Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Colloquium Histone H3 lysine 4 methylation is metliatecl by Set, and promotes maintenance of active chromatin states in fission yeast Ken-ichi Noma and Shiv 1. S. Grewal* Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, NY 11724 Methylation of histone H3 at Iysine 4 (H3 Lys-4) or Iysine 9 (H3 Lys-9) is known to define active and silent chromosomal domains respectively from fission yeast to humans. However, in budding yeast, H3 Lys-4 methylation is also necessary for silent chromatin assembly at telomeres and ribosomal DNA. Here we demonstrate that deletion of sets, which encodes a protein containing an RNA recognition motif at its amino terminus and a SET domain at the carboxy terminus, abolishes H3 Lys-4 methylation in fission yeast. Unlike in budding yeast, Set1-mediated H3 Lys-4 methylation is not required for heterochromatin assembly at the silent mating-type region and centromeres in fission yeast. Our analysis suggests that H3 Lys-4 methylation is a stable histone modification present throughout the cell cycle, including mitosis. The loss of H3 Lys-4 methylation in set11t cells is correlated with a decrease in histone H3 acetylation levels, suggesting a mechanistic link between H3 Lys-4 methylation and acetylation of the H3 tail. We suggest that methylation of H3 Lys-4 primarily acts in the maintenance of transcriptionally poised euchromatic domains, and that this mod- ification is dispensable for heterochromatin formation in fission yeast, which instead utilizes H3 Lys-9 methylation. Dynamic changes in chromatin structure are directly influ- enced by the posttranslational modification of histones. Specific amino acids on histone amino-terminal tails that extend outward from nucleosome core particle are the targets of a number of modifications including acetylation, phosphorylation, ubiquitylation, ADP ribosylation, and methylation (1-5~. The presence of a specific pattern of histone modifications has been linked to various chromosomal processes, such as the mainte- nance of gene expression patterns during development, recom- bination, chromosome condensation, and the proper segregation of chromosomes during mitosis. For example, hyperacetylated regions of chromatin contain active transcription units, whereas hypoacetylated chromatin is transcriptionally silent (6~. How histone modifications participate in the modulation of chromatin structure is not fully understood. It is believed that different combinations of histone modifications can determine the binding affinities of his/one-interacting proteins whose chromo- somal associations lead to discrete downstream events (2, 7~. This is best illustrated by recent studies showing that modifications of the H3 tail by deacetylase and methyltransferase activities likely act in concert to establish the "histone code" essential for heterochro- matin assembly (3, 4, 8~. The chromodomains of heterochromatin proteins Swi6 and HP1 from fission yeast and Drosophila, respec- tively, are specific interaction motifs for the histone H3 amino- terminal tail modified by methylation on lysine 9 (9-11), and localization of these proteins to heterochromatic loci depends on H3 Lys-9 methylation (8, 12~. Similarly, the bromodomain of many transcriptional coactivators binds specifically to the acetylated lysine residues on histone tails (13~. Applying this general concept to other chromatin modulators, it is likely that unique combinations 16438-16445 1 PNAS 1 December 10, 2002 1 vol. 99 1 suppl. 4 of histone modifications serve as marks for the recruitment of different chromatin proteins or protein complexes to initiate the formation of defined chromosomal subdomains with distinct functions. The methylation of H3 and H4 tails can occur on both arginine and lysine residues, and plays a critical role in transcriptional regulation (3, 5~. A specific class of methyltransferases including CARM1 and PRMT1, which act as transcriptional coactivators, catalyze histone arginine methylation (14-16~. However, enzymes that contain an evolutionary conserved SET domain, a 130-residue motif originally identified in Su~var)3-9, _nhancer-of-zeste and _rithorax proteins in Drosophila (17), are implicated in the lysine methylation of histones (5~. The founding members of this class of histone methyltransferases are mammalian SUV39H1 and its fis- sion yeast homolog Clr4, which specifically methylate H3 lysine 9 (H3 Lys-9) at heterochromatic loci (8, 18~. Recently, several SET domain-containing proteins, such as G9a, SETDB1, and ESET, have been shown to methylate H3 Lys-9 in mammals (19-21~. In contrast to the H3 Lys-9 methylation that defines silent chromosomal domains, H3 Lys-4 methylation is specific to tran- scriptionally poised euchromatic regions in fission yeast, Tetrahy- mena, chicken, and mammals (22-26~. Moreover, high-resolution mapping using chromatin immunoprecipitation (CHIP) has shown that distinct patterns of histone methylation marking heterochro- matic and euchromatic domains are separated by boundary ele- ments that protect against spreading of repressive chromatin into neighboring areas (22, 23~. H3 Lys-4 is methylated by SET domain containing proteins- Setl in Saccharomyces cerevisiae (27-29) and Set7/Set9 in mammals (30, 31~. Although the precise function of H3 Lys-4 methylation is not known, it has been hypothesized to facilitate transcription by serving as a mark to recruit necessary transcription machinery components, or alternatively to protect euchromatic regions from the repressive effects of neighboring silent chromatin complexes (22, 23~. Interestingly, the deletion of SET1 in S. cerevisiae leads to defects in silencing at telomeres, ribosomal DNA (rDNA), and the mating- type region (27, 32, 334. It has been suggested that H3 Lys-4 methylation in the context of other histone modifications might have a dual function both in transcriptional activation and silencing (27, 33~. However, it remains to be investigated whether H3 Lys-4 methylation by Setl homologs is also required for silent chromatin assembly in other species. This paper results from the Arthur M. Sackier Coiloquium of the Nationai Academy of Sciences, "Self-Perpetuating Structurai States in Bioiogy, Disease, and Genetics," held March 22-24, 2002, at the Nationai Academy of Sciences in Washington, DC. Abbreviations: SET, Su(var)3-9 enhancer-of-zeste and trithorax; ChlP, chromatin immunopre- cipitation; RRM, RNA recognition motif; YEA, yeast extract adenine; FOA, 5'-fluoroorotic acid; DAPI, 4',6-diamidino-2-phenylindole. *To whom reprint requests should be addressed. E-mail: grewal~cshl.org. www. pnas.org/cg i/doi/ 10.1 073/pnas. 182436399
