Evaluating opportunities for advancing the use of alternative methods in risk assessment through the development of fit-for-purpose in vitro assays

Alternatives to animal testing in toxicity testing: Current status and future perspectives in food safety assessments

The development of alternative methods to animal testing has gained great momentum since Russel and Burch introduced the "3Rs" concept of Reduction, Refinement, and Replacement of animals in safety testing in 1959. Several alternatives to animal testing have since been introduced, including but not limited to in vitro and in chemico test systems, in silico models, and computational models (e.g., [quantitative] structural activity relationship models, high-throughput screens, organ-on-chip models, and genomics or bioinformatics) to predict chemical toxicity. Furthermore, several agencies have developed robust integrated testing strategies to determine chemical toxicity. The cosmetics sector is pioneering the adoption of alternative methodologies for safety evaluations, and other sectors are aiming to completely abandon animal testing by 2035. However, beyond the use of in vitro genetic testing, agencies regulating the food industry have been slow to implement alternative methodologies into safety evaluations compared with other sectors; setting health-based guidance values for food ingredients requires data from systemic toxicity, and to date, no standalone validated alternative models to assess systemic toxicity exist. The abovementioned models show promise for assessing systemic toxicity with further research. In this paper, we review the current alternatives and their applicability and limitations in food safety evaluations.

Polycyclic aromatic hydrocarbons (PAHs): Updated aspects of their determination, kinetics in the human body, and toxicity

Polycyclic aromatic hydrocarbons (PAHs) are legacy pollutants of considerable public health concern. Polycyclic aromatic hydrocarbons arise from natural and anthropogenic sources and are ubiquitously present in the environment. Several PAHs are highly toxic to humans with associated carcinogenic and mutagenic properties. Further, more severe harmful effects on human- and environmental health have been attributed to the presence of high molecular weight (HMW) PAHs, that is PAHs with molecular mass greater than 300 Da. However, more research has been conducted using low molecular weight (LMW) PAHs). In addition, no HMW PAHs are on the priority pollutants list of the United States Environmental Protection Agency (US EPA), which is limited to only 16 PAHs. However, limited analytical methodologies for separating and determining HMW PAHs and their potential isomers and lack of readily available commercial standards make research with these compounds challenging. Since most of the PAH kinetic data originate from animal studies, our understanding of the effects of PAHs on humans is still minimal. In addition, current knowledge of toxic effects after exposure to PAHs may be underrepresented since most investigations focused on exposure to a single PAH. Currently, information on PAH mixtures is limited. Thus, this review aims to critically assess the current knowledge of PAH chemical properties, their kinetic disposition, and toxicity to humans. Further, future research needs to improve and provide the missing information and minimize PAH exposure to humans.

Comparison of Rat Hepatocyte 2D-Monocultures and Hepatocytes Non-Parenchymal Cell Co-Cultures for Assessing Chemical Toxicity

Liver responses are the most common endpoints used as the basis for setting exposure standards. Liver hepatocytes play a vital role in biotransformation of xenobiotics, but non-parenchymal cells (NPCs) in the liver are also involved in certain liver responses. Development of in vitro systems that more faithfully capture liver responses to reduce reliance on animals is a major focus of New Approach Methodology (NAMs). Since rodent regulatory studies are frequently the sole source safety assessment data, mode-of-action data, and used for risk assessments, in vitro rodent models that reflect in vivo responses need to be developed to reduce reliance on animal models. In the work presented in this paper, we developed a 2-D hepatocyte monoculture and 2-D liver cell co-culture system using rat liver cells. These models were assessed for conditions for short-term stability of the cultures and phenotypic and transcriptomic responses of 2 prototypic hepatotoxicants compounds - acetaminophen and phenobarbital. The optimized multi-cellular 2-D culture required use of freshly prepared hepatocytes and NPCs from a single rat, a 3:1 ratio of hepatocytes to NPCs and growth medium using 50% Complete Williams E medium (WEM) and 50% Endothelial Cell Medium (ECM). The transcriptomic responses of the 2 model systems to PB were compared to previous studies from TG-Gates on the gene expression changes in intact rats and the co-culture model responses were more representative of the in vivo responses. Transcriptomic read-outs promise to move beyond conventional phenotypic evaluations with these in vitro NAMs and provide insights about modes of action.

