nusap is essential for chromatin-induced spindle formation … · 2011-03-08 · detrimental to...

3244 Research Article

IntroductionDuring mitosis, replicated molecules of DNA compacted intochromosomes must be partitioned equally between daughter cells.The segregation of chromosomes needs to be precise, because asingle error can lead to aneuploidy or genetic instability, promotingcell death or disease. This is especially true for pluripotentembryonic stem (ES) cells, which have the capacity to propagateand differentiate into a range of specialized cell types, andeventually produce an entire organism.

Assembly of the mitotic spindle apparatus, which has a key rolein ensuring proper chromosome congression and accurate sisterchromatid separation, relies on two partially redundant mechanisms– a centrosome-mediated and a chromatin-induced pathway –whose relative contribution can vary depending on the cell type(O’Connell et al., 2008). With the first mechanism, microtubulesare nucleated around the centrosomes and subsequently radiateinto the mitotic cytoplasm until they are captured and stabilized bykinetochores (Kirschner and Mitchison, 1986). The second processrelies on the establishment of a RanGTP gradient around thechromosomes (Kalab et al., 2006), which regulates the activity ofseveral target proteins and thereby stimulates the nucleation andstabilization of microtubules near chromosomes (for a review, seeKalab and Heald, 2008).

Most RanGTP targets have been extensively characterized inbiochemical experiments, meiotic egg extracts and/or cancer celllines. These rather specific approaches have often suggested uniquefunctions for the proteins involved, which nonetheless proved tobe redundant in more complex experimental settings (Groen et al.,2009). Moreover, their contribution to spindle assembly in vivofrequently turned out to be more limited than anticipated fromthese in vitro experiments. For example, in vitro experiments have

shown that HURP is crucial for kinetochore fiber stabilization inhuman cells (Koffa et al., 2006; Silljé et al., 2006; Wong and Fang,2006) and essential for chromatin-induced microtubule assembly,microtubule stabilization and their organization into a bipolar arrayin Xenopus egg extracts (Casanova et al., 2008). Nevertheless,mice deficient in Hurp develop normally (Tsai et al., 2008),indicating that this protein is dispensable for cell proliferationduring mouse development and in most adult tissues. Instead, Hurpis required for proliferation of the endometrial stroma in the uterus,resulting in infertility of Hurp-deficient females owing toimplantation failure.

Another example is the chromokinesin Kid. This RanGTP target,originally characterized as a plus-end-directed microtubule-basedmotor (Tokai et al., 1996), contributes to the alignment ofchromosome arms at the metaphase plate in Xenopus egg extracts(Antonio et al., 2000; Funabiki and Murray, 2000), as well ashuman cells (Levesque and Compton, 2001; Tokai-Nishizumi etal., 2005) by creating a ‘polar ejection force’. Interestingly,inactivation of Kid in mice resulted in embryonic lethality beforeday 9.5 (E9.5) in 50% of the mice, whereas the other 50%developed into healthy and fertile mice (Ohsugi et al., 2008). Thisin vivo study pointed out that Kid contributes to compaction ofanaphase chromosomes, thereby preventing the formation ofmultinucleated cells. This function of Kid seems to be crucial onlyduring oocyte meiosis II and the first embryonic cleavages, whereasit is essential for neither somatic cell mitosis after the morulastage, nor for male meiosis (Ohsugi et al., 2008).

Together, these studies nicely illustrate the significance ofinvestigating the contribution of RanGTP targets to cell proliferationin vivo, as they unveiled new functions of the proteins concernedand indicated that their action is essential only at specific

NuSAP is essential for chromatin-induced spindleformation during early embryogenesisAn Vanden Bosch1, Tim Raemaekers2, Sarah Denayer3, Sophie Torrekens1, Nico Smets1, Karen Moermans1,Mieke Dewerchin4,5, Peter Carmeliet4,5 and Geert Carmeliet1,*1Laboratory of Experimental Medicine and Endocrinology, 2Laboratory of Membrane Trafficking, 3Laboratory of Molecular Endocrinology and4Vesalius Research Center (VRC), Katholieke Universiteit Leuven, 3000 Leuven, Belgium5Vesalius Research Center (VRC), VIB, B-3000 Leuven, Belgium*Author for correspondence ([email protected])

Accepted 2 June 2010Journal of Cell Science 123, 3244-3255 © 2010. Published by The Company of Biologists Ltddoi:10.1242/jcs.063875

SummaryMitotic spindle assembly is mediated by two processes: a centrosomal and a chromosomal pathway. RanGTP regulates the latterprocess by releasing microtubule-associated proteins from inhibitory complexes. NuSAP, a microtubule- and DNA-binding protein, isa target of RanGTP and promotes the formation of microtubules near chromosomes. However, the contribution of NuSAP to cellproliferation in vivo is unknown. Here, we demonstrate that the expression of NuSAP highly correlates with cell proliferation duringembryogenesis and adult life, making it a reliable marker of proliferating cells. Additionally, we show that NuSAP deficiency in miceleads to early embryonic lethality. Spindle assembly in NuSAP-deficient cells is highly inefficient and chromosomes remain dispersedin the mitotic cytoplasm. As a result of sustained spindle checkpoint activity, the cells are unable to progress through mitosis,eventually leading to caspase activation and apoptotic cell death. Together, our findings demonstrate that NuSAP is essential forproliferation of embryonic cells and, simultaneously, they underscore the importance of chromatin-induced spindle assembly.

Key words: Mitosis, Spindle assembly, Microtubules, Embryogenesis

Jour

nal o

f Cel

l Sci

ence

developmental stages or at defined sites. Thus, although in vitroexperiments can provide interesting clues about their function,RanGTP targets are not necessarily essential in vivo, possiblybecause of compensation by other proteins.

Until now, NuSAP [nucle(ol)ar spindle-associated protein] hasonly been characterized in biochemical reconstitution experimentsand through loss-of-function studies in vitro. In Xenopus eggextracts, NuSAP promotes the stabilization and crosslinking ofmicrotubules near chromatin, activities that are regulated byRanGTP (Ribbeck et al., 2006; Ribbeck et al., 2007). Togetherwith the localization of NuSAP to spindle microtubules and itsenrichment around chromosomes (Raemaekers et al., 2003;Ribbeck et al., 2006), these data suggest a role for NuSAP inchromatin-induced spindle formation. Consistent with this,knockdown of NuSAP in HeLa cells causes severe mitotic defects,i.e. progression through mitosis is delayed as a result of incompletechromosome congression to the metaphase plate. Yet, somecells progress to anaphase, often resulting in chromosomemissegregation or failed cytokinesis and, at later stages, theappearance of binucleated cells and multipolar spindles(Raemaekers et al., 2003). Nevertheless, in vitro systems sufferfrom a few drawbacks. For instance, HeLa cells and other cancercell lines are inherently abnormal and have weakened checkpointactivity. Hence, our aim was to explore whether NuSAP expressionis crucial for in vivo cell proliferation.

We therefore investigated whether the NuSAP expression patterncorrelates with cell proliferation in developing and adult mice, andfound an excellent association at all stages. Based on these findings,we hypothesized that NuSAP deficiency in mice could bedetrimental to embryonic development, a process that relies stronglyon cell proliferation. This hypothesis was verified by generatingmice with a systemic NuSAP inactivation, which resulted in earlyembryonic lethality, underscoring the importance of NuSAP duringin vivo proliferation.

