CSE4 Genetically Interacts With the Saccharomyces cerevisiae Centromere DNA Elements CDE I and CDE II but Not CDE III: Implications for the Path of the Centromere DNA Around a Cse4p Variant Nucleosome

Corresponding author: Molly Fitzgerald-Hayes, Department of Biochemistry and Molecular Biology, University of Massachusetts, Amherst, MA 01003., mollyfh{at}biochem.umass.edu (E-mail)

Communicating editor: M. LICHTEN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Each Saccharomyces cerevisiae chromosome contains a single centromere composed of three conserved DNA elements, CDE I, II, and III. The histone H3 variant, Cse4p, is an essential component of the S. cerevisiae centromere and is thought to replace H3 in specialized nucleosomes at the yeast centromere. To investigate the genetic interactions between Cse4p and centromere DNA, we measured the chromosome loss rates exhibited by cse4 cen3 double-mutant cells that express mutant Cse4 proteins and carry chromosomes containing mutant centromere DNA (cen3). When compared to loss rates for cells carrying the same cen3 DNA mutants but expressing wild-type Cse4p, we found that mutations throughout the Cse4p histone-fold domain caused surprisingly large increases in the loss of chromosomes carrying CDE I or CDE II mutant centromeres, but had no effect on chromosomes with CDE III mutant centromeres. Our genetic evidence is consistent with direct interactions between Cse4p and the CDE I-CDE II region of the centromere DNA. On the basis of these and other results from genetic, biochemical, and structural studies, we propose a model that best describes the path of the centromere DNA around a specialized Cse4p-nucleosome.

CSE4P and CENP-A are variant histone H3 proteins involved in centromere structure and function in Saccharomyces cerevisiae and mammals, respectively. Cse4p and CENP-A have C-terminal histone-fold domains that are >60% identical to the histone-fold domain of H3 (SULLIVAN et al. 1994 ; STOLER et al. 1995 ). Amino acids throughout the histone-fold domains of Cse4p and CENP-A participate in a cooperative manner to specify centromere structure and function, indicating that the two proteins mediate formation of centromeric chromatin by a common mechanism despite the differences in centromere DNA (SHELBY et al. 1997 ; KEITH et al. 1999 ). The highly divergent N termini of Cse4p, CENP-A, and H3 are localized outside the nucleosome and interact with transcription factors and chromatin remodeling factors in the case of H3 and kinetochore proteins in the case of Cse4p.

The histone-fold domain is an evolutionarily conserved protein motif shared by the four core histones and a variety of proteins involved in DNA metabolism (BAXEVANIS et al. 1995 ). The X-ray crystal structure of the core histones shows that the histone-fold domain consists of three -helical structures (helix I, II, and III) separated by ß-loop strands (loop I and II; LUGER et al. 1997 ). The histone-fold domain mediates H3-H4 and H2A-H2B dimer interactions in which the two long central helices (helix II) cross each other in an antiparallel fashion, juxtapositioning loop I of one protein with loop II of the other protein (LUGER et al. 1997 ). The dimers have three major sites, two loop I-loop II sites and a helix I-helix I site, that contact the minor groove of the DNA approximately every 10 bp. The histone-fold domain is involved in H4-H3/H3-H4 tetramer formation that is mediated through a four helix bundle interaction involving the C terminus of helix II, loop II, and helix III of each H3 molecule. The (H3-H4)₂ tetramer mediates initial contact with the DNA at the dyad axis. Two H2A-H2B dimers bind separately to the (H3-H4)₂ tetramer through four helix bundle interactions involving H2B and H4. The resulting core octamer binds 150 bp of DNA.

CENP-A is a mammalian centromere-specific H3 variant that is always associated with active centromeres and is a key component of centromeric chromatin. CENP-A copurifies with core histones (PALMER et al. 1987 ) and forms homodimers in vivo, suggesting that CENP-A replaces both H3 molecules in specialized nucleosomes (SHELBY et al. 1997 ). The histone-fold domain of CENP-A functions to localize the protein to the mammalian centromere where it binds predominantly to -satellite DNA in phased arrays of nucleosomes (VAFA and SULLIVAN 1997 ).

The yeast H3 centromere variant, Cse4p, is essential for cell viability and proper mitotic chromosome transmission (STOLER et al. 1995 ). Mutations in CSE4 can cause cell cycle arrest at G2/M and increased chromosome missegregation. The high amino acid homology among the histone-fold domains of Cse4p, H3, and CENP-A indicates that Cse4p most likely replaces H3 in a subset of yeast nucleosomes. In addition to similarities with H3 and CENP-A, several lines of evidence show that Cse4p functions as a nucleosome core histone in yeast. Biochemically, Cse4p is an integral chromatin protein with physical and chemical properties resembling histone H3 (STOLER et al. 1995 ). Mutations in the histone-fold domain of Cse4p disrupt chromatin structure at the centromere (MELUH et al. 1998 ). Furthermore, evidence also supports interactions between Cse4p and other histones. Overexpression of CSE4 suppresses hhf1-20, a mitosis-specific H4 allele that causes mitotic arrest at the nonpermissive temperature and increased chromosome missegregation (SMITH et al. 1996 ). High dosage of CSE4 also rescues the defective centromeric chromatin structure observed in hhf1-20 cells, further supporting a direct interaction between Cse4p and H4 at the centromere (MELUH et al. 1998 ).

