Map and model-moving from observation to prediction in toxicogenomics

Advancing toxicity studies of per- and poly-fluoroalkyl substances (pfass) through machine learning: Models, mechanisms, and future directions

Perfluorinated and perfluoroalkyl substances (PFASs), encompassing a vast array of isomeric chemicals, are recognized as typical emerging contaminants with direct or potential impacts on human health and the ecological environment. With the complex and elusive toxicological profiles of PFASs, machine learning (ML) has been increasingly employed in their toxicity studies due to its proficiency in prediction and data analytics. This integration is poised to become a predominant trend in environmental toxicology, propelled by the swift advancements in computational technology. This review diligently examines the literature to encapsulate the varied objectives of employing ML in the toxicity studies of PFASs: (1) Utilizing ML to establish Quantitative Structure-Activity Relationship (QSAR) models for PFASs with diverse toxicity endpoints, facilitating the targeted toxicity prediction of unidentified PFASs; (2) Investigating and substantiating the Adverse Outcome Pathway (AOP) through the synergy of ML and traditional toxicological methods, with this refining the toxicity assessment framework for PFASs; (3) Dissecting and elucidating the features of established ML models to advance Open Research into the toxicity of PFASs, with a primary focus on determinants and mechanisms. The discourse extends to an in-depth examination of ML studies, segregating findings based on their distinct application trajectories. Given that ML represents a nascent paradigm within PFASs research, this review delineates the collective challenges encountered in the ML-mediated study of PFAS toxicity and proffers strategic guidance for ensuing investigations.

Investigation of Peroxisome Proliferator-Activated Receptor Genes as Requirements for Visual Startle Response Hyperactivity in Larval Zebrafish Exposed to Structurally Similar Per- and Polyfluoroalkyl Substances (PFAS)

BACKGROUND

Per- and polyfluoroalkyl Substances (PFAS) are synthetic chemicals widely detected in humans and the environment. Exposure to perfluorooctanesulfonic acid (PFOS) or perfluorohexanesulfonic acid (PFHxS) was previously shown to cause dark-phase hyperactivity in larval zebrafish.

OBJECTIVES

The objective of this study was to elucidate the mechanism by which PFOS or PFHxS exposure caused hyperactivity in larval zebrafish.

METHODS

Swimming behavior was assessed in 5-d postfertilization (dpf) larvae following developmental (1-4 dpf) or acute (5 dpf) exposure to 0.43 - 7.86 μ M PFOS, 7.87 - 120 μ M PFHxS, or 0.4% dimethyl sulfoxide (DMSO). After developmental exposure and chemical washout at 4 dpf, behavior was also assessed at 5-8 dpf. RNA sequencing was used to identify differences in global gene expression to perform transcriptomic benchmark concentration-response (BMC T ) modeling, and predict upstream regulators in PFOS- or PFHxS-exposed larvae. CRISPR/Cas9-based gene editing was used to knockdown peroxisome proliferator-activated receptors (ppars) pparaa/ab, pparda/db, or pparg at day 0. Knockdown crispants were exposed to 7.86 μ M PFOS or 0.4% DMSO from 1-4 dpf and behavior was assessed at 5 dpf. Coexposure with the ppard antagonist GSK3787 and PFOS was also performed.

RESULTS

Transient dark-phase hyperactivity occurred following developmental or acute exposure to PFOS or PFHxS, relative to the DMSO control. In contrast, visual startle response (VSR) hyperactivity only occurred following developmental exposure and was irreversible up to 8 dpf. Similar global transcriptomic profiles, BMC T estimates, and enriched functions were observed in PFOS- and PFHxS-exposed larvae, and ppars were identified as putative upstream regulators. Knockdown of pparda/db, but not pparaa/ab or pparg, blunted PFOS-dependent VSR hyperactivity to control levels. This finding was confirmed via antagonism of ppard in PFOS-exposed larvae.

DISCUSSION

This work identifies a novel adverse outcome pathway for VSR hyperactivity in larval zebrafish. We demonstrate that developmental, but not acute, exposure to PFOS triggered persistent VSR hyperactivity that required ppard function. https://doi.org/10.1289/EHP13667.

