Bacillus subtilis is an attractive host for the production of heterologous proteins. Fed-batch fermentation can attain a high cell density by avoiding substrate inhibition and accumulation of inhibitory metabolites such as acetate. In a fed-batch fermentation, the concentration of glucose or the specific growth rate is commonly used as an index in controlling the addition of the substrate (i.e., glucose) in order to maintain a stable, low concentration of the substrate during the entire process. A computer controlled system is required to monitor or control the limiting index when the specific fed-batch strategy is applied to the high cell density culture.
In this work, a computer controlled system using LabVIEW software was established. It achieved flexible and automatic monitoring and control of process parameters.
A dual exponential fed-batch culture strategy was applied to minimize the formation of acetate and to control the glucose concentration and specific growth rate at the predetermined values. Due to low solubilities of tyrosine and tryptophan in Feed Stream 1 containing concentrated glucose and other nutrients, tyrosine and tryptophan were removed from Feed Stream 1 and were dissolved in 14.4% ammonium water to form Feed Stream 2. By dual feeding both Stream 1 and Stream 2 at different exponential feed rates, the cells grew exponentially and a high cell density of 24.2 g/l and a final alpha-amylase activity of 71.4 U/ml were achieved. The overall biomass yield was 0.39 g cell/g glucose. In comparison, for the batch culture with the initial glucose concentration of 8 g/l, the final cell density was 2.3 g/l, and alpha-amylase concentration was 1.5 U/ml. The corresponding biomass yield was 0.28 g cell/g glucose.
A mathematic model was developed to investigate the inherent relationships between growth, substrate consumption, differentiation and product formation. The model includes three distinguishable cell states and the transition from the vegetative phase to sporangium and finally to mature spore. An age-based population balance model was applied to describe the maturity of sporangium toward the formation of spores. The model was able to predict the transient behavior of B. subtilis in both batch and fed-batch cultures satisfactorily.