Schizosaccharomyces Pombe Bub3 Is Dispensable for Mitotic Arrest Following Perturbed Spindle Formation

Schizosaccharomyces pombe Bub3 Is Dispensable for Mitotic Arrest Following Perturbed Spindle Formation

Yoshie Tange and Osami Niwa1 Kazusa DNA Research Institute, Kisarazu, Chiba 292-0818, Japan Manuscript received September 6, 2007 Accepted for publication April 7, 2008

ABSTRACT The core proteins of the spindle assembly checkpoint (SAC), Mads, Bubs, and Mps1, first identified in the budding yeast, are thought to be functionally and structurally conserved through evolution. We found that fission yeast Bub3 is dispensable for SAC, as bub3 null mutants blocked mitotic progression when spindle formation was disrupted. Consistently, the bub3 mutation only weakly affected the stability of minichromosome Ch16 compared with other SAC mutants. Fission yeast Rae1 has sequence homology with Bub3. The bub3 rae1 double mutant and rae1 single mutant did not have defective SAC, suggesting that these genes do not have overlapping roles for SAC. Observations of living cells revealed that the duration of the mitotic prometaphase/metaphase was longer in the bub3 mutant and was Mad2 dependent. Further, the bub3 mutant was defective in sister centromere association during metaphase. Together, these findings suggest that fission yeast Bub3 is required for normal spindle dynamics, but not for SAC.

HE spindle assembly checkpoint (SAC) ensures the proteins to the kinetochores in both perturbed and T accurate segregation of chromosomes in mitosis unperturbed mitoses (Taylor et al. 1998; Kadura et al. by monitoring the presence of kinetochores that are 2005). Bub3 is a stoichiometric component of the not properly linked to spindle microtubules and blocking mitotic checkpoint complex (Hardwick et al. 2000; mitotic progression until all sister kinetochores are Sudakin et al. 2001; Poddar et al. 2005). Bub3 is attached to kinetochore microtubules (kMTs) in a biori- required for SAC function not only in budding yeast, ented manner. The core SAC proteins, Mad1, Mad2, but also in other organisms, including flies and mam- BubR1 (Mad3 in yeast), Bub1, Bub3, and Mps1 (Mph1 mals (Hoyt et al. 1991; Babu et al. 2003; Lopes et al. in fission yeast), were originally identified in budding 2004). Bub3 is a WD repeat-containing protein whose yeast, and are evolutionarily conserved in a wide range structural features are conserved in orthologs from of eukaryotes (Lew and Burke 2003; May and diverse species (Wilson et al. 2005). The fission yeast Hardwick 2006; Musacchio and Salmon 2007). genome contains two protein-coding sequences that We reported that a fission yeast g-tubulin mutant have significant homology to the budding yeast BUB3 requires these SAC genes for near-normal mitosis, but (Wood et al. 2002). One is rae1 (SPBC16A3.05c accord- the requirement differs greatly between individual ing to the systematic naming of genes) and the other is genes (Tange and Niwa 2007). By isolating novel bub3 (SPAC23H3.08c). Because bub3-deleted cells are mad2 alleles, we demonstrated that the SAC gene hypersensitive to microtubule-destabilizing drugs and functions required for mitosis in the g-tubulin mutant also because the bub3 mutant containing a cold-sensitive are genetically discernible from those required for b-tubulin mutation loses viability at a restrictive tem- mitotic arrest when spindle formation is grossly per- perature, fission yeast bub31 is postulated to be a SAC gene turbed. Thus, SAC genes may function in different (Vanoosthuyse et al. 2004). Consistent with this idea, modes in response to various mitotic defects. From this the fission yeast Bub3 interacts with Bub1 and Mad3, and point of view, we examined why a deletion mutation in this interaction is also required for kinetochore localiza- bub3 did not affect the growth of the g-tubulin mutant. tion (Millband and Hardwick 2002; Vanoosthuyse The BUB3 gene was first identified in budding yeast as et al. 2004; Kadura et al. 2005). In this study, however, we a high-copy suppressor of a bub1 mutant (Hoyt et al. found that fission yeast Bub3 was dispensable for SAC 1991). The Bub3 protein physically interacts with Bub1 function. in vivo in several organisms (Roberts et al. 1994; Taylor et al. 1998; Vanoosthuyse et al. 2004). This interaction is required for the localization of these MATERIALS AND METHODS

1Corresponding author: Kazusa DNA Research Institute, 2-6-7 Kazusa- Strains and general genetic methods: The strains are listed kamatari, Kisarazu, Chiba 292-0818, Japan. E-mail: [email protected] in supplemental Table 1. Details of the genetic methods were

