The spindle checkpoint is a surveillance mechanism that regulates the metaphase-anaphase transition during somatic cell division through inhibition of the APC/C ensuring proper chromosome segregation. We show that the conserved spindle checkpoint protein BubR1 is required during early embryonic development. BubR1 is maternally provided and localises to kinetochores from prophase to metaphase during syncytial divisions similarly to somatic cells. To determine BubR1 function during embryogenesis, we generated a new hypomorphic semi-viable female sterile allele. Mutant females lay eggs containing undetectable levels of BubR1 show early developmental arrest,abnormal syncytial nuclear divisions, defects in chromosome congression,premature sister chromatids separation, irregular chromosome distribution and asynchronous divisions. Nuclei in BubR1 mutant embryos do not arrest in response to spindle damage suggesting that BubR1 performs a checkpoint function during syncytial divisions. Furthermore, we find that in wild-type embryos BubR1 localises to the kinetochores of condensed polar body chromosomes. This localisation is functional because in mutant embryos, polar body chromatin undergoes cycles of condensation-decondensation with additional rounds of DNA replication. Our results suggest that BubR1 is required for normal synchrony and progression of syncytial nuclei through mitosis and to maintain the mitotic arrest of the polar body chromosomes after completion of meiosis.

The first stages of Drosophila embryonic development occur within a syncytium into which 10 synchronous and three meta-synchronous nuclear divisions take place. These events are under the control of maternally provided protein and mRNA before cellularisation(Edgar et al., 1994; Foe et al., 1993). During each of the first 10 syncytial cycles, nuclei alternate between S and M phases without intervening gap phases. Until cycle 8, Cdk1 activity and cyclin B levels do not oscillate and the cycles are driven autonomously by the mitotic apparatus rather than by mitotic regulators(Edgar et al., 1994). Subsequently, the cycles gradually slowdown becoming meta-synchronous and from cycles 8 to 13 depend upon cyclin B activity. It has been suggested that the delays induced by cell cycle checkpoints would not be compatible with the rapid and synchronous syncytial divisions and damaged nuclei are though to be eliminated by a mechanism that aims to preserve normal early development(Edgar et al., 1994; Foe et al., 1993).

Current studies of the mitotic checkpoint proteins Mad2 and Bub3 indicate that spindle checkpoint activity is required for proper chromosome segregation within the epiblast of the mouse embryo at the time of rapid nuclear proliferation (Dobles et al.,2000; Kalitsis et al.,2000). In the Drosophila syncytial embryo, global perturbations such as blocking DNA replication, inducing extensive DNA damaged, or blocking spindle assembly are known to arrest or delay mitosis throughout the syncytium (Callaini et al.,1994; Debec et al.,1996; Fogarty et al.,1997; Raff and Glover,1988). Interestingly, embryonic expression of a non-degradable form of cyclin B or increase maternal cyclin B contribution can affect the global behaviour of nuclear cycles (Ji et al., 2002; Su et al.,1998). In addition, syncytial mutant embryos for the mitotic checkpoint protein Mps1 fail to arrest in mitosis in response to hypoxia(Fischer et al., 2004). It has been recently proposed that embryogenesis between mammals and other metazoans such as fly or frog should be aligned not from fertilisation but from the rapid embryonic cell cycles (O'Farrell et al., 2004). Thus, mammalian peri-gastrulation cell cycles should be homologous to the rapid cleavages cycles of other metazoans, such as syncytial cycles in flies. Therefore, mitotic checkpoint proteins could have an important and conserved role in monitoring rapid embryonic cell divisions in different and evolutionary distant metazoans.

Mitotic checkpoint proteins detect microtubule attachment and/or tension applied across kinetochore pairs to ensure that chromatids separate during mitosis only when kinetochores form stable bipolar microtubule attachments(Musacchio and Hardwick,2002). Current studies indicate that the well-conserved Bub and Mad proteins play an essential role in the spindle checkpoint function. These proteins inhibit the activity of the anaphase promoting complex or cyclosome(APC/C) until microtubule attachment/tension is achieved for all chromosomes(Peters, 2002; Zachariae and Nasmyth, 1999). BubR1 is an essential spindle checkpoint protein conserved among higher eukaryotic organisms (Chan et al.,1998; Chen, 2002). Recently, it has been shown that the previously identified Bub1 protein in Drosophila (Basu et al.,1999) should be reclassified as BubR1(Logarinho et al., 2004). In Drosophila somatic cells, BubR1 loss of function mutation causes an accelerated transit through mitosis which leads to abnormal chromosome segregation and apoptosis (Basu et al.,1999).

In order to analyse whether BubR1 is required during the nuclear proliferation at the syncytial stages in Drosophila embryos, we characterised a new semi-viable female sterile hypomorphic allele(bubR1Rev1) that was produced by imprecise excision of a previously characterised P element insertion(Basu et al., 1999). In embryos from homozygous bubR1Rev1 mutant females, syncytial nuclei show defects similar to those observed in neuroblasts homozygous for the bubR11 allele and do not arrest in mitosis in response to microtubule depolymerisation. In addition, whereas in wild-type embryos the 10 first syncytial nuclear divisions are synchronous, embryos from homozygous bubR1Rev1 mutant females show severe nuclear de-synchronisation. Furthermore, BubR1 accumulates at the kinetochores of the highly condensed polar body chromosomes. Analysis of fertilised embryos from homozygous bubR1Rev1 mutant females indicates that BubR1 is required to maintain the condensed conformation of polar body chromosomes and to exclude it from the mitotic oscillation.

