Early invasive growth along specific anatomical structures, especially the white matter tract, is regarded as one of the main causes of poor therapeutic outcome of people with gliomas. We show that some glioma stem cells (GSCs) are preferentially located along white matter tracts, which exhibit a demyelinated phenotype, at the invasive frontier of glioma tissues. These GSCs are CD133+Notch1+, whereas the nerve fibers express the Notch ligand Jagged1. The Notch-induced transcription factor Sox9 promotes the transcription of SOX2 and the methylation level of the NOTCH1 promoter is attenuated by the upregulation of SOX2 to reinforce NOTCH1 expression in GSCs. This positive-feedback loop in a cohort of glioma subjects is correlated with a poor prognosis. Inhibition of Notch signaling attenuates the white-matter-tract tropism of GSCs. These findings provide evidence indicating that the NOTCH1-SOX2 positive-feedback loop controls GSC invasion along white matter tracts.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Data Availability

All routine analysis methods are included in Methods. The data that support the findings of this study are presented in the paper and supplementary materials and all raw data are available from the corresponding author upon reasonable request. The five published mRNA expression data sets (Glioblastoma-TCGA-540; Glioma-French-284, GSE16011; Glioblastoma-TCGA-395; Glioblastoma Stemcells-20, GSE15209 and GSE67089) used in this study were obtained from R2, an Affymetrix analysis and visualization platform developed at the Department of Human Genetics at the Academic Medical Center, University of Amsterdam.

Additional information

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


  1. 1.

    Chen, J., McKay, R. M. & Parada, L. F. Malignant glioma: lessons from genomics, mouse models, and stem cells. Cell 149, 36–47 (2012).

  2. 2.

    Preusser, M., Haberler, C. & Hainfellner, J. A. Malignant glioma: neuropathology and neurobiology. Wien. Med. Wochenschr. 156, 332–337 (2006).

  3. 3.

    Huse, J. T. & Holland, E. C. Targeting brain cancer: advances in the molecular pathology of malignant glioma and medulloblastoma. Nat. Rev. Cancer 10, 319–331 (2010).

  4. 4.

    Cuddapah, V. A., Robel, S., Watkins, S. & Sontheimer, H. A neurocentric perspective on glioma invasion. Nat. Rev. Neurosci. 15, 455–465 (2014).

  5. 5.

    Calabrese, C. et al. A perivascular niche for brain tumor stem cells. Cancer Cell 11, 69–82 (2007).

  6. 6.

    Pietras, A. et al. Osteopontin-CD44 signaling in the glioma perivascular niche enhances cancer stem cell phenotypes and promotes aggressive tumor growth. Cell Stem Cell 14, 357–369 (2014).

  7. 7.

    Bao, S. et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 444, 756–760 (2006).

  8. 8.

    Singh, S. K. et al. Identification of a cancer stem cell in human brain tumors. Cancer Res. 63, 5821–5828 (2003).

  9. 9.

    Stangeland, B. et al. Combined expressional analysis, bioinformatics and targeted proteomics identify new potential therapeutic targets in glioblastoma stem cells. Oncotarget 6, 26192–26215 (2015).

  10. 10.

    Brown, D. V. et al. Coexpression analysis of CD133 and CD44 identifies proneural and mesenchymal subtypes of glioblastoma multiforme. Oncotarget 6, 6267–6280 (2015).

  11. 11.

    Lu, J. et al. Endothelial cells promote the colorectal cancer stem cell phenotype through a soluble form of Jagged-1. Cancer Cell 23, 171–185 (2013).

  12. 12.

    Stidworthy, M. F. et al. Notch1 and Jagged1 are expressed after CNS demyelination, but are not a major rate-determining factor during remyelination. Brain 127, 1928–1941 (2004).

  13. 13.

    Wang, S. et al. Notch receptor activation inhibits oligodendrocyte differentiation. Neuron 21, 63–75 (1998).

  14. 14.

