Exploiting the genomic diversity of bayer-villiger monooxygenases for new industrial oxidation reactions

Enzymes that catalyse the Baeyer-Villiger (BV) reaction, or the insertion of an oxygen atom adjacent to a carbonyl group in a carbon skeleton, have been of increasing interest over the past few years as they are able to catalyse the formation of esters or lactones from aliphatic or alicyclic ketones under mild reaction conditions and with selectivities that cannot be matched by abiotic catalysts. Baeyer-Villiger monooxygenases (BVMOs), as they are termed, use a reduced nicotinamide cofactor (NADPH) to reduce an enzyme-bound flavin (FAD) which reacts with molecular oxygen to form a flavin hydroperoxidate that acts as the oxidant in the enzymatic BV reaction. Whilst the early applications of such enzymes in organic synthesis has proved encouraging, the breadth of application has been limited by access to only comparatively few enzymes, capable therefore of only a limited substrate range and narrow selectivity. The field of genomics has revealed large amounts of open-reading-frames that might encode BVMOs, and the cloning and expression of some of these genes is revealing unprecedented useful activities. The genome sequence of the soil actinomycete Rhodococcus jostii RHA1, published in 2006, has revealed at least nineteen genes in that organism alone, consistent perhaps with the impressive catabolic abilities of that organism. We have recently cloned and expressed twenty-three genes from R. jostii, thirteen of which demonstrate the ability to catalyse the BV reaction when heterologously expressed in E. coli and used as whole cell catalysts. A brief screen of their catalytic properties has revealed unprecedented catalytic behaviour and also a breadth of characteristics which is illustrative perhaps of the catabolic capacity of the organism from which the genes were sourced. Early results suggest that, in most cases, the catalytic characteristics of the enzymes can be grouped along with their primary structure, but that this is not always the case, and that, in related groups, certain properties, such as substrate selectivity and regioselectivity, may be due to short amino acid sequence motifs. This family of diverse BVMOs provides a new reservoir of activities for application in the production of chiral synthetic intermediates. In collaboration with Dr Reddy's Laboratories, substrates of interest will be screened against this new library of catalysts in an effort to produce chiral lactones for further synthetic elaboration. In addition to exploiting new catalysts for asymmetric synthesis, the experiments will provide valuable new information on the molecular determinants of substrate recognition and selectivity in this class of enzymes. We intend to further explore these aspects with structural studies of the enzymes in complex with the industrial substrates, with a view to the optimisation of enzyme activity using structure-guided protein engineering techniques. In addition, successful reactions will be used as models for the optimisation of the biotransformation process itself, including studies on factors such as choice of expression strain, expression conditions, such as temperature and medium formulation, and reaction parameters such as oxygenation, and substrate/catalyst concentrations. Process considerations such as modes of substrate delivery using adsorbent resins and product recovery will also be addressed, so that new developments will have direct relevance to the synthetic reactions in an industrial context.