Characterization of covalently bound enzyme inhibitors as transition-state analogs by protein stability measurements: phosphonate monoester inhibitors of a beta-lactamase.

An experimental method is described for determining whether a covalent enzyme-inhibitor complex has the properties expected of a transition-state analog. The method involves a comparison of the noncovalent interaction energies between the enzyme and the inhibitor on one hand (determined from protein denaturation thermodynamics) and the analogous transition state on the other (determined from kinetic measurements). These two quantities should presumably be large (in comparison with the interaction energies of substrates or reaction intermediates) and close to equal for a good transition state analog; the former is seen dramatically in a large increase in protein stability. The method is absolute in the sense that it does not require a crystal structure of the inhibited enzyme or any preconceptions as to the mechanism of action of the enzyme except those which led to adoption of the potential transition state analog and which might turn out to be right or wrong. In this paper the method is quantitatively applied to the inhibition of the Staphylococcus aureus PC1 beta-lactamase by phosphonate monoesters. It is concluded that the enzyme-inhibitor complex in this case is likely to be a good transition-state mimic. Therefore, mechanistic interpretation of the crystal structure of the complex can be made with more confidence. A semiquantitative assessment of the situation with serine proteinases is also made. It is concluded, in agreement with predictions based on the generally accepted mechanism and on crystal structures, that anionic, but not neutral, phosph(or/on)yl derivatives are good transition-state analogs.