The concept of cancer stem cells – specific subsets of tumour cells that share the regenerative properties of normal stem cells – has gained currency in recent years, but there is still controversy over the prevalence of these cells, particularly in solid tumours.

Two independent studies published in the same issue of Nature have now definitively identified subsets of cells in mouse models of solid tumours that are responsible for the re-growth of those cancers.

The first study, by Luis Parada and co-workers from the University of Texas, Dallas, Texas [1] examined a mouse model of glioblastoma multiforme, which is the most common and deadly form of brain tumour in adults with a five-year survival rate of less than 6%.

Many glioblastomas initially respond well to primary therapy with temozolomide, but resistance often develops quickly. The mechanisms of drug resistance and tumour recurrence are not yet fully known.

Parada and his colleagues studied a transgenic mouse model of glioblastoma in which all mice spontaneously develop glioblastomas deriving from the subventricular zone of the brain.

Assuming that these tumours develop from normal neural stem cells (NSCs), the researchers tested whether a fluorescent nestin transgene, developed as a marker of NSCs, would also mark glioblastoma cells.

This transgene also contained a modified herpes simplex tyrosine kinase gene that enabled dividing cells expressing the construct to be readily killed by administration of ganciclovir.

The researchers discovered that this transgene labelled a subpopulation of relatively quiescent tumour cells in these mice in much the same way as it labels quiescent neural stem cells. When the mice were treated with temozolomide, the drug initially killed proliferating tumour cells but tumour recurrence and re-growth occurred rapidly, and the subset of quiescent tumour cells that were labelled with the transgene (GFP⁺ cells) was not affected by temozolomide.

These results suggested that the GFP⁺ cells showed stem-like properties that might indicate that they were the source of tumour re-growth. To test this, Parada and his colleagues treated the transgenic mice with ganciclovir to remove the labelled cells while the tumours were still very small, which showed a significant survival advantage for the ganciclovir-treated mice.

Similar results were obtained if ganciclovir was administered only when the tumours were larger. Simultaneous treatment with temozolomide to remove the proliferating cells and ganciclovir to remove the quiescent, stem-like cells led to a highly significant reduction in tumour mass.

In the second study, Cédric Blanpain from the Université Libre de Bruxelles, Brussels, Belgium and his colleagues investigated the fate of individual cancer cells during early tumour development in a mouse model of the skin cancer, squamous cell carcinoma.  The mice were treated with chemical carcinogens and mitogens to induce the formation of benign, slow-growing skin papillomas, a fraction of which develop into invasive tumours. 

The researchers used an inducible genetic lineage tracing system to label about 1% of papilloma cells with yellow fluorescent protein (YFP); labelling at this density enabled individual cells to be tracked over time. Labelled cells were found to be able to generate all the cell types found in the tumours over time.

Most interestingly, the researchers noted that when the fate of the tumour cells was followed for seven weeks after labelling, a subset of the labelled cells proliferated to form large clones, some of which occupied a whole lobe of the tumour. Other cells formed much smaller clones or remained quiescent. Quantitative analysis of the growth, size and persistence of the clones suggested that only about 20% were still self-renewing after seven weeks.

Blanpain and his colleagues suggested from their observation and modeling of the fate of individual cells that papilloma proliferation is driven from a sub-population of stem-like cells. Each stem-like cell undergoes two cell divisions per day and each division of a stem-like cell could lead to two dividing cells, two differentiating cells or one of each. The invasive squamous cell carcinomas in the mouse model were also found to contain stem-like cells, but, in contrast to the papilloma cells, these cells undergo clonal expansion without generating differentiating cells. 

Taken together, the results of these studies have shown that, at least in mouse models, glioblastoma multiforme and squamous cell carcinoma – two very different solid tumours – both contain a subset of proliferating and rapidly dividing cells with properties that are very similar to those of normal stem cells.

References

[1]: Chen, J., Li, Y. Yu, T-S., McKay, R.M., Burns, D.K., Kernie, S.G. and Parada, L.F. (2012). A restricted cell population propagates glioblastoma growth after chemotherapy. Nature, published online ahead of print 2 August 2012. doi:10.1038/nature11287

[2]: Driessens, G., Beck, B., Caauwe, A., Simons, B.D. and Blanpain, C. (2012). Defining the mode of tumour growth by clonal analysis. Nature, published online ahead of print 2 August 2012. doi:10.1038/nature11344