In principle there are two ways for the biocatalytic synthesis of α-hydroxyketones. Oxidoreductases may be used to convert diketones or diols into the respective α-hydroxyketones. This kind of reaction is exploited, e.g., in the industrial synthesis of the α-glucosidase inhibitor miglitol ([2R,3R,4R,5S]-1-[2-hydroxyethyl]-2-[hydroxymethyl]-3,4,5-piperidintriole) (1 ). Alternatively, thiamine diphosphate dependent transketolases are capable of synthesizing α-hydroxyketones by catalyzing the transfer of activated glycolaldehyde (2 ). In natural biosyntheses these enzymes are usually found in sugar metabolism. Another thiamine diphosphate-dependent enzyme is used for the industrial production ofR-phenylacetylcarbinol ([1R]-1-hydroxy-1-phenyl-propan-2-one).R-phenylacetylcarbinol (R-PAC) is the first chiral intermediate in the production process of pseudoephedrine and ephedrine. Since 1921, it has been known that yeast (Saccharomyces cerevisiae) are able to catalyze the formation ofR-PAC (3 ,4 ). The further synthetic steps fromR-PAC to pseudoephedrine are carried out by classical chemical synthesis (5 ). The actual enzyme in yeast catalyzing the synthesis ofR-PAC is pyruvate decarboxylase (PDC; EC 4.1.1.1). In vivo this enzyme converts pyruvate to acetaldehyde. 2-α-Hydroxyethyl-thiamine diphosphate (“activated acetaldehyde”) is an intermediate of the catalytic cycle of PDC. Its α-carbanion reacts with several aldehydes in a nucleophilic attack to form the respective acyloins (6 ). In this manner, benzaldehyde and pyruvic acid formR-PAC and CO2
