In budding yeast Saccharomyces cerevisiae, kinetochores are attached by microtubules during most of the cell cycle, but the duplication of centromeric DNA disassembles kinetochores, which results in a brief dissociation of chromosomes from microtubules. Kinetochore assembly is delayed in the presence of hydroxyurea, a DNA synthesis inhibitor, presumably due to the longer time required for centromeric DNA duplication. Some kinetochore mutants are sensitive to stressful DNA replication as these kinetochore proteins become essential for the establishment of the kinetochore–microtubule interaction after treatment with hydroxyurea. To identify more genes required for the efficient kinetochore–microtubule interaction under stressful DNA replication conditions, we carried out a genome-wide screen for yeast mutants sensitive to hydroxyurea. From this screen, cik1 and kar3 mutants were isolated. Kar3 is the minus-end-directed motor protein; Cik1 binds to Kar3 and is required for its motor function. After exposure to hydroxyurea, cik1 and kar3 mutant cells exhibit normal DNA synthesis kinetics, but they display a significant anaphase entry delay. Our results indicate that cik1 cells exhibit a defect in the establishment of chromosome bipolar attachment in the presence of hydroxyurea. Since Kar3 has been shown to drive the poleward chromosome movement along microtubules, our data support the possibility that this chromosome movement promotes chromosome bipolar attachment after stressful DNA replication.
THE attachment of chromosomes by microtubules is essential for chromosome segregation and the kinetochore is a protein complex that associates with centromeric DNA to mediate this attachment. In higher eukaryotic cells, the establishment of kinetochore–microtubule (KT–MT) interaction occurs during metaphase when DNA has been duplicated and compacted; however, budding yeast kinetochores are associated with microtubules during most of the cell cycle (Janke et al. 2002; Li et al. 2002). The KT–MT interaction in budding yeast must be disrupted when centromeric DNA is being duplicated. Indeed, previous data indicate that centromeric DNA shows a transient detachment from microtubules during S-phase (Tanaka et al. 2007). Therefore, the reestablishment of the KT–MT interaction in budding yeast takes place in S-phase just after centromere duplication and the subsequent kinetochore assembly.
The establishment of KT–MT interaction is a multistep process. First, one of the sister kinetochores is captured by the side of a microtubule to establish side-on binding. In budding yeast, the captured kinetochore moves toward the spindle poles with the assistance of the minus-end-directed motor protein Kar3 (Tanaka et al. 2005). During this movement, the side-on binding could be switched to end-on binding, which depends on the DASH/Dam1, a kinetochore complex that associates with microtubules before the establishment of KT–MT interaction (Janke et al. 2002; Li et al. 2002; Tanaka et al. 2007). Recent data suggest that the accumulated Stu1 protein on unattached kinetochores may facilitate chromosome capture, and Stu1 could be the protein responsible for the initial KT–MT interaction on the basis of its microtubule-binding nature (Ortiz et al. 2009). After chromosomes are moved close to the spindle pole, they become bipolar attached, but it is less clear whether this poleward movement facilitates bipolar attachment. One possibility is that chromosomes close to one spindle pole are easily captured by microtubules emanating from the other pole.
The duplication of centromeric DNA results in dissociation of kinetochore proteins from the centromere. Hydroxyurea (HU) slows down DNA synthesis by depleting the pool of dNTPs, the basic unit of DNA. The examination of the interaction of kinetochore proteins with centromeric DNA in budding yeast revealed that HU treatment delays kinetochore reassembly, presumably due to slower centromeric DNA replication (Liu et al. 2008). Because of this delay, we expect more detached chromosomes in HU-treated yeast cells. After kinetochore assembly, one of the sister kinetochores is captured by a microtubule and moved close to the spindle pole (Tanaka et al. 2007). Therefore, this poleward chromosome transport might become critical for the reestablishment of KT–MT interaction after HU treatment.
