Rapid Enzyme-Catalyzed Heterolytic C-H Bond Cleavage by a Base Strength Amplification Mechanism: A Theoretical Examination of the Mechanism of Oxidation of Vitamin K Ya-Jun Zheng and Thomas C. Bruice* Department of Chemistry, UniVersity of California Santa Barbara, California 93106 ReceiVed September 8, 1997 Vitamin K is the essential cofactor for vitamin K-dependent γ-glutamyl carboxylase, which catalyzes the conversion of N-terminal glutamates (Glu) in proteins of the blood clotting cascade to corresponding γ-carboxyglutamates (Gla). 1 The carboxylation catalyzed by γ-glutamyl carboxylase is accompa- nied by concomitant oxidation of the hydroquinone form of vitamin K to its quinone 2,3-epoxide. Recent studies have demonstrated that the catalytic carboxylation occurs via formation of a carbanion intermediate. 1,2 The C-H bond which is ionized is R to the terminal carboxyl group of the involved Glu with a pK a of around 22. 3 Thus, it is neither thermodynamically nor kinetically favorable for any protein base to abstract the methylene hydrogen adjacent to the γ-carboxylate group unless the enolate anion intermediates can be effectively stabilized. 4 According to the reports of Paul Dowd and co-workers, 5-8 the oxidation of vitamin K is supposed to provide the needed strong base for this difficult proton abstraction. The process for the generation of this base has been termed base strength amplification. 8 Questions still remain as to exactly what is the strong base involved in the proton abstraction. There are suggestions that the base is 5 or HO - generated by dissociation of 5 to 6 (Scheme 1). Here we report a theoretical investigation 9 of the possible intermediates involved in the oxidation of vitamin K. On the basis of the present study, an alternative pathway is suggested which is more consistent with recent experimental observations. Scheme 1 displays two possible pathways for the formation of vitamin K 2,3-epoxide from the KH - anion. Pathway A was proposed by Dowd and co-workers on the basis of studies on the model compound 2,4-dimethylnaphthol. 5-8 Pathway B is not feasible with 2,4-dimethylnaphthol but is a reasonable pathway for the oxidation of KH - . Figure 1 shows the calculated geometry for each species involved in the oxidation of the reduced vitamin K. Addition of O 2 to compound 2 could occur at either C-2 or C-4 positions, forming the corresponding peroxides 3 and 7. According to our calculations, peroxide 3 is favored in a vacuum since the negatively charged oxygen in 3 can be stabilized through intramolecular hydrogen bonding. Peroxide 3 is about 21.1 kcal/ mol lower in energy than 7. The initially formed peroxide intermediate 3 is not stable, and it rearranges to a dioxetane intermediate 4, which then forms a much more stable epoxide alkoxide intermediate 5. The dioxetane intermediate 4 is less stable than 3 by 10.4 kcal/mol. Intermediate 4 is separated from the alkoxide intermediate 5 by a small barrier of 3.9 kcal/mol. The alkoxide intermediate 5 is about 54.0 kcal/mol more stable than intermediate 4. At the HF/6-31+G(d) level, intermediate 7 is about 30.5 kcal/mol lower in energy than the transition state (9) for the interconversion of 4 to 5; however, at the B3LYP/6- 31+G(d) level, it is higher in energy than 9 by 6.7 kcal/mol. Clearly, electron correlation has a significant effect on the calculated energetics. Intermediate 7 is probably still accessible even in low dielectric environment. The protonated species of 3 and 7 (3H and 7H) are essentially isoenergetic. Therefore, both species should be accessible in a polar environment. In the peroxide intermediate 3 (Figure 1), there is an intramo- lecular hydrogen bond between the hydroxyl hydrogen and the oxyanion with a hydrogen bonding distance of 1.732 Å; the peroxide O-O distance is 1.450 Å and the C-O(-O) distance is 1.384 Å. Although the peroxide O-O distance in 7 remains about the same as in 3 (1.453 vs 1.450 Å), the C-O(-O) distance becomes shorter in 7 by 0.02 Å. In the transition state (9) for the interconversion of 4 and 5, the breaking O- - -O distance is 1.829 Å and the forming O- - -C distance is 2.100 Å. In 5H (the neutral form of 5), the two C-O(-H) distances are 1.399 and 1.384 Å for the cis and trans hydroxyl groups, respectively. Structural information regarding these reactive intermediates are difficult to obtain experimentally. However, X-ray crystal structures of compounds analogous to 3H and 5H are known. 5 Indeed, our calculated structures for 3H and 5H are very similar to the X-ray crystal structures for analogous compounds inves- tigated previously by Dowd and co-workers. 6 The calculated gas-phase protonation energies for anions 2-5 and 7 are given in Table 1. Interestingly, peroxide intermediate 7 is more basic than 3 owing to the intramolecular hydrogen bonding in 3. According to the present calculations, the gas- * To whom correspondence should be addressed. E-mail: tcbruice@ bioorganic.ucsb.edu. Fax: (805) 893-2229. (1) McTigue, J. J.; Suttie, J. W. J. Biol. Chem. 1983, 258, 12129. Suttie, J. W. FASEB J. 1993, 7, 445. (2) Li, S.; Furie, B. C.; Furie, B.; Walsh, C. T. Biochemistry 1997, 36, 6384. (3) Based on pKa of similar systems such as the R-C-H of mandelic acid and acetic acid: (a) Kenyon, G. L.; Gerlt, J. A.; Petsko, G. A.; Kozarich, J. W. Acc. Chem. Res. 1995, 28, 178. (b) Grabowski, J. J. Chem. Commun. 1997, 255. (4) Gerlt, J. A.; Gassman, P. G. J. Am. Chem. Soc. 1993, 115, 11552. Gerlt, J. A.; Gassman, P. G. Biochemistry 1993, 32, 11943. (5) Ham, S. W.; Dowd, P. J. Am. Chem. Soc. 1990, 112, 1660. (6) Dowd, P.; Ham, S. W.; Geibs, S. J. J. Am. Chem. Soc. 1991, 113, 7734. (7) Dowd, P.; Ham, S. W.; Hershline, R. J. Am. Chem. Soc. 1992, 114, 7613. (8) Dowd, P.; Hershline, R.; Ham, S. W.; Naganathan, S. Science 1995, 269, 1684. (9) Geometry optimization and transition state search were performed at the HF/6-31+G(d) level of theory. Additional energy calculations were performed using a hybrid density functional theory method (B3LYP/6-31+G- (d)). A simpler vitamin K model compound, 2,3-dimethyl-1,4-hydronaphtho- quinone, was used. Unless otherwise indicated, the energetics refer to the B3LYP/6-31+G(d) values. Gaussian 94, Revision B.2: Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery, J. A.; Ragha- vachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian, Inc., Pittsburgh, PA, 1995. Scheme 1 1623 J. Am. Chem. Soc. 1998, 120, 1623-1624 S0002-7863(97)03140-5 CCC: $15.00 © 1998 American Chemical Society Published on Web 02/04/1998