Plc1p is required for proper chromatin structure and activity of the kinetochore in Saccharomyces cerevisiae by facilitating recruitment of the RSC complex
- First Online:
- Cite this article as:
- Desai, P., Guha, N., Galdieri, L. et al. Mol Genet Genomics (2009) 281: 511. doi:10.1007/s00438-009-0427-9
- 106 Views
High-fidelity chromosome segregation during mitosis requires kinetochores, protein complexes that assemble on centromeric DNA and mediate chromosome attachment to spindle microtubules. In budding yeast, phosphoinositide-specific phospholipase C (Plc1p encoded by PLC1 gene) is important for function of kinetochores. Deletion of PLC1 results in alterations in chromatin structure of centromeres, reduced binding of microtubules to minichromosomes, and a higher frequency of chromosome loss. The mechanism of Plc1p’s involvement in kinetochore activity was not initially obvious; however, a testable hypothesis emerged with the discovery of the role of inositol polyphosphates (InsPs), produced by a Plc1p-dependent pathway, in the regulation of chromatin-remodeling complexes. In addition, the remodels structure of chromatin (RSC) chromatin-remodeling complex was found to associate with kinetochores and to affect centromeric chromatin structure. We report here that Plc1p and InsPs are required for recruitment of the RSC complex to kinetochores, which is important for establishing proper chromatin structure of centromeres and centromere proximal regions. Mutations in PLC1 and components of the RSC complex exhibit strong genetic interactions and display synthetic growth defect, altered nuclear morphology, and higher frequency of minichromosome loss. The results thus provide a mechanistic explanation for the previously elusive role of Plc1p and InsPs in kinetochore function.
KeywordsPhospholipase CRSC complexChromatin structureKinetochore function
Chromosome segregation during mitosis is mediated by kinetochores, specialized protein complexes that assemble at centromeric DNA and bind to spindle microtubules (Westermann et al. 2007). Our previous work demonstrated that phospholipase C (Plc1p encoded by PLC1 gene) is important for full kinetochore activity and high-fidelity chromosome segregation (Lin et al. 2000; DeLillo et al. 2003). The aim of this study was to elucidate the mechanism how Plc1p and inositol polyphosphates (InsPs) produced by Plc1p-dependent pathway affect kinetochore activity and high-fidelity transmission of genetic information.
Whereas in animal cells the centromeres span several megabases, have complex organization, and bind several microtubules, in Saccharomyces cerevisiae the centromere (CEN) is 125-bp long and engages only a single microtubule (Winey et al. 1995). CEN of S. cerevisiae includes three conserved elements, termed CDEI, CDEII, and CDEIII (Fitzgerald-Hayes et al. 1982; Hieter et al. 1985). CDEI (8 bp) and CDEIII (26 bp), highly conserved palindromic sequences, are separated by A- and T-rich CDEII (78–87 bp; Hegemann and Fleig 1993). Extensive mutational analysis has shown that complete deletion of CDEI or sequence alterations or partial deletions of CDEII can reduce but not entirely inactivate centromere function (Hegemann and Fleig 1993; Cumberledge and Carbon 1987; Gaudet and Fitzgerald-Hayes 1987; Hegemann et al. 1988). However, point mutations in the central-CCG of CDEIII completely inactivate the centromere (McGrew et al. 1986; Ng and Carbon 1987).
