Tissue plasminogen activator mutants lacking the growth factor domain and the first kringle domain: II: Enzymatic properties in plasma and in vivo thrombolytic activity and clearance rates in rabbits

Abstract

The in vitro fibrinolytic activities and fibrin specificity in human plasma as well as in vivo fibrinolytic and pharmacokinetic properties in rabbits were studied for six genetically engineered variants of human tissue-type plasminogen activator (t-PA). All variants lacked the growth factor (G) domain and the first kringle (K1) domain and varied in the presence or absence of the finger (F) domain. Additional specific site modifications were Lys 277→Val and Asn 448→Gln. The variants were denoted K2P, K2P(Val277), K2P(Gln448), FK2P, FK2P(Val277) and FK2P(G1n448). The highest specific activity in vitro in a plasma clot lysis assay was observed for the wild type t-PA and decreased in the order FK2P(Val277)≈K2P(Val277)>FK2P≈K2P>K2P(Gln448)>FK2P(Gln448). All variants displayed a time lag phase of 30–60 min before fibrinolysis was detected. This lag phase was far less pronounced for the wild type molecule. The fibrin specificity of the variants in vitro was studied by analyses of clottable fibrinogen and consumption of α 2-antiplasmin after incubating human plasma with fibrinolytically equipotent doses of the different activators. Variants K2P(Val277) and FK2P(Val277) were as fibrin specific as t-PA, whereas K2P, FK2P and the non-glycosylated K2P(Gln448) consumed more fibrinogen and α 2-antiplasmin at fibrinolytically equipotent doses. However, all t-PA variants were more fibrin specific than streptokinase. The thrombolytic and pharmacokinetic properties of t-PA and five of the variants were examined in a rabbit jugular vein thrombosis model. The thrombolytic activity in vivo following a bolus injection of 0.3mg/kg was, in order of descent, K2P>FK2P>K2P(Val277)≈FK2P(Val277)>K2P(Gln448)≈rt-PA. Two of the variants and rt-PA were investigated at two additional doses (0.6 and 1.2mg/kg). Their ED 50 (dose required to lyse 50% of the thrombus within 3h), as calculated from the dose-response curve, were 1.4, 1.25 and 0.4 mg/kg for rt-PA, K2P(Val277) and K2P, respectively. When rt-PA and the variants were compared at equivalent doses (weight) the variants were found to cause a significantly higher consumption of fibrinogen and α 2-antiplasmin. However, when the compounds were compared at a thrombolytically equivalent dose the two most potent variants, K2P and FK2P, caused the same haemostatic effect as t-PA, whereas the less potent variants were found to have a reduced fibrin specificity in vivo. All of the variants had a significantly slower clearance and a longer plasma half-life compared to rt-PA. Differences in pharmacokinetic profile were also seen among the five variants, as those containing the finger domain had a faster clearance and a shorter half-life than those without the finger.