pmc logo image
Logo of mbcJournal URL: http://www.molbiolcell.org
Journal List > Mol Biol Cell > v.19(3); Mar 2008

Formats:
Mol Biol Cell. 2008 March; 19(3): 843–854.
doi: 10.1091/mbc.E07-06-0556.
PMCID: PMC2262965
Copyright © 2008 by The American Society for Cell Biology
CENP-O Class Proteins Form a Stable Complex and Are Required for Proper Kinetochore Function
Tetsuya Hori,* Masahiro Okada,* Katsumi Maenaka,† and Tatsuo Fukagawacorresponding author*
*Department of Molecular Genetics, National Institute of Genetics and The Graduate University for Advanced Studies, Mishima 411-8540, Japan; and
†Division of Structural Biology, Medical Institute of Bioregulation, Kyushu University, 3-1-1 Maidashi, Fukuoka 812-8582, Japan
Ted Salmon, Monitoring Editor
corresponding authorCorresponding author.
Address correspondence to: Tatsuo Fukagawa (Email: tfukagaw/at/lab.nig.ac.jp)
Received June 12, 2007; Revised November 21, 2007; Accepted December 10, 2007.
Small right arrow pointing to: This article has been cited by other articles in PMC.
Abstract
We previously identified a multisubunit complex (CENP-H/I complex) in kinetochores from human and chicken cells. We showed that the CENP-H/I complex is divided into three functional classes. In the present study, we investigated CENP-O class proteins, which include CENP-O, -P, -Q, -R, and -50 (U). We created chicken DT40 cell knockouts of each of these proteins, and we found that all knockout lines were viable, but that they showed slow proliferation and mitotic defects. Kinetochore localization of CENP-O, -P, -Q, and -50 was interdependent, but kinetochore localization of these proteins was observed in CENP-R–deficient cells. A coexpression assay in bacteria showed that CENP-O, -P, -Q, and -50 proteins form a stable complex that can associate with CENP-R. Phenotype analysis of knockout cells showed that all proteins except for CENP-R were required for recovery from spindle damage, and phosphorylation of CENP-50 was essential for recovery from spindle damage. We also found that treatment with the proteasome inhibitor MG132 partially rescued the severe mitotic phenotype observed in response to release from nocodazole block in CENP-50–deficient cells. This suggests that CENP-O class proteins are involved in the prevention of premature sister chromatid separation during recovery from spindle damage.
INTRODUCTION
Accurate chromosome segregation during mitosis is essential for the correct transmission of genetic material. A kinetochore is assembled at the centromere of each chromatid of a replicated chromosome, and it forms a dynamic interface with microtubules of the mitotic spindle (Cleveland et al., 2003 blue right-pointing triangle; Fukagawa, 2004 blue right-pointing triangle). Although chromosome segregation errors cause genetic diseases, including some cancers (Yuen et al., 2005 blue right-pointing triangle), the mechanism by which kinetochores interact with microtubules of the spindle apparatus during cell division is not fully understood.
New insight regarding the mechanism of chromosome segregation has come from the discovery and characterization of kinetochore proteins (Chan et al., 2005 blue right-pointing triangle). In recent years, many kinetochore proteins have been identified in vertebrate cells by various approaches (Cheeseman et al., 2004 blue right-pointing triangle; Obuse et al., 2004a blue right-pointing triangle,b blue right-pointing triangle; Minoshima et al., 2005 blue right-pointing triangle; Foltz et al., 2006 blue right-pointing triangle; Izuta et al., 2006 blue right-pointing triangle; McAinsh et al., 2006 blue right-pointing triangle; Meraldi et al., 2006 blue right-pointing triangle; Okada et al., 2006 blue right-pointing triangle). Characterization of the cellular functions of each of these proteins and the protein–protein network within cells will lead to an understanding of how kinetochores assemble and function.
We previously isolated a large protein complex (CENP-H/I complex), which contains at least 11 components, including CENP-H and CENP-I, and we found that this complex can be divided into three functional classes, based on phenotype analyses (Okada et al., 2006 blue right-pointing triangle). Initial studies of the CENP-H/I complex in chicken and human cells suggested that it is required for proper chromosome alignment and segregation (Okada et al., 2006 blue right-pointing triangle). To further understand the function of the CENP-H/I complex, it is necessary to characterize each class in detail. Here, we present results of functional analyses of CENP-O class proteins, which include CENP-O, -P, -Q, -R, and -50 (U). We created chicken DT40 cell knockouts (KOs) of each of these proteins, and we expressed the proteins in Escherichia coli. The results suggest that these proteins form a stable complex that is required for proper chromosome segregation. We previously showed that CENP-50 is essential for recovery from spindle damage (Minoshima et al., 2005 blue right-pointing triangle). In the present study, we found that all CENP-O class proteins except for CENP-R were required for recovery from spindle damage. Furthermore, we found that treatment with the proteasome inhibitor MG132 partially rescued the severe mitotic phenotype that occurs in response to release from nocodazole block in CENP-50–deficient cells. These results suggest that CENP-O class proteins are involved in the prevention of premature sister chromatid separation during the recovery from spindle damage.
MATERIALS AND METHODS
Molecular Biology, Cell Culture, and Transfection
All plasmids were constructed by standard methods. Disruption constructs for each gene were generated such that genomic fragments encoding several exons were replaced with a histidinol-, neomycine-, or puromycin-resistance cassette under control of the β-actin promoter. Target constructs were transfected with a Gene Pulser II electroporator (Bio-Rad, Hercules, CA). All molecular biological experiments, including Southern blot analysis and Western blot analysis, were followed by standard methods.
