. Five candidate loci of MetS are shown

Full size image
Table 2 MetS associated independent SNPs in the KoGES cohort grouped by each locus
Full size table

Based on the COJO-derived variants from GWAS of MetS components, 12 exonic and non-synonymous SNPs were identified (Table S3). For the 11 independent signals derived from MetS GWAS (Table 2), the genetic association of each variant for each of the five MetS components was further illustrated (Figure S3, Table S4). The most significant APOA5 locus in the MetS GWAS was also the most significant in TG and HDL GWAS (Figure S4). Although differences in the magnitude of statistical significance were noticed, HDL and TG levels overlapped most with the results of MetS GWAS, thereby corroborating the findings of a previous study indicating that lipid traits considerably contribute to the MetS phenotype [33]. In addition, among the 11 MetS signals, one overlapped with BP (SNHG32), three overlapped with WC (SNHG32, MTNR1B, PEPD), and two overlapped with FBS (MTNR1B, PEPD), all with the same OR valence except for rs731849 (PEPD) in WC.

Relationship between APOA5 variants and lipid levels

Regarding the APOA5 signal, which was most significant for MetS, TG, and HDL, rs651821 was the most significant variant for MetS and TG, whereas rs662799 was the most significant variant for HDL, with an LD (r2) > 0.99 between the two variants (Figure S4). In addition to rs651821 and rs662799, there were other significant variants within the APOA5 locus, including rs2072560 and rs2266788 (Table 3). The relationship between these variants and TG/HDL levels was also analyzed; when these variants were adjusted for rs651821, their OR directions were reversed for both lipid traits. This particular multilayered phenomenon was further replicated in a separate KoGES rural cohort (N = 8,102), where the OR for rs2266788 was reversed when conditioned with the top rs651821 SNP (rs2266788 unconditioned OR/beta/P-value 1.38/0.14/2.44E-16, rs651821 conditioned rs2266788 OR/beta/P-value 0.78/-0.11/1.63E-04). The LD (r2) between rs651821-rs2072560 and rs651821-rs2266788 was 0.65 each, whereas rs2072560 and rs2266788 were highly correlated with each other(r2 = 0.995). The P-value of epistatic interaction for rs651821-rs2266788 SNPs was 0.067, thereby excluding the possibility of SNP-SNP pairwise epistatic effects. rs2072560 is located in the intron of APOA5, whereas rs2266788 is located in the 3’-untranslated region(UTR) of APOA5; rs2266788 has been reported in ClinVar [34] as a potential risk variant for familial hypertriglyceridemia [35,36,37] (ClinVar Accession number SCV000148905).

Table 3 GWAS and conditioned results of APOA5 major SNPs on TG and HDL
Full size table

Inter-population differences in genetic associations and fine-mapping of KoGES and UK Biobank cohorts

Within the 11 COJO-derived independent signals of the MetS GWAS, no cases were observed where signals remained significant but had reversed ORs after conditioning the top SNPs of each locus, except for rs651821. Within the UK Biobank cohort, conditional analyses of rs651821 and rs2266788 revealed results distinct from those of KoGES, where OR directions were maintained even after conditional analysis. To determine the mechanisms underlying these different results across populations, fine-mapping of each cohort by classifying plasma triglyceride values as a binary variable (150 mg/dL according to the MetS criteria) was performed which revealed distinct characteristics because no overlapping SNPs were observed among SNPs with PIP values > 0.9 (Figure S5). In the KoGES cohort, rs651821 located in APOA5 locus on chromosome 11 was an independent SNP and potential causal variant in Koreans, with a PIP value of 0.998 (Table S5). However, when the same analysis was performed using UK Biobank data, rs964184, located at the 3’-UTR of ZPR1, was identified as the top SNP (Figure S6) with a high probability of causality (Table S6) whereas rs651821 had a PIP value of approximately zero and was not included in the ten derived credible sets. Nevertheless, rs4938311 located in BUD13, was the only overlapping SNP with a PIP > 0.1 in both cohorts. Further, LD blocks of KoGES and UK Biobank cohorts at the APOA5 locus showed the inter-population differences in genetic architecture (Figure S7).

