Liver cancer exhibits limited and heterogeneous responses to immune checkpoint blockade, underscoring the need to elucidate tumour microenvironmental mechanisms sustaining immune dysfunction. Here, we integrated single-cell transcriptomic annotation with BayesPrism deconvolution across multiple bulk cohorts and combined these analyses with in vivo liver cancer models, functional assays, and mechanistic investigations to define tumour-associated neutrophil (TAN) programmes linked to clinical outcome. We identified a neutrophil subset characterized by elevated sphingolipid metabolism as the predominant poor-prognosis signature, with SPTLC2 serving as the key molecular marker. High SPTLC2 expression defining this sphingolipid metabolism-high neutrophil population correlated significantly with advanced disease stage, inferior overall survival, and increased tumour infiltration. Tumour cell co-culture induced SPTLC2 upregulation and enhanced sphingolipid metabolic activity in neutrophils. Functional assays verified that myriocin, an SPTLC2 inhibitor, suppressed the proliferation and metastasis of Liver cancer cells and remodelled the antitumour immune function of neutrophils. Myriocin attenuated neutrophil sphingolipid metabolism and restored ROS production and neutrophil extracellular trap (NETosis) formation, revealing a critical regulatory role for sphingolipid metabolism in neutrophil-mediated innate immunity, as corroborated by single-cell analyses showing impaired innate immune function in sphingolipid metabolism-high neutrophils. Mechanistically, we uncovered an SPTLC2-dependent sphingolipid-lipid raft checkpoint that remodels plasma membrane microdomains to constrain neutrophil anti-tumour effector functions. Specifically, myriocin restored NOX2-dependent ROS generation by reducing ceramide accumulation in lipid rafts, thereby facilitating p47phox-p22phox interaction and productive NOX2 complex assembly in tumour-conditioned neutrophils. Collectively, these findings reveal a previously unrecognized immunometabolic axis governing TAN dysfunction in liver cancer, establish SPTLC2 as a rational combinatorial target for cancer immunotherapy, and highlight the therapeutic potential of targeting sphingolipid metabolism to restore neutrophil-mediated antitumour immunity and restrain hepatocellular carcinoma progression.

Over the past decades, metastasis research has largely emphasized genetic and epigenetic regulators. Emerging evidence indicates that metabolites-including sugars (glucose, fructose), antioxidant vitamins, glutathione, lipids (palmitic acid, cholesterol), and amino acids (glutamine, aspartate, BCAAs)-can critically influence metastasis in a context-dependent manner by driving or restraining progression via metabolic reprogramming and signaling modulation. This review examines how cancer cells re-modulate core metabolic pathways to adapt and metastasize, summarizes how endogenous metabolites, dietary bioactive compounds, metabolic toxins, and pharmacological agents regulate metastasis, and discusses new concepts and perspectives on metabolic regulation of tumor metastasis, and outlines future directions in metabolite-driven tumor metastasis. The metabolic dependencies of metastasis provide a rational basis for intervention through integrated nutritional and therapeutic strategies.

N[6]-methyladenosine (m[6]A), the most prevalent internal RNA modification in mammals, regulates key cellular processes through writer, eraser, and reader proteins. Recent studies have emerged that dysregulation of m[6]A modifiers promotes tumorigenesis by altering RNA metabolism. However, the role of m[6]A in multiple myeloma (MM) remains enigmatic. Here, we identify YTHDC1, a nuclear m[6]A reader, as a key oncogenic factor in MM. YTHDC1 is upregulated in MM cells and associated with poor prognosis. Functional studies reveal that YTHDC1 facilitates MM cell growth and survival in an m[6]A-dependent manner. Mechanistically, YTHDC1 directly binds m[6]A-modified EP300 mRNA and enhances its stability and expression. In turn, p300 activates YTHDC1 transcription via H3K27ac deposition and IRF3/IRF4 recruitment, establishing a self-reinforcing feedback loop. This interplay between RNA m[6]A methylation and histone H3K27 acetylation contributes to MM pathogenesis. Our findings highlight the YTHDC1-EP300 axis as a promising target for therapeutic intervention.

