News & Views | Published:


Stem cells and brain tumours

Stem cells are increasingly implicated in maintaining certain cancers. Studies of an intractable type of brain tumour provide hints as to why such cells may underlie the tumours' resistance to therapy.

Cancers are notorious for their ability to survive treatment and recur. Hopes of understanding how they can do so, however, have grown with the prospective identification of rare populations of cancer stem cells in solid tumours1,2. Two studies in this issue3,4 mark a step towards realizing these hopes, and provide further insight into the stem-cell nature of human glioblastoma, an especially nasty type of brain cancer. Both studies build on the identification2 of a tumour-initiating subpopulation of cells that express a cell-surface marker, CD133, that is a hallmark of neural precursor cells.

On page 756, Bao et al.3 show that glioblastoma cells expressing CD133 (CD133+cells) are resistant to ionizing radiation because they are more efficient at inducing the repair of damaged DNA than is the bulk of the tumour cells. Radiation therapy has been the mainstay of glioblastoma treatment for more than 40 years, but although it is transiently effective, it offers no lasting cure. The implication of these results3 is that radiation treatment fails in the long run because it cannot kill the subpopulation of CD133+ tumour-initiating cells.

On page 761, Piccirillo et al.4 describe their work with bone morphogenetic proteins (BMPs), soluble factors that normally induce neural precursor cells to differentiate into mature astrocytes — a subtype of brain cells called glial cells. The authors show that BMPs can also prompt the differentiation of CD133+ brain tumour cells, critically weakening their tumour-forming ability. The results further imply that tumour populations at least partially retain a developmental hierarchy based on stem cells, and remain able to respond to the normal signals that induce them to mature. These findings should lead to renewed interest in devising therapies that promote the differentiation of cancer cells.

Both groups3,4 arrived at their findings by considering the functional hierarchy of the heterogeneous population of tumour cells. In doing so, they add weight to the importance in this research of dissociating solid-tumour samples into single-cell suspensions, purifying the stem-cell fractions, and testing their response to treatment. Crucially, both groups verified their in vitro results with in vivo studies — a true demonstration that human tumour-initiating cells can act as such requires use of the 'gold standard' assay5 of transplanting them into immunodeficient mice to see if they retain their stem-cell capacity.

In Bao and colleagues' work3 (Fig. 1a), ionizing-radiation treatment of glioblastoma cells grown in vitro or as grafts in mice produced an increased fraction of CD133+ cells in the residual tumour population compared with that in unirradiated tumours. These cells retained the ability to reinitiate heterogeneous tumours when transplanted into other mice, demonstrating the retention of stem-cell ability. Although both CD133+ and CD1331 cells sustained a similar amount of DNA damage, activation of DNA-repair responses was greater in CD133+ cells.

Figure 1: Response of glioblastomas to ionizing radiation and bone morphogenetic proteins (BMPs).

Glioblastomas are heterogeneous tumours that contain a few tumour-initiating CD133+ stem cells among other, more differentiated, CD133 cells, including glioblastoma progenitor cells. a, Following radiation, the bulk glioblastoma responds and the tumour shrinks. But CD133+ cells activate checkpoint controls for DNA repair more strongly than CD133 cells, resist radiation and prompt the tumour to regrow. These cells could be targeted with DNA-checkpoint blockers to render them radiosensitive. b, BMPs normally cause neural stem cells to differentiate into astrocytes. When used to treat isolated glioblastoma CD133+ cells, they weaken the cells' tumorigenicity both in vitro and, when engrafted into mice, in vivo. The knowledge that a tumour retains a developmental hierarchy suggests that targeting different cell populations is a promising therapeutic strategy.

Bao et al. went further, to foreshadow how this finding might be translated into therapy. They found that CD133+ cells could be rendered less resistant to radiation if two agents — the checkpoint kinases Chk1 and Chk2, which control pauses in the cell cycle to allow DNA repair to take place — were inhibited pharmacologically. The authors did not look at whether these cells lost the ability to subsequently initiate tumours in vivo. But their work, together with studies showing that leukaemia stem cells are resistant to chemotherapy6,7, means that the idea that cancer stem cells are important because they are resistant to therapy now carries greater weight.

In their research, Piccirillo and colleagues4 first showed that human glioblastoma cells express BMPs and their cell-surface receptors. Using culture conditions that support the growth of undifferentiated glioblastoma cells, they found that BMP treatment reduced cell proliferation and induced differentiation predominantly into cells resembling mature astrocytes. Treatment of cultured glioblastoma progenitor cells or CD133+ glioblastoma cells in vitro with BMP, or co-treatment of glioblastoma cells transplanted with beads soaked in BMP, reduced the size of the tumours grafted into mice and prolonged the animals' survival (Fig. 1b). The BMP-treated tumour cells engrafted into mice were more mature and less invasive. CD133+ cells could not be recovered from these small tumours, and such tumour populations were not capable of serial engraftment.

Although these results4 are remarkable in showing that a differentiation-promoting agent is a potential treatment for brain tumours, some mice still developed tumours and died 3 months after BMP treatment. So it seems that certain cancer cells — presumably cancer stem cells — still escape BMP treatment. Will these cancer stem cells induce recurrence at a longer latency?

In both cases3,4, definitive demonstration of a specific effect of treatment on glioblastoma stem cells will require improved purification of the tumours — the true stem cells are probably a subpopulation of the CD133+ fraction. But both studies add depth to the cancer-stem-cell hypothesis and illustrate the potential of re-examining cancer in the light of that hypothesis. Given the identification of CD133+ tumour-initiating cells from human colon cancer, published last month8,9, it would seem that these approaches are ripe for testing in other human cancers.


  1. 1

    Al-Hajj, M. et al. Proc. Natl Acad. Sci. USA 100, 3983–3988 (2003).

  2. 2

    Singh, S. K. et al. Nature 432, 396–401 (2004).

  3. 3

    Bao, S. et al. Nature 444, 756–760 (2006).

  4. 4

    Piccirillo, S. G. M. et al. Nature 444, 761–765 (2006).

  5. 5

    Bonnet, D. & Dick, J. E. Nature Med. 3, 730–737 (1997).

  6. 6

    Graham, S. M. et al. Blood 99, 319–325 (2002).

  7. 7

    Guzman, M. L. et al. Blood 105, 4163–4169 (2005).

  8. 8

    O'Brien, C. A., Pollett, A., Gallinger, S. & Dick, J. E. Nature doi:10.1038/nature05372 (2006).

  9. 9

    Ricci-Vitiani, L. et al. Nature doi:10.1038/nature05384 (2006).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

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.