Aging is a major risk factor for metabolic liver disorders, yet the molecular mechanisms driving hepatic aging remain incompletely understood. Here, we identify the Hmgcs2-Pparα signaling pathway that regulates the prolongevity gene Cisd2 and liver aging. Using naturally aged mice, we show that late-life administration of the citrus flavonoid hesperetin restores hepatic Cisd2 expression, improves liver pathology, and partially reverses aging-associated transcriptomic alterations. Mechanistically, we identify Hmgcs2 as a candidate molecular target of hesperetin and demonstrate that Hmgcs2 interacts with Pparα to activate Cisd2 transcription through a Ppar response element (PPRE) in the Cisd2 promoter. Genetic and transcriptomic analyses reveal that the protective effects of hesperetin are largely Cisd2-dependent, as they are significantly attenuated in hepatocyte-specific Cisd2 knockout mice. Consistent with these findings, PPARα and CISD2 expression decline with age in human liver tissues. Together, our results define a signaling axis linking Hmgcs2, Pparα, and Cisd2 that regulates liver aging and suggest that targeting this pathway may represent a strategy to mitigate age-associated hepatic dysfunction.

A series of N-(1,2,3,4-tetrahydro-3-isoquinolinylmethyl)benzamides, which are potent μ-opioid receptor (MOR) agonists, has been discovered. The most promising compound, compound 56 (BPR1M492), is an MOR agonist without a clear signaling bias between cAMP and β-arrestin-2 pathways, a cAMP-biased nociceptin-orphanin FQ opioid peptide agonist, and a weak cAMP-biased δ/κ-opioid receptor agonist. Compound 56 demonstrated potent in vivo antinociception at 0.027 mg/kg, offering rapid pain relief within 5 min of subcutaneous injection. It produced markedly milder withdrawal symptoms than TRV130 in mice, whereas differences in respiratory, gastrointestinal, reward-related, and tolerance-related measures were less pronounced and should be interpreted cautiously in light of the substantially lower dose required for antinociception. Compound 56 is a highly stable, slightly hygroscopic, low-moisture-containing crystalline solid that is safe at its in vivo effective dose.

Drug discovery and development are both time and resource-intensive. The exploration of huge chemical space has increased demand for fast and accurate ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) prediction, necessitating advanced computational frameworks that can balance prediction performance with scalability. To address this issue, a fast and robust model named vADMET was developed for ADMET prediction of small molecules. In this study, a total of 31 binary classification and 10 regression learning tasks of ADMET were evaluated. A biological readacross-inspired multitask learning algorithm, MTForestNet, was utilized to learn the relationship among ADMET tasks and outperformed the state-of-the-art method. Our experimental results demonstrate that vADMET achieves superior predictive accuracy, with an area under the receiver operating characteristic curve (AUROC) of 0.880 and an area under the precision-recall curve (AUPRC) of 0.741 for the classification tasks, and a mean absolute error (MAE) of 8.741 and an R^2 of 0.592 for the regression tasks, respectively. The performance represents 2.56 R^2 ) improvement over the state-of-the-art method. Notably, vADMET achieves a 23-fold speedup over the state-of-the-art method for large-scale screening. Built on the learning across diverse chemical spaces, vADMET is expected to be a valuable tool for drug discovery and development.

Long-term exposure to low-dose food contact materials (FCMs) has raised concerns regarding developmental toxicity. In the present study, we prioritized FCMs with potential developmental toxicity using a weight-of-evidence computational model, which predicted 127 chemicals to be of high concern. From these, we selected 7 chemicals-representing both high and low concern categories-and evaluated their potential embryotoxicity using the mouse embryonic stem cell test (mEST). Among the selected chemicals, thiram most strongly inhibited cardiac differentiation in mEST. We further examined the effects of thiram on morphology and expression of differentiation-related genes in mouse embryonic stem cells (mESCs). Treatment with thiram inhibited 50% early differentiation in mESCs, suppressed the expression of markers associated with the three-germ layers, but increased the expression of neurectoderm markers during early embryogenesis. Additionally, treatment with 20 ng/mL thiram, which was the lowest-observed-effect concentration of cytotoxicity, disrupted neuronal differentiation in both mESCs and human pluripotent embryonal carcinoma NT2 cells. Finally, based on transcriptome analysis, 20 and 30 ng/mL thiram disrupted the neural crest differentiation pathway, altering the expression of genes including homeobox A1 (HOXA1), homeobox B1 (HOXB1), Heart And Neural Crest Derivatives Expressed 1 (HAND1), Distal-Less Homeobox 5 (DLX5), and transcription factor AP-2 alpha (TFAP2A) in NT2 cells. Therefore, disruption of neural crest differentiation is one of the potential mechanisms underlying thiram-induced embryotoxicity. The integrated alternative approach adopted in the present study to identify mechanism-based biomarkers for thiram-induced embryotoxicity in a human-relevant model could facilitate safety assessment for data-poor chemicals in future.

