JON D. STEWART
Department of Chemistry, University of Florida Gainesville, Florida 32611
I. II. III. IV.
Introduction Yeast Dehydrogenase Gene Identiﬁcation Expression and Isolation of Yeast Dehydrogenases Characterization of Yeast Dehydrogenases A. Results from Ethyl Acetoacetates B. Results from Higher Homologs C. Synthetic Applications V. Conclusions and Future Directions References
31 33 34 35 37 40 43 46 49
I. Introduction How well biocatalysis enables novel routes to target molecules ultimately rests on the properties of the available enzymes. In most applications, proteins are called on to catalyze reactions for which they did not speciﬁcally evolve. This means one must ﬁrst identify the best enzyme ‘‘tool’’ for a particular synthetic task, and this phase of bioprocess development has often consumed a signiﬁcant fraction of the total effort expended. Biocatalyst identiﬁcation was originally conducted by screening whole organisms (usually bacteria and fungi) to uncover those that catalyzed the desired transformation. This strategy has proven especially successful with steroid biotransformations in which site‐selective hydroxylation or carbonyl reductions are the typical goals (Mahato and Majumdar, 1993; Warhurst and Fewson, 1994). Many organizations have accumulated large microbial culture collections, and these have provided the ‘‘hits’’ that have been—and continue to be—developed into commercial bioprocesses. Despite the many success stories derived from whole organism screening, this approach suffers from several drawbacks. First, the strategy is purely empirical, and the screening process must be repeated for each new substrate/reaction pair. While accumulated experience can help narrow the search to those organisms known to mediate related reactions, the exploration remains labor‐intensive since cells of each organism must be grown anew for every screening. Another problem with using whole cells is that they often catalyze side‐ reactions in addition to the desired conversion. The enzyme of interest rarely makes up more than a few percent of the total protein in a native 31
ADVANCES IN APPLIED MICROBIOLOGY, VOLUME 59 Copyright 2006, Elsevier Inc. All rights reserved. 0065-2164/06 $35.00 DOI: 10.1016/S0065-2164(06)59002-1
JON D. STEWART
microbial cell, and its activity level is often insufﬁcient to outcompete other enzymes for access to the substrate. The reaction product may also be subject to degradation by the organism. The traditional solution to these problems of low activity and overmetabolism was to identify additional microorganisms that share the same (or closely related) enzyme of interest but with a different complement of competitors and/or expression level. The high degree of horizontal gene transfer among bacteria makes this approach particularly valuable in these organisms. Access to the enzyme of interest in pure form side‐steps the problems outlined earlier. Until recently, however, few synthetically useful enzymes were commercially available. Lipases and other hydrolases were the ﬁrst class for which a variety of enzymes were packaged into kits that could be screened by bench chemists with no particular biological expertise. It is no accident that hydrolase applications blossomed after this development. Other enzyme classes, such as dehydrogenases, have been similarly targeted for expansion (Zhu et al., 2005). Commercial enzymes are nearly always produced by recombinant systems, and these engineered cells can also be used directly as the biocatalytic reagent since the enzyme of interest often makes up !20% of total cellular protein. At such elevated speciﬁc activity levels, competition with other enzymes and product overmetabolism are often insigniﬁcant. Moreover, for those enzymes that require cofactors (e.g., dehydrogenases) intact cells allow regeneration by their native metabolic pathways. Cloned enzymes for biocatalysis have...