Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry,...

14
Copyright Ó 2009 by the Genetics Society of America DOI: 10.1534/genetics.109.105338 Domains of Heterochromatin Protein 1 Required for Drosophila melanogaster Heterochromatin Spreading Karrie A. Hines,* Diane E. Cryderman,* Kaitlin M. Flannery,* Hongbo Yang, Michael W. Vitalini,* Tulle Hazelrigg, Craig A. Mizzen and Lori L. Wallrath* ,1 *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, Department of Cell and Developmental Biology, University of Illinois, Urbana, Illinois 61801 and Department of Biological Sciences, Columbia University, New York, New York 10027 Manuscript received May 21, 2009 Accepted for publication June 1, 2009 ABSTRACT Centric regions of eukaryotic genomes are packaged into heterochromatin, which possesses the ability to spread along the chromosome and silence gene expression. The process of spreading has been challenging to study at the molecular level due to repetitious sequences within centric regions. A heterochromatin protein 1 (HP1) tethering system was developed that generates ‘‘ectopic heterochromatin’’ at sites within euchromatic regions of the Drosophila melanogaster genome. Using this system, we show that HP1 dimerization and the PxVxL interaction platform formed by dimerization of the HP1 chromo shadow domain are necessary for spreading to a downstream reporter gene located 3.7 kb away. Surprisingly, either the HP1 chromo domain or the chromo shadow domain alone is sufficient for spreading and silencing at a downstream reporter gene located 1.9 kb away. Spreading is dependent on at least two H3K9 methyltransferases, with SU(VAR)3-9 playing a greater role at the 3.7-kb reporter and dSETDB1 predominately acting at the 1.9 kb reporter. These data support a model whereby HP1 takes part in multiple mechanisms of silencing and spreading. H ETEROCHROMATIN protein 1 (HP1) was iden- tified in Drosophila as a nonhistone chromosomal protein enriched in centric heterochromatin ( James and Elgin 1986; James et al. 1989). On polytene chromosomes, HP1 localizes near centromeres and telomeres, along the fourth chromosome and at 200 sites within the euchromatic arms ( James et al. 1989; Fanti et al. 2003). Heterochromatin has the ability to ‘‘spread,’’ or propagate in cis, along the chromosome (Weiler and Wakimoto 1995). Spreading is observed when a chromosomal rearrangement places a euchro- matic domain next to a heterochromatic domain. Cytologically, spreading is visualized as densely compact chromatin that emanates from the chromocenter, the structure formed by the fusion of centromeres, and extends into the banded regions of polytene chromo- somes (Belyaeva and Zhimulev 1991). Euchromatic genes brought into juxtaposition with heterochromatin by chromosomal rearrangements exhibit gene silencing, termed position effect variegation (PEV) (Weiler and Wakimoto 1995). Mutations in Su(var)2-5, the gene encoding HP1, suppress silencing, suggesting HP1 plays a key role in spreading (Eissenberg et al. 1990). The molecular processes of spreading are not well un- derstood. Repetitive sequences within heterochromatin make it difficult to study spreading at the molecular level. In addition, specific repetitive elements are thought to function as initiation sites for heterochromatin forma- tion (Sun et al. 2004; Haynes et al. 2006), making it challenging to separate initiation from spreading. To overcome these problems, we generated a system that nucleates small domains (,20 kb) of repressive chromatin that share many properties with centric heterochroma- tin. Here we refer to these as ectopic heterochromatin domains. These domains are generated by expressing a fusion protein, consisting of the DNA binding domain of the Escherichia coli lac repressor (LacI) fused to HP1, in stocks possessing lac operator (lacO) repeats upstream of a reporter gene cassette (Danzer and Wallrath 2004). LacI-HP1 associates with the lacO repeats and causes silencing of the adjacent reporter genes. Silenc- ing correlates with alterations in chromatin structure that include the generation of regular nucleosome arrays similar to those observed in centric heterochro- matin (Sun et al. 2001; Danzer and Wallrath 2004). Chromatin immunoprecipitation (ChIP) experiments demonstrated that HP1 spreads bidirectionally, 5–10 kb from the lacO repeats, encompassing the reporter genes (Danzer and Wallrath 2004). Thus, HP1 is sufficient to nucleate small heterochromatin-like domains at geno- mic locations devoid of repetitious sequences, allowing for molecular studies of spreading. HP1 contains an amino terminal chromo domain (CD) and a carboxy chromo shadow domain (CSD), Supporting information is available online at: http://www.genetics. org/cgi/content/full/genetics.109.105338/DC1. 1 Corresponding author: Department of Biochemistry, 3136 MERF, University of Iowa, Iowa City, IA 52242. E-mail: [email protected] Genetics 182: 967–977 (August 2009)

Transcript of Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry,...

Page 1: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

Copyright � 2009 by the Genetics Society of AmericaDOI: 10.1534/genetics.109.105338

Domains of Heterochromatin Protein 1 Required for Drosophilamelanogaster Heterochromatin Spreading

Karrie A. Hines,* Diane E. Cryderman,* Kaitlin M. Flannery,* Hongbo Yang,†

Michael W. Vitalini,* Tulle Hazelrigg,‡ Craig A. Mizzen† and Lori L. Wallrath*,1

*Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology, University of Illinois,Urbana, Illinois 61801 and ‡Department of Biological Sciences, Columbia University, New York, New York 10027

Manuscript received May 21, 2009Accepted for publication June 1, 2009

ABSTRACT

Centric regions of eukaryotic genomes are packaged into heterochromatin, which possesses the ability tospread along the chromosome and silence gene expression. The process of spreading has been challenging tostudy at the molecular level due to repetitious sequences within centric regions. A heterochromatin protein 1(HP1) tethering system was developed that generates ‘‘ectopic heterochromatin’’ at sites within euchromaticregions of the Drosophila melanogaster genome. Using this system, we show that HP1 dimerization and thePxVxL interaction platform formed by dimerization of the HP1 chromo shadow domain are necessary forspreading to a downstream reporter gene located 3.7 kb away. Surprisingly, either the HP1 chromo domain orthe chromo shadow domain alone is sufficient for spreading and silencing at a downstream reporter genelocated 1.9 kb away. Spreading is dependent on at least two H3K9 methyltransferases, with SU(VAR)3-9playing a greater role at the 3.7-kb reporter and dSETDB1 predominately acting at the 1.9 kb reporter. Thesedata support a model whereby HP1 takes part in multiple mechanisms of silencing and spreading.

HETEROCHROMATIN protein 1 (HP1) was iden-tified in Drosophila as anonhistone chromosomal

protein enriched in centric heterochromatin ( James

and Elgin 1986; James et al. 1989). On polytenechromosomes, HP1 localizes near centromeres andtelomeres, along the fourth chromosome and at �200sites within the euchromatic arms ( James et al. 1989;Fanti et al. 2003). Heterochromatin has the ability to‘‘spread,’’ or propagate in cis, along the chromosome(Weiler and Wakimoto 1995). Spreading is observedwhen a chromosomal rearrangement places a euchro-matic domain next to a heterochromatic domain.Cytologically, spreading is visualized as densely compactchromatin that emanates from the chromocenter, thestructure formed by the fusion of centromeres, andextends into the banded regions of polytene chromo-somes (Belyaeva and Zhimulev 1991). Euchromaticgenes brought into juxtaposition with heterochromatinby chromosomal rearrangements exhibit gene silencing,termed position effect variegation (PEV) (Weiler andWakimoto 1995). Mutations in Su(var)2-5, the geneencoding HP1, suppress silencing, suggesting HP1 playsa key role in spreading (Eissenberg et al. 1990). Themolecular processes of spreading are not well un-derstood.

