Polycyclic aromatic hydrocarbons (PAHs) are well-known for their mutagenic and carcinogenic effects. Benzo[b]fluoranthene (BbF) is one of 16 PAHs prioritized by the US Environmental Protection Agency for toxicological evaluation due to pervasive human exposure. As part of a multi-stakeholder consortium, the genotoxic effects of BbF were evaluated in MutaMouse males exposed to five doses of BbF or a vehicle control via repeated oral gavage for 28, 60, 90, 120, or 180 days, with dose ranges adjusted by duration of exposure. Mutagenesis was evaluated in lung tissue (n = 4) at 28, 90, and 180 days using Duplex Sequencing (DS), and chromosomal damage (n = 8) was evaluated using the micronucleus assay in peripheral blood at all time points. Dose- and time-dependent increases in total mutation frequency (MF) and C:G > A:T mutations were observed in lung tissue after 28, 90, and 180 days of exposure. By 28 days, BbF exposure produced lung cancer-associated mutational signatures linked to tobacco smoking. Mutations accumulated over time in lung, whereas chromosomal damage in peripheral blood erythrocytes reached a steady state by 28 days. Benchmark dose (BMD) confidence intervals (CIs) narrowed with extended exposure only for MF. Collectively, the data demonstrate that BbF is a potent mutagen capable of inducing cancer-relevant mutations in lung, supporting its potential role in human lung carcinogenesis. By distinguishing early mutagenic responses from cumulative mutation effects over time, these findings highlight the value of integrating mutagenicity assessment into extended-duration studies to better inform the potential health effects of chronic genotoxic exposures.

Brain cancers hijack biological systems involved in neural development and synaptic plasticity. Medulloblastoma (MB), the most common malignant brain tumor in children, is thought to arise from disruptions in neurodevelopmental programs. Glutamatergic transmission mediated by α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors (AMPARs) has been implicated in synaptic communication between adult brain tumors and surrounding neurons; however, the possible role of AMPARs in MB remains largely unexplored. Here, we analyzed the expression of genes encoding AMPAR subunits, GRIA1-4, in datasets of MB tumors and cell lines, revealing distinct expression patterns and associations with overall survival (OS) across molecular subgroups and histological variants. Expression levels differed among MB molecular subgroups. Analysis using single-cell RNA sequencing (scRNA-seq) was consistent with enrichment of GRIA1 in Group 3 and GRIA4 in sonic hedgehog (SHH) MB. Higher GRIA1, GRIA2, and GRIA4 transcription was associated with more favorable patient outcomes in specific MB subgroups. In contrast, high expression of GRIA3 in SHH, or of either GRIA3 or GRIA4 in Group 3 MB, was associated with worse prognosis. Particularly robust but opposing associations with patient survival were found for GRIA3 and GRIA4 in SHH MB. Analysis of GRIA mRNA levels in MB cell lines using both quantitative reverse transcription polymerase chain reaction (qRT-PCR) and data from The Human Protein Atlas, supported some of the gene expression patterns observed in tumors. Together, these findings suggest that GRIA genes and their corresponding AMPAR subunits may have subgroup-specific prognostic relevance in MB.

Colorectal cancer (CRC) frequently harbors activating mutations in the WNT and MAPK pathways. While KRAS mutations alone can drive tumor initiation in many tissues, they are insufficient in the intestine. Leveraging allele-specific properties of RAS, we developed a mouse model to investigate MAPK hyperactivation. Here we show that KRAS mutations drive a regenerative state while antagonizing the Lgr5+ intestinal stem cell state; however, this regenerative state cannot initiate tumorigenesis. Instead, tumor initiation requires a stem-like state dependent on mutational activation of the WNT pathway. We identify two aberrant states-a WNT-driven stem-like state for tumor initiation and MAPK-driven transit-amplifying-like state for tumor growth. These plastic states, essential for tumorigenesis, also impact drug response, potentially explaining lower response rates and shorter duration of response to KRAS-G12C inhibitors in CRC compared to non-small cell lung cancer. These findings highlight the need to target both pathways and their associated cell states for effective CRC treatment.

