SEARCH · bioRxiv Science

Results for “pharmacology and toxicology”

Search indexed bioRxiv preprints in genomics, neuroscience, cell biology and bioinformatics. Read source abstracts and check manuscript versions; preprints are not peer reviewed.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 109 records · Page 6Linked to original sources

bioRxiv2026

Human internal exposures to alternariol and its monomethyl ether are predicted below thresholds of in vitro toxicity by physiologically based kinetic modeling

The foodborne mycotoxins alternariol (AOH) and alternariol monomethyl ether (AME) have been associated with several adverse effects, including cytotoxicity, genotoxicity, endocrine disruption, and immunomodulation. As these endpoints are typically observed in vitro at micromolar concentrations, the question arises whether such levels are attainable in exposed humans. To address this data gap in chemical risk assessment, a physiologically based kinetic (PBK) model was developed to predict internal exposure doses to AOH and AME in humans. As input parameters, kinetic constants for hepatic glucuronidation were obtained in vitro by incubating Sprague Dawley rat and human liver S9 fractions with 0.5-50 {micro}M AOH and 0.5-20 {micro}M AME, demonstrating rapid biotransformation in both species. Intestinal absorption of AME and physicochemical parameters were estimated using quantitative structure-activity relationship (QSAR) models. Sensitivity analysis identified parameters describing hepatic glucuronidation and gastrointestinal uptake as among the most influential, confirming the importance of their reliable estimation. The PBK model was evaluated against available rodent toxicokinetic data and subsequently extrapolated to humans. Ultimately, the currently available exposure estimates published by EFSA in 2016 were applied to predict target tissue concentrations, which were compared to points of departure (PoDs) for relevant toxicological endpoints. Even in the most susceptible group of male toddlers, predicted internal concentrations (10{square}{square} {micro}M range) were approximately four orders of magnitude below the respective PoDs. Consequently, under the applied exposure assumptions and considering the compounds as isolated chemicals, AOH and AME are not expected to reach systemic or tissue concentrations associated with the investigated effects.

pharmacology and toxicology↗

bioRxiv2025

Strain and sex effects on blood accumulation of lead, chromium, and cadmium in strains from the Collaborative Cross mouse population

Lead (Pb), chromium (Cr), and cadmium (Cd) are heavy metals that contaminate sites throughout North America. Historically, toxicological effects of Pb, Cr, or Cd compounds have been investigated in a hybrid mouse strain, B6C3F1. However, humans have more genetic diversity and population variability in response to toxicants than is represented in this homogeneous mouse model, which leaves genetic effects on dose response uncertain. Use of the Collaborative Cross (CC) addresses the problem of limited genetic diversity inherent in models like B6C3F1. In previous work, blood Pb levels in panel of female CC lines exposed to high-dose (0.1%) lead acetate showed a strain dependent response. Four strains from the original study with varying Pb blood levels after exposure were selected to determine if strain and sex dependence was exhibited in a two-week acute exposure to Pb, but also to Cr or Cd exposure. To investigate genetic background influence on metal deposition, five animals of each sex from each strain were placed on an American diet for one week prior to dosing high- (0.1%) or low- (0.01%) dose Pb acetate, high- (0.1%) or low- (0.01%) sodium dichromate, or high- (0.1%) or low- (0.01%) cadmium chloride via drinking water ad libitum for 14-days, matching the standard short-term exposure of the National Toxicology Program. Body composition was measured before the start of dosing and prior to necropsy using EcoMRI. Blood Pb at necropsy from this study suggests the strain dependent trends observed in previous exposures is conserved for acute Pb exposure but with different trends for Cr and Cd indicating that even a small panel of strains will not suffice for estimating variation across all toxicants.

