An NADH-Coupled Biosensor for Engineering Redox Metabolism in Saccharomyces cerevisiae

Sammanfattning: Baker’s yeast, Saccharomyces cerevisiae’s potential in industrial biotechnology for producing valuable products e.g. biofuels, bulk chemicals, bio-flavours and pharmaceuticals is established. In order to achieve efficient production, the intracellular balance of the NADH/NAD + redox couple must be monitored and maintained, as this influences the feasibility of over 100 reactions in the cell. My approach based on the inherent ability of S. cerevisiae to control cytosolic NADH/NAD+ homeostasis by regulating the transcription of the glycerol phosphate-3 dehydrogenase 2 (GPD2) gene. It is responsible for glycerol synthesis in the cell, via the concurrent oxidation of NADH. It is known that its promoter is activated when there is a strong need for NADH oxidation. An induction different from that of the homologous gene GPD1, which is induced under osmotic stress. GFP levels were correlated to the need for GPD2 synthesis by cloning a green fluorescent protein (GFP) encoding gene in a plasmid downstream of GPD2p. It showed to be an efficient fluorescent biosensor for monitoring the redox state of the cell, enabling the distinction between strains with different abilities to reoxidize NADH. The sensor was applied to monitor the regulation of GPD2p under various growth conditions. It provided single-cell level information on the mode of metabolism enabling the identification of possible subpopulations. When no aeration was applied, GPD2p could be induced at a level twice that of the constitutively high TDH3 promoter. However, under conditions of high aeration and when the growth rate was maintained at 0.3 h -1 , the promoter was inactive. These findings, together with that a gpd1?gpd2? strain could be cultivated as efficiently as a wild-type strain, demonstrates the possibility of using the gpd1?gpd2? strain as a microbial production platform and the GPD2p as an inducible promoter for recombinant protein production. The technical and physiological boundaries of a gpd1?pgpd2? yeast for FACS-based discovery of xylose or acetophenone reductases were investigated. The reduction of xylose resulted in lower GFP fluorescence, but the opposite was observed when acetophenone was the substrate. Acetophenone was inhibitory to cell growth above 10 mM, and caused significant cell flocculation, challenging FACS-based screening. The results suggest that biosensor-based screening must be performed under well-controlled and substrate-specific conditions, as the substrate-specific responses may have strong and unforeseen effects on the output signal. The highly reducing potential of the gpd1?gpd2? background strain was separately evaluated for driving NADH-dependent whole-cell bioconversion with a transaminase and a ketone reductase for the production of the chiral pharmaceutical building blocks (R)-1-phenylethylamine and (S)-1-phenylethanol. The gpd1?gpd2? strain exhibited a 3-fold higher reduction rate and a 10-fold lower glucose requirement than the wild-type strain. The results provide detailed information on the need for GPD2 activity, and highlight the potential of the reducing environment of a gpd1?gpd2? strain to replace the production of glycerol with that of more economically interesting compounds.