Background: Secondary lymphedema is a common complication of cancer treatment and epidermal changes are recognised as histological hallmarks of secondary lymphedema; however, the role of keratinocytes in the pathophysiology of this disease remains unclear.

Methods: Hyperkeratosis, up-regulation of protease-activated receptor 2 (PAR2) and Th2-inducing cytokines were assessed in biopsy specimens from patients with unilateral breast cancer-related lymphedema (BCRL) and in a mouse model of lymphedema. PAR2 inhibition using global PAR2 knockout, keratinocyte-specific PAR2 KO and bone marrow chimera models, or keratinocyte proliferation inhibition using a topical formulation of Teriflunomide (TF), was analysed in mouse models of lymphedema. We also assessed the direct effects of patient-derived lymphedema lymph fluid (LF) on keratinocyte activation in vitro.

Results: Hyperkeratosis, expression of Th2-inducing cytokines and PAR2 were significantly increased in BCRL patient biopsies and mouse models. Keratinocytes play a primary role in the lymphedema development by producing T helper 2 (Th2)-inducing cytokines. Specifically, keratinocyte proliferation and PAR2 expression are early responses following lymphatic injury and regulate the expression of Th2-inducing cytokines, the migration of Langerhans cells and the infiltration of Th2-differentiated T cells into the skin. Deficiency of PAR2 or topical inhibition of thymic stromal lymphopoietin rescues secondary lymphedema by reducing Th2 inflammation. Inhibition of PAR2 activation with a small-molecule inhibitor, or the proliferation of the inhibitor TF, prevents activation of keratinocytes stimulated with lymphedema fluid. Finally, topical TF is highly effective in reducing swelling, fibrosis and inflammation and the overall pathology of lymphedema.

Conclusions: Our findings suggest that lymphedema is a chronic inflammatory skin disease, and topically targeting keratinocyte inhibition may be a clinically effective therapy for this condition.

Key points:

Background: Precise regulation of mitochondrial function is critical for liver regeneration. However, the underlying regulatory mechanism remains elusive. Here, we aimed to investigate the role of hepatocellular glutathione peroxidase 3 (GPX3) in liver regeneration.

Methods: In a 70% partial hepatectomy (PH) mouse model, immunostaining and single-cell RNA sequencing revealed significant enrichment but down-regulation of mitochondrial oxidative phosphorylation pathways post-PH, along with up-regulated hypoxia-inducible factor 1a (HIF-1a) and GPX3 in hepatocytes. Single-cell analysis confirmed peak GPX3 expression in hepatocytes at day 2 post-PH. Hepatocyte-specific GPX3 knockout impaired mitochondrial function and delayed liver regeneration.

Results: Mechanistically, immunoprecipitation–mass spectrometry and MitoCarta3.0 analysis identified voltage-dependent anion channel 1 (VDAC1) as a direct GPX3-binding partner. GPX3 interacted with VDAC1 via its A2 domain (residues 75–150), suppressing VDAC1 oligomerisation to restore mitochondrial Ca²⁺ homeostasis and preserve mitochondrial quality control (MQC). Notably, GPX3 deficiency promoted mitochondrial DNA (mtDNA) release, activating the cyclic GMP–AMP synthase (cGAS)–stimulator of interferon genes (STING) pathway in macrophages. Persistent STING hyperactivation increased interferon production while suppressing hepatocyte growth factor release, further inhibiting regeneration. Critically, GPX3 overexpression enhanced liver regeneration in both PH and hepatic ischemia–reperfusion injury models, underscoring its central role across regenerative stressors.

Conclusions: In conclusion, GPX3 promotes liver regeneration by inhibiting VDAC1 oligomerisation to stabilise mitochondrial Ca²⁺ dynamics and MQC, while preventing mtDNA-mediated functional and phenotypic alterations in macrophages, positioning it as a therapeutic target for liver regeneration.

Key points:

Background: Growing evidence from multi-cancer cohort studies has positioned the oral pathobiont Fusobacterium nucleatum (F. nucleatum) as an emerging microbial contributor to cancer progression. Increased intratumoral abundance of F. nucleatum has been reported in colorectal, breast, esophageal, pancreatic, oral, and gastric cancers and is frequently associated with adverse clinicopathological features, treatment resistance, metastatic behavior, and poor prognosis. Advances in microbiome profiling, spatial analysis, and single-cell technologies have begun to reveal how F. nucleatum colonizes tumors and interacts with host cells and tumor-associated microbial communities.

Main body: This review summarizes current evidence regarding the tumor-associated activities of F. nucleatum, with emphasis on its routes of tumor entry, spatiotemporal colonization patterns, adhesion- and glycan-dependent tropism, polymicrobial niche formation, and crosstalk with cancer cells and immune components. We discuss how F. nucleatum promotes oncogenic signaling, inflammatory amplification, genomic and epigenetic reprogramming, epithelial–mesenchymal transition, metastatic dissemination, immune evasion, and therapy adaptation. Particular attention is given to its context-dependent effects on chemotherapy, radiotherapy, and immunotherapy responses, as well as emerging strategies aimed at detecting or selectively targeting intratumoral F. nucleatum.

Conclusion: F. nucleatum represents both a biomarker-associated organism and a potentially modifiable component of the tumor microenvironment. Defining its strain-level heterogeneity, spatial ecology, and therapy-specific functions will be essential for translating microbiome-guided precision oncology from mechanistic insight into clinical application.

Key points:

Background: Silicosis is a work-related condition resulting from breathing in crystalline silica particles, marked by persistent inflammation and abnormal healing mechanisms in the lungs. Our previous studies demonstrated that inhibition of Cxcr4 with Plerixafor (AMD3100) markedly attenuates pulmonary fibrosis.

