Prioritizing Unknown Transformation Products from Biologically-Treated Wastewater Using High-Resolution Mass Spectrometry, Multivariate Statistics, and Metabolic Logic

Abstract

Incomplete micropollutant elimination in wastewater treatment plants (WWTPs) results in transformation products (TPs) that are released into the environment. Improvements in analytical technologies have allowed researchers to identify several TPs from specific micropollutants but an overall picture of nontarget TPs is missing. In this study, we addressed this challenge by applying multivariate statistics to data collected with liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS) and subsequent tandem HRMS (MS/MS) in order to characterize peaks detected in the influent and effluent of a WWTP. Known biotransformation reactions were used to link potential parent compounds and TPs, while the structural similarity of these pairs hypothesized by MS/MS similarity was used for further prioritization. The methodology was validated with a set of spiked compounds, which included 25 parent/TP pairs for which analytical standards were available. This procedure was then applied to nontarget data, and 20 potential parent and TP pairs were selected for identification. In summary, primarily a surfactant homologue series, with associated TPs, was detected. Some obstacles still remain, including spectral interferences from coeluting compounds and identification of TPs, whose structures are less likely to be present in compound databases. The workflow was developed using openly accessible tools and, after parameter adjustment, could be applied to any data set with before and after information about various biological or chemical processes.

Nontarget Screening with High Resolution Mass Spectrometry in the Environment: Ready to Go?

“…29 For a full scale plant, overall component characteristics such as retention time and mass changes were indicative of elimination processes of organic compounds during activated sludge treatment, without full identification. 30 The combination of controlled laboratory experiments with real world monitoring often facilitates prioritization by adding new information. 31,32 For example, in lab experiments Kolkmann et al 31 traced mutagenic nitrogenous disinfection byproducts (N-DBPs) formed with reactive N species during UV drinking water treatment by adding 15 N-labeled nitrate.…”

“…Experiment-driven Frequency, signal intensity of masses 14,33 Persistence 34 , elimination/formation 29 over process Component with characteristic isotope pattern (C, Cl, Br, N, O, S) 13,19,35 Reaction-based search of transformation products to link masses before and after treatment 30 Part of homologue series (mass difference, Kendrick mass defect) 13,36,37 Biological 28 , electrochemical 38 , oxidative 32 transformation product formation Suspect screening (looking for "known" or predicted chemicals without standard) 39,40 Reaction with isotopically-labelled reagents 31 Specific functional groups (MS/MS, derivatisation, neutral loss) 41,42 Effect-directed selection of masses in toxic fractions 43,44 Temporal or spatial profile over several samples 24,33 A second limitation is accounting for potential toxicity during prioritization. Approaches such as Effect-Directed Analysis use biological effect tests to prioritize chromatographic fractions with unknown components associated with specific toxic effects for identification.…”

Facilitating high resolution mass spectrometry data processing for screening of environmental water samples: An evaluation of two deconvolution tools

“…Most manufacturers have software specific for their instrument and data, which can automatically extract analytes of interest from the raw data, to facilitate suspect screening approaches. However, despite the tremendous advances in software for metabolite/transformation product detection and further non-target work, sometimes not all required information is available in one platform, leading users to manufacturer-independent software, such as the Eawag open-source R-code packages enviMass, enviPick, nontarget and RMassBank (Schollée et al, 2015;Schymanski et al, 2014a) which can enable the incorporation of additional parameters, such as the steps outlined above. In spite of these problems, non-target screening is necessary to identify new or unknown relevant pollutants, which is why efforts need to be made in developing proper software and efficient identification tools.…”

Advanced analytical tools for the investigation of licit and illicit drug residues in water

“…Most manufacturers have software specific for their instrument and data, which can automatically extract analytes of interest from the raw data, to facilitate suspect screening approaches. However, despite the tremendous advances in software for metabolite/transformation product detection and further non-target work, sometimes not all required information is available in one platform, leading users to manufacturer-independent software, such as the Eawag open-source R-code packages enviMass, enviPick, nontarget and RMassBank (Schollée et al, 2015;Schymanski et al, 2014a) which can enable the incorporation of additional parameters, such as the steps outlined above. In spite of these problems, non-target screening is necessary to identify new or unknown relevant pollutants, which is why efforts need to be made in developing proper software and efficient identification tools.…”

“…Alternatively, to gain more specific information on the compounds present multiple SPE cartridges could be used , or, if, in advance, a screening was made on a more specific set of compounds, different sample treatment can be applied -such as adding activated carbon for more polar compounds (Schollée et al, 2015).…”