Drain the swamp to beat glioma

Efforts to treat brain tumours by targeting cancer cells have had only modest clinical success. It emerges that targeting a protein secreted from neurons adjacent to the tumour might be a useful alternative approach. See Letter p.533

Brain tumours known as malignant gliomas are almost always lethal. Indeed, this cancer is the leading cause of brain-tumour-related death in children and adults1. Despite tremendous advances in understanding the cell biology and genomics underlying glioma development and progression, patient survival rates have remained stagnant, and few effective therapies exist. Venkatesh et al.2 reveal on page 533 that a protein secreted from non-malignant brain cells in the microenvironment surrounding the tumour can be therapeutically targeted. Such intervention might offer a fundamentally new approach to glioma treatment.

Standard therapy for malignant glioma often combines surgical tumour removal, radiation treatment and chemotherapy. Since the start of the twentieth century, a meagre three therapies have been approved for the treatment of malignant gliomas3,4,5. Dozens of clinical trials testing targeted therapies for gliomas have failed over the past two decades, despite being based on sound scientific rationales. The development of new therapies is hampered by the diversity of glioma cells in different patients, as well as the paucity of easily targetable mutations in most such cells.

Current understanding about the interaction between glioma cells and their microenvironment in the central nervous system is limited, particularly with regard to the activity of their neighbouring neurons. Although it is known that the electrical activity of neurons in the brain shapes normal brain development, many questions remain about how the proliferation and growth of glioma cells are governed by electrical stimuli in neurons and neuronal activity.

Gliomas arise from non-neuronal, glial cells. Venkatesh and colleagues have previously shown6,7 that neuronal activity promotes the growth of glial cells in vivo in mice6 and, through a series of in vitro and in vivo experiments, demonstrated that, in response to neuronal activity, neurons secrete a portion of the protein Nlgn3 that promotes glioma growth and proliferation7.

Venkatesh et al. began the present study by investigating the signalling pathways mediating the glioma-supporting effects of secreted Nlgn3 in mice. They found that a key event is the phosphorylation of the kinase enzyme FAK in glioma cells, a modification that results in the activation of downstream signalling targets that include the proteins SRC, PI3K, mTOR and MAPK.

To try to translate their observations of this system into possible therapeutic options, Venkatesh and colleagues used mice containing mutations in certain genes and pharmacological tools to disrupt the signalling between the microenvironment and the cancerous cells. They also transplanted samples of human gliomas into immunodeficient mice that do not reject human tissue, a technique known as xenografting.

Consistent with their hypothesis that secreted Nlgn3 mediates microenvironmental support driven by neuronal activity, the authors found that these transplanted glioma cells can grow in the brains of mice that express wild-type Nlgn3, but not in the brains of mice lacking Nlgn3. This Nlgn3-dependent difference in tumour growth was less prominent in the initial establishment of the tumour, and more pronounced in later stages of growth. However, when the authors transplanted samples of human breast cancer that had migrated to grow in the brain, these tumour cells did not exhibit Nlgn3-dependent growth differences, suggesting that the role of Nlgn3 in the microenvironment is specific to glioma, rather than a general mechanism that affects any type of tumour growing in the brain.

Nlgn3 is expressed on the surface of non-malignant cells surrounding the tumour, and is secreted and cleaved in response to neuronal electrical activity7. Venkatesh and colleagues identified the enzyme responsible for this cleavage by searching for enzymes that cleave proteins at amino acid sites that match the cleavage pattern observed for Nlgn3. The enzyme Adam10 has the expected cleavage capacity, and when the authors tested its role using drugs that inhibit it, the secretion of Nlgn3 from neurons grown in vitro was diminished or prevented. Pharmacological inhibition of Adam10 had no direct effect on the in vitro growth of a cell population consisting solely of malignant glioma cells, but administering Adam10 inhibitors to mice containing tumour xenografts inhibited the growth and progression of their cancers. This finding is consistent with a model in which a positive therapeutic outcome can result from interrupting an interaction between malignant glioma and its microenvironment.

