Sphingosine-1-phosphate (S1P) is a bioactive lipid implicated in e.g. angiogenesis, lymphocyte trafficking, and endothelial barrier function. Erythrocytes are a main source of plasma S1P together with platelets and endothelial cells. Apolipoprotein M (apoM) in HDL carries 70% of plasma S1P, whereas 30% is carried by albumin. The current aim was to investigate the role of apoM in export of S1P from human erythrocytes. Erythrocytes exported S1P more efficiently to HDL than to albumin, particularly when apoM was present in HDL. In contrast, export of sphingosine to HDL was unaffected by the presence of apoM. The specific ability of apoM to promote export of S1P was independent of apoM being bound in HDL particles. Treatment with MK-571, an inhibitor of the ABCC1 transporter, effectively reduced export of S1P from human erythrocytes to apoM, whereas the export was unaffected by inhibitors of ABCB1 or ATPase. Thus, ABCC1 could be involved in export of S1P from erythrocytes to apoM.

Hepatocytes, renal proximal tubule cells as well as the highly specialized endothelium of the blood brain barrier (BBB) express and secrete apolipoprotein M (apoM). ApoM is a typical lipocalin containing a hydrophobic binding pocket predominantly carrying Sphingosine-1-Phosphate (S1P). The small signaling molecule S1P is associated with several physiological as well as pathological pathways whereas the role of apoM is less explored. Hepatic apoM acts as a chaperone to transport S1P through the circulation and kidney derived apoM seems to play a role in S1P recovery to prevent urinal loss. Finally, polarized endothelial cells constituting the lining of the BBB express apoM and secrete the protein to the brain as well as to the blood compartment. The review will provide novel insights on apoM and S1P, and its role in hepatic fibrosis, neuroinflammation and BBB integrity.

Abstract Context Recent studies discovered a role of apolipoprotein M (apoM) in energy metabolism, and observational analyses in humans suggest an association with type 2 diabetes. The causal relationship remained however elusive. Objective To investigate whether reduced plasma apoM concentration are causally linked to increased risk of type 2 diabetes. Design Prospective study design analysed by mendelian randomization. Setting and participants Two cohorts reflecting the Danish general population: The Copenhagen City Heart Study (CCHS, n=8,589) and the Copenhagen General Population Study (CGPS; n=93,857). Observational analyses included a subset of participants from the CCHS with available plasma apoM (n=725). Genetic analyses included the complete cohorts (n=102,446). During a median follow-up of 16 years (CCHS) and 8 years (CGPS), 563 and 2,132 participants developed type 2 diabetes. Main outcome measures Plasma apoM concentration, genetic variants in APOM, and type 2 diabetes. Results First, we identified an inverse correlation between plasma apoM and risk of type 2 diabetes in a subset of participants from the CCHS (hazard ratio between highest vs lowest quartile (reference)=0.32; 95% CI=0.1-1.01; p for trend=0.02). Second, genotyping of specific SNPs in APOM further revealed a 10.8% (p=6.2e-5) reduced plasma apoM concentration in participants with variant rs1266078. Third, a meta-analysis including data from 599,451 individuals showed no association between rs1266078 and risk of type 2 diabetes. Conclusions The present study does not appear to support a causal association between plasma apoM and risk of type 2 diabetes.