IBD2 Encodes a Novel Component of the Bub2p-Dependent Spindle Checkpoint in the Budding Yeast Saccharomyces cerevisiae

Corresponding author: Kiwon Song, College of Science, Yonsei University, Seoul 120-749, Korea., bc5012{at}yonsei.ac.kr (E-mail)

Communicating editor: M. D. ROSE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

During mitosis, genomic integrity is maintained by the proper coordination of mitotic events through the spindle checkpoint. The bifurcated spindle checkpoint blocks cell cycle progression at metaphase by monitoring unattached kinetochores and inhibits mitotic exit in response to the incorrect orientation of the mitotic spindle. Bfa1p is a spindle checkpoint regulator of budding yeast in the Bub2p checkpoint pathway for proper mitotic exit. We have isolated a novel Bfa1p interacting protein named Ibd2p in the budding yeast Saccharomyces cerevisiae. We found that IBD2 (Inhibition of Bud Division 2) is not an essential gene but its deletion mutant proceeded through the cell cycle in the presence of microtubule-destabilizing drugs, thereby inducing a sharp decrease in viability. In addition, overexpression of Mps1p caused partial mitotic arrest in ibd2 as well as in bub2, suggesting that IBD2 encodes a novel component of the spindle checkpoint downstream of MPS1. Overexpression of Ibd2p induced mitotic arrest with increased levels of Clb2p in wild type and mad2, but not in deletion mutants of BUB2 and BFA1. Pds1p was also stabilized by the overexpression of Ibd2p in wild-type cells. The mitotic arrest defects observed in ibd2 in the presence of nocodazole were restored by additional copies of BUB2, BFA1, and CDC5, whereas an extra copy of IBD2 could not rescue the mitotic arrest defects of bub2 and bfa1. The mitotic arrest defects of ibd2 were not recovered by MAD2, or vice versa. Analysis of the double mutant combinations ibd2mad2, ibd2bub2, and ibd2dyn1 showed that IBD2 belongs to the BUB2 epistasis group. Taken together, these data demonstrate that IBD2 encodes a novel component of the BUB2-dependent spindle checkpoint pathway that functions upstream of BUB2 and BFA1.

CELLS ensure genomic integrity in each division through checkpoints that control cell cycle progressions by monitoring the successful completion of preceding processes. One mitotic checkpoint, known as the spindle checkpoint, monitors the assembly and orientation of the mitotic spindle for the equal segregation of replicated chromosomes during mitosis. By sensing defects in the microtubule cytoskeleton, the spindle checkpoint arrests cells at metaphase and prevents exit from mitosis through the regulation of Cdk activity. The spindle checkpoint is vital for maintaining genomic stability during cell division, and deficiency of this checkpoint can lead to genomic instability associated with cancer (CAHILL et al. 1998 ). Genetic studies in budding yeast have identified several components of the spindle checkpoint by isolating mutants that could no longer sense spindle depolymerization and died rapidly in the presence of microtubule-depolymerizing drugs such as nocodazole or benomyl (HOYT et al. 1991 ; LI and MURRAY 1991 ). These components include MAD1, MAD2, MAD3, BUB1, BUB2, BUB3, and MPS1. Overexpression of MPS1 is sufficient to activate the spindle checkpoint in the absence of any spindle damage in wild-type cells but not in other spindle checkpoint mutants, placing MPS1 upstream of other spindle checkpoint genes (HARDWICK et al. 1996 ). The spindle checkpoint bifurcates into two separate signaling pathways, the MAD/BUB spindle assembly checkpoint for metaphase arrest and the BUB2-dependent pathway to block mitotic exit and cytokinesis (ALEXANDRU et al. 1999 ; LI 1999 ).

Bub1p, Bub3p, Mad1p, Mad2p, and Mad3p form a conserved spindle assembly checkpoint that monitors the attachment of bipolar microtubules to the kinetochores of sister chromatids and delays the metaphase-to-anaphase transition in response to spindle assembly defects by inhibiting Cdc20p (FANG et al. 1998 ; HWANG et al. 1998 ). Cdc20p is a key component of the anaphase-promoting complex (APC) that targets the B-type cyclin Clb5p and the anaphase inhibitor Pds1p for degradation by ubiquitin-mediated proteolysis to trigger chromosome separation for anaphase onset (ZACHARIAE et al. 1998 ; SHIRAYAMA et al. 1999 ). Homologs of yeast Mad1p, Mad2p, Bub1p, and Bub3p have been identified and localized to unattached kinetochores in higher eukaryotes, including mammalian cells (CHEN et al. 1996 , CHEN et al. 1998 ; TAYLOR and MCLEON 1997 ; TAYLOR et al. 1998 ; MARTINEZ-EXPOSITO et al. 1999 ), although the mechanism that senses spindle defects is not yet understood. Vertebrate homologs of Cdc20p and Pds1p have also been identified, suggesting that spindle assembly checkpoint mechanisms are also conserved in eukaryotes (ZOU et al. 1999 ).

Bub2p is present in the spindle pole body, the microtubule-organizing center in yeast, and forms a separate branch of the spindle checkpoint pathway that controls mitotic exit and the timing of cytokinesis (FESQUET et al. 1999 ; FRASCHINI et al. 1999 ; LI 1999 ). In the fission yeast Schizosaccharomyces pombe, the Bub2p homolog Cdc16p interacts with Byr4p to form a two-component GTPase activating protein (GAP) that stimulates the GTPase activity of Spg1p to coordinate the onset of cytokinesis (FURGE et al. 1998 ; JWA and SONG 1998 ). BFA1, a putative byr4 homolog in budding yeast that has also been known as IBD1 and BYR4, has been reported to function in the BUB2-dependent spindle checkpoint for mitotic exit and cytokinesis (ALEXANDRU et al. 1999 ; LEE et al. 1999 ; LI 1999 ). The simplest model in budding yeast would thus be that Bub2p and Bfa1p together form a GAP activity that inhibits Tem1p to prevent cells from exiting mitosis, when the checkpoint is activated. Recently the Bub2p checkpoint, including Bub2p, Bfa1p, and Tem1p, has been shown to monitor spindle orientation during anaphase and to inhibit mitotic exit when nuclear migration into the bud is delayed (BARDIN et al. 2000 ; BLOECHER et al. 2000 ; PEREIRA et al. 2000 ). Mitotic exit is ultimately repressed by inhibiting Cdh1p/Hct1p, which targets degradation of the mitotic cyclin Clb2p through the APC (SCHWAB et al. 1997 ; VISINTIN et al. 1997 ; ZACHARIAE et al. 1998 ).

In this study, we discuss the function of IBD2, which we isolated on the basis of its interaction with BFA1 in budding yeast, and present evidence that IBD2 plays a role in the spindle checkpoint pathway and belongs to the BUB2 epistasis group.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains, cultures, cell cycle arrest, and release:
The Saccharomyces cerevisiae strains used in this study are listed in Table 1. Yeast cells were grown in YPD medium (1% yeast extract, 2% bactopeptone, and 2% glucose) or in synthetic complete (SC) dropout media prepared with yeast nitrogen base (YNB) and necessary supplements. To induce expressions from the GAL10-1 or GAL1 promoter, cells grown to midlog phase in 2% glucose were transferred to SC dropout media with raffinose for 4 hr, 2% galactose was added to this culture, and then cells were incubated in 2% raffinose/galactose for 12–14 hr at 29°. Cell cycles of log phase cells (5 x 10⁶ cells/ml) were arrested with a final concentration of 6 µM -factor (Sigma, St. Louis) or 0.1 M hydroxyurea (HU; Sigma) for 3 hr and released from the cell cycle arrest by washing cultures in fresh medium several times.