Toxicogenomic profiling after sublethal exposure to nerve- and muscle-targeting insecticides reveals cardiac and neuronal developmental effects in zebrafish embryos

For specific primary modes of action (MoA) in environmental non-target organisms, EU legislation restricts the usage of active substances of pesticides or biocides. Corresponding regulatory hazard assessments are costly, time consuming and require large numbers of non-human animal studies. Currently, predictive toxicology of development compounds relies on their chemical structure and provides little insights into toxicity mechanisms that precede adverse effects. Using the zebrafish embryo model, we characterized transcriptomic responses to a range of sublethal concentrations of six nerve- and muscle-targeting insecticides with different MoA (abamectin, carbaryl, chlorpyrifos, fipronil, imidacloprid & methoxychlor). Our aim was to identify affected biological processes and suitable biomarker candidates for MoA-specific signatures. Abamectin showed the most divergent signature among the tested insecticides, linked to lipid metabolic processes. Differentially expressed genes (DEGs) after imidacloprid exposure were primarily associated with immune system and inflammation. In total, 222 early responsive genes to either MoA were identified, many related to three major processes: (1) cardiac muscle cell development and functioning (tcap, desma, bag3, hspb1, hspb8, flnca, myoz3a, mybpc2b, actc2, tnnt2c), (2) oxygen transport and hypoxic stress (alas2, hbbe1.1, hbbe1.3, hbbe2, hbae3, igfbp1a, hif1al) and (3) neuronal development and plasticity (npas4a, egr1, btg2, ier2a, vgf). The thyroidal function related gene dio3b was upregulated by chlorpyrifos and downregulated by higher abamectin concentrations. Important regulatory genes for cardiac muscle (tcap) and forebrain development (npas4a) were the most frequently ifferentially expressed across all insecticide treatments. We consider the identified gene sets as useful early warning biomarker candidates, i.e. for developmental toxicity targeting heart and brain in aquatic vertebrates. Our findings provide a better understanding about early molecular events in response to the analyzed MoA. Perceptively, this promotes the development for sensitive and informative biomarker-based in vitro assays for toxicological MoA prediction and AOP refinement, without the suffering of adult fish.

Nextcast: a software suite to analyse and model toxicogenomics data

•

Toxicogenomics is emerging as a valid approach to characterise the mechanism of action of chemicals.

•

Structured pipelines for toxicogenomics increase standardisation and regulatory acceptance.

•

We developed the Nextcast software suite for robust analysis and modelling of toxicogenomic data.

•

Nextcast offers customisable modular pipelines to tackle multiple biological questions.

The recent advancements in toxicogenomics have led to the availability of large omics data sets, representing the starting point for studying the exposure mechanism of action and identifying candidate biomarkers for toxicity prediction. The current lack of standard methods in data generation and analysis hampers the full exploitation of toxicogenomics-based evidence in regulatory risk assessment. Moreover, the pipelines for the preprocessing and downstream analyses of toxicogenomic data sets can be quite challenging to implement. During the years, we have developed a number of software packages to address specific questions related to multiple steps of toxicogenomics data analysis and modelling. In this review we present the Nextcast software collection and discuss how its individual tools can be combined into efficient pipelines to answer specific biological questions. Nextcast components are of great support to the scientific community for analysing and interpreting large data sets for the toxicity evaluation of compounds in an unbiased, straightforward, and reliable manner. The Nextcast software suite is available at: ( https://github.com/fhaive/nextcast).

Gene co-expression network analysis in zebrafish reveals chemical class specific modules

BACKGROUND

Zebrafish is a popular animal model used for high-throughput screening of chemical hazards, however, investigations of transcriptomic mechanisms of toxicity are still needed. Here, our goal was to identify genes and biological pathways that Aryl Hydrocarbon Receptor 2 (AHR2) Activators and flame retardant chemicals (FRCs) alter in developing zebrafish. Taking advantage of a compendium of phenotypically-anchored RNA sequencing data collected from 48-h post fertilization (hpf) zebrafish, we inferred a co-expression network that grouped genes based on their transcriptional response.