We have found that mutation in a kinetochore protein Ask1 leads to sensitivity to HU and ask1-3 mutant cells show difficulty in establishing correct KT–MT interaction after HU treatment (Li et al. 2002; Liu et al. 2008). To identify additional yeast genes required for the efficient KT–MT interaction after HU treatment, we performed a genome-wide screen for HU-sensitive mutants. Interestingly, cik1Δ and kar3Δ mutants were found to be sensitive to HU. Kar3 is one of the six kinesin proteins in budding yeast and it was first found to be essential for yeast nuclear fusion during mating (Meluh and Rose 1990; Molk et al. 2006). Unlike other kinesins, Kar3 protein contains a motor domain at its carboxy terminus that possesses minus-end-directed motility (Endow et al. 1994). Two proteins, Cik1 and Vik1, associate with Kar3. Cik1 is required for the localization of Kar3 to microtubule-associated structures, whereas Vik1 aids in the localization of Kar3 at spindle poles (Page et al. 1994; Manning et al. 1999; Barrett et al. 2000). Kar3 protein forms a heterodimer with either Vik1 or Cik1, which is required for association of Kar3 with microtubules and its motility (Chu et al. 2005; Allingham et al. 2007). Recent evidence indicates that Kar3 is also responsible for the poleward chromosome movement (Tanaka et al. 2005). In addition to its motor function, Cik1/Kar3 complex localizes at the spindle midzone as an interpolar microtubule crosslinker to prevent spindle collapse (Gardner et al. 2008b). In other eukaryotic cells, the homologs of Kar3 promote the bundling of parallel microtubules as well as the sliding of antiparallel microtubules (Braun et al. 2009; Fink et al. 2009). This function may explain the abnormal dot-like spindle structure in cik1 and kar3 mutants (Meluh and Rose 1990; Page and Snyder 1992). Together, these observations indicate that Kar3 forms distinct complexes with Cik1 and Vik1 to participate in different microtubule-mediated events, such as mating, spindle morphogenesis, and chromosome segregation.
Although we found that cik1Δ and kar3Δ mutants are very sensitive to HU, deletion of another Kar3 interacting protein Vik1 and other microtubule motor proteins did not result in any HU sensitivity. Our evidence indicates that the HU sensitivity is not a consequence of impaired DNA replication. Instead, the cik1Δ and kar3Δ mutant cells fail to establish proper KT–MT interaction after exposure to HU, as indicated by abnormally distributed kinetochore proteins. Moreover, we generated cik1HU mutants that exhibit similar HU sensitivity to cik1Δ but show relatively normal spindle structure, indicating the abnormal spindle structure may not be the cause of HU sensitivity. Therefore, Cik1/Kar3-dependent chromosome transport might be required for the efficient establishment of proper KT–MT interaction after stressful DNA replication.
MATERIALS AND METHODS
Yeast strains, growth, and media:
The relevant genotypes and sources of the yeast strains used in this study are listed in Table 1. All the strains listed are isogenic to Y300, a W303 derivative. To arrest yeast cells in G1 phase, 5 μg/ml α-factor was added into cell cultures in midlog phase in YPD (pH = 3.9) for 2.5 hr. The cells were centrifuged and washed once with water to release from G1 arrest. Hydroxyurea was purchased from ACROS Organics.
The preparation of yeast protein samples was described previously (Liu and Wang 2006). Protein samples were resolved by 10% SDS–PAGE. Primary antibodies (anti-myc) were purchased from Covance (Madison, WI), and anti-Pgk1 antibody was from Molecular Probes (Eugene, OR). The HRP-conjugated secondary antibody was purchased from Jackson ImmunoResearch (West Grove, PA).
Collected cells were fixed with 3.7% formaldehyde for 5 min at room temperature. The cells were washed once with 1× PBS (pH 7.2) and then resuspended in 1× PBS buffer to examine fluorescence signals with a microscope (Zeiss Axioplan 2) (Carl Zeiss, Thornwood, NY).
Screen of HU-sensitive mutants:
Yeast cells of the ∼4700 deletion strains from American Type Culture Collection (ATCC; Rockville, MD) were spotted onto YPD plates with and without 100 mm of HU. After a 3-day incubation at 25°, the growth of these yeast cells was examined and the strains that showed slow or no growth were selected. All the HU-sensitive mutants are listed in supporting information, Table S1. “+” means that the mutant cells show slow growth on HU plates. “++” means no growth.
The construction of pRS416-CIK1 plasmid:
Yeast genomic DNA was used as template for the PCR reaction with a pair of CIK1 gene-specific primers (forward primer, 5′-GAT TCC CCG CGG GGA CTG TTA GTC CCG TAA CAT T-3′; reverse primer, 5′-GCT ACC ATC GAT GGT GCG GTT GAT TCG TTT TAT A-3′). The forward and reverse primers contain SacII and ClaI restriction enzyme sites, respectively. The PCR products and pRS416 vector were digested with both SacII and ClaI (Sikorski and Hieter 1989). After recovery from the agarose gel, the fragments were ligated for the transformation to Escherichia coli.
