科技科学

Osteoclasts intercept meningeal lymphatic paths for dural metastasis

Nature作者：Rui Zeng2026年9月30日查看原文 ↗

Data availability

All raw and processed scRNA-seq data generated during this study were deposited with the Gene Expression Omnibus (GEO) database under accession number GSE284281. The RNA sequencing data of meningeal LECs were also deposited to the GEO database under the accession number GSE325223. The scRNA-seq data used for control OCs were obtained from the GEO database under accession number GSE147174 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE147174). Source data are provided with this paper.

Code availability

All data analysis and processing were executed using published software or packages, the details of which were described previously and are referred to in Methods. No new code or mathematical algorithms were generated by this study.

References

  1. Laigle-Donadey, F. et al. Dural metastases. J. Neurooncol. 75, 57–61 (2005).

    Article  PubMed  Google Scholar 

  2. Nayak, L. et al. Intracranial dural metastases. Cancer 115, 1947–1953 (2009).

    Article  PubMed  Google Scholar 

  3. Ahn, S. J. et al. Dural metastasis in breast cancer: MRI-based morphological subtypes and their clinical implications. Cancer Res. Treat. 56, 1105–1112 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Rodriguez-Calero, A. et al. Alterations in homologous recombination repair genes in prostate cancer brain metastases. Nat. Commun. 13, 2400 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  5. Nassiri, F. et al. A clinically applicable integrative molecular classification of meningiomas. Nature 597, 119–125 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  6. Wang, A. Z. et al. Single-cell profiling of human dura and meningioma reveals cellular meningeal landscape and insights into meningioma immune response. Genome Med. 14, 49 (2022).

    Article  PubMed  PubMed Central  Google Scholar 

  7. Rippaus, N. et al. Metastatic site-specific polarization of macrophages in intracranial breast cancer metastases. Oncotarget 7, 41473–41487 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  8. Merlini, A. et al. Distinct roles of the meningeal layers in CNS autoimmunity. Nat. Neurosci. 25, 887–899 (2022).

    Article  CAS  PubMed  Google Scholar 

  9. Koh, B. I. et al. Adult skull bone marrow is an expanding and resilient haematopoietic reservoir. Nature 636, 172–181 (2024).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  10. Kolabas, Z. I. et al. Distinct molecular profiles of skull bone marrow in health and neurological disorders. Cell 186, 3706–3725 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Dobersalske, C. et al. Cranioencephalic functional lymphoid units in glioblastoma. Nat. Med. 30, 2947–2956 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Lad, M. et al. Glioblastoma induces the recruitment and differentiation of dendritic-like ‘hybrid’ neutrophils from skull bone marrow. Cancer Cell 42, 1549–1569 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Rustenhoven, J. et al. Functional characterization of the dural sinuses as a neuroimmune interface. Cell 184, 1000–1016 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Herisson, F. et al. Direct vascular channels connect skull bone marrow and the brain surface enabling myeloid cell migration. Nat. Neurosci. 21, 1209–1217 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Pulous, F. E. et al. Cerebrospinal fluid can exit into the skull bone marrow and instruct cranial hematopoiesis in mice with bacterial meningitis. Nat. Neurosci. 25, 567–576 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Louveau, A. et al. Structural and functional features of central nervous system lymphatic vessels. Nature 523, 337–341 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  17. Aspelund, A. et al. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J. Exp. Med. 212, 991–999 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Zhao, J. et al. Dura immunity configures leptomeningeal metastasis immunosuppression for cerebrospinal fluid barrier invasion. Nat. Cancer 5, 1940–1961 (2024).

    Article  CAS  PubMed  Google Scholar 

  19. Sankowski, R. et al. Multiomic spatial landscape of innate immune cells at human central nervous system borders. Nat. Med. 30, 186–198 (2024).

    Article  CAS  PubMed  Google Scholar 

  20. Hu, X. et al. Meningeal lymphatic vessels regulate brain tumor drainage and immunity. Cell Res. 30, 229–243 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Song, E. et al. VEGF-C-driven lymphatic drainage enables immunosurveillance of brain tumours. Nature 577, 689–694 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  22. Zhou, C. et al. Meningeal lymphatics regulate radiotherapy efficacy through modulating anti-tumor immunity. Cell Res. 32, 543–554 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Salvador, A. F. M. et al. Meningeal lymphatics in central nervous system diseases. Annu. Rev. Neurosci. 47, 323–344 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Dissing-Olesen, L. et al. New brain lymphatic vessels drain old concepts. eBioMedicine 2, 776–777 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  25. Petrova, T. V. et al. Biological functions of lymphatic vessels. Science 369, eaax4063 (2020).

    Article  CAS  PubMed  Google Scholar 

  26. Janardhan, H. P. et al. Lymphatic system in organ development, function, and regeneration. Circ. Res. 132, 1181–1184 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Louveau, A. et al. CNS lymphatic drainage and neuroinflammation are regulated by meningeal lymphatic vasculature. Nat. Neurosci. 21, 1380–1391 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Alves de Lima, K. et al. Meningeal immunity and its function in maintenance of the central nervous system in health and disease. Annu. Rev. Immunol. 38, 597–620 (2020).