Association between genotypes of rs651821 and rs2266788 with the prevalence of MetS, and serum TG and HDL levels

The prevalence of MetS were 21%, 26%, and 34% for TT, CT, and CC genotypes, respectively, within rs651821 (Fig. 2A, Table S7). MetS prevalence was 23%, 26%, and 31% for AA, AG, and GG genotypes, respectively, within rs2266788 (Fig. 2B). However, within the rs651821 CC subgroup, MetS prevalence were 45%, 35%, and 31% for AA, AG, and GG genotypes, respectively, within rs2266788, indicating a multilayered effect between rs651821 and rs2266788 (Fig. 2C). Serum TG levels, according to the rs651821 genotype, increased as the minor allele (C allele) count increased (Fig. 2D). Similarly, serum TG levels, according to the rs2266788 genotype, significantly increased as the minor allele (G allele) count increased (Fig. 2E). In contrast, in the rs651821 CC genotype subgroup (homozygous minor allele subgroup), serum TG levels decreased as the minor allele count of rs2266788 increased (Fig. 2F). Similarly, the HDL levels decreased when the minor allele count of rs651821 increased (from TT to CT to CC) (Fig. 2G) or when the minor allele count of rs2266788 increased (from AA to AG to GG) (Fig. 2H). However, within the rs651821 CC subgroup, the HDL levels significantly increased when the minor allele count of rs2266788 increased (from AA to AG to GG) (Fig. 2I).

Fig. 2
figure 2

Multilayered effects of rs651821 and rs2266788 in MetS prevalence, TG and HDL. Metabolic syndrome prevalence percentage (A-C), plasma TG (D-F) and plasma HDL (G-I) levels according to rs651821 and rs2266788 genotypes. Corresponding values according to rs651821 genotypes (A, D, G), rs2266788 genotypes (B, E, H), rs2266788 genotypes within rs651821 CC subgroup (C, F, I) are depicted.

Full size image

Gene-environment interactions in metabolic factors

Because metabolic factors are influenced by genetic and environmental factors, three representative lifestyle environmental factors, such as smoking, alcohol consumption, and physical activity, were examined to evaluate gene-environment interactions. Regarding smoking habits, TG levels were lowest in non-smokers, followed by ex-smokers and current smokers. When stratified into genotypes, the minor homozygous rs651821 CC subgroup and rs2266788 GG subgroup were associated with the highest TG levels (Fig. 3A and B). However, within the rs651821 CC subgroup, the rs2266788 AA subgroup was associated with the highest TG levels, which were all higher than those of the rs2266788 GG subgroup (Fig. 3C). The overall trend of TG levels increased from moderate to heavy drinkers; however, the rs2266788 correlation of AA-AG-GG was reversed in the rs651821 CC genotype subgroup (Fig. 3D, E and F). This reversal of TG levels for the rs651821 CC subgroup is in line with data presented in Fig. 2, except that the additional analysis of environmental factors strengthened the putative hypothesis. Individuals with rs651821 CC and rs2266788 AA genotypes have significantly elevated TG levels when smoking (Fig. 3C) or excessively drinking (Fig. 3F), but with highly relatively highly increasing TG slopes compared to those of the rs651821 genotypes analyzed as a whole (Fig. 3B and E). Further, within the rs651821 CC subgroup, TG levels of the rs2266788 AA subgroup increased when the extent of physical activity increased (Fig. 3I). HDL levels were also analyzed for gene-environment interactions and showed similar results regarding rs651821 and rs2266788 with respect to lifestyle factors (Figure S8). Particularly, regarding physical activity, ‘beneficial’ HDL cholesterol levels increased with increasing extents of physical activity within rs651821 (Figure S8G) and rs2266788 (Figure S8H) genotypes analyzed as a whole, specifically conspicuous within the rs2266788 AA subgroup. However, within the rs651821 CC and rs2266788 AA subgroups (Figure S8I), high extent of physical activity conversely reduced HDL cholesterol levels, suggesting an interaction between genetic and environmental factors in combination with the multilayered effects.

