A novel neutral loss/product ion scan-incorporated integral approach for the untargeted characterization and comparison of the carboxyl-free ginsenosides from Panax ginseng, Panax quinquefolius, and Panax notoginseng.

Differentiated composition in precursor ions for different subclasses of ginsenosides in the negative electrospray-ionization mode has been reported, which lays a foundation for the sorted and untargeted identification of ginsenosides. Carboxyl-free ginsenosides simultaneously from Panax ginseng, P. quinquefolius, and P. notoginseng, were comprehensively characterized and statistically compared. A neutral loss/product ion scan (NL-PIS) incorporated untargeted profiling approach, coupled to ultra-high performance liquid chromatography, was developed on a linear ion-trap/Orbitrap mass spectrometer for characterizing carboxyl-free ginsenosides. It incorporated in-source fragmentation (ISF) full scan-MS1, mass tag-MS2, and product ion scan-MS3. Sixty batches of ginseng samples were analyzed by metabolomics workflows for the discovery of ginsenoside markers. Using formic acid (FA) as the additive, carboxyl-free ginsenosides (protopanaxadiol-type, protopanaxatriol-type, and octillol-type) gave predominant FA-adducts, while rich deprotonated molecules were observed for carboxyl-containing ginsenosides (oleanolic acid-type and malonylated) when source-induced dissociation (SID) was set at 0 V. Based on the NL transition [M+FA‒H]- > [M-H]- and the characteristic sapogenin product ions, a NL-PIS approach was established. It took advantage of the efficient full-information acquisition of ISF-MS1 (SID: 50 V), the high specificity of mass tag (NL: 46.0055 Da)-induced MS2 fragmentation, and the substructure fragmentation of product ion scan-MS3. We could characterize 216 carboxyl-free ginsenosides, and 21 thereof were potentially diagnostic for the species differentiation. Conclusively, sorted and untargeted characterization of the carboxyl-free ginsenosides was achieved by the established NL-PIS approach. In contrast to the conventional NL or PIS-based survey scan strategies, the high-accuracy MSn data obtained can enable more reliable identification of ginsenosides.