Background: It is well known that ginsenosides are the main active ingredients in ginseng, and they have also been important indexes for assessing the quality of ginseng. However, the absolute contents of ginsenosides in ginseng were shown to be varied with the origin, cultivated type, cultivated year and climate. It is a great challenge to distinguish the commercial types of ginsengs according to the content of one or several ginsenosides.
Methods: The common commercial types of ginsengs are white ginseng (WG), red ginseng (RG), American ginseng (AG). To clearly illustrate the differences among WG, RG and AG at the ginsenosides level, we established a strategy for the detection and identification of ginsenosides based on an optimized LC-Q-Orbitrap MS/MS method coupled with an in-house database of ginsenosides. Before and after the normalization, the ginsenosides datasheet was analyzed and compared using several state-of-the-art multivariate statistical analysis methods.
Results: Here, 81 ginsenosides were identified in different ginseng samples. The majority of the ginsenosides (59 in 81) were all shared by WG, RG and AG. When the shared ginsenosides datasheet was normalized by the level of ginsenoside Ro, our analysis strategy clearly divided the ginseng samples into three groups (i.e., WG, RG and AG groups). We found that the ginsenoside profiles in RG and WG were significantly different from those in AG. The potential markers and multivariate diagnostic models differentiating the three types of ginsengs were also indicated.
Conclusion: Our novel methodology based on ginsenoside profiles is more robust than existing methods, and data normalization is required to improve the efficiency of multivariate statistical analysis.
Keywords: Data normalization; Ginsenosides profiles; High-resolution mass spectrometry; Multivariate analysis; Panax ginseng.
© 2022 The Author(s).