Novel genetic approach to investigate the role of plasma secretory phospholipase A2 (sPLA2)-V isoenzyme in coronary heart disease: modified Mendelian randomization analysis using PLA2G5 expression levels

Abstract

Background: Secretory phospholipase A2 (sPLA2) enzymes are considered to play a role in atherosclerosis. sPLA2 activity encompasses several sPLA2 isoenzymes, including sPLA2-V. Although observational studies show a strong association between elevated sPLA2 activity and CHD, no assay to measure sPLA2-V levels exists, and the only evidence linking the sPLA2-V isoform to atherosclerosis progression comes from animal studies. In the absence of an assay that directly quantifies sPLA2-V levels, we used PLA2G5 mRNA levels in a novel, modified Mendelian randomization approach to investigate the hypothesized causal role of sPLA2-V in coronary heart disease (CHD) pathogenesis.

Methods and results: Using data from the Advanced Study of Aortic Pathology, we identified the single-nucleotide polymorphism in PLA2G5 showing the strongest association with PLA2G5 mRNA expression levels as a proxy for sPLA2-V levels. We tested the association of this SNP with sPLA2 activity and CHD events in 4 prospective and 14 case-control studies with 27 230 events and 70 500 controls. rs525380C>A showed the strongest association with PLA2G5 mRNA expression (P=5.1×10(-6)). There was no association of rs525380C>A with plasma sPLA2 activity (difference in geometric mean of sPLA2 activity per rs525380 A-allele 0.4% (95% confidence intervals [-0.9%, 1.6%]; P=0.56). In meta-analyses, the odds ratio for CHD per A-allele was 1.02 (95% confidence intervals [0.99, 1.04]; P=0.20).

Conclusions: This novel approach for single-nucleotide polymorphism selection for this modified Mendelian randomization analysis showed no association between rs525380 (the lead single-nucleotide polymorphism for PLA2G5 expression, a surrogate for sPLA2-V levels) and CHD events. The evidence does not support a causal role for sPLA2-V in CHD.

Keywords: Mendelian randomization analysis.

Figure 1. Manhattan plot of the association between SNPs in the PLA2G5 region and PLA2G5 mRNA expression by tissue type

rs525380 A>C showed the strongest association with PLA2G5 mRNA expression in the aorta adventitia (P=5.05×10⁻⁶). The black horizontal line above the scale represents the position of PLA2G5. Total number of individuals providing tissue samples for analysis = 272 (samples available for each tissue: Mammary Artery 89, Liver 212, Aorta Med 138, Aorta Adventitia 133, Heart 127).

Figure 2. Overall expression of all probe-sets and the differential expression of PLA2G5

rs525380 C>A with PLA2G5 mRNA in the five tissue types. MMed: Mammary artery intima-media; AMed: dilated and non-dilated ascending aorta intima-media; Aorta ADV: aorta adventitia. For CC/AC/AA the sample sizes are as follows: heart 43/68/16, MMed 21/51/17, AMed 44/70/24, Aorta ADV 38/73/22, Liver 59/120/32.

Figure 3. Forest plot of the association of PLA2G5 rs525380 (per A-allele) with CHD in 27,230 cases in a total of 97,730 individuals

When limited to studies with fewer than 1000 CHD events, the OR was 1.04 (95%CI: 0.97, 1.11) with an I² of 31% (95%CI: 0% to 71%). For studies with more than 1000 CHD events, the OR was 1.01 (95%CI: 0.99, 1.04) with an I² of 0% (95%CI: 0% to 43%).