Cells expressing a mutant allele of α-tubulin, tub1-729, are cold sensitive and arrest as large-budded cells with microtubule defects. The cold sensitivity of tub1-729 is suppressed by extra copies of a subset of the mitotic checkpoint genes BUB1, BUB3, and MPS1, but not MAD1, MAD2, and MAD3. This suppression by checkpoint genes does not depend upon their role in the MAD2-dependent spindle assembly checkpoint. In addition, BUB1 requires an intact kinase domain as well as Bub3p to suppress tub1-729. The data suggest that tub1-729 cells are defective in microtubule-kinetochore attachments and that the products of specific checkpoint genes can act either directly or indirectly to affect these attachments.
DURING cell division, DNA is replicated and then partitioned equally between two newly formed cells. Incorrect partitioning of DNA during mitosis or meiosis causes aneuploidy, which can lead to deleterious consequences such as cancer and birth defects. Microtubules are essential for proper DNA segregation: They position the nucleus for cell division, form the mitotic spindle, and contribute to force generation during anaphase (reviewed in Mitchison 1988; Koshland 1994). At each of these steps microtubules interact with other proteins or protein complexes that regulate, assist, or monitor microtubule function. For example, motors bind to microtubules in mitosis and help generate the force required to separate sister chromatids during anaphase. To understand the process of cell division, it is necessary to understand microtubules and the proteins that interact with them.
To learn more about the various roles of microtubules in the cell, we have taken a genetic approach using the budding yeast, Saccharomyces cerevisiae. In budding yeast, there are two functionally interchangeable α-tubulin genes, TUB1 and TUB3 (Schatz et al. 1986, 1988), and one β-tubulin gene, TUB2 (Thomaset al. 1985). A genetic screen for conditional alleles of the major α-tubulin gene, TUB1, isolated 35 cold-sensitive mutants (Schatzet al. 1988). These mutants fall into three classes: Class 1 mutants have a reduced number of microtubules; class 2 mutants have extra microtubules; and class 3 mutants have abnormal microtubules. The diversity of the phenotypes observed in this mutant collection suggests that these α-tubulin alleles may be deficient in distinct aspects of microtubule function. Analyses of these mutants can lead to an understanding of microtubule processes as well as to the identification of other proteins that interact with tubulin.
When shifted to the nonpermissive temperature (15°), cells expressing tub1-729, a class 1 allele, as its sole source of α-tubulin, accumulate as large-budded cells with either a short spindle or a single dot of microtubule staining (Schatzet al. 1988; Guénetteet al. 1995). A screen for low-copy suppressors of the cold sensitivity of tub1-729 identified only one gene, BUB3 (budding uninhibited by benomyl) eight times (Guénetteet al. 1995). BUB3 encodes a protein involved in the spindle assembly checkpoint that binds to and is phosphorylated by Bub1p (Hoytet al. 1991; Robertset al. 1994). Direct testing revealed that BUB1 also suppresses the cold sensitivity of tub1-729 cells, but another spindle assembly checkpoint gene, MAD2 (mitotic arrest deficient), does not (Guénetteet al. 1995). Bub1p and Bub3p, together with Mps1p (mono polar spindle), Mad1p, Mad2p, and Mad3p, are part of the spindle assembly checkpoint. These proteins monitor microtubule-kinetochore attachments and prevent cells from entering anaphase until all of the chromosomes are attached to spindle microtubules. Mutations in these checkpoint genes—either deletion of BUB1 and -3 and MAD1, -2, and -3 or a point mutation in the essential MPS1—eliminate the spindle assembly checkpoint, allowing cells to enter anaphase when their mitotic spindle is disrupted either by the microtubule depolymerizing drug, benomyl (Hoytet al. 1991; Li and Murray 1991) or by mutations that destabilize the spindle (Pangilinan and Spencer 1996). The higher eukaryotic homologs of these six mitotic checkpoint genes localize to unattached kinetochores during pro-metaphase and all except Mps1p are released from the kinetochore after microtubules bind (Chen et al. 1996, 1998; Li and Benezra 1996; Taylor and McKeon 1997; Tayloret al. 1998; Martinez-Expositoet al. 1999; Abrieuet al. 2001). The kinetochore localization of these proteins supports the idea that they monitor microtubule-kinetochore attachment and help to convey a signal to halt cell-cycle progression through downstream factors.
