Multipotential neural stem cells (NSCs) in the central nervous system (CNS) proliferate indefinitely and give rise to neurons, astrocytes, and oligodendrocytes. As NSCs hold promise for CNS regeneration, it is important to understand how their proliferation and differentiation are controlled. We show here that the expression of sox2 gene, which is essential for the maintenance of NSCs, is regulated by the Gli2 transcription factor, a downstream mediator of sonic hedgehog (Shh) signaling: Gli2 binds to an enhancer that is vital for sox2 expression in telencephalic neuroepithelial (NE) cells, which consist of NSCs and neural precursor cells. Overexpression of a truncated form of Gli2 (Gli2⌬C) or Gli2-specific short hairpin RNA (Gli2 shRNA) in NE cells in vivo and in vitro inhibits cell proliferation and the expression of Sox2 and other NSC markers, including Hes1, Hes5, Notch1, CD133, and Bmi1. It also induces premature neuronal differentiation in the developing NE cells. In addition, we show evidence that Sox2 expression decreases significantly in the developing neuroepithelium of Gli2-deficient mice. Finally, we demonstrate that coexpression of Gli2⌬C and Sox2 can rescue the expression of Hes5 and prevent premature neuronal differentiation in NE cells but cannot rescue its proliferation. Thus these data reveal a novel transcriptional cascade, involving Gli2 3 Sox2 3 Hes5, which maintains the undifferentiated state of telencephalic NE cells.

During development of the cerebral cortex, neural stem cells (NSCs) divide symmetrically to proliferate and asymmetrically to generate neurons. Although faithful segregation of mitotic chromosomes is critical for NSC divisions, its fundamental mechanism remains unclear. A class of evolutionarily conserved protein complexes, known as condensins, is thought to be central to chromosome assembly and segregation among eukaryotes. Here we report the first comprehensive genetic study of mammalian condensins, demonstrating that two different types of condensin complexes (condensins I and II) are both essential for NSC divisions and survival in mice. Simultaneous depletion of both condensins leads to severe defects in chromosome assembly and segregation, which in turn cause DNA damage and trigger p53-induced apoptosis. Individual depletions of condensins I and II lead to slower loss of NSCs compared to simultaneous depletion, but they display distinct mitotic defects: chromosome missegregation was observed more prominently in NSCs depleted of condensin II, whereas mitotic delays were detectable only in condensin I-depleted NSCs. Remarkably, NSCs depleted of condensin II display hyperclustering of pericentric heterochromatin and nucleoli, indicating that condensin II, but not condensin I, plays a critical role in establishing interphase nuclear architecture. Intriguingly, these defects are taken over to postmitotic neurons. Our results demonstrate that condensins I and II have overlapping and non-overlapping functions in NSCs, and also provide evolutionary insight into intricate balancing acts of the two condensin complexes.

Prominin1 (Prom1, also known as CD133 in human) has been widely used as a marker for cancer stem cells (CSCs), which self-renew and are tumorigenic, in malignant tumors including glioblastoma multiforme (GBM). However, there is other evidence showing that Prom1-negative cancer cells also form tumors in vivo. Thus it remains controversial whether Prom1 is a bona fide marker for CSCs. To verify if Prom1-expressing cells are essential for tumorigenesis, we established a mouse line, whose Prom1-expressing cells can be eliminated conditionally by a Cre-inducible DTA gene on the Prom1 locus together with a tamoxifen-inducible CreERTM, and generated glioma-initiating cells (GICs-LD) by overexpressing both the SV40 Large T antigen and an oncogenic H-RasL61 in neural stem cells of the mouse line. We show here that the tamoxifen-treated GICs-LD (GICs-DTA) form tumor-spheres in culture and transplantable GBM in vivo. Thus, our studies demonstrate that Prom1-expressing cells are dispensable for gliomagenesis in this mouse model.

These findings improve our understanding of the mechanism of cell death in Prominin-1-related disease and provide evidence that fenretinide may be worth studying in human disease.

Membrane morphology is an important structural determinant as it reflects cellular functions. The pentaspan membrane protein Prominin-1 (Prom1/CD133) is known to be localised to protrusions and plays a pivotal role in migration and the determination of cellular morphology; however, the underlying mechanism of its action have been elusive. Here, we performed molecular characterisation of Prom1, focussing primarily on its effects on cell morphology. Overexpression of Prom1 in RPE-1 cells triggers multiple, long, cholesterol-enriched fibres, independently of actin and microtubule polymerisation. A five amino acid stretch located at the carboxyl cytosolic region is essential for fibre formation. The small GTPase Rho and its downstream Rho-associated coiled-coil-containing protein kinase (ROCK) are also essential for this process, and active Rho colocalises with Prom1 at the site of initialisation of fibre formation. In mouse embryonic fibroblast (MEF) cells we show that Prom1 is required for chloride ion efflux induced by calcium ion uptake, and demonstrate that fibre formation is closely associated with chloride efflux activity. Collectively, these findings suggest that Prom1 affects cell morphology and contributes to chloride conductance.