Structure-based Determination Of Catalytic Mechanism And Performance Regulation For (R)-specific Alcohol Dehydrogenase From Candida Parapsilosis

Alcohol dehydrogenase (ADH) catalyzes the reversible reactions from alcohols toaldehydes or ketones. ADH as a bicatalyst with unique chemo-, regio-, and stereoselectivitiesis used for biosynthesis of versatile chiral alcohols, which are important chiral building blocksof pharmaceuticals and fine chemicals. However, there are several problems in the researchand application of ADHs, including lack of structural resources, unclear chiral recognition andcatalytic mechanism, limited selectivity and applicability, and bottleneck of cofactorregeneration.(R)-specific alcohol dehydrogenase (RCR) from Candida parapsilosis CCTCCM203011was used as a model enzyme of our research. This study had improved theheterogenous expression level after mRNA optimization, determined the complex structuresof RCR with its ligands and the molecular catalytic mechanism by X-ray crystallography,engineered the stereoselectivity and substrate specificity by rational design, and constructedcoenzyme regeneration system employing isopropanol or glycerol.(1) Optimization of mRNA secondary structure improved heterogenous expression levelof RCR in Escherichia coli. We optimized the mRNA secondary structure in translationinitiation region (from+1to+78), and constructed the corresponding variant with13synonymous mutations. The formation of hairpin structure was significantly reduced and thetheoretical Gibbs free energy was dramatically decreased from-9.5kcal·mol-1to-5.0kcal·mol-1. As a result, the expression level of RCR in the variant was increased by4-5timesand its specific activity in cell-free extract was enhanced by61.9%compared to the WT strain.When using the whole cells as catalyst and2-hydroxyacetophenone (HAP) as substrate, thevariant showed excellent performance to give (R)-1-phenyl-1,2-ethanediol ((R)-PED) withoptical purity of93.1%e.e. and a yield of81.8%, increased by27.5%and40.5%respectivelythan those of WT.(2) Crystal structures and molecular catalytic mechanism of RCR were determined byX-ray crystallography. Firstly, the protein samples were prepared by chromatographypurification. Secondly, through screening and optimization of crystal growth conditions usingthe sitting-drop vapor-diffusion method, large single holo RCR crystals (0.3×0.1×0.1mm)and the variant holo H49A crystals (0.5×0.1mm or0.5×0.3mm) were obtained in thepresence of NAD+after1-2days. The complex structures were obtained by soaking thecrystals with the ligands. Finally, after X-ray diffraction data collection and processing, thestructures of RCR-NAD+, RCR-(R)-PED and RCR_H49A-HAP had been solved bymolecular replacement using a high homologous ADH from Rhodococcus ruber (31.5%identity, PDB code3JV7). The overall structure of RCR revealed as a homo-tetramer. Eachsubunit was composed of a cofactor-binding domain and a catalytic domain, containing astructural and a catalytic zinc ions. Between these two domains lay a NAD+molecule. TheNAD+binding induced a significant conformational change from an open form to a closedform.(R)-PED/HAP was bound in the hydrophobic cavity, and the orientation ensured thatNADH delivered its pro-R hydride to the re-face of HAP to produce (R)-PED, following thePrelog’s rule. (3) Structure-based rational design changed the stereoselectivity of RCR. Combiningamino acid sequences and crystal structures of the relevant medium-chain ADHs, the aminoacid residues of RCR involving to form the large and small pocket and their conservatismwere identified. We considered that F285and W286might play critical roles instereoselectivity and constructed three non-conservative variants F285A, W286A andF285A/W286A. F285A/W286A revealed that the stereoselectivity was changed towards arylketones (1a-8a). Docking models further confirmed that substrate orientation was changedfrom re-face to si-face in the active site of F285A/W286A, leading to reversedstereoselectivity. The cavity-forming amino acids, especially the unconserved residues as‘‘hotspots’’, may contribute to the difference in pocket shape and substrate orientation, andthus direct the diversity of reaction stereoselectivity and stereopreference of medium-chainADHs.(4) Structure-based rational design changed the substrate specificity of RCR. Substratespecificity of RCR-WT was identified by determining the kinetic parameters towards variouscarbonyl compounds, including aryl ketones, β-ketoesters and aliphatic ketones. WTperformed good catalytic efficiency towards the small substrates, while low catalyticefficiency towards bulky substrates. The large and the small substrate-binding pockets weredesigned separately. Compared with WT, the activities of the variant W116A located the largepocket towards the bulky substrates were improved without affecting the stereoselectivity,such as1.8-13.8fold towards2a-7a,>20-fold towards16a,18a and19a. Double-mutantF285A/W286A in the small pocket showed better catalytic properties towards most substrateswith respect to WT, except the small substrates. Docking models further confirmed that theenlarged cavity reduced the steric effect of bulky substrates. WT enzyme preferred the smallsubstrates, while variants preferred the bulky substrates. The complementary substratespectrum of WT and variants expanded the application scope of ADHs.(5) In vivo dual-cosubstrate-coupled system improved the utilization efficiency ofcoenzyme. Employing10%(v·v-1) isopropanol and8%(v·v-1) glycerol as hydrogen donors,(R)-PED was obtained with optical purity of>99.9%e.e. and conversion of85.5%inRCR-mediated the whole-cell reaction. The conversion was increased by13%and21%compared with isopropanol or glycerol addition, respectively. The productivity represented a250-fold increase from0.02g·L-1·h-1to5g·L-1·h-1, and the substrate concentration wasincreased from1g·L-1to10g·L-1. Activity assay showed that isopropanol was oxidized byRCR for NADH regeneration, while glycerol was metabolized by intracellular enzymes forNADH regeneration. Furthermore, isopropanol and glycerol were cosolvents of substrates,and glycerol was protected solvent of cells. In vivo NADH recycling system employingisopropanol and glycerol would make use of the function of engineering enzymes and otherintracellular enzymes inside the cell to drive double coenzyme regeneration way forimproving the utilization efficiency of coenzyme.