Automatic MS/MS Data Mining Strategy for Discovering Target Natural Products: A Case of Lindenane Sesquiterpenoids.

Untargeted liquid chromatography-tandem mass spectrometry (LC-MS/MS) is a widely used method for discovering natural products (NPs); however, automatic MS/MS data mining for the discovery of NPs remains a challenge. In this work, LindenaneExtractor, a program based on characteristic MS/MS ions of lindenane sesquiterpenoids (LSs) was developed to automatically extract the LSs features for target LS discovery in plant extracts. To build this program, fragmentation mechanisms of characteristic ions of LSs were elucidated and confirmed by quantum chemical calculation and deuterium-labeled compounds. Subsequently, the information of characteristic ions was integrated and coded to develop LindenaneExtractor, which was further examined by standards and several public databases. Finally, the target LS features in extract were automatically extracted by LindenaneExtractor and visualized by feature-based molecular networking and two-dimensional (2D) retention time-/ plot, leading to the discovery of 96 target LSs in total, 37 of these compounds were potentially new NPs and one was confirmed by further isolation. This work proposed a new strategy for target NP analysis and discovery based on automatic MS/MS data mining, which could significantly improve the efficiency and accuracy of NP discovery.