Saccharomyces cerevisiae BUB2 Prevents Mitotic Exit in Response to Both Spindle and Kinetochore Damage

Corresponding author: Forrest Spencer, Johns Hopkins University School of Medicine, 720 Rutland Ave./Ross 850, Baltimore, MD 21205., fspencer{at}jhmi.edu (E-mail)

Communicating editor: P. RUSSELL

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The spindle assembly checkpoint-mediated mitotic arrest depends on proteins that signal the presence of one or more unattached kinetochores and prevents the onset of anaphase in the presence of kinetochore or spindle damage. In the presence of either damage, bub2 cells initiate a preanaphase delay but do not maintain it. Inappropriate sister chromatid separation in nocodazole-treated bub2 cells is prevented when mitotic exit is blocked using a conditional tem1^c mutant, indicating that the preanaphase failure in bub2 cells is a consequence of events downstream of TEM1 in the mitotic exit pathway. Using a conditional bub2^tsd mutant, we demonstrate that the continuous presence of Bub2 protein is required for maintaining spindle damage-induced arrest. BUB2 is not required to maintain a DNA damage checkpoint arrest, revealing a specificity for spindle assembly checkpoint function. In a yeast two-hybrid assay and in vitro, Bub2 protein interacts with the septin protein Cdc3, which is essential for cytokinesis. These data support the view that the spindle assembly checkpoint encompasses regulation of distinct mitotic steps, including a MAD2-directed block to anaphase initiation and a BUB2-directed block to TEM1-dependent exit.

THE cell division cycle in most eukaryotes is tightly regulated by checkpoints that monitor and ensure appropriate passage past critical steps essential for successful cell division. The spindle assembly checkpoint serves to arrest cells at prometaphase until each sister chromatid pair has achieved bipolar attachment to the spindle (RIEDER and SALMON 1998 ; SKIBBENS and HIETER 1998 ).

Chromatid attachment to the spindle can be affected by impaired spindle microtubules or kinetochores, and the spindle assembly checkpoint is experimentally induced using antimicrotubule drugs such as nocodazole or conditional alleles that interfere with kinetochore protein function. A single unattached kinetochore appears to emit an inhibitory signal through the spindle assembly checkpoint pathway that prevents the onset of anaphase and subsequent mitotic events (SPENCER and HIETER 1992 ; NICKLAS et al. 1995 ; RIEDER et al. 1995 ). In Saccharomyces cerevisiae, the MAD1-3 (LI and MURRAY 1991 ) and the BUB1-3 (HOYT et al. 1991 ) genes were first identified through genetic screens for the inability to delay or arrest in the presence of microtubule-damaging drugs. When kinetochore structure is compromised S. cerevisiae cells arrest prior to anaphase despite having intact microtubules, and this arrest is dependent on a subset set of genes that are involved in the response to spindle damage induced by antimicrotubule drugs (WANG and BURKE 1995 ; PANGILINAN and SPENCER 1996 ). Mad1, Mad2, Bub1, and Bub3 proteins have been shown to localize to kinetochores in experimental systems that permit immunocytochemical localization of chromosomal structures (reviewed by SKIBBENS and HIETER 1998 ). The kinetochore localization and associated biochemical properties of this group of proteins have suggested a model wherein an unattached kinetochore emits a signal that prevents Cdc20 protein from directing the anaphase promoting complex (APC)-mediated degradation of the anaphase inhibitor Pds1 protein and, consequently, metaphase-to-anaphase transition (COHEN-FIX et al. 1996 ; VISINTIN et al. 1997 ; ELLEDGE 1998 ; LIM et al. 1998 ; SCHOTT and HOYT 1998 ). This aspect of the spindle assembly checkpoint is highly conserved, supported by observations of functional interactions between MAD2 and CDC20 in widely divergent organisms (FANG et al. 1998 ; HWANG et al. 1998 ; KALLIO et al. 1998 ; KIM et al. 1998 ).

In contrast, the role of the Bub2 protein has only recently begun to emerge. Like the other BUB and MAD genes, it is not required for cell viability in budding yeast and its absence causes a hypersensitivity to microtubule-depolymerizing drugs (HOYT et al. 1991 ). Unlike the other MAD and BUB genes, the bub2 mutant exhibits a partial preanaphase arrest (PANGILINAN and SPENCER 1996 ). Several recent studies have implicated BUB2 in the control of a mitotic regulatory step independent of MAD2 in the presence of spindle damage (ALEXANDRU et al. 1999 ; FESQUET et al. 1999 ; FRASCHINI et al. 1999 ; LI 1999 ). The late regulatory step is proposed to be exit from mitosis as defined by a decrease in M-phase-specific cyclin-dependent kinase activity, controlled by an elaborate "mitotic exit network" (JASPERSON et al. 1998 ). The execution of mitotic exit in S. cerevisiae depends on re-gulated GTPase (TEM1), protein kinase (CDC15, CDC5, DBF2, DBF20), and phosphatase (CDC14) activities that ultimately control an HCT1-associated form of the anaphase promoting complex (PETERS 1999 ). APC^HCT1-directed polyubiquitination of CLB2 leads to its destruction and establishment of G1 cyclin-associated forms of cyclin-dependent kinase (MORGAN 1999 ).

A role for S. cerevisiae BUB2 in mitotic exit is supported by evidence from characterization of the Schizosaccharomyces pombe homolog cdc16⁺, which has been shown to regulate both Clb2 protein degradation and cytokinesis (FANKHAUSER et al. 1993 ). The S. cerevisiae BUB2 gene complements the temperature-sensitive allele cdc16-116 in S. pombe (FANKHAUSER et al. 1993 ). The possibility that BUB2 serves to prevent premature exit from mitosis when anaphase is inhibited has served as a working hypothesis.