IntroductionCystoscopy-based diagnosis and surveillance of bladder cancer (BC) remain challenging. This study presents a urine-based assay that enriches DNA-methylation signals via PCR enrichment and applies a machine-learning model to enable cost-effective early detection.MethodsIn a prospective cohort at hospital (May 2022-November 2023), 155 individuals were enrolled, BC was diagnosed and confirmed by cystoscopy-guided biopsy and histopathology. Targeted next-generation sequencing of urine DNA quantified methylation at 44 CpG sites within IRF4, PENK and PXDN. Supervised classifiers trained on these features distinguished tumor from non-tumor urine samples. Systems analyses around IRF4/PENK/PXDN mapped signaling pathways, protein interaction modules and tumor-microenvironment contexts in BC.ResultsUsing 155 urine samples (68 BC, 87 non-BC) with methylation and transcript expression data in TCGA-BLCA, we found significantly increased methylation of IRF4, PENK, and PXDN in BC (P < 0.0001). Corresponding mRNA levels of IRF4 and PENK were significantly downregulated in tumor tissues, with PXDN showing a declining trend. Methylation of IRF4 and PXDN negatively correlated with their expression (P < 0.0001). A 41-CpG-site random forest classifier targeting IRF4, PENK, and PXND (BladderCando model) demonstrated excellent performance in distinguishing BC from non-BC individuals (AUC = 0.9783, F1 = 0.9773), outperforming urine cytology for low-grade BC detection. Co-expression and enrichment analyses identified DCN as a central hub gene, primarily linked to ECM functions. High expression of PXDN, IRF4, and DCN correlated with upregulated immune checkpoint genes and increased immune cell infiltration. Single-cell sequencing revealed PXDN in fibroblasts and endothelial cells, DCN in fibroblasts, and IRF4 in T cells, with expression patterns in urine mirroring tumor tissue profiles.ConclusionThis study establishes IRF4, PENK and PXDN methylation in urine as robust molecular signatures for BC detection, with machine-learning integration markedly enhancing diagnostic precision. Beyond early surveillance, these epigenetic alterations delineate tumor-microenvironment interactions that may inform future therapeutic strategies.

Germline variants in Ribosomal Protein S20 (RPS20) have been reported to predispose to colorectal cancer (CRC) based on occurrence in a total of 3455 CRC cases and co-segregation with disease in 3 families. RPS20 encodes a component of the ribosomal 40 S subunit, but the mechanism by which the variants might influence CRC risk remains unclear. The current study assessed the association of RPS20 variants with CRC predisposition in whole-genome sequencing (WGS) data from 9738 CRC cases and 161,403 cancer-free controls from the COloRectal tumour Gene Identification consortium (CORGI) study, 100,000 Genomes (100kGP) and UK Biobank (UKB). One hundred and sixty-eight CORGI cases and 23 100kGP cases were recruited based on a family history of CRC. After excluding common polymorphisms, we identified one putative loss-of-function RPS20 variant in cases (p.Ile89fs), and none in controls. This individual was not from a multiplex family, and hence co-segregation analysis of the variant and disease was not possible. One in-frame deletion and one missense variant predicted to be probably pathogenic were found in controls. We also reviewed the evidence on RPS20 and CRC risk from the literature. We concluded that the combined data do not show conclusively that RPS20 is a proven CRC predisposition gene. Its inclusion in diagnostic gene panels for CRC cannot be justified until and unless stronger supporting evidence becomes available in the future.

Colorectal cancer (CRC) is the third most diagnosed malignancy and the second leading cause of cancer-related mortality worldwide. A consistent and epidemiologically well-documented feature of CRC is its sexual dimorphism: age-standardized incidence rates are 33-45% higher in men than in women, and mortality rates differ by 43-50%. Beyond epidemiology, biological sex influences tumor location, molecular subtype, and clinical outcome. Women more frequently develop right-sided, microsatellite-unstable tumors driven by the CpG island methylator phenotype pathway, whereas men predominantly present with left-sided, chromosomally unstable tumors harboring APC, KRAS, and TP53 mutations. Sex steroid hormones play a central modulatory role: estrogens, primarily via estrogen receptor β (ERβ), exert tumor-suppressive effects on colonic epithelium, whereas androgens promote pro-inflammatory and pro-tumorigenic signaling through androgen receptor (AR)-dependent pathways. The gut microbiome displays sex-specific compositional profiles ('microgenderome') and contributes to sex-specific CRC susceptibility through bidirectional interactions with sex hormones, shaping distinct immunological and metabolic microenvironments. Finally, sex influences the pharmacokinetics of fluoropyrimidines, the toxicity of targeted agents, and the response to immune checkpoint inhibitors. This review summarizes current evidence on sex-related differences in CRC epidemiology, molecular pathology, hormonal regulation, gut microbiota composition, and treatment outcomes, highlighting the need to systematically incorporate sex as a biological variable in CRC research and clinical practice.

