Background: Abrus cantoniensis Hance. (Ac) and Abrus mollis (Am), two edible and medicinal plants with economic value in southern China, belong to the Abrus genus. Due to its growth characteristics, Am often replaces Ac in folk medicine. However, the latest National Pharmacopeia of China only recommends Ac. The differences in the metabolite composition of the plants are directly related to the differences in their clinical efficacy. Results: The difference in metabolites were analyzed using an untargeted metabolomic approach based on ultrahigh-performance liquid chromatography-electrospray ionization-tandem mass spectrometry (UPLC‒ESI‒MS/MS). The roots (R), stems (S) and leaves (L) of the two varieties were examined, and 635 metabolites belonging to 8 classes were detected. A comparative study revealed clear variations in the metabolic profiles of the two plants, and the AmR group had more active ingredients (flavonoids and terpenoids) than the AcR group. The metabolites classified as flavonoids and triterpene saponins showed considerable variations among the various samples. Both Ac and Am had unique metabolites. Two metabolites (isovitexin-2''-xyloside and soyasaponin V) specifically belong to Ac, and nine metabolites (vitexin-2"-O-galactoside, ethyl salicylate, 6-acetamidohexanoic acid, rhein-8-O-glucoside, hederagenin-3-O-glucuronide-28-O-glucosyl(1,2)-glucoside, methyl dioxindole-3-acetate, veratric acid, isorhamnetin-3-O-sophoroside-7-O-rhamnoside, and isorhamnetin-3-O-sophoroside) specifically belong to Am. Conclusions: The metabolite differences between Ac and Am cause the differences in their clinical efficacy. Our findings serve as a foundation for further investigation of biosynthesis pathways and associated bioactivities and provide guidance for the clinical application of traditional Chinese medicine.

Astragalus membranaceus (AM) is a kind of traditional Chinese herbal medicine that is extensively utilized in vascular cognitive impairment (VCI) treatment. Nevertheless, because of its complex ingredients, the exact molecular mechanism remains unclear. Therefore, the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP), STITCH, and SwissTargetPrediction have been utilized to gather the primary active ingredients of AM. The potential therapeutic targets of VCI were collected through GeneCards, OMIM, and DisGeNET databases. Then, the interactive network was drawn. The potential mechanism of AM in VCI treatment was demonstrated by means of enrichment analysis and network topology analysis of Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genome (KEGG) pathway. Subsequently, molecular docking predicted the binding properties of its potential active ingredients and targets. 20 effective compounds and 733 targets were screened from AM, among which 158 targets were identified as possible targets of AM in treating VCI. Protein-protein interaction (PPI) network and topology analysis revealed that MAPK3 and MMP9 were the key targets of AM intervention in VCI. Molecular docking studies indicated that calycosin and quercetin may be potential active compounds of AM in VCI treatment. Our study contributes to an in-depth understanding of the pharmacological actions of AM in VCI treatment, which also offers a scientific foundation for greater understanding of the molecular mechanism of AM in treating VCI.