Cancer cachexia is a multifactorial syndrome characterized by progressive skeletal muscle wasting that diminishes quality of life, limits responses to anticancer therapies, and increases mortality. Although excessive protein degradation has traditionally been viewed as a primary driver of muscle loss, cachectic muscle undergoes broader disruption of proteostasis involving protein turnover, organelle quality control, stress adaptation, metabolism, and regenerative capacity. Accumulating evidence suggests that diverse tumor- and host-derived signals converge on interconnected transcriptional networks that coordinate these processes. Rather than functioning independently, transcription factors integrate inflammatory, metabolic, hormonal, and cellular-stress signals to generate catabolic, stress-responsive, anabolic, and regenerative programs whose effects vary with disease stage and cellular context. Importantly, these transcriptional responses are not uniformly pathological but can represent causal, adaptive, or compensatory programs. Epigenetic, post-transcriptional, and post-translational mechanisms provide additional layers that shape the activation and magnitude of these transcriptional programs. Emerging single-cell and multi-omic approaches have further revealed myofiber- and cell-specific regulatory networks underlying the heterogeneous response of skeletal muscle to tumor growth. In this review, we argue that transcriptional reprogramming represents a central integrative mechanism through which diverse catabolic signals remodel skeletal muscle proteostasis, and we examine how network-level regulation contributes to muscle adaptation and wasting in cancer cachexia. Defining which regulatory programs are causal, compensatory, or maladaptive may reveal more precise biomarkers and therapeutic targets for preserving skeletal muscle mass and function in patients with cachectic cancer.

The coexistence of chronic obstructive pulmonary disease (COPD) and lung cancer (LC) presents a complex clinical scenario with distinct pathological interactions. Beyond shared epidemiological and genetic risk factors, COPD reshapes the pulmonary immune microenvironment through reprogrammed immune-cell compartments, aberrant programmed cell death, and epithelial-mesenchymal transition (EMT), fostering an immunologically "hot" but pathologically permissive milieu that simultaneously drives carcinogenesis and influences treatment responsiveness. Recent evidence suggests that COPD may confer an altered immunological microenvironment that paradoxically enhances responsiveness to immune checkpoint inhibitors (ICIs) in patients with LC. This review mainly synthesizes clinical evidence on the efficacy of ICI, currently the most widely used immunotherapy, in LC patients concomitant with COPD and elucidates potential immunological and molecular mechanisms underlying observed clinical benefits, including heightened programmed cell death protein ligand 1 (PD-L1) expression, elevated tumor mutational burden, increased human leukocyte antigen I (HLA-I) expression, enriched tissue-resident memory T cells and mucosal-associated invariant T cells, and a macrophage polarization shift toward a pro-inflammatory phenotype, with benefits appearing largely attributable to patients with mild-to-moderate airflow obstruction. The safety profile of ICIs in this population-exacerbation risk, eosinophilic airway disease, checkpoint inhibitor pneumonitis (CIP), and autoimmune thyroiditis-is appraised, alongside emerging strategies such as lymphocyte-activation gene 3 (LAG-3), T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), and T cell immunoreceptor with Ig and ITIM domains (TIGIT) blockade and the immunomodulatory potential of COPD medications (inhaled corticosteroids and long-acting β-agonists). Current evidence remains limited by small retrospective cohorts and heterogeneous stratification. Large-scale prospective studies are warranted. This review aims to clarify the prognostic value of concomitant COPD and guide precision therapeutic strategies for this growing comorbid population.

KRAS is the most frequently mutated oncogene in pancreatic cancer, with over 90% of patients harboring KRAS mutations, approximately 45% of which are G12D substitutions. The development of KRAS[G12D] selective inhibitors, such as HRS-4642, MRTX1133 and RMC-9805, has brought new hope for treating this patient population. However, the clinical efficacy of these inhibitors is limited by the emergence of acquired resistance, the mechanisms of which remain incompletely understood. Our study employed a KRAS saturation mutagenesis library to systematically map the landscape of KRAS mutations conferring resistance to HRS-4642 and MRTX1133. We identified several secondary mutations-including R73P, E63P, D69W, and M72K-that confer resistance to both inhibitors, while M72H was found specifically resistant to HRS-4642. We further characterized the impact of these mutations on KRAS downstream signaling and performed structural modeling of the mutant proteins. Through molecular dynamics simulations, we suggested that the R73P mutation, although distant from the binding pocket, is predicted to exert a global dynamic effect on the allosteric site, thereby disrupting inhibitor-protein interactions. This suggests the existence of a novel class of KRAS resistance mutations and a previously unrecognized mechanism of acquired resistance to KRAS-targeted therapies. Furthermore, we identified Y64W, A59G, and R73P as mutations that concurrently confer resistance to the tri-complex inhibitor RMC-9805, as well as to the allosteric inhibitors MRTX1133 and HRS-4642. This comprehensive analysis delineates the global landscape of secondary resistance mutations to KRAS[G12D]-selective inhibitors.

