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

The vast, diverse universe of organic pollutants is a formidable challenge for environmental sciences, engineering, and regulation. Nontarget screening (NTS) based on high resolution mass spectrometry (HRMS) has enormous potential to help characterize this universe, but is it ready to go for real world applications? In this Feature article we argue that development of mass spectrometers with increasingly high resolution and novel couplings to both liquid and gas chromatography, combined with the integration of high performance computing, have significantly widened our analytical window and have enabled increasingly sophisticated data processing strategies, indicating a bright future for NTS. NTS has great potential for treatment assessment and pollutant prioritization within regulatory applications, as highlighted here by the case of real-time pollutant monitoring on the River Rhine. We discuss challenges for the future, including the transition from research toward solution-centered and robust, harmonized applications.

Rapid Screening for Exposure to "Non-Target" Pharmaceuticals from Wastewater Effluents by Combining HRMS-Based Suspect Screening and Exposure Modeling.

Active pharmaceutical ingredients (APIs) have raised considerable concern over the past decade due to their widespread detection in water resources and their potential to affect ecosystem health. This triggered many attempts to prioritize the large number of known APIs to target monitoring efforts and testing of fate and effects. However, so far, a comprehensive approach to screen for their presence in surface waters has been missing. Here, we explore a combination of an automated suspect screening approach based on liquid chromatography coupled to high-resolution mass spectrometry and a model-based prioritization using consumption data, readily predictable fate properties and a generic mass balance model for activated sludge treatment to comprehensively detect APIs with relevant exposure in wastewater treatment plant effluents. The procedure afforded the detection of 27 APIs that had not been covered in our previous target method, which included 119 parent APIs. The newly detected APIs included seven compounds with a high potential for bioaccumulation and persistence, and also three compounds that were suspected to stem from point sources rather than from consumption as medicines. Analytical suspect screening proved to be more selective than model-based prioritization, making it the method of choice for focusing analytical method development or fate and effect testing on those APIs most relevant to the aquatic environment. However, we found that state-of-the-practice exposure modeling used to predict potential high-exposure substances can be a useful complement to point toward oversights and known or suspected detection gaps in the analytical method, most of which were related to insufficient ionization.

Non-target screening with high-resolution mass spectrometry: critical review using a collaborative trial on water analysis.

In this article, a dataset from a collaborative non-target screening trial organised by the NORMAN Association is used to review the state-of-the-art and discuss future perspectives of non-target screening using high-resolution mass spectrometry in water analysis. A total of 18 institutes from 12 European countries analysed an extract of the same water sample collected from the River Danube with either one or both of liquid and gas chromatography coupled with mass spectrometry detection. This article focuses mainly on the use of high resolution screening techniques with target, suspect, and non-target workflows to identify substances in environmental samples. Specific examples are given to emphasise major challenges including isobaric and co-eluting substances, dependence on target and suspect lists, formula assignment, the use of retention information, and the confidence of identification. Approaches and methods applicable to unit resolution data are also discussed. Although most substances were identified using high resolution data with target and suspect-screening approaches, some participants proposed tentative non-target identifications. This comprehensive dataset revealed that non-target analytical techniques are already substantially harmonised between the participants, but the data processing remains time-consuming. Although the objective of a "fully-automated identification workflow" remains elusive in the short term, important steps in this direction have been taken, exemplified by the growing popularity of suspect screening approaches. Major recommendations to improve non-target screening include better integration and connection of desired features into software packages, the exchange of target and suspect lists, and the contribution of more spectra from standard substances into (openly accessible) databases. Graphical Abstract Matrix of identification approach versus identification confidence.

Strategies to characterize polar organic contamination in wastewater: exploring the capability of high resolution mass spectrometry.

