Development of xylose reductase isozyme system for enhancing xylose metabolism in Saccharomyces cerevisiae

Abstract

Biological production of fuels and chemicals from lignocellulosic biomass is a sustainable and ecofriendly process. Glucose and xylose are the major constituents of lignocellulosic hydrolysates. Although the engineered Saccharomyces cerevisiae can use xylose, its fermentative capacity on xylose has been known to be much lower than that on glucose. Therefore, the efficient consumption of xylose is one of the key steps for economically feasible production of biofuels and chemicals. The overall goal of this thesis is to develop metabolically engineered S. cerevisiae able to ferment xylose efficiently. In this thesis, S. cerevisiae D452-2 was used as a host strain for production of xylitol and bioethanol, and has been chosen in metabolic and evolutionary engineering studies of other chemicals production from xylose. Many S. cerevisiae strains were sequenced, providing additional information of unexplored differences between S. cerevisiae strains. On the other hand, the D452-2 strain was not sequenced yet and its underlying beneficial genetic polymorphisms remain unknown. The whole genome sequencing of the D452-2 strain was performed by PacBio sequencing. Its assembled genome sequence was nearly identical to that of S288c which is a reference strain and was sequenced completely. As a result of comparative analysis, the genome of D452-2 was rearranged by transposon elements and small indels, and had nineteen ORFs that were absent in S288c. Compared to the S288c genome, 6,948 of SNPs were detected to change properties of 313 metabolic enzymes. These genetic variations of the D452-2 strain provide the basis for a forward genetic approach for developing xylose-fermenting yeast strains with enhanced performance. The XYL1, XYL2 and XYL3 genes involved in the xylose assimilation pathway from Scheffersomyces stipitis and a native xylose-fermenting yeast had been introduced into S. cerevisiae D452-2 to assimilate xylose. However, the resulting S. cerevisiae strains often exhibited undesirable phenotypes. The cofactor imbalance generated from different cofactor requirement between NADPH-dependent xylose reductase (XR) and NAD+-dependent xylitol dehydrogenase (XDH) inceased xylitol accumultation. To alleviate cofactor imbalance, a XR mutant with NADH preferency was constructed by protein engineering. The adoption of this XR mutant reduced xylitol accumulation but led to slow xylose consumption rate. The availability and balance of cofactors might be limiting factors for xylose fermentation by engineered S. cerevisiae. In this thesis, a synthetic isozyme system of XRs was designed to overcome the above mentioned problems. To construct this system, NADH-dependent mutant XR was expressed in the S. cerevisiae D452-2 strain expressing NADPH-dependent wild-type XR, XDH and xylulose kinase (XK). While the strains having only one type of XR exhibited XR acvitities which were highly specific for NADPH or NADH, the S. cerevisiae strain having the XR-based isozyme system showed similar XR activities toward NADPH and NADH. The engineered strain exhibited low xylitol accumulation and fast xylose consumption compared to the control strains expressing one type of XR. The xylose fermenting performance was confirmed by fermentations in various conditions. In a batch fermentation using silver grass hydrolysates, the engineered strain produced 50.7 g/L ethanol with 0.43 g/g ethanol yield. Besides ethanol production, the XR-based synthetic isozyme system was applied to improving the production of value-added products such as xylitol. The engineered S. cerevisiae strain having both XRs exhibited higher xylitol productivity than the control strains in both batch and glucose-limited fed-batch fermentations. To supply NADPH and NADH sufficiently, cofactor regeneration enzymes were co-expressed additionally. The coexpression of glucose-6-phosphate dehydrogenase encoded by the ZWF1 gene and acetyl-CoA synthetase by the ACS1 gene increased intracellular concentrations of NADPH and NADH, and improved xylitol productivity. In order to extend the period for the highest xylitol productivity and hence elevate final xylitol concentration, fed-batch fermentation strategies were optimized. Finally, the optimized fed-batch fermentation of the engineered strain resulted in 196.2 g/L xylitol concentration, 4.27 g/L-h productivity and almost the theoretical yield. The synthetic isozym system of XR is a promising strategy to meet the industrial demands for production of ethanol and xylitol.