Bone metastasis affects 30-40% of lung cancer patients, markedly reducing survival and quality of life, yet the underlying molecular mechanisms remain incompletely understood. Comparative transcriptomic profiling of primary lung tumors and paired bone metastases revealed pronounced KIFC1 enrichment in metastatic lesions, which correlated with poor bone metastasis-free survival. Functionally, KIFC1 enhanced lung cancer cell migration and adhesion toward osteoblast-like cells and upregulated the bone-homing factors CXCR4 and OPN. KIFC1 promoted osteoclast recruitment and differentiation while suppressing osteogenic differentiation, thereby disrupting bone homeostasis. Mechanistically, KIFC1 associated with RUNX2 through its C-terminal motor domain and promoted RUNX2 nuclear accumulation in a motor activity-, microtubule-, and importin-dependent manner. Nuclear RUNX2 directly activated TGFB1 transcription, resulting in preferential activation of the TGF-β/SMAD3 signaling pathway. Genetic silencing of RUNX2 or pharmacological inhibition of SMAD3 largely abrogated KIFC1-mediated pro-metastatic and pro-osteolytic effects. In vivo, KIFC1 knockdown reduced intraosseous tumor burden and alleviated trabecular bone loss, whereas KIFC1 overexpression exacerbated osteolytic lesions, which were mitigated by disrupting RUNX2/TGF-β/SMAD3 signaling. Collectively, our findings identify KIFC1-mediated RUNX2 nuclear transport as a previously unrecognized upstream regulatory mechanism that activates the TGF-β/SMAD3 axis during lung cancer bone metastasis, which provides a conceptual framework for the development of targeted therapeutic strategies.

Immunotherapy has shown promising efficacy in esophageal squamous cell carcinoma (ESCC), yet its clinical benefits remain limited due to immune evasion. Lactate accumulation in the tumor microenvironment has been found to facilitate immune evasion through protein lactylation. However, the key lactylated substrates and mechanisms driving immune evasion remain undefined. Here, we show that HSPA1A lactylation at lysine 108 (K108) by KAT5 drives CD8[+] T cell dysfunction and immunotherapy resistance via the STAT1/PD-L1 axis in ESCC. Mechanistically, K108 lactylation enhances HSPA1A binding to STAT1, protecting STAT1 from TRIM25-mediated ubiquitination and proteasomal degradation. This stabilization promotes STAT1 phosphorylation and nuclear translocation, leading to transcriptional upregulation of PD-L1 and subsequent immune evasion. Collectively, these findings highlight the pivotal role of HSPA1A lactylation in facilitating immune evasion via the STAT1/PD-L1 axis, suggesting a potential target for improving immunotherapy efficacy in ESCC.

Characterizing the genealogy of tumor cells and their population dynamics is central to our understanding of the evolution of cancer. While direct monitoring of the tumor evolutionary process is impractical, we leverage advances in phylogenetics and phylodynamics to explore the evolutionary dynamics of non-small cell lung cancer (NSCLC) using single-cell whole exome sequencing data from 351 tumor cells across 11 NSCLC patients, predominantly focusing on the primary tumors and early stages of NSCLC. Our results reveal distinct phylogenetic tree shapes and evolutionary trajectories across different stages of NSCLC evolution. However, this intricate and multi-phase evolutionary process cannot be fully elucidated by current tumor models. Herein, we hypothesize that tumors progress through the following phases: (1) an initial rapid expansion through Big Bang-like growth, (2) a prolonged plateau period of non-progression with possible accumulation of neutral substitutions, (3) a subsequent phase where adaptive substitutions accumulate gradually, (4) a bottleneck event where selectively advantageous subclones acquire invasive capabilities (characterized by a pattern of punctuated evolution), and, potentially, a rapid expansion leading to metastasis. This work highlights the use of single-cell phylogenetics and phylodynamics for interpreting the evolution of NSCLC, with potential in predicting cancer progression, resistance mechanisms, and treatment response.