Assays for sensitivity to microtubule-destabilizing drugs:
For nocodazole treatment, cells were grown to midlog phase (5 x 10⁶ cells/ml) in YPD at 29° and nocodazole was added to a final concentration of 15 µg/ml from a 10 mg/ml stock in DMSO. Cells were incubated at 25° in the presence of nocodazole, if not specified. To assay viability, the same number of cells treated with 15 µg/ml of nocodazole for 0, 3, and 6 hr, respectively, was plated on YPD without any nocodazole, and the percentage viability at each time point was calculated by dividing the number of colonies formed at 0, 3, or 6 hr by the number at time 0 (STRAIGHT and MURRAY 1997 ). For plate assays, equivalent numbers of cells grown in YPD to an OD₆₀₀ of 1.0 were serially diluted, spotted onto both YPD (-) benomyl and YPD (+) benomyl (10 µg/ml) plates, and incubated at 25° for 3–4 days. To assay cell cycle progression, cells with an extra new bud(s) were counted after cells arrested with -factor were released in the presence of nocodazole (15 µg/ml) for 3 hr. Before counting, cells from each time point were briefly fixed in 70% ethanol and sonicated lightly.

Yeast two-hybrid screen:
Yeast two-hybrid screening was performed essentially by following GYURIS et al. 1993 . The full-length BFA1 subcloned into pLex202 + PL was used as the bait for the screen. For the interactor hunt, EGY48 with BFA1/pLex202 + PL was transformed with a cDNA library in pJG4-5, where the expression of each cloned cDNA was under the control of a GAL1-inducible promoter. A total of 50,000 colonies were screened. True positive interaction was verified by leucine assay in EGY48, where chromosomal LEU2 is replaced by LexAop6-LEU2. True positives grew on SC-leu dropout media only in the presence of galactose/raffinose and not glucose, since the cDNA was expressed in galactose/raffinose only. To confirm the interaction between Ibd2p and Bfa1p, the full-length IBD2 open reading frame (ORF) was subcloned into pJG4-5 and a two-hybrid assay was performed with BFA1.

DNA manipulations and strain constructions:
The full BFA1 ORF was amplified by PCR from genomic DNA with the 5' oligonucleotide 5'-TTGGATCCCTATGTCAATTAG-3' and the 3' oligonucleotide 5'-CAATGGATCCGGCTAAAGGGCTAATCTTTTG-3' and subcloned into the BamHI site of pLex202 + PL to be used as a bait for two-hybrid screening. To express hemagglutinin (HA)-tagged BFA1 in a CEN plasmid under its endogenous promoter, the full promoter and ORF of BFA1 were amplified with the 5' oligonucleotide 5'-TGCTCTAGACGGAGCAAGAGATAGTCTGAG-3' containing an XbaI site and the 3' oligonucleotide 5'-ACAGGATCCATCTTTTGTCGAATTGATTACCATGTT-3' containing a BamHI site and were subcloned into pTS903CL (a gift from Dr. Toh-e, Tokyo University). The full ORF of IBD2 was amplified by PCR from genomic DNA with the 5' oligonucleotide 5'-GGGTTTATATGAATC ATAGACTAATATAG-3' and the 3' oligonucleotide 5'-CCTGTTACTATTTAGTCATAATACCGC-3' and was subcloned into the SpeI and SacI sites of pRS316 and pRS315 via T-vector (Promega, Madison, WI). The full ORF of IBD2 was also subcloned into pJG4-5 by amplifying with the 5' oligonucleotide 5'-CCGGAATTCAAGAAAATGACACCTACAAACC-3' containing an EcoRI site and the 3' oligonucleotide 5'-TTCCTCGAGTAATAATGTTGACTATCTATTC-3' containing an XhoI site. For the overexpression of IBD2 under the GAL10-1 promoter, the full ORF of IBD2 amplified with the 5' oligonucleotide 5'-ATCGAATTCAAAATGACACCTACAAACCAATC-3' containing an EcoRI site and the 3' oligonucleotide 5'-TAAGGATCCCTATCTATTCCTTTTCC-3' containing a BamHI site was subcloned into pMW20.

To make a YSK1 strain in which the IBD2 ORF tagged with Myc and His at its 3' end was integrated into the chromosome of an IBD2 knockout strain (CDLY011), the full-length ORF of IBD2 amplified with the 5' oligonucleotide 5'-TATCTGCAGCATAGACTAATATAGATAG-3' and the 3' oligonucleotide 5'-ACTCTGCAGTTCATCTCTTGGTGGATTC-3' was subcloned into the PstI site of pTS905IT (a gift from Dr. Toh-e, Tokyo University). IBD2-Myc-His/pTS905IT was linearized with EcoRI, transformed into CDLY012 (ibd2), and selected on SC-trp plates. YSK2, YSK3 (both ibd2), YSK5 (ibd2bub2), YSK7 (ibd2mad2), YSK9 (ibd2dyn1), and YSK10 (ibd2 bfa1) were constructed by PCR-based gene deletion (WACH et al. 1994 ; LONGTINE et al. 1998 ) using 5' and 3' primers that contained 50 nucleotides spanning the start or stop codons of the targeted genes, respectively. For YSK2, URA3 from pRS316 was amplified with the 5' oligonucleotide 5'-CAAGAAAATGACACCTACAAACCAATCTAGTGGAACGACTAATGCATCTGTGGAGGTACGTACTGAGAGTGCACCACGC-3' and the 3' oligonucleotide 5'-GACTATCTATTCCTTTTCCTACTATCTTGTTCATCTCTTGGTGGATTCGGCATACGCTCCTTACGCATCTGTGCGG-3'. For YSK3, HIS3 from pRS303 was amplified with 5'-GAAGAAAATGACACCTACAAACCAATCTAGTGGAACGACTAATGCATCTGAGCTTGGTGAGCGCTAGGAGTCAC-3' and 5'-GACTATCTATTCCTTTTCCTACTATCTTGTTCATCTCTTGGTGGATTCGGCTCGTTCAGAATGACACGTATAG-3'. PCR products were transformed into W303, selected using URA3 and HIS3 markers, and verified by genomic PCR and Southern blotting. The same PCR product used to construct YSK3 was also transformed into MAY2052 (bub2), RHC 15.1 (mad2), KT1374 (dyn1), and YSK8 (bfa1) to construct YSK5 (ibd2bub2), YSK7 (ibd2mad2), YSK9 (ibd2dyn1), and YSK10 (ibd2bfa1), respectively. YSK4 (ibd2bub2) and YSK6 (ibd2mad2) were made essentially by following SENA et al. 1973 . Diploids were constructed by mating CDLY011 (ibd2) and MAY2052 (bub2) or CDLY012 (ibd2) and RHC 15.1 (mad2) of opposite mating types on the surface of agar plates and were sporulated on sporulation plates at 25° for 6–7 days. Tetrads were resuspended in 1 ml ddH₂O, incubated with 0.25 mg zymolase 20T (Seikagaku, Rockville, MD) overnight at 30° with gentle shaking, sonicated briefly, and spread on YPD. The genotype of each spore colony was determined by replica plating on selective media.

