Introduction

Lipoprotein-associated phospholipase A2 (Lp-PLA2) is an enzyme produced by monocyte-macrophages, T-lymphocytes and other inflammatory cells1. In humans, Lp-PLA2 primarily circulates with low-density lipoproteins. It hydrolyses oxidised phosphatidylcholines and produces two proinflammatory molecules, namely lysophosphatidylcholine and oxidised free fatty acids2. Growing evidence supports a role of Lp-PLA2 in the pathogenesis of atherosclerosis1,2,3,4,5. Although several observational studies have suggested Lp-PLA2 activity or mass as an independent predictor for atherosclerotic cardiovascular disease (CVD)5 and the progression of subclinical atherosclerosis6,7,8,9, two recent large randomised, controlled, prospective clinical trials that selectively inhibited Lp-PLA2 failed to show benefits for clinical outcomes in patients with stable or unstable coronary artery diseases10,11. In light of conflicting findings from previous studies, exploring the determinants of Lp-PLA2 levels may be of clinical importance to identify those subjects with high Lp-PLA2 levels who will be more likely to experience a reduction in cardiovascular disease risk with Lp-PLA2 inhibition.

It is widely recognised that a strong genetic component underlies Lp-PLA2 activity and mass. Twin studies estimated the heritability of Lp-PLA2 activity and mass as 0.54 and 0.37, respectively12. A meta-analysis of genome-wide studies in 13,664 Caucasians revealed that genetic defects in PLA2G7 (6p21.2-p12), the gene encoding Lp-PLA2, were significantly associated with plasma Lp-PLA2 activity and mass13. Studies on variants of the PLA2G7 gene in general Asian populations demonstrated significant heterogeneity compared with Caucasians, and the frequency of alleles varied across different Asian groups14,15,16. Moreover, although multiple individual polymorphisms have been reported to be associated with Lp-PLA2 levels, whether there is a joint effect among these polymorphisms remains an open question. We therefore sought to explore PLA2G7 polymorphisms associated with Lp-PLA2 activity and mass in a general Han Chinese population, and to test the joint effect of these polymorphisms on Lp-PLA2 activity and mass.

Materials and Methods

Study participants

Study participants were recruited from the Chinese Multi-provincial Cohort Study (CMCS)-Beijing Project, which is embedded in the CMCS, a nationwide population-based study investigating the risk factors related to the incidence of CVD17. In 1992, 1982 participants were enrolled for CMCS from a Beijing community using a stratified random sampling for each sex and a 10-year age group. From these, 1511 unrelated participants aged 45–74 years provided demographic characteristics and measurements of traditional risk factors from September to November in 2002. After excluding participants with established CVD (n = 73), hemolytic blood samples (n = 25), unavailable DNA samples (n = 107) and failed genotyping (n = 48), 1258 participants (592 male and 666 female) were analysed.

All participants provided informed consent. This study was reviewed and approved by the Ethics Committee of Beijing An Zhen Hospital, Capital Medical University, and was performed in accordance with standards set forth by the Declaration of Helsinki18.

Risk factor survey

This study complied with the protocol set forth by the World Health Organization-MONICA (Monitoring of Trends and Determinants in Cardiovascular Disease). A standard questionnaire was designed to collect information on demographic characteristics, smoking status and personal medical history. Anthropometric measurements and blood pressure (BP) levels were recorded during physical examination. Body mass index (BMI) was calculated as weight in kilograms divided by height squared in metres. BP was measured in the right arm at a sitting position with a regular mercury sphygmomanometer after resting for at least 5 min. The mean value of two consecutive BP readings was used. Hypertension was defined as a mean systolic BP ≥ 140 mmHg and/or a mean diastolic BP ≥ 90 mmHg and/or currently on antihypertensive therapy19. Diabetes mellitus was defined as fasting blood glucose (FBG) ≥ 7.00 mmol/L and/or currently taking glucose-lowering medical treatments20. Regular smoking of one or more cigarettes per day was defined as current smoking.