DT40 cells were cultured and transfected as described previously (Fukagawa et al., 2004 blue right-pointing triangle). All DT40 cells were cultured at 38°C in Dulbecco's modified medium supplemented with 10% fetal calf serum, 1% chicken serum, 2-mercaptoethanol, penicillin, and streptomycin.
Immunofluorescence and Image Acquisition
Immunofluorescent staining of whole cells was performed as described previously (Okada et al., 2006 blue right-pointing triangle). Cells were collected onto slides with a cytocentrifuge, and then they were fixed in 3% paraformaldehyde in phosphate-buffered saline (PBS) for 15 min at room temperature and permeabilized in 0.5% NP-40 in PBS for 10 min at room temperature or 100% methanol for 15 min at −20°C, rinsed three times in 0.5% bovine serum albumin (BSA) in PBS, and incubated for 1 h at 37°C with primary antibody. Binding of primary antibody was then detected with fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit immunoglobulin (Ig)G (Jackson ImmunoResearch Laboratories, West Grove, PA) diluted to an appropriate concentration in 0.5% BSA in PBS. Affinity-purified rabbit polyclonal antibodies were generated against recombinant chicken CENP-Q and chicken CENP-R as described previously (Fukagawa et al., 1999 blue right-pointing triangle). Other antibodies used in this study were described previously (Nishihashi et al., 2002 blue right-pointing triangle; Hori et al., 2003 blue right-pointing triangle; Minoshima et al., 2005 blue right-pointing triangle; Okada et al., 2006 blue right-pointing triangle). Chromosomes and nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI) at 0.2 μg/ml in Vectashield Antifade (Vector Laboratories, Burlingame, CA). All immunofluorescence images were collected with a cooled charge-coupled device camera (CoolSNAP HQ, Photometrics Image Point; Photometrics, Tucson, AZ) mounted on an IX71 inverted microscope (Olympus, Tokyo, Japan) with a 60× objective lens (PlanApo 60×, numerical aperture [NA] 1.40) together with a filter wheel. All subsequent analysis and processing of images were performed using IPLab software (Signal Analytics, Vienna, VA).
Live-Cell Observations
For live-cell imaging, a histone H2B-red fluorescent protein (RFP) plasmid is transfected into CENP-Q–deficient cells to visualize chromosomes. Fluorescently stained living cells were observed with an IX71 inverted microscope with an oil immersion objective lens (PlanApo 60×, NA 1.40) in temperature-controlled box to keep temperature at 38°C. Time-lapse images were recorded at 3-min intervals with an exposure time of 0.2–0.3 s.
Spindle Damage Recovery Assay
Cells were treated with 1.7 μM nocodazole for 10–12 h, and cells were washed with fresh medium several times. Cells were put into fresh medium in the presence or absence of 10 μM MG132, and they were cultured at 38°C. At the indicated time points, cells were collected and examined by Hoechst33342 staining. Morphology of chromosomes and nuclei of the cells were observed and scored. More than 1000 cells were scored for each time point.
Protein Expression and Analysis
Coexpression of CENP-O class proteins was performed by cloning the cDNAs for each protein into pST39 (a gift from S. Tan, The Pennsylvania State University, University Park, PA; Tan, 2001 blue right-pointing triangle). The cDNA for CENP-R was cloned into pCOLADuet-1 (Novagen, Madison, WI). Protein expression in BL21-codon Plus (DE3)-RIL E. coli was induced with 0.5 mM isopropyl β-d-thiogalactoside for 2 h at 37°C, and the proteins were purified using nickel-nitrilotriacetic acid (Ni-NTA) beads (QIAGEN, Valencia, CA). The eluted protein from the Ni-NTA beads with elution buffer (50 mM Na phosphate, pH 8.0, 0.3 M NaCl, 0.3 M imidazole, 0.1% NP-40, and 1 mM β-mercaptoethanol) was fractionated on a Superose 6 gel filtration column in elution buffer. DT40 cells (1 × 109 cells) that express CENP-50-FLAG or CENP-P-FLAG were lysed in 5 ml of lysis buffer (50 mM Na phosphate, pH 8.0, 0.3 M NaCl, 0.1% NP-40, 5 mM β-mercaptoethanol, complete protease inhibitor [Roche Diagnostics, Mannheim, Germany], and 20 U/ml DNase I [Takara, Kyoto, Japan]), centrifuged at 20,000 × g for 10 min at 4°C, and then a supernatant fraction was collected. Anti-FLAG M2-beads (Sigma, Tokyo, Japan) was incubated with the supernatant fraction for 2 h at 4°C and washed with lysis buffer and eluted with lysis buffer in the presence of 3× FLAG peptide (Sigma). The eluted protein complex was fractionated on a Superose 6 gel filtration column in 50 mM Na phosphate, pH 8.0, 0.3 M NaCl, 0.3 M imidazole, 0.1% NP-40, and 1 mM β-mercaptoethanol at 4°C.
