Semi-automated approach for generation of biological networks on drug-induced cholestasis, steatosis, hepatitis, and cirrhosis.

Drug-induced liver injury (DILI) is one of the leading reasons for discontinuation of a new drug development project. Diverse machine learning or deep learning models have been developed to predict DILI. However, these models have not provided an adequate understanding of the mechanisms leading to DILI. The development of safer drugs requires novel computational approaches that enable the prompt understanding of the mechanism of DILI. In this study, the mechanisms leading to the development of cholestasis, steatosis, hepatitis, and cirrhosis were explored using a semi-automated approach for data gathering and associations. Diverse data from ToxCast, Comparative Toxicogenomic Database (CTD), Reactome, and Open TG-GATEs on reference molecules leading to the development of the respective diseases were extracted. The data were used to create biological networks of the four diseases. As expected, the four networks had several common pathways, and a joint DILI network was assembled. Such biological networks could be used in drug discovery to identify possible molecules of concern as they provide a better understanding of the disease-specific key events. The events can be target-tested to provide indications for potential DILI effects. Supplementary Information: The online version contains supplementary material available at 10.1007/s43188-022-00124-6.

How to Foster 'New Approach Methodology' Toxicologists.

The need to reduce, refine and replace animal experimentation has led to a boom in the establishment of new approach methodologies (NAMs). This promising trend brings the hope that the replacement of animals by using NAMs will become increasingly accepted by regulators, included in legislation, and consequently more-often implemented by industry. The majority of NAMs, however, are still not very well understood, either due to the complexity of the applied approach or the data analysis workflow. A potential solution to this problem is the provision of better educational resources to scientists new to the area - showcasing the added value of NAMs and outlining various ways of overcoming issues associated with knowledge gaps. In this paper, the educational exchange between four institutions - namely, two universities and two SMEs - via a series of video training sessions, is described. The goal of this exchange was to showcase an exemplary event to help introduce scientists to non-animal approaches, and to actively support the development of resources enabling the use of alternatives to laboratory animals.

Skin sensitization in silico protocol.

The assessment of skin sensitization has evolved over the past few years to include in vitro assessments of key events along the adverse outcome pathway and opportunistically capitalize on the strengths of in silico methods to support a weight of evidence assessment without conducting a test in animals. While in silico methods vary greatly in their purpose and format; there is a need to standardize the underlying principles on which such models are developed and to make transparent the implications for the uncertainty in the overall assessment. In this contribution, the relationship between skin sensitization relevant effects, mechanisms, and endpoints are built into a hazard assessment framework. Based on the relevance of the mechanisms and effects as well as the strengths and limitations of the experimental systems used to identify them, rules and principles are defined for deriving skin sensitization in silico assessments. Further, the assignments of reliability and confidence scores that reflect the overall strength of the assessment are discussed. This skin sensitization protocol supports the implementation and acceptance of in silico approaches for the prediction of skin sensitization.

A semi-automated workflow for adverse outcome pathway hypothesis generation: The use case of non-genotoxic induced hepatocellular carcinoma.

The utility of the Adverse Outcome Pathway (AOP) concept has been largely recognized by scientists, however, the AOP generation is still mainly done manually by screening through evidence and extracting probable associations. To accelerate this process and increase the reliability, we have developed an semi-automated workflow for AOP hypothesis generation. In brief, association mining methods were applied to high-throughput screening, gene expression, in vivo and disease data present in ToxCast and Comparative Toxicogenomics Database. This was supplemented by pathway mapping using Reactome to fill in gaps and identify events occurring at the cellular/tissue levels. Furthermore, in vivo data from TG-Gates was integrated to finally derive a gene, pathway, biochemical, histopathological and disease network from which specific disease sub-networks can be queried. To test the workflow, non-genotoxic-induced hepatocellular carcinoma (HCC) was selected as a case study. The implementation resulted in the identification of several non-genotoxic-specific HCC-connected genes belonging to cell proliferation, endoplasmic reticulum stress and early apoptosis. Biochemical findings revealed non-genotoxic-specific alkaline phosphatase increase. The explored non-genotoxic-specific histopathology was associated with early stages of hepatic steatosis, transforming into cirrhosis. This work illustrates the utility of computationally predicted constructs in supporting development by using pre-existing knowledge in a fast and unbiased manner.

