Rational Design And Application Of Mandelate Racemase And Esterase BioH

With remarkable increase of optically pure drugs in the market, production of chiral drugs or their intermediates has been gathering more and more attention. Clopidogrel is one of such drugs. As a platelet aggregation inhibitor, it is wildly administered to the atherosclerotic patients with risk of heart attack and stroke. Worldwide sales of clopidogrel ranked second in the global pharmaceutical market in2009. Methyl (R)-o-chloromandelate (R-CMM) is a key precursor for the synthesis of clopidogrel. To date, there are several routes to produce it. However, a number of drawbacks, such as addition of toxic HCN, expensive NADP+and low yield, limited their extensive application.In this study, we aimed to apply dynamic kinetic resolution to produce R-CMM, where mandelate racemase and esterase BioH were used to catalyze the process of racemization and hydrolysis, respectively. Compared to other methods, dynamic kinetic resolution is more eco-friendly, with lower cost and higher yield.Firstly, to fully understand the factors affecting the enantioselectivity, an quantitative enantioselectivity (E value) prediction model for the lipase CALB was constructed by QSAR analysis. It turned out that the enantioselectivity of CALB was influenced by various molecular interactions, including steric interaction, hydrophobic interaction, electrostatic interaction and hydrogen bond interaction. The model based on the above four molecular interaction fields exhibited high prediction accuracy (Rpred=0.92). Moreover, according to the contour maps, it was found that except four catalytic residues in the active cavity, other amino acids affecting the enantioselectivity of CALB were widely distributed at the acyl-binding pocket, the alcohol-binding pocket and the substrate access tunnel.Secondly, a tailor-made esterase BioH with high S-selectivity was prepared through the structure-based rational design. Guided by the structural information from the molecular dynamic (MD) simulation, the steric and electronic interactions between the enantiomers and the neighboring key amino acids were finely tuned. The enantioselectivity of BioH towards methyl (S)-o-chloromandelate was improved from3.3(wild type) to73.4(L123V/L181A/L207F), making it possible to produce R-CMM through dynamic kinetic resolution.Nextly, we focused on the redesign of mandelate racemase. The catalytic mechanism of racemase was further explored by MD simulation and MM-PBSA analysis. It was found that residue S139contributed to the stabilization of enzyme-substrate complex through hydrogen bond interaction. And there was a synergistic effect between S139and E317on the stabilization of substrate at ground/transition state. Replacement of S139by Ala not only destroyed the hydrogen bond between S139and the substrate, but also influenced the interaction between the residues in the active cavity and the substrate, leading to remarkable reduction of the catalytic efficiency (Kcat/Km) towards R-and S-mandelate (about45-fold and60-fold, respectively). The discovery of the key residue S139was beneficial in understanding the catalytic mechanism of racemase and would be helpful for further redesign of racemase.Then enzyme computational design based on the binding energy of enzyme-substrate complex at transition state was applied to improve the catalytic efficiency of mandelate racemase towards non-natural substrates. Using this strategy, the catalytic activities (kcat/Km) of racemase towards R-3-chloromandelic acid, R-2-naphthylglycolate, R-mandelamide and R-o-chloromandelic acid were improved to1.3×106,1.7×104,8.5×104and2.1×103M-1s-1, which were2-5fold higher than those of the wild type. The results indicated that computational design based on the binding energy was an efficient strategy in enzyme design.Finally, the designed mandelate racemase and esterase BioH were combined and used in the dynamic kinetic resolution for the production of methyl (R)-o-chloromandelate. Addition of racemase into the hydrolytic system not only accelerated the hydrolytic rate of methyl (S)-o-chloromandelate, but also enhanced the apparent E value of esterase BioH towards methyl (S)-o-chloromandelate. Under the optimal conditions of pH7.5,25℃and esterase/racemase molar ratio of1:21, using25mM methyl o-chloromandelate as the initial substrate (reaction volume of500ul), the final yield of product methyl (R)-o-chloromandelate was80%, with ee value of higher than97%after4cycles of dynamic kinetic resolution.