A substrate-based approach to convert SerpinB1 into a specific inhibitor of proteinase 3, the Wegener's granulomatosis autoantigen.

The physiological and pathological functions of proteinase 3 (PR3) are not well understood due to its close similarity to human neutrophil elastase (HNE) and the lack of a specific inhibitor. Based on structural analysis of the active sites of PR3 and HNE, we generated mutants derived from the polyvalent inhibitor SerpinB1 (monocyte/neutrophil elastase inhibitor) that specifically inhibit PR3 and that differ from wt-SerpinB1 by only 3 or 4 residues in the reactive center loop. The rate constant of association between the best SerpinB1 mutant and PR3 is 1.4 × 10⁷ M⁻¹ · s⁻¹, which is ∼100-fold higher than that observed with wt-SerpinB1 and compares with that of α1-protease inhibitor (α1-PI) toward HNE. SerpinB1(S/DAR) is cleaved by HNE, but due to differences in rate, inhibition of PR3 by SerpinB1(S/DAR) is only minimally affected by the presence of HNE even when the latter is in excess. SerpinB1(S/DAR) inhibits soluble PR3 and also membrane-bound PR3 at the surface of activated neutrophils. Moreover, SerpinB1(S/DAR) clears induced PR3 from the surface of activated neutrophils. Overall, these specific inhibitors of PR3 will be valuable for defining biological functions of the protease and may prove useful as therapeutics for PR3-related inflammatory diseases, such as Wegener's granulomatosis.