Computational Toxicology最新文献

Pesticides can transport via food webs and bioaccumulate in birds. Feathers can be used as a biomarker to investigate the bird's pesticide exposure, but there are no modeling methodologies that can link the bird's daily intake of pesticides to their amounts in feathers. To fill this gap, we propose a physiologically based kinetic modeling approach that takes into account the feather compartment and feather growth dynamics to estimate pesticide biotransfer potentials in the bird's body. In comparison with the non-feather model, the feather compartment acted as an additional elimination pathway for pesticides from the bird's body, resulting in a decrease in the overall simulated pesticide concentrations in the bird's body. High-lipophilic or high-volatile pesticides had exceptionally poor biotransfer potentials in feathers due to thermodynamics (e.g., partitioning potentials) or kinetics (e.g., elimination rates). As a result, legacy pesticides (such as persistent organic pollutants) will have limited biotransfer potentials in feathers, and the presence of legacy pesticides in feathers could imply relatively high daily pesticide consumption rates, which could affect bird reproductive health. The sensitivity and variability tests revealed that the metabolic (or biotransformation) kinetics of pesticides in avian livers influenced the biotransfer potential of pesticides in feathers significantly. Given the lack of information on pesticide metabolic kinetics in avian livers, we strongly suggest that future research (e.g., in vivo or in vitro studies) determine the metabolic or biotransformation rates of pesticides in avian livers in order to improve the performance of models. Furthermore, the use of additional biomarkers such as blood and uric acid could be valuable in assessing birds' exposure to pesticides. Hopefully, this study will help ecologists comprehend the fate, transport, and biotransfer of pesticides in bird feathers from a modeling point of view.

Read-across and chemical grouping approaches are increasingly utilized in risk assessment and regulatory submissions. The European Food Safety Authority (EFSA) has established Chemical Group (CG) 14, which contains furfuryl and furan derivatives both with and without side-chain substituents; however, the rationale for this grouping is not available based on current EFSA documentation. Therefore, this study aimed to identify the chemicals belonging to CG14, evaluate the constituent chemicals for metabolic, biological, and toxicological properties via clustering tools, and apply existing chemical grouping workflows for identifying representative chemicals for this group to support testing strategies. Membership to CG14 was difficult to identify, and varied by EFSA source (e.g., published reports, OpenFoodTox database). Based on predictions from the Organisation for Economic Co-operation and Development (OECD) Quantitative Structure Activity Relationship (QSAR) Toolbox, ChemACE, SMARTCyp, and WhichCyp, as well as data extracted from the U.S. Environmental Protection Agency’s (EPA’s) Toxicity Forecaster (ToxCast) on CompTox Chemicals Dashboard and the European Chemicals Agency (ECHA) Dossier, no suitable metabolic, toxicological, structural, or mechanistic clustering method was identified for CG14. Biological effect data were too sparse to refine the chemical subgroupings within CG14 with confidence based on existing read-across principles for chemical grouping. This paucity of data limits the development of a tiered testing strategy in which more complete testing would be conducted for selected representative CG14 compounds only. Therefore, efforts to generate key pieces of data (e.g., mode of action [MOA], metabolism) for chemical grouping and read-across are needed to apply this workflow to EFSA CG14.

Chronic toxicity is one of the most important toxicological endpoints related to human health. Since experimental tests are costly and difficult, in silico methods are crucial to assessing the chronic toxicity of compounds. There are only very few QSAR studies available on chronic toxicity prediction. This study aimed to develop a QSAR model using 650 diverse and complex compounds based on the lowest observed adverse effect level (LOAEL), which was determined in rats by orally exposing them to these compounds. We have extracted important descriptors from a pool of 868 descriptors using stepwise regression and a genetic algorithm. We validated the developed partial least squares (PLS) model statistically, and the results demonstrate the model's reliability, robustness, and predictive ability (R² = 0.60, Q²_(LOO) = 0.58, Q²_F1 = 0.56, and Q²_F2 = 0.56). Our validated models were also used to assess the chronic toxicity of 11,300 pharmaceuticals present in the DrugBank database. It has been found that hydrophobicity, electronegativity, lipophilicity, bulkiness, complex chemical structure, bridgehead atoms, and phosphate group play a crucial role in chronic toxicity. Therefore, these markers can be used to synthesize safe, and eco-friendly organic chemicals.

