Progress, applications, and challenges in high-throughput effect-directed analysis for toxicity driver identification - is it time for HT-EDA?

The rapid increase in the production and global use of chemicals and their mixtures has raised concerns about their potential impact on human and environmental health. With advances in analytical techniques, in particular, high-resolution mass spectrometry (HRMS), thousands of compounds and transformation products with potential adverse effects can now be detected in environmental samples. However, identifying and prioritizing the toxicity drivers among these compounds remain a significant challenge. Effect-directed analysis (EDA) emerged as an important tool to address this challenge, combining biotesting, sample fractionation, and chemical analysis to unravel toxicity drivers in complex mixtures. Traditional EDA workflows are labor-intensive and time-consuming, hindering large-scale applications. The concept of high-throughput (HT) EDA has recently gained traction as a means of accelerating these workflows. Key features of HT-EDA include the combination of microfractionation and downscaled bioassays, automation of sample preparation and biotesting, and efficient data processing workflows supported by novel computational tools. In addition to microplate-based fractionation, high-performance thin-layer chromatography (HPTLC) offers an interesting alternative to HPLC in HT-EDA. This review provides an updated perspective on the state-of-the-art in HT-EDA, and novel methods/tools that can be incorporated into HT-EDA workflows. It also discusses recent studies on HT-EDA, HT bioassays, and computational prioritization tools, along with considerations regarding HPTLC. By identifying current gaps in HT-EDA and proposing new approaches to overcome them, this review aims to bring HT-EDA a step closer to monitoring applications.

Identification of antimicrobial and glucocorticoid compounds in wastewater effluents with effect-directed analysis.

Pharmaceuticals, such as glucocorticoids and antibiotics, are inadequately removed from wastewater and may cause unwanted toxic effects in the receiving environment. This study aimed to identify contaminants of emerging concern in wastewater effluent with antimicrobial or glucocorticoid activity by applying effect-directed analysis (EDA). Effluent samples from six wastewater treatment plants (WWTPs) in the Netherlands were collected and analyzed with unfractionated and fractionated bioassay testing. Per sample, 80 fractions were collected and in parallel high-resolution mass spectrometry (HRMS) data were recorded for suspect and nontarget screening. The antimicrobial activity of the effluents was determined with an antibiotics assay and ranged from 298 to 711 ng azithromycin equivalents·L-1. Macrolide antibiotics were identified in each effluent and found to significantly contribute to the antimicrobial activity of each sample. Agonistic glucocorticoid activity determined with the GR-CALUX assay ranged from 98.1 to 286 ng dexamethasone equivalents·L-1. Bioassay testing of several tentatively identified compounds to confirm their activity revealed inactivity in the assay or the incorrect identification of a feature. Effluent concentrations of glucocorticoid active compounds were estimated from the fractionated GR-CALUX bioassay response. Subsequently, the biological and chemical detection limits were compared and a sensitivity gap between the two monitoring approaches was identified. Overall, these results emphasize that combining sensitive effect-based testing with chemical analysis can more accurately reflect environmental exposure and risk than chemical analysis alone.

Identifying antimicrobials and their metabolites in wastewater and surface water with effect-directed analysis.

This study aimed to identify antimicrobial contaminants in the aquatic environment with effect-directed analysis. Wastewater influent, effluent, and surface water (up- and downstream of the discharge location) were sampled at two study sites. The samples were enriched, subjected to high-resolution fractionation, and the resulting 80 fractions were tested in an antibiotics bioassay. The resulting bioactive fractions guided the suspect and nontargeted identification strategy in the high-resolution mass spectrometry data that was recorded in parallel. Chemical features were annotated with reference databases, assessed on annotation quality, and assigned identification confidence levels. To identify antibiotic metabolites, Phase I metabolites were predicted in silico for over 500 antibiotics and included as a suspect list. Predicted retention times and fragmentation patterns reduced the number of annotations to consider for confirmation testing. Overall, the bioactivity of three fractions could be explained by the identified antibiotics (clarithromycin and azithromycin) and an antibiotic metabolite (14-OH(R) clarithromycin), explaining 78% of the bioactivity measured at one study site. The applied identification strategy successfully identified antibiotic metabolites in the aquatic environment, emphasizing the need to include the toxic effects of bioactive metabolites in environmental risk assessments.

