- Diaminopelargonic acid transaminase from Psychrobacter cryohalolentis is active towards (S)-(-)-1-phenylethylamine, aldehydes and α-diketones.
Diaminopelargonic acid transaminase from Psychrobacter cryohalolentis is active towards (S)-(-)-1-phenylethylamine, aldehydes and α-diketones.
Substrate and reaction promiscuity is a remarkable property of some enzymes and facilitates the adaptation to new metabolic demands in the evolutionary process. Substrate promiscuity is also a basis for protein engineering for biocatalysis. However, molecular principles of enzyme promiscuity are not well understood. Even for the widely studied PLP-dependent transaminases of class III, the reliable prediction of the biocatalytically important amine transaminase activity is still difficult if the desired activity is unrelated to the natural activity. Here, we show that 7,8-diaminopelargonic acid transaminase (synthase), previously considered to be highly specific, is able to convert (S)-(-)-1-phenylethylamine and a number of aldehydes and diketones. We were able to characterize the (S)-amine transaminase activity of 7,8-diaminopelargonic acid transaminase from Psychrobacter cryohalolentis (Pcryo361) and analyzed the three-dimensional structure of the enzyme. New substrate specificity for α-diketones was observed, though only a weak activity towards pyruvate was found. We examined the organization of the active site and binding modes of S-adenosyl-L-methionine and (S)-(-)-1-phenylethylamine using X-ray analysis and molecular docking. We suggest that the Pcryo361 affinity towards (S)-(-)-1-phenylethylamine arises from the recognition of the hydrophobic parts of the specific substrates, S-adenosyl-L-methionine and 7-keto-8-aminopelargonic acid, and from the flexibility of the active site. Our results support the observation that the conversion of amines is a promiscuous activity of many transaminases of class III and is independent from their natural function. The analysis of amine transaminase activity from among various transaminases will help to make the sequence-function prediction for biocatalysis more reliable.