|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 279, Issue 47, 49542-49550, November 19, 2004
The Spindle Pole Body Assembly Component Mps3p/Nep98p Functions in Sister Chromatid Cohesion*![]() ![]() ¶ || **
From the
Received for publication, April 19, 2004 , and in revised form, September 2, 2004.
For successful chromosome segregation during mitosis, several processes must occur early in the cell cycle, including spindle pole duplication, DNA replication, and the establishment of cohesion between nascent sister chromatids. Spindle pole body duplication begins in G1 and continues during early S-phase as spindle pole bodies mature and start to separate. Key steps in spindle pole body duplication are the sequential recruitment of Cdc31p and Spc42p by the nuclear envelope transmembrane protein Msp3p/Nep98p (herein termed Mps3p). Concurrent with DNA replication, Ctf7p/Eco1p (herein termed Ctf7p) ensures that nascent sister chromatids are paired together, identifying the products of replication as sister chromatids. Here, we provide the first evidence that the nuclear envelope spindle pole body assembly component Mps3p performs a function critical to sister chromatid cohesion. Mps3p was identified as interacting with Ctf7p from a genome-wide two-hybrid screen, and the physical interaction was confirmed by both in vivo (co-immunoprecipitation) and in vitro (GST pull-down) assays. An in vivo cohesion assay on new mps3/nep98 alleles revealed that loss of Mps3p results in precocious sister chromatid separation and that Mps3p functions after G1, coincident with Ctf7p. Mps3p is not required for cohesion during mitosis, revealing that Mps3p functions in cohesion establishment and not maintenance. Mutated Mps3p that results in cohesion defects no longer binds to Ctf7p in vitro, demonstrating that the interaction between Mps3p and Ctf7p is physiologically relevant. In support of this model, mps3 ctf7 double mutant cells exhibit conditional synthetic lethality. These findings document a new role for Mps3p in sister chromatid cohesion and provide novel insights into the mechanism by which a spindle pole body component, when mutated, contributes to aneuploidy.
Microtubule-organizing center duplication, separation, and microtubule nucleation are all essential facets of proper chromosome segregation. In budding yeast, microtubule-organizing centers, called spindle pole bodies, function on both sides of the nuclear envelope to nucleate microtubules in both the cytoplasmic and nuclear volumes. Spindle pole body duplication is a multistep process that includes formation of a satellite, expansion into a cytoplasmic duplication plaque, half-bridge growth, and subsequent insertion into the nuclear envelope (1, 2). Evidence from genetic and electron microscopy studies indicates that spindle pole body duplication starts in G1 and continues into early S-phase. Mps3p/Nep98p (herein termed Mps3p) is an essential nuclear envelope protein that is concentrated at the nuclear envelope half-bridge and required for spindle pole body assembly. Early in spindle pole body assembly, Mps3p functions to recruit and anchor Cdc31p to the half-bridge. Later, Mps3p is required for integration of Spc42p into the spindle pole duplication plaque (3, 4). Coincident with the latter portion of spindle pole body maturation and separation are the processes of DNA replication and cohesion establishment (2, 5). During S-phase, each chromosome replicates to produce two identical sister chromatids. To identify the products of DNA replication as sister chromatids, sisters become paired along their entire length. The initial pairing process, or cohesion establishment, ensures that cohesion occurs exclusively between sister chromatids while precluding the pairing of nonsister chromatids, homologous chromosomes, or highly repetitive DNA sequences (68). Whereas the molecular mechanism by which cohesion is established remains unknown, it is clear that establishment is intimately coupled to DNA replication. Previously, we reported that the cohesion establishment factor CTF7/ECO1 (herein termed CTF7) genetically interacts with both POL30 (proliferating cell nuclear antigen) and CHL12/CTF18 (an RFC homolog) (9). Several DNA replication factors now have been identified as functioning in cohesion, including replication factor C subunits, DNA helicases, and DNA polymerases (916). These findings have popularized a model in which cohesion is established as nascent sister chromatids emerge from behind the DNA replication fork (8, 17, 18). Recent findings that DNA helicases are required for sister chromatid cohesion (16, 19, 20) suggest that DNA helicases first encounter and possibly unwind DNA loci marked by cohesins. This model is supported by numerous observations that at least a subset of cohesins are associated with DNA after anaphase onset and prior to S-phase (2123), suggesting that a cohesin subset remains from the previous cell cycle to mark loci destined for cohesion assembly reactions in the subsequent cell cycle (8, 16). Intriguingly, DNA helicases, such as budding yeast Chl1p and Sgs1p, are homologs of human factors (BACH1, Blooms, and Werners) implicated in genome maintenance, aging, and cancer progression (16, 2427). Here, we report on new findings that a nuclear envelope protein required for spindle pole body assembly also functions in sister chromatid cohesion. Several independent lines of evidence reveal that Mps3p physically associates with Ctf7p, a cohesion establishment factor. New mps3 alleles were generated and used to show that Mps3p indeed functions in cohesion and that this activity occurs after the G1 portion of the cell cycle. These findings reveal that a spindle pole body assembly factor performs a function that is critical to sister chromatid cohesion and proper chromosome segregation.
