Soluble Urokinase Plasminogen Activator Receptor Is Increased in Patients with Paroxysmal Nocturnal Hemoglobinuria (PNH) and Thrombosis and Inhibits Plasmin Generation In Vitro and Promotes Thrombosis in the Mouse Model.

Venous thrombotic events produce serious morbidity and are the most frequent cause of death in patients with paroxysmal nocturnal hemoglobinuia (PNH). Urokinase plasminogen activator receptor (uPAR) is a glycosylphosphatidylinositol-linked protein expressed on neutrophils and endothelial cells which mediates endogenous thrombolysis through a urokinase plasminogen activator (uPA)-dependent mechanism. On the cell surface, pro-uPA binds to uPAR, promoting the conversion of pro-uPA to active uPA; uPA then cleaves plasminogen, yielding active plasmin. Fibrinolysis may play an important role in PNH where ongoing hemolysis produces procoagulant-rich red cell microvesicles (Ninomiya h et al, Br J Haematol 106, 224); that thrombotic events frequently occur in the context of a hemolytic event in PNH suggests a role of erythrocyte or platelet microvesicles in clot formation. Membrane GPI-anchored uPAR is decreased or absent on granulocytes and soluble uPAR levels are increased in patients’ plasma. In the current study of 45 patients with PNH and 20 normal controls, we demonstrate that plasma suPAR concentrations correlated with the proportion of neutrophils lacking CD16, another GPI-anchored protein (R²=0.9), and that levels are 4–10-fold greater in PNH than in controls. Patients with a history of thrombosis had a higher soluble uPAR levels (in stored samples pre-dating the thrombosis) than did patients without a history of thrombosis (mean 2.2 ng/mL compared to 0.5 ng/mL; p=0.0001). In vitro suPAR was released from PNH hematopoietic cells during short-term culture in three experiments; suPAR levels were 20 fold greater than those obtained from normal CD34 cells. Activated platelets also appear to be a substantial source of suPAR: uPAR was expressed on the surface of ADP-activated CD59-positive but not of CD59-negative platelets. By ELISA specific for suPAR, we showed that substantial quantities of suPAR were released following activation of PNH but not normal platelets. Immunoblot of released material using suPAR mAb demonstrated a band at 35kDa, consistent with absence of the GPI-anchor. We hypothesized that suPAR might bind to plasma uPA, preventing its interaction with the active membrane-anchored form. In vitro we demonstrated dose-dependent decrease in the activity of single-chain urokinase following the addition of recombinant suPAR, with loss of 25% of the activity after the addition of 0.45nM suPAR. Next we injected suPAR (to achieve a intravascular concentration of 3ng/mL) and autologous mouse red cell microvesicles into three mice for three consecutive days; six additional mice were injected with uPAR or microvesicles alone and three mice injected with PBS served as controls. Mice were sacrificed after 7 days and samples obtained from liver and lung for pathological examination. All 3 mice receiving both suPAR and red cell microvesicles had numerous thrombi in the lungs and liver, while these changes were not observed in the control mouse or in mice receiving either microvesicles or suPAR alone. These data suggest that absent leukocyte and platelet membrane uPAR, high levels of suPAR, and the presence of red cell microvesicles all contribute to produce thrombosis in patients with PNH.