Repetitive sequences within heterochromatin make itdifficult to study spreading at the molecular level. Inaddition, specific repetitive elements are thought tofunction as initiation sites for heterochromatin forma-tion (Sun et al. 2004; Haynes et al. 2006), making itchallenging to separate initiation from spreading. Toovercome these problems, we generated a system thatnucleates small domains (,20 kb) of repressive chromatinthat share many properties with centric heterochroma-tin. Here we refer to these as ectopic heterochromatindomains. These domains are generated by expressing afusion protein, consisting of the DNA binding domainof the Escherichia coli lac repressor (LacI) fused to HP1,in stocks possessing lac operator (lacO) repeats upstreamof a reporter gene cassette (Danzer and Wallrath

2004). LacI-HP1 associates with the lacO repeats andcauses silencing of the adjacent reporter genes. Silenc-ing correlates with alterations in chromatin structurethat include the generation of regular nucleosomearrays similar to those observed in centric heterochro-matin (Sun et al. 2001; Danzer and Wallrath 2004).Chromatin immunoprecipitation (ChIP) experimentsdemonstrated that HP1 spreads bidirectionally, 5–10 kbfrom the lacO repeats, encompassing the reporter genes(Danzer and Wallrath 2004). Thus, HP1 is sufficientto nucleate small heterochromatin-like domains at geno-mic locations devoid of repetitious sequences, allowingfor molecular studies of spreading.

HP1 contains an amino terminal chromo domain(CD) and a carboxy chromo shadow domain (CSD),

Supporting information is available online at: http://www.genetics.org/cgi/content/full/genetics.109.105338/DC1.

1Corresponding author: Department of Biochemistry, 3136 MERF,University of Iowa, Iowa City, IA 52242. E-mail: [email protected]

Genetics 182: 967–977 (August 2009)

Page 2: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

separated by a flexible hinge (Li et al. 2002). The CDforms a hydrophobic pocket implicated in chromosomalassociation through binding to di- and trimethylatedlysine 9 of histone H3 (H3K9me2 and me3, respectively),an epigenetic mark generated by the histone methyl-transferases (HMT) SU(VAR)3-9 and dSETDB1 (alsoknown as Egg) ( Jacobs et al. 2001; Schotta et al. 2002;Schultz et al. 2002; Ebert et al. 2004; Clough et al. 2007;Seum et al. 2007; Tzeng et al. 2007). Association withmethylated H3 is one mechanism of HP1 chromosomeassociation; however, other mechanisms involving inter-actions with DNA and/or partner proteins likely exist(Fanti et al. 1998; Li et al. 2002; Cryderman et al. 2005).In Drosophila HP1, a single amino acid substitutionwithin the CD (V26M) is present in the Su(var)2-502

allele; flies heterozygous for this allele show suppressionof gene silencing by heterochromatin (Eissenberg et al.1990). Furthermore, flies trans-heterozygous forSu(var)2-502 and a null allele of Su(var)2-5 show dra-matic reduction of HP1 near centromeres and do notsurvive past the third larval stage (Fanti et al. 1998).Consistent with these observations, structural studiesshow that V26 plays a critical role in forming thehydrophobic pocket of the CD that binds to H3K9me( Jacobs et al. 2001).

The HP1 CSD dimerizes and mediates interactionswith a variety of nuclear proteins (Cowieson et al. 2000;Yamamoto and Sonoda 2003; Thiru et al. 2004). CSDdimerization sets up an interaction platform for thebinding of proteins possessing a penta-peptide motif,PxVxL (where x represents any amino acid) (Thiru et al.2004; Lechner et al. 2005). Amino acid substitutionswithin HP1 have been identified that disrupt dimeriza-tion, and interaction with PxVxL proteins (Lechner

et al. 2000; Thiru et al. 2004). For example, a singleamino acid substitution within the CSD (I161E) disruptsdimerization of mouse HP1beta (Brasher et al. 2000).The lack of dimerization also caused the loss of inter-actions with nuclear factors containing PxVxL motifsand non-PxVxL partners (Yamamoto and Sonoda 2003;Lechner et al. 2005). In contrast, a single amino acidsubstitution elsewhere in the CSD (W170A) of mouseHP1beta does not prevent dimerization, but disrupts theinteraction with PxVxL partner proteins (Brasher et al.2000). Therefore, the requirement for HP1 dimeriza-tion and binding to the PxVxL proteins can be func-tionally separated. Here, we investigate effects of HP1domain deletions and amino acid substitutions on HP1localization, partner protein interactions, and hetero-chromatin spreading.

MATERIALS AND METHODS

Drosophila stocks: Drosophila stocks were raised on stan-dard corn meal sucrose media at 25� unless otherwise noted.Stocks encoding wild-type (wt) LacI–HP1, GFP–LacI, and thereporter transgenes containing lacO repeats were previously

generated (Danzer and Wallrath 2004). Transgenic stocksexpressing the mutant forms of LacI–HP1 were made usingstandard P-element transformation techniques (Rubin andSpradling 1982). The dSETDB1 mutants egg1473-8 anddmSetDB110.1a were described previously (Clough et al. 2007;Seum et al. 2007).

Plasmid construction: Constructs containing mutant formsof HP1 were generated by PCR using a full-length HP1 cDNAclone as a template (Eissenberg et al. 1990). For the constructpossessing only the CD, a 240-bp fragment encoding aminoacids 1–81 of the HP1 CD was amplified and cloned down-stream of sequences encoding the DNA binding domain of theE. coli LacI repressor (Robinett et al. 1996). The sequence ofthe forward primer was 59-CGGATCCGGAATGGGCAAGAAAATCGAC-39 and the reverse primer 59-TGATCTAGATCAATCCTTCTTGGA-39. For the construct containing only the CSD, a230-bp fragment encoding amino acids 132–208 was cloneddownstream of sequences encoding the LacI DNA bindingdomain. The sequence of the forward primer was 59-TCCGGATCCGGAATGGAGCAGGACACCATT-39 and the reverseprimer 59-TGATCTAGATCAATCTTCATTATC-39. The V26M,I191E, and W200A amino acid substitutions were generatedusing the Quick Change Site Directed Mutagenesis kit(Stratagene). The sequences of the PCR primers used formutagenesis were as follows: V26M forward, 59-GAGGAGTACGCCATGGAAAAGATCATCG-39; V26M reverse, 59-CTCCTCATGCGGTACCTTTTCTAGTAGC-39; I191E forward, 59-CCCACGAATGGTAGAACACTTCTACGAAGAGCGCC -39; I191Ereverse, 59-GGGTGCTTACCATCTTGTGAAGATGCTTCTCGCGG-39; W200A forward, 59-CCACTTCTACGAAGAGCGCCTATCCGCATACTCTGATAATGAAG- 39; and W200A reverse,59-GGTGAAGATGCTTCTCGCGGATAGGCGTATGAGACTATTACTTC-39. The primers contain nucleotide substitutions thatresult in single-amino-acid substitutions within HP1. Amplifiedfragments were fused downstream of sequences encoding theLacI DNA binding domain. The LacI-HP1 fusions were clonedinto the pCaSpeR-hs-act transformation vector downstream ofthe inducible hsp70 heat-shock promoter (http://www.ncbi.nlm.nih.gov/nuccore/1432080report=genbank) and used togenerate transgenic stocks.

