Molecular evolution of plasminogen activator inhibitor-1 functional stability.

Abstract

Plasminogen activator inhibitor‐1 (PAI‐1) is a member of the serine protease inhibitor (serpin) supergene family and a central regulatory protein in the blood coagulation system. PAI‐1 is unique among serpins in exhibiting distinct active and inactive (latent) conformations in vivo. Though the structure of latent PAI‐1 was recently solved, the structure of the short‐lived, active form of PAI‐1 is not known. In order to probe the structural basis for this unique conformational change, a randomly mutated recombinant PAI‐1 expression library was constructed in bacteriophage and screened for increased functional stability. Fourteen unique clones were selected, and shown to exhibit functional half‐lives (T1/2S) exceeding that of wild‐type PAI‐1 by up to 72‐fold. The most stable variant (T1/2 = 145 h) contained four mutations. Detailed analysis of these four mutations, individually and in combination, demonstrated that the markedly enhanced functional stability of the parent compound mutant required contributions from all four substitutions, with no individual T1/2 exceeding 6.6 h. The functional stability of at least eight of the remaining 13 compound mutants also required interactions between two or more amino acid substitutions, with no single variant increasing the T1/2 by > 10‐fold. The nature of the identified mutations implies that the unique instability of the PAI‐1 active conformation evolved through global changes in protein packing and suggest a selective advantage for transient inhibitor function.

A Deletion Mutant of Vitronectin Lacking the Somatomedin B Domain Exhibits Residual Plasminogen Activator Inhibitor-1-binding Activity

“…Multimeric vitronectin was prepared by denaturation of the protein in 8 M urea for 2 h at room temperature followed by dialysis into PBS (140 mM NaCl, 3 mM KCl, 10 mM Na 2 HPO 4 , 2 mM KH 2 PO 4 , pH 7.4). Recombinant human wild-type PAI-1 and recombinant stable PAI-1 mutant, 14-1B (38), used in the ELISA assays were both purchased from Molecular Innovations, Inc. PAI-1 used in the Biacore experiments and latency transition assays was produced according to the method outlined in Jensen et al (28). The PAI-1 protein preparations were tested for activity using a urokinase inhibition assay; all behaved similarly and fully inhibited urokinase.…”

“…The binding of vitronectin to PAI-1 decreases the rate at which PAI-1 converts to this latent form. Stable forms of PAI-1 that do not collapse to the latent form at an appreciable rate have been engineered by incorporating amino acid replacements at a variety of loci, including the core ␤-sheet (16,27,38,54,55). The stable variant of PAI-1 that has been best characterized and has yielded the crystallographic structure of active PAI-1 (55) is the 14-1B mutant with the mutations N152H, K156T, Q321L, and M356I (16,38).…”

Subtle structural differences between human and mouse PAI-1 reveal the basis for biochemical differences

“…6B). The different position of the RCL is also correlated to the different conformation of the C-terminal part of Berkenpas et al (1995). hA: in latent moPAI-1 hA is bent more to the core of the protein compared to hA in latent huPAI-1 (Fig.…”

“…1.9 h vs. 3.6 h for moPAI-1-wt and moPAI-1-D150H-K154T-Q319L, respectively) ( Table 2). Alternatively a single mutation (Ile at position 91 to Leu) resulting in 9-fold increased half-life in huPAI-1 (huPAI-1-I91L, 18.4 h; (Berkenpas et al, 1995) did not affect the stability of mo-PAI-1 (i.e. half-life of moPAI-1-I91L, 2.1 h).…”

“…To test this hypothesis, mutations that stabilize active human PAI-1 (i.e. N150H-K154T-Q319L-M354I) (Berkenpas et al, 1995) were introduced into mouse PAI-1 (resulting in moPAI-1-D150H-K154T-Q319L, since moPAI-1 has already an Ile at position 354). HuPAI-1-N150H-K154T-Q319L-M354I has a 72-fold increased half-life (145 h), compared to huPAI-1-wt (2 h).…”