swr1 complex poises heterochromatin boundaries for

MOLECULAR AND CELLULAR BIOLOGY, May 2010, p. 2391–2400 Vol. 30, No. 100270-7306/10/$12.00 doi:10.1128/MCB.01106-09Copyright © 2010, American Society for Microbiology. All Rights Reserved.

SWR1 Complex Poises Heterochromatin Boundaries forAntisilencing Activity Propagation�†

Bo O. Zhou,1 Shan-Shan Wang,1 Lu-Xia Xu,1 Fei-Long Meng,1 Yao-Ji Xuan,1 Yi-Min Duan,1Jian-Yong Wang,1 Hao Hu,2 Xianchi Dong,1,3 Jianping Ding,1,3 and Jin-Qiu Zhou1*

State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology,1 CAS-MPG Partner Institute forComputational Biology,2 and Research Center for Structural Biology, Institute of Biochemistry and Cell Biology,3

Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences and Graduate School of theChinese Academy of Sciences, Shanghai, China

Received 1 September 2009/Returned for modification 10 October 2009/Accepted 14 March 2010

In eukaryotes, chromosomal processes are usually modulated through chromatin-modifying complexes thatare dynamically targeted to specific regions of chromatin. In this study, we show that the chromatin-remodelingcomplex SWR1 (SWR1-C) uses a distinct strategy to regulate heterochromatin spreading. Swr1 binds in astable manner near heterochromatin to prepare specific chromosomal regions for H2A.Z deposition, which canbe triggered by NuA4-mediated acetylation of histone H4. We also demonstrate through experiments with Swc4,a module shared by NuA4 and SWR1-C, that the coupled actions of NuA4 and SWR1-C lead to the efficientincorporation of H2A.Z into chromatin and thereby synergize heterochromatin boundary activity. Our resultssupport a model where SWR1-C resides at the heterochromatin boundary to maintain and amplify antisilenc-ing activity of histone H4 acetylation through incorporating H2A.Z into chromatin.

Both histone modification and chromatin-remodeling en-zymes affect nucleosome dynamics in Saccharomyces cerevisiae.Histone modification enzymes covalently modify various his-tones through chemical means, such as lysine acetylation,serine phosphorylation, lysine and arginine methylation, ubiq-uitylation, deimination, ADP ribosylation, and proline isomer-ization (18). Chromatin-remodeling enzymes use the energyof ATP hydrolysis to induce nucleosome mobility or disrupthistone-DNA interactions (38). NuA4 is a histone acetyltrans-ferase complex responsible for global histone H4 acetylation(1, 12, 39). The catalytic subunit of NuA4 is Esa1, which asso-ciates with Yng2p and Epl1p to form a smaller complex in vivonamed Piccolo NuA4 (8). Piccolo NuA4 has a nontargetedhistone acetyltransferase activity (8), while the full NuA4 com-plex, united by the Eaf1 protein, carries out site-specific andtargeted acetylation reactions (2). The SWR1 complex(SWR1-C) is an ATP-dependent chromatin-remodeling com-plex that catalyzes the exchange of H2A-H2B with H2A.Z-H2B (H2A.Z is also known as Htz1) (17, 31). Notably, theplatform proteins Eaf1 and Swr1 in NuA4 and SWR1-C, re-spectively, not only exhibit considerable sequence similarity toeach other but also display strong and distinct homologies tohuman p400/Domino (2), suggesting an evolutionary mergingof functions in higher eukaryotes.

Heterochromatin was originally defined cytologically as ge-nome blocks where the structure of chromatin is highly con-densed throughout the cell cycle and hence remains transcrip-tionally silenced (36). The main structural proteins of silent

chromatin in S. cerevisiae are known as silent information reg-ulator (Sir) proteins and include Sir2, Sir3, and Sir4. The Sircomplex propagates along chromatin to form an ordered andcompact structure that is usually restrictive to transcription(35).

Heterochromatin spreading is limited by boundary elementsbetween silenced and active chromatin (35). These barriersappear to fall into two classes (6). The first class of barrierconfers a structural characteristic to the boundary region, suchas gaps in nucleosomes or association with nuclear pores. Thesecond class of barrier is more dynamic and involves modifi-cation of heterochromatin-proximal regions in a wide area.Local chromatin dynamics can counteract the repressive Siractivity and prevent further spread of silencing.

Both SWR1-C and NuA4 have been linked to antisilencingfunctions at heterochromatin boundaries (11, 24, 27), however,their functional interplay in antisilencing is not yet clear. In thisstudy, we dissected the individual and coupled actions ofSWR1-C and NuA4 at antisilencing regions. By studying thefunctions of Swc4, we discovered an intrinsic relationship be-tween NuA4 and SWR1-C in the establishment of a chromatinboundary.

MATERIALS AND METHODS

Yeast strains and reagents. Antibodies, yeast strains, and primers used in thisstudy are listed in Tables S1, S2 and S3 in the supplemental material, respec-tively.

ChIP assay. Most chromatin immunoprecipitation (ChIP) assays were per-formed as described previously (47). For Esa1 ChIP, we followed an optimizedprotocol (20). ChIP products were directly analyzed by real-time PCR usingSYBR green as a label (Toyobo). Enrichment values were the average immu-noprecipitation (IP)/whole-cell extract (WCE) ratios from triplicate samples withstandard errors of the means (SEM). Relative occupancy of protein/modificationat indicated loci was the enrichment value at indicated loci relative to the controlloci (PRP8 for hemagglutinin [HA]-tagged Swr1 [Swr1-HA], Htz1-HA, and tan-dem affinity purification (TAP)-tagged Swc4 [Swc4-TAP], TELVIR for H4 and

* Corresponding author. Mailing address: State Key Laboratory ofMolecular Biology, Institute of Biochemistry and Cell Biology, 320Yue-Yang Road, Shanghai 200031, China. Phone: 86-21-54921076.Fax: 86-21-54921075. E-mail: [email protected].

† Supplemental material for this article may be found at http://mcb.asm.org/.

� Published ahead of print on 22 March 2010.

2391

on April 7, 2018 by guest

http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

AcH4K8, region R for Myc-tagged Esa1 [Esa1-Myc], and ACT1 for Sir2-Myc andSir3-Myc).