Two-hybrid AssayspJL59 containing CEN, TRP1, and GAL4 DNA-binding domain (BD) sequences was obtained by digesting pPC62 (CEN, LEU2, and GAL4(BD) and replacing the CEN-LEU2 cassette for a CEN-TRP1 cassette obtained by PvuI digestion of pRS315 (a generous gift from John Lamb) (28). pJL59 was linearized with ClaI and ligated in the presence of CTF7 obtained by BstBI digestion of pBS2 (9). CTF7 inserted in the reverse orientation was also obtained. Two-hybrid analyses were performed as described (29). Proper functioning of Gal4(BD)-Ctf7p was confirmed by transformation into a ctf7 deletion strain containing (pRS316-CTF7). All transformants remained viable on medium supplemented with 5'-FOA1 to counterselect for the pRS316-based plasmid (30). In contrast, transformants harboring either Gal4(BD) vector alone or vector with CTF7 placed in the reverse orientation were inviable upon 5'-FOA exposure.
GST Pull-downsGST-CTF7 expression in Escherichia coli cells was performed as previously described (14). Briefly, yeast cells containing Mps3-13MYCp, Mps3 Co-immunoprecipitationCells containing Mps3-MYCp were transformed with either Ctf7-HAp or HA only vector control 2-µm plasmids. Transformed cells were grown to log phase and then lysed with glass beads. The lysate was centrifuged, and the resulting pellet was treated with lysis buffer supplemented to 1 M NaCl and centrifuged, and the resulting salt-extracted supernatant was harvested for immunoprecipitation. Co-immunoprecipitation was performed as previously described with minor modifications (14). Briefly, supernatants were diluted in Buffer A (50 mM Tris (pH 7.5), 50 mM NaCl, 0.2% Triton X-100) supplemented with protease inhibitors (Roche Molecular Biochemicals), incubated with A-14 polyclonal anti-MYC (Santa Cruz Biotechnology) at 4 °C, and then incubated with Protein A-Sepharose CL-4B beads (Amersham Biosciences) at 4 °C overnight. A parallel sample was also treated with Protein A-Sepharose beads without anti-MYC antibody. In both cases, beads were washed with Buffer A several times using low speed centrifugation, and bound proteins were eluted with SDS solubilization buffer. Western blot analysis was carried out using Y-11 monoclonal anti-HA antibody (Santa Cruz Biotechnology) followed by goat anti-mouse horseradish peroxidase (Bio-Rad), and ECL-Plus (Amersham Biosciences). Generation of mps3 Mutant AllelesPCR mutagenesis was performed using pSJ140 (pRS316-MPS3; generously provided by Drs. S. Jaspersen and M. Winey) as DNA template and limiting amounts of dNTPs as previously described (9, 31). NcoI and NdeI were used to release a 1.9-kbp fragment from pRS315-MPS3. The remaining vector and PCR product were then co-transformed into a yeast strain containing an MPS3 knockout with KAN covered by pSJ140. Approximately 6000 transformants were plated to medium containing 5'-FOA to counterselect against pSJ140 (30). The remaining transformants (harboring potentially PCR-mutated MPS3 as the sole source of Mps3p function) were placed on rich medium at 23 and 37 °C and scored for temperature sensitivity. From a total of six temperature-sensitive mps3 strains identified, plasmids were recovered from the four tightest temperature-sensitive alleles. These mps3 alleles were cloned into an integrating vector (pRS402), linearized to promote integration at the ADE2 loci, and transformed into the MPS3-KAN knockout yeast strain covered by pSJ140. Transformants were then plated to medium containing 5'-FOA to counterselect against pSJ140, leaving the integrated PCR-mutagenized mps3 alleles as the sole source of Mps3p function. Stably transfected cells were then placed on rich medium at 23 and 37 °C and scored for temperature sensitivity. All four mutants retained temperature sensitivity and were named mps3-2, mps3-3, mps3-4, and mps3-5.
