Biocatalytic strategies for the asymmetric synthesis of alpha-hydroxy ketones

The development of efficient syntheses for enantiomerically enriched alpha-hydroxy ketones is an important research focus in the pharmaceutical industry. For example, alpha-hydroxy ketones are found in antidepressants, in selective inhibitors of amyloid-beta protein production (used in the treatment of Alzheimer's), in farnesyl transferase inhibitors (Kurasoin A and B), and in antitumor antibiotics (Olivomycin A and Chromomycin A3). Moreover, alpha-hydroxy ketones are of particular value as fine chemicals because of their utility as building blocks for the production of larger molecules. They can also be used in preparing many other important structures, such as amino alcohols, diols, and so forth. Several purely chemical synthetic approaches have been proposed to afford these compounds, together with some organocatalytic strategies (thiazolium-based carboligations, proline alpha-hydroxylations, and so forth). However, many of these chemical approaches are not straightforward, lack selectivity, or are economically unattractive because of the large number of chemical steps required (usually combined with low enantioselectivities). In this Account, we describe three different biocatalytic approaches that have been developed to efficiently produce alpha-hydroxy ketones: (i) The use of thiamine diphosphate-dependent lyases (ThDP-lyases) to catalyze the umpolung carboligation of aldehydes. Enantiopure alpha-hydroxy ketones are formed from inexpensive aldehydes with this method. Some lyases with a broad substrate spectrum have been successfully characterized. Furthermore, the use of biphasic media with recombinant whole cells overexpressing lyases leads to productivities of approximately 80-100 g/L with high enantiomeric excesses (up to >99%). (ii) The use of hydrolases to produce alpha-hydroxy ketones by means of (in situ) dynamic kinetic resolutions (DKRs). Lipases are able to successfully resolve racemates, and many outstanding examples have been reported. However, this approach leads to a maximum theoretical yield of 50%. As a means of overcoming this problem, these traditional lipase-catalyzed kinetic resolutions are combined with racemization of remnant substrate, which can be done in situ or in separate compartments. Examples showing high conversions (>90%) and enantiomeric excesses (>99%) are described. (iii) Whole-cell redox processes, catalyzed by several microorganisms, either by means of free enzymes (applying a cofactor regeneration system) or by whole cells. Through the use of redox machineries, different strategies can lead to high yields and enantiomeric excesses. Some enantiopure alpha-hydroxy ketones can be formed by reductions of diketones and by selective oxidations of vicinal diols. Likewise, some redox processes involving sugar chemistry (involving alpha-hydroxy ketones) have been developed on the industrial scale. Finally, the redox whole-cell concept allows racemizations (and deracemizations) as well. These three strategies provide a useful and environmentally friendly synthetic toolbox. Likewise, the field represents an illustrative example of how biocatalysis can assist practical synthetic processes, and how problems derived from the integration of natural tools in synthetic pathways can be efficiently tackled to afford high yields and enantioselectivities.