Genetic toxicology in silico protocol.

In silico toxicology (IST) approaches to rapidly assess chemical hazard, and usage of such methods is increasing in all applications but especially for regulatory submissions, such as for assessing chemicals under REACH as well as the ICH M7 guideline for drug impurities. There are a number of obstacles to performing an IST assessment, including uncertainty in how such an assessment and associated expert review should be performed or what is fit for purpose, as well as a lack of confidence that the results will be accepted by colleagues, collaborators and regulatory authorities. To address this, a project to develop a series of IST protocols for different hazard endpoints has been initiated and this paper describes the genetic toxicity in silico (GIST) protocol. The protocol outlines a hazard assessment framework including key effects/mechanisms and their relationships to endpoints such as gene mutation and clastogenicity. IST models and data are reviewed that support the assessment of these effects/mechanisms along with defined approaches for combining the information and evaluating the confidence in the assessment. This protocol has been developed through a consortium of toxicologists, computational scientists, and regulatory scientists across several industries to support the implementation and acceptance of in silico approaches.

Integrated analysis of in vitro data and the adverse outcome pathway framework for prioritization and regulatory applications: An exploratory case study using publicly available data on piperonyl butoxide and liver models.

The integration of existing knowledge to support the risk assessment of chemicals is an ongoing challenge for scientists, risk assessors and risk managers. In addition, European Union regulations limiting the use of new animal testing in cosmetics makes already existing information even more valuable. Applying a previous SEURAT-1 program framework to derive predictions of in vivo toxicity responses for a compound, we selected piperonyl butoxide (PBO) as a case study for identification of knowledge and methodology gaps in understanding a compound's effects on the human liver. This is investigated through integration of data from human in vitro transcriptomics studies, biological pathway analysis, chemical and disease associations, and adverse outcome pathway (AOP) information. The outcomes of the analysis are used to generate AOPs of liver-related endpoints, identifying areas of concern for risk assessors and regulators. We demonstrate that integration of data through already existing and publicly available tools can produce outcomes comparable to those that may be found through more conventional time- and resource-intensive methods. It is also expected that, with more refinement, this approach could in the future provide evidence to support chemical risk assessment, while also identifying data gaps for which additional testing may be needed.

A novel genotoxin-specific qPCR array based on the metabolically competent human HepaRG™ cell line as a rapid and reliable tool for improved in vitro hazard assessment.

Although the value of the regulatory accepted batteries for in vitro genotoxicity testing is recognized, they result in a high number of false positives. This has a major impact on society and industries developing novel compounds for pharmaceutical, chemical, and consumer products, as afflicted compounds have to be (prematurely) abandoned or further tested on animals. Using the metabolically competent human HepaRG™ cell line and toxicogenomics approaches, we have developed an upgraded, innovative, and proprietary gene classifier. This gene classifier is based on transcriptomic changes induced by 12 genotoxic and 12 non-genotoxic reference compounds tested at sub-cytotoxic concentrations, i.e., IC10 concentrations as determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The resulting gene classifier was translated into an easy-to-handle qPCR array that, as shown by pathway analysis, covers several different cellular processes related to genotoxicity. To further assess the predictivity of the tool, a set of 5 known positive and 5 known negative test compounds for genotoxicity was evaluated. In addition, 2 compounds with debatable genotoxicity data were tested to explore how the qPCR array would classify these. With an accuracy of 100%, when equivocal results were considered positive, the results showed that combining HepaRG™ cells with a genotoxin-specific qPCR array can improve (geno)toxicological hazard assessment. In addition, the developed qPCR array was able to provide additional information on compounds for which so far debatable genotoxicity data are available. The results indicate that the new in vitro tool can improve human safety assessment of chemicals in general by basing predictions on mechanistic toxicogenomics information.

The Vitotox and ToxTracker assays: A two-test combination for quick and reliable assessment of genotoxic hazards.

