Creating a non-native bio-amidation activity in asparagine synthetase via biosensor-guided directed evolution for the synthesis of (S)-2-aminobutyramide.

The direct asymmetric synthesis of chiral amides remains a focal point in green manufacturing, particularly for (S)-2-aminobutyramide whose traditional production is often hampered by low atom economy and arduous chiral resolution steps. Here, we report the creation of a non-native bio-amidation activity in Escherichia coli asparagine synthetase A (AsnA) for the proof-of-concept one-step biosynthesis of (S)-2-aminobutyramide, a critical intermediate for anti-epileptic drugs. To overcome the bottleneck of high-throughput screening, we engineered the AmiC-AmiR transcription factor into a highly sensitive biosensor (RC-4-Y83H) by remodeling its ligand-binding pocket, achieving a 23-fold induction response specifically toward (S)-2-AB. Utilizing this biosensor to screen saturation mutagenesis libraries, we identified an evolved variant (K77L/E120L/R255W) that showed a reversed substrate preference between the two substrates tested, converting (S)-2-aminobutyric acid to (S)-2-AB with retention of configuration. Molecular dynamics simulations and free-energy landscape analysis reveal that the mutations fundamentally remodel the active site, inducing a 180° flip of the substrate binding mode. This structural reorientation positions the non-native carboxyl group in a catalytically productive geometry relative to ATP, a conformation inaccessible in the wild-type enzyme. As a proof of concept, this work does not aim to deliver an industrially competitive biocatalyst; instead, it shows that the specificity of a native amide-forming enzyme can be reprogrammed toward a non-native alpha-amino acid, and it provides a generalizable, biosensor-guided framework for such reprogramming.