Framework for Identifying Chemicals with Structural Features Associated with the Potential to Act as Developmental or Reproductive Toxicants

Advancing systemic toxicity risk assessment: Evaluation of a NAM-based toolbox approach

For many years, a method that allowed systemic toxicity safety assessments to be conducted without generating new animal test data, seemed out of reach. However, several different research groups and regulatory authorities are beginning to use a variety of in silico, in chemico, and in vitro techniques to inform safety decisions. To manage this transition to animal-free safety assessments responsibly, it is important to ensure that the level of protection offered by a safety assessment based on new approach methodologies (NAMs), is at least as high as that provided by a safety assessment based on traditional animal studies. To this end, we have developed an evaluation strategy to assess both the level of protection and the utility offered by a NAM-based systemic safety "toolbox." The toolbox comprises physiologically based kinetic models to predict internal exposures, and bioactivity NAMs designed to give broad coverage across many different toxicity modes of action. The output of the toolbox is the calculation of a bioactivity:exposure ratio (analogous to a margin of internal exposure), which can be used to inform decision-making. In this work, we have expanded upon an initial pilot study of 10 chemicals with an additional 38 chemicals and 70 consumer exposure scenarios. We found that, for the majority of these (>90%), the NAM-based workflow is protective of human health, enabling us to make animal-free safety decisions for systemic toxicity and preventing unnecessary animal use. We have also identified critical areas for improvement to further increase our confidence in the robustness of the approach.

Health risk assessment for dietary exposure to 3-monochloropropane-1,2-diol, 2-monochloropropane-1,2-diol, and glycidol for Italian consumers

3-Monochloropropane-1,2-diol (3-MCPD), 2-monochloropropane-1,2-diol (2-MCPD) and 2,3-epoxy-1-propanol (glycidol), in their free form or esterified to fatty acids, are food contaminants formed during the refinement of oils and fats. We conducted a survey to quantify the levels of these compounds in 130 food items, in order to assess the exposure to them in food and the consequent health risk for consumers. Food samples, including infant formula, were analysed by gas-chromatography mass spectrometry with the indirect method, and we used the latest open access food consumption database for the Italian population for a probabilistic assessment of exposure. We adopted an in silico approach to fill the gap for the toxicity of 2-MCPD. The occurrence values for the three contaminants in food were in most cases lower than or comparable to those reported in previous surveys. Exposure assessment for the most exposed individuals (95thpercentiles of consumers only) of different age groups, gave values below the tolerable daily intake recommended by the European Food Safety Authority for 3-MCPD and below the simulated or predicted toxicity thresholds for 2-MCPD, indicating a negligible risk due to dietary exposure to these contaminants. For glycidol, however, estimated exposure indicated a non-negligible increase in cancer risk, and a margin of exposure <25,000 for younger population groups, indicating a potential health concern.

Assessment of emerging and persistent contaminants in an anthropogenic-impacted watershed: Application using targeted, non-targeted, and in vitro bioassay techniques

Emerging and persistent contaminants (EPC) pose a significant challenge to water quality monitoring efforts. Effect-based monitoring (EBM) techniques provide an efficient and systematic approach in water quality monitoring, but they tend to be resource intensive. In this study, we investigated the EPC distribution for various land uses using target analysis (TA) and non-target screening (NTS) and in vitro bioassays, both individually and integrated, in the upper Ping River Catchment, northern Thailand. Our findings of NTS showed that urban areas were the most contaminated of all land use types, although agriculture sites had high unexpected pollution levels. We evaluated the reliability of NTS data by comparing it to TA and observed varying inconsistencies likely due to matrix interferences and isobaric compound interferences. Integrating NTS with in vitro bioassays for a thorough analysis posed challenges, primary due to a scarcity of concentration data for key compounds, and potentially additive or non-additive effects of mixture samples that could not be accounted for. While EBM approaches place emphasis on toxic sites, this study demonstrated the importance of considering non-bioactive sites that contain toxic compounds with antagonistic effects that may go undetected by traditional monitoring approaches. The present work emphasizes the importance of improving NTS workflows and ensuring high-quality EBM analyses in future water quality monitoring programs.

Knowledge-based machine learning for predicting and understanding the androgen receptor (AR)-mediated reproductive toxicity in zebrafish

Traditional methods for identifying endocrine-disrupting chemicals (EDCs) that activate androgen receptors (AR) are costly, time-consuming, and low-throughput. This study developed a knowledge-based deep neural network model (AR-DNN) to predict AR-mediated adverse outcomes on female zebrafish fertility. This model started with chemical fingerprints as the input layer and was implemented through a five-layer virtual AR-induced adverse outcome pathway (AOP). Results indicated that the AR-DNN effectively and accurately screens new reproductive toxicants (AUC = 0.94, accuracy = 0.85), providing potential toxicity pathways. Furthermore, 1477 and 2448 chemicals that could lead to infertility were identified in the plastic additives list (PLASTICMAP, n = 7112) and the Inventory of Existing Chemical Substances in China (IECSC, n = 17741), respectively. Colourants containing steroid-like structures are the major active plastic additives that might lower female zebrafish fertility through AR binding, DNA binding, and transcriptional activation. While active IECSC chemicals primarily have the same fragments, such as benzonitrile, nitrobenzene, and quinolone. The predicted toxicity pathways were consistent with existing fish evidence, demonstrating the model's applicability. This knowledge-based approach offers a promising computational toxicology strategy for predicting and characterising the endocrine-disrupting effects and toxic mechanisms of organic chemicals, potentially leading to more efficient and cost-effective screening of EDCs.