RESULTS
KO Cell Lines of CENP-O Class Proteins Are Viable but Show Slow Proliferation Rates
We previously proposed the CENP-O class of proteins containing CENP-O, -P, -Q, -R, and -50, and we reported the creation of DT40 cell line KOs for CENP-O, CENP-P, and CENP-50 (Minoshima et al., 2005 blue right-pointing triangle; Okada et al., 2006 blue right-pointing triangle). To further elucidate the function of CENP-O class proteins, we generated DT40 cell line KOs of the other two proteins in the class (CENP-Q and CENP-R), and we compared the phenotypes of all of these KO cell lines. CENP-Q- or CENP-R-KO constructs were generated such that genomic fragments encoding several exons were replaced with drug-resistance cassettes (Supplemental Figure S1, A and B). We then transfected each construct into DT40 cells, and we isolated homozygous CENP-Q−/− or CENP-R−/−/− clones. We confirmed that CENP-Q or CENP-R protein was not expressed by Western blot analysis with anti-CENP-Q or anti-CENP-R antibody (Supplemental Figure S1, Ac and Bc). All KO cell lines of CENP-O class were viable. We then compared the proliferation rates of these KO cell lines with that of wild-type (WT) DT40 cells (Figure 1Figure 1.A). The proliferation rate of CENP-O–, -P–, -Q–, or -50–deficient cell lines was slower than that of WT cells. The doubling times of the KO cell lines were ~1–4 h longer than that of WT cells (WT, 9 h; CENP-50-KO, 10.2 h; CENP-O-KO, 10.4 h; CENP-P-KO, 10.4 h; CENP-Q-KO, 13.0 h; and CENP-R KO, 9.5 h). As discussed below, we think that these slow proliferation rates were due to increases of G2/M and apoptotic populations. The proliferation rate of CENP-R–deficient cells was similar to that of WT cells, suggesting that CENP-R may play a different role from that of other CENP-O class proteins.
Figure 1.
Figure 1.
Figure 1.
KO cells of CENP-O class proteins are viable, but they show mitotic defects. (A) Representative growth curves for WT cells; single-KO cells of CENP-O, -P, -Q, -R, or -50; and double-KO cells of CENP-P and CENP-50. Numbers of cells not stained with trypan (more ...)
To determine relations between CENP-O class proteins, we created a double-KO cell line (4-46), in which both CENP-P and CENP-50 proteins were not expressed by Western blot analysis (Supplemental Figure S1C). Cells from the 4-46 line were viable, and the proliferation rate (11.8 h) was slower than that of WT cells (Figure 1Figure 1.A). We did not observe strong additive effects by creation of a double mutant, suggesting that at least four proteins in the CENP-O class (CENP-O, -P, -Q, and -50) function in a single pathway.
To determine the mechanism for the slow proliferation of CENP-O–, -P–, -Q–, or -50–deficient cells and cells deficient in both CENP-P and CENP-50, we performed cell-cycle analysis of CENP-O class KO cells and WT cells. We measured cellular DNA content and DNA synthesis by fluorescence-activated cell sorting (FACS) after pulse labeling with bromodeoxyuridine (BrdU) (Figure 1Figure 1.B). We did not observe significant differences in the distribution of G1 phase fractions between CENP-O class KO cells and WT cells. However, the number of S phase fractions in each CENP-O class KO cell line was decreased, and the number of G2/M phase fractions was increased compared with that of WT cells. Twelve percent of WT cells were in G2/M phase, whereas ~20% of CENP-O–, -P–, -Q–, or -50–deficient cells were in G2/M phase, suggesting that it takes a longer time for CENP-O–, -P–, -Q–, or -50–deficient cells to progress through G2/M phase. There was also no additive accumulation of G2/M phase in cells deficient in CENP-P and CENP-50, supporting the idea that CENP-O, -P, -Q, and -50 proteins function in a single pathway. The cell cycle distribution of CENP-R–deficient cells was similar to that of WT cells, consistent with the result that the growth curve for CENP-R–deficient cells was similar to that for WT cells (Figure 1Figure 1., A and B).
We previously showed that a slight accumulation of the G2/M phase fraction in CENP-50–deficient cells is due to a defect in normal mitotic progression (Minoshima et al., 2005 blue right-pointing triangle). The doubling time of DT40 cells is ~8–12 h. Therefore, differences in population of G2/M phase between CENP-O class KO cells and WT cells of several percentages points correspond to approximately a 1-h difference on the growth curve. We then measured numbers of mitotic and apoptotic cells in cells with KOs of CENP-O class proteins (Figure 1Figure 1.C), and we observed increases of several percentage points for mitotic and apoptotic cells. These increases are consistent with the mitotic delay observed by live-cell observation described below (Figure 1Figure 1., E and F). We also observed that some CENP-Q–deficient cells (~8.5% observed by live-cell imaging) showed abnormal mitotic behavior, with chromosomes that failed to congress normally to the metaphase plate (Figure 1Figure 1.D), although many of the cells went through the cell cycle. These results also suggest that the G2/M increase observed in CENP-Q–deficient cells is due mainly to mitotic defects. However, we cannot exclude the possibility that cells with KOs of CENP-O class proteins have a defect in mitotic entry.
To analyze mitotic defects in individual cells, we transfected the histone H2B-RFP plasmid into CENP-Q–deficient cells to visualize nuclei and chromosomes and observed the behavior of individual cells by microscopy at 38°C. Microscopic images were typically obtained at 3-min intervals, and chromosomes within individual cells were observed through an entire mitotic cycle. Time-lapse observations of WT and CENP-Q–deficient cells are shown in Figure 1Figure 1.E and the Supplemental Movies. Selected frames of the chromosome dynamics from prophase to telophase are shown. We observed well-ordered metaphase plates (12 min) and normal cell division in WT cells. It took WT cells an average of 22.4 ± 5.6 min (±SD, n = 38) to progress from prophase to telophase. Some CENP-Q–deficient cells did not congress chromosomes to the metaphase plate (63 min in Figure 1Figure 1.E), and these cells took an average of 39.1 ± 19.3 min (±SD, n = 47) to complete mitosis from prophase (Figure 1Figure 1.F), twice that of WT cells and similar to that of CENP-50–deficient cells (Minoshima et al., 2005 blue right-pointing triangle). Thus, cells with KOs of CENP-O class proteins went through the cell cycle, and kinetochore-microtubule interaction and chromosome movements occurred. However, it took longer for KO cells to complete mitosis, and we observed increased numbers of apoptotic cells in KO cells cultures. Considering these data, we concluded that CENP-O class proteins are required for proper kinetochore function, although cells with KOs of CENP-O class proteins are viable.