This study uses synchronized budding yeast strains to show that bub2 mutants are unable to maintain a mitotic arrest when kinetochore damage is induced. This is significant to our current understanding of the role of BUB2, as it has been suggested that BUB2 is not involved in the cellular response to kinetochore damage but rather senses another (as yet undefined) aspect of spindle structure or function (WANG and BURKE 1995 ; TAVORMINA and BURKE 1998 ; LI 1999 ). We show that the inappropriate sister chromatid separation in the presence of spindle damage in bub2 mutants is dependent on the TEM1-mediated signaling of mitotic exit. In addition, we have constructed a conditional BUB2 allele (BUB2^tsd) and used it to show that the continuous presence of Bub2 protein is required for spindle assembly checkpoint-mediated arrest to be maintained. BUB2 is not required for the maintenance of the DNA damage checkpoint-mediated mitotic arrest induced in cdc13-1 mutants, precluding a global role for BUB2 in maintaining mitotic arrest in response to diverse triggers. In a yeast two-hybrid screen with Bub2 protein as bait, the septin CDC3 was identified. Recombinant Bub2 and Cdc3 proteins also associate in vitro. Cdc3 belongs to the highly conserved septin family of proteins essential for cytokinesis (KIM et al. 1991 ; FIELD and KELLOG 1999 ). These results suggest the possibility that the interaction between Cdc3 and Bub2 proteins links the spindle assembly checkpoint to a septin-mediated role in cytokinesis or mitotic exit.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plasmids and strains:
The bub2^tsd allele, a conditional allele with a permisive temperature of 25° and restrictive temperature of 37°, was constructed in plasmid pRaj154 (BUB2 promoter-UBIQUITIN-DHFR^ts-BUB2-pRS304) by combining the following fragments in a four-way ligation: (i) a 600-bp segment from immediately upstream of the BUB2 open reading frame generated by PCR using oligonucleotides OLF208 (5'-gactacgagctcgctgttgacgggggctcta-3') and OLF209 (5'-cggaattc gcaaaagttaacaag-3') and cut with SacI and SalI; (ii) an EcoRI/HindIII fragment from pJW8 (DOHMEN et al. 1994 ) carrying the ubiquitin-DHFR^ts fusion moiety; (iii) a 474-bp NH₂-terminal segment of the BUB2 open reading frame generated by PCR using oligonucleotides OLF210 (5'-cccaagcttatgacctcaatt gaa gat-3') and OLF211 (5'- acgcgtcgacagtgaaaagttga tacg-3') and cut with HindIII and SalI; and (iv) SacI- and SalI-treated pRS304 (SIKORSKI and HIETER 1989 ) whose unique ApaI site had been destroyed. Plasmid pWS103 (SHOU et al. 1999 ), carrying a conditional degron allele of TEM1 (designated as tem1^c in the integrating vector pRS304; SIKORSKI and HIETER 1989 ), was a gift from R. Deshaies. It encodes a UBIQUITIN-LacI-Tem1 fusion gene driven by the galactose responsive GAL1 promoter.

For yeast two-hybrid analysis, the GAL4 DNA-binding domain was fused with a full-length BUB2 open reading frame in plasmid pRaj72, constructed by ligation of a 937-bp PCR amplification product from genomic DNA (using oligonucleotides OLF161 (5'-atccgtcgacccatggagatgacctcaattgaagatctgata-3') and OLF162 (5'-catgccatgggcgtcgacttacggtatatatatgtct gggt-3') into NcoI-SalI-cut pAS2 (gift of S. Elledge) as a NcoI-SalI fragment. Yeast two-hybrid host strains and control plasmids were obtained from S. Elledge (BAI and ELLEDGE 1997 ) and from CLONTECH (Palo Alto, CA). The GST-CDC3 fusion encoded in plasmid pRaj134 was constructed by PCR amplification of the CDC3 open reading frame from YRaj127 using oligonucleotides OLF229 (5'-cgcgccatggaggaattcatgagtttaaag gaggaacacgtg-3') and OLF230 (5'-ggcgtcgacctaacgtaaaaatcccttcttc-3') cloned into EcoRI/SalI-digested pGEX4T-1 (Pharmacia, Piscataway, NJ).

Yeast transformation and genetic manipulation were performed according to published methods (GUTHRIE and FINK 1991 ). Experimental and control strain sets within any single experiment were isogenic derivatives related by DNA-mediated transformation or backcrossing. Yeast strains and genotypes are shown in Table 1. Strain YRaj132, carrying the BUB2^tsd allele, was generated by transforming YRaj127 with ApaI-linearized pRaj154, followed by selection for TRP1. This strain expresses the BUB2^tsd from a natural BUB2 promoter and also carries, in tandem, a nonfunctional truncated BUB2 gene ending at amino acid 158. Strains YRaj144 (cdc13-1 bub2) and YRaj140 (cdc13-1) were generated by transforming XhoI-linearized pVL451 (gift of V. Lundblad) into YRaj130 and YRaj127, respectively, under selection for URA3. 5-Fluoroorotic acid-resistant derivatives were subsequently screened for temperature sensitivity at 37°. The introduction of the bub2::LEU2 deletion has been described previously (PANGILINAN and SPENCER 1996 ). In the case of YFS1214 and YFS1215 used for the green fluorescent protein (GFP) sister chromatid labeling experiments, the bub2::URA3 disruption plasmid pTR24 (HOYT et al. 1991 ) was used for one-step gene replacement in strains AFS173 and AFS387 (STRAIGHT et al. 1996 ; gifts from Straight lab) to introduce bub2::URA3 into wild-type and mad2-1 backgrounds, respectively. The mad2::HIS3 allele was obtained from M. MAYER and P. HIETER (unpublished results). Conditional Tem1 protein expression was achieved in strains YRaj164, YRaj166, and YRaj169 by transforming strains YFS1214, YFS1215, and AFS387, respectively, with EcoRI-linearized pWS103. The TEM1 conditional degron allele, tem1^c, disrupted the resident TEM1 gene. Tem1^c-carrying haploids were inviable in glucose media and arrested with separated nuclei.

Analysis of new bud formation, mitotic arrest, and viability:
Standard cell culture procedures were followed (GUTHRIE and FINK 1991 ). For factor synchronization experiments, log phase cells were incubated with 50 µg/ml factor (Sigma, St. Louis) for 2 hr at 25° followed by washing and release into appropriate medium at the stated temperatures. For visualization of GFP fluorescence, cells were fixed in 4% paraformaldehyde. In experiments using nocodazole, drug concentration was at 15 µg/ml. Cells were prepared for microscopy by fixation in 3.7% formaldehyde in SK (20 mM sorbitol, 50 mM phosphate buffer pH 7.5) and staining with 4',6-diamidino-2-phenylindole (DAPI) at a final concentration of 300 ng/ml. Very large budded, uninucleate cells, indicating a preanaphase delay or arrest morphology, were defined as mother cells containing a single nuclear DNA mass and with a single daughter bud. New bud formation was observed as the presence of a second bud (a total of three connected cell bodies) containing a single nuclear DNA mass.

Cell viability was assayed using microcolony formation as follows. After nonpermissive temperature or nocodazole treatment, cell aliquots were returned either to permissive temperature or nocodazole-free medium, respectively, and plated at a density of 500 cells/cm² on YPD. Inviable cells were scored as those that failed to produce at least 50 cell bodies/microcolony after incubation for 20 hr, at which time wild-type microcolonies comprised >100 cell bodies. In most cases inviable microcolonies contained <10 cell bodies.

Yeast two-hybrid screening:
The plasmid pRaj72 was transformed into the yeast two-hybrid strain CG1945 (CLONTECH). The yeast cDNA library (gift of S. Elledge) was transformed into pRaj72 (GalBD-BUB2 fusion) carrying CG1945, and transformants were directly selected on plates lacking tryptophan, leucine, and histidine and containing 10 mM 3-aminotriazole (3-AT; Sigma). Approximately 10,000 independent transformants were screened, yielding seven colonies with robust and reproducible growth on selective media by 4–7 days. These colonies were tested for ß-galactosidase activity using the filter assay (BAI and ELLEDGE 1997 ). The interacting plasmid from one of three colonies showing ß-galactosidase activity was recovered and retransformed along with pRaj72 (GalBD-Bub2) into strain YJ69-2a (JAMES et al. 1996 ), which carries an ADE2 as well as HIS3 reporter gene. These transformants were selected on medium lacking tryptophan and leucine (for establishment of the plasmids). Subsequent plating on medium also lacking histidine and adenine and containing 20 mM 3-AT confirmed the strength and plasmid dependence of the two-hybrid interaction. Sequence analysis of the insert in the interacting library plasmid revealed the presence of coding sequence from CDC3 (amino acids 315 to 521) cloned in frame C-terminal to the pACT2-GAL4 activation domain (GalAD-Cdc3). The GalBD-Cdc16 fusion expressing the cdc16⁺ of S. pombe was a gift from C. Albright (FURGE et al. 1998 ). The GalBD-p53 and GalAD-SVT plasmids expressing fusions of p53 and SV40 T antigen, respectively, were from CLONTECH.