Coimmunoprecipitation assay. Yeast whole-cell extract was prepared from�150 ml yeast culture (optical density at 600 nm [OD600] � 1.0) and thenincubated with 25-�l of IgG Sepharose beads (Amersham Biosciences) for 90min at 4°C. Beads were then pelleted and washed three times with 1 ml ofphosphate-buffered saline (PBS). After washing, the beads were resuspended inSDS sample buffer and subjected to SDS-PAGE electrophoresis and Westernblotting.

Immunostaining assay. Immunostaining experiments were performed as pre-viously described (47). Confocal microscopy was performed on a Leica TCS SP2microscope with a 63� lambda blue objective (oil). Similar filtration and thresh-old levels were standardized for all images.

RESULTS

SWR1-C is abundant at antisilencing regions as well assilenced HMR loci. To understand the antisilencing mecha-nisms for SWR1-C-mediated H2A.Z deposition and NuA4-

mediated histone H4 acetylation, we first examined theirrelative abundances at regions near heterochromatin. Usingchromatin immunoprecipitation (ChIP) combined with real-time PCR, we detected the relative abundance of HA-taggedSwr1, HA-tagged H2A.Z, and acetylated histone H4 at regionsnear HMR and within the subtelomeric regions of chromo-somes III and XIV (Fig. 1A). To distinguish between NuA4-mediated histone H4 acetylation and histone H4K16 acetyla-tion mediated by SAS complex (26), we used anti-acetyl H4lysine 8 (AcH4K8) antibody instead of anti-tetra-acetylated H4antibody. The specificity of anti-AcH4K8 antibody was vali-dated by Western blotting and ChIP (Fig. 1B and C). Asexpected, the occupancy of Swr1 was high at most of theeuchromatic regions we tested. The occupancy of Swr1 wasundetectable at distal telomeres but was surprisingly high atHMR loci (Fig. 1D, regions D, E, F and G), where the nucleo-

FIG. 1. SWR1-C is abundant at antisilencing regions as well as silenced HMR loci. (A) Schematic diagram of four major antisilencingchromosomal domains. Shown are the HMR silent mating-type cassette and its proximal regions and regions near the ends of chromosome III-R,XIV-R, and IX-R. Depicted are the locations of genes used for mRNA analysis. A to Z are the locations and designations of PCR primers usedfor ChIP analysis. The loci where the lac operator array was integrated are also indicated. (B) Whole-cell extracts from NSY429 strains expressingwild-type histone H4 and an arginine substitution mutant of H4 Lys8 were subjected to Western blot analysis with the specific antibodies indicatedon the right. (C) ChIP assay on chromatin from wild-type and H4K8R mutant strains. IPs were done with increasing amounts of anti-acetyl H4K8antibody: 1:200 (lane 1), 1:200 (lane 2), 1:50 (lane 3), and 1:25 (lane 4). (D to F) ChIP experiments with Swr1-HA (D), Htz1-HA (E), and AcH4K8(F) in cells with various genetic backgrounds, as indicated. Primer sets with the amplified region are denoted in panel A. The gray areas arepredicted silenced regions, typically with hypoacetylated histones and abundant Sir proteins. The AcH4K8 and H2A.Z signals were corrected byH4 density. Error bars represented standard errors of the means for three independent experiments. Values that are greater than 1 indicate moreenrichment at the indicated locus in the mutant compared than in the wild type. (G) ChIP experiments with Htz1-HA, AcH4K8, and Swr1-HAat intergenic regions as indicated. The AcH4K8 and H2A.Z signals were corrected by H4 density.

2392 ZHOU ET AL. MOL. CELL. BIOL.


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

somes are largely hypoacetylated. This result was unexpected,because a previous study had shown that SWR1-C was re-cruited to hyperacetylated chromatin domains through its Bdf1subunit (25, 46). It is possible that a subunit(s) other than Bdf1also functions to stabilize the association of SWR1-C complexwith chromatin.

We next sought to determine if changes in heterochromatinstructure affected chromatin association of Swr1 and H2A.Z.Deletion of SIR2, which encodes one of the structural compo-nents of heterochromatin, disrupted heterochromatin silencingof both HM loci and telomeres (35). Interestingly, in the ab-sence of Sir2, the occupation pattern of Swr1 at most regions,including distal telomeres, was barely altered (Fig. 1D). Basedon these data, we conclude that SWR1-C binds to chromatin ina manner that is independent of the chromatin silencing status.

Preassociation of SWR1-C with chromatin is essential butnot sufficient for H2A.Z incorporation. Although SIR2 dele-tion caused a dramatic elevation of H2A.Z levels at HMR lociwhere Swr1 was present (Fig. 1D and E, regions E, F, G, andH), it did not lead to an increase in H2A.Z levels at distaltelomeres where Swr1 was absent (Fig. 1D and E, regions R, S,T, X, Y, and Z), suggesting that the chromatin binding ofSWR1-C contributed to the replacement of H2A-H2B withH2A.Z-H2B. In addition, we found that Swr1 occupied manymore chromatin regions than H2A.Z. For example, Swr1 washighly enriched at HMR, HML, and regions I and J, whileH2A.Z was not enriched over background levels at those sites(Fig. 1D and E). Thus, the preassociation of SWR1-C withchromatin is not the sole driving force for H2A.Z deposition.

NuA4-mediated H4 acetylation is required for H2A.Z incor-poration at Swr1-enriched regions. Next, we considered ifNuA4-dependent H4 acetylation contributed to H2A.Z depo-sition. Interestingly, the relative abundance of H4K8 acetyla-tion and H2A.Z binding correlated nicely at most of the chro-mosomal loci we examined (Fig. 1E and F), implying apotential requirement of histone H4 acetylation for H2A.Zincorporation into nucleosomes. Additionally, EAF1 deletion,which eliminated H4K8 acetylation (Fig. 1F), greatly reducedH2A.Z levels in the corresponding boundary regions (Fig. 1E).Moreover, when a set of randomly selected intergenic regionswas examined for H2A.Z occupation, we surprisingly foundthat H2A.Z was specifically incorporated into regions wherewe observed significantly high levels of Swr1 and acetylatedH4K8 (Fig. 1G, regions CLB6 and DYN1). Therefore, NuA4-dependent H4 acetylation is likely required for SWR1-C-me-diated H2A.Z replacement of H2A at heterochromatin bound-aries.