Role of CENP-O Class Proteins in Kinetochore Assembly
CENP-O class proteins localize constitutively to the kinetochore throughout the cell cycle, and we are interested in how each CENP-O class protein is involved in kinetochore assembly. We generated antibodies against each CENP-O class protein, and we assessed the localization of these proteins in each KO cell line. We observed typical kinetochore staining with each CENP-O class antibody in WT cells, whereas kinetochore staining was abolished in CENP-O–, -P–, -Q–, or -50–deficient cells (Figure 2Figure 2.A). However, we still observed clear kinetochore staining of CENP-O, -P, -Q, and -50 in CENP-R–deficient cells (Figure 2Figure 2.A). These results indicate that kinetochore localization of CENP-O, -P, -Q, and -50 is interdependent and that localization of CENP-R occurs downstream of that of CENP-O, -P, -Q, and -50. These localization data strongly support our idea that CENP-R has a different role from the other CENP-O class proteins.
Figure 2.
Figure 2.
Figure 2.
Role of CENP-O class proteins in kinetochore assembly. (A) Immunofluorescence analysis of WT and CENP-50–, -O–, -P–, -Q–, or -R–deficient cells. Cells were stained with anti-CENP-50, anti-CENP-O, anti-CENP-P, anti- (more ...)
CENP-O class proteins were originally identified as members of the CENP-H/I complex, and we previously showed that some members of CENP-O class do not localize to kinetochores in CENP-H- or CENP-I–deficient cells (Minoshima et al., 2005 blue right-pointing triangle; Okada et al., 2006 blue right-pointing triangle), suggesting that localization of CENP-O class proteins occurs downstream of that of CENP-H and CENP-I. To test this idea, we assessed the localization of CENP-Q and CENP-R in CENP-H–deficient cells (Figure 2Figure 2.B). We also assessed the localization of CENP-H and CENP-I in CENP-Q– or CENP-R–deficient cells (Figure 2Figure 2.C). We observed that kinetochore staining of CENP-Q and CENP-R were abolished in CENP-H–deficient cells (+tet, Figure 2Figure 2.B), but CENP-H and CENP-I kinetochore staining was clear in CENP-Q– or CENP-R–deficient cells, indicating that localization of CENP-O class proteins occurs downstream of that of CENP-H and CENP-I. Thus, CENP-O class proteins are required for their own kinetochore localization. However, these proteins do not strongly affect the localization of essential kinetochore proteins CENP-A, CENP-C, CENP-H, CENP-I, and Ndc80 (Minoshima et al., 2005 blue right-pointing triangle; Okada et al., 2006 blue right-pointing triangle; unpublished results). These results support the idea that CENP-O class proteins are required for proper kinetochore function.
We also found that the amounts of both CENP-P and CENP-O were decreased in cells deficient in either CENP-O or CENP-P and that the amounts of both CENP-50 and CENP-Q were decreased in cells deficient in either CENP-50 or CENP-Q (Figure 2Figure 2.D). The amount of CENP-R was decreased in all cells with KOs of CENP-O class proteins (Figure 2Figure 2.D). These results suggest that CENP-P is closely associated with CENP-O and that CENP-50 is closely associated with CENP-Q, although kinetochore localization of these proteins is interdependent (Figure 2Figure 2.E). With respect to the stabilities of the CENP-O class proteins, we suggest that CENP-O, -P, -Q, and -50 form a discrete complex that functions downstream of CENP-H class proteins and upstream of CENP-R (Figure 2Figure 2.E).
CENP-O Class Proteins Form a Discrete Complex
KO cells of CENP-O class proteins, with the exception of CENP-R, show similar phenotypes. Our present results in KO cells also suggest that CENP-O class proteins form a discrete complex (Figure 2Figure 2.E). Therefore, we were interested to determine whether these proteins physically form a stable complex. To examine this possibility, we coexpressed CENP-O, CENP-P, and CENP-50 proteins and 6x histidine-tagged (6xHis) CENP-Q protein in bacteria with a polycistronic system (Tan, 2001 blue right-pointing triangle; Kline et al., 2006 blue right-pointing triangle) (Figure 3Figure 3.A). Samples purified with nickel affinity beads were subjected to gel filtration chromatography, and each fraction was analyzed by Western blot analysis with each CENP-O class antibody. The four proteins cofractionated as a single peak (Figure 3Figure 3.B). These results indicated that the four proteins form a stable complex. To test whether the stable complex associates with CENP-R, we coexpressed the above-mentioned four proteins with CENP-R simultaneously. The four proteins and CENP-R protein copurified with nickel affinity beads (Figure 3Figure 3.C), suggesting that CENP-R can associate with the stable complex.
Figure 3.
Figure 3.
Figure 3.
CENP-O class proteins form a discrete complex. (A) Coexpression of CENP-50, CENP-P, 6xHis- CENP-Q, and CENP-O with a polycistronic system in E. coli cells. The CENP-R construct that was cotransformed with the polycistronic constructs was used in C. (B) (more ...)
