Primary structures of platelet aggregation inhibitors (disintegrins) autoproteolytically released from snake venom hemorrhagic metalloproteinases and new fluorogenic peptide substrates for these enzymes.

A hemorrhagic protein (60 kDa), HR1B, present in the venom of Trimeresurus flavoviridis is a mosaic protein consisting of an NH2-terminal metalloproteinase-domain, a disintegrin (platelet aggregation inhibitor)-like domain, and a unique COOH-terminal Cys-rich domain. Since the gross structures of HR1B and protein precursors of disintegrins, trigramin, and rhodostomin, all of which contain the metalloproteinase domain, are similar, many disintegrins so far detected in snake venoms are assumed to be autoproteolytic fragments released from precursors. In ongoing related experiments, the newly purified hemorrhagic metalloproteinases, HR1A from T. flavoviridis venom and HT-1 from Crotalus ruber ruber venom, in addition to HR1B, were autoproteolyzed, in the absence of Ca2+, at 37 degrees C for 3-12 h. Under these conditions, HR1A, HR1B, and HT-1 each released a single major fragment of 32, 34, and 31 kDa, respectively. The entire amino acid sequences of the isolated fragments indicated the presence of disintegrin-like and Cys-rich domains in the COOH-terminal regions of HR1A, HR1B, and HT-1, respectively. It seems likely that so-called disintegrins probably originate from various metalloproteinases present in venom. On the bases of peptide sequences close to the autoproteolytic cleavage sites of these metalloproteinases and the sites of fibrinogen cleaved by these enzymes, we synthesized new intramolecularly quenched fluorogenic peptide substrates. Among the 10 peptides tested, 2-aminobenzoyl (Abz)-Ser-Pro-Met-Leu-2,4-dinitroanilinoethylamide (Dna) proved to be the best substrate for venom metalloproteinase, as deduced from kinetic analyses.