Cancer research has undergone a fundamental transformation in recent decades due to the integration of artificial intelligence (AI) models into the study of tumor biology. However, tumor evolution, driven by genetic and phenotypic alterations leading to heterogeneity, resistance and metastasis, remains a major challenge in oncology. To understand these processes is crucial for developing effective therapeutic strategies and improving patient outcomes. Conventional methods often fail to capture the complexity and dynamics of these processes. In contrast, AI tools have the ability to integrate and analyze large-scale multi-omics, imaging and clinical data, offering the capability to decode tumor complexity. AI-driven methods facilitate multi-modal data integration, enabling the recognition of patterns that connect molecular alterations with phenotypic outcomes. In functional genomics, AI tools predict the effects of genetic variants, identify regulatory elements and map dysregulated pathways, thus clarifying mechanisms underlying tumor development and resistance. In the imaging field, deep learning techniques improve tumor segmentation, characterization and longitudinal monitoring, providing more accurate insights into tumor progression and treatment response. Predictive modeling could allow the anticipation of tumor evolution and drug response, supporting adaptive therapeutic plans and real-time treatment adjustments. Moreover, AI supports biomarker discovery, patient stratification and decision support systems that can improve clinical trial design and accelerate the development of personalized therapies. However, these advances raise important ethical challenges, including data privacy, algorithmic bias and the preservation of patient autonomy. Addressing these concerns is essential to ensure the responsible deployment of AI in oncology.

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancer types with a dismal prognosis, where early metastatic dissemination and rich desmoplastic stroma limit the therapeutic efficacy. Nischarin (NISCH)/Imidazoline-1 receptor (IR-1) is a potential tumor suppressor that is involved in the regulation of cell migration, invasion, and cytoskeletal organization. Importantly, several clinically approved agonists have been shown to target this receptor. This study aimed to examine NISCH expression in PDAC and the effects of its agonists with the intent of drug repurposing. NISCH was expressed in PDAC tumor tissue, in both the epithelial and stromal compartments of tumors, and higher NISCH expression was associated with longer patient survival. Out of the three tested NISCH agonists, moxonidine, clonidine and rilmenidine, rilmenidine was the only one affecting cancer cell viability and at high doses induced cancer cell apoptosis. Transcriptome analysis of rilmenidine-treated PDAC cells revealed changes associated with cytoskeletal organization, translating into decreased adhesion and migration in vitro. In cancer-associated fibroblasts (CAFs), rilmenidine treatment decreased the expression of activation markers and limited cancer cell-CAF cytokine communication in the co-culture. Ultimately, in the in vivo tumor xenograft zebrafish model, rilmenidine reduced the metastatic spread of pancreatic cancer cells. Our results suggest that the NISCH agonist rilmenidine is a promising candidate for drug repurposing as an antimetastatic agent in PDAC, and that NISCH can be a potential target for the development of new PDAC therapeutics.

Intrahepatic cholangiocarcinoma (ICC) is an aggressive liver malignancy with a rising global incidence and limited therapeutic options. Vascular invasion (VI) is a hallmark of advanced disease, correlating with early recurrence and dismal prognosis, yet its tumor microenvironment (TME) drivers remain elusive. We analyzed single-cell RNA sequencing (scRNA-seq) data from 25 ICC samples to systematically characterize the cellular composition and molecular features related to VI. By integrating bulk RNA-seq data, spatial transcriptomics, and multiplex immunofluorescence, we identified a distinct subset of tumor-like cancer-associated fibroblasts (CAFs), termed tCAFs, enriched in VI-positive tumors. Functional enrichment analyses revealed that tCAFs were prominently associated with hypoxia and angiogenesis pathways, findings corroborated by the significant upregulation of tCAF markers (MME and NT5E) in ICC-derived CAFs under hypoxic conditions in vitro. Cell-cell communication analysis and spatial mapping uncovered that tCAFs might promote VI primarily through VEGF signaling interactions with endothelial cells. Integrative bioinformatics and RT-qPCR validation identified three key functional genes in tCAFs: SLC2A1, PTGS2, and PLOD2. In endothelial sprouting assays, pharmacological inhibition of SLC2A1 exerted a pronounced suppressive effect. Consistently, sprouting assays using ICC-derived CAFs with SLC2A1 knockdown confirmed that its downregulation significantly reduced endothelial sprouting capacity. Importantly, administration of the SLC2A1 inhibitor BAY-876 effectively suppressed tumor progression and intrahepatic metastasis in the orthotopic ICC mouse model. Our findings define a VI-associated cellular ecosystem and molecular landscape in ICC, unveiling a novel hypoxia-tCAFs-endothelial cells axis. Furthermore, we identify SLC2A1 as a clinically relevant therapeutic target, offering new insights into tumor VI.