ResultsTissue-specific expression of NuSAP during adult lifeIn cultured cells, the expression of NuSAP is restricted to theproliferation stage (Raemaekers et al., 2003). Based on thesefindings, it was assumed that NuSAP expression in vivo wouldalso be related to cell proliferation. Therefore, NuSAP mRNA andprotein levels were examined in several murine and human tissuesthat vary in their proliferation activity. Northern blot andquantitative RT-PCR (qRT-PCR) analysis of adult mouse tissuesshowed abundant expression of Nusap1 in the spleen, testis, thymus,bone and spine (containing bone marrow), with lower mRNAlevels in the lymph node, muscle, ovary, uterus, small intestine andstomach (Fig. 1A,B). In general, Nusap1 mRNA levels correlatedwell with expression of the histone H4 gene, which is a DNAreplication marker. These data indicated a preferential expressionof NuSAP in organs with persistent cell proliferation in adulthood,including the immune and reproductive systems. A similarexpression pattern was observed when NUSAP1 mRNA levelswere analyzed in human tissues. In particular, NUSAP1 was highlyexpressed in the thymus and testis with lower levels in the ovary,spleen, small intestine and colon (Fig. 1C). Western blot analysisof several adult mouse tissues confirmed this expression patternand showed a close correlation between the expression of NuSAPand proliferating cell nuclear antigen (PCNA). Notably, NuSAPwas abundantly expressed in the thymus, long bones (containingbone marrow) and colon, with lower expression levels in the skin

3245NuSAP is essential for early embryogenesis

(Fig. 1D). Taken together, these data indicate that NuSAP is highlyexpressed in adult tissues that contain a considerable population ofproliferating cells.

NuSAP expression is confined to cells engaged in mitosisor meiosisTo substantiate that the expression of NuSAP is limited toproliferating cells, we examined its expression pattern in murinetissues typified by localized regions of cell proliferation, includingthe small intestine, skin, bone and testis. Immunohistochemistry ofthe small intestine showed strong nuclear NuSAP staining, whichwas restricted to the epithelial cells localized in the crypts ofLieberkühn (Fig. 1E). This region contains the mitotically activecells, as confirmed by the positive staining for phospho-histone H3(Ser10) (H3-P), a marker of mitotic chromatin. By contrast, theabsorptive epithelial cells of the villi and cells of the laminapropria, the supporting layer underlying the intestinal epithelium,showed little or no NuSAP or H3-P expression. High-magnificationconfocal imaging of sections immunofluorescently labeled for bothNuSAP and H3-P showed that cells engaged in mitosis stainedpositive for both proteins (Fig. 1F). Notably, in early mitosis(prophase, prometaphase), NuSAP was found around the chromatin,whereas during anaphase it relocalized to the spindle microtubules,which is consistent with previous in vitro observations (Raemaekerset al., 2003). In addition, NuSAP staining was also detected in asmall subset of interphase cells with immunoreactivity localizingpredominantly to the nucleoli. This finding is in agreement withearlier data showing that NuSAP is expressed during the S-, G2-and M-phase of the cell cycle, but not during G1-phase(Raemaekers et al., 2003). Some interphase cells also displayedfoci of H3-P, which might be associated with transcriptionalactivation, as previously described (Cheung et al., 2000).

Next, we investigated whether NuSAP showed a confinedexpression pattern in murine skin sections. As expected, NuSAPstaining was restricted to the keratinocytes in the basal layer of theepidermis and to the outer root sheath of the hair follicles, whichharbor mitotically active cells (Fig. 1G). Accordingly, the stainingpattern of NuSAP was largely similar to that of Ki-67, a marker ofproliferating cells, albeit that Ki-67 was more abundantly expressedthan NuSAP. Differentiating keratinocytes and most dermal cellsexpressed neither Ki-67 nor NuSAP.

Subsequently, bone sections were analyzed for NuSAPexpression, because proliferation of the hematopoietic cells residingin the bone marrow continues to be important during adult life.NuSAP was abundantly expressed in the bone marrow cells, butnot in the differentiated osteocytes embedded in the bone cortex,an expression pattern that was similar to that of Ki-67, althoughKi-67 staining was more abundant (Fig. 1H).

Finally, we analyzed the expression pattern of NuSAP in thetestis. Surprisingly, little or no NuSAP staining was detected in themitotically active spermatogonia of the testis. By contrast, highexpression was confined to the primary spermatocytes undergoingmeiosis (Fig. 1I). The interstitial cells of the seminiferous tubules,the Sertoli cells and the seminiferous epithelium were negative forNuSAP expression. Additionally, in sections from the ovary, NuSAPexpression was present in meiotic cells, as indicated by co-stainingwith the oocyte marker Stat3 (Murphy et al., 2005), as well as inproliferating follicle cells (supplementary material Fig. S1A).

To conclude, our analysis of NuSAP expression in adult somaticas well as gonadal tissues indicated that its expression is confinedto actively proliferating cells in somatic tissues, as observed in the

Jour

nal o

f Cel

l Sci

ence

small intestine, skin and bone marrow. In the gonads, NuSAPexpression was also found in meiotic cells.

NuSAP is widely expressed during murine embryonicdevelopmentGiven the close correlation of NuSAP expression with cellproliferation in adult tissues, we hypothesized that NuSAP wouldbe abundantly expressed from early embryonic development onand throughout embryonic growth, processes that strongly rely oncell proliferation. Indeed, confocal imaging of mouse blastocyststage embryos clearly showed NuSAP staining in the nuclei ofnumerous interphase cells, with enrichment in the nucleoli (Fig.

3246 Journal of Cell Science 123 (19)

2A, left panel). In mitotic cells, NuSAP was detected around thechromosomes and at the chromosome-proximal microtubules ofthe spindle, as expected (Fig. 2A, right panel and magnifications).

Northern blot analysis of whole mouse embryos at later stagesof development showed high Nusap1 mRNA levels from E10.5until E16.5 followed by a decline in Nusap1 expression, a patternthat was comparable with that of histone H4 gene expression (Fig.2B). This developmentally regulated expression pattern wasconfirmed by qRT-PCR (Fig. 2C).

The tissue distribution of NuSAP expression was evaluated inE13.5 embryos, a developmental stage that is characterized byongoing organogenesis and is accompanied by extensive cell

Fig. 1. Expression of NuSAPcorrelates with cell proliferation inadult murine and human tissues.(A–C) Analysis of Nusap1 mRNAlevels in adult mice. Northern blotanalysis showing high similaritybetween Nusap1 and histone H4(H4) gene expression in murinetissues (A) and a comparableexpression pattern of NUSAP1 inhuman tissues (C). Actb (-actin)and Gapdh levels were used asloading controls (p. bl. lymph.,peripheral blood lymphocytes).(B)Quantification of Nusap1 andHprt mRNA levels in murine tissuesvia qRT-PCR. Levels are expressedrelative to the Nusap1-to-Hprt ratioin the brain, which was set to 1.(D)Western blot analysis of NuSAP,PCNA and -actin in murine tissues.(E–H) Immunohistochemicalanalysis showing that NuSAPexpression is confined toproliferating cells in somatic murinetissues. (E)In the small intestine,NuSAP and phospho-histone H3(Ser10) (H3-P) immunostaining arerestricted to the crypts. (F)Confocalimages show that intestinal cryptcells engaged in mitosis are double-stained for NuSAP and H3-P. Amagnification of cells in interphase(arrowhead), prophase (arrow),prometaphase (*) and anaphase (**)is shown in the right panels.(G,H)NuSAP and Ki-67 display acomparable expression pattern in theskin (G) and tibia (H). Scale bars:50m (E,G,H), 10m (F).(I)NuSAP immunostaining ofmurine testis showing positivity inmeiotic but not mitotic cells(magnification in right panel). Scalebar: 50m.

