pubs.acs.org/biochemistry Published on Web 01/17/2011 r 2011 American Chemical Society Biochemistry 2011, 50, 1505–1513 1505 DOI: 10.1021/bi101493p Mechanism of Mycolic Acid Cyclopropane Synthase: A Theoretical Study † Rong-Zhen Liao, ‡,§ Polina Georgieva, ‡ Jian-Guo Yu, § and Fahmi Himo* ,‡ ‡ Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, SE-10691 Stockholm, Sweden, and § College of Chemistry, Beijing Normal University, Beijing 100875, People’s Republic of China Received September 15, 2010; Revised Manuscript Received December 15, 2010 ABSTRACT: The reaction mechanism of mycolic acid cyclopropane synthase is investigated using hybrid density functional theory. The direct methylation mechanism is examined with a large model of the active site constructed on the basis of the crystal structure of the native enzyme. The important active site residue Glu140 is modeled in both ionized and neutral forms. We demonstrate that the reaction starts via the transfer of a methyl to the substrate double bond, followed by the transfer of a proton from the methyl cation to the bicarbonate present in the active site. The first step is calculated to be rate-limiting, in agreement with experimental kinetic results. The protonation state of Glu140 has a rather weak influence on the reaction energetics. In addition to the natural reaction, a possible side reaction, namely a carbocation rearrangement, is also considered and is shown to have a low barrier. Finally, the energetics for the sulfur ylide proposal, which has already been ruled out, is also estimated, showing a large energetic penalty for ylide formation. Mycobacterium tuberculosis mycolic acid cyclopropane synthases (MACSs) catalyze the cyclopropanation of the double bond of unsaturated mycolic acids at proximal and distal posi- tions, using S-adenosyl-L-methionine (SAM or AdoMet) as the methyl donor (1). The mycolic acid products are lipid components of the cell envelope of mycobacteria (2). This chemical modifica- tion is significant for the pathogenicity, persistence, and virulence of M. tuberculosis, making these enzymes therapeutic targets for antibacterial drugs, like isoniazid and ethionamide (3, 4). Four members of the MACS family have been identified, namely, CmaA1, CmaA2, MmaA2, and PcaA (1b-1e). These enzymes were found to have different selectivities. CmaA1 was found to be responsible for cis cyclopropanation at the distal position of R-mycolate (1b), while CmaA2 was shown to be involved in trans cyclopropanation at the proximal position of oxygenated mycolate (1d). MmaA2 is required for cis cyclopropa- nation at the distal position of R-mycolate or at the proximal position of oxygenated mycolate (1e). PcaA catalyzes cis cyclopro- panation at the proximal position of R-mycolate (1c). The crystal structures of CmaA1, CmaA2, and PcaA have been determined, and they revealed >50% identity in their primary sequences, indicating a conserved reaction mechanism (5). In the structure of CmaA1 in complex with the product S-adenosyl-L-homocysteine (SAH) and didecyldimethylammonium bromide (DDDMAB) (Figure 1), the inhibitor DDDMA cation adopts a U shape and is inserted into the binding pocket. Several aromatic residues, such as Tyr16, Tyr33, and Tyr232, are thought to provide stabilization to the positive charge of the ammonium through cation-π interactions (5). A bicarbonate ion is located close to the ammonium cation and hydrogen-bonded to Glu140, His167, and Tyr232 and the Ser34-Cys35 peptide (5). There is another type of bacterial cyclopropane synthase, namely Escherichia coli cyclopropane fatty acid synthase (CFAS), which uses unsaturated phospholipids as substrates (6). CFAS shares up to 33% sequence identity with MACS and also harbors a bicarbonate ion in the active site, suggesting a similar catalytic mechanism (7). A direct methyl transfer reaction mechanism for the enzymatic cyclopropanation has been proposed (Scheme 1) (8) on the basis of the crystal structure (5) and also on the basis of fluorine substitution (8g), onium chalcogen effect (8i), and deuterium isotope effect (8i) studies. In the first step, the transfer of a methyl from AdoMet to the substrate double bond occurs to form a carbocation intermediate. Then the bicarbonate ion acts as a general base to take a proton from the methyl group, resulting in ring closure. Mutations of three active site residues of E. coli CFAS (E239, H266, and Y317, corresponding to E140, H167, and Y232 in CmaA1, respectively), which form hydrogen bonds to the bicarbonate, result in a dramatic decrease in activity (7). This observation might confirm the role of the bicarbonate in the reaction. The first step is suggested to be rate-limiting on the basis of the onium chalcogen effect study, in which the AdoMet was substituted with SeAdoMet and TeAdoMet and different turn- over numbers were obtained for the three different methyl donors (8i). In addition, substitution of the methyl donor with deuterium shows an inverse isotope effect, further confirming the suggestion that the methyl transfer step is rate-limiting (8i). Previously, a metal-assisted sulfur ylide mechanism was sus- pected, in which the reaction starts by deprotonation of the methyl substituent of AdoMet (8a, 8b). However, this has subsequently been ruled out (8e). In this study, we use density functional theory (DFT) calcula- tions to shed more light on the reaction mechanism of MACS. A large model of the active site is designed on the basis of the crystal structure of CmaA1. The protein surrounding is modeled by homogeneous continuum model. Recent studies of three different kinds of enzymes have shown that at a model size of ∼150-200 atoms, the solvation effects almost vanish and the choice of † This work was supported by The Swedish Research Council (Grants 621-2009-4736 and 622-2009-371 to F.H.) and the National Natural Science Foundation of China (Grants 20733002 and 20873008 to J.-G.Y.). *To whom correspondence should be addressed. Telephone: þ46-8- 161094. Fax: þ46-8-154908. E-mail: himo@organ.su.se.