Drosophila stocks

bubR1Rev1 was obtained by imprecise excision of the P-element in 1(2)K03113 (Basu et al., 1999; Cooley et al.,1988). Histone-GFP transgene 62A was used for live imaging of syncytial nuclei and polar body (Clarkson and Saint, 1999). Fertilised mutant embryos were obtained by crossing homozygous bubR1Rev1 females with wild-type males(w1118).

Southern blotting

Genomic DNA was extracted according to Roberts(Roberts, 1986) from adult flies frozen in liquid nitrogen and digested with HindIII and EcoRI (New England Biolabs). Southern blots were done according to Sambrook et al. (Sambrook et al.,1989) and probed with a complete cDNA (see Fig. 2A).

Western blotting

Third larval instar brains, adult female ovaries and embryos were dissected in 1×PBS, transferred to sample buffer [50 mM Tris (pH 6.8), 100 mM DTT,2% SDS, 0,1% Bromophenol Blue and 10% glycerol] heated 3 minutes at 95°C before SDS-page (7.5%) migration. Proteins were transferred onto Immobilon-P membrane (Schleicher & Schuell) and probed with anti-DmBubR1 rabbit polyclonal antibodies (Rb666 Ab)(Logarinho et al., 2004),diluted 1:2500 in PBST (1× PBS and 0.2% Triton-100) containing 1% BSA(Sigma) and 2% dried non-fat milk. Anti-α-tubulin was detected using mouse mAb B512 (Sigma-Aldrich), diluted 1:2000. Secondary antibodies conjugated to HRP (Vector, UK) were used according to the manufacturer's instructions. Detection of antibody signals was performed with the ECL system(Amersham).

Immunofluorescence of embryos

Embryos were collected and aged at 25°C and processed as previously(Sullivan et al., 2000). Primary antibodies were: anti-BubR1 (Rb666) antibody pre-adsorbed on wild-type 0-2 hours embryos at a 1:10 dilution and used at 1:1000 dilution;anti-α-tubulin (mouse mAb B512), used at 1:3000 (Sigma-Aldrich);anti-phospho H3 rabbit polyclonal, used at 1:500 (Upstate Biotechnology);anti-CID chicken polyclonal, used at 1:1000(Blower and Karpen, 2001);anti-CNN rabbit polyclonal, used at 1:1000(Heuer et al., 1995);anti-BrdU mouse monoclonal primary antibody, used at 1:50 (Sigma-Aldrich). Secondary antibodies were: anti-rabbit Alexa Fluor 488 used at 1:2000(Molecular Probes), anti-mouse Alexa Fluor 568 used at 1:2000 (Molecular Probes) or anti-chicken Cy5 used at 1:200 (Jackson). Samples were mounted in Vectashield (Vector, UK) containing 1 μg/ml of DAPI or 1.0 μg/ml propidium iodide. Observation and images were obtained using Zeiss Axiovert 200M microscope (Zeiss, Germany) or a Leica TCS SP2 AOBS Confocal Microscope(Leica Microsystems, Heidelberg). Images were deconvolved and processed with Adobe Photoshop 7.0. For colchicine treatment, embryos were permeabilised in n-heptane, containing 250 μM colchicine in 1× PBS for 30 minutes before fixation. For BrdU incorporation, the embryos were permeabilised for 4 minutes in n-octane followed by 8-10 minutes BrdU pulse at 1 mg/ml in 1×PBS before fixation.

Live imaging

Chromosomes in wild-type and bubR1Rev1 embryos were visualised by time-lapse recording of the Histone-GFP transgene. Images were taken every 10 or 15 seconds using a BioRad MRC600 Confocal Microscopy with two scans per image. To determine the polar body Histone-GFP signal intensity over time, we repeated three measurements of a restricted area of 16 pixels taken from regions with higher intensity signal and use ImageJ software to obtain an average and standard deviation for each image. Using the same measurements, we calculated the average standard deviation pixel intensity of the recorded period of time.

Immunolocalisation of BubR1 during embryogenesis

BubR1 localisation has been described in somatic cells of different organisms, including Drosophila(Basu et al., 1999), but it has never been characterised during the syncytial stages of Drosophilaembryos. Accordingly, we analysed the localisation of BubR1 at 0-180 minutes after egg laying (AEL) in wild-type embryos(Fig. 1). BubR1 was first detected at kinetochores of mitotic chromosomes during prophase and became stronger in prometaphase. During metaphase, BubR1 kinetochore staining decrease significantly and during anaphase it was undetectable(Fig. 1A-J). BubR1 localisation follows the same pattern during early (Fig. 1A-E) or late (Fig. 1F-J) cycles. Co-localisation with anti-CID antibody shows that BubR1 accumulates at kinetochores of condensed chromosomes(Fig. 1K-K′′′). In addition, we detected BubR1 on spindle microtubules at metaphase (inset in Fig. 1D and Fig. 1L,L′′′). Overall, our results indicate that BubR1 protein has a dynamic pattern of kinetochore localisation, consistent with its role in monitoring microtubule kinetochore interaction during the syncytial nuclear cycles of Drosophila embryos.