A B 0~05 _9~ SpSet1 Setl family Suv39 / Clr4 family Set2 family ~ - - -.- I SpSet5 ~ SpSet6 _ SpSet7 I SpSet8 mSET7 SpSet9 SpSet3 SpSet2 I= 798 aa SpSet3 Clr4 SpSet5 _ 1 319 aa t SpSet6 I ~ 1 483 aa SpSet7 ~ 147 aa SpSet8 I ~ 1 429 aa SpSet9 =_ ~ ~ 441 aa 920 aa 859 aa ~ SET ~ n-SET ~ Cys - rich ~ Ser- rich | In this study, we report identification and structural features of at least nine SET domain proteins present in the fission yeast genome. We show that one of these SET domain proteins that we named Setl appears to be the exclusive H3 Lys-4-specific methyl- transferase. Unlike in S. cerevisiae, Setl is not required for hetero- chromatin assembly in fission yeast, consistent with our previous results showing that H3 Lys-4 methylation is specific to euchromatic regions. H3 Lys-4-methylaton is present throughout the cell cycle and important for the upkeep of transcriptionally poised domains in euchromatic regions, perhaps through the maintenance of his- tone acetylation levels. Materials and Methods Sequence Analysisa Database searches were performed with BLASTP. Multiple amino acid sequences were aligned by using CLUSTALW, Version 1.7. A phylogenetic tree was created by the neighbor- joining method based on the amino acid sequence alignment. The domains were characterized by using the MOTIFSCAN program against the PROSITE database. The resulting dendrogram showing the relationship between different SET domain proteins is shown in Fig. L4, and structural features of different SET domain proteins in fission yeast are shown in Fig. iB. Strains. The genotypes of the Sch~zosaccharomyces pombe strains used in this study are listed in Table 1. The setli\ strain was constructed by a PCR based method as described (34~. Deletion was confirmed by PCR and Southern analysis. A strain containing deletion of a part of the setl ORE encoding RNA recognition motif (RRM) was constructed as follows: DNA fragments from upstream (0.9 kb) and downstream (2.6 kb) of RRM encoding region of sell were amplified by PCR with Pp`Turbo DNA polymerase (Strat- Table 1. Sch. pombe strains used in this study ~ Chromo ~ PHD Fig. 1. SET domain proteins in Sch. pombe genome. (A) Phylogenetic tree of the SET domain proteins. The tree was con- structed by the neighborjoining method based on the amino acid sequences in the SET domains. The scale bar equals a dis- tance of 0.05 ea. The SpSet (Sch. pombe SET) members (GenBank accession nos.: SpSet1, AL049728; SpSet2, Z99164; SpSet3, Z70043; SpSet5, AL03 1540; SpSet6, AL032684; SpSet7, AL049609; SpSet8, Z99568; SpSet9, AL132870) were identified by BLAST using previously identified SET do- mains as query. Abbreviations: Hs, Homo sapiens; m, Mus musculus. (B) Schematic representation of Sch. pombe SET domain proteins. The length of each protein (in aa) is noted on the right. Conserved domains are indicated as follows: RRM, RNA recog- nition motif; Chromo, Chromodomain; PHD, PHD finger; Cys-rich, Cysteine-rich do- main; Ser-rich, Serine-rich domain. agene) and cloned in frame into pCRII-TOPO (Invitrogen) to construct setl ORF minus RRM encoding region (setl~RR~. The resulting 3.5-kb fragment was gel-purified and used for transfor- mation of SPK10 strain carrying setl/~::kanMX6 allele. Transfor- mants were screened for G418 sensitivity, and colonies carrying setl~RRM were confirmed by using PCR analysis. Standard genetic crosses were used to construct all other strains. iodine Staining Assay. Efficiency of mating-type switching was analyzed by the iodine-staining assay. Individual colonies were replicated onto sporulation (PMA+) medium and then grown for 3 days at 26°C before being exposed to iodine vapors. The dark staining indicates efficient mating-type switching, which requires heterochromatin-mediated chromatin organization at the mating- type region (354. Defects in heterochromatin assembly at the mating-type region results in inefficient switching, causing a de- . . . . . crease 1n 1OC .me stammg. Western Analysis of Histones. For isolation of bulk histones, fission yeast cells were grown to mid-log phase. Cells (5 x 108) were washed with 10 ml of NIB buffer (0.25 M sucrose/60 mM KCl/15 mM NaCl/5 mM MgCl2/1 mM CaCl2/15 mM Pipes, pH 6.8/0.8% Triton X-100) and resuspended in 500 ,ul of NIB buffer containing 10 ng/,ul TSA, 2 mM ZnS04, Complete protease inhibitor mixture (Roche, 1 tablet per 10 ml), and 1 mM PMSF. Cells were disrupted by acid-washed glass beads (425-600 ,uM) using a minibeadsbeater (Biospec) for 4-5 min. Cell extracts were centrifuged at 11,000 x g for 10 m~n. The pellets were resuspended in 0.4 M H2SO4 and incubated on ice for 1 h with occasional m~xing. The supernatant was collected by centrifugation at 8,000 x g for 5 min. The H2SO4 extraction was repeated. Pooled supernatants were trichloroacetic Strain ura4+ insertion Genotype FY498 FY648 PG925 SP1 464 SPG 1 236 SPK1 0 SPK1 08 SPK] 1 1 SPK] 21 Noma and Grewal imr1R:: ura4+ otrlR::ura4+ mat3M:: ura4 KA:: ura4+ Kint2:: ura4+ Kint2:: ura4+ imrl R : : ura4+ otrlR::ura4+ Kint2:: ura4+ h+ leul-32 ura4DS/E ade6-210 h+ leu 1-32 ura4DS/E ade6-2 10 h90 leul-32 ura4D18 ade6-210 cir3-735 h90 leul-32 ura4D18 ade6-210 cir6-1 h90 /eu 1-32 ura4DS/E ade6-2 16 his2 h90 leul-32 ura4DS/E ade6-216 his2 setl/\::kanMX6+ h+ /eu 1-32 ura4DS/E ade6-2 16 setl i\:: kanMX6+ h+ leul-32 ura4DS/E ade6-210 setl7\ ::kanMX6+ h90 leul-32 ura4DS/E ade6-216 his2 set1~RRM PNAS | December 10, 2002 1 vol. 99 | suppl. 4 | 16439