Laboratory assays

Venous blood samples were drawn from antecubital veins in the morning after fasting for at least 8 h. Fasting total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and FBG were measured on the day of blood collection. TC, TG and FBG were determined by enzymatic methods; LDL-C and HDL-C were measured by homogeneous assays (Daiichi, Tokyo, Japan). The remaining samples were aliquoted and stored at −80 °C until used. Lp-PLA2 activity and mass were measured in 2012. A previous report confirmed that Lp-PLA2 activity and mass measurements in plasma-EDTA samples stored at −80 °C are stable after more than 10 years of storage21. Plasma Lp-PLA2 activity was measured using the Cayman colorimetric assay kit (Cayman Chemical Company, Ann Arbor, MI, USA). A pre-study validation was performed by analysing 20 samples for low-level control and 20 samples for high-level control in duplicate on consecutive days, and a Levey–Jennings chart was plotted22. The mean duplicate coefficient of variation (CV) was 7.31% for the low-level controls and 6.39% for the high-level controls. Plasma Lp-PLA2 mass was assayed according to the enzyme-linked immunoassay method using the diaDexus PLAC® Test Kit (diaDexus, Inc., South San Francisco, CA, USA). The mean duplicate CV was 6.77% and 5.05% for low- and high-level controls, respectively. Because of limited plasma quantities, Lp-PLA2 mass measurements were conducted in 85% of samples (n = 1084) and run singly. To test whether this selection would lead to any potential bias affecting the validity of our findings, comparisons of major risk factors and levels of Lp-PLA2 activity were performed between samples for which Lp-PLA2 mass assays were and were not performed and no significant differences were observed (Table SI).

Selection and genotyping of PLA2G7 polymorphisms

Genomic DNA was extracted from white blood cells using the phenol/chloroform method and stored in 400 μl TE (10 mMTris-HCl, 1 mM EDTA, pH 8.0). Population-specific tagging polymorphisms with minor allele frequency (MAF) of at least 0.05 were selected from the HapMap PhaseII database using Haploview software (version 4.2) under the criteria r2 ≥ 0.8. Nine candidate polymorphisms were chosen and genotyped in this study according to a previously determined association between PLA2G7 polymorphisms and Lp-PLA2 levels in studies of individual association, genome-wide studies and meta-analysis23,24,25, as well as whether a polymorphism is located in a functional region of PLA2G7. These included rs1805017 (R92H), rs1805018 (I198T), rs16874954 (V279F) and rs1051931 (A379V) in the PLA2G7 coding region, rs10948301, rs1421378 and rs9395208 near the transcription start site, and rs9381475 and rs2216465 in PLA2G7 introns. Genotyping was carried out on the Sequenom MassARRAY genotyping platform, and the call rates were more than 97% for the nine polymorphisms. The accuracy of our genotyping method was further confirmed by direct sequencing (BGI LifeTech, Beijing, China) of amplified DNA from 100 randomly selected samples and the discordance was less than 2% between the two methods. It is worth noting that during sequencing, we additionally detected a polymorphism, rs13218408, in intron 8 that was archived in the 1000GENOMES database on August 16, 2014. The association of this polymorphism with Lp-PLA2 levels had not been previously reported in any population; however, rs13218408 had strong association with plasma Lp-PLA2 activity and mass in the sequenced samples. Accordingly, rs13218408 was genotyped in whole samples by the Taqman genotyping assay kit (Applied Biosystems, Foster City, CA, USA) with call rates of 98%. Therefore, a total of ten selected polymorphisms were included and their pairwise linkage disequilibrium patterns are presented in Fig. 1.

Figure 1
figure 1

Pair-wise linkage disequilibrium among ten polymorphisms.

The numbers inside the squares are D′ × 100.

Full size image

Study power estimation

In view of the partial determinant coefficients (R2) ranging from 0.007 to 0.04 between plasma Lp-PLA2 levels and examined polymorphisms, the estimated sample size in the present study was adequately powered (84.4%) such that the type I error probability (α) for a two sided-test was defined as 0.05, MAF was defined as 5% and partial R2 was estimated at 0.007.

