Large-scale lipidomics profiling reveals characteristic lipid signatures associated with an increased cardiovascular risk

Abstract

Background and aims: Patients with cardiovascular disease (CVD) are at high risk to develop adverse events. The distinct risk of developing adverse cardiovascular (CV) events is not solely explained by traditional risk factors. Platelets are essentially involved in progression of CVD including coronary artery disease (CAD) and platelet hyperreactivity leads to development of adverse CV events. Alterations in the platelet lipidome lead to platelet hyperresponsiveness and thus might alter the individual risk profile. In this study, we investigate the platelet lipidome of CAD patients by untargeted lipidomics and elucidate alterations in the lipid composition of patients with adverse CV events.

Methods: We characterized the platelet lipidome in a large consecutive CAD cohort (n = 1057) by an untargeted lipidomics approach using liquid chromatography coupled to mass spectrometry.

Results: The platelet lipidome in this study identified 767 lipids and characteristic changes occurred in patients with adverse CV events. The most prominent upregulated lipids in patients with cardiovascular events primarily belong to the class of phospholipids and fatty acyls. Further, upregulated platelet lipids are associated with an increased cardiovascular or bleeding risk and independently associated with adverse events. In addition, alterations of the platelet lipidome are associated with modulation of in vitro platelet functions.

Conclusions: Our results reveal that the composition of the platelet lipidome is altered in CVD patients with an increased cardiovascular risk and distinct platelet lipids may indicate adverse events. Results of this study may contribute to improved risk discrimination and classification for cardiovascular events in patients with CVD. Main findings of this study and hypothetical impact of altered platelet lipid signatures in patients with adverse cardiovascular events on platelet function and clinical outcome. LPE lysophosphatidylethanolamines, CAR acylcarnitines, FA fatty acids.

Keywords: Bleeding; Cardiovascular risk; Coronary artery disease; Lipidome; Platelets.

Fig. 1

Assessment of the human platelet lipidome in the prospective CAD patient cohort reveals distinct lipid signatures. A The platelet lipidome in this study comprised 767 lipids from isolated circulating platelets. The lipids are attributable to 40 subclasses and six main lipid categories: glycerophospholipids (55.8% of verified lipids), sphingolipids (17.1%), glycerolipids (15.9%), fatty acyls (9.2%), sterol (1.6%) and prenol (0.4%) lipids. Platelet lipid categories are coloured accordingly and significantly altered species among patients with adverse cardiovascular events are highlighted (red). B Hexagonal bin plot of structural lipid analysis. The entirety of platelet lipids is depicted according to number of double bonds (Y-axis) and fatty acid side chain length (X-axis) and quantity of distinct lipid species is coloured accordingly. A predominance of lipids with carbon chain length between 25 and 45 with mainly poly-unsaturated fatty acid (PUFA) sidechain and less monounsaturated (MUFA) or saturated sidechains (SFA) is highlighted

Fig. 2

Characteristic changes in the platelet lipidome occur in patients with adverse cardiovascular events. A Volcano plot of all identified platelet lipids in patients with CAD. The X-axis (fold change, FC) displays the base 2 logarithm of the concentration ratios between patients with cardiovascular events including cardiac death, myocardial infarction or stroke (ischaemic endpoint) to those without adverse events during the 3-year follow-up (ischaemic endpoint n = 51, no endpoint n = 1006). Values > 1 indicate upregulation in patients with cardiovascular events; < 1 downregulation compared to patients without incident CVD. Y-axis showing negative log10-transformed p values (BH p < 0.05). Significantly upregulated lipids with a > 1.5-fold change are labelled and mainly comprised lysophosphatidylethanolamines and lysophosphatidylserines. B Volcano plot of all identified platelet lipids comparing patients with major bleeding events (n = 21) to those without bleeding (n = 1036) during the follow-up period. Significantly (BH p < 0.05) upregulated lipids with a > 1.5-fold increase in patients with a major bleeding are displayed and exclusively belong to the category of fatty acyls