Mutagenesis PCR of CIK1:
The PCR reaction mixture with less dATP (−A) is 25 μl that contains 0.5 μl of 20 ng/μl pRS416-CIK1 plasmid template; 0.5 μl 20-μm primers (the same primers used for the construction of pRS416-CIK1 plasmid); 1 μl 12.5 mm dTTP, dGTP, and dCTP and 1 μl 2.5 mm dATP; 2.5 μl 10× Mg2+ free Taq DNA polymerase buffer; 2.5 μl 25 mm MgCl2; 0.94 μl 20 mm MnCl2; 0.5 μl Taq DNA polymerase; and H2O. Similarly, the −T, −G, and −C PCR mixtures contain less concentrated dTTP, dGTP, and dCTP, respectively. After PCR reaction, the four PCR products were mixed together. The constructed pRS416-CIK1 plasmid was digested with PacI (one single cut within the CIK1 gene) and followed by the treatment with calf intestinal phosphatase (CIP) to prevent self ligation. The treated pRS416-CIK1 plasmid was transformed into a cik1Δ strain along with the PCR mixture and the transformants were selected on URA dropout plates. The mutated cik1 genes through the biased PCR reaction would insert into the linearized plasmid by recombination. The growing colonies were replica copied onto HU (100 mm) and YPD plates and incubated at 25° and 37°, respectively. The HU- and temperature-sensitive mutants were selected. The plasmids containing the mutated cik1 gene were recovered and reintroduced into cik1Δ mutant to confirm the phenotype. Then the plasmids were sequenced to determine the mutation sites.
To construct cik1HU mutant strains, the pRS416-cik1HU plasmids were digested with ClaI and SacII and the cik1HU genes were inserted into an integration vector pRS403 (Sikorski and Hieter 1989). The resulting pRS403-cik1HU plasmids were digested with NdeI and transformed into a cik1Δ strain. The transformants were selected on HIS dropout plates and the HU sensitivity was confirmed after transformation.
cik1Δ and kar3Δ mutants exhibit HU sensitivity:
ask1-1 was isolated as a HU-sensitive mutant and Ask1 turned out to be a kinetochore protein (Alcasabas et al. 2001; Li et al. 2002). To identify more kinetochore mutants that exhibit HU sensitivity, we carried out a genome-wide screen for mutants that are sensitive to HU from the ATCC collection of yeast deletion mutants, and >100 HU-sensitive mutants were identified (Table S1). The cik1Δ mutant was found to exhibit dramatic HU sensitivity (Figure 1A). Kar3 is one of the kinesin proteins in budding yeast and both Cik1 and Vik1 form a complex with Kar3 (Manning et al. 1999), but our ATCC collection does not include the kar3Δ deletion strain. Nevertheless, deletion of VIK1 did not show any HU sensitivity. We also examined the HU sensitivity of other kinesin and dynein mutants, cin8Δ, kip1Δ, kip2Δ, kip3Δ, and dyn1Δ (Saunders et al. 1995), and none of them showed HU sensitivity (Figure 1A). To determine the HU sensitivity of kar3 mutants, we generated cik1Δ and kar3Δ mutants with the Y300 strain, a W303 derivative, by using a one-step PCR approach (Longtine et al. 1998). Interestingly, both cik1Δ and kar3Δ mutants exhibited HU sensitivity. Moreover, a kar3-1 mutant was also sensitive to HU (Figure 1A). Therefore, we conclude that the loss of function of the Cik1/Kar3 complex leads to sensitivity to stressful DNA synthesis.
One possibility for the HU sensitivity of cik1 and kar3 mutants is that DNA replication is compromised and the treatment with HU exacerbates the defect. Thus, we first compared the DNA replication kinetics in synchronous wild-type (WT) and cik1Δ mutant cells by fluorescence-activated cell sorting (FACS) analysis, but no difference was observed (data not shown). Then we compared the DNA synthesis kinetics in WT and cik1Δ mutant cells after HU treatment. G1-arrested cells were released into 200 mM HU medium for 100 min and DNA replication was monitored by using pulsed-field gel electrophoresis (PFGE) after HU was washed off (Liu and Wang 2006). The chromosomal DNA from both WT and cik1Δ cells was able to run into the agarose gel after release from HU for 80 min, indicating the completion of DNA synthesis (Figure 1B). Therefore, cik1Δ mutants showed no DNA synthesis defects with or without HU treatment.