Jour

nal o

f Cel

l Sci

ence

proliferation. In situ hybridization and immunohistochemistrydemonstrated that NuSAP was ubiquitously expressed at thisdevelopmental stage (Fig. 2D,E). Positive signals were observedin several tissues including brain, heart, liver, lung, kidneys andspine. NuSAP expression seemed therefore not to be confined totissues originating from a specific germ layer. To explore the linkwith cell proliferation, the expression pattern of NuSAP wascompared with sites of BrdU incorporation and Ki-67 expression,which are two established markers of cell proliferation. In thegastrointestinal tract, NuSAP, BrdU and Ki-67 were detected in theendodermal epithelium, which is still devoid of cytodifferentiation


at this embryonic stage. Positive staining for these markers wasalso observed in the mesodermal cells that give rise to the laminapropria, the muscularis mucosae, the submucosa, the circular andlongitudinal muscle layers and the outer mesothelial layer(oesophagus, Fig. 2F; stomach, Fig. 2G). In the brain, NuSAP, aswell as BrdU and Ki-67 staining, was found adjacent to theventricles, where proliferating cells are located that originate fromthe ectoderm (Fig. 2H). In the growth plate of E18.5 pups, NuSAPexpression was restricted to the region of the periarticular andcolumnar chondrocytes, and was absent from the hypertrophiczone, which contains terminally differentiated cells (Fig. 2I). BrdU

Fig. 2. NuSAP is widely expressed during murine development. (A)Confocal images of NuSAP immunostaining of blastocysts isolated at embryonic day 3.5(E3.5) or cultured in vitro for 1 day (E3.5+1). Note that NuSAP redistributes from the nucleoli in interphase cells (late S- and G2-phase) to the spindle microtubulesin the vicinity of the chromosomes in mitotic cells (arrows and magnifications). Scale bar: 20m. (B,C)Analysis of Nusap1 mRNA levels in developing mouseembryos. (B)Northern blot analysis showing that Nusap1 expression correlates with that of histone H4 (H4) at different stages of development. 18S was used as aloading control. (C)Quantification of Nusap1 and Hprt mRNA levels via qRT-PCR. Levels are expressed relative to the Nusap1-to-Hprt ratio at E9.5, which wasset to 1. (D)In situ hybridization of Nusap1 on a sagittal section showing extensive Nusap1 expression at E13.5. (E)Immunostaining of NuSAP on a thoracic (top)and abdominal (bottom) transversal section showing extensive NuSAP expression at E13.5. Scale bar: 500m. (F–J) NuSAP localizes to regions of Ki-67expression and BrdU incorporation. Positive staining in E13.5 embryos is found in the endodermal epithelium and mesodermal cells of the trachea, oesophagus (F)and stomach (G) and adjacent to the ventricles in the brain (H), among other sites. Oe, Oesophagus; Tr, trachea; Ve, ventricles. (I,J) Immunostaining of NuSAP,Ki-67 and BrdU in the growth plate of E18.5 embryos. (I)Immunoreactivity is found in the periarticular and columnar region and is absent from hypertrophicchondrocytes. (J)Localization of NuSAP- and BrdU-positive cells in subsequent sections (some of the double-positive cells are indicated with arrows).Pe, periarticular; Co, columnar; Hy, hypertrophic region. Scale bars: 100m (E), 200m (F–H).

Jour

nal o

f Cel

l Sci

ence

and Ki-67 staining showed an identical distribution. Detailedanalysis of subsequent sections stained for NuSAP and BrdUshowed that nearly all cells displaying NuSAP positivity were alsoBrdU-positive cells, supporting previous observations that NuSAPexpression in interphase is restricted to S- and G2-phase (Fig. 2J).Overall, the pattern of NuSAP expression correlated better with theextent of BrdU incorporation than with Ki-67 expression.

Taken together, NuSAP expression during murine developmentwas not confined to a specific germ layer, as tissues originatingfrom the ectoderm (brain, Fig. 2H), endoderm (gastrointestinaltract, Fig. 2F,G; lungs and pancreas, supplementary material Fig.S1B,C) and mesoderm (cartilage, Fig. 2I; heart and kidney,supplementary material Fig. S1D,E) all contained NuSAP-positivecells. Rather, its expression correlated with regions of cell


proliferation, as shown by the close parallel of its expression withthe pattern of BrdU incorporation and Ki-67 expression.

NuSAP inactivation results in early embryonic lethalityWe next explored whether NuSAP is required during cellproliferation in vivo by disrupting the gene in murine ES cells.Homologous recombination and Cre–LoxP technology enabled usto generate mutant ES cells lacking exon 3 of the Nusap1 locus inone allele (Fig. 3A). Correct targeting of the mutant construct andgermline transmission was shown by Southern blot analysis of EScell and mouse tail DNA, respectively (Fig. 3B). Mice heterozygousfor Nusap1 were morphologically indistinguishable from wild-type (WT) animals (Fig. 3C). They appeared healthy, gained weightsimilarly to WT mice (supplementary material Fig. S2) and were

Fig. 3. Targeted disruption of Nusap1 induces early embryonic lethality. (A)Schematic representation of Nusap1 exon 3 inactivation strategy. Exons aredepicted as small boxes, loxP sites as triangles, the 3� external b-probe as a red bar and primers for genotyping as numbered arrows. At the bottom, DNA fragmentsgenerated by NcoI digestion of either the wild-type (9 kb) or mutant allele (>12 kb) are indicated. (B)Southern blot analysis of offspring from heterozygousmatings by digestion of genomic tail DNA with NcoI and detection with the 3� external b-probe. (C)WT and heterozygous Nusap1 littermates are phenotypicallyindistinguishable at 13 weeks. (D)PCR genotyping of WT and mutant alleles. (E)Phase-contrast microscopy shows that Nusap1–/– embryos cultured in vitro growslower than WT embryos and show internal disorganization. White arrowheads indicate cells of various sizes that are not connected to the inner cell mass (ICM,dotted lines) or trophectoderm layer (TE). Black arrowheads indicate very small cells or fragments inside the ICM (magnifications in the insets). After 3 days of invitro culture (E3.5+3), WT and heterozygous embryos develop outgrowths composed of an ICM on top of a layer of trophoblastic giant cells (TGC). Nusap1–/–

embryos that reach this stage consist either of a very small ICM, remnant TGC or a combination of both. Scale bars: 50m.

Jour

nal o

f Cel

l Sci

ence

fertile. However, intercrossing Nusap1+/– mice failed to yield anyNusap1-null mice among the 478 live births, whereas othergenotypes were present at the expected frequency (Table 1). Thesedata indicate that one functional Nusap1 allele is sufficient fornormal development, whereas inactivation of both alleles leads toembryonic lethality, suggesting that Nusap1 is essential forembryogenesis.

To determine the stage when NuSAP is crucial for embryonicdevelopment, embryos from Nusap1+/– intercrosses were collectedat different times of gestation and genotyped by PCR (Fig. 3D).Analysis at E10.5 did not reveal any Nusap1-null embryos; yet, alarger number of empty decidua was found in the uterus (38%)compared with the amount observed after mating WT mice (19%)(Table 1). The presence of these empty decidua suggests thatNusap1-null embryos could implant into the uterine wall, but failedto develop into gastrulated embryos and were therefore completelyresorbed at E10.5. Accordingly, all genotypes were present at theexpected frequency in embryos isolated before implantation atE3.5 (Table 1). In addition, no difference in the morphology of theblastocysts was observed between genotypes at this stage (Fig. 3E,left panels). Together, these data indicate that Nusap1-null embryosdied between E3.5 and E10.5.

Nusap1-null mice die at the peri-implantation stageTo define more accurately the stage of embryonic lethality, embryoswere collected at the blastocyst stage (E3.5), cultured in vitro andclosely monitored by phase-contrast microscopy (Fig. 3E). It shouldbe noted that embryos develop more slowly in vitro than in utero.WT and heterozygous Nusap1 embryos progressively increased insize and hatched from their zona pellucida after two to three daysin culture. They subsequently attached to the chamber slide byspreading of the trophectoderm cells, a behavior that is analogousto in vivo implantation. However, Nusap1-null embryos wereclearly distinguishable from WT and heterozygous embryos fromE3.5+1 onwards. They were smaller, and the size of their inner cellmass (ICM) was reduced, suggesting a decreased number of cells.These differences became more pronounced at E3.5+2 (Fig. 3E,dotted lines). In addition, the internal structure of the Nusap1-nullembryos was disorganized, because we could detect cells or cellularfragments that were part of neither the ICM, nor the surroundingtrophectoderm layer (Fig. 3E, white arrowheads). These cells orfragments were of various sizes, with an excess of very smallstructures. Small fragments were also found in the ICM (Fig. 3E,black arrowheads). Furthermore, after 3 days of culture, the numberof Nusap1-null outgrowths decreased (Table 2), indicating that alarge fraction of them died or were in such an advanced stage ofdisintegration that we were unable to retrieve sufficient geneticmaterial for genotyping.

