Increasing n-butanol production with Saccharomyces cerevisiae by optimizing acetyl-CoA synthesis, NADH levels and trans-2-enoyl-CoA reductase expression

Virginia Schadeweg, Eckhard Boles

n-Butanol can serve as an excellent gasoline substitute. Naturally, it is produced by some Clostridia species which, however, exhibit only limited suitability for industrial n-butanol production. The yeast Saccharomyces cerevisiae would be an ideal host due to its high robustness in fermentation processes. Nevertheless, n-butanol yields and titers obtained so far with genetically engineered yeast strains are only low. In our recent work, we showed that n-butanol production via a clostridial acetoacetyl-CoA-derived pathway in engineered yeast was limited by the availability of coenzyme A (CoA) and cytosolic acetyl-CoA. Increasing their levels resulted in a strain producing up to 130 mg/L n-butanol under anaerobic conditions. Here, we show that under aerobic conditions. this strain can even produce up to 235 mg/L n-butanol probably due to a more efficient NADH re-oxidation. Nevertheless, expression of a bacterial water-forming NADH oxidase (nox) significantly reduced n-butanol production although it showed a positive effect on growth and glucose consumption. Screening for an improved version of an acetyl-CoA forming NAD(+)-dependent acetylating acetaldehyde dehydrogenase, adhE(A267T/E568K/R577S), and its integration into n-butanol-producing strain further improved n-butanol production. Moreover, deletion of the competing NADP(+)-dependent acetaldehyde dehydrogenase Ald6 had a superior effect on n-butanol formation. To increase the endogenous supply of CoA, amine oxidase Fms1 was overexpressed together with pantothenate kinase coaA from Escherichia coli, and could completely compensate the beneficial effect on n-butanol synthesis of addition of pantothenate to the medium. By overexpression of each of the enzymes of n-butanol pathway in the n-butanol-producing yeast strain, it turned out that trans-2-enoyl-CoA reductase (ter) was limiting n-butanol production. Additional overexpression of ter finally resulted in a yeast strain producing n-butanol up to a titer of 0.86 g/L and a yield of 0.071 g/g glucose. By further optimizing substrate supply and redox power in the form of coenzyme A, acetyl-CoA and NADH, n-butanol production with engineered yeast cells could be improved to levels never reached before with S. cerevisiae via an acetoacetyl-CoA-derived pathway in synthetic medium. Moreover, our results indicate that the NAD(+)/NADH redox balance and the trans-2-enoyl-CoA reductase reaction seem to be bottlenecks for n-butanol production with yeast.