Genetics 179: 785–792 ( June 2008) 786 Y. Tange and O. Niwa previously described in Tange and Niwa (2007). The mad31 TABLE 1 gene and bub11 gene on Ch16 were disrupted by the G418- resistant gene (kanr) and the nourseothricin-resistant gene Effect of SAC mutations on minichromosome stability (natr), respectively (Bahler et al. 1998; Sato et al. 2005). Nourseothricin (clonNAT) was purchased from Werner Bio- %of Agents ( Jena, Germany) and used at 100 mg/ml. For the half-sectored Total no. of minichromosome stability assay, log phase culture of cells con- Genotype colonies colonies Minichromosome taining a minichromosome was prepared in a synthetic medium lacking adenine and appropriate dilutions were made before Wild type 0.011 161,438 Ch16 a plating on YE plates. The green fluorescent protein (GFP)- mad2D 0.15 53,651 tagged rad211 gene was obtained from M. Yanagida via the mad1D 0.86 29,250 Yeast Genetic Resource Center. A temperature-sensitive rad21 mad1D mad2D 1.16 19,016 mutant (rad21-K1) was obtained from H. Ikeda (Tatebayashi bub3D 0.008 171,077 et al. 1998). The bub3DTura41 mutant used in previous studies bub3D mad2D 1.35 9,180 (Vanoosthuyse et al 2004) was obtained from K. Hardwick bub3D mad1D 2.15 3,962 (University of Edinburgh) via T. Matsumoto (Kyoto Univer- mph1D 8.2 1,602 sity). We verified that the minichromosome stability and Wild type 0.14 40,454 Ch10-CN2 mitotic-arrest activity of this bub3 mutant were identical with mad2D 0.29 44,166 that of the bub3 mutant that we created and used in this study mad3D 0.17 46,007 (see supplemental Table 2). A rae1 temperature-sensitive mu- bub1D 1.6 8,175 1 tant strain, h ade6-216 rae1-167, was obtained from R. Dhar bub3D 0.36 44,408 (National Institutes of Health) via T. Tani (Kumamoto Univer- Wild type 0.013 113,166 Ch16Dmad3Dbub1d 1 sity). This mutant formed extremely poor colonies at 30°. ade6 bub3D 0.040 113,911 cells had similar temperature sensitivity. Thus, we set 28° as the mad2D 0.17 83,060 semipermissive temperature for the minichromosome stability bub3D mad2D 3.8 4,414 assay and for all other assays. mad3D 0.090 42,181 Chromatin immunoprecipitation: Chromatin immunopre- bub3D mad3D 1.6 5,271 cipitation (ChIP) was performed essentially as described by D b c D e Saitoh et al. (1997). Rad21-GFP was immunoprecipitated with bub1 3.5 NA Ch16 bub1 an anti-GFP antibody (Living Colors full length A.v. polyclonal a Tange and Niwa (2007). antibody; Clontech, Mountain View, CA) and Dynabeads b Bernard et al. (1998). protein G (Invitrogen, Oslo). Coprecipitated DNA fragments c NA, not applicable. were quantified by real-time polymerase chain reaction (PCR) d Both mad31 and bub11 genes on Ch16 are deleted. using the Applied Biosystems (Foster City, CA) 7500 Real-Time- e bub11 gene on Ch16 is deleted. PCR system. PCR primers used were for inner centromere sequences (cnt and imr), outer centromere sequences (dg and dh), lys1,andmes1, which were described in Yokobayashi et al. (2003). Where indicated, the cell culture was synchronized by rate in bub3D cells was approximately twice that in wild the hydroxyurea treatment-release method, as described in ange iwa type (supplemental Table 3). Thus, the presence of one T and N (2007). 1 Other methods: Microscopy methods, including staining extra copy of the bub1 gene ameliorated the Ch16 with 49,6-diamidino-2-phenylindole and live observation of instability induced by the the lack of Bub3. The extra cells, are described in detail in Tange and Niwa (2007). Cdc2 copy of mad31 might also contribute to this improve- kinase activity assay is also described in detail in Tange and ment. Further, disruption of the mad31 and bub11 genes iwa N (2007), but the reaction in this study was performed at on Ch16 did not affect the minichromosome stability in 25° for 25 min. mad2D cells. It remains to be investigated how Bub1 and Mad3, when produced from extra genes, recover chromosome stability in the absence of the Bub3 protein. RESULTS AND DISCUSSION It was reported that Ch16 with the bub11 gene Minichromosome stability in SAC-related mutants: deletion in bub3D cells was lost at a frequency of 0.2% We compared the stability of the minichromosome (Vanoosthuyse et al 2004), a value different from that Ch16 in different SAC mutants. The minichromosome obtained in this study. The discrepancy might be due to stabilities in different mutants determined by the half- differences in experimental procedures; for example, in sector method (Allshire et al. 1995) are shown in Table 1 the previous study, cells from colonies were used while together with published data. Ch16 stability was the in this study log phase cultures were used. Because we most impaired in mph1D and bub1D, followed by mad1D, used the same experimental condition for every SAC mad2D, mad3D, and bub3D. Because Ch16 contains the mutant, it was evident that individual SAC genes have mad31 and bub11 genes (Chikashige et al. 2007), we also highly differential contributions to chromosome stabil- tested the stability of Ch16 with both of these genes ity. Thus, different (sets of) SAC genes are required to deleted. This minichromosome was three times more monitor various defects occurring in unperturbed mi- unstable in the bub3D background than in wild-type toses, or each SAC gene has a function in mitosis in cells, although it was still more stable than in mad3D cells addition to the checkpoint function, or both. It should and in other SAC mutants (Table 1). When we used be noted that the stability was different between mad1D Ch16 with only bub11 deleted, the minichromosome loss and mad2D, particularly between bub1D and bub3D.By Mitotic Arrest in S. pombe Bub3 Mutant 787

