S. cerevisiae Bir1 Localizes as a Chromosomal Passenger
Localization of full-length Bir1 has proven difficult, although it seems to associate with microtubules (Uren et al., 1999
; Yoon and Carbon, 1999
; Huh et al., 2003
). We successfully tagged and localized endogenous full-length Bir1 with the YFP variant Venus. Bir1-Venus shows a dynamic localization through the cell cycle that is similar to chromosomal passenger proteins in vertebrates (). Bir1 colocalizes with Sli15 at all stages of the cell cycle. Bir1 seems to move from kinetochores to the spindle at anaphase and concentrates at the midzone late in the cell cycle.
Figure 1. Bir1 localizes as a chromosomal passenger. (A–C) Strain PWY19, expressing Bir1 with a C-terminal Venus tag and Sli15 with a C-terminal CFP tag expressed under control of their endogenous promoters, was imaged as described in Materials and Methods (more ...)
We compared the localization of Bir1-Venus and Ndc10-CFP. Bir1 and Ndc10 colocalize at kinetochores before anaphase. During anaphase, Bir1 leaves the kinetochore, whereas the majority of Ndc10 remains. Bir1 and a pool of Ndc10 colocalize to the spindle during anaphase and at the midzone during telophase.
BIR1 Is Essential
There has been some dispute as to whether BIR1
is essential for viability. Two groups have reported BIR1
is not essential (Uren et al., 1999
; Yoon and Carbon, 1999
), and one group reported that BIR1
is essential (Li et al., 2000
). We found that BIR1
is essential. We replaced one copy of the BIR1
ORF with the hygromycin B resistance gene, hph, in a diploid. Colony PCR using primers flanking the ORF confirmed the presence of a deleted copy and a wild-type copy in the diploid. Sporulation and tetrad dissection of the diploid isolates gave two live and two dead spores. The two live spores in each of 24 tetrads were hygromycin B sensitive. Li and coworkers proposed that any viable bir1
Δ haploids isolated from a heterozygous diploid contain a wild-type copy of BIR1
. We tested this possibility directly. Our BIR1/bir1
heterozygous diploid is also heterozygous for the recessive cyhr
allele. We sporulated this strain and selected for haploids on medium containing cycloheximide. Forty-eight of 276 cycloheximide-resistant spores isolated were resistant to hygromycin B indicating the presence of a disrupted copy of BIR1
. Colony PCR of six of these viable cells containing bir1
confirmed the presence of both a wild-type and a disrupted copy of BIR1
Finally, we constructed a plasmid shuffle strain for BIR1, where a haploid has the endogenous copy of BIR1 disrupted and carries a multicopy plasmid carrying BIR1 and ADE3. If BIR1 is not essential, then the strain should form colonies with white sectors, the white sectors being cells that have lost the only copy of BIR1, which is on the ADE3 plasmid. However, this strain forms solid red colonies, indicating that BIR1 is essential for viability.
The C Terminus of Bir1 Is Essential
Because BIR1 is essential, we performed a deletion analysis to define the essential regions of BIR1. Deletions and truncations of Bir1 were expressed from the BIR1 promoter in an ARS/CEN vector (). These constructs were transformed into the plasmid shuffle strain PWY16-4D to assess their ability to support the growth of yeast cells. Only the C-terminal truncation (bir1-876stop) and the largest deletion (bir1-ΔMN) were unable to support growth ().
Figure 2. Bir1 deletions and truncations. (A) Internally deleted and truncated versions of BIR1 were constructed as described in Materials and Methods. Sectoring is shown as the percentage of colonies that formed white sectors in red colonies when transformed with (more ...)
We localized the two inactive deletions. Bir1-876stop is present in the nucleus but does not localize to the kinetochore or spindle (). Bir1-ΔMN seems uniformly spread throughout the cell and is not found in the nucleus (). Addition of a nuclear localization sequence to Bir1-ΔMN partially restored proper localization (). The NLS-Bir1-ΔMN is also able to support growth, indicating that Bir1-ΔMN lacks an NLS but is otherwise functional. Therefore, the C terminus is the essential region of Bir1 and is required for kinetochore and spindle localization.
Figure 3. Localization of Bir1, Sli15, and Ndc10 in bir1 deletion mutants. (A) Two bir1 mutants that could not independently support growth, Bir1-876stop and Bir1-ΔMN, were tagged with Venus and integrated into a wild-type haploid and imaged as described (more ...)
The Essential C Terminus of Bir1 Interacts with Sli15
We purified TAP-tagged Bir1 and identified the copurifying proteins by mass spectrometry. In six of six purifications, Sli15 copurified with Bir1. Ipl1 was not detected even when Bir1 was purified under low stringency. These results are consistent with a model in which Bir1 forms a complex with Sli15 that does not contain Ipl1.
The region of Bir1 responsible for interaction with Sli15 was mapped by directed two-hybrid experiments. Full-length Bir1 showed a strong interaction with full-length Sli15 (). The C-terminal 297 amino acids (C terminus of bir1-ΔMB) and the C-terminal 80 amino acids (C terminus of bir1-ΔMN) also showed interaction with full-length Sli15 ().
