Equations for estimating binary mixture toxicity: 3-methyl-2-butanone with a series of electrophiles.

Mixture toxicity was determined for 32 binary combinations. One chemical was the non-reactive, non-polar narcotic 3-methyl-2-butanone (always chemical A) and the other was a potentially reactive electrophile (chemical B). Bioluminescence inhibition in Allovibrio fischeri was measured at 15-, 30-, and 45-minutes of exposure for A, B, and the mixture (MX). Concentration-response curves (CRCs) were developed for each chemical and used to develop predicted CRCs for the concentration addition (CA) and independent action (IA) mixture toxicity models. Also, MX CRCs were generated and compared with model predictions using the 45-minute data. Classification of observed mixture toxicity used three specific criteria: 1) predicted IA EC50 vs. CA EC50 values at 45-minutes, 2) consistency of 45-minute MX CRC fit to IA, CA, or otherwise at three effect levels (EC25, EC50 and EC75), and 3) the known/suspected mechanism of toxicity for chemical B. Mixture toxicity was then classified into one of seven groupings. As a result of the predicted IA EC50 being more toxic than the predicted CA EC50, IA represented the greater toxic hazard. For this reason, non-sham MXs having toxicity consistent with CA were classified as being "coincident" with CA rather than mechanistically-consistent with CA. Multiple linear regression analyses were performed to develop equations that can be used to estimate the toxicity of other 3M2B-containing binary mixtures. These equations were developed from the data for both IA and CA, at each exposure duration and effect level. Each equation had a coefficient of determination (r2) above 0.950 and a variance inflation factor <1.2. This approach can potentially reduce the need for mixture testing and is amenable to other model systems and to assays that evaluate toxicity at low effect levels.

Uncertainty in the results from oral repeated dose toxicity tests: Impact on regulatory classifications.

The Lowest Observed (Adverse) Effect Level (LO(A)EL) values are point-of-departure (PoD) values that quantify repeat dose toxicity (RDT). Here, the uncertainty in the regulatory classification of these PoDs is investigated. In the application stage, the dose-response was approximated for a large set of series, giving an account of the possible presence of a hormesis zone. The minimal effect dose (MED) or dose was computed, and the ratio MED/LO(A)EL was used to represent the two components of the experimental uncertainty. The uncertainty estimations were calculated for any combination of gender and reported examination item. Subsequently, how this uncertainty affects the possible classifications was analyzed, and the percentage of the chemicals receiving ambiguous classification was determined. It was shown that more than 40% of the investigated chemicals cannot be classified unambiguously in the Globally Harmonized System (GHS) classification scheme and bear a potential for misclassification when a regulatory decision is based on a single LO(A)EL value. A table containing grey zones for different risk levels and a table with GHS classification distributions for various LO(A)EL values were prepared to facilitate the use of the RDT uncertainty in the practice.

Estimating uncertainty in LLNA EC3 data and its impact on regulatory classifications.

The murine Local Lymph Node Assay (LLNA) is a test that produces numerical results (EC3 values) quantifying the sensitization potency of chemicals. These results are broadly used in toxicology and serve as a basis for various classifications, which determine subsequent regulatory decisions. The continuing interest in LLNA data and the diminished likelihood of new experimental EC3 data being generated sparked this investigation of uncertainty. Instead of using the Gaussian distribution as a default choice for assessing variability in a data set, two strictly positive distributions were proposed and their performance over the available experimental EC3 values was tested. In the application stage, how the uncertainty in EC3 values affects the possible classifications was analyzed, and the percentage of the chemicals receiving ambiguous classification was determined. It was shown that this percentage is high, which increases the risk of improper classification. Two approaches were suggested in regulatory practice to address the uncertainty in the EC3 data: the approaches based on "grey zones" and the classification distribution. If a chemical cannot be classified unambiguously, the latter appears to be an acceptable means to assess the level of sensitization potency of chemicals and helps provide better regulatory decisions.

An End Point-Specific Framework for Read-Across Analog Selection for Human Health Effects.

Integrating computational chemistry and toxicology can improve the read-across analog approach to fill data gaps in chemical safety assessment. In read-across, structure-related parameters are compared between a target chemical with insufficient test data and one or more materials with sufficient data. Recent advances have focused on enhancing the grouping or clustering of chemicals to facilitate toxicity prediction via read-across. Analog selection ascertains relevant features, such as physical-chemical properties, toxicokinetic-related properties (bioavailability, metabolism, and degradation pathways), and toxicodynamic properties of chemicals with an emphasis on mechanisms or modes of action. However, each human health end point (genotoxicity, skin sensitization, phototoxicity, repeated dose toxicity, reproductive toxicity, and local respiratory toxicity) provides a different critical context for analog selection. Here six end point-specific, rule-based schemes are described. Each scheme creates an end point-specific workflow for filling the target material data gap by read-across. These schemes are intended to create a transparent rationale that supports the selected read-across analog(s) for the specific end point under study. This framework can systematically drive the selection of read-across analogs for each end point, thereby accelerating the safety assessment process.

