Recombinant protein production in high cell density fed-batch cultures of E. Coli: application of dynamic flux balance analysis

KIKOT, PAMELA; CASTAÑEDA, MARÍA TERESITA; DE BATTISTA, HERNÁN; GRASSELLI, MARIANO

Congreso; XVIII Congreso Argentino de Microbiología General (SAMIGE); 2023

Due to several advantages, E. coli is the preferred choice for expressing nonglycosylated proteins in the biotech industry. Most expressed recombinant proteins in this organism accumulate intracellularly, and productivity is proportional to cell density. To produce high cell-density culture (HCDC), fed-batch techniques are performed. However, it can lead to enhanced acetate secretion (and consequently reduced yield and inhibited cell growth) due to carbon overflow metabolism and the metabolic burden. Metabolic engineering strategies, such as dynamic flux balance analysis (dFBA), can be applied for HCDC optimization. This analysis describes the kinetics of substrates and products concentrations which depend on intracellular metabolic flux distribution.In this study, experimental results from fed-batch cultures of E. coli BL21(DE3) and E. coli BL21(DE3) pLysS were compared with dFBA predictions. The strains harbored different plasmids: pET-28a(+)-SpA and pET-21a(+)-mAV; for the production of a Staphylococcal protein A and monomeric avidin, respectively. The cultures were conducted in a 5 L stirred tank bioreactor (BIOSTAT Aplus, Sartorius) with 20 g/l glucose mineral medium and a 300 g/l glucose feed media. dFBA analysis was performed using the metabolic models iHK1487 and iECBD_1354 for describing E. coli BL21(DE3) and E. coli BL21(DE3) pLysS metabolism, respectively. dFBA implementation was performed in the DFBAlab toolbox for MATLAB, using the Gurobi solver. DFBAlab obtains unique exchange fluxes by applying lexicographic optimization; therefore, objective functions were maximized in the following order: biomass production (X), glucose consumption (G), ammonia consumption (N), oxygen consumption (O), and acetate production (A). The exchange reaction fluxes of substrates and products were described by Michaelis-Menten expressions for G, N, O, andA. Parameters were taken from the bibliography and experimental data. The initial conditions of volume, substrates, and products of dFBA analysis were in concordance with the experimental conditions, and their dynamics were represented by ordinary differential equations (ODEs).The experimental cultures showed differences between both strains. While BL21(DE3) generates higher amounts of biomass (70 g/l), BL21(DE3) pLysS has lower yield and higher oxygen demand. It is demonstrated that metabolic models and dFBA analysis are adequate to predict both strains' in vivo phenotypes. Predictions were highly dependent on dFBA implementation, such as objective function selection and kinetic parameters. The methodology has the potential as a simulation and optimization platform to define rational strategies for E. coli cultivation in HCDC, thus saving time and resources.