The Utilization of Alcohol Dehydrogenase (ADH) Enzymes in Stereocontrolled Synthesis

Clostridium acetobutylicum alcohol dehydrogenase (CaADH) has been shown to catalyze the stereoselective reduction of a broad array of carbonyl compounds. This study uses a Hammett linear free-energy relationship (LFER) analysis to gain insight into CaADH enzyme mechanism across three substrate classes: aryl aldehydes, aryl β-keto esters, and aryl trifluoromethyl ketones. Kinetic data were collected by monitoring NADPH fluorescence, and Michaelis-Menten kinetics were applied. Hammett plots reveal that hydride transfer is rate-limiting for aldehydes and β-keto esters (ρ > 0), while it is proposed that dehydration is rate-limiting for trifluoromethyl ketones (ρ < 0). 13C NMR confirmed significant hydration in trifluoromethyl ketones. The study also reports high enantioselectivity (99% ee) in the reduction of α,α,α-trifluoroacetophenone. The principal study in this thesis describes the use of the CaADH enzyme to synthesize 20 aryl isoserine side chains for Taxotère-like chemotherapeutics via dynamic reductive kinetic resolution (DYRKR) of α-chloro-β-keto esters, giving in most cases, largely the D-syn stereoisomer of four possible products from the corresponding (±)-α- chloro-β-keto ester. Among the side chains obtained are the thienyl and furyl systems for the known Taxotère analogues Milataxel and Simotaxel. Cross-coupling on the p-bromophenyl isoserine side chain further expanded the taxoid library giving rise to an additional 16 new Taxotère side chain structures. These studies also include the first X-ray crystal structure of CaADH (1.59 Å), solved in the laboratory of collaborator Professor Mark Wilson. The CaADH structure features the NADP+ cofactor bound in the catalytically non-productive anti-conformation, a feature that also has the advantage of revealing the substrate carbonyl binding pocket, as this is where the nicotinamide carbonyl is bound. The structure also reveals a flexible loop near the active site, that is illuminated through MDS studies, providing possible insight into the remarkable substrate promiscuity seen. As Frank Westheimer noted in his classic piece, “Why Nature Chose Phosphates,” phosphate esters are ubiquitous in Nature, offering binding handles, facilitating compartmentalization and providing for biological signaling. α-Monofluorophosphonates have long been studied in the Berkowitz group as tools for chemical biology, as they are excellent ‘isoacidic’ mimics of native phosphate esters, as they provide on constitutive ‘phosphorylation-ON’ phenotype, being inert to phosphatase enzymes. They also contain an additional stereocenter that can be ‘tuned’ for optimal target-binding, raising interest in their stereoselective synthesis. This work reveals a promising new method to achieve this; namely the use of a yeast alcohol dehydrogenase-mediated DYRKR entry into D-anti-α-monofluoro-β-hydroxy phosphonates. Success with a variety of targets is complemented by structural biological studies here by the Wilson lab, providing the first ‘pseudo-ternary’ structure of this yeast alcohol dehydrogenase enzyme, providing a framework from which to rationalize and predict substrate tolerance and selectivity. Billions of years of evolution have produced (hyper)thermophiles that thrive at high temperatures. Their enzymes maintain stability and functionality even near boiling water temperature, making them valuable for biocatalysis and hybrid bio/chemocatalytic synthesis. However, these enzymes are underutilized in biocatalysis. Operating at higher temperatures offers benefits such as improved reaction rates, better substrate solubility, and easier separation of products. This concept paper highlights the potential of archaeal enzymes in non-natural synthesis and notes that many remain unexplored. These ancient enzymes could drive future innovations in biocatalytic chemistry. Advisor: David B. Berkowitz