Metabolic engineering of Saccharomyces cerevisiae for enhanced production of branched-chain higher alcohols

Abstract

Branched-chain higher alcohols (BCHAs) are promising advanced biofuels with properties that enable their use as complete gasoline substitutes in spark-ignition engines or as precursors to jet fuel. Of particular interest are isobutanol, isopentanol, and 2-methyl-1-butanol, which are naturally produced by the ethanol workhorse Saccharomyces cerevisiae as products of amino acid degradation. Using a combination of metabolic engineering, synthetic biology, systems biology, and protein engineering, this dissertation describes the engineering of S. cerevisiae for the production of isobutanol and isopentanol at high yields and titers. In order to optimize S. cerevisiae for isobutanol production, we elucidated the roles of the mitochondrial and cytosolic branched-chain amino acid transaminases involved in isobutanol biosynthesis from glucose. Our results demonstrate that deletion of the gene encoding the mitochondrial branched-chain amino acid transaminase increases isobutanol production by 14.2-fold in the absence of externally supplied valine. We next sought to enhance tolerance of yeast to isobutanol, as product toxicity remains a major barrier to achieving high titers. Screening the haploid S. cerevisiae gene deletion library revealed that deletion of GLN3, a gene involved in nitrogen utilization, increases yeast tolerance to isobutanol and boosts isobutanol production in engineered strains as much as 4.9-fold. Transcriptomic analyses suggest a mechanism by which isobutanol induces a nitrogen starvation response dependent on GLN3. In order to expand the diversity of BCHA products we can efficiently produce in S. cerevisiae, we harnessed mitochondria to enhance isopentanol production at the expense of isobutanol. Mitochondrial localization of the three enzymes that comprise the recursive 2-ketoacid carbon-elongation pathway enabled us to achieve the highest isopentanol titer and isopentanol to isobutanol product ratio as of yet reported in S. cerevisiae. Lastly, we develop genetically encoded biosensors to enable rapid phenotyping of S. cerevisiae strains for isobutanol or isopentanol production, and describe future work applying these biosensors to select high producers from diverse strain libraries. Overall, this work has led to the improvement of BCHA production in S. cerevisiae by developing novel tools, strategies, and strains to enable the economically viable production of isobutanol and isopentanol.