Microscopic techniques:
To observe nuclei and cell shape, cells were fixed in 70% ethanol for 5 sec, washed twice with PBS, briefly sonicated, and mounted with 1 µg/ml 4',6-diamidino-2-phenylindole (DAPI). For immunostaining, cells were fixed by incubating in 3.7% formaldehyde for 2 hr, washed with KPO₄ (pH 6.5) twice, and resuspended in KPO₄ (pH 6.5) containing 1.2 M sorbitol. Fixed cells were permeabilized with a final concentration of 300 µg/ml zymolase (Seikagaku) and 62.5 mM ß-mercaptoethanol for 90 min at 30°. Microtubules were stained with 1:100 diluted anti--tubulin mouse monoclonal antibody (Sigma) followed by 1:50 diluted Texas red-conjugated anti-mouse IgG (Molecular Probes, Eugene, OR) and were mounted with DAPI. For microscopy, cells were observed with a x100 objective on a Zeiss (Thornwood, NY) Axioskop or on a Leica (Deerfield, IL) fluorescence DMR. Photographs were taken with Tmax 400 film and negatives were scanned with a Proimager 8200 (Pixelcraft) for figures, or alternatively images were taken directly using a Leica DC200 and a Leica DC viewer.

Fluorescence-activated cell sorter analysis:
For flow cytometry, budding yeast cells were prepared essentially as described by PAULOVICH and HARTWELL 1995 . Briefly, cells were fixed in 70% ethanol overnight at 4° and washed with 50 mM sodium citrate (pH 7.5). A total of 4 x 10⁶ cells were resuspended in 0.5 ml of 50 mM sodium citrate and incubated with 250 µg/ml RNase A followed by 1 mg/ml proteinase K for 1 hr at 50°, respectively. After adding 0.5 ml of propidium iodide in 50 mM sodium citrate (final concentration 8 µg/ml), samples were incubated in the dark for 12–24 hr at 4° and were sonicated briefly just before fluorescence-activated cell sorter (FACS) analysis. For each sample, 20,000 cells were analyzed with a Becton Dickinson (San Jose, CA) fluorescence-activated cell analyzer. For the standard of the DNA content of each FACS analysis, -factor-treated haploid wild-type (W303) cells [G1 (1N)], wild-type cells in log phase [G1 (1N) and G2 (2N)], and wild-type diploid (FY1679C) cells in log phase (2N and 4N) were used.

Coprecipitation and immunoblot:
To detect the coprecipitation of Bfa1p/Ibd1p and Ibd2p, YSK1 cells transformed with BFA1-HA/pTS903CL were grown to 5 x 10⁶ cells/ml and proteins were extracted in H-buffer [25 mM Tris-HCl pH 7.4, 15 mM EGTA pH 7.5, 15 mM MgCl₂, 0.1% Triton X-100, 10% glycerol, 1 mM NaN₃, 0.6 mM sodium vanadate pH 7.0, 1x protease cocktail in DMSO (Boehringer Mannheim, Indianapolis), 1 mM dithiothreitol, and 5 mg/ml phenylmethylsulfonyl fluoride] by beadbeating (MARK et al. 1990 ). The cell lysate was incubated with anti-HA antibody (Novagen) for 2 hr followed with Protein A-Sepharose (Sigma) for 1 hr at 4°, and the Protein A-Sepharose was washed twice with H-buffer. SDS sample buffer (6x) was added to the resin, the resin was boiled for 5 min, and the proteins were resolved using 10% SDS-PAGE. The resolved proteins were transferred onto nitrocellulose membrane and then incubated with polyclonal rabbit anti-HA antisera and polyclonal mouse anti-Myc antisera (Upstate Laboratories), respectively. Bound antibodies were detected with anti-rabbit or anti-mouse IgG-HRP (Jackson Immunochemicals, West Grove, PA) and enhanced chemiluminescence (ECL) reagents (Amersham, Arlington Heights, IL).

For immunoblots, cell extracts were prepared as described and the protein concentration of each yeast extract was determined by a dye-binding assay (Pierce, Rockford, IL ). A total of 35 µg of total protein from each extract was separated by 8% SDS-PAGE, transferred to a membrane, and incubated with polyclonal goat anti-Clb2 antisera (Santa Cruz Biotechnology), polyclonal rabbit anti-Pds1 antisera (Santa Cruz Biotechnology), or monoclonal rabbit anti-actin antisera (Calbiochem-Novabiochem, La Jolla, CA) followed by anti-goat or anti-rabbit IgG-HRP and was detected with ECL.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

IBD2 was isolated by its interaction with BFA1 in budding yeast:
Loss of BFA1 results in reduced viability in the presence of microtubule-destabilizing drugs, demonstrating its function in the spindle checkpoint of budding yeast (ALEXANDRU et al. 1999 ; LI 1999 ). To identify new components of the BFA1-dependent spindle checkpoint, we performed a yeast two-hybrid screen. An open reading frame YNL164C that encodes a potential Bfa1p interacting protein was isolated, which we named IBD2 (Inhibition of Bud Division 2) on the basis of its overexpression phenotypes described below. Ibd2p exhibited specific interactions with Bfa1p in the yeast two-hybrid LEU assay, in which IBD2 under the GAL1 promoter interacted with BFA1 only when its expression was induced by galactose/raffinose (Fig 1A). To verify the direct interaction of Bfa1p and Ibd2p, we constructed strain YSK1 in which IBD2 tagged with Myc and His at its C terminus was integrated into the chromosome of ibd2 and was expressed under its endogenous promoter. YSK1 and wild-type cells were transformed with a low copy plasmid carrying BFA1-HA under its endogenous promoter. When HA-tagged Bfa1p was purified with anti-HA antibody, HA-tagged Bfa1p exclusively coprecipitated with Myc-tagged Ibd2p expressed in YSK1 (Fig 1B).

View larger version (39K): In this window In a new window Download PPT slide   Figure 1. Ibd2p was isolated as a Bfa1p-interacting protein in a yeast two-hybrid screen. (A) Ibd2p directly interacts with Bfa1p in the yeast two-hybrid assay. The specific positive interaction of Ibd2p and Bfa1p was verified by a LEU assay in which true positives should grow on galactose/raffinose SC-ura, his, leu, trp (left) but not on glucose SC-ura, his, leu, trp (right), since the expression of IBD2 cDNA was under control of the GAL1 promoter. Lanes 1 and 2 are positive and negative controls, respectively. Lane 3 shows the specific interaction of Ibd2p with Bfa1p, a bait of the screen, on SC gal/raff-ura, his, leu, trp but not on SC glu-ura, his, leu, trp. (B) Ibd2p coprecipitates with Bfa1p. Strain YSK1, in which IBD2 tagged with Myc and His was integrated into the chromosome of ibd2, was transformed with a low copy plasmid carrying BFA1-HA (pTS903CL/BFA1-HA), and Myc-tagged Ibd2p and HA-tagged Bfa1p were coexpressed from their endogenous promoters (lane 2). Wild-type cells were transformed with a low copy plasmid carrying BFA1-HA (pTS903CL/BFA1-HA; lane 1). Strain YSK1 was also transformed with pTS903CL vector only as a negative control (lane 3). HA-tagged Bfa1p was purified with anti-HA antibody and Protein A-Sepharose. Lane 2 shows that HA-tagged Bfa1p exclusively coprecipitated with Myc-tagged Ibd2p. Lane 1 shows that Ibd2p in wild type was not detected by anti-Myc and lane 3 shows that only Myc-tagged Ibd2p and pTS903CL vector did not interact. The top and bottom show purified HA-tagged Bfa1p detected by immunoblotting with anti-HA antibody and Myc-tagged Ibd2p detected with anti-Myc antibody, respectively. (C) The predicted protein structure of Ibd2p containing a motif of CCPHHHYENLS between amino acids 262 and 273 that is frequently found in chitinase families by e-MOTIF database search. Boldface letters indicate amino acids conserved between Ibd2p and other proteins with this motif.

