Chemical Research in Toxicology最新文献

Exposure to environmental and pharmaceutical chemicalspresentsa significant risk for neurotoxicity. To reduce reliance on in vivotesting, the U.S. Toxicology in the 21st Century (Tox21) consortiumhas generated quantitative high-throughput screening data for approximately10,000 compounds (the Tox21 10K library). This study evaluates thepredictive performance of Tox21 assay data and chemical structureinformation acting simultaneously as input descriptors to determinean independent target variable of potential compound neurotoxicity.We constructed predictive models using five machine learning algorithms:Random Forest, Naïve Bayes, eXtreme Gradient Boosting, SupportVector Machine, and Neural Networks. Model performance was evaluatedusing the area under the receiver operating characteristic curve (AUC-ROC).Two models were built on individual types of chemical fingerprints(ECFP4 and ToxPrint), which achieved AUC-ROC values of 0.72–0.77and 0.67–0.71, respectively, while those using Tox21 assaydata ranged from 0.66–0.73. Notably, models that integratedall three data types outperformed single-source models, achievingAUC-ROC values of 0.78–0.84, indicating that multimodal integrationof biological and chemical features reliably enhances predictive accuracy.These models were then applied to virtually screen the Tox21 10K libraryand to prioritize compounds with the highest predicted neurotoxicityfor experimental validation. Key contributors to potential neurotoxicityprediction included assay data on glucocorticoid receptor activationand structural features such as pyridine rings and tertiary aliphaticamines. Overall, this study demonstrates the effectiveness of integratinghigh-throughput assay data with chemical structure analysis to enablereliable in silico prediction of neurotoxicity, offering a scalableand animal-free approach for assisting chemical safety assessment.

In vitro toxicity data can be used forcomparativehazard assessment of the problematic plastic chemical bisphenol A(BPA) and its replacement chemicals. Hazard assessment is exclusivelybased on nominal effect concentrations of the BPA alternatives, butdue to their diversity of physicochemical properties and stability,this may lead to inaccurate hazard ranking. In the present study,chemical stability as well as freely dissolved concentrations (C_free) were determined for BPA and 24 of itsalternatives. BPA and 22 alternatives were found to be stable in bioassaymedium for up to 48 h. The diglycidyl ether of BPA (BADGE) was completelydegraded in medium. For BPPH, a loss of up to 37% was observed. Amass balance model was used to derive C_free of the stable chemicals in the bioassay medium. For model validation, C_free in bioassay medium, as well as bindingto proteins and membrane lipids, was measured for BPA and 15 of thealternatives. Experimentally determined C_free agreed with model predictions within a factor of 10. Relative effectpotencies (REPs) of the BPA alternatives compared to BPA were calculatedbased on nominal (REP_nom) and freely dissolved effect concentrations(REP_free) for two representative end points (activationof estrogen receptor α and inhibition of mitochondrial membranepotential). For hydrophilic and moderately hydrophobic BPA alternatives,REP_nom and REP_free agreed well. REP_free was up to 147× higher compared to REP_nom for hydrophobicand highly protein-bound alternatives like TCBPA or BPAF. BP-MIBKhad the highest estrogenic potential based on REP_nom, butthe highly protein-bound BPAF was identified as most potent basedon REP_free. TCBPA has a lower estrogenic potential thanBPA based on REP_nom (0.11), but based on REP_free (12.8), it is more potent than BPA. The results of the present studyhighlight the need to correct for different in vitro exposure conditions to allow accurate hazard ranking.

Trichloroethylene (TCE), a widespreadenvironmental and occupationalhazardous chemical, exerts potent nephrotoxic effects, while its specificmolecular mechanism remains largely elusive. In this investigation,we revealed that the exposure of HK-2 cells to environmentally relevantconcentrations of TCE resulted in a significantly enhanced expressionof Golgi phosphoprotein 3 (GOLPH3). Meanwhile, TCE treatment couldsignificantly hinder the autophagic flux of HK-2 cells. Further, TCEexposure significantly reduced intracellular lysosomal acidity andimpaired their functional activity. Notably, genetic silencing ofGOLPH3 via RNA interference (RNAi) strategy effectively relieved autophagicobstruction and restored lysosomal acidity and function activity.In summary, this study indicates that TCE can induce severe autophagicflux disturbance and lysosomal acidity and function activity in HK-2cells, in which GOLPH3 plays a key regulatory role and is expectedto serve as a potential biomarker and intervention target for TCE-inducednephrotoxic damage.