Cohesion AssayOur modified cohesion assay strain was previously described (13, 14, 32). Briefly, Tet operator repeats integrated
Spindle Pole Body Duplication AssayLog phase wild type and ctf7 mutant cells were placed in 23 °C rich medium supplemented with nocodazole for 3 h. The mitotic arrested cells were washed two times in 37 °C medium and allowed to proceed through the cell cycle at this temperature (restrictive for ctf7-203). Samples were harvested every 30 min and processed for flow cytometry and immunofluorescence. Time points in which both wild type and ctf7 cells were judged to have traversed the cell cycle and reenter mitosis were further processed to detect Tub4p and -tubulin (anti-Tub4p a generous gift from S. Jaspersen and M. Winey; B512 anti- -tubulin was purchased from Sigma). Only large budded cells (bud size roughly one-half of mother) were counted. Cells in which Tub4p was not detected were also not counted. Results shown represent averages for three independent time courses in which at least 50 cells were counted for each time point.
Genome-wide Two-hybrid Screen Identifies Mps3p as Associating with Ctf7pCtf7p is an essential yeast protein that functions to establish cohesion between sister chromatids (9, 21). To identify novel cohesion establishment factors, we performed a genome-wide two-hybrid screen. The CTF7 open reading frame (encoding for amino acids 26281) was fused in-frame 3' to the GAL4 DNA binding domain (BD) sequence (28). When expressed as the sole source of Ctf7p function in yeast, Gal4(BD)-Ctf7p rescues ctf7 strain viability. These results indicate that the Gal4(BD)-Ctf7p chimera is functional in vivo and further reveals that amino acids 125 are not essential for Ctf7p function. Prior to performing the two-hybrid screen, we determined that Gal4(BD)-Ctf7p did not promote expression of any of the three reporter genes (HIS3, ADE2, and lacZ) utilized in the two-hybrid assay (28). Nor did Gal4(BD)-Ctf7p activate reporter gene expression in the presence of a Gal4 activation domain (AD) vector. Yeast matings were used to introduce the GAL4(BD)-CTF7 construct into strains that harbor a genome-wide array of full-length open reading frame sequences fused to GAL4(AD) (29). Expressed AD-gene fusions that interact with GAL4(BD)-Ctf7p induce HIS3 expression, promoting growth on the appropriate selection plates. From this genome-wide screen, only two gene products exhibited strong interactions with Ctf7p. At the time that the screen was performed, one of these candidate interacting gene products was listed as an unidentified open reading frame (YJL019W). More recently, YJL019W was characterized as MPS3, a gene that encodes for an essential nuclear envelope spindle pole body assembly component. Defects in MPS3 result in a monopolar spindle phenotype (3, 4). Investigations into the candidate second interactor are currently under way. Mps3p Associates with Ctf7p under in Vitro Assembly ConditionsIt became important to independently test for a physical interaction between Mps3p and Ctf7p. We first employed a GST pull-down strategy to test for a physical association in vitro. One advantage of GST-based purification is that physical interactions are successfully detected even for Ctf7p, which occurs in vivo at very low copy (14). In addition, GST pull-downs test for proteins that are competent to associate in the absence of in vivo assembly reactions. The entire CTF7 open reading frame was inserted in-frame 3' to the GST DNA sequence to produce GST-CTF7 (14). Gst-Ctf7p is readily expressed in bacteria and retains functional activity in vitro (16, 33). We then obtained a yeast strain in which the sole source of Mps3p function is derived from MPS3-13MYC stably integrated into the genome (a generous gift from Drs. S. Jaspersen and M. Winey). Extracts from this stable integrated cell line contained a protein band of the appropriate molecular weight detectable by Western blot. A similar protein band was not detectable in the parental nontransformed strain.
