Login / Signup

Substrate scope expansion of 4-phenol oxidases by rational enzyme selection and sequence-function relations.

Daniel EggerichsNils WeindorfHeiner G WeddelingInja M Van der LindenDirk Tischler
Published in: Communications chemistry (2024)
Enzymes are natures' catalysts and will have a lasting impact on (organic) synthesis as they possess unchallenged regio- and stereo selectivity. On the downside, this high selectivity limits enzymes' substrate range and hampers their universal application. Therefore, substrate scope expansion of enzyme families by either modification of known biocatalysts or identification of new members is a key challenge in enzyme-driven catalysis. Here, we present a streamlined approach to rationally select enzymes with proposed functionalities from the ever-increasing amount of available sequence data. In a case study on 4-phenol oxidoreductases, eight enzymes of the oxidase branch were selected from 292 sequences on basis of the properties of first shell residues of the catalytic pocket, guided by the computational tool A 2 CA. Correlations between these residues and enzyme activity yielded robust sequence-function relations, which were exploited by site-saturation mutagenesis. Application of a peroxidase-independent oxidase screening resulted in 16 active enzyme variants which were up to 90-times more active than respective wildtype enzymes and up to 6-times more active than the best performing natural variants. The results were supported by kinetic experiments and structural models. The newly introduced amino acids confirmed the correlation studies which overall highlights the successful logic of the presented approach.
Keyphrases
  • amino acid
  • copy number
  • structural basis
  • gene expression
  • machine learning
  • deep learning
  • big data
  • artificial intelligence
  • metal organic framework