Submitted to: Biotechnology and Bioengineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 15, 2008
Publication Date: May 1, 2009
Citation: Bischoff, K.M., Liu, S., Leathers, T.D., Worthington, R.E., Rich, J.O. 2009. Modeling bacterial contamination of fuel ethanol fermentation. Biotechnology and Bioengineering. 103(1):117-122. Interpretive Summary: Bacterial contamination of commercial fermentation cultures is a common and costly problem to the fuel ethanol industry. Antibiotics may be used to control contamination but the emergence of resistant bacteria may limit their effectiveness. There is a need for tools to test the development of new antibacterial agents for application in the fuel ethanol industry. In the present study, a simple system using yeast growing on corn mash was developed to simulate the effects of bacterial infection. Specific strains of bacteria were used to experimentally infect the model, causing the fermentation to become "stuck" and not proceed to completion. Treatment with the antibiotic virginiamycin was effective in treating the infection, demonstrating the potential of this model for testing new antibacterial agents. This information will be of interest to ethanol producers and researchers attempting to design effective intervention strategies to control bacterial contamination in commercial fermentation cultures.
Technical Abstract: The emergence of antibiotic resistant bacteria may limit the effectiveness of antibiotics to treat bacterial contamination in fuel ethanol plants, and therefore, new antibacterial intervention methods and tools to test their application are needed. Using shake-flask cultures of Saccharomyces cerevisiae grown on saccharified corn mash and strains of lactic acid bacteria isolated from a dry-grind ethanol facility, a simple model to simulate bacterial contamination and stuck fermentation was developed. Challenging the model with 10**8 CFU/ml Lactobacillus fermentum decreased ethanol yield by 27% and increased residual glucose from 13.6 g/l to 61.7 g/l. The magnitude of the effect was proportional to the initial bacterial load, with 10**5 CFU/ml L. fermentum still producing a 7% decrease in ethanol and a 3.2-fold increase in residual glucose. Stuck fermentation was also dependent on the bacterial species used to challenge the fermentation, as neither L. delbrueckii ATCC 4797 nor L. amylovorus 0315-7B produced a significant decrease in ethanol when inoculated at a density of 10**8 CFU/ml. In the shake-flask model, treatment with 2 ug/ml virginiamycin mitigated the stuck fermentation when challenged with a susceptible strain of L. fermentum (MIC for virginiamycin <- 2 ppm), but treatment was ineffective at treating infection by a resistant strain of L. fermentum (MIC = 16 ppm). The model may find application in developing new antibacterial agents and