The surviving Nusap1-null embryos were still able to hatchfrom their zona pellucida and attach to the chamber slide with anormal frequency. Indeed, 33% of the surviving Nusap1-nullembryos attached to the slide at E3.5+2, compared with 29% forWT embryos (n18 and 28, respectively), and this number increasedto 77% and 80%, respectively, at E3.5+3 (n26 and n45,respectively). Nevertheless, the Nusap1-null outgrowths remainedsmaller than the WT or heterozygous outgrowths and mainlyconsisted of residual ICM cells and/or remnant trophectodermcells, called trophoblastic giant cells at this stage, because theybecome polyploid, non-dividing cells after implantation (Fig. 3E,right panels). The aberrant appearance of Nusap1-null outgrowthsat E3.5+3 was associated with a strongly reduced mitotic activity,


assessed by immunofluorescence staining for H3-P, indicating thatthey were devoid of cell proliferation, whereas WT embryosdisplayed numerous H3-P-positive cells (supplementary materialFig. S3). Collectively, these results demonstrate that NuSAPinactivation induces embryonic death at the peri-implantation stage.

Absence of NuSAP leads to increased apoptosisThe reduced growth of Nusap1-null embryos, their rapiddisintegration and the presence of small cellular fragmentssuggested that lack of NuSAP resulted in apoptosis. To explore thispossibility, we closely inspected DNA integrity in embryos thathad been cultured for 3 days (Fig. 4A), a stage at which theNusap1-null outgrowths were severely affected. Confocalmicroscopy of TO-PRO-3 iodide-labeled WT outgrowths showedthat cells of the ICM had regularly shaped nuclei, were of uniformsize (assessed by phalloidin staining) and grew on top oftrophoblastic giant cells. By contrast, the smaller Nusap1-nulloutgrowths lacked a discernible ICM and the remaining cellscontained nuclei of various sizes. Large, irregular nuclei (probablythe remnants of trophoblastic giant cells) as well as very small,dense DNA fragments were detected in the Nusap1-null outgrowths.The latter morphological changes are suggestive of nuclearfragmentation during apoptosis.

To confirm that the apoptotic pathway in Nusap1-null embryoswas turned on, we assessed the activation of caspase-3 byimmunostaining. Quantitative analysis was performed on confocalimages through the entire embryo at E3.5+2, the time point whenthe phenotype became noticeable (Fig. 4B). Very few cells in theWT embryos stained positive for active caspase-3 (0 or 1 perembryo; n6). However, the number of caspase-3-positive cells inNusap1-null embryos was strongly increased (8–20 per embryo;n5; P<0.001) (Fig. 4C). These data indicate that the apoptoticpathway is triggered at an early stage in Nusap1-null embryos andthat massive apoptosis ultimately results in the disintegration ofthe entire embryo.

NuSAP deficiency impairs mitotic progressionA potential explanation for the increased apoptosis is that NuSAPdeficiency leads to mitotic defects. To investigate this possibility,

Table 1. Genotypes of progeny from Nusap1+/– matings

Empty Total Nusap1+/+ Nusap1+/– Nusap1–/– decidua number

Neonatal Expected 25% 50% 25% Detected 40% 60% 0% 478E10.5 Expected 20% 41% 20% 19%a 138 Detected 20% 42% 0% 38% 173E3.5 Expected 25% 50% 25% Detected 24% 56% 21% 102

aResult obtained after mating Nusap1+/+ mice.

Table 2. Genotypes of progeny from Nusap1+/– intercrossesharvested at E3.5 and cultured in vitro

Nusap1+/+ Nusap1+/– Nusap1–/– Total number

Expected 25% 50% 25%E3.5 + 1 day 22% 53% 25% 292E3.5 + 2 day 29% 50% 21% 84E3.5 + 3 day 26% 60% 14% 125E3.5 + 4 day 33% 55% 12% 33

Jour

nal o

f Cel

l Sci

ence

we analyzed the number of mitotic cells and mitotic progression inE3.5+1 embryos immunostained for H3-P by confocal imaging(Fig. 5A). The mitotic cells were identified based on both H3-Ppositivity and their DNA staining pattern. Only mitotic cells withclearly condensed chromosomes (prometaphase until telophase)were counted, because the speckled H3-P staining pattern inprophase cells was not clearly discernible from the H3-P stainingthat indicates transcriptional activation in interphase cells (Cheunget al., 2000). Likewise, apoptotic cells, as judged by the presenceof strongly condensed DNA fragments, or cells with extremelyaltered chromosome arrangements, were not included.

Mitotic cells were occasionally detected in WT and heterozygousblastocysts (0–9 per embryo; n10) and the majority of them(63%) were found in the ICM. By contrast, Nusap1-null embryosshowed a pronounced number of mitotic cells (4–22 per embryo;n10) (Fig. 5B). This finding could either result from increasedcell proliferation or a mitotic delay. Several observations supportthe latter possibility. First, quantification of the total cell numberrevealed that Nusap1-null embryos have a decreased total cellnumber at this stage compared with WT and heterozygouslittermates, making increased cell proliferation less likely (Fig.5C). Second, the distribution of the mitotic phases in these embryosis different between genotypes. In WT and heterozygous embryos,


19% of the mitotic cells was in prophase, 54% in prometaphase,18% in metaphase and 9% in anaphase or telophase, a distributionthat is comparable with the one described by Le Cam and colleagues(Le Cam et al., 2004). In Nusap1-null embryos, more cells werefound in prometaphase (72%), whereas the fraction of cells inmetaphase and in anaphase or telophase was strongly decreased(10% and 2%, respectively) (Fig. 5D). These data suggest thatNusap1-null cells can enter mitosis normally but get retarded inprometaphase.

NuSAP is required for proper chromosome alignmentTo investigate whether this impaired mitotic progression resultedfrom spindle abnormalities, we immunofluorescently stainedembryos for -tubulin and DNA to visualize the mitotic spindles.As expected, the mitotic spindles appeared properly arrangedduring the different stages of mitosis in WT and heterozygousembryos (Fig. 5E, left panel). More specifically, chromosomeswere precisely aligned at the equatorial plane during metaphaseand were correctly segregated during anaphase.

In Nusap1-null embryos, however, the majority of mitotic cellsresided in a stage that was reminiscent of prometaphase, i.e. amitotic spindle was assembled but chromosomes remainedunaligned (Fig. 5E, right panel). Spindle structures were oftendistorted and chromosomes were severely dispersed throughoutthe cell, suggesting that not all chromosomes could be capturedproperly by spindle microtubules. These observations, togetherwith the localization of NuSAP at chromosome-proximalmicrotubules, are consistent with the suggested role of NuSAP inlinking microtubules to mitotic chromosomes, a model that wasproposed by Ribbeck and co-workers (Ribbeck et al., 2007). Notethat we did not observe any multinuclear cell or multipolar spindlein the ICM of WT, heterozygous or Nusap1-null embryos. Takentogether, these in vivo findings indicate that NuSAP is required formitotic progression by ensuring proper chromosome alignment.