S. cerevisiae accomplishes very high fidelity chromosome segregation using just 125 bp of centromere DNA present on each chromosome. All 16 S. cerevisiae centromeres contain three conserved centromere DNA elements (CDE), CDE I, CDE II, and CDE III (FITZGERALD-HAYES et al. 1982 ; FLEIG et al. 1995 ). CDE I is bound by a homodimer of CP1 (also called Cbf1p, Cpf1p; BRAM and KORNBERG 1987 ; BAKER et al. 1989 ; CAI and DAVIS 1989 ), but neither CDE I nor the bound CP1 protein is essential for centromere function and abolishing either one increases chromosome missegregation 10- to 30-fold (BAKER and MASISON 1990 ). Apparently, the CDE I/CP1 complex functions to enhance centromere function and promote high fidelity mitotic chromosome transmission, probably by maintaining favorable chromatin structures at the centromere. CDE III is a 26-bp DNA element containing a highly conserved central region that is absolutely essential, since a single base-pair mutation in CDE III can completely abolish centromere function (MCGREW et al. 1986 ). CDE III is bound by the CBF3 complex, which consists of three structural proteins, p110, p64, and p58, and a regulatory protein, Skp1p. The CBF3 complex, along with at least three other proteins, Ctf19p, Mcm21p, and Okp1p, are thought to mediate a connection between the centromere DNA and microtubule (ORTIZ et al. 1999 ). Recently, we have shown that a unique domain in the N terminus of Cse4p interacts directly with the Ctf19p-Mcm21p-Okp1p outer kinetochore complex (CHEN et al. 2000 ) The CDE II DNA elements from all 16 yeast chromosomes have similar characteristics, including conserved length (78–87 bp), intrinsic bends, high A + T base composition (>90%), and short repeated tracts of A or T residues. All four features are important determinants that contribute to the function of CDE II in chromosome segregation (CUMBERLEDGE and CARBON 1987 ; GAUDET and FITZGERALD-HAYES 1987 ; MURPHY et al. 1991 ).

Yeast centromere DNA is organized into a unique chromatin structure, where 160–220 bp of DNA, including CDE I, II, and III, are protected from nuclease digestion and are flanked by arrays of phased nucleosomes (SCHULMAN and BLOOM 1991 ). Histone proteins are directly involved in this unique chromatin structure, since depletion of either histone H4 or H2B renders the centromere DNA sensitive to nuclease digestion (SAUNDERS et al. 1990 ).

Current models of the yeast kinetochore propose that the yeast centromere DNA is wrapped around a Cse4p variant nucleosome. Here we present evidence that the histone-fold domain of Cse4p interacts specifically with CDE I and CDE II centromere DNA. We made a series of mutations distributed throughout the histone-fold domain of Cse4p that alter amino acids that, by analogy with H3 in the nucleosome crystal structure, contact or are adjacent to regions of the protein that directly interact with the DNA (LUGER et al. 1997 ). The cse4 histone-fold domain mutations were analyzed for synthetic chromosome loss phenotypes when combined in the same cell with chromosomes carrying mutant cen3 centromere DNA. Our results show genetic interactions between Cse4p and CDE I and CDE II, but no detectable interactions between Cse4p and CDE III DNA. We propose a model supported by current genetic and biochemical evidence that predicts the most likely position of the centromere DNA around a specialized Cse4p variant nucleosome. The implications of this novel chromatin structure on centromere function and chromosome segregation in yeast and humans are discussed.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains and plasmids:
To integrate mutant centromeres into chromosome III, cen3 mutants cloned in pJUP [3B14, BCT1 (MCGREW et al. 1986 ); CAT1, P130-3, X78, X69 (GAUDET and FITZGERALD-HAYES 1987 ); and GA/TG] or pJII (MOI, MO4, and MO4B, MURPHY et al. 1991 ) were linearized with EcoRI and used to transform KC99 or KC140 cells (Table 1). Proper integrants, verified by Southern blot analysis (data not shown), were mated with KC148 or KC151 to generate the 10 strains used in the chromosome loss assays. Mutant alleles of CSE4 were made by site-directed mutagenesis using the Clontech transformer site-directed mutagenesis kit (CLONTECH, Palo Alto, CA) following the vendor's instructions and using pCSE4HA as template DNA (STOLER et al. 1995 ). Plasmids were transformed into the chromosome loss strains and transformants were cured of pHCC4 or pC4/RS318 (YEp351 and pRS318, respectively, containing the ClaI-DraI fragment of CSE4) by isolating LEU- auxotrophs. Plasmid pCSE4HA contains the HA epitope-tagged allele of wild-type CSE4 cloned into pRS314 (CEN4/ARS/TRP1; STOLER et al. 1995 ).

Mitotic chromosome loss assays:
The rate of mitotic loss of marked copies of chromosome III was quantified using fluctuation analysis as described previously (HEGEMANN et al. 1988 ; KEITH et al. 1999 ). A pink colony was picked from nonselective color indicator plates [0.6% yeast nitrogen base without amino acids (Difco, Detroit), 0.5% casamino acids, 2% glucose, and 30 mg of uracil and 4.5 mg of adenine per liter] and grown 4–6 hr at 30° in media that selects for the cse4 mutant plasmid (TRP1) and the marked chromosome (URA3). Cultures were diluted and 50 cells were plated onto color indicator plates and incubated at 30° for 18–20 hr. Six to eight equally sized test colonies for each mutant tested were picked with a Pasteur pipette on an agar plug, resuspended in water, vortexed, and half the volume was spread onto color indicator plates (150 x 15 mm). Plates were incubated at 30° for 5 days and at 4° for 4–6 days to allow colony colors to fully develop. The total number of cells plated onto the large indicator plates represents the number of cell divisions that occurred during test colony growth and the number of red colonies represents the number of cells that lost the marked chromosome prior to plating. The total number of colonies and the total number of red colonies were counted and the values doubled to reflect the total number of cells originally picked. From the total number of colonies and the number of red colonies the median number of cells without the marked chromosome was calculated and used to determine the mean number of chromosome loss events during growth of the test colony (LEA and COULSON 1949 ). The mean number of chromosome loss events divided by the total number of cell divisions that occurred during test colony growth is the rate of chromosome loss per cell division.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Analysis of cse4-cen3 double mutants:
To investigate the functional relationship between Cse4p and centromere DNA, we employed a genetic approach in which we tested the effect of cse4 mutations on the segregation of chromosomes carrying mutations in either CDE I, CDE II, or CDE III DNA. It has been previously shown that mutant centromere proteins that have little or no effect on the function of wild-type centromeres can significantly increase the missegregation events involving chromosomes carrying mutant centromeres (MCGREW et al. 1989 ; MELUH and KOSHLAND 1995 ; STRUNNIKOV et al. 1995 ; BAKER et al. 1998 ). Such synergistic effects can result when the mutations alter regions usually involved in protein-centromere DNA interactions.