In vitro protein interaction assays:
GST-CDC3 fusion and GST proteins were expressed in Escherichia coli DH5-alpha (Life Technologies) from plasmid pRaj134 and pGEX4T-1 (Pharmacia), respectively, and purified according to manufacturer's instructions (Pharmacia). Cell lysis and binding reactions were performed in ice-cold phosphate-buffered saline pH 7.4, 1% Triton X-100. ³⁵S-labeled BUB2 protein was synthesized in vitro using a coupled transcription, translation rabbit reticulo-lysate system (Promega, Madison, WI). Reactions were carried out according to manufacturer's instructions. Equal volumes of plain glutathione beads or beads carrying GST or GST-CDC3 were incubated with 35 µl out of a total of 150 µl of in vitro-translated extract in a total volume of 100 µl on ice for 1 hr. The bead samples were then pelleted, and unbound material was removed in four successive washes with ice-cold binding buffer. After the final wash, proteins were eluted in 50 µl binding buffer containing 10 mM glutathione and 10 µl of 5x SDS polyacrylamide gel electrophoresis sample buffer was added. The samples were boiled and analyzed on duplicate 10% SDS polyacrylamide gels (25 µl volume/lane). One gel was processed for autoradiography and the other for silver staining.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Kinetochore-induced preanaphase arrest requires MAD2 for establishment and BUB2 for maintenance:
CTF13 (DOHENY et al. 1993 ) encodes a component of CBF3, an essential kinetochore protein complex (RUSSEL et al. 1999 ). The temperature-sensitive allele ctf13-30 induces a preanaphase delay, which is morphologically visible as an accumulation of very large budded uninucleate cells. This class of delaying cells occurs at a frequency of 1–2% in wild-type cultures, compared to 80% in ctf13-30 cultures at the nonpermissive temperature (SPENCER and HIETER 1992 ; DOHENY et al. 1993 ). The preanaphase arrest in ctf13-30 cells depends on the spindle assembly checkpoint (WANG and BURKE 1995 ; PANGILINAN and SPENCER 1996 ) and requires MAD1, MAD2, BUB1, and BUB3 gene functions. Although it has been suggested that BUB2 is dispensable for the checkpoint arrest induced by kinetochore damage (WANG and BURKE 1995 ), we have observed that the ctf13-30-induced delay is reduced in ctf13-30 bub2 populations (PANGILINAN and SPENCER 1996 ). This reduction might reflect a smaller proportion of cells entering a delay or, alternatively, the presence of a shorter delay.

Examination of synchronous cultures supports the latter interpretation. ctf13-30 kinetochore damage was induced at a semipermissive temperature (34°) sufficient to induce a preanaphase delay without a significant loss of viability, allowing observation of the delay dynamics as well as the associated impact on viability. ctf13-30 strains lacking MAD2 or BUB2 were compared to isogenic relatives with an intact checkpoint. -Factor-arrested cultures were released (at t = 0) into pheromone-free medium at 34° and followed for two cell cycles. At each time point, preanaphase delay was measured as the frequency of very large budded uninucleate cells, and viability was measured as the frequency of microcolony-forming units plated at permissive temperature (25°). As expected, strains without a ctf13-30 defect never produce delaying cells and suffer no loss of viability on shift to 34° (Fig 1A and Fig B). In cells with an intact checkpoint, ctf13-30 produces a sustained delay that does not completely subside but accumulates in the next cell cycle (Fig 1A). In contrast, mad2-1 ctf13-30 cells are completely unable to delay and this is accompanied by a steady loss of viability beginning when cells with an intact checkpoint normally start to delay (compare to ctf13-30 at 60–80 min; Fig 1A and Fig B). Interestingly, the bub2 ctf13-30 strain initially produces delaying cells to approximately the same level as ctf13-30 cells with an intact checkpoint, but the frequency of delayed cells declines sharply at 120 min (Fig 1A). The population remains synchronous and repeats this pattern through a second cell cycle. The decline at 120 min is accompanied by onset of loss of viability, which appears later in bub2 ctf13-30 than in mad2-1 ctf13-30 mutants (Fig 1A and Fig B). This suggests that BUB2 has a role in delay maintenance while MAD2 is important for delay establishment, and both genes are important for the viability of cells containing kinetochore damage.

View larger version (25K): In this window In a new window Download PPT slide   Figure 1. Damaged kinetochore-induced mitotic arrest in wild-type and bub2 cells. (A) Preanaphase delay in synchronous ctf13-30 cultures at 34°. Indicated strains were arrested in -factor at 25° and released from arrest into pheromone-free medium at 34° (t = 0 min). Aliquots taken at 20-min intervals were fixed in formaldehyde, stained with DAPI, and scored for the presence of anaphase delay. The frequency (%) of large budded uninucleate cells is shown for each strain. Morphological parameters were as previously defined (SPENCER and HIETER 1992 ), where scored cells are those with bud diameter >75% that of the mother sphere and a single DAPI-staining chromosomal mass at the mother-bud junction. This criterion identifies cells that are very rarely observed in wild-type cell cycles in the absence of mitotic checkpoint induction. (B) Loss of viability in synchronous ctf13-30 cultures at 34°. Aliquots of samples removed for assay in (A) were diluted, sonicated, and plated on YPD for 24 hr at 25° and scored for microcolony formation. Strains were wild type (YFS824), mad2-1 (YFP451), mad2-1 ctf13-30 (YFS449), bub2 (YFS822), bub2 ctf13-30 (YFS961), and ctf13-30 (YFS884). Nearly identical time course results were obtained in two independent experiments, and representative results from a single experiment are shown.

MAD2 and BUB2 control distinct steps within mitosis, and both control sister chromatid separation:
Exposure of cells to microtubule-destabilizing drugs like nocodazole also induces the spindle assembly checkpoint. At high concentration (15 µg/ml), nocodazole completely blocks formation of a mitotic spindle, and cells arrest in a prometaphase state. Failure of the spindle assembly checkpoint causes these cells to exit mitosis without segregating chromosomes to the spindle poles; they are unable due to the absence of a spindle. In such cells, the appearance of a new cycle can be very simply assayed as new (second) bud formation (HOYT et al. 1991 ). Although the reason for retention of the first daughter cell is not well understood, the appearance of a second bud serves as an important marker of inappropriate cell cycle progress in spindle assembly checkpoint mutants.