Long considered neutral, synonymous codon usage is increasingly recognized as a regulatory layer of mRNA translation. Proteins expressed during proliferation are translated from mRNAs with codon compositions distinct from those encoding proteins in differentiated states, suggesting the existence of specialized translational programs. In hormone-receptor-positive breast cancer, estrogen receptor-alpha (ERα/ESR1) acts both as a driver of proliferation and as a maintainer of luminal identity, making its translational regulation particularly relevant. Here, we show that the transition of breast cancer cells from quiescent, differentiated states towards proliferative, less differentiated phenotypes, is accompanied by a systematic shift from G/C3- to A/U3-biased codon usage. ERα coding sequence architecture mirrors these translational programs through both global codon composition and position-dependent codon patterns. These features influence translational efficiency and likely modulate co-translational folding dynamics, thereby adjusting receptor abundance and functional outputs according to the prevailing translational program. Consistent with their functional importance, synonymous codon choices are subject to increased selective constraints within regions encoding structural and functional domains. Using engineer synonymous ERα variants, we demonstrate that codon usage fine-tunes receptor function across distinct cellular states. Together, our findings identify synonymous codon usage as a previously unrecognized mechanism linking translational programs to ERα expression and function.

The circadian rhythm is an intrinsic, self-regulatory physiological system in organisms that is hierarchically governed by core clock genes. Dysregulation of this rhythm is closely associated with tumor progression. However, the specific molecular mechanisms underlying circadian rhythm disruption-mediated colorectal cancer (CRC) remain largely elusive. The histone methyltransferase SET domain bifurcated 1 (SETDB1) exerts oncogenic functions in multiple solid tumors, prompting this study to investigate the molecular mechanisms through which SETDB1 regulates CRC progression. Our results demonstrate that SETDB1 acts as an oncogene in CRC by interacting with the Cullin 4B-RING E3 ligase (CRL4B) complex to enhance cancer cell proliferation, invasion, and stemness maintenance. Specifically, SETDB1 and the CRL4B complex assemble into a transcriptional repressor complex that binds to the promoters of core clock genes, such as PER3 and CRY1, to repress their transcription, thereby disrupting cellular circadian rhythms and ultimately accelerating CRC cell proliferation and metastasis. Collectively, our findings elucidate the pivotal role of the SETDB1/CRL4B complex in CRC pathogenesis and provide a potential therapeutic strategy targeting SETDB1 for patients with CRC with high CUL4B expression.

Integrated single-cell transcriptomics and clinical analyses reveal elevated DDR1 expression in pancreatic ductal adenocarcinoma (PDAC) tissues, which correlates with poor prognosis. Mechanistically, collagen-activated Discoidin Domain Receptor 1 (DDR1) recruits SHC1 to activate the MAPK/ERK pathway, thereby driving transcriptional upregulation of SLC40A1 via the ERK-MYC axis. SLC40A1, encoding an iron exporter, reduces intracellular labile iron pools, thereby suppressing lipid peroxidation and ferroptosis. Consequently, DDR1 overexpression confers resistance to dihydroartemisinin (DHA)-induced ferroptosis in PDAC cells, characterized by diminished ROS accumulation, mitochondrial shrinkage, and cristae loss. Conversely, DDR1 knockdown or pharmacological inhibition (Dasatinib) sensitizes PDAC cells to DHA. Crucially, combining Dasatinib and DHA treatment synergistically inhibits tumor growth in vivo by reactivating ferroptosis, as evidenced by increased 4-HNE accumulation and decreased Ki67 expression. These findings identify DDR1 as a key regulator of iron metabolism and ferroptosis in PDAC, suggesting that dual targeting of DDR1 and ferroptosis represents a promising therapeutic strategy.