acid precipitated and pellets were washed twice in 500 ,ul of cold acetone, air-dried and resuspended in 10 mM Tris HCl, pH 8.0. Bulk histone samples were kept in -70°C freezer until use. Ten micrograms of crude histone samples were resolved on an SDS/ 18% PAGE, transferred to a poly~vinylidene difluoride) mem- brane, and probed with site-specific acetyl- or methyl-his/one antibodies. Antibodies to H3 Lys-4-methyl, H3 Lys-9-acetyl, and H3 Lys-14-acetyl were purchased from Upstate Biotechnology, whereas antibodies specific to acetylated Lys-5, Lys-8, Lys-12, or Lys-16 of histone H4 were purchased from Serotec. The band intensities were quantified with NIH IMAGE 1.62 software. Immunofluorescence Analysis. Cells were grown to mid-log phase in yeast-extract adenine (YEA) medium. An equal volume of YEA + 2.4 M sorbitol was added and the culture was incubated further at 18°C for 5 min. For using o`-H3 Lys-4-methyl, cells were fixed by adding paraformaldehyde to a final concentration of 1.7% and incubated at 18°C for 45 min. For staining by or-H3 Ser-10-phospho, cells were fixed in 3.0% paraformaldehyde for 30 min. Fixed cells were treated with Zymolyase to permeabilize the cell wall and then incubated overnight with primary antibodies, such as mouse ~x- tubulin TAT1 (1:150 dilution) or a-Nopl (1:1,000 dilution) and rabbit cY-H3 Lys-4-methyl (1:1,500) or a-H3 Ser-10-phospho (1:5004. After extensive washing, ceils were incubated for 6-8 h with Alexa Fluor 594 anti-rabbit IgG and Oregon Green 488 anti-mouse IgG (Molecular Probes) at a 1:2,000 dilution. After washing, cells were stained with 4',6-diamidino-2-phenylindole (DAPI), mounted in Vectashield mounting medium (Vector Laboratories), and an- alyzed by a Zeiss Axioplan 2 fluorescence microscope. ChlP. ChIP analysis was performed as described (36, 37~. Fission yeast cells grown at 32°C in YEA (5 x 108 cells at 1 x 107 cells per ml for each reaction) were shifted to 18°C for 2 h before 30-min fixation in 3~o paraformaldehyde. Soluble chromatin fractions prepared from fixed cells were sheared to ~0.5- to 0.8-kb DNA fragments by sonication before immunoprecipitating by using an- tibodies to H3 Lys-4-methyl, H3 Lys-9-methyl, H3 Lys-14-acetyl, and Swi6. DNA fragments recovered from immunoprecipitated chromatin fractions or from whole cell crude extracts were sub- jected to PCR analyses (94°C for 30 s, 55°C for 30 s, 72°C for 1 min. 30 cycles). PCR products were labeled by including 0.25 ,ul of tor-32Pideoxycytidine triphosphate (10 mCi/ml; 1 Ci = 37 GBq) in each reaction. PCR products were separated on a 4% polyacryl- amide gel, and band intensities were auantified bv usin~ a Fuii PhosphoImager. Results SET Domain Proteins in Sch. pombe and Their Structural Features. A database search of the Sch. pombe genome (http://www.sanger. ac.uk/Projects/S_pombe/) was performed to identify the total number of SET domain proteins. The Sch. pombe genome has been sequenced, and contains 4,824 genes (38~. A BLAST search with the SET domain of the Clr4 histone methyltransferase revealed that at least nine SET domain proteins, including Clr4, reside in the fission yeast genome (Fig. 1A). All of the Sch. pombe SET domain proteins contain a conserved NHSC motif, which, when mutated, has been shown to abolish histone methyltransferase activity (18~. SpSetl, SpSet2, and SpSet3 were named after their S. cerevisiae counter- parts based on homology. However, Clr4, SpSetS, SpSet6, SpSet7, SpSet8, and SpSet9 did not share significant similarities to S. cerevisiae proteins outside of their SET domain. The absence of a Clr4 homolog in S. cerevisiae is consistent with the fact that bulk histones isolated from budding yeast cells lack detectable levels of H3 Lys-9 methylation (24, 27~. We next carried out phylogenetic analysis to assess the relation- ships between different SET domain proteins (Fig. 1A). The SET domain sequences of Sch. pombe proteins and previously charac- terized proteins from other species were aligned together and a 16440 1 www.pnas.org/cgi/doi/10.1073/pnas.182436399 phylogenetic tree was constructed. Based on the dendrogram (Fig. 1A), Sch. pombe SET domain proteins were not highly related to one another. As expected, SpSetl and Clr4 proteins cluster with the previously described Setl and SUV39 family of histone methyl- transferases that have the capacity to methylate H3 Lys-4 and H3 Lys-9, respectively (5~. The Setl proteins from budding and fission yeast share 26% identity throughout their length and 63~o in the SET domain region. Our analysis suggests that SpSetl is more closely related to Setl in Homo sapiens than it is to Setl in S. cerevisiae. Another SET domain protein, SpSet2, is closely related to a H3 Lys-36-specific methyltransferase from S. cerevisiae (39~. Careful examination of the SET domains in Sch. pombe proteins revealed distinct structural features. As reported previously, the SET domain in Clr4 is surrounded by two cysteine-rich regions, referred to as preSET and postSET, that are essential for its catalytic activity (8~. The preSET and postSET domains are also present in SpSet2, but only the postSET domain is present in SpSetl, SpSetS, SpSet6, and SpSet9 (Fig. 1B). SpSetl contains a highly conserved 160-aa motif called n-SET at the preSET location. The SET domains of SpSet3, SpSet7, and SpSet8 do not have pre- or postSET motifs. It is possible that differences in structural features surrounding the SET domains might account for the altered substrate specificity of these proteins. Further analysis using the MOTIFSCAN program revealed that in addition to the presence of a chromodomain in Clr4, as reported (40), SpSetl and SpSet3 contain an RRM and PHD finger at their amino-terminal regions, respectively. Both the PHD finger that is involved in protein-protein interactions, and the RRM, which is known to interact with both RNA and proteins, are conserved motifs shown to be present in subunits of chromatin-modifying activities (41, 42~. Sch. pombe Set1 Is Required for H3 Lys-4 Methylation in Vivo. As mentioned above, Setl mediates H3 Lys-4 methylation in S. cer- evisiae (27-29~. Considering that Setl proteins from budding and fission yeasts share considerable homology within and outside of their SET domains, it was possible that SpSetl might also be involved in H3 Lys-4 methylation. To test this possibility, we constructed a strain containing a complete deletion of setl gene, replacing the entire ORF with the K:41V1~6 gene. The resulting setl /\ mutant was viable, suggesting that SpSetl is dispensable for cell growth. To study the biological effects of setl ~ on H3 Lys-4 methylation, we performed immunofluorescence analysis using an antiserum