To confirm that the stable complex forms in DT40 cells, we created a cell line (3xFLAG-CENP-50) in which expression of CENP-50 was replaced with that of CENP-50-FLAG. Whole-cell extract was prepared from 3xFLAG-CENP-50 cells, and immunoprecipitation was performed with anti-FLAG antibody. Immunoprecipitates were subjected to gel filtration chromatography, and each fraction was analyzed by Western blot analysis with several antibodies against CENP-O class proteins. CENP-O class proteins cofractionated with one peak nearly coincident with that observed for the recombinant protein complex purified from bacterial cells (Figure 3Figure 3., B–D). CENP-50 was detected in both the peak fraction (Figure 3Figure 3.D, 14) and smaller fractions (Figure 3Figure 3.D, 15 and 16). Excess CENP-50-FLAG by overexpression in DT40 cells may not form a complex. Therefore, we created another cell line (3xFLAG-CENP-P) in which expression of CENP-P was replaced with that of CENP-P-FLAG, and we performed similar experiments. The strongest CENP-50 band was detected in the peak fraction (Figure 3Figure 3.E, 14). Thus, biochemical experiments in both E. coli and DT40 cells showed that CENP-O class proteins form a stable complex.
CENP-O Class Proteins Are Required for Recovery from Spindle Damage
Although we propose that CENP-O class proteins are required for proper kinetochore function, mutant cells lacking expression of CENP-O class proteins are viable. However, we found that some KO cells of CENP-O class proteins died after chromosome missegregation (Figure 1Figure 1.C). We previously showed that CENP-50 deficiency adversely affects mitotic exit after prolonged mitotic delay. If CENP-O class proteins, including CENP-50, function cooperatively, other CENP-O class KO cells should show a similar phenotype. We then treated CENP-O–, -P–, -Q–, -R–, or -50–deficient cells and WT cells with nocodazole for 12 h, washed the cells, and placed them into fresh media. Then, we examined the entry of cells into the next cell cycle (Figure 4Figure 4.A). As shown previously (Minoshima et al., 2005 blue right-pointing triangle), WT cells progressed to G1 phase immediately, and <10% were mitotic at 6 h after release from nocodazole block (Figure 4Figure 4.A). In contrast, ~50% of CENP-O–, -P–, -Q–, or -50–deficient cells showed mitotic figures at 4 h after release, and 30–50% of these cells were still mitotic at 6 h after release (Figure 4Figure 4.A). We also observed that most cells that did not exit mitosis after several hours eventually died (data not shown). The percentage of mitotic CENP-R–deficient cells at 4 h after release was ~30%, slightly higher than that of WT cells but lower than that of CENP-O–, -P–, -Q–, or -50–deficient cells (Figure 4Figure 4.A).
Figure 4.
Figure 4.
Figure 4.
CENP-O class proteins are required for recovery from spindle damage. (A) WT cells and KO cells of CENP-O class proteins were treated with nocodazole for 12 h. After removal of nocodazole by washing, cells were cultured, and the percentages of mitotic, (more ...)
The observed phenotype indicates that CENP-O, -P, -Q, and -50 are required for recovery from spindle damage. We propose that CENP-50 functions cooperatively with CENP-O, -P, and -Q (and in part with CENP-R). An important question is how CENP-50 and other CENP-O class proteins are involved in the recovery from spindle damage. We previously reported that the distance between sister kinetochores in CENP-50–deficient cells is increased after long nocodazole treatment (Minoshima et al., 2005 blue right-pointing triangle). To gain further information regarding the function of CENP-O class proteins, we added the proteasome inhibitor MG132 in the culture of cells released from nocodazole block. In the absence of MG132, we observed severe mitotic defects 4 h after release from nocodazole block in CENP-50-KO cells (Figure 4Figure 4.B, right middle). We observed many chromosomes (~10 in each mitosis) that did not align at the metaphase plate and they were scattered (Figure 4Figure 4.B, right middle). In the presence of MG132, most chromosomes aligned at the metaphase plate 4 h after release from nocodazole block in CENP-50-KO cells (Figure 4Figure 4., B, right, bottom, and C). We also quantified aligned chromosomes in the presence or absence of MG132 (Figure 4Figure 4.C). These results indicate that MG132 treatment rescued the chromosome misalignment phenotype that occurred during release from nocodazole block in CENP-50–deficient cells. MG132 is a proteasome inhibitor that acts to protect proteins involved in sister chromatid adhesion from degradation. Our present data suggest that CENP-50 is involved in the prevention of premature sister chromatid separation during the release from nocodazole block. We observed that mitotic cells showed positive mitotic checkpoint signals during the release from nocodazole block (Figure 4Figure 4.D). Therefore, CENP-50-KO cells do not show silencing of the spindle checkpoint pathway during the release from nocodazole block. We would like to note that although MG132 treatment rescued the chromosome misalignment phenotype that occurred during release from nocodazole block in CENP-50–deficient cells, cells were still mitotic even after removing of MG132 (data not shown). This suggests that that some irreversible mitotic defects also occurred during the nocodazole block.