Temporal analysis of the appearance of a new bud in bub2, mad2, and mad2 bub2 strains underscored the difference in the roles of the two genes in cells exposed to nocodazole. The onset of new bud formation in mad2 is 30 min earlier than in bub2 cells (Fig 2A). The lag in new bud formation seen in bub2 mutants (relative to mad2 mutants) is consistent with a partial function of the checkpoint. New bud formation in the bub2 mad2 double mutant preceded that in mad2 by 30 min (Fig 2A).

View larger version (27K): In this window In a new window Download PPT slide   Figure 2. Kinetics of new bud formation and sister chromatid separation in nocodazole. (A) A time course analysis of arrest failure as exhibited by the kinetics of second bud formation was performed. -Factor-synchronized cells of the indicated strains were released into YPD at 25° containing 15 µg/ml nocodazole at t = 0 min. Aliquots were removed at indicated time points, sonicated, and fixed in formaldehyde for DAPI staining and microscopy. New (second) bud formation during nocodazole arrest was detected as the appearance of three or more associated cell bodies despite the persistence of a single DAPI-stained nucleus. Similar data were obtained in three independent experiments, and representative results are shown. Strains were wild type (YRaj127), mad2 (YFS1280), bub2 (YRaj121), and mad2 bub2 (YRaj158). (B) Cell cultures of the indicated strains in logarithmic phase were synchronized in G1 with -Factor (see MATERIALS AND METHODS). To enhance GFP-based visualization of sister chromatids, the expression of the Lac repressor-GFP fusion was induced during the last 30 min of -factor arrest by transferring cells to synthetic complete dextrose medium lacking histidine and containing 10 mM 3-AT (Sigma) according to STRAIGHT et al. 1996 . Synchronized cells were then released into pheromone-free YPD containing 15 µg/ml nocodazole at 25°, and aliquots were removed at indicated time points, sonicated, fixed in 4% paraformaldehyde, and visualized for GFP fluorescence. Similar data were obtained in two independent experiments, and representative results are shown. Strains were wild type (AFS173), mad2-1 (AFS387), bub2 (YFS1214), and mad2-1 bub2 (YFS1215).

Our results differ somewhat from others recently reported (FESQUET et al. 1999 ; LI 1999 ) in which the kinetics of new bud formation in bub2 and mad2 cell populations is inverted or indistinguishable, respectively. Appearance of a new bud as an indicator of a new cell cycle measures an event located far downstream of the control point of interest at the metaphase/anaphase transition. Perhaps differences in laboratory strain genetic background create different rate-limiting steps in mitotic progression. Alternatively, closely spaced time points may be required to reproducibly discern differences between mutants. In agreement with LI 1999 and FESQUET et al. 1999 , the double mutant preceded the single mutants, providing strong evidence for the hypothesis that MAD2 and BUB2 control distinct steps in mitotic progress with additive timing. The lack of epistasis for bud formation contrasts with analysis of sister chromatid separation in the double mutant (below).

A direct comparison of MAD2 and BUB2 function in the control of the metaphase/anaphase transition is provided by the visualization of GFP-marked sister chromatids (STRAIGHT et al. 1996 ). The kinetics of sister chromatid separation in mad2, bub2, and mad2 bub2 double mutants was assayed in synchronous populations. -Factor-arrested cells were released into medium containing nocodazole at 15 µg/µl and aliquots removed every 30 min for analysis. Sister chromatid separation occurred with identical kinetics in mad2 and mad2 bub2 double mutants (Fig 2B), indicating that for this phenotype MAD2 gene function is epistatic to BUB2. Sister chromatid separation occurred 90 min later in bub2 cells (Fig 2B), indicating that BUB2 plays a role in preventing sister chromatid separation at prometaphase arrest. Wild-type cells maintain arrest throughout the duration of this experiment. Our observations confirm and extend results presented by FRASCHINI et al. 1999 .

This analysis is consistent with the working hypothesis that bub2 cells establish but do not maintain a prometaphase arrest in the presence of nocodazole. Interestingly, this assay directly measures an effect on the morphological transition known to be controlled by the MAD2-dependent checkpoint step and indicates that sister chromatid separation accompanies the mitotic exit allowed in cells that lack BUB2.

The sister chromatid separation phenotype of bub2 cells requires the mitotic exit network:
The sister chromatid separation seen in bub2 cells exposed to nocodazole may be due either to activation of the normal CDC20-directed mechanism for Pds1 protein degradation or to an indirect effect of premature mitotic exit. Anaphase normally accompanies a CDC20-mediated degradation of Pds1 protein. However, it has been demonstrated that mitotic exit, when induced in cells lacking CDC20 function, by overexpression of Hct1 protein leads to degradation of Pds1 protein (VISINTIN et al. 1997 ). We therefore tested the effect of bub2 on sister chromatid separation in cells exposed to nocodazole, when the mitotic exit pathway is blocked.

Tem1 protein is a GTPase essential for the activation of the phosphatase CDC14 and the mitotic exit pathway (SHOU et al. 1999 ). A conditional Tem1 protein expression system has been developed (SHOU et al. 1999 ) based on the "N-terminal degron" concept (DOHMEN et al. 1994 ), wherein a destabilizing residue at the N terminus of a protein greatly reduces its half-life. Despite the very short half-life, TEM1 function is provided by overexpression from a galactose responsive promoter. When the promoter is shut off on incubation in glucose, Tem1 protein levels are depleted rapidly and tem1 phenotype is evident within a single cell cycle (SHOU et al. 1999 ). The tem1 degron allele (tem1^c) was introduced into the strains used for analysis of sister chromatid separation in the previous section. Single cycle shutoff was achieved as judged by appearance of a typical tem1 mutant phenotype (cells arrested prior to cytokinesis after completed nuclear division; data not shown). The strains compared in Fig 3 were propagated in galactose-containing, glucose-free medium (TEM1 function present) and brought to a G1 arrest using -factor. Following arrest, cells were released into galactose-free, glucose-containing medium to extinguish Tem1 protein function in cells that have the conditional allele (tem1^c) and aliquots were removed at various time points for analysis. Whereas mad2-1 tem1 cells showed separated sister chromatids, bub2 tem1 cells maintained sister chromatid cohesion for the duration of the experiment (Fig 3). Thus, the mitotic exit pathway functions downstream of TEM1 are responsible for the sister chromatid separation in bub2 mutants but are not required for sister chromatid separation in mad2 mutants.

View larger version (20K): In this window In a new window Download PPT slide   Figure 3. Effect of TEM1 on sister chromatid separation in nocodazole-treated bub2 cells. Asynchronous cultures grown in YPG (2% galactose, 2% raffinose) were arrested in -factor and then released into pheromone-free YPD (2% dextrose) containing 15 µg/ml nocodazole at 25°. Aliquots were removed at the indicated time points, sonicated, and fixed in 4% paraformaldehyde for visualization of GFP fluorescence. Similar sister chromatid separation kinetics were observed in two independent experiments, and representative results are shown. Strains were wild type (AFS173), mad2-1 tem1c (YRaj169), bub2 tem1c (YRaj164), and bub2 (YFS1214).

