Targeted discovery of aromatic glycosides with dual detoxification effects via a highly customized molecular...

Natural products embedded within complex metabolomes are valuable sources of drug leads. Untargeted metabolomics using cosine-based MS² comparisons is widely used for discovering bioactive molecules. To improve annotation accuracy and resolution, alternative algorithms have been developed to complement cosine-based MS² comparison. Here, we present MSanalyst, a user-friendly platform that integrates 46 distinct mass spectral similarity algorithms. Benchmarking with microbial metabolite datasets and over three million MS² spectral pairs demonstrated that complementary algorithms substantially enhance the detection of metabolite-metabolite spectral connections. Applying MSanalyst to Kutzneria viridogrisea DSM 43850 led to the discovery of a class of aromatic glycosides, the kutznaposides. Biological assays and multi-omics analyses revealed that kutznaposides C-F arise from a previously unrecognized menaquinone shunt pathway, enabling the host to mitigate oxidative stress and avoid self-toxicity. Collectively, these findings highlight the potential of MSanalyst to uncover hidden metabolites, metabolic pathways, and their biological functions.