Polytene chromosome staining: Third instar larval salivaryglands were fixed, squashed, and stained with mouse anti-bodies to LacI (Upstate Biotechnology no. 05-501, clone 9A5,1:300 dilution), followed by goat anti-mouse FITC ( JacksonImmunoresearch Laboratories, 1:200) or Alexa Fluor 546donkey anti-mouse (Invitrogen, 1:200) after a 1-hr heat shocktreatment according to published procedures (Li et al. 2003).Hip was detected using previously described rabbit antibodiesto Hip (Schwendemann et al. 2008; 1:50) followed by AlexaFluor 488 goat anti-rabbit (Invitrogen, 1:300). dSETDB1detection was carried out using previously described rat anti-bodies (Clough et al. 2007; 1:150) followed by donkey anti-ratFITC ( Jackson Immunoresearch Laboratories, 1:200). Suv4-20rabbit antibodies were generated against the first 100 residuesof Drosophila Suv4-20 expressed as a 63 His-tag fusion in E.coli. dsRNA corresponding to nucleotides 1088–1597 ofCG13363 was used for Suv4-20 knockdown in S2 cellsas described previously (Yang et al. 2008) and resulted in theelimination of the major band corresponding to the predictedmolecular weight of Suv4-20 by Western analysis (data notshown). Polytene chromosomes were stained with antibodiesto Suv4-20 (1:100) followed by Alexa Fluor 488 goat anti-rabbit(Invitrogen, 1:300).

Northern analysis: Total RNA was isolated from 20 larvae oradults using Trizol Reagent (Invitrogen) as described by themanufacturer after daily heat-shock treatment at 37� for45 min. Total RNA (25 mg) was used in Northern analysesand hybridized with radio-labeled fragments corresponding to

968 K. A. Hines et al.

Page 3: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

the unique tag (sequences from the SIP1 gene of barley) fusedto hsp26 or sequences corresponding to the white transgene.Hybridization with sequences corresponding to the ribosomalprotein gene, rp49, served as a loading control. Radioactivecounts from each hybridization signal were quantitated usingan Instant Imager (Packard). A single hsp26 transcript of 1.2-kbnucleotides was observed upon heat-shock treatment. Twotranscripts, 6.0 kb and 2.6 kb in size, were observed upon hsp70heat-shock-induced expression. The 6.0-kb transcript repre-sents unspliced message; at least 15 min of recovery from heatshock were required for full processing. The 2.6-kb transcriptrepresents the processed message. The rp49 transcript of�0.6 kb was observed.

Chromatin immunoprecipitation: For ChIP experimentsDrosophila stocks were raised in bottles at 18�. Larvae wereheat shocked once for 45 min at 37�. One hundred salivaryglands (50 pairs) per sample were dissected from third instarlarvae in Ringer’s solution (8.0 g NaCl, 0.2 g KCl, 1 g NaHCO3,0.04 g NaH2PO4�2H2O), 0.2 g CaCl2�2H2O, 0.05 g MgCl2�6H2O,and 1 g glucose in 1L H2O) within 1 hr following heat shock.Chromatin immunoprecipitation was performed as previouslydescribed (Danzer and Wallrath 2004; Cryderman et al.2005). Four microliters of polyclonal HP1 antibody, 4 ml of GFPpolyclonal antibody (negative control; A-6455, MolecularProbes), or 4 ml of polyclonal dimethyl K9 antibody (07-441,Upstate Biotechnology) was used for immunoprecipitation.Primers for PCR were as follows: hsp26-forward 59-CGAGGAAGAGCGTGTTGTAGG-39, hsp26-reverse 59-ACAACACCGACATGCTCTACAG-39 (region 1954 to 11082 relative to tran-scription start); hsp70-white-forward 59-GCAACCAAGTAAATCAACTGC-39, hsp70-white-reverse 59-GTTTTGGCACAGCACTTTGTG-39 (region 1149 to 1250); rp49-forward 59-ATCGGTTACGGATCGAACAAGC-39, rp49-reverse 59-GTAAACGCGGTTCTGCATGAGC-39, (region 12 to 1190); and cdc2-forward 59-GTAGCTAGCTTAGCATCGTT-39, cdc2-reverse 59-CCATATGTGCCCTCGCCAAT-39 (region �21 to 1143).

Real-time PCR: RNA was isolated using TRIzol reagent(Invitrogen) from adult flies treated with a daily 45-min heat-shock regimen (37�). Approximately 15 mg of total RNA wastreated with 2 mg of amplification grade DNase I (Invitrogen)in 20 ml total volume and incubated at room temperature for15 min. The reaction was stopped by addition of 2 ml of 25 mm

EDTA and heated to 70� for 10 min. cDNA was produced from6 ml of each DNase I reaction via the SuperScript First-StrandcDNA Synthesis kit (Invitrogen) as described in the kitprotocol with one exception: after RNaseH treatment, 20 mlof DEPC-treated water was added to each tube for a finalvolume of 40 ml. Real-time PCR was then performed usingSyber Green PCR master mix (Applied Biosystems). Primersused for RT–PCR were the same as those used in ChIPexperiments. PCR and fluorescent analyses were performedin an iCycler (Bio-Rad) running the iCycler IQ program(version 3.1.7050); threshold cycle numbers (Ct) were calcu-lated automatically. Normalized DCt values were determinedby subtracting the Ct obtained for the housekeeping ribosomalprotein gene rp49 from the Ct obtained for hsp26-tag and hsp70-white. The same PCR cycle parameters were used for allprimers: 50� for 2 min; 95� for 2 min; 35 cycles of 95� for 15sec, 60� for 1 min, and 72� for 1 min. Melting curve analysis wasrun following PCR to exclude the possibility of primer-dimeramplification products.