High-Performance Data Processing Workflow Incorporating Effect-Directed Analysis for Feature Prioritization in Suspect and Nontarget Screening.

Effect-directed analysis (EDA) aims at the detection of bioactive chemicals of emerging concern (CECs) by combining toxicity testing and high-resolution mass spectrometry (HRMS). However, consolidation of toxicological and chemical analysis techniques to identify bioactive CECs remains challenging and laborious. In this study, we incorporate state-of-the-art identification approaches in EDA and propose a robust workflow for the high-throughput screening of CECs in environmental and human samples. Three different sample types were extracted and chemically analyzed using a single high-performance liquid chromatography HRMS method. Chemical features were annotated by suspect screening with several reference databases. Annotation quality was assessed using an automated scoring system. In parallel, the extracts were fractionated into 80 micro-fractions each covering a couple of seconds from the chromatogram run and tested for bioactivity in two bioassays. The EDA workflow prioritized and identified chemical features related to bioactive fractions with varying levels of confidence. Confidence levels were improved with the in silico software tools MetFrag and the retention time indices platform. The toxicological and chemical data quality was comparable between the use of single and multiple technical replicates. The proposed workflow incorporating EDA for feature prioritization in suspect and nontarget screening paves the way for the routine identification of CECs in a high-throughput manner.

Characterisation of (anti-)progestogenic and (anti-)androgenic activities in surface and wastewater using high resolution effectdirected analysis.

The quality of surface waters is threatened by pollution with low concentrations of bioactive chemicals, among which those interfering with steroid hormone systems. Induced by reports of anti-progestogenic activity in surface waters, a two-year four-weekly survey of (anti-)progestogenic activity was performed at three surface water locations in the Netherlands that serve as abstraction points for the production of drinking water. As certain endogenous and synthetic progestogenic compounds are also potent (anti-)androgens, these activities were also investigated. Anti-progestogenic and anti-androgenic activities were detected in the majority of the monitoring samples, sometimes in concentrations exceeding effect-based trigger values, indicating the need for further research. To characterize the compounds responsible for the activities, a high resolution Effect-Directed Analysis (hr-EDA) panel was combined with PR and AR CALUX bioassays, performed in agonistic and antagonistic modes. The influent and effluent of a domestic wastewater treatment plant (WWTP) were included as effluent is a possible emission source of active compounds. As drivers for androgenic and progestogenic activities several native and synthetic steroid hormones were identified in the WWTP samples, namely androstenedione, testosterone, DHT, levonorgestrel and cyproterone acetate. The pesticides metolachlor and cyazofamid were identified as contributors to both the anti-progestogenic and anti-androgenic activities in surface water. In addition, epiconazole contributed to the anti-progestogenic activities in the rivers Rhine and Enclosed Meuse. This study showed the strength of hr-EDA for the identification of bioactive compounds in environmental samples and shed light on the drivers of (anti-)progestogenic and (anti-)androgenic activities in the aquatic environment.

Development of a framework to derive effect-based trigger values to interpret CALUX data for drinking water quality.

Bioassays are increasingly being implemented for water quality monitoring as targeted chemical analyses are not always sufficient for the detection of all emerging chemicals or transformation products. However, the interpretation of bioassay results remains challenging, in particular because a positive response does not necessarily indicate that there may be an increased risk. For this purpose, effect-based trigger (EBT) values have been introduced as thresholds above which action needs to be undertaken to determine the cause of the response. The goals of this study were to (i) evaluate various approaches used to determine EBT values and (ii) based on the findings, derive human health EBT values for Chemical Activated LUciferase gene eXpression (CALUX) in vitro bioassays used for routine monitoring of water quality in the Netherlands. Finally, (iii) an uncertainty analysis was carried out to determine the protective power of the derived EBT values and the chance that potentially harmful substances might not be detected. EBT values that can be implemented in routine monitoring could be determined for four of eight selected bioassays. These EBT were compared to bioassay results from routine water quality monitoring carried out in the Netherlands. Furthermore, a framework for the calculation and evaluation of derived EBT values for routine application to monitor drinking water and its sources is proposed.

High resolution effect-directed analysis of steroid hormone (ant)agonists in surface and wastewater quality monitoring.