The extensive utilization of copper oxide nanoparticles(Nano-CuO)in industrial applications has heightened the concerns about humanhealth effects after occupational or nonoccupational exposure. Previousstudies demonstrated that Nano-CuO exposure caused pulmonary injury,inflammation, and fibrosis. However, the underlying mechanisms havenot been fully understood. Here, we proposed that matrix metalloproteinase-3(MMP-3) might play an important role in Nano-CuO-induced disruptionof tight junction-associated proteins. Our results demonstrated thatexposure of BEAS-2B cells to Nano-CuO, but not Nano-TiO₂, caused a significant dose-dependent increase in MMP-3 productionand activity, and MMP-3 activity reached the highest level at 12 hafter exposure. In addition, Nano-CuO exposure also disrupted theexpression of tight junction-associated proteins (ZO-1, occludin,and claudin-1), whereas MMP-3 siRNA transfection effectively reversedthis downregulation in BEAS-2B cells. In vivo, Nano-CuOexposure caused increased expression of MMP-3 protein and reducedexpression of tight junction-associated proteins (ZO-1 and occludin)in mouse lungs, thereby leading to epithelial barrier dysfunctionand increased lung permeability. Knocking down MMP-3 significantlyattenuated Nano-CuO-induced lung permeability through restorationof ZO-1 and occludin expression in mouse lungs. These findings suggestthat Nano-CuO exposure significantly upregulates MMP-3 in BEAS-2Bcells and mouse lungs, which mechanistically disrupts tight junction-associatedproteins and compromises epithelial barrier integrity. These findingsprovide critical insights into the pathogenesis of pulmonary inflammationand fibrosis caused by metal nanoparticle exposure.

The adverse health effects of fine particulate matter(PM_2.5) remain a pressing public-health concern. Untargetedmetabolomics,which provides global and relatively unbiased profiling of small-moleculemetabolites, offers unique opportunities to discover new exposurebiomarkers and clarify toxicity mechanisms beyond those of conventionalapproaches. This systematic review synthesizes 42 untargeted metabolomicsstudies in PM_2.5-exposed rodent models. Integrative analysisidentifies 1,276 unique exposure-associated metabolites under thepredefined lenient screening criteria, with 24 high-frequency metabolites(e.g., taurine, citric acid, alanine) and 31 recurrently perturbedpathways (e.g., purine metabolism, linoleic acid metabolism, and tricarboxylicacid cycle). These convergent alterations suggest three recurringmechanistic themes: an energy-crisis-driven redox imbalance, oxidative-stress-mediatedepigenetic dysregulation, and inflammatory signaling via polyunsaturatedfatty acids. Tissue-specific metabolic alterations were prominentin lungs (e.g., altered arachidonic acid), liver (e.g., purine catabolism),and brain (e.g., lipid homeostasis disruption), demonstrating multiorganpathophysiological responses. Nevertheless, substantial methodologicalheterogeneity across exposure protocols, animal models, and analyticalplatforms complicates cross-study comparisons. These findings supportuntargeted metabolomics as a high-throughput approach for characterizingPM_2.5-induced metabolic toxicity and offer an applicableframework that may also inform studies of other emerging contaminantsand mixtures. Further progress will require more standardized analyticalworkflows, complementary platform integration, and cross-species validation.

The atmospheric aging of nicotine-containing e-cigaretteaerosolsin the presence of nitrogen dioxide (NO₂) and ozone (O₃) may significantly alter their chemical composition and toxicologicalproperties. However, the chemical transformation of nicotine withinthe complex e-cigarette aerosol matrix and its interaction with atmosphericoxidants remain largely uncharacterized. This study investigated theformation of tobacco-specific nitrosamines (TSNAs) from the agingof nicotine in e-cigarette aerosols exposed to NO₂ andO₃. We observed that NO₂ facilitates the formationof N′-nitrosonornicotine (NNN) through direct nitrosation pathways,whereas O₃ alone does not significantly contribute to TSNAproduction. In contrast, the presence of both oxidants enhanced theformation of nicotine-derived nitrosamine ketone (NNK), suggestingthat oxidative transformation of nicotine plays an important rolein its formation. FIGAERO-ToF-CIMS measurements identified severalnicotine transformation products in aged e-cigarette aerosols thatare not present in the fresh emissions, indicating chemical transformationduring aging. Furthermore, long-amplicon qPCR analysis revealed thatBEAS-2B cells exposed to aged aerosols experienced increased DNA damage,demonstrating enhanced genotoxic potential relative to their freshcounterparts. These findings indicate that atmospheric aging can convertnicotine into carcinogenic TSNAs and other reactive intermediates,increasing the potential health risks associated with secondhand andthirdhand vaping exposure. This work highlights the importance ofconsidering chemically aged aerosols when evaluating the toxicityof e-cigarette emissions.