Molecular and phenotypic characterisation of a hypomorphic bubR1 allele

To study the role of BubR1 in embryogenesis, we characterised a new hypomorphic semi-viable female sterile bubR1 allele(bubR1Rev1) obtained by imprecise excision of the lethal P element in bubR1 (Basu et al., 1999; Cooley et al.,1988). bubR1Rev1 was found to be semi-lethal and ∼10% of the homozygous reach adult stages. Homozygous mutant females are sterile and lay fewer eggs than wild-type flies. Molecular analysis indicated that the P element present in bubR1Rev1 has an internal deletion of ∼2 kb (Fig. 2A,B). PCR with a combination of appropriate primers of the ends of the P-element confirm this conclusion(Fig. 2A and data not shown). In the course of the reversion screen, we also identified a perfect excision allele (rev74) for which southern blot, homozygous viability and fertility were indistinguishable from the wild-type control(Fig. 2A,B and data not shown),indicating that the semi-lethality of bubR1Rev1 homozygote adults and the lethality and phenotype of embryos derived from bubR1Rev1/bubR1Rev1 females results from the P element present in this line.

BubR1 protein levels in wild-type and mutant alleles at various developmental times were determined by western blotting(Fig. 2C). High levels of BubR1 were detected in wild-type third instar larval brains and ovaries. In bubR11/bubR11 brains, no protein is detected,while in bubR1Rev1/bubR1Rev1 brains and ovaries there is a strong decrease in protein levels. BubR1 protein in ovaries results from somatic expression during the cell proliferation stages, the germline expression during cyst formation and the maternal contribution (data not shown). BubR1 could also be detected by western blot of embryo extracts at 30-90 minutes AEL using two wild-type embryos(Fig. 2C). However, in embryos derived from bubR1Rev1 homozygous females, the level of BubR1 was undetectable (Fig. 2C).

We also analyzed the mitotic behaviour and spindle checkpoint response of neuroblasts carrying different bubR1 allelic combinations(Fig. 2D). bubR11/bubR11,bubR11/bubR1Rev1 and bubR1Rev1/bubR1Rev1 mutant neuroblasts show a significant decrease in prophase-metaphase mitotic figures and a significant increase in sister chromatid separation(Fig. 2D). Furthermore, bubR1Rev1/bubR1Rev1 neuroblasts fail to show a mitotic arrest in response to spindle damage (compare Fig. 2E with 2F). Accordingly,we suggest that bubR1Rev1 is a new hypomorphic allele and embryos derived from bubR1Rev1 homozygous females,fertilised by wild-type males, will, from now on, be designated as bubR1Rev1 embryos.

Analysis of nuclear divisions in bubR1Rev1embryos

To determine whether BubR1 was required for nuclear divisions during early embryogenesis, we analysed bubR1Rev1 embryos(Fig. 3A-J). Quantitative analysis of fixed bubR1Rev1 embryos (n=80) at 120 to 150 minutes AEL showed that ∼25% of the embryos stop development before cycle 10 (when nuclei reach the cortex); 62% developed up to cycle 10-13,although nuclei are not found evenly distributed throughout the cortex and they did not cellularise; 10% initiated cellularisation; and only 3%gastrulate but show pyknotic nuclei (Fig. 3H). bubR1Rev1 embryos accumulate defective nuclei throughout the syncytial stages and after cycle 10, most had many regions without cortical nuclei, extensive areas contained micronuclei,abnormally larger sized nuclei or paired nuclei with extensive DNA bridges(Fig. 3A-G). bubR1Rev1 embryos also showed abnormal mitotic progression, including precocious sister chromatid separation, lagging chromosomes, DNA bridges, irregular chromosome segregation and aneuploidy(Fig. 3E-G). A similar range of abnormalities was also observed in nuclei within mitotic domains of cellularised embryos (Fig. 3H).

To obtain a more accurate description of mitotic progression in bubR1Rev1 embryos, we quantified the relative frequency of mitotic figures in 30-90 minutes AEL embryos after fixation(Fig. 3K). We also visualised in vivo the behaviour of chromosomes using a GFP-Histone transgene(Clarkson and Saint, 1999)(Fig. 3I,J; see Movies 1 and 2 in the supplementary material). Quantification of the mitotic figures on fixed material reveals that, whereas in wild-type embryos all nuclei undergo synchronous division, almost half of the bubR1Rev1 embryos show asynchronous nuclear behaviour following cycle 5. Among the synchronous embryos, we observed a significant decrease in interphase nuclei and we never observed proper metaphase alignment. Live recording shows that the divisions in mutant embryos are highly asynchronous compared with the meta-synchronous divisions in wild-type controls (see Movies 1 and 2 in the supplementary material). In bubR1Rev1 embryos, nuclei in anaphase(Fig. 3J at 9 minutes), in prophase (Fig. 3J at 9 minutes)and nuclei initiating sister chromatid separation(Fig. 3J at 9 minutes) can be found at the same time. At later times, some nuclei show unequal chromosome segregation with lagging chromatids (Fig. 3J at 11 minutes) and chromatin bridges(Fig. 3J at 14 minutes). In addition, during mitotic progression, cytoplasmic movements are strongly affected in mutant embryos (compare Movie 1 with 2 in the supplementary material). Taken together these results indicate that BubR1 activity is required during early embryonic development to maintain nuclear synchrony and cytoplasmic flux, and to prevent abnormal chromosome segregation.

