Portal de Periódicos da CAPES

Bmi1 and Cell of Origin Determinants of Brain Tumor Phenotype

2007; Cell Press; Volume: 12; Issue: 4 Linguagem: Inglês

10.1016/j.ccr.2007.10.003

1878-3686

Peter B. Dirks,

Developmental Biology and Gene Regulation

Glioblastomas frequently express oncogenic EGFR and loss of the Ink4a/Arf locus. Bmi1, a positive regulator of stem cell self renewal, may be critical to drive brain tumor growth. In this issue of Cancer Cell, Bruggeman and colleagues suggest that brain tumors with these molecular alterations can be initiated in both neural precursor and differentiated cell compartments in the absence of Bmi1; however, tumorigenicity is reduced, and tumors contain fewer precursor cells. Surprisingly, tumors appear less malignant when initiated in precursor cells. Bmi1-deficient tumors also had fewer neuronal lineage cells, suggesting a role for Bmi1 in determination of cell lineage and tumor phenotype. Glioblastomas frequently express oncogenic EGFR and loss of the Ink4a/Arf locus. Bmi1, a positive regulator of stem cell self renewal, may be critical to drive brain tumor growth. In this issue of Cancer Cell, Bruggeman and colleagues suggest that brain tumors with these molecular alterations can be initiated in both neural precursor and differentiated cell compartments in the absence of Bmi1; however, tumorigenicity is reduced, and tumors contain fewer precursor cells. Surprisingly, tumors appear less malignant when initiated in precursor cells. Bmi1-deficient tumors also had fewer neuronal lineage cells, suggesting a role for Bmi1 in determination of cell lineage and tumor phenotype. Glioblastomas remain among the most aggressive human cancers. The application of the conceptual and methodological framework of neural stem cell biology to brain cancer (Bachoo et al., 2002Bachoo R.M. Maher E.A. Ligon K.L. Sharpless N.E. Chan S.S. You M.J. Tang Y. DeFrances J. Stover E. Weissleder R. et al.Cancer Cell. 2002; 1: 269-277Abstract Full Text Full Text PDF PubMed Scopus (547) Google Scholar, Holland et al., 1998Holland E.C. Hively W.P. DePinho R.A. Varmus H.E. Genes Dev. 1998; 12: 3675-3685Crossref PubMed Scopus (439) Google Scholar) and the identification of human brain tumor initiating cells (Singh et al., 2004Singh S.K. Hawkins C. Clarke I.D. Squire J.A. Bayani J. Hide T. Henkelman R.M. Cusimano M.D. Dirks P.B. Nature. 2004; 432: 396-401Crossref PubMed Scopus (6063) Google Scholar) has opened up fresh approaches to interrogate the cell of origin for brain tumors. Particularly in mouse model systems, investigators can address the effect of expression of oncogenes or loss of tumor suppressors in normal precursor or differentiated cell compartments. Current understanding of the mechanisms of tumor progression and initiation remain limited, particularly the cell context of recognized molecular signaling pathways implicated in the disease, such as aberrant EGFR signaling. Does oncogene expression drive a stem cell or a progenitor compartment in the tumor? Determinants of tumor phenotype and relationship to prognosis are also poorly understood. How do distinct molecular alterations specify the ultimate histopathologic tumor picture? How does the expression of neural precursor or differentiated lineages in the tumor correlate with tumor behavior? Are tumors that express more markers of differentiation less aggressive? As well, the relationship of tumor behavior to the putative cell of origin is not understood. Do tumors arise in a stem cell or a more differentiated cell compartment, and does the behavior and phenotype of the tumor depend on the cell compartment of origin? Are tumors that arise in stem cell compartments more malignant than those arising in progenitors, or vice versa? These questions come in to focus in the study by Bruggeman and colleagues (Bruggeman et al., 2007Bruggeman S.W.M. Hulsman D. Tanger E. Buckle T. Blom M. Zevenhoven J. van Tellingen O. van Lohuizen M. Cancer Cell. 2007; (this issue)PubMed Google Scholar). Bmi1 has been implicated in control of stem cells in multiple tissues, particularly as a positive regulator of self renewal, and Bmi1-deficient mice have deficiencies in their stem cell compartments, including the brain (Molofsky et al., 2003Molofsky A.V. Pardal R. Iwashita T. Park I.K. Clarke M.F. Morrison S.J. Nature. 2003; 425: 962-967Crossref PubMed Scopus (1121) Google Scholar, Park et al., 2003Park I.K. Qian D. Kiel M. Becker M.W. Pihalja M. Weissman I.L. Morrison S.J. Clarke M.F. Nature. 2003; 423: 302-305Crossref PubMed Scopus (1610) Google Scholar). Bmi1 promotion of proliferation and self renewal is thought to relate to suppression of the Ink4a/Arf locus (Bruggeman et al., 2005Bruggeman S.W. Valk-Lingbeek M.E. van der Stoop P.P. Jacobs J.J. Kieboom K. Tanger E. Hulsman D. Leung C. Arsenijevic Y. Marino S. van Lohuizen M. Genes Dev. 2005; 19: 1438-1443Crossref PubMed Scopus (282) Google Scholar), although other loci have recently been shown to be targeted as well (Fasano et al., 2007Fasano C.A. Dimos J.T. Ivanova N.B. Lowry N. Lemischka I.R. Temple S. Cell Stem Cell. 2007; 1: 87-100Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar). Ink4a/Arf loss itself, consistent with its tumor suppressor role, causes increased neural stem cell activity in vivo (Molofsky et al., 2006Molofsky A.V. Slutsky S.G. Joseph N.M. He S. Pardal R. Krishnamurthy J. Sharpless N.E. Morrison S.J. Nature. 