RESULTS

Genes responding to the FRCs and AHR2 Activators localized to distinct regions of the network, with FRCs inducing a broader response related to neurobehavior. AHR2 Activators centered in one region related to chemical stress responses. We also discovered several highly co-expressed genes in this module, including cyp1a, and we subsequently show that these genes are definitively within the AHR2 signaling pathway. Systematic removal of the two chemical types from the data, and analysis of network changes identified neurogenesis associated with FRCs, and regulation of vascular development associated with both chemical classes. We also identified highly connected genes responding specifically to each class that are potential biomarkers of exposure.

CONCLUSIONS

Overall, we created the first zebrafish chemical-specific gene co-expression network illuminating how chemicals alter the transcriptome relative to each other. In addition to our conclusions regarding FRCs and AHR2 Activators, our network can be leveraged by other studies investigating chemical mechanisms of toxicity.

Transcriptome-Wide Prediction and Measurement of Combined Effects Induced by Chemical Mixture Exposure in Zebrafish Embryos

BACKGROUND

Humans and environmental organisms are constantly exposed to complex mixtures of chemicals. Extending our knowledge about the combined effects of chemicals is thus essential for assessing the potential consequences of these exposures. In this context, comprehensive molecular readouts as retrieved by omics techniques are advancing our understanding of the diversity of effects upon chemical exposure. This is especially true for effects induced by chemical concentrations that do not instantaneously lead to mortality, as is commonly the case for environmental exposures. However, omics profiles induced by chemical exposures have rarely been systematically considered in mixture contexts.

OBJECTIVES

In this study, we aimed to investigate the predictability of chemical mixture effects on the whole-transcriptome scale.

METHODS

We predicted and measured the toxicogenomic effects of a synthetic mixture on zebrafish embryos. The mixture contained the compounds diuron, diclofenac, and naproxen. To predict concentration- and time-resolved whole-transcriptome responses to the mixture exposure, we adopted the mixture concept of concentration addition. Predictions were based on the transcriptome profiles obtained for the individual mixture components in a previous study. Finally, concentration- and time-resolved mixture exposures and subsequent toxicogenomic measurements were performed and the results were compared with the predictions.

RESULTS

This comparison of the predictions with the observations showed that the concept of concentration addition provided reasonable estimates for the effects induced by the mixture exposure on the whole transcriptome. Although nonadditive effects were observed only occasionally, combined, that is, multicomponent-driven, effects were found for mixture components with anticipated similar, as well as dissimilar, modes of action.

DISCUSSION

Overall, this study demonstrates that using a concentration- and time-resolved approach, the occurrence and size of combined effects of chemicals may be predicted at the whole-transcriptome scale. This allows improving effect assessment of mixture exposures on the molecular scale that might not only be of relevance in terms of risk assessment but also for pharmacological applications. https://doi.org/10.1289/EHP7773.

Assessing the Ecological Risks of Per- and Polyfluoroalkyl Substances: Current State-of-the Science and a Proposed Path Forward

Per- and poly-fluoroalkyl substances (PFAS) encompass a large, heterogenous group of chemicals of potential concern to human health and the environment. Based on information for a few relatively well-understood PFAS such as perfluorooctane sulfonate and perfluorooctanoate, there is ample basis to suspect that at least a subset can be considered persistent, bioaccumulative, and/or toxic. However, data suitable for determining risks in either prospective or retrospective assessments are lacking for the majority of PFAS. In August 2019, the Society of Environmental Toxicology and Chemistry sponsored a workshop that focused on the state-of-the-science supporting risk assessment of PFAS. The present review summarizes discussions concerning the ecotoxicology and ecological risks of PFAS. First, we summarize currently available information relevant to problem formulation/prioritization, exposure, and hazard/effects of PFAS in the context of regulatory and ecological risk assessment activities from around the world. We then describe critical gaps and uncertainties relative to ecological risk assessments for PFAS and propose approaches to address these needs. Recommendations include the development of more comprehensive monitoring programs to support exposure assessment, an emphasis on research to support the formulation of predictive models for bioaccumulation, and the development of in silico, in vitro, and in vivo methods to efficiently assess biological effects for potentially sensitive species/endpoints. Addressing needs associated with assessing the ecological risk of PFAS will require cross-disciplinary approaches that employ both conventional and new methods in an integrated, resource-effective manner. Environ Toxicol Chem 2021;40:564-605. © 2020 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC. This article has been contributed to by US Government employees and their work is in the public domain in the USA.

