The acylation and deacylation stages of the hydrolysis of N -acetyl-Phe-Gly methyl thionoester catalysed by papain and actinidin were investigated by stopped-flow spectral analysis. Differences in the forms of pH-dependence of the steady-state and pre-steady-state kinetic parameters support the hypothesis that, whereas for papain, in accord with the traditional view, the rate-determining step is the base-catalysed reaction of the acyl-enzyme intermediate with water, for actinidin it is a post-acylation conformational change required to permit release of the alcohol product and its replacement in the catalytic site by the key water molecule. Possible assignments of the kinetically influential p K (a) values, guided by the results of modelling, including electrostatic-potential calculations, and of the mechanistic roles of the ionizing groups, are discussed. It is concluded that Asp(161) is the source of a key electrostatic modulator (p K (a) 5.0+/-0.1) in actinidin, analogous to Asp(158) in papain, whose influence is not detected kinetically; it is always in the 'on' state because of its low p K (a) value (2.8+/-0.06).

The possibility of a slow post-acylation conformational change during catalysis by cysteine proteinases was investigated by using a new chromogenic substrate, N-acetyl-Phe-Gly methyl thionoester, four natural variants (papain, caricain, actinidin and ficin), and stopped-flow spectral analysis to monitor the pre-steady state formation of the dithioacylenzyme intermediates and their steady state hydrolysis. The predicted reversibility of acylation was demonstrated kinetically for actinidin and ficin, but not for papain or caricain. This difference between actinidin and papain was investigated by modelling using QUANTA and CHARMM. The weaker binding of hydrophobic substrates, including the new thionoester, by actinidin than by papain may not be due to the well-known difference in their S2-subsites, whereby that of actinidin in the free enzyme is shorter due to the presence of Met211. Molecular dynamics simulation suggests that during substrate binding the sidechain of Met211 moves to allow full access of a Phe sidechain to the S2-subsite. The highly anionic surface of actinidin may contribute to the specificity difference between papain and actinidin. During subsequent molecular dynamics simulations the P1 product, methanol, diffuses rapidly (over<8 ps) out of papain and caricain but 'lingers' around the active centre of actinidin. Uniquely in actinidin, an Asp142-Lys145 salt bridge allows formation of a cavity which appears to constrain diffusion of the methanol away from the catalytic site. The cavity then undergoes large scale movements (over 4.8 A) in a highly correlated manner, thus controlling the motions of the methanol molecule. The changes in this cavity that release the methanol might be those deduced kinetically.

The effects of increasing the content of the aprotic dipolar organic co-solvent acetonitrile on the observed first-order rate constant (k(obs)) of the pre-steady state acylation phases of the hydrolysis of N-acetyl-Phe-Gly methyl thionester catalysed by the cysteine proteinase variants actinidin and papain in sodium acetate buffer, pH 5.3, were investigated by stopped-flow spectral analysis. With low acetonitrile content, plots of k(obs) against [S]0 for the actinidin reaction are linear with an ordinate intercept of magnitude consistent with a five-step mechanism involving a post-acylation conformational change. Increasing the acetonitrile content results in marked deviations of the plots from linearity with a rate minimum around [S]0=150 microM. The unusual negative dependence of k(obs) on [S]0 in the range 25-150 microM is characteristic of a rate-determining isomerization of the free enzyme before substrate binding, additional to the five-step mechanism. There was no evidence for this phenomenon nor for the post-acylation conformational change in the analogous reaction with papain. For this enzyme, however, acetonitrile acts as an inhibitor with approximately uncompetitive characteristics. Possible mechanistic consequences of the differential solvent-perturbed kinetics are indicated. The free enzyme isomerization of actinidin may provide an explanation for the marked difference in sensitivity between this enzyme and papain of binding site-catalytic site signalling in reactions of substrate-derived 2-pyridyl disulphide reactivity probes.

The possibility of a slow post-acylation conformational change during catalysis by cysteine proteinases was investigated by using a new chromogenic substrate, N-acetyl-Phe-Gly methyl thionoester, four natural variants (papain, caricain, actinidin and ficin), and stopped-flow spectral analysis to monitor the pre-steady state formation of the dithioacylenzyme intermediates and their steady state hydrolysis. The predicted reversibility of acylation was demonstrated kinetically for actinidin and ficin, but not for papain or caricain. This difference between actinidin and papain was investigated by modelling using QUANTA and CHARMM. The weaker binding of hydrophobic substrates, including the new thionoester, by actinidin than by papain may not be due to the well-known difference in their S2-subsites, whereby that of actinidin in the free enzyme is shorter due to the presence of Met211. Molecular dynamics simulation suggests that during substrate binding the sidechain of Met211 moves to allow full access of a Phe sidechain to the S2-subsite. The highly anionic surface of actinidin may contribute to the specificity difference between papain and actinidin. During subsequent molecular dynamics simulations the P1 product, methanol, diffuses rapidly (over < 8ps) out of papain and caricain but ‘lingers’ around the active centre of actinidin. Uniquely in actinidin, an Asp142–Lys145 salt bridge allows formation of a cavity which appears to constrain diffusion of the methanol away from the catalytic site. The cavity then undergoes large scale movements (over 4.8 Å) in a highly correlated manner, thus controlling the motions of the methanol molecule. The changes in this cavity that release the methanol might be those deduced kinetically.