We have examined the effect of thrombin on the activity of plasminogen activator (PA) and plasminogen activator-inhibitor (PA-I) in medium conditioned by primary cultures of human umbilical vein endothelial cells. PA activity was measured by fibrinolytic and esterolytic assays, and total tissue-type PA (tPA) antigen by radioimmunoassay. Net PA-I activity was assayed by titration of human urokinase esterolytic activity. Incubation of confluent endothelial cell cultures with thrombin for 24 h caused a sixfold increase in PA-I activity. The effect of thrombin was half-maximal at -0.4 U/ml (<4 nM), and required concomitant RNA and protein synthesis. The stimulation of PA-I activity required active a-thrombin and was not obtained with 'y-thrombin nor with thrombin catalytically inactivated with hirudin. Because of the excess of PA-I, PA activity was not measurable in either control or thrombin-treated cells. Thrombin did, however, increase medium concentration of tPA antigen by approximately fourfold. The thrombin-induced PA-I inhibited both tPA and urokinase, did not lose activity upon acidification, and was stable to sodium dodecyl sulfate and thiol reduction. We conclude that physiologic concentrations of thrombin increase both PA-I activity and tPA antigen in medium conditioned by human umbilical vein endothelial cells. Because there was always a several-fold increase in the net activity of free PA-I, these observations suggest that the net effect of thrombin is to decrease fibrinolytic activity in human endothelial cells. Thus, thrombin, in addition to its role in coagulation, may protect clots from premature lysis by increasing the amount of a specific fibrinolytic inhibitor.

Incubation of HTC rat hepatoma cells with the synthetic glucocorticoid dexamethasone rapidly inhibits plasminogen activator (PA) activity and reveals the presence of a specific PA inhibitor (PAI-1). To determine whether the hormonal inhibition of PA activity reflects a decrease in the amount of PA or an increased amount of the inhibitor, or both, we have assayed PA and PAI-1 immunologically. HTC PA was determined to be entirely of the tissue type (tPA), and both free and complexed antigen was quantified by a RIA using rabbit antirat tPA, with rat insulinoma tPA as tracer and standard. PAI-1 was quantified by a Western blot assay using rabbit anti-HTC PAI-1 antibody and purified HTC PAI-1 as standard. Under conditions in which dexamethasone inhibited PA activity by 90%, there was no decrease in the cellular content of tPA antigen. Paradoxically, dexamethasone increased tPA antigen approximately 1.5-fold. Under these same conditions, dexamethasone increased PAI-1 antigen 4- to 5-fold. We conclude that the glucocorticoid inhibition of tPA activity in HTC cells is not secondary to a decrease in the amount of tPA but is secondary to the induction of a specific PA inhibitor.

No abstract