specific to methyllysine 4 of histone H3 (22, 27~. In wild-type cells, immunofluorescence signal corresponding to H3 Lys-4-methyl was preferentially enriched at the chromatin in the DAPI-stained areas but seemed to be excluded from the nucleolus containing rDNA repeats, as indicated by the staining of nucleolar marker protein Nopl (Fig. 2A; ref. 43~. In comparison to a high level of H3 Lys-4 methylation in the nuclei of wild-type cells, strikingly, methylation of histone H3 at lysine 4 was completely abolished insetl i\ cells (Fig. 2C), suggesting that SpSetl is responsible for H3 Lys-4 methylation. This result was further confirmed by Western analysis of bulk histones prepared from wild-type and setlA cells by using H3 Lys-4-methyl-specific antibodies. Histones prepared from wild-type cells exhibit high levels of methylated H3 Lys-4, but we could not detect any signal in the setl ~ cells (Fig. 2B). These analyses suggest that Setl mediates H3 Lys-4 methyaltion in Sch. pombe. Although it remains a possibility that low levels of H3 Lys-4 methylation exist in the nucleolus, histone H3 methylated at lysine 4 is clearly not enriched in the nucleolar compartment. Previous studies in S. cerevisiae have suggested that H3 Lys-4 methylation is important for normal growth, and that mutations in SET1 can result in a number of phenotypes including morphological abnormalities, perturbed DNA distribution, growth and sporula- tion defects (27, 32~. We therefore investigated whether setl ~ cells display similar phenotypes in Sch. pombe. We noticed that colonies Noma and Grewa~
A H3K4Me B H3K4Me CBB D DAPI H3K4Me + None C H3K4Me vat set1^ 10 8 6 4 2 Art - set1^ ' Generation time ~ wt ~ 2~2 h ' , ,¢ ·--o- setup ~ 2.6 h ~ 4` , , ,' I/ ~ ~ ~ ~ ~ "I ~ ~ - - 1 - - ~ t r f 7 T A t Jo ,'~ f ~ ~ 1 1 1 /t,~ 1 1 1 1 1 1 ,i,,Cp 1 1 1 ~ 1 1 ALL,, 1 1 1 1 1 1 1 ',Si 1 ~ 1 1 1 - - - 1- - J=^ T- ~ - -. - - T - - r - 1 /O 1 1 1 1 1 1 ,1,~,~ 1 1 1 1 1 1 1 ~ 1 1 1 1 1 1 1 in' 1 1 ~ ~ ~ ~ ~ ~ 0 1 2 3 4 5 6 7 8 9 Time (h) DAPI Merged E 30°C 37°C wet setup formed by setl /\ mutants are smaller in size when compared with their wild-type counterparts (data not shown). Furthermore, setl l\ cells have a slightly longer doubling time and exhibit temperature- sensitive growth defects (Fig. 2D andE). However, deletion of Sch. pombe setl did not cause any obvious morphological abnormalities, sporulation defects, or abnormal DAPI staining patterns. RRM of SpSet1 is Necessary for Its Role in H3 Lys-4 Methylation. As described above, the amino terminus of the SpSetl contains a canonical RNA binding domain called RRM (Fig. 1B). Interest- ingly, this motif is also present in orthologs of SpSetl in S. cerevisiae, Caenorhabditis elegans, Drosophila, and humans. We sought to investigate whether the RRM domain is essential for SpSetl- mediated H3 Lys-4 methylation. For this purpose, we constructed a strain in which a small part of the setl ORE, encoding the RRM, was deleted at its endogenous chromosomal location. The expres- sion of setl~RRM is under the control of the native setl regulatory elements, so as to achieve wild-type levels of expression. Interest- ingly, Western analysis of bulk histones revealed that H3 Lys-4 methylation is severely defective in the setl PROM strain as compared with wild-type (Fig. 34. Moreover, setl ERRS strain showed temper- ature-sensitive growth defects (data not shown). Although a pos- sibility remains that deletion of RRM affects steady-state Setl levels in the cells, we suggest that the RRM domain of SpSetl might be required for its role in H3 Lys-4 methylation. Deletion of set1 Does Not Affect Silencing at the Mating-Type Region and Centromeres. Modifications of histone tails are known to play a critical role in heterochromatin assembly through their role in Noma and Grewal Fig. 2. Deletion of setl abolishes H3 Lys-4 methylation. (A) Wild-type cells (SPG1236) were stained with anti-H3 Lys-4- methyl (H3K4Me) antibody (red), anti-Nop1 antibody (green), and DAPI (blue). (B) Western blot with H3K4Me antibody against crude bulk histones prepared from wild-type (SPG 1236) and set1A (SPK10) strains. Identical samples were examined in parallel by Coomassie staining to show histone loading. (C) Wild-type and setlA (SPK10) cells were stained with anti- H3K4Me and DAPI. (D) Comparison in growth rate of wild-type (SPG1236) and set1A (SPK10) strain. Wild-type and mutant strains were grown in YEA-rich medium at 30°C. Cell numbers at fixed time points are plotted on a semilogarithmic graph. (E) Temperature sensitivity of set1A strain. The strains previously grown under permissive growth conditions (30°C) were repli- cated onto YEA plates and incubated overnight at 30°C or 37°C. dictating the interactions between nucleosome arrays and non- histone chromatin proteins (44~. Setl-mediated H3 Lys-4 methyl- ation has been shown to be required for transcriptional silencing of the silent mating-type loci, telomeres, and rDNA in S. cerevisiae. To investigate whether setl is also required for transcriptional silencing in Sch. pombe, we combined setli\ with a ura4+ marker gene A RRM n-S\T SET Cys- rich Sets | ~ ~ - ERRS =----- ..... ~ 00 aa B H3K4Me CBB wt set1A sets Fig. 3. RRM of SpSet1 is required for its role in H3 Lys-4 methylation. (A) Schematic representation of the Set1 and Set1~RRM. (B) Western blot with H3K4Me antibody against crude bulk histones prepared from wild-type (SPG1236), Seth\ (SPK10), and sets (SPK121) strains. Coomassie brilliant blue (CBB) staining is shown as loading controls. PNAS 1 December 10, 2002 1 vol. 99 1 suppl. 4 1 16441