Another possibility for the HU sensitivity is that Cik1 and Kar3 function in the S-phase checkpoint and the impaired checkpoint function contributes to the HU sensitivity. Thus, we determined the viability of cik1Δ and kar3Δ mutant cells in the presence of HU. After cells were incubated in 200 mm HU for 6 hr, we did not observe significant viability loss; 60% of cik1Δ and 80% of kar3Δ cells were able to form colonies, suggesting that the S-phase checkpoint is proficient in cik1Δ and kar3Δ mutant cells (Figure 1C). As a control, most of the rad53-21 mutant cells lost viability after incubation for 2 hr (Sanchez et al. 1996; Wang and Elledge 1999). Therefore, it is unlikely that the HU sensitivity of cik1Δ and kar3Δ mutants is a consequence of impaired S-phase checkpoint.
Although we did not observe obvious DNA synthesis defects with FACS and PFGE methods, we could not exclude the possibility that cik1 and kar3 mutant cells show specific defects in the duplication of centromeric DNA, which could contribute to their HU sensitivity. If that is the case, we expect delayed centromere separation in cik1Δ mutants as the separation of centromeric DNA depends on centromere duplication. A CEN5-GFP strain was generated in the Nasmyth laboratory, wherein the tetO array was inserted into the genome that is 1.4 kb away from the centromere on chromosome V. The expressed tetR-GFP fusion protein binds to the tetO array to mark CEN5 (Tanaka et al. 2000; Zhang et al. 2006). Therefore, the separation of Cen5-GFP will indicate the completion of CEN5 duplication. cik1Δ mutant cells exhibit mitotic defects, which delay the separation of Cen5-GFP by activating of the spindle assembly checkpoint, but the absence of sister chromatid cohesion will allow the generation of two Cen5-GFP dots immediately after CEN5 duplication even when the checkpoint is active (Michaelis et al. 1997). Thus, we compared the separation of Cen5-GFP in scc1-73 and scc1-73 cik1Δ cells after G1 release into 37° medium, which inactivates cohesin Scc1 and prevents the generation of sister chromatid cohesion. We found that the kinetics of the appearance of two GFP dots was identical in WT and cik1Δ mutant cells (Figure 1D), a strong indication that there is no delay in centromere duplication in cik1Δ cells. Therefore, we conclude that the HU sensitivity of cik1 and kar3 mutants is not due to the defects in DNA replication.
cik1 and kar3 mutants exhibit anaphase entry delay after HU treatment:
To further address why cik1Δ and kar3Δ mutants are sensitive to HU treatment, we examined the cell cycle progression in synchronous WT, cik1Δ, and kar3Δ cells by determining the budding index and Pds1 protein level, as the degradation of Pds1 marks anaphase entry (Cohen-Fix et al. 1996). After G1 release for 140 min, the majority of WT cells divided and became unbudded, indicating the completion of the cell cycle. However, ∼40% of cik1Δ and kar3Δ cells were still large budded after G1 release for 180 min. Pds1 protein disappeared 80 min after G1 release in WT cells, but cik1Δ and kar3Δ mutant cells displayed high levels of Pds1 even after G1 release for 180 min (Figure 2A). The failure of Pds1 protein degradation in cik1Δ and kar3Δ mutant cells indicates an anaphase entry defect in these mutants.
Because cik1Δ and kar3Δ mutant cells are HU sensitive, we speculated that cik1Δ and kar3Δ mutants would show more significant cell cycle delay after HU treatment. Therefore, we examined Pds1 protein levels in cik1Δ and kar3Δ mutants after HU treatment. G1-arrested cells were released into medium containing 200 mm HU and incubated for 2 hr. After HU was washed off, the cells were released into YPD medium to determine the budding index and Pds1 protein levels. WT cells exited mitosis after release from HU for 140 min; however, almost 80% of cik1Δ and kar3Δ mutants were still large budded at 180 min. Consistently, Pds1 protein disappeared after 120 min release from HU in WT cells, but cik1Δ and kar3Δ mutants showed persistent Pds1 protein levels (Figure 2B), suggesting the failure of anaphase entry in cik1Δ and kar3Δ mutants after HU treatment.
