Apolipoproteins in High-Density Lipoproteins (HDL) and their modifications determine the Sphingosine 1-Phosphate (S1P) content of HDL: effects on the HDL-mediated, S1P-dependent signaling

Abstract

Purpose: The bioactive sphingolipid sphingosine 1-phosphate (S1P) binds in plasma mainly to high-density lipoproteins (HDL) via apolipoprotein M (apoM) and mediates many of the atheroprotective and cardioprotective properties of HDL. Recently, we demonstrated that the S1P-content of plasma and HDL is reduced in patients with coronary artery disease (CAD). In the current study, we tested the hypothesis that the reduction of S1P in CAD-HDL is due to modifications of apolipoproteins in HDL. In addition, we tested whether the extent of cellular responses upon stimulation with HDL was related to the S1P content of HDL. Methods: HDL were isolated by density gradient ultracentrifugation from plasma of patients with CAD (n=65) and of controls (n=68). The contents of apoM, apoAI and S1P were detected by western blotting, ELISA and mass spectroscopy, respectively. The uptake capacity of HDL for S1P was measured in native and oxidized HDL as well as in the subfractions HDL2 and HDL3. Induction of the phosphorylation of p42/44 Erks upon stimulation with HDL was determined in S1P-receptor 1-overexpressing CHO-cells. The same assays were performed in HDL isolated from plasma of apoM-knock out, apoAI knock-out, and apoAI-transgenic mice. Results: While the S1P content was lower in CAD-HDL, the contents of apoM and apoAI were similar in CAD-HDL and control HDL. Oxidation of HDL reduced their baseline S1P content by 44% and their uptake capacity for S1P by 69%. HDL3 had less S1P than HDL2 not only at baseline (122 pmol/mg of protein [102-171] vs. 225 pmol/mg of protein [216-448], P=0.004) but also after loading HDL with S1P (1510 pmol/mg of protein [930-1990] vs. 5732 pmol/mg of protein [5092-7252], P=0.004). HDL isolated from apoM-deficient and apoAI-transgenic mice had a significantly reduced content of S1P (6.3% and 20.2%, respectively). In addition, intracellular phosphorylation of p42/44 Erks after stimulation with apoM-deficient HDL was greatly diminished. An enormous capacity of HDL to uptake S1P was found (20 fold of baseline). Loading apoM-deficient HDL with S1P normalized the defective activation of p42/44 Erks. HDL mediated cell signaling was completely blocked by a S1P neutralizing antibody. Conclusion: Changes in the apolipoprotein composition of HDL, posttranslational modifications of apolipoproteins, or the disarranged ratio of HDL2/HDL3 in patients with CAD might be causes for the reduction of S1P in plasma and in HDL in CAD. As the content of S1P in HDL seems to determine the extent of HDL-mediated effects, quality and function of HDL might be improved by targeting their S1P content.