Wastewater effluents contain a multitude of organic contaminants and transformation products, which cannot be captured by target analysis alone. High accuracy, high resolution mass spectrometric data were explored with novel untargeted data processing approaches (enviMass, nontarget, and RMassBank) to complement an extensive target analysis in initial "all in one" measurements. On average 1.2% of the detected peaks from 10 Swiss wastewater treatment plant samples were assigned to target compounds, with 376 reference standards available. Corrosion inhibitors, artificial sweeteners, and pharmaceuticals exhibited the highest concentrations. After blank and noise subtraction, 70% of the peaks remained and were grouped into components; 20% of these components had adduct and/or isotope information available. An intensity-based prioritization revealed that only 4 targets were among the top 30 most intense peaks (negative mode), while 15 of these peaks contained sulfur. Of the 26 nontarget peaks, 7 were tentatively identified via suspect screening for sulfur-containing surfactants and one peak was identified and confirmed as 1,3-benzothiazole-2-sulfonate, an oxidation product of a vulcanization accelerator. High accuracy, high resolution data combined with tailor-made nontarget processing methods (all available online) provided vital information for the identification of a wider range of heteroatom-containing compounds in the environment.

Suspect and nontarget screening approaches to identify organic contaminant records in lake sediments.

Sediment cores provide a valuable record of historical contamination, but so far, new analytical techniques such as high-resolution mass spectrometry (HRMS) have not yet been applied to extend target screening to the detection of unknown contaminants for this complex matrix. Here, a combination of target, suspect, and nontarget screening using liquid chromatography (LC)-HRMS/MS was performed on extracts from sediment cores obtained from Lake Greifensee and Lake Lugano located in the north and south of Switzerland, respectively. A suspect list was compiled from consumption data and refined using the expected method coverage and a combination of automated and manual filters on the resulting measured data. Nontarget identification efforts were focused on masses with Cl and Br isotope information available that exhibited mass defects outside the sample matrix, to reduce the effect of analytical interferences. In silico methods combining the software MOLGEN-MS/MS and MetFrag were used for direct elucidation, with additional consideration of retention time/partitioning information and the number of references for a given substance. The combination of all available information resulted in the successful identification of three suspect (chlorophene, flufenamic acid, lufenuron) and two nontarget compounds (hexachlorophene, flucofuron), confirmed with reference standards, as well as the tentative identification of two chlorophene congeners (dichlorophene, bromochlorophene) that exhibited similar time trends through the sediment cores. This study demonstrates that complementary application of target, suspect, and nontarget screening can deliver valuable information despite the matrix complexity and provide records of historical contamination in two Swiss lakes with previously unreported compounds.

Exploring the Behaviour of Emerging Contaminants in the Water Cycle using the Capabilities of High Resolution Mass Spectrometry.

To characterize a broad range of organic contaminants and their transformation products (TPs) as well as their loads, input pathways and fate in the water cycle, the Department of Environmental Chemistry (Uchem) at Eawag applies and develops high-performance liquid chromatography (LC) methods combined with high-resolution tandem mass spectrometry (HRMS/MS). In this article, the background and state-of-the-art of LC-HRMS/MS for detection of i) known targets, ii) suspected compounds like TPs, and iii) unknown emerging compounds are introduced briefly. Examples for each approach are taken from recent research projects conducted within the department. These include the detection of trace organic contaminants and their TPs in wastewater, pesticides and their TPs in surface water, identification of new TPs in laboratory degradation studies and ozonation experiments and finally the screening for unknown compounds in the catchment of the river Rhine.

Paradise lost? Pesticide pollution in a European region with considerable amount of traditional agriculture.

Pesticide contamination of agricultural streams has widely been analysed in regions of high intensity agriculture such as in Western Europe or North America. The situation of streams subject to low intensity agriculture relying on human and animal labour, as in parts of Romania, remains unknown. To close this gap, we determined concentrations of 244 pesticides and metabolites at 19 low-order streams, covering sites from low to high intensity agriculture in a region of Romania. Pesticides were sampled with two passive sampling methods (styrene-divinylbenzene (SDB) disks and polydimethylsiloxane (PDMS) sheets) during three rainfall events and at base flow. Using the toxic unit approach, we assessed the toxicity towards algae and invertebrates. Up to 50 pesticides were detected simultaneously, resulting in sum concentrations between 0.02 and 37 µg L-1. Both, the sum concentration as well as the toxicities were in a similar range as in high intensity agricultural streams of Western Europe. Different proxies of agricultural intensity did not relate to in-stream pesticide toxicity, contradicting the assumption of previous studies. The toxicity towards invertebrates was positively related to large scale variables such as the catchment size and the agricultural land use in the upstream catchment and small scale variables including riparian plant height, whereas the toxicity to algae showed no relationship to any of the variables. Our results suggest that streams in low intensity agriculture, despite a minor reported use of agrochemicals, exhibit similar levels of pesticide pollution as in regions of high intensity agriculture.