Manually curated transcriptomics data collection for toxicogenomic assessment of engineered nanomaterials

Toxicogenomics (TGx) approaches are increasingly applied to gain insight into the possible toxicity mechanisms of engineered nanomaterials (ENMs). Omics data can be valuable to elucidate the mechanism of action of chemicals and to develop predictive models in toxicology. While vast amounts of transcriptomics data from ENM exposures have already been accumulated, a unified, easily accessible and reusable collection of transcriptomics data for ENMs is currently lacking. In an attempt to improve the FAIRness of already existing transcriptomics data for ENMs, we curated a collection of homogenized transcriptomics data from human, mouse and rat ENM exposures in vitro and in vivo including the physicochemical characteristics of the ENMs used in each study.

Aspirin Eugenol Ester Attenuates Paraquat-Induced Hepatotoxicity by Inhibiting Oxidative Stress

Aspirin eugenol ester (AEE) is a new potential drug with anti-inflammatory and antioxidant stress pharmacological activity. Paraquat (PQ) is an effective and commercially important herbicide that is widely used worldwide. However, paraquat is highly toxic and can cause various complications and acute organ damage, such as liver, kidney and lung damage. The purpose of this study was to investigate whether AEE has a protective effect on hepatotoxicity induced by PQ in vivo and in vitro. Cell viability, apoptosis rate, mitochondrial function and intracellular oxidative stress were detected to evaluate the protective effect of AEE on PQ-induced BRL-3A (normal rat hepatocytes) cytotoxicity in vitro. In vivo, AEE pretreatment could attenuate oxidative stress and histopathological changes in rat liver induced by PQ. The results showed that AEE could reduce the hepatotoxicity induced by PQ in vivo and in vitro. AEE reduced PQ-induced hepatotoxicity by inhibitingoxidative stress and maintaining mitochondrial function. This study proved that AEE is an effective antioxidant and can reduce the hepatotoxicity of PQ.

A multi-omics concentration-response framework uncovers novel understanding of triclosan effects in the chlorophyte Scenedesmus vacuolatus

In aquatic ecosystems, the biocide triclosan represents a hazard for the non-target microalgae. So far, algal responses were mainly investigated at apical levels hampering the acquisition of a holistic view on primary, adaptive, and compensatory stress responses. We assessed responses of the chlorophyte Scenedesmus vacuolatus to triclosan at apical (growth, photosynthesis) and molecular (transcriptome, metabolome) levels for comparative pathway sensitivity analysis. For each responsive signal (contigs, metabolites), a concentration-response curve was modeled and effect concentrations were calculated leading to the setting of cumulative sensitivity distributions. Molecular responses showed higher sensitivity than apical observations. The functional annotation of contigs and metabolites revealed 118 metabolic pathways putatively impaired by triclosan, highlighting a wide repercussion on the algal metabolism. Metabolites involved in the lipid metabolism showed decreasing trends along the concentration gradient and a globally highest sensitivity, pointing to the primary target of triclosan. The pathways involved in xenobiotic degradation and membrane transporters were mainly regulated in the transcriptome with increasing response trends comprising compensatory responses. The suggested novel approach, combining apical and multi-omics analyses in a concentration-response framework improves mechanistic understanding and mode of action analysis on non-targeted organisms and is suggested to better implement high-throughput multi-omics data in environmental risk assessment.