Figure 4. The C terminus of Bir1 interacts with Sli15. Full-length Sli15 was fused to the Gal4 DNA binding domain (DBD). Full-length and fragments of Bir1 were fused to the Gal4 activation domain. Combinations of Sli15 and Bir1 fragments were cotransformed into (more ...)
The IAP Repeats of Bir1 Are Not Required for the Spindle Checkpoint
The function of the IAP repeats in Bir1 is not known. Deletion of the repeats did not affect Bir1's localization or expression (Figures and ). The IAP repeats were not required for Sli15 localization or Ndc10 localization (). Deletion of the IAP repeats did not alter the rate of chromosome loss as measured by a fluctuation analysis (). Because Bir1 has been linked to Ipl1 function and Ipl1 homologues can phosphorylate IAP repeats in vitro (Leverson et al., 2002
; Wheatley et al., 2004
), we examined whether the IAP repeats were required for the tension checkpoint. Cells carrying a temperature-sensitive mutation in cohesin, mcd1-1
, delay in the cell cycle, and this delay is dependent on Ipl1 (Biggins and Murray, 2001
). We found that deletion of the IAP repeats did not alter the cell cycle delay caused by the mcd1-1
Figure 5. The IAP repeats of Bir1 are not required for the tension checkpoint. Cells carrying the bir1-ΔIAP allele and a temperature-sensitive cohesin mutation, mcd1-1, were arrested in G1 with α-factor and released to the restrictive temperature (more ...)
Deletions of the Middle Region of Bir1 Alter Ndc10 Localization and Increase the Rate of Chromosome Loss
Deletions in the middle region of Bir1 could support growth. These deletions localized properly except the bir1-ΔSN and the bir1-ΔMB mutants showed weaker spindle localization (). All deletions also expressed well (). The deletions of the middle region did not alter localization of Sli15 (). Localization of Ndc10 to the kinetochore was not altered. However, localization of Ndc10 to the anaphase spindle and the midzone was prevented (). Moreover, the middle region deletions caused a moderate-to-severe increase in the rate of chromosome loss as analyzed by a fluctuation analysis () (Runge et al., 1991
; Davis, 1992
Bir1 Is Cell Cycle Regulated
Bir1 expression levels fluctuate through the cell cycle (). Bir1 is barely expressed in cells arrested with α-factor. Protein levels increase after release from the arrest. Bir1 is also phosphorylated in a cell cycle-dependent manner. Phosphorylation occurs as the cells enter S phase and is removed late in the cell cycle (). Immunoprecipitated Bir1 migrates on SDS-PAGE as multiple bands, which can be condensed to one band by treatment with phosphatase, demonstrating that the mobility shift is because of phosphorylation (). Bir1 is heavily modified at a cdc15 cell cycle arrest, indicating phosphorylation is maintained into telophase ().
Figure 6. Bir1 is phosphorylated in a cell cycle-dependent manner. (A) Haploid cells expressing TAP-tagged Bir1 were synchronized by treatment with α-factor and released as described previously (Yoder et al., 2005 ). At each time point, extracts were prepared (more ...)
Phosphorylation of Bir1 Is Required for Interaction with Ndc10 and Efficient Spindle Elongation
We identified which sites were phosphorylated in vivo by analyzing purified Bir1 by mass spectrometry (). The two sites, T735 and T747, were minimal consensus sites for Cdc28 phosphorylation.
Mutation of the two identified sites diminished but did not abolish phosphorylation (, bir1-2A). Mutation of all nine cdc28 consensus sites in the central region (S383A, S395A, S552A, S587A, S667A, T684A, S688A, T735A, and T747A) dramatically decreased the shift (, bir1-9xA). Mutation of these nine sites prevents localization of Ndc10 to the spindle at anaphase (). Moreover, treatment of Bir1-13xMyc with phosphatase disrupts the in vitro interaction between Ndc10 and Bir1 ().
Figure 7. Phosphorylation of Bir1 is required for interaction with Ndc10. (A) The bir1-9xA mutant was transformed into the modified plasmid shuffle strain PWY91-10c. Ndc10-YFP was localized in the presence of this bir1 mutant as described in Materials and Methods (more ...)
Finally, we examined the consequence of loss of Ndc10 from the anaphase spindle and midzone by characterizing the bir1-9xA mutant. Cells were synchronized by treatment with α-factor and then released and progression through the cell cycle was analyzed. Bud emergence and cytokinesis occurred at the same time in mutant and wild-type cultures (). Next, we measured the length of the mutant and wild-type anaphase spindles in a cycling population of cells. The mutant anaphase spindles were shorter with an average length of 4.9 μm compared with an average length of 5.4 μm for wild-type spindles. This was a statistically significant difference to a 99% confidence level with Student's t test. The main difference between wild-type and mutant was a lack of long spindles greater than 7 μm in length in the mutant cells. Less than 1% of the mutant anaphase spindles were above 7 μm in length compared with 12% of wild-type anaphase spindles ().
Figure 8. Full spindle elongation is prevented in the bir1-9xA mutant. (A) Cultures of strains BESY49-6c and PWY194 were synchronized by treatment with α-factor as described previously (Yoder et al., 2005 ). At the given times, samples were removed and (more ...)