Eicosapentaenoic acid (EPA) limits rapid oxidation of lipoprotein(a) [Lp(a)] yielding favorable changes in endothelial cell protein expression that may contribute to clinical benefits

Background In REDUCE-IT, the omega-3 fatty acid EPA (20:5, n-3) delivered as icosapent ethyl reduced CV events in high-risk patients, including those with elevated Lp(a). Elevated Lp(a) is associated with increased cardiovascular (CV) risk and enriched with oxidized phospholipid that may accelerate oxidation reactions. EPA has multiple alkene bonds which neutralize free radicals. Whether the lipoprotein species (i.e., Lp(a)) or oxidative modification has a greater impact on protein expression is unknown. Purpose We compared the antioxidant effects of EPA, α-linolenic acid (ALA 18:3, n-3), and probucol on the oxidation rate of Lp(a)-enriched plasma and non-modified LDL. We then assessed the effects of Lp(a)±EPA or LDL±EPA undergoing oxidation on protein expression in endothelial cells (ECs) using proteomics. Methods Lp(a) was enriched by isopycnic centrifugation to 62% of total ApoB particles from subjects with elevated levels. Lp(a) and LDL samples were matched for total protein and incubated with EPA or ALA (50 µM) or probucol (10 µM). Samples underwent copper sulfate-induced oxidation monitored by formation of lipid hydroperoxides (LOOH) and malondialdehyde (MDA) through 4 h. For proteomic analysis, Lp(a) and LDL samples were incubated with EPA (50 µM) or equivolume vehicle followed by copper sulfate-initiated oxidation for 2 h. Human umbilical vein ECs (HUVECs) were then incubated with 150 µg/mL Lp(a)±EPA, LDL±EPA, or untreated controls for 8 h. Global proteomics was performed by LC/MS. Results Lp(a) samples had higher baseline LOOH levels compared with LDL and underwent more rapid oxidation: 25% more LOOH and 89% more MDA production by 0.5 h (all p<0.01). At peak oxidation levels, EPA attenuated oxidation of both Lp(a) and LDL >65% (p<0.001) unlike ALA or probucol. Based on unadjusted p-value, proteomics showed 210 and 212 proteins significantly modulated by Lp(a) and LDL following oxidation compared with untreated controls, including similar increases in matrix metalloproteinase-1, heat shock protein (HSP) 70 kDa 1B, HSP105/110 kDa, and Bcl-2-associated athanogene 3 which together facilitate protein folding. These effects were reversed in cells receiving EPA-treated Lp(a) and/or EPA-treated LDL. When restricted based on false-discovery rate-adjusted p-value, only the changes with Lp(a) remained significant, including the reversals with EPA, though the trends were similar in LDL samples. Conclusions Lp(a) underwent rapid oxidation compared with LDL - possibly due to higher oxidized lipid content - that was inhibited by EPA but not by ALA or probucol. Lp(a), but not LDL, significantly modulated certain stress response proteins under oxidized conditions in ECs that were reversed by EPA. The beneficial effects of EPA on Lp(a)/LDL oxidation and subsequent protein expression may contribute to mechanisms of CV event reduction, including in patients with elevated Lp(a) and/or LDL (REDUCE-IT).