Targeted H4 acetylation triggers H2A.Z deposition at Swr1-enriched regions. To investigate further the interplay betweenNuA4-mediated H4 acetylation and SWR1-C-mediated H2A.Zdeposition, we adopted an artificial recruitment system (34)where an array of lac operator sites (lacO) was integrated intoa locus �4 kb from HMR or a locus �4 kb from the end ofchromosome XIV-R (Fig. 1A). Esa1 or Swr1 was artificiallytargeted to the lacO loci by being fused to the LacI repressor.Western blot analysis with anti-LacI antibody showed that theexpression levels of fusion proteins were comparable to eachother (Fig. 2A). As expected, tethered Esa1 acetylated chro-

FIG. 2. NuA4-mediated H4 acetylation triggers SWR1-C-mediated H2A.Z deposition. (A) Western blotting was performed with whole-cellextracts from strains with exogenous LacI fusion genes indicated on the top. Antibodies used are indicated on the right. (B to G) The lacO arraywas inserted into a locus �4 kb from the end of chromosome XIV (B to D) and a locus �4 kb from HMR (E to G), as depicted in Fig. 1A.Overexpression plasmids harboring lacI (wild type) or the ESA1-lacI or SWR1-lacI fusion were introduced into the strains that are indicated in eachpanel. The relative abundance of AcH4K8 (B and E), Htz1-HA (C and F), or Sir2-Myc (D and G) was then analyzed by the ChIP experiment asfor Fig. 1E and F.

VOL. 30, 2010 SWR1-C PROPAGATES ANTISILENCING ACTIVITY 2393


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

matin near lacO (Fig. 2B and E). In contrast, tethered Swr1increased H2A.Z incorporation into chromatin in an inefficientway (Fig. 2C and F). Consistently, tethered Esa1 but not Swr1attenuated Sir2 binding at regions near lacO (Fig. 2D and G).This result coincided with the above statement that chromatinrecruitment of SWR1-C per se was insufficient for H2A.Z depo-sition. Notably, recruited Esa1 also gave rise to a markedincrease in H2A.Z levels at HMR-proximal regions (Fig. 2F).The increase was apparently Swr1 dependent, because tar-geted Esa1 failed to boost H2A.Z levels at telomere XIV-R,where Swr1 was absent (Fig. 1D and 2C). Hence, Esa1 target-ing was able to directly induce H2A.Z incorporation at Swr1-enriched regions.

Since both histone H4 and H2A.Z were substrates for NuA4(1, 3), we wondered which event, either H4 acetylation orH2A.Z acetylation, was the signal that led to SWR1-C-dependent H2A.Z deposition. To this end, we mutated spe-cific lysines on H4 or H2A.Z to arginines (H4K5,8,12R andHTZ1K14R). A lysine-to-arginine mutation was unable to beacetylated and thereby mimicked an unacetylated state. Inter-estingly, in H4K5,8,12R cells, tethered Esa1 reduced H2A.Zincorporation to a level much lower than that observed inwild-type cells (Fig. 2F). In contrast, the HTZ1K14R mutationhad a modest influence on H2A.Z incorporation induced byEsa1 tethering (Fig. 2F). These data support the idea thatH2A.Z K14 acetylation is not required for its deposition atboundaries (3), and NuA4-mediated H4K5,8,12 acetylationtriggers H2A.Z deposition at heterochromatin boundaries.

Swc4 is required for both NuA4-mediated histone acetyla-tion and SWR1-C-mediated H2A.Z incorporation at antisi-lencing regions. We next considered the genetic interactionbetween NuA4 and SWR1-C in antisilencing. Unfortunately,double mutation of the key components of NuA4 and SWR1-C(e.g., eaf1� swr1�) caused synthetic lethality (17), thereby in-troducing challenges to the direct assessment of the geneticinteraction between these two complexes. In a large-scalescreening for mother-selfless genes in the BY4743 strain back-ground (28), we coincidentally discovered that cells carrying adeletion of the SWC4 gene, which encodes a shared subunit ofNuA4 and SWR1-C, were viable, although exhibiting a slow-growth phenotype (Fig. 3A to D). The Swc4 protein containsan N-terminal SANT domain and a C-terminal Yaf9-interact-ing domain (YID) (7) (Fig. 3E). A SANT domain deletionmutant (swc4sant�) phenocopied swc4� cells in yeast growth(Fig. 3F), suggesting an essential role for the SANT domain inmediating Swc4’s function. In addition, comparison of the tran-scriptional profiles of swc4yid� cells to that of htz1� cells re-vealed a marked overlap (Fig. 3G), suggesting that the YIDdomain mediates Swc4’s functions in SWR1-C-dependentH2A.Z deposition. Using analytical-scale affinity purification,we observed that Swc4 stably associated with Esa1 and Swr1(Fig. 3H and I). Deletion of the SANT domain or the YIDdomain did not dissociate Swc4 from either of the main com-plexes (Fig. 3H and I).

Proteins containing SANT domains had central roles inmany chromatin-modifying complexes (9, 10). SWC4 deletiondid not detectably alter global acetylation level of histone H4(Fig. 3J), suggesting that Swc4, as is the case for Eaf1 (2),regulates the site-specific roles of the NuA4 complex. Wetherefore asked whether Swc4 contributed to the actions of

NuA4 and SWR1-C specifically at boundaries. ChIP experi-ments were performed to monitor the chromatin changes inswc4� cells. Interestingly, deletion of SWC4 resulted inhypoacetylation of histone H4K8 and dramatic reduction ofH2A.Z levels at all the boundary regions we tested (Fig. 4Aand B), most likely because deletion of SWC4 reduced bindingof Swr1 and Esa1 with chromatin (Fig. 4C and D). Addition-ally, Swc4 itself was associated with regions (Fig. 4E) that werealso bound by Esa1 or Swr1 (Fig. 4E). These data indicate thatSwc4 directly regulates the chromatin association of bothNuA4 and SWR1-C and thereby is indispensable for histoneH4 acetylation and H2A.Z incorporation at heterochromatinboundaries.