Glioblastoma (GBM) remains a major challenge in neuro-oncology, associated with a high rate of mortality despite decades of intensive research and therapeutic advancements, underscoring the urgent need for innovative preclinical platforms that can more accurately recapitulate the biological and pathological features of human disease. While conventional animal models have contributed to our understanding of GBM biology and the evaluation of treatment efficacy, they fail to capture the full complexity and heterogeneity of the tumour microenvironment (TME). Ex vivo models are associated with certain advantages in this context; however, they can not mirror the complex dynamic and multicellular interactions present in living organisms, particularly the critical treatment barriers unique to the central nervous system: the blood-brain barrier (BBB), blood-cerebrospinal fluid barrier (BCSFB) and blood-meningeal barrier (BMB). In response to these limitations, humanised mouse models have emerged as an advanced platform capable of faithfully mimicking the molecular, pathological and immunological features of human GBM. These models enable the replication of complex in vivo crosstalk between the immune system and the TME, while preserving the relevant treatment barriers that govern drug delivery to the brain. Accumulating evidence indicates that humanised mouse models closely reproduce the infiltration of human immune components into the TME, enabling the study of clinically relevant interactions that contribute to therapeutic resistance and treatment failure in GBM. This review aims to provide a comprehensive and systematic overview of the currently employed humanised mouse models in GBM research, highlighting their applications and comparative advantages. Finally, we evaluate the opportunities and challenges associated with each model and discuss future directions to increase the translational relevance and predictive power of preclinical GBM research. Humanised mouse models provide a valuable translational platform combining the human immune system and PDX orthotopic engraftment. Compared to conventional models and ex vivo models, these models can reproduce the complex cross-talk between tumour cells and the immune system, tumour heterogeneity, immunosuppressive TME, as well as complex in vivo interactions such as brain-specific barriers, including BBB, BCSFB and BMB. Future implementation of the human gut microbiome in these models has the potential to further increase translational relevance and precision in GBM research. Created in BioRender.com.

Dysregulation of three-dimensional (3D) genome architecture is a hallmark of cancer, yet the mechanisms by which its disruption drives tumorigenesis remain incompletely understood. Here, we demonstrated that SMAD4 regulated 3D genome organization through its canonical transcription factor activity by directly binding chromatin, thereby suppressing lung tumorigenesis. Hi-C analyses of human lung tumors identify SMAD4 as a potential regulator of 3D genome integrity. Using PTEN-deficient human bronchial epithelial cells and genetically engineered mouse models, we showed that SMAD4 loss induced widespread reorganization of chromatin compartments, topologically associating domain (TAD) boundaries, and chromatin loops, leading to oncogenic transcriptional rewiring during early tumorigenesis. Mechanistically, SMAD4 directly bound chromatin loops and spatially connected gene promoters with active or repressive regulatory elements to control transcription. Upon SMAD4 ablation, enhancer-promoter rewiring dysregulated critical oncogenic drivers, including ELF3. Functional manipulation of an SMAD4-regulated ELF3-associated chromatin loop altered ELF3 expression and cell survival both in vitro and in vivo. Collectively, our findings establish SMAD4 as a direct regulator of 3D genome architecture through classical transcription factor binding to chromatin loops and reveal enhancer-promoter rewiring as a key mechanism driving lung tumorigenesis and a potential therapeutic vulnerability in SMAD4-deficient lung cancer.

