Impaired Cardioprotection by HDL in CAD and Diabetes in Ischemia/Reperfusion Injury: role of S1P and SR-BI

Abstract Background HDL dysfunction rather than HDL-cholesterol concentration is involved in the pathogenesis of coronary artery disease (CAD) and type-2 diabetes (T2DM). While causes and consequences of HDL dysfunction are manifold, reduced concentrations of HDL-sphingosine-1-phsophate (S1P) are partially responsible for impaired vasodilation and suppression of inflammation by CAD-HDL. Administration of healthy human HDL prior to coronary ischemia/reperfusion (I/R) in mice reduced infarct size (IS) due to HDL-S1P. The role of major HDL receptor SR-BI is unexplored in this context. Purpose In this study, we, a) investigated the cardioprotective properties of healthy versus CAD-HDL and T2DM-HDL in a murine model of acute myocardial infarction (AMI), b) tested the effect of S1P loading of CAD-HDL in cardioprotection, and c) detected the relevant HDL receptor of this cardioprotection. Methods HDL were isolated from plasma of healthy volunteers, CAD, and T2DM patients by density ultra-centrifugation and injected (43 mg HDL protein/KG) in the tail vein of C57Bl/6J mice 5 minutes prior to 30 minutes of ischemia. Cardiac function was assessed after 24 hours of reperfusion by echocardiography. IS was analyzed by TTC staining and S1P concentration measured by LC-MS/MS, respectively. Results Administration of human healthy HDL reduced IS by 23% and increased ejection fraction (EF) by 22% 24 hours after I/R (IS: Control 43.8±6.9% [n=17] vs. healthy HDL 32.9±3.6% [n=9]; EF: Control 34.5±5.7% vs. healthy HDL 41.9±4.1%). In contrast, CAD-HDL in the same dosage had no protective effect (IS: 40.1±5.7% [n=12]; EF: 31.9±8.4%). As HDL-S1P concentrations were 33% lower in CAD-HDL compared to healthy HDL, we tested whether S1P-loading may correct CAD-HDL's defective cardioprotection. Indeed, S1P-loading (38 μg S1P/kg in 43 mg HDL protein/KG) completely restored CAD-HDL cardioprotection to levels achieved by healthy HDL. S1P-loading of healthy HDL had no additional benefit. Administration of T2DM-HDL prior to I/R led to 28% larger IS and 22% worse EF compared to healthy HDL (IS: healthy HDL 33.0±4.3% [n=6] vs. T2DM-HDL 42.4±8.9% [n=13]; EF: healthy HDL 38.5±5.8% [n=6] vs. T2DM-HDL 33.1±.4.0%). Compared to healthy HDL, T2DM-HDL exhibited a 19% decrease in S1P content. We thus tested whether the major HDL receptor SR-BI is involved in HDL-S1P mediated cardioprotection using global SR-BI deficient mice (Scarb1−/−). Remarkably, cardioprotection by HDL administration was completely absent in Scarb1−/− mice but intact in wild type controls. (IS Scarb1++: Vehicle 38.6±8.3% [n=12] vs. HDL 29.9±8.5% [n=11], Scarb1−/−: Vehicle 31.4±5.6% [n=12] vs. HDL 31.1±6.9% [n=17]) Conclusion We have identified: (a) impaired cardioprotection after I/R as new characteristic of HDL dysfunction in CAD and T2DM; (b) low HDL-S1P as its cause and the possibility of its therapeutic correction by S1P loading, and (c) SR-BI as the HDL receptor responsible for HDL-S1P-mediated cardioprotection. Funding Acknowledgement Type of funding sources: None.