An effective strategy for soluble bovine enterokinase expression in Escherichia coli.

Abstract

Bovine enterokinase light chain (bEkL) is a serine protease, widely used for the specific cleavage of affinity tags from various recombinant proteins. However, getting soluble expression in Escherichia coli is a challenging task given the presence of multiple cysteines and four disulfide bonds. Strategies that have only been partially successful involve mutating the gene or covalent attachment of solubility tags. We demonstrate a simpler and more efficient production method that combines different strategies like co-expressing the GroES-GroEL chaperone in E. coli SHuffle cells, lowering temperatures to 18 °C post-induction, and ensuring the sufficient accumulation of GroES-GroEL in the cytoplasm before inducing the bEkL gene. A trade-off exists between producing too little GroES-GroEL and promoting inclusion body formation (of bEkL) or expressing too much GroES-GroEL thereby reducing bEkL yields. This optimum level was determined by varying the time difference between the two inductions, and the best results obtained when the co-expressed pGro7 plasmid was induced 2 h before bEkL induction, thus avoiding inclusion body formation. Interestingly, when we delayed the induction of GroES-GroEL to an OD₆₀₀ of 4, which in turn further delayed the induction of bEkL to an OD₆₀₀ of 10, we observed a slowdown in expression rates, but a further improvement in soluble yields. These yields increased over a 36 h long period post-induction at 18 °C in TB medium, where nutrient starvation was prevented by the addition of a concentrated pulse of substrate 20 h post-induction. This slow and steady buildup of soluble bEkL in the cellular cytoplasm allowed us to reach a concentration of 10 mg L^-1 with a high specific activity of approximately 5,000 AU µg^-1. Finally, Ni-NTA affinity chromatography was used to purify the soluble bEkL, and we obtained > 90% homogenous bEkL protein product. The enzymatic activity of this protein was tested using a fusion protein, containing an enterokinase recognition site, as a substrate which showed that the net increase in activity was around 20-fold compared to the initial expression levels obtained with SHuffle cells.