Lawrence Berkeley National Laboratory

Repurposing the intrinsic properties of natural enzymes can offer a viable solution to current synthetic challenges through the development of novel biocatalytic processes. Although amino acid racemases are ubiquitous in living organisms, an amine racemase (AR) has not yet been discovered despite its synthetic potential for producing chiral amines. Here, we report the creation of an AR based on the serendipitous discovery that amine transaminases (ATAs) can perform stereoinversion of 2-aminobutane. Kinetic modeling revealed that the unexpected off-pathway activity results from stereochemically promiscuous futile cycles due to incomplete stereoselectivity for 2-aminobutane. This finding motivated us to engineer an S-selective ATA through in silico alanine scanning and empirical combinatorial mutations, creating an AR with broad substrate specificity. The resulting AR, carrying double point mutations, enables the racemization of both enantiomers of diverse chiral amines in the presence of a cognate ketone. This strategy may be generally applicable to a wide range of transaminases, paving the way for the development of new-to-nature racemases.