The traditional model of a tumour as a homogenous mass of abnormal cells has now largely been replaced by the view that tumours contain a subset of cells with stem cell-like properties. These cancer stem cells share with normal stem cells the properties of self-renewal and of generating progeny that can differentiate into a variety of cell types. They also divide more slowly than other tumour cells. Therefore, drugs that are active against "normal" tumour tissue – particularly those that target rapidly dividing cells – may not be active against cancer stem cells. As tumours may re-grow from a few stem cells left behind after apparently successful chemotherapy, it will be important to develop drugs to target these cells. Understanding the genetics and molecular biology of cancer stem cells and how they differ both from other cells of the same tumour type and from normal stem cells is crucial to this. In a recent paper in Cell , Pier Giuseppe Pelicci and colleagues at the IFOM-IEO campus, Milan, Italy, have now shown that normal and cancer stem cells differ in the way they divide, and implicated the loss of the tumour suppressor p53 in the development of the cancer stem cell phenotype.

Stem cells are known to divide asymmetrically, with one daughter cell remaining a largely quiescent stem cell and the other undergoing further rapid divisions as the progeny differentiate. This process of “asymmetric self-renewing division” achieves the key stem cell properties of differentiation and self-renewal in a single set of mitotic divisions. Normally, this keeps stem cell numbers low and fairly constant. However, at certain times, during development and after tissue injury in adulthood, stem cell populations must increase. We know that this is achieved, at least in invertebrates, through “symmetrical self-renewing division”, in which each stem cell division produces two more. Pelicci and colleagues have investigated the mechanism of self-renewing cell division in mammalian normal and tumour stem cells using an ErbB2 oncogene-activated mouse mammary tumour model.

Primary cells from wild type and ErbB2 positive tumour mouse mammary glands were cultured in suspension as floating colonies known as mammospheres. Each mammosphere was shown to derive from a single self-renewing stem cell and to contain both daughter stem cells and differentiated cells, with the intact spheres indistinguishable in cell make-up from normal breast epithelium and tumour tissue respectively. Pellici’s group first showed that the tumour-derived mammospheres contained a higher proportion of stem cells than the wild type ones. They then used a fluorescent dye, PKH-26, which binds to cell membranes and segregates in daughter cells, to determine the division history of each type of cell. In this assay, fluorescence intensity decreases with each cell division. In the wild type, those groups of cells with the highest fluorescence – and so the slowest dividing – were the only ones found to express stem cell specific markers and to be able to form new cultures, indicating that only these are enriched in self-renewing stem cells. In the tumour-derived mammospheres, however, groups of cells with fluorescence levels covering the whole range were shown to have stem cell potential, showing that tumour-derived stem cells proliferate more rapidly.

The group then used time-lapse video microscopy to monitor stem cell divisions and determine the proportion that were asymmetric (giving rise to one quiescent and one rapidly dividing cell) and the proportion in which both daughter cells divided to produce equal numbers of progeny. After three days, an asymmetrically-dividing cell will produce five and a symmetrically-dividing one eight progeny. In tumour-derived mammospheres, the majority of cells’ first division was symmetric, whereas the reverse applied in the wild type, showing that tumour stem cells undergo increased numbers of symmetrical cell divisions. This correlated with increased replicative potential of the tumour cells, which were found through serial re-plating to be almost immortal. In contrast, wild type stem cells rapidly lost their self-renewal ability.

The tumour suppressor protein p53 has been implicated in the regulation of stem cells’ potential for self-renewal through down-regulation of the transcription factor Nanog. Pelicci et al. investigated the influence of p53 on symmetrical and asymmetrical self-renewing stem cell division by analysing p53 negative cells using the same PKH-26 fluorescence assay. Pre-malignant P53 negative primary cells formed mammospheres that resembled those from tumours in containing increased numbers of stem cells that proliferated rapidly through symmetrical self-replicating division, and that stem cell proportion increased with time. Furthermore, the number of stem cells, and the proportion dividing symmetrically, was reduced in these cells when p53 was restored.

Together, these results demonstrate that cancer stem cells differ from normal adult stem cells in containing a larger proportion of cells that proliferate symmetrically, with both daughter cells showing the stem cell phenotype, and that these cells have increased replicative potential. A similar, and reversible, stem cell phenotype was observed in p53 null cells, even though p53 is not mutated in the ErbB2 positive cells used. The fact that loss of p53 favours cancer-promoting symmetrical stem cell division indicates that treatments that restore the activity of this key tumour suppressor may be effective against at least some types of cancer stem cells.

Reference

1. A. Cicalese, et al: The tumour suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell 138, 1083-1095 (2009)