Transformation of beta-lactam antibacterial agents during aqueous ozonation: reaction pathways and quantitative bioassay of biologically-active oxidation products.

Reactions of ozone (O3) with the beta-lactam antibiotics penicillin G (PG) and cephalexin (CP) have previously been found to yield products retaining antibacterial activities. These products are unequivocally identified here as the stereoisomeric (R)-sulfoxides of each parent molecule and characterized by a combination of chemical analysis and an antibacterial activity assay. PG-(R)-sulfoxide, which is approximately 15% as potent as PG itself, is formed in approximately 55% yield, whereas CP-(R)-sulfoxide, which is 83% as active as CP, is formed with a maximum 34% yield. PG-(R)-sulfoxide is recalcitrant toward further oxidation by O3, but readily transformed by hydroxyl radical (HO*) (kHO*,app"=7.4x10(9) M(-1) s(-1), pH 7), resulting in quantitative elimination of its antibacterial activity. In contrast, CP-(R)-sulfoxide is degraded by both O3 and HO* (kO3,app"=2.6x10(4) M(-1) s(-1) and kHO*,app"= 7.6x10(9) M(-1) s(-1), pH 7), leading to quantitative elimination of its antibacterial activity. During ozonation of a secondary municipal wastewater effluent sample (pH 8.1, CDOC=4.0 mg/L, [alkalinity]=3.6 mM as HCO3-) spiked with [PG]0=1 microM, PG-(R)-sulfoxide yields did not exceed 0.15 microM for O3 doses up to 100 microM (4.8 mg/L), but reached 0.47 microM with 10-mM t-BuOH added as a HO* scavenger. In contrast, CP-(R)-sulfoxide yields did not exceed 0.1 microM for the same wastewater spiked with [CP]0=1 microM in either the presence or absence of t-BuOH, indicating that CP-(R)-sulfoxide transformation is governed primarily by direct reaction with O3. These findings suggest that, for a given degree of parent compound transformation, PG-(R)-sulfoxide yields would likely be greatest during ozonation of wastewaters characterized by low O3 demands and high HO* scavenging rates, whereas CP-(R)-sulfoxide yields would be less matrix-dependent. In general, complete deactivation of penicillins during wastewater treatment will likely require higher O3 exposures than necessary for deactivation of cephalosporins.

Wide-scope target screening of >2000 emerging contaminants in wastewater samples with UPLC-Q-ToF-HRMS/MS and smart evaluation of its performance through the validation of 195 selected representative analytes.

This study presents the development and validation of a comprehensive quantitative target methodology for the analysis of 2316 emerging pollutants in water based on Ultra-Performance Liquid Chromatography Quadrupole-Time-Of-Flight Mass Spectrometry (UPLC-Q-ToF-HRMS/MS). Target compounds include pesticides, pharmaceuticals, drugs of abuse, industrial chemicals, doping compounds, surfactants and transformation products, among others. The method was validated for 195 analytes, chosen to be representative of the chemical space of the target list, enabling the assessment of the performance of the method. The method involves a generic sample preparation based on mixed mode solid phase extraction, a UPLC-QTOF-MS/MS screening method using Data Independent Acquisition (DIA) mode, which provides MS and MS/MS spectra simultaneously and an elaborate strong post-acquisition evaluation of the data. The processing method was optimized to provide a successful identification rate >95 % and to minimize the number of false positive results (< 5 %). Decision limit (CCα) and detection capability (CCß) were also introduced in the validation scheme to provide more realistic metrics on the performance of a HRMS-based wide-scope screening method. A new system of identification points (IPs) based on the one described in the Commission Decision 2002/657/EC was applied to communicate the confidence level in the identification of the analytes. This system considers retention time, mass accuracy, isotopic fit and fragmentation; taking full advantage of the capacities of the HRMS instruments. Finally, 398 contaminants were detected and quantified in real wastewater.

