Informed Feature-Based Molecular Networking as a Complementary Approach to Identify Bioactive Phyllobilins in Senescent Plant Extracts

Phyllobilins are chlorophyll-derived metabolites that represent a distinct class of bioactive natural products. Although widespread in nature, they remain largely underexplored in phytochemical research. In this study, we present a workflow for the dereplication and identification of phyllobilins in complex plant extracts. The approach combines RP-MPLC fractionation, UHPLC-VWD-HRMS² analysis, antioxidant activity testing, and data analysis using feature-based molecular networking. Methanolic extracts of senescent <i>Echinacea purpurea</i> leaves, known to be rich in structurally diverse chlorophyll metabolites, were used to generate experimental MS² spectra for a set of phylloxanthobilins, which were subsequently uploaded to the GNPS community database. The workflow was then applied to senescent <i>Salvia officinalis</i> leaves, a pharmaceutically important plant with an unknown phyllobilin profile, leading to the annotation of several phylloxanthobilins. One metabolite featuring a malonic acid ester moiety was isolated and structurally confirmed using 1D- and 2D-NMR spectroscopy. Future applications of this workflow may enable the systematic annotation and prioritization of previously overlooked phyllobilins across diverse plant species, addressing a critical gap in the analytical profiling of bioactive plant extracts. By leveraging public MS² spectral libraries and network-based data analysis, it offers a scalable tool to address a long-standing blind spot of natural product research.