MicroRNAs: shortcuts in dealing with molecular complexity?

Recent studies from Clarke's group published in the journal Cell indicate that miRNAs may be the elusive universal stem cell markers that the field of cancer stem cell biology has been seeking. Distinct profiles of miRNAs appear to reflect the state of cell differentiation not only in breast cancer cells, but also in normal mammary epithelial cells. Moreover, they are conserved across tissues and species. The authors of this work also show evidence that downregulation of miRNA-200c in normal and malignant breast stem cells and in embryonal carcinoma cells has functional relevance, being responsible for the proliferative potential of these cells in vitro and in vivo.

MicroRNAs are able to control complex programs by regulating the expression of hundreds of genes simultaneously. Since their discovery almost three decades ago, numerous alterations in miRNA expression with varying underlying mechanisms were associated with human malignancies [1]. Th e study by Shimono and colleagues now shows that certain miRNAs may control the molecular makeup of stemness, and may be a shared trait of stem cells from various origins: embryonal and adult stem cells, normal and malignant stem cells [2]. Th is molecular similarity between normal and malignant stem cells re-enforces the concept put forward by the cancer stem cell model, according to which stem cells and early progenitor cells are more susceptible to transformation than their diff erentiated counterparts [3]. Th is may be due in part to a molecular intracellular context that sustains self-renewal and/or high proliferative potential.
Shimono and colleagues performed a comparative analysis of purifi ed CD44 + CD24lincancer stem cell populations from three diff erent breast cancers, which revealed diff erential expression of 37 miRNAs [2]. Among these, three clusters of miRNAs were consistently downregulated in an additional eight breast cancer samples: miRNA-183-96-182, miRNA-200c-141 and miRNA-200b-200a-429. Th e latter two clusters have the same seed sequence, suggesting that they may have overlapping targets. Remarkably, this downregulation appeared to be conserved in embryonal carcinoma cells (Tera-2 cells), in normal and malignant mammary stem cells of mouse origin defi ned by the CD24 -CD49f + linphenotype [4], and in normal mammary stem/progenitor cells defi ned by the CD49f + EpCAM neg/low CD31 -CD45phenotype [5]. When miRNA-200c levels were restored in any of these cells, they lost the ability to proliferate in vitro, as demonstrated by a dramatic decline in clonogenicity, and they lost the ability to proliferate in vivo, as demonstrated by an inability to generate tumors or normal outgrowths upon orthotopic implantation in mice.
In a long list of genes potentially regulated by miRNA-200c, the authors focused on BMI-1 for further validation, because of its recognized role in self-renewal. Bmi-1 is a polycomb group protein that, in a variety of experimental systems, appeared to be necessary for selfrenewal and proliferation of stem cells and appeared able to repress diff erentiation, senes cence and apoptosis. Impres sively, BMI-1 expression restored the clono genicity of MMTV-Wnt 1 breast cancer cells expressing miRNA-200c. Th e MMTV-Wnt 1 cell line was used in the study as an experimental model of mouse tumors with an expanded stem cell population [4]. Expression of miRNA-200c in these cells dramatically reduced clonogenicity, which was restored to levels seen in uninfected cells by lentiviral-driven expression of Bmi-1.
Th e implications of these fi ndings are several-fold. First, these results suggest the potential use of miRNAs as stem cell markers. Fairly simple phenotypes have so far been used as stem cell markers, defi ned by the presence of a maximum of 10 to 12 antigens or by the presence of a particular cell function, such as transmembrane effl ux (SP population) [6] or enzymatic activity (aldehyde dehydro genase) [7]. Since miRNAs are regulators of large molecular programs, they defi ne much more complex phenotypes. Moreover, they appear to confer specifi c

Abstract
Recent studies from Clarke's group published in the journal Cell indicate that miRNAs may be the elusive universal stem cell markers that the fi eld of cancer stem cell biology has been seeking. Distinct profi les of miRNAs appear to refl ect the state of cell diff erentiation not only in breast cancer cells, but also in normal mammary epithelial cells. Moreover, they are conserved across tissues and species. The authors of this work also show evidence that downregulation of miRNA-200c in normal and malignant breast stem cells and in embryonal carcinoma cells has functional relevance, being responsible for the proliferative potential of these cells in vitro and in vivo. developmental identities to cells. It would be very interesting to see whether the upregulation of the miRNA clusters miRNA-214, miRNA-127, miRNA-142-3p and miRNA-199a, identifi ed in the same study, is involved in promoting stem-cell-specifi c functions, such as selfrenewal and maintenance of an undiff erentiated state.
Another potential implication is developing cancer therapies by targeting miRNAs, as discussed in the commentary that accompanied Shimono and colleagues' paper [8]. Conceptually identical with cancer therapy through diff erentiation, miRNA targeting puts a molecular face to this old notion. By changing the intracellular molecular context, by interfering with the cells' stemness, we may be able to annihilate the consequences of cancer-initiating and cancer-promoting events without directly targeting them. If clusters of miRNAs with key roles in this cell-fate determination are identifi ed, it may be possible to circumvent the challenging task of elucidat ing networks of molecular interactions responsible for cell-fate determination and the complexity related to redundancy, feedback regulatory and compensatory mechanisms.
What would be the caveats of such approaches? Th e same characteristics that make miRNA appealing targets may represent important limitations. As the authors of this study mention, the number of miRNAs targets is typically large. Moreover, it includes genes that encode for molecules with opposing functions. For example, the TargetScan analysis of miRNA-200c indicates about 800 possible targets -some of them, such as Bmi-1, Notch1 and SOX2, whose upregulation was associated with selfrenewal; and other targets, such as PTEN, whose downregulation was associated with an undiff erentiated state and self-renewal [9,10]. Th is is consistent with previous observations that both oncogenes and tumor supressors, both genes promoting and suppressing cell proliferation, and both proapoptotic and antiapoptotic genes can be targets of a certain miRNA [2]. From this perspective, the large number of targets may not be advanta geous when developing miRNA-targeted strategies.
In conclusion, elucidating the role of miRNAs in cellfate determination would be an important step for understanding the basic biology of stem cells and their role during malignant transformation and tumor progression. Important applications may be developed based on this knowledge, such as using miRNAs as stem cell markers. Targeting miRNA also emerges as an opportunistic shortcut to circumvent the complexity resulting from feedback regulatory and compensatory mechanisms when aiming to eff ectively change cellular programs that dictate cell fate. Developing therapeutic approaches based on this concept should be considered with extreme caution, however, given the considerable potential for side eff ects.