The mitotic loss rates of marked (URA3, SUP11) copies of chromosome III containing either wild-type CEN3 or mutant cen3 DNA were determined by fluctuation assays performed in cse4 null diploid cells with the lethal phenotype covered by wild-type Cse4p expressed from a plasmid (MATERIALS AND METHODS). Fluctuation assays are highly sensitive, allowing recognition of loss rates as low as two- to threefold (HEGEMANN et al. 1988 ). The loss rate for chromosomes with wild-type CEN3 was 0.17 x 10^-3 (approximately two loss events per 10,000 cell divisions; Fig 1). The cen3 centromere DNA mutations studied cause a range of increases in chromosome loss. For instance, the X78 centromere, which has a small deletion mutation in CDE II, causes only a slight increase in chromosome loss (0.60 x 10^-3), while BCT1, which has a single base-pair change in CDE III, causes dramatic chromosome loss (47.5 x 10^-3).

View larger version (58K): In this window In a new window Download PPT slide   Figure 1. Cis-acting mutations in CEN3 DNA. The sequences of the nine different cen3 mutations tested in the mitotic chromosome loss assays are shown compared to the wild-type CEN3 sequence and the centromere consensus sequence derived from all 16 budding yeast centromeres. The shaded regions indicate the base pairs altered by each mutation. The chromosome loss rates determined using fluctuation analysis are shown at the right (chromosome loss events per cell division). The loss rates of chromosome III copies containing mutant cen3 centromeres were measured in wildtype Cse4HAp cells.

cse4 mutants have little or no effect on the segregation of chromosomes with wild-type centromeres:
We made several cse4 alleles containing mutations that alter regions containing potential Cse4p-centromere DNA contact sites (Fig 2; LUGER et al. 1997 ). To preserve the nucleosome structure and limit disruption of histone-histone interactions, we replaced the Cse4p residues with the specific amino acids found at the analogous position in yeast H3. The two cse4 alleles with H3 residue substitutions in the N-helix, cse4-52 and cse4-53, overlap, as do the N-loop and helix I substitution alleles, cse4-49 and cse4-50. The cse4 loop I alleles contain a point mutation or a single amino acid deletion in cse4W-F and cse4D, respectively. Three residues in loop II/helix III are changed in cse4-56 and a single helix III amino acid is substituted in cse4-101. Immunoblot analysis shows that the cse4 mutant and wild-type proteins have identical steady-state expression levels in yeast cells containing wild-type or mutant centromeres (data not shown; KEITH et al. 1999 ).

View larger version (37K): In this window In a new window Download PPT slide   Figure 2. Cse4p histone-fold domain mutants. Residues in the histone-fold domain of Cse4p were substituted with amino acids from analogous positions in histone H3. Sequence alignment of the histone-fold domains of S. cerevisiae H3 and Cse4p proteins is shown at the top of the figure; amino acids that are identical between the two proteins are shaded. The positions of the -helices and ß-loops determined from the nucleosome X-ray crystal structure are noted above the Cse4p and H3 amino acid sequences.

We tested whether the mutant cse4 alleles affect the segregation of chromosomes carrying wild-type CEN3 DNA (Table 2). Diploid yeast shuttle strains were constructed containing a marked chromosome III copy carrying either a wild-type CEN3 or mutant cen3 centromeres and with both CSE4 genes disrupted (cse4::HIS3/cse4::HIS3) and the lethal phenotype covered with a plasmid containing CSE4. These strains were transformed with low-copy plasmids carrying a mutant cse4 allele, after which the wild-type CSE4 plasmid was removed so that the function of the mutant Cse4 protein could be studied in the absence of wild-type Cse4 protein. The chromosome loss rates per cell division for the mutant and wild-type CEN3 chromosomes were measured in cells that express either a wild-type CSE4 or mutant cse4 allele. The fold increases in chromosome loss rates caused by the mutant cse4 alleles were obtained by comparison to congenic cells expressing wild-type CSE4 (Fig 3). Notably, none of the cse4 alleles caused a significant increase in the loss of chromosomes with wild-type CEN3 DNA (1.1- to 1.9-fold; Table 2).

View larger version (30K): In this window In a new window Download PPT slide   Figure 3. The fold increases in loss rates of chromosomes with mutant centromeres in different cse4 mutant cells. The fold increases were determined by fluctuation tests (MATERIALS AND METHODS). Values for each cse4 allele are normalized in each column by comparison to the loss rate of wild-type chromosomes carrying the same centromere in CSE4HA cells.

