RÉSUMÉ
The inherent evolvability of promiscuous enzymes endows them with great potential to be artificially evolved for novel functions. Previously, we succeeded in transforming a promiscuous acylaminoacyl peptidase (apAAP) from the hyperthermophilic archaeon Aeropyrum pernix K1 into a specific carboxylesterase by making a single mutation. In order to fulfill the urgent requirement of thermostable lipolytic enzymes, in this paper we describe how the substrate preference of apAAP can be further changed from p-nitrophenyl caprylate (pNP-C8) to p-nitrophenyl laurate (pNP-C12) by protein and solvent engineering. After one round of directed evolution and subsequent saturation mutagenesis at selected residues in the active site, three variants with enhanced activity towards pNP-C12 were identified. Additionally, a combined mutant W474V/F488G/R526V/T560W was generated, which had the highest catalytic efficiency (k (cat)/K (m)) for pNP-C12, about 71-fold higher than the wild type. Its activity was further increased by solvent engineering, resulting in an activity enhancement of 280-fold compared with the wild type in the presence of 30% DMSO. The structural basis for the improved activity was studied by substrate docking and molecular dynamics simulation. It was revealed that W474V and F488G mutations caused a significant change in the geometry of the active center, which may facilitate binding and subsequent hydrolysis of bulky substrates. In conclusion, the combination of protein and solvent engineering may be an effective approach to improve the activities of promiscuous enzymes and could be used to create naturally rare hyperthermophilic enzymes.