Cheeseman et al. (2002) proposed classification of individual kinetochore proteins and protein complexes according to whether they function at the interface with centromeric DNA (inner kinetochore proteins), at the interface with spindle microtubules (outer kinetochore proteins), or at the interface between the inner and outer kinetochore proteins (central kinetochore proteins). CBF3, a four-protein inner kinetochore complex that binds the CDEIII sequence is absolutely essential for kinetochore assembly and function in vivo (Lechner and Carbon 1991; Doheny et al. 1993). In the absence of CBF3, the association of all known kinetochore proteins with the centromere is disrupted (Ortiz et al. 1999). Kinetochore function requires also correctly configured chromatin structure. Specific mutations in histone H4 or H2A alter chromatin structure of the centromeric locus and affect kinetochore function (Smith et al. 1996; Pinto and Winston 2000). Mutations in the histone H2A variant Htz1p as well as histone acetyltransferase NuA4 also result in chromosome missegregation (Krogan et al. 2004). Chromatin-remodeling complex remodels structure of chromatin (RSC) and chromatin assembly factor I (CAF-I) together with histone regulatory (HIR) proteins are required for building centromeric chromatin and kinetochore activity (Tsuchiya et al. 1998; Hsu et al. 2003; Sharp et al. 2002). In human cells, the SWI/SNF-B chromatin-remodeling complex related to yeast RSC localizes at kinetochores during mitosis (Xue et al. 2000).
In yeast, RSC is a 15 subunit complex that belongs to the Swi/Snf family of ATP dependent chromatin-remodeling complexes. Sth1p and Sfh1p in RSC are homologs of Swi2p/Snf2p and Snf5p, respectively, of Swi/Snf complex. RSC is almost tenfold more abundant than Swi/Snf complex and is essential for the mitotic growth (Cairns et al. 1996). Importantly, the catalytic and structural subunits of RSC, Sth1p and Sfh1p, respectively, are required for maintaining the structure of the centromere and the flanking chromatin that supports proper kinetochore function (Hsu et al. 2003). In addition, Rsc2p, a nonessential subunit of the RSC complex, is required for high-fidelity chromosome segregation as well as chromosome cohesion (Baetz et al. 2004). Several components of RSC demonstrate synthetic phenotypes with kinetochore genes CSE4,MIF2, NDC10, and CTF13 (Hsu et al. 2003).
Inositol polyphosphates were shown to regulate the activity of chromatin-remodeling complexes in vivo and in vitro (Shen et al. 2003; Steger et al. 2003). The induction of the phosphate-responsive PHO5 gene, chromatin remodeling of its promoter, as well as recruitment of Swi/Snf and Ino80 chromatin-remodeling complexes require InsPs (Steger et al. 2003). InsPs are produced by a pathway that includes Plc1p and four inositol polyphosphate kinases (Ipk2p/Arg82p, Ipk1p, Kcs1p, and Vip1p) that convert InsP3 into InsP4, InsP5, InsP6, and inositol pyrophosphates PP-InsP5 and PP2-InsP4 (York et al. 1999, 2005; Odom et al. 2000; Saiardi et al. 2004, 2005; Lee et al. 2007; Mulugu et al. 2007). Our previous work demonstrated that deletion of PLC1 causes higher frequency of minichromosome loss, mild mitotic delay, and reduced binding of minichromosomes to microtubules (Lin et al. 2000). In addition, plc1Δ mutation displays strong genetic interactions with components of the inner kinetochore and activates mitotic checkpoint (DeLillo et al. 2003). These results implicated InsPs as being important for full kinetochore activity, possibly by modulating centromeric chromatin structure.
In this study we wanted to determine the molecular mechanism of InsPs function at the kinetochore. This mechanism was not initially obvious; however, a testable hypothesis emerged with the discovery of the role of InsPs in the regulation of chromatin-remodeling complexes (Shen et al. 2003; Steger et al. 2003). In addition, the RSC chromatin-remodeling complex was found to associate with kinetochores and to affect centromeric chromatin structure (Tsuchiya et al. 1998; Hsu et al. 2003). In human cells, the SWI/SNF-B chromatin-remodeling complex, which is related to yeast RSC, localizes to kinetochores during mitosis (Xue et al. 2000), raising the possibility that regulation of centromeric chromatin structure by InsPs and chromatin-remodeling complexes is evolutionarily conserved. In this paper, we provide a mechanistic explanation for the role of Plc1p in kinetochore function and demonstrate that Plc1p and InsPs are required for recruitment of the RSC complex to kinetochores, which is required for establishing proper centromeric chromatin structure and kinetochore activity.