    Article  CAS  PubMed  Google Scholar 

  29. Graham, M. S. et al. Meningeal lymphatics prime tumor immunity in glioblastoma. Cancer Cell 39, 304–306 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Kovacs, M. A. et al. Meningeal lymphatic drainage promotes T cell responses against Toxoplasma gondii but is dispensable for parasite control in the brain. eLife 11, e80775 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Castellani, G. et al. Transforming the understanding of brain immunity. Science 380, eabo7649 (2023).

    Article  CAS  PubMed  Google Scholar 

  32. Oliveira, G. et al. Dynamics and specificities of T cells in cancer immunotherapy. Nat. Rev. Cancer 23, 295–316 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. John, L. et al. Resolving the spatial architecture of myeloma and its microenvironment at the single-cell level. Nat. Commun. 14, 5011 (2023).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  34. Monroy, M. et al. Lymphatics in bone arise from pre-existing lymphatics. Development 147, dev184291 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Zhang, D. et al. Reduced expression of semaphorin 3A in osteoclasts causes lymphatic expansion in a Gorham–Stout disease (GSD) mouse model. J. Zhejiang Univ. Sci. B 25, 38–50 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Wang, W. et al. Lymphatic endothelial cells produce M-CSF, causing massive bone loss in mice. J. Bone Miner. Res. 32, 939–950 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Sato, K. et al. RANKL synthesized by both stromal cells and cancer cells plays a crucial role in osteoclastic bone resorption induced by oral cancer. Am. J. Pathol. 182, 1890–1899 (2013).

    Article  CAS  PubMed  Google Scholar 

  38. Renema, N. et al. RANK–RANKL signalling in cancer. Biosci. Rep. 36, e00366 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  39. Li, E. C. et al. Zoledronic acid: a new parenteral bisphosphonate. Clin. Ther. 25, 2669–2708 (2003).

    Article  CAS  PubMed  Google Scholar 

  40. Russell, R. G. Bisphosphonates: the first 40 years. Bone 49, 2–19 (2011).

    Article  CAS  PubMed  Google Scholar 

  41. Wang, L. et al. Mechanical sensing protein PIEZO1 regulates bone homeostasis via osteoblast-osteoclast crosstalk. Nat. Commun. 11, 282 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  42. Nakamura, T. et al. Estrogen prevents bone loss via estrogen receptor alpha and induction of Fas ligand in osteoclasts. Cell 130, 811–823 (2007).

    Article  CAS  PubMed  Google Scholar 

  43. Yang, W. et al. Ptpn11 deletion in a novel progenitor causes metachondromatosis by inducing hedgehog signalling. Nature 499, 491–495 (2013).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  44. Feng, H. et al. Tendon-derived cathepsin K-expressing progenitor cells activate Hedgehog signaling to drive heterotopic ossification. J. Clin. Invest. 130, 6354–6365 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Debnath, S. et al. Discovery of a periosteal stem cell mediating intramembranous bone formation. Nature 562, 133–139 (2018).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  46. Han, Y. et al. Lkb1 deletion in periosteal mesenchymal progenitors induces osteogenic tumors through mTORC1 activation. J. Clin. Invest. 129, 1895–1909 (2019).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  47. McDonald, M. M. et al. Osteoclasts recycle via osteomorphs during RANKL-stimulated bone resorption. Cell 184, 1330–1347 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Tsukasaki, M. et al. Stepwise cell fate decision pathways during osteoclastogenesis at single-cell resolution. Nat. Metab. 2, 1382–1390 (2020).

    Article  CAS  PubMed  Google Scholar 

  49. Dieterich, L. C. et al. Lymphatic vessels in cancer. Physiol. Rev. 102, 1837–1879 (2022).

    Article  CAS  PubMed  Google Scholar 

  50. Zheng, W. et al. Lymphangiogenic factors, mechanisms, and applications. J. Clin. Invest. 124, 878–887 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Souma, T. et al. Context-dependent functions of angiopoietin 2 are determined by the endothelial phosphatase VEPTP. Proc. Natl Acad. Sci. USA 115, 1298–1303 (2018).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  52. Khosraviani, N. et al. Angiopoietin-2: an emerging tie to pathological vessel enlargement. Arterioscler. Thromb. Vasc. Biol. 42, 3–5 (2022).

    Article  CAS  PubMed  Google Scholar 

  53. Lu, X. et al. VCAM-1 promotes osteolytic expansion of indolent bone micrometastasis of breast cancer by engaging α4β1-positive osteoclast progenitors. Cancer Cell 20, 701–714 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Sethi, N. et al. Tumor-derived Jagged1 promotes osteolytic bone metastasis of breast cancer by engaging Notch signaling in bone cells. Cancer Cell 19, 192–205 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Hirbe, A. C. et al. Granulocyte colony-stimulating factor enhances bone tumor growth in mice in an osteoclast-dependent manner. Blood 109, 3424–3431 (2006).

    Article  PubMed  PubMed Central  Google Scholar 

  56. Yue, Z. et al. RSPO2 and RANKL signal through LGR4 to regulate osteoclastic premetastatic niche formation and bone metastasis. J. Clin. Invest. 132, e144579 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Cheng, J.-N. et al. Bone metastases diminish extraosseous response to checkpoint blockade immunotherapy through osteopontin-producing osteoclasts. Cancer Cell 43, 1093–1107 (2025).

    Article  CAS  PubMed  Google Scholar 

  58. Hua, X. Y. et al. Structural and functional characteristics of leptomeningeal lymphatic vessels in leptomeningeal metastases from lung cancer patients. Preprint at Research Square https://doi.org/10.21203/rs.3.rs-5244229/v1 (2024).

  59. Schulz, P. et al. Angiopoietin-2 drives lymphatic metastasis of pancreatic cancer. FASEB J. 25, 3325–3335 (2011).

    Article  CAS  PubMed  Google Scholar 

  60. Fagiani, E. et al. Angiopoietin-1 and -2 exert antagonistic functions in tumor angiogenesis, yet both induce lymphangiogenesis. Cancer Res. 71, 5717–5727 (2011).

    Article  CAS  PubMed  Google Scholar 

  61. Chi, Y. et al. Cancer cells deploy lipocalin-2 to collect limiting iron in leptomeningeal metastasis. Science 369, 276–282 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  62. Boire, A. et al. Complement component 3 adapts the cerebrospinal fluid for leptomeningeal metastasis. Cell 168, 1101–1113 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Choi, D. et al. Piezo1 regulates meningeal lymphatic vessel drainage and alleviates excessive CSF accumulation. Nat. Neurosci. 27, 913–926 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Butler, A. et al. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Korsunsky, I. et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat. Methods 16, 1289–1296 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Gu, Z. et al. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847–2849 (2016).

    Article  CAS  PubMed  Google Scholar 

  67. Yu, G. et al. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284–287 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank all members of the Chi laboratory for valuable discussions. We thank W. Zhang for her preliminary exploratory efforts on the DM mouse model construction method. We acknowledge the staff of the Department of Laboratory Animal Science in Fudan University. We thank all members of the Innovative Research Team of High-Level Local University in Shanghai at each of the participating sites for their contributions to this study. We extend our appreciation to the Institute of Brain Research of Fudan University for their two-photon microscopy platform. We also thank the support by the Medical Science Data Center in Shanghai Medical College of Fudan University.

Funding

The research received funding from the Ministry of Science and Technology of China (2022YFA1106400 to Y.C.) and the National Natural Science Foundation of China (82573109/82621102 to Y.C. and 82403445 to Y.H.).

Author information

Author notes

  1. These authors contributed equally: Rui Zeng, Jiaxu Zhao, Haibao Peng

Authors and Affiliations

  1. Department of Neurosurgery of Huashan Hospital, State Key Laboratory of Brain Function and Disorders and MOE Frontiers Center for Brain Science, Institute for Translational Brain Research, Fudan University, Shanghai, China

    Rui Zeng, Haibao Peng, Jiadong Ma, Youming Zeng, Xiaohui Li, Yang Huang, Cheng Li, Jiahui Yang, Liang Chen & Yudan Chi

  2. Department of Radiology, Huadong Hospital, Shanghai Key Laboratory of Clinical Geriatric Medicine, Shanghai Institute of Geriatrics and Gerontology, Fudan SANS Neuroscience Center, Fudan University, Shanghai, China

    Rui Zeng, Haibao Peng, Jiadong Ma, Shuyi Hua, Youming Zeng, Xiaohui Li, Yang Huang, Cheng Li, Jiahui Yang, Hongru Chen & Yudan Chi

  3. School of Chemistry and Molecular Engineering, East China Normal University, Shanghai, China

    Jiaxu Zhao

  4. Laboratory Animal Center, Fudan University, Shanghai, China

    Xue Feng

  5. Department of Biotechnology and Biomedicine, Advanced Model Animal Research Center, Zhejiang Key Laboratory of Multiomics and Molecular Enzymology, Yangtze Delta Region Institute of Tsinghua University, Zhejiang, China

    Qiangqiang Zhang

  6. Tianqiao and Chrissy Chen Institute Clinical Translational Research Center, Shanghai, China

    Liang Chen

  7. Shanghai Key Laboratory of Brain Function and Restoration and Neural Regeneration, Shanghai, China

    Liang Chen

Authors

  1. Rui Zeng
  2. Jiaxu Zhao
  3. Haibao Peng
  4. Jiadong Ma
  5. Shuyi Hua
  6. Youming Zeng
  7. Xiaohui Li
  8. Yang Huang
  9. Cheng Li
  10. Jiahui Yang
  11. Hongru Chen
  12. Xue Feng
  13. Qiangqiang Zhang
  14. Liang Chen
  15. Yudan Chi

Contributions

R.Z. and Y.C. conceptualized the study. R.Z. carried out clinical image and micro-CT analysis. R.Z., J.M., Y.Z., X.L. and C.L. performed immunofluorescence and TRAP staining. J.Z. and J.M. carried out scRNA-seq analysis. R.Z., J.Y. and H.C. performed flow cytometry analysis and cell sorting. R.Z. and S.H. performed bioluminescence imaging and survival analysis. H.P. carried out two-photon intravital imaging. R.Z., X.F., L.C. and Y.C. provided resources. R.Z., J.Z., H.P., J.M., Y.Z., L.C. and Y.C. wrote the manuscript with advice from Q.Z., X.L., C.L., J.Y. and Y.H. All authors reviewed and approved the manuscript. The work was supervised by J.Z., L.C. and Y.C.

Corresponding authors

Correspondence to Jiaxu Zhao, Liang Chen or Yudan Chi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks Jonathan Kipnis and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Skull erosion is associated with dural metastasis but absent in non-tumour and leptomeningeal metastasis.

(a) Magnetic resonance imaging of heads from patients diagnosed with meningioma (n = 14 patients), breast cancer brain metastasis (BRCA BM, n = 10 patients), breast cancer leptomeningeal metastasis (BRCA LM, n = 7 patients), and breast cancer dural metastasis (BRCA DM, n = 8 patients). Tumour lesions were pointed by white arrowheads. (b) Magnetic resonance imaging of heads from patients diagnosed with lung cancer brain metastasis (LUCA BM, n = 10 patients), lung cancer leptomeningeal metastasis (LUCA LM, n = 8 patients), and lung cancer dural metastasis (LUCA DM, n = 6 patients). The tumour lesions were pointed out by white arrowheads. (c) Statistical analysis of distance between tumour to skull from data in Extended Data Fig. 1a,b. (d) CT of human cranium from patients in b. Skull bone thinning was pointed out by pink arrowheads. (e) Bioluminescent images with CT from 4T1-DM mice, depicting the specific dissemination and habitation of tumour cells in the dura mater. (f) Statistical proportional analysis of locations of tumour lesions. n = 5 mice. (g) Three-dimensional reconstruction and two-dimensional sections of CT showing cranium from non-tumour and 4T1-LM mice. White arrowheads mark the corresponding regions in non-tumour or LM skulls as pink arrowheads in Fig. 1d. (h) Bioluminescent and H&E staining images of coronal head sections from non-tumour, 4T1-LM and 4T1-DM mice. Tumour lesions were pointed out by black arrowheads. Scale bar, 1 mm (upper staining image) and 200 μm (lower staining image). (i) Statistical analysis of skull thickness above tumour lesions in h. n = 5 mice per group. Rem., remote to lesion. Adj., adjacent to lesion. Data represent means ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparison test in (c, f, i). NS, not significant.

Source data

Extended Data Fig. 2 DM exerts no influence on lymphangiogenesis in dura mater and validation of the lymphatic morphology and drainage in carotid-ligated non-tumour mice.

(a) Whole-mount confocal fluorescent images of dura from non-tumour mice and 4T1-DM mice. Regions adjacent to or remote from tumour lesions were presented respectively. The samples were stained for LYVE1 (lymphatic endothelial cell marker, green), CD31 (endothelial cell marker, magenta) and DAPI (gray). Red, tdTomato+ tumour cells. Scale bar, 1 mm (low magnification) and 200 μm (inset). (b) Statistical analysis of LYVE1+ area from a. n = 5 mice per group. Data were normalized to non-tumour. TS, transverse sinus. SSS, superior sagittal sinus. (c) Intravital two photon fluorescent images of dura from non-tumour mice and 4T1-DM mice at 12 days post-tumour inoculation. Original Qdot signals (top) and surface reconstructions (bottom) in regions adjacent to or remote from tumour lesions were presented respectively. The dilation was labeled with a white rectangle. The bleb was labeled with a blue rectangle. The local injuries were labeled with green arrowheads. Scale bar, 20 μm. (d-f) Statistical analysis of diameter (d), bleb (e) and local injuries (f) of Qdot+ lymphatic vessels. n = 5 mice per group. (g) Whole-mount confocal fluorescent images of dura mater from control (Ctrl) Balb/c mice and carotid artery ligated Balb/c mice. The samples were stained for LYVE1 (green), CD31 (magenta) and DAPI (blue). Scale bar, 200 μm. (h) Statistical analysis of LYVE1+ area from g. n = 4 per group. (i) Experimental design of MLVs drainage evaluation. Balb/c mice were given an intracisternal injection of 70 kDa FITC-dextran or OVA-AF488 tracer. The dCLNs were harvested 1 h after tracer injection. (j) Whole-mount fluorescent images of dCLNs from Ctrl mice and carotid artery ligated mice. The samples were stained for DAPI (blue). BF, bright field; Green, dextran. Scale bar, 2 mm. (k) Statistical analysis of normalized 70 kDa FITC-dextran fluorescent signals from j, data were normalized to Ctrl. n = 4 per group. (l) Whole-mount fluorescent images of dCLNs from Ctrl mice and carotid artery ligated mice. The samples were stained for DAPI (blue). BF, bright field; Green, OVA-AF488. Scale bar, 2 mm. (m) Statistical analysis of normalized OVA-AF488 fluorescent signals from l, data were normalized to Ctrl. n = 4 mice per group. Data represent means ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparison test in (b, d, e, f), and two-sided unpaired t-test in (h, k, m). NS, not significant.

Source data

Extended Data Fig. 3 Signature genes in scRNA-seq from non-tumour or LM dura and cluster-highly-expressing genes in scRNA-seq data of DM dura maters.

(a) UMAP of color-coded clusters based on manually annotated cell types from dura maters of non-tumour mice. (b) Dot plot depicting the expression of marker genes across all clusters from scRNA-seq data of cells in a. T+NK cell, Cd3e and Ncr1; ILC2, Gata3 and Il1rl1; B cell, Cd19 and Cd79b; monocyte-macrophage, Aif1, Adgre1, Ly6c2 and Lyz2; neutrophil, Csf3r and Ly6g; DC, Itgax and Flt3; endothelial cell, Cdh5 and Pecam1; Schwann cell, Sox10 and Plp1; Habenula neuron contamination, Pde6g, Tph1; fibroblast, Col1a1 and Dcn. (c) UMAP of color-coded clusters based on manually annotated cell types from dura maters of LM mice. (d) Dot plot depicting the expression of marker genes across all clusters from scRNA-seq data of cells in c. T+NK cell, Cd3e and Ncr1; ILC2, Gata3 and Il1rl1; B cell, Cd19 and Cd79b; monocyte-macrophage, Aif1, Adgre1, Ly6c2 and Lyz2; neutrophil, Csf3r and Ly6g; DC, Itgax and Flt3; endothelial cell, Cdh5 and Pecam1; Schwann cell, Sox10 and Plp1; Habenula neuron contamination, Pde6g, Tph1; fibroblast, Col1a1 and Dcn; epithelial cell, Epcam; Proliferative cell, Mki67. (e) Heatmap representing the expression of the top 5 signature genes from each cell type defined in Fig. 2b.

Extended Data Fig. 4 Analyses of dural immune compositions from non-tumour, leptomeningeal metastasis and dural metastasis models.

(a) Statistical analysis of dural immune cell proportions from scRNA-seq data of non-tumour, LM and DM mice. n = 3 samples (each sample contains mixture of dura maters from 6 mice) per group. (b) Flow cytometry analysis of dura maters from non-tumour and 4T1-DM mice. (c, d) Statistical analysis of absolute number per sample (c) and proportions of each WBC subsets (d) from b. n = 5 mice per group, from three independent experiments. Data represent means ± SEM. Statistical analysis was performed using two-way ANOVA with Tukey’s multiple comparison test in (a), and two-sided multiple t-test with two-stage step-up procedure in (c, d). NS, not significant.

Source data

Extended Data Fig. 5 Representing gating strategy for dural T cell exhaustion.

(a) Signature genes expression in T cell subsets of non-tumour, 4T1-LM or 4T1-DM dura maters. Related to Fig. 2e. (b) Flow cytometry analysis of CD45+CD3e+ T cells in dura maters from non-tumour and 4T1-DM mice. (c) Statistical analysis of data from b. n = 6 mice per group, from three independent experiments. (d) Flow cytometry gating strategy of CD8+ T cells and CD4+ T cells in dura maters from non-tumour and 4T1-DM mice. Related to Fig. 2h,j. (e) UMAP of cytotoxicity score (based on Prf1, Gzma, Gzmb, Ifng, Tnf) in T cells of non-tumour, 4T1-LM or 4T1-DM mice. Data represent means ± SEM. Statistical analysis was performed with two-sided unpaired t-test in (c).

Source data

Extended Data Fig. 6 Validation of T cells in carotid-ligated mice and representing gating strategy for cytokines secretion in dural T cells.

(a) Flow cytometry analysis of CD45+CD3e+ T cells in dura maters from Ctrl Balb/c mice and carotid artery ligated Balb/c mice. (b) Statistical analysis of data from a. n = 5 mice per group, from three independent experiments. (c) Flow cytometry gating strategy of CD8+ T cells and CD4+ T cells in dura maters from non-tumour and 4T1-DM mice. Related to Fig. 2l,n. (d-f) UMAP showing the expression of Tnfsf11 from scRNA-seq data of dural cells of non-tumour (d), 4T1-LM (e) and 4T1-DM (f) mice. Data represent means ± SEM. Statistical analysis was performed with two-sided unpaired t-test in (b). NS, not significant.

Source data

Extended Data Fig. 7 Osteoclasts depletion rescues MLVs morphology and drainage ability in dural metastasis.

(a) Whole-mount confocal fluorescent images of dura maters from non-tumour Balb/c mice that were treated with PBS or ZOL every 3 days. Experiments were performed on day 12. The samples were stained for LYVE1 (green), CD31 (magenta) and DAPI (blue). Scale bar, 400 μm. (b) Statistical analysis of LYVE1+ region area from a. n = 4 per group. (c) Intravital two photon fluorescent images of dura from 4T1-DM mice that were treated with PBS or ZOL every 3 days until day 12 after tumour cell inoculation. Original Qdot signals (top) and surface reconstructions (bottom) were presented respectively. The dilation was labeled with a white rectangle. The bleb was labeled with a blue rectangle. The local injuries were labeled with green arrowheads. Scale bar, 20 μm. (d-f) Statistical analysis of diameter (d), bleb (e) and local injuries (f) of Qdot+ lymphatic vessels. n = 5 mice per group. (g) TRAP staining of coronal head section in 4T1-DM mice at 12 days post-tumour inoculation. Mice were treated with or without ZOL. Scale bar, 50 μm (large images) and 10 μm (inset). (h-j) Statistical analysis of TRAP+ cell densities from skull-tumour interface (h) and tumour core (i), and resorption cavities (j) per mm bone surface in (g). n = 5 mice per group. (k) CCK8 assay at varied concentrations of ZOL in 4T1-DM cells. n = 3 per group. Data passed Shapiro-Wilk normality test. Data represent means ± SEM. Statistical analysis was performed with two-sided unpaired t-test in (b, d, e, f, h, i, j), and one-way ANOVA with Tukey’s multiple comparison test in (k). NS, not significant.

Source data

Extended Data Fig. 8 Quantitative analyses reveal osteoclasts depletion boosts T cell infiltration.

(a) Flow cytometry analysis of dural cells from 4T1-DM mice that were treated with PBS or ZOL every 3 days since tumour inoculation. Experiments were performed on day 12 since tumour inoculation. (b, c) Statistical analysis of absolute number per sample (b) and proportions of each WBC subsets (c) in a. n = 5 mice per group, from three independent experiments. (d) Flow cytometry analysis of CD45+CD3e+ T cells in dura maters from 4T1-DM mice that were treated with PBS or ZOL every 3 days since tumour inoculation. (e) Statistical analysis of data from d. n = 6 mice per group, from three independent experiments. (f) Flow cytometry analysis of dural CD8+ T cells and CD4+ T cells in PBS or ZOL-treated 4T1-DM mice at 12 days post-tumour inoculation, showing the PD1- and TIM3-expressing cells. (g) Experimental set-up related to h. (h) Left: quantification of tumour growth in 4T1-DM mice treated with PBS+sham surgery, PBS+dCLNs ligation, ZOL+ sham surgery or ZOL+ dCLNs ligation at 12 days post-tumour inoculation. Right: representative BLI image by IVIS assay from n = 6 mice per group, from three independent experiments. Data represent means ± SEM. Statistical analysis was performed with multiple t-test with two-stage step-up procedure in (b, c), two-sided unpaired t-test in (e), and one-way ANOVA with Tukey’s multiple comparison test in (h). NS, not significant.

Source data

Extended Data Fig. 9 Validation of CtskCre:Ai9 transgenic mice for excluding mislabeling of non-osteoclast cells.

(a) Confocal fluorescent images and adjacent H&E staining image of femur bone marrow from CtskCre:Ai9 mice. The samples were stained for F4/80 (green) and DAPI (blue). Red, tdTomato+ cells. Scale bar, 50 μm. (b) The proportion of F4/80+ in tdTomato+ cells in femur bone marrow from a. n = 5 mice. (c) Dot plot showing the relative expression of Dlx5, Fgfr3, Scx, Tnmd, Tnc, Runx2, Sox9 and Col1a1 in all clusters from scRNA-seq data of cells in Fig. 2b. (d) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for DLX5 (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (e) The proportion of DLX5+ in tdTomato+ OCs or ITCs from d. n = 5 mice per group. (f) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for SCX (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (g) The proportion of SCX+ in tdTomato+ OCs or ITCs from f. n = 5 mice per group. (h) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for TNMD (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (i) The proportion of TNMD+ in tdTomato+ OCs or ITCs from h. n = 5 mice per group. (j) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for RUNX2 (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (k) The proportion of RUNX2+ in tdTomato+ OCs or ITCs from j. n = 5 mice per group. (l) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for COL1A1 (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (m) The proportion of COL1A1+ in tdTomato+ OCs or ITCs from l. n = 5 mice per group. (n) Violin plot showing the relative expression of Acp5 in all clusters from scRNA-seq data of cells in Fig. 2b. (o) Time-lapse 2-photon fluorescent images of skull-dural interface in non-tumour CtskCre:Ai9 mice. Red, tdTomato+ cells; Blue, second harmonic generation (SHG) signals. Scale bar, 100 μm. (p) Statistical analysis of tdTomato+ cell density in skull-dural interface from o. n = 5 mice. (q) Confocal fluorescent images of dura maters from LLC-DM-CopGFP models of CtskCre:Ai9 mice. The samples were stained for F4/80 (magenta) and DAPI (gray). Red, tdTomato+ cells; Green, CopGFP+ tumour cells. Scale bar, 20 μm. (r) The proportion of F4/80+ in tdTomato+ OCs or ITCs from q. n = 5 mice per group. (s) UMAP of color-coded clusters from integration of DM associated OCs (Fig. 4j) and control OCs (GSE147174). (t, u) GO biological processes analysis enriched from up-regulated genes in OC1 versus control OC (t) and up-regulated genes in OC2 versus control OC (u), cutoff = log2FC > 1. Data represent means ± SEM. Statistical analysis was performed with two-sided hypergeometric distribution with Benjamini-Hochberg-adjusted P (t, u).

Source data

Extended Data Fig. 10 PDGF-BB does not affect meningeal lymphatic drainage and tumour progression in dural metastasis.

(a) Experimental set-up related to b-l. 4T1-DM mice were treated with ZOL (every 3 days), or the combination of ZOL (every 3 days) and recombinant PDGF-BB (every day) since tumour inoculation. Experiments were performed at 12 days post-tumour inoculation. (b) Confocal fluorescent images of dura mater from ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. The samples were stained for LYVE1 (magenta) and DAPI (blue). Red, tdTomato+ tumour cells. Scale bar, 400 μm. (c) Statistical analysis of LYVE1+ vessel area from b. n = 5 per group. Data were normalized to ZOL group. (d) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of 40 kDa FITC-dextran, 70 kDa FITC-dextran or OVA-AF488 tracer, respectively. The dCLNs were harvested 1 h after tracer injection. (e) Whole-mount fluorescent images of dCLNs from ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, dextran; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (f) Statistical analysis of relative 40 kDa FITC-dextran fluorescent signals from e, data were normalized to ZOL-treated DM mice. n = 5 mice per group. (g) Whole-mount fluorescent images of dCLNs from ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, dextran; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (h) Statistical analysis of relative 70 kDa FITC-dextran fluorescent signals from g, data were normalized to ZOL-treated DM mice. n = 5 mice per group. (i) Whole-mount fluorescent images of dCLNs from ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, OVA-AF488; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (j) Statistical analysis of relative OVA-AF488 fluorescent signals from i, data were normalized to ZOL-treated DM mice. n = 5 mice per group. (k) Left: quantification of tumour growth in ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. Right: two representative BLI images by IVIS assay from n = 5 mice per group, from three independent experiments. (l) Kaplan-Meier survival curve of ZOL-treated or ZOL + PDGF-BB-treated 4T1-DM mice. n = 10 per group, from two independent experiments. (m) Violin plot showing the relative expression of Angpt2 in each cell population in DM dura. (n) Confocal fluorescent images of femur bone from CtskCre:Ai9 mice. The samples were stained for ANG2 (green) and DAPI (gray). Red, tdTomato+ cells. Scale bar, 20 μm. (o) The proportion of ANG2+ in tdTomato+ cells from n. n = 5 mice. Data represent means ± SEM. Statistical analysis was performed with two-sided unpaired t-test in (c, f, h, j, k), and log-rank test (l). NS, not significant.

Source data

Extended Data Fig. 11 ANG2 staining in non-tumour mice dura and MLVs drainage evaluation in ANG2-treated non-tumour mice.

(a) Confocal fluorescent images of dura maters from non-tumour CtskCre:Ai9 mice. The samples were stained for ANG2 (green) and DAPI (blue). Red, tdTomato+ cells. Scale bar, 40 μm. (b) Statistical analysis of ANG2+ cells in DAPI+ cells from a. n = 5 mice. (c) Intravital two photon fluorescent images of dura from ZOL-treated 4T1-DM mice with or without ANG2 administration. Original Qdot signals (top) and surface reconstructions (bottom) were presented respectively. The dilation was labeled with a white rectangle. The bleb was labeled with a blue rectangle. The local injuries were labeled with green arrowheads. Scale bar, 20 μm. (d-f) Statistical analysis of diameter (d), bleb (e) and local injuries (f) of Qdot+ lymphatic vessels. n = 5 mice per group. (g) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of 70 kDa FITC-dextran or OVA-AF488 tracer, respectively. The dCLNs were harvested 1 h after tracer injection. (h, i) Whole-mount fluorescent images of dCLNs from ZOL-treated or ZOL+ANG2-treated 4T1-DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, tracers; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (j) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of 70 kDa FITC-dextran and OVA-AF488 tracers. The dCLNs were harvested 1 h after tracer injection. (k, m) Whole-mount fluorescent images of dCLNs from non-tumour Balb/c mice that were treated with PBS or ANG2 daily. The samples were stained for DAPI (blue). BF, bright field; Green, tracers. Scale bar, 2 mm. (l, n) Statistical analysis of relative tracer signals from k, m. Data were normalized to PBS group. n = 5 per group. (o) Whole-mount confocal fluorescent images of dura maters from non-tumour Balb/c mice that were treated with PBS or ANG2 every day. Experiments were performed on day 12. The samples were stained for CD31 (magenta) and DAPI (blue). Scale bar, 400 μm. (p) Statistical analysis of CD31+ region area from o. n = 5 mice per group. Data represent means ± SEM. Statistical analysis was performed using two-sided unpaired t-test in (d, e, f, l, n, p).

Source data

Extended Data Fig. 12 ANG2 knock-down in osteoclasts rescued osteoclasts-related MLVs impairments and provide therapeutic efficacy.

(a) Experimental set-up related to b-n. LLC-DM or 4T1-DM mice were treated with pAAV-Ctsk-Angpt2-shRNA to induce Angpt2 knockdown in Ctsk-expressing cells at 3 days post-tumour inoculation. Experiments were performed at 12 days post-tumour inoculation. (b) Confocal fluorescent images of dura maters from Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) treated LLC-DM model of CtskCre:Ai9 mice. The samples were stained for ANG2 (green) and DAPI (gray). Red, tdTomato+ cells. Scale bar, 20 μm. (c) The mean fluorescence intensity of ANG2 in tdTomato+ OCs or ITCs from b. Data were normalized to ShCtrl group. n = 5 mice per group. (d) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of 40 kDa FITC-dextran. The dCLNs were harvested 1 h after tracer injection. (e) Whole-mount fluorescent images of dCLNs from Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, dextran; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (f) Statistical analysis of relative 40 kDa FITC-dextran fluorescent signals from e. Data were normalized to ShCtrl group. n = 5 mice per group. (g) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of 70 kDa FITC-dextran. The dCLNs were harvested 1 h after tracer injection. (h) Whole-mount fluorescent images of dCLNs from Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, dextran; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (i) Statistical analysis of relative 70 kDa FITC-dextran fluorescent signals from h. Data were normalized to ShCtrl group. n = 5 mice per group. (j) Experimental design of MLVs drainage evaluation. Mice were given an intracisternal injection of OVA-AF488. The dCLNs were harvested 1 h after tracer injection. (k) Whole-mount fluorescent images of dCLNs from Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) DM mice. The samples were stained for DAPI (blue). BF, bright field; Green, OVA; Red, tdTomato+ tumour cells. Scale bar, 2 mm. (l) Statistical analysis of relative OVA-AF488 fluorescent signals from k. Data were normalized to ShCtrl group. n = 5 mice per group. (m) Left: quantification of tumour growth in Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) DM mice. Right: two representative BLI images by IVIS assay. n = 5 mice per group, from three independent experiments. (n) Kaplan-Meier survival curve of Ctsk-shRNA(NC) or Ctsk-shRNA(Angpt2) DM mice. n = 10 per group, from two independent experiments. Data represent means ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparison test in (c, f, i, l, m), and log-rank test in (n). NS, not significant.

Source data

Extended Data Fig. 13 RNA-seq of meningeal lymphatic endothelial cells in dural metastasis mice after osteoclasts depletion and ANG2 administration.

(a) Schematic depiction for experimental setup related to b-e. Dura maters from PBS-treated, ZOL-treated or ZOL+ANG2-treated 4T1-DM mice were harvested for FACS at 12 days post-tumour inoculation. Meningeal LECs were isolated for RNA-seq. Each sample contains mixture of samples from 6 mice. (b) Flow cytometry of dural cells in a, LECs were gated as CD45−CD31+PDPN+ cells. (c) PCA plots from the transcriptome of isolated LECs from b. (d, e) GO biological processes analysis enriched from up-regulated genes in LECs from ZOL versus PBS-treated DM mice (d) and down-regulated genes in LECs from ZOL+ANG2 versus ZOL-treated DM mice (e), cutoff = log2FC > 1. (f) Flow cytometry analysis of CD45+CD3e+ T cells in dura maters from ZOL-treated or ZOL+ANG2-treated 4T1-DM mice at 12 days post-tumour inoculation. (g) Statistical analysis of data from f. n = 5 mice per group, from three independent experiments. (h, i) Flow cytometry analysis of CD8+ T cells (h) and CD4+ T cells (i) from dura maters in ZOL-treated or ZOL+ANG2-treated 4T1-DM mouse model, showing the PD1- and TIM3-expressing cells. Related to Fig. 5l,m. Data represent means ± SEM. Statistical analysis was performed with two-sided hypergeometric distribution with Benjamini-Hochberg-adjusted P (d, e), and two-sided unpaired t-test in (g).

Source data

Supplementary information

Source data

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zeng, R., Zhao, J., Peng, H. et al. Osteoclasts intercept meningeal lymphatic paths for dural metastasis. Nature (2026). https://doi.org/10.1038/s41586-026-11082-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1038/s41586-026-11082-6