TABLE 2 Stability of minichromosome Ch16 in bub3D under different conditions

%of half-sectored Total no. of Genotype colonies colonies TBZ Wild type 0.011a 161,438 0 mg/ml mad2D 0.15b 53,651 mad2-56Tkanr 0.035b 33,853 mad2-64Tkanr 0.020b 64,886 bub3D 0.008a 171,077 bub3D mad2D 1.35a 9,180 bub3D mad2-56Tkanr 0.80 9,788 bub3D mad2-64Tkanr 0.25 9,788 bub3D mad21Tkanr 0.011 34,960 Wild type 0.007c 69,069 rae1-167 0.076c 53,950 rae1-167 bub3D 0.070c 55,408 Wild type 0.034 154,885 7.5 mg/ml bub3D 0.31 126,342 a The same data presented in Table 1 for comparison. b Tange and Niwa (2007). c Assayed at 28°; others were at 30°.

the Mad2 function required for chromosome stability was only weakly supplied by the mad2-56 and mad2-64 mutants in the bub3D mutant (Table 2). These mutants igure lack the SAC activity to arrest mitotic progression, but F 1.—TBZ-sensitivity test. Log phase cultures of the ange indicated strains in YE medium were serially diluted (fivefold) remain functional for chromosome stability (T and spotted on (a) YEA, (b) YEA 1 TBZ (7.5 mg/ml), and (c) and Niwa 2007). Together, the results suggested that YEA 1 TBZ (12.5 mg/ml) plates, aiming to have 10 cells in the the bub3D mutation induces a certain spindle defect that most diluted spot. Incubation was at 30° for 3 days. All mu- only slightly decreases chromosome stability if it is tants had the deletion mutation. WT, wild type. appropriately monitored by SAC function. Live analysis of spindle dynamics in the bub3 mutant: contrast, in budding yeast, minichromosome stability is To determine if there is a mitotic defect in the bub3D almost identical between these respective pairs, consis- mutant, we performed live analyses of the spindle tent with the distinct functional complex formation dynamics and chromosome behavior. Spindle dynamics from the respective pairs of the gene products (Warren in wild-type fission yeast are divided into three phases et al. 2002). (Nabeshima et al. 1998; Mallavarapu et al. 1999). In We then compared the sensitivity of each SAC mutant phase 1, a bipolar spindle is formed; in phase 2, which to thiabendazole (TBZ), a microtubule-destabilizing corresponds to prometaphase and metaphase, spindle agent. The TBZ sensitivity of individual SAC mutants length remains unchanged or slightly elongates; and in roughly paralleled that of the minichromosome in- phase 3, the spindle elongates. The duration of phases 1 stability in these mutants, with the exception of bub3D and 2 in wild type was 7.9 6 1.0 min (n ¼ 31), whereas it (Figure 1). This mutant was more sensitive to TBZ than was 10.8 6 1.4 min (n ¼ 28) in the bub3D mutant (Figure mad1D, mad2D, and mad3D, and yet it only weakly 2). This 37% increase in the phase length in the bub3D affected minichromosome stability. There were modest mutant was almost completely abolished in the bub3D synergetic effects of mad2D and mad3D mutations with mad2D mutant (8.1 6 1.3 min; n ¼ 31), indicating that bub3D (Figure 1). the increase was dependent on Mad2. This finding also We examined why minichromosome stability was not suggested that the Mad2-dependent extension of phase impaired in bub3D as much as in other SAC mutants. In 2 allowed for the time necessary for accurate chromo- accordance with the hypersensitivity of bub3D to TBZ, some segregation. In addition, we analyzed the spindle Ch16 was 10-fold more unstable in bub3D than in wild- dynamics in a SAC defective mutant, mad3D, (Figure 2, type cells in the presence of a low concentration of TBZ bottom right). In this mutant, the duration was 7.6 6 2.1 (Table 2). Further, when bub3D was combined with min (n ¼ 22). More samples need to be observed before mad2D, mad1D,ormad3D, Ch16 became more unstable drawing a conclusion, but the duration might be shorter than in any of the single mutants (Table 1). In addition, in the mad3 mutant, similar to our previous finding that 788 Y. Tange and O. Niwa

