Combinatorial engineering of Saccharomyces cerevisiae for efficient pentose catabolism

The efficient fermentation of lignocellulosic biomass would enable more economically and environmentally friendly production of biofuels and biochemicals. Yet, Saccharomyces cerevisiae, a platform organism for biofuels and biochemicals production, is unable to convert all of the sugars in lignocellulosic biomass into biofuels and biochemicals mainly due to the lack of a pentose catabolic pathway. Though the advance of genetic engineering enabled S. cerevisiae to utilize pentose sugars, the efficiency of pentose sugar catabolism in S. cerevisiae is still limited. Here, the goal of this research was to confer efficient pentose sugar catabolism to S. cerevisiae by combinatorial and evolutionary engineering. To this end, pentose catabolic pathways were 1) constructed by heterologous expression of pentose catabolic genes, 2) optimized through rational engineering, and 3) further improved through evolutionary engineering. Through these efforts, we reported the highest ethanol yield (0.45 g ethanol / g xylose) and the second highest xylose consumption and ethanol production rates (0.98 g xylose g cell⁻¹ h⁻¹ and 0.44 g ethanol g cell⁻¹ h⁻¹, respectively) in xylose fermentation reported to date. The high performance in xylose fermentation was achieved based on the mutant xylose isomerase (xylA3), which showed 77% increased enzyme activity, engineered through directed evolution. In addition, we have established the first cells capable of growing on arabinose in mimimal medium and demonstrated ethanol production from xylan in minimal medium. The arabinose and xylan catabolic pathways were constructed in S. cerevisiae by expressing novel pentose catabolic genes from a strain with remarkable pentose catabolic potential that we isolated and named Ustilago bevomyces. In doing so, a complete workflow of bioprospecting to pathway engineering and evolution was detailed as an effective way to transfer a desired phenotype from a non-model organism to a model organism. This study substantially improved the prospect of biofuels and biochemicals production from lignocellulosic biomass by developing efficient pentose utilizing strains, finding new pentose catabolic genes, and suggesting alternative pentose catabolic pathway. Furthermore, the general tools for metabolic engineering demonstrated in this study would also advance microbial strain engineering.