We verified the expression of IBD2 in S. cerevisiae by Northern blot (data not shown). IBD2 encodes a protein predicted to contain 352 amino acids. A BLASTP search of sequences in GenBank using the full-length Ibd2p sequence revealed no meaningful similarity, but a motif search showed that Ibd2p contains a conserved sequence between amino acids 263 and 273 (CCPHHHYENLS) that is found in chitinase families (Fig 1C) (HUANG and BRUTLAG 2001 ). A similar motif has also been found in BIK1 and DFG10, as shown in Fig 1C (BERLIN et al. 1990 ; MOSCH and FINK 1997 ).

IBD2 encodes a component of the spindle checkpoint:
The deletion of IBD2 was not lethal and showed no growth defects, as is also the case for the BFA1 deletion mutant (DUENAS et al. 1999 ; data not shown). To determine whether Ibd2p functions in the spindle checkpoint, we tested the sensitivity of ibd2 cells to a sublethal concentration of nocodazole. After incubation with 15 µg/ml nocodazole, the viability of ibd2 cells decreased as sharply as that of mad2, bub2, and bfa1 cells (Fig 2A). As reported, mad2 cells were more sensitive to the drug than were bub2 cells (FESQUET et al. 1999 ). To confirm these results, the sensitivity of ibd2 cells to benomyl was examined in conventional plating experiments. In the plate assay shown in Fig 2B, ibd2 cells serially diluted in the presence of 10 µg/ml benomyl exhibited a similar degree of hypersensitivity as mad2, bub2, and bfa1 cells. For the experiments described in Fig 2 Fig 3 Fig 4 Fig 5, we used spindle checkpoint mutants with isogenic W303 background except bub2. We also made a bub2 in W303 (YSK11) and confirmed that there is no obvious difference in physiological responses between the two bub2 strains of different backgrounds.

View larger version (42K): In this window In a new window Download PPT slide   Figure 2. The deletion mutant of IBD2 is sensitive to microtubule-destabilizing drugs. (A) Viability of the IBD2 deletion mutant cells decreases sharply in the presence of nocodazole. Wild-type (W303), bfa1 (YSK8), ibd2 (YSK2), bub2 (MAY2052), and mad2 (RHC 15.1) cells were grown to midlog phase (5 x 106 cells/ml), incubated with 15 µg/ml nocodazole for 0, 3, and 6 hr, respectively, and plated on YPD without any nocodazole to measure viability. The percentage of viable cells (%) was calculated relative to the number of viable cells at time 0. Three independent experiments were performed and the average was plotted with standard deviations. (B) ibd2 cells are sensitive to benomyl in plate assays. Actively growing wild-type (W303), bfa1 (YSK8), ibd2 (YSK2), bub2 (MAY2052), and mad2 (RHC 15.1) cells were serially diluted 10-fold and spotted onto either YPD plates (left) or YPD plates containing 10 µg/ml benomyl (right).

View larger version (50K): In this window In a new window Download PPT slide   Figure 3. IBD2 encodes a component of the spindle checkpoint. (A) ibd2 cells produce a new bud in the presence of nocodazole. Logarithmically growing wild-type (W303), bfa1 (YSK8), ibd2 (YSK2), bub2 (MAY2052), and mad2 (RHC 15.1) cells were arrested by -factor and released in the presence of 15 µg/ml nocodazole. The number of cells with more than one bud was counted every hour for 3 hr after fixation and brief sonication. A total of 500 cells were counted at each time point and the percentage of cells with more than one bud was calculated relative to total counted cells. Three independent experiments were performed and the average was plotted with standard deviations. (B) The phenotypes of ibd2 cells in the presence of nocodazole. Wild-type and ibd2 cells arrested with -factor were released in the presence of 15 µg/ml nocodazole for 5 hr. Cells taken every hour were fixed, briefly sonicated, and stained with DAPI. The top shows wild-type cells incubated either without nocodazole (left) or with nocodazole for 5 hr (right). The bottom shows the phenotypes of ibd2 cells treated with nocodazole at each hour as depicted. All panels are at the same scale. Bar, 5 µm. (C) ibd2 cells progress to the subsequent cell cycle in the presence of nocodazole. Wild-type (W303), ibd2 (YSK2), bfa1 (YSK8), and bub2 (MAY2052) cells arrested with -factor were released from the cell cycle arrest in the presence of nocodazole for 3 hr and their DNA contents were measured by flow cytometry. Cells are shown either without (left) or after nocodazole treatment (right). (D) ibd2 cells do not accumulate Clb2p when exposed to a microtubule-destabilizing drug. Wild-type (W303) and ibd2 (YSK2) cells were arrested in G1 by -factor and released in the presence of nocodazole at 25°. Cells were collected every 20 min for 140 min and the cells with a grown large bud and the nucleus in the neck were counted in each sample, as shown in the top graph. Cell extracts from each sample were immunoblotted with anti-Clb2 and anti-actin antibody (as a loading control).

View larger version (67K): In this window In a new window Download PPT slide   Figure 4. The overexpression of IBD2 induces mitotic arrest. (A) Cells overexpressing IBD2 are arrested in M phase. Wild-type (W303), bfa1 (YSK8), bub2 (MAY2052), and mad2 (RHC 15.1) cells were transformed with pMW20/IBD2 or pMW20L/IBD2 and the overexpression of IBD2 under the GAL10-1 promoter was induced for 12 hr. Cells in M phase were counted after fixation and staining with DAPI. The same strains transformed with only pMW20 were used as negative controls. The average of three independent experiments is shown with standard deviation. (B) Cells overexpressing IBD2 show no mitotic spindle defects. The mitotic spindle was visualized by immunostaining with antitubulin in cells overexpressing IBD2. Among cells in M phase by Ibd2p overexpression, each number of cells with a short mitotic spindle or with an elongated spindle was counted after antitubulin staining and displayed in the top. In the bottom section, the top shows a cell in metaphase with a short spindle while the bottom displays a cell with an elongated spindle. DNA (left) and microtubule (right) staining are shown. Bar, 5 µm. (C) Clb2p accumulates in cells overexpressing IBD2. Cell extracts were prepared from (1) wild type (W303) transformed with the vector pMW20 only, (2) wild type cells in which IBD2 overexpression was induced, and (3) wild-type cells incubated with nocodazole for 3 hr. Extracts were subjected to immunoblottings using anti-Clb2p (top) and anti-actin (bottom) as a loading control. (D) Clb2p does not accumulate when IBD2 is overexpressed in bfa1 and bub2 mutant cells. IBD2 overexpression was induced in wild-type (W303), bfa1 (YSK8), bub2 (MAY2052), and mad2 (RHC 15.1) cells, and Clb2p in extracts from each sample was analyzed by immunoblotting using anti-Clb2p (top). Actin is shown as a loading control (bottom). Lane 1, wild-type cells transformed with vector only as a control; lanes 2–5, cells overexpressing Ibd2p [(2) wild type, (3) bfa1, (4) bub2, and (5) mad2]. (E) Pds1p is stabilized in wild-type cells overexpressing IBD2. Wild-type cells (W303) were arrested with HU and released in the presence of 2% galactose to induce overexpression of IBD2. Cells were taken every 30 min for 210 min, and Pds1p and Clb2p were examined in each extract by immunoblots. The top shows wild-type cells transformed with vector pMW20 only; the bottom shows wild-type cells in which overexpression of IBD2 was induced. Pds1p and Clb2p are both stabilized in cells overexpressing IBD2, while Pds1p and Clb2p in wild-type cells degrade as cells exit mitosis. As a loading control, an actin blot for each sample is shown.

