Substrate specificity engineering of leucine dehydrogenase for efficient synthesis of aromatic L-homophenylalanine

Leucine dehydrogenase (LeuDH)-catalyzed asymmetric reductive amination of α-keto acids represents a promising approach for synthesizing L-amino acids. However, the stringent substrate specificity restricts its application primarily to aliphatic amino acid synthesis. Herein, we conducted substrate-specific engineering of Bacillus subtilis leucine dehydrogenase (BsLeuDH) for bulky aromatic L-homophenylalanine (L-HPA) production. Through four rounds of iterative steric-attenuated pocket remodeling, five active mutants with expanded substrate acceptance toward the non-natural precursor 2-oxo-4-phenylbutyric acid (2-OPBA) were identified. The optimal mutant, M3-2 (L40V/A113G/V294G), exhibited a 438-fold increase in catalytic efficiency (k_cat/K_m) toward 2-OPBA and shifted substrate preference to the bulky aromatic α-keto acids. Preparative-scale synthesis confirmed its potential for L-HPA production, achieving 215.0 g/L∙h space-time yields (STY), >99% ee. Molecular docking and MD simulations indicate that mutations in M3-2 enlarged the substrate-binding pocket and improved conformational transition efficiency, thereby facilitating accommodation of the large aromatic group of 2-OPBA and the formation of the pre-reaction state. More importantly, introducing M3-2 mutations into five homologous LeuDHs (64%-87% identity to BsLeuDH) similarly enhanced 2-OPBA acceptance, further revealing the universality of this steric-attenuation strategy in shifting LeuDH substrate specificity toward bulky aromatic substrates.