Hepatocellular carcinoma (HCC) is a highly heterogeneous malignancy characterized by poor prognosis and limited therapeutic response. Cancer stem cells (CSCs) contribute to tumor progression, therapeutic resistance, and tumor recurrence. Among transcriptional regulators potentially involved in these processes, Specificity Protein 1 (SP1) has emerged as a candidate integrator of oncogenic and epigenetic signaling networks. However, its contribution to CSC-associated phenotypes and drug resistance in HCC remains incompletely defined. In this study, we combined transcriptomic analyses of TCGA datasets with functional experiments in HCC cell lines (Huh7 and HepG2). SP1-associated transcriptional programs were targeted pharmacologically using mithramycin A (MIT-A) and genetically using siRNA-mediated knockdown. The effects were assessed by RNA sequencing, RT-qPCR, Western blotting, flow cytometry, and functional assays evaluating proliferation, migration, CSC-associated properties, and response to sorafenib. MIT-A treatment markedly reduced the expression of stemness-associated transcription factors (NANOG, OCT4, SOX2) and CSC markers (CD133, CD24), impaired CSC-related functions including ALDH activity and the Side Population phenotype, and inhibited cell proliferation and migration. MIT-A also sensitized both parental and sorafenib-resistant HCC cells to sorafenib, associated with modulation of apoptotic regulators and reduced transporter-mediated efflux activity. SP1 knockdown partially reproduced several of these effects, supporting a contribution of SP1-dependent transcriptional programs to these phenotypes. Overall, these findings identify SP1-associated transcriptional networks as potential regulators of CSC features and therapeutic resistance in HCC and support targeting SP1-associated transcriptional programs as a strategy to enhance sorafenib efficacy.

Loss-of-function mutations in BCOR, a subunit of the non-canonical Polycomb Repressive Complex 1.1 (PRC1.1), are frequently observed in acute myeloid leukemia (AML) and associate with adverse risk. Paradoxically, leukemic stem cell viability in BCOR wild-type AMLs strongly depends on PRC1.1 activity. Here, we use BCOR and KDM2B degron models to study PRC1.1 dependency in leukemic cells and find that BCOR is a bridging factor tethering the catalytic and chromatin-binding moieties of the PRC1.1 complex. BCOR degradation induces a quick localized loss of H2AK119ub at PRC1.1 target genes, whereas PRC2-induced H3K27me3 remains unaffected. Degron-mediated depletion of BCOR or KDM2B induces a rapid but time-dependent transcriptional induction, whereby late-upregulated genes are more heavily decorated with H3K27me3 compared to early-upregulated genes. Combined PRC1.1 inactivation and PRC2 inhibition further amplifies gene induction, suggesting collaborative yet distinct control over target genes. Strikingly, both JARID2/AEBP2 and SUZ12 knockout cells, devoid of PRC2.2 or PRC2.1/PRC2.2 respectively, retain PRC1.1 loss-induced transcriptional activation, underscoring that PRC1.1 can repress target genes independently of a downstream PRC2.2-canonical PRC1 repressive axis. Finally, combined targeting of PRC1.1 and PRC2 induces differentiation of leukemic cells, emphasizing that co-targeting PRC1.1 and PRC2 represents a promising strategy to improve treatment of AML patients.

Neuroblastoma (NB) is the most common extracranial pediatric solid tumor, and the prognosis of high-risk patients remains poor, while early diagnosis is still challenging. Although PANoptosis plays an important role in tumor development and progression, its role in NB has not been systematically investigated.