Statistical analysis

Continuous variables, expressed as the mean (standard deviation) for normal distributions or as medians (inter-quartile ranges), were compared by the unpaired Student’s t-test or the Mann–Whitney test between two groups and by one-way analysis of variance (ANOVA) across three or more groups. Categorical variables, expressed as numbers (percentages), were compared by the χ2 test. Spearman correlation coefficients were adopted to quantify the relationship between Lp-PLA2 activity and mass.

Deviation from Hardy–Weinberg equilibrium was tested by a Pearson goodness-of-fit test for all polymorphisms examined. After multiple comparisons, each polymorphism that showed significant association with Lp-PLA2 activity and mass was further adjusted for the known CVD risk factors, including age, sex, BMI, FBG, systolic BP, LDL-C, HDL-C and current smoking status by analysis of covariance. The percentage of variance of Lp-PLA2 activity and mass explained by each significant polymorphism was expressed as determinant coefficients in linear regression analyses. Considering linkage disequilibrium, five polymorphisms, rs1421378, rs1805017, rs13218408, rs16874954 and rs2216465, were selected in four respective haplotype blocks, and they were incorporated in a multiple linear regression model to evaluate their independent association with Lp-PLA2 levels. Interactive effects on Lp-PLA2 activity and mass between polymorphisms that are independently associated with Lp-PLA2 levels were further analysed by the general linear model.

Lipid-lowering medication has been reported to affect Lp-PLA2 levels. Therefore, the relationships of the examined polymorphisms with Lp-PLA2 activity and mass after adjusting for known CVD risk factors were also analysed with lipid-lowering treatment as a covariate in regression models or after excluding users of lipid-lowering medication.

The statistical analyses were computed with SPSS software (version 13.0; SPSS Inc, Chicago, IL, USA) and Haploview software (version 4.2; http://www.broad.mit.edu/mpg/haploview)26. Statistical power was calculated with Quanto software (version 1.2.3; http://hydra.usc.edu/gxe)27. All statistical tests were two-tailed, and P < 0.05 was considered statistically significant unless otherwise indicated.

Results

Characteristics of the study participants

Characteristics of the study participants are shown and compared between sexes in Table 1. Male participants had significantly higher plasma Lp-PLA2 activity and mass than female participants. There was a positive correlation between Lp-PLA2 activity and mass (Spearman correlation coefficient = 0.32, P < 0.001).

Table 1 Characteristics of the study participants.
Full size table

The MAF of rs13218408 was 8.1% in the present study, which was lower than that in East Asian samples recently described in the 1000GENOMES database. However, the MAFs of the other nine polymorphisms were similar to those in East Asian samples from the 1000GENOMES database (Table 2). For all polymorphisms examined, the genotype distributions were in Hardy–Weinberg equilibrium, and the genotype/allele distributions were comparable between sexes (Table SII).

Table 2 Distribution of PLA2G7 gene polymorphisms.
Full size table

Effect of individual PLA2G7 polymorphisms on Lp-PLA2 activity and mass

Five of the ten examined polymorphisms (rs1421378, rs1805018, rs13218408, rs16874954 and rs2216465) exhibited a significant association with Lp-PLA2 activity and mass after Bonferroni correction (P < 0.005) (Fig. 2). Carriers homozygous for the minor alleles of these five polymorphisms had the lowest levels of Lp-PLA2 activity and mass compared with the major allele homozygotes. The association of these polymorphisms with Lp-PLA2 activity and mass remained significant after adjustment for known CVD risk factors (Table 3). The proportion of variance of Lp-PLA2 activity and mass explained by each significant polymorphism ranged from 1.7% to 8.1% for activity, and from 1.4% to 16.9% for mass; the highest observed for rs16874954 (8.1% for activity and 16.9% for mass), followed by rs13218408 (7.2% for activity and 13.3% for mass). After excluding participants taking lipid-lowering medication or modelling lipid-lowering treatment as a covariate in regression analyses, there were still significant associations of these five polymorphisms with Lp-PLA2 activity and mass (data not shown).