BUB2 is not required for maintenance of DNA damage-induced arrest:
The arrests induced by spindle damage or DNA damage are mediated by stabilization of Pds1p (YAMAMOTO et al. 1996 ; COHEN-FIX and KOSHLAND 1997 ; GARDNER et al. 1999 ) in response to distinct pathways. For example, HARDWICK et al. 1999 have used the cdc13-1 allele to demonstrate that the DNA damage checkpoint-mediated arrest does not require the spindle assembly checkpoint genes MAD1, MAD2, and MAD3. However, the emerging role of Bub2 protein in preventing premature mitotic exit is one that could be recruited in response to either type of damage. We therefore asked whether BUB2 is required for the maintenance of a DNA damage-induced mitotic arrest.

Cells with the temperature-sensitive cdc13-1 mutation accumulate single-stranded DNA under nonpermissive conditions and exhibit a DNA damage checkpoint-dependent arrest (GARVIK et al. 1995 ; POLOTNIANKA et al. 1998 ; GARDNER et al. 1999 ), mediated through the stabilization of Pds1p (LIM and SURANA 1996 ; COHEN-FIX and KOSHLAND 1997 ). To study the role of BUB2, logarithmically growing cdc13-1 and cdc13-1 bub2 double mutant strains were synchronized in -factor and then released under restrictive (37°) or permissive (25°) conditions for cdc13-1. Under restrictive conditions, arrest was evidenced by accumulation of large budded cells with undivided nuclei (Fig 4A). The cdc13-1 bub2 double mutant behaved just like the cdc13-1 single mutant. Thus, while both the DNA damage- and spindle damage-induced checkpoints act through stabilization of Pds1p to prevent anaphase, only the spindle damage-induced checkpoint breaks down prematurely in the absence of Bub2 protein.

View larger version (28K): In this window In a new window Download PPT slide   Figure 4. Effect of bub2 on cdc13-1-induced arrest. (A) Exposure to 37°. Indicated strains were synchronized in -factor and released into pheromone-free YPD at either 25° or 37°. After 4 hr, samples were sonicated, fixed in formaldehyde, stained with DAPI, and scored for the presence of large budded uninucleate cells as defined in Fig 1. (B) Exposure to both 37° and nocodazole. Each strain was arrested in -factor and released into pheromone-free YPD containing 15 µg/ml nocodazole at 25° or 37°. After 4 hr, samples were sonicated, fixed in formaldehyde, stained with DAPI, and scored for the presence of arrest failure as indicated by new bud formation. New (second) bud formation during nocodazole arrest was detected as the appearance of three or more associated cell bodies despite the persistence of a single DAPI-stained nucleus. Stains were wild type (YRaj127), cdc13-1 (YRaj140), bub2 (YRaj121), and cdc13-1 bub2 (YRaj144). Similar results were obtained in three independent experiments, and representative data are shown.

Interestingly, the DNA damage arrest phenotype is epistatic to the spindle damage arrest phenotype. This was tested in cells subjected to both cdc13-1-induced DNA damage as well as nocodazole-induced spindle damage in the presence and absence of BUB2 function. Comparison of arrest competence was measured by scoring the frequency of cells exhibiting new bud formation. cdc13-1 and cdc13-1 bub2 strains were released from -factor arrest at restrictive (37°) or permissive (25°) temperature into rich medium containing 15 µg/ml nocodazole. The cdc13-1-induced arrest was maintained in the bub2 background in the presence of nocodazole. The bub2 strain showed new bud formation at both 25° and 37° as expected while the cdc13-1 bub2 double mutant showed new bud formation only at 25° and not at 37° (Fig 4B).

Taken together, these data confirm that the arrest mechanisms employed in response to DNA and spindle damage are distinct and that the BUB2-controlled negative regulation of mitotic exit is not essential for the development of a robust DNA damage-mediated arrest.

Continuous function of Bub2 protein is required for maintenance of spindle assembly checkpoint-mediated arrest:
To test whether Bub2 protein is required continuously once the spindle assembly checkpoint-mediated arrest has been established, we constructed a conditional bub2 allele (BUB2^tsd) using a "temperature-sensitive degron" (DOHMEN et al. 1994 ). This was constructed by fusing a temperature-sensitive derivative of the dihydrofolate reductase (DHFR) protein, whose stability is controlled by the N-end rule pathway of protein degradation (DOHMEN et al. 1994 ) to the N terminus of Bub2 protein. This resulting fusion protein was expected to be stable at 25° and to be degraded rapidly at 37°. The construct replaced the wild-type BUB2 gene by integration into the chromosomal BUB2 locus (see MATERIALS AND METHODS) such that expression was driven by the BUB2 promoter. The Bub2^tsd fusion protein provides BUB2 function at 25° as evidenced by sustained arrest in the presence of nocodazole (Fig 5, left). However at 37° there is a clear loss of Bub2 protein function as cells fail to arrest in the presence of nocodazole (Fig 5, center). To test whether Bub2 protein function is required continuously once the arrest is established, cells that were exposed to nocodazole at 25° for 4 hr (achieving 85% population arrest) were then shifted to 37°. These exhibited a breakdown of the arrest as evidenced by the appearance of new (second) bud formation (Fig 5, right). Thus the Bub2 protein is required continuously for maintenance of the spindle assembly checkpoint-mediated block even when its function is manipulated well after establishment of prometaphase arrest.

View larger version (26K): In this window In a new window Download PPT slide   Figure 5. Conditional expression of BUB2. Log phase populations grown in YPD at 25° were transferred to YPD containing 15 µg/ml nocodazole at the indicated temperatures and cultured for the indicated time periods. Samples were sonicated, fixed in formaldehyde, stained with DAPI, and scored for the presence of arrest failure as indicated by new (second) bud formation. Strains were BUB2 (YRaj127), bub2tsd (YRaj132), and bub2 (YRaj121). Similar results were obtained in two independent experiments, and representative data are shown.

Bub2 protein exhibits a physical interaction with the septin protein Cdc3:
To better understand the role of BUB2 through its interactions with other genes, a two-hybrid screen (BAI and ELLEDGE 1997 ) using a yeast cDNA library has been initiated. One interacting plasmid obtained (Fig 6) has been analyzed in depth. The library insert encodes an in-frame GAL4 fusion containing the C-terminal 206 amino acids of the CDC3 gene product. CDC3 encodes a member of the septin family of proteins that play important roles in cytokinesis and cell morphology (FIELD and KELLOG 1999 ). An in vitro test was performed to further explore the two-hybrid association between Cdc3 and Bub2 proteins. A bacterially expressed GST-CDC3 fusion protein was used in an affinity trap for ³⁵S-methionine-labeled Bub2 protein obtained after coupled in vitro transcription/translation. Bub2 protein was specifically retained on beads carrying GST-CDC3 protein but not on beads carrying GST protein alone (Fig 6B). The in vitro affinity between Bub2 and Cdc3 proteins indicates that the yeast two-hybrid interaction may represent a direct contact between these proteins in vivo.