NuSAP deficiency causes sustained spindle assemblycheckpoint activityWe next tried to determine why the cells in Nusap1-null embryosaccumulated in prometaphase. We hypothesized that the spindleassembly checkpoint (SAC), which retains cells in mitosis bypreventing anaphase onset in the presence of unattachedkinetochores (Musacchio and Salmon, 2007), could not be satisfiedin NuSAP-deficient embryos because of impaired chromosomecongression. We tested this hypothesis in several ways. First, wereasoned that in embryos with a properly functioning SAC, cellswould accumulate in mitosis in response to artificial induction ofsustained SAC activation. In other words, when the SAC isoperative in Nusap1-null embryos, the mitotic index will (further)increase when SAC activity is stimulated. This was verified byculturing WT and Nusap1-null embryos in the presence of eithernocodazole or taxol. Both drugs prevent normal spindle assemblyand thus cause prolonged SAC activity as soon as the cells entermitosis. Interestingly, after 18 hours of culture in the presence ofeither of these compounds, the mitotic index of the embryosincreased markedly, irrespective of their genotype (Fig. 6A,B).More specifically, following nocodazole treatment, the mitoticindex in WT and heterozygous embryos increased 20-fold comparedwith 13-fold for Nusap1-null embryos. Taxol treatment caused asixfold increase of the mitotic index in WT and heterozygousembryos and a threefold increase in Nusap1-null embryos. Thesefindings demonstrated that the SAC was indeed capable of arresting

Fig. 4. Nusap1-null blastocysts display increased apoptosis. (A)Confocalimages of DNA integrity in blastocysts harvested at E3.5 and cultured in vitrofor 3 days (E3.5+3). The majority of nuclei in Nusap1–/– outgrowths displayfragmentation (magnification in inset). Note that these embryos consist mainlyof remnant trophoblastic giant cells (TGC) (Nusap1–/–, left panel) and/or asmall residual inner cell mass (ICM) (Nusap1–/–, right panel). Scale bar:10m. (B)Maximum intensity Z-projection of confocal images of E3.5+2blastocysts stained for cleaved caspase-3. Scale bar: 10m. (C)Thepercentage of cells positive for caspase-3 cleavage is increased in Nusap1–/–

embryos. Values are means ± s.e.m., n5–6; **P<0.01.

Jour

nal o

f Cel

l Sci

ence

cells in mitosis in Nusap1-null embryos and thus that NuSAPinactivation did not affect the integrity of the SAC.

Second, we examined the presence of a crucial SAC component,BubR1, on the kinetochores of unattached chromosomes. Despitetechnical limitations, BubR1 staining was found on the kinetochoresin pro(meta)phase cells of control as well as Nusap1-null embryos(Fig. 6C). These data suggested that NuSAP is not essential forrecruitment of BubR1 to unattached chromosomes. Additionally,the SAC was shown to be activated in at least a subset of thepro(meta)phase cells found in Nusap1-null embryos at E3.5+1.

Finally, we treated embryos with SP600125, a compound thatdisrupts the SAC by inhibiting Mps1 (Schmidt et al., 2005). Wefound that the mitotic index in WT and heterozygous embryos wasnot altered, but it strongly decreased (–47%) in Nusap1-nullembryos, indicating that sustained SAC activity was indeedresponsible for retaining the cells in Nusap1-null embryos inmitosis (Fig. 6D,E).

To conclude, our data support the following sequence of eventsin Nusap1-null embryos: incomplete chromosome congression tothe metaphase plate delays cells in prometaphase and keeps theSAC activated, causing a steep rise in the mitotic index; thisprolonged stay in mitosis eventually triggers the apoptotic pathway


and leads to cell death. Together, our findings show that NuSAP isessential for embryogenesis.

DiscussionIn the present study, we demonstrate that NuSAP, a protein shownto contribute to chromatin-induced spindle formation in vitro bygenerating microtubules near chromatin, is essential for cellproliferation during murine embryonic development. NuSAPdeficiency impairs mitotic progression in early embryos becausechromosomes fail to align at the metaphase plate, resulting insustained SAC activity and ultimately, cell death (Fig. 7). Loss ofNuSAP function clearly cannot be compensated by any otherprotein, underscoring the unique function of NuSAP in chromatin-induced spindle assembly.

Apart from validating that NuSAP is essential for mitotic spindleassembly in vivo, our study confirms the requirement of chromatin-induced spindle assembly for cell proliferation in vivo. NuSAP isone of the few proteins, in addition to Rae1 and Cdk11 (Yokoyamaet al., 2008; Blower et al., 2005), shown thus far to be crucial toRanGTP-mediated spindle assembly in vivo, as demonstrated bythe early lethality of mice lacking either of these proteins (Babu etal., 2003; Li et al., 2004). Indeed, other RanGTP targets seem to

Fig. 5. Impaired mitotic progression in NuSAP deficient embryos. (A)Confocal images of phospho-histone H3 (Ser10) (H3-P) staining in E3.5+1 blastocysts,showing a high number of mitotic cells in Nusap1–/– blastocysts. Scale bar: 10m. (B)The mitotic index is increased in Nusap1–/– embryos. Values are means ±s.e.m.; n10; **P<0.01. (C)The total cell number of E3.5+1 embryos is comparable between genotypes. Values are means ± s.e.m.; n10. (D)Cells inprometaphase are enriched in E3.5+1 Nusap1–/– embryos. Values are means ± s.e.m.; n14–30; *P<0.05. (E)Confocal images showing mitotic spindles in E3.5+1embryos (Nusap1+/+ and Nusap1+/– left to right: prometaphase, metaphase, anaphase, telophase, cytokinesis). In Nusap1–/– embryos, the majority of the cellsdisplay abnormal spindle organization and chromosome dispersion (arrows). Scale bar: 10m.

Jour

nal o

f Cel

l Sci

ence

be dispensable for spindle assembly in vivo, as exemplified by thesurvival of mice lacking either Hurp or Kid (Ohsugi et al., 2008;Tsai et al., 2008). The fact that NuSAP is crucial for chromatin-induced spindle assembly might be explained by its ability tointeract with microtubules and DNA (Ribbeck et al., 2006; Ribbecket al., 2007), whereas other proteins, such as Hurp, only act onmicrotubules (Koffa et al., 2006; Silljé et al., 2006; Wong andFang, 2006).

The chromosome dispersion and inefficient chromosomecongression observed in Nusap1-null embryos is consistent withits proposed role in the generation of microtubules nearchromosomes, thereby increasing the likelihood that a chromosomegets incorporated into the spindle (Ribbeck et al., 2007). The exactmolecular mechanism underlying the interaction between NuSAP,chromatin and microtubules remains unresolved.

When chromosomes fail to align at the metaphase plate inNusap1-null cells, the SAC remains active for a prolonged period,after which caspase-mediated apoptosis is induced. Although our


data are not conclusive about the mechanism that triggers celldeath in early embryos, we speculate that cells lacking NuSAPeither die in mitosis, or in G1-phase after exiting mitosis, withoutsatisfying the spindle checkpoint, referred to as ‘mitotic slippage’(Weaver and Cleveland, 2005; Rieder and Maiato, 2004). In thelatter case, the cells would bypass the SAC as a result of gradual,proteasome-mediated degradation of cyclin B1 (Brito and Rieder,2006).

An intriguing, yet not fully explained observation is that Nusap1-null embryos are able to reach the blastocyst stage withoutdisplaying morphological abnormalities or obvious cell proliferationdefects. Analogous findings of peri-implantation lethality are alsonoticed when cell-cycle-related genes are inactivated (Artus andCohen-Tannoudji, 2008). A possible explanation is the plasticity ofthe mammalian pre-implantation embryo, which has the ability toadapt its development in response to various perturbations.Alternatively, this phenotype might indicate that NuSAP is stillpresent in early Nusap1-null embryos, either in the form of protein

Fig. 6. Analysis of spindle checkpoint activity in NuSAP-deficient embryos. (A)Confocal images of E3.5 blastocysts cultured in the presence of nocodazole ortaxol and stained for phospho-histone H3 (Ser10) (H3-P). Scale bar: 10m. (B)The mitotic index of the embryos is increased by nocodazole or taxol treatmentirrespective of the genotype of the embryo. Values are mean ± s.e.m.; n10 (untreated), n>9 (nocodazole), n3 (taxol); ***P<0.001 vs untreated of same genotype.(C)Confocal images showing BubR1 staining in several pro(meta)phase cells in control as well as Nusap1–/– blastocysts cultured in vitro for 1 day (E3.5+1). Scalebar: 10m. (D)H3-P staining in E3.5 embryos cultured overnight in the presence of SP600125. Scale bar: 10m. (E)The mitotic index of Nusap1–/– embryos isstrongly decreased after treatment with SP600125. Values are mean ± s.e.m.; n3–6.

