Characterisation and Engineering of New Biocatalysts

Abstract

Biocatalysis - using natures catalysts to catalyse chemical reactions - is an interesting and expanding field in protein biochemistry. Biocatalysis typically makes use of enzymes, the proteinogenic catalysts present throughout nature. Enzymes have numerous advantages over traditional small molecule catalysts in synthesis, namely their environmentally friendly conditions and waste, and their high selectivity, comprising of stereo-, regio-, and enantio-selectivity. However, enzymes have two major disadvantages: their instability under industrial synthesis conditions, especially elevated temperatures and organic solvents, and their slow speed, especially when utilising non-natural substrates. The limitations of enzymes can be overcome using enzyme engineering. Enzyme engineering approaches can range from rationally designed mutations, relying on a comprehensive understanding of the enzyme, to completely random, resulting in thousands of enzyme variants. Furthermore, computationally driven techniques can be phylogeny, physics, or machine learning based. These engineered variants, along with wildtype enzymes, can be used for the stereoselective synthesis of small molecule drugs, the degradation of waste materials, and the treatment of textiles in manufacturing. For these applications, both wildtype and engineered enzymes need to be characterised to understand their biochemistry, mechanism, and kinetics. In this thesis I characterised wildtype and engineered enzymes structurally using X-ray crystallography, biochemically using liquid chromatography-mass spectrometry based assays, and kinetically using colorimetric kinetic assays. Across each chapter, I integrated protein biochemistry knowledge with experimental approaches to characterise these enzymes. The engineered enzyme variants in this thesis were made using two approaches: a rational approach involving the swapping of natural metal ion cofactors for non-natural metal ion cofactors; and a phylogenetics based approach involving the statistical inference of ancestral enzyme sequences at nodes of a phylogenetic tree. Chapter 1 outlines the foundational concepts and ideas that underpin this thesis. Chapter 2 investigates the rational design of a model metalloenzyme by swapping the natural bimetallic centre with a non-natural monometallic centre. Chapter 3 investigates the dye degrading properties of a newly isolated bacterial strain, using a combined microbiology, genetics, and biochemistry approach. An enzyme capable of turning over malachite green was characterised. Chapter 4 investigates the ancestral protein reconstruction of a family of epoxyketone synthases to generate an enzyme variant library, which catalyse the synthesis of a,b-epoxyketones. These enzymes were screened against a range of substrates, including two therapeutically relevant substrates. Chapter 5 investigates the same enzyme variant library generated in Chapter 4, but investigates their ability to synthesise a precursor to another a,b-epoxyketone. Chapter 5 also investigates the suitability of using a photocleavable protecting group during a,b-epoxyketone synthesis, and attempted to generate X-ray crystallography standard protein crystals of active epoxyketone synthases. Overall, this work advances our understanding of biocatalysis and biodegradation, and also generated an enzyme capable of producing a therapeutically relevant a,b-epoxyketone. Show more