Deciphering rationales that determine the resistance of a lipase towards non-conventional media

The naturally existing enzymes have been evolved to work efficiently in aqueous environment of their hosts. Very often, their activity gets dramatically lost in the presence of non-conventional media (e.g. ionic liquids (ILs) and organic solvents (OS)) which are often present in industrial processes. Therefore, it is crucial to maintain/improve the enzyme resistance towards non-conventional media for the efficient operation of enzymes in industrial process conditions. Despite of the many successful stories of improving enzyme resistance towards many non-conventional media by protein engineering, there have been no geneneral principles for guiding researchers for efficient reengineering of enzymes. The aim of this work was to obtain protein engineering principles for improving IL and OS resistance of enzymes. A systematic study including site saturation mutagenesis (SSM) library generation of the complete sequence of a model enyzme Bacillus subtilis LipA (BSLA) and screening in presence of four ILs ([BMIM][Cl], [BMIM][Br], [BMIM][I], [BMIM][TfO]) and three OS (DMSO, TFE, 1,4-dioxane) were performed. The generation of 181 SSM-BSLA libraries was carried out by two PhD fellows in close cooperation (91 SSM-BSLA libraries by Victorine Josiane Frauenkron Machedjou and 90 by Alexander Fulton). Overall, 18547 clones were generated and screened which contained all possible 3620 single mutations as confirmed by sequencing. The number of beneficial positions and substitutions, type of beneficial amino acid exchange and location of positions was investigated. To the best of our knowledge, this is the first study based on such a complete set of experimental data to investigate IL/OS resistance of an enzyme.In a first study (described in chapter 3), the resistance of SSM-BSLA variants towards four [BMIM]-based ILs (Cl-, Br-, I-, and TfO-) was investigated. The highest number of improved variants was obtained for [BMIM][Cl]: 69% of all positions and 13% of all substitutions showed improvements. For the four ILs, at 50-69% of all positions at least one substitution led to improvement whereas only 6 13% of all substitutions displayed increased resistance in all four ILs. It was also noteworthy that higher improvements from single substitution (>2.5-fold) were difficult to achieve by screening random mutagenesis library since only 0.3% of all the substitutions resulted in improvement >2.5-fold. Furthermore, amino acid substitutions with polar and charged amino acids were demonstrated to considerably enhance ILs resistance of BSLA. Among the best variants, the highest improvement (11.3-fold) was obtained for [BMIM][Cl] for a substitution from charged acidic to aromatic (Asp91Trp). Two main lessons were obtained: a) resistance improvements were obtainable at 50-69% of all amino acid positions, explaining the success rate of small sized random mutant libraries; b) in total 6-13% of substitutions led to improved resistance. Among the beneficial substitutions 66-95% were substituted by chemically different amino acids (e.g. aromatic to polar/aliphatic/charged amino acids), indicating that mutagenesis methods introducing transversions should at least for lipases like BSLA be favored to improve IL resistance.In a second study (described in chapter 4), the resistance pattern of SSM-BSLA variants towards three OS (DMSO, TFE, and 1,4-dioxane) was investigated. For 1,4-dioxane and TFE, 4-5% substitutions at 41% positions led to improved resistance. As observed for IL resistance of BSLA, improvements >2.5-fold were difficult to achieve with single point mutation, only 0.03% and 0.17% variants were obtained for DMSO / 1,4-dioxane and TFE, respectively. In addition, substitutions of BSLA WT amino acid residues with charged amino acids residues predominantly led to improved OS resistance towards TFE and 1,4-dioxane, whereas polar substitutions were preferred for DMSO.Overall, two features for BSLA resistance in non-conventional media were observed: 1) resistance improvements were obtainable at >41% (in OS) and >50% (in ILs) of all amino acid positions of BSLA, indicating why improved enzyme variants can often be identified from small sized random mutant libraries (1000-2000 clones) and 2) often only few substitutions per amino acid positions led to improved resistance (4-13% of all substitutions) for OS and ILs. Among the beneficial substitutions, 66-95% and 58-95% were exchanged with chemically different amino acids (e.g. aromatic amino acids substituted by polar/aliphatic/charged amino acid) for ILs and OS, respectively. The latter indicates that mutagenesis methods that introduce transversions should, at least for lipases like BSLA be favored to generate variants with improved OS and/or IL resistance. Furthermore, the highest number of improved BSLA variants were obtained for [BMIM][Cl] and DMSO among the four ILs and the three OS, respectively