Abstract

Neural stem cells are defined by their unique ability to undergo self-renewal divisions. By dividing asymmetrically, a stem cell simultaneously produces a daughter cell that retains stem cell identity, whereas the other starts to differentiate and contributes to a continuous supply of neural cell types. Drosophila neuroblasts provide an excellent model system to study asymmetric stem cell divisions. The first part of this thesis will concentrate on the important adaptor protein Miranda which ensures the asymmetric segregation of cell fate determinants to the differentiating ganglion mother cell during neuroblast mitosis. The dynamic apical-then-basal localization pattern and the requirement for both Myosin II and Myosin VI suggested that Miranda is actively transported to the basal pole as a myosin cargo. However, immunofluorescence studies combined with time-lapse confocal microscopy and FRAP analyses revealed that Miranda reaches the basal cortex by passive diffusion throughout the cell rather than by long range myosin-directed transport. Instead, myosins play an indirect role in asymmetric Miranda localization. The formation of active Myosin II filaments in early prophase results in the exclusion of Miranda from the apical cortex. In the cytoplasm, Miranda diffuses three-dimensionally through the cell and becomes restricted to the basal half of the metaphase neuroblast by Myosin VI to facilitate its interaction with a putative basal cortical anchor. There is growing evidence that deregulation of the self-renewing process of stem cells may be an early event in tumorigenesis and that many cancers contain a small population of so called cancer stem cells which are responsible for maintenance and growth of tumors. The second part of the thesis will report on the isolation of cells with stem-like features from a murine mouse model of oligodendroglioma with activated EGFR signaling and loss of the tumor suppressor p53 in the postnatal stem cell lineage. Although oligodendroglioma-derived progenitor cells share many similarities with normal neural stem cells, they have increased self-renewing and proliferation capacities and in addition, undergo aberrant differentiation. They are multipotential, however, when induced to differentiate they preferentially generate cells of the oligodendrocytic lineage recapitulating the properties of the tumor they originate from. Brain cancer derived stem-like cells generate new tumors following intracranial injections that faithfully reproduce the phenotype of the parental tumor qualifying them as cancer stem cells. Interestingly, neural stem cells isolated from tumor prone mice long before oligodendroglioma occurrence show similar, but less severe alterations in their self-renewing and differentiation capacities. Importantly, they never form orthotopic tumors and thus were referred to as premalignant stem cells. The overproduction of oligodendrocytic cells is caused by a defect in asymmetric cell division that is very likely accompanied with genetic instabilities and epigenetic alterations. This results strengthen the hypothesis that early defects in neural stem cells, together with additional genetic alterations lead to the progression to a more malignant stem cell type which is responsible for tumor growth and maintenance.