To test for protein-protein interactions, GST-Ctf7p or GST alone were bacterially expressed and coupled to glutathione-Sepharose beads. Extracts from yeast cells expressing Mps3-13MYCp were obtained as previously described (14). The resulting soluble fraction was divided into four aliquots; one aliquot was incubated with Sepharose beads, the second aliquot was incubated with beads coupled to GST, and the third aliquot was incubated with beads coupled to GST-Ctf7p. The fourth aliquot served as a marker to identify Mps3p during Western blot analyses. The three bead matrices were copiously washed before eluting bound proteins off of the bead matrices. Western blot analyses of the eluted proteins reveal that Mps3p specifically associated with GST-Ctf7p, but that Mps3p did not associate with GST or beads alone (Fig. 1). These results provide independent confirmation of the two-hybrid interaction and demonstrate that Mps3p and Ctf7p associate in vitro. We repeated the GST pull-down using MYC-tagged Mps3p in which the C-terminal 60 amino acids are deleted (Mps3
Mps3p Associates with Ctf7p in VivoTo test that Mps3p associates with Ctf7p in vivo, we next performed co-immunoprecipitations in which epitope-tagged Ctf7p and Mps3p were co-expressed in yeast cells. To circumvent detection problems associated with low expression levels, we used a Ctf7-HAp construct that expresses Ctf7p at elevated levels. Importantly, this level of Ctf7-HAp expression is fully functional in vivo and provides for normal cell growth in otherwise ctf7 cells (9). Cells co-expressing Ctf7-HAp and Mps3 6013MYCp were processed for co-immunoprecipitations as previously described with minor modifications (16). Cells were lysed by mechanical disruption and centrifuged, and extracted soluble supernatants were incubated with MYC-directed antibodies and Protein A-Sepharose beads. Protein-bead complexes were washed several times prior to Western blot analysis. Anti-MYC staining reveals that Mps3p-13MYCp is efficiently immunoprecipitated under these conditions and that Mps3p immunoprecipitation requires both the MYC tag and MYC-immunoprecipitating antibody (data not shown). We then tested for the ability of Ctf7p-HA to co-immunoprecipitate with Mps3p. Parallel protein membranes containing immunoprecipitated Mps3p were probed using HA-directed antibodies. The results show that Ctf7p indeed co-immunoprecipitates with Mps3p, revealing that Ctf7p associates with Mps3p in vivo (Fig. 2). Importantly, Ctf7p co-immunoprecipitation with Mps3-MYCp depended on the presence of the MYC-precipitating antibody; Ctf7p remained in the supernatant and did not co-immunoprecipitate with Mps3p in the absence of the MYC-precipitating antibody (Fig. 2). In summary, the combined data obtained from two-hybrid, GST pull-down, and co-immunoprecipitation analyses reveal that Mps3p associates directly with Ctf7p.
Generation of New MPS3 AllelesGiven the physical association of Mps3p and Ctf7p, we hypothesized that Mps3p may function in sister chromatid cohesion. Often, haploid cells harboring defects in spindle pole body components such as MPS3 diploidize to produce cells with 2C and 4C DNA contents (3, 4), thereby complicating the quantification of cohesion defects. To circumvent this obstacle, we used a PCR strategy to generate conditional mps3 alleles that exhibit wild type DNA contents (31). Briefly, mps3 cells harboring pRS316-MPS3 were co-transformed with pooled PCR-mutagenized MPS3 products and pRS315-MPS3 with the majority of the MPS3 open reading frame deleted. Of 6000 transformants, six scored positive as temperature-sensitive after pRS316-MPS3 counterselection (34). The four tightest temperature-sensitive alleles were stably integrated into the genome of mps3 cells, providing the only source of Mps3p function in these cells. All four mutant strains (mps3-2, mps3-3, mps3-4, and mps3-5) exhibited temperature sensitivities relative to a wild type control strain (Fig. 3). DNA sequence analyses for each of these new mps3 mutations are summarized in Fig. 4.
