1. 2-(N'-Acetyl-D-phenylalanyl)hydroxyethyl 2'-pyridyl disulphide (compound IV) (m.p. 59 degrees C; [alpha]D20 -6.6 degrees (c 1.2 in methanol)) was synthesized. 2. The results of a study of the pH-dependence of the second-order rate constant (k) for its reaction with the catalytic-site thiol group (Cys-25) of papain (EC 3.4.22.2) together with analogous kinetic data for the reactions of related time-dependent inhibitors, notably the L-enantiomer of compound (IV) (compound III) and the L- and D-enantiomers of 2-(N'-acetylphenylalanylamino)ethyl 2'-pyridyl disulphide (compounds I and II respectively), were used to assess the contributions of the (P1)-NH ... O = C < (Asp-158) and (P2) > C = O ... H-N-(Gly-66) hydrogen bonds and enantiomeric P2-S2 hydrophobic contacts in two manifestations of dynamic molecular recognition in papain-ligand association: (a) signalling to the catalytic-site region to provide for a (His-159)-IM(+)-H-assisted transition state and (b) the dependence of P2-S2 stereoselectivity on hydrogen-bonding interactions outside the S2 subsite. The analysis involved determination of the reactivities of individual ionization states of the reactions (pH-independent rate constants, k) and associated macroscopic pKa values and difference kinetic specificity energies (delta delta GKS = -RT1n(k1/k2), where k1 is the pH-independent second-order rate constant for reaction with one inhibitor and k2 is the analogous rate constant in the same ionization state for reaction with another inhibitor so that, when the structural change provides that k2 > k1, delta delta GKS is positive. 3. The kinetic data further illuminate the nature of the interdependence of binding interactions in papain first noted by Kowlessur, Topham, Thomas, O'Driscoll, Templeton & Brocklehurst [(1989) Biochem. J. 258, 755-764] in the S2 subsite, S1-S2 intersubsite and catalytic-site regions. Of particular note is the apparent dependence of the binding of the N-Ac-D-Phe moiety on the binding of the leaving group to (His-159)-Im+H and the fact that the resulting rate enhancement is more effective when (P1)-N-H is absent than when it is present. This result revealed by kinetic analysis goes beyond the conclusion suggested by model building that it is possible to make all of the binding contacts in complexes involving the D-enantiomers [(II) and (IV)] as in those involving the L-enantiomers [(I) and (III)].(ABSTRACT TRUNCATED AT 400 WORDS)

1. Values of the kinetic specificity constant, kcat./Km, for the hydrolysis of N-acetyl-L-phenylalanylglycine 4-nitroanilide (I) and of its D-enantiomer (II) catalysed by ficin (EC 3.4.22.3) and by actinidin (EC 3.4.22.14) at pH 6.0, I 0.1 mol/l, 8.3% (v/v) NN-dimethylformamide and 25 degrees C were determined by using initial-rate data with [S] much less than Km and weighted nonlinear regression analysis as: for ficin, (kcat./Km)L = 271 +/- 6 M-1.s-1, (kcat./Km)D = 2.9 +/- 0.1 M-1.s-1, and for actinidin (kcat./Km)L = 13.3 +/- 0.7 M-1.s-1, (kcat/Km)D = 0.34 +/- 0.01 M-1.s-1.2. These data and analogous values for the corresponding reactions catalysed by papain (EC 3.4.22.2), (kcat./Km)L = 2064 +/- 31 M-1.s-1, (kcat./Km)D = 5.5 +/- 0.1 M-1.s-1, demonstrate marked variation in stereochemical selectivity for substrates (I) and (II) among the three cysteine proteinases with the following values for the index of stereochemical selectivity Iss = (kcat./Km)L/(kcat./Km)D: for papain, 375; for ficin 93; for actinidin 39. 3. Model building suggests ways in which, for the papain-catalysed reactions, binding interactions involving the extended acyl groups of the substrates may need to change as the reaction proceeds from adsorptive complex (ES) to tetrahedral intermediate (THI) before its rate-determining, general acid-catalysed collapse to acylenzyme intermediate. In particular, satisfactory alignment in the catalytic site at the THI stage of the acylation process appears to demand rotation of the substrate moiety about its long axis. 4. The different consequences of this rotation for the L- and D-enantiomers suggest that for closely related systems the greater the extent of this rotational adjustment the greater would be the value of Iss.5. For the actinidin-substrate combinations, model building suggests that even at the ES complex stage of catalysis it is not possible to approach optimized P2-S2 contacts and the three hydrogen-bonding interactions deduced for papain-ligand complexes in the absence of significant movement of protein conformation. Possible binding modes in which some of the interactions deduced for papain are relaxed are discussed. Consideration of postulated binding modes in the various transition states is shown to account for the order of reactivity reflected in values kcat./Km for the four reactions involving papain (Pap) and actinidin (Act) with the L- and D-enantiomeric substrates: Pap-L much greater than Act-L greater than Pap-D much greater than Act-D.(ABSTRACT TRUNCATED AT 400 WORDS)

To understand molecular recognition in enzyme-substrate and enzyme-time-dependent inhibitor systems, it is necessary to investigate not only binding per se but also the possibility that binding may control covalency changes that take place in the catalytic site. The cysteine proteinase family of enzymes (reviewed in [l]) provides valuable opportunities for investigation of the coupling of different binding interactions (electrostatic, hydrogen bonding, van der Waals and hydrophobic) with each other and with catalytic site chemistry [2,3]. Such studies are facilitated by the existence of (a) significant differences in both structure and chemical behaviour throughout members of the enzyme family and (b) the common catalytic site (Cys)-S-/(His)-Im+H ion pair. The latter is the device used to provide the nucleophilic and general acid or hydrogen bonding/electrostatic characteristics used in the transformation of the reaction centre of substrate or inhibitor to products.We are continuing to develop 2-pyridyl disulphide reactivity probes (R-S-S-2-Py), in which R may provide a variety of binding opportunities, for the study of the binding site-catalytic site coupling in the cysteine proteinases. In the absence of signalling to the catalytic site occasioned by binding of R, the pHk profile contains a bell-shaped component with k maximal at pH3-4. This characteristic profile shape results from (a) the coexistence in acidic media of significant concentrations of Cys-S~ /His-Im+H ion pair and the formally hydronated (protonated) form of the probe (R-S-S-2-Py+H) and (b) the lack of substantial association of the 2-pyridyl N atom with the His-Im+H component of the ion pair in neutral media. In reactions of papain, binding of the P,-P, amide bond of 2-(acetamido)-ethyIL2'-pyridyl disulphide [I] and particularly of the L-Phe side chain and the (PJ > C = O of 2-(N'-acetyl-L-phenylalanylamino)-ethyl-2'-pyridyldisulphide (11) results in substantial reaction via an imidazolium-ion assisted transition state (111), recognised by marked changes in the shape of the pH-k profile with k maximal at pH6-7 instead of at pH3-4. It is a striking observation that these marked changes in profile shape do not occur also for the analogous reactions of actinidin [2,3]. This was unexpected because the crystal structures of papain Table 1. Second-order raie Constanis (8 for lhe reactbns o! ficin %i!h ihrec Fpyridyl disulphides at 25%. 10.1M a1 (a) pH3.0 (b) pH4.0 (c) pti6 0 and (01 pH1O.O Prdbe k(u-'s-')