Automated read-across workflow for predicting acute oral toxicity: I. The decision scheme in the QSAR toolbox.

A decision-scheme outlining the steps for identifying the appropriate chemical category and subsequently appropriate tested source analog(s) for data gap filling of a target chemical by read-across is described. The primary features used in the grouping of the target chemical with source analogues within a database of 10,039 discrete organic substances include reactivity mechanisms associated with protein interactions and specific-acute-oral-toxicity-related mechanisms (e.g., mitochondrial uncoupling). Additionally, the grouping of chemicals making use of the in vivo rat metabolic simulator and neutral hydrolysis. Subsequently, a series of structure-based profilers are used to narrow the group to the most similar analogues. The scheme is implemented in the OECD QSAR Toolbox, so it automatically predicts acute oral toxicity as the rat oral LD50 value in log [1/mol/kg]. It was demonstrated that due to the inherent variability in experimental data, classification distribution should be employed as more adequate in comparison to the exact classification. It was proved that the predictions falling in the adjacent GSH categories to the experimentally-stated ones are acceptable given the variation in experimental data. The model performance estimated by adjacent accuracy was found to be 0.89 and 0.54 while based on R2. The mechanistic and predictive coverages were >0.85.

Determination of "fitness-for-purpose" of quantitative structure-activity relationship (QSAR) models to predict (eco-)toxicological endpoints for regulatory use.

In silico models are used to predict toxicity and molecular properties in chemical safety assessment, gaining widespread regulatory use under a number of legislations globally. This study has rationalised previously published criteria to evaluate quantitative structure-activity relationships (QSARs) in terms of their uncertainty, variability and potential areas of bias, into ten assessment components, or higher level groupings. The components have been mapped onto specific regulatory uses (i.e. data gap filling for risk assessment, classification and labelling, and screening and prioritisation) identifying different levels of uncertainty that may be acceptable for each. Twelve published QSARs were evaluated using the components, such that their potential use could be identified. High uncertainty was commonly observed with the presentation of data, mechanistic interpretability, incorporation of toxicokinetics and the relevance of the data for regulatory purposes. The assessment components help to guide strategies that can be implemented to improve acceptability of QSARs through the reduction of uncertainties. It is anticipated that model developers could apply the assessment components from the model design phase (e.g. through problem formulation) through to their documentation and use. The application of the components provides the possibility to assess QSARs in a meaningful manner and demonstrate their fitness-for-purpose against pre-defined criteria.

Clustering a Chemical Inventory for Safety Assessment of Fragrance Ingredients: Identifying Read-Across Analogs to Address Data Gaps.

A valuable approach to chemical safety assessment is the use of read-across chemicals to provide safety data to support the assessment of structurally similar chemicals. An inventory of over 6000 discrete organic chemicals used as fragrance materials in consumer products has been clustered into chemical class-based groups for efficient search of read-across sources. We developed a robust, tiered system for chemical classification based on (1) organic functional group, (2) structural similarity and reactivity features of the hydrocarbon skeletons, (3) predicted or experimentally verified Phase I and Phase II metabolism, and (4) expert pruning to consider these variables in the context of specific toxicity end points. The systematic combination of these data yielded clusters, which may be visualized as a top-down hierarchical clustering tree. In this tree, chemical classes are formed at the highest level according to organic functional groups. Each subsequent subcluster stemming from classes in this hierarchy of the cluster is a chemical cluster defined by common organic functional groups and close similarity in the hydrocarbon skeleton. By examining the available experimental data for a toxicological endpoint within each cluster, users can better identify potential read-across chemicals to support safety assessments.

Mechanistic relationship between biodegradation and bioaccumulation. Practical outcomes.

