Transaminases show potential for the industrial synthesis
of important
pharmaceutical ingredients. However, these naturally occurring enzymes
show poor activity toward bulky N-heterocyclic compounds.
To produce a catalyst with enhanced catalytic efficiency, this study
redesigned an (R)-selective transaminase from Rhodobacter sp. 140A (RbTA). Key residues
for substrate binding were identified by molecular docking and molecular
dynamics simulations. A “simplified amino acid alphabet,”
consisting of amino acids of different sizes (Phe, Asn, Val, and Ala),
was then used to fine-tune the substrate-binding pocket by producing
a small but smart variant library. Residue Y125 was found to be critical
for substrate binding, and variant RbTAM1(Y125A), exhibiting a remarkable activity enhancement, was obtained.
Through combined mutation, the most active variant, RbTAM2(Y125A/I6A/L7A/L158V), was constructed, exhibiting
1064-fold greater catalytic efficiency (k
cat/K
m) toward substrate N-Boc-3-piperidone (7a) than the wild-type enzyme. This
variant also exhibited significantly improved activity (4–110-fold)
toward a series of cyclic and bulky heterocyclic ketones. Structure-guided
analysis of variant Y125A and molecular simulations revealed that
the introduction of residue A125 enlarged the substrate-binding pocket
volume and enabled additional hydrophobic interactions with the substrate,
facilitating binding in a more favorable conformation for catalysis.
The activity of variant RbTAM2 was verified
in the gram-scale synthesis of chiral N-heterocyclic
amine (R)-1-Boc-3-piperidinamine (7b), achieving 99% conversion and a space-time yield of 222 g L–1 d–1.