Monitoring of chemical water quality is extremely challenging due to the large variety of compounds and the presence of biologically active compounds with unknown chemical identity. Previously, we developed a high resolution Effect-Directed Analysis (EDA) platform that combines liquid chromatography with high resolution mass spectrometry and parallel bioassay detection. In this study, the platform is combined with CALUX bioassays for (anti)androgenic, estrogenic and glucocorticoid activities, and the performance of the platform is evaluated. It appeared to render very repeatable results, with high recoveries of spiked compounds and high consistency between the mass spectrometric and bioassay results. Application of the platform to wastewater treatment plant effluent and surface water samples led to the identification of several compounds contributing to the measured activities. Eventually, a workflow is proposed for the application of the platform in a routine monitoring context. The workflow divides the platform into four phases, of which one to all can be performed depending on the research question and the results obtained. This allows one to make a balance between the effort put into the platform and the certainty and depth by which active compounds will be identified. The EDA platform is a valuable tool to identify unknown bioactive compounds, both in an academic setting as in the context of legislative, governmental or routine monitoring.

Development of a high-throughput bioassay for screening of antibiotics in aquatic environmental samples.

The goal of the present study was to select a Gram-positive (Gram+) and Gram-negative (Gram-) strain to measure antimicrobial activity in environmental samples, allowing high-throughput environmental screening. The sensitivity of eight pre-selected bacterial strains were tested to a training set of ten antibiotics, i.e. three Gram+ Bacillus subtilis strains with different read-outs, and five Gram- strains. The latter group consisted of a bioluminescent Allivibrio fischeri strain and four Escherichia coli strains, i.e. a wild type (WT) and three strains with a modified cell envelope to increase their sensitivity. The WT B. subtilis and an E. coli strain newly developed in this study, were most sensitive to the training set. This E. coli strain carries an open variant of an outer membrane protein combined with an inactivated multidrug efflux transport system. The assay conditions of these two strains were optimized and validated by exposure to a validation set of thirteen antibiotics with clinical and environmental relevance. The assay sensitivity ranged from the ng/mL to µg/mL range. The applicability of the assays for toxicological characterization of aquatic environmental samples was demonstrated for hospital effluent extract. A future application includes effect-directed analysis to identify yet unknown antibiotic contaminants or their transformation products.

Identification of mutagenic and endocrine disrupting compounds in surface water and wastewater treatment plant effluents using high-resolution effect-directed analysis.

Effect-directed analysis (EDA) has shown its added value for the detection and identification of compounds with varying toxicological properties in water quality research. However, for routine toxicity assessment of multiple toxicological endpoints, current EDA is considered labor intensive and time consuming. To achieve faster EDA and identification, a high-throughput (HT) EDA platform, coupling a downscaled luminescent Ames and cell-based reporter gene assays with a high-resolution fraction collector and UPLC-QTOF MS, was developed. The applicability of the HT-EDA platform in the analysis of aquatic samples was demonstrated by analysis of extracts from WWTP influent, effluent and surface water. Downscaled assays allowed detection of mutagenicity and androgen, estrogen and glucocorticoid agonism following high-resolution fractionation in 228 fractions. From 8 masses tentatively identified through non-target analysis, 2 masses were further investigated and chemically and biologically confirmed as the mutagen 1,2,3-benzotriazole and the androgen androstenedione. The compatibility of the high-throughput EDA platform with analysis of water samples and the incorporation of mutagenic and endocrine disruption endpoints allow for future application in routine monitoring in drinking water quality control and improved identification of (emerging) mutagens and endocrine disruptors.

Statistical analysis of a large set of semi-quantitative GC-MS screening data to evaluate and prioritize organic contaminants in surface and drinking water of the Netherlands.