According to the REACH Regulation, for all substances manufactured or imported in amounts of 10 or more tons per year, that are not exempted from the registration requirement, a Chemical Safety Assessment (CSA) must be conducted. According to CSA criteria, for these substances persistent, bioaccumulative and toxic (PBT), and very persistent and very bioaccumulative (vPvB) assessment is requested. In order to reduce the number of applications of the expensive bioaccumulation test it seems useful to search thresholds for other related parameters above which no bioaccumulation is observed. Given the known relationship between ready biodegradability and bioaccumulation, one such parameter is biodegradation. This article addresses this relationship in searching for BOD threshold above which no vB and B chemicals could be observed. It was found that the regulatory criteria for persistency could be used for identification of not vB and B chemicals. In addition, fish liver metabolism is determined as the most significant factor in reducing of maximum bioaccumulation potential of the chemicals. It was found that parameters associated with the models simulating fish metabolism could be also used for identification of not vB and B chemicals.

The implementation of RAAF in the OECD QSAR Toolbox.

The Read-Across Assessment Framework (RAAF) was developed by the European Chemicals Agency (ECHA) as an internal tool providing a framework for a consistent, structured and transparent assessment of grouping of chemicals and read-across. Following a RAAF-based evaluation, also developers and users of read-across predictions outside ECHA can judge whether their read-across rationale is sufficiently robust from a regulatory perspective. The aim of this paper is to describe the implementation of RAAF functionalities in the OECD QSAR Toolbox report. These can be activated in the prediction report after performing a readacross prediction. Once the user manually selects the appropriate scenario, the RAAF assessment elements appear and are automatically aligned with the suitable category elements of the Toolbox report. Subsequently, these are evaluated as part of the category consistency assessment functionality. The implementation of the RAAF functionality is illustrated in practice with two examples.

Identification and description of the uncertainty, variability, bias and influence in quantitative structure-activity relationships (QSARs) for toxicity prediction.

Improving regulatory confidence in, and acceptance of, a prediction of toxicity from a quantitative structure-activity relationship (QSAR) requires assessment of its uncertainty and determination of whether the uncertainty is acceptable. Thus, it is crucial to identify potential uncertainties fundamental to QSAR predictions. Based on expert review, sources of uncertainties, variabilities and biases, as well as areas of influence in QSARs for toxicity prediction were established. These were grouped into three thematic areas: uncertainties, variabilities, potential biases and influences associated with 1) the creation of the QSAR, 2) the description of the QSAR, and 3) the application of the QSAR, also showing barriers for their use. Each thematic area was divided into a total of 13 main areas of concern with 49 assessment criteria covering all aspects of QSAR development, documentation and use. Two case studies were undertaken on different types of QSARs that demonstrated the applicability of the assessment criteria to identify potential weaknesses in the use of a QSAR for a specific purpose such that they may be addressed and mitigation strategies can be proposed, as well as enabling an informed decision on the adequacy of the model in the considered context.

Validation of the performance of TIMES genotoxicity models with EFSA pesticide data.

This study validates the performance of the TIssue MEtabolism Simulator (TIMES) genotoxicity models with data on pesticide chemicals included in a recently released European Food Safety Authority (EFSA) genotoxicity database. The EFSA database is biased towards negative chemicals. A comparison of substances included in the EFSA database and TIMES genotoxicity databases showed that the majority of the EFSA pesticides is not included in the TIMES genotoxicity databases and, thus, out of the applicability domains of the current TIMES models. However, the EFSA genotoxicity database provides an opportunity to expand the TIMES models. Where there is overlap of substances, consistency between EFSA and TIMES databases for the chemicals with documented data is found to be high (>80%) with respect to the Ames data and lower than the Ames data with respect to chromosomal aberration (CA) and mouse lymphoma assay (MLA) data. No conclusion for consistency with respect to micronucleus test and comet genotoxicity data can be provided due to the limited number of overlapping substances. Specificity of the models is important, given the prevalence of negative genotoxicity data in the EFSA database. High specificity (>80%) is obtained for prediction of the EFSA pesticides with Ames data. Moreover, this high specificity of the TIMES Ames models is not dependant on pesticides being within the domains. Specificity of the TIMES CA and MLA models is lower (>40%) to pesticides for out of domain. Sensitivity of TIMES in vitro and in vivo models cannot be properly estimated due to the small number of positive chemicals in the EFSA database.

Development of a Decision Tree for Mitochondrial Dysfunction: Uncoupling of Oxidative Phosphorylation.

Mitochondrial dysfunction is the result of a number of processes including the uncoupling of oxidative phosphorylation. This study outlines the development of a decision tree-based profiling scheme capable of assigning chemicals to one of six confidence-based categories. The decision tree is based on a set of structural alerts and physicochemical boundaries identified from a detailed study of the literature. The physicochemical boundaries define a chemical relationship with both log P and p Ka. The study also outlines how the decision tree can be used to profile databases through an analysis of the publically available databases in the OECD QSAR Toolbox. This analysis enabled a set of additional structural alerts to be identified that are of concern for protonophoric ability. The decision tree will be incorporated in the OECD QSAR Toolbox V4.3. The intended usage is to group the chemicals into categories of chronic human health and environmental toxicological end points.