In recent years, significant progress has been made in developing targeted therapy for non-small cell lung cancer (NSCLC). However, these therapies are only applicable to specific NSCLC groups with corresponding gene mutations. Therefore, developing novel targeted therapies to benefit more patients remains an important task. Our prior work demonstrated that LKB1-mutant NSCLC cells are highly susceptible to cell death induced by glucose starvation. In this study, we investigated a novel combinational therapeutic strategy utilizing a sodium-glucose co-transporter-2 (SGLT2) inhibitor canagliflozin and a covalent KRAS[G12C] inhibitor adagrasib, two clinically approved therapeutics, to selectively target LKB1-KRAS[G12C] co-mutated NSCLC. First, we verified the co-occurrence of LKB1 and KRAS mutations in clinical NSCLC cohorts. Second, compared to other SGLT2 inhibitors, canagliflozin induced higher cytotoxicity in LKB1-mutant cells, and SGLT2 knockdown partially attenuated this effect. Mechanistically, the high susceptibility to canagliflozin was associated with impaired activation of AMP-activated protein kinase (AMPK). Furthermore, canagliflozin suppressed mammalian target of rapamycin (mTOR) activity and robustly inhibited cell growth in LKB1-mutant models in vitro. Third, adagrasib exhibited selective cytotoxicity in KRAS[G12C]-mutated NSCLC. Finally, the combination of canagliflozin and adagrasib synergistically reduced cell viability in LKB1-KRAS[G12C] co-mutated H23 and H2122 cells in vitro. Importantly, canagliflozin attenuated the adaptive MAPK signaling rebound induced by adagrasib in vitro. The therapeutic potential of this combination was confirmed in an H2122 nude mouse xenograft model in vivo. Collectively, our findings demonstrate that a combination of canagliflozin and adagrasib suppresses LKB1-KRAS[G12C] co-mutated NSCLC, laying a solid preclinical foundation for developing a new therapeutic strategy for this group of NSCLC.

To elucidate the mechanisms underlying therapy-induced resistance, we investigated the temporal dynamics of the polyploid giant cancer cell (PGCC) life cycle following treatment with the microtubule-destabilizing agent vincristine (VCR). Longitudinal live-cell fluorescence imaging revealed that VCR-induced mitotic suppression activated a distinct PGCC life cycle characterized by nuclear amplification and progressive cellular reprogramming. At the cellular level, PGCCs entered a senescence-like state accompanied by an enhanced senescence-associated secretory phenotype (SASP), increased epithelial-mesenchymal transition (EMT), acquisition of early embryonic-like features, and multilineage differentiation potential into ectoderm, mesoderm, and endoderm. PGCC-derived progeny exhibited dose- and time-dependent changes in proliferative behavior, aggressive phenotypes in vitro, along with time-dependent tumorigenic potential in vivo. Remarkably, at the tissue level, PGCCs also underwent spatial nuclear reorganization and self-organization into morphologically defined glandular epithelial structures exhibiting varying degrees of tissue maturation, reminiscent of early embryonic development. Silencing IL1β, IL6, or IL8, or pharmacologically inhibiting their receptors, suppressed PGCC formation, budding, EMT, and embryonic-like reprogramming associated with tissue morphogenesis. Furthermore, NAMPT emerged as a key regulator of embryonic-like stemness, glandular morphogenesis, and structural expansion. Together, these findings reveal that therapeutic stress activates a PGCC life cycle capable of generating two divergent outcomes: resistant tumor regeneration through clonal selection of undifferentiated tumor cells and organized tissue differentiation through spatial nuclear reprogramming. Our study establishes a conceptual framework linking therapy-induced polyploidization, developmental plasticity, and resistance generation at both cellular and tissue levels.

