![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 19, Issue 2, 682-690, February 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Cell Biology, Institute of Life Science, Kurume University, Fukuoka 839-0864, Japan
Submitted May 29, 2007;
Revised November 12, 2007;
Accepted November 27, 2007
Monitoring Editor: Kerry Bloom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ams2, CENP-A fails to retain during S, but it reaccumulates onto centromeres via the G2 deposition pathway, which is down-regulated by Hip1, a homologue of HIRA histone chaperon. Reducing the length of G2 in ams2 results in failure of CENP-A accumulation, leading to chromosome missegregation. N-terminal green fluorescent protein-tagging reduces the centromeric association of CENP-A, causing cell death in ams2 but not in wild-type cells, suggesting that the N-terminal tail of CENP-A may play a pivotal role in the formation of centromeric nucleosomes at G2. These observations imply that CENP-A is normally localized to centromeres in S phase in an Ams2-dependent manner and that the G2 pathway may salvage CENP-A assembly to promote genome stability. The flexibility of CENP-A incorporation during the cell cycle may account for the plasticity of kinetochore formation when the authentic centromere is damaged. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Lack of a kinetochore or formation of multiple kinetochores on a chromosome may have deleterious effects on mitosis (Karpen and Allshire, 1997; Amor and Choo, 2002; Henikoff and Dalal, 2005). CENP-A represents the most likely candidate for the epigenetic "mark" responsible for the maintenance of centromere identity (Black et al., 2004, 2007). Because reformation of CENP-A–containing nucleosomes after DNA synthesis is thought to be a prerequisite for mitotic kinetochore assembly, precise targeting of CENP-A into a single, restricted locus on each chromosome before cell division is essential for cell viability (Takahashi et al., 2005). At least three components that affect CENP-A localization, the Mis16–Mis18 complex (Hayashi et al., 2004; Fujita et al., 2007), the Mis6–Sim4 complex (Takahashi et al., 2000; Pidoux et al., 2003), and Ams2 GATA-type transcription factor (Chen et al., 2003a), have been identified in the fission yeast Schizosaccharomyces pombe, which is an ideal model organism in which to study complex centromere structure and function (Takahashi et al., 1992; Karpen and Allshire, 1997).
Which phase of the cell cycle is used for CENP-A incorporation remains controversial. During S phase, canonical core histones have been suggested to be deposited into duplicated DNAs in a semiconservative manner (Tagami et al., 2004; Natsume et al., 2007). Experiments using fluorescence recovery after photobleaching demonstrated that CENP-A of the budding yeast Saccharomyces cerevisiae is recruited to centromeres coincident with DNA synthesis (Pearson et al., 2004), presumably reflecting disassembly and reassembly of centromeric nucleosomes at the replication fork. In contrast, studies performed in human cells (Shelby et al., 2000; Jansen et al., 2007) and in Drosophila melanogaster (Ahmad and Henikoff, 2001; Sullivan and Karpen, 2001; Schuh et al., 2007) indicated that CENP-A is incorporated in a replication-independent manner, although the molecular components and physiological significance of this pathway remain elusive. Here, we report that ams2 cells are defective in the retention of Cnp1, S. pombe CENP-A, at S phase, but they are able to survive through Cnp1 incorporation using the replication-independent pathway at G2. Our observations indicated that the G2 deposition of Cnp1 is mechanically distinct from the S deposition and could act as a salvage pathway that enables unincorporated Cnp1 to reassociate to the centromeres before cells go through subsequent lethal mitosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Chen et al., 2003a). The genotypes of strains and procedures used for gene disruption are described in Supplemental Table S1. Anti-Cnp1 polyclonal antibody was generated by immunizing rabbits with glutathione S transferase-tagged Cnp1. The minichromosome loss assay was performed as described previously (Takahashi et al., 1994). To construct pRep41 based-plasmids carrying the N-terminally deleted derivative of the C-terminal tagged Cnp1-green fluorescent protein (GFP) gene, the open reading frame (ORF) of the deleted derivatives of Cnp1 gene was amplified by PCR and inserted into the SalI–NotI cloning sites of pGP4110 in frame. pGP4110 (a derivative of pGP110; Saitoh et al., 2005) is a multicopy plasmid for the C-terminal GFP tagging under the control of the nmt41 promoter. Polymerase chain reaction (PCR) primer sequences used for the PCR amplification were as follows: full length, 5'-GTCGACATGGCAAAGAAATC-3'/5'-GCGGCCGCTAGCACCACGAATCC-3'; D5, 5'-GTCGACATGGCTGAGCCTGG-3'/5'-GCGGCCGCTAGCACCACGAATCC-3'; D10, 5'-GTCGACATGGATCCTATTCCAC-3'/5'-GCGGCCGCTAGCACCACGAATCC-3'; D15, 5'-GTCGACATGCCACGTAAAAAG-3'/5'-GCGGCCGCTAGCACCACGAATCC-3'; and D20, 5'-GTCGACATGTATCGTCCAGGTAC-3'/5'-GCGGCCGCTAGCACCACGAATCC-3'.
