Molecular modeling of substrate-enzyme reactions for the cysteine protease papain.

AM1 quantum mechanical reaction coordinate (RC) calculations were run to simulate the rate-limiting deacylation (hydrolysis) reaction for a series of para-X-PhC(O)NHCH2-C(Y)-S-papain intermediates, where X = OCH3, CH3, H, Cl, NO2 and Y = O (thioester) or S (dithioester), for which a large body of structural, kinetic, and spectroscopic data is available. Several reaction zones, in particular the so-designated Large Zone and Small Zone, were extracted for these RC simulations from the fully solvated and energy-minimized X-ray crystal structure of papain (pdb9pap) bound to the appropriate substrate moiety. The major structural difference between these two zones was the absence of the oxyanion hole in the latter. For both the thioester and dithioester cases, the calculated Ea value associated with the parent (X = H) acyl-enzyme intermediate was lower by ca. 10 kcal/mol for the Large Zone than for the Small Zone. The magnitude of this difference suggests that the oxyanion hole plays a functional if not essential role in stabilizing the anionic tetrahedral intermediate with the cysteine proteases. The calculated Ea value was lower by ca. 10 kcal/mol for the thioester [-C(O)-S-] than for the corresponding dithioester [-C(S)-S-], in qualitative agreement with kinetic data for this series of substrates which reveal that the specific rate constant for deacylation k3 is ca. 60 times larger for the former. This difference is also consistent with both AM1 and 6-31G* calculations on model intermediates, which indicate that the weaker polarity of the dithioester compared with the thioester [i.e., -C(<--S)-S-versus-C(-->O)-S-] renders the former a much poorer site for nucleophilic attack. The anionic tetrahedral intermediate is energetically more stable for the dithioester than for the corresponding thioester, a finding that is discussed in terms of its kinetic and mechanistic implications. The mode of attack by the H2O nucleophile is "concerted" rather than "sequential" in terms of the order of proton abstraction by His-159 and nucleophilic attack on the acyl-enzyme intermediate. While the presumably key Sthiol . . . N nonbonded contact remained almost constant (ca. 2.90 A) up to formation of the [TS] structure, the substrate torsion angles phi and psi rotated significantly as the hybridization around the reaction site transforms from sp2 to sp3 during formation of the tetrahedral intermediate. The AM1-calculated frontier molecular orbitals for model thioester and dithioester acyl-enzyme intermediates generally associate the HOMOs with the reaction site and the LUMOs with the benzamide moiety. Computer graphics images corroborate our view that, in relation to the Sthiol . . . N interaction, the HOMOs and LUMOs should be identified, respectively, with Sthiol and N rather than the reverse, as suggested by other workers.