The most reliable class of metabolite annotation is provided when chemical information of an observed metabolite matches known compounds in natural product databases or descriptions from the literature. Moco etal. (2006) reported an analysis of fruit extracts pooled from 96 cultivars representing three major fruit types (cherry, Dutch beef, normal round) using LC-Q-TOF-MS. They annotated metabolites using m/z, retention time, absorption spectra, and MS/MS data and performed database searches against the Dictionary of Natural Products (http://www.chemnetbase.com/ scripts/dnpweb.exe?welcome-main) and SciFinder (http://www.cas.org/SCIFINDER/). In addition, they compared MS data of the observed metabolites with the 60 tomato metabolites that had been reported in the literature. Consequently, they identified 26 metabolites based on literature reports, and 14 novel metabolites that were not previously reported in tomato fruit extracts. A large fraction of the identified metabolites belonged to the phenylpropanoid pathway. The authors created a tomato metabolite database MoToDB (http://appliedbioinformatics.wur.nl) based on MS data of metabolites that have been reported in literature (compiled data of 110 metabolites as of February 2007).
Iijima et al. (unpublished data) annotated metabolites from fruit of the cultivar Micro-Tom based on LC Fourier transform ion cyclotron resonance MS (LC-FTICR-MS) data. From peels and flesh of four maturation stages (immature green, mature green, orange, red ripe), approximately 700 peak groups that had both representative ion and isotopic ions in a series of consecutive scans were subjected to molecular formula prediction and database searches. Based on metabolite annotation, putative metabolic pathways between annotated metabolites were predicted. All MS data and annotations will be compiled into a database (Iijima et al. in prep).
Was this article helpful?