The OECD QSAR Toolbox Starts Its Second Decade.

The OECD QSAR Toolbox is a computer software designed to make pragmatic qualitative and quantitative structure-activity relationship methods-based predictions of toxicity, including read-across, available to the user in a comprehensible and transparent manner. The Toolbox, provide information on chemicals in structure-searchable, standardized files that are associated with chemical and toxicity data to ensure that proper structural analogs can be identified. This chapter describes the advantages of the Toolbox, the aims, approach, and workflow of it, as well as reviews its history. Additionally, key functional elements of it use are explained and features new to Version 4.1 are reported. Lastly, the further development of the Toolbox, likely needed to transform it into a more comprehensive Chemical Management System, is considered.

Lessons learned from read-across case studies for repeated-dose toxicity.

A series of case studies designed to further acceptance of read-across predictions, especially for chronic health-related endpoints, have been evaluated with regard to the knowledge and insight they provide. A common aim of these case studies was to examine sources of uncertainty associated with read-across. While uncertainty is related to the quality and quantity of the read across endpoint data, uncertainty also includes a variety of other factors, the foremost of which is uncertainty associated with the justification of similarity and quantity and quality of data for the source chemical(s). This investigation has demonstrated that the assessment of uncertainty associated with a similarity justification includes consideration of the information supporting the scientific arguments and the data associated with the chemical, toxicokinetic and toxicodynamic similarity. Similarity in chemistry is often not enough to justify fully a read-across prediction, thus, for chronic health endpoints, toxicokinetic and/or toxicodynamic similarity is essential. Data from New Approach Methodology(ies) including high throughput screening, in vitro and in chemico assay and in silico tools, may provide critical information needed to strengthen the toxicodynamic similarity rationale. In addition, it was shown that toxicokinetic (i.e., ADME) similarity, especially metabolism, is often the driver of the overall uncertainty.

Skin Sensitization QMM for HRIPT NOEL Data: Aldehyde Schiff-Base Domain.

The general chemistry principles underlying skin sensitization for Schiff base (SB) electrophiles may be used to develop a quantitative mechanistic model (QMM), based on reactivity supplemented with a hydrophobicity parameter for some but not all structures within the SB reaction domain. For aliphatic Schiff base electrophiles, the log of the no observed effect level (NOEL) values (pNOEL) from the human repeated insult patch test (HRIPT) can be calculated by the reactivity parameter summation of sigma star values (Σσ*) and a hydrophobicity parameter (logP). Specifically, the QMM, pNOEL = 2.34(±0.33) Σσ* + 0.19(±0.07) logP - 2.62(±0.22), n = 19, R2 = 0.77, R2(adj) = 0.74, s = 0.20, F = 27, was developed. Not all parts of the Schiff base domain are modeled with one equation. Particularly, predicting aromatic aldehydes and ketones appears to require a separate equation. Interestingly, the same physical organic chemical properties originally applied to modeling the local lymph node assay potency of Schiff base electrophiles apply to human potency as represented by the HRIPT.

Validation of a Fragment-Based Profiler for Thiol Reactivity for the Prediction of Toxicity: Skin Sensitization and Tetrahymena pyriformis.

This study outlines the use of a recently developed fragment-based thiol reactivity profiler for Michael acceptors to predict toxicity toward Tetrahymena pyriformis and skin sensitization potency as determined in the Local Lymph Node Assay (LLNA). The results showed that the calculated reactivity parameter from the profiler, -log RC50(calc), was capable of predicting toxicity for both end points with excellent statistics. However, the study highlighted the importance of a well-defined applicability domain for each end point. In terms of Tetrahymena pyriformis, this domain was defined in terms of how fast or slowly a given Michael acceptor reacts with thiol leading to two separate quantitative structure-activity models. The first, for fast reacting chemicals required only -log RC50(calc) as a descriptor, while the second required the addition of a descriptor for hydrophobicity. Modeling of the LLNA required only a single descriptor, -log RC50(calc), enabling potency to be predicted. The applicability domain excluded chemicals capable of undergoing polymerization and those that were predicted to be volatile. The modeling results for both end points, using the -log RC50(calc) value from the profiler, were in keeping with previously published studies that have utilized experimentally determined measurements of reactivity. These results demonstrate that the output from the fragment-based thiol reactivity profiler can be used to develop quantitative structure-activity relationship models where reactivity toward thiol is a driver of toxicity.

