Search

Angiostatin was first discovered as a plasminogen fragment with antitumor/antiangiogenic property. One of the angiostatin isoforms, that is, angiostatin 4.5 (AS4.5), consisting of plasminogen kringle 1 to 4 and a most part of kringle 5, is produced by autoproteolysis and present in human plasma. beta2-glycoprotein I (beta2GPI) is proteolytically cleaved by plasmin in its domain V (nicked beta2GPI), resulting in binding to plasminogen. Antiangiogenic properties have been recently reported in nicked beta2GPI as well as in intact beta2GPI at higher concentrations. In the present study, we found significant binding of nicked beta2GPI to AS4.5 (K(D) = 3.27 x 10(6) M(-1)). Via this binding, nicked beta2GPI attenuates the antiangiogenic functions of AS4.5 in the proliferation of arterial/venous endothelial cells, in the extracellular matrix invasion and the tube formation of venous endothelial cells, and in vivo angiogenesis. In contrast, intact beta2GPI does not bind to AS4.5 or inhibit its antiangiogenic activity. Thus, nicked beta2GPI exerts dual effects on angiogenesis, that is, nicked beta2GPI promotes angiogenesis in the presence of AS4.5, whereas nicked beta2GPI inhibits angiogenesis at concentrations high enough to neutralize AS4.5. Our data suggest that plasmin-nicked beta2GPI promotes angiogenesis by interacting with plasmin-generated AS4.5 in sites of increased fibrinolysis such as thrombus.

BEta(2)-glycoprotein I (beta(2)-GPI) is proteolytically cleaved by plasmin in domain V (nicked beta(2)-GPI), being unable to bind to phospholipids. This cleavage may occur in vivo and elevated plasma levels of nicked beta(2)-GPI were detected in patients with massive plasmin generation and fibrinolysis turnover. In this study, we report higher prevalence of elevated ratio of nicked beta(2)-GPI against total beta(2)-GPI in patients with ischemic stroke (63%) and healthy subjects with lacunar infarct (27%) when compared to healthy subjects with normal findings on magnetic resonance imaging (8%), suggesting that nicked beta(2)-GPI might have a physiologic role beyond that of its parent molecule in patients with thrombosis. Several inhibitors of extrinsic fibrinolysis are known, but a negative feedback regulator has not been yet documented. We demonstrate that nicked beta(2)-GPI binds to Glu-plasminogen with K(D) of 0.37 x 10(-6) M, presumably mediated by the interaction between the fifth domain of nicked beta(2)-GPI and the fifth kringle domain of Glu-plasminogen. Nicked beta(2)-GPI also suppressed plasmin generation up to 70% in the presence of tissue plasminogen activator, plasminogen, and fibrin. Intact beta(2)-GPI lacks these properties. These data suggest that beta(2)-GPI/plasmin-nicked beta(2)-GPI controls extrinsic fibrinolysis via a negative feedback pathway loop.

beta(2)-Glycoprotein I (beta(2)GPI) is a major antigen for antiphospholipid antibodies, and its multiple in vitro functions have been reported. This glycoprotein not only down-regulates thrombin formation by inhibiting contact activation or prothrombinase activity, but also up-regulates coagulation by reducing protein C anticoagulant activity. However, the in vivo roles of beta(2)GPI remain obscure. Coagulation and fibrinolytic characteristics were investigated in individuals with beta(2)GPI deficiency. An apparently healthy woman and her brother are homozygotes for beta(2)GPI deficiency. In these patients, Russell viper venom time was shortened (40.4 seconds; normal range, 47.8 +/- 4.95 seconds), but all markers of thrombin generation and fibrin turnover were within normal ranges. Exogenous activated protein C adequately prolonged the clotting time of the beta(2)GPI-deficient plasma, and euglobulin lysis time was also normal. Thus, elevated thrombin generation, enhancement of activated protein C response, and an altered fibrinolytic system were not found in congenitally beta(2)GPI-deficient plasma. (Blood. 2000;96:1594-1595)