Next generation risk assessment for occupational chemical safety - A real world example with sodium-2-hydroxyethane sulfonate

Next Generation Risk Assessment (NGRA) is an exposure-led approach to safety assessment that uses New Approach Methodologies (NAMs). Application of NGRA has been largely restricted to assessments of consumer use of cosmetics and is not currently implemented in occupational safety assessments, e.g. under EU REACH. By contrast, a large proportion of regulatory worker safety assessments are underpinned by toxicological studies using experimental animals. Consequently, occupational safety assessment represents an area that would benefit from increasing application of NGRA to safety decision making. Here, a workflow for conducting NGRA under an occupational safety context was developed, which is illustrated with a case study chemical; sodium 2-hydroxyethane sulphonate (sodium isethionate or SI). Exposures were estimated using a standard occupational exposure model following a comprehensive life cycle assessment of SI and considering factory-specific data. Outputs of this model were then used to estimate internal exposures using a Physiologically Based Kinetic (PBK) model, which was constructed with SI specific Absorption, Distribution, Metabolism and Excretion (ADME) data. PBK modelling indicated a worst-case plasma maximum concentration (Cmax) of 0.8 μM across the SI life cycle. SI bioactivity was assessed in a battery of NAMs relevant to systemic, reproductive, and developmental toxicity; a cell stress panel, high throughput transcriptomics in three cell lines (HepG2, HepaRG and MCF-7 cells), pharmacological profiling and specific assays relating to developmental toxicity (Reprotracker and devTOX quickPredict). Points of Departure (PoDs) for SI ranged from 104 to 5044 µM. Cmax values obtained from PBK modelling of occupational exposures to SI were compared with PoDs from the bioactivity assays to derive Bioactivity Exposure Ratios (BERs) which demonstrated the safety for workers exposed to SI under current levels of factory specific risk management. In summary, the tiered and iterative workflow developed here represents an opportunity for integrating non animal approaches for a large subset of substances for which systemic worker safety assessment is required. Such an approach could be followed to ensure that animal testing is only conducted as a "last resort" e.g. under EU REACH.

Where developmental toxicity meets explainable artificial intelligence: state-of-the-art and perspectives

INTRODUCTION

The application of Artificial Intelligence (AI) to predictive toxicology is rapidly increasing, particularly aiming to develop non-testing methods that effectively address ethical concerns and reduce economic costs. In this context, Developmental Toxicity (Dev Tox) stands as a key human health endpoint, especially significant for safeguarding maternal and child well-being.

AREAS COVERED

This review outlines the existing methods employed in Dev Tox predictions and underscores the benefits of utilizing New Approach Methodologies (NAMs), specifically focusing on eXplainable Artificial Intelligence (XAI), which proves highly efficient in constructing reliable and transparent models aligned with recommendations from international regulatory bodies.

EXPERT OPINION

The limited availability of high-quality data and the absence of dependable Dev Tox methodologies render XAI an appealing avenue for systematically developing interpretable and transparent models, which hold immense potential for both scientific evaluations and regulatory decision-making.

TISBE: A Public Web Platform for the Consensus-Based Explainable Prediction of Developmental Toxicity

Despite being extremely relevant for the protection of prenatal and neonatal health, the developmental toxicity (Dev Tox) is a highly complex endpoint whose molecular rationale is still largely unknown. The lack of availability of high-quality data as well as robust nontesting methods makes its understanding even more difficult. Thus, the application of new explainable alternative methods is of utmost importance, with Dev Tox being one of the most animal-intensive research themes of regulatory toxicology. Descending from TIRESIA (Toxicology Intelligence and Regulatory Evaluations for Scientific and Industry Applications), the present work describes TISBE (TIRESIA Improved on Structure-Based Explainability), a new public web platform implementing four fundamental advancements for in silico analyses: a three times larger dataset, a transparent XAI (explainable artificial intelligence) framework employing a fragment-based fingerprint coding, a novel consensus classifier based on five independent machine learning models, and a new applicability domain (AD) method based on a double top-down approach for better estimating the prediction reliability. The training set (TS) includes as many as 1008 chemicals annotated with experimental toxicity values. Based on a 5-fold cross-validation, a median value of 0.410 for the Matthews correlation coefficient was calculated; TISBE was very effective, with a median value of sensitivity and specificity equal to 0.984 and 0.274, respectively. TISBE was applied on two external pools made of 1484 bioactive compounds and 85 pediatric drugs taken from ChEMBL (Chemical European Molecular Biology Laboratory) and TEDDY (Task-Force in Europe for Drug Development in the Young) repositories, respectively. Notably, TISBE gives users the option to clearly spot the molecular fragments responsible for the toxicity or the safety of a given chemical query and is available for free at https://prometheus.farmacia.uniba.it/tisbe.

Interlaboratory Study on Zebrafish in Toxicology: Systematic Evaluation of the Application of Zebrafish in Toxicology's (SEAZIT's) Evaluation of Developmental Toxicity

Embryonic zebrafish represent a useful test system to screen substances for their ability to perturb development. The exposure scenarios, endpoints captured, and data analysis vary among the laboratories who conduct screening. A lack of harmonization impedes the comparison of the substance potency and toxicity outcomes across laboratories and may hinder the broader adoption of this model for regulatory use. The Systematic Evaluation of the Application of Zebrafish in Toxicology (SEAZIT) initiative was developed to investigate the sources of variability in toxicity testing. This initiative involved an interlaboratory study to determine whether experimental parameters altered the developmental toxicity of a set of 42 substances (3 tested in duplicate) in three diverse laboratories. An initial dose-range-finding study using in-house protocols was followed by a definitive study using four experimental conditions: chorion-on and chorion-off using both static and static renewal exposures. We observed reasonable agreement across the three laboratories as 33 of 42 test substances (78.6%) had the same activity call. However, the differences in potency seen using variable in-house protocols emphasizes the importance of harmonization of the exposure variables under evaluation in the second phase of this study. The outcome of the Def will facilitate future practical discussions on harmonization within the zebrafish research community.

