Engineering <i>O</i> -Methyltransferase for the Biocatalytic Synthesis of Ribosome-Targeting Anisomycin Derivatives

Enzyme engineering provides a powerful route to diversify the chemical space of complex natural products for drug discovery. Here, we reprogram the ribosome-targeting antibiotic anisomycin through the engineering of its late-stage O-methyltransferase AniK. By integrating structure-guided rational design with artificial intelligence (AI)-assisted computational screening, we generated AniK variants with expanded cofactor promiscuity, enabling the transfer of diverse alkyl groups from noncanonical S-adenosylmethionine analogues. Concurrently, a structure-based generative algorithm (DELETE) was employed to design a focused library of phenolic hydroxyl modified anisomycin derivatives, which were synthesized via a one-pot enzymatic cascade using the engineered halide methyltransferase AclHMT-v16 and the optimized AniK mutant M144A-A133D. This integrated platform delivered 14 structurally diverse anisomycin derivatives, among which compounds 7 and 15 exhibited enhanced antidermatophyte activity while maintaining potent ribosome targeting capability. Our work establishes a reprogrammable enzymatic platform for the late-stage diversification of anisomycin, providing a modular strategy to access bioactive analogues with optimized therapeutic potential.