(180ah) The Use of a Parallel Fermentation System for Optimization of Saccharomyces Cerevisiae Ethanol Production in Batch Culture

Batch cell culture remains an important technology for the large-scale production of many biochemical products, including commodity biochemicals such as ethanol. We have used dynamic extensions of metabolic flux balance models to show that the optimal operating policy for Saccharomyces cerevisiae ethanol production in batch culture consists of a fully aerobic phase that maximizes biomass production followed by an anaerobic phase that maximizes ethanol synthesis. Our computational results suggest that ethanol productivity is sensitive to the switching time between the aerobic and anaerobic growth phases. Validation of these predictions requires time consuming fermentation experiments in which the ethanol productivity is determined for different aerobic-anaerobic switching times to establish the optimal point as well as the curvature near the optima.

In this contribution, we demonstrate the use of a 4-bioreactor parallel system to establish and validate optimal fermentation operating policies using S. cerevisiae ethanol production from glucose media as a model system. Each bioreactor has a 250 mL working volume and independent gas pumps, dissolved oxygen, temperature and pH probes, and stirring system for the execution of parallel batch experiments in precisely controlled growth environments. The four bioreactors are shown to yield reproducible biomass, glucose and ethanol concentration measurements when inoculated from the same shake flask and grown under the same aerobic conditions. The parallel fermentation data are further validated with experiments performed on a standard laboratory bioreactor with 1.25 L working volume and by comparison to predictions from a dynamic flux balance model. Having established that the parallel system produces reliable data, our computational results for the optimal aerobic-anaerobic switching time are evaluated by performing parallel experiments in which each bioreactor is subjected to a different switching time. With the continuing miniaturization and parallelization of bioreactor technology, our results indicate that parallel systems hold great promise for fermentation process optimization.