Estimating provisional margins of exposure for data-poor chemicals using high-throughput computational methods

Current computational technologies hold promise for prioritizing the testing of the thousands of chemicals in commerce. Here, a case study is presented demonstrating comparative risk-prioritization approaches based on the ratio of surrogate hazard and exposure data, called margins of exposure (MoEs). Exposures were estimated using a U.S. EPA’s ExpoCast predictive model (SEEM3) results and estimates of bioactivity were predicted using: 1) Oral equivalent doses (OEDs) derived from U.S. EPA’s ToxCast high-throughput screening program, together with in vitro to in vivo extrapolation and 2) thresholds of toxicological concern (TTCs) determined using a structure-based decision-tree using the Toxtree open source software. To ground-truth these computational approaches, we compared the MoEs based on predicted noncancer TTC and OED values to those derived using the traditional method of deriving points of departure from no-observed adverse effect levels (NOAELs) from in vivo oral exposures in rodents. TTC-based MoEs were lower than NOAEL-based MoEs for 520 out of 522 (99.6%) compounds in this smaller overlapping dataset, but were relatively well correlated with the same (r² = 0.59). TTC-based MoEs were also lower than OED-based MoEs for 590 (83.2%) of the 709 evaluated chemicals, indicating that TTCs may serve as a conservative surrogate in the absence of chemical-specific experimental data. The TTC-based MoE prioritization process was then applied to over 45,000 curated environmental chemical structures as a proof-of-concept for high-throughput prioritization using TTC-based MoEs. This study demonstrates the utility of exploiting existing computational methods at the pre-assessment phase of a tiered risk-based approach to quickly, and conservatively, prioritize thousands of untested chemicals for further study.

NAM-based Prediction of Point-of-contact Toxicity in the Lung: A Case Example With 1,3-dichloropropene

Time, cost, ethical, and regulatory considerations surrounding in vivo testing methods render them insufficient to meet existing and future chemical safety testing demands. There is a need for the development of in vitro and in silico alternatives to replace traditional in vivo methods for inhalation toxicity assessment. Exposures of differentiated airway epithelial cultures to gases or aerosols at the air-liquid interface (ALI) can assess tissue responses and in vitro to in vivo extrapolation can align in vitro exposure levels with in-life exposures expected to give similar tissue exposures. Because the airway epithelium varies along its length, with various regions composed of different cell types, we have introduced a known toxic vapor to five human-derived, differentiated, in vitro airway epithelial cell culture models-MucilAir of nasal, tracheal, or bronchial origin, SmallAir, and EpiAlveolar-representing five regions of the airway epithelium-nasal, tracheal, bronchial, bronchiolar, and alveolar. We have monitored toxicity in these cultures 24hours after acute exposure using an assay for transepithelial conductance (for epithelial barrier integrity) and the lactate dehydrogenase (LDH) release assay (for cytotoxicity). Our vapor of choice in these experiments was 1,3-dichloropropene (1,3-DCP). Finally, we have developed an airway dosimetry model for 1,3-DCP vapor to predict in vivo external exposure scenarios that would produce toxic local tissue concentrations as determined by in vitro experiments. Measured in vitro points of departure (PoDs) for all tested cell culture models were similar. Calculated rat equivalent inhaled concentrations varied by model according to position of the modeled tissue within the airway, with nasal respiratory tissue being the most proximal and most sensitive tissue, and alveolar epithelium being the most distal and least sensitive tissue. These predictions are qualitatively in accordance with empirically determined in vivo PoDs. The predicted PoD concentrations were close to, but slightly higher than, PoDs determined by in vivo subchronic studies.