View larger version (91K): In this window In a new window Download PPT slide   Figure 5. IBD2 belongs to the BUB2 epistasis group. (A) New bud formation in ibd2bub2 and ibd2mad2 in the presence of nocodazole. Logarithmically growing wild-type (W303), ibd2 (YSK3), bub2 (MAY2052), mad2 (RHC15.1), ibd2bub2 (YSK5), ibd2bfa1 (YSK10), and ibd2mad2 (YSK7) cells were incubated in the presence of 15 µg/ml nocodazole, and cells with new buds were counted every 20 min for 160 min after fixation and brief sonication. A total of 300 cells were counted at each time point and the percentage of cells with new buds at each time point was calculated relative to the total number of cells counted. Three independent experiments were performed and the average was plotted with standard deviations. (B) Cell cycle progression of ibd2mad2 and ibd2bub2 in the presence of nocodazole by the analysis of DNA content. (a) ibd2bub2 (YSK5), (b) ibd2mad2 (YSK7), (c) bub2 (MAY2052), and (d) mad2 (RHC15.1) were arrested with -factor and released in the presence of 15 µg/ml nocodazole for 110 min. Cells were taken every 10 min and their DNA contents were analyzed by flow cytometry as described in MATERIALS AND METHODS. (C) dyn1ibd2 cells proceed through mitosis with a misoriented mitotic spindle. The dyn1 ibd2 double mutant was constructed and its phenotypes compared with that of dyn1 after DAPI and antitubulin immunostaining. The number of cells with each phenotype was counted as shown in the top table. In the images, top, dyn1 single mutant; bottom, dyn1ibd2 double mutant. DAPI staining for DNA (left) and antitubulin staining (right) are shown. The scale is the same in all panels. Bar, 5 µm.

Another phenotype of spindle checkpoint-defective mutants is repeated budding without mitotic arrest in the presence of microtubule-destabilizing drugs. To investigate whether the hypersensitivity of ibd2 cells to nocodazole was due to their inability to arrest the cell cycle in mitosis, we examined rebudding by counting cells that formed a new bud when mitotic spindle formation was inhibited by nocodazole. ibd2 cells arrested in G1 by -factor were released in the presence of 15 µg/ml nocodazole and the number of cells with more than one bud was compared with that of wild-type and other spindle checkpoint mutant cells. The number of cells with a new bud was not much changed in wild-type cells (W303), but was increased in ibd2 as well as in mad2, bub2, and bfa1 cells (Fig 3A).

We also compared by microscopy the phenotypes of ibd2 cells with those of wild type every hour for 6 hr, after cells arrested with -factor were released in 15 µg/ml nocodazole. Wild-type cells progressed through the cell cycle until mitosis, at which time they arrested with a large bud and a condensed nucleus in the bud neck (Fig 3B, top right). The ibd2 cells became enlarged at 2 hr compared with wild type (suggesting a transient mitotic delay), but produced aberrant extra bud(s) within 3 hr in the presence of nocodazole (Fig 3B, bottom). Transient mitotic delay and similar enlarged phenotypes are also observed when deletion mutants of BUB2 and BFA1 are treated with microtubule-destabilizing drugs (HOYT et al. 1991 ; FESQUET et al. 1999 ; LI 1999 ; data not shown). This transient enlarged phenotype observed in ibd2 as well as in bub2 and bfa1 mutants following exposure to nocodazole could be explained as temporary mitotic arrest due to the presence of a functional MAD2 spindle assembly checkpoint pathway. In the nocodazole-treated ibd2 cells, the nucleus also failed to become properly divided between the mother and the bud(s) (Fig 3B, bottom). These observations show that ibd2 cells could not ultimately arrest the cell cycle in mitosis and progressed to the next cell cycle without proper nuclear division and cytokinesis, as reported in other spindle checkpoint mutants including mad2, bub2, and bfa1 (HOYT et al. 1991 ; FESQUET et al. 1999 ; LI 1999 ), although microtubule-destabilizing drugs inhibited mitotic spindle formation.

To verify that the ibd2 mutant progressed through the cell cycle in the presence of mitotic spindle defects as other spindle checkpoint mutants, ibd2 cells arrested with -factor were released from cell cycle arrest into medium containing nocodazole for 3 hr, and their DNA contents were analyzed by flow cytometry. Wild-type, ibd2, bub2, bfa1, and mad2 cells all contained either a G1 (1N) or a G2 (2N) DNA content in the absence of nocodazole (Fig 3C, left). However, a larger proportion of ibd2, bub2, and bfa1 cells contained a G2 (2N) DNA content compared to wild type, suggesting a possible delay in mitosis (Fig 3C, left). After incubation with nocodazole for 3 hr, the ibd2, bub2, bfa1, and mad2 cells all displayed a G2 (2N) or higher DNA content, while most wild-type cells contained only a G2 (2N) DNA content (Fig 3C, right). The DNA content of ibd2 cells in the presence or absence of nocodazole was very similar to that observed for bub2 and bfa1. The distribution in DNA content of ibd2 was consistent with the new bud formation phenotypes described previously and demonstrated that the ibd2 cells proceeded to the next cell cycle and replicated their DNA without proper nuclear division and cytokinesis despite the loss of a mitotic spindle. These characteristics were also reported in other spindle checkpoint mutants including mad2, bub2, and bfa1 (HOYT et al. 1991 ; FESQUET et al. 1999 ; LI 1999 ).

When exposed to microtubule-destabilizing drugs such as nocodazole, wild-type yeast cells arrest in mitosis through the stabilization of B-type cyclins and the elevated activity of cyclin B-associated Cdc28 kinase. Thus, the level of Clb2 protein serves as an excellent marker for mitotic arrest. To further assess cell cycle progression in ibd2 cells, we compared Clb2 protein levels in ibd2 and wild-type cells incubated with nocodazole. Cells synchronized at G1 with -factor were released in fresh medium with nocodazole and collected every 20 min for 140 min to investigate the level of Clb2p. The percentage of cells with a large bud and the nucleus in the neck was also counted in each sample as an index for mitosis (Fig 3D, top). Wild-type cells gradually accumulated and maintained Clb2p at high levels in the presence of nocodazole, since cells were arrested in mitosis. ibd2 cells accumulated Clb2p with similar kinetics as wild type, but Clb2p began to degrade at 100 min, was completely degraded at 120 min, and started to reaccumulate at 140 min (Fig 3D). This observed profile of Clb2p in ibd2 cells was consistent with the above results that ibd2 cells exited mitosis and progressed into the next cell cycle in the presence of nocodazole. Taken together, these data demonstrate that Ibd2p functions in the spindle checkpoint of budding yeast.

