Abstracts Biological Sciences

Whole genome engineering, evolution, and high-throughput screening for propionic acid production and tolerance in yeast

by Xin Xu

Metabolic engineering of the cell factory Saccharomyces cerevisiae (yeast) has enabled bioproduction of a wide variety of chemicals, fuels, and pharmaceuticals. However, strain engineering is time-consuming, expensive, and limited by biological knowledge. With the rapid development of synthetic biology, novel approaches can be developed to revolutionise the problematic strain engineering processes. The global consortium 'Saccharomyces cerevisiae version 2.0 (Sc2.0)' is constructing the world's first synthetic eukaryotic genome with whole genome design and synthesis. One of the design features of Sc2.0 is the inclusion of a synthetic evolution system termed 'SCRaMbLE' (Synthetic Chromosome Rearrangement and Modification by LoxP-mediated Evolution). This can generate high rates of genome rearrangements, which has great potential to be applied in yeast metabolic engineering by generating diverse and novel genomes with superior phenotypes. Propionic acid (PA), a top platform chemical, has been widely used as a food preservative and a chemical intermediate in many industries. PA production via engineering yeast is potentially a sustainable alternative to current petroleum derived production. In this thesis, novel approaches have been established, combined with synthetic biology approaches and high throughput screening, for the generation and selection of genetic diversity to improve PA production and PA tolerance. Firstly, to build the ultimate Sc2.0 platform, I contributed to the construction and de-bugging of synthetic Chromosome XIV. The full synXIV was functional with comparable growth to the wild-type. Secondly, to facilitate the high-throughput screening of PA producers, I contributed to the development of an organic acid biosensor in yeast. The biosensor consists of the nativeWar1p transcriptional regulator and PDR12 promoter, which transduce the concentrations of para-hydroxybenzoic acid (PHBA) and PA to a GFP signal. The dynamic range has been further increased by the positive feed-back expression of WAR1, and the GFP signal has been normalised to the constitutive expression of mCherry within each cell to control for intrinsic noise. Thirdly, to improve PA production in yeast, SCRaMbLE of the PA-producing Wood-Werkman cycle was performed in a haploid strain and a synthetic-wild-type hybrid diploid yeast containing synthetic chromosomes III, VI, and IXR. SCRaMbLE of a diploid strain was found to be superior to SCRaMbLE of a haploid strain due to reduced SCRaMbLE-mediated viability loss and a subsequent higher frequency of cells with improved phenotypes. A novel strategy has been developed for SCRaMbLE and biosensor facilitated FACS screening. With this strategy, two strains with 1.7-fold to 2.4-fold improved PA titres were isolated and characterised using whole-genome sequencing. However, their parental diploid strain had all but one of the genes in the synthetic Wood-Werkman cycle deleted before the induction of SCRaMbLE. The two isolates also had lost the synthetic chromosomes. Improved production of PA…