Accurate prediction of chemical properties and toxicological profiles is a critical challenge in cheminformatics and drug discovery. The Simplified Molecular Input Line Entry System (SMILES) provides a textual representation of molecular structures, enabling machine learning models to analyze large-scale datasets efficiently. Deep learning models, such as graph neural networks and sequence models, have demonstrated powerful capabilities in extracting features from SMILES and improving accuracy across various molecular prediction tasks. This study leverages advanced large language models (LLMs), including GPT3.5 Turbo, GPT-4o, Gemini 1.5 Flash, and Gemini 1.5 pro, to perform zero-shot and few-shot learning on SMILES data. By comparing these approaches to traditional deep learning methods and prior work, we explore their potential in capturing contextual and sequential information for chemical property prediction. Our experiments demonstrate that zero-shot and few-shot learning enable high predictive accuracy in scenarios with limited labeled data, providing significant improvements over earlier methods. By systematically evaluating the performance of different LLMs on toxicity and physicochemical datasets, we highlight their strengths in enhancing prediction efficiency while addressing key challenges, such as dataset variability and interpretability. This work underscores the potential of LLMs to transform cheminformatics by offering scalable and versatile tools for molecular property prediction and accelerating advancements in drug discovery.

Betaine, a methylated glycine derivative, shows promise in treating neuropsychiatric and neurological disorders, but its mechanisms remain unclear. This study investigates betaine's concentration-dependent effects on NMDA receptor-mediated excitatory field potentials (EFPs) in mouse medial prefrontal cortex slices and calcium influx in cultured cortical neurons. We further assessed subtype specificity using HEK293 cells stably expressing GluN1/GluN2A, GluN1/GluN2B, or GluN1/GluN2C NMDA receptor subtypes, measuring calcium flux and single-channel activity. In cortical slices, betaine alone did not alter EFPs but, when co-applied with glutamate, significantly increased EFP frequency and amplitude. Conversely, co-application with glutamate and glycine markedly reduced EFPs. Similarly, in cultured cortical neurons, betaine enhanced NMDA receptor-dependent calcium influx with glutamate alone but attenuated it when both glutamate and glycine were present. In HEK293 cells, betaine exhibited dual modulatory effects on NMDA receptor-mediated calcium influx, varying with GluN2 subtype and glutamate/glycine concentrations. Cell-attached patch-clamp recordings confirmed that betaine with glutamate induced NMDA receptor currents, which were reduced by glycine co-application. Molecular docking and dynamics simulations proposed that betaine binds effectively to the glycine-binding site on the GluN1 subunit, with stable interactions. These findings identify betaine as a context-dependent modulator of NMDA receptors via its interaction with the glycine-binding site, offering a novel mechanism that may underlie its therapeutic potential in disorders involving NMDA receptor dysfunction.

RNA modifications are critical in regulating gene expression and cell functions by affecting RNA stability, splicing, translation, and degradation. The catalytic core of N6-adenosine-methyltransferase catalytic subunit METTL3 has emerged as a key enzyme in tumorigenesis by enhancing the translation efficiency of oncogenic transcripts, which is a promising therapeutic target for cancers, including acute myeloid leukemia. In this study, we presented a novel METTL3 inhibitory bioactivity (pIC50) prediction model (ML3-mix-DPLIFE) by combining machine learning, protein-ligand docking, and protein-ligand interaction analysis, through encoding the conventional physicochemical properties, chemical fingerprint, and the docking-based protein-ligand interaction features (DPLIFE) with leveraging auto-stacking 6 algorithms. A feature selection algorithm further optimized the model (ML3-mix-DPLIFE-FS) and obtained a promising mean squared error (MSE) of 0.261 and a Pearson's correlation coefficient (CC) of 0.853 on an independent test dataset, and identified 8 residues critical for ligand interactions with METTL3. To further test the model, the pIC50s of recently reported inhibitors were predicted using the ML3-mix-DPLIFE-FS model, and a good MSE of 0.418 and CC of 0.727 were obtained. This innovative strategy seamlessly integrates machine learning prediction with structural biology insights and reveals a novel way to identify key protein-ligand interactions for further structural rational drug design.