    Patel, A. P. et al. Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science 344, 1396–1401 (2014).

  15. 15.

    Gangemi, R. M. et al. SOX2 silencing in glioblastoma tumor-initiating cells causes stop of proliferation and loss of tumorigenicity. Stem Cells 27, 40–48 (2009).

  16. 16.

    Ehm, O. et al. RBPJkappa-dependent signaling is essential for long-term maintenance of neural stem cells in the adult hippocampus. J. Neurosci. 30, 13794–13807 (2010).

  17. 17.

    Candy, P. A. et al. Notch-induced transcription factors are predictive of survival and 5-fluorouracil response in colorectal cancer patients. Br. J. Cancer 109, 1023–1030 (2013).

  18. 18.

    Taranova, O. V. et al. SOX2 is a dose-dependent regulator of retinal neural progenitor competence. Genes Dev. 20, 1187–1202 (2006).

  19. 19.

    Terragni, J. et al. Notch signaling genes: myogenic DNA hypomethylation and 5-hydroxymethylcytosine. Epigenetics 9, 842–850 (2014).

  20. 20.

    Bulstrode, H. et al. Elevated FOXG1 and SOX2 in glioblastoma enforces neural stem cell identity through transcriptional control of cell cycle and epigenetic regulators. Genes Dev. 31, 757–773 (2017).

  21. 21.

    Teodorczyk, M. & Schmidt, M. H. Notching on cancer’s door: Notch signaling in brain tumors. Front. Oncol. 4, 341 (2014).

  22. 22.

    Krop, I. et al. Phase I pharmacologic and pharmacodynamic study of the gamma secretase (Notch) inhibitor MK-0752 in adult patients with advanced solid tumors. J. Clin. Oncol. 30, 2307–2313 (2012).

  23. 23.

    Darsalia, V. et al. Cell number and timing of transplantation determine survival of human neural stem cell grafts in stroke-damaged rat brain. J. Cereb. Blood Flow Metab. 31, 235–242 (2011).

  24. 24.

    Aharonowiz, M. et al. Neuroprotective effect of transplanted human embryonic stem cell-derived neural precursors in an animal model of multiple sclerosis. PLoS One 3, e3145 (2008).

  25. 25.

    Giese, A., Bjerkvig, R., Berens, M. E. & Westphal, M. Cost of migration: invasion of malignant gliomas and implications for treatment. J. Clin. Oncol. 21, 1624–1636 (2003).

  26. 26.

    Gritsenko, P. G., Ilina, O. & Friedl, P. Interstitial guidance of cancer invasion. J. Pathol. 226, 185–199 (2012).

  27. 27.

    Iwadate, Y., Fukuda, K., Matsutani, T. & Saeki, N. Intrinsic protective mechanisms of the neuron-glia network against glioma invasion. J. Clin. Neurosci. 26, 19–25 (2016).

  28. 28.

    Belien, A. T., Paganetti, P. A. & Schwab, M. E. Membrane-type 1 matrix metalloprotease (MT1-MMP) enables invasive migration of glioma cells in central nervous system white matter. J. Cell Biol. 144, 373–384 (1999).

  29. 29.

    Beadle, C. et al. The role of myosin II in glioma invasion of the brain. Mol. Biol. Cell 19, 3357–3368 (2008).

  30. 30.

    Yeo, S. Y. & Chitnis, A. B. Jagged-mediated Notch signaling maintains proliferating neural progenitors and regulates cell diversity in the ventral spinal cord. Proc. Natl Acad. Sci. U.S.A. 104, 5913–5918 (2007).

  31. 31.

    Corn, B. W. et al. White matter changes are correlated significantly with radiation dose. Observations from a randomized dose-escalation trial for malignant glioma (Radiation Therapy Oncology Group 83-02). Cancer 74, 2828–2835 (1994).

  32. 32.

    Milic, M. & Rees, J. H. Acute demyelination following radiotherapy for glioma: a cautionary tale. Pract. Neurol. 17, 35–38 (2017).