In this article, we show that genetic interactions with tub1-729 distinguish among the mitotic checkpoint genes. The results demonstrate that, like BUB1 and BUB3 (Guénetteet al. 1995), MPS1 suppresses the cold sensitivity of tub1-729. However, BUB2 and MAD1, 2, and 3 do not suppress. Strikingly, the suppression by BUB1, BUB3, or MPS1 is independent of Mad2p and therefore independent of the Mad2p-dependent mitotic checkpoint pathway. Additional genetic analyses indicate that BUB1 and BUB3 require one another to suppress the cold sensitivity of tub1-729, while MPS1 requires neither BUB1 nor BUB3. Further experiments suggest that the mutant tub1-729p α-tubulin is defective in its interactions with the kinetochore, as assayed by arrest phenotype, frequency of chromosome loss, and genetic interactions with mutations in genes encoding kinetochore proteins. The data suggest that Bub3p contributes to stabilizing microtubule-kinetochore attachment and that it is this activity that suppresses the phenotypes of tub1-729.
MATERIALS AND METHODS
Strains and media: All yeast strains used in this study are derivatives of FSY182, -183, or -185 (Weinstein and Solomon 1990) except where noted (Table 1). We used standard methods for yeast manipulations (Guthrie and Fink 1991; Solomonet al. 1992).
Plasmid construction: To construct a LYS2-marked plasmid expressing BUB3 under the control of its own promoter, the BUB3 open reading frame was amplified by PCR and ligated into either the pRS317 vector (Sikorski and Hieter 1989) or the YEp426 vector (Maet al. 1987). The upstream primer used to amplify BUB3 (5′-ATAGCGGCCGCGTGACAACCA AGC-3′) contains a NotI site (underlined). The downstream primer (5′-TCTGTCTTCTTGCGTATAGG-3′) contains a SalI site (underlined). These two primers generate a PCR product containing the BUB3 gene as well as ~250 bp of upstream and downstream sequence (~1500 bp). The PCR product was digested with NotI and SalI and ligated into a NotI/SalI-digested pRS317 or YEp426 vector to make pKC52 and pKC74. Both of these vectors rescue the benomyl supersensitivity of bub3Δ cells.
To construct a LYS2-marked plasmid expressing either tub1-729 or TUB1 under the control of its own promoter, the TUB1 coding region was amplified by PCR from pRB629 or pRB539, respectively. The upstream TUB1 primer (5′-GCGCGCCGGCGGCCGCTGCTCACACCAAGCATCA-3′) contains a NotI site (underlined). The downstream TUB1 primer (5′-CCGCCGGCCCTGCAGCAATACGATGAGACTCAT-3′) contains a PstI site (underlined). These two primers amplified a region including the α-tubulin open reading frame as well as 500 bp of upstream sequence and 260 bp of downstream sequence (~2200 bp). The PCR products were digested with NotI and PstI and ligated into NotI/PstI-digested pRS317 (Sikorski and Hieter 1989) to generate pKC43 (tub1-729) and pKC44 (TUB1).
Gene disruptions: To delete the MAD2 open reading frame with KANR, we digested the mad2::KANR knockout plasmid (pKS198-MAD2; a gift of K. Simons and P. Sorger) with NdeI and SacII and transformed this fragment into KCY1955 to generate SKY294. Proper integration was verified using PCR. Mad2::KANR is benomyl supersensitive and this sensitivity can be rescued by pRC4.
BUB3 was disrupted using the LEU2 disruption construct, pTR27 (a gift from M. A. Hoyt). pTR27 was digested with StuI and SacI and transformed into the wild-type diploid FSY185 to generate KCY576. Proper integration of the LEU2 gene at the BUB3 locus was verified by PCR. KCY576 was sporulated to generate bub3::LEU2 haploids sensitive to the microtubule depolymerizing drug, benomyl. This benomyl supersensitivity can be rescued by a low-copy BUB3 plasmid.
To delete the TUB3 open reading frame with KANR, we used the PCR-based knock-out strategy described by Wach et al. (1997). Using primers containing both 40 bp of TUB3-flanking DNA and 20 bp of DNA found in the pFA6a-KANMX6 plasmid (a gift from P. Phillipsen), we amplified a fragment containing the KANR gene surrounded by TUB3 flanking DNA. We purified this fragment and transformed it into KCY2011 to generate KCY2024. Proper integration was verified using PCR.