Table 3 Levels of Lp-PLA2 activity and mass by genotypes of PLA2G7 polymorphisms, after adjusting for cardiovascular risk factors.
Full size table
Figure 2
figure 2

Levels of Lp-PLA2 activity (A) and mass (B) by genotypes of PLA2G7 gene polymorphisms†. Data are presented as mean ± standard error. Differences were tested by the analysis of covariance (ANCOVA) after Bonferroni correction. †Lp-PLA2 mass was measured in 1084 participants.

Full size image

Furthermore, to evaluate whether the association between the polymorphisms and Lp-PLA2 levels was independent of the other variants, a multiple linear regression analysis was conducted. Only rs13218408 and rs16874954 were independently associated with Lp-PLA2 activity and mass (Table SIII).

Joint effect of PLA2G7 polymorphisms on Lp-PLA2 activity and mass

Genotype combination analysis was conducted for rs13218408 and rs16874954, the two polymorphisms that showed the strongest and independent association with Lp-PLA2 activity and mass (Table SIV). Interaction analysis showed that rs13218408 and rs16874954 had a significantly interactive effect on Lp-PLA2 mass (P = 0.003), and a marginally significant interactive effect on Lp-PLA2 activity (P = 0.058), as shown in Fig. 3. Participants who were simultaneous carriers for the minor alleles of these two polymorphisms (8.1% of study participants) had the lowest levels of Lp-PLA2 activity and mass compared with the other participants. This effect was independent of the known CVD risk factors.

Figure 3
figure 3

Joint effects of PLA2G7 polymorphisms rs13218408 and rs16874954on Lp-PLA2 activity and mass.

Abbreviations: Lp-PLA2, lipoprotein-associated phospholipase A2; MM, participants with minor allele homozygotes; MW, participants with heterozygotes; WW, participants with major allele homozygotes; BMI, body mass index; BP, blood pressure; FBG, fasting blood glucose; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol. Data are expressed as mean ± standard deviation for continuous variables. Data were compared by analysis of covariance between subgroups with both the minor allele homozygotes and other subgroups. Association was adjusted for age, sex, BMI, systolic BP, FBG, LDL-C, HDL-C, and current smoking status. **P value < 0.001. †Lp-PLA2 mass was measured in 1084 participants.

Full size image

Discussion

The major finding of this study was that we found a significant association of the PLA2G7 polymorphism, rs13218408, with the level of Lp-PLA2 activity and mass, which has not been reported previously in any population. Moreover, we also found a significant joint effect between this polymorphism and a widely validated coding polymorphism, rs16874954, on the level of Lp-PLA2.

Among the ten polymorphisms examined in this study, four were in the PLA2G7 coding region, and two of them exhibited a marked association with Lp-PLA2 activity and mass in this population. Notably, polymorphism rs16874954 (V279F) is a widely evaluated locus in exon 9 with replacement of C by A, resulting in transversion of valine to phenylalanine. As indicated in our results, the heterozygous carriers of the rs16874954 minor allele had a significant reduction of 32.4% in Lp-PLA2 activity and 34.4% in mass, and almost no detectable enzyme activity and mass were found in the homozygous carriers. This loss of activity or mass caused by rs16874954 is supported by many previous reports14,15,16,28,29. Consistent with these observations, functional expression studies of the V279F mutation by Miwa et al.28 and Stafforini et al.28,29 demonstrated complete abolition of enzymatic activity and the molecular basis of an autosomal recessive form of Lp-PLA2 deficiency. Ishihara et al.30 and Zhang et al.31 reported that a complete absence of Lp-PLA2 activity in homozygous carriers of the rs16874954 minor allele was caused by a defect in enzyme secretion. Importantly, this polymorphism was mainly observed in Asians, with the highest frequency of the mutant allele found in Japanese (17.8%)15, and a lower frequency found in Koreans (12.6%)14 and Han Chinese (4.8% in this study, 5.4% in the study of Hou et al.16 and 6.5% in the study of Liu et al.32), whereas the polymorphism is rare in Europeans13. These findings highlight the genetic heterogeneity across ethnicities. In contrast to the Asian-specific nature of rs16874954 frequency, the other three non-synonymous PLA2G7 polymorphisms, rs1805018 (I198T), rs1805017 (R92H) and rs1051931 (A379V), have been reported in multiple ethnic populations, and remarkable ethnic differences in the association with Lp-PLA2 activity and mass were found. For example, the minor allele 379 A of rs1051931 (A379V) was significantly associated with higher Lp-PLA2 activity in Caucasians33, but was not related to Lp-PLA2 levels in the present study or was significantly associated with lower Lp-PLA2 activity in Asians in a previous study34. Together, these findings suggest that the frequency and effect of PLA2G7 coding polymorphisms on Lp-PLA2 levels might be influenced by divergent ethnic-specific genetic profiles.