We tested whether these four mps3 mutant strains exhibited a diploidization phenotype. All four strains were allowed to proceed through several cell divisions in order to accentuate the diploidization phenotype. Flow cytometry analyses of log phase wild type and mps3 strains reveal an allelic series (Fig. 3). mps3-2 and mps3-4 both exhibit a completely diploidized DNA content, consistent with previously described mps3 alleles (3, 4). In contrast, mps3-5 mutant cells exhibit a low to intermediate level of diploidization within the cell population. mps3-3 mutant cells exhibit predominantly wild type haploid levels of DNA that can be maintained for extended periods of time. Thus, the mps3-3 allele provides an important tool with which to test for cohesion defects. Mps3p Functions in Sister Chromatid CohesionTo investigate the possibility that Mps3p might play a role in sister chromatid cohesion, mps3-3 cells exhibiting wild type levels of DNA were crossed with a cohesion assay strain and sporulated, and appropriate spores were obtained. Diploidized cells mated to haploid cells fail to produce viable progeny. Thus, any mps3-3 parental cells that might have diploidized prior to mating failed to contribute to the resulting progeny upon sporulation; all spores exhibited only wild type DNA contents. The cohesion assay strain contains Tet operator repeats integrated 40 kb proximal to the centromere of chromosome V and also expresses GFP-tagged Tet repressor protein, allowing for visualization of the centromere-proximal locus for each sister chromatid (14, 32). Indirect immunofluorescence further allows for nuclear detection of Pds1p, a biochemical marker of preanaphase cells (35, 36). Thus, cell morphology, GFP-tagged chromosomal loci, and coincident DNA and Pds1p staining were simultaneously assessed on a cell-by-cell basis to map the disposition of sister chromatids in preanaphase cells.
To assay for cohesion defects, log phase wild type and mps3-3 mutant strains were synchronized in G1 at 23 °C with Mps3p Functions in Cohesion Establishment, Not MaintenanceGiven Mps3p association with Ctf7p and role in cohesion, it became important to test whether Mps3p plays a role in cohesion maintenance or establishment. We directly tested this model using methods previously employed to reveal a role for structural cohesins in cohesion maintenance (23, 32). Briefly, log phase wild type and mps3-3 cells were arrested in mitosis at the permissive temperature using medium supplemented with nocodazole. After 2 h, the cultures were split, and half of each culture was shifted to 37° for an additional 1 h. This regimen results in loss of cohesion for structural cohesin mutants but not establishment factors (9). As before, parallel cell samples were harvested and assessed for DNA content, cell morphology, Pds1p content, and disposition of sister chromatid loci via GFP. Despite the absence of presynchronization in G1, nocodazole treatment produced predominantly mitotic cultures over this 3-h time course (wild type and mps3-3 cultures containing 76.9 versus 79.9% 2C DNA contents, respectively). We normalized to 100% the frequency of wild type cells that were large budded and contained a 2C DNA content and tightly paired sisters. In comparison, mps3-3 mutant strains treated identically exhibited a frequency of paired sister chromatids that was equal to or greater than that of wild type cells; no cohesion defect was observed (Fig. 6). These results reveal that Mps3p does not function in cohesion maintenance but instead that Mps3p participates in cohesion establishment, consistent with its association with Ctf7p.
Mps3p Association with Ctf7p Is Physiologically Relevant Mps3p both binds Ctf7p and functions in cohesion. One prediction of these findings is that mps3-3 strains may fail in cohesion due to loss of Ctf7p-Mps3p interactions. To test this model directly, we epitope-tagged the mps3-3 temperature allele and compared its association with Ctf7p with that of wild type Mps3p. GST-Ctf7p, GST alone, and glutathione-Sepharose beads were incubated with extracts from yeast cells expressing either Mps3-13MYCp or Mps3-3-13MYCp as described above. In this case, the binding reaction was performed for 30 min at 37 °C. Afterward, the three bead matrices were copiously washed before eluting the bound proteins. Western blot analyses of the eluted proteins confirmed again that Mps3p specifically associated with GST-Ctf7p but not with GST or beads alone (Fig. 7). In contrast, Mps3-3-13MYCp failed to bind to GST-Ctf7p at 37 °C (Fig. 7). Loss of binding was temperature-dependent such that binding reactions performed at 23 °C resulted in Mps3-3-13MYCp associating with Ctf7p (data not shown). These findings reveal that the association of Mps3p with Ctf7p is physiologically relevant; Mps3p competent to promote cohesion binds to Ctf7p, whereas Mps3-3p that is deficient in cohesion fails to associate with Ctf7p.
