Slugging their way to immortality: driving mammary epithelial cells into a stem cell-like state

Delineating the molecular factors that define and maintain the mammary stem cell state is vital for understanding normal development and tumourigenesis. A recent study by Guo and colleagues identifies two master transcriptional regulators of mammary stem cells, Slug and Sox9, ectopic expression of which confers stem cell attributes on differentiated mammary epithelial cells. Slug and Sox9 expression was also shown to determine in vivo metastatic potential of human breast cancer cell lines. Understanding these factors in the context of normal lineage differentiation is an important step toward elucidating the mammary epithelial cell hierarchy and the origins of cancer stem cells.


Background
Th e mammary gland is a highly dynamic tissue, undergoing signifi cant morphological change during puberty, pregnancy, lactation and involution. Flow cytometry using multiple markers has enabled the prospective isolation of mammary epithelial subpopulations suggesting a hierarchical organisation of stem, progenitor and diff erentiated cells. Utilising the cleared mammary fat pad transplantation assay, developed from the pioneering work of Deome and colleagues [1], mammary stem cells (MaSCs) have been defi ned as highly enriched in the CD24 +/Low Sca1 -CD49f High CD29 High population [2][3][4][5]. A single cell from this population is able to recapitulate the entire gland upon trans plantation [3,5], showing defi ning stem cell charac teristics of multidiff erentiative potential and self-renewal. Despite much work to defi ne factors necessary for MaSC function, molecular mechanisms regu lating MaSCs are poorly defi ned. Epithelial-mesen chymal transition (EMT), a key developmental programme in the embryo that confers mesenchymal cell traits on epithelial cells, has recently been linked to the MaSC state [6,7]. EMT has also been implicated in tumour invasion and metastasis, and a potential role in self-renewal strengthens known associations between normal tissue stem cells and cancer cells [8].
Despite ongoing controversy over the existence and origin of cancer cells, the cancer stem cell model -in which tumours are maintained by a population of stemlike cancer cells -provides an attractive framework to understand metastatic potential and tumour heterogeneity in response to treat ment. Irrespective of the origins of these cells, their exact nature and nomenclature and their operational detection, it is unlikely that neoplastic (stem-like) cell populations invent a novel programme to drive their sustained proliferation. Instead, cancer cells adopt the self-renewal programme active in the antecedent stem cell population and exploit this to organise the complex tissues observed at various stages of neoplastic progression. Delineating the molecular factors that defi ne the mammary epithelial cell hierarchy and maintain the MaSC state is therefore an important step towards under standing both normal development and tumouri genesis.

Slug and Sox9 as master regulators of the mammary stem cell state
A recent study by Guo and colleagues identifi ed two master transcriptional regulators of the mammary stem cell state: the EMT-associated transcription factor Slug, and the SRY-box transcription factor Sox9 [9]. Ectopic co-expression of Slug and Sox9 in primary mouse mammary cells in vitro for 5 days was suffi cient to convert diff erentiated luminal cells into MaSCs, as demonstrated by a competitive cleared mammary fat pad transplantation assay. Diff erentiated luminal cells expressing Slug and Sox9 were able to reconstitute the entire mammary gland upon serial transplantation, exhibiting defi ning stem cell characteristics of multi-potency and self-renewal. Th ese factors were shown to act cooperatively as expression of each factor individually was