Detection of lenalidomide metabolites in urine to discover drug-resistant compounds

Lenalidomide is the first-line drug for the clinical treatment of multiple myeloma. However, its efficacy differs significantly among patients. Clinically, after lenalidomide treatment, few patients' conditions worsened, whereas others remained stable or improved. To clarify the reasons for this difference in efficacy, 20 patients with multiple myeloma who received maintenance treatment with lenalidomide were retrospectively included in this study. Lenalidomide metabolic compounds were detected in patient urine using mass spectrometry. A rapid and accurate ultra-performance liquid chromatography-time-of-flight tandem mass spectrometry (UPLC-TOF-MS/MS) method was used to characterize metabolites in the urine of different patients. Eleven metabolites, including four new compounds, were identified and characterized in all the samples. Among these, two metabolites were found to have obvious discrepancies in different groups of patients. One metabolite named Denitrified-2 glutarimide, a new potential compound, was only detected in the urine of ineffective and stable patients, whereas the other metabolite named 5-Hydroxy-lenalidomide was found only in the urine of effective patients.

Making sense of chemical space network shows signs of criticality

Chemical space modelling has great importance in unveiling and visualising latent information, which is critical in predictive toxicology related to drug discovery process. While the use of traditional molecular descriptors and fingerprints may suffer from the so-called curse of dimensionality, complex networks are devoid of the typical drawbacks of coordinate-based representations. Herein, we use chemical space networks (CSNs) to analyse the case of the developmental toxicity (Dev Tox), which remains a challenging endpoint for the difficulty of gathering enough reliable data despite very important for the protection of the maternal and child health. Our study proved that the Dev Tox CSN has a complex non-random organisation and can thus provide a wealth of meaningful information also for predictive purposes. At a phase transition, chemical similarities highlight well-established toxicophores, such as aryl derivatives, mostly neurotoxic hydantoins, barbiturates and amino alcohols, steroids, and volatile organic compounds ether-like chemicals, which are strongly suspected of the Dev Tox onset and can thus be employed as effective alerts for prioritising chemicals before testing.

Advancing Computational Toxicology by Interpretable Machine Learning

Chemical toxicity evaluations for drugs, consumer products, and environmental chemicals have a critical impact on human health. Traditional animal models to evaluate chemical toxicity are expensive, time-consuming, and often fail to detect toxicants in humans. Computational toxicology is a promising alternative approach that utilizes machine learning (ML) and deep learning (DL) techniques to predict the toxicity potentials of chemicals. Although the applications of ML- and DL-based computational models in chemical toxicity predictions are attractive, many toxicity models are "black boxes" in nature and difficult to interpret by toxicologists, which hampers the chemical risk assessments using these models. The recent progress of interpretable ML (IML) in the computer science field meets this urgent need to unveil the underlying toxicity mechanisms and elucidate the domain knowledge of toxicity models. In this review, we focused on the applications of IML in computational toxicology, including toxicity feature data, model interpretation methods, use of knowledge base frameworks in IML development, and recent applications. The challenges and future directions of IML modeling in toxicology are also discussed. We hope this review can encourage efforts in developing interpretable models with new IML algorithms that can assist new chemical assessments by illustrating toxicity mechanisms in humans.

Refine and Strengthen SAR-Based Read-Across by Considering Bioactivation and Modes of Action

Structure-activity relationship (SAR)-based read-across is an important and effective method to establish the safety of a data-poor target chemical (structure of interest (SOI)) using hazard data from structurally similar source chemicals (analogues). Many methods use quantitative similarity scores to evaluate the structural similarity for searching and selecting analogues as well as for evaluating analogue suitability. However, studies suggest that read-across based purely on structural similarity cannot accurately predict the toxicity of an SOI. As mechanistic data become available, we gain a greater understanding of the mode of action (MOA), the relationship between structures and metabolism/bioactivation pathways, and the existence of "activity cliffs" in chemical chain length, which can improve the analogue rating process. For this purpose, the current work identifies a series of classes of chemicals where a small change at a key position can result in a significant change in metabolism and bioactivation pathways and may eventually result in significant changes in chemical toxicity that have a big impact on the suitability of analogues for read-across. Additionally, a series of SAR-based read-across case studies are presented, which cover a variety of chemical classes that commonly link to different toxic endpoints. The case study results indicate that SAR-based read-across can be refined and strengthened by considering MOAs or proposed reactive metabolite formation pathways, which can improve the overall accuracy, consistency, transparency, and confidence in evaluating analogue suitability.

A review on potential approach for in silico toxicity analysis of respirable fraction of ambient particulate matter

Epidemiological and toxicological studies have shown the adverse effect of ambient particulate matter (PM) on respiratory and cardiovascular systems inside the human body. Various cellular and acellular assays in literature use indicators like ROS generation, cell inflammation, mutagenicity, etc., to assess PM toxicity and associated health effects. The presence of toxic compounds in respirable PM needs detailed studies for proper understanding of absorption, distribution, metabolism, and excretion mechanisms inside the body as it is difficult to accurately imitate or simulate these mechanisms in lab or animal models. The leaching kinetics of the lung fluid, PM composition, retention time, body temperature, etc., are hard to mimic in an artificial experimental setup. Moreover, the PM size fraction also plays an important role. For example, the ultrafine particles may directly enter systemic circulations while coarser PM10 may be trapped and deposited in the tracheo-bronchial region. Hence, interpretation of these results in toxicity models should be done judiciously. Computational models predicting PM toxicity are rare in the literature. The variable composition of PM and lack of proper understanding for their synergistic role inside the body are prime reasons behind it. This review explores different possibilities of in silico modeling and suggests possible approaches for the risk assessment of PM particles. The toxicity testing approach for engineered nanomaterials, drugs, food industries, etc., have also been investigated for application in computing PM toxicity.