Fig. 3
figure 3

Effects of multilayered SNPs in MetS component TG levels and their interactions with lifestyle factors. TG levels according to rs651821, rs2266788 genotypes associated with stratified levels of smoking (A-C), alcohol drinking (D-F) and physical activity (G-I). The purple arrows referring to the boxes (C, F, I) indicate minor homozygous subgroup of rs651821 CC genotype. All error bars represent standard errors. PA: physical activity

Full size image

Discussion

In this study, MetS GWAS of 58,600 Koreans from the KoGES database was performed and 11 independent signals with genome-wide significance were identified. Additional analysis uncovered a complex relationship between the two SNPs (rs651821 and rs2266788), which was not apparent in a single GWAS, and revealed the actual direction of the effect of rs2266788 on MetS and lipid traits. These two SNPs located within the APOA5 locus, each 5’-UTR and 3’-UTR variants respectively, have been identified as possible controllers of fasting and postprandial TG levels and recently came into the spotlight in genetic studies involving humans and mice [38]. In addition, this study considered environmental factors such as alcohol consumption, smoking status, and exercise frequency and assessed the relationship between these factors and different genotypes. Previously, MetS GWAS has been performed on the Korean population [39, 40]. Although these studies uncovered previously unidentified SNPs, they had a limited sample size of a few thousand individuals, resulting in limited study scope. Furthermore, a previous study [41] has also investigated the interaction between APOA5 rs662799 and rs2266788, which collectively influenced TG concentrations, and lifestyle factors such as carbohydrates, calcium intake, and smoking; however, they did not explore the intricate relationship between the two SNPs.

The APOA5 locus located in chromosome 11 was the most significant loci for all MetS, TG, and HDL GWAS. The underlying mechanism, through which increased APOA5 expression leads to decreased TG levels has been previously described [36, 42,43,44]. Regarding apoav, which is the mouse equivalent of APOA5, transgenic knockout mice have shown 4-fold higher TG levels than those in control mice, whereas APOA5-overexpressing mice had one-third lower TG concentrations than those of the controls [36]. The apoav lipid complexes bind to heparin, thereby indirectly affecting lipoprotein lipase (LPL) activity in vitro [44]. APOA5 stimulates the LPL-mediated hydrolysis of lipoproteins, for instance very low-density lipoprotein (VLDL) and chylomicrons, which are inhibited by APOE [43]. A study using immunoprecipitation and mass spectrometry reported that APOA5 lowers TG levels by suppressing ANGPTL3/8-mediated inhibition of LPL in the human serum [42].

In addition to APOA5, GWAS on MetS and its diagnostic components revealed numerous significant signals associated with two or more MetS components that should be functionally mentioned. CETP, associated with MetS/HDL, encodes a cholesteryl ester transfer protein that transfers cholesteryl ester and TG between HDL, LDL, and VLDL [45]. Furthermore, rs10830963 located within MTNR1B, a gene encoding melatonin receptor 1B, has been associated with MetS/FBS [46]. MTNR1B is related to circadian rhythms and is implicated in mediating the inhibitory effect of melatonin on insulin secretion. BRAP, associated with HDL/FBS, encodes the breast cancer suppressor protein-associated protein (BRAP) and is associated with obesity and metabolic factors in an East Asian population [47]. GCKR is highly associated with TG among MetS components [48], which is in line with the current study’s conclusion regarding the association of GCKR with TG/FBS levels. Increased activity of the GCKR influences glucokinase to regulate the first step of glycolysis. Glycolysis provides acetyl-CoA intermediates, thereby fueling the synthesis of fatty acids, which leads to increased lipogenesis in the liver and, in this case, TG [49].