Mutations in both CSE4 and CDE I or CDE II cause synergistic increases in chromosome loss:
The CAT1 cen3 mutation changes a highly conserved C in CDE I to TT, which severely decreases CP1 binding (BAKER et al. 1989 ) and results in a chromosome loss rate of 1.6 x 10^-3 in wild-type CSE4 cells (Fig 1). Some CAT1 CDE I, cse4 double mutants exhibited synthetically high chromosome loss rates compared to the rate for the identical cen3 mutant chromosome in wild-type CSE4 cells. Three cse4 alleles, cse4D, cse4W-F, and cse4-101, cause large increases in mitotic loss of the CAT1 chromosome (12- to 29-fold), while cse4-49 causes a moderate increase (6.8-fold) and cse4-53 had a small effect (2.9-fold; Table 2 and Fig 3). The MOI cen3 mutation contains an inversion of CDE I relative to CDE II and CDE III and causes a chromosome loss rate of 0.53 x 10^-3 (Fig 1; MURPHY et al. 1991 ). Moderate increases in loss rates of chromosomes with MOI mutant centromeres were observed for cells with cse4-49 (4.0-fold), cse4W-F (4.1-fold), and cse4-101 (2.8-fold) alleles, while the other cse4 alleles had small or no detectable effects (Table 2 and Fig 3).

Five CDE II cen3 mutations were tested that either alter the length or reduce the A + T base composition of CDE II, or both (Fig 1). The P130-3 cen3 mutation increases the length of CDE II from 84 to 130 bp, resulting in a chromosome loss rate of 1.5 x 10^-3 (Fig 1; GAUDET and FITZGERALD-HAYES 1987 ). The cse4W-F and cse4-101 alleles caused high increases in P130-3 chromosome loss (16.6- and 14.2-fold, respectively), while cse4-49 and cse4-53 caused a moderate increase (4.6-fold; Table 2 and Fig 3). Centromere mutants X78 and X69 have shortened CDE II regions that cause chromosome loss rates of 0.60 x 10^-3 and 8.24 x 10^-3, respectively (Fig 1; GAUDET and FITZGERALD-HAYES 1987 ). The cse4-49 and cse4W-F allele had a moderate effect on X78 cen3 chromosome loss (3.9- and 4.4-fold, respectively), while cse4-52, cse4-53, cse4-49, cse4-50, cse4D, cse4W-F, and cse4-101 all caused moderate increases in the loss rate of X69 cen3 chromosomes (4.3- to 7.1-fold; Table 2 and Fig 3). The MO4 and MO4B cen3 mutations maintain the length of CDE II while changing the A + T composition or the intrinsic bend of CDE II (MURPHY et al. 1991 ). The MO4 mutation reduces the A + T composition from 93 to 53% and results in a chromosome loss rate of 27 x 10^-3 (Fig 1). Alleles cse4-52, cse4-53, cse4-49, and cse4D caused moderate increases in the chromosome loss rates of MO4 cen3 chromosomes (4.2- to 5.0-fold; Table 2 and Fig 3). The MO4 cen3 chromosomes exhibit high loss rates even in wild-type CSE4 cells and were too unstable for fluctuation assays when combined with cse4 mutations that cause increases >7- to 8-fold, such as cse4W-F and cse4-101. MO4B enhances the intrinsic bend in CDE II, resulting in a chromosome loss rate of 3.9 x 10^-3 (Fig 1). Alleles cse4-49, cse4D, and cse4W-F caused moderate increases and cse4-101 caused a high increase (12.4-fold) in the loss rate of MO4B cen3 chromosomes (Table 2 and Fig 3).

Mutations in CSE4 do not increase missegregation of CDE III mutant chromosomes:
Two CDE III mutations were tested, BCT1 and GA/TG cen3. The integrity of CDE III is essential for centromere function because specific CDE III sequences are required for the recognition and assembly of the CBF3 kinetochore complex. Mutations that change bases in the central region of CDE III have drastic effects on centromere function even in CSE4 wild-type cells. For instance, the BCT1 mutation, which changes the essential central CCG triplet to TCG in CDE III (MCGREW et al. 1986 ), causes a chromosome loss rate of 47.5 x 10^-3 in CSE4 cells (Fig 1). Mutations that change bases flanking the central region, such as the GA/TG CDE III mutation, have relatively low chromosome loss rates (1.33 x 10^-3). In contrast to CDE I and CDE II, none of the cse4 mutant alleles significantly affected the loss rates of chromosomes carrying either the BCT1 or GA/TG CDE III centromere mutations (Table 2 and Fig 3). In fact, all of the CDE I and CDE II mutant centromeres we tested showed increased chromosome loss rates of up to 28-fold when combined with the various cse4 mutant alleles, while the largest increase observed for the CDE III mutants was 1.4-fold.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Cse4p has two distinct domains, an essential, unique N terminus necessary for interactions with the Ctf13p/Mcm21p/Okp1p/CBF3 complex at CDE III and the highly conserved C-terminal histone-fold domain that is thought to be involved in a unique centric nucleosome (STOLER et al. 1995 and CHEN et al. 2000 ). Our genetic results reported here indicate that the histone-fold domain of Cse4p engages in critical interactions with the CDE I and CDE II regions of the centromere DNA, but does not interact with CDE III DNA. Synthetic lethal phenotypes can occur when two proteins that usually interact are mutated and coexpressed in the same cell (HUFFAKER et al. 1987 ). The single mutants alone are partially functional and viable, but the double mutant exhibits a strongly enhanced phenotype. In a similar manner, synthetic phenotypes can be elicited by combining mutations that affect interactions involving proteins and DNA.

Increases in chromosome loss phenotypes are exhibited by cells carrying certain combinations of mutant proteins and chromosomes with mutant centromere DNA (MCGREW et al. 1989 ; MELUH and KOSHLAND 1995 ; STRUNNIKOV et al. 1995 ; BAKER et al. 1998 ). If a DNA-binding protein is mutated so that it can no longer bind to a specific site in the DNA, then a subsequent mutation in the DNA is unlikely to cause a further loss of function. However, if the first mutation confers partial function, then a second mutation altering an appropriate site in the DNA might significantly exacerbate the chromosome loss phenotype. The synthetic phenotypes between cse4 mutants and mutations in CDE I and CDE II, but not CDE III, support a model where the histone-fold domain of Cse4p associates with CDE I and CDE II DNA and wraps the centromere DNA elements into a nucleosome-like structure, with CDE III positioned in the linker DNA.