Increased H2A.Z incorporation at subtelomeric regions re-inforces the antisilencing activity of histone H4 acetylation.We have shown above that H2A.Z deposition at chromosomalregions adjacent to heterochromatin required the coordinatedaction of SWR1-C and NuA4. Since Swc4 stabilized bothSWR1-C and NuA4 at boundaries, tethered Swc4 might conferan even higher activity for H2A.Z incorporation at these sites.To test this idea, we utilized the lacO-LacI system again totarget LacI-fused Swc4 to the subtelomeric region of chromo-some XIV, where both SWR1-C and NuA4 were physiologi-cally absent (Fig. 4C and D). As expected, targeted Swc4 in-creased subtelomeric H4K8 acetylation (Fig. 4F). Strikingly,Swc4 targeting also boosted H2A.Z levels to greater than 10-fold above the basal level at the regions near lacO (Fig. 4G),which was hardly seen when Esa1 or Swr1 was targeted (Fig.2C). Thus, Swc4 likely recruits both NuA4 and SWR1-C andcouples the actions of SWR1-C and NuA4 toward H2A.Zdeposition.

To examine whether Swc4 targeting conferred higher anti-silencing activity than Esa1 or Swr1 targeting did, we examinedSir2 association at chromosomal regions near lacO and foundthat subtelomeric association of Sir2 was markedly reducedwhen Swc4-LacI was tethered to the lacO loci (Fig. 4H). Theabundance of Sir2 at region Z in the Swc4-tethering strain waseven lower than that in the Esa1-tethering strain (Fig. 4H and2D). Additionally, the mRNA levels of PAU6, a gene locatedwithin the subtelomeric region of chromosome XIV, werehighest in the Swc4-targeting strain (Fig. 4I). Thus, tetheredSwc4 conferred higher antisilencing activity than tethered Esa1or tethered Swr1. The increased antisilencing activity was likelyattributable to the increased H2A.Z deposition, because teth-ered Swc4 elevated histone H4 acetylation to a level similar tothat observed in the Esa1-LacI strain (Fig. 4F and 2B). Takentogether, these data support the argument that incorporationof H2A.Z at the subtelomeric region enhances antisilencingactivity normally attributed to histone H4 acetylation.

Synergistic action of NuA4 and SWR1-C confers broaderantisilencing activity. To further elucidate the functional rela-tionship between NuA4 and SWR1-C in antisilencing, we car-ried out ChIP experiments to examine the distribution of Sir2and Sir3 in eaf1�, swr1�, and swc4� cells. Under physiologicalconditions, chromatin-bound Sir proteins were primarily local-ized to silent chromatin regions, with limited spreading outsidethe boundaries (Fig. 5A and B, gray bars). In eaf1� cells, thebinding of Sir proteins was increased in both heterochromatinand its nearby regions (Fig. 5A and B, pink bars), suggestingthat NuA4 generally counteracts the association of Sir com-



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

FIG. 3. Swc4 is a nonessential subunit shared by NuA4 and SWR1-C. (A and B) Shown is the tetrad dissection analysis of heterozygousSWC4/swc4� diploid yeast strains in a BY4743 background. Specific genetic backgrounds are shown on the top. Spores were grown on a yeastextract-peptone-dextrose (YPD) plate at 30°C for �60 h prior to imaging. Black circles were used to indicate the swc4� spores that had beenvalidated by PCR and restreaking on selection plates. (C) The heterozygous SWC4/swc4�::HIS3 strain was obtained by mating wild-type BY4742with the BY4741-derived swc4�::HIS3 strain. Tetrad dissection was then performed as for panel A. (D) Wild-type SWC4 carried on a URA3plasmid was introduced into the BY4743 (SWC4/swc4�::HIS3) strain, and tetrads were dissected. Spores were then restreaked on minimummedium containing 5-fluoroorotic acid (5-FOA). Cells were grown for �60 h at 25°C prior to imaging. URA3 plasmids loss was then verified byrestreaking colonies on a YC plate lacking uracil (data not shown). (E) Schematic diagram of Swc4 domain structure. Dark boxes indicate theSANT domain and the Yaf9-interacting domain. Numbers indicate the positions of truncation mutations for swc4sant� and swc4yid�. (F) Restreak-ing of isogenic strains on a YC plate. Cells were grown for �60 h at 30°C prior to imaging. (G) Color representation of all genes that weresignificantly reduced in expression (�1.5-fold) in mutant cells. Each column represents data from an independent microarray experiment thatcompared genome-wide expression in mutant cells of the indicated genotype to that in wild-type cells. Each row represents the changes inexpression of a single gene across the experiments. Change in expression, measured as the log2 mutant/wild-type expression ratio, is indicatedaccording to the color scale shown. (H and I) Analytical-scale TAP purifications. Captured TAP-tagged Swc4 was detected with anti-TAP antibody.Swc4-, Swc4sant�-, or Swc4yid�-associated Myc-tagged Esa1 (H) or Swr1 (I) was detected with anti-Myc antibody. (J) Nucleosomal histones fromwild-type and mutant strains (labeled on top) were purified and analyzed by Western blotting using antibodies indicated at left.



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

plexes with chromatin. In swr1� cells, the relative enrichmentof both Sir2 and Sir3 was specifically increased at regionsadjacent to heterochromatin, and this increase was accompa-nied by a modest reduction of Sir protein levels within hetero-chromatin (Fig. 5A and B, blue bars), a result that is typical ofa Sir complex overspreading phenotype (27). Notably, in swc4�cells, Sir proteins were now enriched in the adjacent euchro-matin, and the levels of Sir proteins in this euchromatin weremuch higher than we observed in swr1� or eaf1� cells (Fig. 5Aand B). Since all three mutants we tested did not alter theexpression of any of the major silencing-related genes, includ-ing RAP1, SIR1, SIR2, SIR3, SIR4, and HHF1 (data notshown), we concluded that these mutations had not likely af-fected Sir spreading in a nonspecific and indirect manner.Collectively, these data support the idea that Swc4 coordinatesand integrates the effects of NuA4 and SWR1-C on the asso-ciation of the Sir complex with chromatin.