To further confirm the anaphase entry delay in cik1Δ and kar3Δ mutants after HU exposure, we examined kinetochore distribution in WT and cik1Δ mutant cells after HU treatment by visualizing GFP-tagged Mtw1, a kinetochore protein (Goshima and Yanagida 2000; Pinsky et al. 2003). WT and cik1Δ cells in log phase were treated with 200 mm HU for 2 hr at 25°, and we then washed off HU. Both WT and cik1Δ cells exhibited an unseparated cluster of Mtw1-GFP in HU-arrested cells. After 3 hr release from HU, the majority of WT cells already finished chromosome segregation as judged by the appearance of separated Mtw1-GFP clusters in two daughter cells. However, most of the cik1Δ mutant cells still showed a single Mtw1-GFP cluster (Figure 3A). Because the establishment of chromosome bipolar attachment results in two separated kinetochore clusters (Goshima and Yanagida 2000; He et al. 2000), this observation indicates the potential failure of chromosome bipolar attachment. We also compared the spindle morphology in WT and cik1Δ mutant cells after HU exposure. Most WT cells either were in G1 phase or had an elongated spindle after 3 hr release from HU, while the majority of cik1Δ cells showed a dot-like spindle structure (Figure 3B), raising a possibility that the abnormal spindle morphology could contribute to the single kinetochore cluster in cik1Δ mutants after exposure to HU. It is also possible that the defect in Cik1/Kar3-dependent chromosome movement contributes to the single kinetochore cluster phenotype after HU treatment.
Isolation of HU-sensitive and temperature-sensitive cik1 mutants:
Previous data demonstrate the spindle defects in cik1Δ and kar3Δ cells during vegetative growth. Moreover, both cik1Δ and kar3Δ mutants are temperature sensitive for growth and the mutant cells arrest as large-budded cells with a dot-like spindle structure (Meluh and Rose 1990; Page and Snyder 1992; Manning et al. 1999). We noted that cik1Δ mutant cells showed dot-like spindle morphology after HU treatment, but it remains unclear whether this abnormal spindle structure leads to their HU sensitivity. Therefore, we performed a PCR-based mutagenesis in an attempt to isolate temperature-sensitive (TS) and HU-sensitive cik1 mutants. With these HU-sensitive mutants, we might be able to address whether the HU sensitivity of cik1 and kar3 mutants correlates with the abnormal spindle structure.
Mutated cik1 genes in a centromere plasmid were introduced into cik1Δ mutant cells and the transformants were copied onto HU (100 mm) and YPD plates. The HU and YPD plates were incubated at 25° and 37°, respectively, to screen HU-sensitive and TS mutants. Some mutants (cik1HU) were sensitive to HU but grew well at 37°, while others (cik1TS) grew well on HU plates but failed to grow at 37° (Figure 4A). The plasmids with mutated cik1 genes were recovered from the transformants and sequenced, and the mutation sites are illustrated in Figure 4B. Interestingly, the two HU-sensitive mutants are frameshift mutations, which result in truncation of part of the carboxyl-terminal end. In contrast, the mutations in the two cik1TS mutants localize within the coiled-coil domain, which is required for Cik1-Kar3 interaction (Barrett et al. 2000). As cik1HU5 exhibited similar HU sensitivity to cik1HU3, but with a less truncated C terminus, only this strain was used to characterize the mutant phenotype and it was named cik1HU.
We first compared the cell cycle progression and the spindle morphology in WT, cik1HU, and cik1Δ cells. Compared to that in WT cells, we observed a slightly increased accumulation of large-budded cells in the cik1HU mutant (from 34 to 40%), but cik1Δ deletion mutants showed many more large-budded cells (58%) (Figure 4C, left). Among the large-budded cells, 72% of cik1Δ mutant cells showed a dot-like spindle structure, but this number reduced to 26% in cik1HU mutants and to 3% in WT cells (Figure 4C, right). Therefore, the cik1HU mutant exhibited much less spindle defect than the cik1Δ mutant. Although cik1HU mutants grew better on YPD plates than cik1Δ mutants did, they exhibited similar sensitivity to various concentrations of HU, ranging from 5 to 50 mm (Figure 4D). On the basis of these observations, we reason that defects other than the abnormal spindle structure contribute to the HU sensitivity in cik1HU mutants. The truncation in cik1HU mutants may compromise the motor activity of the Cik1/Kar3 complex, which is likely the cause of HU sensitivity.