Although rs651821 and rs2266788 were both identified in previous TG GWAS, rs651821 has received much attention owing to its high significance. GATA4 inclines to bind to the protective T allele of rs651821, thereby increasing APOA5 expression in liver cells and in individuals with the TT genotype, which ultimately leads to decreased TG levels [50]. Each genomic variant may affect the expression of a certain gene through the enhancer/promoter activities or mRNA stability to influence a certain phenotype [51]. Whether rs2266788 had enhancer activity towards APOA5 remained controversial [52]. Cui et al. [53] have reported that the G allele of this 3’-UTR variant of APOA5 increases APOA5 expression by maintaining more stability of APOA5 mRNA than does the A allele. Alternatively, the A allele of rs2266788 enhanced the degradation of APOA5 mRNA, and APOA5 mRNA with the G allele has a longer half-life than that with the A allele. Therefore, these results concur with those of the present study, rendering the G allele of rs2266788 protective towards both MetS and lipid levels. Based on these insights, both variants rs651821 and rs2266788 appear to be functional and to independently contribute to MetS.

Strengths

This is the first report of multilayered effects of rs651821 and rs2266788, SNPs of APOA5, regarding MetS, TG and HDL. Additionally, we identified ethnic differences of the APOA5 region. To elaborate, comparing the APOA5 locus for TG levels in the UK Biobank cohort, we found that the most significantly associated SNP and overall genetic structure were different between populations. Furthermore, we performed gene-environment interaction analysis of lifestyle factors, such as smoking, alcohol consumption, and physical activity, and identified that multilayered effects were relatively strong in current smokers and heavy drinkers. These results suggest that the risk assessment of metabolic disease development should be tailored according to personal genotypes and lifestyle factors.

Limitations

The limitation of this study is that although GWAS on MetS and its components was performed, previously undiscovered loci could not be identified. Instead, this study focused on the most significant loci of MetS, TG, and HDL and identified the multilayered effect of two SNPs (rs651821 and rs2266788). Furthermore, the gene-environment interaction analysis performed herein reports observational characteristics; therefore establishing causal relationships is difficult, and accuracy issues remain owing to the self-reported nature of the three main lifestyle factors. Other environmental factors such as dietary habits or sleep patterns should also be explored in future studies. Last, functional validation studies, such as allele-specific differences in APOA5 expression for relevant cell types and tissues and in vivo experiments using mouse models associated with MetS, are necessary to support the results presented herein.

Conclusions

GWAS on MetS and its components revealed the most influential SNPs regarding MetS and its components. The effect of APOA5 SNPs rs651821 and rs2266788 on MetS and serum lipid levels was analyzed, and gene-environment interaction observations underscored the multilayered effects between the APOA5 SNPs. From the perspective of precision medicine, personalized genotyping and counseling of the general population could be beneficial and could inform patients how to enhance lifestyle factors and ultimately prevent or overcome MetS.

Data availability

This study was conducted using bioresources from the National Biobank of Korea and the Centers for Disease Control and Prevention Agency (KDCA), Republic of Korea (KBN-2020-029). The KoGES data utilized for this study are not available to the public, but can be requested from the KDCA website www.kdca.go.kr with an appropriate research application form and separate institutional review board consent for the study. This research has also been conducted using the UK Biobank Resource under Application Number 97642.

Abbreviations

APOA5 :

Apolipoprotein A5

BMI:

Body mass index

BP:

Blood pressure

FBS:

Fasting blood sugar

GCTA:

COJO-Genome-wide complex trait analysis-conditional and Joint multiple SNP association analysis

GWAS:

Genome-wide association study

HDL:

High-density lipoprotein cholesterol

HEXA:

Health examinee study

KoGES:

Korean genome and epidemiology study

LD:

Linkage disequilibrium

MetS:

Metabolic syndrome

PIP:

Posterior inclusion probability

SNP:

Single nucleotide polymorphism

TG:

Triglycerides

UTR:

Untranslated region

WC:

Waist circumference

References

  1. Expert Panel on Detection E. Treatment of high blood cholesterol in A. Executive Summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA. 2001;285(19):2486–97.