To further examine heterochromatin spreading in swr1�,eaf1�, and swc4� cells, the subcellular localization of Myc-tagged Sir2 was examined by immunostaining using an anti-Myc antibody. In wild-type cells, we detected both a strongsignal within a restricted nuclear subdomain and a weakerpunctate pattern of staining (Fig. 5C). The punctate Sir2 stain-ing likely represented telomere-localized Sir2 (14). In eithereaf1� or swr1� cells, there were no distinguishable changes tothe Sir2-Myc staining pattern compared with that in wild-typecells (Fig. 5C). However, in the swc4� mutant, the Sir2 stainingpattern differed from that of wild-type cells by the presence of

more nuclear foci (Fig. 5C). Simultaneous immunostaining ofSir2 and Nop1, the yeast fibrillarin homolog, showed theircolocalization and indicated an intact nucleolus in all the mu-tant strains (Fig. 5C). Please note for Fig. 5C that the weakNop1 staining of one of the swc4� cells was not a result ofdispersed or diminished nucleolar Nop1 but occurred becausethe nucleolus was slightly out of focus. These data, coupledwith our ChIP data, have led us to postulate that the increasein punctate Sir2 staining in swc4� cells was most likely a resultof the ectopic spread of the Sir2 protein into a wider range ofeuchromatic regions.

Previous work had shown that the ectopic spread of Sirproteins resulted in a transcriptional repression of genes adja-cent to heterochromatin (27). We also examined the transcrip-tional changes in genes adjacent to heterochromatin in ourmutant strains. The mRNA levels of representative genes prox-imal to or distant from heterochromatin (Fig. 1A) were ana-lyzed by real-time quantitative PCR. SIR2 deletion elevatedthe expression of the heterochromatin-proximal, but not thedistal, genes (Fig. 5D), revealing a limited ectopic spreading ofSir-mediated silencing proteins under normal physiologicalconditions. In swr1�, eaf1�, or swc4� mutant cells, the mRNAlevels expressed by most genes we tested were decreased (Fig.5D), suggesting that both NuA4 and SWR1-C are required fornormal expression of genes near heterochromatin. Notably,the SWC4 deletion strain showed the highest repressive effectson genes adjacent to heterochromatin. For example, the ex-pression of the GTT1 gene, which was �20 kb from the chro-

FIG. 4. Swc4 couples the actions of both NuA4 and SWR1-C on H2A.Z incorporation at boundaries. (A to E) Relative enrichments, as assayedfor Fig. 1F, of AcH4K8 (A), Htz1-HA (B), Esa1-Myc (C), Swr1-HA (D), or Swc4-TAP (E) in cells with the indicated genetic backgrounds. (F toI) The lacO array was inserted into a locus �4 kb from the end of chromosome XIV, as depicted in Fig. 1A. Overexpressing plasmids harboringSWC4-lacI were introduced into the modified strain. The relative abundances of AcH4K8 (F), Htz1-HA (G), and Sir2-myc (H) in the SWC4-lacIstrain was analyzed alongside the strains shown in Fig. 2 in a parallel ChIP experiment. mRNA levels of the PAU6 gene in different geneticbackgrounds was analyzed by real-time PCR. Shown are the average expression ratios (normalized by that of ACT1) relative to wild-type expressionof the indicated genotypes.



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

FIG. 5. Swc4 mediates a synergistic action of NuA4 and SWR1-C toward preventing Sir-dependent heterochromatin spreading. (A and B)Relative enrichments of Sir2-Myc (A, left panel) and Sir3-Myc (B, left panel) for wild-type (WT) and mutant cells. Average mutant/WT ratios areshown plotted on logarithmic scales (right panels). Values greater than 1 indicate more enrichment at the indicated locus in the mutant than withthe wild type. (C) Immunolocalization of Sir2-Myc and Nop1 in wild-type and mutant cells. Sir2-Myc and Nop1 were stained by rabbit anti-Mycantibody and mouse anti-Nop1 antibody, respectively. (D) Real-time PCR of relative mRNA levels in a cluster of genes adjacent to heterochro-matin. Shown are the average expression ratios, which are normalized to ACT1 levels and are relative to the wild-type level for the indicatedgenotypes. A log2 ratio of less than zero indicates repression of transcription, whereas a greater-than-zero ratio indicates enhancement oftranscription. Error bars represent standard errors of the means for three independent RNA purifications.



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

mosome end, was slightly reduced in swr1� cells and eaf1�cells but was strongly repressed in swc4� cells (Fig. 5D), sug-gesting that Swc4 has the greatest role in boundary activity ofthe three proteins we tested. Double mutations, eaf1� sir2�,swr1� sir2�, and swc4� sir2�, largely relieved the repressiveeffects on the transcription profile of the genes tested (Fig.5D), suggesting that the transcriptional repression in the mu-tant strains was Sir dependent. Since SWC4 deletion repre-sented a synthetic defect of both NuA4 and SWR1-C, weconcluded that NuA4 and SWR1-C synergistically work toantagonize Sir complex-mediated transcriptional silencing.

DISCUSSION

Chromatin-modifying complexes are usually targeted to spe-cific regions in a dynamic manner to transiently modulate chro-mosomal processes like transcription and DNA damage repair

(5, 41). In this study, we established that SWR1-C works atregions near heterochromatin to propagate an antisilencingsignal that involves histone H4 acetylation and the promotionof H2A.Z incorporation. Genetic and functional assays exam-ining the role of Swc4 in antisilencing have led us to proposethat it is the synergistic actions of NuA4 and SWR1-C that leadto the efficient incorporation of H2A.Z at regions near hetero-chromatin and thereby contribute substantively to the levels ofheterochromatin-euchromatin boundary activity (Fig. 6).

SWC4 is a nonessential gene in the BY4743 background. Inthis study, we unexpectedly identified SWC4 as a nonessentialgene in the BY4743 background (Fig. 3A to D). The severeslow-growth phenotype of swc4� cells probably led others toconclude that SWC4 was an essential gene (2, 4, 23, 29). Ourgenetic and biochemical data have revealed that Swc4 is acoregulator of both NuA4 and SWR1-C (Fig. 4). Since Swc4 isa member of both NuA4 and SWR1-C (2, 17, 19, 30, 43), the

FIG. 6. Model for the actions of NuA4 and SWR1-C at heterochromatin boundaries. Step 1: SWR1-C associates with chromatin through theshared module (Arp4, Act1, Yaf9, and Swc4). Without histone acetylation, Sir proteins tend to spread along the chromatin fiber. Step 2: NuA4is recruited onto and acetylates heterochromatin-proximal regions. SWR1-C recognizes the acetylated histone with the help of the Bdf1 subunit.Step 3: SWR1-C deposits histone variant H2A.Z into acetylated chromatin. NuA4 subsequently acetylates the newly deposited histone variantH2A.Z. NuA4-mediated histone H4 and H2A.Z acetylation efficiently antagonizes Sir proteins-dependent heterochromatin spreading.