To ensure safety for humans, it is essential to characterize the genotoxic potential of new chemical entities, such as pharmaceutical and cosmetic substances. In a first tier, a battery of in vitro tests is recommended by international regulatory agencies. However, these tests suffer from inadequate specificity: compounds may be wrongly categorized as genotoxic, resulting in unnecessary, time-consuming, and expensive in vivo follow-up testing. In the last decade, novel assays (notably, reporter-based assays) have been developed in an attempt to overcome these drawbacks. Here, we have investigated the performance of two in vitro reporter-based assays, Vitotox and ToxTracker. A set of reference compounds was selected to span a variety of mechanisms of genotoxic action and applicability domains (e.g., pharmaceutical and cosmetic ingredients). Combining the performance of the two assays, we achieved 93% sensitivity and 79% specificity for prediction of gentoxicity for this set of compounds. Both assays permit quick high-throughput analysis of drug candidates, while requiring only small quantities of the test substances. Our study shows that these two assays, when combined, can be a reliable method for assessment of genotoxicity hazard.

In silico tools and transcriptomics analyses in the mutagenicity assessment of cosmetic ingredients: a proof-of-principle on how to add weight to the evidence.

Prior to the downstream development of chemical substances, including pharmaceuticals and cosmetics, their influence on the genetic apparatus has to be tested. Several in vitro and in vivo assays have been developed to test for genotoxicity. In a first tier, a battery of two to three in vitro tests is recommended to cover mutagenicity, clastogenicity and aneugenicity as main endpoints. This regulatory in vitro test battery is known to have a high sensitivity, which is at the expense of the specificity. The high number of false positive in vitro results leads to excessive in vivo follow-up studies. In the case of cosmetics it may even induce the ban of the particular compound since in Europe the use of experimental animals is no longer allowed for cosmetics. In this article, an alternative approach to derisk a misleading positive Ames test is explored. Hereto we first tested the performance of five existing computational tools to predict the potential mutagenicity of a data set of 132 cosmetic compounds with a known genotoxicity profile. Furthermore, we present, as a proof-of-principle, a strategy in which a combination of computational tools and mechanistic information derived from in vitro transcriptomics analyses is used to derisk a misleading positive Ames test result. Our data shows that this strategy may represent a valuable tool in a weight-of-evidence approach to further evaluate a positive outcome in an Ames test.

Linking existing in vitro dermal absorption data to physicochemical properties: Contribution to the design of a weight-of-evidence approach for the safety evaluation of cosmetic ingredients with low dermal bioavailability.

To characterize the risk of cosmetic ingredients when threshold toxicity is assumed, often the "margin of safety" (MoS) is calculated. This uncertainty factor is based on the systemic no observable (adverse) effect level (NO(A)EL) which can be derived from in vivo repeated dose toxicity studies. As in vivo studies for the purpose of the cosmetic legislation are no longer allowed in Europe and a validated in vitro alternative is not yet available, it is no longer possible to derive a NO(A)EL value for a new cosmetic ingredient. Alternatively, cosmetic ingredients with a low dermal bioavailability might not need repeated dose data, as internal exposure will be minimal and systemic toxicity might not be an issue. This study shows the possibility of identifying compounds suspected to have a low dermal bioavailability based on their physicochemical properties (molecular weight, melting point, topological polar surface area and log P) and their in vitro dermal absorption data. Although performed on a limited number of compounds, the study suggests a strategic opportunity to support the safety assessor's reasoning to omit a MoS calculation and to focus more on local toxicity and mutagenicity/genotoxicity for ingredients for which limited systemic exposure is to be expected.

Measurement of Cytochrome P450 Enzyme Induction and Inhibition in Human Hepatoma Cells.

Cytochrome P450 enzymes are a diverse group of catalytic enzymes in the liver that are mainly responsible for the biotransformation of organic substances. Cytochrome P450 activity as well as both its induction and inhibition are key factors in drug biotransformation and can be involved in deactivation, activation, detoxification and toxification processes. Thus, the modulation of cytochrome P450 activity is an important parameter when evaluating the potential toxicity of chemical compounds using an in vitro system. The cytochrome P450 3A subfamily proteins are among the most important drug-metabolizing enzymes in human liver and are responsible for about half of all cytochrome P450-dependent drug oxidations. In vitro, these enzymes are active not only in primary human hepatocyte cultures, but also in differentiated human hepatoma HepaRG cells. The present protocol describes the culture of cryopreserved differentiated HepaRG cells and the evaluation of its cytochrome P450 activity upon exposure to a chemical compound using a commercially available luminogenic cytochrome P450 assay. This in vitro model can be used to monitor the induction and inhibition of cytochrome P450 3A following exposure to a particular test compound.