Wastewater treatment plant resistomes are shaped by bacterial composition, genetic exchange, and upregulated expression in the effluent microbiomes.

Wastewater treatment plants (WWTPs) are implicated as hotspots for the dissemination of antibacterial resistance into the environment. However, the in situ processes governing removal, persistence, and evolution of resistance genes during wastewater treatment remain poorly understood. Here, we used quantitative metagenomic and metatranscriptomic approaches to achieve a broad-spectrum view of the flow and expression of genes related to antibacterial resistance to over 20 classes of antibiotics, 65 biocides, and 22 metals. All compartments of 12 WWTPs share persistent resistance genes with detectable transcriptional activities that were comparatively higher in the secondary effluent, where mobility genes also show higher relative abundance and expression ratios. The richness and abundance of resistance genes vary greatly across metagenomes from different treatment compartments, and their relative and absolute abundances correlate with bacterial community composition and biomass concentration. No strong drivers of resistome composition could be identified among the chemical stressors analyzed, although the sub-inhibitory concentration (hundreds of ng/L) of macrolide antibiotics in wastewater correlates with macrolide and vancomycin resistance genes. Contig-based analysis shows considerable co-localization between resistance and mobility genes and implies a history of substantial horizontal resistance transfer involving human bacterial pathogens. Based on these findings, we propose future inclusion of mobility incidence (M%) and host pathogenicity of antibiotic resistance genes in their quantitative health risk ranking models with an ultimate goal to assess the biological significance of wastewater resistomes with regard to disease control in humans or domestic livestock.

Pesticides drive risk of micropollutants in wastewater-impacted streams during low flow conditions.

Micropollutants enter surface waters through various pathways, of which wastewater treatment plants (WWTPs) are a major source. The large diversity of micropollutants and their many modes of toxic action pose a challenge for assessing environmental risks. In this study, we investigated the potential impact of WWTPs on receiving ecosystems by describing concentration patterns of micropollutants, predicting acute risks for aquatic organisms and validating these results with macroinvertebrate biomonitoring data. Grab samples were taken upstream, downstream and at the effluent of 24 Swiss WWTPs during low flow conditions across independent catchments with different land uses. Using liquid chromatography high resolution tandem mass spectrometry, a comprehensive target screening of almost 400 organic substances, focusing mainly on pesticides and pharmaceuticals, was conducted at two time points, and complemented with the analysis of a priority mixture of 57 substances over eight time points. Acute toxic pressure was predicted using the risk assessment approach of the multi-substance potentially affected fraction, first applying concentration addition for substances with the same toxic mode of action and subsequently response addition for the calculation of the risk of the total mixture. This toxic pressure was compared to macroinvertebrate sensitivity to pesticides (SPEAR index) upstream and downstream of the WWTPs. The concentrations were, as expected, especially for pharmaceuticals and other household chemicals higher downstream than upstream, with the detection frequency of plant protection products upstream correlating with the fraction of arable land in the catchments. While the concentration sums downstream were clearly dominated by pharmaceuticals or other household chemicals, the acute toxic pressure was mainly driven by pesticides, often caused by the episodic occurrence of these compounds even during low flow conditions. In general, five single substances explained much of the total risk, with diclofenac, diazinon and clothianidin as the main drivers. Despite the low predicted acute risk of 0%-2.1% for affected species, a significant positive correlation with macroinvertebrate sensitivity to pesticides was observed. However, more effect data for pharmaceuticals and a better quantification of episodic pesticide pollution events are needed for a more comprehensive risk assessment.

Exhaustive extraction of sulfonamide antibiotics from aged agricultural soils using pressurized liquid extraction.