Centrosome replication in bubR1Rev1 embryos

Coordination between local events of centrosome duplication, microtubule dynamics, changes in nuclear structure and positioning are essential for the correct spatial organisation of the early embryo. It has been recently proposed that regulation of DNA replication/segregation and centrosome function are coupled during the late syncytial divisions through a checkpoint mechanism designed to eliminate abnormal nuclei from the embryo(Fogarty et al., 1997; Sibon et al., 2000; Takada et al., 2003). However,an alternative model suggests that the grape gene coordinates nuclear envelope breakdown with chromosome condensation(Yu et al., 2000). In bubR1Rev1 embryos, in vivo analysis of the syncytial nuclear divisions show abnormal mitotic progression and cytoplasmic flux,suggesting that the coordination between these events was severely perturbed(see Movie 2 in the supplementary material). Cortical nuclei in the wild-type blastoderm embryos show a correlation between centrosomes pairs and nuclei(Fig. 4A). However, in bubR1Rev1 embryos, more than 90% of the embryos observed at a similar stage (cycles 10-13) showed cortical areas devoid of nuclei with abundant free centrosomes probably left at the cortex after nuclei had fallen into the embryo (Fig. 4C). In addition, during the syncytial stage many nuclei appear to loose their attachment to centrosomes as we observed centrosomes migration to the cortex in the absence of DNA (Fig. 4B).

As free centrosomes can be detected in bubR1Rev1embryos, we analysed the relation between the nuclear cycle and the centrosome number over time. For this, we quantify centrosomes in individual bubR1Rev1 and wild-type embryos collected throughout the syncytial stages and plotted centrosome numbers by unit of space as a function of developmental time (Fig. 4D). Because at the time of cellularisation the centrosome number in bubR1Rev1 embryos is 30-40% higher relative to the control, our results indicate that centrosomes initiate an extra round of replication, which is never fully completed. We also found a small number of embryos (<5%) that had very few nuclei and centrosomes, suggesting that development had stopped completely.

Owing to the abnormal mitotic progression of bubR1Rev1embryos, we looked at the ability of centrosomes to nucleate microtubules and to organise proper mitotic spindles. Although all the centrosomes we observed were able to nucleate microtubule asters, a detailed analysis showed several abnormalities in spindle organisation and distribution with strong variation within each embryo (Fig. 4E-H). Spindles were abnormally distributed within the syncytium and some spindles were shorter in length (Fig. 4Earrowhead and arrow). In other cases, we observed fused spindles while neighbouring nuclei display an almost normal anaphase onset with traces of chromosome bridges (Fig. 4F,arrow and arrowhead). Free centrosomes were observed to replicate asynchronously from the cycle of DNA replication and were able to nucleate microtubules (arrowhead and arrow in Fig. 4G,H). In addition, chromosomes from neighbouring nuclei often shared the same spindle pole (asterisk in Fig. 4K,H). These results indicate that the abnormal mitotic progression observed in bubR1Rev1 embryos uncoupled DNA and centrosomes cycle.

Mitotic progression in bubR1Rev1 embryos after spindle damage

Previous findings in somatic cells of many organisms showed that treatment with colchicine, a microtubule depolymerising agent, causes cells to arrest in a prometaphase-like state dependent upon BubR1 accumulation at kinetochores(Basu et al., 1999; Chan et al., 1998). However,although colchicine treatment of Drosophila embryos during syncytial stages causes a mitotic arrest (Zalokar and Erk, 1976), the role of spindle checkpoint proteins in this arrest has never been analyzed. Thus, 0-2 hours AEL, wild-type and bubR1Rev1 embryos were treated with colchicine at a concentration of 250 μM for 30 minutes and the mitotic arrest was monitored by immunodetection of Histone H3 phosphorylated at S10, which is diagnostic of mitotic chromosome condensation (Fig. 5). For quantitative analysis, we divided embryos into three categories, cycles 3 to 6, cycles 7 to 9 and cycles 10 to 13. The results show that after colchicine incubation, all nuclei in wild-type embryos arrest in a prometaphase-like state, with Phospho-Histone H3-positive chromatin(Fig. 5A,A′′,E) and strong accumulation of BubR1 at kinetochores(Fig. 5C). However, in bubR1Rev1 embryos, we observed a decrease in the proportion of nuclei showing mitotic arrest over time. Although 63% of the embryos at cycle 3-6 arrest, only 40% at cycle 7-9 and 23% at cycle 10-13 are Phospho-Histone H3 positive (Fig. 5E). The remaining embryos had mostly post-mitotic nuclei,Phospho-Histone H3 negative (Fig. 5B) with no BubR1 accumulation at kinetochores(Fig. 5D). Accordingly, our results demonstrate that BubR1 checkpoint activity is required during embryogenesis to allow mitotic arrest in response to spindle damage. In addition, our results suggest that BubR1 maternal supply from bubR1Rev1 allele although undetectable by western blotting of individual embryos is sufficient to support partial checkpoint activity during early nuclear cycles but as development proceeds it becomes limiting.

Polar body organisation in bubR1Rev1 embryos

The mature oocyte is arrested in metaphase of meiosis I and meiosis activation occurs during passage through the oviduct. Soon after activation,meiosis is resumed, even if fertilisation has not occurred. After meiosis II,the inner-most female pro-nucleus participates in zygotic development, while the three remaining haploid nuclei stay at the egg cortex and form the polar body after one round of replication and DNA condensation(Fig. 6A′′)(Foe et al., 1993). The polar body does not participate in further nuclear division cycles and maintains a mitotic-arrest configuration until cycle 8-10, when it falls into the interior of the embryo. In wild-type unfertilised (data not shown) and fertilised embryos, BubR1 accumulates strongly at the kinetochores of the polar body chromosomes where it co-localises with the centromere marker CID(Fig. 6A,A′′′). In order to assess whether organisation and/or maintenance of the mitotic arrest is dependent upon BubR1 localisation at polar body kinetochores, we analyzed its organisation in bubRRev1 embryos. In unfertilised and fertilised embryos,we failed to detect any BubR1 accumulation at polar body kinetochores(Fig. 6B,C) and instead of a well-organised structure, we observed large masses of decondensed DNA with a spherical interphase-like conformation(Fig. 6B′′,C′′), presenting an increase number of CID-positive dots and suggesting that they undergo extra rounds of DNA replication (compare Fig. 6A′ with 6C′).