IBD2 functions downstream of MPS1:
To further examine the function of IBD2 in the spindle checkpoint pathway, MPS1 overexpression phenotypes in ibd2 mutants were analyzed and compared with those in bub2 and mad2 checkpoint mutants by introducing a plasmid that overexpressed Mps1p under the GAL1 promoter. Cells transformed with a plasmid vector only were used as negative controls. As shown previously, the overexpression of Mps1p alone was sufficient to cause a mitotic delay by activating the MAD- and BUB-dependent spindle checkpoints in the absence of spindle damage, locating MPS1 upstream of the checkpoint composed of these genes (HARDWICK et al. 1996 ). However, it is arguable whether Mps1p overexpression affects both bifurcated spindle checkpoint pathways or only the MAD/BUB pathway and not the BUB2-dependent pathway, since the observed mitotic delay by Mps1p overexpression was completely eliminated in mad2 but not in bub2 mutants (ALEXANDRU et al. 1999 ; FESQUET et al. 1999 ; LI 1999 ). As displayed in Table 2, Mps1p overexpression in galactose-containing media for 12 hr resulted in mitotic arrest in 76% of wild-type cells, while only 19% of cells in the negative control were in a mitotic stage. This mitotic delay by Mps1p overexpression was abolished in mad2, when compared with the mad2 negative control (Table 2). Overexpression of Mps1p in ibd2 and bub2 also caused a mitotic delay, but the delay in ibd2 and bub2 was less than that in wild type and more than that in mad2 (Table 2). A similar percentage of ibd2 and bub2 mutant cells in mitotic delay in response to Mps1p overexpression implies that IBD2 and BUB2 are likely to function in the same pathway. This incomplete mitotic delay in response to Mps1p overexpression in ibd2 and bub2 also suggests that IBD2 and BUB2 are downstream of MPS1 and that MPS1 functions more effectively through the MAD/BUB pathway than through the BUB2-dependent pathway.

View this table: In this window In a new window   Table 2. MPS1 overexpression in ibd2 cells

Overexpression of Ibd2p induces mitotic arrest:
The function of Ibd2p in the spindle checkpoint was further verified by studying the overexpression of Ibd2p. Overexpression of Ibd2p under the GAL10-1 promoter in wild-type cells blocked the cell cycle in mitosis, as revealed by DAPI staining and nuclear morphology counts. A total of 78% of wild-type cells overexpressing Ibd2p were arrested either at metaphase or at anaphase with a large bud, while 29% of wild-type cells with the vector control were in M phase (Fig 4A). These increased mitotic arrest phenotypes by Ibd2p overexpression strongly suggest that Ibd2p functions in the spindle checkpoint.

To confirm that the overexpression of Ibd2p does not activate the spindle checkpoint indirectly by causing a spindle defect, we inspected microtubule structures in cells overexpressing Ibd2p by antitubulin immunofluorescence microscopy. As can be seen in Fig 4B, cells overexpressing Ibd2p did not show any defect in the orientation of microtubules and displayed the normal microtubule patterns of M-phase cells in metaphase or anaphase. When the cells arrested in mitosis by Ibd2p overexpression were further analyzed by their spindle patterns, 67.1% of cells had a short spindle and 32.9% contained an elongated spindle (Fig 4B, top). These observations demonstrated that the overexpression of Ibd2p does not generate mitotic spindle defects but induces mitotic arrest directly through its role as a component of the spindle checkpoint. The mitotic arrest caused by the overexpression of Ibd2p in wild type was also validated by measuring Clb2p levels. The amount of Clb2p in wild-type cells overexpressing Ibd2p was elevated relative to actively growing wild-type cells, although a similar amount of actin was present, and was comparable to that of wild-type cells arrested in mitosis as a result of nocodazole treatment (Fig 4C and Fig D).

To examine in which of the bifurcated spindle pathways and where in the pathway Ibd2p functions, Ibd2p overexpression phenotypes were observed in wild-type as well as in mad2, bfa1, and bub2 cells, and these were compared with negative controls in each strain that overexpressed only the vector control. The increased mitotic arrest caused by Ibd2p overexpression was observed in mad2 cells but much less in bfa1 and not at all in bub2 (Fig 4A). When compared with each vector control, the lack of mitotic arrest in response to Ibd2p overexpression in bfa1 and bub2 suggests that Ibd2p functions through Bfa1p and Bub2p in the BUB2 pathway. No mitotic arrest by Ibd2p overexpression was observed in bub2 cells and only a minor mitotic arrest was detected in bfa1. This could be explained if a regulator of Bub2p other than Bfa1p interacts with Ibd2p for mitotic arrest. As the spindle checkpoint is bifurcated into Mad/Bub-dependent and Bub2-dependent pathways, the mitotic arrest by Ibd2p overexpression in mad2 is more likely due to the functional independence of IBD2 and MAD2 in different pathways and supports the hypothesis that IBD2 functions in the BUB2 pathway.

To gain further insight into the functional pathway of Ibd2p, we also measured the level of Clb2p in wild-type as well as in mad1, bfa1, and bub2 cells overexpressing Ibd2p. As described above, the overexpression of IBD2 in wild type resulted in mitotic arrest with increased Clb2p (Fig 4A and Fig 4C and Fig D). Overexpression of Ibd2p resulted in an increased level of Clb2p in mad1 but not in bfa1 and bub2, reinforcing the conclusion that Ibd2p is a component of the spindle checkpoint and that overexpression of Ibd2p induces mitotic arrest through the BUB2 pathway upstream of BFA1 and BUB2.

We analyzed the stability of Pds1p in wild-type cells overexpressing Ibd2p to further verify the mitotic arrest caused by Ibd2p overexpression. For this experiment, wild-type cells were synchronized in S phase with 0.1 M HU and released in the presence of 2% galactose to induce the overexpression of IBD2. In control cells expressing vector alone, Pds1p accumulated gradually and was completely degraded at 210 min after the release when cells exit mitosis. Clb2p, whose degradation is inhibited by Pds1p (COHEN-FIX and KOSHLAND 1999 ), began to decrease at 90 min when Pds1p started to diminish (Fig 4E). In contrast, in cells where the overexpression of Ibd2p was induced, Pds1p accumulated and was maintained at a constant high level. Clb2p was also maintained stably in these cells (Fig 4E). The observed stability of Pds1p in wild-type cells overexpressing Ibd2p confirms that the overexpression of Ibd2p induces mitotic arrest and that Ibd2p functions as a component of the spindle checkpoint.

IBD2 belongs to the BUB2 epistasis group:
The above results, including the phenotypes of ibd2 in the presence of nocodazole as well as Ibd2p overexpression, strongly suggest that IBD2 functions in the BUB2 branch of the spindle checkpoint. The function of IBD2 in the BUB2 pathway was further validated by genetic epistasis analysis. Blockage of both MAD2 and BUB2 pathways in the mad2bub2 double mutant appeared to ablate the checkpoint completely, so that these cells progressed into the next cell cycle without any mitotic delay and rebudded more rapidly than either single mutant (mad2 or bub2) in the presence of nocodazole (FESQUET et al. 1999 ). If IBD2 functions in the BUB2 branch, we would expect that the checkpoint would be blocked more effectively in the ibd2mad2 mutant than in ibd2bub2. We constructed an ibd2bub2 strain (YSK2) and an ibd2mad2 strain (YSK3) and investigated whether increased nocodazole sensitivity was detected by an enhanced checkpoint defect in each double mutant. First, we counted the rebudding of these double mutants in the presence of nocodazole, after cells were released from the arrest by -factor. As can be seen in Fig 5A, while the proportion of cells with new buds in the ibd2bub2 double mutant was similar to that in the bub2 single mutant, the increase occurred a little more rapidly in ibd2mad2 in comparison with ibd2, mad2, and ibd2bub2 mutants. Cell cycle progression as assayed by new bud formation in the ibd2mad2 double mutant was especially remarkable during the early stages of incubation with nocodazole (Fig 5A).