    Article  Google Scholar 

  2. Gurka MJ, Filipp SL, DeBoer MD. Geographical variation in the prevalence of obesity, metabolic syndrome, and diabetes among US adults. Nutr Diabetes. 2018;8(1):14.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Ranasinghe P, Mathangasinghe Y, Jayawardena R, Hills AP, Misra A. Prevalence and trends of metabolic syndrome among adults in the asia-pacific region: a systematic review. BMC Public Health. 2017;17(1):101.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Saklayen MG. The global epidemic of the metabolic syndrome. Curr Hypertens Rep. 2018;20(2):12.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Grundy SM, Hansen B, Smith SC Jr., Cleeman JI, Kahn RA, American Heart A et al. Clinical management of metabolic syndrome: report of the American Heart Association/National Heart, Lung, and Blood Institute/American Diabetes Association conference on scientific issues related to management. Circulation. 2004;109(4):551-6.

  6. van Dongen J, Willemsen G, Chen WM, de Geus EJ, Boomsma DI. Heritability of metabolic syndrome traits in a large population-based sample. J Lipid Res. 2013;54(10):2914–23.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Lind L. Genome-wide Association study of the metabolic syndrome in UK Biobank. Metab Syndr Relat Disord. 2019;17(10):505–11.

    Article  CAS  PubMed  Google Scholar 

  8. Roman TS, Marvelle AF, Fogarty MP, Vadlamudi S, Gonzalez AJ, Buchkovich ML, et al. Multiple hepatic Regulatory variants at the GALNT2 GWAS Locus Associated with High-Density Lipoprotein Cholesterol. Am J Hum Genet. 2015;97(6):801–15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Abell NS, DeGorter MK, Gloudemans MJ, Greenwald E, Smith KS, He Z, et al. Multiple causal variants underlie genetic associations in humans. Science. 2022;375(6586):1247–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Lee HS, Kim Y, Park T. New Common and Rare variants Influencing Metabolic Syndrome and its Individual Components in a Korean Population. Sci Rep. 2018;8(1):5701.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Povel CM, Boer JM, Reiling E, Feskens EJ. Genetic variants and the metabolic syndrome: a systematic review. Obes Rev. 2011;12(11):952–67.

    Article  CAS  PubMed  Google Scholar 

  12. Brown AE, Walker M. Genetics of insulin resistance and the metabolic syndrome. Curr Cardiol Rep. 2016;18(8):75.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Kirby JB, Liang L, Chen HJ, Wang Y. Race, place, and obesity: the complex relationships among community racial/ethnic composition, individual race/ethnicity, and obesity in the United States. Am J Public Health. 2012;102(8):1572–8.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Gujral UP, Kanaya AM. Epidemiology of diabetes among South asians in the United States: lessons from the MASALA study. Ann N Y Acad Sci. 2021;1495(1):24–39.

    Article  PubMed  Google Scholar 

  15. Zilbermint M, Hannah-Shmouni F, Stratakis CA. Genetics of Hypertension in African americans and others of African descent. Int J Mol Sci. 2019;20(5).

  16. Kim Y, Han BG, Ko GES. Cohort Profile: the Korean Genome and Epidemiology Study (KoGES) Consortium. Int J Epidemiol. 2017;46(2):e20.

    Article  PubMed  Google Scholar 

  17. Moon S, Kim YJ, Han S, Hwang MY, Shin DM, Park MY, et al. The Korea Biobank array: design and Identification of Coding Variants Associated with blood biochemical traits. Sci Rep. 2019;9(1):1382.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Benedict M. NCEP’s new guidelines at a glance. National Cholesterol Education Program. Health Care Food Nutr Focus. 2002;18(12):8–9.

    PubMed  Google Scholar 

  19. Skilton MR, Moulin P, Serusclat A, Nony P, Bonnet F. A comparison of the NCEP-ATPIII, IDF and AHA/NHLBI metabolic syndrome definitions with relation to early carotid atherosclerosis in subjects with hypercholesterolemia or at risk of CVD: evidence for sex-specific differences. Atherosclerosis. 2007;190(2):416–22.

