Hepatocellular carcinoma (HCC) exhibits pronounced cellular heterogeneity and dynamic state transitions during tumour progression, yet the emergence of tumour cell states along fate trajectories and their transcriptional regulation remain unclear. Here, we established an integrated framework linking cell fate dynamics with transcriptional regulatory networks using publicly available single-cell RNA sequencing data. Malignant cells were identified by inferred copy number variation, followed by CellRank-based fate inference and SCENIC-based regulatory network reconstruction. We identified three terminal HCC cell states-immune-activated, metabolic, and proliferative hepatocytes-forming a differentiation continuum from stem-like to more mature phenotypes. Each state was governed by a distinct regulatory network with specific core regulators, including IRF3, PPARA, and JUN. Integration with TCGA-LIHC bulk transcriptomic and clinical data revealed that a proliferative state-derived transcriptional signature was associated with poorer overall survival. Together, our study provides a mechanistic framework linking tumour cell fate trajectories, regulatory heterogeneity and clinical outcomes in HCC.

Succinate dehydrogenase (SDH), also termed complex II or succinate-ubiquinone oxidoreductase, is a crucial biological enzyme in the process of mitochondrial oxidative phosphorylation. Succinate dehydrogenase inhibitors (SDHIs) are a promising class of fungicides targeting the energy production pathway of pathogenic fungi. However, overuse has resulted in the emergence of resistance, underscoring the need for novel and effective SDHIs. This study utilized the Transformer model to generate a customized virtual library of potential SDHIs. These candidates were then meticulously screened based on expert knowledge and synthetic feasibility, ultimately yielding several pyrazole carboxamide derivatives as the promising leads. Subsequent synthesis, antifungal activity assessment, and structural optimization further refined these leads into potent SDHI candidates. This work represents the first application of a generative model to SDHI design, establishing a robust workflow encompassing virtual library generation, screening, activity evaluation, and structure optimization. This study paves the way for the rational design of future SDHIs, not only against fungi, but potentially other agricultural pathogens as well.

Single-cell perturbation sequencing technologies (e.g., Perturb-seq, CROP-seq), which integrate CRISPR-based gene editing with single-cell transcriptome profiling, have revolutionized the analysis of transcriptomic changes induced by genetic perturbations at single-cell resolution. These technologies serve as a powerful tool for identifying key genes that inhibit tumor growth or reverse cancer cell phenotypes. However, they face two major challenges: data explosion with high experimental costs, and data complexity characterized by high dimensionality, noise, sparsity, and heterogeneity. To address these challenges, we developed the single-cell Rank-based Genetic Perturbation predictor (scRGP), the first deep learning framework leveraging gene expression rank-order information for this task. scRGP demonstrates superior performance in terms of robustness, cross-cell-line perturbation prediction, and high-throughput screening. Specifically, scRGP achieves an approximately 10-16 percentage points improvement in Pearson correlation coefficient (PCC) over state-of-the-art methods (e.g., GEARS and scFoundation) for single- and double-gene perturbation predictions, while also extending prediction capability to triple-gene perturbations. Furthermore, it outperforms these methods by approximately 5-9 percentage points in cross-cell-line predictions. These advancements promise to shift the paradigm of single-cell perturbation studies from experiment-driven to computation-driven approaches, providing new support for functional genomics and precision medicine. ### Competing Interest Statement The authors have declared no competing interest. National Natural Science Foundation of China, ZX2200521 Anti-tumor New Drug Rapid Translation Public Service Platform of Jiangsu Province, BM2023002

Cancer is a highly heterogeneous disease, with different patients’ cancers potentially having varying genetic mutations, phenotypic characteristics, and molecular mechanisms. Therefore, accurately predicting drug responses is crucial in precision personalized medicine. However, existing research usually focuses solely on the basic information of the drugs themselves, potentially overlooking the impact of interactions between genes and gene pathway-specific combinatorial implications on biological processes or drug responses. In this work, we propose a multi-module hybrid neural framework for drug response prediction (DRP) that learns global and local feature information, called Graph-Pathway-Transformer Drug Response Prediction (GPT-DRP). In GPT-DRP, drugs are represented by molecular graphs using two types of graph neural networks to capture drug structural information, while cell lines are described by gene pathway activity scores employing a fully connected network and a convolutional neural network to capture cell line features. In addition, Transformer is used for the extracted drug and cell line representations to integrate the features learned from the drugs and the cell lines. Experimental results on the CCLE/GDSC datasets show that GPT-DRP outperforms the state-of-the-art models.