Phosphorylation of CENP-50 Is Essential for Recovery from Spindle Damage
Our data suggested that CENP-50 is involved in the prevention of premature sister chromatid separation and for recovery from spindle damage. We found that CENP-50 is highly phosphorylated in cells treated with nocodazole (Figure 5Figure 5.Aa). Therefore, the recovery from spindle damage may be related to phosphorylation of CENP-50. There are many potential phosphorylation sites in the amino acid CENP-50 sequence (Supplemental Figure S2). Recently, Kang et al. (2006) blue right-pointing triangle reported that human CENP-50 (referred as PBIP1) is a target of polo-like kinase 1 (PLK1) in human HeLa cells. We also found that CENP-O class proteins copurified with PLK1 by a proteomics approach (Hori, Cheeseman, and Fukagawa, unpublished observations). Kang et al. (2006) blue right-pointing triangle also reported that S77 and T78 in human CENP-50 are critical sites for association or phosphorylation by PLK1. We found that S77 and T78 in human CENP-50 are highly conserved sites, corresponding to S62 and T63 in chicken CENP-50 (Figure 5Figure 5., A–C). We created a mutant cDNA of chicken CENP-50 in which two potential serine/threonine phosphorylation sites are substituted with alanine (62.63A cDNA) and transfected it into tetracycline-repressible conditional-KO cells of CENP-50 (17-15). We isolated a CENP-50 KO cell line expressing 62.63A cDNA (62.63A/17-15 cells), and we confirmed that phosphorylation of CENP-50 was decreased in these cells (Figure 5Figure 5., A and B). We also transfected WT cDNA into 17-15 cells and isolated a cell line expressing WT cDNA (WT/17-15 cells). We then analyzed the recovery of these cells from spindle damage. More than 70% of 62.63A/17-15 cells showed mitotic figures at 4 h after release from nocodazole block (Figure 5Figure 5.B, right), similar to 17-15 cells (+tet; Figure 5Figure 5.B, left), but WT/17-15 cells rescued by WT cDNA expression recovered completely from spindle damage (Figure 5Figure 5.B, middle), suggesting that phosphorylation of S62 and T63 in chicken CENP-50 is essential for complete recovery.
Figure 5.
Figure 5.
Figure 5.
Phosphorylation of CENP-50 by PLK1 is essential for recovery from spindle damage. (Aa) Low-mobility bands were detected by Western blot analysis with anti-CENP-50 antibody. The amounts of the low-mobility bands (P) were increased in cells enriched in (more ...)
CENP-50 functions in association with other CENP-O class proteins. Therefore, it is possible that 62.63A mutation of CENP-50 affects kinetochore localization or protein amounts of CENP-O class proteins. Degradation or mislocalization of CENP-O class proteins may affect the ability to recover from spindle damage. Therefore, we performed Western blot analysis to examine the stability of CENP-O class proteins in 62.63A/17-15 cells (Figure 5Figure 5.C). No significant change in the amounts of CENP-O class proteins was observed in 62.63A/17-15 cells compared with WT cells. Interestingly, we detected a slow-mobility band for CENP-Q, which may correspond to phosphorylated protein. Because we CENP-Q is associated with CENP-50, PLK1 may phosphorylate CENP-Q. The amount of phosphorylated CENP-Q band was reduced in 62.63A/17-15 cells, suggesting that PLK1 localization may be altered in these cells. We also examined the localization of CENP-O class proteins in 62.63A/17-15 cells. Mutant CENP-50 (62.63A) proteins localized to the kinetochore throughout the cell cycle, and kinetochore localization of other CENP-O class proteins was not changed in 62.63A/17-15 cells (Figure 5Figure 5.D), indicating that kinetochore localization of CENP-O class proteins was not affected by 62.63A mutation of CENP-50.
As mentioned, there are many potential phosphorylation sites in the CENP-50 sequence (Supplemental Figure S2). Because we found that CENP-50 phosphorylation was reduced by treatment of a cyclin-dependent kinase (CDK) inhibitor, we focused on potential CDK phosphorylation sites (T99, S112, S145, S148, and S151) and predicted additional phosphorylation sites around the CDK sites by using the NetPhos 2.0 software (http://www.cbs.dtu.dk/services/NetPhos/). In total, 16 potential serine/threonine phosphorylation sites in the chicken CENP-50 sequence were predicted (Supplemental Figure S2). We created a mutant cDNA of chicken CENP-50 in which all 16 of these sites were substituted with alanine (16×A cDNA), and we transfected it into tetracycline-repressible 17-15 cells. We isolated a cell line expressing 16×A cDNA(16×A/17-15 cells), and we confirmed that the amount of the phosphorylated band of CENP-50 was reduced (Figure 5Figure 5.E). However, 16×A/17-15 cells showed recovery from spindle damage, similar to WT cells (Figure 5Figure 5.E). We added the 62.63A mutations to the 16×A cDNA (16×A + 62.63A cDNA) and isolated a cell line expressing 16×A + 62.63A cDNA (16×A + 62.63A/17-15 cells), in which we detected defects in the recovery from spindle damage. These data suggest that S62 and T63 in CENP-50, which are putative phosphorylation sites, are essential for the recovery from spindle damage.
Kinetochore Localization of PLK1 Is Dependent on CENP-O Class Proteins, Especially in the G2 Phase
We showed that the amount of phosphorylated CENP-Q was reduced in 62.63A/17-15 cells expressing CENP-50 cDNA with mutated phosphorylation sites (Figure 5Figure 5.C). Therefore, PLK1 localization may be altered in cells lacking phosphorylation of CENP-50. We created a GFP-fused PLK1 construct under the control of the cytomegalovirus promoter and transfected it into WT cells, CENP-50-KO cells, and 62.63A/17-15 cells (Figure 6Figure 6.). PLK1 localizes to both the centrosomes and kinetochores. However, we focused on the kinetochore localization of PLK1 in this article, because CENP-O class proteins localize to kinetochores not to the centrosomes. PLK1 initially localized to the kinetochores in G2 phase and continued to localize to the kinetochores in prometaphase, but kinetochore-associated signals were very faint when the chromosomes aligned at the metaphase plate (WT in Figure 6Figure 6.). In CENP-50–deficient cells, kinetochore localization of PLK1 in G2 nuclei was not visible. In CENP-50–deficient prometaphase cells, kinetochore localization of PLK1 was weaker that in WT prometaphase cells, but it was visible (Figure 6Figure 6.). In 62.63A/17-15 cells, kinetochore localization of PLK1 was not visible in G2 nuclei and faint in prometaphase (Figure 6Figure 6.). These data suggest that phosphorylation of S62 and T63 in CENP-50 is required for the kinetochore localization of PLK1, especially in the G2 phase. Our data also suggest that there are several localization pathways of PLK1 to kinetochore in prometaphase and that one pathway is dependent on phosphorylation of S62 and T63 in CENP-50.