  33. 33.

    Esiri, M. M. The interplay between inflammation and neurodegeneration in CNS disease. J. Neuroimmunol. 184, 4–16 (2007).

  34. 34.

    Dos Santos, C. D., Picanco-Diniz, C. W. & Gomes-Leal, W. Differential patterns of inflammatory response, axonal damage and myelin impairment following excitotoxic or ischemic damage to the trigeminal spinal nucleus of adult rats. Brain Res. 1172, 130–144 (2007).

  35. 35.

    Sinha, S., Bastin, M. E., Whittle, I. R. & Wardlaw, J. M. Diffusion tensor MR imaging of high-grade cerebral gliomas. Am. J. Neuroradiol. 23, 520–527 (2002).

  36. 36.

    Yen, P. S. et al. White matter tract involvement in brain tumors: a diffusion tensor imaging analysis. Surg. Neurol. 72, 464–469 (2009).

  37. 37.

    Givogri, M. I. et al. Central nervous system myelination in mice with deficient expression of Notch1 receptor. J. Neurosci. Res. 67, 309–320 (2002).

  38. 38.

    Bonini, S. A. et al. Nuclear factor κB-dependent neurite remodeling is mediated by Notch pathway. J. Neurosci. 31, 11697–11705 (2011).

  39. 39.

    Morokoff, A., Ng, W., Gogos, A. & Kaye, A. H. Molecular subtypes, stem cells and heterogeneity: Implications for personalised therapy in glioma. J. Clin. Neurosci. 22, 1219–1226 (2015).

  40. 40.

    Verhaak, R. G. et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell 17, 98–110 (2010).

  41. 41.

    Venkatesh, H. S. et al. Neuronal activity promotes glioma growth through neuroligin-3 secretion. Cell 161, 803–816 (2015).

  42. 42.

    Venkatesh, H. S. et al. Targeting neuronal activity-regulated neuroligin-3 dependency in high-grade glioma. Nature 549, 533–537 (2017).

  43. 43.

    Lathia, J. D. et al. Integrin α6 regulates glioblastoma stem cells. Cell Stem Cell 6, 421–432 (2010).

  44. 44.

    Haas, T. L. et al. Integrin α7 is a functional marker and potential therapeutic target in glioblastoma. Cell Stem Cell 21, 35–50.e9 (2017).

  45. 45.

    Zhu, X., Tollkuhn, J., Taylor, H. & Rosenfeld, M. G. Notch-dependent pituitary SOX2(+) stem cells exhibit a timed functional extinction in regulation of the postnatal gland. Stem Cell Rep. 5, 1196–1209 (2015).

  46. 46.

    Phi, J. H. et al. Upregulation of SOX2, NOTCH1, and ID1 in supratentorial primitive neuroectodermal tumors: a distinct differentiation pattern from that of medulloblastomas. J. Neurosurg. Pediatr. 5, 608–614 (2010).

  47. 47.

    Yin, J. et al. Pigment epithelium-derived factor (PEDF) expression induced by EGFRvIII promotes self-renewal and tumor progression of glioma stem cells. PLoS Biol. 13, e1002152 (2015).

  48. 48.

    Garros-Regulez, L. et al. mTOR inhibition decreases SOX2-SOX9 mediated glioma stem cell activity and temozolomide resistance. Expert Opin. Ther. Targets 20, 393–405 (2016).

  49. 49.

    Yu, S. C. et al. Isolation and characterization of cancer stem cells from a human glioblastoma cell line U87. Cancer Lett. 265, 124–134 (2008).

  50. 50.

    Zhou, W. et al. Periostin secreted by glioblastoma stem cells recruits M2 tumour-associated macrophages and promotes malignant growth. Nat. Cell Biol. 17, 170–182 (2015).