Construction of GFP-Cnp1 Strains
To construct a pBluescript KS-based plasmid carrying the N-terminal tagged GFP-Cnp1 gene (PS1), HindIII-BamHI cloning sites were created behind the first Met of the ORF of the Cnp1 gene. The ORF of the enhanced green fluorescent protein gene (Clontech Laboratories, Heidelberg, Germany) amplified by PCR by using primers containing HindIII and BamHI sites was inserted into the cloning sites in frame. A cnp1-1ts strain (SP1786, ura4 cnp1::ura4+ lys1+::cnp1-1) was transformed with PS1, and the transformants were grown at 22°C (permissive temperature for cnp1-1) on Edinburgh minimal medium (EMM)2 plates in the presence of 1 mg/ml 5-fluoroorotic acid, permitting identification of colonies where recombination between the GFP-Cnp1 gene and the disrupted Cnp1 gene at the authentic locus has led to loss of the Ura4 marker and return to uracil auxotrophy. All transformants were viable at 36°C. Replacement of the GFP-Cnp1 gene with the authentic gene in the transformants was confirmed by genomic Southern analysis. The transformants were then crossed with wild-type strain (SP171) to remove the cnp1-1 gene at the lys1 locus, and the resultant GFP-Cnp1 strain (SP1769) was used for the experiments shown in Figures 3 and 4. To generate the lys1+::GFP-Cnp1 strain (SP1468), a 2.4-kb KpnI–EcoRI fragment containing the GFP-Cnp1 gene was subcloned into the plasmid pTK2 designed for integration of the inserted DNAs into the lys1+ locus. A lys1– strain (SP91) was transformed with the resultant integration plasmid (PS2) and plated on EMM2 lacking lysine to select the integrants. The additional integration of the GFP-Cnp1 gene at the lys1+ locus in the transformants was confirmed by genomic Southern analysis.
Microscopy
For 4,6-diamidino-2-phenylindole (DAPI) staining in Figure 5, cells were cultured in EMM2 containing the appropriate supplements with or without 2 µM thiamine at 33°C. Cells were fixed in methanol at –80°C, washed with PBS, and mixed with 200 ng/ml DAPI. Images were collected using VB-6000/6010 (Keyence, Osaka, Japan) with a 100x 1.45 numerical aperture 
-Plan-FLUAR objective (Carl Zeiss, Jena, Germany). For Live cell analyses, cells were cultured in EMM2 containing the appropriate supplements and then embedded in 1.5% low melting point agarose in EMM2 with supplements on glass-bottomed dishes. Images of living cells were obtained using an ASMDW live cell imaging system microscope with a 100x 1.40 numerical aperture ACX PL Apo objective (Leica Microsystems, Wetzlar, Germany). Images collected every 0.3 µm along the z-axis were processed with the nonblind three-dimensional deconvolution algorithm by using ASMDW software. The projected images were converted to kymographs with MetaMorph software (Molecular Devices, Sunnyvale, CA). The fluorescent intensity of the centromeric GFP-dots and that of the nuclear GFP background were calculated by Image Quant IL (GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom) after the background (signals outside of the cell) titration. For the experiments shown in Figure 2, 513 videos in total were recorded and analyzed individually. Cells were classified into five cell cycle stages: stage I, unseptated binucleate cells (M/G1); stage II, septated binucleate cells (G1/S); and stages III–V, single nuclear cells <7.1 µm (III, S/early G2), 7.1–10.5 µm (IV, mid-G2), or >10.5 µm (V, late G2) in length. Because the average septated cell length of hip1 was 36% longer than that of wild-type controls, we used the following values as the criteria of G2 classification for hip1: stage III–V, single nuclear cells <9.6 µm (III), 9.6–14.3 µm (IV), or >14.3 µm (V) in length.
Chromatin Immunoprecipitation (ChIP)
For the experiments shown in Figure 1C, cells in the following genetic backgrounds (lys1+::Cnp1-GFP represents the native promoter-driven Cnp1-GFP gene additionally integrated at the lys1 locus) were cultured in YES at 26°C, and they were then shifted to 36°C for 3.5 h (cdc25-22 lys1+::Cnp1-GFP; SP1234, cdc25-22; YTP379, cdc25-22 ams2 lys1+::Cnp1-GFP; SP1751 and cdc25-22 ams2; YTP355 for late G2), 4.5 h (cdc10-129 lys1+::Cnp1-GFP; SP1628, cdc10-129; SP2959 for G1, and cdc22-C11 lys1+::Cnp1-GFP; SP1633, cdc22-C11; SP2962 for S) or 5.5 h (cdc10-129 ams2 lys1+::Cnp1-GFP; SP1699, cdc10-129 ams2; SP2960 for G1, and cdc22-C11 ams2 lys1+::Cnp1-GFP; SP1704, cdc22-C11 ams2; SP2963 for S). Wild-type (SP91), lys1+::Cnp1-GFP (SP92), ams2 (YTP155), and ams2 lys1+::Cnp1-GFP (SP75) cells were cultured in YES at 33°C for asynchronous samples, and then the early G2 cells were prepared by centrifugal elutriation (Avanti HP-20XP, JE-5.0 elutriation rotor; Beckman Coulter, Fullerton, CA). ChIP assays were performed using anti-GFP monoclonal antibody (mAb) (Roche, Indianapolis, IN) and Dynabeads anti-mouse immunoglobulin (Ig)G (Dynal Biotech, Oslo, Norway), or anti-Cnp1 polyclonal antibody and Dynabeads anti-rabbit IgG (Dynal Biotech) as described previously (Takayama and Takahashi, 2007). The DNA samples prepared from 3% formaldehyde-fixed chromatin solutions or immunoprecipitated fractions were analyzed by real-time PCR by using an ABI 7000 Sequence Detection System and Power SYBR Green PCR master mix (Applied Biosystems, Foster City, CA). The ratios of ChIP signals for cnt2 in the ams2 background were quantified as intensities relative to those in the ams2+ background. In all samples, intensities of ChIP signals for otr (Saitoh et al., 1997) and act1 probes (background controls) were <3.6 and 0.2% of those for the cnt2 probe, respectively. PCR primer sequences were as follows: cnt2, 5'-AAAGCAAACAGCAGTAACCTTGTAA-3'/5'-TGCGTCCTTATATGCGGCTTA-3'; imr1, 5'-CCTTTACTGGAAAATTGTCG-3'/5'-GCTGAGGCTAAGTATCTGTT-3'; and otr (dg), 5'-CATGGAACTACGTCAGGAGTGG-3'/5'-TGCCCTGTTCACTTATCTAATTCG-3'; and act1, 5'-CTTTCTACAACGAGCTTCGTGTTG-3'/5'-GAGTCATCTTCTCACGGTTGGAT-3'.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ams2). Recently, we demonstrated that Ams2 promotes S phase-specific activation of core histone genes (Takayama and Takahashi, 2007). Periodic histone transcription may be important for appropriate Cnp1 incorporation. Surprisingly, Ams2 was shown to be dispensable for cell viability, although its depletion causes severe mislocalization of Cnp1. To clarify the behavior of Cnp1, we performed live cell analyses by using wild-type or ams2 carrying the C-terminal GFP-tagged Cnp1 gene driven by the native promoter (Figure 1, A and B, Supplemental Figure S1, and Supplemental Videos 1–4). In ams2, we found that Cnp1-GFP is located on centromeres in a cell cycle-dependent manner. Cnp1-GFP began to accumulate onto centromeres from the latter half of G2, and signal intensities peaked around late G2 and M phases. The GFP signals then disappeared after cell division. Under these conditions, S phase occurs during septum formation and cell division in both wild-type (Rustici et al., 2004) and ams2 cells (Supplemental Figure S2). Thus, Ams2 is required for appropriate centromeric localization of Cnp1-GFP from late S phase to mid-G2 phase.
