Therapeutic monoclonal antibodies are mainly produced in mammalian cells to date. However, unglycosylated antibody fragments can also be produced in the bacterium Escherichia coli which brings several advantages, like growth on cheap media and high productivity. One of the most popular E. coli strains for recombinant protein production is E. coli BL21(DE3) which is usually used in combination with the pET expression system. However, it is well known that induction by isopropyl β-D-1-thiogalactopyranoside (IPTG) stresses the cells and can lead to the formation of insoluble inclusion bodies. In this study, we revisited the pET expression system for the production of a novel antibody single-chain variable fragment (scFv) with the goal of maximizing the amount of soluble product. Thus, we (1) investigated whether lactose favors the recombinant production of soluble scFv compared to IPTG, (2) investigated whether the formation of soluble product can be influenced by the specific glucose uptake rate (q s,glu) during lactose induction, and (3) determined the mechanistic correlation between the specific lactose uptake rate (q s,lac) and q s,glu. We found that lactose induction gave a much greater amount of soluble scFv compared to IPTG, even when the growth rate was increased. Furthermore, we showed that the production of soluble protein could be tuned by varying q s,glu during lactose induction. Finally, we established a simple model describing the mechanistic correlation between q s,lac and q s,glu allowing tailored feeding and prevention of sugar accumulation. We believe that this mechanistic model might serve as platform knowledge for E. coli.
Keywords: Antibody fragment; Escherichia coli BL21(DE3); Lactose induction; Mechanistic model; Soluble protein; pET expression system.