Download references


We thank C.-J. Zhao, S.-M. Yang and X.-D. Liu for their constructive suggestions. This study was supported by grants from the National Key Research and Development Program of China (2016YFA0202104, 2016YFA0101200), the National Natural Science Foundation of China (81570131, 81572880, 61327902), and the Key Clinical Research Program of Southwest Hospital, Army Medical University (Third Military Medical University) of China (SWH2016ZDCX1005, SWH2017ZDCX1003).

Author information

Author notes

  1. These authors contributed equally: Jun Wang, Sen-Lin Xu, Jiang-Jie Duan, Liang Yi.


  1. Institute of Pathology and Southwest Cancer Center, Key Laboratory of the Ministry of Education, Southwest Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Jun Wang
    • , Sen-Lin Xu
    • , Jiang-Jie Duan
    • , Yu-Feng Guo
    • , Yu Shi
    • , Lin Li
    • , Ze-Yu Yang
    • , Xue-Mei Liao
    • , Jiao Cai
    • , Li Yin
    • , Xia Zhang
    • , Xiu-Wu Bian
    •  & Shi-Cang Yu
  2. Department of Stem Cell and Regenerative Medicine, Southwest Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Jun Wang
    • , Jiang-Jie Duan
    • , Yu-Feng Guo
    • , Lin Li
    • , Ze-Yu Yang
    • , Xue-Mei Liao
    • , Jiao Cai
    • , Li Yin
    •  & Shi-Cang Yu
  3. Department of Neurosurgery, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Liang Yi
  4. Department of Medical Statistics, Military Preventive Medicine Academy, Army Medical University (Third Military Medical University), Chongqing, China

    • Yan-Qi Zhang
  5. Department of Pathology, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Hua-Liang Xiao
  6. Department of Radiology, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Hao Wu
  7. Department of Medical Imaging, College of Biomedical Engineering, Army Medical University (Third Military Medical University), Chongqing, China

    • Jing-Na Zhang
  8. Department of Neurosurgery, Xinqiao Hospital, Army Medical University (Third Military Medical University), Chongqing, China

    • Sheng-Qing Lv
  9. Institute of Cancer Stem Cell, Cancer Center, Second Affiliated Hospital, Dalian Medical University, Dalian, China

    • Qing-Kai Yang
  10. Center for Neuroscience, Medical College, Shantou University, Shantou, China

    • Xiao-Jun Yang
  11. Department of Neurosurgery, TianTan Hospital, Capital Medical University, Beijing, China

    • Tao Jiang


  1. Search for Jun Wang in:

  2. Search for Sen-Lin Xu in:

  3. Search for Jiang-Jie Duan in:

  4. Search for Liang Yi in:

  5. Search for Yu-Feng Guo in:

  6. Search for Yu Shi in:

  7. Search for Lin Li in:

  8. Search for Ze-Yu Yang in:

  9. Search for Xue-Mei Liao in:

  10. Search for Jiao Cai in:

  11. Search for Yan-Qi Zhang in:

  12. Search for Hua-Liang Xiao in:

  13. Search for Li Yin in:

  14. Search for Hao Wu in:

  15. Search for Jing-Na Zhang in:

  16. Search for Sheng-Qing Lv in:

  17. Search for Qing-Kai Yang in:

  18. Search for Xiao-Jun Yang in:

  19. Search for Tao Jiang in:

  20. Search for Xia Zhang in:

  21. Search for Xiu-Wu Bian in:

  22. Search for Shi-Cang Yu in:


J.W. acquired, analyzed and interpreted data and drafted the manuscript. S.L.X., J.J.D. and L.Y. acquired, analyzed and interpreted data. Y.F.G., Y.S., L.L., Z.Y.Y., X.M.L., J.C., H.L.X. and L.Y. acquired data. Y.Q.Z. performed statistical analysis. H.W., J.N.Z., S.Q.L., Q.K.Y., X.J.Y., T.J. and X.Z. analyzed and interpreted data. X.W.B. analyzed and interpreted data, critically revised the manuscript for intellectual content, obtained funding and supervised the study. S.C.Y. initiated the study concept and design, acquired, analyzed and interpreted data, drafted of the manuscript, critically revised the manuscript for intellectual content, performed statistical analysis, obtained funding, and supervised the study. All authors read and approved the manuscript for publication.

