PlantMAT: A Metabolomics Tool for Predicting the Specialized Metabolic Potential of a System and for Large-Scale Metabolite Identifications

Abstract

Custom software entitled Plant Metabolite Annotation Toolbox (PlantMAT) has been developed to address the number one grand challenge in metabolomics, which is the large-scale and confident identification of metabolites. PlantMAT uses informed phytochemical knowledge for the prediction of plant natural products such as saponins and glycosylated flavonoids through combinatorial enumeration of aglycone, glycosyl, and acyl subunits. Many of the predicted structures have yet to be characterized and are absent from traditional chemical databases, but have a higher probability of being present in planta. PlantMAT allows users to operate an automated and streamlined workflow for metabolite annotation from a user-friendly interface within Microsoft Excel, a familiar, easily accessed program for chemists and biologists. The usefulness of PlantMAT is exemplified using ultrahigh-performance liquid chromatography-electrospray ionization quadrupole time-of-flight tandem mass spectrometry (UHPLC-ESI-QTOF-MS/MS) metabolite profiling data of saponins and glycosylated flavonoids from the model legume Medicago truncatula. The results demonstrate PlantMAT substantially increases the chemical/metabolic space of traditional chemical databases. Ten of the PlantMAT-predicted identifications were validated and confirmed through the isolation of the compounds using ultrahigh-performance liquid chromatography-mass spectrometry-solid-phase extraction (UHPLC-MS-SPE) followed by de novo structural elucidation using 1D/2D nuclear magnetic resonance (NMR). It is further demonstrated that PlantMAT enables the dereplication of previously identified metabolites and is also a powerful tool for the discovery of structurally novel metabolites.

Comprehensive mass spectrometry-guided plant specialized metabolite phenotyping reveals metabolic diversity in the cosmopolitan plant family Rhamnaceae

“…Considering that most species used in this study have not been engaged in any phytochemical studies, we expect that our method will accelerate chemical identification of uncharted plant metabolite space. There have been other approaches for accelerating plant metabolite identification, such as candidate substrate-product pair (CSPP) network (Morreel et al , 2014), ISDB-molecular networking (Allard et al , 2016), MatchWeiz (Shahaf et al , 2016), or PlantMAT (Qiu et al , 2016). However, all these approaches not only rely on compound database content like our approach but also previous knowledge such as reported phytochemical composition or metabolic pathways.…”

UHPLC‐MS Analyses of Plant Flavonoids

“…UHPLC-MS has been successfully adopted Table. for flavonoid analyses in many plant species including a number of economically important crops (Cho, Howard, Prior, & Clark, 2004;Farag et al, 2007;Li et al, 2016;Wang et al, 2017). The utility of UHPLC-MS in plant flavonoid analyses is further enhanced by the development of mass spectral libraries as well as informatics tools (Akimoto et al, 2017;Qiu, Fine, Wherritt, Lei, & Sumner, 2016). To date, over 7000 flavonoids have been identified from various plant species and a database of probable mass fragments for 6867 known flavonoids (FsDatabase) was manually constructed (Akimoto et al, 2017).…”

From chromatogram to analyte to metabolite. How to pick horses for courses from the massive web resources for mass spectral plant metabolomics

“…Metabolite Identification Package (MI-Pack) [131], implemented in python, calculates differences in mass between all molecular formulas annotated from HRMS and compares them to known substrate/product pairs from KEGG, but matches are considered based on the error between experimental and theoretical masses compared to a threshold defined by a calculated mass error surface. Plant Metabolite Annotation Toolbox (PlantMAT) [132] is a particularly interesting tool designed specifically for the investigation of plant specialized metabolism, which uses an approach based on common metabolic building blocks to predict combinatorial possibilities of phytochemical structures used for annotation and as such is a highly effective way to search the chemical space surrounding a (set of) metabolite(s).…”