Fast Evaluation of Herbal Substance Class Composition by Relative Mass Defect Plots

• Published in: Analytical chemistry (Washington) 2020-10, Vol.92 (19), p.12909-12916
• Main Authors: Waldner, Birgit J, Machalett, Ramona, Schönbichler, Stefan, Dittmer, Martin, Rubner, Moritz M, Intelmann, Daniel
• Format: Article
• Language: English
• Subjects: Algorithms; Alkaloids - analysis; Alkylation; Amino Acids - analysis; Analytical chemistry; Chemistry; Clustering; Composition; Glycosylation; Herbal Medicine; High resolution; Mapping; Mass Spectrometry; Mathematical analysis; Metabolites; Oligosaccharides - analysis; Phenols - analysis; Polarity; Retention time; Spectrometry; Terpenes - analysis; Visualization
• ISSN: 0003-2700; 1520-6882
• DOI: 10.1021/acs.analchem.0c01447

A holistic, nontargeted mass spectrometric analysis of any herbal material and preparation is intimately connected to fast chemical profiling and visualization of secondary plant metabolite classes or single compounds. High-resolution mass spectral data enable a broad variety of analytical possibilities. Often a fast and comprehensive overview on compound classes (phytochemical profiling) is needed before single-substance considerations. We present a fast approach for the initial characterization and substance class profiling using relative mass defect plots for the visualization of herbal compositions. From a dataset of 1160 common plant metabolites that represent a varied mixture of molecular classes in polarity, glycosylation, and alkylation, manually annotated for substance classes, the relative mass defects were calculated using theoretical molecular masses. For the calculation of the relative mass defect, a new approach incorporating two correction functions to obtain correct relative mass defect results also for large hydrocarbons, and a multitude of polyhalogenated molecules was developed. Using the Khachyan algorithm, elliptical areas clustering substance classes within the relative mass defect plots were calculated. The resulting novel relative mass defect plots provide a quick way of two-dimensional substance class mapping directly from high-resolution mass spectral data and may be considered as a unique fingerprint for herbals, part of them or herbal preparations. We show that adding the retention time as a third dimension improves the resolution power of the two-dimensional relative mass defect plot and offers the possibility for a more detailed substance class mapping.