Figure 6.
Figure 6.
Figure 6.
Kinetochore localization of PLK1 is dependent on CENP-O class proteins in G2 phase. PLK1-GFP was expressed in WT, CENP-50-KO, and 62.63A/17-15 cells. Kinetochore localization is denoted by arrowheads, and centrosome localization is denoted by arrows. (more ...)
DISCUSSION
We previously isolated a CENP-H/I complex containing at least 11 components, and we found that this complex can be divided into three functional classes based on phenotype analysis (Okada et al., 2006 blue right-pointing triangle). We defined CENP-O, -P, -Q, -R, and -50 as CENP-O class proteins. In the present study, we created single KO cell lines for CENP-Q and CENP-R and a double-KO cell line for CENP-P and CENP-50. We analyzed all KO cell lines for CENP-O class proteins. With a combination of genetic and biochemical approaches, we showed that CENP-O class proteins form a stable complex. Protein stability experiments with KO cells indicated that CENP-P is closely associated with CENP-O, CENP-50 is closely associated with CENP-Q, and CENP-R functions downstream of all other CENP-O class proteins (Figure 2Figure 2.E). The elucidation of molecular structure of CENP-O class proteins will likely aid in the understanding of kinetochore assembly.
CENP-O class proteins are not required for cell proliferation under normal conditions. However, they seem to be required for recovery from spindle damage. We previously reported that CENP-50 is involved in sister chromatid adhesion during spindle checkpoint activation (Minoshima et al., 2005 blue right-pointing triangle). In the present study, we showed that all CENP-O class proteins except for CENP-R are essential for recovery from spindle damage, and that treatment with the proteasome inhibitor MG132 partially rescued the severe mitotic phenotype observed in response to release from nocodazole block in CENP-50–deficient cells. We also showed that phosphorylation of S62 and T63 in CENP-50 is essential for the recovery from spindle damage. We therefore propose a model for the function of CENP-O class proteins in the recovery from spindle damage (Figure 7Figure 7.). In this model, CENP-O class proteins are phosphorylated by PLK1 in response to spindles damage. When cells are cultured in fresh media, phosphorylated CENP-O class proteins by PLK1 protect several factors involved in sister chromatid adhesion from degradation by proteasome activity. We think that CENP-O class proteins function to protect against premature sister chromatid separation until the microtubules attach properly to the kinetochores. In CENP-O class KO cells or cells expressing mutant CENP-50, proteasomes degrade several factors before the microtubules attach properly to the kinetochores (Figure 7Figure 7.). This model may explain why we observed an increased distance between sister kinetochores in CENP-50–deficient cells after long nocodazole treatment (Minoshima et al., 2005 blue right-pointing triangle). There are reports in which the pathway of cohesin removal, which requires PLK1, is proteolysis independent (Tang et al., 2004 blue right-pointing triangle; Kitajima et al., 2005 blue right-pointing triangle; Resnick et al., 2006 blue right-pointing triangle). Our findings related to CENP-O class proteins are distinct from the proteolysis-independent pathway of cohesin removal.
Figure 7.
Figure 7.
Figure 7.
Model of the functional role of CENP-O class proteins during release from spindle damage. CENP-O class proteins are phosphorylated by PLK1 in response to spindle damage. When cells are cultured in fresh media, phosphorylated CENP-O class by PLK1 protect (more ...)
We sometimes observed mitotic defects in nonnocodazole treated CENP-Q–deficient cells (Figure 1Figure 1.). We think that these defects were similar defects observed in nocodazole-treated/washout cells with KO of CENP-O class proteins. Although spindle damage does not happen frequently in fresh media, cells occasionally could have some spindle damage. At that time, CENP-O class proteins function in recovery from spindle damage. CENP-O class proteins seem to fine-tune proper chromosome segregation rather than to act as essential structural components for kinetochore assembly. Therefore, CENP-O class KO cells are viable when cultured in fresh media. However, RNA interference analysis for CENP-50 (U) shows defects in chromosome alignment and colony formation in human HeLa cells (Foltz et al., 2006 blue right-pointing triangle). We have created CENP-50-KO mice, and we observed that these mice die during early embryogenesis (our unpublished data). Cell division early during embryogenesis must be performed under strict conditions for completion of several developmental steps such as differentiation and proliferation. The function of CENP-O class proteins may be essential under these conditions. It will be interesting to determine how the mechanism of chromosome segregation differs between differentiated and undifferentiated cells. We think that CENP-O class proteins are key players in relation to this question. Determination of the molecular and biochemical properties of the CENP-O class complex will be necessary to understand the biological function of this complex.
Supplementary Material
[Supplemental Materials]
Click here to view.
ACKNOWLEDGMENTS
We thank K. Suzuki, M. Takahashi, and K. Kita for technical assistance. We also thank S. Tan for pST39 plasmid and I. M. Cheeseman for unpublished information about PLK1. This work was supported by grants-in-aid for scientific research from the Ministry of Education, Science, Sports and Culture of Japan.
Footnotes
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/10.1091/mbc.E07-06-0556) on December 19, 2007.
