Exploitation of active site flexibility-low temperature activity relation for engineering broad range temperature active enzymes

Abstract

Differences in the structural and thermodynamic properties of enzymes adapted to different temperatures indicate that broad range temperature active enzymes can be designed by incorporating cold activity in thermophilic enzymes. This is based on a concept that the cold activity and thermostability are not mutually exclusive and that cold activity in psychrophilic enzymes is associated with active site flexibility. In Wang et al. Biochem. Eng. J. 2021, 174, 10803, we identified two point mutants of Geobacillus thermocatenulatus lipase (GTL) which were screened to improve active site flexibility. Even though the identified thermophilic mutants had psychrophilic traits, we observed complex trends such as higher kinetic stability and substrate-dependent activity–temperature relation on further analysis. In this work, we apply molecular dynamics simulations and network theory to show that the changes in GTL properties with the selected mutations cannot be directly associated with active site flexibility. Our computational results indicate the mutations resulted in residues with both higher and lower flexibility, which are both proximal and away (>1.5 nm) from the active site. We show that the intricate changes in the flexibility of residues distal from the mutation site can be rationalized by the altered dependency between residue–residue fluctuations with mutation. These alterations in residue–residue flexibility dependency are a consequence of the redistribution of the inter-residue interactions from the mutation site to other residues, which are driven by several tightly connected charged residues. This indicates design rules associated with residue–residue flexibility correlations are critical in applying site-directed mutagenesis to successfully exploit active site flexibility–activity relation for incorporating low temperature activity in thermophilic enzymes. Similarly, such correlations can be valuable in minimizing false positives in high-throughput screening methods based on directed evolution and/or machine learning-based engineering of enzyme activity–temperature relation.