CITE-seq provides a powerful method for simultaneously measuring RNA and protein expression at the single-cell level. The integrated analysis of RNA and protein expression in identical cells is crucial for revealing cellular heterogeneity. However, the high experimental costs associated with CITE-seq limit its widespread application. In this paper, we propose scTEL, a deep learning framework based on Transformer encoder layers, to establish a mapping from sequenced RNA expression to unobserved protein expression in the same cells. This computation-based approach significantly reduces the experimental costs of protein expression sequencing. We are now able to predict protein expression using single-cell RNA sequencing (scRNA-seq) data, which is well-established and available at a lower cost. Moreover, our scTEL model offers a unified framework for integrating multiple CITE-seq datasets, addressing the challenge posed by the partial overlap of protein panels across different datasets. Empirical validation on public CITE-seq datasets demonstrates scTEL significantly outperforms existing methods.

Serotonin receptor subtype 1 A (5-HTR1A) is a critical therapeutic target for neuropsychiatric disorders. Crataegus pinnatifida Bge. (hawthorn) exhibits potent antidepressant-like effects via 5-HTR1A activation, yet its bioactive constituents remain unidentified. Herein, we engineered 5-HTR1A-haloalkane dehalogenase fusion protein and immobilized it on 6-bromohexanoic acid-functionalized microspheres to establish affinity-based screening platform. Bioactivity-guided isolation identified malic acid and chlorogenic acid as 5-HTR1A ligands, and their binding energies were-12.9 kcal/mol and-44.7 kcal/mol, respectively. In vitro assays demonstrated that chlorogenic acid significantly protected PC12 cells from corticosterone-induced damage at concentrations as low as 2.5 mu M (p = 0.0305), whereas malic acid exhibited comparable effects at 20 mu M. Both compounds effectively counteracted corticosterone-induced reductions in serotonin, dopamine, and BDNF levels likely through interactions with 5-HTR1A. Collectively, these findings suggest that malic acid and chlorogenic acid are stable, high-affinity 5-HTR1A ligands with significant neuroprotective potential, making them promising lead compounds for developing novel 5-HTR1A-targeted therapeutics.

The binding of transcription factors (TFs) to TF binding sites plays a vital role in the process of regulating gene expression and evolution. With the development of machine learning and deep learning, some successes have been achieved in predicting transcription factors and binding sites. In this paper, we develop a model, BTFBS, which predicts whether the bacterial transcription factors and binding sites combine or not. The model takes both the amino acid sequences of bacterial transcription factors and the nucleotide sequences of binding sites as inputs, and extracts features through convolutional neural network and MultiheadAttention. For the model inputs, we use two negative sample sampling methods: RS and EE. On the test dataset of RS, the accuracy, sensitivity, specificity, F1-score, and MCC of BTFBS are 0.91446, 0.89746, 0.93134, 0.91264, and 0.82946, respectively. Furthermore, on the test dataset of EE, the accuracy, sensitivity, specificity, F1-score and MCC of BTFBS are 0.87868, 0.89354, 0.86394, 0.87996, and 0.75796, respectively. Meanwhile, our findings indicate that the optimal approach for obtaining negative samples in the context of bacterial research is to utilize the whole genome sequences of the corresponding bacteria, as opposed to the shuffling method. The above results on the test dataset have shown that the proposed BTFBS model has a good performance and it can provide an experimental guide.

