Rapid recognition and targeted isolation of potential anti-breast cancer xanthones in Hypericum bellum Li by "seed" mass spectra-based molecular networking and in silico MS/MS fragmentation.

Abstract

Instruction: Hypericum bellum Li is rich in xanthones with various bioactivities, especially in anti-breast cancer. While the scarcity of mass spectral data of xanthones in Global Natural Products Social Molecular Networking (GNPS) libraries have challenged the rapid recognition of xanthones with similar structures.

Objective: This study is aimed to enhance the molecular networking (MN)-based dereplication and visualisation ability of potential anti-breast cancer xanthones from H. bellum to overcome the scarcity of xanthones mass spectral data in GNPS libraries. Separating and purifying the MN-screening bioactive xanthones to verify the practicality and accuracy of this rapid recognition strategy.

Methodology: A combined strategy of "seed" mass spectra-based MN, in silico annotation tools, substructure identification tools, reverse molecular docking, ADMET screening, molecular dynamics (MDs) simulation experiments, and an MN-oriented separation procedure was first introduced to facilitate the rapid recognition and targeted isolation of potential anti-breast cancer xanthones in H. bellum.

Results: A total of 41 xanthones could only be tentatively identified. Among them, eight xanthones were screened to have potential anti-breast cancer activities, and six xanthones that were initially reported in H. bellum were obtained and verified to have good binding abilities with their paired targets.

Conclusion: This is a successful case study that validated the application of "seed" mass spectral data could overcome the drawbacks of GNPS libraries with limited mass spectra and enhance the accuracy and visualisation of natural products (NPs) dereplication, and this rapid recognition and targeted isolation strategy can be also applicable for other types of NPs.