Colorectal cancer (CRC) is one of the most common malignancies worldwide and remains a major cause of cancer-related mortality. Increasing evidence suggests that aberrant RNA metabolism contributes to tumour initiation, progression and therapeutic resistance. In this study, we integrated single-cell RNA sequencing and bulk transcriptomic data to identify RNA metabolism-related genes (RMRGs) associated with CRC progression and evaluate their clinical significance. Through comprehensive bioinformatics analyses, 39 differentially expressed RMRGs were identified, with high RMRG activity predominantly enriched in epithelial cell populations. Cell-cell communication analysis revealed enhanced interactions between epithelial cells and other cell types within the tumour microenvironment. Further integration of single-cell and bulk RNA-sequencing datasets identified PCBP3 and NGRN as key prognostic genes. A two-gene prognostic model based on PCBP3 and NGRN was established and validated in independent cohorts, demonstrating favourable predictive performance. Human Protein Atlas data further confirmed the expression of both proteins in colorectal cancer tissues. Immune infiltration analyses indicated that PCBP3 and NGRN were associated with distinct immune-cell infiltration patterns and immunotherapy-related immune status. Functional experiments demonstrated that silencing PCBP3 or NGRN significantly inhibited the proliferation and invasion of HCT116 colorectal cancer cells. Moreover, Western blot analysis suggested that these effects may be mediated, at least in part, through regulation of the PI3K/AKT signalling pathway. Collectively, our findings identify PCBP3 and NGRN as promising prognostic biomarkers and potential therapeutic targets in colorectal cancer and provide new insights into the role of RNA metabolism in colorectal cancer progression and tumour immune regulation.

Radiation resistance severely impairs the therapeutic efficacy of radiotherapy in colorectal cancer (CRC). Lipid metabolic reprogramming, particularly the activation of fatty acid oxidation (FAO), has been well recognized to be closely implicated in radioresistance, whereas its underlying regulatory mechanism remains poorly elucidated. This study aimed to explore the functional role of the transporter SLC16A9 in CRC radioresistance and its correlation with carnitine metabolism as well as the FAO signaling pathway.

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin's lymphoma. Although standard immunochemotherapy is available, 30%-40% of patients develop refractory or relapsed disease, underscoring the need to elucidate the underlying mechanisms of therapy resistance. High mobility group box 2 (HMGB2) is overexpressed in DLBCL; however, its role in the pathogenesis of DLBCL and in mediating therapy resistance remains unclear.

Birt-Hogg-Dubé (BHD) syndrome is a rare autosomal dominant disorder characterised by pulmonary cysts, recurrent spontaneous pneumothorax, cutaneous lesions and an increased risk of renal tumours. Pulmonary manifestations often precede other features, leading to delayed recognition. We report the case of a non-smoking male in his early 60s who presented with recurrent left-sided spontaneous pneumothorax and exertional dyspnoea. Cutaneous examination revealed multiple acrochordons over the neck. Chest imaging demonstrated pneumothorax, and high-resolution CT of the thorax showed a left upper lobe bulla with multiple bilateral thin-walled pulmonary cysts with basal predominance. The patient underwent bullectomy and surgical pleurodesis. Histopathology revealed thin-walled cysts lined by normal respiratory epithelium without smooth muscle proliferation, and HMB-45 immunostaining was negative, excluding lymphangioleiomyomatosis. Serum alpha-1 antitrypsin levels were normal. Genetic analysis identified a heterozygous pathogenic frameshift variant (c.1285delC) in the FLCN gene, confirming the diagnosis of BHD syndrome. The postoperative course was uneventful, and renal surveillance and family counselling were initiated. This case underscores the importance of considering BHD syndrome in non-smokers presenting with recurrent pneumothorax and cystic lung disease, as early diagnosis facilitates appropriate surveillance and genetic evaluation.