Figure 2.—Live observation of spindle dynamics. Spindle length was determined as the distance between two separating Sid4- GFP signals, taking the difference in focal plane into account. Living cells of indicated strains were observed every 30 sec, and at each time point seven images were taken at serial focal planes at 0.3-mm steps. The time of the onset of phase 3 was set as time 0 and all graphs obtained for each strain are assembled in a single figure. Red triangles and horizontal bars indicate average time of phase 1 onset and standard deviation, respectively. the duration was shorter in the mad2D mutant (Tange aberrant feature was that the distances between the two and Niwa 2007). cen2-GFP signals in phase 2 (d4 in Figure 3A) tended to To further investigate bub3 mutant mitosis, we per- be greater in the mutant (Figure 3C). The average formed a live observation of a GFP-labeled locus 5 kb distances in the same windows used above were 0.32 6 apart from centromere 2 (cen2-GFP;Yamamoto and 0.14 mm(n ¼ 340) and 0.47 6 0.21 mm(n ¼ 595), Hiraoka 2003). There were two apparently abnormal respectively. Perhaps related to this observation, mini- features in the mutant mitosis (Figure 3A; movies are chromosome Ch10-CN2, a derivative of Ch16, in which available as supplemental Figures 1–12). One was that both of the arms are deleted (Niwa et al. 1989), was the cen2-GFP pair frequently oscillated within a range of more unstable in bub3D than in mad2D and mad3D the spindle length during phase 2; that is, the length of (Table 1). These results suggest that the Schizosacchar- the kMTs (d2 and d3 in Figure 3A) frequently changed omyces pombe Bub3 function may be more important for between elongated and shortened. For a quantitative sister centromere association in metaphase, but less analysis, we set a 5-min window immediately before the important for chromosome arms. It is also possible that onset of anaphase A and calculated the difference be- sister centromeres are further apart in the mutant between two successive d2 and d3 values from the window. cause they are pulled apart by a stronger force. Bub3 The rate of elongation/shrinkage was significantly might directly function in the regulation of the kMT higher in the bub3 mutant (Figure 3B). The frequency attachment/assembly. Regardless of the primary reason of the time intervals with difference values of more than for the apparent defect in centromere association, the 0.4 mm next to intervals with an opposite elongation/ establishment and/or maintenance of bipolar spindle shrinkage phase was 5.8% (37/636) and 13.8% (155/ formation may become less efficient in the absence 1122) in wild type and mutant, respectively, indicating of Bub3, which may result in the extended phase 2 that swift switching of the elongation/shrinkage phases duration in a SAC-dependent manner. Consistent with occurred more frequently in the mutant. The other this idea, fission yeast Bub3 is localized on the kinet- Mitotic Arrest in S. pombe Bub3 Mutant 789

Figure 3.—Live analysis of sister chromatid separation. Fission yeast cells carrying both Sid4- GFP and cen2-GFP were observed every 15 sec with 10 steps (0.2 mm) along the z-axis at each time point. (A) Distances (d1–d4) were measured and plotted. (B) Distributions of the rates of elongation and shrinkage of kMTs in phase 2 (kMT lengths,d2 and d3, changed in the 15-sec intervals) (see text for details). (C) Distribu- tions of the distances between two cen2 signals (d4).

