Engineering the activities and specificities of a glycerol dehydratase and an alcohol dehydrogenase

There is growing interest in finding renewable, environmentally friendly routes to chemicals that are traditionally derived from crude oil. The use of bacteria as living factories for the biosynthesis of these chemicals is one such alternative. Of particular interest for this application are the acetogenic members of the genus Clostridium, due to their ability to utilise carbon monoxide and carbon dioxide gas, as their sole carbon and energy source. Not only are these cheap and plentiful feedstocks, but they also both contribute to the greenhouse gas effect, so the ability to capture atmospheric carbon and turn it into useful chemicals makes this strategy especially promising. In this thesis, I have used enzyme engineering in an attempt to expand the chemical-production repertoire of Clostridium autoethanogenum to include butanone and 2-butanol, two important industrial solvents. C. autoethanogenum natively produces 2,3-butanediol, a precursor to butanone, which, in turn, is a precursor to butanol. The glycerol dehydratase from Klebsiella pneumoniae (KpGDHt) can promiscuously catalyse the dehydration of 2,3-butanediol to butanone. However, it does so with low catalytic efficiency, and with strict stereoselectivity for the meso- isomer. While C. autoethanogenum produces a small amount of meso-2,3-butanediol, it predominantly produces (2R,3R)-butanediol. Therefore, in this study, enzyme engineering was used to try and create an enzyme that can turn over (2R,3R)-butanediol, and can catalyse the dehydration of both 2,3-butanediol isomers with high catalytic efficiency. KpGDHt is a dimer of an αβγ heterotrimer. Previous attempts to purify this enzyme had shown a propensity of the β subunit to disassociate during purification, yielding an inactive (αγ)2 complex. Modification of the enzyme to include a peptide linker joining the α and β subunits together (creating KpGDHt-L) prevented the loss of the β subunit during purification and increased the stability of the enzyme complex, with no detrimental effect on enzyme activity. Two strategies were used to modify the catalytic activity of KpGDHt-L: combinatorial active site saturation testing (CASTing), and consensus-guided mutagenesis. While neither strategy resulted in an enzyme that could turn over (2R,3R)-butanediol, and all variants built using CASTing showed significant losses in activity towards meso-2,3-butanediol, consensus mutagenesis yielded many variants with activity equal to, or greater than wild type towards meso-2,3-butanediol. Ultimately, the best variant enzyme contained a Thr200Ser mutation, which increased activity towards meso-2,3-butanediol by 3.6-fold. While further engineering will be required to bring activity with meso-2,3-butanediol to levels comparable to activity with native substrates, this study represents the first step on the path towards an efficient 2,3-butanediol dehydratase. The reduction of butanone to 2-butanol can be catalysed by the native primary-secondary alcohol dehydrogenase present in C. autoethanogenum (CaADH). This enzyme is strictly NADPH-dependent. As NADH is typically found at higher cellular concentrations, I used enzyme engineering to attempt to switch the cofactor specificity from NADPH to NADH. Site-saturation mutagenesis was used to mutate Ser199, the residue most proximal to the phosphate moiety differentiating NADPH and NADH. While no mutants were found that could utilise NADH, CaADH(S199A) showed a 3-fold increase in activity towards butanone, and CaADH(S199R) showed increased activity towards both acetone and acetoin. This work highlights the role that distal mutations can play in determining substrate specificity. This work represents the first attempt at mutagenesis of CaADH, and the most through mutagenesis study of KpGDHt. Insights from this work will help inform further mutagenesis of the two enzymes studied in this work, as well as contributing to our overall knowledge regarding enzyme engineering