Computational Toxicology最新文献

Traditional QSAR models predict mutagenicity solely based on structural alerts for the interaction of parent chemicals or their metabolites with target macromolecules. In the present work, it is demonstrated that the presence of an alert is necessary to identify damage but it is not always sufficient to assess mutagenic potential. This is addressed by accounting for the kinetics of simulating metabolism and formation of adducts with macromolecules. The mutagenic potential of chemicals is related to the degree to which selected macromolecules are altered. This extent is estimated by the amount of formed DNA/protein adducts. Here the effect of modelling kinetic factors is investigated for chemicals having documented in vitro negative and in vivo positive data in mutagenicity and clastogenicity tests of similar capacity - in vitro Ames vs in vivo TGR and in vitro CA vs in vivo MN tests. Two factors justify the conflict in mutagenicity data: the differences in enzyme expression in the in vitro vs in vivo metabolism and the difference in exposure time for in vitro and in vivo tests. Addressing these factors required simulating the formation of DNA/protein adducts and introducing empirically-defined thresholds for the amounts of the adducts leading to mutagenic potential.

The platform of OECD Toolbox version 4.5 was used for building an automated decision tree for filling data gaps for rat acute oral toxicity (AOT) by read-across (RA). Our previous publications have described the workflow of the AOT tree and conducted verification and validation studies on it. The overall uncertainty in the AOT workflow is low as the similarity in mechanistic probability, metabolism and 2D structure are maximized in the RA analogue selection process. The endpoint, rat oral LD50, is well-defined and has universal regulatory acceptance. Since OECD test guidelines are followed in generating the database, the data are widely recognized to be of the highest quality. The credibility of the workflow is high as it meets the critical factors of being based on confirmed assumptions, having demonstrated concordance and consistency, permitting the ability to explain AOT-related mechanisms and modes of action, and being simple in design. Additionally, the Z-score and probability distribution methods of assessing the uncertainty of a particular RA are discussed. Two examples of numerical and classification uncertainty are presented. These cases represent the extremes observed in a series of target chemical-based predictions that the authors observed when testing the workflow. The reliability and relevance associated with the workflow are high. However, the completeness and weights-of-evidence varied markedly among possible RA scenarios and particular target substances.

Context

An apparent deviation from nonlinearity in cancer dose-response was reported for various carcinogens. In particular, some studies hypothesized that in mesothelioma, the exposure-response relationship can be modelled as a power function with exponent from 0.6 to 1. However, various factors can affect the shape of the dose-response, producing the apparent supralinear trend.

Objective

To develop a mathematical model that would demonstrate a relationship of mesothelioma lifetime risk and exposure duration, with various assumptions about a hazard rate function.

Methods

Two different hazard rate functions – the Peto model and the two-stage clonal expansion (TSCE) model – were considered. The analytical formulas for lifetime risk were developed, with and without a lifetable correction. Standard calculus methods were applied to test the shape of the lifetime risk curve.

Results

For both Peto and TSCE models, it was shown that mesothelioma lifetime risk changes supralinearly with duration; the exponent of the power function was ranging from 0.68 to 0.89. However, the dose-response curve by cumulative exposure is close to linearity and is linear if the exposure duration would be constant. The model has been tested for chrysotile asbestos cohorts, with a good agreement demonstrated with published mesothelioma excess mortality (R=0.88, p<0.0041).

Conclusion

For mesothelioma, the observed deviation from linearity in the dose-response relationship can be potentially explained by the impact of a change in the duration of exposure. In a meta-analysis, this deviation can be eliminated by standardizing the mortality data for various cohorts by duration of exposure.

Short Abstract

An apparent deviation from nonlinearity in cancer dose-response was reported for various carcinogens. We applied two different hazard rate equations – the Peto model and the two-stage clonal expansion (TSCE) model – to pleural mesothelioma mortality. The analytical formulas for lifetime risk were developed. For both the Peto and TSCE models, it was shown that mesothelioma lifetime risk changes supralinearly with duration. However, the dose-response curve for cumulative exposure is close to linearity.