The adverse outcome pathway for skin sensitisation: Moving closer to replacing animal testing.

This article outlines the work of the Organisation for Economic Co-operation and Development (OECD) that led to being jointly awarded the 2015 Lush Black Box Prize. The award-winning work centred on the development of 'The Adverse Outcome Pathway for Skin Sensitisation Initiated by Covalent Binding to Proteins'. This Adverse Outcome Pathway (AOP) has provided the mechanistic basis for the integration of skin sensitisation-related information. Recent developments in integrated approaches to testing and assessment, based on the AOP, are summarised. The impact of the AOP on regulatory policy and on the Three Rs are discussed. An overview of the next generation of the skin sensitisation AOP module in the OECD QSAR Toolbox, based on more-recent work at the Laboratory of Mathematical Chemistry, is also presented.

Chemical applicability domain of the Local Lymph Node Assay (LLNA) for skin sensitisation potency. Part 3. Apparent discrepancies between LLNA and GPMT sensitisation potential: False positives or differences in sensitivity?

The Local Lymph Node Assay (LLNA) is the gold standard regulatory toxicology test for skin sensitisation along with the guinea pig maximisation test (GPMT). Compared with the GPMT, LLNA uses fewer animals, it is quantitative, and it gives a numerical prediction of potency. However several concerns have been raised with this assay, mainly related to false positives and false negatives. Over the years, many authors, including the developers of the assay, have presented cases where there have been discrepancies between the GMPT and LLNA results. Several theories have been put forward for these discrepancies, the main one being the "over-sensitivity" of the GPMT. This paper analyses the data from a systematic study, published in three papers from 2008 to 2011, covering several classes of chemicals, in particular unsaturated fatty acids, sugar surfactants and ethoxylated alcohols, with many cases of chemicals testing positive in the LLNA being negative in the GPMT. Based on consideration of reaction chemistry and structural alerts, it is concluded that these discrepancies are not LLNA false positives, but can be rationalised in terms of the different protocols of the assays.

Kinetic-Based Reactivity for Michael Acceptors: Structural Activity Relationships and Its Relationship to Excess Acute Fish Toxicity.

Acute aquatic toxicity is divided into the "physical" mode governed by weak, non-covalent interactions and the "chemical" mode governed by covalent reactions. The potency of chemical interactions is typically expected to be greater than that for physical ones. This enhanced potency is called "excess" toxicity. As databases have become complex, substances thought to elicit a chemical mode reveal a lack of excess toxicity. One mechanism where the latter is prevalent is Michael-type addition. A series of α-ß-unsaturated substances were evaluated for reactivity. Second order rate constants (k') were calculated (M-1 s-1) and found to vary from >4000 to <0.0003. The electron-withdrawing capacity of the polar group impacts k' values; the sequence is nitro > carbonyl or sulfone â‰« sulfoxide, nitrile or amide. When the α-carbon and/or the ß-carbon of the π-system are substituted, the k' value is sharply reduced. Excess toxicity is associated with k' values >0.01 (M-1 s-1).

Integrated approach to testing and assessment for predicting rodent genotoxic carcinogenicity.

We investigated the performance of an integrated approach to testing and assessment (IATA), designed to cover different genotoxic mechanisms causing cancer and to replicate measured carcinogenicity data included in a new consolidated database. Genotoxic carcinogenicity was predicted based on positive results from at least two genotoxicity tests: one in vitro and one in vivo (which were associated with mutagenicity categories according to the Globally Harmonized System classification). Substances belonging to double positives mutagenicity categories were assigned to be genotoxic carcinogens. In turn, substances that were positive only in a single mutagenicity test were assigned to be mutagens. Chemicals not classified by the selected genotoxicity endpoints were assigned to be negative genotoxic carcinogens and subsequently evaluated for their capability to elicit non-genotoxic carcinogenicity. However, non-genotoxic carcinogenicity mechanisms were not currently included in the developed IATA. The IATA is docked to the OECD Toolbox and uses measured data for different genotoxicity endpoints when available. Alternatively, the system automatically provides predictions by SAR genotoxicity models using the OASIS Tissue Metabolism Simulator platform. When the developed IATA was tested against the consolidated database, its performance was found to be high, with sensitivity of 74% and specificity of 83%, when measured carcinogenicity data were used along with predictions falling within the models' applicability domains. Performance of the IATA would be slightly changed to a sensitivity of 80% and specificity of 72% when the evaluation by non-genotoxic carcinogenicity mechanisms was taken into account. Copyright © 2016 John Wiley & Sons, Ltd.