Methylglucosylation of aromatic amino and phenolic moieties of drug-like biosynthons by combinatorial biosynthesis

Glycosylation is a prominent strategy to optimize the pharmacokinetic and pharmacodynamic properties of drug-like small-molecule scaffolds by modulating their solubility, stability, bioavailability, and bioactivity. Glycosyltransferases applicable for “sugarcoating” various small-molecule acceptors have been isolated and characterized from plants and bacteria, but remained cryptic from filamentous fungi until recently, despite the frequent use of some fungi for whole-cell biocatalytic glycosylations. Here, we use bioinformatic and genomic tools combined with heterologous expression to identify a glycosyltransferase–methyltransferase (GT–MT) gene pair that encodes a methylglucosylation functional module in the ascomycetous fungus Beauveria bassiana. The GT is the founding member of a family nonorthologous to characterized fungal enzymes. Using combinatorial biosynthetic and biocatalytic platforms, we reveal that this GT is a promiscuous enzyme that efficiently modifies a broad range of drug-like substrates, including polyketides, anthraquinones, flavonoids, and naphthalenes. It yields both O- and N-glucosides with remarkable regio- and stereospecificity, a spectrum not demonstrated for other characterized fungal enzymes. These glucosides are faithfully processed by the dedicated MT to afford 4-O-methyl-glucosides. The resulting “unnatural products” show increased solubility, while representative polyketide methylglucosides also display increased stability against glycoside hydrolysis. Upon methylglucosi-dation, specific polyketides were found to attain cancer cell line-specific antiproliferative or matrix attachment inhibitory activities. These findings will guide genome mining for fungal GTs with novel substrate and product specificities, and empower the efficient combinatorial biosynthesis of a broad range of natural and unnatural glycosides in total biosynthetic or biocatalytic formats.