In standard nucleosomes, the H3/H4 tetramer interacts with 60 bp of DNA, centered at the dyad axis, and makes a series of interactions that act cooperatively to stabilize the overall structure. We found a similar pattern of potential centromere DNA contact sites distributed across the histone-fold domain of Cse4p as revealed by mutagenesis studies (KEITH et al. 1999 ) and by the synthetic interactions between cse4 histone-fold domain mutants and cen3 CDE II mutants. If the histone-fold domain of Cse4p interacts at multiple sites with the DNA helix, then the synthetic effects we observed most likely result from changes in the DNA-protein interactions. It is possible that specific protein-DNA contacts between Cse4p in a nucleosome-like structure and the centromere DNA have been disrupted, causing synthetic phenotypes. Alternatively, the two mutations might affect separate sites that usually interact cooperatively, resulting in a general weakening of the multicomponent structure. By analogy with H3, these results suggest that the CDE II region of the centromere DNA might lie across the dyad axis where Cse4p would replace both copies of H3 in a centric nucleosome. Furthermore, the nonspecific DNA-binding properties of H3 indicate that Cse4p probably interacts nonspecifically with CDE II DNA. Analysis of CDE II DNA from 16 different S. cerevisiae budding yeast centromeres shows no primary consensus sequence, but reveals three common features: conserved length, high A + T base composition, and an intrinsic bend, all of which are important features for centromere function in vivo (MURPHY et al. 1991 ). The CDE II mutants studied here affect these characteristics, suggesting that structural properties conferred by DNA composition and not the primary DNA sequence per se is important to Cse4p-CDE II recognition and interactions.

On the basis of nuclease protection experiments, BLOOM and CARBON 1982 proposed that CDE I, CDE II, and CDE III are assembled into a nucleosome in vivo. Further mapping studies demonstrated that several different yeast centromeres have similar nuclease-resistant chromatin structures, which are characterized by hypersensitive sites flanking a core region of 220 bp of centromere DNA beginning upstream of CDE I and extending past CDE III (reviewed by SCHULMAN and BLOOM 1991 ). The CDE II DNA is the length (80 bp) required to wrap once around a histone octamer, which would position the CDE I and CDE III DNA elements on the same side of the nucleosome. This early model is further supported by evidence that the CP1/CDE I and CBF3/CDE III protein-DNA complexes interact in vivo (BAKER et al. 1998 ; ORTIZ et al. 1999 ) and that core histones are important for centromere structure (SAUNDERS et al. 1990 ). The similarities between Cse4p and H3 and the genetic interactions between Cse4p and histone H4 provide strong support for a specialized Cse4p-nucleosome(s) at the centromere. MELUH et al. 1998 modified this model on the basis of chromatin immunoprecipitation experiments that placed Cse4p at the centromere and genetic interactions between Cse4p and H4. They proposed that CDE I DNA is located outside the Cse4p nucleosome structure while CDE III is placed across the dyad axis where Cse4p replaces histone H3. The synthetic interactions between cse4 and cen3 reported here are not easily accommodated by the DNA path around the specialized Cse4p nucleosome shown in MELUH et al. 1998 . Furthermore, recent studies show that two CBF3 subunits bind to at least three distinct sites on opposite sides of the CDE III DNA helix (RUSSELL et al. 1999 ). If the CDE III DNA is in continuous physical contact with the core histones, access to regions of the CDE III helix by proteins should be significantly limited.

Our genetic results are consistent with the model shown in Fig 4, where the CDE III DNA is located in the linker region, permitting the CBF3 proteins to have access to the CDE III DNA helix (RUSSELL et al. 1999 ). CDE I and CDE II DNA are located within the 60 bp where Cse4p would interact if it replaces H3, with CDE II across the dyad axis where it could interact with Cse4p proteins that replace H3. In this configuration, interactions between the CBF3 proteins and the Cse4p histone-fold domain and other histones in the nucleosome are minimized. One potential discrepancy stems from the location of the CDE I DNA element relative to an upstream nuclease hypersensitive site in the centromeric chromatin. One study reports a shorter nuclease-resistant core of 150–160 bp for CEN14, beginning 11 bp upstream of CDE I (FUNK et al. 1989 ), compared to the 220 bp reported for other centromeres (SCHULMAN and BLOOM 1991 ). Our model is most consistent with a nuclease-resistant core region containing 220 bp of centromere DNA, including CDE I. Nuclease accessibility upstream might be affected during the cell cycle or by the bend in the CDE I DNA helix caused by CP1 binding. Alternatively, CDE may be within the first 20 bp that enter the centric nucleosome. In this configuration CDE II would remain across the dyad axis of the nucleosome in contact with Cse4p, and the part of CDE II at the dyad axis would be shifted 30–40 bp toward CDE III. This would place CDE III within the nucleosome, where CDE III would interact mainly with histones H2A and H2B, not Cse4p. This alternative model does accommodate our results, however; as we argued above, it is unlikely that the CBF3 subunits and the nucleosomal histones could interact at the same sites simultaneously.

