Engineered Alcohol Dehydrogenases for Stereoselective Chemical Transformations

Sammanfattning: Enzymes are biomolecules built from amino acids and catalyze the chemical transformations in a cell. Enzymes are by nature stereoselective, biodegradable, environmentally friendly, and can perform catalysis in aqueous solutions and at ambient temperatures. Due to these advantages the use of enzymes as biocatalysts for chemical transformations has emerged as an attractive “greener” alternative to conventional chemical synthesis strategies. And, if naturally occurring enzymes cannot carry out the desired chemical transformations, the functional properties of enzymes can be modified by directed evolution or protein engineering techniques. Since enzymes are genetically encoded they can be optimized for desired traits such as substrate selectivity or improved catalytic efficiency. Considering these advantages and also keeping the synthetic and industrial application in mind, we have employed alcohol dehydrogenase-A (ADH-A) from Rhodococcus ruber DSM 44541 as a study object in engineering for new catalytic properties. ADH-A tolerates water miscible organic solvents, accepts a relatively wide range of aromatic sec-alcohols/ketones as substrates and is therefore a potentially useful biocatalyst for asymmetric synthesis of organic compounds. Presented research work in this thesis has been primarily focused on engineering of ADH-A and characterization of resulting enzyme variants. The engineering efforts have aimed for altered substrate scope, as well as stereo- and regioselectivities. Furthermore, possible substrate promiscuity in engineered enzyme variants has also been addressed. In short, i). Paper I: three sub sites, each consisting of two-three amino acid residues within the active-site cavity were exposed to saturation mutagenesis in step-wise manner, coupled to an in vitro selection for improved catalytic activity with the unfavored (R)-1-phenylethanol. The observed stereoselectivity could be explained partly by a shift in nonproductive substrate binding. ii). Paper II is aimed specifically towards the improving the catalytic activity with aryl-substituted vicinal diols, such as (R)-1-phenylethane-1,2-diol, and the possibility to link the ADH-A reaction with a preceding epoxide hydrolysis to produce the acyloin 2-hydroxyacetophenone from rac-styrene oxide. iii). Paper III is mainly focused towards studies of regioselectivity. Here, ADH-A and engineered variants were challenged with a substrate containing two sec-alcohol functions and the cognate di-ketone. The regioselectivity in wild type as well as in engineered variants could in part be explained by a combination of experimental and computer simulations. iv). Paper IV is focused on elucidating possible effects on substrate promiscuities in engineered variants as compared to the wild type parent enzyme, when challenged with a spectrum of potential previously untested substrates.