A drug’s skin permeability is a key quantity within dermatological risk assessment as it quantifies the maximal rate of dermal absorption, under steady state conditions, per unit concentration difference across the skin of a topically applied drug. Given descriptors of the permeant, the vehicle, and skin conditions, this paper adopts a systematic approach to efficiently calculate estimates of a topically applied chemical’s skin permeability based on mechanistic knowledge of permeability across the elements of a fine spatial discretization of the dermal strata. The permeability estimates are obtained by solving a system of linear equations constructed using an electrical resistor network analogy. Being mechanism based, these estimates can account for skin conditions, heterogenous dermal penetration pathways, and the chemical-dependence and spatial dependence of partitioning and diffusivity.

The contribution of this work can be viewed as a mechanistic, numerical, extension of the Potts-Guy relation that augments an open-source dermal PBPK model. Moreover, rather than requiring model simulations of steady state conditions, the approach centers on a direct calculation of permeability based on the mechanistic descriptors of the permeation scenario. The calculation method may therefore be directly integrated into parameter identification, optimization, and sensitivity analysis algorithms. We demonstrate the validity of the method by comparing its permeability predictions with previously reported in silico estimates and in vitro measurements. We additionally illustrate the utility of the method towards the analysis of dermal PBPK models using a minimal example that relates overall skin permeability to the interaction between multiple permeation pathways.

The accumulation of toxic perfluoroalkyl and polyfluoroalkyl substances (PFAS), also known as forever chemicals, in umbilical cord blood calls for an urgent need to explore PFAS kinetics at the maternal-fetal interface placenta. Therefore, this study modeled the possible effects of ten PFAS on two enzymes (glutathione S-transferase (GST) and N-acetyltransferase (NAT2) that are active in the placenta and can protect the fetus from xenobiotics. Molecular docking was used to determine the binding affinities of some common PFAS at two placental enzyme targets. Density functional theory (DFT) analysis and artificial neural networks (ANN) on the PFAS were performed to identify their chemical reactivity descriptors and the most important one responsible for binding, respectively. The molecular docking studies showed that perfluorooctanesulphonamide (PFOSA) and perfluorodecanoic acid (PFDA) consistently had higher binding affinities on the two placental enzymes than the controls, glutathione, and coenzyme A. DFT revealed that out of the ten PFAS analyzed, PFDA had the lowest binding affinity and chemical softness, making it the most reactive and as such toxic PFAS in the group. At normalized importance of >80 %, the ANN analysis predicted that the molecular weight and total energy were the primary reactivity descriptors of the PFAS responsible for their binding on the GST. In contrast, their binding energy was responsible for binding at the NAT2. The results from these simulations indicate that PFAS, especially PFDA, have the potential to inhibit placental enzyme activity in humans. This may have far-reaching consequences for placental functions and fetal development, which needs to be clarified in future studies.

Accurate in-silico models of human skin are required to obtain the uptake/release of molecules across the skin layers to supplement the in-vivo/in-vitro experiments for faster development/testing of cosmetics and drugs. We aim to develop an in-silico skin permeation model by extending the multiscale modeling framework developed earlier for skin’s top layer to deeper layer and compared the outcomes with in-vitro experimental permeation data of 43 cosmetic-relevant molecules across human skin.

In this study, we have extended a multiscale modeling framework, with realistic heterogeneous stratum corneum (SC) comprising of network of permeable lipids and corneocytes, followed by homogeneous viable epidermis and dermis. The diffusion coefficients of molecules in lipid layer were determined using molecular dynamics simulations, whereas the diffusion coefficients in other layers and all the partition coefficients were calculated from correlations reported in literature. These parameters were then used in the macroscopic models to predict the release profiles of drugs through the deeper skin layers. The obtained release profiles were in good agreement with available experimental data for most of the molecules. The reported model could provide insight into cosmetics/drugs skin permeation and act as a time-saving and efficient guiding tool for performing targeted experiments.

Macitentan is a dual endothelin receptor antagonist indicated for the treatment of pulmonary arterial hypertension, a chronic and complex disease. Under different stress conditions, such as changes in pH and temperature, the drug can generate a large number of degradation products, while many process-related impurities can occur during the four main synthetic routes. The assessment of the potential toxicity of these impurities is an essential regulatory requirement for the quality and safety of drugs. The goal of this study was to identify all metabolites and potential impurities for macitentan and evaluate their in silico toxicity. Thirty-five compounds related to macitentan were found reported in the literature, two of which were described simultaneously as metabolites, degradation products, and process-related impurities. In the present study, the main degradation products and the conditions under which they could be formed, and the major impurities according to the synthetic route, are discussed. The types and amounts of process-related impurities were dependent on the synthesis route and process controls, while macitentan was found to be more susceptible to degradation in acidic media resulting in the most different types of degradation products. The structure of each compound was generated and the potential risk for mutagenicity and carcinogenicity were determined using three different in silico platforms, in addition the metabolic substrate/inhibition profile for each compound was assessed. Overall, five compounds were considered critical as they had a possible toxicity risk in terms of mutagenicity, tumorigenicity, irritation, and reproductive effects. These data support the current legislation for raw materials and pharmaceutical products containing macitentan as to prevent any adverse effects from this drug.

