NEDD1

Abstract: Microtubules in eukaryotic cells are nucleated from ring-shaped complexes that contain γ-tubulin and a family of homologous γ-tubulin complex proteins (GCPs), but the subunit composition of the complexes can vary among fungi, animals and plants. Arabidopsis GCP3-interacting protein 1 (GIP1), a small protein with no homology to the GCP family, interacts with GCP3 in vitro, and is a plant homolog of vertebrate mitotic-spindle organizing protein associated with a ring of γ-tubulin 1 (MOZART1), a recently identified component of the γ-tubulin complex in human cell lines. In this study, we characterized two closely related Arabidopsis GIP1s: GIP1a and GIP1b. Single mutants of gip1a and gip1b were indistinguishable from wild-type plants, but their double mutant was embryonic lethal, and showed impaired development of male gametophytes. Functional fusions of GIP1a with green fluorescent protein (GFP) were used to purify GIP1a-containing complexes from Arabidopsis plants, which contained all the subunits (except NEDD1) previously identified in the Arabidopsis γ-tubulin complexes. GIP1a and GIP1b interacted specifically with Arabidopsis GCP3 in yeast. GFP-GIP1a labeled mitotic microtubule arrays in a pattern largely consistent with, but partly distinct from, the localization of the γ-tubulin complex containing GCP2 or GCP3 in planta. In interphase cortical arrays, the labeled complexes were preferentially recruited to existing microtubules, from which new microtubules were efficiently nucleated. However, in contrast to complexes labeled with tagged GCP2 or GCP3, their recruitment to cortical areas with no microtubules was rarely observed. These results indicate that GIP1/MOZART1 is an integral component of a subset of the Arabidopsis γ-tubulin complexes.