ochores until the onset of anaphase in normal mitosis by the hydroxyurea treatment method (supplemental (Vanoosthuyse et al. 2004; Kadura et al. 2005). Figure 13, and data not shown). Using this method, it Genetic interaction of bub3D with a cohesin mutant: was not possible to determine whether the amount of Because of the possibility that the bub3 mutant had Rad21 protein bound to the centromeric region in the impaired centromeric cohesion, we examined whether bub3 mutant relative to that bound to the arm regions the bub3 mutant genetically interacts with a cohesin differed from that in wild type. Further detailed studies mutant, rad21-K1 (Tatebayashi et al. 1998). The bub3D are needed to determine whether Bub3 is involved in rad21-K1 double mutant did not make colonies at 33°,a regulating the proper assembly of cohesin complexes, temperature at which the rad21 mutant formed visible both temporally and spatially, in the centromeric re- colonies (Figure 4). In contrast, the bub1D mutant gions, or if there is only an indirect relation between slightly affected the restrictive temperature, but not as them. much as the bub3D mutant, the mad2D mutant did not SAC activity is retained in bub3D: The results de- affect at all (Figure 4), indicating that the deleterious scribed above led us to reinvestigate whether bub3D has effect of the bub3 mutation on the rad21 mutant was not defective SAC function. For quantitative evaluation of due to a functional defect in SAC. SAC activity, we used the cut7-24 and nda3-KM311 We then investigated whether the amount of chro- mutants in which mitotic spindle formation is severely matin-bound cohesin complex was altered in the bub3 impaired. Under the restrictive condition in these mu- mutant. ChIP was performed using a GFP-tagged Rad21. tants, cells with hypercondensed chromosomes accu- Consistent with previous results (Tomonaga et al. 2000; mulate when SAC function is active, whereas those with Yokobayashi et al. 2003), Rad21-GFP in wild-type cells the cut phenotype increase when the function is was enriched in the outer centromere regions (dg and impaired (He et al. 1997; Kim et al. 1998). Both the wild dh sequences, see supplemental Figure 13). Rad21 type and bub3 mutant were active in SAC function, enrichment in the dg and dh sequences was also ob- whereas the mad2 and bub1 mutants almost completely served in bub3D mutant extracts from a logarithmic lacked the activity (Table 3). To verify this result, we phase culture, metaphase-arrested cut9 cells, and syn- examined Cdc2 kinase activity. Cdc2 kinase activity chronized cell culture at G2 or mitotic phases produced increased after shifting to restrictive temperatures both 790 Y. Tange and O. Niwa

TABLE 3 Mitotic arrest activity

% of cells with hypercondensed % of cells showing chromosomes the cut phenotype Genotype (no. of cells)a (no. of cells) cut7 56.5 (124) NDb cut7 bub1D 1.2 (172) cut7 bub3D 68.7 (115) cut7 mad2D 0.7 (138) cut7 mad3D 0.0 (144) nda3 35.0 (203) 0.7 (303) nda3 bub1D 1.0 (203) 6.4 (346) nda3 bub3D 57.5 (153) 1.2 (404) nda3 mad2D 3.3 (210) 6.6 (336) a Cells were incubated at 36° for 3 hr and at 18° for 9 hr for the cut7 and nda3 mutants, respectively. b Not performed.

ever, are involved in the spindle checkpoint in mice (Babu et al. 2003). Therefore, we examined whether fission yeast Rae1 and Bub3 have overlapping roles in SAC. To this end, we investigated the rae1-167 mutant at a semipermissive temperature of 28° (see below and materials and methods for temperature settings). We determined the stability of Ch16 in rae1-167 and in rae1-167 bub3D mutants (Table 2). In both mutants, Ch16 was 10 times more unstable than in wild type, indicating that Rae1 activity was reduced at 28°. Why the minichromosome was destabilized in the rae1 mutant Figure 4.—Genetic interaction of rad21-K1 mutation with SAC-related mutations. Strains with the indicated genotypes is not known, but the destabilization was not due to were inoculated on YES plates and incubated for 4 days at defective SAC function (see below). the indicated temperatures. The TBZ sensitivity of the rae1-167 single mutant was almost the same as that of the wild type, and the rae1 in the wild type and bub3 mutants, whereas there was bub3D double mutant was only marginally more sensitive only a small increase in mad2D and bub1D (Figure 5A). than the bub3D single mutant (supplemental Figure 14). These results indicated that the fission yeast bub31 gene Because increased sensitivity to microtubule destabiliz- is dispensable for SAC function. The fission yeast Bub3 ing agents is a common property of all known SAC protein is required for the localization of Bub1 and Mad3 mutants, this finding argues that the rae1 gene is un- to kinetochores in perturbed mitoses (Vanoosthuyse likely to be a SAC gene. We then directly analyzed et al. 2004; Kadura et al. 2005). Although this localization whether the rae11 gene is involved in SAC function. hadbeenpostulatedtoberequiredforSACfunction, First, cells were treated with carbendazim (CBZ), an our results indicated that the Bub3-dependent localiza- inhibitor of microtubule polymerization, and the fre- tion of these SAC proteins is not necessary, at least for quency of cells with hypercondensed chromosomes was SAC function. determined (Saitoh et al. 2005). Cells with hyper- In fission yeast, the rae11 gene has sequence homol- condensed chromosomes accumulated after the addi- ogy with the bub31 gene (29% identity and 51% tion of CBZ in both rae1 single and rae1 bub3 double similarity in .80% of the whole sequence). The rae11 mutants. Importantly, the effect of CBZ was more gene is essential for cell viability and is required for prominent in the rae1 bub3 double mutant (Figure mRNA export (Yoon et al. 1997). The gene is not in- 5B). The Cdc2 kinase activity data were consistent with volved in SAC function, on the basis of observations that this result (Figure 5B). Thus, in the rae1 mutant where a temperature-sensitive mutant, rae1-167, is slightly Rae1 function was sufficiently impaired to appreciably more resistant to a microtubule polymerization inhibi- affect minichromosome stability, there was no indica- tor, the Rae1 protein does not form any complexes with tion of any SAC defect, even in the absence of Bub3. We known checkpoint proteins, and the overproduction therefore conclude that fission yeast bub3D is not de- of Rae1 does not rescue a bub3 mutant phenotype fective in SAC function, and this is not due to the (Vanoosthuyse et al. 2004). Both Rae1 and Bub3, how- presence of Rae1, a protein structurally similar to Bub3. Mitotic Arrest in S. pombe Bub3 Mutant 791