In vitro airway models from mice, rhesus macaques, and humans maintain species differences in xenobiotic metabolism and cellular responses to naphthalene

The translational value of high-throughput toxicity testing will depend on pharmacokinetic validation. Yet, popular in vitro airway epithelia models were optimized for structure and mucociliary function without considering the bioactivation or detoxification capabilities of lung-specific enzymes. This study evaluated xenobiotic metabolism maintenance within differentiated air-liquid interface (ALI) airway epithelial cell cultures (human bronchial; human, rhesus, and mouse tracheal), isolated airway epithelial cells (human, rhesus, and mouse tracheal; rhesus bronchial), and ex vivo microdissected airways (rhesus and mouse) by measuring gene expression, glutathione content, and naphthalene metabolism. Glutathione levels and detoxification gene transcripts were measured after 1-h exposure to 80 µM naphthalene (a bioactivated toxicant) or reactive naphthoquinone metabolites. Glutathione and glutathione-related enzyme transcript levels were maintained in ALI cultures from all species relative to source tissues, while cytochrome P450 monooxygenase gene expression declined. Notable species differences among the models included a 40-fold lower total glutathione content for mouse ALI trachea cells relative to human and rhesus; a higher rate of naphthalene metabolism in mouse ALI cultures for naphthalene-glutathione formation (100-fold over rhesus) and naphthalene-dihydrodiol production (10-fold over human); and opposite effects of 1,2-naphthoquinone exposure in some models-glutathione was depleted in rhesus tissue but rose in mouse ALI samples. The responses of an immortalized bronchial cell line to naphthalene and naphthoquinones were inconsistent with those of human ALI cultures. These findings of preserved species differences and the altered balance of phase I and phase II xenobiotic metabolism among the characterized in vitro models should be considered for future pulmonary toxicity testing.

IVIVE: Facilitating the Use of In Vitro Toxicity Data in Risk Assessment and Decision Making

During the past few decades, the science of toxicology has been undergoing a transformation from observational to predictive science. New approach methodologies (NAMs), including in vitro assays, in silico models, read-across, and in vitro to in vivo extrapolation (IVIVE), are being developed to reduce, refine, or replace whole animal testing, encouraging the judicious use of time and resources. Some of these methods have advanced past the exploratory research stage and are beginning to gain acceptance for the risk assessment of chemicals. A review of the recent literature reveals a burst of IVIVE publications over the past decade. In this review, we propose operational definitions for IVIVE, present literature examples for several common toxicity endpoints, and highlight their implications in decision-making processes across various federal agencies, as well as international organizations, including those in the European Union (EU). The current challenges and future needs are also summarized for IVIVE. In addition to refining and reducing the number of animals in traditional toxicity testing protocols and being used for prioritizing chemical testing, the goal to use IVIVE to facilitate the replacement of animal models can be achieved through their continued evolution and development, including a strategic plan to qualify IVIVE methods for regulatory acceptance.

A face-to-face comparison of claudin-5 transduced human brain endothelial (hCMEC/D3) cells with porcine brain endothelial cells as blood-brain barrier models for drug transport studies

BACKGROUND

Predictive in vitro models of the human blood-brain barrier (BBB) are essential in early drug discovery and development. Among available immortalized human brain capillary endothelial cell lines (BCECs), the hCMEC/D3 cell line has become the most widely used in vitro BBB model. However, monolayers of hCMEC/D3 cells form only moderately restrictive barriers, most likely because the major tight junction protein, claudin-5, is markedly downregulated. Thus, hCMEC/D3 monolayers cannot be used for vectorial drug transport experiments, which is a major disadvantage of this model.

METHODS

Here we transduced hCMEC/D3 cells with a claudin-5 plasmid and compared the characteristics of these cells with those of hCMEC/D3 wildtype cells and primary cultured porcine BCECs.

RESULTS

The claudin-5 transduced hCMEC/D3 exhibited expression levels (and junctional localization) of claudin-5 similar to those of primary cultured porcine BCECs. The transduced cells exhibited increased TEER values (211 Ω cm²) and reduced paracellular mannitol permeability (8.06%/h), indicating improved BBB properties; however, the barrier properties of porcine BCECs (TEER 1650 Ω cm²; mannitol permeability 3.95%/h) were not reached. Hence, vectorial transport of a selective P-glycoprotein substrate (N-desmethyl-loperamide) was not observed in claudin-5 transduced hCMEC/D3 (or wildtype) cells, whereas such drug transport occurred in porcine BCECs.