Figure 1B shows the data quantifying the changes in fluorescence intensity of Cnp1-GFP throughout the cell cycle in the mixture of wild-type and ams2 in the same frame. In the image of both of the wild-type and the ams2 cell, the centromere fluorescence of Cnp1 signals seemed to rapidly drop as chromosomes segregate (Figure 1B and Supplemental Figure S1, asterisks), but it seemed to increase in fluorescence shortly thereafter. We found that the intensities decreased roughly half at the onset of M phase, which may reflect the instability of Cnp1 at (Mellone and Allshire, 2003) and/or declustering of the mitotic centromeres (Nabeshima et al., 1998). We also found that, in the ams2 cell, a short uptake of Cnp1 incorporation occurred at G1/S phase (Figure 1B and Supplemental Figure S1B, arrows). Because Cnp1 is transcribed at G1/S phase before core histone transcription (Takahashi et al., 2000) in the Ams2-independent manner (Takayama and Takahashi, 2007), the uptake may correspond to G1/S deposition of Cnp1. Full boost of Cnp1 may require the following Ams2-dependent activation of the core histone expression (Takayama and Takahashi, 2007). The signal intensity reached to the maximum from mid-G2 to late G2 phase in wild-type cell, whereas it reached just before M phase in ams2 cells.
We performed ChIP analysis to confirm that the dynamics of GFP actually reflect the association of Cnp1-GFP with the DNA of the centromere (Figure 1C). The amount of centromere-bound Cnp1-GFP at late G2 in the ams2 background was almost equivalent to that in wild-type controls, whereas those at G1, early S, and early G2 phases were reduced to 49, 27, and 5%, respectively. In the ams2 background, Cnp1-GFP was associated correctly with the central centromere region (cnt). These observations indicate that, in ams2, the centromeric Cnp1-GFP is diminished during S phase, and then it disappears. Inhibition of DNA synthesis by addition of hydroxyurea (HU) blocked the disappearance of Cnp1-GFP signals in ams2 (Supplemental Figure S3), indicating that the dissociation of Cnp1-GFP in ams2 requires completion of DNA replication. By ChIP analysis using anti-Cnp1 polyclonal antibody, we next examined the cell cycle change in the centromeric association of nontagged Cnp1 that is expressed from the authentic locus. The total amount of centromere-bound Cnp1 in ams2 relative to that in wild-type controls was reduced to 49% in early S and 46% in early G2 phase, whereas there were smaller changes in late G2 (71%) and G1 (80%) phases (Figure 1C). Therefore, Ams2 is required to prevent the reduction of Cnp1 from S phase to early G2 phase. Considering the inconsistency with the drastic change of the centromeric Cnp1-GFP, C-terminal GFP-tagging may reduce the affinity of Cnp1 with the centromeric DNA to some extent, and thus it may somewhat exaggerate the impacts on Cnp1 localization by deleting Ams2.
Two Distinct Phases of Cnp1-GFP Incorporation during the Cell Cycle: S and Late G2
The dynamic behavior of Cnp1-GFP in ams2 cells highlights substantial Ams2-independent Cnp1 localization activity during the later half of G2 (Figure 1 and Supplemental Figure S1B). There are likely to be at least two distinct cell cycle phases in which centromeric Cnp1 localization in fission yeast. To confirm that Cnp1 incorporation during G2 phase actually occurs in the presence of Ams2, we performed Cnp1 reloading assay using C-terminal GFP-tagged Cnp1ts protein (Chen et al., 2003a). The Cnp1ts-GFP gene driven by the native promoter was integrated into the genome of wild-type or ams2 cells, and the resultant cells carried both the Cnp1ts gene and the authentic gene, but only the ts protein was visible by GFP-tagging. The Cnp1ts protein with a mutation (L87Q) in the middle of the 2 helix, corresponding to the CENP-A targeting domain in humans (Black et al., 2004; Black et al., 2007), showed a temperature-dependent localization defect (Chen et al., 2003a). To examine the timing of Cnp1-GFP incorporation, we followed Cnp1ts-GFP dynamics after shifting from 36°C to a low temperature of 22°C. For comparison between wild-type and ams2, the cells were classified into five morphological categories (stages I–V; see Materials and Methods) based on the number of nuclei, the appearance of the septum and cell length. Under the condition used, the cell cycle distribution of ams2 cell culture was roughly comparable with that of wild-type controls (Figure 2C, top). Mass analyses of Cnp1 reloading assay indicated that Cnp1ts-GFP protein was incorporated around S phase in 75.4% of cases in wild-type cells (181 of 240 cells examined; Figure 2, A and C, bottom: stages II–III, Supplemental Video 5). In the remaining 24.6% of the wild-type cells, Cnp1ts-GFP accumulated at late G2 (Figure 2, B and C, bottom: stage V, Supplemental Video 6). These observations indicated that G2-loading of Cnp1-GFP is not an exceptional phenomenon occurring only in Ams2-deficient cells. Therefore, wild-type cells exhibit at least two distinct peaks of Cnp1ts-GFP incorporation across the cell cycle, at S and late G2 phases, with no Cnp1ts-GFP incorporation in M/G1 (Figure 2C, bottom: stage I) or mid-G2 (stage IV).