pharmacology and toxicology↗

bioRxiv2023

Beyond the nucleus: Plastic chemicals activate G protein-coupled receptors

O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=84 SRC="FIGDIR/small/560665v1_ufig1.gif" ALT="Figure 1"> View larger version (23K): org.highwire.dtl.DTLVardef@e3b5edorg.highwire.dtl.DTLVardef@1bcf3aeorg.highwire.dtl.DTLVardef@224012org.highwire.dtl.DTLVardef@9c61ea_HPS_FORMAT_FIGEXP M_FIG C_FIG G protein-coupled receptors (GPCRs) are central mediators of cell signaling and physiological function. Despite their biological significance, GPCRs have not been widely studied in the field of toxicology. Herein, we investigated these receptors are novel targets of plastic chemicals using a high-throughput drug screening assay with 126 human non-olfactory GPCRs. In a first-pass screen, we tested the activity of triphenol phosphate, bisphenol A, and diethyl phthalate as well as three real-world mixtures of chemicals extracted from plastic food packaging covering all major polymer types. We found 11 GPCR-chemical interactions, of which the chemical mixtures exhibited the most robust activity at Adenosine receptor 1 (ADORA1) and Melatonin receptor 1 (MTNR1A) in a confirmatory secondary screen. We further confirmed that polyvinyl chloride and polyurethane products contain ADORA1 or MTNRA1 agonists using pharmacological knockdown experiments. Finally, an analysis of the associated gene ontology terms suggests that ADORA1 and MTNR1A activation may be linked to downstream effects on circadian and metabolic processes. Our findings exemplify the diversity of endpoints plastic chemicals can target and demonstrate the relevance of non-genomic pathways which have thus far remained unexplored.

pharmacology and toxicology↗

bioRxiv2025

Network toxicology, AlphaFold 3 and molecular docking for analysing the toxicity of heavy metal ions and their organic complexes in environmental pollutants: A case of Cobalt(II) ion & Bis-(2,4-pentanedionato)cobalt(II)

The aim of this study is to advance the application of network toxicology, AlphaFold 3 and molecular docking techniques in the toxicity analysis of heavy metal ions (HMIs) and heavy metal ions organic complexes (HMIOCs) in environmental pollutants. In this study, the potential toxicity and molecular mechanisms of HMIs and HMIOCs in environmental pollutants were effectively investigated using Cobalt(II) ion & Bis-(2,4-pentanedionato)cobalt(II) (Co(II) & BPCo(II)) as examples. It was found that Co(II) & BPCo(II) regulates multiple signaling pathways by modulating hub targets such as APP, NR3C1, ESR1, CASP3, MMP9, and PTGS2. These pathways include Serotonergic synapse, Cocaine addiction, Pathways of neurodegeneration-multiple diseases, Calcium signaling pathway, Neuroactive ligand-receptor interactions, Alzheimer disease and other pathways, which in turn affects the Blood-Brain Barrier (BBB) in order to destroy neuronal cells, ultimately leading to a variety of mental illnesses (MIs) and nervous system diseases (NSDs), such as Schizophrenia, Depressive disorder, nervous system disorder and Alzheimer Disease, Late Onset. In this study, the toxic effects and molecular mechanisms of Co(II) and BPCo(II) were elucidated, while the limitations of conventional toxicological methods--such as safety issues, time constraints, high expenses, ethical concerns regarding animal use, and limited predictive power--were effectively addressed. These results contribute to the foundation for studying disease diagnosis associated with environmental exposure to HMIs and HMIOCs. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=120 SRC="FIGDIR/small/642058v1_ufig1.gif" ALT="Figure 1"> View larger version (42K): org.highwire.dtl.DTLVardef@1147414org.highwire.dtl.DTLVardef@1d0d7dborg.highwire.dtl.DTLVardef@ba53dorg.highwire.dtl.DTLVardef@8d6bb8_HPS_FORMAT_FIGEXP M_FIG C_FIG

pharmacology and toxicology↗

bioRxiv2021

Pharmacokinetics of Orally Administered GS-441524 in Dogs

Despite being FDA-approved for COVID-19, the clinical efficacy of remdesivir (Veklury(R)) remains contentious. We previously pointed out pharmacokinetic, pharmacodynamic and toxicology reasons for why its parent nucleoside GS-441524, is better suited for COVID-19 treatment. Here, we assess the oral bioavailability of GS-441524 in beagle dogs and show that plasma concentrations [~]24-fold higher than the EC50 against SARS-CoV-2 are easily and safely sustained. These data support translation of GS-441524 as an oral agent for COVID-19.