Pharmacological inhibition of the cGAS-STING-controlled innate immune pathway is an emerging therapeutic strategy for a myriad of inflammatory diseases. Here, we report GHN105 as an orally bioavailable covalent STING inhibitor. Late-stage diversification of the briarane-type diterpenoid excavatolide B allowed the installation of solubility-enhancing functional groups while enhancing its activity as a covalent STING inhibitor against multiple human STING variants, including the S154 variant responsible for a genetic autoimmune disease. Selectively engaging the membrane-proximal Cys91 residue of STING, GHN105 dose-dependently inhibited cGAS-STING signaling and type I interferon responses in cells and in vivo. Moreover, orally administered GHN105 exhibited on-target engagement in vivo and markedly reversed key pathological features in a delayed treatment of the acute colitis mouse model. Our study provided proof of concept that the synthetic briarane analog GHN105 serves as a safe, site-selective, and orally active covalent STING inhibitor and devises a regimen that allows long-term systemic administration.

Zebrafish is an effective model organism for toxicological investigations due to their tiny size, quick reproduction, and conserved vertebrate biology. As environmental pollutants continue to increase, it becomes challenging to detect all chemical-related hazards using zebrafish models. In silico models can facilitate prioritizing chemicals prior to further experimental evaluations and providing potential underlying mechanisms. Chemical-phenotype inference system for zebrafish (ZFinfer), an enrichment analysis tool that can predict affected endpoints, was developed by integrating chemical-protein interaction data from the Search Tool for Interacting Chemicals (STITCH) database and gene-phenotype annotation data from the Zebrafish Information Network (ZFIN). Currently, 419,328 chemicals, 23,180 zebrafish proteins, and 3,104 phenotypes for zebrafish were curated and included in the system. ZFinfer has been validated using 777 ToxCast chemicals and 51 priority pollutants from the USEPA. The inference results demonstrated a sensitivity of 0.72 in critical morphological endpoints and a 93 % rediscovery rate for known toxicity records in the ECOTOX knowledgebase. Furthermore, the affected endpoints of 5,195 PFAS chemical exposures were inferred to fill data gaps. ZFinfer could be useful to prioritize chemicals that should be further evaluated and may be applicable in drug discovery and environmental chemical hazard prediction.

Pulmonary absorption is an important route for drug delivery and chemical exposure. To streamline the chemical assessment process for the reduction of animal experiments, several animal-free models were developed for pulmonary absorption research. While Calu-3 and Caco-2 cells and their derived computational models were used in estimating pulmonary permeability, the ex vivo isolated perfused lung (IPL) models are considered more clinically relevant measurements. However, the IPL experiments are resource-consuming making it infeasible for the large-scale screening of potential inhaled toxicants and drugs. In silico models are desirable for estimating pulmonary absorption. This study presented a novel machine learning method that employed an extratrees-based multitask learning approach to predict the IPL absorption rate constant (kaIPL) of various chemicals. The shared permeability knowledge was extracted by simultaneously learning three relevant tasks of Caco-2 and Calu-3 cell permeability and IPL absorption rate. Seven informative physicochemical descriptors were identified. A rigorous evaluation of the developed prediction model showed good performance with a high correlation between predictions and observations (r = 0.84) in the independent test dataset. Two case studies of inhalation drugs and respiratory sensitizers revealed the potential application of this model, which may serve as a valuable tool for predicting pulmonary absorption of chemicals.

Indole-3-acetic acid (IAA), a protein-bound uremic toxin resulting from gut microbiota-driven tryptophan metabolism, increases in hemodialysis (HD) patients. IAA may induce endothelial dysfunction, inflammation, and oxidative stress, elevating cardiovascular and cognitive risk in HD patients. However, research on the microbiome–IAA association is limited. This study aimed to explore the gut microbiome’s relationship with plasma IAA levels in 72 chronic HD patients aged over 18 (August 2016–January 2017). IAA levels were measured using tandem mass spectrometry, and gut microbiome analysis utilized 16s rRNA next-generation sequencing. Linear discriminative analysis effect size and random forest analysis distinguished microbial species linked to IAA levels. Patients with higher IAA levels had reduced microbial diversity. Six microbial species significantly associated with IAA levels were identified; Bacteroides clarus, Bacteroides coprocola, Bacteroides massiliensi, and Alisteps shahii were enriched in low-IAA individuals, while Bacteroides thetaiotaomicron and Fusobacterium varium were enriched in high-IAA individuals. This study sheds light on specific gut microbiota species influencing IAA levels, enhancing our understanding of the intricate interactions between the gut microbiota and IAA metabolism.