TinderMIX: Time-dose integrated modelling of toxicogenomics data

BACKGROUND

Omics technologies have been widely applied in toxicology studies to investigate the effects of different substances on exposed biological systems. A classical toxicogenomic study consists in testing the effects of a compound at different dose levels and different time points. The main challenge consists in identifying the gene alteration patterns that are correlated to doses and time points. The majority of existing methods for toxicogenomics data analysis allow the study of the molecular alteration after the exposure (or treatment) at each time point individually. However, this kind of analysis cannot identify dynamic (time-dependent) events of dose responsiveness.

RESULTS

We propose TinderMIX, an approach that simultaneously models the effects of time and dose on the transcriptome to investigate the course of molecular alterations exerted in response to the exposure. Starting from gene log fold-change, TinderMIX fits different integrated time and dose models to each gene, selects the optimal one, and computes its time and dose effect map; then a user-selected threshold is applied to identify the responsive area on each map and verify whether the gene shows a dynamic (time-dependent) and dose-dependent response; eventually, responsive genes are labelled according to the integrated time and dose point of departure.

CONCLUSIONS

To showcase the TinderMIX method, we analysed 2 drugs from the Open TG-GATEs dataset, namely, cyclosporin A and thioacetamide. We first identified the dynamic dose-dependent mechanism of action of each drug and compared them. Our analysis highlights that different time- and dose-integrated point of departure recapitulates the toxicity potential of the compounds as well as their dynamic dose-dependent mechanism of action.

Transcriptomics in Toxicogenomics, Part III: Data Modelling for Risk Assessment

Transcriptomics data are relevant to address a number of challenges in Toxicogenomics (TGx). After careful planning of exposure conditions and data preprocessing, the TGx data can be used in predictive toxicology, where more advanced modelling techniques are applied. The large volume of molecular profiles produced by omics-based technologies allows the development and application of artificial intelligence (AI) methods in TGx. Indeed, the publicly available omics datasets are constantly increasing together with a plethora of different methods that are made available to facilitate their analysis, interpretation and the generation of accurate and stable predictive models. In this review, we present the state-of-the-art of data modelling applied to transcriptomics data in TGx. We show how the benchmark dose (BMD) analysis can be applied to TGx data. We review read across and adverse outcome pathways (AOP) modelling methodologies. We discuss how network-based approaches can be successfully employed to clarify the mechanism of action (MOA) or specific biomarkers of exposure. We also describe the main AI methodologies applied to TGx data to create predictive classification and regression models and we address current challenges. Finally, we present a short description of deep learning (DL) and data integration methodologies applied in these contexts. Modelling of TGx data represents a valuable tool for more accurate chemical safety assessment. This review is the third part of a three-article series on Transcriptomics in Toxicogenomics.

A Self-Organizing Map of the Fathead Minnow Liver Transcriptome to Identify Consistent Toxicogenomic Patterns across Chemical Fingerprints

Lack of consistent findings in different experimental settings remains a major challenge in toxicogenomics. The present study investigated whether consistency between findings of different microarray experiments can be improved when the analysis is based on a common reference frame ("toxicogenomic universe"), which can be generated using the machine learning algorithm of the self-organizing map (SOM). This algorithm arranges and clusters genes on a 2-dimensional grid according to their similarity in expression across all considered data. In the present study, 19 data sets, comprising of 54 different adult fathead minnow liver exposure experiments, were retrieved from Gene Expression Omnibus and used to train a SOM. The resulting toxicogenomic universe aggregates 58 872 probes to 2500 nodes and was used to project, visualize, and compare the fingerprints of these 54 different experiments. For example, we could identify a common pattern, with 14% of significantly regulated nodes in common, in the data sets of an interlaboratory study of ethinylestradiol exposures. Consistency could be improved compared with the 5% total overlap in regulated genes reported before. Furthermore, we could determine a specific and consistent estrogen-related pattern of differentially expressed nodes and clusters in the toxicogenomic universe by applying additional clustering steps and comparing all obtained fingerprints. Our study shows that the SOM-based approach is useful for generating comparable toxicogenomic fingerprints and improving consistency between results of different experiments. Environ Toxicol Chem 2020;39:526-537. © 2019 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals, Inc. on behalf of SETAC.