A B ,:et1 C 0.10 42.818.4 - set1 IimrlR ::ura4+ A, otr1R | ::ura4+ | '2, Mati ng-type reg ion heterochrOmatin- _ _ 2 5 kb mat1 JR-L mat2P cenH mat3M IR-R ......... ; K`nt2::ura4+ WT set1A K4 K9 Swi6 WCE K4 K9 Swi6 WCE 0.01 19.4 12.2 - NE 45.8 12.4 - Enrichment (-fold) .RPf1 Kint2 ::ura4+ Centromere FOA AA-URA 2.5 kb otrl L imr1L cnt' imr1R otr1R . . imr1R::uradf otr1R::ura4+ WT set1A K4 K9 Swi6 WCE K4 K9 Swi6 WCE WT set1A K4 K9 Swi6 WCE K4 K9 Swi6 WCE NE 35.4 14.1 - 0.06 15.3 18.0 - NE 39.2 17.7 - Enrichment (-fold) FOA AA-URA __ a is_ inserted at either the silent mating-type region (Kint2::ura4+) or at two different sites within cenl, one each at the outer (otrlR::ura4+) and inner (~imrlR::ura4+) centromeric repeats that flank the central (cnt) domain (Fig. 4; ref. 45~. Previous studies have shown that marker genes inserted within or adjacent to these heterochromatic locations are subject to transcriptional repression, as a consequence of repressive chromatin complexes spreading into the marker genes (46~. Wild-type yeast cells carrying repressed ura4+ at a hetero- chromatic site cannot form colonies on medium lacking uracil (URA-) but grow efficiently on a counterselective medium con- taining 5'-fluoroorotic acid (FOA). However, cells defective in silencing grow on URA- medium and are FOA sensitive. Dilution analysis of the wild-type and setli\ strains carrying ura4+ marker gene at the silent mating-type region or at the centromeric locations revealed that SpSetl is not required for silencing at these loci, as indicated by comparable growth of wild-type and setlA cells on URA- and FOA media (Fig. 4 B and C). We also tested whether setli\ affects expression of the KA ::ura4+ marker gene. The KA::ura4+ cells containing a substitution of part the interval between silent mat~ng-type loci with the ura4+ marker gene exhibit variegated ura4+ expression (37~. ura4-off end ura4-on epigenetic states are mitotically metastable. We used fluctuation 16442 1 www.pnas.org/cgi/doi/10.1 073/pnas.182436399 Fig. 4. Effects of set1 deletion on silencing and heterochromatin assembly. (A) iodine-staining assay. Wild-type and set1A colonies were sporu- lated on PMA+ medium at 25°C and exposed to iodine vapors before photography. (B and C) De- letion of set1 does not affect silencing at the mating-type region and centromere. The strains that contain ura4+ genes inserted at the mating- type locus (Kint2::ura4+) or centromere region of chromosome I (imr1R::ura4+ and otr1R::ura4+) were used for ChlP assays with antibodies to H3 Lys-4-methyl (K4), H3 Lys-9-methyl (K9), or Swi6 protein. DNAfrom ChlPorWCE (wholecell crude extract) was analyzed by a competitive PCR strat- egy, whereby one set of primers amplifies differ- ent-sized products from the ura4+ marker gene and the control ura4DS/E minigene at the en- dogenous euchromatic location. The ratios of ura4+ and control ura4DS/E signals present in ChlP and WCE were used to calculate relative enrichment, shown beneath each lane. NE indi- cates no enrichment observed. The ura4+ expres- sion levels at the mating-type and cen1 region were evaluated by dilution analysis. Cells were suspended in water, and 10-fold serial dilutions were spotted onto nonselective (N/S), counter- selective FOA, AA-URA medium, and grown 3 days before being photographed. analysis to measure the effect of setllk on stability of the epigenetic states. The setlA had a subtle effect on the ura4-off to ura4-on transition (7.9 x 10-4 per cell division), as compared with wild-type cells (8.4 x 10-4 per cell division). Moreover, we observed that setli\ caused a slight decrease (9.6 x 10-4 to 4.6 x 10-4 per cell division) in ura4-on to ura4-off conversion. This decrease in the "on" to "off" state conversion in setlA background could be caused by changes in the levels of trans-acting factors critical for establish- ment of the silenced state. In addition to silencing, the efficiency of mating-type switching is also regulated by formation of a heterochromatic structure at the mat locus. Mutations that affect silencing and heterochromatin assembly at the mating-type region adversely affect mating-type interconversion (47, 48~. We monitored the efficiency of mating- type switching in wild-type and setl i\ strains at the colony level by iodine-staining as described in the Materials and Methods section. We found that setl ~ has no effect on the efficiency of mating-type switching (Fig. 4A), consistent with our results that expression of the marker gene inserted at the silent mating-type region is not affected. SpSet1 Is Dispensable for Heterochromatin Assembly. Silencing at the mating-type region and centromeres depends on Swi6 protein, Noma ancl Grewa~
which is recruited to these loci through its binding to methylated H3 Lys-9 (46~. We therefore explored whether sell /\ affects H3 Lys-9 methylation and/or Swi6 at heterochromatic loci. ChIP assays with antibodies to Swi6 or H3 Lys-9-methyl were used. DNA recovered from immunoprecipitated chromatin fractions was quantitated by using competitive PCR, whereby one primer pair amplifies 694-bp and 426-bp products from full-length ura4+ inserted at the hetero- chromatic location and a mini-ura4 (ura4DS/E) at its endogenous euchromatic location. Our analysis revealed that the levels of H3Lys9 methylation and Swi6 at the silent mating-type region (`Kint2::ura4+) and centromeric repeats (~imrlR::ura4+ and otrlR::ura4+) of setl/\ cells were comparable to their wild-type counterparts (Fig. 4 B and C). As expected, setll\ abolished the preferential enrichment of H3 Lys-4 methylation at the euchro- matic ura4DS/E locus. Taken together, the results presented above suggest that Setl-mediated H3 Lys-4 methylation is not required for heterochromatin assembly in Sch. pombe. H3 Lys-4 Methylation Is Present Throughout the Cell Cycle. In com- parison to the acetylation of histones, lysine methylation is believed to be a relatively stable histone modification, which might serve as an epigenetic imprint for the long-term maintenance of chromatin states. If H3 Lys-4 methylation indeed serves as a molecular booLmark for inheritance of the active chromatin state, it is likely to be present throughout the cell cycle, even during mitosis. To address this issue, we performed immunofluorescence with anti- bodies to methylated H3 Lys-4 and tubulin used to visualize microtubules. As a control, we also studied H3 Ser-10 phosphor- ylation, which correlates with chromosome condensation during mitosis (49~. At the G2/M boundary, H3 Ser-10 phosphorylation was mainly localized to one or two discrete foci. However, Ser-10 phosphorylation spread throughout chromosomes by metaphase, the intensity of the signal diminished as cells enter anaphase, and almost all staining had disappeared in G~/S cells (Fig. 5~. In contrast to H3 Ser-10 phosphorylation, we found that H3 Lys-4 methylation levels remained unchanged throughout the cell cycle, including mitosis when chromosomes