Cell morphology and nuclear position: To assay the budding phenotypes of TUB3/p-tub1-729, mad2Δ cells, we transformed these cells with MAD2- or BUB3-expressing plasmids (KCY-2072, KCY2070, KCY2077, and KCY2080) and grew these cells to mid-log phase (1 × 107 cells/ml) at the permissive temperature (30°). After shifting them to restrictive temperature (13°), we counted the cells at varying intervals and determined the number of large-budded cells (Figure 2C). For the purpose of this assay, large-budded cells were defined as cells with buds greater than half the size of their mothers' buds. To assay the budding phenotypes and nuclear positioning in tub1-729 (with or without excess BUB3), tub1-724, and wild-type cells, the cells were grown for 20 hr at the permissive (30°) or restrictive (15°) temperatures. Nuclear positioning and bud size were analyzed as described in Schatz et al. (1988) and Table 2 and Figure 4. The identity of each of the samples in both assays was hidden during these analyses.
Chromosome loss assays: To construct a strain that would enable us to test the frequency of chromosome loss in tub1-729 cells, we crossed KCY2024 to YPH278 (a gift from P. Hieter) to generate KCY2045. We sporulated and dissected this diploid to generate KCY2098, which expresses tub1-729 as the sole source of α-tubulin and contains the ade2-101 mutation and a mini-chromosome expressing SUP11. The colonies produced by these cells are white as long as the mini-chromosome expressing the ochre suppressor SUP11 is maintained. To measure the frequency of chromosome loss in tub1-729 cells, we used the method described by Christianson et al. (1992). We grew cells at the permissive temperature in media that selects for the maintenance of the mini-chromosome and then plated these cells to YEPD media. We determined the viable number of cells in each culture (N1) and the percentage of cells that contained the mini-chromosome (P2). We diluted these cells 20-fold and grew the cells at a semipermissive temperature (20°) for 36 hr in medium that does not select for mini-chromosome maintenance. We plated these cells to YPD at the permissive temperature and determined the number of viable cells per milliliter (N2) and the number of cells that still contained the mini-chromosome (i.e., number of white colonies, P2). The frequency of mini-chromosome loss (m) is described by m = 1 − eln(P2/P1)/g, where g represents the number of cell divisions at the semipermissive temperature and is calculated by g = ln(N2/N1)/ln 2. Over 1200 colonies per strain were analyzed to produce the data shown in Table 3.
Synthetic lethality of tub1-729 with kinetochore mutants: To test whether cells expressing tub1-729 as the sole source of α-tubulin are viable in combination with kinetochore mutations, we crossed KCY2103 (tub1::HIS3, tub3::KANR, ptub1-729-LEU2-CEN, pTUB1-URA3-CEN) to four different kinetochore mutants: ndc10-1 (JK418); ndc80-1; nuf2-61 (PSY455); and dam1-1 (MJYM14-pb 889). The ndc10-1 and ndc80-1 alleles are in the W303 yeast background; both of these strains were backcrossed at least three times into the S288C yeast background before being crossed to KCY2103. We isolated tub1-729 kinetochore double mutants from these crosses and struck them to selective media or selective media containing 5-fluoroorotic acid (5-FOA) to select for the loss of the wild-type α-tubulin plasmid expressing URA3. After 5 days of growth at 30°, the plates were photographed (Figure 5).
A subset of the mitotic checkpoint genes suppresses the cold sensitivity of tub1-729: Cells expressing the mutant tub1-729 as the sole source of α-tubulin are cold sensitive and arrest with no microtubules (Schatzet al. 1988; Guénetteet al. 1995). These phenotypes are suppressed partially by extra copies of either of two mitotic checkpoint genes, BUB1 or BUB3 (Guénetteet al. 1995). The interaction is specific; other α-tubulin mutants with the same arrest phenotype are not suppressed by these genes. Extra copies of a third checkpoint gene, MPS1, suppress the cold sensitivity of tub1-729 as efficiently as BUB1 or BUB3 (Figure 1A). However, the other known mitotic checkpoint genes—BUB2, MAD1, MAD2 (Guénetteet al. 1995), or MAD3—do not suppress the cold sensitivity of tub1-729 (Figure 1A).