Retrospective analysis of the mutagenicity/genotoxicity data of the cosmetic ingredients present on the Annexes of the Cosmetic EU legislation (2000-12).

To evaluate the mutagenicity/genotoxicity of cosmetic ingredients at the regulatory level, usually a battery of three in vitro tests is applied. This battery, designed to be very sensitive, produces a high number of positive results, imposing the need for in vivo follow-up testing to clear the substance under study. In Europe, the use of experimental animals has become impossible for cosmetic ingredients due to the implementation of animal testing and marketing bans. Consequently, the possibility to 'de-risk' substances with positive in vitro results disappear and potentially safe cosmetic substances will be lost for the EU market unless currently used in vitro assays can be adapted or new non-animal mutagenicity/genotoxicity studies become available. Described strategies to improve the specificity of existing in vitro assays include optimisation of the used cell type and cytotoxicity assay and lowering of the applied top concentration. A reduction of the number of tests in the battery from three to two also has been suggested. In this study, the performance of the 'standard' in vitro mutagenicity/genotoxicity testing battery is analysed for a number of cosmetic ingredients. We composed a database with toxicological information on 249 cosmetic ingredients, mainly present on the Annexes of the European cosmetic legislation. Results revealed that the in vitro mutagenicity/genotoxicity tests showed a low specificity for the cosmetic ingredients concerned, comparable to the specificity published for chemicals. Non-confirmed or 'misleading' positive results amounted up to 93% for the in vitro test batteries. The cell type and top concentrations did not have a major impact on the specificity. With respect to cytotoxicity determinations, different end points were used, potentially leading to different testing concentrations, suggesting the need for a consensus in this matter. Overall, the results of this retrospective analysis point to an urgent need of better regulatory strategies to assess the potential mutagenicity/genotoxicity of cosmetic ingredients.

Way forward in case of a false positive in vitro genotoxicity result for a cosmetic substance?

The currently used regulatory in vitro mutagenicity/genotoxicity test battery has a high sensitivity for detecting genotoxicants, but it suffers from a large number of irrelevant positive results (i.e. low specificity) thereby imposing the need for additional follow-up by in vitro and/or in vivo genotoxicity tests. This could have a major impact on the cosmetic industry in Europe, seen the imposed animal testing and marketing bans on cosmetics and their ingredients. Afflicted, but safe substances could therefore be lost. Using the example of triclosan, a cosmetic preservative, we describe here the potential applicability of a human toxicogenomics-based in vitro assay as a potential mechanistically based follow-up test for positive in vitro genotoxicity results. Triclosan shows a positive in vitro chromosomal aberration test, but is negative during in vivo follow-up tests. Toxicogenomics analysis unequivocally shows that triclosan is identified as a compound acting through non-DNA reactive mechanisms. This proof-of-principle study illustrates the potential of genome-wide transcriptomics data in combination with in vitro experimentation as a possible weight-of-evidence follow-up approach for de-risking a positive outcome in a standard mutagenicity/genotoxicity battery. As such a substantial number of cosmetic compounds wrongly identified as genotoxicants could be saved for the future.

Transcriptomic responses generated by hepatocarcinogens in a battery of liver-based in vitro models.

As the conventional approach to assess the potential of a chemical to cause cancer in humans still includes the 2-year rodent carcinogenicity bioassay, development of alternative methodologies is needed. In the present study, the transcriptomics responses following exposure to genotoxic (GTX) and non-genotoxic (NGTX) hepatocarcinogens and non-carcinogens (NC) in five liver-based in vitro models, namely conventional and epigenetically stabilized cultures of primary rat hepatocytes, the human hepatoma-derived cell lines HepaRG and HepG2 and human embryonic stem cell-derived hepatocyte-like cells, are examined. For full characterization of the systems, several bioinformatics approaches are employed including gene-based, ConsensusPathDB-based and classification analysis. They provide convincingly similar outcomes, namely that upon exposure to carcinogens, the HepaRG generates a gene classifier (a gene classifier is defined as a selected set of characteristic gene signatures capable of distinguishing GTX, NGTX carcinogens and NC) able to discriminate the GTX carcinogens from the NGTX carcinogens and NC. The other in vitro models also yield cancer-relevant characteristic gene groups for the GTX exposure, but some genes are also deregulated by the NGTX carcinogens and NC. Irrespective of the tested in vitro model, the most uniformly expressed pathways following GTX exposure are the p53 and those that are subsequently induced. The NGTX carcinogens triggered no characteristic cancer-relevant gene profiles in all liver-based in vitro systems. In conclusion, liver-based in vitro models coupled with transcriptomics techniques, especially in the case when the HepaRG cell line is used, represent valuable tools for obtaining insight into the mechanism of action and identification of GTX carcinogens.

