Aromatic aldehydes find extensive applications in food,
perfume,
pharmaceutical, and chemical industries. However, a limited natural
enzyme selectivity has become the bottleneck of bioconversion of aromatic
aldehydes from natural phenylpropanoid acids. Here, based on the original
structure of feruloyl–coenzyme A (CoA) synthetase (FCS) from Streptomyces sp. V-1, we engineered five substrate-binding
domains to match specific phenylpropanoid acids. FcsCIAE407A/K483L, FcsMAE407R/I481R/K483R, FcsHAE407K/I481K/K483I, FcsCAE407R/I481R/K483T, and FcsFAE407R/I481K/K483R showed 9.96-, 10.58-, 4.25-, 6.49-, and 8.71-fold enhanced catalytic
efficiency for degrading CoA thioesters of cinnamic acid, 4-methoxycinnamic
acid, 4-hydroxycinnamic acid, caffeic acid, and ferulic acid, respectively.
Molecular dynamics simulation illustrated that novel substrate-binding
domains formed strong interaction forces with substrates’ methoxy/hydroxyl
group and provided hydrophobic/alkaline catalytic surfaces. Five recombinant E. coli with FCS mutants were constructed with the
maximum benzaldehyde, p-anisaldehyde, p-hydroxybenzaldehyde, protocatechualdehyde,
and vanillin productivity of 6.2 ± 0.3, 5.1 ± 0.23, 4.1
± 0.25, 7.1 ± 0.3, and 8.7 ± 0.2 mM/h, respectively.
Hence, our study provided novel and efficient enzymes for the bioconversion
of phenylpropanoid acids into aromatic aldehydes.