The drug discovery of G protein-coupled receptors (GPCRs) superfamily using computational models is often limited by the availability of protein three-dimensional (3D) structures and chemicals with experimentally measured bioactivities. Orphan GPCRs without known ligands further complicate the process. To enable drug discovery for human orphan GPCRs, multitask models were proposed for predicting half maximal effective concentrations (EC50) of the pairs of chemicals and GPCRs. Protein multiple sequence alignment features, and physicochemical properties and fingerprints of chemicals were utilized to encode the protein and chemical information, respectively. The protein features enabled the transfer of data-rich GPCRs to orphan receptors and the transferability based on the similarity of protein features. The final model was trained using both agonist and antagonist data from 200 GPCRs and showed an excellent mean squared error (MSE) of 0.24 in the validation dataset. An independent test using the orphan dataset consisting of 16 receptors associated with less than 8 bioactivities showed a reasonably good MSE of 1.51 that can be further improved to 0.53 by considering the transferability based on protein features. The informative features were identified and mapped to corresponding 3D structures to gain insights into the mechanism of GPCR-ligand interactions across the GPCR family. The proposed method provides a novel perspective on learning ligand bioactivity within the diverse human GPCR superfamily and can potentially accelerate the discovery of therapeutic agents for orphan GPCRs.

Background Fipronil (FPN) is a broad-spectrum pesticide and commonly known as low toxicity to vertebrates. However, increasing evidence suggests that exposure to FPN might induce unexpected adverse effects in the liver, reproductive, and nervous systems. Until now, the influence of FPN on immune responses, especially T-cell responses has not been well examined. Our study is designed to investigate the immunotoxicity of FPN in ovalbumin (OVA)-sensitized mice. The mice were administered with FPN by oral gavage and immunized with OVA. Primary splenocytes were prepared to examine the viability and functionality of antigen-specific T cells ex vivo. The expression of T cell cytokines, upstream transcription factors, and GABAergic signaling genes was detected by qPCR. Results Intragastric administration of FPN (1–10 mg/kg) for 11 doses did not show any significant clinical symptoms. The viability of antigen-stimulated splenocytes, the production of IL-2, IL-4, and IFN-γ by OVA-specific T cells, and the serum levels of OVA-specific IgG 1 and IgG 2a were significantly increased in FPN-treated groups. The expression of the GABAergic signaling genes was notably altered by FPN. The GAD67 gene was significantly decreased, while the GABAR β2 and GABAR δ were increased. Conclusion FPN disturbed antigen-specific immune responses by affecting GABAergic genes in vivo. We propose that the immunotoxic effects of FPN may enhance antigen-specific immunity by dysregulation of the negative regulation of GABAergic signaling on T cell immunity.

Hair analysis is a crucial method in forensic toxicology with potential applications in revealing doping histories in sports. Despite its widespread use, knowledge about detectable substances in hair is limited. This study systematically assessed the detectability of prohibited substances in sports using a multifaceted approach. Initially, an animal model received a subset of 17 model drugs to compare dose dependencies and detection windows across different matrices. Subsequently, hair incorporation data from the animal experiment were extrapolated to all substances on the World Anti-Doping Agency’s List through in-silico prediction. The detectability of substances in hair was further validated in a proof-of-concept human study involving the consumption of diuretics and masking agents. Semi-quantitative analysis of substances in specimens was performed using ultra-performance liquid chromatography–tandem mass spectrometry. Results showed plasma had optimal dose dependencies with limited detection windows, while urine, faeces, and hair exhibited a reasonable relationship with the administered dose. Notably, hair displayed the highest detection probability (14 out of 17) for compounds, including anabolic agents, hormones, and diuretics, with beta-2 agonists undetected. Diuretics such as furosemide, canrenone, and hydrochlorothiazide showed the highest hair incorporation. Authentic human hair confirmed diuretic detectability, and their use duration was determined via segmental analysis. Noteworthy is the first-time reporting of canrenone in human hair. Anabolic agents were expected in hair, whereas undetectable compounds, such as peptide hormones and beta-2 agonists, were likely due to large molecular mass or high polarity. This study enhances understanding of hair analysis in doping investigations, providing insights into substance detectability.