CONCLUSIONS

The claudin-5 transduced hCMEC/D3 cells provide a tool to studying the contribution of claudin-5 to barrier tightness and how this can be further enhanced by additional transfections or other manipulations of this widely used in vitro model of the BBB.

Novel biomarker genes which distinguish between smokers and chronic obstructive pulmonary disease patients with machine learning approach

Background

Chronic obstructive pulmonary disease (COPD) is combination of progressive lung diseases. The diagnosis of COPD is generally based on the pulmonary function testing, however, difficulties underlie in prognosis of smokers or early stage of COPD patients due to the complexity and heterogeneity of the pathogenesis. Computational analyses of omics technologies are expected as one of the solutions to resolve such complexities.

Methods

We obtained transcriptomic data by in vitro testing with exposures of human bronchial epithelial cells to the inducers for early events of COPD to identify the potential descriptive marker genes. With the identified genes, the machine learning technique was employed with the publicly available transcriptome data obtained from the lung specimens of COPD and non-COPD patients to develop the model that can reflect the risk continuum across smoking and COPD.

Results

The expression levels of 15 genes were commonly altered among in vitro tissues exposed to known inducible factors for earlier events of COPD (exposure to cigarette smoke, DNA damage, oxidative stress, and inflammation), and 10 of these genes and their corresponding proteins have not previously reported as COPD biomarkers. Although these genes were able to predict each group with 65% accuracy, the accuracy with which they were able to discriminate COPD subjects from smokers was only 29%.

Furthermore, logistic regression enabled the conversion of gene expression levels to a numerical index, which we named the “potential risk factor (PRF)” index. The highest significant index value was recorded in COPD subjects (0.56 at the median), followed by smokers (0.30) and non-smokers (0.02). In vitro tissues exposed to cigarette smoke displayed dose-dependent increases of PRF, suggesting its utility for prospective risk estimation of tobacco products.

Conclusions

Our experimental-based transcriptomic analysis identified novel genes associated with COPD, and the 15 genes could distinguish smokers and COPD subjects from non-smokers via machine-learning classification with remarkable accuracy. We also suggested a PRF index that can quantitatively reflect the risk continuum across smoking and COPD pathogenesis, and we believe it will provide an improved understanding of smoking effects and new insights into COPD.

The role of fit-for-purpose assays within tiered testing approaches: A case study evaluating prioritized estrogen-active compounds in an in vitro human uterotrophic assay

Chemical risk assessment relies on toxicity tests that require significant numbers of animals, time and costs. For the >30,000 chemicals in commerce, the current scale of animal testing is insufficient to address chemical safety concerns as regulatory and product stewardship considerations evolve to require more comprehensive understanding of potential biological effects, conditions of use, and associated exposures. We demonstrate the use of a multi-level new approach methodology (NAMs) strategy for hazard- and risk-based prioritization to reduce animal testing. A Level 1/2 chemical prioritization based on estrogen receptor (ER) activity and metabolic activation using ToxCast data was used to select 112 chemicals for testing in a Level 3 human uterine cell estrogen response assay (IKA assay). The Level 3 data were coupled with quantitative in vitro to in vivo extrapolation (Q-IVIVE) to support bioactivity determination (as a surrogate for hazard) in a tissue-specific context. Assay AC₅₀s and Q-IVIVE were used to estimate human equivalent doses (HEDs), and HEDs were compared to rodent uterotrophic assay in vivo-derived points of departure (PODs). For substances active both in vitro and in vivo, IKA assay-derived HEDs were lower or equivalent to in vivo PODs for 19/23 compounds (83%). Activity exposure relationships were calculated, and the IKA assay was as or more protective of human health than the rodent uterotrophic assay for all IKA-positive compounds. This study demonstrates the utility of biologically relevant fit-for-purpose assays and supports the use of a multi-level strategy for chemical risk assessment.