cik1HU mutants exhibit a chromosome biorientation defect in the presence of low concentrations of HU:
Since WT and cik1HU mutant cells show distinct sensitivity to 20 mm HU (Figure 4D), we examined the chromosome segregation in WT and cik1HU mutant cells with mCherry-tagged Tub1 and GFP-marked Mtw1 after treatment with 20 mm HU (Pinsky et al. 2003; Khmelinskii et al. 2007). After incubation in 20 mm HU medium for 6 hr at 25°, 63% cik1HU mutant cells arrested as large-budded cells, compared to 44% in WT cells. Among the large-budded cells, 45% of cik1HU mutant cells exhibited a dot-like spindle structure. Therefore, we compared the distribution of Mtw1-GFP in WT and cik1HU cells with a normal short spindle structure. Interestingly, 78% of WT cells showed two separated Mtw1-GFP clusters, indicating the establishment of chromosome bipolar attachment (Goshima and Yanagida 2000; He et al. 2000). In contrast, only 39% cik1HU mutant cells with a short spindle displayed two clearly separated Mtw1-GFP clusters (Figure 5A, middle panel). In some cik1HU mutant cells, the Mtw1-GFP signals distributed along the entire spindle (Figure 5A, bottom panel, arrow), while some cells showed multiple Mtw1 clusters.
We have shown that the treatment of yeast cells with HU slows down kinetochore reassembly (Liu et al. 2008), which may result in more detached chromosomes. We reason that the Kar3-mediated chromosome transport becomes more critical for the reestablishment of KT–MT interaction after HU treatment. Therefore, we examined the centromere localization relative to the spindle by using strains with mCherry-tagged TUB1 and GFP-marked centromere of chromosome IV (Cen4-GFP) (D'Amours et al. 2004; Tang and Wang 2006). The relative localization of Cen4-GFP to the spindle can be grouped into three categories (Liu et al. 2008): cells with one or two close GFP dots colocalized with the middle part of the spindle (middle), a single GFP dot at one end of the spindle (end), and a single GFP dot that is away from the spindle (out). After 6 hr incubation in 20 mm HU, the accumulation of large-budded cells in the cik1HU mutant was more dramatic compared to that of WT cells. In 76% of WT cells with a short spindle, the Cen4-GFP was localized at the middle of the spindle, but the number was reduced to 39% in cik1HU cells. More Cen4-GFP dots localized at the end of the spindle in cik1HU mutant cells (53%) compared to WT cells (23%), and 8% of mutant cells showed a Cen4-GFP dot that was away from the spindle (Figure 5B). One possibility is that some cik1HU mutant cells fail to move the captured chromosomes to the spindle. Moreover, even after a chromosome has been moved close to the spindle pole through one of the attached sister kinetochores, cik1 and kar3 mutant cells may have difficulty in moving the sister kinetochore toward the other pole. Therefore, we observed more cik1HU mutant cells showing a Cen4-GFP dot close to one end of the spindle.
cik1HU mutants exhibit delayed anaphase entry after exposure to HU:
In addition to the analysis of cell cycle defects in cik1HU mutant cells in a low concentration of HU, we also examined cell cycle progression in cik1HU mutant cells after a short exposure to a high concentration of HU. First, we compared undisturbed cell cycle progression in WT, cik1Δ, and cik1HU mutant cells. Without HU treatment, the cell cycle progression of the cik1HU mutant was relatively normal as indicated by the budding index and Pds1 protein level, which was in clear contrast to that of the cik1Δ mutant (Figure 6A). We next analyzed the cell cycle progression in these cells after exposure to a high concentration of HU. G1-synchronized cells with Pds1-myc were grown in 200 mm HU medium for 2 hr at 25°. After we washed off HU, the cells were released into YPD medium and collected to analyze cell cycle progression. After HU exposure, cik1HU mutant cells showed an obvious cell cycle delay because more mutant cells remained large budded, although the delay was not as dramatic as that of the cik1Δ mutant. Consistently, the Pds1 protein level was stabilized in cik1HU mutant cells after release from HU arrest (Figure 6B), suggesting that some cik1HU cells exhibit delay in anaphase onset after HU treatment.
