"BMI-1 ist ein direktes Zielgen von E2F-1 in primären Neuroblastomen"

Summary The neuroblastoma is the third most common cancer in childhood (it amounts to 9% of all malignant tumors in this age) (Kaatsch et Zambon, 2006. The neuroblastoma shows a huge clinical and molecularbiological heterogeneity. The prognosis basically depends on the age of patients when the diagnosis is confirmed and on the amplification of the MYCN-oncogene (Berthold et Zischnag, 1997). MYCN shows a higher expression in fast-growing tumours than in slow-growing tumours (Schwab, 1995). The members of the MYC-family also play a significant role in the development of many other tumors (for example: breast-carcinoma, prostate-carcinoma, melanomas) (reviewed by Dang, 2005; Kent, 2006). The function of MYCN involves the blockade of cell differentiation, the induction of apoptosis and the immortalisation and malignant transformation of the cell. E2F1 induces the expression of MYCN in neuroblastomas (Strieder et Lutz, 2003). E2F1 is a member of the E2F-gene family, that regulates 5% of all genes (Pearson et Wang, 1991). In neuroblastoma E2F1 is allocated to two different genetic programmes: 1. to abide in s-phases 2. the induction of apoptosis E2F1 induces the polycombprotein Bmi1 in neuroblastomas. This interaction between these two oncogenes is also known from several other tumors. The polycombgroup proteins basically form two different complexes: 1. PcGi = initiating complex consists of: Eed (embryonic ectoderm development), Enx/EzH2 and Enx/EzH1 (Francis et Kingston, 2001; Satjin et Otte, 2001). 2. PcGm= maintainance complex consists: PSC (posterior sex complex), PH (polyhomeotic) and RING1 (Francis et Kingston, 2001; Satjin et Otte, 2001). In human cells this second complex contains also Bmi-1, Mel18, Mph1, M33, Mpc2. These two complexes play a redundant role in the regulation of the HOX-genes, which are important for the differentiation of the cell (van der Lugt, 1994; Owens et Hawley, 2002; Guthrie, 2004). For the proliferation of the haematopoetic stem cell both complexes acts antagonistally: the EED/EZH-complex inhibits, while the Bmi1-containig complex stimulates the proliferation of the haematopoetic stem cell (Franke, 1992; Alkema, 1997; Gunster, 1997). There are several advices for that Bmi1 acts as an oncogene in different tumors: 1. In lymphomas Bmi1 cooperates with CMYC (Haupt et Adam, 1991; van Lohuizen et Berns, 1991; Jacobs, 1999) and regulates INK4a/ARF negatively. Thereby p16 and p19 are activated. P16 inhibits the proliferation of the cells by regulation of cyclinD-dependent kinases and blocks the degradation and inactivation of the tumorsuppressor p53. 2. In mouse embryonic fibroblasts the overexpression of Bmi1 blocks senescence. Bmi1 regulates the INK4a/ARF locus. 3. Bmi1 is overexpressed in several tumors: non- small-cell lung cancer, colon carcinoma, multiple myeloma, medulloblastoma (Vonlanthen, 2001; Molofsky et Morisson, 2003; Kim, 2004). In neuroblastoma the role that Bmi1 plays in tumourgenesis is mainly unclear. Bmi1 is overexpressed in neuroblastoma and is induced by the overexpression of E2F1 (Kramps et al, 2003). It is known that E2F1 regulates Bmi1 on RNA level. It was part of my work to determine the specific interaction between E2F1 and Bmi1 and the effects that the knock-out of Bmi1 has in relation to the cell cycle. In my first experiment I demonstrated that Bmi1 is induced by E2F1 even on protein level. In another experiment protein biosynthesis was blocked by cycloheximid and Bmi1 was induced by E2F1 in 1A3 cells. In this experiment I showed that the induction of Bmi1 by E2F1 is a direct effect. I generated a point mutation of the potential E2F1 binding site in the Bmi1 promotor and showed in a luciferase assay, that this binding site is responsible for the Bmi1 induction by E2F1. To analyse the functional relevance of a gene, it is most effective to knock out this gene. I used two approaches to knock out the BMI1 gene. The first approach was performed by using RNAi. Cells showed a growth arrest after stable transfection of RNAi against BMI1. In order to overcome this problem I changed the strategy for the loss of function approaches by using a tetracycline inducible system as well for the RNAi system as for the dominant negative system. I analysed two of these Bmi1 knock out clones in the FACS analyses. In these approaches the loss of function of Bmi1 had neither an effect on S-phase induction nor on appoptosis. Cui et al. 2006 described similar effects by knocking out Bmi1 in neuroblastomas. There are no effects seen on cell cycle regulation, but they also described a growth arrest. This growth arrest is caused by a loss of the neuroblastoma cells for the capacity for self renewal. To summarize my results: 1. E2F1 regulates Bmi1 on nucleic and protein level. 2. The induction of Bmi1 by E2F1 is a direct effect. 3. The potential E2F1 binding site is responsible for the induction of Bmi1 by E2F1. 4. The loss of function of Bmi1 has no effect on the cell cycle profile in neuroblastoma cell, but causes a growth arrest in these cells. One possible reason for this growth arrest is a loss of the capacity of neuroblastoma cells for a self renewal.