View larger version (60K): In this window In a new window Download PPT slide   Figure 4. Proposed role of Cse4p in the structure of yeast centric chromatin. Using the nucleosome structure determined by X-ray crystallography (LUGER et al. 1997 ) and the positions of known nuclease sensitive sites flanking the yeast centromere DNA (SCHULMAN and BLOOM 1991 ) as a template, we used our genetic results to predict the path of the CEN DNA around a novel Cse4p nucleosome. The model is drawn with the variant nucleosome split in half down the superhelical axis, with the top half on the left and the bottom half shown on the right. The dyad axis is marked with an arrow and is the point where the two halves are divided. H3 is replaced by Cse4p (gray) and the histone-fold domains of the three other core histones are shown with H4 in white and H2A and H2B shown together in black. The N-terminal region of Cse4p has been drawn only on the Cse4p molecule in the left half of the split variant nucleosome (light gray) to show that the N terminus extends out from the core nucleosome, near CDE III and the CBF3 complex. Helix I, helix II, and helix III of Cse4p are labeled in both halves of the nucleosome as h1, h2, and h3, respectively. The centromere CDE DNA regions are shaded. By analogy to H3 in a standard nucleosome, the region containing potential Cse4p-centromere DNA interactions is indicated with a heavy black line. The arrow marked "Cse4p" shows the position of the four helix bundle interaction predicted between the two Cse4p molecules. The second Cse4p molecule in each half of the nucleosome is truncated, showing the C-terminal half of helix II and loop II and helix III of Cse4p. The diagram at the upper right shows the centromere DNA path around an octamer represented by the cylinder.

The CP1 protein binds to CDE I elements in centromeres and also mediates the regulation of some biosynthetic genes by inducing bends and changing the chromatin structures associated with promoter DNA (KENT et al. 1994 ). Interestingly, the cse4 histone-fold domain mutant, cse4^csl2, is synthetically lethal in cep1 null cells, indicating a functional connection between the histone-fold domain of Cse4p and CP1 (BAKER et al. 1998 ). In this study we observed synthetic chromosome loss phenotypes when the mutant cen3 CDE I CAT1 centromere was combined in cells with a cse4 allele. The CAT1 centromere DNA mutation alone dramatically lowers the binding affinity of the CP1 protein to the altered CDE I DNA and causes an increase in loss events involving the chromosome with the mutant centromere (BAKER et al. 1989 ). CP1 may play an important role in chromatin remodeling at the centromere, similar to its function at promoters, which excludes standard H3 containing nucleosomes and promoting Cse4p specialized nucleosomes.

Positioning the CBF3/CDE III complex in the linker region does not preclude interactions between Cse4p and proteins assembled on the CDE III DNA. The crystal structure of the histone proteins in nucleosomes (LUGER et al. 1997 ) suggests that the N terminus of Cse4p should pass between the gyres of adjacent DNA helices and extend away from the nucleosome core. In this configuration, the essential region in the N terminus of the two Cse4 proteins would be on either side of the core CDE III sequence where both would be available for interactions with CBF3 subunits (ORTIZ et al. 1999 ). We have recently shown that the N terminus of Cse4p has an essential domain, which is responsible for interactions with the centromere complex containing Ctf19p, Mcm21p, and Okp1p (CHEN et al. 2000 ). In addition, cse4 N-terminal mutants show synthetic lethality with mutant alleles of three out of four of the CBF3 components (p110, p64, and p58).

Why S. cerevisiae centromeres require the function of a specialized core histone protein is not yet clear, but the requirement may be universal as proteins similar to Cse4p have now been discovered in flies and worms as well as mammals and yeast (PALMER et al. 1987 and HENIKOFF et al. 2000 ). In budding yeast, Cse4p may play a critical role by arranging the centromere DNA sequence elements so that the kinetochore-binding site (CDE III) is accessible to microtubules in the higher-order chromatin structures in the chromosome. Indeed, CDE II DNA is clearly required for microtubule binding (KINGSBURY and KOSHLAND 1991 ). The tight association afforded by the integration of nucleosome structure with the kinetochore microtubule binding site may be critically important to ensure that the centromere-spindle connection has the strength necessary to withstand the physical forces of anaphase chromosome movement as well as the flexibility to change with the cell cycle.

	FOOTNOTES

¹ Present address: Department of Molecular Genetics and Cell Biology, University of Chicago, 1103 E. 57th St.-EBC Rm. 304, Chicago, IL 60637.

	ACKNOWLEDGMENTS

We thank Richard Baker and Sam Stoler for many helpful discussions and critical reading of the manuscript and members of the Fitzgerald-Hayes lab for comments and suggestions during the research. This work was supported by a grant to M.F.-H. from the National Institutes of Health (GM54766).

Manuscript received November 29, 1999; Accepted for publication July 14, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BAKER, R. E. and D. C. MASISON, 1990  Isolation of the gene encoding the Saccharomyces cerevisiae centromere-binding protein CP1. Mol. Cell. Biol. 10:2458-2467[Abstract/Free Full Text].

BAKER, R. E., M. FITZGERALD-HAYES, and T. C. O'BRIEN, 1989  Purification of the yeast centromere binding protein CP1 and a mutational analysis of its binding site. J. Biol. Chem. 264:10843-10850[Abstract/Free Full Text].

BAKER, R. E., K. HARRIS, and K. ZHANG, 1998  Mutations synthetically lethal with cep1 target S. cerevisiae kinetochore components. Genetics 149:73-85[Abstract/Free Full Text].

BAXEVANIS, A. D., G. ARENTS, E. N. MOUDRIANAKIS, and D. LANDSMAN, 1995  A variety of DNA-binding and multimeric proteins contain the histone fold motif. Nucleic Acids Res. 23:2685-2691[Abstract/Free Full Text].

BLOOM, K. S. and J. CARBON, 1982  Yeast centromere DNA is in a unique and highly ordered structure in chromosomes and small circular minichromosomes. Cell 29:305-317[Medline].

BRAM, R. J. and R. D. KORNBERG, 1987  Isolation of a Saccharomyces cerevisiae centromere DNA-binding protein, its human homolog, and its possible role as a transcription factor. Mol. Cell. Biol. 7:403-409[Abstract/Free Full Text].