We further assayed the relative localization of Cen4-GFP to the spindle in WT and cik1HU mutants after HU exposure. Similarly, G1-arrested WT and cik1HU cells with Cen4-GFP and Tub1-mCherry were first grown in 200 mm HU for 2 hr and then released into YPD medium. After HU release, cik1HU mutant cells showed an obvious anaphase entry delay as indicated by the budding index and the appearance of elongated spindles (Figure 6, C and D). After release from HU for 80 min, 59% of WT cells showed one or two Cen4-GFP dots that localized in the middle part of the spindle, but this number was only 28% in cik1HU mutant cells (Figure 6E). There were more cik1HU mutant cells with a Cen4-GFP dot that was either close to one spindle end or away from the spindle (Figure 6E, arrow), indicating a possible failure of the poleward chromosome transport. These results support the notion that cik1HU mutants have difficulty in establishing chromosome bipolar attachment after HU treatment. The impaired chromosome transport in cik1HU mutants may contribute to this defect.
In budding yeast, kinetochores are attached to microtubules throughout the cell cycle except when the DNA replication machinery encounters the kinetochore that assembles on centromeric DNA. The replication-induced kinetochore disassembly allows the migration of the detached centromeres away from the spindle pole. More chromosomes will be detached during stressful DNA replication because of the longer time required to finish the duplication of centromeric DNA. After the kinetochore reassembles, a poleward transport mechanism moves the captured chromosome close to the spindle. Thus, chromosome transport may become more critical when yeast cells are treated with HU that slows down DNA synthesis. Kar3, a minus-end-directed motor protein, has been found to be required for poleward chromosome transport (Tanaka et al. 2005). Here, we show that yeast cells deficient in the Cik1/Kar3 motor complex are very sensitive to HU. On the basis of the similarity of the mitotic defects in cik1 and kar3 mutants (Page and Snyder 1992; Manning et al. 1999), we reason that the Cik1/Kar3 complex is responsible for chromosome transport. Our data suggest that the chromosome transport defect in cik1 and kar3 mutants contributes to the HU sensitivity.
Because cik1Δ and kar3Δ mutants exhibit a dot-like spindle structure, especially when incubated at higher temperatures (Meluh and Rose 1990; Page and Snyder 1992), either the abnormal spindle structure or the chromosome transport defect could contribute to the HU sensitivity. To distinguish these possibilities, we isolated HU-sensitive cik1HU mutants that were not temperature sensitive and showed relatively normal spindle structure. Interestingly, these mutants exhibited similar HU sensitivity to cik1Δ. Although some cik1HU mutant cells had abnormal spindles in the presence of HU, we found that many cik1HU cells with a normal short spindle failed to form two Mtw1-GFP foci after incubation in 20 mm HU, an indication of the failure of chromosome bipolar attachment. We also noted that the relative localization of the GFP-marked centromere to the spindle in cik1HU cells was different from that in WT cells. There were increased numbers of cells with a Cen4-GFP dot either at one end of the spindle or away from the spindle in cik1HU mutants after HU treatment. Therefore, we reason that a spindle-independent defect, likely the failure of chromosome transport, contributes to the HU sensitivity in cik1 and kar3 mutants, although we cannot exclude the possibility that the abnormal spindle structure contributes to their HU sensitivity as well.
The failure to transport captured chromosomes to the spindle pole can explain the increased population of cik1HU cells with a Cen4-GFP dot away from the spindle. Nevertheless, we also observed that many cik1HU cells showed a Cen4-GFP dot at one end of the spindle. One possibility is that, after one of the sister kinetochores is captured and transported toward the spindle, the other sister kinetochore will be captured by a microtubule from the other spindle pole and the Cik1/Kar3 motor complex moves it to the spindle midzone. Therefore, defects in this process may lead to the accumulation of cells with a Cen4-GFP dot at one end of the spindle. If Cik1/Kar3 is required to move sister kinetochores toward opposite directions, this movement may generate tension across sister kinetochores, but further research is needed to test this possibility.
We noted the abnormal distribution of kinetochore proteins in cik1HU mutant cells incubated in medium containing HU and we speculate that the chromosome transport defect leads to this phenotype. A plus-end-directed motor protein, Cin8, is required for chromosome congression by regulating the spindle microtubule dynamics, and the defect in this regulation leads to the failure of formation of two kinetochore clusters before anaphase entry (Gardner et al. 2008a). The similar kinetochore distribution pattern in cik1HU and cin8 mutants raises the possibility that the Cik1/Kar3 complex also plays a role in chromosome congression. Indeed, the Cik1/Kar3 complex binds to the end of the microtubule and promotes its depolymerization (Maddox et al. 2003; Sproul et al. 2005). However, we found that cin8 mutants are not sensitive to HU. Although we cannot exclude the role of Cik1/Kar3 in chromosome congression, the HU sensitivity of cik1 and kar3 mutants is unlikely due to the defect in chromosome congression.