Modern fungicides have made significant contributions to crop disease management, but the development of resistant fungal strains has caused their failure in disease control. Therefore, developing fungicides with novel action mechanisms is the most effective measure to manage resistance. Quinofumelin, a novel quinoline fungicide, exhibits exceptional antifungal activity against phytopathogens. However, there is currently no available information on its mechanism of action. Here, we used transcriptome and metabolome analysis to observe a co-enrichment pattern of differentially expressed genes (DEGs) and differentially accumulated metabolites (DAMs) within pyrimidine biosynthesis pathway (PBP), identifying down-regulation of dihydroorotate dehydrogenase (DHODH). Exogenous uridine monophosphate (UMP), uridine, or uracil (metabolites in PBP) successfully restored quinofumelin-induced inhibition of mycelial growth in Fusarium graminearum and Fusarium asiaticum. Additionally, the deletion of FgDHODHII was determined to be lethal; however, mycelial growth of ΔFgDHODHII mutants could be restored by adding UMP, uridine, or uracil. These findings indicate that the deficiencies in FgDHODHII are functionally equivalent to complete inhibition of its activity by quinofumelin. Finally, molecular docking, surface plasmon resonance (SPR), and microscale thermophoresis (MST) results strongly support the precise interaction between quinofumelin and FgDHODHII. Collectively, these findings provide compelling evidence for the involvement of de novo uracil biosynthesis as a mechanism of action for quinofumelin while identifying FgDHODHII as its specific target.

Abstract Background Glioma is a brain tumor and glioblastoma represents its most common type, accounting for almost 40% of primary tumors of the central nervous system, and is associated with the worst prognosis. Glioblastoma remains one of the major challenges in neurooncology. Recently, the Food and Drug Administration (FDA) has approved tumor treating fields (TTF) for the treatment of glioblastoma. In this analysis, we aimed to systematically assess the adverse outcomes associated with TTF when combined to temozolamide versus temozolamide alone for the treatment of glioblastoma. Methods MEDLINE (subset PubMed), EMBASE, Web of Science, Cochrane databases, Google scholar, and http://www.ClinicalTrials.gov were searched for publications comparing TTF in combination with temozolomide versus temozolomide alone for the treatment of glioblastoma. The adverse outcomes were considered as the endpoints in this analysis. The statistical analysis was carried out by the latest version of the RevMan software (version 5.4). Risk ratios (RR) with 95% confidence intervals (CI) were used to represent the data following assessment. A random effect model was used during the statistical analysis. Results A total number of 994 participants were included in this analysis whereby 572 participants were assigned to TTF + temozolomide, whereas 422 participants were assigned to temozolomide alone. All the studies were randomized trials. The majority of the participants were males (68.0–75.0%) with a mean age of the varying from 54.0 to 69.0 years. Results of this analysis showed that TTF in combination with temozolomide was not associated with significantly higher adverse outcomes related to blood and lymphatic system disorders (RR:0.87, 95% CI 0.30–2.48; P = 0.79), gastrointestinal disorders (RR: 0.83, 95% CI 0.14–4.72; P = 0.83), general deterioration and malaise (RR: 1.43, 95% CI 0.90–2.25; P = 0.13), infections and infestations (RR: 1.21, 95% CI 0.33–4.45; P = 0.78), injury/poison and procedural complications (RR: 1.78, 95% CI 0.88–3.61; P = 0.11), metabolic and nutritional disorders (RR: 0.61, 95% CI 0.34–1.08; P = 0.09), central nervous system disorders (RR: 1.14, 95% CI 0.89–1.46; P = 0.31), musculoskeletal and connective tissue disorders (RR: 0.99, 95% CI 0.48–2.07; P = 0.99), respiratory/thoracic/mediastinal disorders (RR: 0.85, 95% CI 0.44–1.62; P = 0.62), and psychiatric disorders (RR: 1.27, 95% CI 0.42–3.79; P = 0.67). Similar adverse effects were observed. Conclusion TTF in combination with temozolomide was not associated with significantly worse adverse outcomes for the treatment of patients with glioblastoma. Hence, future trials should focus on the cost-effectiveness of TTF for the treatment of glioblastoma.

MTDH-SND1 protein-protein interaction (PPI) plays an important role in the initiation and development of tumors, and it is a target for the treatment of breast cancer. In this study, we identified and synthesized a series of novel small-molecule inhibitors of MTDH-SND1 PPI. The representative compound C19 showed potent activity against MTDH-SND1 PPI with an IC50 of 487 ± 99 nM and tight binding to the SND1-purified protein with a Kd value of 279 ± 17 nM. Compound C19 significantly degraded SND1 and downregulated downstream at the protein level. Further biological evaluations suggested that compound C19 exhibited potent activity against the proliferation of breast cancer MCF-7 cells with an IC50 value of 626 ± 27 nM, significantly inhibited invasion and migration, and induced cell apoptosis. In addition, compound C19 exhibited promising tumor growth inhibition in the xenograft model. Our study provides a potential candidate targeting MTDH-SND1 PPI for the treatment of breast cancer.

