(499a) Screening of a Combinatorial Pathway Library to Improve Isobutanol Production in Saccharomyces Cerevisiae

The yeast Saccharomyces cerevisiae is a well-studied and tractable microbe with considerable potential for use in production of next-generation biofuels such as isobutanol. S. cerevisiae is also tolerant to a variety of stressors present in lignocellulosic hydrolysate, a feedstock that does not compete with food resources. To maximize isobutanol production from hydrolysate with S. cerevisiae, we built a platform strain of S. cerevisiae that can ferment both glucose and xylose, in which we additionally deleted several genes to prevent formation of unwanted byproducts and increase flux towards isobutanol. In parallel, we sought to optimize the isobutanol biosynthesis pathway by screening a combinatorial library of diverse homologs of the five isobutanol pathway genes (>15 homologs per gene), each driven by either a strong, medium, or weak promoter. This variation of homologs as well as promoter strength, while increasing the library complexity to a final size of ~10⁹, allows for the simultaneous identification of active homologs and optimal expression levels. Preliminary screening of this library relied on the ability of an active isobutanol pathway to restore growth in a fermentation-deficient strain and has already provided insights into pathway bottlenecks and enabled production of isobutanol at yields close to the highest reported in academic literature. Additionally, the presence of several false-positives in our growth selection revealed promiscuous homologs of 2-ketoisovalerate decarboxylase that act on pyruvate and restore ethanol fermentation, deepening our understanding of substrate specificity of these enzymes. Further screening of the library will give us a more complete picture of the combinatorial space, and this data can be used to build machine learning models to elucidate trends in the most active isobutanol cassettes and predict designs that maximize isobutanol production.