pharmacology and toxicology↗

bioRxiv2026

The Nematicide Tioxazafen Disrupts Proteasome Function via Cytochrome P450 Bioactivation

Tioxazafen is an effective nematicide whose commercialization was halted because handlers reportedly developed rashes after working with seeds coated with a tioxazafen-laced cocktail. Here, we show that tioxazafen is bioactivated into toxic products by nematode and human cytochrome P450s. Through systematic analyses, we found that bioactivated tioxazafen disrupts proteasome function, leading to the accumulation of the NRF1 transcription factor ortholog SKN-1A in the nematode C. elegans, and a bounce-back transcriptional up-regulation of proteasome components. Genetic upregulation of the C. elegans proteasome supresses tioxazafens lethality, indicating that proteotoxicity is a key contributor to death. A survey of human P450s revealed that skin-expressed CYP1A1 toxifies tioxazafen and may account for the reaction to tioxazafen-coated seeds. We also found that rabbit CYP1A1 fails to bioactivate tioxazafen, which may explain the pre-market failure to detect robust adverse skin reactions. Our work highlights vulnerabilities in pre-market toxicological assays and a provides potential solution to prospectively identifying P450 toxication events. One-Sentence SummaryTioxazafen is bioactivated by cytochrome P450s into a proteasome disruptor.

pharmacology and toxicology↗

bioRxiv2022

Proteome integral solubility alteration assay combined with multi-criteria decision-making analysis for developing adverse outcome pathways

Understanding the biological impact of chemicals is hindered by the high number and diversity of compounds in the market. To simplify the chemical risk assessment, the adverse outcome pathway (AOP) method has arisen as a framework to predict the impact of chemical exposure on human and environmental health. The development of this predictive tool requires knowledge of the molecular interaction between chemicals and protein targets. Those molecular initiating events connect alterations of cellular function with physiological impairment. This strategy aims to focus on the complex biological interaction to predict the impact on health. The high-throughput identification of all chemical targets can be obtained by a proteomics-based thermal shift assay, however, selecting the priority target candidate is a biased process strongly dependent on expert knowledge and literature. Here, we unravel new molecular initiating event from a tested chemical combining the target deconvolution by the proteome integral solubility alteration (PISA) assay, and the target selection by an analytical hierarchy process (AHP) approach. In the proof-of-concept study, we identified by PISA assay 8 protein targets for 2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD) from the soluble proteome from hepatic cells containing 2824 proteins. The definition of the AHP approach facilitates the selection of heat shock protein beta-1 (Hspb1) as the most suitable protein for developing AOPs. Our results demonstrated that the process of target identification is independent from a chemical characterization, and that the process of data curation and target selection is less sensitive to lack of toxicological information. We anticipate that this innovative integration of methods could decipher the chemical-protein interactions from new chemicals including the new alternative chemicals designed for chemical replacement and that would discover new molecular initiating events to support more sustainable methodologies to gain time and resources in chemicals assessment. SYNOPSISOur combined methodologies can determine the most suitable target to develop adverse outcome pathways from the proteome-wide protein target identification.

pharmacology and toxicology↗

bioRxiv2026

Network Toxicology, Molecular Docking, and Molecular Dynamics Simulations Reveal the Mechanism of Tetrabromobisphenol A in Bullous Pemphigoid.