RESULTS

Wild-type and mutant HP1 fusion proteins showdistinct localization patterns: To determine the functionof the different domains of HP1, wild-type and mutant

forms of HP1 were expressed as LacI fusion proteinsunder control of a heat-shock inducible promoter(hsp26) in stocks containing a lacO-reporter gene cassette(Figure 1). The LacI–HP1 fusion protein rescues lethalityof Su(var)2-5 mutants (Li et al. 2003). Mutant forms in-clude HP1 truncations and amino acid substitutions thatdisrupt specific functions (Figure 1A). The LacI–CDconstruct encodes amino acids 1–81, representing theCD domain only. The LacI–CSD construct encodesamino acids 131–206, representing the CSD only. V26Mcontains an amino acid substitution within the CD thatdisrupts interaction with H3K9me ( Jacobs et al. 2001). Inmouse HP1beta, an amino acid substitution in the alpha-helical region disrupts HP1 dimerization (Brasher et al.2000); therefore, the analogous I191E amino acid sub-stitution was generated in Drosophila HP1 (LacI–I191E).This mutant form of HP1 failed to dimerize in vitro underconditions in which wild-type HP1 dimerized (Brower-Toland et al. 2007). In mouse HP1beta, a W170A aminoacid substitution disrupts interactions with PxVxL-containing partner proteins, yet retains the ability todimerize (Brasher et al. 2000). The analogous W200Aamino acid substitution in Drosophila HP1 was generated(LacI–W200A), which retained the ability to dimerizein vitro (Brower-Toland et al. 2007). Transgenic Dro-sophila stocks expressing each mutant form of LacI–HP1were generated. Western analyses of protein extracts fromheat-shocked larvae revealed that the LacI–HP1 proteinswere expressed 1.0- to 2.4-fold over endogenous HP1levels and persisted for at least 2 hr post-heat-shockinduction (data not shown). To determine the require-ments for heterochromatin formation and spreading, thewild-type and mutant forms of HP1 fusion proteins wereexpressed in a stock possessing lacO repeats upstream of atagged hsp26 gene (hsp26) and an hsp70-white gene (hsp70)(Figure 1B). The distance between the lacO repeats andthe hsp26 transcription start site is 1.9 kb, and the distancefrom the lacO repeats to the hsp70 transcription start is 3.7kb (Figure 1B).

Drosophila stocks expressing the LacI–HP1 fusionproteins were independently mated to stocks containingthe reporter gene cassette inserted at cytological loca-tions 4D5, 79D1, and 87C1 (Danzer and Wallrath

2004). The LacI–HP1 proteins were localized on salivarygland chromosomes of resulting progeny using anti-bodies specific to LacI (Figure 1C). LacI–HP1 showedthe anticipated localization pattern with association atthe lacO repeats as well as the chromocenter, telomeres,and several euchromatic regions, including region 31(Li et al. 2003; Danzer and Wallrath 2004). In con-trast, the mutant forms of LacI–HP1 showed abnormalpatterns of localization. LacI antibodies used to detectthe LacI–CD protein showed negligible staining at thechromocenter and other locations, except for at the lacOrepeats. Similarly, antibodies detecting the LacI–CSD,LacI–V26M, and LacI–W200A showed weak staining atthe chromocenter in addition to a strong signal at the

Drosophila Heterochromatin Spreading 969

Page 4: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

lacO repeats. LacI–W200A showed weak staining ateuchromatic region 31, in contrast to other mutantforms that showed barely detectable to no signal at thissite. Taken together, these data suggest that full-lengthHP1 is needed for proper association at the chromocenter.

Tethered HP1 recruits Hip and Suv4-20: HP1 hasnumerous interaction partners, some of which coloc-alize with HP1 at heterochromatin (Pak et al. 1997;Schotta et al. 2002; Schwendemann et al. 2008). Todetermine whether tethered HP1 is able to maintainthese interactions at sites of ectopic heterochromatin,colocalization studies on polytene chromosomes wereperformed. HP1-interacting protein (Hip), shows coloc-alization with HP1 in heterochromatic regions of poly-tene chromosomes and requires HP1 for chromosomeassociation (Schwendemann et al. 2008). Immunofluo-rescence microscopy of polytene chromosomes showedthat LacI–HP1 recruited Hip to sites of ectopic hetero-chromatin (Figure 2). Of the mutant forms of HP1, onlythe LacI–CSD displayed colocalization with Hip, dem-onstrating that in vivo interaction requires a functionalCSD (Figure 2). These results agree with in vitro data andsuggest that HP1 maintains partner–protein interac-tions at sites of ectopic heterochromatin.

Interaction between HP1 and a H4K20 HMT, Suv4-20,has been demonstrated by in vitro immunoprecipitationassays using mammalian orthologs (Schotta et al.

2004). This interaction is supported by genetic inter-actions in Drosophila (Yang et al. 2008) and suppressionof PEV in a Drosophila Suv4-20 mutant (Schotta et al.2004). However, a recent study reported that a nullallele of Suv4-20 does not suppress PEV (Sakaguchi

et al. 2008). To determine whether HP1 recruits Suv4-20to chromatin in vivo, immunofluorescence microscopyof polytene chromosomes was performed. Antibody toSuv4-20 colocalized to sites of ectopic heterochromatin(Figure 3), supporting the link between HP1 and methy-lation of H4K20. None of the mutant forms of HP1,including the LacI–CSD fusion protein, displayed co-localization with Suv4-20. These results suggest that full-length HP1 is necessary to recruit Suv4-20 in vivo.

Mutant forms of HP1 abolish silencing, spreading,and methylation at the hsp70 promoter: The effects ofexpressing mutant forms of LacI–HP1 on heat-shock-induced expression of the hsp70 reporter were deter-mined by Northern analyses using sequences specific forthe white transgene (Danzer and Wallrath 2004). Theresults of three independent Northerns are shown foreach insertion site (Figure 4). LacI–HP1 silenced thehsp70 promoter, while GFP–LacI had no effect on heat-shock-induced expression, similar to published findings(Figure 4A; Danzer and Wallrath 2004). None of themutant forms of HP1 tested showed silencing of thehsp70 reporter gene. These data imply that association

Figure 1.—The LacI–HP1/lacO tethering sys-tem and localization of LacI–HP1 fusion proteinsto polytenechromosomes. (A)TheLacI DNAbind-ing domain was fused to the amino terminus ofHP1. The chromo domain (CD), hinge, andchromo shadow domain (CSD) of HP1 are indi-cated. The amino acid numbers are designated be-low each construct. Locations of amino acidsubstitutions are marked by asterisks and the func-tion(s) affected by each mutation are given at right:D, dimerization; PxVxL, interactions with PxVxLpartner proteins; and H3K9me, binding of dime-thylated lysine 9 of histone H3. (B) The taggedhsp26 and hsp70-white transgenes are positioned1.9 and 3.7 kb from the 256 lacO repeats,respectively. (C) Chromosomes were fixed,squashed, and stained with antibodies to LacI. LacIDNA binding sites at position 4D5 are denoted byan arrowhead, region 31 is indicated by an arrow,and the chromocenter is indicated by a ‘‘C.’’

970 K. A. Hines et al.

Page 5: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

with H3K9me, dimerization of HP1, and interactionswith PxVxL-containing proteins are required for silenc-ing at the hsp70 reporter gene.

Previously we demonstrated by chromatin immuno-precipitation (ChIP) analyses that wild-type HP1 spreads�5–10 kb from the lacO repeats and associates with thehsp70 transgene (Danzer and Wallrath 2004). Here,we used ChIP to test whether mutant forms of HP1retained the ability to spread. The cdc2 gene, locatedwithin cytological region 31, was used as a positivecontrol as it is enriched for HP1 (Cryderman et al.2005). The rp49 gene was used as a negative controlbecause it does not associate with HP1 (Danzer andWallrath 2004; Cryderman et al. 2005). ChIP analysesdemonstrated that HP1 was not enriched at hsp70 in theabsence of LacI–HP1 as compared to the negative control(Figure 5 and supporting information, Table S1). Forexample, at position 87C, only 0.12% of input wasobserved at hsp70 in the absence of LacI–HP1 expression,

similar to the enrichment observed at the negativecontrol. In contrast, 2.18% of input was observed forhsp70 upon expression of LacI–HP1, similar to theamount of enrichment observed for the positive control.Expression of LacI–I191E and LacI–W200A showed 0.13and 0.07% of input at hsp70 integrated at position 87C,similar to the negative control. These data demonstratea lack of spreading to hsp70 upon expression of themutant HP1 proteins and suggest a requirement fordimerization and PxVxL interactions in spreading.