Jour

nal o

f Cel

l Sci

ence

originating from the oocyte or as the product of maternally inheritedmRNA. Generally, this maternal mRNA persists only until the midtwo-cell stage, however, the proteins encoded by this mRNA canbe present beyond this time point (Bachvarova and De Leon,1980). Nevertheless, this possibility is less likely in the case ofNuSAP, because it is degraded at the end of mitosis. In agreement,we could not demonstrate its presence in Nusap1-null blastocystsat E3.5 by confocal microscopy, whereas it was clearly detectablein wild-type embryos (supplementary material Fig. S4), indicatingthat the amount of maternally derived protein is negligible at thisstage in Nusap1-null embryos. An alternative explanation is thatchromatin-induced microtubule nucleation is not essential for theearly cleavage divisions or that, because of the slow cycling speedof ES cells before implantation (Ciemerych and Sicinski, 2005),chromosomes have sufficient time to become aligned at themetaphase plate, even though chromosome congression isinefficient.

The present study also highlights the parallels and contradictionsbetween in vivo inactivation and in vitro depletion of NuSAP(Raemaekers et al., 2003). Several similarities are prominent; forinstance, in both cell types, chromosome congression to themetaphase plate is impaired, leading to accumulation of cells inmitosis. However, differences in the way HeLa cells and ES cellsreact to NuSAP depletion are even more pronounced. WhereasHeLa cells accumulate in mitosis without becoming fully blockedin the cell cycle, the progression through the cell cycle of virtuallyall embryonic cells is precluded once they have passed a crucialstage, presumably around E3.5. As a result, no abnormal anaphases,binucleated interphase cells or multipolar spindles are detected in


the ICM of Nusap1-null embryos, whereas their number is greatlyincreased among NuSAP-depleted HeLa cells. Rather thanprogressing through mitosis and producing abnormal daughtercells, the ES cells are delayed in prometaphase and eventually diefrom apoptosis. A likely explanation is that HeLa cells are onlypartially depleted by RNAi, whereas the ES are completely devoidof NuSAP. Alternatively, these differences might reflect the intrinsiccapability of HeLa cells to bypass the SAC in the presence of alimited number of unaligned chromosomes (Wong and Fang, 2006).

The crucial contribution of NuSAP to cell proliferation in vivois also suggested by its expression pattern in vivo, which weextensively characterized. NuSAP expression in vivo is confinedto proliferating cells, as evidenced by its restricted localization toareas where proliferating cells are present, such as the crypts of theintestine in adult mice and the cells lining the ventricles in thebrains of developing embryos. The strong correlation of the patternof NuSAP expression with that of BrdU incorporation and Ki-67staining confirms the proliferation-restricted expression of NuSAP.Moreover, its colocalization with the mitotic marker H3-P, togetherwith its presence in only a subset of interphase cells, stronglysuggest that the cell-cycle-dependent expression of NuSAP issimilar in vivo and in vitro. Consistent with this, NuSAP expressionis generally less abundant than Ki-67 expression. We argue thatNuSAP more accurately predicts the number of proliferating cells,because it is only expressed from late S-phase, in cells that haveundergone DNA replication and are therefore committed to entermitosis (Raemaekers et al., 2003), whereas Ki-67 is also expressedduring G1-phase (Gerdes et al., 1984). These G1-phase cellsrepresent a large fraction of the total cell population; however,their fate can range from entering G0-phase and staying quiescentfor a longer period, over proceeding to S-phase, to becomingapoptotic. Consequently, quantification of Ki-67-positive cellsgenerally leads to an overestimation of the number of proliferativecells.

Remarkably, in the gonads, NuSAP expression was also foundin meiotic cells. More precisely, in the testis, NuSAP was notexpressed in the proliferating spermatogonia, but instead, it waspresent in spermatocytes undergoing meiosis. By contrast, in theovary, NuSAP expression was found both in proliferating follicularepithelial cells and in oocytes. The role NuSAP has in meiosisremains to be elucidated.

Proliferation is not only essential during development and tissuehomeostasis during adult life, but it is also a hallmark ofmalignancy. Consistent with this, NuSAP was detected in fourindependent studies aiming to identify genes related to clinicaloutcome of breast cancer (Lauss et al., 2008). Thus, apart frombeing a reliable proliferation marker, NuSAP holds the potential ofbecoming a diagnostic marker for breast cancer and potentiallyalso other tumor types.

To conclude, the expression of NuSAP is restricted toproliferating cells in developing as well as adult mice, establishingits potential as a reliable proliferation marker. Importantly, weshow that NuSAP is an essential protein in vivo and herebyemphasize the importance of the chromatin-induced mechanism ofspindle assembly for cell proliferation in vivo.

Materials and MethodsImmunohistochemistryDissected bones (E18.5), soft tissues and whole embryos (E13.5) were fixed in 2%paraformaldehyde in PBS. Bones were decalcified in 0.5 M EDTA-PBS (pH 7.4) for3 days, before paraffin embedding. To detect NuSAP and Ki-67, antigens wereretrieved in Tris–EDTA solution (10 mM Tris-HCl, 1 mM EDTA, pH 9) at 98°C.

Fig. 7. Schematic interpretation of the effects of NuSAP inactivation onmitosis in embryonic stem cells. In wild-type (Nusap1+/+) cells, bothcentrosome- and chromatin-induced microtubule formation contribute tospindle assembly, resulting in efficient spindle assembly and chromosomecongression. The spindle assembly checkpoint (SAC) ensures accuratechromosome segregation by preventing anaphase onset until all chromosomesare properly attached to microtubules from both spindle poles. One of theproteins involved in this checkpoint, BubR1, localizes to unattachedkinetochores. In NuSAP-deficient (Nusap1–/–) cells, the chromatin-inducedpathway of spindle assembly is severely affected. Consequently, chromosomecapture by spindle microtubules is highly inefficient and the chromosomesremain dispersed in the mitotic cytoplasm. The SAC therefore cannot besatisfied, as witnessed by the presence of BubR1 on the kinetochores.Eventually, caspases are activated and cell death is induced.

Jour

nal o

f Cel

l Sci

ence


Endogenous peroxidase activity was blocked with 3% H2O2 in methanol andunspecific binding with TNB blocking buffer (TSA Biotin System kit, Perkin Elmer,Waltham, MA). Sections were incubated overnight with rabbit anti-NuSAP (1:1000)(Raemaekers et al., 2003) or rabbit anti-Ki-67 antibody (Novocastra, Newcastleupon Tyne, UK; 1:2000). Antibody binding was visualized with the Dako EnvisionSystem (Dako, Glostrup, Denmark). For H3-P immunostaining (rabbit anti-H3-Pantibody; Upstate, Charlottesville, VA; 1:100), antigens were retrieved in TargetRetrieval Solution (Dako; pH 6) at 90°C; endogenous peroxidases were blockedwith 0.3% H2O2 and unspecific binding was blocked with 20% normal goat serum(Dako) in TNB blocking buffer. BrdU injection and staining were performed asdescribed (Maes et al., 2004). Immunofluorescent labeling of NuSAP, H3-P andStat3 (rabbit anti-Stat3 antibody; Santa Cruz Biotechnology, Santa Cuz, CA; 1:100)was performed essentially as described for immunohistochemistry. Antibody bindingwas detected with Alexa-Fluor-488- or -546-labeled secondary antibodies (1:500;Invitrogen, Carlsbad, CA). Stat3 detection required amplification with the TyramideSignal Amplification (TSA; Cy3) System (Perkin Elmer). Hoechst 33342 was appliedto stain nuclei. Bright-field images were acquired on an Axioplan 2 microscope(Zeiss, Jena, Germany) and fluorescence images on a FV1000 confocal microscope(Olympus, Tokyo, Japan).