are highly condensed. Al- though changes in Lys-4 methylation at individual loci cannot be ruled out, H3 methylated at Lys-4 can be detected during different stages of the cell cycle. Because covalent modification by one enzyme can positively or negatively influence the efficiency of other enzymes responsible for modifying residues on the same histone tail, we also tested the effects of setl i\ on H3 Ser-10 phosphoryla- tion. As shown in Fig. 5, the loss of H3 Lys-4 methylation in setl lY strains did not affect H3 Ser-10 phosphorylation during mitosis. Interplay Between H3 Lys-4 Methylation and Histone Acetylation in Vivo. It has been shown that Lys-4 methylation of H3 is preferen- tially associated with H3 acetylation in S. cerevisiae, chicken, and HeLa cells (23, 24~. Here we examined whether SpSetl-mediated H3 Lys-4 methylation affects histone acetylation in vivo. Bulk histones prepared from wild-type and setll\ were subjected to Westem blot analysis with acetylation-site-specific H3 or H4 anti- bodies (Fig. 6~. Interestingly, we found that the acetylation levels of histone tails, in particular of H3 Lys-9 and H3 Lys-14, were significantly decreased in setl ~ cells when compared with wild-type cells (Fig. 6~. The setlZ\ also results in a subtle but consistent decrease in H4 Lys-5 and H4 Lys-12 acetylation. We also investi- gated the possible effects of mutations in the histone deacetylases clr3 and clr6 on H3 Lys-4 methylation in bulk histones in vivo. As shown recently, clr3 specifically affects H3 Lys-14 acetylation, whereas mutation in clr6 results in elevated acetylation levels at all residues tested on the histone H3 and H4 tails (Fig. 6; ref. 50~. Although H3 Lys-4 was slightly more methylated in clr6 mutant, mutation in clr3 had no effect on Lys-4 methylation. These data suggest that H3 Lys-4 methylation might help promote acety- lation of histones in the transcriptionally poised regions of the chromosomes. Noma and Grewal Fig. 5. H3 Lys4 methylation is present throughout the cell cycle. (A) Wild-type cells were stained with anti-H3K4Me (red), anti-tubulin TAT-1 antibody (green), and DAPI (blue). For each cell, the corresponding panels are placed below each other. (B) Wild-type and setli\ cells were stained with antibodies to phosphory- lated H3 Ser-10 (H3S10P). We also analyzed the effects of H3 Lys-4 methylation on H3 acetylation levels by ChIP assays. The loss of H3 Lys-4 methylation in setll\ cells causes 50-60% reduction in H3 Lys-14 acetylation levels at the constitutively expressed loci ura4, actl, and ade6, as compared with wild-type background cells (Fig. 7~. Consistent with decreased histone acetylation, we also observed that setl/\ causes subtle changes in ade6 expression. Because H3 Lys-4 methylation seems to globally affect active chromatin regions, changes in ade6 expression could not be quantified using Northern analysis because of the lack of appropriate controls. However, we observed that setl l\ cells carrying the ade6-216 allele at its endogenous chromo- somal location formed deep red colonies on adenine-limiting medium, as compared with the pink colonies formed by their wild-type counterparts (Fig. 7~. This phenotype indicating a de- crease in ade6 expression consistently segregated with the setl ~ in more than 30 tetrads. Taken together, these analyses suggest that H3 Lys-4 methylation might be involved in the maintenance of active chromatin configurations, presumably by facilitating the acetylation of histones. D· . Iscusslon Epigenetic control of higher-order chromatin assembly has been linked to the posttranslational covalent modifications of the histones tails. It has been formally suggested that distinct modifications on one or more of the histone tails act sequentially or in combination to form a histone code that is recognized by other chromatin-associated proteins (2, 7~. Although histone- modifying enzymes such as acetyltransferases and deacetylases have been identified and characterized from a number of organisms, the factors that regulate methylation of histones are only now being discovered. Recent studies have identified a novel class of protein methyltransferases defined by an evolu- tionarily well-conserved structure, the SET domain (5, 18~. In this paper, we report the structural features of at least nine SET domain proteins present in fission yeast. The data presented demonstrate that a highly conserved SET domain protein named PNAS | December '0, 2002 | vof. 99 | supp~. 4 | 16443
< ~'<3 \<6' ~ 66 Go< G\~t ura4 act! ade6 ura4 act1 ade6 IP W IP W IP W IP W IP W IP W 1~ Hi- ~ t ~ :~: :~ ~~;~ :~ - I : :::: : ~~ ~ ~ ~ ~ _ _____ ;~: __ ~ . . Enrlchment 11.0 3.63 7.83 5~87 2~03 5~27 ( fold) H3K4Me H3K9Ac H3K1 4Ac H4K5Ac H4K8Ac H4K1 2Ac H4K1 6Ac ~? T--?iSiSE?SE?~ T ~~: ?~: ~~ ? ?~ COYOTE? ~ 2 W?~ ?? 1~00 0.00 1~16 1.42 H3K4Me I ~ 1.00 0.34 1.13 1.88 H3K9Ac H3K1 4Ac H4K5Ac H4K8Ac H4K1 2Ac 1.00 0.34 2.65 ~ .15 1.00 0.65 0.76 1.86 1.00 0.76 0.87 2~05 1.00 0.50 0.84 3.82 1.00 0.76 1.32 1.17 Fig. 6. set1A strain exhibits decreased levels of acetylation at Iysine residues in H3 tail. (A) Bulk histones were prepared from wild-type, set1A, cir3-735, and cir6-1 strains. Ten micrograms of crude histoneswere separated by SDS/PAGE and subjected to Western analysis with antibodies specific for H3K4me, or histone H3 or H4 acetylated at the indicated Iysine residues. Coomassie staining (CBB) is shown as the loading control. Similar results were obtained in at least three independent experiments. (B) The intensity of the bands shown in A, quantified by using NIH IMAGE software, is summarized. SpSetl likely catalyzes the H3 Lys-4 methylation present at euchro- matic regions. Based on genetic and biochemical studies, the SET domain and its flanking pre- and postSET domains are required for catalytic activity of SUV39H1 and Clr4 methyltransferases (8, 18~. However, recent studies argue against the general requirement of the pre- and postSET domains for enzymatic activity. For example, human Set7/Set9 contains a SET domain but is devoid of the pre- and postSET motifs, and can efficiently methylate H3 Lys-4 in vitro and in vivo (30, 31), though it is possible that other sequences surrounding the SET domain might promote methyltransferase activity. The general consensus