Abstract
Delineating the molecular factors that defi ne and maintain the mammary stem cell state is vital for understanding normal development and tumourigenesis. A recent study by Guo and colleagues identifi es two master transcriptional regulators of mammary stem cells, Slug and Sox9, ectopic expression of which confers stem cell attributes on diff erentiated mammary epithelial cells. Slug and Sox9 expression was also shown to determine in vivo metastatic potential of human breast cancer cell lines. Understanding these factors in the context of normal lineage diff erentiation is an important step toward elucidating the mammary epithelial cell hierarchy and the origins of cancer stem cells. Ectopic Sox9 expression alone in diff erentiated cells produced cells with luminal progenitor characteristics, being able to proliferate in vitro and form hollow acini in three-dimensional culture conditions. Ectopic expression of Slug was able to convert luminal progenitor cells, which endogenously express Sox9, into cells with stem cell activity. In contrast, stem cell activity was enhanced in basal cells, which endogenously express Slug, by forced expression of Sox9. Interestingly, when Sox9 was ectopically expressed in basal cells at the same time as Slug was knocked down, the cells acquired a luminal progenitorlike phenotype in vitro. Gene expression analysis of diff erentiated cells ectopically expressing Slug or Sox9 showed upregulation of basal or luminal progenitorassociated genes, respectively, and both signatures were upregulated upon Slug/Sox9 co-expression. Th e authors postulated that Slug and Sox9 regulate basal and luminal lineage programmes, respectively. Each confers distinct biological properties on the cell, but both are required for MaSC function.
Given the important potential link between MaSCs and breast cancer stem cells, the study next looked at the eff ects of Slug and Sox9 on tumour-initiating potential and metastasis. Th e human breast cancer cell line MDA-MB-231 expresses both Slug and Sox9, forms tumours upon subcutaneous injection and metastasises to the lung upon tail vein injection into NOD/SCID mice. Knockdown of Sox9 reduced the tumour-initiating potential of the cells by over 70-fold following subcutaneous injection. In con trast, knockdown of Slug did not aff ect tumour initiation but the resulting tumours were sixfold smaller than those in controls. In a metastasis assay, Slug or Sox9 knockdown reduced lung metastases by fi vefold and 40-fold, respectively. In nonmetastatic MCF7ras human breast cancer cells implanted orthotopically into NOD/SCID mice, there was a signifi cant increase in macrometastases from primary tumours ectopically co-expressing Slug and Sox9. Finally, in a tissue microarray of 306 clinical breast cancer samples, patients with primary tumours expressing high levels of both Slug and Sox9 had a signifi cantly lower overall survival rate. Th e authors conclude that human breast cancer stem cells are controlled by the same master regulators governing the murine MaSC state.

Stem cell enhancement or altered cell fate?
Many studies have now demonstrated that transplantation potential in the cleared mammary fat pad assay is most highly enriched in the basal cell layer of the mammary epithelium [2][3][4]. If this transplantation potential can be serially propagated, we call this mammary epithelial stem cell potential. However, both the potential plasticity of the normal mammary epithelium and the inductive potential of the mammary environment are increasingly well recognised.
Th e work of Guo and colleagues is a very important step forward in understanding the molecular regulation of this plasticity and how it can promote mammary stem cell potential, but it will be diffi cult to disentangle these phenomena. Slug promotes the basal phenotype when ectopically expressed in luminal progenitors. Sox9 promotes luminal progenitor activity in diff erentiated luminal cells and stem cell activity in diff erentiated basal cells. Together, Slug and Sox9 determine stem cell-like activity in the mammary epithelium. Whether the basal phenotype is required for this activity (perhaps by determining response to the microenvironment) or whether changes in underlying self-renewal processes are independent of phenotypic changes remains unclear. Perhaps both scenarios are occurring, with Sox9 priming cellular self-renewal to respond to niche interactions that are made possible by the transcriptional programme driven by Slug activity.
Finally, as Guo and colleagues note themselves, in the mammary epithelium 'basal' does not necessarily mean 'stem cell'; and indeed the majority of basal mammary cells are not stem cells but rather are myoepithelial cells. Myoepithelial cells are, in terms of both function and identity, a cross between epithelial cells and mesenchymal cells and they express numerous EMT markers as part of their normal biology [10]. It is therefore wrong to characterise a luminal to basal phenotypic switch as EMT. Rather, expression of 'mesenchymal markers' is part of the normal diff erentiative programme of the mammary epithelium, and this paper must be understood in that context.

Competing interests
The authors declare that they have no competing interests.