Cancer is the second leading cause of death worldwide. Among various anticancer drug targets, mTOR is noteworthy. Numerous first-generation mTOR inhibitors are already approved and few second-generation mTOR inhibitors targeting the kinase domain are in the clinical trials, but yet to reach the market, and many lead to serious toxicities. Here we are focused to discover some novel kinase inhibitors from the ZINC database which may effectively inhibit mTOR kinase. For this, computational chemistry and pharmacophore-based ZINC database search has been adopted. Series of virtual screening analysis lead to the discovery of 5 active hits. Among these 5, compound 4 (ZINC79476038) having binding energy of −8.9 Kcal/mol shows maximum interactions within the binding pocket. Study proved that all these compounds can potentially inhibit mTOR kinase and can be successfully developed as anticancer agents. We further proved that these compounds are not only active for general cancers like lung, breast, colon, and other peripheral cancers but also equally active in CNS, targeting numerous brain cancers.

The Analog Identification Methodology (AIM) was developed over 20 years ago to identify analogues to support read-across at the US Environmental Protection Agency. However, the current public version of the standalone tool, released in 2012, is no longer usable on Windows operating systems supported by Microsoft. Additionally, the structural logic for analogue selection is based on older, customised Simplified molecular-input-line-entry system (SMILES)-type features that are incompatible with modern cheminformatics tools. Given these limitations, a case study was undertaken to explore a more transparent, extensible method of implementing the AIM fragments using Chemical Subgraphs and Reactions Mark-up Language (CSRML). A CSRML file was developed to codify the original AIM fragments, and the extent to which AIM fragments were faithfully replicated was assessed using the AIM Database. The overall mean performance of the CSRML-AIM across all fragments in terms of sensitivity, specificity, and Jaccard similarity was 89.5%, 99.9%, and 82.2%, respectively. Comparing the AIM fragments with public ToxPrints using a large set of ∼25,000 substances of regulatory interest to EPA found them to be dissimilar, with an average maximum Jaccard score of 0.24 for AIM and 0.29 for ToxPrint fingerprints. Both fragment sets were then used as inputs in the automated read-across approach, Generalised Read-Across (GenRA), to evaluate the quality of fit in predicting rat acute oral toxicity LD₅₀ values with the coefficient of determination (R²) and root mean squared error (RMSE). The performance of AIM fragments was R²=0.434 and RMSE=0.663 whereas that of ToxPrints was R²=0.477 and RMSE=0.638. A bootstrap resampling using 100 iterations found the mean and the 95th confidence interval of R² to be 0.349 [0.319, 0.379] for AIM fragments and 0.377 [0.338, 0.412] for ToxPrints. Although AIM and ToxPrints performed similarly in predicting LD_50, they differed in their performance at a local level, revealing that their features can offer complementary insights.

Exposure to chemicals generally occurs in the form of mixtures. However, the great majority of the toxicity data, upon which chemical safety decisions are based, relate only to single compounds. It is currently unfeasible to test a fully representative proportion of mixtures for potential harmful effects and, as such, in silico modelling provides a practical solution to inform safety assessment. Traditional methodologies for deriving estimations of mixture effects, exemplified by principles such as concentration addition (CA) and independent action (IA), are limited as regards the scope of chemical combinations to which they can reliably be applied. Development of appropriate quantitative structure-activity relationships (QSARs) has been put forward as a solution to the shortcomings present within these techniques – allowing for the potential formulation of versatile predictive tools capable of capturing the activities of a full contingent of possible mixtures. This review addresses the current state-of-the-art as regards application of QSAR towards mixture toxicity, discussing the challenges inherent in the task, whilst considering the strengths and limitations of existing approaches. Forty studies are examined within – through reference to several characteristic elements including the nature of the chemicals and endpoints modelled, the form of descriptors adopted, and the principles behind the statistical techniques employed. Recommendations are in turn provided for practices which may assist in further advancing the field, most notably with regards to ensuring confidence in the acquired predictions.