Isolation of RNA and real-time quantitative RT-PCR analysisRNA isolation from adult mouse tissues and whole embryos as well as qRT-PCRanalysis was performed essentially as described (Maes et al., 2004). Primer andprobe sequences are outlined in supplementary material Table S2.

Targeted disruption of the Nusap1 locusA mouse bacterial artificial chromosome library was screened with Nusap1 cDNAas a probe. One clone, encompassing exons 3, 4 and 5 of the Nusap1 gene, wassubjected to restriction mapping, subcloning and DNA sequencing. To delete exon3 (containing the start codon), a targeting construct was generated by insertinghomologous regions of the Nusap1 gene into a pNT-lox-2b vector (Fig. 3A).Following electroporation into 129/Sv mouse ES cells, G418- and gancyclovir-resistant clones were screened for homologous recombination. Correctly recombinedclones were transiently transfected with a Cre-expressing plasmid to remove the neocassette together with exon 3. Selected clones were introduced into Swiss mouseembryos via morula aggregation, implanted into pseudopregnant foster mothers(129/Sv) and bred for germline transmission with Swiss mice, to eventually obtainNusap1+/– mice. All mice were bred in our animal housing facilities(Proefdierencentrum Leuven). Experiments were approved by the ethical committeeof the Katholieke Universiteit Leuven.

Genomic DNA isolation and PCR genotypingGenomic DNA from mouse-tail biopsies was prepared as described (Maes et al.,2006). Genomic DNA from early mouse embryos was obtained by incubating themin 20 l lysis buffer (10 mM Tris–HCl pH 8.3, 50 mM KCl, 2.5 mM MgCl2, 0.1 g/lgelatin, 0.45% NP-40, 0.45% Tween-20, 200 g/ml proteinase K) for 6 hours at55°C. Semi-nested PCR primers (arrows in Fig. 3A; sequences in supplementarymaterial Table S1) were designed to generate a product specific for either the wild-type or the mutant Nusap1 allele.

Isolation of preimplantation embryos and in vitro blastocyst cultureNusap1+/– mice were mated overnight and successful matings were detected by thepresence of a vaginal plug. Pregnant mice were killed at E3.5, their uterus wasremoved and embryos were flushed out of the oviducts with ES cell medium(Schoonjans et al., 2003). For in vitro culture, embryos were transferred to gelatin-coated chamber slides containing ES cell medium. Where mentioned, nocodazole,taxol (Sigma) or SP600125 (Enzo Life Sciences) was added to the medium to a finalconcentration of 200 ng/ml, 200 nM and 20 mM respectively. The embryos wereinspected and photographed regularly with a Zeiss Axiovert 25 inverted microscope.

Immunofluorescence analysis of embryosWhen needed, embryos were treated briefly with acidic Tyrode’s solution (Sigma)to remove their zona pellucida. They were fixed in 4% paraformaldehyde in PBSand permeabilized in 1% Triton X-100 in PBS. Unspecific binding was blocked with0.5% Tween-20 in PBS containing 5% bovine serum albumin. Subsequently, theembryos were incubated overnight with either rabbit anti-NuSAP (1:500), rabbitanti- H3-P (1:500), rabbit anti-caspase-3 (Cell Signaling Technology, Danvers, MA;1:500) or sheep anti-BubR1 antibody (Abcam, Cambridge, UK; 1:200). Unspecificbinding was blocked with 50% normal goat serum, followed by incubation with anAlexa-Fluor-488- or -546-labeled secondary antibody (Invitrogen; 1:500). Finally,the embryos were incubated with Alexa-Fluor-488- or -546-conjugated phalloidin(Invitrogen; 1/100) and mounted in fluorescence mounting medium (Dako) containingTO-PRO-3 iodide (Invitrogen; 1:200). For -tubulin staining (mouse anti--tubulin;Sigma, clone DM1A; 1:200), embryos were incubated in prewarmed (37°C)extraction buffer (80 mM PIPES pH 6.8, 1 mM MgCl2, 5 mM EGTA, 0.5% Tween-20, 25% glycerol) before permeabilization in 1% Tween-20 in TBS.

Embryos were imaged at several z-positions by confocal microscopy, on averageevery 1.5 m. Images were acquired either on an inverted Diaphot 300 microscope(Nikon, Tokyo, Japan) (Plan Apo 60�/1.40 oil) connected to an MRC1024 confocal

imaging system (Bio-Rad Laboratories, Hercules, CA) or on an inverted EclipseE800 microscope (Nikon) (Plan Apo 60�A/1.40 oil) connected to a Bio-RadRadiance 2100 confocal imaging system using LaserSharp software. Quantificationof cell numbers was performed using the Cell Counter plug-in of ImageJ.

StatisticsComparison between quantitative data of two groups was done by two-tailedStudent’s t-test following an F-test using statistical software (NCSS, Kaysville, UT).For multiple comparisons, Fisher’s LSD was performed after ANOVA.

The authors thank T. Van de Putte for advice on the isolation andculture of preimplantation embryos, L. Le Cam for helpful suggestionson blastocyst manipulations and immunofluorescent staining, L.Schoonjans for providing ES cell medium, M. Depypere, I. Stockmans,R. Van Looveren and M. Van Camp for technical assistance. This workwas supported by grants from the Fund for Scientific Research Flanders(FWO; G.0508.05 and G.0587.09) and fellowships from the Institutefor the Promotion of Innovation through Science and Technology inFlanders (IWT) (A.V.) and FWO (T.R.).

Supplementary material available online athttp://jcs.biologists.org/cgi/content/full/123/19/3244/DC1

ReferencesAntonio, C., Ferby, I., Wilhelm, H., Jones, M., Karsenti, E., Nebreda, A. R. and

Vernos, I. (2000). Xkid, a chromokinesin required for chromosome alignment on themetaphase plate. Cell 102, 425-435.

Artus, J. and Cohen-Tannoudji, M. (2008). Cell cycle regulation during early mouseembryogenesis. Mol. Cell. Endocrinol. 282, 78-86.

Babu, J. R., Jeganathan, K. B., Baker, D. J., Wu, X., Kang-Decker, N. and vanDeursen, J. M. (2003). Rae1 is an essential mitotic checkpoint regulator that cooperateswith Bub3 to prevent chromosome missegregation. J. Cell Biol. 160, 341-353.

Bachvarova, R. and De Leon, V. (1980). Polyadenylated RNA of mouse ova and loss ofmaternal RNA in early development. Dev. Biol. 74, 1-8.

Blower, M. D., Nachury, M., Heald, R. and Weis, K. (2005). A Rae1-containingribonucleoprotein complex is required for mitotic spindle assembly. Cell 121, 223-234.

Brito, D. A. and Rieder, C. L. (2006). Mitotic checkpoint slippage in humans occurs viaCyclin B destruction in the presence of an active checkpoint. Curr. Biol. 16, 1194-1200.

Casanova, C. M., Rybina, S., Yokoyama, H., Karsenti, E. and Mattaj, I. W. (2008).Hepatoma up-regulated protein is required for chromatin-induced microtubule assemblyindependently of TPX2. Mol. Biol. Cell 19, 4900-4908.

Cheung, P., Allis, C. D. and Sassone-Corsi, P. (2000). Signaling to chromatin throughhistone modifications. Cell 103, 263-271.

