Threonine aldolases as tools for stereoselective synthesis

Enzymes are biological catalysts capable of an enormous array of reactions. They demonstrate the ability to be reshaped and repurposed as synthetic tools by the process of enzyme engineering. Industrially, enzymes have been exploited for their ability to synthesise high value chiral molecules with excellent stereoselectivity. They are capable of functioning at low temperatures, near neutral pH and are completely biodegradable. Methods to harness the synthetic potential of enzymes have been extensively developed and applied over recent years. Directed evolution is one such approach that mimics natures evolution strategy but over a significantly shorter time frame. It involves iterative cycles of mutagenesis and screening until a required function is achieved. Advances in structural biology have enabled a huge number of enzyme structures to become available. These can help to implicate important catalytic residues in an enzyme, in turn, focussing engineering efforts to likely mutagenic hotspots. This can lead to greater chances of obtaining a desirable enzyme function if paired with an effective mutagenesis and screening strategy.
In this thesis we use combinatorial active site saturation testing, a focused directed evolution method, to engineer a low-specificity L-threonine aldolase from Escherichia coli for improved stereoselectivity. We target the reversible retro-aldol cleavage reaction of phenylserine, an industrially interesting compound with four possible stereoisomers. Stereoselective variants are identified following the development of an effective high-throughput screen. Further, we use this screen to identify and produce rational combinations of variants. This achieves an enzyme that is 500-fold stereoselective for a single phenylserine stereoisomer. We also aim to provide insight into the poor stereoselectivity of the wild-type enzyme using molecular mechanics and quantum mechanics/molecular mechanics simulations. These are performed on an obtained 1.6 Å crystal structure of Escherichia coli threonine aldolase. We further propose plausible mechanisms for the stereogenic step of the aldol-condensation reaction.