Shedding of endothelial protein C receptor contributes to vasculopathy and renal injury in lupus: In vivo and in vitro evidence

Background. Candidate biomarkers for vasculopathy in systemic lupus erythematosus (SLE) include circulating endothelial cells and the recently identified endothelial protein C receptor (EPCR) which, when shed, promotes a thrombotic diathesis. This study sought correlation between plasma levels of soluble EPCR and disease manifestation/severity, with a focus on lupus nephritis. Methods. In 81 SLE patients (evaluated cross-sectionally and longitudinally) and 59 healthy controls, levels of soluble EPCR and soluble E-selectin were assessed by sandwich enzyme-linked immunosorbent assay (ELISA), circulating endothelial cells isolated by immunomagnetic separation, and EPCR gene polymorphisms determined. Mechanisms of vascular injury were addressed in vitro in human aortic endothelial cells (HAEC) cultured in the presence and absence of interferon-gamma (IFN-gamma). Results. The mean level of soluble EPCR was significantly higher in SLE patients (263 +/- 13 ng/mL) than controls (174 +/- 11 ng/mL) (P < 0.0001). Patients with active or past renal involvement had significantly higher mean soluble EPCR levels (306 +/- 21 ng/mL) (N= 40) than patients without nephritis (228 +/- 14 ng/mL) (N= 41) (P= 0.0033). Mean soluble EPCR correlated positively with serum creatinine (R= 0.3429, P < 0.0001). The prevalence of the enhanced-shedding EPCR polymorphism A6936G was higher in SLE (41%) (N= 27) than controls (7%) (N= 29) (P= 0.0039). Patient and control plasma were also interrogated for soluble E-selectin, a comparator plasma marker. The results suggest that soluble E-selectin and soluble EPCR are not equivalent end points of vasculopathy and endothelial perturbation in SLE. Although in SLE patients the absence or diminished expression of membrane EPCR on circulating endothelial cells varied, the rare circulating endothelial cells detected in controls invariably expressed membrane-bound EPCR. IFN-gamma-treated HAEC expressed less membrane-bound EPCR [133 relative fluorescence units (rfu)] than untreated HAEC (275 rfu); more soluble EPCR was detected in IFN-gamma-treated (1.1 ng/10(6) cells) than untreated HAEC (0.65 ng/10(6) cells) (P= 0.027). Conclusion. The results obtained from this cross-sectional/longitudinal study support the hypothesis that the vascular dysfunction characteristic of SLE may be related to a dramatically altered distribution of EPCR, both soluble and membrane-bound forms.