|
ams2, reloading of Cnp1ts-GFP around S phase was reduced to 17.7% (31 of 175 cells examined; Figure 2C, bottom: stages II and III), whereas that during late G2 increased to 82.3% (stage V). In most cases in ams2, Cnp1ts-GFP occurred as dispersed signals in the nucleus but failed to accumulate into centromeres at S phase, which was then reloaded at late G2 (Supplemental Video 7). Thus, Ams2 is critical for S phase-dependent but not -independent deposition of Cnp1-GFP onto centromeres. As shown in Figure 1C, the amount of authentic Cnp1 on centromeres in ams2 cells was reduced to approximately half of that in wild-type cells at S and early G2. This reduction may be a consequence of the failure in the centromere loading of newly synthesized Cnp1 proteins in ams2 cells in which pre-existing Cnp1 is retained on centromeres.
Hip1-dependent Repression of Cnp1 Deposition during G2 Phase
The above-mentioned observation raises questions regarding the components of the G2 deposition pathway. We examined whether the fission yeast homologue of HIRA, Hip1 (Blackwell et al., 2004), is involved in Cnp1 deposition outside S phase; Xenopus HIRA is critical for the nucleosome assembly pathway independent of DNA synthesis (Ray-Gallet et al., 2002), and human HIRA was identified as a chaperone for H3.3-nucleosome (Tagami et al., 2004), which is deposited in a replication-independent manner (Ahmad and Henikoff, 2002). We generated a Hip1-null strain, which was viable as reported previously (Blackwell et al., 2004). Although the doubling time of hip1 cells (4.4 h) was 1.5 times longer than that of wild-type cells (3.0 h) in YES at 26°C, the cell cycle progression of hip1 cells seems comparable with that of wild-type by using the five morphological categories; judging from the timing of the appearance of Mrc1 protein, the S phase in hip1 cells seemed to occur during the septum formation at same as in wild-type and ams2 (data not shown). This shows that the G2 phase starts with the normal timing in hip1 cells and the cell length distribution and the septum formation can be used as good hallmarks for the cell cycle progression in the absence of Hip1. Native promoter-driven Cnp1-GFP protein was localized at centromeres throughout the cell cycle in the hip1 background (data not shown). However, we found that Cnp1ts reloading occurred throughout G2 in cells lacking Hip1 (Figure 2C). In addition to Cnp1ts-GFP incorporation in late G2 (stage V), that in the first half of G2 (stages III-IV) occurred. Thus, in wild-type cells, the replication-independent G2 deposition activity of Cnp1-GFP during the normal cell cycle is likely under active repression and could be controlled by Hip1, an S. pombe homologue of HIRA protein. Alternatively, Hip1 may promote the refolding, reassembly of Cnp1ts protein, or both to facilitate the reloading. It is also possible that the lack of Hip1 activity could just create additional loading sites for Cnp1 as a secondary consequence, not an active mechanism to exclude Cnp1. The additional deletion of Hip1 in ams2 cells did not suppress the growth retardation, but rather it showed the synthetic growth defect (data not shown), suggesting that Hip1 and Ams2 have the additional roles in cell growth besides regulating Cnp1 loading.
N-terminal GFP-tagged Cnp1 Is Functional in Wild Type but Not in ams2
Because replacement of the authentic Cnp1 gene with the C-terminal tagged Cnp1-GFP gene causes cell growth retardation (data not shown), we generated wild-type cells expressing N-terminal tagged GFP-Cnp1 at the authentic locus, and we tested that it is fully functional. This GFP-Cnp1 integrant grew normally at any temperatures tested, and its cell viability was comparable with that of wild-type controls (Figure 3A; data not shown). The sensitivity of the GFP-Cnp1 strain to the spindle poison carbendazim (CBZ) was the same as that of the wild-type control (Figure 3B). Fluorescence microscopy indicated colocalization of the signal of GFP-Cnp1 with that of the centromere-marker SpMif2-CFP throughout the cell cycle (data not shown). The results of ChIP analysis using anti-GFP antibody indicated that GFP-Cnp1 binds to the central core DNAs (cnt and imr) but not the outer repeats (otr) of the centromere as does authentic Cnp1 (Figure 3C), suggesting that the N-terminal GFP tag does not prevent the localization of Cnp1 to the centromere. However, the mitotic loss rate of a linear minichromosome, CN2, in the GFP-Cnp1 integrant cells is threefold greater than that in wild-type cells (Figure 3D). The total expression level of Cnp1-GFP protein additionally integrated at lys1 locus was severalfold greater than that of endogenous Cnp1, whereas that of GFP-Cnp1 expressed from the lys1 locus or from the native locus was comparable with that of authentic Cnp1 protein (Figure 3E). The level of expression of each tagged Cnp1 protein in the ams2 background was comparable with that in wild-type cells (Figure 3E).
