Computational Toxicology最新文献

Context

Rock dusts often contain minerals called amphiboles. Elongate mineral particles produced by excavation, crushing, or grinding amphibole-containing rock can belong to different morphological groups, or habits: asbestiform or non-asbestiform. Some asbestiform particles are highly potent for causing mesothelioma, but non-asbestiform elongate structures have not been implicated in elevated cancer risk. Computational analysis and modelling of the dimensional characteristics of the elongate mineral particles is needed to develop efficient criteria for their differentiation, and also for determining the parameters driving their carcinogenic potential.

Objectives

To develop conceptual and quantitative models allowing reliable distinctions between asbestiform and non-asbestiform amphibole particles that are based on particle dimensions and are consistent with observed disease outcome following human exposure.

Methods

For modelling, the unique database including 56 datasets designated as dominantly asbestiform (67,876 amphibole particles), 37 designated as dominantly non-asbestiform (235,247 amphibole particles), and 12 as inhomogeneous or anomalous (35,277 amphibole particles) was utilized. The discriminant analysis was used to determine functions that separate elongate mineral particles by their habit based on length and width. Linear regression and cluster analysis were applied to determine the relationship between values of the selected discriminant function and relevant toxicological parameters.

Results

For particles longer than 5 µm, the function $Y=2.99{\log }_{10}Length-5.82{\log }_{10}Width-3.80$ was selected as the best discriminator of particles for their asbestiform and non-asbestiform habits, with a misclassification rate of about 15% total. The value of the discriminant function derived for each particle correlates with the particle’s calculated aerodynamic diameter (R = −0.859, p < 0.00001) and with its specific surface area (R = 0.857, p < 0.00001). The cluster analysis demonstrated that subdivision of particles by two groups according to their length and width closely reconstructs the pre-defined habits.

Conclusion

The proposed methodology of differentiating between asbestiform and non-asbestiform particles can be used for analytical, toxicological, and regulatory purposes.

Baicalin is a main active ingredient of the dried root of Scutellaria and has been extensively employed in Traditional Chinese Medicine for the treatment of asthma, fever, and psoriasis. Based on the reports of antioxidant, anti-inflammatory, anti-infection, and anti-tumor activities of baicalin, we have explored its radioprotective efficacy using in vitro and in vivo experimental model systems. In the present study, we have investigated the radioprotective, immunomodulatory, and anti-inflammatory properties of baicalin using wet lab and in silico approaches. It was observed that pre-treatment of murine splenic lymphocytes with baicalin protected cells against radiation-induced cell death possibly by decreasing the cellular reactive oxygen species levels. Prophylactic oral administration of baicalin offered significant increase in endogenous spleen colony counts and an enhancement in the survival of mice. We have also observed that baicalin suppressed mitogen-induced splenic lymphocyte proliferation and IL-2 production. It also inhibited the production of nitric oxide in RAW 264.7 cells in response to elicitation of lipopolysaccharide. Further, in silico study was performed to evaluate the possible mechanism of radioprotection and immunomodulation by selecting different pro-inflammatory mediators such as COX2, Lck, NIK, and IKK-β which have a significant role in radioprotection, lymphocyte activation, and inflammation. Our molecular docking and molecular dynamics study show that baicalin has a significant predicted binding affinity with COX2, Lck, NIK, and IKK-β. These in silico results can explain the experimentally observed radioprotective, immunosuppressive, and anti-inflammatory properties of baicalin. Thus, radioprotection offered by baicalin may be because of its antioxidant, anti-inflammatory, and immunomodulatory properties.