View larger version (58K): In this window In a new window Download PPT slide   Figure 6. Bub2 and Cdc3 proteins interact in the yeast two-hybrid system and in vitro. (A) Yeast reporter strain J69-2A was cotrans- formed with GAL4-binding domain (BD) or activation domain (AD) fusions as indicated (top). Transformants were patched onto synthetic complete medium lacking tryptophan and leucine (middle) or synthetic complete medium lacking tryptophan, leucine, adenine, and histidine that also contained 20 mM 3-AT (bottom). GalAD-SVT and GalBD-p53 contain a GAL4-activation domain/SV40 large T antigen fusion gene and GAL4/DNA-binding domain/human p53 fusion gene, respectively. These well-characterized constructs (CLONTECH) serve as negative controls not expected to exhibit interaction with CDC3, BUB2, or CDC16 fusions. (B) Interaction between Bub2 and Cdc3 proteins in vitro. (Top) An autoradiogram of 35S-labeled in vitro-translated Bub2 protein and its binding properties. Lane 1, 35S-labeled in vitro transcription/translation product without supplied plasmid template (negative control); lane 2, 35S-labeled in vitro transcription/translation product from BUB2 template (35S-Bub2 protein); lane 3, eluate from glutathione beads incubated with 35S-Bub2 protein; lane 4, eluate from GST-bound glutathione beads incubated with 35S-Bub2 protein; lane 5, eluate from GST-CDC3-bound beads incubated with 35S-Bub2 protein. (Bottom) Relevant lanes (4 and 5) from a duplicate but silver-stained gel. GST and GST-CDC3 proteins are as indicated. Molecular sizes were confirmed using prestained marker from Bio-Rad (not shown).

cdc3-1 cells appear to have an intact spindle assembly checkpoint and show no synthetic lethality in bub2 or BUB2 overexpression backgrounds (not shown). BUB2 has a well-conserved homologue in the distantly related fungus S. pombe (FANKHAUSER et al. 1993 ). S. cerevisiae Bub2 and S. pombe cdc16⁺ proteins exhibit 39% identity and 58% similarity (data not shown) in BLASTP alignment using default parameters (TATUSOVA and MADDEN 1999 ). Cdc16⁺ stabilizes p34^CDC2 kinase activity and has been shown to negatively regulate the timing of cytokinesis in fission yeast, as well as placement and number of the septa formed on mitotic exit (MINET et al. 1979 ; FANKHAUSER et al. 1993 ). Furthermore, S. cerevisiae BUB2 complements the temperature-sensitive allele cdc16-116 in S. pombe, although it is unable to complement a cdc16^- null mutation (FANKHAUSER et al. 1993 ). We therefore asked whether the Cdc16 protein of S. pombe interacts with Cdc3 protein of S. cerevisiae in a two-hybrid assay and found that it does (Fig 6A). The conservation of this interaction, together with known late mitotic functions for each of these proteins, suggests that the association indicated may have significant biological consequences.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The spindle assembly checkpoint controls the timing of the metaphase-to-anaphase transition, inhibiting sister chromatid separation until all chromatid pairs have achieved bipolar spindle attachment. However, mitotic progression is also controlled at a later point prior to the execution of events leading to exit and cytokinesis (JASPERSON et al. 1998 ). Emerging evidence suggests the existence of an aspect of the spindle assembly checkpoint that regulates a late mitotic control step and depends on the BUB2 protein (ALEXANDRU et al. 1999 ; FESQUET et al. 1999 ; FRASCHINI et al. 1999 ; LI 1999 ; this work). These two distinct aspects of the spindle assembly checkpoint are controlled by MAD2- and BUB2-dependent pathways and are both required for full prometaphase arrest in response to spindle or kinetochore damage.

A key issue in the description of these two activities of the spindle assembly checkpoint is whether or not they represent a branched response following a common initiation signal. The MAD2-dependent pathway is clearly induced by an unattached or defective kinetochore (WANG and BURKE 1995 ; PANGILINAN and SPENCER 1996 ). It has been generally suggested that BUB2 may not have a role to play when the spindle assembly checkpoint responds to defective kinetochores but rather responds to a distinct spindle damage structure (WANG and BURKE 1995 ; LI 1999 ). However, the presence of a single chromosome containing a centromere DNA mutation (one that generates a mitotic delay in checkpoint proficient cells) induces the occasional production of cells with extra buds in cycling bub2 strains (PANGILINAN and SPENCER 1996 ). This phenotype, albeit occurring at a low frequency, represents inappropriate mitotic exit despite the presence of a functional spindle. In agreement with this result, a synchronous bub2 ctf13-30 double-mutant population exhibits a transient accumulation of delayed cells indicating that kinetochore damage-induced arrest is established but is not maintained. Thus a kinetochore damage-induced checkpoint arrest requires Bub2 protein for its full function in two independent assays. We suggest therefore that the MAD2- and BUB2-dependent pathways are both triggered by a signal from an unattached kinetochore.

Analysis of mad2 bub2 double mutants indicates that these two genes act in additive functions within mitosis, in general agreement with previously published work (ALEXANDRU et al. 1999 ; FESQUET et al. 1999 ; FRASCHINI et al. 1999 ; LI et al. 1999). This conclusion is established by the observation of additive effects of mutant alleles on a postmitotic phenotype and the appearance of new (second) bud formation in a nocodazole block. Interestingly, the timing of inappropriate sister chromatid separation exhibits epistasis in the mad2 bub2 double mutant, with the effect of the mad2 mutant rendering the bub2 defect irrelevant. This result is consistent with the hypothesis that the MAD2-dependent prevention of sister chromatid separation in the presence of spindle damage requires a later BUB2-dependent step for maintenance.

The sister chromatid separation seen in nocodazole-exposed bub2 cells could be explained either as an indirect consequence of onset of mitotic exit and late mitotic APC activity or as a direct consequence of in-ability to hold off anaphase. This latter possibility would be supported by the observation that under normal conditions the degradation of Pds1p is essential for subsequent degradation of Clb2p and for mitotic exit (COHEN-FIX and KOSHLAND 1999 ; TINKER-KULBERG and MORGAN 1999 ). On the other hand, the ability of late mitotic APC activity to nonspecifically degrade Pds1 protein, when the latter has been stabilized through lack of Cdc20 protein function, has been demonstrated by VISINTIN et al. 1997 . Under conditions of spindle damage in mad2 bub2 mutants, degradation kinetics of Pds1 and Clb2 proteins appear indistinguishable (FRASCHINI et al. 1999 ). Moreover, when the kinetics of sister chromatid separation are followed, there is no difference between mad2 and mad2 bub2 mutants, suggesting that BUB2 does not influence a rate-limiting step in sister chromatid separation. Under conditions of spindle damage, we asked whether the sister chromatid separation in bub2 mutants is largely a consequence of onset of the mitotic exit pathway. Our results (Fig 3) that bub2 tem1 cells, which do not exit mitosis, do maintain sister chromatid cohesion in nocodazole is direct evidence that inappropriate sister chromatid separation seen in bub2 cells is due to the onset of the exit pathway. This is consistent with the observation that overexpression of Tem1 protein mimics a bub2 phenotype in nocodazole-treated cells (ALEXANDRU et al. 1999 ). Taken together these results demonstrate a situation wherein the inhibitory effect of Pds1 protein on exit (COHEN-FIX and KOSHLAND 1999 ; TINKER-KULBERG and MORGAN 1999 ) is apparently prevented in cells lacking BUB2 function, raising the intriguing possibility that Pds1 protein acts through a BUB2-dependent pathway to control exit. This in turn would be consistent with the finding by TAVORMINA and BURKE 1998 that the mitotic arrest in cdc20 mutants requires BUB2 function.