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

cellular dysfunctions in swc4� cells are likely a result of asynthetic defect of both NuA4 and SWR1-C. Our findingsindicate that subunits shared between NuA4 and SWR1-C arenot coincident but instead underlie the cooperation betweenNuA4 and SWR1-C. It remains unclear how swc4� cells sur-vive with the synthetic defect of inactivation of both NuA4 andSWR1-C. Since all swc4� spores were viable (Fig. 3A to D), itseems unlikely that a suppressor mutation has occurred inevery isolate of swc4� cells. One possibility is that the Swc4regulates the chromosomal functions of NuA4 and SWR1-C atlimited regions. It will be intriguing to explore the potentialroles of Swc4 in mediating these complexes’ collaboration inother chromatin domains.

Shared subunits at the center of circuitry between SWR1-Cand NuA4. One attractive mechanism for the synergism be-tween SWR1-C and NuA4 involves the differing association ofunique subunits of the two complexes with the four sharedsubunits Act1, Arp4, Swc4, and Yaf9. Yaf9 contributes to thefunctions of both NuA4 and SWR1-C (45). Since there isphysical interaction between Swc4 and Yaf9 (7, 44), it is likelythat Swc4 links Yaf9 to NuA4 and SWR1-C and thereforemediates the function of Yaf9 in these two complexes. Theactin protein Act1 and the actin-related protein Arp4 are twoessential skeletal components that link the shared module withthe main respective components of the complexes (40, 44).Interestingly, Arp4 is capable of direct interaction with bothhistones and nucleosomes (16). We propose that these fourshared subunits cooperatively facilitate the general associationof SWR1-C and NuA4 with chromatin.

The recruitment of SWR1-C to chromatin also involvesbinding of the Bdf1 subunit to acetylated histones through itsbromodomain (21, 46). Likewise, the association of full NuA4with promoters also involves the recognition of methylatedhistone H3 lysine 4 by the Yng2 subunit through its PHDdomain (32). We propose that the shared subunits act as adocking platform for chromatin while other subunits of thecomplexes provide for the binding specificity.

Chromatin-modifying complexes synergize boundary activ-ity. The synergistic relationship between NuA4-dependent hi-stone H4 acetylation and SWR1-C-dependent incorporation ofH2A.Z has been previously described (13, 24, 33, 46). In thisstudy, we have built on these pioneering studies and demon-strated that the coupled action of NuA4 and SWR1-C effi-ciently antagonizes Sir protein-dependent heterochromatinspreading. Taking advantage of Swc4, the versatile coregulatorof NuA4 and SWR1-C, we found that heterochromatin be-came much more pervasive when both NuA4 and SWR1-Cwere deregulated (Fig. 5). Using the LacI-lacO system, weshowed that corecruitment of NuA4 and SWR1-C conferredhigher antisilencing activity than either module anchoring didalone (Fig. 2D and 4H and I).

Synergistic antisilencing actions of H2A.Z incorporationand other chromatin modifications, such as SET1-mediatedhistone H3K4 trimethylation and SAS-mediated histoneH4K16 acetylation, have also been reported (37, 42). SinceH3K4 trimethylation is proposed to regulate NuA4 recruit-ment (32), it will be interesting to know whether H3K4 tri-methylation contributes to boundary function through recruit-ment of the NuA4 complex.

SWR1-C prepares boundary regions to propagate antisi-lencing activities. The genome-wide chromatin associationpattern of H2A.Z has been mapped (15, 22, 33, 46), but theblueprint for SWR1-C has not been well drawn. One specula-tion is that the binding profile of SWR1-C overlaps that ofH2A.Z. Our results demonstrated that Swr1 covered a widerrange of chromosomal domains than H2A.Z did. Swr1 washighly enriched at some heterochromatic regions, but H2A.Zwas not (Fig. 1D, E, and G). However, targeted Swr1 did notboost H2A.Z levels in hypoacetylated regions (Fig. 2C). Wepropose that recruitment of SWR1-C provides potential but byitself is not sufficient for specific chromatin regions to incor-porate H2A.Z.

H2A.Z incorporation appeared to take place at histoneH4K8-hyperacetylated regions (Fig. 1E, F, and 1G). Theseresults are in agreement with an earlier finding that the ge-nome-wide occupancy of H2A.Z positively correlated withAcH4K8 and AcH4K12 (46). Esa1 tethering enhanced histoneH4 acetylation and induced chromatin-bound SWR1-C to ex-change H2A-H2B with H2A.Z-H2B. The incorporated H2A.Zthen efficiently counteracted the encroachment of heterochro-matin silencing (27). Since histone H4 acetylation per se was anatural antagonist of silencing, H2A.Z incorporation can fur-ther amplify its antisilencing activity (Fig. 6).

It remains unclear how NuA4-mediated histone acetyla-tion induces SWR1-C-mediated H2A.Z deposition. Onepossibility is that histone acetylation facilitates the chroma-tin-remodeling process by neutralizing the positive chargeson the histone surface and thereby leading to a looser config-uration of the chromatin fiber. Interestingly, histone H4K16,catalyzed by the SAS complex, is also required for H2A.Zincorporation (37). It is possible that NuA4 and SAS cooper-atively define the nucleosomal surface used for SWR1-C-me-diated H2A.Z deposition.

ACKNOWLEDGMENTS

We thank Brian A. Lenzmeier (Buena Vista University, Storm Lake,IA) for his critical reading of our manuscript. We thank Susan M.Gasser, who generously provided the plasmids for the lacO targetingsystem.

This project was supported by grants from the Ministry of Scienceand Technology of China (2005CB522402) and the National NaturalScience Foundation of China (90919027).

We declare that no competing interests exist.