mps3-3 Exhibits Allele-specific Conditional Synthetic Lethality When Combined with ctf7-203The physical association of Mps3p with Ctf7p raised the possibility that mps3 and ctf7 alleles also might interact genetically. To test this prediction, mps3-3 was crossed with ctf7-203, and the resulting diploids were sporulated. For the mps3-3 ctf7-203 cross, 30 of 31 tetrads produced viable spores. Of the 120 possible spores from the 30 tetrads, six mps3 ctf7 double mutant strains were obtained, nearly identical to the expected yield of seven spores (Table I). We then tested whether mps3-3 ctf7-203 double mutant cells would exhibit growth characteristics different from strains harboring either allele alone. Importantly, all 6 mps3-3 ctf7-203 double mutant strains exhibited a robust conditional synthetic lethal interaction. At 23 °C on YPD-rich medium, a dilution series of single mutant mps3-3 or ctf7-203 cells and double mutant mps3-3 ctf7-203 cells produced dense growth plaques. At 29 °C, single mutant strains ctf7-203 and mps3-3 remained viable. In contrast, however, all six of the double mutant mps3-3 ctf7-203 strains were inviable (Fig. 8). All single and double mutant strains were inviable at 37 °C. A similar interaction was not found for mps3-5 ctf7-203 double mutants. These findings that mps3-3 and ctf7-203 exhibit allele-specific synthetic lethality support a model in which Mps3p and Ctf7p physically associate to perform roles in cohesion establishment.
Ctf7p Is Not Required for Spindle Pole Body DuplicationWe next tested whether Ctf7p plays a role in spindle pole body duplication. Ctf7p is known to establish cohesion during S-phase (9, 21). Thus, we first tested whether S-phase Ctf7p activity was required for spindle pole body duplication. Log phase wild type and ctf7 mutant cells were arrested in G1 at 23 °C with -factor, washed, and released in 37 °C rich medium. These cells were then allowed to proceed through the cell cycle at 37 °C, abrogating Ctf7-203p function during S-phase and G2. Upon release from G1, cell samples were harvested every 15 min and assayed for DNA content and cell morphology. The disposition of spindle pole bodies was determined in cell fractions in which the majority of cells were in G2/M, as assayed by flow cytometry, and for cells in which the bud size was at least equal to one-half of the mother cell when viewed microscopically. To monitor for spindle pole body duplication, we visualized individual spindle poles by indirect immunofluorescence using a Tub4p-specific antibody (a generous gift from Drs. S. Jaspersen and M. Winey) (Fig. 9). Assessing spindle poles on a cell-by-cell basis, the results show that both wild type and ctf7 mutant cells exhibited similar low levels of monopolar spindles (Fig. 9). Occasionally, we observed that Tub4p staining was unequal between the spindle pole bodies. Thus, whereas our results probably overestimate the number of monopolar spindles in both wild type and ctf7 mutant cells (cells in which the weaker spindle pole staining was not detected), these data support a model in which Ctf7p does not perform an essential function in spindle pole duplication during S-phase. We also tested whether Ctf7p functions in spindle pole body duplication during G1. In this case, log phase wild type and ctf7 mutant cells were arrested in preanaphase at 23 °C with nocodazole, released into rich medium held at 37 °C, and then allowed to progress through the cell cycle at 37 °C (abolishing Ctf7-203p activity through G1, S-phase, and G2). Quantification of Tub4p staining by immunofluorescence revealed that Ctf7p inactivation early in the cell cycle did not abrogate spindle pole duplication; both wild type and ctf7 mutant cells exhibited similarly low levels of monopolar spindles (Fig. 9).
In this study, we used the well established cohesion establishment factor Ctf7p as a platform to identify novel cohesion proteins. Mps3p was found to interact with Ctf7p by a genome-wide two-hybrid screen. This physical interaction was confirmed both in vivo and in vitro using GST pull-down and co-immunoprecipitation strategies. The association of Ctf7p and Mps3p may not be surprising given that Ctf7p is a nuclear protein. Mps3p is a nuclear envelope protein that also associates with spindle pole body half-bridges (3, 4). Thus, Ctf7p and Mps3p may associate at the interface of the nucleoplasm and nuclear envelope especially near spindle pole bodies. Mps3p may also interact directly with chromatin components, consistent with the interactions between Ctf7p and Mps3p reported here and by previous observations that Mps3p associates with several DNA associated factors including Est1p, Hta3p, and Ndj1p (29, 37). We used a cohesion assay strain to demonstrate that Mps3p plays a functional role in cohesion establishment. The cohesion defect (23 versus 3% after 5 h and 14.2 versus 2.75% after 3 h of nocodazole) for mps3 mutant cells, compared with wild type, is similar to levels exhibited by numerous nonessential cohesion factors, including trf4 (20%), chl1 (23%), ctf18 (25%), ctf8 (30%), and ctf18 (35%) (12, 13, 