Ciemerych, M. A. and Sicinski, P. (2005). Cell cycle in mouse development. Oncogene24, 2877-2898.

Funabiki, H. and Murray, A. W. (2000). The Xenopus chromokinesin Xkid is essentialfor metaphase chromosome alignment and must be degraded to allow anaphasechromosome movement. Cell 102, 411-424.

Gerdes, J., Lemke, H., Baisch, H., Wacker, H. H., Schwab, U. and Stein, H. (1984).Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined bythe monoclonal antibody Ki-67. J. Immunol. 133, 1710-1715.

Groen, A. C., Maresca, T. J., Gatlin, J. C., Salmon, E. D. and Mitchison, T. J. (2009).Functional overlap of microtubule assembly factors in chromatin-promoted spindleassembly. Mol. Biol. Cell 20, 2766-2773.

Kalab, P. and Heald, R. (2008). The RanGTP gradient-a GPS for the mitotic spindle. J.Cell Sci. 121, 1577-1586.

Kalab, P., Pralle, A., Isacoff, E. Y., Heald, R. and Weis, K. (2006). Analysis of aRanGTP-regulated gradient in mitotic somatic cells. Nature 440, 697-701.

Kirschner, M. and Mitchison, T. (1986). Beyond self-assembly: From microtubules tomorphogenesis. Cell 45, 329-342.

Koffa, M. D., Casanova, C. M., Santarella, R., Köcher, T., Wilm, M. and Mattaj, I.W. (2006). HURP is part of a Ran-dependent complex involved in spindle formation.Curr. Biol. 16, 743-754.

Lauss, M., Kriegner, A., Vierlinger, K., Visne, I., Yildiz, A., Dilaveroglu, E. andNoehammer, C. (2008). Consensus genes of the literature to predict breast cancerrecurrence. Breast Cancer Res. Tr. 110, 235-244.

Le Cam, L., Lacroix, M., Ciemerych, M. A., Sardet, C. and Sicinski, P. (2004). TheE4F protein is required for mitotic progression during embryonic cell cycles. Mol. Cell.Biol. 24, 6467-6475.

Levesque, A. A. and Compton, D. A. (2001). The chromokinesin Kid is necessary forchromosome arm orientation and oscillation, but not congression, on mitotic spindles.J. Cell Biol. 154, 1135-1146.

Li, T., Inoue, A., Lahti, J. M. and Kidd, V. J. (2004). Failure to proliferate and mitoticarrest of CDK11p110/p58-null mutant mice at the blastocyst stage of embryonic celldevelopment. Mol. Cell. Biol. 24, 3188-3197.

Maes, C., Stockmans, I., Moermans, K., Van Looveren, R., Smets, N., Carmeliet, P.,Bouillon, R. and Carmeliet, G. (2004). Soluble VEGF isoforms are essential forestablishing epiphyseal vascularization and regulating chondrocyte development andsurvival. J. Clin. Invest. 113, 188-199.

Jour

nal o

f Cel

l Sci

ence


Maes, C., Coenegrachts, L., Stockmans, I., Daci, E., Luttun, A., Petryk, A.,Gopalakrishnan, R., Moermans, K., Smets, N., Verfaillie, C. M. et al. (2006).Placental growth factor mediates mesenchymal cell development, cartilage turnover,and bone remodeling during fracture repair. J. Clin. Invest. 116, 1230-1242.

Murphy, K., Carvajal, L., Medico, L. and Pepling, M. (2005). Expression of Stat3 ingerm cells of developing and adult mouse ovaries and testes. Gene Expr. Patterns 5,475-482.

Musacchio, A. and Salmon, E. D. (2007). The spindle-assembly checkpoint in space andtime. Nat. Rev. Mol. Cell Biol. 8, 379-393.

O’Connell, C. B., Loncarek, J., Hergert, P., Kourtidis, A., Conklin, D. S. andKhodjakov, A. (2008). The spindle assembly checkpoint is satisfied in the absence ofinterkinetochore tension during mitosis with unreplicated genomes. J. Cell Biol. 183,29-36.

Ohsugi, M., Adachi, K., Horai, R., Kakuta, S., Sudo, K., Kotaki, H., Tokai-Nishizumi,N., Sagara, H., Iwakura, Y. and Yamamoto, T. (2008). Kid-mediated chromosomecompaction ensures proper nuclear envelope formation. Cell 132, 771-782.

Raemaekers, T., Ribbeck, K., Beaudouin, J., Annaert, W., Van Camp, M., Stockmans,I., Smets, N., Bouillon, R., Ellenberg, J. and Carmeliet, G. (2003). NuSAP, a novelmicrotubule-associated protein involved in mitotic spindle organization. J. Cell Biol.162, 1017-1029.

Ribbeck, K., Groen, A. C., Santarella, R., Bohnsack, M. T., Raemaekers, T., Kocher,T., Gentzel, M., Gorlich, D., Wilm, M., Carmeliet, G. et al. (2006). NuSAP, a mitoticRanGTP target that stabilizes and cross-links microtubules. Mol. Biol. Cell 17, 2646-2660.

Ribbeck, K., Raemaekers, T., Carmeliet, G. and Mattaj, I. W. (2007). A role forNuSAP in linking microtubules to mitotic chromosomes. Curr. Biol. 17, 230-236.

Rieder, C. L. and Maiato, H. (2004). Stuck in division or passing through: what happenswhen cells cannot satisfy the spindle assembly checkpoint. Dev. Cell 7, 637-651.

Schmidt, M., Budirahardja, Y., Klompmaker, R. and Medema, R. H. (2005). Ablationof the spindle assembly checkpoint by a compound targeting Mps1. EMBO Rep. 6, 866-872.

Schoonjans, L., Kreemers, V., Danloy, S., Moreadith, R. W., Laroche, Y. and Collen,D. (2003). Improved generation of germline-competent embryonic stem cell lines frominbred mouse strains. Stem Cells 21, 90-97.

Silljé, H. H. W., Nagel, S., Korner, R. and Nigg, E. A. (2006). HURP is a Ran-importinbeta-regulated protein that stabilizes kinetochore microtubules in the vicinity ofchromosomes. Curr. Biol. 16, 731-742.

Tokai, N., Fujimoto-Nishiyama, A., Toyoshima, Y., Yonemura, S., Tsukita, S., Inoue,J. and Yamamoto, T. (1996). Kid, a novel kinesin-like DNA binding protein, islocalized to chromosomes and the mitotic spindle. EMBO J. 15, 457-467.

Tokai-Nishizumi, N., Ohsugi, M., Suzuki, E. and Yamamoto, T. (2005). Thechromokinesin kid is required for maintenance of proper metaphase spindle size. Mol.Biol. Cell 16, 5455-5463.

Tsai, C. Y., Chou, C. K., Yang, C. W., Lai, Y. C., Liang, C. C., Chen, C. M. and Tsai,T. F. (2008). Hurp deficiency in mice leads to female infertility caused by an implantationdefect. J. Biol. Chem. 283, 26302-26306.

Weaver, B. A. A. and Cleveland, D. W. (2005). Decoding the links between mitosis,cancer, and chemotherapy: The mitotic checkpoint, adaptation, and cell death. CancerCell 8, 7-12.

Wong, J. and Fang, G. (2006). HURP controls spindle dynamics to promote properinterkinetochore tension and efficient kinetochore capture. J. Cell Biol. 173, 879-891.

Yokoyama, H., Gruss, O. J., Rybina, S., Caudron, M., Schelder, M., Wilm,M., Mattaj, I. W. and Karsenti, E. (2008). Cdk11 is a RanGTP-dependentmicrotubule stabilization factor that regulates spindle assembly rate. J. Cell Biol.180, 867-875.

Jour

nal o

f Cel

l Sci

ence

nusap is essential for chromatin-induced spindle formation … · 2011-03-08 · detrimental to...

Documents