Competing interests

The authors declare no competing interests.

Corresponding authors

Correspondence to Xiu-Wu Bian or Shi-Cang Yu.

Integrated supplementary information

  1. Supplementary Figure. 1 Spatial relationship between GSCs and nerve fibers.

    (a) Schematic diagram of the spatial distribution analysis. (b-c) Spatial relationship between GSCs (PBK+, CDC6+, and CD44+, red). and nerve fibers (NF200+, green) at the invasive frontier of glioma patients as detected by immunofluorescence staining. Three. samples, 3 male, aged from 46 to 57, 3 sections/sample, 10 high-power fields/section. Representative laser confocal microscopy. images of nuclei stained with DAPI (blue) are shown. Scale bar=20 μm (b). Quantitative analysis of the spatial orientation of every GSC. (CDC6, n=445 independent positive cells; PBK, n=426 independent positive cells; CD44, n=450 independent positive cells) to the. nearest nerve fibers (c)

  2. Supplementary Figure 2 Migration, adherence, and invasion of GSCs in vitro and in vivo.

    (a) Schematic diagram of the migration assay in vitro. (b) Quantitative analysis of GSC (CD133+, derived from U87, GBM1,and GBM2). migration under high-power field light microscopy showing representative results from 4 independent expriments, the data shown are. the mean±SD. A two-sided Student’s t test was used to generate P values (CS vs. CM. U87-GSC: n=10 independent high-power fields, df=18, t=0.44; GBM1-GSC: n=10 independent high-power fields, df=18, t=0.42; GBM1-GSC: n=10 independent high-power fields, df=18, t=0.79;). CS, culture supernatants; CM, culture media. (c) Adherence of GSCs (CD133+) and non-GSCs (CD133-) to the neurites of neurons cultured in vitro. Representative results from 4 independent expriments, the data shown are the mean±SD. A twosided Student’s t test was used to generate P values (CD133- vs. CD133+, n=25 independent high-power fields, df=48, t=-13). (d, e). The adherence of GSCs (CD133+) to the neurites of neurons cultured in vitro after treatment of Decoy (d) and Notch1-siRNA (e). Representative results from 4 independent exprements, the data shown are the mean±SD. A two-sided Student’s t test was used to. generate P values (PBS vs. Decoy, n=30 independent high-power fields, df=58, t=16; Mock vs. NOTCH1-siRNA, n=30 independent. high-power fields, df=58, t=11). (f) Immunofluorescent staining of CD44 (green) in xenografts of the control group to reveal the invasive. growth of glioma cells (red) along nerve fibers (blue). Two weeks after orthotopic implantation of 1×105 mCherry-labeled GL261 GSCs. into the right striatum adjacent to the corpus callosum of Thy1-EGFP transgenic mice. Five mice per group, 3 sections per mouse, 3-6. high-power fields per section, n=54 high-power fields. Representative laser confocal microscopy images of nuclei stained with DAPI. (gray) are shown. Scale bar=20 μm. (G) 3D diagram of the orthotopic xenograft. View of the horizontal plane and view angle with an. 80º horizontal plane. 3D diagram of mouse brain derived from Brain explore 2 (Allen Institute for Brain Science). The red circle denotes. the site of implantation

  3. Supplementary Figure. 3 Relationship between SOX2 and Notch receptors (NOTCH1–4) and transcription factors (SOX9, HEY1, HEY2, and HES1) expression.