    Article  CAS  PubMed  Google Scholar 

  20. Hong AR, Lim S. Clinical characteristics of metabolic syndrome in Korea, and its comparison with other Asian countries. J Diabetes Investig. 2015;6(5):508–15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Yoo SK, Kim CU, Kim HL, Kim S, Shin JY, Kim N, et al. NARD: whole-genome reference panel of 1779 Northeast asians improves imputation accuracy of rare and low-frequency variants. Genome Med. 2019;11(1):64.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Choi J, Kim S, Kim J, Son HY, Yoo SK, Kim CU, et al. A whole-genome reference panel of 14,393 individuals for east Asian populations accelerates discovery of rare functional variants. Sci Adv. 2023;9(32):eadg6319.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Yang J, Ferreira T, Morris AP, Medland SE, Genetic Investigation of ATC, Replication DIG et al. Conditional and joint multiple-SNP analysis of GWAS summary statistics identifies additional variants influencing complex traits. Nat Genet. 2012;44(4):369 – 75, S1-3.

  24. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010;38(16):e164.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Zou Y, Carbonetto P, Wang G, Stephens M. Fine-mapping from summary data with the Sum of single effects model. PLoS Genet. 2022;18(7):e1010299.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Robinson JT, Thorvaldsdottir H, Winckler W, Guttman M, Lander ES, Getz G, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29(1):24–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Pruim RJ, Welch RP, Sanna S, Teslovich TM, Chines PS, Gliedt TP, et al. LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics. 2010;26(18):2336–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Su X, Kong Y, Peng DQ. New insights into apolipoprotein A5 in controlling lipoprotein metabolism in obesity and the metabolic syndrome patients. Lipids Health Dis. 2018;17(1):174.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Sylvers-Davie KL, Davies BSJ. Regulation of lipoprotein metabolism by ANGPTL3, ANGPTL4, and ANGPTL8. Am J Physiol Endocrinol Metab. 2021;321(4):E493–508.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Yang LG, March ZM, Stephenson RA, Narayan PS. Apolipoprotein E in lipid metabolism and neurodegenerative disease. Trends Endocrinol Metab. 2023;34(8):430–45.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Zhang L, You Y, Wu Y, Zhang Y, Wang M, Song Y, et al. Association of BUD13 polymorphisms with metabolic syndrome in Chinese population: a case-control study. Lipids Health Dis. 2017;16(1):127.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Wendt FR, Pathak GA, Vahey J, Qin X, Koller D, Cabrera-Mendoza B, et al. Modeling the longitudinal changes of ancestry diversity in the million veteran program. Hum Genomics. 2023;17(1):46.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Marzoog B. Lipid behavior in metabolic syndrome pathophysiology. Curr Diabetes Rev. 2022;18(6):e150921196497.

    Article  CAS  PubMed  Google Scholar 

  34. Landrum MJ, Lee JM, Riley GR, Jang W, Rubinstein WS, Church DM, et al. ClinVar: public archive of relationships among sequence variation and human phenotype. Nucleic Acids Res. 2014;42(Database issue):D980–5.

    Article  CAS  PubMed  Google Scholar 

  35. Caussy C, Charriere S, Marcais C, Di Filippo M, Sassolas A, Delay M, et al. An APOA5 3’ UTR variant associated with plasma triglycerides triggers APOA5 downregulation by creating a functional mir-485-5p binding site. Am J Hum Genet. 2014;94(1):129–34.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Pennacchio LA, Olivier M, Hubacek JA, Cohen JC, Cox DR, Fruchart JC, et al. An apolipoprotein influencing triglycerides in humans and mice revealed by comparative sequencing. Science. 2001;294(5540):169–73.

    Article  CAS  PubMed  Google Scholar 

  37. Pennacchio LA, Olivier M, Hubacek JA, Krauss RM, Rubin EM, Cohen JC. Two independent apolipoprotein A5 haplotypes influence human plasma triglyceride levels. Hum Mol Genet. 2002;11(24):3031–8.

