Harnessing aldo-keto reductases for biocatalysis

The aldo-keto reductases (AKRs) are a large protein superfamily of mainly NAD(P)-dependent oxidoreductases involved in carbonyl metabolism. A characteristic feature of many AKRs is their extremely broad substrate specificity when assayed in vitro. Considering the use of AKRs for the production of chiral alcohols, the broad range of substrates that these enzymes reduce is a clear advantage and offers scope for biocatalysis. But which AKR would be suited best for a particular transformation? Answers to this question, which would be important from basic and applied points of view, could come from an improved understanding of how substrate specificity is achieved in AKRs. This project proposes to carry out a structure-based systematic examination of Candida tenuis xylose reductase (CtXR) and Saccharomyces cerevisiae glycerol dehydrogenase (Gcy1), with respect to substrate specificity and stereochemical selectivity during ketone reduction. We will expand the spectrum of ketone substrates that are stereoselectively converted by CtXR, using a protein engineering approach based on detailed analysis of relationships between structure and substrate specificity of selected AKRs. While CtXR exhibits no C-C double bond reductase activity, other AKRs and short chain dehdydrogenases/reductases (SDRs) accept enoates as substrates. In this project, we will engineer C-C double bond activity into CtXR by mimicking the active site of AKR 1D in which active-site histidine is replaced by a glutamate and the particular activity is present. Likewise, an SDR-type catalytic triad will be realized in a Tyr51Phe mutant. AKRs, and reductases in general, can be used in either of two ways to catalyze the reduction of ketones: as whole cell systems, or as purified enzymes. In both cases, the coenzyme (NADH or NADPH) is present in substoichiometric amounts and has to be recycled during the course of the reaction. Baker`s yeast mediated reductions offer the advantage that the supply of NADH or NADPH is accomplished by the metabolism of a co- substrate. Unfortunately, bioreductions often go along with poor stereoselectivities, which can be traced to the presence of multiple NADPH-dependent reductases with divergent enantio- and diastereomeric preferences in the cell. We introduce an integrated approach of process development to overcome these limitations, which combines (i) structure based enzyme engineering, (ii) cell engineering and (iii) process engineering.