    The data were derived from 3 Affymetrix DNA microarray datasets (Tumor Glioblastoma - TCGA - 540, n= 540 biologically independent. patients; Tumor Glioma - French - 284, n= 284 biologically independent patients; and Tumor Glioblastoma - TCGA – 395, n= 395. biologically independent patients). Pearson correlation analysis (2-tailed) was used to generate P values and correlation coefficients (r)

  4. Supplementary Figure 4 The mechanism underlying NOTCH1 upregulation of the transcription of SOX2.

    (a) Luciferase reporter assay of the 5’-promoter of SOX2. NOTCH1OE, GBM2 non-GSC cells stably overexpressing NOTCH1 via. pcDNA 3.1-NOTCH1; Vector, GBM2 non-GSC cells stably transfected with pcDNA3.1 empty vector; n=3 biologically independent. samples, the data are the mean±SD. (b) The similarity of the Sox2 promoter region between mouse and human. (c) The putative. binding sites of RBP-Jκ located at the promoter of the human SOX2 gene. (d) Chromatin. immunoprecipitation (ChIP) analysis in GSCs. derived from GBM2 cells. PCR primers were designed to surround predicted RBP-Jκ-binding sites 1 and 2 on the SOX2 promoter. Nonspecific IgG and anti-RNA polymerase II were used as controls. Three times these experiments were repeated independently with similar results. m, DNA ladder

  5. Supplementary Figure 5 Immunofluorescent staining of Notch1, Sox9, and Sox2 in GSCs derived from GBM 1-3.

    Representative lase confocal microscopy images of nuclei stained with hoechst33342 are shown. Scale bar=20 μm. (A) Double. immunofluorescent staining of Notch1 (red), Sox9 (red or green), and Sox2 (red or green) in GSCs derived from GBM 2 and GBM 3 cell. lines. Right panel: expanded view of the inset regions (white box) shown in the left panel. Inset1-3, scale bar=10 μm; Inset4-6, scale. bar=2 μm._Three times these experiments were repeated independently with similar results. (B) Triple immunofluorescent staining of. Notch1 (green), Sox9 (blue), and Sox2 (red) in GSCs derived from GBM 1 and GBM3. Scale bar=20 μm. Three times these. experiments were repeated independently with similar results

  6. Supplementary Figure 6

    Schematic diagram of orthotopic xenografts

  7. Supplementary Figure 7 Luciferase reporter assays of the enhancers of NOTCH1.

    (A) The putative enhancer binding sites of Sox2 located in the introns and exons of the NOTCH1 gene. (B) Luciferase activity of GBM2. non-GSC cells after different fragments containing the putative enhancers were cloned into a minimal promoter of the pGL-4 luciferase. vector and then co-transfected with pcDNA 3.1-SOX2, n=3 biologically independent samples. Center line represents the median, lower. and upper limits of the box represent the 25th and 75th percentiles, and whiskers show maximum and minimum

  8. Supplementary Figure 8 Schematic diagram of the reciprocal activation between SOX2 and Notch signaling determining the WM tract tropism of GSCs.

    CD133 and Notch1 double-positive GSCs were preferentially located along the Notch ligand Jagged1 expressed WM tracts, which. exhibited a demyelinated phenotype, at the invasive frontier of glioma tissues. The Notch-induced transcription factor Sox9 promoted. the transcription of SOX2 and the methylation level of the NOTCH1 promoter was attenuated by the upregulation of SOX2 to reinforce. NOTCH1 expression in GSCs

  9. Supplementary Figure 9 Distribution of Sox9+ cells in different regions of human glioma samples.

    Seven samples, 3 sections per sample, 1-20 high-power fields per section. Upper panel: Representative image of IHC staining. Dottedline. boxes (1-3) indicate the center, invasive frontier, and distant site of the tumor mass. Scale bar=200 μm. Lower panel: expanded. view of the inset regions (red box) shown in the upper panel. Scale bar=50 μm

  10. Supplementary Figure 10 Full-length western blots.

    Full-length western blots for cropped images in Figs. 4h,j, 4p, 6b,d, and 7f

Supplementary information

  1. Supplementary Figures 1–10

    Supplementary Figures 1–10, Supplementary Tables 1–6, and Supplementary Note

  2. Reporting Summary

About this article

Publication history