Fig. 3

The platelet lipidome is altered in patients with adverse cardiovascular events. Principal component analysis (PCA) of the platelet lipidome (767 lipids). Clustering of patients with adverse ischaemic events (orange) or bleeding events (purple) reveals a clear separation compared to patients without adverse cardiovascular events (light blue). Box plots of first two components (PC1/2) display first and third quartiles, and whiskers extend from each quartile to the minimum or maximum values. Data were based on normalized concentrations of all identified lipids and Mann–Whitney U-test was used to analyse significant differences between cardiovascular endpoints

Fig. 4

Physicochemical attributes of platelet lipid species vary among patients with adverse cardiovascular events. A Dot plot of identified lipids in patients with symptomatic CAD. Each dot represents a unique lipid species. X-axis represents fatty acyl carbon chain number and y-axis the number of double bonds. Size is scaled by negative Log-transformed p-value of comparison between lipid levels in patients with and without cardiovascular events. Colour continuously represents fold change between patients subgroups. Upregulation of glycerophospholipids mainly with low fatty acyl carbon number (C < 25) and poly-unsaturated side chain (DB ≥ 2) is observed in patients with cardiovascular endpoints compared to those without incident CVD. B The right plot depicts upregulation of mainly fatty acyls and glycerophospholipids in patients with major bleeding in contrast to those without bleeding events. C Mean ratio of platelet lipid abundance between patients with cardiovascular events compared to patients without adverse events by fatty acid side chain length and degree of saturation in different lipid classes. Significant (p < 0.05) Spearman correlations (R_s) are shown. Ratio > 1 indicates a higher abundance in patients with adverse cardiovascular events. D Spearman correlations (R_s) of lipid abundance in patients with bleeding events and degree of saturation and side chain length

Fig. 5

Association of distinct platelet lipid signatures with an increased hazard of developing adverse cardiovascular events. A Forest plot of the estimated regression coefficient (95% CI) from Cox proportional hazard models on the association of significantly upregulated lipids in patients with ischaemic endpoints and the cardiovascular risk. All lipids were found to be independently associated with an increased cardiovascular risk and hazard ratios (HR) were calculated for a 1-SD unit higher lipid concentration and adjusted for age, gender and statin treatment. Thus, especially lysophosphatidylethanolamines might be independently suitable to predict the cardiovascular risk in patients with CVD. B Associations of significantly upregulated platelet lipids in patients with an increased bleeding risk. High levels of acylcarnitines and fatty acids FA18:1 and FA 18:2 might indicate an increased hazard to develop major bleeding events in patients with CVD

Fig. 6

Alterations of the platelet lipidome in patients with adverse cardiovascular events modulate platelet functions and thus, might influence disease progression and outcome of patients with incident CVD. A Patients were divided into quartiles (upper quartile = high/lower quartile = low) according to the concentration of distinct lipids found to be significantly upregulated in patients with cardiovascular events when compared to those without adverse events. We found, that collagen-mediated platelet aggregation was significantly (Student’s t test, p < 0.05) increased under high concentrations of LPE. In contrast, high concentrations of acylcarnitines were found to be significantly (p < 0.05) associated with a decrease of collagen, arachidonic acid and ADP-mediated platelet aggregation. B Correlation of lipids found to be significantly increased in patients with cardiovascular or bleeding events compared to patients without adverse events. Collagen-mediated platelet aggregation (AUC) significantly (p < 0.05) correlated with increased concentrations of LPE. In contrast, collagen-mediated platelet aggregation was inversely associated with enhanced concentrations of acylcarnitines

Fig. 7

Platelet functions are altered in patients with distinct changes in the platelet lipidome. A The lipid class of lysophosphatidylethanolmines (LPE) was summarized by mean concentrations. All patients were divided into quartiles according to the LPE concentration (1st quartile < 25%, 2nd quartile < 50%, 3rd quartile < 75%, 4th quartile ≥ 75%). We found that collagen-mediated platelet aggregation was significantly (Student’s t-test, p < 0.05) increased under high (≥ 75%) and intermediate (< 50%) concentrations of LPE in contrast to low (< 25%) concentrations. B In contrast, collagen-mediated platelet aggregation was significantly (Student’s t test, p < 0.05) reduced under high (≥ 75%) concentrations of acylcarnitines (CAR) in contrast to intermediate (< 75%, < 50%) and low (< 25%) concentrations