Abiraterone acetate is a standard-of-care therapy for prostate cancer (PCa). However, resistance frequently emerges, often characterized by the progression to AR-independent phenotypes. Employing a genome-wide CRISPR/Cas9 library screening strategy, we identified 523 long non-coding RNAs (lncRNAs) and 2,183 protein-coding genes as potential candidates associated with abiraterone resistance. Notably, a pair of sense-antisense genes, BIRC6-AS1/BIRC6, was identified as a significant contributor to abiraterone resistance, serving as a critical survival factor in AR-independent contexts. BIRC6-AS1 depletion led to a reduction in both the mRNA and protein levels of BIRC6. Moreover, depletion of either BIRC6-AS1 or BIRC6 enhanced the sensitivity of PCa cells to abiraterone in both in vitro and in vivo settings. Further investigation revealed that BIRC6-AS1 stabilized the mRNA of BIRC6 through interaction with ILF2. Suppression of either BIRC6-AS1 or BIRC6 attenuated non-homologous end joining (NHEJ) repair activity, resulting in the disassembly of 53BP1 foci at DNA damage sites and an increased accumulation of DNA damage, thereby exposing a vulnerability in AR-independent resistant cells. Mechanistically, BIRC6 interacted with A20 and facilitated the K48-linked ubiquitination and subsequent degradation of A20 at the K337 residue. Additionally, A20 knockdown effectively reversed the abiraterone sensitivity induced by BIRC6-AS1 depletion. Collectively, our study provides a landscape of lncRNAs and protein-coding genes associated with abiraterone resistance and suggests that targeting the BIRC6-AS1/BIRC6 axis represents a potential therapeutic strategy to eradicate AR-independent resistant tumors in prostate cancer.

Neuroendocrine bladder carcinoma (NEBC) is a highly aggressive malignancy with unresolved lineage determinants and limited preclinical models, hindering mechanistic investigation and therapeutic development. Here, we sought to assess whether bladder epithelial-derived models are competent to acquire neuroendocrine lineage programs under defined tumor suppressor alterations and to identify candidate therapeutic vulnerabilities in these systems. We integrated genomic and transcriptomic analyses of human NEBC with genetically engineered mouse models, epithelial-derived bladder organoids, and patient-derived NEBC models. Human NEBC exhibited dominant RB1 and TP53 alterations and an epithelial transcriptional continuum consistent with lineage plasticity. In vivo, intravesical Adeno-Cre-mediated tumor suppressor deletion predominantly generated sarcoma-like tumors, whereas epithelial-restricted organoid models recapitulated the molecular and neuroendocrine features of human NEBC, supporting epithelial lineage competence for neuroendocrine differentiation. Patient-derived models and human NEBC specimens further supported epithelial identity in NEBC. Using these complementary platforms, drug screening identified homoharringtonine (HHT) as a candidate therapeutic vulnerability in the tested NEBC systems. HHT suppressed neuroendocrine marker expression, induced apoptosis, and attenuated IL6-JAK-STAT3 signaling. Together, these findings describe complementary epithelial-derived NEBC models and support further investigation of HHT as a candidate therapeutic vulnerability.

CDC73 is a well-characterized tumor suppressor regulated by stress stimuli, governing progression of diverse human malignancies. Although previous studies have shown that E3 ubiquitin ligase UBR5 drives CDC73 ubiquitination and degradation to modulate tumorigenesis, the mechanisms by which stress-responsive pathways regulate UBR5-mediated CDC73 inactivation and transcriptional reprogramming remain elusive. Here, via integrated analyses of public datasets, multi-omics profiling (assay for transposase-accessible chromatin with sequencing [ATAC-seq], cleavage under targets and tagmentation [CUT&Tag], mRNA sequencing [mRNA-seq]), in vitro/in vivo assays, and molecular approaches including co-immunoprecipitation (Co-IP) and molecular docking, we demonstrate that UBR5 depletion profoundly alters chromatin accessibility and genome-wide transcriptional profiles in a CDC73-dependent manner. UBR5 ablation markedly suppresses osteosarcoma malignant phenotypes in cultured cells and xenograft models, with these effects fully rescued by concurrent CDC73 silencing. Mechanistically, we identify the JNK cascade as the critical upstream regulator: JNK activation sustains CDC73 stability by antagonizing UBR5-mediated CDC73 polyubiquitination, and map Lys257 as the key residue for UBR5-dependent CDC73 ubiquitination and degradation. Collectively, our findings define a novel JNK-dependent UBR5-CDC73 axis that acts as a molecular brake of the CDC73 positive feedback loop to orchestrate transcriptional programs, providing new mechanistic insights into CDC73 post-translational regulation in tumorigenesis and promising therapeutic targets for CDC73-dysregulated diseases.