Abstract The high catalytic efficiency of enzymes under reaction conditions is one of the main goals in biocatalysis. Despite the dramatic progress in the development of more efficient biocatalysts by protein design, the search for natural enzymes with useful properties remains a promising strategy. The pyridoxal 5'-phosphate (PLP)-dependent transaminases represent a group of industrially important enzymes due to their ability to stereoselectively transfer amino groups between diverse substrates; however, the complex mechanism of substrate recognition and conversion makes the design of transaminases a challenging task. Here we report a detailed structural and kinetic study of thermostable transaminase from the bacterium Thermobaculum terrenum (TaTT) using the methods of enzyme kinetics, X-ray crystallography and molecular modeling. TaTT can convert L-branched-chain and L-aromatic amino acids as well as (R)-(+)-1-phenylethylamine at a high rate and with high enantioselectivity. The structures of TaTT in complex with the cofactor pyridoxal 5′-phosphate covalently bound to enzyme and in complex with its reduced form, pyridoxamine 5′-phosphate, were determined at resolutions of 2.19 A and 1.5 A, and deposited in the Protein Data Bank as entries 6GKR and 6Q8E, respectively. TaTT is a fold type IV PLP-dependent enzyme. In terms of structural similarity, the enzyme is close to known branched-chain amino acid aminotransferases, but differences in characteristic sequence motifs in the active site were observed in TaTT compared to canonical branched-chain amino acid aminotransferases, which can explain the improved binding of aromatic amino acids and (R)-(+)-1-phenylethylamine. This study has shown for the first time that high substrate specificity towards both various l -amino acids and (R)-primary amines can be implemented within one pyridoxal 5′-phosphate-dependent active site of fold type IV. These results complement our knowledge of the catalytic diversity of transaminases and indicate the need for further biochemical and bioinformatic studies to understand the sequence-structure-function relationship in these enzymes.
Ekaterina Yu. Bezsudnova, Konstantin M. Boyko, Konstantin M. Boyko