|
ams2 strain. Tetrad analysis indicated that spores of ams2 expressing GFP-Cnp1 formed microcolonies but soon ceased growing. Microscopic observation of the microcolonies revealed that GFP-Cnp1 signals failed to accumulate on centromere-like dots in ams2 and that frequent chromosome missegregation occurred (Figure 4A). The intact N-terminal tail may be important for association of Cnp1 with the centromere in Ams2-deficient cells. Alternatively, it is formally possible that the cell viability is apt to be influenced by a small decrease in GFP-Cnp1 protein in the ams2 background. Additional integration of the GFP-Cnp1 gene at the lys1 locus in ams2 cells resulted in recovery of cell viability to the level in ams2 cells expressing nontagged Cnp1 (data not shown). However, the centromeric localization activity of GFP-Cnp1 was clearly defective (Figure 4A). Therefore, the growth arrest observed in ams2 spores expressing GFP-Cnp1 is not due to a dominant-negative effect of GFP-Cnp1 on Ams2-deletion, but it is presumably due to the failure of GFP-Cnp1 retention at centromeres in ams2 cells. One interesting possibility suggested by the results shown in Figures 3 and 4A is that the N-terminal tagging specifically inhibits G2 deposition but not S deposition of Cnp1. To test this possibility, we next attempted to generate wild-type and ams2 cells expressing N-terminal tagged GFP-Cnp1ts protein expressed from the lys1 locus for reloading assay. However, because the centromeric localization activity of GFP-Cnp1ts protein coexpressed with authentic Cnp1 was too low for quantitative analysis in the reloading assay, even in the wild-type background (data not shown), it was not possible to examine the above-mentioned possibility in this study.
|
ams2 cells, whereas D15 derivative exhibited the nuclear localization. Because D20 derivative lost the nuclear localization activity, the five amino acids sequences (PRKKR) may contain the nuclear localization signal (NLS), in which the basic cluster RKKR is an apparent candidate matching to the consensus sequence of NLS (Jans et al., 2000). Interestingly, although D10 derivatives were imported into the nucleus both in wild-type cells and ams2 cells, they were located on the centromeres in wild-type cells but not in ams2 cells, indicative of the function of the N-terminal tail of Cnp1 for Ams2-independent centromeric localization.
G2 Deposition Pathway Ensures Stable Chromosome Transmission in Cell Division When S Deposition Is Impaired
The above-mentioned results indicated that wild-type cells exhibit biphasic incorporation of histone variant Cnp1 during the cell cycle, which could be regulated differently. The next question is why an S. pombe cell has evolved two distinct deposition pathways. ams2 cells are viable, although Cnp1 could not retain efficiently at the centromere during S phase. Therefore, the subsequent incorporation during G2 phase could function as another opportunity for replenishing the duplicated centromere DNA with newly synthesized Cnp1 before cell division. If this hypothesis is correct, shortening of G2 phase by introduction of wee1 mutation (Hayles and Nurse, 1992) should reduce cell viability in the absence of Ams2. To test this possibility, we generated wee1-50 cells carrying the ams2+ gene under the control of a repressible promoter. When Ams2 was expressed, the cells showed the wee phenotype, but they grew normally with equal chromosome segregation in M phase (Figure 5A, ams2-ON). In contrast, the cells exhibited a high frequency of unequal chromosome segregation with the wee phenotype (Figure 5A, ams2-OFF), and they did not survive when Ams2 was repressed (Figure 5B). The centromeric localization of Cnp1-GFP was markedly reduced in ams2-deficient wee1-50 mutant in comparison with wee1-50, ams2-deficient or wild-type controls (Figure 5C). In contrast, ams2 cdc25-22 double mutant with prolonged G2 phase (Hayles and Nurse, 1992) showed much better growth than the single ams2 mutant (Figure 5D). These observations indicated that the viability of Ams2-deficient cells is influenced by the length of the G2 phase. Therefore, when S deposition is impaired, G2 phase could function as the recovery period for Cnp1 deposition, and this safety mechanism based on the biphasic incorporation could ensure high fidelity of chromosome transmission in mitosis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Holmes and Mitchell Smith, 2001; Maruyama et al., 2007). The results of the present study indicated that Ams2 GATA factor prompting core histone genes' activation at S phase (Takayama and Takahashi, 2007) ensures simultaneous Cnp1 deposition onto centromeres in fission yeast. Although depletion of Ams2 results in reduction of the binding of Cnp1 to centromeres and chromosome missegregation, increasing its dosage restores the association of Cnp1 mutant protein (Cnp1ts) with centromeres (Chen et al., 2003a). Supply of histone H4, a binding partner of Cnp1, by transformation with a multicopy plasmid partially restored the centromeric localization of Cnp1 in ams2 cells, whereas overproduction of histone H3, which competes with Cnp1 for H4 binding, is toxic to ams2 cell growth (Chen et al., 2003a; Takahashi et al., 2005). These observations indicated that appropriate regulation of histone transcription at S phase, especially up-regulation of H4 transcription, is required for efficient incorporation of Cnp1 onto centromeres. Because the supply of histone H4 in ams2 cell does not fully suppress the defect of Cnp1-GFP localization (Takahashi et al., 2005), in addition to up-regulation of histone transcription, Ams2 may have other roles in stabilizing the centromeric retention of Cnp1. For example, binding of Ams2 to the centromere may remodel the centromeric nucleosomes, which is a prerequisite for suitable incorporation of Cnp1 (Chen et al., 2003b), or other as yet unidentified transcriptional targets of Ams2 may be important for the formation/retention of Cnp1-containing nucleosomes at S phase, or a combination.