Advancing the use of new approach methodologies for assessing teratogenicity: Building a tiered approach

Many New Approach Methodologies (NAMs) have been developed for the safety assessment of new ingredients. Research into reproductive toxicity and teratogenicity is a particularly high priority, especially given their mechanistic complexity. Forty-six non-teratogenic and 39 teratogenic chemicals were screened for teratogenic potential using the in silico DART model from the OECD QSAR Toolbox; the devTox quickPredict™ (devTox assay) test and the Zebrafish Embryotoxicity Test (ZET). The sensitivity and specificity were 94.7% and 84.1%, respectively, for the DART tree (83 chemicals), 86.1% and 35.6% for the devTox (81 chemicals) and 77.8% and 76.7% for the ZET (57 chemicals). Fifty-three chemicals were tested in all three assays and when results were combined and based on a "2 out of 3 rule", the sensitivity and specificity were 96.0% and 71.4%, respectively. The specificity of the devTox assay for a sub-set of 43 chemicals was increased from 26.1% to 82.6% by incorporating human plasma concentrations into the assay interpretation. When all 85 chemicals were assessed in a decision tree approach, there was an excellent predictivity and assay robustness of 90%. In conclusion, all three models exhibited a good sensitivity and specificity, especially when outcomes from all three were combined or used in "2 out of 3" or a tiered decision tree approach. The latter is an interesting predictive approach for evaluating the teratogenic potential of new chemicals. Future investigations will extend the number of chemicals tested, as well as explore ways to refine the results and obtain a robust Integrated Testing Strategy to evaluate teratogenic potential.

High Throughput Read-Across for Screening a Large Inventory of Related Structures by Balancing Artificial Intelligence/Machine Learning and Human Knowledge

Read-across is an in silico method applied in chemical risk assessment for data-poor chemicals. The read-across outcomes for repeated-dose toxicity end points include the no-observed-adverse-effect level (NOAEL) and estimated uncertainty for a particular category of effects. We have previously developed a new paradigm for estimating NOAELs based on chemoinformatics analysis and experimental study qualities from selected analogues, not relying on quantitative structure-activity relationships (QSARs) or rule-based SAR systems, which are not well-suited to end points for which the underpinning data are weakly grounded in specific chemical-biological interactions. The central hypothesis of this approach is that similar compounds have similar toxicity profiles and, hence, similar NOAEL values. Analogue quality (AQ) quantifies the suitability of an analogue candidate for reading across to the target by considering similarity from structure, physicochemical, ADME (absorption, distribution, metabolism, excretion), and biological perspectives. Biological similarity is based on experimental data; assay vectors derived from aggregations of ToxCast/Tox21 data are used to derive machine learning (ML) hybrid rules that serve as biological fingerprints to capture target-analogue similarity relevant to specific effects of interest, for example, hormone receptors (ER/AR/THR). Once one or more analogues have been qualified for read-across, a decision theory approach is used to estimate confidence bounds for the NOAEL of the target. The confidence interval is dramatically narrowed when analogues are constrained to biologically related profiles. Although this read-across process works well for a single target with several analogues, it can become unmanageable when, for example, screening multiple targets (e.g., virtual screening library) or handling a parent compound having numerous metabolites. To this end, we have established a digitalized framework to enable the assessment of a large number of substances, while still allowing for human decisions for filtering and prioritization. This workflow was developed and validated through a use case of a large set of bisphenols and their metabolites.

Identifying chemicals based on receptor binding/bioactivation/mechanistic explanation associated with potential to elicit hepatotoxicity and to support structure activity relationship-based read-across

The liver is the most common target organ in toxicology studies. The development of chemical structural alerts for identifying hepatotoxicity will play an important role in in silico model prediction and help strengthen the identification of analogs used in structure activity relationship (SAR)- based read-across. The aim of the current study is development of an SAR-based expert-system decision tree for screening of hepatotoxicants across a wide range of chemistry space and proposed modes of action for clustering of chemicals using defined core chemical categories based on receptor-binding or bioactivation. The decision tree is based on ∼ 1180 different chemicals that were reviewed for hepatotoxicity information. Knowledge of chemical receptor binding, metabolism and mechanistic information were used to group these chemicals into 16 different categories and 102 subcategories: four categories describe binders to 9 different receptors, 11 categories are associated with possible reactive metabolites (RMs) and there is one miscellaneous category. Each chemical subcategory has been associated with possible modes of action (MOAs) or similar key structural features. This decision tree can help to screen potential liver toxicants associated with core structural alerts of receptor binding and/or RMs and be used as a component of weight of evidence decisions based on SAR read-across, and to fill data gaps.

Quantifying Analogue Suitability for SAR-Based Read-Across Toxicological Assessment

Structure activity relationship (SAR)-based read-across often is an integral part of toxicological safety assessment, and justification of the prediction presents the most challenging aspect of the approach. It has been established that structural consideration alone is inadequate for selecting analogues and justifying their use, and biological relevance must be incorporated. Here we introduce an approach for considering biological and toxicological related features quantitatively to compute a similarity score that is concordant with suitability for a read-across prediction for systemic toxicity. Fingerprint keys for comparing metabolism, reactivity, and physical chemical properties are presented and used to compare these attributes for 14 case study chemicals each with a list of potential analogues. Within each case study, the sum of these nonstructural similarity scores is consistent with suitability for read-across established using an approach based on expert judgment. Machine learning is applied to determine the contributions from each of the similarity attributes revealing their importance for each structure class. This approach is used to quantify and communicate the differences between a target and a potential analogue as well as rank analogue quality when more than one is relevant. A numerical score with easily interpreted fingerprints increases transparency and consistency among experts, facilitates implementation by others, and ultimately increases chances for regulatory acceptance.

Structure-activity relationship read-across and transcriptomics for branched carboxylic acids

The purpose of this study was to use chemical similarity evaluations, transcriptional profiling, in vitro toxicokinetic data, and physiologically based pharmacokinetic (PBPK) models to support read-across for a series of branched carboxylic acids using valproic acid (VPA), a known developmental toxicant, as a comparator. The chemicals included 2-propylpentanoic acid (VPA), 2-ethylbutanoic acid, 2-ethylhexanoic acid (EHA), 2-methylnonanoic acid, 2-hexyldecanoic acid, 2-propylnonanoic acid (PNA), dipentyl acetic acid or 2-pentylheptanoic acid, octanoic acid (a straight chain alkyl acid), and 2-ethylhexanol. Transcriptomics was evaluated in 4 cell types (A549, HepG2, MCF7, and iCell cardiomyocytes) 6 h after exposure to 3 concentrations of the compounds, using the L1000 platform. The transcriptional profiling data indicate that 2- or 3-carbon alkyl substituents at the alpha position of the carboxylic acid (EHA and PNA) elicit a transcriptional profile similar to the one elicited by VPA. The transcriptional profile is different for the other chemicals tested, which provides support for limiting read-across from VPA to much shorter and longer acids. Molecular docking models for histone deacetylases, the putative target of VPA, provide a possible mechanistic explanation for the activity cliff elucidated by transcriptomics. In vitro toxicokinetic data were utilized in a PBPK model to estimate internal dosimetry. The PBPK modeling data show that as the branched chain increases, predicted plasma Cmax decreases. This work demonstrates how transcriptomics and other mode of action-based methods can improve read-across.