Models for heterochromatin spreading involve re-cruitment of SU(VAR)3-9 by HP1 and subsequent methy-lation of H3K9 (Bannister et al. 2001; Elgin andGrewal 2003). In Drosophila, HP1 shows significantcolocalization with H3K9me2 (Li et al. 2002; Ebert et al.2004). To determine whether tethered LacI–HP1 causeddimethylation of H3K9 at the sites of ectopic hetero-chromatin,ChIPanalyses wereperformedusingantibodiesthat recognize H3K9me2. In the absence of tethered

Figure 2.—Colocalization of LacI–HP1 pro-teins and Hip on larval polytene chromosomes.Chromosomes were fixed, squashed, and stainedwith antibodies to LacI and Hip. The location ofthe lacO reporter transgene cassette inserted atcytological position 4D5 is indicated by an arrow;Co., colocalization.

Figure 3.—Colocalization of LacI-HP1 pro-teins and Suv4-20 on larval polytene chromo-somes. Chromosomes were fixed, squashed andstained with antibodies to LacI and Suv4-20.The location of the lacO reporter transgene cas-sette inserted at cytological position 4D5 or 87Cis indicated by an arrow; Co., colocalization.

Drosophila Heterochromatin Spreading 971

Page 6: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

LacI–HP1, H3K9me2 was not detected at hsp70, with0.16% of input at position 87C1 (Figure 5A and TableS1). In contrast, histone methylation at hsp70 wascomparable to the cdc2 positive control upon expressionof LacI–HP1, yielding 2.18% of input (Figure 5A andTable S1). These studies demonstrate that tethered HP1led to dimethylation of H3K9 at distances of at least3.7 kb.

The SU(VAR)3-9 HMTrequires HP1 dimerization forinteraction (Yamamoto and Sonoda 2003). Therefore,we tested whether the LacI–I191E dimerization mutantwould support histone methylation. Upon expression ofLacI–I191E, only 0.16% of input was observed for thehsp70 promoter(Figure 5A).Similar resultswere obtainedfollowing expression of LacI–W200A, which resulted in0.13% of input at hsp70 (Figure 5A and Table S1). Ex-periments performed with the reporter transgenesinserted at position 4D5 yielded similar values, display-ing a lack of methylation at hsp70 following expression ofLacI–I191E or LacI–W200A (Figure 5B and Table S2).Taken together, these findings suggest that HP1 di-merization and interaction with PxVxL-containing pro-teins are needed for HP1 to spread, recruit H3K9 HMTs,and silence gene expression at the hsp70 promoter.

Mutant forms of HP1 support silencing, spreading,and methylation at the hsp26 promoter: In contrast toresults obtained for silencing at hsp70, all mutant formsof LacI–HP1 silenced the hsp26 transgene (Figure 4B).Expression levels of the transgene were 25% or less thanthat obtained upon expression of LacI–GFP and com-parable to that obtained with LacI–HP1. Thus, non-overlapping domains of HP1 were sufficient to achievesilencing at the hsp26 reporter and the inability to dimerizeor interact with a PxVxL partner did not hinder silencing.

In agreement with the ability of the mutant HP1proteins to silence expression of the hsp26 promoter,ChIP analyses revealed that these proteins were associ-ated with hsp26. Expression of LacI–I191E and LacI–W200A showed 2.69 and 3.38% of input for HP1 athsp26, values similar to that for wild-type LacI–HP1(Figure 5A and Table S1). No enrichment was observedat the hsp26 reporter in the absence of fusion-proteinexpression (0.18% of input). Similar results were ob-tained for the reporter gene cassette inserted at position4D5 (Figure 5B and Table S2). These data indicate thatHP1 association with the hsp26 reporter occurs in theabsence of dimerization or PxVxL interactions.

Surprisingly, expression of LacI–I191E led to H3K9dimethylation at hsp26, showing 2.69% of input (Figure5A). Similar results were obtained following expressionof LacI–W200A, where 4.00% of input was observed athsp26 (Figure 5A and Table S1). H3K9me was notenriched at the hsp26 reporter observed in the absenceof fusion-protein expression (0.17% of input). Experi-ments performed with the reporter transgenes insertedat position 4D5 yielded similar values, showing H3K9me2at hsp26 upon tethering either mutant form of HP1(Figure 5B and Table S2). Together, these data suggestthat HP1 is capable of associating with the hsp26 reporter,recruiting an H3K9 HMT, and silencing gene expressionvia a mechanism that is independent of dimerizationand PxVxL interactions. Given that dimerization of mam-malian HP1 is required for interaction with SUV39-H1(Yamamoto and Sonoda 2003), methylation athsp26 is likely due to another histone methyl transferase.

dSETDB1 functions in silencing the hsp26 promoter:The gene that encodes the dSETDB1 (Egg) H3K9-specificHMT, named dsetdb1, eggless, and dEset, is required forfemale fertility in Drosophila with mutants displayingreduced levels of H3K9me1, me2 and me3, and partialloss of PEV (Clough et al. 2007; Seum et al. 2007; Tzeng

et al. 2007; Yoon et al. 2008; Brower-Toland et al. 2009).To begin to define the contribution of dSETDB1 toheterochromatin gene silencing and spreading, we firstdetermined the ability of HP1 to recruit dSETDB1 tosites of ectopic heterochromatin. Immunofluorescencemicroscopy of polytene chromosomes revealed thatdSETDB1 colocalized with LacI–HP1 at the lacO tether-ing site (Figure 6A). These results suggested thatdSETDB1 may be responsible for the observed methyl-ation and silencing of the hsp26 reporter. To address this

Figure 4.—Effects of mutant forms of lacI–HP1 on reportergene expression. Heat-shock-induced expression of hsp26-tagand hsp70-white in the presence or absence of LacI–HP1 orGFP–LacI (negative control). (A) Graphical representationof results from three independent Northerns for hsp70 insertedat three different cytological locations: 4D5 (solid bars), 79D1(dark shaded bars), and 87C1 (light shaded bars). Percentageof relative expression was determined by setting the expressionvalue for the reporter gene without tethered HP1 at 100%(-, open bar). Error bars represent standard error of the mean.(B) Graphical representation of results obtained from threeindependent Northerns for hsp26. Designations and calcula-tions were the same as described for B.

972 K. A. Hines et al.

Page 7: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

possibility, we tested the ability of HP1 to silence thereporter genes at sites of ectopic heterochromatin in eggmutants. Transheterozygous egg mutants displayed asmall, but statistically significant reduction in silencingat the hsp70 transgene (Figure 6B). Conversely, silencingat the hsp26 transgene was severely compromised, withlevels of expression similar to the GFP negative control(Figure 6C). These results suggest that dSETDB1 isresponsible for silencing the hsp26 promoter, and to alesser extent, the hsp70 promoter.

