In the conversion of bovine plasminogen to bovine plasmin not only the expected urokinase-catalysed cleavage of Arg-557-Val-558, and the following autocatalytic cleavage separating the N-terminal peptide 1-77 from the heavy chain of plasmin, but also a cleavage at Arg-342-Met-343 between kringles 3 and 4 is seen. Here, kinetic studies of the interaction of bovine alpha(2)-antiplasmin with bovine plasmin were performed on isolated bovine midiplasmin (lacking kringles 1-3) and on bovine plasmin containing ah of the activation products from the bovine plasminogen. A series of experiments using stopped-flow fluorescence fast kinetics as well as conventional techniques suggests a reaction model in accordance with the one known for the human system. First, a tight complex (K-1 in the nanomolar range) is formed in a fast reaction step; and second, a tightening of this complex occurs in a slow reaction step. The final complex is indeed so tight (K-i less than or equal to pM), that the reaction for many practical purposes is legitimately considered irreversible. The stopped-flow method allows for the determination of reliable values of the second-order rate constant for the fast association step. At pH 7.4 and 25 degrees C, k(+1) = 1.7 x 10(6) M(-1) s(-1) was obtained in the absence and k(+1) = 0.9 x 10(6) M(-1) s(-1) in the presence of the kringles 1-3 domain of bovine plasmin. In contrast to this, substantial reductions of k(+1) were seen in the presence of concentrations of 6-amino-hexanoic acid corresponding to lysine-binding-site interactions and far too low to be attributed to active-site interactions with the bovine plasmins (for each, K-i = 42 mM). All in all, the data indicated that the lysine-binding site(s) not of kringle 1, but of midiplasmin (those of kringles 4 and 5) are regulating the inhibition reaction.