Workshop on the validation and regulatory acceptance of innovative 3R approaches in regulatory toxicology - Evolution versus revolution

At a joint workshop organized by RIVM and BfR, international experts from governmental institutes, regulatory agencies, industry, academia and animal welfare organizations discussed and provided recommendations for the development, validation and implementation of innovative 3R approaches in regulatory toxicology. In particular, an evolutionary improvement of our current approach of test method validation in the context of defined approaches or integrated testing strategies was discussed together with a revolutionary approach based on a comprehensive description of the physiological responses of the human body to chemical exposure and the subsequent definition of relevant and predictive in vitro, in chemico or in silico methods. A more comprehensive evaluation of biological relevance, scientific validity and regulatory purpose of new test methods and assessment strategies together with case studies that provide practical experience with new approaches were discussed as essential steps to build up the necessary confidence to facilitate regulatory acceptance.

Bridging the Data Gap From in vitro Toxicity Testing to Chemical Safety Assessment Through Computational Modeling

Chemical toxicity testing is moving steadily toward a human cell and organoid-based in vitro approach for reasons including scientific relevancy, efficiency, cost, and ethical rightfulness. Inferring human health risk from chemical exposure based on in vitro testing data is a challenging task, facing various data gaps along the way. This review identifies these gaps and makes a case for the in silico approach of computational dose-response and extrapolation modeling to address many of the challenges. Mathematical models that can mechanistically describe chemical toxicokinetics (TK) and toxicodynamics (TD), for both in vitro and in vivo conditions, are the founding pieces in this regard. Identifying toxicity pathways and in vitro point of departure (PoD) associated with adverse health outcomes requires an understanding of the molecular key events in the interacting transcriptome, proteome, and metabolome. Such an understanding will in turn help determine the sets of sensitive biomarkers to be measured in vitro and the scope of toxicity pathways to be modeled in silico. In vitro data reporting both pathway perturbation and chemical biokinetics in the culture medium serve to calibrate the toxicity pathway and virtual tissue models, which can then help predict PoDs in response to chemical dosimetry experienced by cells in vivo. Two types of in vitro to in vivo extrapolation (IVIVE) are needed. (1) For toxic effects involving systemic regulations, such as endocrine disruption, organism-level adverse outcome pathway (AOP) models are needed to extrapolate in vitro toxicity pathway perturbation to in vivo PoD. (2) Physiologically-based toxicokinetic (PBTK) modeling is needed to extrapolate in vitro PoD dose metrics into external doses for expected exposure scenarios. Linked PBTK and TD models can explore the parameter space to recapitulate human population variability in response to chemical insults. While challenges remain for applying these modeling tools to support in vitro toxicity testing, they open the door toward population-stratified and personalized risk assessment.

OPERA models for predicting physicochemical properties and environmental fate endpoints

The collection of chemical structure information and associated experimental data for quantitative structure–activity/property relationship (QSAR/QSPR) modeling is facilitated by an increasing number of public databases containing large amounts of useful data. However, the performance of QSAR models highly depends on the quality of the data and modeling methodology used. This study aims to develop robust QSAR/QSPR models for chemical properties of environmental interest that can be used for regulatory purposes. This study primarily uses data from the publicly available PHYSPROP database consisting of a set of 13 common physicochemical and environmental fate properties. These datasets have undergone extensive curation using an automated workflow to select only high-quality data, and the chemical structures were standardized prior to calculation of the molecular descriptors. The modeling procedure was developed based on the five Organization for Economic Cooperation and Development (OECD) principles for QSAR models. A weighted k-nearest neighbor approach was adopted using a minimum number of required descriptors calculated using PaDEL, an open-source software. The genetic algorithms selected only the most pertinent and mechanistically interpretable descriptors (2–15, with an average of 11 descriptors). The sizes of the modeled datasets varied from 150 chemicals for biodegradability half-life to 14,050 chemicals for logP, with an average of 3222 chemicals across all endpoints. The optimal models were built on randomly selected training sets (75%) and validated using fivefold cross-validation (CV) and test sets (25%). The CV Q² of the models varied from 0.72 to 0.95, with an average of 0.86 and an R² test value from 0.71 to 0.96, with an average of 0.82. Modeling and performance details are described in QSAR model reporting format and were validated by the European Commission’s Joint Research Center to be OECD compliant. All models are freely available as an open-source, command-line application called OPEn structure–activity/property Relationship App (OPERA). OPERA models were applied to more than 750,000 chemicals to produce freely available predicted data on the U.S. Environmental Protection Agency’s CompTox Chemistry Dashboard.