A pressurized liquid extraction (PLE) method was developed for the quantification of five sulfonamide antibiotics (sulfadiazine, sulfadimethoxine, sulfamethazine, sulfamethoxazole, and sulfathiazole) in aged soil samples. To account for sequestration effects the extraction was optimized using a composite grassland soil sample collected 11 days after the application of manure containing these substances. The optimized method uses a mixture of buffered water (pH 8.8) and acetonitrile (85:15) as solvent for the extraction at 200 degrees C and 100 bar during five min. The most important parameter for the extraction efficiency was temperature whereas the pH of the extraction solvent did hardly influence extraction efficiency between pH 4.1 and 8.8. A temperature increase from 100 to 200 degrees C improved the extraction efficiency up to a factor of six for aged residues in soils. In contrast, no temperature dependence was observed during short-term spike experiment. After 90 min exposure in these spike experiments we recovered 62-93% of the sulfonamides, except for sulfamethoxazole with only 41%. These percentages decreased substantially after a contact time of 6 and 17 days. The reasons for this decline remained unknown. Inter-day precision of the method was very satisfactory: relative standard deviations from the average were below 10%. Limits of detection for the extraction procedure were lower than 15 microg/kg. The performance of the developed extraction method was demonstrated by measuring the decrease of sulfonamide concentration in a top soil after manure application. Within 3 months the concentration of sulfadiazine dropped from 450 to 150 microg/kg.

Fully automated online solid phase extraction coupled directly to liquid chromatography-tandem mass spectrometry. Quantification of sulfonamide antibiotics, neutral and acidic pesticides at low concentrations in surface waters.

A fully automated online solid phase extraction-liquid chromatography-tandem mass spectrometry (SPE-LC-MS/MS) instrumental setup has been developed for the quantification of sulfonamide antibiotics and pesticides in natural water. The direct coupling of an online solid phase extraction cartridge (Oasis HLB) to LC-MS/MS was accomplished using column switching techniques. High sensitivity in the low ng/L range was achieved by large volume injections of 18 mL with a combination of a tri-directional auto-sampler and a dispenser system. This setup allowed high sample throughput with a minimum of investment costs. Special emphasis was placed on low cross contamination. The chosen approach is suitable for research as well as for monitoring applications. The flexible instrumental setup was successfully optimised for different important groups of bioactive chemicals resulting in three trace analytical methods for quantification of (i) sulfonamide antibiotics and their acetyl metabolites; (ii) neutral pesticides (triazines, phenylureas, amides, chloracetanilides) and (iii) acidic pesticides (phenoxyacetic acids and triketones). Absolute extraction recoveries from 85 to 112% were obtained for the different analytes. More than 500 samples could be analyzed with one extraction cartridge. The inter-day precision of the method was excellent indicated by relative standard deviations between 1 and 6%. High accuracy was achieved by the developed methods resulting in maximum deviation relative to the spiked amount of 8-15% for the different analytes. Detection limits for various environmental samples were between 0.5 and 5 ng/L. Matrix induced ion suppression was in general smaller than 25%. The performance of the online methods was demonstrated with measurements of concentration dynamics of sulfonamide antibiotics and pesticides concentrations in a little creek during rain fall events.

Quantitative target and systematic non-target analysis of polar organic micro-pollutants along the river Rhine using high-resolution mass-spectrometry--Identification of unknown sources and compounds.