Recent data from the Tanaka laboratory suggest that kinetochores either can slide along the microtubule lateral surface, which is driven by Kar3, or are tethered at the microtubule ends through the DASH/Dam1 kinetochore complex and pulled poleward as microtubules shrink (end-on pulling) (Tanaka et al. 2007). Ask1, a component of the DASH/Dam1 complex, was originally identified as a protein required for growth in the presence of HU (Alcasabas et al. 2001; Li et al. 2002). ask1-3 mutants also exhibit HU sensitivity and we speculated that the partially elongated spindle contributes to the HU sensitivity, because deletion of CIN8 suppressed the spindle elongation and the HU sensitivity of the ask1-3 mutant (Liu et al. 2008). As both Kar3 and the DASH/Dam1 complex are required for the poleward chromosome transport, it is possible that the defective chromosome transport in ask1-3 mutants contributes to the HU sensitivity as well. Our observation that ask1-3 and cik1Δ mutants are synthetically lethal supports this possibility (data not shown). It will be interesting to determine the chromosome transport defect in ask1-3 and another HU-sensitive mutant, ask1-1.
We isolated two cik1HU mutants and both mutant genes are truncated at the C termini. It remains unclear how the truncation of the C terminus affects the function of the Cik1/Kar3 motor complex. One possibility is that the truncated fraction of the Cik1 protein mediates the interaction of the Cik1/Kar3 complex with the kinetochore. We can examine the association of Kar3 with centromeric DNA in the cik1HU mutants to test this possibility. Alternately, the C terminus of the Cik1 protein binds to microtubules and allows the motility of the Kar3 motor, like Vik1, the other partner of Kar3 (Chu et al. 2005; Allingham et al. 2007). Therefore, the truncation of the C terminus of Cik1 may abolish the motility of the Cik1/Kar3 complex, but more experiments are needed to clarify this issue.
In mammals, dynein and dynactin are responsible for the minus-end-directed chromosome movement (Rieder and Alexander 1990; King et al. 2000). In fission yeast, the homologs of Kar3 and Dam1 are required for chromosome transport through lateral sliding or the end-on pulling mechanism (Gachet et al. 2008), suggesting that the mechanism is conserved. Here we characterized the role of the Cik1/Kar3 motor complex in response to stressful DNA replication, but this function is likely specific to budding yeast because only budding yeast cells establish KT–MT interaction during S-phase. Nevertheless, this poleward chromosome movement may facilitate chromosome bipolar attachment in all eukaryotic cells. First, the poleward movement might be important to orient sister kinetochores to favor the establishment of chromosome bipolar attachment. It has been shown that the abrogation of the poleward chromosome transport in higher eukaryotic cells leads to misoriented kinetochores (Varma et al. 2008). Moreover, after a chromosome is attached by microtubules from opposite spindle poles, poleward movement of sister kinetochores could generate tension on this chromosome, which further regulates KT–MT interaction and cell cycle progression. Further, we are interested in characterizing the role of Cik1/Kar3 in the establishment of chromosome bipolar attachment.
We thank Steve Elledge, Mark Rose, Sue Biggins, Uttam Surana, and Elmar Schiebel for yeast strains. We also thank Akash Gunjan for advice on PCR mutagenesis and Ruth Didier for the FACS analysis. We are grateful to Daniel Richmond and Kelly McKnight who read this manuscript. This work was supported by a Research Scholar Grant (RSG-08-104-010CCG) from the American Cancer Society, a Multi-Disciplinary Grant from the Florida State University (FSU) Council on Research and Creativity, and a Research Enhancement Grant from the FSU College of Medicine (to Y.W.).
Supporting information is available online at http://www.genetics.org/cgi/content/full/genetics.110.125468/DC1.
↵1 These authors contributed equally to this work.
↵2 Present address: Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, TX 75390.
Communicating editor: N. M. Hollingsworth
- Received October 20, 2010.
- Accepted November 30, 2010.
- Copyright © 2011 by the Genetics Society of America