Unlocking the Substrate Acceptance of Phenylalanine Amine Dehydrogenase Enables the Asymmetric Synthesis of Pharmaceutical N-Heterocyclic Primary Amines

N-Heterocyclic primary amines are highly valuable and recurrent synthons in the pharmaceutical industry. Amine dehydrogenase (AmDH)-catalyzed direct asymmetric reductive amination of the readily available N-heterocyclic ketones represents a promising approach for synthesizing N-heterocyclic primary amines. However, the limited substrate acceptance of AmDHs restricts their application in reductive amination. In this endeavor, we unlocked the substrate acceptance of phenylalanine amine dehydrogenase from Bacillus badius (F-BbAmDH) to access a panel of N-Boc-substituted heterocyclic ketones with a broad range of structural features. Based on the combination of a reverse substrate design strategy and structural-guided steric hindrance mutagenesis, two active F-BbAmDH mutants with expanded N-heterocyclic ketone specificity toward N-Boc-3-piperidone were generated, and two rounds of additional iterative site mutagenesis further increased the catalytic activity by 143.4-fold. The optimal triple mutant M3-2 displayed a significantly extended scope of N-Boc-substituted heterocyclic ketones, and its practical asymmetric synthesis performance was confirmed in the gram-scale synthesis of the key intermediate of alogliptin, (R)-1-Boc-3-piperidinamine, achieving >99% ee, 86% conversion, and 80% (4.0 g) isolated yield. This work lays the foundation for the biocatalytic synthesis of structurally diverse N-heterocyclic primary amines and gives referable guidance for engineering oxidoreductases sharing an analogous pocket into versatile biocatalysts.