2006; 443: 448-452Crossref PubMed Scopus (781) Google Scholar). Although Ink4a/Arf is lost genetically in a large fraction of human glioblastoma samples, mice deficient for Ink4a/Arf rarely develop spontaneous brain tumors. The current study by Bruggeman et al., 2007Bruggeman S.W.M. Hulsman D. Tanger E. Buckle T. Blom M. Zevenhoven J. van Tellingen O. van Lohuizen M. Cancer Cell. 2007; (this issue)PubMed Google Scholar attempts to further probe the functional role of Bmi1 together with Ink4a/Arf locus in an ex vivo oncogenic transduction model of glioma. Are tumors that can be initiated in the context of Ink4a/Arf deficiency dependent on Bmi1? The authors transduced either mouse adult-derived adherent neural precursor cells or early postnatal (day 7)-derived cortical mouse astrocytes with a brain tumor-associated oncogene to test the effects of tumorigenicity in the presence of a Bmi1-deficient background. The main finding of this study is that in the presence of Ink4a/Arf and Bmi1 deficiency, transformation by oncogenic EGFR results in different growth kinetics and tumor cell lineages depending on whether neural precursor cells or astrocyte cells are transformed. Importantly, both cell compartments are permissive for transformation, and therefore, Bmi1 expression is not required for tumor initiation. However, loss of Bmi1 attenuates overall tumorigenicity (in both cell compartments) in an Ink4a/Arf-deficient background and also affects lineage determination in resulting tumors; Bmi1 seems to be required for specification of neuronal lineages in the resulting tumors. In addition, Bmi1-deficient tumors have fewer cells expressing the neural precursor marker nestin, suggesting that one reason they may be less aggressive is that they have fewer stem cells. It remains uncertain whether in vivo tumors that lack Bmi1 are critically impaired in their self renewal ability, as defined by serial transplantation studies (Lessard and Sauvageau, 2003Lessard J. Sauvageau G. Nature. 2003; 423: 255-260Crossref PubMed Scopus (1298) Google Scholar). The results suggest most strongly that the brain tumors that emerge in these models depends on the cell context, and perhaps surprisingly, tumors originating from cultures highly enriched for neural stem cells appear less malignant than those arising from astrocytes, although mice died at about the same time (Figure 1). As normal neural stem cells are thought to be predominantly quiescent in vivo, could transformation events in these cells cause a more slowly proliferating (although not shown here) and less aggressive tumor? These findings are very interesting, but there are questions concerning the purity of cell types that are transformed, which reflects our generally poor understanding of the normal neural stem cell hierarchy and possible problems with cultures. The normal neural stem cell hierarchy is largely undefined, and a paucity of cell surface markers makes progress challenging. The markers used to identify cell populations in the brain are also not specific; GFAP marks stem cells as well as differentiated astrocytes. Consequently, the populations tested for transformation ability in this study are not directly purified from the brain but are derived from cultures, which may alter the function of the stem cells and more differentiated cells. The precursor and differentiated populations are not matched for age; could the astrocyte cultures derived from P7 brain contain more primitive cells than the adult derived precursors? Also, the effect that Bmi1 deficiency has on the normal neural stem cell compartment suggests that the starting cell populations in wild-type versus the deficient Bmi1 cells may be different. Despite these concerns, the authors do use state-of-the-art techniques to obtain enriched populations of stem cells or differentiated cells. This study also suggests that Bmi1 may have a role outside control of neural stem cells in the brain. Bmi1 is widely expressed in the postnatal brain, and it is expressed in the bulk population of the human brain tumor samples. This data would suggest that Bmi1 has a functional role in both normal stem cells and differentiated cells of the brain or different functions in these distinct cell populations. Bmi1 may be important for specifying differentiated neuronal lineages in the brain as shown by data in vitro and in the tumors in vivo. In summary, these studies, which are challenging to perform, raise further interesting questions about cell context of molecular alterations that play a role in brain tumors and do provide important further insight into the role of Bmi1 in brain tumorigenesis. The finding that distinct tumor phenotypes that arise from transformation of different cell populations in the brain suggests that knowing the cell of origin may be important for understanding the prognosis of the tumor. It will likely also be important to determine the signaling mechanisms that subsequently become operational in the different cell contexts to devise new therapies. Bmi1 Controls Tumor Development in an Ink4a/Arf-Independent Manner in a Mouse Model for GliomaBruggeman et al.Cancer CellOctober 16, 2007In BriefThe Polycomb group and oncogene Bmi1 is required for the proliferation of various differentiated cells and for the self-renewal of stem cells and leukemic cancer stem cells. Repression of the Ink4a/Arf locus is a well described mechanism through which Bmi1 can exert its proliferative effects. However, we now demonstrate in an orthotopic transplantation model for glioma, a type of cancer harboring cancer stem cells, that Bmi1 is also required for tumor development in an Ink4a/Arf-independent manner. Full-Text PDF Open Archive