Nanomaterial development is one of the most significant technological advances of the 21st century, with considerable impact in several fields. However, nanomaterials can pose risks to human health and the environment. Therefore, it is imperative to perform toxicological tests; nonetheless, identification and analysis of all preparations is laborious. In this regard, in silico approaches facilitate nanotoxicity assessment at low cost and without involving animal testing. In this paper we review the use of computational approaches for nanotoxicology prediction. First, we present computational nanotoxicology in a regulatory context. Next, we discuss the primary computational methods used in toxicology, such as (quantitative) structure–activity relationship models, physiologically based pharmacokinetic models, and molecular modeling, and address the singularities of each method for nanomaterial analyses. Lastly, we describe several integrative approaches for computational nanotoxicology. Various database analyses combined with complementary computational approaches can lead to creative solutions for predicting toxicological effects during the design of new nanomaterials. Therefore, data-integration methods promote understanding of complex nanotoxicological events and can be used to develop successful precision models.

Neurotoxicology is the study of adverse effects on the structure or function of the developing or mature adult nervous system following exposure to chemical, biological, or physical agents. The development of more informative alternative methods to assess developmental (DNT) and adult (NT) neurotoxicity induced by xenobiotics is critically needed. The use of such alternative methods including in silico approaches that predict DNT or NT from chemical structure (e.g., statistical-based and expert rule-based systems) is ideally based on a comprehensive understanding of the relevant biological mechanisms. This paper discusses known mechanisms alongside the current state of the art in DNT/NT testing. In silico approaches available today that support the assessment of neurotoxicity based on knowledge of chemical structure are reviewed, and a conceptual framework for the integration of in silico methods with experimental information is presented. Establishing this framework is essential for the development of protocols, namely standardized approaches, to ensure that assessments of NT and DNT based on chemical structures are generated in a transparent, consistent, and defendable manner.

BSEP inhibition is one risk factor for Drug-Induced Liver Injury (DILI). While in vitro screening of BSEP inhibition may prevent compounds with BSEP liability from progressing into the clinic, these in vitro data alone can result in false-positives and as such a specific in vivo biomarker would further enhance our BSEP inhibition de-risking strategy. Measurement of endogenous bile acids as biomarkers of BSEP inhibition in vivo is complicated by several factors, including drugs that inhibit BSEP can also inhibit other bile acid transporters such as NTCP. Here, we developed a novel translational framework, including an in vivo biomarker with a corresponding mechanistic model, and attempted to decouple the effect of liver sinusoidal uptake inhibition from efflux inhibition on bile acid disposition in the beagle dog. Specifically, we hypothesized that the change of a stable isotope-labeled (SIL) bile acid tracer’s exposure would yield a toxicodynamic signal that can provide insight into BSEP inhibition and ensuing bile salt accumulation. For this purpose we dosed the stable isotope-labeled cholic acid (¹³C-CA) and taurocholic acid (D4-TCA) as biomarker tracers in dogs, with and without the liver transporter inhibitor simeprevir, and determined the plasma and bile exposure of ¹³C-CA, ¹³C-TCA, D4-CA and D4-TCA in vivo. Key bile acid clearance and transporter inhibition parameters were determined in vitro. We developed a novel Physiologically Based Pharmacokinetic model (PBPK) to integrate the mechanistic physiological understanding, literature knowledge, and in vitro laboratory data to model bile acid disposition. Using modeling and simulation, we provided an increased mechanistic understanding of how to use plasma bile acid tracer data to inform on potential liver transporters inhibition and limitations to in vivo translation. The novel translational framework can enhance the future BSEP inhibition de-risking strategy, particularly if the experimental confounders to studying kinetics in dog hepatocytes in vitro models are solved.