Subsequently, using Phospho-Histone H3 antibodies, we quantified the state of condensation of polar bodies in 0-90 minutes AEL wild-type and bubR1Rev1 embryos (Fig. 6H). In wild-type embryos either unfertilised (data not shown) or fertilised, polar bodies are always Phospho-Histone H3 positive(Fig. 6D′,H). However, in unfertilised bubR1Rev1 embryos, we found that in 80% of the embryos (n=30) (Fig. 6H), polar bodies are Phospho-Histone H3 positive and their DNA appear condensed but without a normal chromosome organisation (compare Fig. 6D with 6E), whereas in fertilised embryos (n=41) the chromatin appeared to follow a cycle of condensation-decondensation, which seems to be in phase with the division cycles of the neighbouring nuclei (Fig. 6F,H). The cycle in polar body condensation-decondensation was not always complete and some Phospho-Histone H3 signal can be detected within the chromatin of decondensed polar body (Fig. 6G′).

Our observations of condensation-decondensation cycles, variation in Phospho-Histone H3 signal and increase number of CID-positive dots lead us to investigate whether in the absence of BubR1, the polar body became under the control of the mitotic oscillator that drives the nuclear cycles during the syncytial stages. Using a Histone-GFP transgene and time-lapse confocal microscopy, we recorded for a period of 40-80 minutes the polar body structure in wild type and bubR1Rev1 fertilised embryos collected at 0-30 min AEL (Fig. 7A,B, Movies 3 and 4 in the supplementary material). Then we quantified the relative Histone-GFP signal intensity over time(Fig. 7C) and calculated the pixel intensity±s.d. for the recorded period of time(Fig. 7D). As expected, in fertilised wild-type embryos the polar body always maintains a mitotic-arrest configuration during the recording period(Fig. 7A). The relative Histone-GFP signal over time remains constant with a low variation in pixel intensity (Fig. 7C,D). By contrast, bubR1Rev1 embryos display cycles of chromatin condensation and decondensation during the recorded period of time(Fig. 7B; Movie 4 in the supplementary material). Quantification of the relative Histone-GFP signal shows oscillation over time with a strong variation in pixel intensity for the recorded period of time, suggesting extra cycles of DNA replication(Fig. 7C,D). Indeed, although in wild type embryos, the polar body never incorporates BrdU, in bubR1Rev1 embryos, BrdU signal can be easily detected within the polar body chromosomes at 30-90 minutes AEL(Fig. 7D,E). Taken together,our results show that BubR1 is essential to maintain the mitotic arrest of the polar body and to prevent extra rounds of DNA replication.

BubR1 is required during early embryonic development

Western blot analysis and immunodetection on stage 10 egg chambers revealed a maternal contribution for the highly conserved spindle checkpoint protein BubR1 to ensure normal progression through early stages of embryo development prior to the onset of zygotic gene expression. During syncytial divisions,BubR1 localisation correspond to previous description in Drosophilalarval neuroblast and S2 cells (Basu et al., 1999), except for BubR1 localisation on spindle microtubules at metaphase. Dynein-dependent redistribution of BubR1 and other checkpoint proteins was previously observed in mammalian cells and Drosophilaembryos (Basto et al., 2004; Howell et al., 2001; Wojcik et al., 2001). Accordingly, our observations of BubR1 localisation along spindle microtubules at metaphase suggests that during syncytial division, it could be partially removed from kinetochores via microtubule transport. As we only observe BubR1 spindle localisation in one third of the metaphases, we suspect that the removal and redistribution of BubR1 is likely to occur within a very short period of time at this stage. Further studies using live imaging in different mitotic cell populations in Drosophila will elucidate if our observation reflects different mechanism of BubR1 checkpoint regulation between the syncytial and somatic divisions.

In order to study BubR1 checkpoint activity during the syncytial nuclear divisions, we characterised bubR1Rev1, a new hypomorphic female sterile allele. Analysis of nuclear proliferation in bubR1Rev1 embryos shows that starting at cycle 4-5 all embryos present abnormal mitotic progression as described for bubR11 mutant allele(Basu et al., 1999). It has been previously shown that BubR1 is involved in chromosome alignment/congression (Ditchfield et al.,2003; Lampson and Kapoor,2005) and in the inhibition of APC/C activity by signalling lack of tension at the kinetochore (Logarinho et al., 2004). The lagging chromatids, DNA bridges and aneuploidy observed in bubR1Rev1 embryos indicate precocious anaphase onset without stable microtubule-kinetochore attachment and an increase local degradation of APC/C targets. Therefore, our observations of mitotic progression on fix and live embryos, in addition to the failure to delay mitosis exit in the presence of colchicine, support the notion that BubR1 spindle checkpoint activity is required during the rapid nuclear proliferation of Drosophila embryo. In the absence of BubR1, nuclear mitotic exit is initiated too rapidly, resulting in nuclear cycle asynchrony and developmental failure.