TIRESIA: An eXplainable Artificial Intelligence Platform for Predicting Developmental Toxicity

Herein, a robust and reproducible eXplainable Artificial Intelligence (XAI) approach is presented, which allows prediction of developmental toxicity, a challenging human-health endpoint in toxicology. The application of XAI as an alternative method is of the utmost importance with developmental toxicity being one of the most animal-intensive areas of regulatory toxicology. In this work, the established CAESAR (Computer Assisted Evaluation of industrial chemical Substances According to Regulations) training set made of 234 chemicals for model learning is employed. Two test sets, including as a whole 585 chemicals, were instead used for validation and generalization purposes. The proposed framework favorably compares with the state-of-the-art approaches in terms of accuracy, sensitivity, and specificity, thus resulting in a reliable support system for developmental toxicity ensuring informativeness, uncertainty estimation, generalization, and transparency. Based on the eXtreme Gradient Boosting (XGB) algorithm, our predictive model provides easy interpretative keys based on specific molecular descriptors and structural alerts enabling one to distinguish toxic and nontoxic chemicals. Inspired by the Organisation for Economic Co-operation and Development (OECD) principles for the validation of Quantitative Structure-Activity Relationships (QSARs) for regulatory purposes, the results are summarized in a standard report in portable document format, enclosing also details concerned with a density-based model applicability domain and SHAP (SHapley Additive exPlanations) explainability, the latter particularly useful to better understand the effective roles played by molecular features. Notably, our model has been implemented in TIRESIA (Toxicology Intelligence and Regulatory Evaluations for Scientific and Industry Applications), a free of charge web platform available at http://tiresia.uniba.it.

Practical application of the interim internal threshold of toxicological concern (iTTC): a case study based on clinical data

We present a case study that provides a practical step-by-step example of how the internal Threshold of Toxicological Concern (iTTC) can be used as a tool to refine a TTC-based assessment for dermal exposures to consumer products. The case study uses a theoretical scenario where there are no systemic toxicity data for the case study chemicals (avobenzone, oxybenzone, octocrylene, homosalate, octisalate, octinoxate, and ecamsule). Human dermal pharmacokinetic data following single and repeat dermal exposure to products containing the case study chemicals were obtained from data published by the US FDA. The clinical studies utilized an application procedure that followed maximal use conditions (product applied as 2 mg/cm² to 75% of the body surface area, 4 times a day). The case study chemicals were first reviewed to determine if they were in the applicability domain of the iTTC, and then, the human plasma concentrations were compared to an iTTC limit of 1 µM. When assessed under maximum usage, the external exposure of all chemicals exceeded the external dose TTC limits. By contrast, the internal exposure to all chemicals, except oxybenzone, was an order of magnitude lower than the 1 µM interim iTTC threshold. This work highlights the importance of understanding internal exposure relative to external dose and how the iTTC can be a valuable tool for assessing low-level internal exposures; additionally, the work demonstrates how to use an iTTC, and highlights considerations and refinement opportunities for the approach.

Understanding and overcoming the technical challenges in using in silico predictions in regulatory decisions of complex toxicological endpoints - A pesticide perspective for regulatory toxicologists with a focus on machine learning models

There are many challenges that must be overcome before in silico toxicity predictions are ripe for regulatory decision-making. Today, mandates in the United States of America and the European Union to avoid animal usage in toxicity testing is driving the need to consider alternative technologies, including Quantitative Structure Activity Relationship (QSAR) models, and read across approaches. However, when adopting new methods, it is critical that both new approach developers as well as regulatory users understand the strengths and challenges with these new approaches. In this paper, we identify potential sources of bias in machine learning methods specific to toxicity predictions, that may impact the overall performance of in silico models. We also discuss ways to mitigate these biases. Based on our experiences, the most prevalent sources of bias include class imbalance (differing numbers of "toxic" vs "nontoxic" compounds), limited numbers of chemicals within a particular chemistry, and biases within the studies that make up the database used for model building, as well as model evaluation biases. While this is already complex for repeated dose toxicity, in reproduction and developmental toxicity a further level of complexity is introduced by the need to evaluate effects on individual animal and litter basis (e.g., a hierarchal structure). We also discuss key considerations developers and regulators need to make when they use machine learning models to predict chemical safety. Our objective is for our paper to serve as a desk reference for model developers and regulators as they evaluate machine learning models and as they make decisions using these models.

Human-Based New Approach Methodologies in Developmental Toxicity Testing: A Step Ahead from the State of the Art with a Feto-Placental Organ-on-Chip Platform

Developmental toxicity testing urgently requires the implementation of human-relevant new approach methodologies (NAMs) that better recapitulate the peculiar nature of human physiology during pregnancy, especially the placenta and the maternal/fetal interface, which represent a key stage for human lifelong health. Fit-for-purpose NAMs for the placental-fetal interface are desirable to improve the biological knowledge of environmental exposure at the molecular level and to reduce the high cost, time and ethical impact of animal studies. This article reviews the state of the art on the available in vitro (placental, fetal and amniotic cell-based systems) and in silico NAMs of human relevance for developmental toxicity testing purposes; in addition, we considered available Adverse Outcome Pathways related to developmental toxicity. The OECD TG 414 for the identification and assessment of deleterious effects of prenatal exposure to chemicals on developing organisms will be discussed to delineate the regulatory context and to better debate what is missing and needed in the context of the Developmental Origins of Health and Disease hypothesis to significantly improve this sector. Starting from this analysis, the development of a novel human feto-placental organ-on-chip platform will be introduced as an innovative future alternative tool for developmental toxicity testing, considering possible implementation and validation strategies to overcome the limitation of the current animal studies and NAMs available in regulatory toxicology and in the biomedical field.