emerging is that the SET domain constitutes the catalytic motif, whereas flanking sequences might facilitate folding of the histone tails, providing specificity for a particular lysine residue. In this regard, the highly conserved n-SET motif and cysteine rich sequences flanking the SET domain in Setl orthologs are likely to be critical for their specificity to H3 Lys-4. (Fig. 1; ref. 28~. Based on conservation of the SET domain in yet uncharacterized SET domain proteins in fission yeast, it is probable that these proteins also act as methyltransferases. In many cases, SET domain proteins also contain other con- served motifs such as the chromodomain, PHD finger, and RRM domain. For example, the SUV39 family of proteins including Clr4 is known to contain a chromodomain at their amino terminus (40~. Although the SET domain of Clr4 and its surrounding sequences are sufficient for methyltransferase activity in vitro, both SET domains and chromodomains are required in vivo (8~. We found that the RRM domain of Sch. pombe Setl is required for its role in 16444 1 www.pnas.org/cgi/doi/10.1073/pnas.182436399 B WT set1^ ade6-216 - ade6 - ura4 act] K- region Fig. 7. setup results in a decrease of H3 Lys-14 acetylation levels at genes. (A) ChlP analysiswith antibodiesto acetylated H3 Lys-14 (H3K14Ac). DNAfragments from the immunoprecipitated (IP) fraction and whole cell crude extract (W) were analyzed by using multiplex PCR. Relative enrichment of DNA fragments corre- sponding to the coding regions of ura4, act1, or ade6 was examined. A DNA fragment from the silent mating-type region (K-region), which is known to lack H3 Lys-4-methyl and H3 Lys-14-acetyl modifications, was used as a control to normalize and calculate the relative enrichment of ura4, act1, and ade6 se- quences in IPed chromatin fractions. (B) Analysis of ade6-216 phenotype. Wild- type and set16 cells were streaked onto adenine-limiting yeast extract medium and incubated at 30°C for 2-3 days. Deep red or pink color of colonies indicates ade6-Z16 expression states. Colonies of setup strains were deeper red than wild type, indicating decrease in ade6 expression in mutant cells. H3 Lys-4 methylation in viva. The precise function of the RRM is not known, but it is possible that RRMs and chromodomains have related functions. These domains might help promote chromo- somal targeting of their respective proteins, either through protein- protein interactions or through their binding to RNA. Supporting the possible role for RNAs in chromatin assembly, our recent work suggests that RNA interference (RNAi) mechanisms, through which small RNAs silence cognate genes, might be required for the targeting of his/one-modifying activities to specific chromosomal domains in fission yeast. Specifically, RNAi machinery is essential for histone deacetylation and Clr4-mediated H3 Lys-9 methylation at centromeric repeats (I. Hall and S.I.S.G., unpublished data). Considering that certain chromodomains act as an RNA interaction module, it can be imagined that the binding of chrome- and/or RRM domains to RNA might guide the his/one-modifying activi- ties to homologous genomic sequences. In this scenario, RNA might provide specificity for the chromosomal targeting of these enzymes. Of course, it remains a possibility that chromodomains and RRM domains associated with these SET domain proteins are protein- protein interaction motifs. Future studies are necessary to address these possibilities. Recent work suggests that H3 Lys-4 methylation can perform dual functions in S. cerevisiae. Interestingly, Setl-mediated H3 Lys-4 methylation at rDNA and telomeres in S. cerevisiae is required for transcriptional silencing (27, 32, 33~. It has been suggested that H3 Lys-4 methylation in combination with other histone modifications could have a negative or positive effect on transcription at different chromosomal locations (27, 33~. In this regard, our analysis suggests that Sch. pombe Sell is not required for transcriptional silencing and heterochromatin assembly at centromeres or the silent mating-type interval. The differences in species of H3 Lys-4 that are mono-, di-, or tri-methylated might explain these seemingly contradictory results. A more likely explanation is that the histone code for silent chromatin assembly in S. cerevisiae and Sch. pombe is fundamentally different. H3 Lys-4 methylation is used in the process of silent Noma and Grewal
chromatin assembly in budding yeast, whereas H3 Lys-9 methyl- ation is used for formation of repressive chromatin structures in fission yeast and higher eukaryotes. Consistent with this idea, H3 Lys-9 methylation has not been detected in S. cerevisiae, and homologs for neither the H3 Lys-9 methyltransferase nor the Swi6/HP1 protein that recognizes this modification are present (27~. The mechanism by which H3 Lys-4 methylation modulates chromatin structure is not clear. The budding yeast Sell is found in a complex with homologs of the Drosophila protein Ash2 and Trithorax, which are essential for the stable maintenance of active gene expression states during development (28, 29~. Furthermore, recent studies have shown that H3 Lys-4 methylation is preferen- tially associated with euchromatic regions and is correlated with H3 acetylation (22-25~. Our analysis suggests that H3 Lys-4 methyl- ation and acetylation at the H3 tail interact in cis at euchromatic regions in vivo. We found that sell l\ results in a decrease in H3 acetylation levels. Furthermore, we also observed that a mutation in the Clr6 histone deacetylase, which displays broad specificity to lysine residues at H3 and H4 tails (ref. 50 and this study), results in 1. van Holde, K. E. (1989) Chromatin (Spunger, New York). 2. Strahl, B. D. & Allis, C. D. (2000) Nature (London) 403, 41-45. 3. Zhang, Y. & Reinberg, D. (2001) Genes Dev. 15, 2343-2360. 4. Berger, S. L. (2002) Curr. Opin. Genet. Dev. 12,142-148. 5. Kouzarides, T. (2002) Curr. Opin. Genet. Dev. 12,198-209. 6. Grunstein, M. (1997) Nature (London) 389, 349-352. 7. Turner, B. M. (2000) BioEssays 22, 836-845. 8. Nakayama, J., Rice, J. C., Strahl, B. D., Allis, C. D. & Grewal, S. I. S. (2001) Science 292, 110-113. 9. Bannister, A. J., Zegerman, P., Partridge, J. F., Miska, E. A., Thomas, J. O., Allshire, R. C. & Kouzarides, T. (2001) Nature (London) 410, 120-124. 