The prediction of drug-target binding affinity (DTA) plays an important role in drug discovery. Computerized virtual screening techniques have been used for DTA prediction, greatly reducing the time and economic costs of drug discovery. However, these techniques have not succeeded in reversing the low success rate of new drug development. In recent years, the continuous development of deep learning (DL) technology has brought new opportunities for drug discovery through the DTA prediction. This shift has moved the prediction of DTA from traditional machine learning methods to DL. The DL frameworks used for DTA prediction include convolutional neural networks (CNN), graph convolutional neural networks (GCN), and recurrent neural networks (RNN), and reinforcement learning (RL), among others. This review article summarizes the available literature on DTA prediction using DL models, including DTA quantification metrics and datasets, and DL algorithms used for DTA prediction (including input representation of models, neural network frameworks, valuation indicators, and model interpretability). In addition, the opportunities, challenges, and prospects of the application of DL frameworks for DTA prediction in the field of drug discovery are discussed.

Ethnopharmacological relevanceThroughout Chinese history, Hydrangea paniculata Siebold has been utilized as a traditional medicinal herb to treat a variety of ailments associated to inflammation. In a number of immune-mediated kidney disorders, total coumarins extracted from Hydrangea paniculata (HP) have demonstrated a renal protective effect.Aim of the studyTo investigate renal beneficial effect of HP on experimental Adriamycin nephropathy (AN), and further clarify whether reversing lipid metabolism abnormalities by HP contributes to its renoprotective effect and find out the underlying critical pathways.Materials and methodsAfter establishment of rat AN model, HP was orally administrated for 6 weeks. Biochemical indicators related to kidney injury were determined. mRNAs sequencing using kidney tissues were performed to clarify the underlying mechanism. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis, western blot, molecular docking, and drug affinity responsive target stability (DARTS) assay was carried out to further explore and confirm pivotal molecular pathways and possible target by which HP and 7-hydroxylcoumarin (7-HC) played their renal protection effect via modulating lipid metabolism.ResultsHP could significantly improve renal function, and restore renal tubular abnormal lipid metabolism and interstitial fibrosis in AN. In vitro study demonstrated that HP and its main metabolite 7-HC could reduce ADR-induced intracellular lipid deposition and fibrosis characteristics in renal tubular cells. Mechanically, HP and 7-HC can activate AMP-activated protein kinase (AMPK) via direct interaction, which contributes to its lipid metabolism modulation effect. Moreover, HP and 7-HC can inhibit fibrosis by inhibiting CCAAT/enhancer binding protein beta (C/EBPβ) expression in renal tubular cells. Normalization of lipid metabolism by HP and 7-HC further provided protection of mitochondrial structure integrity and inhibited the nuclear factor kappa-B (NF-κB) pathway. Long-term toxicity using beagle dogs proved the safety of HP after one-month administration.ConclusionCoumarin derivates from HP alleviate adriamycin-induced lipotoxicity and fibrosis in kidney through activating AMPK and inhibiting C/EBPβ.

Doa10 (MARCHF6 in metazoans) is a large polytopic membrane-embedded E3 ubiquitin ligase in the endoplasmic reticulum (ER) that plays an important role in quality control of cytosolic and ER proteins. Although Doa10 is highly conserved across eukaryotes, it is not understood how Doa10 recognizes its substrates. Here, we define the substrate recognition mechanism of Doa10 by structural and functional analyses on Saccharomyces cerevisiae Doa10 and its model substrates. Cryo-EM analysis shows that Doa10 has unusual architecture with a large lipid-filled central cavity, and its conserved middle domain forms an additional water-filled lateral tunnel open to the cytosol. Our biochemical data and molecular dynamics simulations suggest that the entrance of the substrate’s degron peptide into the lateral tunnel is required for efficient polyubiquitination. The N- and C-terminal membrane domains of Doa10 seem to form fence-like features to restrict polyubiquitination to those proteins that can access the central cavity and lateral tunnel. Our study reveals how extended hydrophobic sequences at the termini of substrate proteins are recognized by Doa10 as a signal for quality control.

