Global profiling combined with predicted metabolites screening for discovery of natural compounds: Characterization of ginsenosides in the leaves of Panax notoginseng as a case study.

Liquid chromatography-mass spectrometry (LC-MS) guided isolation is a favored strategy to quickly and efficiently explore the chemical diversity of herbal medicines. In this study, two methods were adopted to improve the performance of the strategy, including offline two-dimensional (2D) LC to extend the peak capacities and predicted metabolites screening (PMS) to automatically screen the targets with expanded databases. Ginsenosides in the leaves of Panax notoginseng (PNL) were taken as a case. An offline 2D LC system was constructed with an orthogonality of 0.69 and peak capacity of 8925. Quadrupole time-of-flight mass spectrometry-fast data directed analysis (QTOF-Fast DDA) was employed for detection of the ginsenosides in the fractioned samples. Four modified groups, including glucose, xylose, rhamnose and malonyl, were adopted and markedly extended the screening coverage. The combined strategy showed about 7.5 times improvement in the screening capability. PMS is conveniently and automatically implemented in UNIFI. Using this strategy, 945 ginsenosides were discovered from PNL, including 662 potentially novel ginsenosides. Furthermore, two new ginsenosides were purified, and unambiguously identified by NMR analysis, partially demonstrating the LC-MS guided isolation. The combined strategy can also be applied in characterizing and discovering new bioactive constituents from other herbal medicines.