Efficient discovery of eight unreported compounds in Tall Gastrodia Tuber (Tianma) by stratified-precursor-list-based...

Discovering previously unreported chemical compounds can significantly expand the utilization of plants. Although molecular network (MN) accelerates the discovery/prediction of unreported compounds, the current MN methods have some shortcomings in predicting unreported compounds. A new method named as stratified-precursor-list (SPL)-based MN analysis was therefore developed to efficiently predict unreported compounds in plant materials in this study. In this method, LC-MS/MS data are collected based on SPLs containing a representative precursor ion for each compound. The retention time range for the MS² data acquisition of each precursor does not overlap with that of other co-eluted compounds resulting in high data quality and more reliable MN analysis. The developed MN is more intuitive because each node represents a different compound. Moreover, unreported compounds are predicted based on the node connection, molecular weight and MS² fragmentation patterns, which comprehensively utilizes the information in the MN and avoids the excessive reliance on MS/MS database. Tall Gastrodia Tuber (tuber of Gastrodia elata Blume, Tianma), a widely used herb and healthy food, was chosen as an example to discover unreported compounds. By integrating the developed prediction method and LC-MS-guided targeted isolation, eight unreported phenolic compounds were accurately predicted, targetedly isolated and identified including five phenolic glucoside lactates (gastrodilactate A-E) and three phenolic glucosides (gastrotetrabenzyloside A-C). Phenolic glucoside lactates are reported for the first time in this herb. It is demonstrated that the developed method can efficiently predict unreported compounds in plants.