Proc. Natl. Acad. Sci. USA 110, E4407–E4416 (2013)

Secondary metabolites are produced by a broad range of microorganisms, and many of these 'natural products' are used clinically as antibiotics or anticancer agents. The discovery and characterization of a therapeutically useful natural product can be challenging, especially if it is glycosylated, as there are >100 different sugars found in microbial glycosylated natural products (GNPs). Most GNPs are inactive if the carbohydrates are missing, so knowing the structures of these sugars and understanding how they are attached to the natural product scaffold can help unravel how the GNP elicits its biological effect. Kersten et al. now report a new method— 'glycogenomics'—that can be used to identify GNPs and the gene clusters responsible for their biosynthesis. The authors first used tandem MS to generate fragmentation data for 83 microbial monosaccharides; using the resulting glycogenetic code, they were able to link 18 of the 20 GNPs in a 'test set' to the correct biosynthetic gene cluster. The methodology was then used to show that an anthracycline polyketide was produced by a strain of Streptomyces that was not previously known to do so and to determine that Salinispora arenicola CNB-527, a marine actinobacterium, makes a new arenimycin derivative that can kill multidrug-resistant strains of Staphylococcus aureus. This approach should facilitate the rapid identification of other biologically active GNPs and the biosynthetic gene clusters that assemble them.