Planta Med 2013; 79 - PN6
DOI: 10.1055/s-0033-1348687

A New LC-MS-Based Strategy to Establish the Solanum Metabolic Database: Integrating Chemistry, Morphology, and Evolution

SB Wu 1, RS Meyer 2, 3, BD Whitaker 4, EJ Kennelly 1
  • 1Department of Biological Sciences, Lehman College, and The Graduate Center, The City University of New York, Bronx, NY, 10468
  • 2Department of Biology, New York University, New York, NY 10003
  • 3The New York Botanical Garden, Bronx, NY 10458
  • 4Food Quality Laboratory, USDA, Beltsville, MD 20705

The economically valuable plant genus Solanum, which contains dozens of functional food species such as eggplant and tomato, affords an excellent system to compare and correlate metabolic chemistry with species morphology and evolution. We have devised a strategy based on repeatable reversed-phase LC-TOF-MS methods and statistical tools, including untargeted PCA and targeted PLS/OPLS-DA models, to create a Solanum metabolic database (SMD). As part of establishing this new SMD, 31 accessions representing 24 species were analyzed, including eight species whose metabolic profiles were studied for the first time. Sixty-two Solanum metabolic compounds were identified after detailed analysis of UV spectra, mass spectral fragmentation patterns, NMR spectra, and/or co-injection experiments with authentic standards, which included two new 5-O-caffeoylquinic acid derivatives identified by analyzing their MS/MS fragmentation. Based on our SMD, a detailed biosynthetic pathway of the metabolites in Solanum was created. In addition, three statistical models (PCA, PLS and OPLS-DA) were used to classify the origin of eggplant species and find the differences between groups of phylogenetically-related species and species sharing the same use (food or medicine). As a result, seven marker metabolites were identified to distinguish four Solanum sections. This is the first metabolic study of the genus Solanum that provides a means to investigate the origin of compounds in the evolutionary process. This new strategy combining an LC-MS database with multivariate statistical analyses was proven effective in studying the metabolic network within a large genus; integrating complicated chemistry, morphology, and evolutionary problems.