The mutation in the solute carrier 6 (SLC6A19) gene causes the Hartnup disorder, affecting the absorption of non-polar amino acids. Recent DNA sequencing advances have increased the identification of single nucleotide polymorphisms (SNPs) in the SLC6A19 gene, but no further information regarding their deleterious probability is available. Hence, this study aims to comprehensively analyze and identify the potentially deleterious non-synonymous-SNPs of the SLC6A19 gene with a computational approach using openly accessible online software tools including SIFT, PolyPhen2, SAVES 5.0, SPIDER, etc. and also to determine effective lead compound for its treatment by docking. The SLC6A19 gene translates to B⁰AT1 tetramer protein, amongst chain A was taken into consideration. The analysis revealed mutation G490S (chain A) of the said protein as the candidate ns-SNP among the screened 539 missense mutations, retrieved from the National Centre for Biotechnology Information (NCBI). Moreover, the binding energy of the candidate ns-SNP had a higher affinity for benztropine over conventional drugs such as nicotinamide and niacin. Yet, clinical validation is required to support the above findings.

Toxicity prediction using linear QSAR models typically show good predictivity when trained on a small-scale, local level of similar chemicals, but not on a global level spanning a chemical library. We hypothesize that large chemical toxicity datasets generally have a locally-linear data structure, and propose the locality-sensitive deep learner (LSDL), a deep neural network with attention mechanism [1] and an optional instance-based feature weighting component, to tackle the challenges of heterogeneous classification space with locally-varying noise features. On carefully-constructed synthetic data with extremely unbalanced classes (10% positive), the locality-sensitive deep learner with learned feature weights retained high test performance (AUC > 0.9) in the presence of 60% cluster-specific feature noise, while feed-forward neural network appeared to over-fit the data (AUC < 0.6). For the Tox21 dataset [2], locality-sensitive deep learner out-performed feed-forward neural network in 9 out of 12 labels. For acetylcholinesterase inhibition (AChEi) [3], Collaborative Modeling Project for Androgen Receptor Activity (CoMPARA) [4], and Acute Oral Toxicity (AOT) [5] datasets, we observed that the combination of locality-sensitive deep learner with feed-forward neural network showed improved test performance than individual models in almost all cases. Generalizing machine learning models to fit locally-linear data may potentially improve predictivity of chemical toxicity models. The proposed modeling approach could potentially complement and add diversity to the current suite of predictive toxicity algorithms for use in ensemble and/or consensus models.

Traditional in vivo strategies for investigating toxicokinetics can be time consuming, expensive, and often do not directly address species of interest, e.g., humans. As such, conventional approaches for addressing emerging human health risk assessment concerns that rely on toxicokinetic information have been slow and suboptimal. Alternatives to rodent in vivo toxicokinetic studies include in vitro and in silico approaches for estimating toxicokinetic parameters. This paper focuses on quantitative structure-activity relationships (QSARs) that predict both maximal capacity for metabolism (Vmax) and KM (Michaelis constant, or half-maximal concentration for metabolism). The QSARs, identified from four publications, were evaluated using a previously published 10-step work flow. None of the evaluated QSARs in their published forms could be fully validated. Literature review, finding alternative sources of descriptors and identifiers, substitution of correlated descriptors, and use of graphical information allowed the deficiencies to be addressed for QSARs describing alkylbenzenes, volatile organic compounds (VOCs), and substrates of alcohol dehydrogenase (ADH), aldehyde dehydrogenase (ALDH), cytochrome P450 (CYP), and flavin containing monooxygenases (FMO). Ultimately, reliable, well-documented, updated expressions for Vmax and KM (or Vmax/KM) were derived for each source/data set. The smaller data sets tended to have better predictivity, and Vmax was generally more accurately predicted than KM. Comparisons of the QSARs’ source chemicals found limited overlap in source chemicals, but substantial overlap in descriptor domains. In a feasibility case study, applicability of these QSARs to jet fuel components with limited toxicokinetic parameterization was assessed to determine the potential utility for investigation of mixture toxicokinetics. The VOC QSARs and alkylbenzene QSARs were identified as having greater potential to accurately predict in vivo toxicokinetics of the selected jet fuel components than the CYP QSARs, due to the physicochemical characteristics of the chemicals used in their development.