DISCUSSION

Domain-dependent localization of HP1: The LacI–lacO tethering system has proven useful for studies ofchromosome pairing (Vazquez et al. 2002), chromatincompaction (Verschure et al. 2005; Brink et al. 2006;Deng et al. 2008; Strukov and Belmont 2009), andchromatin-mediated gene silencing (Danzer andWallrath 2004). Previously, we demonstrated thatLacI–HP1 rescues lethality in Su(var)2-5 mutants (Li

et al. 2003). Tethering of LacI–HP1 caused gene silenc-ing and altered chromatin structure (Danzer andWallrath 2004). Here, we employed the LacI–lacOsystem to characterize the ability of independent do-mains of HP1 to support gene silencing and spreading.As expected, each of the LacI fusion proteins localized tothe lacO repeats; however, fusion-protein specific local-

ization to endogenous sites was also observed, providinginsight into the requirements for localization to centro-meric and euchromatic regions. For example, only thewild-type fusion protein, LacI–HP1, showed strong local-ization to euchromatic region 31, suggesting a mecha-nism of association with this region that requires thefunction of all three domains of HP1. However, weakassociation at region 31 by LacI–W200A suggests thatPxVxL interactions may be dispensable for recruitmentof HP1 to this region. Surprisingly, the LacI–CD proteincompletely failed to associate with the chromocenter,indicating that the H3K9me binding domain alone isnot sufficient for this association. LacI–I191E, which isdeficient in dimerization and PxVxL interactions, dis-played strong localization to the chromocenter. Takentogether, these data support a role for the hinge regionin this association as suggested by studies of hingephosphorylation-mutant proteins (Badugu et al. 2005).

Mechanisms of silencing by HP1: Previously, wedetermined that SU(VAR)3-9 was required for silencingat the hsp70 reporter, 3.7 kb from the lacO repeats, andcontributed to silencing at the hsp26 reporter (Danzer

and Wallrath 2004). Here we show that silencing at thehsp70 promoter occurred only upon tethering of wild-type LacI–HP1, containing an intact CD and CSD. Giventhat the HP1 CSD interacts directly with SU(VAR)3-9(Schotta et al. 2002) and that this interaction requiresdimerization of HP1 (Yamamoto and Sonoda 2003), it

Figure 5.—ChIP analysis of the reportergenes. (A) Graphical representations ofthree independent ChIP experiments fol-lowing expression of LacI–HP1, mutantforms of HP1 or without tethering (�) forreporter genes inserted at cytological posi-tion 87C1. Chromatin was immunoprecipi-tated with antibodies to GFP (negativecontrol), HP1 and H3K9me2. Primers cor-responding to hsp26 and hsp70-white wereused to amplify precipitated material. Pri-mers to rp49 were used as a negative controland primers corresponding to cdc2 as a pos-itive control (see materials and methods

for details). (B) Same as in A, for reportersinserted at cytological position 4D5.

Drosophila Heterochromatin Spreading 973

Page 8: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

is likely that the loss of silencing results from lack of inter-action with SU(VAR)3-9. The LacI–W200A mutant alsofails to silence at hsp70, suggesting that other factorscontaining PxVxL motifs are likely needed for spreading.

Silencing observed with full-length proteins such asLacI–HP1, LacI–I191E, and LacI–W200A might involvethe hinge region, which remains intact in these fusionproteins. The hinge interacts with numerous partners(Nielsen et al. 2001; Maison et al. 2002; Muchardt et al.2002; Zhang et al. 2002; Hale et al. 2006). For example,the hinge region of HP1Hs-alpha binds histone H1b in aphosphorylation-dependent manner (Hale et al. 2006).Hypophosphorylated H1 is postulated to be involved inrecruitment of HP1 to centric heterochromatin. Thehinge region of mouse HP1a has been implicated incontrolling muscle differentiation transcriptional pro-grams through interactions with complexes containinghistone deacetylases (HDACs) (Zhang et al. 2002),suggesting possible mechanisms of gene silencing.

The HP1 CD and CSD support silencing of the hsp26promoter: The LacI tethering system allowed for a

determination of the requirements for silent chromatinformation and spreading. Surprisingly, both the HP1 CDand CSD were sufficient to cause silencing at 1.9 kb.Consistent with this finding, the CD and CSD of mouseHP1beta caused gene silencing when tethered 200 bpupstream of a reporter gene (Lechner et al. 2000). Incontrast, the CD of human HP1Hs-alpha was unable to silenceat short distances. Differences in the amino acid sequence,structure, and/or function between the human and mouseHP1 CDs might account for the contrasting results. In-terestingly, mouse HP1beta shows a greater percentageof identity to Drosophila HP1 than HP1Hs-alpha, suggestingthat amino acids conserved between Drosophila HP1and mouse HP1beta, but absent from HP1Hs-alpha mightbe key residues involved in short-distance silencing.

Tethering of the CD alone was sufficient to silence thehsp26 promoter, but not hsp70. There are known partnerproteins that interact with the CD of HP1, such as theorigin recognition complex proteins (Pak et al. 1997;Huang et al. 1998) and the retinoblastoma protein(Williams and Grafi 2000). Silencing mediated by theCD alone may be the result of interactions with these orother proteins. Similarly, the LacI–CSD protein was suffi-cient for silencing of the hsp26 reporter, but not hsp70.Silencing across a short distance has also been observedupon tethering the murine HP1beta CSD 200 bp up-stream of a reporter gene (Lechner et al. 2000). TheCSD interacts with a wide variety of nuclear proteinsincluding SU(VAR)3-9. The fact that silencing at hsp26was not abolished in a Su(var)3-9 null (Danzer andWallrath 2004) suggests that partners other thanSU(VAR)3-9 are involved. The ability of these non-overlapping domains of HP1 to induce gene silencingwithin 1.7 kb suggests multiple, independent mecha-nisms for the initiation of gene silencing by HP1.

Histone methylation plays a role in gene silencing atboth hsp26 and hsp70: Upon tethering LacI–HP1,H3K9me2 is detected at both the hsp26 and hsp70 re-porter transgenes by ChIP analyses (Figure 5, Table S1,and Table S2). Similarly, targeting HP1Hs-alpha, HP1Hs-beta,or a CSD to reporter genes in mammalian cell culturestudies caused H3K9me3, in this case by SETDB1(Verschure et al. 2005; Brink et al. 2006). Likewise,targeting HP1Hs-alpha or HP1Hs-gamma upstream of a luciferasetransgene in mammalian cells caused recruitment ofSETDB1 (Ayyanathan et al. 2003). In Drosophila HP1tethering studies, two key findings implicate involve-ment of SU(VAR)3-9 in H3K9 methylation associatedwith silencing of the hsp70 promoter. First, Su(var)3-9null mutants show a dramatic reduction of silencing atthe hsp70 promoter (Danzer and Wallrath 2004).Second, a loss of silencing at hsp70 was observed for theI191E mutant predicted to disrupt interactions withSU(VAR)3-9.