Bullous pemphigoid (BP) is an autoimmune blistering disease with a growing incidence, and environmental factors are receiving increasing attention. Tetrabromobisphenol A (TBBPA), a widely used brominated flame retardant, is a significant environmental pollutant. However, the molecular mechanisms by which TBBPA contributes to BP pathogenesis remain unclear. This study integrated network toxicology, molecular docking, and molecular dynamics (MD) simulations to systematically investigate the molecular mechanisms of TBBPA-induced BP. Using network toxicology, we identified 797 potential targets of TBBPA and 446 BP-related targets. A Venn diagram analysis revealed 48 common targets. Protein-protein interaction (PPI) network and topological analyses further identified five core hub targets: TNF, CXCL8, MMP9, ICAM1, and ITGB1. Gene enrichment analysis indicated that these targets were significantly enriched in immune-inflammatory pathways, such as leukocyte migration, inflammatory responses, and the IL-17 signaling pathway, as well as in various pathogen infection and cancer-related pathways. Molecular docking revealed that TBBPA stably binds to all five core targets with binding energies [≤] -5 kcal/mol, driven primarily by hydrophobic interactions and {pi}-{pi} stacking. Subsequent MD simulations confirmed that TBBPA complexes with TNF, CXCL8, and MMP9 remained stable throughout the 100 ns simulation. The overall protein structures remained compact, and the ligands were effectively encapsulated within the binding pockets, forming stable networks of hydrogen bonds and hydrophobic interactions. In conclusion, this study, for the first time, proposes a systematic molecular framework using integrated computational biology. Our findings suggest that the environmental pollutant TBBPA may act as a potential risk factor in BP pathogenesis by targeting core proteins (TNF, CXCL8, and MMP9). These interactions potentially disrupt critical signaling pathways related to immune inflammation, cell migration, and tissue remodeling. This study offers a novel mechanistic hypothesis regarding environmental chemical exposure in autoimmune blistering diseases, although further experimental validation is required. HighlightsO_LINetwork toxicology identified 48 common targets linking Tetrabromobisphenol A(TBBPA) exposure to Bullous Pemphigoid (BP). C_LIO_LIFive core targets (TNF, CXCL8, MMP9, ICAM1, ITGB1) were screened as potential mediators. C_LIO_LITBBPA stably binds to TNF, CXCL8, and MMP9 with binding energies [≤] -5 kcal/mol. C_LIO_LIMolecular dynamics simulations confirm stable binding and structural integrity of complexes. C_LIO_LIThis study provides a mechanistic framework for TBBPA as an environmental risk factor in BP. C_LI

pharmacology and toxicology↗

bioRxiv2025

Validation of a Microsampling-Compatible LC-MS/MS Method for Cannabinoid Quantitation in Whole Blood

Recently developed dried blood sampling methods for cannabinoid quantitation require small blood volumes, making them microsampling-compatible, but have notable limitations including hematocrit-related bias for dried blood spots (DBS) and higher consumable costs for volumetric absorptive microsampling (VAMS(R)). To address these issues, we developed a highly sensitive liquid chromatography-tandem mass spectrometry (LC-MS/MS) method capable of quantifying cannabinoids in 50 {micro}L of liquid whole blood, providing a practical microsampling alternative to dried blood approaches. Using liquid-liquid extraction (LLE) with sodium hydroxide alkalinization and acetonitrile precipitation, followed by quantitative analysis on an Agilent 6495 liquid chromatography-triple quadrupole (LC-TQ) mass spectrometer, we achieved lower limits of quantitation (LLOQs) of 0.10 ng/mL for {Delta}9-tetrahydrocannabinol (THC) and cannabinol (CBN), 0.20 ng/mL for cannabigerol (CBG), 0.30 ng/mL for cannabidiol (CBD), 0.50 ng/mL for 11-hydroxy-THC (11-OH-THC), and 5.0 ng/mL for 11-nor-9-carboxy-THC (THC-COOH). Calibration was linear from the LLOQ to 300 ng/mL for all analytes. To our knowledge, this is the first validated approach for cannabinoid quantitation in less than 100 {micro}L of liquid whole blood with an LLOQ for THC comparable to that with the most sensitive LC-MS/MS methods using standard blood volumes. The achieved LLOQs for other cannabinoids are also suitable for forensic toxicology applications. We anticipate particular utility for obtaining evidence from suspected impaired drivers at the roadside when paired with finger-prick sampling and liquid blood microcollection tubes.This approach enables measurement of THC levels at the time of driving and thereby overcoming current limitations, including the decrease in THC levels that occurs if blood sampling is delayed, the requirement for larger sample volumes ([≥]100 {micro}L), and dependence on trained phlebotomists for venipuncture.