Next-generation risk assessment (NGRA) involves the combination of in vitro and in silico models for more human-relevant, ethical, and sustainable human chemical safety assessment. NGRA requires a quantitative mechanistic understanding of the effects of chemicals across human biology (be they molecular, cellular, organ-level or higher) coupled with a quantitative understanding of the uncertainty in any experimentally measured or predicted values. These values with their uncertainties can then be considered as a probability distribution, which can then be compared to exposure estimates to establish the presence or absence of a margin of safety. We have constructed Bayesian learning neural networks to provide such quantitative predictions and uncertainties for 20 pharmacologically important human molecular initiating events. These models produce high quality quantitative estimates (p(IC50), p(EC50), p(Ki), p(Kd)) of biochemical activity at a molecular initiating event (MIE) with average mean absolute errors (in Log units) of 0.625 ± 0.048 in test data and 0.941 ± 0.215 in external validation data. The key advantage of these models is their ability to also produce standard deviations and credible intervals (CIs) to quantify the uncertainty in these predictions, which we show to be able to distinguish between molecules close to the training data in chemical structure, those less similar to the training data, and decoy compounds drawn from the wider ChEMBL database. These uncertainty values mean that when a prediction is made a user can understand the certainty of the prediction, similar to a quantitative applicability domain, aiding prediction usefulness in NGRA. The ability for in silico methods to produce quantitative predictions with these kinds of probability distributions will be vital to their further use in NGRA, and here clear first steps have been taken.

In silico (computational) methods continue to evolve as part of a robust 21st century public health strategy in risk assessment, relevant to all sectors of chemical safety including preclinical drug discovery, industrial chemicals testing, food and cosmetics. Alongside in vitro methods as components of intelligent testing and pathway driven strategies, in silico models provide the potential for more human relevant solutions to the use of animals in safety testing and biomedical research. These are often termed ‘New Approach Methodologies’ (NAMs). Some NAMs incorporate the use of ‘big data’, for example the information provided from high throughput or high content in vitro screening assays or ‘omics’ technologies. Big data has increasing relevance to predictive toxicology but must be appropriately defined, particularly with regard to ‘quality vs quantity’. The purpose of this article is to provide a commentary on the progress of in silico human-based research methods within the context of NAMs, as well as discussion of the emerging use of big data with relevance to safety assessment. The current status of in silico methods is discussed, with input from researchers in the field. Scientific and legislative drivers for change are also considered, along with next steps to address challenges in funding and recognition, to achieve regulatory acceptance and uptake within the research community. To provide some wider context, the use of in silico methods alongside other relevant approaches (e.g., human-based in vitro) is also discussed.

Apoptosis signal kinase 1 (ASK 1) is a member of the mitogen-activated protein kinase (MAPK) family that induces cells apoptosis including colorectal cancer (CRC). CRC is the second most common type of malignancy globally. Hence, ASK 1 plays an essential role in the pathogenesis of CRC and therefore, is an exclusive target in drug design and discovery for CRC. Herein, applied computational approaches including molecular docking, molecular mechanics/generalized born surface area calculation and pharmacokinetic models were performed to propose putative ASK 1 antagonists from natural compounds. Seven (7) ligands were identified as potent inhibitors and two top hit compounds were validated using molecular dynamics (MD) simulation studies. The density function theory (DFT) of the hits were performed at the Becke three Lee Yang Parr/6-31G(d) level of theory to understand their molecular reactivity. Seven compounds identified as ASK 1 antagonists have docking score ranging from −9.10 to −8.14 kcal/mol which is comparable to the reference ligand camptosar (-7.03 kcal/mol). One of the compounds, odoratin has finally emerged as the structurally stable compound with −9.10 kcal/mol and MD simulation over 50 ns indicated that odoratin forms stable interaction with key amino acid residues such as LEU 686, VAL 757 and PRO 758. DFT study showed that the studied compounds have proton donating and accepting ability hence, potent inhibitory and solubility effects. The findings from this study suggest that, odoratin could be considered a potent ASK 1 inhibitor and could be experimentally verified as a lead compound for search of MAPK inhibitors for colorectal cancer.

Statistical modeling approaches for dose-response or concentration-response analyses are often required in toxicological applications, especially for cytotoxicity assays. By fitting a concentration-response curve, one can derive target concentrations, such as the ${\mathrm{EC}}_{50}$. In practice, concentration-response data for different exposure durations might be available and the target concentration for each or some exposure duration(s) are of interest. In this work, we propose a statistical modeling approach that improves the precision of the target concentration estimation at a given exposure duration by extrapolating the concentration-response data from other exposure durations. The method further enables target concentration estimation at exposure durations that were not conducted in the experiment. For practitioners, the proposed model yields additional complexity compared to the simple approach of a single concentration-response curve for all exposure durations. It would only be used if it improves the estimation of the target concentration compared to the simple approach. We propose a two-step pipeline to decide between using the complex and the simple approach to result in a precise target concentration estimation.