Neoadjuvant immunotherapy has improved clinical outcomes in resectable non-small cell lung cancer (NSCLC); however, a substantial proportion of patients fail to achieve durable responses, and the mechanisms underlying therapeutic resistance remain incompletely understood. Here, we identified protein phosphatase 1 regulatory inhibitor subunit 1B (PPP1R1B) as a metabolic driver of immunotherapy resistance in NSCLC through RNA sequencing of tumors with differential responses to neoadjuvant anti-PD-1 therapy. PPP1R1B was significantly upregulated in non-responders and correlated with poor prognosis in NSCLC patients. Mechanistically, PPP1R1B directly interacted with IRF3 through the Arg29 residue and promoted dephosphorylation of IRF3 at Ser396, thereby suppressing IRF3 activation and downstream interferon-β (IFN-β) signaling. PPP1R1B-mediated inhibition of IFN-β signaling reduced CD8[+] T-cell infiltration and impaired antitumor immunity, ultimately conferring resistance to anti-PD-1 therapy in vivo. Blockade of IFN-β signaling abrogated the antitumor effects induced by PPP1R1B depletion, supporting an immune-dependent mechanism. Notably, tumors resistant to neoadjuvant immunotherapy exhibited impaired oxidative phosphorylation and excessive lactate accumulation. Lactate induced PPP1R1B expression through NF-κB-dependent transcriptional activation, linking metabolic reprogramming to innate immune suppression. Collectively, our findings identify a lactate/NF-κB/PPP1R1B/IRF3 signaling axis that drives immune evasion and resistance to neoadjuvant immunotherapy in NSCLC, and nominate PPP1R1B as a potential biomarker and therapeutic target for overcoming anti-PD-1 resistance.

Mucin-16 (MUC16) is overexpressed in >60% of pancreatic ductal adenocarcinoma (PDAC) cases and correlates with poor outcomes, yet its role in regulating tumor-immune interactions remains incompletely defined. Using KPC and KPC-MUC16 knockout (KPC-MUC16[KO]) genetically engineered and orthotopic mouse models, we show that loss of MUC16 markedly reduces tumor burden and reprograms the tumor microenvironment (TME) toward a less immunosuppressive state. MUC16 deficiency significantly decreased CD206[+] and Ym1/2[+] M2-like macrophages, increased the M1/M2 ratio, and improved survival. Cytokine profiling identified CCL2 and Serpin-E1 as MUC16- dependent secreted factors that promote macrophage recruitment and polarization; neutralization of these cytokines reduced macrophage migration in vitro. Bone marrow-derived macrophages (BMDM) exposed to MUC16-expressing tumor cells or conditioned media adopted a CD206[+]Arg1[+] M2-like phenotype, whereas MUC16[KO] tumor signals favored inflammatory CD86[+]iNOS [+] M1-like polarization. Macrophage depletion using clodronate liposomes significantly impaired tumor growth in MUC16-competent tumors but had minimal effect in KPC-MUC16[KO] tumors, indicating dependence on macrophage-mediated immunosuppression in the former group. Spatial transcriptomics revealed a globally distinct TME in MUC16[KO] tumors, including enhanced antigen-presentation programs, increased complement activation, and fibroblast states skewed toward pro-inflammatory signaling. These findings establish MUC16 as a key orchestrator of macrophage-driven immunosuppression in PDAC and support therapeutic strategies combining MUC16-targeted interventions with TME-reprogramming approaches.

Pancreatic ductal adenocarcinoma (PDAC) responds poorly to immune checkpoint blockade (ICB), largely because of its immunosuppressive and T-cell-excluded tumor microenvironment. While inducing immunogenic cell death (ICD) offers a promising strategy to enhance tumor immunogenicity, redox-targeted ICD inducers with defined molecular targets remain limited in PDAC. Here, a functional screen of natural compounds identified erianin as a potent ICD inducer in PDAC. Erianin triggered canonical ICD hallmarks, including cell-surface calreticulin exposure, extracellular release of HMGB1 and ATP, thereby promoting dendritic-cell maturation, antigen cross-presentation and CD8[+] T-cell activation. Target-engagement assays together with genetic perturbation supported glutathione reductase (GSR) as a functionally important erianin target. Erianin decreased the GSH/GSSG ratio, accompanied by increased oxidative stress and preferential activation of the PERK-eIF2α-ATF4-CHOP branch of the endoplasmic reticulum (ER) stress response. Mutational and rescue experiments identified ARG-233 as an important residue required for GSR-dependent protection against erianin-induced redox stress and ICD. In syngeneic PDAC models, erianin-treated tumor cells elicited protective antitumor immunity in a tumor-challenge setting, while a therapeutic regimen combining erianin with anti-PD-1 antibody suppressed tumor growth and was associated with increased dendritic-cell activation and enhanced cytotoxic CD8[+] T-cell infiltration. These findings define a GSR-associated redox-immune axis that links glutathione metabolism to immunogenic ER stress and provide a rationale for combining erianin with PD-1 blockade in immunologically cold PDAC.