BubR1 is essential to maintain the mitotic arrest of the polar body

In wild-type Drosophila embryo, three out of the four haploid meiotic products form the polar body which is formed by Phospho-H3-positive condensed chromosomes containing radiating microtubules without centrosomes(Foe et al., 1993). Although it has been hypothesised that F-actin and myosin II are involved into pulling and anchoring the polar body to the embryonic cortex with the plus ends of microtubules attached to the kinetochores and the minus ends facing outwards(Foe et al., 2000), the establishment and maintenance of the polar body mitotic arrest imply that specific factors are required to exclude it from the mitotic oscillation as it occurs in the same cytoplasm where neighbouring nuclei undergo synchronous divisions. Surprisingly, our immunodetection study revealed that BubR1 accumulates at polar body kinetochores in unfertilised and fertilised wild-type embryo. In bubR1Rev1 embryos; the polar body fail to establish and maintain its structure. Instead, polar bodies undergo cycle of DNA condensation-decondensation in phase with the mitotic cycle of the neighbouring nuclei. Thus, our results strongly suggest that at least the maintenance of the mitotic arrest and the exclusion of the polar body from the mitotic oscillator activity in the cytoplasm are dependent on normal BubR1 levels. Accordingly, BubR1 localisation at the polar body kinetochores appears to have a dominant effect by allowing maintenance of the arrest.

Similarly, it has been shown that the conserved mitotic checkpoint protein Mps1 is require for polar body mitotic arrest(Fischer et al., 2004) and while BubR1 polar body localisation is affected by Mps1 mutation, its nuclear syncytial localisation remains unaffected. Moreover, gnu, png and plu gene products which form a multi-protein complex have been shown to be required for entry and exit into mitosis during early syncytial cycles in Drosophila embryos through Cyclin B stabilisation(Fenger et al., 2000; Lee et al., 2003; Zhang et al., 2004). Mutant alleles of these genes induce nuclei to undergo DNA replication in the absence of chromosome segregation and polar body de-condensation. Furthermore, it has been shown that increase Cyclin B levels in a png mutant genetic background can restore polar body chromosome condensation(Lee et al., 2001). These observations lead us to speculate a potential genetic interaction between BubR1, Mps1 and gnu-png-plu during the syncytial stages. However, it remains to be determined whether Mps1 and BubR1 are involved in the same pathway during early syncytial cycles and polar body structure. Moreover, while polar body structure differs between bubR1and gnu mutant embryos, it is possible that they interact genetically to establish and maintain polar body structure through local cyclin B stabilisation.

BubR1 checkpoint activity is required for synchrony of nuclear divisions

It has been recently proposed that syncytial nuclear proliferation should be divided in three phases: a first phase of synchronous proliferation (cycle 1 to 6), a second phase being Cdk1/cyclinB dependent with local variation in metaphase-anaphase duration (cycle 7 to 10), and a third phase of meta-synchronous divisions that is DNA damage checkpoint dependent and shows an increase in the duration of interphase and M phases(Ji et al., 2004). Within each embryo, the total nuclear cycle time between cycles 7 until 10 remains equal,but those located at the centre undergo prolonged metaphase, which is compensated by a shorter anaphase/telophase. These variations are regulated by local Cdk1/cyclin B activity and the total cycle length can be modified and influenced by variation in cyclin B maternal gene dose(Ji et al., 2002; Ji et al., 2004; Stiffler et al., 1999). It has been proposed that the meta-synchronisation observed between cycle 7-10 is solely induced by cytoplasmic flux and local oscillation in Cdk1/cyclin B activity (Ji et al., 2004) and that early cycles are driven by the dynamics of the mitotic apparatus(Edgar et al., 1994). However,our observations that in bubR1Rev1 embryos almost half of the embryos show extensive loss of nuclear synchrony as early as cycle 4-5,suggests that BubR1 checkpoint activity could be directly involved in the process of nuclear synchrony.

As all nuclei share a common cytoplasm, a key factor in the regulation of mitotic progression is the transduction of a global state to the local nuclear level so as to ensure proper synchrony. Although the abnormal mitotic progression in bubR1Rev1 embryos can be easily explained in terms of spindle checkpoint activity, the nuclear cycle asynchrony suggests that BubR1 regulates the mitotic apparatus at a local level by timing the proper progression of chromosome congression and anaphase onset. It has been proposed that higher Cdk1/cyclin B activity decreases microtubule stability and increases sister chromatids velocity at anaphase to maintain constant nuclear cycle length within the embryo (Ji et al., 2004; Stiffler et al.,1999). Our observations suggest that local variation in BubR1 checkpoint activity can provide a feedback mechanism to ensure proper chromosome segregation and local variation in the timing of metaphase/anaphase transition through regulation of microtubule attachment/tension at kinetochore pairs and APC/C inhibition. However, our analysis of local cyclin B levels by immunodetection did not provide conclusive results, probably owing to the severe abnormalities we observed in bubR1Rev1 embryos(data not shown). Furthermore, as an increase in cyclin B copy number induces local changes in nuclear progression and only a global embryonic phenotype at six extra copies (Ji et al.,2002; Ji et al.,2004), our observations lead us to speculate that variation in BubR1/cyclin B copy number should induce local variation in timing mitotic progression without affecting the total embryonic phenotype. Accordingly, a decrease in BubR1 protein level should result in a decrease in metaphase delay during cycle 7, as well as a reduced metaphase delay induced by higher cyclin B levels.

In summary, our analysis of the requirement of BubR1 indicates that during syncytial development the spindle checkpoint appears to operate at various levels. It ensures that nuclei respond to global perturbations by imposing a mitotic arrest especially during later cycles. BubR1 also appears to work at a local level during early and late cycles to ensure proper synchrony of nuclear divisions. In parallel, BubR1 is required to sustain the mitotic arrest of the polar body so as to exclude it from undergoing further rounds of DNA replication during embryonic development.