Figure 5.—(A) The Cdc2 kinase assay. Log phase cultures of cut7 (a) and nda3 (b) were shifted to a restrictive temperature (36° in (a) and 18° in (b) at time 0. Cells were harvested at the indicated time for cell-extract preparation. 32P-Phosphorylated histone H1 is shown in gel images (a and b) and the increase of incorporated 32P from the time 0 sample is shown with arbitrary units (c and d). (B) Mitotic arrest activity. CBZ (50 mg/ml) was added to log phase cultures (in YES medium at 26°) of wild type, bub3, rae1- 167, and rae1-167 bub3 at time 0 and shifted to 28°. At the indicated time, cells with hypercondensed chromosomes were scored and plotted. The histogram shows the increase in H1 kinase activity from the CBZ-treated cells. bub1 mutant is included as a control.

How can our findings be reconciled with previously as suggested above. The known interaction of Bub3 with published results (Rajagopalan et al. 2004; Tournier Bub1 and Mad3 and its requirement for their localiza- et al. 2004; Vanoosthuyse et al. 2004; Asakawa et al. tion to kinetochores might be intimately related to the 2005, 2006; Kadura et al. 2005), when all the findings function of Bub3. Further comparisons of the fission are consistent with the idea that fission yeast Bub3 is a yeast spindle checkpoint system with those in other or- SAC protein? One of the most crucial results leading to ganisms not only should elucidate the function of Bub3 this conclusion was that the Bub3 activity is required for in fission yeast, but also will contribute to determining the viability of mitotically arrested b-tubulin mutant the nature of the spindle checkpoint in general. anoosthuyse cells (V et al. 2004). As shown in this study, We thank A. Yamamoto, H. Ikeda, T. Tani, R. Dhar, M. Yanagida, T. the fission yeast bub3 mutant has a defect in spindle Matsumoto, and K. Hardwick for the yeast strains. We are also grateful dynamics. Therefore, it is possible that the bub3 muta- to an anonymous reviewer who suggested the use of a Ch16 derivative tion affects the recovery process or spindle reformation containing the mad3 and bub1 deletions. This work was supported by after mitotic arrest in the tubulin mutant. It was recently the Kazusa DNA Research Institute and partly by a grant from Japan Society for the Promotion of Science (no. 16370083) to O.N. reported that the recovery process is greatly impaired in the absence of Shugoshin 2, which is not a SAC protein (Vanoosthuyse et al. 2007). Likewise, all of the pre- LITERATURE CITED vious results might be reevaluated without postulating llshire immo kwall averzat Bub3 to be a SAC protein. At present, the cellular func- A , R. C., E. R. N ,K.E , J.-P. J and G. Cranston, 1995 Mutations derepressing silent centromeric do- tion of the S. pombe Bub3 is not known, although it is mains in fission yeast disrupt chromosome segregation. Genes possible that Bub3 has a role in centromeric cohesion, Dev. 9: 218–233. 792 Y. Tange and O. Niwa