Identifying the landscape of developmental toxicity new approach methodologies

BACKGROUND

The dynamics and complexities of in utero fetal development create significant challenges in transitioning from lab animal-centric developmental toxicity testing methods to assessment strategies based on new approach methodologies (NAMs). Nevertheless, considerable progress is being made, stimulated by increased research investments and scientific advances, such as induced pluripotent stem cell-derived models. To help identify developmental toxicity NAMs for toxicity screening and potential funding through the American Chemistry Council's Long-Range Research Initiative, a systematic literature review was conducted to better understand the current landscape of developmental toxicity NAMs.

METHODS

Scoping review tools were used to systematically survey the literature (2010-2021; ~18,000 references identified), results and metadata were then extracted, and a user-friendly interactive dashboard was created.

RESULTS

The data visualization dashboard, developed using Tableau® software, is provided as a free, open-access web tool. This dashboard enables straightforward interactive queries and visualizations to identify trends and to distinguish and understand areas or NAMs where research has been most, or least focused.

CONCLUSIONS

Herein, we describe the approach and methods used, summarize the benefits and challenges of applying the systematic-review techniques, and highlight the types of questions and answers for which the dashboard can be used to explore the many different facets of developmental toxicity NAMs.

Development, Validation, and Application of a Human Reproductive Toxicity Prediction Model Based on Adverse Outcome Pathway

A growing number of environmental contaminants have been proved to have reproductive toxicity to males and females. However, the unclear toxicological mechanism of reproductive toxicants limits the development of virtual screening methods. By consolidating androgen (AR)-/estrogen receptors (ERs)-mediated adverse outcome pathways (AOPs) with more than 8000 chemical substances, we uncovered relationships between chemical features, a series of pathway-related effects, and reproductive apical outcomes─changes in sex organ weights. An AOP-based computational model named RepTox was developed and evaluated to predict and characterize chemicals' reproductive toxicity for males and females. Results showed that RepTox has three outstanding advantages. (I) Compared with the traditional models (37 and 81% accuracy, respectively), AOP significantly improved the predictive robustness of RepTox (96.3% accuracy). (II) Compared with the application domain (AD) of models based on small in vivo datasets, AOP expanded the ADs of RepTox by 1.65-fold for male and 3.77-fold for female, respectively. (III) RepTox implied that hydrophobicity, cyclopentanol substructure, and several topological indices (e.g., hydrogen-bond acceptors) were important, unbiased features associated with reproductive toxicants. Finally, RepTox was applied to the inventory of existing chemical substances of China and identified 2100 and 7281 potential toxicants to the male and female reproductive systems, respectively.

Integration of in silico methods to determine endocrine-disrupting tobacco pollutants binding potency with steroidogenic genes: comprehensive QSAR modeling and ensemble docking strategies

A myriad of tobacco-associated chemicals may have possibilities to developmental/reproductive axis and endocrine-disruption impacts. Mostly they breach the biotransformation of cholesterol in mitochondria by interfering with steroidogenic pathway genes, prompting to adverse effects in steroid biosynthesis; however, studies are scanty. The quantitative structure-activity relationship (QSAR) modeling and comparative docking strategies were used to understand structural features of dataset compounds that influence developmental/reproductive toxicity and estrogen and androgen receptor-binding abilities, and to predict binding levels of toxicants with steroidogenic acute regulatory protein (StAR) and cholesterol side-chain cleavage enzyme (CYP11A1) active sites. Developed QSAR models presented good robustness and predictive ability that were determined from the applicability domain and, clustering and classification of chemicals by performing self-organizing maps. Accordingly, the exorbitant amount of polycyclic aromatic hydrocarbons (PAHs) and a limited number of other chemicals including N-nitrosamines and nicotine was represented as potential developmental/reproductive toxicants as well as estrogen and androgen receptor binders. From the docking analysis, hydrogen bonding, nonpolar, atomic π-stacking, and π-cation interactions were found between PAHs (bay and fjord structural pockets) and functional hotspot residues of StAR and CYP11A1, which strengthened the subtle structural changes at domains. These govern barrier effects to cholesterol binding and/or locking cholesterol to complicate its ejection from the Ω1 loop of StAR, and further mitigates steroid biosynthesis through cholesterol by CYP11A1; therefore, they are presumably considered as block-cluster mechanisms. These outcomes are significant to be hopeful to estimate developmental/reproductive toxicity and endocrine-disruption activities of other environmental pollutants, and could be useful for further assessment to discover binding mechanisms of PAHs with other steroidogenesis pathway genes.

Are Non-animal Systemic Safety Assessments Protective? A Toolbox and Workflow

An important question in toxicological risk assessment is whether non-animal new approach methodologies (NAMs) can be used to make safety decisions that are protective of human health, without being overly conservative. In this work, we propose a core NAM toolbox and workflow for conducting systemic safety assessments for adult consumers. We also present an approach for evaluating how protective and useful the toolbox and workflow are by benchmarking against historical safety decisions. The toolbox includes physiologically based kinetic (PBK) models to estimate systemic Cmax levels in humans, and 3 bioactivity platforms, comprising high-throughput transcriptomics, a cell stress panel, and in vitro pharmacological profiling, from which points of departure are estimated. A Bayesian model was developed to quantify the uncertainty in the Cmax estimates depending on how the PBK models were parameterized. The feasibility of the evaluation approach was tested using 24 exposure scenarios from 10 chemicals, some of which would be considered high risk from a consumer goods perspective (eg, drugs that are systemically bioactive) and some low risk (eg, existing food or cosmetic ingredients). Using novel protectiveness and utility metrics, it was shown that up to 69% (9/13) of the low risk scenarios could be identified as such using the toolbox, whilst being protective against all (5/5) the high-risk ones. The results demonstrated how robust safety decisions could be made without using animal data. This work will enable a full evaluation to assess how protective and useful the toolbox and workflow are across a broader range of chemical-exposure scenarios.

