Mathematical Modelling of the pH-induced Metabolic Shift in C. aceobutylicum Unravels a Heterogeneous Phase Transition

The acetone-butanol-ethanol (ABE) fermentation of Clostridium acetobutylicum attracts new attention because it provides a potential alternative for the synthesis of value added chemicals to petroleum and other fossil reserves. This fermentative metabolic process comprises two distinct metabolic states that differ in their product spectrum. Growing on starch or sugars the predominant fermentation products are acetate and butyrate during acidogenesis (high pH). In contrast, C. acetobutylicum produces the solvents acetone and butanol during solventogenesis (low pH). In a continuous culture under phosphate limitation, the shift between both metabolic states can be induced by changes of the external pH level. Existing models were unable to reproduce the dynamics of this phase transition. Here, we present a kinetic model of ABE fermentation in continuous culture that incorporates metabolic reactions, changes of the proteomic composition, pH-dependent kinetics, and population growth. Our analyses of three independent shift experiments led us to the hypothesis that the pH-induced metabolic shift is a heterogeneous process governed by two sub-populations. The extension of our kinetic model with respect to this hypothesis better reproduces the experimental data and led us to the conclusion that an acidogenic and a solventogenic phenotype coexist during the transition from acidogenesis and solventogenesis and vice versa. Furthermore, we show that the measured evolution of the optical density (OD600) represents a transition between these two phenotypes. Further analysis revealed that the metabolic shift is governed by three different time scales under the experimental conditions. The rapid decline of the external pH level affects the kinetic properties of the ABE fermentation. Interestingly, the acidogenic population seems to be able to respond to these environmental changes until the pH drops below a critical value. Then, the fermentation pathway is seriously disturbed, so that the acidogenic cells are incapable to grow further. Simultaneously, a low level solventogenic population is increasing. The rise of the solventogenic phenotype starts around 20 hours after the pH-shift and is completed after approximate 50 hours. The origin of the solventogenic cells is still unsettled. We suggest two potential explanations: First, the clostridial population is heterogeneous during both metabolic states. Second, few acidogenic cells, selected by unknown criteria, are able to shift their phenotype. Intracellular variability inherent to microbial populations might cause this phenomenon.