REFERENCES

1. Allard, S., R. T. Utley, J. Savard, A. Clarke, P. Grant, C. J. Brandl, L. Pillus,J. L. Workman, and J. Cote. 1999. NuA4, an essential transcription adaptor/histone H4 acetyltransferase complex containing Esa1p and the ATM-re-lated cofactor Tra1p. EMBO J. 18:5108–5119.

2. Auger, A., L. Galarneau, M. Altaf, A. Nourani, Y. Doyon, R. T. Utley, D.Cronier, S. Allard, and J. Cote. 2008. Eaf1 is the platform for NuA4 molec-ular assembly that evolutionarily links chromatin acetylation to ATP-depen-dent exchange of histone H2A variants. Mol. Cell. Biol. 28:2257–2270.

3. Babiarz, J. E., J. E. Halley, and J. Rine. 2006. Telomeric heterochromatinboundaries require NuA4-dependent acetylation of histone variant H2A.Zin Saccharomyces cerevisiae. Genes Dev. 20:700–710.

4. Ben-Aroya, S., C. Coombes, T. Kwok, K. A. O’Donnell, J. D. Boeke, and P.Hieter. 2008. Toward a comprehensive temperature-sensitive mutant repos-itory of the essential genes of Saccharomyces cerevisiae. Mol. Cell 30:248–258.

5. Berger, S. L. 2007. The complex language of chromatin regulation duringtranscription. Nature 447:407–412.

6. Bi, X., and J. R. Broach. 2001. Chromosomal boundaries in S. cerevisiae.Curr. Opin. Genet. Dev. 11:199–204.

7. Bittner, C. B., D. T. Zeisig, B. B. Zeisig, and R. K. Slany. 2004. Direct



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

physical and functional interaction of the NuA4 complex components Yaf9pand Swc4p. Eukaryot. Cell 3:976–983.

8. Boudreault, A. A., D. Cronier, W. Selleck, N. Lacoste, R. T. Utley, S. Allard,J. Savard, W. S. Lane, S. Tan, and J. Cote. 2003. Yeast enhancer of poly-comb defines global Esa1-dependent acetylation of chromatin. Genes Dev.17:1415–1428.

9. Boyer, L. A., M. R. Langer, K. A. Crowley, S. Tan, J. M. Denu, and C. L.Peterson. 2002. Essential role for the SANT domain in the functioning ofmultiple chromatin remodeling enzymes. Mol. Cell 10:935–942.

10. Boyer, L. A., R. R. Latek, and C. L. Peterson. 2004. The SANT domain: aunique histone-tail-binding module? Nat. Rev. Mol. Cell Biol. 5:158–163.

11. Chiu, Y. H., Q. Yu, J. J. Sandmeier, and X. Bi. 2003. A targeted histoneacetyltransferase can create a sizable region of hyperacetylated chromatinand counteract the propagation of transcriptionally silent chromatin. Genet-ics 165:115–125.

12. Clarke, A. S., J. E. Lowell, S. J. Jacobson, and L. Pillus. 1999. Esa1p is anessential histone acetyltransferase required for cell cycle progression. Mol.Cell. Biol. 19:2515–2526.

13. Durant, M., and B. F. Pugh. 2007. NuA4-directed chromatin transactionsthroughout the Saccharomyces cerevisiae genome. Mol. Cell. Biol. 27:5327–5335.

14. Gotta, M., S. Strahl-Bolsinger, H. Renauld, T. Laroche, B. K. Kennedy, M.Grunstein, and S. M. Gasser. 1997. Localization of Sir2p: the nucleolus as acompartment for silent information regulators. EMBO J. 16:3243–3255.

15. Guillemette, B., A. R. Bataille, N. Gevry, M. Adam, M. Blanchette, F. Robert,and L. Gaudreau. 2005. Variant histone H2A.Z is globally localized to thepromoters of inactive yeast genes and regulates nucleosome positioning.PLoS Biol. 3:e384.

16. Harata, M., Y. Oma, S. Mizuno, Y. W. Jiang, D. J. Stillman, and U. Win-tersberger. 1999. The nuclear actin-related protein of Saccharomyces cer-evisiae, Act3p/Arp4, interacts with core histones. Mol. Biol. Cell 10:2595–2605.

17. Kobor, M. S., S. Venkatasubrahmanyam, M. D. Meneghini, J. W. Gin, J. L.Jennings, A. J. Link, H. D. Madhani, and J. Rine. 2004. A protein complexcontaining the conserved Swi2/Snf2-related ATPase Swr1p deposits histonevariant H2A.Z into euchromatin. PLoS Biol. 2:E131.

18. Kouzarides, T. 2007. Chromatin modifications and their function. Cell 128:693–705.

19. Krogan, N. J., K. Baetz, M. C. Keogh, N. Datta, C. Sawa, T. C. Kwok, N. J.Thompson, M. G. Davey, J. Pootoolal, T. R. Hughes, A. Emili, S. Buratowski,P. Hieter, and J. F. Greenblatt. 2004. Regulation of chromosome stability bythe histone H2A variant Htz1, the Swr1 chromatin remodeling complex, andthe histone acetyltransferase NuA4. Proc. Natl. Acad. Sci. U. S. A. 101:13513–13518.

20. Kurdistani, S. K., and M. Grunstein. 2003. In vivo protein-protein andprotein-DNA crosslinking for genomewide binding microarray. Methods31:90–95.

21. Ladurner, A. G., C. Inouye, R. Jain, and R. Tjian. 2003. Bromodomainsmediate an acetyl-histone encoded antisilencing function at heterochromatinboundaries. Mol. Cell 11:365–376.

22. Li, B., S. G. Pattenden, D. Lee, J. Gutierrez, J. Chen, C. Seidel, J. Gerton,and J. L. Workman. 2005. Preferential occupancy of histone variant H2AZat inactive promoters influences local histone modifications and chromatinremodeling. Proc. Natl. Acad. Sci. U. S. A. 102:18385–18390.

23. Lin, Y. Y., Y. Qi, J. Y. Lu, X. Pan, D. S. Yuan, Y. Zhao, J. S. Bader, and J. D.Boeke. 2008. A comprehensive synthetic genetic interaction network govern-ing yeast histone acetylation and deacetylation. Genes Dev. 22:2062–2074.