In this study, the contamination by polar organic pollutants was investigated along the Rhine River, an important source of drinking water for 22 million people in central Europe. Following the flow of the river, a traveling water mass was sampled using weekly flow-proportional composite samples at ten different downstream sites, including main tributaries. Using a broad analytical method based on solid phase extraction and high-resolution mass spectrometry, the water was analyzed for more than 300 target substances. While the water in Lake Constance contained only 83 substances in often low concentrations, the number of detects found in the water phase increased to 143 substances and a weekly load of more than 7 tons at the last sampling site, the Dutch-German border. Mostly present were chemicals originating from wastewater treatment plants, especially the artificial sweetener Acesulfam and two pharmaceuticals, Metformin and Gabapentin, which dominate the weekly load up to 58%. Although the sample campaign was performed in a dry period in early spring, a large variety of pesticides and biocides were detected. Several industrial point sources were identified along the waterway's 900 km journey, resulting in high concentrations in the tributaries and loads of up to 160 kg. Additionally, an unbiased non-target analysis was performed following two different strategies for the prioritization of hundreds of potentially relevant unknown masses. While for the first prioritization strategy, only chlorinated compounds were extracted from the mass spectrometer datasets, the second prioritization strategy was performed using a systematic reduction approach between the different sampling sites. Among others, two substances that never had been detected before in this river, namely, the muscle relaxant Tizanidine and the solvent 1,3-Dimethyl-2-imidazolidinone (DMI), were identified and confirmed, and their loads were roughly estimated along the river.

Quantification of the new triketone herbicides, sulcotrione and mesotrione, and other important herbicides and metabolites, at the ng/l level in surface waters using liquid chromatography-tandem mass spectrometry.

The LC/ESI/MSMS method allows the trace quantification (ng/l) of the new triketone herbicides, i.e. sulcotrione and mesotrione, and important herbicides and metabolites, in natural waters. Solid phase extraction (SPE) for sample enrichment is performed with OASIS (recoveries 94-112% for parent herbicides). Neutral and acidic compounds were analyzed separately with ESI in positive and negative mode, respectively. Quantification limits varied between 0.5 and 10 ng/l. The acidic herbicides detection was improved by a neutralizing post-column addition solution. The influence of ion suppression on quantification is discussed in detail. It is shown that we could overcome this problem and achieve reliable quantification using isotope labeled internal standards (ILIS) for every single analyte. The methods performance is illustrated with samples from a lake depth profile.

Sources of pesticides in surface waters in Switzerland: pesticide load through waste water treatment plants--current situation and reduction potential.

Concentrations of pesticides in Swiss rivers and lakes frequently exceed the Swiss quality goal of 0.1 microg/l for surface waters. In this study, concentrations of various pesticides (e.g., atrazine, diuron, mecoprop) were continuously measured in the effluents of waste water treatment plants and in two rivers during a period of four months. These measurements revealed that in the catchment of Lake Greifensee, farmers who did not perfectly comply with 'good agricultural practice' caused at least 14% of the measured agricultural herbicide load into surface waters. Pesticides, used for additional purposes in urban areas (i.e. protection of materials, conservation, etc.), entered surface waters up to 75% through waste water treatment plants.

Automatic recalibration and processing of tandem mass spectra using formula annotation.

High accuracy, high resolution tandem mass spectrometry (MS/MS) is becoming more common in analytical applications, yet databases of these spectra remain limited. Databases require good quality spectra with sufficient compound information, but processing, calibration, noise reduction and retrieval of compound information are time-consuming tasks that prevent many contributions. We present a comprehensive workflow for the automatic processing of MS/MS using formula annotation for recalibration and cleanup to generate high quality spectra of standard compounds for upload to MassBank (www.massbank.jp). Compound information is retrieved via Internet services. Reference standards of 70 pesticides were measured at various collision energies on an LTQ-Orbitrap XL to develop and evaluate the workflow. A total of 944 resulting spectra are now available on MassBank. Evidence of nitrogen adduct formation during MS/MS fragmentation processes was found, highlighting the benefits high accuracy MS/MS offers for spectral interpretation. A database of recalibrated, cleaned-up spectra resulted in the most correct spectra ranked in first place, regardless of whether the search spectra were recalibrated or not, whereas the average rank of the correct molecular formula was improved from 2.55 (uncalibrated) to 1.53 when using recalibrated MS/MS data. The workflow is available as an R package RMassBank capable of generating MassBank records from raw MS and MS/MS data and can be adjusted to process data acquired with different settings and instruments. This workflow is a vital step towards addressing the need for more high quality, high accuracy MS/MS spectra in spectral databases and provides important information for spectral interpretation.