Review of the state of science and evaluation of currently available in silico prediction models for reproductive and developmental toxicity: A case study on pesticides

BACKGROUND

In silico methods for toxicity prediction have increased significantly in recent years due to the 3Rs principle. This also applies to predicting reproductive toxicology, which is one of the most critical factors in pesticide approval. The widely used quantitative structure-activity relationship (QSAR) models use experimental toxicity data to create a model that relates experimentally observed toxicity to molecular structures to predict toxicity. Aim of the study was to evaluate the available prediction models for developmental and reproductive toxicity regarding their strengths and weaknesses in a pesticide database.

METHODS

The reproductive toxicity of 315 pesticides, which have a GHS classification by ECHA, was compared with the prediction of different in silico models: VEGA, OECD (Q)SAR Toolbox, Leadscope Model Applier, and CASE Ultra by MultiCASE.

RESULTS

In all models, a large proportion (up to 77%) of all pesticides were outside the chemical space of the model. Analysis of the prediction of remaining pesticides revealed a balanced accuracy of the models between 0.48 and 0.66.

CONCLUSION

Overall, predictions were only meaningful in rare cases and therefore always require evaluation by an expert. The critical factors were the underlying data and determination of molecular similarity, which offer great potential for improvement.

Pluripotent stem cell assays: Modalities and applications for predictive developmental toxicity

This manuscript provides a review focused on embryonic stem cell-based models and their place within the landscape of alternative developmental toxicity assays. Against the background of the principles of developmental toxicology, the wide diversity of alternative methods using pluripotent stem cells developed in this area over the past half century is reviewed. In order to provide an overview of available models, a systematic scoping review was conducted following a published protocol with inclusion criteria, which were applied to select the assays. Critical aspects including biological domain, readout endpoint, availability of standardized protocols, chemical domain, reproducibility and predictive power of each assay are described in detail, in order to review the applicability and limitations of the platform in general and progress moving forward to implementation. The horizon of innovative routes of promoting regulatory implementation of alternative methods is scanned, and recommendations for further work are given.

Predicting Prenatal Developmental Toxicity Based On the Combination of Chemical Structures and Biological Data

For hazard identification, classification, and labeling purposes, animal testing guidelines are required by law to evaluate the developmental toxicity potential of new and existing chemical products. However, guideline developmental toxicity studies are costly, time-consuming, and require many laboratory animals. Computational modeling has emerged as a promising, animal-sparing, and cost-effective method for evaluating the developmental toxicity potential of chemicals, such as endocrine disruptors, without the use of animals. We aimed to develop a predictive and explainable computational model for developmental toxicants. To this end, a comprehensive dataset of 1244 chemicals with developmental toxicity classifications was curated from public repositories and literature sources. Data from 2140 toxicological high-throughput screening assays were extracted from PubChem and the ToxCast program for this dataset and combined with information about 834 chemical fragments to group assays based on their chemical-mechanistic relationships. This effort revealed two assay clusters containing 83 and 76 assays, respectively, with high positive predictive rates for developmental toxicants identified with animal testing guidelines (PPV = 72.4 and 77.3% during cross-validation). These two assay clusters can be used as developmental toxicity models and were applied to predict new chemicals for external validation. This study provides a new strategy for constructing alternative chemical developmental toxicity evaluations that can be replicated for other toxicity modeling studies.

Optimization of the TeraTox assay for preclinical teratogenicity assessment

Current animal-free methods to assess teratogenicity of drugs under development still deliver high numbers of false negatives. To improve the sensitivity of human teratogenicity prediction, we characterized the TeraTox test, a newly developed multilineage differentiation assay using 3D human-induced pluripotent stem cells. TeraTox produces primary output concentration-dependent cytotoxicity and altered gene expression induced by each test compound. These data are fed into an interpretable machine-learning model to perform prediction, which relates to the concentration-dependent human teratogenicity potential of drug candidates. We applied TeraTox to profile 33 approved pharmaceuticals and 12 proprietary drug candidates with known in vivo data. Comparing TeraTox predictions with known human or animal toxicity, we report an accuracy of 69% (specificity: 53%, sensitivity: 79%). TeraTox performed better than 2 quantitative structure-activity relationship models and had a higher sensitivity than the murine embryonic stem cell test (accuracy: 58%, specificity: 76%, and sensitivity: 46%) run in the same laboratory. The overall prediction accuracy could be further improved by combining TeraTox and mouse embryonic stem cell test results. Furthermore, patterns of altered gene expression revealed by TeraTox may help grouping toxicologically similar compounds and possibly deducing common modes of action. The TeraTox assay and the dataset described here therefore represent a new tool and a valuable resource for drug teratogenicity assessment.

ZZS similarity tool: The online tool for similarity screening to identify chemicals of potential concern

Screening and prioritization of chemicals is essential to ensure that available evaluation capacity is invested in those substances that are of highest concern. We, therefore, recently developed structural similarity models that evaluate the structural similarity of substances with unknown properties to known Substances of Very High Concern (SVHC), which could be an indication of comparable effects. In the current study the performance of these models is improved by (1) separating known SVHCs in more specific subgroups, (2) (re‐)optimizing similarity models for the various SVHC‐subgroups, and (3) improving interpretability of the predicted outcomes by providing a confidence score. The improvements are directly incorporated in a freely accessible web‐based tool, named the ZZS similarity tool: https://rvszoeksysteem.rivm.nl/ZzsSimilarityTool. Accordingly, this tool can be used by risk assessors, academia and industrial partners to screen and prioritize chemicals for further action and evaluation within varying frameworks, and could support the identification of tomorrow's substances of concern.

Dose Addition in the Induction of Craniofacial Malformations in Zebrafish Embryos Exposed to a Complex Mixture of Food-Relevant Chemicals with Dissimilar Modes of Action

BACKGROUND

Humans are exposed to combinations of chemicals. In cumulative risk assessment (CRA), regulatory bodies such as the European Food Safety Authority consider dose addition as a default and sufficiently conservative approach. The principle of dose addition was confirmed previously for inducing craniofacial malformations in zebrafish embryos in binary mixtures of chemicals with either similar or dissimilar modes of action (MOAs).