pharmacology and toxicology↗

bioRxiv2025

Preclinical and Toxicology Assessment of ISFP10, an Inhibitor of Fungal Phosphoglucomutase (PGM)

Background and ObjectivePreviously, the novel small molecule ISFP10 has been shown to inhibit fungal phosphoglucomutase (PGM) activity in Aspergillus fumigatus and Pneumocystis spp. With 50-fold selectivity over the human PGM molecule due to the presence of a unique yet conserved cysteine residue present in a number pathogenic fungal PGMs, use of this compound may provide a novel broad-spectrum approach to treating fungal infections. Accordingly, we sought to determine the tolerability in test animals receiving this compound, as well as the potential antifungal activity of ISFP10 on cultures of the common fungal pathogens Candida albicans and Candida glabrata. MethodsC57BL6 mice received once daily intraperitoneal (IP) injections of 100 mL of vehicle control (DMSO) or ISFP10 at a concentration of 10 mg/kg. Body weights were recorded daily for 7 days of treatment. On the final day, mice were weighed and euthanized. Postmortem blood collection was conducted via cardiac puncture and distributed to EDTA and lithium heparin tubes for complete blood count (CBC) and comprehensive blood chemistry panels, respectively. Liver, kidney, and lung tissue were also harvested and placed in 10% formalin for H&E staining and blinded histopathologic scoring. Lung samples were further analyzed for proinflammatory cytokines using enzyme-linked immunosorbent assays (ELISA) and quantitative PCR (qPCR). Furthermore, ISPF10 was tested for antifungal activity via 8-hour growth curve analysis in a concentration-dependent fashion against Candida albicans and Candida glabrata. ResultsThere was no significant difference in the daily or final body weights of the mice receiving 10 mg/kg of ISFP10 compared to those of the vehicle control group. Extracellular matrix (ECM) transcripts for IL-6 and TNF were statistically similar via qPCR. ELISA results of proinflammatory cytokines for IL-6 was not significant whereas TNF levels in lung tissue from the ISFP10 treatment group were significantly reduced, indicating a potential anti-inflammatory effect of ISFP10 at this dosage. Overall, blood chemistry and CBC analysis revealed no overall significant differences between the two groups, except for increased neutrophil counts and decreased potassium levels in samples collected from ISFP10 treated animals compared to the vehicle control group. These laboratory abnormalities were not of clinical significance to the test animals. Blinded histopathological examination revealed no abnormalities or evidence of critical organ toxicity from all groups. Inhibition of C. albicans and C. glabrata culture growth by ISFP10 was concentration-dependent in YPD liquid media containing the ISFP10 compared to vehicle control. ConclusionsOur preliminary testing of ISFP10 revealed no inherent safety or toxicology concerns within the observed parameters. These data further support significant culture suppressive activity against C. albicans and C. glabrata. Taken together, these observations of ISFP10 further indicate that targeting PGM might be a novel and viable therapeutic strategy for serious fungal infections. Key pointsAn inhibitor specific to fungal PGM enzymes, termed ISFP10, was generally well tolerated when administered via intraperitoneal (IP) injection in mice. ISFP10 displays antifungal activity in a concentration-dependent manner against C. albicans and C. glabrata.

pharmacology and toxicology↗

bioRxiv2023

Small molecule nitroalkenes inhibit RAD51-mediated homologous recombination and amplify triple-negative breast cancer cell killing by DNA-directed therapies.