Because the intracellular levels of histone proteins are not markedly affected in ams2 (Takayama and Takahashi, 2007), significant reduction of histone gene transcription at S phase may alter the relative ratio of free histones carrying the modification pattern suitable for Cnp1 formation/retention at S phase (Hayashi et al., 2004; Fujita et al., 2007). Recently, we showed that the level of S. pombe histone mRNAs is likely to be controlled as a combination of Ams2-dependent transcriptional activation at S phase and Ams2-independent constitutive transcription throughout the cell cycle (Takayama and Takahashi, 2007). Intriguingly, this basal histone transcription is normally down-regulated by Hip1 (Blackwell et al., 2004; Takayama and Takahashi, 2007), which we showed silences the G2 deposition pathway for Cnp1 in this study. Because the net retention of Cnp1 at the centromere would be regulated by a dynamic equilibrium between incorporation and releasing of Cnp1, it is possible that Ams2 and Hip1 are playing roles in the equilibrium but not as loading factors per se.
Dynamics of Cnp1 Retention at the Centromere
The incorporation of newly synthesized and/or misloaded CENP-A onto the centromere and releasing of pre-existing CENP-A would affect the centromeric localization. Therefore, the steady-state distribution of Cnp1 we observed in this study should be dissected into incorporation processes (formation/stabilization of Cnp1-nucleosomes and/or accessing/loading of Cnp1 to the centromere) and releasing processes (disassembly/destabilization of Cnp1-nucleosomes and/or degradation of Cnp1 at the centromere) in the future study. Here, we demonstrated that there are at least two separate phases of Cnp1 deposition onto the centromere across the cell cycle in S. pombe: one during DNA synthesis, where newly synthesized DNA is rechromatinized after the passage of the replication fork, and a second phase that likely occurs during late G2, just before chromosome segregation in mitosis. Importantly, both deposition pathways may be mechanically distinct, because N-terminal GFP-tagging impairs the centromeric retention of Cnp1 in ams2 cells but not in wild-type cells. One plausible explanation is that N-terminal tagging inhibits Cnp1 loading onto the centromere at G2 but not S phase. Alternatively, GFP tagging may not inhibit Cnp1 deposition itself, but it may prevent formation/stabilization of the functional centromeric nucleosomes after G2 deposition. It is also formally possible that C-terminal GFP tagging specifically inhibits Cnp1 deposition from late M to G1 phase, at which stage newly synthesized GFP-CENP-A has been shown to be loaded in human cells (Jansen et al., 2007). Our results suggest that GFP tagging may mask some of the functional CENP-A incorporation/releasing pathways even though the GFP-tagged protein can be replaced successfully with the authentic gene product in the wild-type background.
In addition to G2 deposition, the quantification of the centromeric Cnp1-GFP in ams2 cells revealed that a short uptake of Cnp1 deposition in G1 and/or early S phase exists (Figure 1B and Supplemental Figure S1B, arrows). This may correspond to G1 deposition of CENP-A documented in human tissue culture cells (Jansen et al., 2007). Although we could not detect G1/early S deposition of Cnp1ts protein in our reloading assay (Figure 2), the behavior of the Cnp1ts-GFP carrying a specific mutation (L87Q) might not be accurately reflecting that of endogenous wild-type Cnp1. Because Cnp1 signals in ams2 cells seemed to disappear shortly after the incorporation in G1/S phase, there may be the active destabilization of Cnp1 in S phase; in the budding yeast, the degradation of Cse4 protein was shown to be an important determinant for the centromeric localization (Collins et al., 2004).
Evolutional Conservation of CENP-A Dynamics
The cell cycle timing of CENP-A loading seems not to be conserved evolutionarily. To date, the following have been proposed: G1 (Jansen et al., 2007) or G2 deposition (Shelby et al., 2000) in human cells, G2 deposition (Ahmad and Henikoff, 2001; Sullivan and Karpen, 2001) in Drosophila cells in culture, anaphase deposition in Drosophila syncytial nuclear division (Schuh et al., 2007), and S deposition in budding yeast (Pearson et al., 2004) and the primitive red alga (Maruyama et al., 2007). Pioneering work using human cells (Shelby et al., 2000) indicated that the incorporation of CENP-A occurs primarily during G2 phase. However, recent pulse-chase experiments of CENP-A protein by using SNAP tag technology clearly demonstrated that, in human cells, loading of newly synthesized CENP-A occurs in a discrete cell cycle window in early G1 (Jansen et al., 2007). hMis18, hMis18β, and M18BP1/hsKNL2 proteins, essential components of CENP-A loading pathways, display a pattern of centromeric localization coincident with CENP-A assembly in G1 (Fujita et al., 2007; Maddox et al., 2007). It is noteworthy that the pattern of fission yeast Mis18-GFP was reported to be diffused just before mitosis and restored as the centromere-like dots in late anaphase (Fujita et al., 2007), indicating that the fission yeast Mis18 is located at centromeres throughout most of the cell cycle, including telophase, G1, S, and almost all G2 phase. This differs from the localization pattern of human Mis18 proteins, which are associated with centromeres for only a short period in the cell cycle coincident with the timing of CENP-A deposition (Jansen et al., 2007). The prolonged centromeric localization of the Mis18 complex in fission yeast may enable Cnp1 to be loaded to centromeres throughout most of the cell cycle.
The stable maintenance of pre-existing CENP-A after passage of S phase in human cells (Jansen et al., 2007) and in Drosophila cells (Sullivan and Karpen, 2001) are also in sharp contrast with the behavior of Cse4, the budding yeast CENP-A (Pearson et al., 2004); the properties of the centromeric localization of Cse4 are rather dynamic, with most pre-existing Cse4 at centromeres replaced with noncentromeric Cse4 during S phase (Bloom, 2007). In the early Drosophila blastoderm embryo, in <2 min after photobleaching, >100% of the initial fluorescent intensity of GFP-tagged centromere identifier (CID), the Drosophila homologue of CENP-A, was shown to be recovered into centromeres in anaphase (Schuh et al., 2007). The observation suggests the dynamic behavior of centromeric histone in some situations; a complete exchange of pre-existing CID takes place in a very short period during anaphase. Even in the same species, CENP-A may behave differently according to developmental stages or specific environmental conditions.