Inhibitors of the human ADAM family of enzymes have been used in clinical trials to prevent cleavage of other protein targets of these enzymes that have been linked to cancers such as lymphoma8 and breast cancer9. These inhibitors could potentially be repurposed to treat individuals with malignant glioma. This approach would not have been immediately obvious from results using existing tools that identify potential biological targets for therapy, such as in vitro high-throughput drug screening. These tools are normally used to test whether drugs can limit the growth of cancer cells in vitro rather than whether they can inhibit interactions between cancer cells and their non-cancerous neighbours. However, some malignant gliomas treated with ADAM10 inhibitors in vivo in this study in mice showed a reduction in tumour growth, but not complete tumour destruction, suggesting that this type of therapy will probably need to be combined with other approaches.

In considering the unique physiology of the brain microenvironment, Venkatesh and colleagues offer a new approach to treating malignant gliomas by focusing on the surrounding cells. Most of the current targeted therapies for malignant gliomas are applicable to only a small subset of patients, but the idea of interrupting the molecular support that the glioma receives from its normal microenvironment raises the possibility of targeting a diverse range of glioma subtypes and patient ages. Rather than approaching the daunting glioma 'monster' head-on, clinicians might instead consider an indirect approach of 'draining the swamp' around the malignant cells (Fig. 1).

Figure 1: Targeting brain tumours by focusing on their microenvironment.

Venkatesh et al.2 investigated brain cancer using mouse models (not shown) in which they transplanted cells from a type of tumour called a glioma, which is derived from non-neuronal brain cells. The growth of glioma cells is promoted by a cleavage product (yellow) of a protein called Nlgn3, which is expressed by neurons7. Venkatesh et al. identified Adam10 as the protease enzyme that cleaves Nlgn3, and showed that treatment with small-molecule inhibitors of Adam10 decreased glioma growth in their model system. Other approaches exist for targeting factors in tumour microenvironments that support cancer growth; these include targeting tumour blood-vessel formation by inhibiting the protein VEGF, or using antibody treatments to prevent cancer-targeting immune cells from being suppressed. Future treatment options for glioma might include a combination of such approaches to combat the tumour indirectly.

Such an environmental-blockade strategy might, in addition to the approach outlined by Venkatesh and colleagues, also include the inhibition of other microenvironmental factors that promote tumour growth. For example, combination therapies might also target the immunosuppressive environment of tumour cells that blunts the body's immune response targeting the tumour, or they might inhibit factors that aid the formation of tumour blood vessels. Such combinations could lead to improved outcomes that are urgently needed for these patients.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Ostrom, Q. T. et al. Neuro-Oncology 18, Suppl. 5 (2016).

  2. 2

    Venkatesh, H. S. et al. Nature 549, 533–537 (2017).

    ADS  Article  Google Scholar 

  3. 3

    Stupp, R. et al. J. Am. Med. Assoc. 314, 2535–2543 (2015).

    CAS  Article  Google Scholar 

  4. 4

    Stupp, R. et al. N. Engl. J. Med. 352, 987–996 (2005).

    CAS  Article  Google Scholar 

  5. 5

    Vredenburgh, J. J. et al. J. Clin. Oncol. 25, 4722–4729 (2007).

    CAS  Article  Google Scholar 

  6. 6

    Gibson, E. M. et al. Science 344, 1252304 (2014).

    Article  Google Scholar 

  7. 7

    Venkatesh, H. S. et al. Cell 161, 803–816 (2015).

    CAS  Article  Google Scholar 

  8. 8

  9. 9

    Witters, L. et al. Cancer Res. 68, 7083–7089 (2008).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Michael D. Taylor.

Related links

Related links

Related links in Nature Research

Cancer: Keeping it real to kill glioblastoma

Cancer: A precision approach to tumour treatment

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Taylor, M., Ramaswamy, V. Drain the swamp to beat glioma. Nature 549, 460–461 (2017).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.