Proteomic pathways linking plasma ceramides to coronary artery disease

Background

There is increasing evidence that plasma ceramide lipids (Cers) reflect cardiovascular risk beyond traditional risk predictors. In this context C16, C24, and the C16-C24 ratio have previously been associated with coronary plaque characteristics, cardiovascular events, and all-cause mortality. However, underlying mechanisms remain incompletely understood. Investigating the plasma proteome in relation to Cers may uncover key biological drivers of these associations.

Purpose

We investigated the pathways linking plasma Cers and coronary artery disease using a targeted proteomic approach.

Methods

We examined 581 patients with chronic or acute coronary syndrome. Normalized protein expression levels(NPX) of 369 cardiometabolic plasma proteins were measured with proximity extension assays. Cers, quantified by mass spectrometry, were log2-transformed for analysis. Possible associations between Cers and NPX were assessed with univariable linear regression, with correction for age, sex, and multiple testing (Benjamini-Hochberg), and multivariable analyses with regularisation-based protein selection using Mini Concave Penalty (MCP). Effect sizes (ß), expressed as the difference in NPX per two-fold increase in Cers level, their 95% confidence intervals (CI), and corresponding marginal false discovery rate (mFDR) were depicted. Furthermore, Reactome pathway over-representation analysis (ORA) was performed on significant Cers-related proteins using a panel-restricted background with fold enrichment (FE) and p-values (p) derived from hypergeometric testing and multiple testing correction. Results were visualized as bubble plots, where bubble size represents FE and color encodes −log10 (p).

Results

Mean age was 62 (standard deviation: 11) years and 76% were men. Univariable analyses revealed C16, C24 and the C16-C24 ratio significantly associated with 255, 175 and 151 proteins, respectively. In MCP-based analyses, vascular cell adhesion molecule 1 ([VCAM1], ß: 0.29, CI [0.23, 0.35], p <0.0001) and inter-alpha-trypsin inhibitor heavy chain H3 ([ITIH3], ß: 0.12, CI [0.09, 0.16], p <0.0001) exhibited the strongest positive associations with the C16-C24 ratio, while tyrosine-protein kinase ([KIT], ß: -0.14, CI [-0.19, -0.10], p <0.0001) and integrin beta-2 ([ITGB2], ß: -0.18, CI [-0.24, -0.12], p <0.0001) demonstrated the most negative association (Figure 1).

Figure 2 shows ORA predicted biological processes, with the strongest C16-C24 ratio enrichment observed in interleukin-4 / interleukin-13 inflammatory signaling pathways (FE 2.18, p < 0,001). C24- and C16-associated proteins were respectively enriched mostly in hemostasis (FE 1.58, p < 0.001) and broader immune system pathways like the adaptive immune system (FE 1.29, p < 0.01).

Conclusion

The C16-C24 ratio was associated with a novel and distinct proteomic signature - including VCAM1, ITIH3, ITGB2, and KITLG - enriched in inflammatory, metabolic, and extra cellular matrix pathways.Figure 1Figure 2