Neurons display a highly polarized microtubule network that mediates trafficking throughout the extensive cytoplasm and is crucial for neuronal differentiation and function. In newborn migrating neurons, the microtubule network is organized by the centrosome. During neuron maturation, however, the centrosome gradually loses this activity, and how microtubules are organized in more mature neurons remains poorly understood. Here, we demonstrate that microtubule organization in post-mitotic neurons strongly depends on non-centrosomal nucleation mediated by augmin and by the nucleator γTuRC. Disruption of either complex not only reduces microtubule density but also microtubule bundling. These microtubule defects impair neurite formation, interfere with axon specification and growth, and disrupt axonal trafficking. In axons augmin does not merely mediate nucleation of microtubules but ensures their uniform plus end-out orientation. Thus, the augmin-γTuRC module, initially identified in mitotic cells, may be commonly used to generate and maintain microtubule configurations with specific polarity.

Regulation of the γ-tubulin ring complex (γTuRC) through targeting and activation restricts nucleation of microtubules to microtubuleorganizing centers (MTOCs), aiding in the assembly of ordered microtubule arrays. However, the mechanistic basis of this important regulation remains poorly understood. Here, we show that, in human cells, γTuRC integrity, determined by the presence of γ-tubulin complex proteins (GCPs; also known as TUBGCPs) 2-6, is a prerequisite for interaction with the targeting factor NEDD1, impacting on essentially all γ-tubulin-dependent functions. Recognition of γTuRC integrity is mediated by MZT1, which binds not only to the GCP3 subunit as previously shown, but cooperatively also to other GCPs through a conserved hydrophobic motif present in the N-termini of GCP2, GCP3, GCP5 and GCP6. MZT1 knockdown causes severe cellular defects under conditions that leave γTuRC intact, suggesting that the essential function of MZT1 is not in γTuRC assembly. Instead, MZT1 specifically binds fully assembled γTuRC to enable interaction with NEDD1 for targeting, and with the CM1 domain of CDK5RAP2 for stimulating nucleation activity. Thus, MZT1 is a 'priming factor' for γTuRC that allows spatial regulation of nucleation.

We compared the composition of γ-tubulin ring complexes tandem-affinity purified from asynchronous and mitotic human cells by mass spectrometry. We identified various interactors including the novel core subunit GCP8. GCP8 is the first subunit with an interphase-specific role in centrosomal γ-tubulin recruitment and microtubule nucleation.

Centriole biogenesis and maintenance are crucial for cells to generate cilia and assemble centrosomes that function as microtubule organizing centers (MTOCs). Centriole biogenesis and MTOC function both require the microtubule nucleator γ-tubulin ring complex (γTuRC). It is widely accepted that γTuRC nucleates microtubules from the pericentriolar material that is associated with the proximal part of centrioles. However, γTuRC also localizes more distally and in the centriole lumen, but the significance of these findings is unclear. Here we identify spatially and functionally distinct subpopulations of centrosomal γTuRC. Luminal localization is mediated by augmin, which is linked to the centriole inner scaffold through POC5. Disruption of luminal localization impairs centriole integrity and interferes with cilium assembly. Defective ciliogenesis is also observed in γTuRC mutant fibroblasts from a patient suffering from microcephaly with chorioretinopathy. These results identify a non-canonical role of augmin-γTuRC in the centriole lumen that is linked to human disease.

Gemcitabine shows a marked antitumor effect as a result of its cytotoxic action toward proliferative cells. In this article, we aim to investigate the potential antitumor and antiangiogenic effect of gemcitabine following a metronomic schedule that involves the regular administration of cytotoxic drugs at doses lower than standard treatment.In vitro results showed that human endothelial cells are more sensitive to gemcitabine (IC 50 3 nmol/L) than pancreatic tumor cells (IC 50 20 nmol/L). For in vivo studies, we used an orthotopic implantation model of human pancreatic carcinoma in nude mice. Gemcitabine was administered i.p. following a low-dose schedule (1 mg/kg/d for a month) and compared with the conventional schedule (100 mg/kg days 0, 3, 6, and 9 postimplantation). Metronomic treatment effect on established tumor was equivalent to standard administration. The measure of CD31 endothelial marked area allowed us to show an in vivo antiangiogenic effect of this drug that was further enhanced by using metronomic administration. This effect correlated with an induction of thrombospondin-1, a natural inhibitor of angiogenesis. Our results allow us to hypothesize that, in addition to a direct antiproliferative or cytotoxic antiendothelial cell effect, a secondary effect involving thrombospondin-1 induction might provide an explanation for the specificity of the effects of metronomic gemcitabine treatment. [Mol Cancer Ther 2008;7(3):638 -47]