Physiological Significance of Biphasic Incorporation of Cnp1
Because the wee1 mutant with a shortened G2 phase showed no apparent defects in Cnp1 localization (Figure 5C), S deposition seems to be the primary pathway for loading in S. pombe, and G2 deposition has likely evolved as a secondary pathway to increase fidelity in this organism. Interestingly, GFP tagging or a partial deletion of the N-terminal tail prevented the centromeric association of Cnp1 protein in ams2 but not in wild type (Figure 4), suggesting that Cnp1 may use its N-terminal tail for correct centromere targeting and/or formation of stable centromeric nucleosomes at G2 phase. The normal cell cycle progression occurred in wild-type cells expressing the N-terminal tagged GFP-Cnp1 gene (Figure 3), suggesting that the G2 deposition pathway may not be essential when S deposition is functional. The second gap phase after DNA replication may act as a "rescue" period for kinetochore reassembly before cell division when an authentic centromere has been damaged or deregulated. What are the components in the G2 deposition pathway is a significant question to be answered. The increased chromosome missegregation in ams2 mis6 ts double mutant cells suggested that Mis6–Sim4 centromere subcomplex may be involved in the G2 loading pathway and the S phase pathway (Takahashi et al., 2005). We speculate that both Mis6 and Mis18 complexes are involved in two loading pathways, because Cnp1 signals completely disappear in the mutants (Takahashi et al., 2000; Hayashi et al., 2004; Fujita et al., 2007).
This is the first in vivo report that the replication-uncoupled deposition of Cnp1 occurs under physiological conditions in fission yeast, along with a description of its functional significance in chromosome segregation. Although the existence of multiple CENP-A loading pathways was shown to be unlikely in transformed human cell lines (Jansen et al., 2007), further studies are required to determine whether the salvage pathway(s) exists in nontransformed normal cells in higher eukaryotes. The G2 deposition pathway of CENP-A may provide an evolutionarily conserved safeguard for the maintenance of genome integrity, and it may assist in preventing aneuploid formation during the cell cycle in nontransformed normal cells.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
Address correspondence to: Kohta Takahashi (takahash{at}lsi.kurume-u.ac.jp)
Abbreviations used: ChIP, chromatin immunoprecipitation.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ahmad, K., and Henikoff, S. (2002). The histone variant H3.3 marks active chromatin by replication-independent nucleosome assembly. Mol. Cell 9, 1191–1200.[CrossRef][Medline]
Amor, D. J., and Choo, K. H. (2002). Neocentromeres: role in human disease, evolution, and centromere study. Am. J. Hum. Genet 71, 695–714.[CrossRef][Medline]
Black, B. E., Foltz, D. R., Chakravarthy, S., Luger, K., Woods, V. L., Jr, and Cleveland, D. W. (2004). Structural determinants for generating centromeric chromatin. Nature 430, 578–582.[CrossRef][Medline]
Black, B. E., Jansen, L. E., Maddox, P. S., Foltz, D. R., Desai, A. B., Shah, J. V., and Cleveland, D. W. (2007). Centromere identity maintained by nucleosomes assembled with histone H3 containing the CENP-A targeting domain. Mol. Cell 25, 309–322.[CrossRef][Medline]
Blackwell, C., Martin, K. A., Greenall, A., Pidoux, A., Allshire, R. C., and Whitehall, S. K. (2004). The Schizosaccharomyces pombe HIRA-like protein Hip1 is required for the periodic expression of histone genes and contributes to the function of complex centromeres. Mol. Cell. Biol 24, 4309–4320.
Bloom, K. (2007). Centromere dynamics. Curr. Opin. Genet. Dev 17, 151–156.[CrossRef][Medline]
Chen, E. S., Saitoh, S., Yanagida, M., and Takahashi, K. (2003a). A cell cycle-regulated GATA factor promotes centromeric localization of CENP-A in fission yeast. Mol. Cell 11, 175–187.[CrossRef][Medline]
Chen, E. S., Yanagida, M., and Takahashi, K. (2003b). Does a GATA factor make the bed for centromeric nucleosomes? Cell Cycle 2, 277–278.[Medline]
Cleveland, D. W., Mao, Y., and Sullivan, K. F. (2003). Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell 112, 407–421.[CrossRef][Medline]
Collins, K. A., Furuyama, S., and Biggins, S. (2004). Proteolysis contributes to the exclusive centromere localization of the yeast Cse4/CENP-A histone H3 variant. Curr. Biol 14, 1968–1972.[CrossRef][Medline]
Fujita, Y., Hayashi, T., Kiyomitsu, T., Toyoda, Y., Kokubu, A., Obuse, C., and Yanagida, M. (2007). Priming of centromere for CENP-A recruitment by human hMis18alpha, hMis18beta, and M18BP1. Dev. Cell 12, 17–30.[CrossRef][Medline]
Hayashi, T., Fujita, Y., Iwasaki, O., Adachi, Y., Takahashi, K., and Yanagida, M. (2004). Mis16 and Mis18 are required for CENP-A loading and histone deacetylation at centromeres. Cell 118, 715–729.[CrossRef][Medline]
Hayles, J., and Nurse, P. (1992). Genetics of the fission yeast Schizosaccharomyces pombe. Annu. Rev. Genet 26, 373–402.[CrossRef][Medline]
Henikoff, S., and Dalal, Y. (2005). Centromeric chromatin: what makes it unique? Curr. Opin. Genet. Dev 15, 177–184.[CrossRef][Medline]
Holmes, S. G., and Mitchell Smith, M. (2001). Replication of minichromosomes in Saccharomyces cerevisiae is sensitive to histone gene copy number and strain ploidy. Yeast 18, 291–300.[CrossRef][Medline]
Jans, D. A., Xiao, C. Y., and Lam, M. H. (2000). Nuclear targeting signal recognition: a key control point in nuclear transport? Bioessays 22, 532–544.[CrossRef][Medline]
Jansen, L. E., Black, B. E., Foltz, D. R., and Cleveland, D. W. (2007). Propagation of centromeric chromatin requires exit from mitosis. J. Cell Biol 176, 795–805.
