Untargeted metabolite profiling: A comprehensive study using data analysis workflow in Salvia L. species (Lamiaceae)

Abstract

The genus Salvia L., a member of the Lamiaceae family, possesses remarkable medicinal properties for treating various disorders due to its rich natural product resources. Despite the genus's significance, there is a dearth of reports on complete metabolite characterization at intraspecific levels within Salvia. This study endeavors to bridge this gap by conducting a comprehensive investigation encompassing mass spectral library matching, biomarker assessment, chemotype specification, metabolite enrichment detection, and computational metabolite annotation. Dried aerial parts from three distinct Salvia species served as the primary material for untargeted quadrupole linear/ion trap (QTRAP) liquid chromatography (LC)- mass spectrometry (MS/MS)-based metabolite analysis. The metabolite datasets underwent rigorous processing, utilizing multivariate analysis methods through MetaboAnalyst 5.0 and the Global Natural Products Social Molecular Networking (GNPS) Web Platform. Our investigation identified 160 metabolites through library matching, and an additional 57 annotated nodes were obtained through automated spectral matching. Employing annotation tools such as DEREPLICATOR+ and substructure discovery tools like MS2LDA (a Bayesian Latent Dirichlet Allocation (LDA) model) led to the discovery of four metabolites and three Mass2Motifs. Furthermore, 48 metabolites were explored for the first time using GNPS libraries. The investigation unveiled specific chemotypes and identified 98 potential biomarkers. Using library and GNPS spectral matching proved to be high-throughput supplementary strategies for annotating and identifying various metabolites. Moreover, integrating QTRAP LC-MS/MS with a multivariate statistical method grounded in biomarker and functional analysis provided differential features at specific and intraspecific levels. This comprehensive overview of metabolites in Salvia species offers valuable insights for quality profiling.