In theory, the importance of a regulatory block to mitotic exit when anaphase is delayed would be encountered in other preanaphase blocks. The initial characterization of bub2 mutants indicated that BUB2 was not required for maintenance of -factor or hydroxyurea-induced arrests in G1 or S phase, respectively (HOYT et al. 1991 ). Interestingly, recent analysis of the roles of components of the DNA damage checkpoint has revealed partially separable controls at metaphase and mitotic exit (SANCHEZ et al. 1999 ). We determined whether BUB2 played a role in maintenance of the DNA damage checkpoint, which, like the spindle assembly checkpoint, acts predominantly through the stabilization of Pds1 protein. The cdc13-1 allele induces the DNA damage checkpoint, at the restrictive temperature of 37° (LIM and SURANA 1996 ; COHEN-FIX and KOSHLAND 1997 ). The results demonstrate that a cdc13-1-induced arrest is maintained in cells lacking Bub2 protein and indicate that the role of the BUB2 pathway in preventing premature exit when anaphase is blocked is not a global one.

While the mechanism by which BUB2 maintains a spindle damage-induced arrest is not apparent, the bub2^tsd allele indicates the presence of an active and continuous role for the Bub2 protein in the maintenance of the mitotic arrest induced by nocodazole. Recent work on the roles of DBF2 kinase in late mitosis (FESQUET et al. 1999 ) and on CLB2 and CLB3 stability throughout mitosis (ALEXANDRU et al. 1999 ; FRASCHINI et al. 1999 ) suggest that these are regulated directly or indirectly by BUB2 when spindle damage is present. An analogy can be drawn between BUB2 and the function of its S. pombe homologue cdc16⁺ (FANKHAUSER et al. 1993 ). In S. pombe, cdc16 and byr4 proteins form a two-component GTPase-activating protein that maintains spg1 protein in its GDP-bound form until it is appropriate for the GTP-bound form, which facilitates mitotic exit, to prevail (FURGE et al. 1998 ). We found that Bub2 protein of S. cerevisiae interacts with byr4 protein of S. pombe in the yeast two-hybrid system (data not shown), adding another piece of evidence that BUB2 and cdc16⁺ are functional counterparts in S. cerevisiae and S. pombe, respectively. LI 1999 has shown that BFA1, the S. cerevisiae homologue of S. pombe byr4⁺, is required for the spindle assembly checkpoint. The TEM1 gene of S. cerevisiae is a homologue of spg1⁺ of S. pombe, and it has been demonstrated that Tem1 protein-dependent signaling facilitates mitotic exit by activating the Cdc14 phosphatase and, consequently, Clb2p degradation (SHOU et al. 1999 ). These results are consistent with the observation that the continuous presence of Bub2 protein and its biochemical activity is required for its role in the maintenance of the spindle assembly checkpoint-mediated arrest.

In a yeast two-hybrid screen for proteins that interact with BUB2 we isolated the Cdc3 protein, a septin essential for cytokinesis (KIM et al. 1991 ). Bub2 and Cdc3 proteins also interact in vitro, suggesting that they are able to make direct contact with each other in the cell. Furthermore, cdc16⁺, the S. pombe homologue of BUB2, interacts with Cdc3 protein of S. cerevisiae in a two-hybrid test.

Although the biological steps affected by a candidate BUB2/CDC3 interaction have not yet been defined, a rapidly growing body of circumstantial evidence suggests several interesting views. While Cdc3 protein localizes predominantly to the mother-bud neck (LONGTINE et al. 1999 ), Bub2 protein localizes to the spindle pole bodies (FRASCHINI et al. 1999 ; LI 1999 ). LI 1999 has proposed that the localization of Bub2 protein at the spindle pole body may suggest a role in monitoring passage of the spindle pole through the mother-bud neck. If so, such a role would be facilitated by an interaction between a neck-localized Cdc3 protein and Bub2 protein. It is interesting to note that in S. pombe, sid2⁺ encodes a spindle pole body-associated kinase whose septin-dependent activity peaks during onset of cytokinesis (SPARKS et al. 1999 ). DBF2 is the S. cerevisiae homologue of sid2⁺, and its kinase activity is influenced by BUB2 under conditions of spindle damage (FESQUET et al. 1999 ; SPARKS et al. 1999 ). This provides intriguing indirect evidence for a biological significance of the observed yeast two-hybrid interaction between BUB2 and the septin CDC3.

In summary, we demonstrate that BUB2 acts in a pathway that is required for maintenance of the spindle assembly checkpoint-mediated arrest and that this pathway is required even when kinetochore damage is the inducer of arrest. Furthermore, the continuous presence of the Bub2 protein is required for this role and such a requirement is consistent with the proposed role for BUB2 in regulating GTPase activity of Tem1 protein. The premature breakdown of metaphase in bub2 cells is a consequence of activation of the TEM1-directed mitotic exit pathway and does not precede the TEM1-directed step. We have noted that the BUB2-controlled step is irrelevant in cells arrested at metaphase by the DNA damage checkpoint, suggesting that Bub2 protein function is not a universal requirement for preventing mitotic exit until anaphase has occurred. Finally, we have uncovered an intriguing physical association of Bub2 protein with Cdc3, a septin that functions in cytokinesis and cell morphogenesis. Experimental analysis of the relationships among essential mitotic events and their control will likely continue to reveal layers in a network of coordinate regulation governing the many ordered processes that comprise the division of one cell into two.

	ACKNOWLEDGMENTS

C. Albright, R. Deshaies, J. Dohmen, S. Elledge, P. Hieter, M. Longtine, V. Lundblad, M. Mayer, A. Murray, and A. Straight kindly provided plasmids and strains. We thank A. Hoyt, O. Cohen-Fix, and an anonymous reviewer for helpful comments on the manuscript. This work was supported by National Institutes of Health grant GM-50842 to F.S.

Manuscript received March 8, 2000; Accepted for publication May 30, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALEXANDRU, G., W. ZACHARIE, A. SCHLEIFFER, and K. NASMYTH, 1999  Sister chromatid separation and chromosome re-duplication are regulated by different mechanisms in response to spindle damage. EMBO J. 18:2707-2721[Medline].

BAI, C. and S. J. ELLEDGE, 1997  Gene identification using the yeast two-hybrid system. Methods Enzymol. 283:141-156[Medline].

COHEN-FIX, O. and D. KOSHLAND, 1997  The anaphase inhibitor of Saccharomyces cerevisiae Pds1p is a target of the DNA damage checkpoint pathway. Proc. Natl. Acad. Sci. USA 94:14361-14366[Abstract/Free Full Text].