The methods were evaluated using a simulation study and a real data set. The models are accessible for practitioners through the R package td2pLL.

A novel modelling platform, the Merck Explosive Prioritisation Scheme, is introduced. It’s dependent on neither atom types nor the chemical class of the assessed molecule. The thermodynamic layer includes simulation of chemical decomposition with further estimation of the enthalpy of the decomposition and the volume of released gases. A new algorithm, the “Greedy” method, is used in calculating decomposition enthalpy. The heats of formation are estimated by quantum-chemical calculations. The enthalpy of decomposition and volume of released gases are used to predict the Power Index (PI) of the chemicals estimated as the ratio of the explosive power of the analysed substance towards the reference chemical picric acid. Based on regulatory defined thresholds for the enthalpy of explosion and volume of released gases, a threshold for the PI is defined. Chemicals are classified as “explosive” if their PI values are higher than the threshold. The performances of the algorithms in the thermodynamic layer showed good predictability. Given the nature of the Greedy algorithm, the thermodynamic model tends to slightly overpredict experimental power indices but never underestimates the explosive properties of chemicals. The kinetic layer estimates the explosive sensitivity by applying the COREPA model based on quantum-chemical and physicochemical parameters. The COmmon REactivity PAttern (COREPA) modelling system performs well for the impact sensitivity dataset. Given the limited number of chemicals used to derive the model, its current applicability domain is narrow.

Due to their interesting hydrogen-bonding properties, Schiff bases are known for their variety of applications in chemistry and medicinal chemistry. In this work, the interaction between symmetrical Schiff base ligand (L: bis [4-hydroxy-6-methyl-3-{(1E)-N-[2 (ethylamino) ethyl] ethanimidoyl}-2H-pyran-2-one]) and cancer cells, neurological, viruses and bacteria receptors was studied theoretically. Density functional theory (DFT) was used to determine the geometry, reactivity and electronic properties of this ligand. Homology modeling and molecular docking were performed to check their biological and medicinal properties, including anticancer, antiviral, antibacterial and neurological activities. DFT revealed that the mulliken charges, the molecular orbitals (HOMO and LUMO) and MEP results are in a good agreement to the localization of electrophilic and nucleophilic attack sites. The theoretical study showed a high chemical reactivity and a low kinetic stability of the ligand. The docking study results revealed that the ligand exhibits a good biological activity against leukemia, breast cancer, Alzheimer and Covid-19 with binding energy values of −7.36 kcal/mol, −6.35 kcal/mol, −6.19 kcal/mol and −5.58 kcal/mol, respectively. These results are explained by the low values of binding energy and inhibition constant and multiple H-bonds.

The definition, characterisation and assessment of the similarity between target and source molecules are cornerstones of the acceptance of a read-across prediction to fill a data gap for a toxicological endpoint. There is much guidance and many frameworks which are applicable in a regulatory context, but as yet no formalised process exists by which to determine whether or not the properties of an analogue (or chemicals within a category) fall within an appropriate domain from which a reliable read-across prediction can be made. This investigation has synthesised much of the existing knowledge in this area into a practical strategy to enable the domain of a read-across prediction to be defined, in terms of chemistry (structure and properties), toxicodynamics and toxicokinetics. The strategy is robust, comprehensive, flexible, and can be implemented readily. It enables the relative similarity and dissimilarity, between target and source molecules, for both the analogue and category approaches, to be analysed and provides a basis for alternative scenarios such as read-across based on formation of a common metabolite or biological profile to be defiend. Herein, the read-across domains for the repeated dose toxicity of a group of triazoles and imidazoles have been evaluated. The most challenging aspect to this approach will continue to be determining what is an “acceptable” degree of similarity when performing read-across for a specific purpose.