Due to anthropogenic activities in the catchments, surface waters are contaminated with a large variety of chemical compounds. Drinking water companies in the Netherlands use surface water from the rivers Rhine, and Meuse, Lake IJssel and water from a reclaimed land area as sources for the production of drinking water. Samples from the abstraction points and the produced drinking waters were investigated using chemical screening with gas chromatography coupled to mass spectrometry to detect an as wide as possible range of organic contaminants, generating enormous data sets. This study aimed to evaluate and interpret five and a half years of screening data to get insight in the variety of known and new less polar compounds in surface and drinking waters, and to investigate if there were spatial patterns in the detection of compounds. Compounds from a wide variety of applications were detected. The vast majority of detected compounds was found only in a few samples. Certain compounds, however, e.g. organophosphate flame retardants, were detected with prevalences up to 100% per location. Most compounds were detected in samples from the rivers Rhine and Meuse, less in those from Lake IJssel and the reclaimed land area and only few in drinking water. Principal component and Hierarchical Cluster Analyses helped to detect patterns in the presence of contaminants on particular locations and to prioritize compounds for further investigation of their emission sources, and -in case of unknown compounds - their identification.

Compound Identification Using Liquid Chromatography and High-Resolution Noncontact Fraction Collection with a Solenoid Valve.

We describe the development of a high-resolution, noncontact fraction collector for liquid chromatography (LC) separations, allowing high-resolution fractionation in high-density well plates. The device is based on a low-dead-volume solenoid valve operated at 1-30 Hz for accurate collection of fractions of equal volume. The solenoid valve was implemented in a modified autosampler resulting in the so-called FractioMate fractionator. The influence of the solenoid supply voltage on solvent release was determined and the effect of the frequency, flow rate, and mobile phase composition was studied. For this purpose, droplet release was visually assessed for a wide range of frequencies and flow rates, followed by quantitative evaluation of a selection of promising settings for highly accurate, repeatable, and stable fraction collection. The potential of the new fraction collector for LC-based bioactivity screening was demonstrated by fractionating the LC eluent of a mixture of estrogenic and androgenic compounds, and a surface water sample (blank and spiked with bioactives) combining mass spectrometric detection and two reporter gene assays for bioactivity detection of the fractions. Additionally, a mixture of two compounds was repeatedly LC separated and fractionated to assess the feasibility of the system for analyte isolation followed by nuclear magnetic resonance analysis.

Steroid hormonal bioactivities, culprit natural and synthetic hormones and other emerging contaminants in waste water measured using bioassays and UPLC-tQ-MS.

Emission of compounds with biological activities from waste water treatment plant (WWTP) effluents into surface waters is a topic of concern for ecology and drinking water quality. We investigated the occurrence of hormone-like activities in waste water sample extracts from four Dutch WWTPs and pursued to identify compounds responsible for them. To this aim, in vitro reporter gene bioassays for androgenic, anti-androgenic, estrogenic, glucocorticoid and progestogenic activity and a UPLC-tQ-MS target analysis method for 25 steroid hormones used in high volumes in pharmacy were applied. Principal component analysis of the data was performed to further characterize the detected activities and compounds. All five types of activities tested were observed in the WWTP samples. Androgenic and estrogenic activities were almost completely removed during WW treatment, anti-androgenic activity was only found in treated WW. Glucocorticoid and progestogenic activities persisted throughout the treatment. The androgenic activity in both influent could predominantly be attributed to the presence of androstenedione and testosterone. Anti-androgenic activity was explained by the presence of cyproterone acetate. The glucocorticoid activity in influent was fully explained by prednicarbate, triamcinolone acetonide, dexamethasone and amcinonide. In effluent however, detected hormones could only explain 10-32% of the activity, indicating the presence of unknown glucocorticoids or their metabolites in effluent. Progesterone and levonorgestrel could explain the observed progestogenic activity. The principle component analysis confirmed the way in which hormones fit in the spectrum of other emerging contaminants concerning occurrence and fate in WWTPs.

High-Throughput Effect-Directed Analysis Using Downscaled in Vitro Reporter Gene Assays To Identify Endocrine Disruptors in Surface Water.

Effect-directed analysis (EDA) is a commonly used approach for effect-based identification of endocrine disruptive chemicals in complex (environmental) mixtures. However, for routine toxicity assessment of, for example, water samples, current EDA approaches are considered time-consuming and laborious. We achieved faster EDA and identification by downscaling of sensitive cell-based hormone reporter gene assays and increasing fractionation resolution to allow testing of smaller fractions with reduced complexity. The high-resolution EDA approach is demonstrated by analysis of four environmental passive sampler extracts. Downscaling of the assays to a 384-well format allowed analysis of 64 fractions in triplicate (or 192 fractions without technical replicates) without affecting sensitivity compared to the standard 96-well format. Through a parallel exposure method, agonistic and antagonistic androgen and estrogen receptor activity could be measured in a single experiment following a single fractionation. From 16 selected candidate compounds, identified through nontargeted analysis, 13 could be confirmed chemically and 10 were found to be biologically active, of which the most potent nonsteroidal estrogens were identified as oxybenzone and piperine. The increased fractionation resolution and the higher throughput that downscaling provides allow for future application in routine high-resolution screening of large numbers of samples in order to accelerate identification of (emerging) endocrine disruptors.