Nitro fatty acids (NO2-FAs) are endogenously generated lipid signaling mediators from metabolic and inflammatory reactions between conjugated diene fatty acids and nitric oxide or nitrite-derived reactive species. NO2-FAs undergo reversible Michael addition with hyperreactive protein cysteine thiolates to induce posttranslational protein modifications that can impact protein function. Herein, we report a novel mechanism of action of natural and non-natural nitroalkenes structurally similar to (E) 10-nitro-octadec-9-enoic acid (CP-6), recently de-risked by preclinical Investigational New Drug-enabling studies and Phase 1 and Phase 2 clinical trials and found to induce DNA damage in a TNBC xenograft by inhibiting homologous-recombination (HR)-mediated repair of DNA double-strand breaks (DSB). CP-6 specifically targets Cys319, essential in RAD51-controlled HR-mediated DNA DSB repair in cells. A nitroalkene library screen identified two structurally different nitroalkenes, a non-natural fatty acid [(E) 8-nitro- nonadec-7-enoic acid (CP-8)] and a dicarboxylate ester [dimethyl (E)nitro-oct-4-enedioate (CP- 23)] superior to CP-6 in TNBC cells killing, synergism with three different inhibitors of the poly ADP-ribose polymerase (PARP) and {gamma}-IR. CP-8 and CP-23 effectively inhibited {gamma}-IR-induced RAD51 foci formation and HR in a GFP-reported assay but did not affect benign human epithelial cells or cell cycle phases. In vivo, CP-8 and CP-23s efficacies diverged as only CP-8 showed promising anticancer activities alone and combined with the PARP inhibitor talazoparib in an HR-proficient TNBC mouse model. As preliminary preclinical toxicology analysis also suggests CP-8 as safe, our data endorse CP-8 as a novel anticancer molecule for treating cancers sensitive to homologous recombination-mediated DNA repair inhibitors. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=172 SRC="FIGDIR/small/552990v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@1dcec56org.highwire.dtl.DTLVardef@1202a09org.highwire.dtl.DTLVardef@a7e7b6org.highwire.dtl.DTLVardef@15d8925_HPS_FORMAT_FIGEXP M_FIG C_FIG

pharmacology and toxicology↗

bioRxiv2026

vToxiNet: a biologically constrained deep learning framework for interpretable prediction of drug-induced hepatotoxicity

Hepatotoxicity remains a leading cause of drug attrition and post-marketing withdrawal, resulting from diverse and complex toxicity mechanisms. Traditional in vitro models can only capture a limited subset of toxicity pathways, and animal studies face translational and ethical limitations. Regulatory agencies have therefore promoted new approach methodologies, including human-relevant assays, omics technologies, and computational models to improve predictive toxicology and support evidence-based decision-making. However, most machine learning models for hepatotoxicity either rely solely on chemical structure or operate as black boxes, limiting mechanistic interpretability and broader applicability. Here, we introduce the virtual toxicity network (vToxiNet), a biologically constrained deep learning framework that embeds systems toxicology knowledge directly into neural network architecture for interpretable hepatotoxicity prediction. vToxiNet integrates chemical descriptors, high-throughput assay responses, transcriptomic signatures, and Reactome pathway hierarchy to construct a virtual adverse outcome pathway network. Across cross-validation and multiple external validation datasets, vToxiNet demonstrates robust predictive performance and generalizes to previously unseen chemicals. Importantly, interpretation of vToxiNet enables gene and pathway-level attribution, supporting mechanism-informed hazard characterization and chemical prioritization. These results demonstrate that encoding biological hierarchy as architectural constraints enables both predictive accuracy and mechanistic insight, establishing a generalizable framework for modeling complex biological outcomes.

