The recombinant xylose-fermenting Saccharomyces cerevisiae strain harboring xylose reductase (XR) and xylitol dehydrogenase (XDH) from Scheffersomyces stipitis requires NADPH and NAD ؉ , creates cofactor imbalance, and causes xylitol accumulation during growth on D-xylose. To solve this problem, noxE, encoding a water-forming NADH oxidase from Lactococcus lactis driven by the PGK1 promoter, was introduced into the xylose-utilizing yeast strain KAM-3X. A cofactor microcycle was set up between the utilization of NAD ؉ by XDH and the formation of NAD ؉ by water-forming NADH oxidase. Overexpression of noxE significantly decreased xylitol formation and increased final ethanol production during xylose fermentation. Under xylose fermentation conditions with an initial D-xylose concentration of 50 g/liter, the xylitol yields for of KAM-3X(pPGK1-noxE) and control strain KAM-3X were 0.058 g/g xylose and 0.191 g/g, respectively, which showed a 69.63% decrease owing to noxE overexpression; the ethanol yields were 0.294 g/g for KAM-3X(pPGK1-noxE) and 0.211 g/g for the control strain KAM-3X, which indicated a 39.33% increase due to noxE overexpression. At the same time, the glycerol yield also was reduced by 53.85% on account of the decrease in the NADH pool caused by overexpression of noxE.

Glycerol is a major by-product in bioethanol fermentation by the yeast Saccharomyces cerevisiae, and decreasing glycerol formation for increased ethanol yield has been a major research effort in the bioethanol field. A new strategy has been used in the present study for reduced glycerol formation and improved ethanol fermentation performance by finely modulating the expression of GPD1 in the KAM15 strain (fps1⌬ pPGK1-GLT1 gpd2⌬). The GPD1 promoter was serially truncated from the 3= end by 20 bp to result in a different expression strength of GPD1. The two engineered promoters carrying 60-and 80-bp truncations exhibited reduced promoter strength but unaffected osmostress response. These two promoters were integrated into the KAM15 strain, generating strains LE34U and LE35U, respectively. The transcription levels of LE34U and LE35U were 37.77 to 45.12% and 21.34 to 24.15% of that of KAM15U, respectively, depending on osmotic stress imposed by various glucose concentrations. In very high gravity (VHG) fermentation, the levels of glycerol for LE34U and LE35U were reduced by 15.81% and 30.66%, respectively, compared to KAM15U. The yield and final concentration of ethanol for LE35U were 3.46% and 0.33% higher, respectively, than those of KAM15U. However, fermentation rate and ethanol productivity for LE35U were reduced. On the other hand, the ethanol yield and final concentration for LE34U were enhanced by 2.28% and 2.32%, respectively, compared to those of KAM15U. In addition, a 2.31% increase in ethanol productivity was observed for LE34U compared to KAM15U. These results verified the feasibility of our strategy for yeast strain development.