CAI, M. and R. W. DAVIS, 1989  Purification of a yeast centromere-binding protein that is able to distinguish single base-pair mutations in its recognition site. Mol. Cell. Biol. 9:2544-2550[Abstract/Free Full Text].

CHEN, Y., R. E. BAKER, K. C. KEITH, K. HARRIS, and S. STOLER et al., 2000  The N-terminus of the centromere H3-like protein Cse4p performs an essential function distinct from that of the histone fold domain. Mol. Cell. Biol. 20:7037-7048[Abstract/Free Full Text].

CUMBERLEDGE, S. and J. CARBON, 1987  Mutational analysis of meiotic and mitotic centromere function in Saccharomyces cerevisiae.. Genetics 117:203-212[Abstract/Free Full Text].

FITZGERALD-HAYES, M., L. CLARKE, and J. CARBON, 1982  Nucleotide sequence comparisons and functional analysis of yeast centromere DNAs. Cell 29:235-244[Medline].

FLEIG, U., J. D. BEINHAUER, and J. H. HEGEMANN, 1995  Functional selection for the centromere DNA from yeast chromosome VII. Nucleic Acid Res. 23:922-924[Abstract/Free Full Text].

FUNK, M., J. H. HEGEMANN, and P. PHILIPPSEN, 1989  Chromatin digestion with restriction endonucleases reveals 150-160 bp of protected DNA in the centromere of chromosome XIV in Saccharomyces cerevisiae.. Mol. Gen. Genet. 219:153-160[Medline].

GAUDET, A. and M. FITZGERALD-HAYES, 1987  Alterations in the adenine-plus-thymine-rich region of CEN3 affect centromere function in Saccharomyces cerevisiae.. Mol. Cell. Biol. 7:68-75[Abstract/Free Full Text].

HEGEMANN, J. H., J. H. SHERO, G. COTTAREL, P. PHILIPPSEN, and P. HIETER, 1988  Mutational analysis of centromere DNA from chromosome VI of Saccharomyces cerevisiae.. Mol. Cell. Biol. 8:2523-2535[Abstract/Free Full Text].

HENIKOFF, S., K. AHMAD, J. S. PLATERO, and B. VAN STEENSEL, 2000  Heterochromatic deposition of centromeric histone H3-like proteins. Proc. Natl. Acad. Sci. USA 97:716-721[Abstract/Free Full Text].

HUFFAKER, T. C., M. A. HOYT, and D. BOTSTEIN, 1987  Genetic analysis of the yeast cytoskeleton. Annu. Rev. Genet. 21:259-284[Medline].

KEITH, K. C., R. E. BAKER, Y. CHEN, K. HARRIS, and S. STOLER et al., 1999  Analysis of primary structural determinants that distinguish the centromere-specific function of histone variant Cse4p from H3. Mol. Cell. Biol. 19:6130-6139[Abstract/Free Full Text].

KENT, N. A., J. S. H. TSANG, D. J. CROWTHER, and J. MELLOR, 1994  Chromatin structure modulation in Saccharomyces cerevisiae by centromere and promoter factor I. Mol. Cell. Biol. 14:5229-5241[Abstract/Free Full Text].

KINGSBURY, J. and D. KOSHLAND, 1991  Centromere-dependent binding of yeast minichromosomes to microtubules in vitro. Cell 66:483-495[Medline].

LEA, D. E. and C. A. COULSON, 1949  The distribution of the number of mutants in bacterial populations. J. Genet. 49:264-285.

LUGER, K., A. W. MADER, R. K. RICHMOND, D. F. SARGENT, and T. J. RICHMOND, 1997  Crystal structure of the nucleosome core particle at 2.8A resolution. Nature 389:251-260[Medline].

MCGREW, J., B. DIEHL, and M. FITZGERALD-HAYES, 1986  Single base-pair mutations in centromere element III cause aberrant chromosome segregation in Saccharomyces cerevisiae.. Mol. Cell. Biol. 6:530-538[Abstract/Free Full Text].

MCGREW, J. T., Z. XIAO, and M. FITZGERALD-HAYES, 1989  Saccharomyces cerevisiae mutants defective in chromosome segregation. Yeast 5:271-284[Medline].

MELUH, P. B. and D. KOSHLAND, 1995  Evidence that the MIF2 gene of Saccharomyces cerevisiae encodes a centromere protein with homology to the mammalian centromere protein, CENP-C. Mol. Cell. Biol. 6:793-807.

MELUH, P. B., P. YANG, L. GLOWCZEWSKI, D. KOSHLAND, and M. M. SMITH, 1998  Cse4p is a component of the core centromere of Saccharomyces cerevisiae.. Cell 94:607-613[Medline].

MURPHY, M. R., D. M. FOWLKES, and M. FITZGERALD-HAYES, 1991  Analysis of centromere function in Saccharomyces cerevisiae using synthetic centromere mutants. Chromosoma 101:189-197[Medline].

ORTIZ, J., O. STEMMANN, S. RANK, and J. LECHNER, 1999  A putative protein complex consisting of Ctf19, Mcm21, and Okp1 represents a missing link in the budding yeast kinetochore. Genes Dev. 13:1140-1155[Abstract/Free Full Text].

PALMER, D. K., K. O'DAY, M. H. WENER, B. S. ANDREWS, and R. L. MARGOLIS, 1987  A 17-kD centromere protein (CENP–A) copurifies with nucleosome core particles and with histones. J. Cell Biol. 104:805-815[Abstract/Free Full Text].

RUSSELL, I. D., A. S. GRANCELL, and P. K. SORGER, 1999  The unstable F-box protein p58-Ctf13 forms the structural core of the CBF3 kinetochore complex. J. Cell Biol. 145:684-689.