Development of a luminescent mutagenicity test for high-throughput screening of aquatic samples.

The Salmonella reversion based Ames test is the most widely used method for mutagenicity testing. For rapid toxicity assessment of e.g. water samples and for effect-directed analysis, however, the Ames test suffers from lack of throughput and is regarded as a laborious, time consuming method. To achieve faster analysis, with increased throughput, a (downscaled) luminescent derivative of the Ames Salmonella/microsome fluctuation test has been developed through expression of the Photorhabdus luminescens luciferase in the Salmonella TA98 and TA100 strains. The applicability of this test is demonstrated by analysis of environmentally relevant compounds, a suspended particulate matter extract and an industrial effluent sample. Use of the luminescent reporter reduced the required detection time from 48 to 28h with a specificity of 84% for responses reported in the literature to a set of 14 mutagens as compared to 72% in the unmodified fluctuation test. Testing of the same compounds in a downscaled luminescent format resulted in an 88% similarity with the response found in the regular luminescent format. The increase in throughput, faster analysis and potential for real-time bacterial quantification that luminescence provides, allows future application in the high-throughput screening of large numbers of samples or sample fractions, as required in effect-directed analysis in order to accelerate the identification of (novel) mutagens.

Improved androgen specificity of AR-EcoScreen by CRISPR based glucocorticoid receptor knockout.

The AR-EcoScreen is a widely used reporter assay for the detection of androgens and anti-androgens. Endogenous expression of glucocorticoid receptors and their affinity for the androgen responsive element that drives reporter expression, however, makes the reporter cells sensitive to interference by glucocorticoids and less specific for (anti-)androgens. To create a glucocorticoid insensitive derivative of the AR-EcoScreen, CRISPR/Cas9 genome editing was used to develop glucocorticoid receptor knockout mutants by targeting various sites in the glucocorticoid gene. Two mutant cell lines were further characterized and validated against the unmodified AR-EcoScreen with a set of 19 environmentally relevant chemicals and a series of environmental passive sampler extracts with (anti-)androgenic activity. Sequencing of the targeted sites revealed premature stop codons following frame-shift mutations, leading to an absence of functional glucocorticoid receptor expression. The introduced mutations rendered cell lines insensitive to glucocorticoid activation and caused no significant difference in the responsiveness towards (anti-)androgens, compared to the unmodified AR-EcoScreen cells, allowing the selective, GR-independent, determination of (anti-)androgenicity in environmental passive sampler extracts. The increase in selectivity for (anti-)androgens improves reliability of the AR-EcoScreen and will provide higher accuracy in determining (anti-)androgenic potential when applied in toxicity screening and environmental monitoring of both single compounds and mixtures.

Rapid activity-directed screening of estrogens by parallel coupling of liquid chromatography with a functional gene reporter assay and mass spectrometry.

In this study we developed a new LC nanofractionation platform that combines a human cell (BG1.Luc) gene reporter assay with a high resolution mass spectrometer for the detection and identification of estrogenic and anti-estrogenic compounds in environmental waters. The selection of this assay was based on its high sensitivity and selectivity, which is required for environmental trace level detection. We modified an autosampler and controlled it with in-house developed software to collect fractions in the low second range in microtiter plates. This ensured that chromatographic separation was maintained and allowed straightforward hyphenation with the bioassay. After bioassay testing, bioassay chromatograms were reconstructed and directly correlated with MS chromatograms that were obtained in parallel. This enabled to pinpoint bioactives in the MS chromatogram within a single fractionation cycle and results in a significant increase in throughput compared to traditional EDA studies. The sensitivity of the platform was low enough for environmental waters (80nM for bisphenol A and 320pM and 3.2nM for estradiol and estriol, respectively). In addition, the ability of the platform to detect anti-estrogens was successfully demonstrated as well. Finally, real samples were analysed.