In contrast to the loss of silencing at the hsp70 reporterin a Su(var)3-9 mutant background, silencing and H3K9meat the hsp26 reporter remained. These data suggest histone

Figure 6.—Loss of dSETDB1 affects silencing by tetheredHP1. (A) Chromosomes were fixed, squashed, and stainedwith antibodies to LacI and dSETDB1. (B) Graphical repre-sentation of results obtained from four independent RT–PCR experiments for hsp26 upon tethering of HP1 in the pres-ence or absence of dSETDB1 (wt and mutant, respectively).Expression from the transgene with no tether (�) or withGFP–LacI serve as controls. Error bars represent standarderror of the mean. (C) Same as in B except primers for thehsp70 reporter were used for RT–PCR analysis.

974 K. A. Hines et al.

Page 9: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

methyltransferases other than SU(VAR)3-9 are recruitedby HP1. Many histone methyltransferases share a com-mon SET domain (Alvarez-Venegas and Avramova

2002). On the basis of in silico analyses,�31 SET domain-containing proteins are encoded by the Drosophila ge-nome (Yang et al. 2008), providing multiple candidatesfor an involvement in hsp26 reporter silencing accompa-nied by H3K9 dimethylation. Brower-Toland et al.(2009) recently reported on the spatial and temporalseparation of the function of three of these: SU(VAR)3-9, dSETDB1, and dG9a. Their data suggest thatSU(VAR)3-9 and dSETDB1 both function in methyla-tion of H3K9 in pericentric heterochromatin, withSU(VAR)3-9 acting during early developmental stages,and dSETDB1 maintaining the pattern of methylationthrough metamorphosis. In contrast, dG9a does not ap-pear to act upon pericentric heterochromatin. Our re-sults suggest an additional level of spatial organizationand cooperation between SU(VAR)3-9 and dSETDB1.dSETDB1 appears to be involved in the H3K9 methyla-tion observed at the hsp26 promoter, as loss of dSETDB1results in increased gene expression from the hsp26reporter gene (Figure 6). Conversely, expression of thehsp70 reporter displays a modest increase upon loss ofdSETDB1. Previously, we reported that loss of SU(-VAR)3-9 has the converse effect, loss of silencing at thehsp70 promoter and minimal effect on the hsp26 pro-moter (Danzer and Wallrath 2004). Together, theseresults suggest a model in which initiation of hetero-chromatin at the hsp26 reporter occurs through meth-ylation by dSETDB1, whereas the majority of spreadingand silencing at the hsp70 reporter requires methylationby SU(VAR)3-9.

In Drosophila, H3K9me and H4K20me are both pre-sent in centric heterochromatin and loss of SU(VAR)3-9,an H3K9 HMT, results in loss of H4K20 methylation(Schotta et al. 2004). Furthermore, HP1 is required forwild-type levels of H4K20me (Yang et al. 2008). Similarly,several lines of evidence have pointed to an interactionbetween HP1 and the H4K20 HMT Suv4-20 in mamma-lian systems, including in vitro pull down assays usingmammalian proteins and a requirement for Suvar3-9hfor proper distribution of H4K20me (Schotta et al.2004). Suv4-20 has been reported to act as a suppressorof PEV in Drosophila; however, this activity appears to bedependent upon the allele and reporter combinationbeing examined (Schotta et al. 2004; Sakaguchi et al.2008). Our results demonstrate an in vivo interactionbetween HP1 and Suv4-20 in Drosophila and support ayet-to-be-defined role for H4K20 in heterochromaticregions.

Mechanisms of silent chromatin spreading: Thespread of silent chromatin has been observed in manyorganisms, yet molecular mechanisms that account forspreading are not well understood. At S. cerevisiae telo-meres, silent chromatin is initiated by the DNA bindingprotein Rap1 that recruits Sir proteins, including the

histone deacetylase Sir2 (Perrod and Gasser 2003). Inwild-type cells, Sir3 spreads 2.5 to 6 kb from the telo-mere. In strains mutant for histone acetyltransferasessuch as gcn5, elp3, and sas2, Sir3 proteins spread up to15 kb from the telomere (Suka et al. 2002; Kristjuhan

et al. 2003). Therefore, spreading is controlled throughthe balance of histone acetylation and deacetylation.

In Schizosaccharomyces pombe, the spread of silent chro-matin involves the recruitment of the siRNA machinery(Irvine et al. 2006). Silencing of a ura4 reporter insertedwithin the transcribed repetitive element was moreefficient than silencing the same reporter insertedupstream. Processing of transcripts derived from therepetitive element by Argonaute (Ago1) correlated withthe recruitment of RNA-induced initiation of transcrip-tional silencing (RITS) complex (Verdel et al. 2004)and H3K9 methylation. Paradoxically, these data dem-onstrate a role for RNA polymerase II transcription ingene silencing (Irvine et al. 2006). Consistent with thesefindings, factors involved in processing small RNAs playroles in Drosophila gene silencing (Pal-Bhadra et al.2004; Kotelnikov et al. 2009; Malone et al. 2009).Mutations in homeless, aubergine, and piwi have dominanteffects on silencing due to centric heterochromatin inDrosophila (Pal-Bhadra et al. 2004). A genetic analysisto definitively test the role of RNAi in gene silencing bytethered HP1 has not yet been conducted.

In cases of heterochromatin spreading, genes close toa heterochromatic breakpoint typically exhibited morefrequent and/or severe inactivation than genes locatedmore distally (Weiler and Wakimoto 1995). Molecularmodels involving the linear spread of heterochromatinhave been proposed to explain these data (Locke et al.1988; Tartof et al. 1989). However, there are examplesof discontinuous spreading in which genes are ‘‘skip-ped.’’ In such cases the gene more distal to theheterochromatic breakpoint is silenced to a greaterextent than a proximal gene (Belyaeva and Zhimulev

1991; Talbert and Henikoff 2000). To account for theseobservations a ‘‘coalescence’’ model of heterochromatinspreading has been proposed in which spreading occursthrough pairing of nonadjacent chromosomal regions,frequently containing repetitive DNA sequences(Talbert and Henikoff 2000). Data presented herefor silencing at hsp26 and hsp70 are consistent with alinear spreading model. At the three genomic posi-tions tested, silencing of the hsp70 reporter was notobserved in the absence of silencing at hsp26.

The different response of hsp26 and hsp70 reportergenes to the mutant forms of HP1 might reflect differ-ences in these two promoters. However, both possessheat-shock elements, GAGA factor binding sites (Rougvie

and Lis 1990), and similar amounts of paused RNA poly-merase II (Rougvie and Lis 1990). Previously we demon-strated that both promoters show similar sensitivity toHP1-mediated heterochromatic silencing (Wallrath

and Elgin 1995). If promoter strength were an explana-

Drosophila Heterochromatin Spreading 975

Page 10: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

tion for the differences observed, hsp26 wouldbe predictedto be the weaker of the two, since it is silenced by themutant forms of HP1, whereas hsp70 remains active. Onthe contrary, hsp26 exhibited a greater fold induction thanhsp70 in the absence of HP1 tethering, suggesting hsp26was the stronger of the two promoters (Danzer andWallrath 2004). Thus, the different response of thepromoters to mutant forms of LacI–HP1 is likely due todistance-dependent HP1 spreading, a theory supported byChIP data.