Materials and methods
Strains and media
Yeast strains used in this study
MATa ade2-1 his3-11,15 leu2-3,112 trp1-1 ura3-1ssd1-d2 can1-100
MATα ade2-1 his3-11,15 leu2-3,112 trp1-1 ura3-1ssd1-d2 can1-100
Lin et al. (2000)
DeLillo et al. (2003)
York et al. (1999)
Odom et al. (2000)
York et al. (2005)
Nguyen et al. (2005)
W303-1α sjl1::URA3 sjl2::HIS3
Nguyen et al. (2005)
MATα sth1-2ts his3-Δ200 ura3-52 lys2-801
Du et al. (1998)
Hsu et al. (2003)
Hsu et al. (2003)
Hsu et al. (2003)
W303-1a plc1::URA3 sth1-2ts
W303-1a plc1::URA3 sfh1-1::HIS3
W303-1a plc1::URA3 rsc2::HIS3
W303-1a plc1::URA3 rsc1::HIS3
MATa ura3-52 trp1Δ63 his3Δ200 leu2::PET56 STH1-9myc::TRP1
Ng et al. (2002)
Kobor et al. (2004)
Kobor et al. (2004)
MATa ura3-52 trp1Δ63 his3Δ200 leu2::PET56 RSC1-9myc::TRP1
Ng et al. (2002)
W303-1a plc1::URA3 RSC1-9myc::TRP1
MATa ura3-52 trp1Δ63 his3Δ200 leu2::PET56 RSC2-9myc::TRP1
Ng et al. (2002)
W303-1a plc1::URA3 RSC2-9myc::TRP1
MATa ura3-52 trp1Δ63 his3Δ200 leu2::PET56 RSC3-9myc::TRP1
Ng et al. (2002)
W303-1a plc1::URA3 RSC3-9myc::TRP1
MATa ura3-52 trp1Δ63 his3Δ200 leu2::PET56 RSC8-9myc::TRP1
Ng et al. (2002)
W303-1 plc1::URA3 RSC8-9myc::TRP1
W303-1a plc1::URA3 STH1-9myc::TRP1
W303-1a plc1::URA3 HTZ1-3HA::HIS5
W303-1a plc1::URA3 SWR1-3HA::KanMX6
W303-1α ipk2::HIS3 STH1-9myc::TRP1
W303-1α kcs1::HIS3 STH1-9myc::TRP1
W303-1a cse4:::KanMX + pCSE4-HA-TRP
Sharp et al. (2002)
W303-1 plc1::URA3 + pCSE4-HA-TRP
W303-1 rsc2::HIS3 + pCSE4-HA-TRP
W303-1 plc1::TRP1 rsc2::HIS3 CSE4HA::URA3
W303-1α rsc2::KanMX6 CSE4-HA-TRP
Baetz et al. (2004)
Chromatin immunoprecipitation and quantitative real-time PCR analysis
In vivo chromatin crosslinking and immunoprecipitation were performed essentially as described (Geng et al. 2001; Guha et al. 2007) with several minor modifications. Briefly, yeast cells were grown in 600 ml YPD to an A600 nm = 1.0, at which point they were fixed for 15 min by the addition of formaldehyde to a concentration of 1%. Subsequently, the cells were converted to spheroplasts with zymolyase. Spheroplasts were washed in 40 ml of ice-cold TBS (25 mM Tris–HCl at pH 7.4, 137 mM NaCl) and 1 ml of FA lysis buffer (50 mM HEPES at pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 0.1% DOC, 1 mM phenylmethylsulfonyl fluoride containing protease inhibitors (Roche; Complete protease inhibitors) for each aliquot of original 50 ml culture. Finally, the spheroplasts from 50 ml of the original culture were resuspended in 200 μl of FA lysis buffer, 300 μl of glass beads were added, and then the samples were vortexed 20 times in 15 s bursts at the highest setting. The samples were pooled, and the suspension was then sonicated ten times for 10 s each to fragment chromosomal DNA to an average size ~500 bp. The suspension was centrifuged 1 h at 12,000×g and the supernatant was diluted with FA buffer to provide 1 ml aliquots of the resultant solubilized chromatin solution per immunoprecipitation and 100 μl for total input DNA. Each aliquot was precleared by adding 50 μl of 50% protein A/G-agarose slurry (Santa Cruz Biotechnology) and incubating 30 min at 4°C with gentle rocking. Beads were then harvested by centrifugation and the supernatant was incubated with 100 μl of 25% protein A/G slurry, which had been previously