Non-animal assessment of skin sensitization is a global trend. Recently, scientific efforts have been focused on the integration of multiple evidence for decision making with the publication of OECD Guideline No. 497 for defined approaches to skin sensitization. The integrated testing strategy (ITS) methods reported by the guideline integrates in chemico, in vitro, and in silico testing to assess both hazard and potency of skin sensitization. The incorporation of in silico methods achieved comparable performance with fewer experiments compared to the traditional two-out-of-three (2o3) method. However, the direct application of current ITSs to agrochemicals can be problematic due to the lack of agrochemicals in the training data of the incorporated in silico methods. To address the issue, we present ITS-SkinSensPred 2.0 for agrochemicals and agrochemical formulations using a reconfigured in silico model SkinSensPred for pesticides. Compared to ITSv2, the proposed ITS-SkinSensPred 2.0 achieved an 11% and 16% improvement in the accuracy and correct classification rate for hazard identification and potency classification, respectively. In addition, an online ITS tool was implemented and available on the SkinSensDB website. The tool is expected to be useful for evaluating skin sensitization of substances.

Data scarcity is one of the most critical issues impeding the development of prediction models for chemical effects. Multitask learning algorithms leveraging knowledge from relevant tasks showed potential for dealing with tasks with limited data. However, current multitask methods mainly focus on learning from datasets whose task labels are available for most of the training samples. Since datasets were generated for different purposes with distinct chemical spaces, the conventional multitask learning methods may not be suitable. This study presents a novel multitask learning method MTForestNet that can deal with data scarcity problems and learn from tasks with distinct chemical space. The MTForestNet consists of nodes of random forest classifiers organized in the form of a progressive network, where each node represents a random forest model learned from a specific task. To demonstrate the effectiveness of the MTForestNet, 48 zebrafish toxicity datasets were collected and utilized as an example. Among them, two tasks are very different from other tasks with only 1.3% common chemicals shared with other tasks. In an independent test, MTForestNet with a high area under the receiver operating characteristic curve (AUC) value of 0.911 provided superior performance over compared single-task and multitask methods. The overall toxicity derived from the developed models of zebrafish toxicity is well correlated with the experimentally determined overall toxicity. In addition, the outputs from the developed models of zebrafish toxicity can be utilized as features to boost the prediction of developmental toxicity. The developed models are effective for predicting zebrafish toxicity and the proposed MTForestNet is expected to be useful for tasks with distinct chemical space that can be applied in other tasks.Scieific contributionA novel multitask learning algorithm MTForestNet was proposed to address the challenges of developing models using datasets with distinct chemical space that is a common issue of cheminformatics tasks. As an example, zebrafish toxicity prediction models were developed using the proposed MTForestNet which provide superior performance over conventional single-task and multitask learning methods. In addition, the developed zebrafish toxicity prediction models can reduce animal testing.

Pharmacological inhibition of cGAS-STING-controlled innate immune pathway is an emerging therapeutic strategy for a myriad of inflammatory diseases, including autoimmune disease, ulcerative colitis, non-alcoholic fatty liver disease and aging-related neurodegeneration. Here we report GHN105 as an orally bioavailable covalent STING inhibitor. Late-stage diversification of the briarane-type diterpenoid excavatolide B allowed the installation of solubility-enhancing functional groups while enhancing its activity as a covalent STING inhibitor against multiple human STING variants, including the S154 variant responsible for a genetic autoimmune disease. Selectively engaging the membrane-proximal Cys91 residue of STING, GHN105 dose-dependently inhibited cGAS-STING signaling and type I interferon responses in cells and in vivo. Orally administered GHN105 exerted marked therapeutic efficacy and reversed key pathological features in a delayed-treatment acute colitis mouse model. Notably, we also showed that GHN105 covalently engaged STING in the colon tissues. Our study provided proof of concept that synthetic briarane analog GHN105 serves as a safe and orally active covalent STING inhibitor. With a growing number of chronic inflammatory diseases linked to aberrant STING activation, orally bioavailable STING inhibitors would benefit patients by lowering the infection risk from frequent injections while allowing long-term systemic administration.