Metabolomics-based profiling of five Salvia L. (Lamiaceae) species using untargeted data analysis workflow.

ABSTRACT

INTRODUCTION: The genus Salvia L., a member of the family Lamiaceae, is a keystone genus with a wide range of medicinal properties. It possesses a rich metabolite source that has long been used to treat different disorders. OBJECTIVES: Due to a deficiency of untargeted metabolomic profiling in the genus Salvia, this work attempts to investigate a comprehensive mass spectral library matching, computational data annotations, exclusive biomarkers, specific chemotypes, intraspecific metabolite profile variation, and metabolite enrichment by a case study of five medicinal species of Salvia. MATERIAL AND METHODS: Aerial parts of each species were subjected to QTRAP liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis workflow based on untargeted metabolites. A comprehensive and multivariate analysis was acquired on the metabolite dataset utilizing MetaboAnalyst 6.0 and the Global Natural Products Social Molecular Networking (GNPS) Web Platform. RESULTS: The untargeted approach empowered the identification of 117 metabolites by library matching and 92 nodes annotated by automated matching. A machine learning algorithm as substructural topic modeling, MS2LDA, was further implemented to explore the metabolite substructures, resulting in four Mass2Motifs. The automated library newly discovered a total of 23 metabolites. In addition, 87 verified biomarkers of library matching, 58 biomarkers of GNPS annotations, and 11 specific chemotypes were screened. CONCLUSION: Integrative spectral library matching and automated annotation by the GNPS platform provide comprehensive metabolite profiling through a workflow. In addition, QTRAP LC-MS/MS with multivariate analysis unveiled reliable information about inter and intraspecific levels of differentiation. The rigorous investigation of metabolite profiling presents a large-scale overview and new insights for chemotaxonomy and pharmaceutical studies.