pharmacology and toxicology↗

bioRxiv2022

KNOWLEDGE GRAPH AIDS COMPREHENSIVE EXPLANATION OF DRUG TOXICITY

In computational toxicology, prediction of complex endpoints has always been challenging, as they often involve multiple distinct mechanisms. State-of-the-art models are either limited by low accuracy, or lack of interpretability due to their black-box nature. Here we introduce AIDTox, an interpretable deep learning model which incorporates curated knowledge of chemical-gene connections, gene-pathway annotations, and pathway hierarchy. AIDTox accurately predicts cytotoxicity outcomes in HepG2 and HEK293 cells. It also provides comprehensive explanations of cytotoxicity covering multiple aspects of drug activity including target interaction, metabolism, and elimination. In summary, AIDTox provides a computational framework for unveiling cellular mechanisms for complex toxicity endpoints.

pharmacology and toxicology↗

bioRxiv2021

Fluid Menisci and In Vitro Particle Dosimetry of Submerged Cells

Understanding the mechanisms of interaction between cells and particulate nanomaterials lies in the heart of assessing the hazard associated with nanoparticles. The paradigm of toxicology requires quantifying and interpreting dose-response relationships, and cells cultured in vitro and exposed to particle dispersions rely on mathematical models that estimate the received nanoparticle dose. Yet, none of these models acknowledges the fact that aqueous cell-culture media wet the inner surface of hydrophilic open wells, which results in curved fluid-air interface called meniscus. We show that omitting this phenomenon leads to a nontrivial but systematic error and twists the fundamental concept of nanotoxicology. Given that reproducibility and harmonization between meta analyses, in vitro, in silico, and in vivo studies must be improved, we present an adequate mathematical model that greatly advances such efforts.

pharmacology and toxicology↗

bioRxiv2022

Knowledge-guided deep learning models of drug toxicity improve interpretation

In drug development, a major reason for attrition is the lack of understanding of cellular mechanisms governing drug toxicity. The black-box nature of conventional classification models has limited their utility in identifying toxicity pathways. Here we developed DTox (Deep learning for Toxicology), an interpretation framework for knowledge-guided neural networks, which can predict compound response to toxicity assays and infer toxicity pathways of individual compounds. We demonstrate that DTox can achieve the same level of predictive performance as conventional models with a significant improvement in interpretability. Using DTox, we were able to rediscover mechanisms of transcription activation by three nuclear receptors, recapitulate cellular activities induced by aromatase inhibitors and PXR agonists, and differentiate distinctive mechanisms leading to HepG2 cytotoxicity. Virtual screening by DTox revealed that compounds with predicted cytotoxicity are at higher risk for clinical hepatic phenotypes. In summary, DTox provides a framework for deciphering cellular mechanisms of toxicity in silico.

pharmacology and toxicology↗

bioRxiv2025

TICToK: A comprehensive knowledgebase of tattoo ink chemicals and investigation of their associated toxicities and regulations

Tattoos and permanent make-up, involving dermal injection of chemicals, are increasingly popular, with 30-40% of young adults in North America and Europe having at least one tattoo. Recent studies have suggested a possible association between tattooing and health risks, including skin cancers and lymphomas, although no definitive link has yet been established. This study comprehensively catalogs tattoo ink chemicals and investigates their potential adverse effects, addressing the urgent need for greater understanding of tattoo-related health concerns. First, 364 unique tattoo chemicals were identified from various scientific and regulatory sources, with nearly half functioning as pigments. Hazardous chemicals were identified, revealing carcinogens, endocrine disruptors, neurotoxicants, and dermal toxicants. A regulatory analysis based on key EU regulations, including harmonised classifications under Classification, Labelling and Packaging (CLP) regulation, Restriction Entry 75 of the REACH regulation, Cosmetic Products Regulation (CPR), and SVHC candidate list, revealed existing regulatory coverage of tattoo ink chemicals. Curated chemical-disease associations highlighted that some tattoo chemicals are known to cause dermatitis. Further, diverse toxicological information, including experimental results from REACH dossiers, were integrated to construct stressor-AOP network linking 151 chemicals to 362 AOPs, revealing potential carcinogenic mechanisms associated with tattoo ink chemicals. A systems biology approach revealed potential immunomodulatory effects associated with these chemicals. Finally, all findings have been made available through the online database Tattoo Ink Chemicals and associated Toxicities Knowledgebase (TICToK; https://cb.imsc.res.in/tictok), which can support risk assessment and sustainable tattoo practices. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=81 SRC="FIGDIR/small/668261v2_ufig1.gif" ALT="Figure 1"> View larger version (37K): org.highwire.dtl.DTLVardef@6ce8e3org.highwire.dtl.DTLVardef@2dc5aborg.highwire.dtl.DTLVardef@10c67e6org.highwire.dtl.DTLVardef@1795281_HPS_FORMAT_FIGEXP M_FIG C_FIG