We also examined the mitotic arrest defect and cell cycle progression of ibd2bub2 and ibd2mad2 by measuring the DNA content. The cell cycle progression of bub2 and mad2 single mutants as well as wild type was also monitored as controls. In this experiment, ibd2-bub2, ibd2mad2, bub2, and mad2 cells in log phase were arrested with -factor and released in the presence of 15 µg/ml nocodazole for 110 min, and their DNA content was analyzed every 10 min by flow cytometry. Approximately 80% of both wild-type and mutant cells were arrested at G1 by -factor with 1N DNA content. Wild-type cells were gradually arrested with 2N DNA content in the presence of nocodazole as cells reached mitosis (Fig 3C and data not shown). Similarly, both ibd2bub2 and ibd2mad2 double mutants proceeded through the cell cycle and displayed a 2N DNA content after release. However, as shown in Fig 5B, Fig B, a fraction of the ibd2mad2 cells began to display a 4N DNA content at 80 min after release and this proportion was highly increased at 110 min without noticeable mitotic delay. Conversely, most of the ibd2bub2 cells retained a 2N DNA content at 80 min and only a small portion of cells with a 4N DNA content started to appear at 110 min, suggesting a temporal mitotic delay of this double mutant in the presence of nocodazole (Fig 5B, a). As ibd2bub2, both bub2 and mad2 single mutant cells showed a temporal mitotic delay with a 2N DNA content until 90 min and only a small portion of cells began to have a 4N DNA content at 110 min (Fig 5B, Fig C and Fig D). Therefore, the deletion of both IBD2 and MAD2 appeared to ablate the spindle checkpoint more severely than that of IBD2 and BUB2, suggesting that IBD2 and MAD2 function in separate checkpoint pathways and that IBD2 and BUB2 likely belong to the same epistasis group.

The BUB2 checkpoint pathway has been reported to monitor anaphase spindle position. BUB2 is required for the prevention of spindle breakdown and mitotic exit in yeast cells lacking dynein or dynactin, in which both poles of the spindle are in the mother cell (BLOECHER et al. 2000 ). Therefore, 25% of the bub2dyn1 double mutant cells with misaligned spindles completed anaphase and initiated new bud formation, while 95% of the dyn1 mutant cells with misaligned spindles were arrested at mitosis (BLOECHER et al. 2000 ). If IBD2 functions in the BUB2 pathway, then ibd2dyn1 cells would also be expected to undergo nuclear division within the mother and enter the next cell cycle, as is observed for bub2dyn1. To confirm that IBD2 functions in the BUB2 branch of the spindle checkpoint pathway, we constructed a ibd2dyn1 double mutant and compared the progression of mitosis in ibd2dyn1 with that in dyn1. Consistent with previous reports on dyn1 (LI et al. 1993 ; BLOECHER et al. 2000 ), only 9% of dyn1 single mutant cells that contained misaligned anaphase spindles exited mitosis and formed an extra bud. On the other hand, 38% of ibd2dyn1 double mutant cells gave rise to binuclear mother cells and anucleate cells and often contained an extra bud (Fig 5C). ibd2dyn1 double mutant cells also showed a sharp decrease in viability compared with each single mutant, possibly due to progression of the cell cycle in the absence of proper nuclear division, as in the case of bub2dyn1 (data not shown). These ibd2dyn1 phenotypes are very similar to those reported for bub2dyn1 and strongly suggest that IBD2 belongs to the BUB2 epistasis group and functions in the BUB2-dependent spindle checkpoint.

Mitotic arrest defects of ibd2 can be rescued by extra copies of BUB2 and BFA1:
As described above, the lack of mitotic arrest by Ibd2p overexpression in bfa1 and bub2 as well as the similar Mps1p overexpression phenotypes observed in ibd2 and bub2 mutants suggests that IBD2 functions upstream of BFA1 and BUB2 in the BUB2 pathway. To further address where IBD2 functions in the BUB2 spindle checkpoint pathway, we investigated whether the loss of mitotic arrest in ibd2 in the presence of microtubule-destabilizing drugs could be compensated by extra copies of BFA1 and BUB2 on a CEN plasmid, as summarized in Table 3. We also examined whether the presence of IBD2 on a CEN plasmid could suppress the lack of mitotic arrest in bfa1 and bub2 upon exposure to nocodazole (Table 4). A total of 82% of negative controls (ibd2 cells transformed with the CEN plasmid pRS316 only) failed to undergo mitotic arrest in the presence of nocodazole, while 75% of positive controls (ibd2 cells transformed with IBD2 on a CEN plasmid) were able to restore mitotic arrest. When compared with ibd2 and mad2 cells transformed with pRS316 alone, the mitotic arrest defects of ibd2 were not much improved by the presence of MAD2 and vice versa (Table 3 and Table 4), supporting the idea that IBD2 is not in the MAD2 pathway. Extra copies of either BUB2 or BFA1 were able to restore mitotic arrest in 73–75% of ibd2 cells, a level comparable to that of IBD2 itself (Table 3). In contrast, while BFA1 and BUB2 were each able to compensate for the inability of bfa1 and bub2 cells to undergo mitotic arrest, extra copies of IBD2 could not (Table 3). These genetic interactions of IBD2 are consistent with the data described previously indicating that IBD2 functions upstream of BUB2 and BFA1 in the BUB2-dependent spindle checkpoint pathway. Interestingly, an extra copy of CDC5, which encodes a polo-like kinase functioning in mitotic exit and cytokinesis, could rescue the deficiency of mitotic arrest in ibd2 (SONG and LEE 2001 ). It is hard to define a relationship between IBD2 and CDC5 solely on the basis of this result, but at least it suggests a possible genetic interaction between these two genes that could link the checkpoint function of IBD2 with mitotic exit.

View this table: In this window In a new window   Table 3. Rescue of the mitotic arrest defects in ibd2 by other spindle checkpoint genes

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

IBD2 is a new component of the BUB2-dependent spindle checkpoint in budding yeast:
Bfa1p, Bub2p, and Tem1p are associated with the spindle pole body (SPB) of budding yeast and together comprise the Bub2p-dependent spindle checkpoint (ALEXANDRU et al. 1999 ; LI 1999 ). The homologies between Bfa1p and Bub2p in budding yeast and Byr4p and Cdc16p in fission yeast suggest that Bfa1p and Bub2p might form a two-component GAP at the SPB that inhibits the Tem1p GTPase, thus blocking mitotic exit (FURGE et al. 1998 ). Recent reports suggest a new model in which the inhibition of Tem1p by Bfa1p/Bub2p GAP is a molecular switch that acts on the Bub2p-dependent spindle checkpoint to prevent spindle breakdown, nuclear division, and cytokinesis in response to a failure of nuclear migration into the bud, which typically results from spindle orientation defects (BARDIN et al. 2000 ; BLOECHER et al. 2000 ; PEREIRA et al. 2000 ). In this model, when the divided nucleus and progeny SPB migrate to the new bud, Lte1p, a guanine nucleotide exchange factor (GEF) that is present only in the bud, activates the Tem1p GTPase to stimulate mitotic exit and cytokinesis (BARDIN et al. 2000 ; PEREIRA et al. 2000 ). The interactions of Bim1p and Kar9p with cortical sites have been reported to properly orient the mitotic spindle in yeast (KORINEK et al. 2000 ; LEE et al. 2000 ; MILLER et al. 2000 ). However, it still remains to be answered whether Bfa1p and Bub2p together are sufficient to act as a molecular switch and how spindle orientation defects are transmitted to Bfa1p/Bub2p.

