An explanation for rapid enzyme-catalyzed proton abstraction from carbon acids: Importance of late transition states in concerted mechanisms

The rates of enzyme-catalyzed abstraction of protons from carbons adjacent to carbonyl or carboxylic acid groups (α-protons of carbon acids) are rapid (k_cats ∼ 10¹-10⁴ s^-1). We recently proposed that these rates can be understood if proton abstraction by an active site general basic catalyst is concerted with protonation of the carbonyl group by an active site general acidic (electrophilic) catalyst to generate an enol intermediate instead of an enolate anion (aldehyde, ketone, or thioester substrate) or dianion (carboxylic acid substrate): Gerlt, J. A.; Kozarich, J. W.; Kenyon, G. L.; Gassman, P. G. J. Am. Chem. Soc. 1991, 113, 9667. Gerlt, J. A.; Gassman, P. G. J. Am. Chem. Soc. 1992, 114, 5928. We now analyze concerted general acid-general base catalyzed enolization reactions in terms of Marcus formalism that partitions the activation energy barrier for the reaction, ΔG*, into (1) a thermodynamic component, ΔG°, associated with both the conversion of the keto tautomer of the carbon acid into its enol tautomer and the transfer of a proton from the general acidic catalyst to the general basic catalyst, and (2) an intrinsic kinetic component, ΔG*_int, the activation energy for the reaction in the absence of a thermodynamic barrier. We propose that in enzyme active sites both ΔG° and ΔG*_int, are reduced from the values that describe nonenzymatic reactions. The transition states for the enzymatic reactions are "late", i.e., the transition states resemble the enol tautomers of the substrate carbon acids, as suggested by the observation that the pK_as of the OH groups of the enols are similar to the pK₂s of the uncharged active site general acidic catalysts. The late transition states in enzyme-catalyzed reactions can be explained best by reductions in ΔG*_int, from the values that describe nonenzymatic reactions. We propose that the reduction in ΔG*_int is achieved by propositioning the electrophilic catalyst adjacent to the carbonyl group of the substrate carbon acid, thereby negating development of a negative charge on the carbonyl oxygen as the α-proton is abstracted. We propose that the reduction in ΔG° is accomplished by stabilization of an enolic intermediate via the formation of a negatively charged, short, strong hydrogen bond between the anionic conjugate base of the active site general acidic catalyst and the OH group of the enol tautomer of the substrate. Such hydrogen bonds can be formed in the bulk solvent-excluded environments of enzyme active sites since the pK₂s of the OH groups of the enol tautomers of the substrate carbon acids are similar to the pK₂s of the active site general acidic catalysts. Taken together, the possible reductions in ΔG‡_int, and ΔG° are quantitatively sufficient to explain the rapid rates of the enzyme-catalyzed reactions. We propose a "late transition state rule" that describes the requirements for concerted general acid-general base catalysis of enolization of carbon acids in enzyme active sites. This rule differs from the "libido rule" proposed by Jencks (Jencks, W. P. J. Am. Chem. Soc. 1972, 94, 4731) to describe the requirements for concerted general acid-general base catalysis of enolization of carbon acids in nonenzymatic reactions. The application of our rule to enzyme-catalyzed reactions involving successive enolization and reketonization reactions of carbon acids reveals that, while the ΔG° between bound substrate carbon acid and bound product carbon acid should approach 0, the ΔG° for formation of the bound enolic intermediate from either bound substrate carbon acid or bound product carbon acid must be endergonic. This description of the energetics of enzyme-catalyzed reactions differs in significant detail from that proposed by Albery and Knowles (Albery, W. J.; Knowles, J. R. Biochemistry 1976, 15, 5631).