Methods: By integrating single-cell RNA sequencing with spatial transcriptomics, we analysed lung tissues from a mouse model of pneumoconiosis. Using the Robust Cell Type Decomposition algorithm to deconvolve spatial transcriptomic data, we identified Cxcr4⁺ macrophages and Cxcl12⁺ fibroblasts as central drivers of pulmonary fibrosis progression, revealing a distinct spatial co-localisation pattern between these cell populations. To further delineate macrophage heterogeneity and functional specialisation during silicosis progression, we focused on key macrophage subpopulations.

Results: AMD3100 dynamically remodels the alveolar macrophages (AMs) niche, promoting the restoration of AM homeostasis and significantly reducing both co-expression and spatial co-localisation of Cxcr4/transforming growth factor-β (TGF-β) signalling within macrophages, thereby modulating the fibrotic immune microenvironment. Mechanistically, silica dust stimulation in vitro upregulates Cxcr4 expression in the AM cell line MH-S, which in turn promotes the release of TGF-β and pro-inflammatory factors, driving fibroblast activation. Activated fibroblasts further enhance the pro-fibrotic phenotype of macrophages via secretion of Cxcl12, reinforcing the Cxcr4 signalling axis and establishing a stable positive-feedback loop.

Conclusion: Our findings suggest that silicosis-associated fibrosis progresses through a positive feedback loop involving interactions between Cxcr4⁺ AM macrophages and Cxcl12⁺ fibroblasts. These findings highlight the therapeutic promise of targeting the Cxcl12/Cxcr4 axis with AMD3100 as an innovative approach for silicosis treatment.

Background: Ferroptosis, an iron-dependent form of regulated cell death driven by lipid peroxidation, has emerged as a potential therapeutic vulnerability in breast cancer. However, increasing evidence indicates that ferroptosis sensitivity is not solely determined by tumour-intrinsic factors, but is dynamically regulated by the tumour microenvironment (TME), particularly through interactions among adipocytes, immune cells and iron metabolism.

Main body: Recent studies provide mechanistic evidence for this context dependence. Adipocyte-derived monounsaturated fatty acids such as oleic acid suppress lipid peroxidation and increase resistance to ferroptosis induction in triple-negative breast cancer, whereas ACSL4-driven polyunsaturated phospholipid remodelling enhances ferroptosis susceptibility. In parallel, CD8⁺ T-cell-derived interferon-γ promotes ferroptosis by suppressing SLC7A11-mediated cystine uptake, while tumour-associated macrophages buffer oxidative stress through iron sequestration and glutathione-dependent antioxidant programs. These opposing forces indicate that ferroptosis is governed by a coordinated adipocyte–immune–iron regulatory network rather than a single pathway. Unlike previous reviews focused mainly on tumour–intrinsic mechanisms or general TME effects, this review integrates adipocyte-derived lipid metabolism, immune-mediated redox regulation, iron handling and spatial heterogeneity into a unified ‘ferroptosis ecosystem’ framework. Based on this concept, we propose eco-ferrotherapy, a translational strategy aimed at simultaneously targeting tumour-intrinsic pathways and microenvironmental buffering systems. This framework may support subtype-specific therapeutic prioritisation, biomarker-guided patient stratification and rational combination strategies involving immunotherapy and nanomedicine.

Conclusion: Ferroptosis in breast cancer should be understood as an ecosystem-level vulnerability shaped by metabolic, immune and spatial factors. Defining and therapeutically targeting this ferroptosis ecosystem provides a conceptual and translational roadmap for improving precision treatment strategies.

Background: Hepatocellular carcinoma (HCC) is one of the leading causes of tumour-related death. T cells and cytokines play a critical role in tumour progression, but the T cell landscape correlated with HCC prognosis remains undepicted.

Methods: The prognostic significance of intra-tumoural immune cells, chemokines and cytokines were analysed using mass cytometry, bulk RNA sequencing and scRNA-seq data with survival information. The signature of CD4⁺CD8⁺ double positive T (DPT) cells was constructed using scRNA-seq and quantified by ssGSEA scores, whose association with the response to atezolizumab plus bevacizumab was evaluated. Cellular cross-talk and spatial patterns were analysed by scRNA-seq and spatial transcriptomics. Flow cytometry, gene knockdown, transwell migration, co-culture assays, qPCR and wound healing assay were performed to further validate the DPT-associated niche.

Findings: Higher intra-tumoural levels of DPT cells, CD45RA⁺CD4⁺ conventional T cells, HBEGF and CX3CR1 were associated with unfavourable prognosis in HCC. In contrast, higher infiltration of CD161⁺CD45RA⁻CD4⁺ conventional T cells and CD8⁺ T cells correlated with prolonged survival. CD45⁺EpCAM⁺ and CD45⁺α-SMA⁺ cells were more frequent in short-term survivors. DPT infiltration was identified across HCC multi-cohorts and syngeneic mouse models. In patients receiving atezolizumab plus bevacizumab, responders exhibited higher DPT ssGSEA scores than non-responders. Multi-omics analyses indicated cross-talk and spatial association of DPT cells with capillary-associated endothelial cells, supporting a pro-tumour niche. HBEGF was positively correlated with DPT cells and highly expressed in endothelial compartments. Endothelial-derived HBEGF knockdown reduced DPT migration. Moreover, DPT co-culture increased expression of signatures associated with immunosuppressive checkpoints, chemokine signalling, epithelial–mesenchymal transition and stemness in Hep3B cells and promoted their migration.

Conclusion: Our findings depicted the prognostic immune landscape of HCC by identifying distinct T cell populations and molecular interactions. DPT cells emerged as a critical biomarker for poor prognosis, and the endothelial-derived HBEGF–DPT axis could represent a potential therapeutic target.