24. Lu, P. Y., N. Levesque, and M. S. Kobor. 2009. NuA4 and SWR1-C: twochromatin-modifying complexes with overlapping functions and compo-nents. Biochem. Cell Biol. 87:799–815.

25. Matangkasombut, O., and S. Buratowski. 2003. Different sensitivities ofbromodomain factors 1 and 2 to histone H4 acetylation. Mol. Cell 11:353–363.

26. Meijsing, S. H., and A. E. Ehrenhofer-Murray. 2001. The silencing complexSAS-I links histone acetylation to the assembly of repressed chromatin byCAF-I and Asf1 in Saccharomyces cerevisiae. Genes Dev. 15:3169–3182.

27. Meneghini, M. D., M. Wu, and H. D. Madhani. 2003. Conserved histone

variant H2A.Z protects euchromatin from the ectopic spread of silent het-erochromatin. Cell 112:725–736.

28. Meng, F. L., Y. Hu, N. Shen, X. J. Tong, J. Wang, J. Ding, and J. Q. Zhou.2009. Sua5p a single-stranded telomeric DNA-binding protein facilitatestelomere replication. EMBO J. 28:1466–1478.

29. Micialkiewicz, A., and A. Chelstowska. 2008. The essential function ofSwc4p—a protein shared by two chromatin-modifying complexes of the yeastSaccharomyces cerevisiae—resides within its N-terminal part. Acta Biochim.Pol. 55:603–612.

30. Mitchell, L., J. P. Lambert, M. Gerdes, A. S. Al-Madhoun, I. S. Skerjanc, D.Figeys, and K. Baetz. 2008. Functional dissection of the NuA4 histoneacetyltransferase reveals its role as a genetic hub and that Eaf1 is essentialfor complex integrity. Mol. Cell. Biol. 28:2244–2256.

31. Mizuguchi, G., X. Shen, J. Landry, W. H. Wu, S. Sen, and C. Wu. 2004.ATP-driven exchange of histone H2AZ variant catalyzed by SWR1 chroma-tin remodeling complex. Science 303:343–348.

32. Morillon, A., N. Karabetsou, A. Nair, and J. Mellor. 2005. Dynamic lysinemethylation on histone H3 defines the regulatory phase of gene transcrip-tion. Mol. Cell 18:723–734.

33. Raisner, R. M., P. D. Hartley, M. D. Meneghini, M. Z. Bao, C. L. Liu, S. L.Schreiber, O. J. Rando, and H. D. Madhani. 2005. Histone variant H2A.Zmarks the 5� ends of both active and inactive genes in euchromatin. Cell123:233–248.

34. Rohner, S., S. M. Gasser, and P. Meister. 2008. Modules for cloning-freechromatin tagging in Saccharomyces cerevisiae. Yeast 25:235–239.

35. Rusche, L. N., A. L. Kirchmaier, and J. Rine. 2003. The establishment,inheritance, and function of silenced chromatin in Saccharomyces cerevisiae.Annu. Rev. Biochem. 72:481–516.

36. Schultz, J. 1936. Variegation in Drosophila and the inert chromosome re-gions. Proc. Natl. Acad. Sci. U. S. A. 22:27–33.

37. Shia, W. J., B. Li, and J. L. Workman. 2006. SAS-mediated acetylation ofhistone H4 Lys 16 is required for H2A.Z incorporation at subtelomericregions in Saccharomyces cerevisiae. Genes Dev. 20:2507–2512.

38. Sif, S. 2004. ATP-dependent nucleosome remodeling complexes: enzymestailored to deal with chromatin. J. Cell Biochem. 91:1087–1098.

39. Smith, E. R., A. Eisen, W. Gu, M. Sattah, A. Pannuti, J. Zhou, R. G. Cook,J. C. Lucchesi, and C. D. Allis. 1998. ESA1 is a histone acetyltransferase thatis essential for growth in yeast. Proc. Natl. Acad. Sci. U. S. A. 95:3561–3565.

40. Szerlong, H., K. Hinata, R. Viswanathan, H. Erdjument-Bromage, P.Tempst, and B. R. Cairns. 2008. The HSA domain binds nuclear actin-related proteins to regulate chromatin-remodeling ATPases. Nat. Struct.Mol. Biol. 15:469–476.

41. van Attikum, H., and S. M. Gasser. 2005. The histone code at DNA breaks:a guide to repair? Nat. Rev. Mol. Cell Biol. 6:757–765.

42. Venkatasubrahmanyam, S., W. W. Hwang, M. D. Meneghini, A. H. Tong,and H. D. Madhani. 2007. Genome-wide, as opposed to local, antisilencingis mediated redundantly by the euchromatic factors Set1 and H2A.Z. Proc.Natl. Acad. Sci. U. S. A. 104:16609–16614.

43. Wu, W. H., S. Alami, E. Luk, C. H. Wu, S. Sen, G. Mizuguchi, D. Wei, andC. Wu. 2005. Swc2 is a widely conserved H2AZ-binding module essential forATP-dependent histone exchange. Nat. Struct. Mol. Biol. 12:1064–1071.

44. Wu, W. H., C. H. Wu, A. Ladurner, G. Mizuguchi, D. Wei, H. Xiao, E. Luk,A. Ranjan, and C. Wu. 2009. N terminus of Swr1 binds to histone H2AZ andprovides a platform for subunit assembly in the chromatin remodeling com-plex. J. Biol. Chem. 284:6200–6207.

45. Zhang, H., D. O. Richardson, D. N. Roberts, R. Utley, H. Erdjument-Bromage, P. Tempst, J. Cote, and B. R. Cairns. 2004. The Yaf9 componentof the SWR1 and NuA4 complexes is required for proper gene expression,histone H4 acetylation, and Htz1 replacement near telomeres. Mol. Cell.Biol. 24:9424–9436.

46. Zhang, H., D. N. Roberts, and B. R. Cairns. 2005. Genome-wide dynamics ofHtz1, a histone H2A variant that poises repressed/basal promoters for acti-vation through histone loss. Cell 123:219–231.

47. Zhou, J., B. O. Zhou, B. A. Lenzmeier, and J. Q. Zhou. 2009. Histonedeacetylase Rpd3 antagonizes Sir2-dependent silent chromatin propagation.Nucleic Acids Res. 37:3699–3713.



http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

swr1 complex poises heterochromatin boundaries for

Documents