Linking process and metabolic modelling for the estimation of carbon flux distribution in Corynebacterium glutamicum growth in spent sulfite liquor

Abstract

Process monitoring in microbial cultures became feasible thanks to the development of accurate measurement devices, including in-situ probes to monitor biomass growth, oxygen, carbon dioxide and sugar consumption. In comparison, estimating the metabolic fluxes of the cell factories still rely on analysis based on an under-determined set of equations, and requires an expensive and time consuming methods of verification. This problem intensifies in the presence of complex substrates, in which different sugars are utilized in parallel by the cell factories. In the present study, a growth experiment of Corynebacterium glutamicum in spent sulfite liquor was studied. The bioprocess was monitored during batch and fed-batch phases, and a parameter estimation routine was conducted to define a process model and the corresponding uptake rates. A tracking optimization algorithm minimized the error between the measured process fluxes and the equivalent fluxes of the elementary flux modes. The results indicate that the optimization technique obtained a set of elementary modes that are closer to reality than the computed from the metabolic analysis. Taken together, we show that an online estimation of metabolic flux distribution of C glutamicum based on a set of process measurement signals was possible with an optimization function that links the process and metabolic model. The procedure can be complementary to the sophisticated and expensive C NMR experimental analytical technique.