Asakawa, K., M. Toya,M.Sato,M.Kanai,K.Kume et al., Rajagopalan,S.,A.Bimbo,M.K.Balasubramanian and S. Oliferenko, 2005 Mal3, the fission yeast EB1 homologue, cooperates with 2004 A potential tension-sensing mechanism that ensures timely Bub1 spindle checkpoint to prevent monopolar attachment. anaphase onset upon metaphase spindle orientation. Curr. Biol. EMBO Rep. 6: 1194–1200. 14: 69–74. Asakawa, K., K. Kume,M.Kanai,T.Goshima,K.Miyahara et al., Roberts, B. T., K. A. Farr and M. A. Hoyt, 1994 The Saccharomyces 2006 The V260I mutation in fission yeast a-tubulin Atb2 affects cerevisiae checkpoint gene BUB1 encodes a novel protein kinase. microtubule dynamics and EB1-Mal3 localization and activates Mol. Cell. Biol. 14: 8282–8291. the Bub1 branch of the spindle checkpoint. Mol. Biol. Cell 17: Saitoh, S., K. Takahashi and M. Yanagida, 1997 Mis6, a fission 1421–1435. yeast inner centromere protein, acts during G1/S and forms spe- Babu, J. R., K. B. Jeganathan,D.J.Baker,X.We,N.Kang-Decker cialized chromatin required for equal segregation. Cell 90: 131– et al., 2003 Rae1 is an essential mitotic checkpoint regulator 143. that cooperates with Bub3 to prevent chromosome missegrega- Saitoh, S., K. Ishii,Y.Kobayashi and K. Takahashi, 2005 Spindle tion. J. Cell Biol. 160: 341–353. checkpoint signaling requires the mis6 kinetochore subcomplex, Bahler, J., J. Q. Wu,M.S.Longtine,N.G.Shah,A.McKenzie III which interacts with mad2 and mitotic spindles. Mol. Biol. Cell et al., 1998 Heterologous modules for efficient and versatile 16: 3666–3677. PCR-based gene targeting in Schizosaccharomyces pombe. Yeast 14: Sato, M., S. Dhut and T. Toda, 2005 New drug-resistant cassettes 943–951. for gene disruption and epitope tagging in Schizosaccharomyces Bernard, P., K. Hardwick and J.-P. Javerzat, 1998 Fission yeast pombe. Yeast 22: 583–591. bub1 is a mitotic centromere protein essential for the spindle Sudakin, V., G. K. Chan and T. J. Yen, 2001 Checkpoint inhibition checkpoint and the preservation of correct ploidy through mito- of the APC/C in HeLa cells is mediated by a complex of BUBR1, sis. J. Cell Biol. 143: 1775–1787. BUB3, CDC20, and MAD2. J. Cell Biol. 154: 925–936. Chikashige, Y., C. Tsutsumi,K.Okamasa,M.Yamane,J.Nakayama Tange, Y., and O. Niwa, 2007 Novel mad2 alleles isolated in a Schiz- et al., 2007 Gene expression and distribution of Swi6 in partial osaccharomyces pombe g-tubulin mutant are defective in metaphase aneuploids of the fission yeast Schizosaccharomyces pombe. Cell arrest activity, but remain functional for chromosome stability in Struct. Funct. 32: 149–161. unperturbed mitosis. Genetics 175: 1571–1584. Hardwick, K. G., R. C. Johnson,D.L.Smith and A. W. Murray, Tatebayashi, K., J. Kato and H. Ikeda, 1998 Isolation of a Schizosac- 2000 MAD3 encodes a novel component of the spindle check- charomyces pombe rad21ts mutant that is aberrant in chromosome point which interacts with Bub3p, Cdc20p, and Mad2p. J. Cell segregation, microtubule function, DNA repair, and sensitive to hy- Biol. 148: 871–882. droxyurea: possible involvement of Rad21 in ubiquitin-mediated He, X., T. E. Patterson and S. Sazer, 1997 The Schizosaccharomyces proteolysis. Genetics 148: 49–57. pombe spindle checkpoint protein mad2p blocks anaphase and Taylor, S. S., E. Ha and F. McKeon, 1998 The human homologue genetically interacts with the anaphase-promoting complex. of Bub3 is required for kinetochore localization of Bub1 and a Proc. Natl. Acad. Sci. USA 94: 7965–7970. Mad3/Bub1-related protein kinase. J. Cell Biol. 142: 1–11. Hoyt, M. A., L. Totis and B. T. Roberts, 1991 S. cerevisiae genes Tomonaga, T., K. Nagao,Y.Kawasaki,K.Furuya,A.Murakami required for cell cycle arrest in response to loss of microtubule et al., 2000 Characterization of fission yeast cohesin: essential function. Cell 66: 507–517. anaphase proteolysis of Rad21 phosphorylated in the S phase. Kadura, S., X. He,V.Vanoosthuyse,K.G.Hardwick and S. Sazer, Genes Dev. 14: 2757–2770. 