We quantitated the BUB3 suppression by measuring plating efficiency of mutant and suppressed cells plated at permissive (30°) and restrictive (18°) temperatures. BUB3 on a CEN-plasmid increases the colony formation at 18° by more than three orders of magnitude (0.0001–0.3%; see Figure 1A). Higher levels of BUB3, present on a high-copy 2μ plasmid, increase plating efficiency to 19.4%. Similar results were obtained for BUB1. However, higher levels of MPS1 overexpression can arrest the cell cycle (Hardwicket al. 1996), and we have found that 2μ plasmids expressing either MAD1 or MAD3 have a detrimental effect on tub1-729 cells (data not shown). Therefore, although high-copy plasmids expressing either BUB1 or BUB3 are more efficient suppressors of the phenotypes of tub1-729, we focused on the effects of a lower level of expression derived from CEN plasmids in order to compare the effects of BUB1 and BUB3 with those of other mitotic checkpoint genes.
The kinase activity of BUB1 is required to suppress the cold sensitivity of tub1-729: Bub1p is a serine/threonine kinase whose enzymatic activity is required for the spindle assembly checkpoint (Robertset al. 1994). A mutant allele of BUB1, bub1K733R, which contains a single amino acid substitution in the conserved kinase domain, does not rescue the benomyl sensitivity of cells deleted for BUB1 (Robertset al. 1994). Although the kinase activity of Bub1K733Rp is lost, the mutant protein still binds to Bub3p (Robertset al. 1994). To determine if the kinase activity of Bub1p is required for the suppression of tub1-729, we transformed tub1-729 cells with low-copy plasmids expressing either wild-type BUB1 or bub1K733R. The overexpression of mutant BUB1, bub1K-733R, does not suppress the cold sensitivity of tub1-729 (Figure 1B). This result suggests that the kinase activity of BUB1 is required for the suppression of tub1-729.
MPS1, BUB1, and BUB3 can suppress the cold sensitivity of tub1-729 in the absence of MAD2: BUB1, BUB3, and MPS1 are each components of the mitotic checkpoint pathway that prevents cells from entering anaphase until all of the chromosomes are attached to the microtubules of the mitotic spindle. We tested whether the suppression of tub1-729 by MPS1, BUB1, or BUB3 depends upon that checkpoint function by making a strain (SKY294) that expresses tub1-729 and in which the mitotic checkpoint is disrupted by the deletion of MAD2 (Li and Murray 1991). However, all of the known components of the mitotic checkpoint pathway (BUB1, BUB3, MAD1, MAD2, and MAD3) are essential in tub1-729 cells in which tub1-729p is the sole source of α-tubulin (tub1Δ, tub3Δ, and ptub1-729; data not shown). Therefore, we generated KCY2010 cells, which express tub1-729 as well as a chromosomal copy of the minor α-tubulin gene, TUB3. The KCY2010 cells are still cold sensitive, but at 13° instead of 18° (Figure 2A). Their cold sensitivity is also suppressed by the same subset of mitotic checkpoint genes (Figure 2A). We used TUB3/p-tub1-729 cells to test whether BUB1, BUB3, and MPS1 could rescue the phenotypes of tub1-729 cells that were deleted for MAD2.
TUB3/p-tub1-729, mad2Δ cells (KCY2042) are viable at 30°. However, they are more cold sensitive than TUB3/p-tub1-729 cells since the checkpoint pathway is disrupted, and they fail to grow at 13°–15° (Figure 2B). Cold sensitivity is rescued by extragenomic copies of BUB1, BUB3, and MPS1 (Figure 2B), despite the abrogation of the mitotic checkpoint. Extra copies of BUB2 or MAD3 did not effect the cold sensitivity of TUB3/p-tub1-729 cells. We have observed an occasional slight suppression by MAD1, but this result is not consistent.
It is possible that overexpression of specific checkpoint genes could suppress tub1-729 cold sensitivity by strengthening the mitotic checkpoint. To test that possibility, we monitored the progression of TUB3/p-tub1-729, mad2Δ cells through the cell cycle after shift to the restrictive temperature (13°). We conducted this experiment at 13° rather than 15° (Figure 2B) because we observed a tighter cell cycle arrest at the lower temperature. Our first observation was that KCY2042 cells (TUB3/p-tub1-729, mad2Δ) failed to arrest as large-budded cells when shifted to the restrictive temperature (Figure 2C, triangles). In contrast, KCY2042 cells in which the mitotic checkpoint has been restored by transformation with a CEN plasmid bearing MAD2 accumulate significantly at the large-bud stage after one to two cell cycles (Figure 2C, diamonds; Table 2). This suggests that tub1-729 cells accumulate as large-budded cells due to the MAD2-dependent spindle assembly checkpoint. Although a plasmid expressing MAD2 can restore the mitotic checkpoint in KCY2042 cells, plasmids expressing BUB3 did not. KCY2042 cells containing extra copies of BUB3 on either a low-copy CEN plasmid (Figure 2C, squares) or a high-copy 2μ plasmid (Figure 2C, circles) continue to grow and do not accumulate as large-budded cells. We conclude that BUB3 does not suppress the phenotypes of tub1-729 by restoring the mitotic checkpoint and that the suppression by BUB3 requires a function independent of its role in the mitotic checkpoint.
Genetic requirements for BUB1, BUB3, and MPS1 suppression of tub1-729 cold sensitivity: To learn more about how BUB1, BUB3, and MPS1 suppress the cold sensitivity of tub1-729, we examined the suppressing activity of these genes in the absence of BUB1 or BUB3. We constructed TUB3/p-tub1-729, bub1Δ cells (KCY2262) and TUB3/p-tub1-729, bub3Δ cells (KCY2202) and transformed them with a panel of mitotic checkpoint genes (Figure 3, A and B). As expected, restoration of wild-type BUB1 rescues the cold sensitivity of TUB3/p-tub1-729, bub1Δ cells, and restoration of wild-type BUB3 rescues the cold sensitivity of TUB3/p-tub1-729, bub3Δ cells. BUB1 and BUB3 require one another to suppress the cold sensitivity of tub1-729; BUB1 cannot rescue TUB3/p-tub1-729, bub3Δ cells and BUB3 cannot rescue TUB3/p-tub1-729, bub1Δ cells. An extra genomic copy of MPS1 can suppress the cold sensitivity of tub1-729 cells in the absence of either BUB3 or BUB1. As previously observed, we saw no suppression by BUB2, MAD1, MAD2, or MAD3. This experiment suggests that Bub1p and Bub3p function together to suppress the cold sensitivity of tub1-729 and that Mps1p functions independently.
Tub1-729 cells accumulate as large-budded cells with their nucleus adjacent to or straddling the mother-bud neck: To investigate the nature of the suppression of tub1-729 by a subset of the mitotic checkpoint genes, we focused on the effects of BUB3 and conducted a more careful examination of the morphology and nuclear positioning of tub1-729 cells at the permissive (30°) and restrictive temperatures (15°). At the permissive temperature, tub1-729 cells resemble wild-type cells; they have a wild-type distribution of large- and small-budded cells, and the position of the nuclei in the large-budded cells is typical for cells progressing through mitosis (Table 2, A and B). When tub1-729 cells are shifted to the restrictive temperature (15°) for 20 hr, the percentage of cells that are large-budded cells almost doubles (74 vs. 41%). DNA staining of tub1-729 cells grown at the restrictive temperature revealed that the nuclei of most of the large-budded cells are either in the mother cell or adjacent to or spanning the mother-bud neck (Table 2B). This suggests that at the restrictive temperature, tub1-729 cells accumulate prior to the onset of anaphase. We compared this arrest phenotype to that of another class 1 α-tubulin mutant, tub1-724. That mutant also arrests as large-budded cells at the restrictive temperature, but almost all of the large-budded cells accumulate with their nucleus in a random position in the mother cell. The contrast between the nuclear positioning phenotypes of tub1-729 and tub1-724 cells suggests that the nuclear positioning phenotypes observed for tub1-729 cells are not a simple consequence of loss of microtubules at restrictive temperature.
Tub1-729 cells overexpressing Bub3p no longer accumulate as large-budded cells with their nuclei positioned at the mother-bud neck: Since a high-copy plasmid expressing BUB3 rescues the cold sensitivity of tub1-729, we examined the budding morphology and nuclear positioning in tub1-729 cells containing either a control plasmid or a 2μ plasmid expressing BUB3. Tub1-729 cells overexpressing BUB3 still delay in mitosis as large-budded cells (76 vs. 74%); however, there is a dramatic decrease in the number of cells that accumulate as large-budded cells with their nuclei near or spanning the mother-bud neck (Figure 4) and an increase in the number of large-budded cells that accumulate with their nuclei in a random position in the mother cell (27–60%) (data not shown). The overexpression of BUB3 causes a shift in the point at which cells accumulate in the cell cycle; tub1-729 cells overexpressing BUB3 accumulate as large-budded cells with their nuclei almost exclusively in a random position in the mother cell. The overexpression of BUB3 causes tub1-729 cells to more closely resemble other class 1 α-tubulin mutants that have general microtubule defects such as tub1-724 (see Table 2).
Tub1-729 cells exhibit a higher frequency of chromosome loss that is reduced by BUB3 overexpression: That tub1-729 cells accumulate at the metaphase-to-anaphase transition at the restrictive temperature suggested that these cells might have defects in proper chromosome segregation. To test this hypothesis, we followed minichromosome loss (Spenceret al. 1990). We constructed a yeast strain (KCY2098) that expresses tub1-729 as the sole source of α-tubulin, the ade2-101 gene, and the ochre suppressor, SUP11, from a mini-chromosome. We transformed this strain with CEN-plasmid expressing BUB3; a 2μ plasmid expressing BUB3; and, as controls, plasmids expressing wild-type α-tubulin (TUB1) or with no insert. We monitored the number of viable cells that contain the mini-chromosome both before and after 36 hr of growth at the semipermissive temperature (20°) in media that do not select for the mini-chromosome. Using the methods described by Christianson and colleagues we determined the frequency of chromosome loss (Christiansonet al. 1992; see materials and methods). The results (Table 3) demonstrate that the higher frequency of chromosome loss in tub1-729 cells relative to wild-type cells (3.8 × 10−2 vs. 0.2 × 10−2) is suppressed by increasing levels of BUB3.
Mutations in the kinetochore components DAM1 and NUF2 are lethal in combination with tub1-729: As shown above, the arrest phenotype of tub1-729 suggests that the mutation affects the ability of cells to progress from metaphase to anaphase. This defect is suppressed by overexpression of three mitotic checkpoint proteins whose mammalian homologs localize to the kinetochore (Chen et al. 1996, 1998; Li and Benezra 1996; Taylor and McKeon 1997; Tayloret al. 1998; Martinez-Expositoet al. 1999; Abrieuet al. 2001). Therefore, we tested whether tub1-729 cells are affected by mutations in kinetochore components. We generated tub1-729 cells that contain a URA3-CEN plasmid expressing wild-type α-tubulin (KCY2103) and each of four different temperature-sensitive kinetochore mutants (ndc10-1, ndc80-1, nuf2-61, or dam1-1; see materials and methods). Two of these double mutants—tub1-729, dam1-1 and tub1-729, nuf2-61—are not viable in the absence of the plasmid expressing wild-type α-tubulin (Figure 5).
Accurate segregation of chromosomes between mother and daughter cells requires that microtubules attach to the kinetochores of all of the chromosomes prior to the initiation of anaphase. The spindle assembly checkpoint arrests cells and prevents entry into anaphase until all the microtubule-kinetochore attachments are made. In S. cerevisiae six genes are essential for the spindle assembly checkpoint: BUB1 and BUB3 (Hoytet al. 1991); MAD1, 2, and 3 (Li and Murray 1991); and MPS1 (Weiss and Winey 1996). In higher eukaryotes, the products of the highly conserved homologs of these genes (Chen et al. 1996, 1998; Li and Benezra 1996; Taylor and McKeon 1997; Cahillet al. 1998; Jinet al. 1998; Tayloret al. 1998) all localize to unattached kinetochores, and all except Mps1p disappear from kinetochores after microtubules have attached (Chen et al. 1996, 1998; Li and Benezra 1996; Taylor and McKeon 1997; Tayloret al. 1998; Martinez-Expositoet al. 1999; Abrieuet al. 2001). When kinetochores are not attached to microtubules, the proteins involved in the checkpoint bind to and inhibit the ubiquitin ligase specificity factor, Cdc20p, which is a component of the anaphase-promoting complex (APC; reviewed in Amon 1999; Burke 2000). The APC degrades Pds1p, an inhibitor of sister chromatid separation. When Cdc20p function is inhibited, Pds1p levels remain high and cells remain in metaphase.
The experiments presented here focus on the interactions between a subset of the mitotic checkpoint genes and a mutant allele of α-tubulin, tub1-729. The data suggest that MPS1, BUB1, and BUB3 have a role at the kinetochore that is distinct from their function in the mitotic checkpoint pathway. We propose that tub1-729 microtubules are deficient in kinetochore binding and that excess MPS1, BUB1, and BUB3 suppress the phenotypes of tub1-729 by affecting this attachment.
A subset of the mitotic checkpoint genes rescue the cold sensitivity of tub1-729: BUB1, BUB3, and MPS1 are partial suppressors of the defects of tub1-729. Modest overexpression of each rescues the cold sensitivity of tub1-729 cells, but not the sensitivity to the microtubule-destabilizing drug benomyl (Guénetteet al. 1995; data not shown). We conclude that tub1-729p has multiple defects and that BUB1, BUB3, or MPS1 suppresses a subset of them.
The ability of BUB1, BUB3, and MPS1 to suppress tub1-729 is specific. Extra copies of four other mitotic checkpoint genes—BUB2, MAD1, MAD2, or MAD3—have no effect on tub1-729. Conversely, BUB1, BUB3, and MPS1 are not general suppressors of microtubule defects; they do not suppress all other cold-sensitive α-tubulin mutants (such as tub1-724; see Table 2; Guénetteet al. 1995; data not shown) nor do they suppress mutants that effect the levels of tubulin polypeptides such as a deletion of the minor α-tubulin gene, TUB3 (data not shown). These results suggest that excess MPS1, BUB1, and BUB3 suppress a specific microtubule defect found in tub1-729.
A potential structural role for kinase components of the mitotic checkpoint: BUB1 and MPS1 encode serine/threonine kinases. The kinase activities of these two proteins are required for the spindle assembly checkpoint. Mutations in the kinase domain of either BUB1 (bub1K733R) or MPS1 (mps1-1) disrupt the spindle assembly checkpoint (Robertset al. 1994; Hardwicket al. 1996; Weiss and Winey 1996). Similarly, BUB1's kinase domain is required for the suppression of tub1-729 (Figure 1B). MPS1 requires both BUB1 and BUB3 to activate the spindle assembly checkpoint when overexpressed. This suggests that MPS1 and BUB1 have partially overlapping functions at the beginning of a signaling pathway (Hardwicket al. 1996; Farr and Hoyt 1998). Interestingly, MPS1 requires neither BUB1 nor BUB3 to suppress the cold sensitivity of tub1-729 (Figure 3). This observation suggests that BUB1 and MPS1 do not suppress the cold sensitivity of tub1-729 by performing partially overlapping functions at the top of a signaling cascade.
The suppression of tub1-729 may depend upon a structural role for the BUB1 and MPS1 gene products. That possibility is supported by several findings. First, the suppression depends upon overexpression of either gene (Figure 1A) and the ability of these genes to suppress tub1-729 is additive (data not shown). Second, excess Bub3p—which forms a complex with Bub1p (Robertset al. 1994; Brady and Hardwick 2000) and is phosphorylated by it—also suppresses tub1-729. In fact, co-overexpression of BUB1 and BUB3 from low-copy plasmids suppresses the cold sensitivity of tub1-729 better than either one alone (data not shown). Third, Bub1p and Bub3p require one another to suppress the cold sensitivity of tub1-729 (Figure 3). The Bub1p-Bub3p complex forms independently of any of the other spindle assembly checkpoint genes and persists throughout the cell cycle. In animal cells, Bub3p may regulate the kinase activity of Bub1p and may help Bub1p localize to unattached kinetochores (Robertset al. 1994; Tayloret al. 1998). These results suggest that the BUB1/BUB3 complex could be important for the suppression of tub1-729.
MPS1, BUB1, and BUB3 suppression of tub1-729 does not act through the MAD2-dependent checkpoint pathway: Our results demonstrate that MPS1, BUB1, and BUB3 do not require an intact mitotic checkpoint pathway to suppress the phenotypes of tub1-729. Recent work on the mammalian spindle assembly checkpoint suggests that BUBR1—a protein bearing homology to both BUB1 and MAD3 (Tayloret al. 1998)—and BUB3 can bind to CDC20 in a MAD2-independent manner and inhibit the APC, thereby preventing the onset of anaphase (Tanget al. 2001). This finding raises the possibility that overexpression of MPS1, BUB1, or BUB3 could act by similarly restoring the checkpoint in cells deleted for MAD2. We performed two experiments to address this concern. First, we showed that the overexpression of BUB3 does not rescue the benomyl supersensitivity of mad2Δ cells (data not shown), indicating that they were unable to arrest their cell cycle when microtubules were depolymerized. Second, we shifted KCY2042 (tub1Δ, TUB3, ptub1-729, mad2Δ) cells overexpressing BUB3 to the restrictive temperature (13°). Unlike MAD2 cells, the KCY2042 cells overexpressing BUB3 did not arrest as large-budded cells (Figure 2C). These two experiments illustrate that excess BUB3 does not restore the spindle assembly checkpoint in mad2Δ cells. They offer further evidence that BUB3 suppresses the phenotypes of tub1-729 using a novel mechanism that is independent of its known role in the mitotic checkpoint.
Tub1-729 cells exhibit phenotypes similar to kinetochore mutants: The mutation in tub1-729p changes alanine 422 to valine (Guénetteet al. 1995). Residue 422 is in C-terminal helix 12, which protrudes from the surface of the microtubule (Nogales et al. 1998, 1999) and may mediate interactions of microtubules with protein ligands (Paschalet al. 1989). This model is compatible with several results above, suggesting that tub1-729p affects microtubule-kinetochore attachment. The arrest phenotype of tub1-729 cells—large budded, with their nuclei predominantly in the mother or adjacent to or spanning the mother-bud neck (Guénetteet al. 1995; Table 2)—is similar to that of kinetochore mutants, such as ctf19 (Hylandet al. 1999), ctf13-30, ndc10-42 (Dohenyet al. 1993), cep3-1 and -2 (Strunnikovet al. 1995), and skp1-4 (Connelly and Hieter 1996). That the tub1-729 arrest phenotype is a consequence of defective microtubule-kinetochore attachment is supported by the finding that the mutant cells at permissive temperature show a rate of plasmid loss higher than that of wild-type cells, an effect substantially rescued by excess BUB3.
Finally, we find that tub1-729 cells require DAM1 and NUF2 function. Nuf2p and Dam1p are both part of protein complexes that are present at the kinetochore. Available data suggest that they may be involved in facilitating microtubule-kinetochore interactions. Nuf2p is part of a complex containing Ndc80p, Spc24p, and Spc25p. This complex is present at the centromere and is localized to a subset of short spindle microtubules (Jankeet al. 2001; Wigge and Kilmartin 2001). DAM1 was originally found in a screen for genes that are synthetically lethal with mps1-1 (Joneset al. 1999). Dam1p forms a complex with Duo1p and Dad1p, and this complex localizes to mitotic spindles (Joneset al. 1999; Cheesemanet al. 2001; Enquist-Newmanet al. 2001) and to kinetochores (Cheesemanet al. 2001). Interestingly, dad1-1 interacts with the same subset of spindle assembly checkpoint genes—MPS1, BUB1, and BUB3—but not with MAD1, -2, and -3. Perhaps the Dam1p, Duo1p, and Dad1p complex and tub1-729p act in a similar way and consequently interact genetically with a subset of the mitotic checkpoint genes including MPS1, BUB1, and BUB3. Although it is possible that tub1-729 cells require NUF2 and DAM1 for some microtubule function that is independent of the kinetochore, the observation that tub1-729 interacts with a specific set of kinetochore mutants provides additional evidence that tub1-729 may be deficient in microtubule-kinetochore interactions.
The data presented above are consistent with a model in which MPS1, BUB1, and BUB3, in addition to their participation in the spindle assembly checkpoint, play a second and distinct role stabilizing microtubule-kinetochore interactions. We propose that this role accounts for their ability to suppress the cold sensitivity of tub1-729. In this capacity, it is possible that Bub1p, Bub3p, and Mps1p could act as the “sensor” in the spindle assembly pathway. Bub1p, Bub3p, and Mps1p may localize to the kinetochore and mediate microtubule kinetochore attachment. Once microtubules attach to the kinetochore, Bub1p, Bub3p, and Mps1p could signal to downstream factors that would consequently modulate cell-cycle progression.
We thank G. Fink (MIT), M. A. Hoyt (Johns Hopkins), J. Kilmartin (MRC, Cambridge, UK), A. Murray (Harvard), P. Philippsen (Universitat Basel, Switzerland), P. Silver (Harvard Medical School), P. Sorger (MIT), M. Winey (Colorado), and R. Young (MIT) for strains and plasmids. We thank A. Amon, A. Grossman, D. Rines, P. Sorger, S. Sanders (MIT), R. Li (Harvard), and members of MIT M&M for valuable discussion. K.C.A. was supported in part by a training grant from the National Institute of General Medical Sciences (NIGMS) to the department of Biology, MIT. This work was supported by a grant to F.S. from NIGMS.
Communicating editor: M. D. Rose
- Received December 20, 2001.
- Accepted March 21, 2002.
- Copyright © 2002 by the Genetics Society of America