Human health risk assessment of the mixture of pharmaceuticals in Dutch drinking water and its sources based on frequent monitoring data.

The presence of pharmaceuticals in drinking water is a topic of concern. Previous risk assessments indicate that their low concentrations are very unlikely to pose risks to human health, however often conclusions had to be based on small datasets and mixture effects were not included. The objectives of this study were to a) investigate if pharmaceuticals in surface and polder water penetrate in drinking water, b) assess the lifelong exposure of consumers to pharmaceuticals via drinking water and c) assess the possible individual and mixture health risks associated with this exposure. To fulfill these aims, a 2-year set of 4-weekly monitoring data of pharmaceuticals was used from three drinking water production plants. The 42 pharmaceuticals that were monitored were selected according to their consumption volume, earlier detection, toxicity and representation of the most relevant therapeutic classes. Lifelong exposures were calculated from concentrations and compared with therapeutic doses. Health risks were assessed by benchmarking concentrations with provisional guideline values. Combined risks of mixtures of pharmaceuticals were estimated using the concept of Concentration Addition. The lifelong exposure to pharmaceuticals via drinking water was calculated to be extremely low, i.e. a few mg, in total corresponding to <10% of the dose a patient is administered on one day. The risk of adverse health effects appeared to be negligibly low. Application of Concentration Addition confirmed this for the mixture of pharmaceuticals simultaneously present. The investigated treatment plants appeared to reduce the (already negligible) risk up to 80%. The large available monitoring dataset enabled the performance of a realistic risk assessment. It showed that working with maximum instead of average concentrations may overestimate the risk considerably.

A multicomponent snapshot of pharmaceuticals and pesticides in the river Meuse basin.

The river Meuse serves as a drinking-water source for more than 6 million people in France, Belgium, and The Netherlands. Pharmaceuticals and pesticides, both designed to be biologically active, are important classes of contaminants present in this river. The variation in the presence of pharmaceuticals in time and space in the Dutch part of the Meuse was studied using a multicomponent analytical method for pharmaceuticals combined with univariate and multivariate statistical analyses of the results. Trends and variation in time in the presence of pharmaceuticals were investigated in a dead-end side stream of the Meuse that serves as an intake point for the production of drinking water, and 93% of the selected compounds were detected. Highest concentrations were found for the antidiabetic metformin. Furthermore, a spatial snapshot of the presence of pharmaceuticals and pesticides was made along the river Meuse. Principal component analysis was successfully applied to reveal that wastewater-treatment plant effluent and water composition at the Belgian border were the main factors determining which compounds are found at different locations. The Dutch part of the river basin appeared responsible for approximately one-half of the loads of pharmaceuticals and pesticides discharged by the Meuse into the North Sea. The present study showed that multicomponent monitoring in combination with principal component analysis is a powerful tool to provide insight into contamination patterns in surface waters.

Effect-directed analysis of municipal landfill soil reveals novel developmental toxicants in the zebrafish Danio rerio.

Effect-directed analysis (EDA) is an approach used to identify (unknown) contaminants in complex samples which cause toxicity, using a combination of biology and chemistry. The goal of this work was to apply EDA to identify developmental toxicants in soil samples collected from a former municipal landfill site. Soil samples were extracted, fractionated, and tested for developmental effects with an embryotoxicity assay in the zebrafish Danio rerio. Gas chromatograph mass selective detection (GC-MSD) chemical screening was used to reveal candidate developmental toxicants in fractions showing effects. In a parallel study, liquid chromatography-hybrid linear ion trap Orbitrap mass spectrometry was also applied to one polar subfraction (Hoogenboom et al. J. Chromatogr. A2009, 1216, 510-519). EDA resulted in the identification of a number of previously unknown developmental toxicants, which were confirmed to be present in soil by GC-MS. These included 11H-benzo[b]fluorene, 9-methylacridine, 4-azapyrene, and 2-phenylquinoline, as well as one known developmental toxicant (retene). This work revealed the presence of novel contaminants in the environment that may affect vertebrate development, which are not subject to monitoring or regulation under current soil quality assessment guidelines.