Abstract: Institucio´ Catalana de Recerca i Estudis Avanc¸ats (ICREA),Passeig de Lluis Companys 23, 08010 Barcelona, SpainSummaryBackground:Theg-tubulinringcomplex(gTuRC)isamultisu-bunit complex responsible for microtubule (MT) nucleationin eukaryotic cells. During mitosis, its spatial and temporalregulation promotes MT nucleation through different path-ways. One of them is triggered around the chromosomes byRanGTP. Chromosomal MTs are essential for functional spin-dle assembly, but the mechanism by which RanGTP activatesMT nucleation has not yet been resolved.Results: We used a combination of Xenopus egg extractsand in vitro experiments to dissect the mechanism by whichRanGTP triggers MT nucleation. In egg extracts, NEDD1-coated beads promote MT nucleation only in the presence ofRanGTP. We show that RanGTP promotes a direct interactionbetween one of its targets, TPX2, and XRHAMM that definesa specific gTuRC subcomplex. Through depletion/add-backexperiments using mutant forms of TPX2 and NEDD1, weshow that the activation of MT nucleation by RanGTP requiresboth NEDD1 phosphorylation on S405 by the TPX2-activatedAurora A and the recruitment of the complex through aTPX2-dependent mechanism.Conclusions: The XRHAMM-gTuRC complex is the target foractivation by RanGTP that promotes an interaction betweenTPX2 and XRHAMM. The resulting TPX2-RHAMM-gTuRCsupracomplex fulfills the two essential requirements fortheactivationofMTnucleationbyRanGTP:NEDD1phosphor-ylation on S405 by the TPX2-activated Aurora A and therecruitment of the complex onto a TPX2-dependent scaffold.Our data identify TPX2 as the only direct RanGTP targetand NEDD1 as the only Aurora A substrate essential for theactivation of the RanGTP-dependent MT nucleation pathway.IntroductionDuring mitosis, different microtubule (MT) nucleation path-waysdrive theassembly of dynamic MTs to organizeabipolarspindle [1]. One of them relies on a gradient of guanosinetriphosphate (GTP)-bound Ran (RanGTP) centered on thechromosomes that promotes the dissociation of nuclearlocalization signal (NLS)-containing proteins from karyopher-ins and drives MT nucleation, stabilization, and organization[2]. Several spindle assembly factors have been identifiedas direct targets of RanGTP during mitosis [3–6]. However,the precise role of these factors in chromosomal-dependentMTassemblyisstillinmanycasesilldefined,andwethereforestill do not understand the mechanism at play. Because theseMTs are essential for building a functional bipolar spindlewhether centrosomes are present or not [7], understandingthis pathway is essential to fully understand cell division.In higher eukaryotic cells, MT nucleation relies on theg-tubulin ring complex ( TuRC), a multisubunit complex con-stitutedbymultiplecopiesofg-tubulinandanumberofassoci-ated proteins called gamma tubulin complex proteins (GCPs)[8, 9]. The gTuRC and its adaptor protein NEDD1 are requiredfor all the known pathways of MT nucleation in animal mitosis[10, 11]. These pathways are defined spatially in the mitoticcell by occurring at the centrosomes, on preexisting MTs, andaround the chromosomes. Work from different groups hasshown thatan important targetfor gTuRCregulationin mitosisis NEDD1 through its phosphorylation by different kinases[12–15]. We recently showed that chromosomal MT assemblyrequires NEDD1 phosphorylation on S405 by Aurora A [16].Another factor that is essential for RanGTP-dependent MTnucleation is TPX2 [17], but this RanGTP-regulated proteinwas shown to play multiple roles during spindle assembly[18, 19]. Interestingly, TPX2 is a RanGTP-dependent activatorof Aurora A [20–23]. However, the link between this activationand the phosphorylation of NEDD1 by Aurora A has not beenestablished yet. Although TPX2 can nucleate MTs in vitro[24], experiments performed in Xenopus egg extracts showedthat RanGTP-dependent MT nucleation does not occur inthe absence of g-tubulin, suggesting that the mechanism atplay is far more complex and involves more componentsin addition to TPX2 [25]. The mechanism by which RanGTPtriggers MT nucleation is therefore still unresolved [26–28].Here,weusedtheXenopuslaeviseggextractsystemincom-binationwithinvitroexperimentstoidentifythecomponentsofthe MT nucleation machinery involved in the RanGTP-depen-dent MT assembly pathway and the mechanism driving itsactivation in the M-phase cytoplasm.ResultsRanGTP Activates the MT Nucleation Activity ofNEDD1-gTuRCTo investigate the mechanism by which RanGTP drives MTnucleation in M phase, we used Xenopus laevis egg extractsin which the whole pathway can be triggered by addition of aGTP hydrolysis-deficient form of Ran (RanQ69L) bound toGTP (RanGTP) [29, 30]. Previous data showed that the gTuRCadaptor protein NEDD1 is essential for RanGTP-dependentMT nucleation [13, 16]. To determine whether the associa-tion of NEDD1 with the gTuRC is regulated by RanGTP, weimmunoprecipitated NEDD1 from extracts in the presenceorabsenceofRanGTP.Inbothconditions,NEDD1coimmuno-precipitated the components of the gTuRC XGrip109 (GCP3),g-tubulin, and XGrip195 (GCP6; Figure 1A). These results sug-gest that, in extracts, NEDD1 associates with intact gTuRCsin a RanGTP-independent manner. We then addressedthe functionality of this complex by examining whether theNEDD1 beads were associated with MTs in the extract.Surprisingly, the NEDD1 beads only generated MT asters in

Abstract: Mouse embryonic fibroblasts (MEFs) are commonly grown in cell culture and are known to enter senescence after a low number of passages as a result of oxidative stress. Oxidative stress has also been suggested to promote centrosome disruption; however, the contribution of this organelle to senescence is poorly understood. Therefore, this study aimed to assess the role of the centrosome in oxidative stress induced-senescence using MEFs as a model. We demonstrate here that coincident with the entry of late-passage MEFs into senescence, there was an increase in supernumerary centrosomes, most likely due to centrosome fragmentation. In addition, disrupting the centrosome in early-passage MEFs by depletion of neural precursor cell expressed developmentally downregulated gene 1 (NEDD1) also resulted in centrosomal fragmentation and subsequent premature entry into senescence. These data show that a loss of centrosomal integrity may contribute to the entry of MEFs into senescence in culture, and that centrosomal disruption can cause senescence.