α2-Antiplasmin (α2AP) is the primary physiological plasmin inhibitor in human plasma. The inhibition is rapid (second order rate constants (k1) are expressed as M−1 s−1 ) (k1 = 2 × 107) and occurs as the consequence of an irreversible 1:1 stoichiometric complex formation; the exact nature of and the forces involved in complex formation are not fully understood. In fact, what makes α2AP an inhibitor, rather than simply a substrate remains unresolved. Recently, we deduced the primary structure of α2 AP from the sequence of its cDNA. 95%of this sequence was confirmed by amino acid (aa) sequence analysis of naturalα2 AP (α2 AP)? The 452 aa molecule contains 2 disulfide bonds and 4 glycosylated Asn residues, aa sequence alignment confirmed α2AP's membership in the Serpin family. The reactive site sequence as determined by NH2 - and COOH-terminal aa sequence analysis of the plasmin-modified inhibitor and the released M−r ∼ 8000 peptide is Met362-Ser363-Arg364-Met365-Ser366, P3-P2-P1-P'1-P'2, respectively.Natural and engineered P1 residue substitutions in the Serpin α2 -antitrypsin ( α2 AT) have shown altered specificities and efficiencies. To further examine the role of P and P' residues in determining Serpin specificity, in the present study we have by site-directed mutagenesis, deleted (△) the P'l-Met365 residue of a AP thereby producing a recombinant (r) inhibitor (r α2 AP△Met365) whose putative new reactive site mimics that of antithrombin III (ATIII) and a AT-Pittsburgh (Pl-Arg-P'1-Ser). A second variant was constructed (ra2AP△Arg364) in which the Pl-Arg364 residue was deleted, producing the new sequence Met362-Ser363-Met364-Ser365, containing 2 potential sites analogous to the Pl-P'l, Met-Ser reactive site of α2 AT. The variants and r α2 AP were expressed in CH0 cells, purified and compared with n α2 AP, α2AT and ATIII for the ability to inhibit plasmin, thrombin, trypsin and elastase. n α2 AP and r α2 AP had nearly identical inhibition constants and like ATIII did not inhibit neutrophil elastase. Without heparin both α2 APs and ATIII inhibited thrombin moderately (k1 = 2 to 4× 103 ). Bovine trypsin was neutralized by the α2 APs with k1 = 3 × 106 and by ATIII with k1 = 1 × 105. The α2APs inhibited plasmin (k1 = 2 ×107 ) much more efficiently than ATIII (K1 =2 × 103 ). In contrast, was a highly effective antielastase (k1 = 1 × 107 ), a poor plasmin and thrombin inhibitor ancl inhibited bovine trypsin with = 2 × 10. As reported by others, α2 AT-Pittsburg has greatly reduced antielastase activity and greatly enhanced antithrombin activity. Analysis of ra APAMet365 revealed little change in activity toward plasmin, trypsin and elastase. Thus, α2 AP has no absolute requirement for Met .in the P'l position in order to effectively inhibit plasmin and trypsin. The other P^ subsites appear to be spatially flexible as deletion of the natural P'l residue must displace them. Contrary to prediction a 20-fold decrease in antithrombin activity was observed rather than an enhanced activity. Analysis of rα2 AP△Arg364 showed that it is unreactive with plasmin, trypsin and thrombin, but that it has acquired a significant antielastase activity (k1 = 1.5 × 105). The exact PI residue(s) has not been determined but removal of the bulky basic Arg364 may have resulted in accessibility of the predicted reactive site(s) peptide bond(s) Met362-Ser363 or Met364-Ser365 to the active site cleft of elastase. α2AP'Enschede', a natural mutant with deficient antiplasmin activity, was shown to contain an Ala insertion between aa 353 and 357, 7 to 10 positions NH2-terminal to its reactive site (Holmes et al., this meeting). This mutation results in conversion of α2 AP'Enschede' from an inhibitor to a substrate that retains a high affinity for the active site of plasmin.
Tailoring the effect of antithrombin-targeting therapy in haemophilia A using in silico thrombin generation