    Article  CAS  PubMed  Google Scholar 

  38. May-Zhang L, Liu M, Black D, Tso P. Apolipoprotein A5, a unique modulator of fasting and postprandial triglycerides. Biochim Biophys Acta Mol Cell Biol Lipids. 2022;1867(9):159185.

    Article  CAS  PubMed  Google Scholar 

  39. Jeong SW, Chung M, Park SJ, Cho SB, Hong KW. Genome-wide association study of metabolic syndrome in koreans. Genomics Inf. 2014;12(4):187–94.

    Article  Google Scholar 

  40. Oh SW, Lee JE, Shin E, Kwon H, Choe EK, Choi SY, et al. Genome-wide association study of metabolic syndrome in Korean populations. PLoS ONE. 2020;15(1):e0227357.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Park S, Kang S, Alcohol. Carbohydrate, and Calcium intakes and Smoking interactions with APOA5 rs662799 and rs2266788 were Associated with elevated plasma triglyceride concentrations in a cross-sectional study of Korean adults. J Acad Nutr Diet. 2020;120(8):1318–29. e1.

    Article  PubMed  Google Scholar 

  42. Chen YQ, Pottanat TG, Zhen EY, Siegel RW, Ehsani M, Qian YW, et al. ApoA5 lowers triglyceride levels via suppression of ANGPTL3/8-mediated LPL inhibition. J Lipid Res. 2021;62:100068.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Kersten S. Physiological regulation of lipoprotein lipase. Biochim Biophys Acta. 2014;1841(7):919–33.

    Article  CAS  PubMed  Google Scholar 

  44. Lookene A, Beckstead JA, Nilsson S, Olivecrona G, Ryan RO. Apolipoprotein A-V-heparin interactions: implications for plasma lipoprotein metabolism. J Biol Chem. 2005;280(27):25383–7.

    Article  CAS  PubMed  Google Scholar 

  45. Wong SK, Ramli FF, Ali A, Ibrahim N. Genetics of cholesterol-related genes in metabolic syndrome: a review of current evidence. Biomedicines. 2022;10(12).

  46. Prokopenko I, Langenberg C, Florez JC, Saxena R, Soranzo N, Thorleifsson G, et al. Variants in MTNR1B influence fasting glucose levels. Nat Genet. 2009;41(1):77–81.

    Article  CAS  PubMed  Google Scholar 

  47. Imaizumi T, Ando M, Nakatochi M, Yasuda Y, Honda H, Kuwatsuka Y, et al. Effect of dietary energy and polymorphisms in BRAP and GHRL on obesity and metabolic traits. Obes Res Clin Pract. 2018;12(Suppl 2):39–48.

    Article  PubMed  Google Scholar 

  48. Zahedi AS, Akbarzadeh M, Sedaghati-Khayat B, Seyedhamzehzadeh A, Daneshpour MS. GCKR common functional polymorphisms are associated with metabolic syndrome and its components: a 10-year retrospective cohort study in Iranian adults. Diabetol Metab Syndr. 2021;13(1):20.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Shen Y, Wu L, Xi B, Liu X, Zhao X, Cheng H, et al. GCKR variants increase triglycerides while protecting from insulin resistance in Chinese children. PLoS ONE. 2013;8(1):e55350.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Chou WC, Chen WT, Shen CY. A common variant in 11q23.3 associated with hyperlipidemia is mediated by the binding and regulation of GATA4. NPJ Genom Med. 2022;7(1):4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Shastry BS. SNPs: impact on gene function and phenotype. Methods Mol Biol. 2009;578:3–22.

    Article  CAS  PubMed  Google Scholar 

  52. Talmud PJ, Palmen J, Putt W, Lins L, Humphries SE. Determination of the functionality of common APOA5 polymorphisms. J Biol Chem. 2005;280(31):28215–20.