Three non-coding polymorphisms were related to Lp-PLA2 levels in this study. It is notable that the polymorphism rs13218408 was found to be significantly associated with Lp-PLA2 levels for the first time. This polymorphism (in intron 8) has been identified in East Asian, African and European samples, with MAFs of 12.9%, 31.2% and 6.5%, respectively. A literature search did not reveal any evidence regarding its association with Lp-PLA2 levels or CVDs. Our study demonstrated a contributory role of rs13218408 to the variation of Lp-PLA2 activity and mass. This contribution (7.2% for activity and 13.3% for mass) was second only to rs16874954 (8.1% for activity and 16.9% for mass), calling for further validation in other populations. Further genetic combination analyses revealed significant interaction effects of rs13218408 and rs16874954 on Lp-PLA2 levels, and these effects were independent of traditional CVD risk factors. Although the linkage disequilibrium analyses showed that rs13218408 was in moderate linkage disequilibrium with rs16874954 and rs1805018, it is perhaps linked with one as-yet-unidentified functional polymorphism in or flanking PLA2G7. Another possible explanation for the biological relevance of rs13218408 is that it is located in a PLA2G7 intron that may contain an enhancer element acting on PLA2G7 or other genes in the vicinity, as supported by the HaploReg database. Despite all this, it is difficult to determine the underlying molecular mechanism causing these effects, and further studies are warranted.

The present study has several strengths. This is the first large-scale study conducted in a general Chinese population to investigate polymorphisms associated with two measures of Lp-PLA2 level (activity and mass). Moreover, for the first time, we identified a significant association of a PLA2G7 polymorphism with Lp-PLA2 levels, and its combined influence with other PLA2G7 polymorphisms on plasma levels of Lp-PLA2 activity and mass. Furthermore, to yield robust estimates, the known potential confounders that might affect Lp-PLA2 levels, such as CVD or lipid-lowering medications, were excluded or adjusted in a sensitive manner. Nevertheless, a number of potential limitations of this investigation merit careful consideration. First, we only focused on ten PLA2G7 polymorphisms. These polymorphisms were chosen by a systematic review of the HapMap database and previous reports, and further strengthened by sequencing. However, we cannot exclude the impact of other unknown rare variants in the Chinese population. Second, the fact that our study participants were of Chinese descent may limit the generality of our findings, calling for further confirmation in other ethnic populations.

In summary, our findings identified a significant association of a PLA2G7 polymorphism with the levels of Lp-PLA2, a potential risk factor of atherosclerotic CVD. Future studies that investigate the biological or clinical implications of this genetic variation are warranted.

Additional Information

How to cite this article: Qi, Y. et al. A previously unreported impact of a PLA2G7 gene polymorphism on the plasma levels of lipoprotein-associated phospholipase A2 activity and mass. Sci. Rep. 6, 37465; doi: 10.1038/srep37465 (2016).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.