(636e) Optimal Induction of Saccharomyces Cerevisiae for Enhanced Metabolite Production in Batch Culture

Genetic engineering has become a standard tool for increasing the cellular production of targeted metabolites. A common problem is that genetic alterations designed to increase single cell metabolite production tend to decrease the cellular growth rate, negatively impacting the overall productivity of batch cell cultures. To overcome this limitation, chemically inducible promoters can be used to initiate the genetically engineered alteration during the batch. Because chemical inducers are prohibitively expensive for large-scale production and entail additional operating complexities, researchers are exploring inducible promoters based on bacterial quorum sensing to initiate gene expression. The quorum sensing components can be engineered to induce the expression of targeted genes once a certain cell density is achieved, providing a simple and cheap induction mechanism that can be transferred to other organisms.

In this contribution, we explore the potential utility of quorum sensing induced gene expression for increasing metabolite production in Saccharomyces cerevisiae. A dynamic flux balance model based on a genome-scale reconstruction of S. cerevisiae metabolism was used to compute optimal induction times for engineered mutants designed to increase ethanol or xylitol production in batch culture. Because the production of growth associated products such as ethanol is enhanced by a fully aerobic phase that maximizes biomass production followed by an anaerobic phase that maximizes ethanol synthesis, we considered simultaneous optimization of the gene induction time and the aerobic-anaerobic switching time. We found that gene induction can be beneficial for mutants in which there is a tradeoff between the biomass and/or ethanol yields under aerobic and/or anaerobic growth conditions compared to the wild type. For example, deletion of the aac1 gene substantially enhanced ethanol yield but decreased biomass yield under aerobic conditions. In the absence of gene induction, this mutant had a substantially lower ethanol productivity than the wild type. By simultaneously optimizing the gene induction time and the aerobic-anaerobic switching time, the aac1 deletion mutant had a significantly enhanced ethanol productivity compared to wild type and even other engineered mutants that appeared to be far superior in the absence of induction. These computational results provide motivation for continued development of quorum sensing based gene induction.