BACKGROUND Neoadjuvant chemotherapy can cause hepatic sinusoidal obstruction syndrome (SOS) in patients with colorectal cancer liver metastases and increases postoperative morbidity and mortality. AIM To evaluate T1 mapping based on gadoxetic acid-enhanced magnetic resonance imaging (MRI) for diagnosis of hepatic SOS induced by monocrotaline. METHODS Twenty-four mice were divided into control (n = 10) and experimental (n = 14) groups. The experimental groups were injected with monocrotaline 2 or 6 days before MRI. MRI parameters were: T1 relaxation time before enhancement; T1 relaxation time 20 minutes after enhancement (T1post); a reduction in T1 relaxation time (△T1%); and first enhancement slope percentage of the liver parenchyma (ESP). Albumin and bilirubin score was determined. Histological results served as a reference. Liver parenchyma samples from the control and experimental groups were analyzed by western blotting, and organic anion transporter polypeptide 1 (OATP1) was measured. RESULTS T1post, △T1%, and ESP of the liver parenchyma were significantly different between two groups (all P < 0.001) and significantly correlated with the total histological score of hepatic SOS (r = -0.70, 0.68 and 0.79; P < 0.001). △T1% and ESP were positively correlated with OATP1 levels (r = 0.82, 0.85; P < 0.001), whereas T1post had a negative correlation with OATP1 levels (r = -0.83; P < 0.001). CONCLUSION T1 mapping based on gadoxetic acid-enhanced MRI may be useful for diagnosis of hepatic SOS, and MRI parameters were associated with OATP1 levels.

To explore the value of T1 mapping-based whole-liver histogram analysis on gadoxetic acid-enhanced MRI for predicting post-hepatectomy liver failure (PHLF). Consecutive patients from March 2016 to March 2018 who underwent gadoxetic acid-enhanced MRI in our hospital were retrospectively analyzed, and 37 patients were enrolled. Whole-liver T1 mapping-based histogram analysis was performed. The indocyanine green (ICG) clearance tests were performed, and albumin–bilirubin (ALBI) scores were calculated. Univariate and multivariate binary logistic analyses were performed to identify independent predictors for PHLF. Diagnostic performance was evaluated with ROC analysis. Histogram-extracted parameters were also associated with the ICG test and ALBI scoring system. In enrolled 37 patients (age 57.19 ± 12.28 years), 28 were male. 35.1

Background Epstein-Barr virus (EBV) is a double-stranded DNA oncogenic virus. Several types of solid tumors, such as nasopharyngeal carcinoma, EBV-associated gastric carcinoma, and lymphoepithelioma-like carcinoma of the lung, have been linked to EBV infection. Currently, several TCR-T-cell therapies for EBV-associated tumors are in clinical trials, but due to the suppressive immune microenvironment of solid tumors, the clinical application of TCR-T-cell therapy for EBV-associated solid tumors is limited. Figuring out the mechanism by which EBV participates in the formation of the tumor immunosuppressive microenvironment will help T cells or TCR-T cells break through the limitation and exert stronger antitumor potential.Methods Flow cytometry was used for analyzing macrophage differentiation phenotypes induced by EBV-infected and EBV-uninfected tumors, as well as the function of T cells co-cultured with these macrophages. Xenograft model in mice was used to explore the effects of M2 macrophages, TCR-T cells, and matrix metalloprotein 9 (MMP9) inhibitors on the growth of EBV-infected tumors.Results EBV-positive tumors exhibited an exhaustion profile of T cells, despite the presence of a large T-cell infiltration. EBV-infected tumors recruited a large number of mononuclear macrophages with CCL5 and induced CD163+M2 macrophages polarization through the secretion of CSF1 and the promotion of autocrine IL10 production by mononuclear macrophages. Massive secretion of MMP9 by this group of CD163+M2 macrophages induced by EBV infection was an important factor contributing to T-cell exhaustion and TCR-T-cell therapy resistance in EBV-positive tumors, and the use of MMP9 inhibitors improved the function of T cells cocultured with M2 macrophages. Finally, the combination of an MMP9 inhibitor with TCR-T cells targeting EBV-positive tumors significantly inhibited the growth of xenografts in mice.Conclusions MMP9 inhibitors improve TCR-T cell function suppressed by EBV-induced M2 macrophages. TCR-T-cell therapy combined with MMP9 inhibitors was an effective therapeutic strategy for EBV-positive solid tumors.

Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are devastating neurodegenerative diseases with no effective cure. GGGGCC repeat expansion in C9orf72 is the most common genetic cause of both ALS and FTD. A key pathological feature of C9orf72 related ALS/FTD is the presence of abnormal dipeptide repeat proteins translated from GGGGCC repeat expansion, including poly Glycine-Arginine (GR). In this study, we observed that (GR)50 conferred significant mitochondria damage and cytotoxicity. Metformin, the most widely used clinical drug, successfully relieved (GR)50 induced mitochondrial damage and inhibited (GR)50 related cytotoxicity. Further research revealed metformin effectively restored mitochondrial function by upregulating AKT phosphorylation in (GR)50 expressed cells. Taken together, our results indicated restoring mitochondrial function with metformin may be a rational therapeutic strategy to reduce poly(GR) toxicity in C9orf72 ALS/FTD patients.

Background: Vasculogenic mimicry (VM) induced by Epstein-Barr virus (EBV) infection plays an important role in resistance to anti-vascular endothelial growth factor (VEGF) therapy in EBV-associated epithelial cancers; however, the interaction between VM and the immune microenvironment has not been systematically investigated. Methods: IHC and multiplex IHC analysis the relationships among tumour-associated macrophage (TAM), VM and EBV infection in EBV-associated epithelial cancer biopsies. In vitro and in vivo evidence using CRISPR-Cas9 system engineered EBV-infected epithelial cancer cells and mouse models support functional role and mechanism for M2c-like macrophages in the VM formation. The prediction of VM in the effectiveness of anti-angiogenic agent was analysed using clinical datasets. Results: EBV-associated epithelial cancer biopsies revealed that infiltration of the TAM surrounding the VM is closely associated with EBV infection. AKT/mTOR/HIF-1 alpha pathway in EBV-infected epithelial cancer cells control the secretion of CCL5 and CSF-1, enabling the recruitment of monocytes and their differentiation into M2c macrophages which promote VM formation by MMP9. Combination of anti-angiogenesis agents and HIF-1 alpha alpha inhibitor caused marked decreases in CD31-positive micro-vessels, VM, and M2c-like macrophages. VM scores can be used as biomarkers to predict the efficacy of anti-angiogenic agent therapy in EBV-associated epithelial cancers. Conclusions: Our findings define a secretory cross-talk between tumour cells and the immune microenvironment in EBV-associated epithelial cancer, revealing an unexpected role of EBV in epithelial cancer cells, controlling VM formation via M2c-like macrophages.

BACKGROUND:Circular RNAs are highly enriched in the synapses of the mammalian brain and play important roles in neurological function by acting as molecular sponges of microRNAs. circAnk3 is derived from the 11th intron of the ankyrin-3 gene, Ank3, a strong genetic risk factor for neuropsychiatric disorders; however, the function of circAnk3 remains elusive. In this study, we investigated the function of circAnk3 and its downstream regulatory network for target genes in the hippocampus of mice.METHODS:The DNA sequence from which circAnk3 is generated was modified using CRISPR/Cas9 (clustered regularly interspaced short palindromic repeats/Cas9) technology, and neurobehavioral tests (anxiety and depression-like behaviors, social behaviors) were performed in circAnk3+/- mice. A series of molecular and biochemical assays were used to investigate the function of circAnk3 as a microRNA sponge and its downstream regulatory network for target genes.RESULTS:circAnk3+/- mice exhibited both anxiety-like behaviors and social deficits. circAnk3 was predominantly located in the cytoplasm of neuronal cells and functioned as a miR-7080-3p sponge to regulate the expression of Iqgap1. Inhibition of miR-7080-3p or restoration of Iqgap1 in the hippocampus ameliorated the behavioral deficits of circAnk3+/- mice. Furthermore, circAnk3 deficiency decreased the expression of the NMDA receptor subunit GluN2a and impaired the structural plasticity of dendritic synapses in the hippocampus.CONCLUSIONS:Our results reveal an important role of the circAnk3/miR-7080-3p/IQGAP1 axis in maintaining the structural plasticity of hippocampal synapses. circAnk3 might offer new insights into the involvement of circular RNAs in neuropsychiatric disorders.