Mapping the extended substrate binding site of cathepsin G and human leukocyte elastase. Studies with peptide substrates related to the (α1)-protease inhibitor reactive site

Abstract

The kinetic constants for the hydrolysis of a series of 4-nitroanilide substrates by human leukocyte (HL) elastase and cathepsin G, porcine pancreatic elastase, and bovine chymotrypsin at pH 7.50 are reported. HL elastase and cathepsin G are currently thought to be the agents responsible for destruction of the lung in the disease emphysema. MeO-Suc-Ala-Ala-Pro-Val-NA is an excellent substrate for HL elastase and is not hydrolyzed by cathepsin G. The MeO-Suc-group increases the solubility of a substrate relative to the acetyl group. With HL elastase, this structural change increases the reactivity of the enzyme toward both 4-nitroanilide substrates and chloromethyl ketone inhibitors. This indicates that HL elastase is interacting with at least 5 residues of a substrate (or inhibitor). Cathepsin G prefers P5 groups which are negatively charged such as Suc-, Suc(4F)-, Glt-, or Mal-. This enzyme, in common with many other serine proteases, cannot accept a Pro residue at its S3 subsite. One of the better substrates for cathepsin G, Suc-Ala-Ala-Pro-Phe-NA, was not hydrolyzed by HL elastase. These tools should be useful in the study of the biological function of HL elastase and cathepsin G. Two tetrapeptide 4-nitroanilide substrates related to the reactive site of the plasma α1-protease inhibitor (α1-antitrypsin) were studied. Both have a P1 Met residue and one, MeO-Suc-Ala-Ile-Pro-Met-NA, has the exact sequence of the P4 to P1 residues at the proteolysis site of α1-PI (Johnson, D.A., and Travis, J. (1978) J. Biol. Chem. 253, 7142-7144). Both MeO-Suc-Ala-Ala-Pro-Met-NA and MeO-Suc-Ala-Ile-Pro-Met-NA react with cathepsin G, HL elastase, and bovine chymotrypsin. The former is in fact the best 4-nitroanilide substrate of cathepsin G yet reported. Oxidation of MeO-Suc-Ala-Ala-Pro-Met-NA yielded two diastereomeric sulfoxides. Neither are bound to or was hydrolyzed by HL elastase or cathepsin G. Both reacted poorly with bovine chymotrypsin. In the preceding paper, Johnson and Travis (Johnson, D., and Travis, J. (1979) J. Biol. Chem. 254, 4022-4026) show that oxidation of α1-PI destroys its inhibitory activity. In concert, our results indicate that oxidation of the P1 Met of α1-PI is capable of destroying its reactivity toward most serine proteases. Oxidation of α1-PI by some component in cigarette smoke would offer one explanation in molecular terms for the link between smoking and emphysema.