Norman M. Schechter

RECOMBINANT EXPRESSION OF HUMAN MAST CELL PROTEASES CHYMASE AND TRYPTASE

Expression of recombinant human chymase and tryptase was achieved in a baculovirus-insect cell system using a fusion protein construct. Recombinant baculovirus was produced with DNA coding for a NH2-ubiquitin-chymase-COOH or NH2-ubiquitin-tryptase-COOH fusion protein inserted immediately downstream of the signal sequence for the secreted envelope protein, glycoprotein 67. In each construct, the natural prepropeptide sequence of the protease was replaced by the amino acid sequence for the enterokinase cleavage site of trypsinogen. High Five insect cells infected with either of the modified baculovirus produced mg quantities of each fusion protein per liter of culture. Treatment of the chymase-fusion protein with enterokinase or the tryptase-fusion protein with enterokinase in the presence of a highly charged polysaccharide (dextran sulfate or heparin) produced enzymatically active proteases with properties of the native enzymes. A procedure for the purification of mg quantities of recombinant chymase from infected-cell medium is presented.

Purification and identification of two serine class proteinases from dog mast cells biochemically and immunologically similar to human proteinases tryptase and chymase

Serine class proteinases with trypsin-like and chymotrypsin-like specificity were purified from dog mastocytoma tissue. An antiserum was produced against the chymotrypsin-like proteinase. The antiserum reacted with mast cells in skin sections prepared from normal dogs consistent with the proteinase being a mast cell constituent. The antiserum also cross-reacted with the major chymotrypsin-like proteinase isolated from normal dog skin and partially cross-reacted with human skin chymase. No cross-reaction was detected with rat chymase. The trypsin-like proteinase from dog mastocytoma tissue was similar to tryptase isolated from human skin. It had a similar subunit structure, was not inhibited by many protein proteolytic enzyme inhibitors, bound to heparin, and reacted strongly with antiserum against human tryptase. Antiserum against human tryptase also reacted with mast cells in skin sections prepared from normal dog skin. No immunocytochemical labeling of rat skin mast cells was observed with anti-human tryptase. These studies establish the presence of a trypsin-like and chymotrypsin-like proteinase in dog skin mast cells and provide immunological evidence which suggests that both proteinases are more closely related to human than rat mast cell proteinases. These immunological and biochemical relationships are important when comparing the roles of these proteinases in different animals.

Heterogeneity in serpin-protease complexes as demonstrated by differences in the mechanism of complex breakdown

Serpins trap their target proteases in the form of an acyl-enzyme complex. The trap is kinetic, however, and thus serpin-protease complexes ultimately break down, releasing a cleaved inactive serpin and an active protease. The rates of this deacylation process vary greatly depending on the serpin-protease pair with half-lives ranging from minutes to months. The reasons for the diversity in breakdown rates are not clearly understood. In the current study, pH and solvent isotope effects were utilized to probe the mechanism of breakdown for an extremely stable complex and several unstable complexes. Two different patterns for the pH dependence of k(bkdn), the first-order rate constant of breakdown, were found. The stable complex, which breaks down at neutral pH with a half-life of approximately 2 weeks, exhibited a pH-k(bkdn) profile consistent with solvent-hydroxide ion mediated ester hydrolysis. There was no evidence for the participation of the catalytic machinery in the breakdown of this complex, suggesting extensive distortion of the active site. The unstable complexes, which break down with half-lives ranging from minutes to hours, exhibited a bell-shaped pH profile for k(bkdn), typical of the pH-rate profiles of free serine proteases. In the low to neutral pH range k(bkdn) increased with increasing pH in a manner characteristic of His57-mediated catalysis. In the alkaline pH range a decrease in k(bkdn) was observed, consistent with the titration of the Ile16-Asp194 salt bridge (chymotrypsinogen numbering). The alkaline pH dependence was not exhibited in pH-rate profiles of free or substrate-bound HNE, indicating that the salt bridge was significantly destabilized in the complexed protease. These results indicate that breakdown is catalytically mediated in the unstable complexes although, most likely, the protease is not in its native conformation and the catalytic machinery functions inefficiently. However, a mechanism in which breakdown is determined by the equilibrium between distorted and undistorted forms of the complexed protease cannot be completely dismissed. Overall, the results of this study suggest that the protease structure in unstable complexes is distorted to a lesser extent than in stable complexes.

Potent Bivalent Inhibition of Human Tryptase-? by a Synthetic Inhibitor

Abstract Human tryptase-? (HT?) is a unique serine protease exhibiting a frame-like tetramer structure with four active sites directed toward a central pore. Potent inhibition of HT? has been attained using CRA-2059. This compound has two phenylguanidinium head groups connected via a linker capable of spanning between two active sites. The properties of the CRA-2059:HT? interaction were defined in this study. Tightbinding reversible inhibition was observed with an inhibition constant (Ki) of 620 pM, an association rate constant of 7×07 M -1s-1 and a relatively slow dissociation rate constant of 0.04 s-1. Bivalent inhibition was demonstrated by displacement of paminobenzamidine from the primary specificity pocket with a stoichiometry, [CRA-2059]0/[HT?]0, of 0.5. The potency of the bivalent interaction was illustrated by CRA-2059 inhibition of HT?, 24% or 53% inhibited by preincubation with an irreversible inhibitor. Two interactions were observed consistent with mono and bivalent binding; the Ki value for bivalent inhibition was at least 104-fold lower than that for monovalent inhibition. Comparison of the affinities of CRA-2059 and phenylguanidine for HT? finds an approximate doubling of the free energy change upon bivalent binding. This doubling suggests that the linker portion minimally hinders the binding of CRA-2059 to HT?. The potency of CRA-2059 is thus attributable to effective bivalent binding.