incubated for 8 h with 6 μg of antibody (anti-myc polyclonal antibody A-14, anti-HA monoclonal antibody F7 from Santa Cruz Biotechnology, or anti-histone H4 antibody from Cell Signaling). Beads were then harvested, washed, and the DNA released and extracted as described. Total input DNA and coimmunoprecipitated DNA was then analyzed by real-time PCR (25 μl reaction mixture) using the iQ SYBR Green Supermix and the Bio-Rad MyIQ Single Color Real-Time PCR Detection System (Bio-Rad). Each PCR reaction mixture was used to detect the presence of a protein at a particular locus. The MyIQ software generates a threshold count for each reaction mixture, which can be used to determine the enrichment of a protein at a given locus. Each immunoprecipitation was performed at least three times using different chromatin samples and the occupancy was calculated using the POL1 coding sequence as a negative control and corrected for the efficiency of the primers. The levels of the tagged proteins used in this work were identical in wild-type and plc1Δ cells as determined by western blotting. Primers used for real-time PCR analysis are as follows: CEN3 (5′-GCGATCAGCGCCAAACAATATGG-3′ and 5′-AAACTTCCACCAGTAAACGTTTC-3′), CEN4 (5′-GAGCCAGAAATAGTAACTTTTGCC-3′ and 5′-CAGTTTCTTTTTCTTGAGCAGG-3′), POL1 (5′-TCCTGACAAAGAAGGCAATAGAAG-3′ and 5′-TAAAACACCCTGATCCACCTCTG-3′).
Nucleosome-scanning analysis was performed essentially as described (Sekinger et al. 2005) with several minor modifications. Yeast cells were grown in 200 ml YPD to an A600nm = 1.5. The cells were converted to spheroplasts with zymolyase and the spheroplasts were washed and resuspended in 1 ml of ice-cold SPC buffer (1 M sorbitol, 20 mM PIPES, 0.1 mM CaCl2) and then treated with 9% Ficoll (Sharp et al. 2002). Aliquots (200 μl) of the nuclear fraction were treated with 0, 4, 8, 10, and 12 U/ml of MNase (Worthington) for 10 min at 37°C. The reaction was terminated with 1/10 volume of 250 mM EDTA and 5% SDS, phenol–choloroform extracted, ethanol precipitated, and RNase treated. Digested DNA was run on a 2% agarose gel and the reaction that yielded a predominantly mononucleosomal DNA was scaled up 15 times and processed as described above. The mononucleosome-sized (140–220 bp) fragment from the scaled-up reaction was gel purified and used for quantitative real-time PCR analysis with a set of overlapping primer pairs, each of which generates 100 ± 8 bp PCR product. The primers were located in 30 ± 10 bp intervals and the PCR efficiency for each primer pair was normalized with purified genomic DNA. Quantitative PCR analysis was performed by using real-time PCR (25 μl reaction mixture) using the iQ SYBR Green Supermix and the Bio-Rad MyIQ Single Color Real-Time PCR Detection System. The nucleosomal DNA enrichment level of a given DNA region was calculated as the ratio between the amounts of PCR product obtained from the purified mononucleosomal DNA and the genomic DNA. The results were normalized with the SUC2 TATA control promoter region that was arbitrarily set to 1, and all other in vivo nucleosomal DNA enrichment values are presented relative to this standard. Sequences of oligonucleotides used for nucleosome-scanning assay are available as Supplementary Material.
