Human alpha-1-proteinase inhibitor (A1-PI), which is a serine protease inhibitor and which in vivo inhibits the activity of neutrophil elastase, undergoes guanidine-HCl (Gu) induced biphasic unfolding with an intermediate state in 1.5 M Gu at 25 deg C. The intermediate state appears transiently stable (over a 1-2 h period) primarily with the formation of monomeric and dimeric intermediates that further polymerize more slowly. Dr. Marszal has folded these smaller intermediates by dilution with buffer and observed that in the mixture of folded species, folded dimer, but not folded monomer, further polymerizes. She has confirmed that this species is dimer by MALDI-TOF measurements and shown that it is disulfide linked (through the single sulfhydryl in each monomer). Dr. Du has established that ca. 80% of the A1-PI monomer has free sulfhydryl and that the dimer exhibits no inhibitory activity. Importantly, Dr. Marszal has demonstrated that the species generated by polymerization of the purified dimer appear to be multiples of it, i.e. no trimer is formed. Furthermore, a reactive center loop (RCL) peptide, which has the same sequence at that of the RCL of A1-PI, which normally targets the active site of neutrophil elastase, binds to the monomer, but appears not to bind to the dimer, thereby preventing the polymerization of the monomeric intermediate but not of the folded dimer. The inhibition of polymerization of monomer by the peptide has been previously reported and used, in part, to support a model of polymerization involving the insertion of the RCL of one A1-PI molecule between the central strands of a beta-sheet on an adjacent molecule thereby forming a linear chain (i.e. a loop-sheet model for polymerization). Electron micrograph data demonstrate that polymers formed from purified dimer are linear bead-like structures that are similar to those formed from monomer (by heating at elevated temperature) and those isolated from inclusion bodies in hepatocytes from patients with a mutant A1-PI that is poorly secreted. The structural constraints of the disulfide linked dimer and its ability to form linear polymers are difficult to reconcile with the proposed loop-sheet model. We propose that formation of the disulfide linked dimer may involve RCL burial, not insertion, and that polymers of the dimer may form by interaction of beta-sheets exposed on the outer surfaces of the dimer. The model proposed here, involving disulfide linked dimer, is an alternate mechanism since under certain reaction conditions, different mechanisms of polymerization may compete.