pharmacology and toxicology↗

bioRxiv2024

In vitro genotoxic and mutagenic potentials of combustion particles from marine fuels with different sulfur contents

Ship emissions cause serious environmental impacts and adverse effects toward human health. Therefore, the International Maritime Organization (IMO) restricted the fuel sulfur content (FSC) of marine fuels: FSC must be <0.5% m/m or <0.1% m/m in sulfur emission control areas, covering a range of fuels from distillate diesel-like fuels to low-sulfur heavy fuel oils (HFOs). As a result, ship emissions, e.g., sulfur oxides and particulate matter (PM) have been reduced. However, how FSC correlates with the toxicological potential of ship emissions is still uncertain. The objective of this study was to understand how the physical and chemical properties of particulate emissions from a marine engine operating on five marine fuels with different FSCs influence their toxicological outcome. For this scope, cytotoxic, genotoxic, mutagenic, and pro-inflammatory potentials of collected particles were evaluated in lung cell model systems. The involvement of intracellular reactive oxygen species and xenobiotic metabolism was also explored. While PM from different fuels combustion resulted in up to approximately 20% of reduction of cytotoxicity at the highest concentration, other toxicological outcomes, including clonogenic and genotoxic potentials, showed a stronger trend with the polycyclic aromatic hydrocarbon contents in PM compared with FSC. This trend was supported by evidence of a significant increase in gene mutation frequency and alterations in cellular mechanisms induced by an aromatic-rich HFO with an intermediate FSC. In conclusion, apart from reducing FSC in marine fuels, additional particle abatement systems should be considered to reduce the adverse effects of particulate emissions from shipping operations on human health.

pharmacology and toxicology↗

bioRxiv2025

Mouse precision-cut liver and kidney slices: an optimized ex vivo model for acute toxicity testing

As excretion organs, the kidneys and liver are exposed to high concentrations of potentially toxic substances. While animal models remain the gold standard for organ-specific toxicity testing, alternative ex vivo approaches are essential to align with the 3R principles (refinement, reduction, replacement). Precision-cut tissue slices (PCTS) retain native tissue architecture, cellular heterogeneity, the interplay of different cell types, and metabolic capacity, offering a promising link between in vitro and in vivo models. Here, we aimed to establish an optimized protocol for preparing and culturing precision-cut kidney and liver slices (PCKS and PCLS) from mice for use in substance-oriented toxicological tests. Key parameters - including slice thickness, media composition, oxygenation, glucose levels, and incubation time - were refined to maintain tissue viability and metabolic function. Five known toxins - acetaminophen, cyclosporin A, cisplatin, arsenic trioxide, and aristolochic acid I - were tested. While PCKS showed comparable sensitivity to established kidney cell lines, PCLS achieved IC50 values closely matching in vivo toxicity data. High reproducibility across different experimenters was achieved, highlighting the robustness of the model. In conclusion, this ex vivo system provides a valuable, reproducible, and ethically approved platform for acute nephrotoxicity and hepatotoxicity testing, supporting preclinical drug screening and potentially reducing reliance on animal experiments.

pharmacology and toxicology↗