Karpen, G. H., and Allshire, R. C. (1997). The case for epigenetic effects on centromere identity and function. Trends Genet 13, 489–496.[CrossRef][Medline]
Maddox, P. S., Hyndman, F., Monen, J., Oegema, K., and Desai, A. (2007). Functional genomics identifies a Myb domain-containing protein family required for assembly of CENP-A chromatin. J. Cell Biol 176, 757–763.
Maruyama, S., Kuroiwa, H., Miyagishima, S. Y., Tanaka, K., and Kuroiwa, T. (2007). Centromere dynamics in the primitive red alga Cyanidioschyzon merolae. Plant J 49, 1122–1129.[Medline]
Meeks-Wagner, D., and Hartwell, L. H. (1986). Normal stoichiometry of histone dimer sets is necessary for high fidelity of mitotic chromosome transmission. Cell 44, 43–52.[CrossRef][Medline]
Mellone, B. G., and Allshire, R. C. (2003). Stretching it: putting the CEN(P-A) in centromere. Curr. Opin. Genet Dev 13, 191–198.[CrossRef][Medline]
Nabeshima, K., Nakagawa, T., Straight, A. F., Murray, A., Chikashige, Y., Yamashita, Y. M., Hiraoka, Y., and Yanagida, M. (1998). Dynamics of centromeres during metaphase-anaphase transition in fission yeast: Dis1 is implicated in force balance in metaphase bipolar spindle. Mol. Biol. Cell 9, 3211–3225.
Natsume, R., Eitoku, M., Akai, Y., Sano, N., Horikoshi, M., and Senda, T. (2007). Structure and function of the histone chaperone CIA/ASF1 complexed with histones H3 and H4. Nature 446, 338–341.[CrossRef][Medline]
Pearson, C. G., Yeh, E., Gardner, M., Odde, D., Salmon, E. D., and Bloom, K. (2004). Stable kinetochore-microtubule attachment constrains centromere positioning in metaphase. Curr. Biol 14, 1962–1967.[CrossRef][Medline]
Pidoux, A. L., Richardson, W., and Allshire, R. C. (2003). Sim 4, a novel fission yeast kinetochore protein required for centromeric silencing and chromosome segregation. J. Cell Biol 161, 295–307.
Ray-Gallet, D., Quivy, J. P., Scamps, C., Martini, E. M., Lipinski, M., and Almouzni, G. (2002). HIRA is critical for a nucleosome assembly pathway independent of DNA synthesis. Mol. Cell 9, 1091–1100.[CrossRef][Medline]
Rustici, G., Mata, J., Kivinen, K., Lio, P., Penkett, C. J., Burns, G., Hayles, J., Brazma, A., Nurse, P., and Bahler, J. (2004). Periodic gene expression program of the fission yeast cell cycle. Nat. Genet 36, 809–817.[CrossRef][Medline]
Saitoh, S., Ishii, K., Kobayashi, Y., and Takahashi, K. (2005). Spindle checkpoint signaling requires the mis6 kinetochore subcomplex, which interacts with mad2 and mitotic spindles. Mol. Biol. Cell 16, 3666–3677.
Saitoh, S., Takahashi, K., and Yanagida, M. (1997). Mis6, a fission yeast inner centromere protein, acts during G1/S and forms specialized chromatin required for equal segregation. Cell 90, 131–143.[CrossRef][Medline]
Schuh, M., Lehner, C. F., and Heidmann, S. (2007). Incorporation of Drosophila CID/CENP-A and CENP-C into centromeres during early embryonic anaphase. Curr. Biol 17, 237–243.[CrossRef][Medline]
Shelby, R. D., Monier, K., and Sullivan, K. F. (2000). Chromatin assembly at kinetochores is uncoupled from DNA replication. J. Cell Biol 151, 1113–1118.
Sullivan, B., and Karpen, G. (2001). Centromere identity in Drosophila is not determined in vivo by replication timing. J. Cell Biol 154, 683–690.
Tagami, H., Ray-Gallet, D., Almouzni, G., and Nakatani, Y. (2004). Histone H3.1 and H3.3 complexes mediate nucleosome assembly pathways dependent or independent of DNA synthesis. Cell 116, 51–61.[CrossRef][Medline]
Takahashi, K., Chen, E. S., and Yanagida, M. (2000). Requirement of Mis6 centromere connector for localizing a CENP-A-like protein in fission yeast. Science 288, 2215–2219.
Takahashi, K., Murakami, S., Chikashige, Y., Funabiki, H., Niwa, O., and Yanagida, M. (1992). A low copy number central sequence with strict symmetry and unusual chromatin structure in fission yeast centromere. Mol. Biol. Cell 3, 819–835.[Abstract]
Takahashi, K., Takayama, Y., Masuda, F., Kobayashi, Y., and Saitoh, S. (2005). Two distinct pathways responsible for the loading of CENP-A to centromeres in the fission yeast cell cycle. Philos. Trans. R. Soc. Lond. B Biol. Sci 360, 595–606. discussion 606–597.[CrossRef][Medline]
Takahashi, K., Yamada, H., and Yanagida, M. (1994). Fission yeast minichromosome loss mutants mis cause lethal aneuploidy and replication abnormality. Mol. Biol. Cell 5, 1145–1158.[Abstract]
Takayama, Y., and Takahashi, K. (2007). Differential regulation of repeated histone genes during the fission yeast cell cycle. Nucleic Acids Res 35, 3223–3237.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