15, 16). In contrast, mutations in essential cohesion factor genes, including ctf7, mcd1/scc1, irr1/scc3, pds5, smc1, and smc3, typically exhibit twice this level (5060%) of sister separation (9, 21, 23, 32, 3841). One interpretation of these findings is that Mps3p may play an essential role in cohesion that is redundant to other proteins, diminishing the cohesion defect observed in mps3 alleles. Alternatively, Mps3p may perform a key but nonessential role in cohesion in addition to its essential role in spindle pole body duplication. This latter model is supported by findings that mps3-1 mutant cells exhibit several phenotypes (including a mitotic delay) at the permissive temperature (3) and that elevated levels of Mps3p fail to rescue the essential function of Ctf7p (this study). Precedence for this model comes from studies involving Rfc5p, Rfc2p, Rfc4p, and Pol2p, which all perform essential roles in DNA replication but, when mutated, exhibit levels of cohesion loss similar to both mps3 mutant strains and nonessential cohesion factors (10, 11, 13, 14). Our finding that a spindle pole body protein functions in cohesion is further supported by new evidence that Kar3p, a minus-end-directed spindle pole microtubule motor, participates in cohesion at levels similar to those found for Mps3p (20). Thus, our data both provide a new function for Mps3p and greatly extend a novel link between cohesion establishment and spindle pole body assembly. Given that spindle pole body duplication and cohesion establishment both occur early in the cell cycle, it is not surprising that these fundamental chromosome segregation processes are coordinated. Formally, the physical and functional interactions between Mps3p and Ctf7p can be envisioned in numerous ways. For instance, Ctf7p may perform a dual function in both cohesion establishment and spindle pole body duplication. At the level of detection used here, however, spindle pole body duplication appears normal in ctf7 mutant cells. Genetic assays further suggest that Ctf7p does not function in spindle pole body assembly; elevated levels of Mps3p did not rescue ctf7 temperature sensitivity, and elevated levels of Ctf7p did not rescue mps3 phenotypes. Another model, based on the role of Mps3p in cohesion, is that the nuclear envelope or spindle pole body serves as a staging or assembly platform for cohesion processes. In this scenario, abrogation of spindle pole body assembly coordinately results in misassembly of cohesion complexes. The identification of two spindle pole body components, Mps3p (this study) and Kar3p (20), as cohesion factors provides novel and independent lines of evidence that spindle pole body components function in cohesion. There is strong precedence that the nuclear envelope plays a role in chromatin metabolism. For instance, perinuclear proximity is a key factor in establishing transcriptionally silent loci such as the HMR (42). The reciprocal is also true; mutations in silencing genes result in loss of both nuclear pore and nucleolar architectures (43). Recently, components required for chromatin anchorage to the nuclear envelope have been identified. Esc1p, an inner face nuclear envelope protein (along with both Ku80 and Sir4p) is required to anchor chromatin to the nuclear envelope, whereas Esc1p-dependent anchorage occurs independent of silencing (44). Furthermore, Mps1p is required for both spindle pole body duplication (including Spc42p incorporation) and mitotic checkpoint function (4547). Consistent with these dual activities, Mps1p is located at both the spindle pole and kinetochores. Similar kinetochore/spindle pole body co-localization has been reported for Mad2p, anaphase-promoting complex subunits, and a host of centromere-binding proteins (8, 48). In parallel, numerous nuclear pore proteins now appear to associate with kinetochores in both yeast and mammalian cells and function in transcriptional repression (4951). The functional interaction between Ctf7p and Mps3p reported here suggests that spindle pole body proteins play an additional and key role in coordinating assembly reactions required for establishing sister chromatid cohesion.
* This material is based upon work supported by National Science Foundation Grant MCB-0212323 (to R. V. S.) and by National Center for Research Resources, National Institutes of Health, Grant P41 RR11823 (to S. F.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ Present address: Institut fur Genetik, Forschunsgszentrum Karlsruhe, Germany. ** To whom correspondence should be addressed. Tel.: 610-758-6162; E-mail: rvs3{at}Lehigh.edu.
1 The abbreviations used are: 5'-FOA, 5'-fluoroorotic acid; GST, glutathione S-transferase; DAPI, 4',6-diamidino-2-phenylindole; HA, hemagglutinin; BD, binding domain; AD, activation domain.
We are especially indebted to Drs. Sue Jaspersen and Mark Winey for advice, expertise, and sharing of reagents during the course of this study. We also thank Skibbens laboratory members Dr. A. Brands and B. Satish for critical reading of the manuscript.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||