Cloning, characterization, and computer- aided evolution of a thermostable laccase of the DUF152 family from Klebsiella michiganensis

Bacterial laccases exhibit relatively high optimal reaction temperatures and possess a broad substrate spectrum, thereby expanding the range of potential applications for laccase enzymes. This study aims to investigate the molecular evolution of bacterial laccases using computational 3D-structure prediction and molecular docking tools such as AlphaFold2, Metal3D, AutoDockVina, and Rosetta. We isolated a bacterium with laccase activities from fecal samples from a Hermann´s tortoise (Testudo hermanni), identified it as Klebsiella michiganensis using 16S rRNA sequencing and nanopore genome sequencing, and then identified a sequence encoding a laccase with a predicted molecular weight of approximately 27.5 kDa. Expression of the corresponding, chemically synthesized DNA fragment resulted in the isolation of an active laccase. The enzyme showed considerable thermostability, retaining 21% of its activity after boiling for 30 min. Using state-of-the-art information technology and machine learning techniques, we conducted 3D-structure prediction on this sequence, predicted its copper-ion binding sites, and validated these predictions through site-directed mutagenesis and expression. Subsequently, we performed computer-aided evolution studies on this sequence and found that 90% of the results from the selected mutations exhibited improved performance. In summary, this study not only revealed a novel laccase but also demonstrated an efficient approach for advancing research on the molecular evolution of bacterial laccases using cutting-edge machine learning, next-generation sequencing, traditional bioinformatics approaches, and laboratory techniques, providing an effective strategy for discovering and design new bacterial laccases.