Comparison of genotoxicant-modified transcriptomic responses in conventional and epigenetically stabilized primary rat hepatocytes with in vivo rat liver data.

The concept of mechanistic toxicogenomics implies that compound-induced changes in gene expression profiles provide valuable information about their mode of action. A growing number of research groups have presented evidence that whole-genome gene expression profiling techniques might be used as tools for in vivo and in vitro generation of gene signatures and elucidation of molecular mechanisms after exposure to toxic compounds. An important issue to be investigated is the in vivo relevance of in vitro-obtained data. In the current study, we compare the gene expression profiles generated in vitro, after exposing conventional and epigenetically stabilized primary rat hepatocytes to well-known genotoxic hepatocarcinogens (aflatoxin B1, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone and 2-nitrofluorene) with those derived in vivo after oral exposure of rats to these compounds. Similar statistical tools were applied on both sets of data. The major molecular pathways affected in the in vivo setting were DNA damage, detoxification and cell survival response, as previously described. In the conventional hepatocyte cultures, two of the three genotoxicants showed quite similar responses as in vivo with respect to these pathways. The third compound (2-nitrofluorene) revealed in vitro response which was not observed in vivo. In the epigenetically stabilized hepatocytes, in contrast to what was expected, the responses were less relevant for the in vivo situation. This study highlights the importance of in vitro/in vivo comparison of data that are generated using in vitro models and shows that conventional primary rat hepatocyte cultures represent an appropriate in vitro model to retrieve mechanistic information on the exposure to genotoxicants.

Comparison of hepatocarcinogen-induced gene expression profiles in conventional primary rat hepatocytes with in vivo rat liver.

At present, substantial efforts are focused on the development of in vitro assays coupled with "omics" technologies for the identification of carcinogenic substances as an alternative to the classical 2-year rodent carcinogenicity bioassay. A prerequisite for the eventual regulatory acceptance of such assays, however, is the in vivo relevance of the observed in vitro findings. In the current study, hepatocarcinogen-induced gene expression profiles generated after the exposure of conventional cultures of primary rat hepatocytes to three non-genotoxic carcinogens (methapyrilene hydrochloride, piperonyl butoxide, and Wy-14643), three genotoxic carcinogens (aflatoxin B1, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, and 2-nitrofluorene), and two non-carcinogens (nifedipine and clonidine) are compared with previously obtained in vivo data after oral administration for up to 14 days of the same hepatocarcinogens to rats. In addition to the comparison of deregulated genes and functions per compound between in vivo and in vitro models, the major discriminating cellular pathways found in vivo in livers of exposed rats were examined for deregulation in vitro. Further, in vivo-derived gene signatures for the identification of genotoxic versus non-genotoxic carcinogens are used to classify in vitro-tested hepatocarcinogens and non-carcinogens. In the primary hepatocyte cultures, two out of the three tested genotoxic carcinogens mimicked the in vivo-relevant DNA damage response and were correctly assessed. Exposure to the non-genotoxic hepatocarcinogens, however, triggered a relatively weak response in the in vitro system, with no clear similarities to in vivo. This study contributes to the further optimization of toxicogenomics predictive tools when applied in in vitro settings.

Opportunities for an alternative integrating testing strategy for carcinogen hazard assessment?