Plc1p is required for the recruitment of the RSC complex at the kinetochore
To eliminate the possibility that the decreased recruitment of Sth1p to CEN3 and CEN4 is an indirect consequence of a slow growth of plc1Δ cells, we determined recruitment of Sth1p to CEN3 and CEN4 also in pik1ts and sjl1Δsjl2Δ cells. PIK1 encodes essential phosphatidylinositol 4-kinase that plays a role in membrane trafficking (Strahl and Thorner 2007). Generation time of pik1ts cells is 4.5 h (Nguyen et al. 2005), which is very similar to the generation time of 5.0 h for plc1Δ cells (Lin et al. 2000). However, recruitment of Sth1p to CEN3 and CEN4 is not affected in pik1ts cells (Fig. 1a, b). SJL1 and SJL2 encode phosphatidylinositol polyphosphate 5-phosphatases, highly homologous to the mammalian synaptic vesicle-associated phosphatidylinositol polyphosphate 5-phosphatase, synaptojanin (Strahl and Thorner 2007). Again, slower growth of sjl1Δsjl2Δ strain (Singer-Kruger et al. 1998; Nguyen et al. 2005) does not affect the recruitment of Sth1p to CEN3 and CEN4 (Fig. 1a, b). To see if Plc1p is required for recruitment of other RSC subunits, we determined occupancy of nonessential subunits Rsc1p and Rsc2p and essential subunits Rsc3p and Rsc8p to CEN3 (Fig. 1c) and CEN4 (Fig. 1d). Similarly to Sth1p, plc1Δ cells display a defect in recruitment of these subunits to both CEN3 and CEN4. These results suggest that Plc1p affects the recruitment of RSC as a whole complex and does not differentially affect individual subunits. The decreased recruitment of the RSC complex to CEN3 and CEN4 in plc1Δ cells is not likely to be due to the lower protein level of RSC subunits, as indicated by Western blot analysis of Sth1p and Rsc8p (Fig. 1e). Cumulatively, these results show that the absence of Plc1p and InsPs causes defect in RSC recruitment to the centromere; this may in turn result in a structural alteration of centromeric chromatin, causing functional defect of the kinetochore and decreased fidelity of chromosome segregation.
Different InsPs do not appear to have completely overlapping roles in the regulation of recruitment and activity of chromatin-remodeling complexes in vivo and in vitro (Shen et al. 2003; Steger et al. 2003). The induction of the phosphate-responsive PHO5 gene, chromatin-remodeling of its promoter, as well as recruitment of Swi/Snf and Ino80 chromatin-remodeling complexes require InsP4 and InsP5 (Steger et al. 2003). In vitro, nucleosome mobilization by the yeast Swi/Snf complex is stimulated by InsP4 and InsP5, while InsP6 inhibits nucleosome mobilization by yeast Isw2 and Ino80 complexes and the Drosophila NURF complex (Shen et al. 2003). To identify which of the InsPs (InsP3, InsP4 + InsP5, InsP6, or PP-InsP5) is/are required for the recruitment of the RSC complex to the centromeres, we performed similar ChIP experiments as described above also in ipk2Δ, ipk1Δ, and kcs1Δ strains. The results show that the defect in Sth1p recruitment at CEN3 in ipk2Δ, ipk1Δ, and kcs1Δ cells is less severe than in plc1Δ cells (Fig. 1f), suggesting that it is the complete lack of all InsPs in plc1Δ cells rather than the synthesis of one specific InsPs that is responsible for the defect in Sth1p recruitment.
