(667a) A Deuterium Labeling Strategy for Untargeted Metabolomics Analyses

Untargeted metabolomics analyses are hindered by the difficulty of assigning chemical formulae to unknown features in raw mass spectral data. For that reason, many metabolomics analyses rely on curated databases of authentic standards or known metabolites. Such search-based strategies cannot detect metabolites of unknown structure and function. Here we show that a novel stable isotope labeling strategy ? post-extraction deuterium labeling of labile protons ? enabled the identification of a family of novel small molecule metabolites that were not present in KEGG, LipidMaps, Metlin, or Metacyc databases. Our stable isotope labeling technique can be applied after metabolite extraction, making it amenable to use even in systems where traditional stable isotope labeling strategies are not possible.

The method was applied to a chemoautotrophic mixed-species microbial biofilm and enabled the identification of a family of highly abundant, unusual methylated phosphatidylethanolamine lipids. Unusual methylation patterns suggest that C1 metabolic pathways play an important role in the biosynthesis of this lipid family. Metabolomic and proteomic analyses of a suite of natural biofilm samples revealed intriguing correlations between these lipids and several rapidly evolving proteins of unknown function. Untargeted, stable isotope labeling-based metabolomic analyses thus have the potential to identify new, functionally important small molecules in microbial systems.