Expression of a lipase in prokaryote and eukaryote host systems allowing engineering

University dissertation from Stockholm : KTH

Abstract: Pseudozyma (Candida) antarctica lipase B (PalB) was expressed in Escherichia coli facilitating protein engineering. The lack of glycosylation was evaluated for a deeper understanding of the difficulties in expressing PalB in E. coli. Different systems were tested: periplasmic expression in Rosetta (DE3), cytosolic expression in Rosetta-gami 2(DE3), Origami 2(DE3), and coexpression of groES and groEL. Periplasmic expression resulted 5.2 mg/L active PalB at 16 °C in shake flasks. This expression level was improved by using the EnBase technology, enabling fed-batch cultivation in 24-deep well scale. The feed rate was titrated with the addition of ?-amylase, which slowly releases glucose as energy source. Different media were evaluated where the EnBase mineral salt medium resulted in 7.0 mg/L of active PalB.Protein secreted directly into the media was obtained using the constitutive glyceraldehyde-3-phosphate dehydrogenase (GAP) promoter for screening and production of PalB in P. pastoris. A protease sensitive fusion protein CBM-PalB (cellulose-binding module) was used as a model system. When optimised, the expression system resulted in 46 mg/L lipase in 72 hours in shake flask, 37 mg/L lipase in 28 hours in 96-deep-well plate format, and 2.9 g PalB per 10 L bioreactor cultivation.The E. coli expression system was used to express a small focused library of PalB variants, designed to prevent water from entering the active site through a hypothesised tunnel. Screening of the library was performed with a developed assay, allowing for simultaneous detection of both transacylation and hydrolytic activity. From the library a mutant S47L, in which the inner part of the tunnel was blocked, was found to catalyse transacylation of vinyl butyrate in 20 mM butanol 14 times faster than hydrolysis. Water tunnels, assisting water in reaching the active sites, were furthermore found by molecular modelling in many hydrolases. Molecular modelling showed a specific water tunnel in PalB. This was supported by experimental data, where the double mutant Q46A S47L catalysed transacylation faster than hydrolysis compared to the wild type PalB.