In addition to centromeric loci, where the RSC complex resides throughout the cell cycle, Sth1p localizes to cohesin binding sites in a cell cycle dependent manner, facilitating the loading of cohesins onto chromosome arms (Huang et al. 2004). However, we did not find any significant difference in Sth1p occupancy between wild-type and plc1Δ cells at two chromosomal cohesin binding sites, chV-549.7 kb and chV-556 kb (Tanaka et al. 1999), which are 549.7 and 556 kb, respectively, from the left telomere of chromosome V (data not shown). This result suggests that InsPs are not universally required at all chromosomal loci for the occupancy of the RSC complex. It also appears that InsPs do not affect integrity of the RSC complex but are required for the interaction of the RSC complex with specific chromosomal loci.
plc1Δ Genetically interacts with rsc mutations
Structure of chromatin flanking the centromere is affected by rsc and plc1Δ mutations
PLC1 was initially identified in a genetic screen for mutants that showed an increased frequency of chromosome missegregation (Payne and Fitzgerald-Hayes 1993). The plc1-1 mutant isolated in this screen displayed 32-fold increase in chromosome missegregation events. However, since plc1-1 cells appeared to have normal nuclei and spindle morphologies, and were not supersensitive to the microtubule-destabilizing drug benomyl, the authors reasoned that it was unlikely that the chromosome missegregation phenotype results from a defect in the components of the mitotic segregation apparatus. Rather, they concluded that plc1-1 affects chromosome segregation indirectly, possibly by interfering with proper timing of the cell cycle or by altering intracellular calcium concentrations. We have subsequently shown that Plc1p is important for kinetochore activity (Lin et al. 2000) and plc1Δ cells display alterations in chromatin structure in the core centromere (DeLillo et al. 2003). Despite the fact that the existence of nuclear phosphatidylinositol pathway is now well documented (Bunce et al. 2006; York 2006), it was not clear how Plc1p and InsPs affect chromatin structure of centromeres. However, two important discoveries allowed us to speculate about a possible mechanism. First, InsPs were shown to be required for recruitment of chromatin-remodeling complexes to specific DNA loci (Steger et al. 2003). Second, the RSC complex was demonstrated to be recruited to centromeres and to be required for kinetochore activity (Hsu et al. 2003). The major objective of this study was to test a hypothesis that Plc1p and InsPs are required for kinetochore activity by facilitating recruitment of the RSC complex to the kinetochore which is an important step for assembly of the kinetochore and the flanking nucleosomes. Our results show that Plc1p and InsPs are required for the recruitment of the RSC complex to the centromeric region. Since the RSC complex is essential, we could not test a synergy between plc1Δ mutation and complete inactivation of RSC that would indicate whether Plc1p and InsPs function only through RSC or whether they affect centromeric chromatin structure and kinetochore activity also by a RSC-independent mechanism.
Two models of kinetochore assembly and architecture have been proposed (Espelin et al. 2003; McAinsh et al. 2006). In the nucleosome-centric model, the kinetochore is organized around specialized centromeric nucleosome that assembles on the CDEI and CDEII region and contains histone H3 variant Cse4p (Cheeseman et al. 2002; Meluh et al. 1998). In the CBF3-centric model the CDE region is not organized in a nucleosome. The CBF3 complex binds to CDEIII region, while additional molecules of Ndc10p bind to CDEII (Espelin et al. 2003). In this model, Cse4p is assembled into nucleosomes flanking CDE region. Since ChIP assays have limited resolution of more than 200 bp and the CDE region is only ~125 bp, it is almost impossible to conclusively determine where Cse4p and Ndc10p bind. Recent reports identified non-histone protein Scm3p to be required for the recruitment of Cse4p to centromeres and for normal kinetochore assembly. It appears that Scm3p, Cse4p, and histone H4 form nucleosome-like structure that assembles at the centromere (Mizuguchi et al. 2007; Stoler et al. 2007; Camahort et al. 2007).
