Metabolomics and computational tools to characterize the chemical space of Momordica | MSA Symposium
Welcome to the second MSA symposium 2021 hosted in collaboration with the ACTG and DIPLOMICS.
The Symposium was sponsored by LECO, BIOCRATES, MICROSEP, WATERS, SCIEX, SEPARATIONS AND SHIMADZU
Speaker: Ramabulana Anza
Authors: Anza T. Ramabulana, Ntakadzeni E. Madala, Daniel Petras, Fidele Tugizimana Email: [email protected]
Abstract
Illuminating the chemistries of medicinal plants holds promise as valuable sources of lead compounds for drug development. Mass spectrometry (MS)-based untargeted metabolomics offers unique opportunities for analysis and characterization of these plant chemical spaces. However, due to the inherent complexity of the plant metabolome, it remains challenging to annotate most of the chemical signatures detected by untargeted MS analyses. Furthermore, the integrated high-resolution MS-based data acquisition protocols, such as data-dependent acquisition (DDA), define the quality of the spectral data obtained from such analyses. Thus, in this study, MS-based untargeted metabolomics and computational tools, particularly molecular networking (MN) approaches, were applied to characterize the metabolomes of a medicinal plant, Momordica species. Mass spectral molecular networking, performed in the Global Natural Product Social Molecular Networking (GNPS) ecosystem, would enable a broad overview of molecular information that can be inferred from MS/MS data. The analysis of chemical relationships between every MS/MS spectrum, visualizing the entire metabolome detected in a sample, would reveal structurally related molecular families in Momordica plants. The key DDA parameters that impact the formation and quality of MNs were firstly explored. The influence of collision energy and intensity thresholds on generated MNs were demonstrated. Hereby it was indicated that the use of low DDA thresholds and a combination of multiple collision energies generated high quality MS spectra. Furthermore, classical MN, in-silico fragmentation tools (Network Annotation Propagation, NAP and DEREPLICATOR) and an unsupervised substructure identification tool (MS2LDA) were applied to annotate the detected metabolomes of four Momordica species (M. cardiospermoides, M. balsamina, M. charantia and M. foetida). These computational approaches allowed for the visualisation of chemical classes (e.g. saponins, flavonoids and fatty acids) and the variety of substructures within the molecular families in Momordica metabolomes. This also holds great potential in the discovery of novel metabolites that are related to knowns within the molecular families. Overall, the studied species are highlighted as phytochemically rich plants consisting of many biologically active metabolites differentially distributed within the different species, with the metabolome of M. cardiospermoides dereplicated here for the first time.