Introduction: The importance of quality evaluation and control of crude drugs using metabolomics is increasingly recognized as a way of providing herbal medicines with predictably reliable quality and efficacy. In crude drugs, most metabolomics studies are conducted using a single analytical method. However, each analytical method has advantages and disadvantages that can influence the analytical coverage of the metabolome. In this study, we used both NMR and LC-MS analyses for a more thorough analysis of the components of Citrus-type crude drugs. Furthermore, we compared NMR analysis with LC-MS analysis for metabolic profiling.
Methods: Five Citrus-type crude drugs (39 samples); Kijitsu, Touhi, Chimpi, Kippi and Seihi, were used. The methanolic extracts of samples were analyzed using 1H- and 13C-NMR and High Resolution LC-ESI-MS. The NMR data were processed using the Alice2 for metabolome (JEOL). LC-MS data were processed using Progenesis (Waters). The resulting data sets were then imported into SIMCA ver. 14.0 (Umetrics) for further multivariate statistical analysis.
Results: The PCA score plots indicated the classification of the Citrus-type crude drugs into the same four groups in both NMR and LC-MS analyses. In the loading plots, differences were found in components contributing to the discrimination of the groups. The metabolites that were identified by the NMR and the LC-MS were three flavonoids such as naringin as contributors. The metabolites that were identified by the NMR only were primary metabolites; α- and β-glucose and sucrose, limonene and synephrine. However, the metabolites that were identified by the LC-MS only were three flavonoids such as hesperidin. The high dynamic range of the NMR provided broad coverage of the metabolome in a single experiment. LC-MS may be superior to detect secondary metabolites with high sensitivity. The combined NMR- and LC-MS-based metabolomics may provide useful information for the quality evaluation and control of crude drugs.