Our data (Fig. 3) and previously published results (Hsu et al. 2003; Tsuchiya et al. 1998) show that mutations in the RSC complex affect both centromeric and flanking chromatin structure. Our data are equally compatible with both models of kinetochore architecture. Both models agree on the importance of the ordered array of nucleosomes flanking the core centromere. Our data and results published previously (Hsu et al. 2003) suggest that the structure of these nucleosomes is affected by the RSC complex. The specialized nucleosome that assembles at the CDEI and CDEII region according to the nucleosome-centric model is also a likely substrate of the RSC complex. However, according to the CBF3-centric model, this region is not organized in a nucleosome and thus one would expect that the structure of this region would not be affected by the RSC complex, in apparent disagreement with previously published results (Hsu et al. 2003). However, we cannot exclude a possibility that the RSC complex is able to reconfigure non-nucleosomal chromatin structures. Mutually non-exclusive is a possibility that the Plc1p- and RSC-dependent assembly of the flanking nucleosomes facilitates correct deposition of kinetochore components and is important for the structural integrity of the core centromeric region. Similar situation was found at origins of DNA replication, where nucleosomes flanking the site for the origin recognition complex affect the efficiency of replication initiation (Simpson 1990; Lipford and Bell 2001).
The nucleosome-scanning data show that neither plc1Δ nor rsc mutations affect positioning of the flanking nucleosomes but they do affect the nucleosome assembly or stability as reflected by the height of the peaks corresponding to nucleosomes A and B (Fig. 3). Nucleosomes are diverse and dynamic structures with varied protein composition as well as covalent modification states (for review, see Cairns 2005). We hypothesize that the RSC complex is involved directly or indirectly in recruitment and assembly of histone and non-histone proteins characteristic for fully active kinetochore and flanking nucleosomes. This model is in agreement with decreased histone H4 occupancy at centromere-flanking regions (Fig. 4a). RSC complex was reported not to be required for the deposition of Cse4p to the centromeric chromatin (Baetz et al. 2004), in apparent disagreement with our results that show decrease in the recruitment of Cse4p to the centromere in rsc2Δ cells (Fig. 4b). However, close examination of the reported results (Fig. 2 in Baetz et al. 2004) suggests that the recruitment of Cse4p is somewhat decreased in rsc2Δ cells. The above study also used cells arrested at 37°C for 3 h which may affect the results. In addition, we used real-time PCR analysis of the immunoprecipitated DNA that allows more accurate quantification.
The requirement for chromatin-remodeling complex for proper assembly of the centromeric chromatin is underscored by the finding that the centromeric chromatin undergoes reversible oscillations caused by tension created by the spindle (He et al. 2000; Pearson et al. 2001). It was calculated that this “stretching” of centromeric DNA would result in loss of the existing nucleosomes. The centromeric chromatin has to be subsequently reassembled and we speculate that this process requires continuous presence of the RSC complex and other factors at the centromere. The need for chromatin remodeling and assembly factors at the centromere is also demonstrated by the requirement of CAF-I and HIR proteins for assembly of active kinetochore (Sharp et al. 2002).
Inositol polyphosphates and phosphatidylinositols facilitate recruitment of chromatin-remodeling complexes to specific chromosomal sites in both yeast and mammalian cells by a mechanism that is not well understood (Steger et al. 2003; Zhao et al. 1998). Our results suggest that Plc1p and InsPs affect chromatin structure of centromeres and kinetochore function by facilitating recruitment of the RSC complex. However, it is also possible that Plc1p and InsPs function partly by a RSC-independent mechanism. We envision two possible, mutually non-exclusive mechanisms for the role of InsPs in recruitment of the RSC complex to kinetochores. InsPs may bind to other kinetochore or chromatin protein(s) which increases affinity of these sites for RSC binding. Alternatively, one or more subunits of the RSC complex could bind InsPs and this binding perhaps makes the RSC complex more competent for interaction with kinetochores and possibly other chromatin sites. Further experiments should distinguish between these two possibilities and should identify mechanism of InsPs-facilitated recruitment of chromatin-remodeling complexes to specific chromosomal loci.
We thank Drs Baetz, Hieter, Laurent, Kaufman, Measday, Rine, Stillman, Struhl, Wente, and York for strains and plasmids and members of Vancura laboratory and Dr Vancurova for helpful comments. This work was supported by grants from the National Institutes of Health (GM076075) to A. Vancura.