Supplementary material

We would like to thank G. Karpen and T. C. Kaufman for antibodies. We also thank Madalena Costa for help with colchicine treatment of embryos and all the members of the laboratory for comments and suggestions. D.P.-M. received a fellowship from l'ARC, France and N.M. was financed by a Training and Mobility grant from the European Union. The work in the laboratory of C. E. Sunkel is financed by grants from the Fundação para a Ciência e Tecnologia of Portugal and the European Union.

Basto, R., Scaerou, F., Mische, S., Wojcik, E., Lefebvre, C.,Gomes, R., Hays, T. and Karess, R. (
2004
). In vivo dynamics of the rough deal checkpoint protein during Drosophila mitosis.
Curr. Biol.
14
,
56
-61.
Basu, J., Bousbaa, H., Logarinho, E., Li, Z., Williams, B. C.,Lopes, C., Sunkel, C. E. and Goldberg, M. L. (
1999
). Mutations in the essential spindle checkpoint gene bub1 cause chromosome missegregation and fail to block apoptosis in Drosophila.
J. Cell Biol.
146
,
13
-28.
Blower, M. D. and Karpen, G. H. (
2001
). The role of Drosophila CID in kinetochore formation, cell-cycle progression and heterochromatin interactions.
Nat. Cell. Biol.
3
,
730
-739.
Callaini, G., Riparbelli, M. G. and Dallai, R.(
1994
). The distribution of cytoplasmic bacteria in the early Drosophila embryo is mediated by astral microtubules.
J. Cell Sci.
107
,
673
-682.
Chan, G. K., Schaar, B. T. and Yen, T. J.(
1998
). Characterization of the kinetochore binding domain of CENP-E reveals interactions with the kinetochore proteins CENP-F and hBUBR1.
J. Cell Biol.
143
,
49
-63.
Chen, R. H. (
2002
). BubR1 is essential for kinetochore localization of other spindle checkpoint proteins and its phosphorylation requires Mad1.
J. Cell Biol.
158
,
487
-496.
Clarkson, M. and Saint, R. (
1999
). A His2AvDGFP fusion gene complements a lethal His2AvD mutant allele and provides an in vivo marker for Drosophila chromosome behavior.
DNA Cell Biol.
18
,
457
-462.
Cooley, L., Kelley, R. and Spradling, A.(
1988
). Insertional mutagenesis of the Drosophila genome with single P elements.
Science
239
,
1121
-1128.
Debec, A., Kalpin, R. F., Daily, D. R., McCallum, P. D.,Rothwell, W. F. and Sullivan, W. (
1996
). Live analysis of free centrosomes in normal and aphidicolin-treated Drosophila embryos.
J. Cell Biol.
134
,
103
-115.
Ditchfield, C., Johnson, V. L., Tighe, A., Ellston, R., Haworth,C., Johnson, T., Mortlock, A., Keen, N. and Taylor, S. S.(
2003
). Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores.
J. Cell Biol.
161
,
267
-280.
Dobles, M., Liberal, V., Scott, M. L., Benezra, R. and Sorger,P. K. (
2000
). Chromosome missegregation and apoptosis in mice lacking the mitotic checkpoint protein Mad2.
Cell
101
,
635
-645.
Edgar, B. A., Sprenger, F., Duronio, R. J., Leopold, P. and O'Farrell, P. H. (
1994
). Distinct molecular mechanism regulate cell cycle timing at successive stages of Drosophila embryogenesis.
Genes Dev.
8
,
440
-452.
Fenger, D. D., Carminati, J. L., Burney-Sigman, D. L.,Kashevsky, H., Dines, J. L., Elfring, L. K. and Orr-Weaver, T. L.(
2000
). PAN GU: a protein kinase that inhibits S phase and promotes mitosis in early Drosophila development.
Development
127
,
4763
-4774.
Fischer, M. G., Heeger, S., Hacker, U. and Lehner, C. F.(
2004
). The mitotic arrest in response to hypoxia and of polar bodies during early embryogenesis requires Drosophila Mps1.
Curr. Biol.
14
,
2019
-2024.
Foe, V. E., Odell, G. M. and Edgar, B. A.(
1993
). Mitosis and morphogenesis in the Drosophila embryo: point and counterpoint. In
The Development of Drosophila
(ed. M. Bate and A. M. Arias), pp.
149
-300. New York:Cold Spring Harbor Laboratory Press.
Foe, V. E., Field, C. M. and Odell, G. M.(
2000
). Microtubules and mitotic cycle phase modulate spatiotemporal distributions of F-actin and myosin II in Drosophila syncytial blastoderm embryos.
Development
127
,
1767
-1787.
Fogarty, P., Campbell, S. D., Abu-Shumays, R., Phalle, B. S.,Yu, K. R., Uy, G. L., Goldberg, M. L. and Sullivan, W.(
1997
). The Drosophila grapes gene is related to checkpoint gene chk1/rad27 and is required for late syncytial division fidelity.
Curr. Biol.
7
,
418
-426.
Heuer, J. G., Li, K. and Kaufman, T. C. (
1995
). The Drosophila homeotic target gene centrosomin (cnn) encodes a novel centrosomal protein with leucine zippers and maps to a genomic region required for midgut morphogenesis.
Development
121
,
3861
-3876.