10. Lachner, M., O'Carroll, D., Rea, S., Mechtler, K. & Jenuwein, T. (2001) Nature (London) 410, 116-120. 11. Jacobs, S. A., Tave~na, S. D., Zhang, Y., Briggs, S. D., Li, J., Eissenberg, J. C., Allis, C. D. & Khorasan~zadeh, S. (2001) EMBO J. 20, 5232-5241. 12. Schotta, G., Ebert, A., Krauss, V., Fischer, A., Hoffmann, J., Rea, S., Jenuwe~n, T., Dorn, R. & Reuter, G. (2002) EMBO J. 21, 1121-1131. Dhalluin, C., Carlson, J. E., Zeng, L., He, C., Aggarwal, A. K. & Zhou, M. M. (1999) Nature (London) 399, 491-496. 14. Ma, H., Baumann, C. T., Li, H., Strahl, B. D., Rice, R., Jelinek, M. A., Aswad, D. W., Allis, C. D., Hager, G. L. & Stallcup, M. R. (2001) Cu~r. Biol. 11, 1981-1985. 15. Strahl, B. D., Briggs, S. D., Brame, C. J., Caldwell, J. A., Koh, S. S., Ma, H., Cook, R. G., Shabanowitz, J., Hunt, D. F., Stallcup, M. R. & Allis, C. D. (2001) Cu~r. B'ol. 11, 996-1000. 16. Wang, H., Huang, Z. Q., Xia, L., Feng, Q., Erdjument-Bromage, H., Strahl, B. D., Briggs, S. D., Allis, C. D., Wong, J., Tempst, P. & Zhang, Y. (2001) Science 293, 853-857. 17. Tschiersch, B., Hofmann, A., Krauss, V., Dorn, R., Korge, G. & Reuter, G. (1994) EMBO J. 13, 3822-3831. 18. Rea, S., Eisenhaber, F., O'Carroll, D., Strahl, B. D., Sun, Z. W., Schmid, M., Opravil, S., Mechtler, K., Ponting, C. P., Allis, C. D. & Jenuwe~n, T. (2000) Nature (London) 406, 593-599. 19. Tachibana, M., Sugimoto, K., Fukushima, T. & Sh~nkai, Y. (2001) J. Biol. Chem. 276, 25309-25317. 20. Yang, L., Xia, L., Wu, D. Y., Wang, H., Chansky, H. A., Schubach, W. H., Hickstein, D. D. & Zhang, Y. (2002) Oncogene 21,148-152. 21. Schultz, D. C., Ayyanathan, K., Negorev, D., Maul, G. G. & Rauscher, F. J., III. (2002) Genes Dev. 16, 919-932. 22. Noma, K., Allis, C. D. & Grewal, S. I. (2001) Science 293, 1150-1155. 23. Litt, M. D., Simpson, M., Gaszner, M., Allis, C. D. & Felsenfeld, G. (2001) Science 293, 2453-2455. 24. Strahl, B. D., Ohba, R., Cook, R. G. & Allis, C. D. (1999) Proc. Natl. Acad. Sci. USA 96, 14967-14972. 25. Boggs, B. A., Cheung, P., Heard, E., Spector, D. L., Ch~nault, A. C. & Allis, C. D. (2002) Nat. Genet. 30, 73-76. Noma and Grewal a slight increase in H3 Lys-4 methylation. Based on these results, it is possible that H3 Lys-4 methylation either plays a facilitatory role in histone acetylation or it protects transcriptionally active regions from the effects of repressive chromatin remodelling activities such as histone deacetylases, or both. Supporting our in vivo analysis, it has been shown recently that H3 Lys-4 methylation facilitates subsequent acetylation of the histone tail by acetyltransferases in vitro (30~. Furthermore, methylation of H3 Lys-4 interferes with interactions between the NuRD histone deacetylase and H3 tail, as well as precludes H3 Lys-9 methylation by SUV39H1 (31, 51~. In conclusion, the results presented in this paper further extend the histone code hypothesis and suggest that Setl-mediated H3 Lys-4 methylation pr~mar~ly acts in the maintenance of active chromatin configurations at euchromatic chromosomal domains in fission yeast. We thank John Aris for providing Nopl antibody, and Keith Gull for TAT1 antibody. We also thank Rui-Ming Xu, Ira Hall, Asra Malikzay, and Gurumurthy Shankaranarayana for critical reading of the manu- script, and Winship Herr for helpful discussions. This work was sup- ported by National Institutes of Health Research Grant GM59772. 26. Heard, E., Rougeulle, C., Arnaud, D., Avner, P., Allis, C. D. & Spector, D. L. (2001) Cell 107, 727-738. 27. Briggs, S. D., Bryk, M., Strahl, B. D., Cheung, W. L., Davie, J. K, Dent, S. Y., Winston, F. & Allis, C. D. (2001) Genes Dev. 15, 3286-3295. 28. Roguev, A., Schaft, D., Shevchenko, A., Pijnappel, W. W., Wilm, M., Aasland, R. & Stewart, A. F. (2001) EMBO J. 20, 7137-7148. 29. Nagy, P. L., Griesenbeck, J., Kornberg, R. D. & Cleary, M. L. (2002) Proc. Natl. Acad. Sci. USA 99, 90-94. 30. Wang, H., Cao, R., Xia, L., Erdjument-Bromage, H., Borchers, C., Tempst, P. & Zhang, Y. (2001) Mol. Cell 8, 1207-1217. 31. Nishioka, K, Chuikov, S., Sarma, K., Erdjument-Bromage, H., Allis, C. D., Tempst, P. & Reinberg, D. (2002) Genes Dev. 16, 479-489. 32. Nislow, C., Ray, E. & Pillus, L. (1997) Mol. Biol. Cell 8, 2421-2436. 33. Bryk, M., Briggs, S. D., Strahl, B. D., Curcio, M. J., Allis, C. D. & Winston, F. (2002) Curr. Biol. 12,165-170. 34. Bahler, J., Wu, J. Q., Longtine, M. S., Shah, N. G., McKenzie, A., Steever, A. B., Wach, A., Philippsen, P. & Pringle, J. R. (1998) Yeast 14, 943-951. 35. Grewal, S. I. & Klar, A. J. (1997) Genetics 146, 1221-1238. 36. Ekwall, K. & Partridge, J. F. (1999) in Chromosome StructuralAnalysis:A Practical Approach, ed. Bickmore, W. (Oxford Univ. Press, Oxford), pp. 47-57. 37. Nakayama, J., Klar, A. J. & Grewal, S. I. (2000) Cell 101, 307-317. 38. Wood, V., Gwilliam, R., Rajandream, M. A., Lyne, M., Lyne, R., Stewart, A., Sgouros, J., Peat, N., Hayles, J., Baker, S., et al. (2002) Nature (London) 415, 871-880. 39. Strahl, B. D., Grant, P. A., Briggs, S. D., Sun, Z. W., Bone, J. R., Caldwell, J. A., Mollah, S., Cook, R. G., Shabanowitz, J., Hunt, D. F. & Allis, C. D. (2002) Mol. Cell. Biol. 22,1298-1306. 40. Ivanova, A. V., Bonaduce, M. J., Ivanov, S. V. & Klar, A. J. (1998) Nat. Genet. 19, 192-195. 41. Burd, C. G. & Dreyfuss, G. (1994) Science 265, 615-621. 42. Aasland, R., Gibson, T. J. & Stewart, A. F. (1995) Trends Biochem. Sci. 20, 56-59. 43. Aris, J. P. & Blobel G. (1988) J. Cell Biol. 107, 17-31. 44. Grewal, S. I. (2000) J. Cell. Physiol. 184, 311-318. 45. Allshire, R. C., Nimmo, E. R., Ekwall, K, Javerzat, J. P. & Cranston, G. (1995) Genes Dev. 9, 218-233. 46. Grewal, S. I. S. & Elgin, S. C. (2002) Curr. Opin. Genet. Dev. 12,178-187. 47. Ekwall, K. & Ruusala, T. (1994) Genetics 136, 53-64. 48. Thon, G., Cohen, A. & Klar, A. J. (1994) Genetics 138, 29-38. 49. Hendzel, M. J., Wei, Y., Mancini, M. A., Van Hooser, A., Ranalli, T., Brinkley, B. R., Bazett-Jones, D. P. & Allis, C. D. (1997) Chromosoma 106, 348-360. 50. Bjerling, P., Silverstein, R. A., Thon, G., Caudy, A., Grewal, S. & Ekwall, K. (2002) Mol. Cell. Biol. 22, 2170-2181. 51. Zegerman, P., Canas, B., Pappin, D. & Kouzarides, T. (2002) J. Biol. Chem. 277, 11621-11624. PNAS | December 10, 2002 1 vol. 99 | suppl. 4 | 16445