    Article  CAS  PubMed  Google Scholar 

  53. Cui G, Li Z, Li R, Huang J, Wang H, Zhang L, et al. A functional variant in APOA5/A4/C3/A1 gene cluster contributes to elevated triglycerides and severity of CAD by interfering with microRNA 3201 binding efficiency. J Am Coll Cardiol. 2014;64(3):267–77.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

This research was supported by a grant from the MD-PhD/Medical Scientist Training Program through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea, Phase III (Postdoctoral fellowship) grant from the SPST (SNU-SNUH Physician Scientist Training) program, the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (2020R1A6A1A03047972 and 2020R1I1A1A01074772), and 2023 Research Grant from Kangwon National University (202304840001).

Funding

The funding sources had no role in the design and conduction of the study; analysis and interpretation of the data; and preparation, review, or approval of the manuscript.

Author information

Author notes

  1. Young Jun Park and Sungji Moon contributed equally to this work.

Authors and Affiliations

  1. Genomic Medicine Institute, Medical Research Center, Seoul National University, Seoul, 03080, Republic of Korea

    Young Jun Park, Jaeyong Choi, Ho-Young Son & Jong-Il Kim

  2. Department of Translational Medicine, Seoul National University College of Medicine, Seoul, 03080, Republic of Korea

    Young Jun Park

  3. Interdisciplinary Program in Cancer Biology, Seoul National University College of Medicine, Seoul, 03080, Republic of Korea

    Sungji Moon

  4. Cancer Research Institute, Seoul National University, Seoul, 03080, Republic of Korea

    Sungji Moon & Jong-Il Kim

  5. Department of Biomedical Sciences, Seoul National University College of Medicine, Seoul, 03080, Republic of Korea

    Juhyun Kim & Jong-Il Kim

  6. National Cancer Control Institute, National Cancer Center, Goyang-Si, Gyeonggi-Do, 10408, Republic of Korea

    Hyun-Jin Kim

  7. Department of Biochemistry and Molecular Biology, Seoul National University College of Medicine, 103 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea

    Ho-Young Son & Jong-Il Kim

  8. Department of Biochemistry and Molecular Biology, Kangwon National University School of Medicine, One Kangwondaehak-gil, Chuncheon-si, Gangwon-do, 24341, Republic of Korea

    Sun-Wha Im

Authors
  1. Young Jun Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Sungji Moon
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Jaeyong Choi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Juhyun Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Hyun-Jin Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ho-Young Son
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Sun-Wha Im
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Jong-Il Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

J-I.K, S-W.I, and H-Y.S had full access to all data in the study and take responsibility for the integrity and accuracy of the data analysis. They also co-supervised this study. Y.J.P and S.M are co-first authors. Concept and design: Y.J.P, S.M, H-Y.S, and S-W.I. Acquisition, analysis, or interpretation of data: Y.J.P, S.M, and J.C. Drafting of the manuscript: Y.J.P, H-Y.S, and S-W.I. Critical revision of the manuscript for important intellectual content: S.M, J.C, J.K, H-J.K, H-Y.S, S-W.I., and J-I.K. Statistical analysis: Y.J.P, S.M, J.C., and H-Y.S. Obtained funding: Y.J.P, J.C., H-Y.S, S-W.I, and J-I.K. Administrative, technical, or material support: S.M, J.C, J.K, and H-J.K.

Corresponding authors

Correspondence to Ho-Young Son, Sun-Wha Im or Jong-Il Kim.

Ethics declarations

Ethical approval

Institutional Review Board (IRB) approval was obtained and confirmed by the Center for Disease Control of the Korean government and the Seoul National University Hospital Clinical Research Institute (IRB No. C-2004-080-1117). The genotyping and phenotyping methods were performed per the Declaration of Helsinki.

Consent to participate

Informed consent was waived due to the anonymization of the KoGES database, ensuring no personal identifications. The Institutional Review Board of Seoul National University Hospital granted consent for database utilization.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 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 in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Park, Y.J., Moon, S., Choi, J. et al. Genome-wide association study for metabolic syndrome reveals APOA5 single nucleotide polymorphisms with multilayered effects in Koreans. Lipids Health Dis 23, 272 (2024). https://doi.org/10.1186/s12944-024-02248-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12944-024-02248-0

Keywords

Lipids in Health and Disease

ISSN: 1476-511X