COHEN-FIX, O. and D. KOSHLAND, 1999  Pds1p of budding yeast has dual roles: inhibition of anaphase initiation and regulation of mitotic exit. Genes Dev. 13:1950-1959[Abstract/Free Full Text].

COHEN-FIX, O., J. M. PETERS, M. W. KIRSCHNER, and D. KOSHLAND, 1996  Anaphase initiation in Saccharomyces cerevisiae is controlled by the APC-dependent degradation of the anaphase inhibitor Pds1p. Genes Dev. 10:3081-3093[Abstract/Free Full Text].

DOHENY, K. F., P. K. SORGER, A. A. HYMAN, S. TUGENDREICH, and F. SPENCER et al., 1993  Identification of essential components of the S. cerevisiae kinetochore. Cell 73:761-774[Medline].

DOHMEN, R. J., P. WU, and A. VARSHAVSKY, 1994  Heat-inducible degron: a method for constructing temperature sensitive mutants. Science 263:1273-1276[Abstract/Free Full Text].

ELLEDGE, S. J., 1998  Mitotic arrest: Mad2 prevents sleepy from waking up the APC. Science 279:999-1000. [comment][Free Full Text].

FANG, G., H. YU, and M. W. KIRSCHNER, 1998  The checkpoint protein MAD2 and the mitotic regulator CDC20 form a ternary complex with the anaphase-promoting complex to control anaphase initiation. Genes Dev. 12:1871-1883[Abstract/Free Full Text].

FANKHAUSER, C., J. MARKS, A. REYMOND, and V. SIMANIS, 1993  The S. pombe cdc16 gene is required both for maintenance of p34cdc2 kinase activity and regulation of septum formation: A link between mitosis and cytokinesis? EMBO J. 12:2697-2704[Medline].

FESQUET, D., P. J. FITZPATRICK, A. L. JOHNSON, K. M. KRAMER, and J. H. TOYN et al., 1999  A Bub2p-dependant spindle checkpoint pathway regulates the Dbf2p kinase in budding yeast. EMBO J. 18:2424-2434[Medline].

FIELD, C. M. and D. KELLOG, 1999  Septins: Cytoskeletal polymers or signalling GTPases? Trends Cell Biol. 10:387-394.

FRASCHINI, R., E. FORMENTI, G. LUCCHINI, and S. PIATTI, 1999  Budding yeast Bub2 is localized at spindle pole bodies and activates the mitotic checkpoint via a different pathway from MAD2. J. Cell Biol. 145:979-991[Abstract/Free Full Text].

FURGE, K. A., K. WONG, J. ARMSTRONG, M. BALASUBRAMANIAN, and C. F. ALBRIGHT, 1998  Byr4 and Cdc16 form a two component GTPase activating protein for Spg1 GTPase that controls septation in fission yeast. Curr. Biol. 8:947-954[Medline].

GARDNER, R., C. W. PUTNAM, and T. W. WEINERT, 1999  RAD53, DUN1 and PDS1 define two parallel G2/M checkpoint pathways in budding yeast. EMBO J. 18:3173-3185[Medline].

GARVIK, B., M. CARSON, and L. HARTWELL, 1995  Single-stranded DNA arising at telomeres in cdc13 mutants may constitute a specific signal for the RAD9 checkpoint. Mol. Cell. Biol. 15:6128-6138[Abstract].

GUTHRIE, C., and G. R. FINK (Editors), 1991. Methods in Enzymology: Guide to Yeast Genetics and Molecular Biology, Vol. 194. Academic Press, San Diego, CA.

HARDWICK, K. G., R. LI, C. MISTROT, R. H. CHEN, and P. DANN et al., 1999  Lesions in many different spindle components activate the spindle checkpoint in the budding yeast Saccharomyces cerevisiae.. Genetics 152:509-518[Abstract/Free Full Text].

HOYT, M. A., L. TOTIS, and B. T. ROBERTS, 1991  S. cerevisiae genes required for cell cycle arrest in response to loss of microtubule function. Cell 66:507-517[Medline].

HWANG, L. H., L. F. LAU, D. L. SMITH, C. A. MISTROT, and K. G. HARDWICK et al., 1998  Budding yeast Cdc20: a target of the spindle checkpoint. Science 279:1041-1044. [see comments][Abstract/Free Full Text].

JAMES, P., J. HALLADAY, and E. A. CRAIG, 1996  Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144:1425-1436[Abstract].

JASPERSON, S. L., J. F. CHARLES, R. L. TINKER-KULBERG, and D. O. MORGAN, 1998  A late mitotic regulatory network controlling cyclin destruction in Saccharomyces cerevisiae.. Mol. Biol. Cell 9:2803-2817[Abstract/Free Full Text].

KALLIO, M., J. WEINSTEIN, J. R. DAUM, D. J. BURKE, and G. J. GORBSKY, 1998  Mammalian p55CDC mediates association of the spindle checkpoint protein Mad2 with the cyclosome/anaphase-promoting complex, and is involved in regulating anaphase onset and late mitotic events. J. Cell Biol. 141:1393-1406[Abstract/Free Full Text].

KIM, H. B., B. K. HAARER, and J. R. PRINGLE, 1991  Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. J. Cell Biol. 112:535-544[Abstract/Free Full Text].

KIM, S. H., D. P. LIN, S. MATSUMOTO, A. KITAZONO, and T. MATSUMOTO, 1998  Fission yeast Slp1: an effector of the Mad2-dependent spindle checkpoint. Science 279:1045-1047. [see comments][Abstract/Free Full Text].

LI, R., 1999  Bifurcation of the mitotic checkpoint pathway in budding yeast. Proc. Natl. Acad. Sci. USA 96:4989-4994[Abstract/Free Full Text].

LI, R. and A. W. MURRAY, 1991  Feedback control of mitosis in budding yeast. Cell 66:519-531. (erratum: Cell 79: following 388).[Medline].

LIM, H. H. and U. SURANA, 1996  Cdc20, a beta-transducin homologue, links RAD9-mediated G2/M checkpoint control to mitosis in Saccharomyces cerevisiae.. Mol. Gen. Genet. 253:138-148[Medline].

LIM, H. H., P. Y. GOH, and U. SURANA, 1998  Cdc20 is essential for the cyclosome-mediated proteolysis of both Pds1 and Clb2 during M phase in budding yeast. Curr. Biol. 8:231-234[Medline].

LONGTINE, M., H. FARES, and J. R. PRINGLE, 1999  Role of the yeast Gin4p protein kinase in septin assembly and the relationship between septin assembly and septin function. J. Cell Biol. 143:719-736[Abstract/Free Full Text].

MINET, M., P. NURSE, P. THURIAUX, and J. M. MITCHISON, 1979  Uncontrolled septation in a cell division cycle mutant of the fission yeast Schizosaccharomyces pombe.. J. Bacteriol. 137:440-446[Abstract/Free Full Text].

MORGAN, D. O., 1999  Regulation of the APC and the exit from mitosis. Nat. Cell Bi