MitoNEET is a [2Fe-2S]-containing enzyme proposed tofunction asa cellular redox-stress sensor. Here we found that purified mitoNEETbinds to and directly acts on the cell’s central redox regulator,glutathione (GSH), to its corresponding disulfide (GSSG) without requiringoxygen. The conversion of GSH was monitored by HPLC using an optimizedisocratic elution protocol and by spectroscopy with Ellman’sreagent. Additionally, mitoNEET’s novel enzyme activity isconsidered in the context of reactive electrophiles. Taken together,the discovery of mitoNEET’s GSH reactivity opens the path towarda deeper mechanistic understanding of how mitoNEET senses oxidativestress in the cell.

Ongoing investigation into the effects of thallium (Tl)on thecentral nervous system (CNS) yields new insights into the mechanismsunderlying its toxicity. Tl is a heavy metal possessing specific chemicalfeatures that can induce severe acute and chronic intoxications inhumans upon environmental, occupational, accidental, and intentionalexposures, with adverse effects in all organs, and particularly thebrain. It readily crosses the blood-brain barrier (BBB) and triggersa series of cellular, molecular, and biochemical events, leading toalterations at the central level that compromise cognitive and motorfunctions. Experimental studies have shown that Tl alters intracellularredox status to induce oxidative damage, mitochondrial dysfunction,apoptotic cell signaling, and major motor changes. In this reviewwe update, in an integrative manner, the mechanisms of Tl toxicityat several functional levels. This state-of-the-art review compilesand discusses recent and past evidence to integrate a general toxicscheme to explain Tl’s neurotoxic effects, identifying effectivetherapeutic approaches against the neurological sequelae of Tl exposure.The review also describes the use of Tl in nanoparticles for severalindustrial and pharmacological processes, highlighting it as a potentialsource of human exposure. Combined, the evidence reviewed in thisdocument led us to suggest an integrative mechanistic hypothesis onthe toxic effects of Tl.

Regulatory agencies require comprehensive genotoxicityassessmentsfor novel small-molecule therapeutics prior to human trials. Developersoften delay these studies until a candidate is nearing regulatorysubmission because they are expensive and secondary to bioactivity.This timing creates a bottleneck where late-stage failures can jeopardize>$10 million in capital and multiple years of developmental progressper candidate. The Ames assay detects a molecule’s mutagenicpotential. Regulators now explicitly support the use of in silicoAmes mutagenicity models through enabling legislation, dedicated FDAAI toxicology programs, internationally harmonized guidelines, andbenchmark challenges. However, current Ames models suffer from a sensitivitydrop-off when they evaluate molecules outside their training domain.The out-of-domain (OOD) test set from the Second Ames/QSAR InternationalChallenge Project yielded a participant average sensitivity of 0.46.Sensitivity is the most important metric in Ames prediction becausefalse negatives allow mutagenic compounds to advance undetected andtrigger the most costly late-stage failures. Attempts to fix thissensitivity drop-off often reduce overall model performance, whichcan be represented by balanced accuracy (BA). For example, DeepAmesreports high levels of sensitivity (0.87) by sacrificing its BA (0.52).We introduce AmesNet, a task-conditioned Ames model that achievesclass-leading performance in novel chemical spaces. AmesNet utilizesa dual branch architecture containing a molecular encoder and a dedicatedchannel to condition Ames assay context such as metabolic activationand bacterial strain type. In comparative benchmarks, AmesNet reacheda sensitivity of 0.72 (95% CI: 0.68–0.76) and a simultaneousBA of 0.81 (95% CI: 0.78–0.83) on the OOD test data. This performanceimproves discrimination without the sensitivity trade-off limitingexisting approaches. Structural analysis demonstrates that AmesNetrecovers difficult-to-detect mutagenic compounds missed by existingmodels. This framework provides a high-confidence filtering mechanismthat enables drug developers to turn a costly late-stage safety bottleneckinto an earlier-stage mutagenicity triage.