REFERENCES
  • Chan G. K., Liu S. T., Yen T. J. Kinetochore structure and function. Trends Cell Biol. 2005;15:589–598. [PubMed]
  • Cheeseman I. M., Niessen S., Anderson S., Hyndman F., Yates J.R., III, Oegema K., Desai A. A conserved protein network controls assembly of the outer kinetochore and its ability to sustain tension. Genes Dev. 2004;18:2255–2268. [PubMed]
  • Cleveland D. W., Mao Y., Sullivan K. F. Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell. 2003;112:407–421. [PubMed]
  • Foltz D. R., Jansen L. E., Black B. E., Bailey A. O., Yates J. R., III, Cleveland D. W. The human CENP-A centromeric nucleosome-associated complex. Nat. Cell Biol. 2006;8:458–469. [PubMed]
  • Fukagawa T. Assembly of kinetochore in vertebrate cells. Exp. Cell Res. 2004;296:21–27. [PubMed]
  • Fukagawa T., Nogami M., Yoshikawa M., Ikeno M., Okazaki T., Takami Y., Nakayama T., Oshimura M. Dicer is essential for formation of the heterochromatin structure in vertebrate cells. Nat. Cell Biol. 2004;6:784–791. [PubMed]
  • Fukagawa T., Pendon C., Morris J., Brown W. CENP-C is necessary but not sufficient to induce formation of functional centromere. EMBO J. 1999;18:4196–4209. [PubMed]
  • Hori T., Haraguchi T., Hiraoka Y., Kimura H., Fukagawa T. Dynamic behavior of Nuf2-Hec1 complex that localizes to the centrosome and centromere and is essential for mitotic progression in vertebrate cells. J. Cell Sci. 2003;116:3347–3362. [PubMed]
  • Izuta H., Ikeno M., Suzuki N., Tomonaga T., Nozaki N., Obuse C., Kisu Y., Goshima N., Nomura F., Nomura N., Yoda K. Comprehensive analysis of the ICEN (interphase centromere complex) components enriched in the CENP-A chromatin of human cells. Genes Cells. 2006;11:673–684. [PubMed]
  • Kang Y. H., et al. Self-regulated Plk1 recruitment to kinetochores by the Plk1-PBIP1 interaction is critical for proper chromosome segregation. Mol. Cell. 2006;24:409–422. [PubMed]
  • Kitajima T. S., Hauf S., Ohsugi M., Yamamoto T., Watanabe Y. Human Bub1 defines the persistent cohesion site along the mitotic chromosome by affecting Shugoshin localization. Curr. Biol. 2005;15:353–359. [PubMed]
  • Kline S. L., Cheeseman I. M., Hori T., Fukagawa T., Desai A. The human Mis12 complex is required for kinetochore assembly and proper chromosome segregation. J. Cell Biol. 2006;173:9–17. [PubMed]
  • McAinsh A. D., Meraldi P., Draviam V. M., Tosa A., Sorger P. K. The human kinetochore proteins Nnf1R and Mcm21R are required for accurate chromosome segregation. EMBO J. 2006;25:4033–4049. [PubMed]
  • Meraldi P., McAinsh A. D., Rheinbay E., Sorger P. K. Phylogenetic and structural analysis of centromeric DNA and kinetochore proteins. Genome Biol. 2006;7:R23. [PubMed]
  • Minoshima Y., Hori T., Okada M., Kimura H., Haraguchi T., Hiraoka Y., Bao Y. C., Kawashima T., Kitamura T., Fukagawa T. The constitutive centromere component CENP-50 is required for recovery from spindle damage. Mol. Cell. Biol. 2005;25:10315–10328. [PubMed]
  • Nishihashi A., Haraguchi T., Hiraoka Y., Ikemura T., Regnier V., Dodson H., Earnshaw W. C., Fukagawa T. CENP-I is essential for centromere function in vertebrate cells. Dev. Cell. 2002;2:463–476. [PubMed]
  • Obuse C., Yang H., Nozaki N., Goto S., Okazaki T., Yoda K. Proteomics analysis of the centromere complex from HeLa interphase cells: UV-damaged DNA binding protein 1 (DDB-1) is a component of the CEN-complex, while BMI-1 is transiently co-localized with the centromeric region in interphase. Genes Cells. 2004a;9:105–120. [PubMed]
  • Obuse C., Iwasaki O., Kiyomitsu T., Goshima G., Toyoda Y., Yanagida M. A conserved Mis12 centromere complex is linked to heterochromatic HP1 and outer kinetochore protein Zwint-1. Nat. Cell Biol. 2004b;6:1135–1141. [PubMed]
  • Okada M., Cheeseman I. M., Hori T., Okawa K., McLeod I. X., Yates J.R., III, Desai A., Fukagawa T. The CENP-H-I complex is required for the efficient incorporation of newly synthesized CENP-A into centromeres. Nat. Cell Biol. 2006;8:446–457. [PubMed]
  • Resnick T. D., Satinover D. L., MacIsaac F., Stukenberg P. T., Earnshaw W. C., Orr-Weaver T. L., Carmena M. INCENP and Aurora B promote meiotic sister chromatid cohesion through localization of the Shugoshin MEI-S332 in Drosophila. Dev. Cell. 2006;11:57–68. [PubMed]
  • Tan S. A modular polycistronic expression system for overexpressing protein complexes in Escherichia coli. Protein Expr. Purif. 2001;21:224–234. [PubMed]
  • Tang Z., Sun Y., Harley S. E., Zou H., Yu H. Human Bub1 protects centromeric sister-chromatid cohesion through Shugoshin during mitosis. Proc. Natl. Acad. Sci. USA. 2004;101:18012–18017. [PubMed]
  • Yuen K. W., Montpetit B., Hieter P. The kinetochore and cancer: what's the connection? Curr. Opin. Cell Biol. 2005;17:576–582. [PubMed]
Articles from Molecular Biology of the Cell are provided here courtesy of
American Society for Cell Biology

PubMed articles by these authors
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph
See more articles cited in this paragraph