Immunotherapy has revolutionized cancer treatment, yet substantial inter-patient response heterogeneity limits therapeutic benefit to specific patient subsets. Here, we present PathTIGR, a pathway topology-informed graph representation learning framework that systematically integrates biological pathway network topology knowledge with genome variation information for immunotherapy response prediction. PathTIGR uses a three-component design: (i) pathway graph encoder with multihead attention embeding pathway topology knowledge and cancer genomic variants to pathway representation, (ii) transformer module capturing pathway regulatory dependencies, and (iii) multilayer perceptron synthesizing pathway-level representations to predict immunotherapy response. This architecture enables PathTIGR to capture complex molecular interactions underlying immunotherapy response. Comprehensive validation across multiple independent immunotherapy cohorts demonstrates that PathTIGR achieves superior predictive performance compared to established biomarkers and state-of-the-art deep learning approaches while maintaining biological interpretability through identification of key signatures underlying response heterogeneity. PathTIGR represents an interpretable graph-based learning framework that enhances immunotherapy response prediction and elucidates molecular determinants of therapeutic efficacy, thereby facilitating the advancement of precision cancer immunotherapy.

The mechanisms by which Wnt/β-catenin signaling regulates gene expression in a tissue- and context-specific manner remain poorly understood, limiting our ability to target the aberrant cell growth typical of many Wnt-driven cancers. Here, we focus on malignant liver tumors driven by activating CTNNB1 (β-catenin) mutations that nevertheless display distinct phenotypic states and Wnt outputs. By profiling patient-derived organoids via single-cell transcriptomics and chromatin dynamics, we identify subtype-specific transcriptional and epigenetic profiles. Using CUT&RUN, we show that β-catenin engages distinct genomic regions, dictated by differential association with TCF/LEF family transcription factors. Specifically, we define a sequence-specific regulatory element engaged by β-catenin only upon interaction with TCF7L2, revealing that partner choice, independent of CTNNB1 mutational status, ultimately determines cell fate. Our findings, validated across multiple tumor models and patient tissues, offer a framework for understanding how differential β-catenin-TCF/LEF interaction orchestrates context-specific Wnt signaling outcomes.

Lung adenocarcinoma (LUAD) is the most common and deadly subtype of lung cancer, with limited therapeutic options and poor prognosis. Ubiquitination is an important posttranslational modification (PTM) that controls protein turnover, localization and interaction. The dysregulated ubiquitination machinery is a hallmark of cancer, contributing to tumor development. Identifying key E3 ubiquitin ligases and deubiquitinases in LUAD could lead to new biomarkers and treatments. Through RNA sequencing, bioinformatics, and clinical studies, SPOP has been identified as a promising E3 ligase target in LUAD. SPOP is downregulated in LUAD samples and associated with poorer patient outcomes. Functional analyses demonstrate that SPOP suppresses LUAD cell migration, proliferation, and in vivo tumor growth. Mechanistically, SPOP targets poly(A)-binding protein cytoplasmic 1 (PABPC1) for nonproteolytic ubiquitination, leading to its nuclear retention and inhibition of global protein synthesis, thereby suppressing the oncogenic effects of PABPC1. Furthermore, the deubiquitinase HAUSP is able to counteract SPOP-mediated ubiquitination of PABPC1, enhancing its oncogenic role in LUAD cells. Collectively, our findings reveal a SPOP/HAUSP-PABPC1 regulatory axis in LUAD, where SPOP and HAUSP oppositely modulate PABPC1 ubiquitination to control LUAD cell proliferation and migration. This axis represents a promising therapeutic target in LUAD.

Joint effect of cancer-associated fibroblasts (CAFs) and regulatory T cells (Tregs) can drive immunosuppression in high-grade serous ovarian cancer (HGSOC), thereby promoting tumor metastasis and limiting the clinical benefit of PD-1/PD-L1 blockade. However, the mechanistic crosstalk between CAFs and Tregs remains unclear. By integrating bulk RNA-seq, single-cell RNA-seq datasets, and spatial transcriptomics datasets with functional assays, we identified a population of TGF-β1-driven Gremlin1 (GREM1)+ myofibroblastic CAFs (myCAFs) enriched in HGSOC metastatic lesions. These GREM1+ myCAFs activated the FGFR1-MAPKs-NFκB-TDO2-kynurenine axis via autocrine signaling, promoting tumor growth, metastasis, and immune evasion characterized by the infiltration of CD4+ Tregs and dysfunctional CD8+ T cells. GREM1 knockdown suppressed tumor metastasis, restored antitumor T cells response, remodeled CAF subtype composition and potentiated anti-PD-1 therapy efficacy. This study highlights the pro-metastatic role of GREM1+ myCAFs across multiple tumors and provides a rationale for combining anti-PD-1 therapy for metastatic HGSOC.