Metabolite fingerprinting: A powerful metabolomics approach for marker identification and functional gene annotation

Abstract

Non-targeted metabolome approaches aim to detect metabolite markers related to stress, disease, developmental or genetic perturbation. In the later context, it is also a powerful means for functional gene annotation. A prerequisite for non-targeted metabolome analyses are methods for comprehensive metabolite extraction. We present three extraction protocols for a highly efficient extraction of metabolites from plant material with a very broad metabolite coverage. The presented metabolite fingerprinting workflow is based on liquid chromatography high resolution accurate mass spectrometry (LC-HRAM-MS), which provides suitable separation of the complex sample matrix for the analysis of compounds of different polarity by positive and negative electrospray ionization and mass spectrometry. The resulting data sets are then analyzed with the software suite MarVis and the web-based interface MetaboAnalyst. MarVis offers a straightforward workflow for statistical analysis, data merging as well as visualization of multivariate data, while MetaboAnalyst is used in our hands as complementary software for statistics, correlation networks and figure generation. Finally, MarVis provides access to species-specific metabolite and pathway data bases like KEGG and BioCyc and to custom data bases tailored by the user to connect the identified markers or features with metabolites. In addition, identified marker candidates can be interactively visualized and inspected in metabolic pathway maps by KEGG pathways for a more detailed functional annotation and confirmed by mass spectrometry fragmentation experiments or coelution with authentic standards. Together this workflow is a valuable toolbox to identify novel metabolites, metabolic steps or regulatory principles and pathways.

Keywords: Custom databases; Data mining; Extraction; Metabolite fingerprinting; Metabolome; Non-targeted metabolomics.