Feature-based molecular networking for identification of organic micropollutants including metabolites by non-target analysis applied to riverbank filtration

Daniela Oberleitner,Wolfgang Schulz,Christine Achten,Axel Bergmann,Robin Schmid

doi:10.1007/s00216-021-03500-7

Daniela Oberleitner, Wolfgang Schulz + Show 3 more

Open Access

https://doi.org/10.1007/s00216-021-03500-7

Copy DOI

Abstract

Due to growing concern about organic micropollutants and their transformation products (TP) in surface and drinking water, reliable identification of unknowns is required. Here, we demonstrate how non-target liquid chromatography (LC)-high-resolution tandem mass spectrometry (MS/MS) and the feature-based molecular networking (FBMN) workflow provide insight into water samples from four riverbank filtration sites with different redox conditions. First, FBMN prioritized and connected drinking water relevant and seasonally dependent compounds based on a modification-aware MS/MS cosine similarity. Within the resulting molecular networks, forty-three compounds were annotated. Here, carbamazepine, sartans, and their respective TP were investigated exemplarily. With chromatographic information and spectral similarity, four additional TP (dealkylated valsartan, dealkylated irbesartan, two oxygenated irbesartan isomers) and olmesartan were identified and partly verified with an authentic standard. In this study, sartans and TP were investigated and grouped regarding their removal behavior under different redox conditions and seasons for the first time. Antihypertensives were grouped into compounds being well removed during riverbank filtration, those primarily removed under anoxic conditions, and rather persistent compounds. Observed seasonal variations were mainly limited to varying river water concentrations. FBMN is a powerful tool for identifying previously unknown or unexpected compounds and their TP in water samples by non-target analysis.Graphical abstract

Highlights

IntroductionDue to rising concern about the increasing number of chemical compounds present in our surface waters, the identification of organic micropollutants (OMP) and their transformation ratio (m/z), retention time, and intensity) offers the possibility to focus on a predefined compound group (e.g., formation during a process [4]) and enables a pre-selection of features of interest
Due to rising concern about the increasing number of chemical compounds present in our surface waters, the identification of organic micropollutants (OMP) and their transformation ratio (m/z), retention time, and intensity) offers the possibility to focus on a predefined compound group and enables a pre-selection of features of interest
molecular networking (MN) creates networks of nodes, i.e., fragmentation spectra (MS/mass spectrometry (MS)), which are connected based on a pairwise spectral alignment and similarity scoring of all experimental MS/MS spectra in a study, which are often acquired in data-dependent fragmentation [8]

Summary

Introduction

Due to rising concern about the increasing number of chemical compounds present in our surface waters, the identification of organic micropollutants (OMP) and their transformation ratio (m/z), retention time, and intensity) offers the possibility to focus on a predefined compound group (e.g., formation during a process [4]) and enables a pre-selection of features of interest. MN creates networks of nodes, i.e., fragmentation spectra (MS/MS), which are connected based on a pairwise spectral alignment and similarity scoring of all experimental MS/MS spectra in a study, which are often acquired in data-dependent fragmentation [8]. Nodes are annotated by matching against the public GNPS spectral libraries. This workflow was applied to different non-target MS studies in various fields, including ocean water [9], agricultural [10], forensic [11], and biomedical research [12]. FBMN has so far mainly been used in natural product research and has not been extended towards surface and drinking water TP identification which is shown in this study. It has been used to identify TP of anthropogenic source in ocean water [20]

Objectives

Methods

Results