OBJECTIVES

In this study, we explored a workflow to select and experimentally test multiple compounds as a complex mixture with each of the compounds at or below its no observed adverse effect level (NOAEL), in the same zebrafish embryo model.

METHODS

Selection of candidate compounds that potentially induce craniofacial malformations was done using in silico methods-structural similarity, molecular docking, and quantitative structure-activity relationships-applied to a database of chemicals relevant for oral exposure in humans via food (EuroMix inventory, n = 1,598 ). A final subselection was made manually to represent different regulatory fields (e.g., food additives, industrial chemicals, plant protection products), different chemical families, and different MOAs.

RESULTS

A final selection of eight compounds was examined in the zebrafish embryo model, and craniofacial malformations were observed in embryos exposed to each of the compounds, thus confirming the developmental toxicity as predicted by the in silico methods. When exposed to a mixture of the eight compounds, each at its NOAEL, substantial craniofacial malformations were observed; according to a dose-response analysis, even embryos exposed to a 7-fold dilution of this mixture still exhibited a slight abnormal phenotype. The cumulative effect of the compounds in the mixture was in accordance with dose addition (added doses of the individual compounds after adjustment for relative potencies), despite different MOAs of the compounds involved.

DISCUSSION

This case study of a complex mixture inducing craniofacial malformations in zebrafish embryos shows that dose addition can adequately predicted the cumulative effect of a mixture of multiple substances at low doses, irrespective of the (expected) MOA. The applied workflow may be useful as an approach for CRA in general. https://doi.org/10.1289/EHP9888.

Beyond AOPs: A Mechanistic Evaluation of NAMs in DART Testing

New Approach Methodologies (NAMs) promise to offer a unique opportunity to enable human-relevant safety decisions to be made without the need for animal testing in the context of exposure-driven Next Generation Risk Assessment (NGRA). Protecting human health against the potential effects a chemical may have on embryo-foetal development and/or aspects of reproductive biology using NGRA is particularly challenging. These are not single endpoint or health effects and risk assessments have traditionally relied on data from Developmental and Reproductive Toxicity (DART) tests in animals. There are numerous Adverse Outcome Pathways (AOPs) that can lead to DART, which means defining and developing strict testing strategies for every AOP, to predict apical outcomes, is neither a tenable goal nor a necessity to ensure NAM-based safety assessments are fit-for-purpose. Instead, a pragmatic approach is needed that uses the available knowledge and data to ensure NAM-based exposure-led safety assessments are sufficiently protective. To this end, the mechanistic and biological coverage of existing NAMs for DART were assessed and gaps to be addressed were identified, allowing the development of an approach that relies on generating data relevant to the overall mechanisms involved in human reproduction and embryo-foetal development. Using the knowledge of cellular processes and signalling pathways underlying the key stages in reproduction and development, we have developed a broad outline of endpoints informative of DART. When the existing NAMs were compared against this outline to determine whether they provide comprehensive coverage when integrated in a framework, we found them to generally cover the reproductive and developmental processes underlying the traditionally evaluated apical endpoint studies. The application of this safety assessment framework is illustrated using an exposure-led case study.

Exploration of the DARTable Genome- a Resource Enabling Data-Driven NAMs for Developmental and Reproductive Toxicity Prediction

The identification of developmental and reproductive toxicity (DART) is a critical component of toxicological evaluations of chemical safety. Adverse Outcome Pathways (AOPs) provide a framework to describe biological processes leading to a toxic effect and can provide insights in understanding the mechanisms underlying toxicological endpoints and aid the development of new approach methods (NAMs). Integrated approaches to testing and assessment (IATA) can be developed based on AOP knowledge and can serve as pragmatic approaches to chemical hazard characterization using NAMs. However, DART effects remain difficult to predict given the diversity of biological mechanisms operating during ontogenesis and consequently, the considerable number of potential molecular initiating events (MIEs) that might trigger a DART Adverse Outcome (DART AO). Consequently, two challenges that need to be overcome to create an AOP-based DART IATA are having sufficient knowledge of relevant biology and using this knowledge to determine the appropriate selection of cell systems that provide sufficient coverage of that biology. The wealth of modern biological and bioinformatics data can be used to provide this knowledge. Here we demonstrate the utility of bioinformatics analyses to address these questions. We integrated known DART MIEs with gene-developmental phenotype information to curate the hypothetical human DARTable genome (HDG, ∼5 k genes) which represents the comprehensive set of biomarkers for DART. Using network analysis of the human interactome, we show that HDG genes have distinct connectivity compared to other genes. HDG genes have higher node degree with lower neighborhood connectivity, betweenness centralities and average shortest path length. Therefore, HDG is highly connected to itself and to the wider network and not only to their local community. Also, by comparison with the Druggable Genome we show how the HDG can be prioritized to identify potential MIEs based on potential to interact with small molecules. We demonstrate how the HDG in combination with gene expression data can be used to select a panel of relevant cell lines (RD-1, OVCAR-3) for inclusion in an IATA and conclude that bioinformatic analyses can provide the necessary insights and serve as a resource for the development of a screening panel for a DART IATA.

In Silico Models for Developmental Toxicity

Modeling developmental toxicity has been a challenge for (Q)SAR model developers due to the complexity of the endpoint. Recently, some new in silico methods have been developed introducing the possibility to evaluate the integration of existing methods by taking advantage of various modeling perspectives. It is important that the model user is aware of the underlying basis of the different models in general, as well as the considerations and assumptions relative to the specific predictions that are obtained from these different models for the same chemical. The evaluation on the predictions needs to be done on a case-by-case basis, checking the analogues (possibly using structural, physicochemical, and toxicological information); for this purpose, the assessment of the applicability domain of the models provides further confidence in the model prediction. In this chapter, we present some examples illustrating an approach to combine human-based rules and statistical methods to support the prediction of developmental toxicity; we also discuss assumptions and uncertainties of the methodology.