In the rabbit jugular vein thrombosis model (Collen et al., 1983) the thrombolytic properties of melanoma tissue-type plasminogen activator (mtPA), recombinant tPA (rtPA; Activase), and a tPA mutant that lacked the growth factor- and kringle 1-domains (designated FK2P) were compared.In vitro, FK2P had a specific activity similar to mtPA or rtPA in a spectrophotometric plasminogen activation assay in the presence of cyanogen bromide-digested fibrinogen.In a plasma clot lysis assay the threshold concentration of FK2P was, however, about 8-fold higher.In vivo, over the dose range tested (1.75 to 30 nmoles/kg), a similar, dose-dependent, thrombolytic efficacy was found for the three enzymes when given as bolus injections.When given as a 90-min infusion, FK2P was superior to rtPA at the tested dose of 15 nmoles/kg, whether given with or without a concomitant infusion of prostaglandin E1.The effects on systemic plasma fibrinogen, plasminogen, and α2-antiplasmin consumption were much larger for rtPA and FK2P than for mtPA.Similar differences were found for plasminogen and antiplasmin.The analog FK2P was cleared more slowly than mtPA or rtPA, despite a similar initial α half-life.In liver perfusion experiments, the extraction of FK2P was decreased compared to that of rtPA.During a single passage, FK2P was extracted for 22% and rtPA for 72.The data suggested that in vivo thrombolytic agent may moreover critically depend on the model chosen, and on the way the agent is administered.