2005 The A78V mutation in the Mad3-like domain of Schizosac- Tournier, S., Y. Gachet,V.Buck,J.S.Hyams and J. B. A. Millar, charomyces pombe Bub1p perturbs nuclear accumulation and kinet- 2004 Disruption of astral microtubule contact with the cell cor- ochore targeting of Bub1p, Bub3p, and Mad3p and spindle tex activates a Bub1, Bub3, and Mad3-dependent checkpoint in assembly checkpoint function. Mol. Biol. Cell 16: 385–395. fission yeast. Mol. Biol. Cell 15: 3345–3356. Kim, S. H., D. P. Lin,S.Matsumoto,A.Kitazono and T. Matsumoto, Vanoosthuyse, V., R. Valsdotti,J.P.Javerzat and K. G. Hardwick, 1998 Fission yeast Slp1: an effector of the Mad2-dependent spin- 2004 Kinetochore targeting of fission yeast Mad and Bub pro- dle checkpoint. Science 279: 1045–1047. teins is essential for spindle checkpoint function but not for Lew, D. J., and D. J. Burke, 2003 The spindle assembly and spindle all chromosome segregation roles of Bub1p. Mol. Cell. Biol. position checkpoints. Annu. Rev. Genet. 37: 251–282. 24: 9786–9801. Lopes,C.S.,P.Sampaio,B.Williams,M.Goldberg and C. E. Sunkel, Vanoosthuyse,V.,S.Prykhozhij and K. G. Hardwick, 2004 The Drosophila bub3 protein is required for the mitotic 2007 Shugoshin 2 regulates localization of the chromosomal checkpoint and for normal accumulation of cyclins during G2 passenger proteins in fission yeast mitosis. Mol. Biol. Cell 18: and early stages of mitosis. J. Cell Sci. 118: 187–198. 1657–1669. Mallavarapu, A., K. Sawin and T. Mitchison, 1999 A switch in mi- Warren, C. D., D. M. Brady,R.C.Johnston,J.S.Hanna,K.G. crotubule dynamics at the onset of anaphase B in the mitotic Hardwick et al., 2002 Distinct chromosome segregation roles spindle of Schizosaccharomyces pombe. Curr. Biol. 9: 1423–1426. for spindle checkpoint proteins. Mol. Biol. Cell 13: 3029–3041. May, K. M., and K. G. Hardwick, 2006 The spindle checkpoint. Wilson, D. K., D. Cerna and E. Chew, 2005 The 1.1-A˚ structure of J. Cell Sci. 119: 4139–4142. the spindle checkpoint protein Bub3p reveals functional regions. Millband, D. N., and K. G. Hardwick, 2002 Fission yeast Mad3p is J. Biol. Chem 280: 13944–13951. required for Mad2p to inhibit the anaphase-promoting complex Wood, V., R. G. William,M.A.Rajandream,M.Lyne,R.Lyne, et al., and localizes to kinetochores in a Bub1p-, Bub3p-, and Mph1p- 2002 The genome sequence of Schizosaccharomyces pombe. Na- dependent manner. Mol. Cell. Biol. 22: 2728–2742. ture 415: 871–880. Musacchio, A., and E. D. Salmon, 2007 The spindle-assembly Yokobayashi, S., M. Yamamoto and Y. Watanabe, 2003 Cohesins checkpoint in space and time. Nat. Rev. Mol. Cell Biol. 8: 379– determine the attachment manner of kinetochores to spindle mi- 393. crotubules at meiosis I in fission yeast. Mol. Cell Biol. 23: 3965– Nabeshima, K., T. Nakagawa,A.F.Straight,A.Murray,Y. 3973. Chikashige et al., 1998 Dynamics of centromeres during meta- Yamamoto, A., and Y. Hiraoka, 2003 Monopolar spindle attach- phase-anaphase transition in fission yeast: Dis1 is implicated in ment of sister chromatids is ensured by two distinct mechanisms force balance in metaphase bipolar spindle. Mol. Biol. Cell 9: at the first meiotic division in fission yeast. EMBO J. 22: 2284– 3211–3225. 2296. Niwa, O., T. Matsumoto,Y.Chikashige and M. Yanagida, Yoon, J. H., W. A. Whalen,A.Bharathi,R.Shen and R. Dhar, 1989 Characterization of Schizosaccharomyces pombe minichromo- 1997 Npp106p, a Schizosaccharomyces pombe nucleoporin some derivatives and a functional allocation of their centromere. similar to Saccharomyces cerevisiae Nic96p, functionally inter- EMBO J. 8: 3045–3052. acts with Rae1p in mRNA export. Mol. Cell. Biol. 17: 7047–7060. Poddar, A., P. T. Stukenberg and D. J. Burke, 2005 Two complexes of spindle checkpoint proteins containing Cdc20 and Mad2 as- semble during mitosis independently of the kinetochore in Sac- charomyces cerevisiae. Eukaryot. Cell 4: 867–878. Communicating editor: P. Russell