In proliferative glioblastoma, a subset of tumor cells survives early temozolomide (TMZ) exposure and drives eventual-resistance, yet the underlying mechanism remains unclear. We found that MYC target activation was associated with attenuated TMZ survival benefit in CGGA cohorts. In patient-derived MYC-high glioma cells, TMZ increased apoptosis but left a larger residual viable population due to enhanced proliferation; we defined these surviving cells as pioneer resistant (P-R) cells, which exhibited reduced TMZ sensitivity and a competitive survival advantage. Mechanistically, P-R cells adopted a senescence-like state via two coordinated arms of chromatin remodeling. On one hand, H3K27ac-dependent super-enhancers were gained at senescence-associated secretory genes; the CBP/p300 inhibitor A485 reverted the senescence-like state and restored TMZ cytotoxicity; On another hand, chromatin closed, reducing MYC occupancy at cell-cycle genes despite unchanged MYC protein, and suppressed nascent transcription and m[6]A modification, accompanied by cell-cycle arrest in some P-R cells. Furthermore, time-course analysis revealed that chromatin closure at cell-cycle genes is likely a temporally ordered epigenetic remodeling process, initiated by DNA damage-associated G2/M arrest mediated by WEE1 activation, followed by γH2AX-mediated recruitment of SUV39H1, which deposits H3K9me3 at MYC-bound loci to close chromatin. Functionally, early, but not delayed, inhibition of WEE1 with the novel brain-penetrant inhibitor TSL2304 blocked senescence-like epigenetic reprogramming, synergized with TMZ in vitro, and significantly prolonged survival in an orthotopic glioblastoma model. Together, these findings reveal a temporally ordered epigenetic remodeling in MYC-high glioblastoma that promotes survival after early TMZ exposure, and identify early WEE1 inhibition as a strategy to synergize with TMZ.

Radiotherapy (RT) combined with PD-1/PD-L1 blockade can induce systemic antitumor responses, but robust control of distant lesions remains uncommon. We identified a quadruple regimen combining immunogenic hypofractionated RT (hRT; 8 Gy × 3) to the primary tumor, low-dose RT (LDRT; 3 Gy × 3) to the secondary tumor, anti-PD-1 (αPD-1), and fenofibrate (FFB), a clinically approved PPARα agonist, that enhanced antitumor efficacy in two bilateral tumor models. Only this regimen substantially improved secondary-tumor control, eradicating LDRT-treated secondary tumors in all mice in the immunologically cold C51 model and inducing prolonged control in the aggressive, αPD-1-poorly responsive B16-CD133 model. This effect required CD8[+] T cells. Tumor-specific CD8[+] TILs showed enhanced effector function, reduced exhaustion, increased mitochondrial fitness, and a shift toward mitochondrial fatty acid oxidation (FAO), with scRNA-seq supporting PPARα/FAO-associated remodeling, including upregulation of Cpt1a, encoding the FAO enzyme CPT1A, in effector-associated, proliferative, and clonally expanded subsets. Both antitumor efficacy and T-cell metabolic remodeling were reversed by the CPT1 inhibitor etomoxir. By contrast, LDRT impaired tumor-cell mitochondrial fitness, while FFB shifted tumor cells from FAO toward glycolytic reliance and induced only slight apoptosis. Consistently, in vivo data argued against direct LDRT-mediated tumor-cell debulking. The quadruple regimen increased vascular MECA-79 expression, while LDRT enhanced endothelial ICAM-1/VCAM-1 and tumor-cell CXCL9/CXCL10 expression in secondary tumors, consistent with local conditioning favoring T-cell recruitment. These findings highlight FFB as a clinically tractable metabolic modulator and link FFB-enhanced hRT/LDRT/αPD-1-induced systemic antitumor efficacy to metabolic reprogramming of tumor-specific CD8[+] TILs and conditioning of the LDRT-treated tumor microenvironment.

