Control of Proteolysis of Recombinant Proteins in Escherichia coli

Proteolysis was shown to have a significant impact on the accumulation and the final yield of recombinant proteins. Much smaller intracellular concentrations of the proteins SpA and ZZT2 compared to their stable versions SpA-βgalactosidase and ZZT0 were explained by their high proteolysis rate. The reduction of proteolysis rate in Clp protease-deficient strain enabled to increase the intracellular ZZT2 concentration two-fold, but impaired the cell growth. A method to calculate a hypothetical accumulation of the product assuming complete stabilization was described, which reveals a potential to increase the productivity by eliminating proteolysis. The effects of scale-up on the production of the ZZT2 protein were investigated in reactors of different configuration and scale: a well-mixed lab-scale bioreactor, an industrial scale 12-m3 bioreactor, and a two-compartment scale-down bioreactor. The production of the recombinant protein ZZT2 was highest in the lab-scale cultivation (50 mg/g), however in the large-scale the ZZT2 concentration declined 8 hours after induction from its highest value of 45 to a final value of 35 mg/g. The final yield of the ZZT2 protein in the scale-down cultivations was even lower than in the large-scale. The higher proteolysis rate and declining synthesis rate are thought to be the reasons for the lower yields of the protein ZZT2 in the large-scale and scale-down cultivations. The heterogeneities in the large-scale and scale-down reactors appeared to prevent a decline in viable cell count and cell death, which are usually observed at the end of fed-batch cultivations in lab-scale reactors. Two mixtures of amino acids were fed together with glucose in fed-batch cultures aiming to improve a biomass yield and a concentration of the ZZT2 protein. One mixture contained all twenty amino acids in quantities, which would be sufficient for a synthesis of a half of biomass proteins, including the ZZT2 protein. The other mixture contained only five socalled “protein amino acids”, which were reported to be primarily incorporated into biomass: alanine, arginine, methionine, histidine and phenylalanine. Although in both cases feeding of amino acids improved the biomass yield and slightly reduced proteolysis rate, the concentration of recombinant protein ZZT2 was less than half compared to the control. A decline of synthesis rate of ZZT2 is thought to be the reason for this. The mixture of five “protein amino acids” in shake flask culture even inhibited the cellular growth, which is seen as an indicator of disturbed metabolism caused by added amino acids. A novel cultivation technique, a temperature-limited fed-batch culture, was applied for a production of the ZZT2 protein. Low temperature and surplus of glucose enabled to limit endotoxin release and decrease the degradation rate of ZZT2. More work is needed to maintain glucose concentration in a desired range to avoid acetate accumulation during this type of cultivation. The absence of both oxygen and glucose in the transition period from cultivation to downstream processing resulted in a decrease of the ATP pool to less than 0.5 mM and almost complete stabilisation of protein A. Neither interruption of the feeding under aerobic conditions nor anaerobic conditions in presence of glucose could stabilise protein A completely and the intracellular ATP pool did not decrease to less than 0.75-1 mM by this treatment. Therefore, protein stabilisation and decrease of the ATP pool were correlated in experiments in vivo. The concentrations of ADP and AMP increased during starvation and could also play a role in stabilisation of the protein. ATP may be a limiting factor of proteolysis also during further steps of downstream processing. Its concentration decreased by 80-90% during harvesting and centrifugation of biomass and even further during disruption of cells.