SAUNDERS, M. J., E. YEH, M. GRUNSTEIN, and K. BLOOM, 1990  Nucleosome depletion alters the chromatin structure of Saccharomyces cerevisiae centromeres. Mol. Cell. Biol. 10:5721-5727[Abstract/Free Full Text].

SCHULMAN, I. and K. S. BLOOM, 1991  Centromeres: an integrated protein/DNA complex required for chromosome movement. Annu. Rev. Cell Biol. 7:311-336.

SHELBY, R. D., O. VAFA, and K. F. SULLIVAN, 1997  Assembly of CENP-A into centromeric chromatin requires a cooperative array of nucleosomal DNA contact sites. J. Cell Biol. 136:501-513[Abstract/Free Full Text].

SMITH, M. M., P. YANG, M. S. SANTISTEBAN, P. W. BOONE, and A. T. GOLDSTEIN et al., 1996  A novel histone H4 mutant defective for nuclear division and mitotic chromosome transmission. Mol. Cell. Biol. 16:1017-1026[Abstract].

STOLER, S., K. C. KEITH, K. E. CURNICK, and M. FITZGERALD-HAYES, 1995  A mutation in CSE4, an essential gene encoding a novel chromatin-associated protein in yeast, causes chromosome nondisjunction and cell cycle arrest at mitosis. Genes Dev. 9:573-586[Abstract/Free Full Text].

STRUNNIKOV, A. V., J. KINGSBURY, and D. KOSHLAND, 1995  CEP3 encodes a centromere protein of Saccharomyces cerevisiae.. J. Cell Biol. 128:749-760[Abstract/Free Full Text].

SULLIVAN, K. F., M. HECHENBERGER, and K. MASRI, 1994  Human CENP-A contains a histone H3 related histone fold domain that is required for targeting to the centromere. J. Cell Biol. 127:581-592[Abstract/Free Full Text].

VAFA, O. and K. F. SULLIVAN, 1997  Chromatin containing CENP-A and alpha-satellite DNA is a major component of the inner kinetochore plate. Curr. Biol. 7:897-900[Medline].

This article has been cited by other articles:

				W.-C. Au, M. J. Crisp, S. Z. DeLuca, O. J. Rando, and M. A. Basrai Altered Dosage and Mislocalization of Histone H3 and Cse4p Lead to Chromosome Loss in Saccharomyces cerevisiae Genetics, May 1, 2008; 179(1): 263 - 275. [Abstract] [Full Text] [PDF]

				S. Hajra, S. K. Ghosh, and M. Jayaram The centromere-specific histone variant Cse4p (CENP-A) is essential for functional chromatin architecture at the yeast 2-{micro}m circle partitioning locus and promotes equal plasmid segregation J. Cell Biol., September 11, 2006; 174(6): 779 - 790. [Abstract] [Full Text] [PDF]

				H. S. Malik A hitchhiker's guide to survival finally makes CENs J. Cell Biol., September 11, 2006; 174(6): 747 - 749. [Abstract] [Full Text] [PDF]

				H. Kanta, L. Laprade, A. Almutairi, and I. Pinto Suppressor Analysis of a Histone Defect Identifies a New Function for the Hda1 Complex in Chromosome Segregation Genetics, May 1, 2006; 173(1): 435 - 450. [Abstract] [Full Text] [PDF]

				M. D. Cervantes, X. Xi, D. Vermaak, M.-C. Yao, and H. S. Malik The CNA1 Histone of the Ciliate Tetrahymena thermophila Is Essential for Chromosome Segregation in the Germline Micronucleus Mol. Biol. Cell, January 1, 2006; 17(1): 485 - 497. [Abstract] [Full Text] [PDF]

				K. A. Collins, A. R. Castillo, S. Y. Tatsutani, and S. Biggins De Novo Kinetochore Assembly Requires the Centromeric Histone H3 Variant Mol. Biol. Cell, December 1, 2005; 16(12): 5649 - 5660. [Abstract] [Full Text] [PDF]

				N. S. Edwards and A. W. Murray Identification of Xenopus CENP-A and an Associated Centromeric DNA Repeat Mol. Biol. Cell, April 1, 2005; 16(4): 1800 - 1810. [Abstract] [Full Text] [PDF]

				T. Stoyan and J. Carbon Inner Kinetochore of the Pathogenic Yeast Candida glabrata Eukaryot. Cell, October 1, 2004; 3(5): 1154 - 1163. [Abstract] [Full Text] [PDF]

				C. Capiaghi, T. V. Ho, and F. Thoma Kinetochores Prevent Repair of UV Damage in Saccharomyces cerevisiae Centromeres Mol. Cell. Biol., August 15, 2004; 24(16): 6907 - 6918. [Abstract] [Full Text] [PDF]

				G. Wieland, S. Orthaus, S. Ohndorf, S. Diekmann, and P. Hemmerich Functional Complementation of Human Centromere Protein A (CENP-A) by Cse4p from Saccharomyces cerevisiae Mol. Cell. Biol., August 1, 2004; 24(15): 6620 - 6630. [Abstract] [Full Text] [PDF]

				C. W. Espelin, K. T. Simons, S. C. Harrison, and P. K. Sorger Binding of the Essential Saccharomyces cerevisiae Kinetochore Protein Ndc10p to CDEII Mol. Biol. Cell, November 1, 2003; 14(11): 4557 - 4568. [Abstract] [Full Text] [PDF]

				J.-m. Hsu, J. Huang, P. B. Meluh, and B. C. Laurent The Yeast RSC Chromatin-Remodeling Complex Is Required for Kinetochore Function in Chromosome Segregation Mol. Cell. Biol., May 1, 2003; 23(9): 3202 - 3215. [Abstract] [Full Text] [PDF]