We have isolated a novel Bfa1p interacting protein named Ibd2p. We have presented evidence that IBD2 functions as a component of the spindle checkpoint in the BUB2-dependent branch of this pathway upstream of Bfa1p and Bub2p. IBD2 deletion mutants proceeded through the cell cycle in the presence of microtubule-destabilizing drugs, thereby causing a sharp decrease in viability. IBD2 overexpression induced mitotic arrest in wild-type cells, as verified by increased levels of Clb2p and the stabilization of Pds1p, but the pattern of microtubule structures was normal in these cells. This mitotic arrest in response to IBD2 overexpression was not observed in BUB2 and was slightly detected in BFA1 deletion mutant cells. These results strongly suggest that IBD2 functions as a component of the spindle checkpoint pathway upstream of BUB2 and BFA1. The minor difference of Ibd2p overexpression phenotypes in BUB2 and BFA1 deletion mutants could be explained if a regulator of Bub2p other than Bfa1p interacts with Ibd2p for mitotic arrest. Further study of proteins interacting with Ibd2p would answer this possibility. Interestingly, we also observed that Ibd2p overexpression caused M-phase arrest with either a short bipolar spindle (67.1%) or an elongated bipolar anaphase B spindle (32.9%), while overexpression of Bfa1p was reported by LI 1999 to induce a cell cycle arrest mainly with an elongated anaphase B spindle. It was reported previously that MPS1 functions upstream of both MAD- and BUB-dependent spindle checkpoints and that its overexpression leads to a cell cycle arrest predominantly with short bipolar spindles (HARDWICK et al. 1996 ). Pds1p was shown to control both the initiation of anaphase and mitotic exit and to interact with mitotic exit network (MEN) proteins by unknown mechanisms (COHEN-FIX and KOSHLAND 1999 ; TINKER-KULBERG and MORGAN 1999 ). Our observation that the overexpression of Ibd2p caused M-phase arrest either with a short bipolar spindle or with an elongated bipolar anaphase B spindle suggests the possibility that communication between Pds1p and Ibd2p may occur upstream of Bfa1p and Bub2p when there are defects in spindle formation and orientation.

The results that the mitotic arrest caused by Ibd2p overexpression was not observed in bub2 and bfa1/ibd1, and that the mitotic arrest defects of ibd2 in the presence of nocodazole were restored by additional copies of BUB2 and BFA1 while an extra copy of IBD2 could not rescue the mitotic arrest defects of bub2 and bfa1, are consistent with a model in which IBD2 functions upstream of BUB2 and BFA1. Considering that Ibd2p and Bfa1p directly interact and that an extra copy of either BFA1 or BUB2 rescues the mitotic defect of ibd2, we could expect that Ibd2p, Bfa1p, and Bub2 form a complex to transmit the spindle orientation defect to block mitotic exit. We observed that Ibd2p physically interacts with Bfa1p but does not interact directly with Bub2p (H. HWANG, unpublished data), although Ibd2p and Bub2p still can communicate through Bfa1p by forming a complex together. The observation that Ibd2p is localized as a dot near the nuclear periphery when expressed as a green fluorescent protein (GFP) fusion on a multicopy plasmid suggests that endogenous Ibd2p, like Bfa1p, is localized to the spindle pole body and supports the possibility that they form a complex in the spindle pole body (H. HWANG, unpublished data).

We also observed that the mitotic arrest defect of ibd2 in the presence of nocodazole was restored by additional copies of CDC5. CDC5 encodes a polo-like kinase in S. cerevisiae and functions as a component of the MEN, but how CDC5 participates in the mitotic exit pathway is not known (KITADA et al. 1993 ; JASPERSEN et al. 1998 ). Cdc5p also plays a role in regulating cytokinesis and is proposed to coordinate mitotic exit with cytokinesis (SONG and LEE 2001 ). A recent report showed that the phosphorylation of Bfa1 by Cdc5p regulates the checkpoint function of Bfa1p (HU et al. 2001 ). Our data showing that the mitotic arrest defect of ibd2 is rescued by additional copies of CDC5 suggest that IBD2 is likely to function upstream of CDC5 in the mitotic exit pathway to regulate the function of Bfa1p. However, the genetic interaction between IBD2 and CDC5 seems to be confined in the control of mitotic exit, since the deletion or overexpression of IBD2 does not affect cytokinesis.

Possible functions of IBD2 in the BUB2-dependent pathway:
Ibd2p is the first component of the Bub2-dependent spindle checkpoint to be identified that acts directly upstream of Bfa1p and Bub2p. Since the Bub2p-dependent spindle checkpoint prevents mitotic exit and cytokinesis in response to spindle orientation defects, Ibd2p is likely to function in transmitting signals about spindle integrity to Bfa1p and Bub2p for regulating the exit from mitosis. Ibd2p is a novel protein with no homology to proteins of known function, but it contains a conserved sequence (CCPHHHYENLS) found in chitinase families (HUANG and BRUTLAG 2001 ). This motif is also found in Bik1p, which is required for proper microtubule function during mitosis and which shows both genetic and direct physical interactions with Bim1p and -tubulin in budding yeast (BERLIN et al. 1990 ; SCHWARTZ et al. 1997 ). The conservation of this sequence in Ibd2p thus points toward possible functions of Ibd2p in monitoring spindle orientation defects during mitosis. Our unpublished data suggest that the domain of Ibd2p containing this motif is important for eliciting the mitotic arrest observed upon overexpression of Ibd2p. Further analysis of genetic and physical interactions between Ibd2p and proteins involved in proper spindle orientation such as Bim1p, Kar9p, Bik1p, and Nud1p is underway and should give clues about the precise mechanism by which Ibd2p functions in the spindle checkpoint.

Since the overexpression of Ibd2p induced mitotic arrest upstream of Bfa1p and Bub2p, Ibd2p could act as a negative regulator of Tem1 GTPase, possibly by activating Bfa1 and Bub2 GAP or by inhibiting the GEF Lte1. We observed that Ibd2p does not physically interact with Lte1p but is localized as a dot near the nuclear periphery when expressed as a GFP fusion, suggesting that it localizes on the spindle pole body, as does Bfa1p (H. HWANG, unpublished data). These observations strongly suggest that Ibd2p functions negatively on Tem1 GTPase by direct interaction with Bfa1p. Although there is no direct interaction between Ibd2p and Lte1p, a genome-wide two-hybrid analysis reported that Ibd2p is connected to Lte1p through YNL091W and Msl1p (UETZ et al. 2000 ). Therefore, the possibility that Ibd2p acts to negatively influence Lte1p cannot be excluded. Conceivably, Ibd2p might function indirectly to inactivate Tem1p GTPase both by activating Bfa1p and by inactivating Lte1p to block mitotic exit when there are defects in spindle formation and orientation.

	ACKNOWLEDGMENTS

We thank Drs. J. Kim, K. Choi, M. Winey, C. Rodriquez, F. Galibert, A. Hoyt, A. Toh-e, M. Lonetine, and F. Rey for providing yeast strains and plasmids. We are also indebted to Dr. Kris Gunsalus of New York University for English editing. This work was supported by a grant from the Korean Ministry of Science and Technology (Critical Technology 21 on "Life Phenomena and Function Research"; 00-J-LF-01-B-60) donated to K. Song and partly by a grant from the Korean Science & Engineering Foundation (KOSEF 1999-2-207-003-3).

Manuscript received November 30, 2001; Accepted for publication March 25, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS