A New Role for Skull Marrow Immune Microenvironment in Brain Tumors: Interpretation Based on Multicenter Studies in Mice and Humans

1. Academic Background and Research Significance

Brain tumors, especially glioblastoma (GBM), are among the most aggressive malignant tumors in the central nervous system. Traditionally, they were considered localized diseases, but recent evidence shows that GBM exhibits widespread systemic effects, including remodeling of primary and secondary immune organs such as the spleen, thymus, bone marrow, etc. Recent studies have found that skull marrow (SM), as the “immune reservoir” for brain tissue, can replenish monocytes and neutrophils into the brain during brain injury and disease (such as autoimmune encephalomyelitis, stroke, etc.), but its specific role in brain cancers (especially in GBM) remains unclear.

The uniqueness of the skull marrow microenvironment lies in: its direct communication pathways to the brain surface via vascularized ossified channels traversing the skull to the dura, enabling immune cells and molecules to shuttle between the skull and brain tissue. This mechanism has been confirmed in various neurological disorders, but whether GBM can alter this marrow microenvironment and its bone structure, affecting immune cell dynamics, thereby interfering with brain tumor development and immunotherapy efficacy, remains an unsolved scientific question. This study addresses this knowledge gap, systematically analyzing the impact of GBM on skull bone structure, bone marrow immune microenvironment, and responses to immunotherapy, opening new directions for basic brain tumor research and clinical interventions.

2. Source of the Paper and Introduction of Authors

This paper, titled “brain tumors induce widespread disruption of calvarial bone and alteration of skull marrow immune landscape,” was published in the November 2025 issue of Nature Neuroscience (Volume 28, Pages 2231–2246), and released online on October 3, 2025. The first and corresponding authors are Abhishek Dubey and Jinan Behnan (contact: jinan.behnan@einsteinmed.edu). The team involves world-renowned research institutions, including Albert Einstein College of Medicine, Karolinska Institutet, Duke University, Osaka University, University of California San Francisco, among others.

3. Study Design and Experimental Process Details

1. Overall Research Framework

This study utilizes two murine glioma models (SB28 and GL261), supplemented by human GBM imaging data, to perform a systematic, multi-level analysis of how GBM impacts the skull’s bone structure and its marrow immune microenvironment. The experimental process consists of several main components:

• Establishment of animal models: Mouse intracranial tumor models were created by inoculation with SB28 and GL261 glioma cells, with controls and various non-tumor brain injury controls (such as stroke, mechanical injury, etc.).
• High-resolution micro-CT (microCT) analysis: Imaging of the skull structure at different tumor progression stages in mice, calculating bone density, thickness, and quantifying marrow channel number and diameter.
• Human GBM imaging analysis: Utilizing public databases like TCIA to collect CT imaging data of GBM patients and age- and gender-matched non-tumor controls, analyzing changes in skull thickness and bone density, as well as correlations between tumor volume and bone change.
• Bone cell dynamics analysis: Using Trap-tdTomato transgenic mice, with whole-bone 3D light-sheet imaging and intravital multiphoton microscopy, quantitatively analyzing osteoclast (OC) distribution, dynamic changes, and aggregation.
• Single-cell RNA sequencing and high-throughput flow cytometry analysis: Isolating skull marrow (SM) and femoral marrow (BM) cells from SB28 and GL261 mouse models, performing single-cell transcriptomics (scRNA-seq) and protein expression profiling to precisely characterize changes in immune cell lineages and functions.
• Drug intervention experiments: Using FDA-approved osteoclast inhibitors (Zoledronic Acid, ZOL, and Rankl antibody), with or without anti-PD-L1 immune checkpoint inhibitors, to observe their effects on tumor progression and immune response.

2. Experimental Design Details

a) Changes in Skull Structure and Bone Marrow Channels

• Subjects: Glioma-bearing mice (SB28 and GL261) at different stages, including both adult and aged groups, sample size 3–8 per group.
• Methods: Key regions such as inner/outer bone plates and suture edges were assessed by 9 μm high-resolution microCT. Key measures included bone volume (BV), trabecular thickness (Tb.Th), trabecular number, and marrow channel numbers/diameters.
• Results: GBM mice displayed bone volume reduction as early as early tumor stages (e.g., significant decrease at day 8 in SB28), and both trabecular thickness and the number/diameter of marrow channels increased notably with tumor progression. These changes were skull-specific and did not occur to a similar extent in the femur. Notably, suture-adjacent and distal occipital bone regions suffered most significant damage, reflecting tumor-induced widespread bony abnormalities.

b) Human Skull Imaging Analysis

• Subjects and Data: 26 GBM patients and 22 age- and gender-matched controls without cancer history from TCIA and other databases, using CT and MRI imaging.
• Methods: Skull thickness at defined anatomical points (e.g., occipital bone, lambda, bregma, etc.) was measured and analyzed for correlation with tumor volume.
• Results: GBM patients had significantly reduced calvarial thickness at lambda and mid-occipital sites. This reduction did not correlate with tumor volume or position, suggesting GBM broadly affects skull structure, not limited only to regions adjacent to the tumor—in line with a systemic effect.

c) Osteoclast (OC) Dynamics

• Subjects: Trap-tdTomato transgenic mice with SB28 and GL261 tumors.
• Methods: Whole bone light-sheet imaging and intravital multiphoton fluorescence microscopy, combined with 3D image analysis algorithms, quantitatively assessed osteoclast numbers, distribution, and aggregation.
• Results: Tumors induced OC loss and abnormal distribution. The SB28 model showed marked loss of osteoclasts, while GL261 had some rebound at late stages. Intercellular distances increased significantly, indicating that OC structure dynamics are closely tied to tumor progression.

d) Immune Microenvironment Lineage Analysis

• Subjects: SM and BM cells from SB28, GL261, and control mice; single-cell sample sizes from thousands to tens of thousands.

• Methods: Optimized bone marrow isolation and single-cell preparation, followed by scRNA-seq and FACS. Data analyses included UMAP for dimensionality reduction, PCA, DEG analysis (using pyDEseq2), and GSEA pathway enrichment.

• Results:

  • The myeloid lineage proportion in SM increased from 41% to 72–90% in GBM, dominated by rising neutrophil fractions, indicating a myelopoiesis bias.
  • Neutrophils, including mature and precursor subtypes, were greatly expanded under tumor conditions, while certain myeloid progenitors and acp5+ macrophages rose in SM but decreased in BM.
  • BM showed distinct gene expression and pathway regulation from SM; tumor-induced upregulated genes in SM were immune-activation, inflammatory, and proliferation related, while BM mostly showed downregulation of inflammatory pathways, indicating contrasting responses to tumor in the two sites.
  • Lymphoid lineage changes were even more dramatic: All B cell subsets in SM were massively lost (decreased by 60–94%), whereas T cells, regulatory T cells, and NKT cells increased, especially in the SB28 model.
  • Cross-model analyses revealed that SB28 caused broader and more intense immune disruption in SM than GL261, echoing clinical differences in GBM subtypes’ immunotherapy responses.

e) Drug Intervention and Immunotherapy Experiments

• Subjects: Mouse models of SB28 and GL261, with intervention and control groups, n=5–10/group.
• Interventions: Osteoclast inhibitors ZOL and Rankl antibody, used alone or in combination with anti-PD-L1 immune checkpoint inhibitors.
• Results: Osteoclast inhibitors robustly prevented bone loss yet paradoxically accelerated tumor progression in SB28 models and diminished the therapeutic effect of anti-PD-L1; anti-PD-L1 monotherapy increased activated tumor-infiltrating T cells and eliminated inflammatory neutrophils, but when combined with osteoclast inhibitors, the pro-inflammatory neutrophil population rebounded and T cell responses were dampened.

4. Key Results and Reasoning

This study is the first to systematically demonstrate that GBM activates extensive skull bone destruction, induces major changes in osteoclast distribution, and increases both the number and diameter of marrow-dura channels, thereby engaging remote immune reservoirs. This regulation was observed in both mouse models and supported by human GBM patient imaging/data. Immune cell profiling revealed that SM and BM, as distinct immune organs, respond differently to brain tumors, especially with myeloid bias and massive B cell depletion in SM determining tumor immunoreactivity and inflammatory state.

The intervention results extend this finding: Pharmacological inhibition of osteoclast function, while blocking bone loss, exacerbates GBM progression (notably in neutrophil-rich mesenchymal subtypes), and directly offsets the benefits of immune checkpoint blockade, offering critical cautionary implications for clinical therapy strategy.

5. Conclusions and Scientific Value

In summary, this study reveals that brain tumors cause widespread bone destruction and immune microenvironment disorder through direct channels between distant marrow reservoirs and the brain surface, presenting a new theory on GBM systemic pathology and immunotherapy response. Skull marrow is not only an “immune reserve” but also a key regulatory node influencing tumor immunotherapy efficacy. This provides new markers and directions for early diagnosis, stratification, and precision immunotherapy of brain tumors.

In practical clinical terms, this study directly explains the mechanistic basis for previously observed reduced survival with osteoporosis treatments (such as ZOL, Denosumab, etc.) in GBM patients, providing important insights for future GBM bone health management and immunotherapy clinical design.

6. Research Highlights and Innovations

• First demonstration of the extensive remodeling effect of GBM on skull bone and marrow immune microenvironment, breaking the previous focus limited to the tumor site.
• Established and validated the comparability and consistency between mouse models and human GBM in terms of bone and immune microenvironment changes.
• Innovatively integrates high-resolution microCT, Trap-tdTomato genetic labeling, multiplex immunoflow cytometry, and single-cell RNA-seq multimodality to systematically depict the dynamic changes in bone structure and marrow immune environment.
• Discovered the counteractive effects of osteoclast inhibitor drugs on the benefits of immunotherapy, sounding a caution on immunotherapy combined with bone drugs in brain tumor patients.

7. Other Important Information

• This study involves 27 world-leading scientists and institutions across neurosurgery, immunology, radiology, and other interdisciplinary fields, reflecting a new fusion trend in neuro-oncology research.
• The appendix methods section details animal management, cell line culture, transgenic mouse phenotyping, etc., providing reproducibility foundation for researchers in relevant fields.

8. Value and Prospects of This Paper

This study introduces the concept of the brain–marrow immune axis into the mechanism and clinical management of brain tumors, emphasizing that both bone health and immune microenvironment should be considered in the comprehensive treatment strategy for brain tumor patients. Further exploration of marrow immune microenvironment modulation and optimization of combined immunotherapy strategies are needed to improve overall prognosis and quality of life for GBM patients.

As an important work published in Nature Neuroscience, this study will have a profound impact on neuroscience, tumor immunology, and translational medicine, and its systematic, multidimensional, and clinically oriented research approach is worthy of broad reference and emulation.