Howell, B. J., McEwen, B. F., Canman, J. C., Hoffman, D. B.,Farrar, E. M., Rieder, C. L. and Salmon, E. D. (
2001
). Cytoplasmic dynein/dynactin drives kinetochore protein transport to the spindle poles and has a role in mitotic spindle checkpoint inactivation.
J. Cell Biol.
155
,
1159
-1172.
Ji, J. Y., Haghnia, M., Trusty, C., Goldstein, L. S. and Schubiger, G. (
2002
). A genetic screen for suppressors and enhancers of the Drosophila cdk1-cyclin B identifies maternal factors that regulate microtubule and microfilament stability.
Genetics
162
,
1179
-1195.
Ji, J. Y., Squirrell, J. M. and Schubiger, G.(
2004
). Both cyclin B levels and DNA-replication checkpoint control the early embryonic mitoses in Drosophila.
Development
131
,
401
-411.
Kalitsis, P., Earle, E., Fowler, K. J. and Choo, K. H.(
2000
). Bub3 gene disruption in mice reveals essential mitotic spindle checkpoint function during early embryogenesis.
Genes Dev.
14
,
2277
-2282.
Lampson, M. A. and Kapoor, T. M. (
2005
). The human mitotic checkpoint protein BubR1 regulates chromosome-spindle attachments.
Nat. Cell. Biol.
7
,
93
-98.
Lee, L. A., Elfring, L. K., Bosco, G. and Orr-Weaver, T. L.(
2001
). A genetic screen for suppressors and enhancers of the Drosophila PAN GU cell cycle kinase identifies cyclin B as a target.
Genetics
158
,
1545
-1556.
Lee, L. A., Van Hoewyk, D. and Orr-Weaver, T. L.(
2003
). The Drosophila cell cycle kinase PAN GU forms an active complex with PLUTONIUM and GNU to regulate embryonic divisions.
Genes Dev.
17
,
2979
-2991.
Logarinho, E., Bousbaa, H., Dias, J. M., Lopes, C., Amorim, I.,Antunes-Martins, A. and Sunkel, C. E. (
2004
). Different spindle checkpoint proteins monitor microtubule attachment and tension at kinetochores in Drosophila cells.
J. Cell Sci.
117
,
1757
-1771.
Musacchio, A. and Hardwick, K. G. (
2002
). The spindle checkpoint: structural insights into dynamic signalling.
Nat. Rev. Mol. Cell. Biol.
3
,
731
-741.
O'Farrell, P. H., Stumpff, J. and Su, T. T.(
2004
). Embryonic cleavage cycles: how is a mouse like a fly?
Curr. Biol.
14
,
R35
-R45.
Peters, J. M. (
2002
). The anaphase-promoting complex: proteolysis in mitosis and beyond.
Mol. Cell
9
,
931
-943.
Raff, J. W. and Glover, D. M. (
1988
). Nuclear and cytoplasmic mitotic cycles continue in Drosophila embryos in which DNA synthesis is inhibited with aphidicolin.
J. Cell Biol.
107
,
2009
-2019.
Roberts, D. B. (
1986
). Basic Drosophila care and techniques. In
Drosophila: A Practical Approach
(ed. D. B. Roberts), pp.
1
-38. Oxford, England: IRL Press.
Sambrook, J., Fritsch, F. and Maniatis, T.(
1989
).
Molecular Cloning: A Laboratory Manual,2nd edn.
Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
Sibon, O. C., Kelkar, A., Lemstra, W. and Theurkauf, W. E.(
2000
). DNA-replication/DNA-damage-dependent centrosome inactivation in Drosophila embryos.
Nat. Cell. Biol.
2
,
90
-95.
Stiffler, L. A., Ji, J. Y., Trautmann, S., Trusty, C. and Schubiger, G. (
1999
). Cyclin A and B functions in the early Drosophila embryo.
Development
126
,
5505
-5513.
Su, T. T., Sprenger, F., DiGregorio, P. J., Campbell, S. D. and O'Farrell, P. H. (
1998
). Exit from mitosis in Drosophila syncytial embryos requires proteolysis and cyclin degradation, and is associated with localized dephosphorylation.
Genes Dev.
12
,
1495
-1503.
Sullivan, W., Ashburner, M. and Hawley, R. S.(
2000
).
Drosophila Protocols
. Cold Spring Harbor: Cold Spring Harbor Laboratory Press.
Takada, S., Kelkar, A. and Theurkauf, W. E.(
2003
). Drosophila checkpoint kinase 2 couples centrosome function and spindle assembly to genomic integrity.
Cell
113
,
87
-99.
Wojcik, E., Basto, R., Serr, M., Scaerou, F., Karess, R. and Hays, T. (
2001
). Kinetochore dynein: its dynamics and role in the transport of the Rough deal checkpoint protein.
Nat. Cell. Biol.
3
,
1001
-1007.
Yu, K. R., Saint, R. B. and Sullivan, W.(
2000
). The Grapes checkpoint coordinates nuclear envelope breakdown and chromosome condensation.
Nat. Cell. Biol.
2
,
609
-615.
Zachariae, W. and Nasmyth, K. (
1999
). Whose end is destruction: cell division and the anaphase-promoting complex.
Genes Dev.
13
,
2039
-2058.
Zalokar, M. and Erk, I. (
1976
). Division and migration of nuclei during early embryogenesis.
J. Microbiol. Cell
25
,
97
-106.
Zhang, X. H., Axton, J. M., Drinjakovic, J., Lorenz, L.,White-Cooper, H. and Renault, A. D. (
2004
). Spatial and temporal control of mitotic cyclins by the Gnu regulator of embryonic mitosis in Drosophila.
J. Cell Sci.
117
,
3571
-3578.