Quantitative molecular networking to profile marine cyanobacterial metabolomes

Untargeted liquid chromatography-MS (LC-MS) is used to rapidly profile crude natural product (NP) extracts; however, the quantity of data produced can become difficult to manage. Molecular networking based on MS/MS data visualizes these complex data sets to aid their initial interpretation. Here, we developed an additional visualization step for the molecular networking workflow to provide relative and absolute quantitation of a specific compound in an extract. The new visualization also facilitates combination of several metabolomes into one network, and so was applied to an MS/MS data set from 20 crude extracts of cultured marine cyanobacteria. The resultant network illustrates the high chemical diversity present among marine cyanobacteria. It is also a powerful tool for locating producers of specific metabolites. In order to dereplicate and identify culture-based sources of known compounds, we added MS/MS data from 60 pure NPs and NP analogs to the 20-strain network. This dereplicated six metabolites directly and offered structural information on up to 30 more. Most notably, our visualization technique allowed us to identify and quantitatively compare several producers of the bioactive and biosynthetically intriguing lipopeptide malyngamide C. Our most prolific producer, a Panamanian strain of Okeania hirsuta (PAB10FEB10-01), was found to produce at least 0.024 mg of malyngamide C per mg biomass (2.4%, w/dw) and is now undergoing genome sequencing to access the corresponding biosynthetic machinery.