It remains largely unclear whether oncogenic super-enhancers (SEs) can drive expression of circular RNAs (circRNAs). In this study, we for the first time profiled the comprehensive circRNA landscapes of esophageal squamous cell carcinoma (ESCC) after suppressing genome-wide SE activities. We identified a novel SE-driven circPRRC2B, as the product of back-splicing of exons 2, 3, 4 and 5 of PRRC2B pre-mRNA, which was controlled by RNA binding protein HuR. Indeed, SE drives high levels of PRRC2B transcription, promotes production of PRRC2B pre-mRNAs, boosts back-splicing of PRRC2B pre-mRNAs, and elevates circPRRC2B expression. ETS1 acts as a key transcriptional factor in SE-driven PRRC2B transcription and circPRRC2B biogenesis. Interestingly, circPRRC2B, but not PRRC2B, could promote proliferation, invasion and metastasis of ESCC in vitro and in vivo. Mechanistically, circPRRC2B binds to TALDO1 and enhances BAP1-mediated TALDO1 de-ubiquitination in ESCC cells. Therefore, circPRRC2B stabilizes and sustains high expression of TALDO1, activates the pentose phosphate pathway (PPP), elevates the glutathione levels via supplying more NADPH via the PPP, inhibits production of reactive oxygen species in mitochondrion, maintains cellular redox equilibrium, and, thus, leads to cancer progression. Importantly, we developed antisense oligonucleotides targeting circPRRC2B which exhibited remarkable inhibition of ESCC growth. Our findings highlighted circPRRC2B as a key target preserving the PPP metabolic flux and redox homeostasis of malignant cells, thereby offering insight into developing novel therapeutics for cancers.

Tissue normalisation has emerged as an interesting therapeutic concept in oncology, aiming to correct the structural, metabolic, and immunological tissue abnormalities that enable tumour progression and treatment resistance. Rather than targeting cancer cells directly, this approach aims to re-establish homeostatic tissue architecture, immune surveillance, and metabolic balance. Interventions that target stromal components, such as cancer-associated fibroblasts (CAFs), immune cells, or stiffness of extracellular matrix (ECM), have been shown to shift tumour cells toward a less aggressive state, underscoring the plasticity of cancer cells and the importance of their interactions with the microenvironment. Extracellular vesicles (EVs) are increasingly recognised as both contributors to tissue dysfunction in tumours and promising tools for restoring tissue homeostasis. EVs circulate across most biofluids and can cross biological barriers to reach distant tissues. They naturally carry nucleic acids, including regulatory miRNAs, along with proteins, lipids, and metabolites that influence cell behaviour, making them attractive tools for tissue normalisation. Moreover, engineered EVs could be designed to deliver targeted "normalisation signals" and hold strong clinical potential as next-generation delivery systems. Here, we explore how EVs, whether native or engineered, can intervene in key tumour-driving processes, including senescence, fibrosis, inflammation, and immune evasion, to shift the microenvironment toward a normalised, less malignant state.

AMP-activated protein kinase (AMPK) is a highly conserved serine/threonine kinase regarded as an indispensable energy sensor during metabolic stress, coordinating diverse biological processes, including glycolysis, the tricarboxylic acid (TCA) cycle, mitochondrial dynamics, and multiple stress-responsive signaling pathways. Growing evidence shows that dysregulated AMPK signaling plays closely linked roles in tumorigenesis. Depending on the context, AMPK signaling can suppress or promote tumor growth. Mechanistically, dysregulated AMPK kinase activity contributes to metabolic reprogramming, mitochondrial dysfunction, and epigenetic alterations, all recognized hallmarks of cancer. Consequently, targeting AMPK signaling has emerged as a promising strategy for cancer therapy. In this review, we summarize the molecular mechanisms regulating AMPK signaling, discuss its dual functions in tumor biology, highlight recent advances in AMPK-targeted therapeutic strategies, and address current challenges and future perspectives for the clinical translation of AMPK-targeted cancer therapy.

Cancer-associated adipocytes (CAAs) are increasingly recognized as active and heterogeneous components of the tumor microenvironment. Initially defined as tumor-adjacent adipocytes characterized by lipid depletion, reduced expression of mature adipocyte markers, and inflammatory activation, CAAs are now understood to encompass a spectrum of context-dependent states. These states are shaped by anatomical location, preexisting adipocyte properties, tumor-derived signals, and therapeutic exposure. Through metabolic exchange, paracrine signaling, direct cell contact, extracellular-vesicle communication, and matrix remodeling, CAAs influence tumor-cell plasticity, metastatic colonization, immune dysfunction, stromal organization, and treatment response. In turn, signals from tumor, immune, and stromal cells continuously reshape adipocyte identity and function, placing CAAs within reciprocal multicellular networks. This review examines the evolving criteria used to identify CAAs, their lineage origins and plasticity, their spatial organization, and their interactions with tumor, immune, and stromal compartments. The contribution of CAAs to therapeutic adaptation is also discussed, together with potential intervention points involving CAA-derived outputs and the metabolic dependencies induced in recipient cells. Finally, major methodological challenges are considered. Adipocyte-resolved profiling, lineage tracing, spatial multiomics, lipid flux analysis, and physiologically relevant models will be needed to distinguish transient remodeling from stable state transitions and to identify context-specific therapeutic vulnerabilities.