The 2-year rodent carcinogenicity bioassay evolved more than 40 years ago, and although it is complex, long lasting, expensive, and animal consuming, it is still the only generally accepted test for assessing the carcinogenicity of chemicals. Over time, different alternative approaches have been developed with the final goal to replace the bioassay. Unfortunately, at present, none of these strategies alone provides sufficient assurance of accurate prediction. In this review paper, we discuss the major advantages and pitfalls of the existing alternative methodologies to the carcinogenicity bioassay. Finally, based on the available scientific data in the public domain, we propose what we would like to call a "feasible integrated testing strategy" which incorporates some promising alternatives, providing at the same time information on the mechanism of action and the toxic nature of the compounds tested. It is, however, clear that the adoption of whatever "new" testing scheme should be considered with caution and its effectiveness should be experimentally demonstrated in advance by addressing a reasonable number of chemical carcinogens and non-carcinogens from a variety of structural and functional classes.

Characterization of spontaneous cell death in monolayer cultures of primary hepatocytes.

Monolayer cultures of primary hepatocytes, isolated from freshly removed livers, represent widely used in vitro tools in the area of liver physiology and pathology, pharmacology and toxicology. However, a major shortcoming of these systems is that they cope with dedifferentiation, which is accompanied by spontaneous cell death. The goal of the present study was to elucidate the mechanisms that drive the process of self-generated cell demise in primary hepatocyte cultures. For this purpose, isolated rat hepatocytes were cultivated under conventional conditions, and the occurrence of apoptosis and necrosis was monitored during 4 days by performing a set of acknowledged cell death assays. These included examination of cell morphology by light microscopy, quantification of apoptotic and necrotic cell populations by Hoechst 33342 and propidium iodide in situ staining, assessment of apoptotic and necrotic activities by measuring caspase 3-like activity and extracellular leakage of lactate dehydrogenase, and studying the expression of apoptosis regulators through immunoblot analysis. In essence, two cell death peaks were observed, namely shortly after cell seeding and in the final stages of the cultivation period, both involving apoptotic and necrotic actions. The outcome of this study not only sheds new light onto the molecular processes that underlie spontaneous cell death in primary hepatocyte cultures, but also opens perspectives for the establishment of strategies to increase cell survival in these popular in vitro systems.

Effect of the histone deacetylase inhibitor trichostatin a on the metabolome of cultured primary hepatocytes.

Trichostatin A (TSA) is a histone deacetylase inhibitor that has antiproliferative and differentiation-inducing effects on cancer cells, and in cultures of primary hepatocytes has been shown to maintain xenobiotic metabolic capacity. Using an NMR-based metabolic profiling approach, we evaluated if the endogenous metabolome was stabilized and the normal metabolic phenotype retained in this model. Aqueous soluble metabolites were extracted from isolated rat hepatocytes after 44 and 92 h exposure to TSA (25 muM) together with time-matched controls and measured by (1)H NMR spectroscopy. Multivariate analysis showed a clear difference in the global metabolic profile over time in control samples, while the TSA treated group was more closely clustered at both time points, suggesting that treatment reduced the time related effect on metabolism that was observed in the control. TSA treatment was associated with decreases in glycerophosphocholine, 3-hydroxybutyric acid, glycine and adenosine, an increase in glycogen, and a reduction in the decrease of inosine, hypoxanthine, and glutathione over time. Collectively, our data suggest that TSA treatment reduces the loss of a normal metabolic phenotype in cultured primary hepatocytes, improving the model as a tool to study endogenous liver metabolism, xenobiotic metabolism, and potentially affecting the accuracy of all biological assays in this system.

Gap junctional intercellular communication as a target for liver toxicity and carcinogenicity.

Direct communication between hepatocytes, mediated by gap junctions, constitutes a major regulatory platform in the control of liver homeostasis, ranging from hepatocellular proliferation to hepatocyte cell death. Inherent to this pivotal task, gap junction functionality is frequently disrupted upon impairment of the homeostatic balance, as occurs during liver toxicity and carcinogenicity. In the present paper, the deleterious effects of a number of chemical and biological toxic compounds on hepatic gap junctions are discussed, including environmental pollutants, biological toxins, organic solvents, pesticides, pharmaceuticals, peroxides, metals and phthalates. Particular attention is paid to the molecular mechanisms that underlie the abrogation of gap junction functionality. Since hepatic gap junctions are specifically targeted by tumor promoters and epigenetic carcinogens, both in vivo and in vitro, inhibition of gap junction functionality is considered as a suitable indicator for the detection of nongenotoxic hepatocarcinogenicity.