A model for silent chromatin spreading was developedfrom the studies described here (Figure 7). It appears thatmultiple HMTs cooperate with HP1; silencing at shortdistances is predicted to involve dSETDB1 and likelyother H3K9 methyltransferases, whereas, spreading atlong distance requires the CD and CSD and H3K9 methy-lation by SU(VAR)3-9. In addition, our data stronglysupport the involvement of factor(s) possessing PxVxLmotifs in the spread of silent chromatin.

We thank members of the Wallrath lab for comments on the manu-script. We are grateful for research support from the National Institutesof Health (NIH) (GM61513) to L.L.W., a predoctoral fellowship toK.A.H. from the American Heart Association (0415328Z), an NIHRuth L. Kirschtein National Research Service Award postdoctoralfellowship (GM08574) to M.W.V., a National Science Foundation(NSF) grant (MCB-0821893) to C.A.M., and an NSF grant to T.H.

LITERATURE CITED

Alvarez-Venegas, R., and Z. Avramova, 2002 SET-domain proteinsof the Su(var)3–9, E(z) and trithorax families. Gene 285: 25–37.

Ayyanathan, K., M. S. Lechner, P. Bell, G. G. Maul, D. C. Schultz

et al., 2003 Regulated recruitment of HP1 to a euchromaticgene induces mitotically heritable, epigenetic gene silencing: amammalian cell culture model of gene variegation. Genes Dev.17: 1855–1869.

Badugu, R., Y. Yoo, P. B. Singh and R. Kellum, 2005 Mutations inthe heterochromatin protein 1 (HP1) hinge domain affect HP1protein interactions and chromosomal distribution. Chromosoma113: 370–384.

Bannister, A. J., P. Zegerman, J. F. Partridge, E. A. Miska, J. O.Thomas et al., 2001 Selective recognition of methylated lysine9 on histone H3 by the HP1 chromo domain. Nature 410: 120–124.

Belyaeva, E. S., and I. F. Zhimulev, 1991 Cytogenetic and molec-ular aspects of position effect variegation in Drosophila III. Con-tinuous and discontinuos compaction of chromosomal material asa result of position effect variegation. Chromosoma 100: 453–466.

Brasher, S. V., B. O. Smith, R. H. Fogh, D. Nietlispach, A. Thiru

et al., 2000 The structure of mouse HP1 suggests a unique modeof single peptide recognition by the shadow chromo domaindimer. EMBO J. 19: 1587–1597.

Brink, M. C., Y. van der Velden, W. de Leeuw, J. Mateos-Langerak,A. S. Belmont et al., 2006 Truncated HP1 lacking a functionalchromodomain induces heterochromatinization upon in vivotargeting. Histochem. Cell Biol. 125: 53–61.

Brower-Toland, B., S. D. Findley, L. Jiang, L. Liu, H. Yin et al.,2007 Drosophila PIWI associates with chromatin and interactsdirectly with HP1a. Genes Dev. 21: 2300–2311.

Brower-Toland, B., N. C. Riddle, H. Jiang, K. L. Huisinga and S. C.R. Elgin, 2009 Multiple SET methyltransferases are required tomaintain normal heterochromatin domains in the genome ofDrosophila melanogaster. Genetics 181: 1303–1319.

Clough, E., W. Moon, S. Wang, K. Smith and T. Hazelrigg,2007 Histone methylation is required for oogenesis in Dro-sophila. Development 134: 157–165.

Cowieson, N. P., J. F. Partridge, R. C. Allshire and P. J. McLaughlin,2000 Dimerisation of a chromo shadow domain and distinctionsfrom the chromodomain as revealed by structural analysis. Curr.Biol. 10: 517–525.

Cryderman, D. E., S. K. Grade, Y. Li, L. Fanti, S. Pimpinelli et al.,2005 Role of Drosophila HP1 in euchromatic gene expression.Dev. Dyn. 232: 767–774.

Danzer, J. R., and L. L. Wallrath, 2004 Mechanisms of HP1-mediated gene silencing in Drosophila. Development 131:3571–3580.

Deng, H., X. Bao, W. Cai, M. J. Blacketer, A. S. Belmont et al.,2008 Ectopic histone H3S10 phosphorylation causes chromatinstructure remodeling in Drosophila. Development 135: 699–705.

Ebert, A., G. Schotta, S. Lein, S. Kubicek, V. Krauss et al.,2004 Su(var) genes regulate the balance between euchromatinand heterochromatin in Drosophila. Genes Dev. 18: 2973–2983.

Eissenberg, J. C., T. C. James, D. M. Foster-Hartnett, T. Hartnett,V. Ngan et al., 1990 Mutation in a heterochromatin-specificchromosomal protein is associated with suppression of posi-tion-effect variegation in Drosophila melanogaster. Proc. Natl.Acad. Sci. USA 87: 9923–9927.

Elgin, S. C., and S. I. Grewal, 2003 Heterochromatin: silence isgolden. Curr. Biol. 13: R895–R898.

Fanti, L., G. Giovinazzo, M. Berloco and S. Pimpinelli, 1998 Theheterochromatin protein 1 prevents telomere fusions in Dro-sophila. Mol. Cell 2: 527–538.

Fanti, L., M. Berloco, L. Piacentini and S. Pimpinelli,2003 Chromosomal distribution of heterochromatin protein1 (HP1) in Drosophila: a cytological map of euchromatic HP1binding sites. Genetica 117: 135–147.

Hale, T. K., A. Contreras, A. J. Morrison and R. E. Herrera,2006 Phosphorylation of the linker histone H1 by CDK regu-lates its binding to HP1alpha. Mol. Cell 22: 693–699.

Haynes, K. A., A. A. Caudy, L. Collins and S. C. R. Elgin,2006 Element 1360 and RNAi components contribute toHP1-dependent silencing of a pericentric reporter. Curr. Biol.16: 2222–2227.

Huang, D. W., L. Fanti, D. T. Pak, M. R. Botchan, S. Pimpinelli

et al., 1998 Distinct cytoplasmic and nuclear fractions of Dro-sophila heterochromatin protein 1: their phosphorylation levelsand associations with origin recognition complex proteins. J. CellBiol. 142: 307–318.

Irvine, D. V., M. Zaratiegui, N. H. Tolia, D. B. Goto, D. H. Chit-

wood et al., 2006 Argonaute slicing is required for heterochro-matic silencing and spreading. Science 313: 1134–1137.

Jacobs, S. A., S. D. Taverna, Y. Zhang, S. D. Briggs, J. Li et al.,2001 Specificity of the HP1 chromo domain for the methylatedN-terminus of histone H3. EMBO J. 20: 5232–5241.

James, T. C., J. C. Eissenberg, C. Craig, V. Dietrich, A. Hobson

et al., 1989 Distribution patterns of HP1, a heterochromatin-associated nonhistone chromosomal protein of Drosophila.Eur. J. Cell Biol. 50: 170–180.

Figure 7.—A model of ectopic heterochromatin spreadingby HP1. Binding of the LacI DNA binding domain to the lacOrepeats tethers HP1 to the reporter transgene. Tethered HP1 iscapable of maintaining interactions with partner proteins suchas Hip and Suv4-20. Spreading of silent heterochromatininvolves the action of dSETDB1 at the hsp26 promoter andSU(VAR)3-9 downstream at the hsp70 promoter. Interactionswith PxVxL-containing proteins are also predicted to play arole in silencing at the hsp70 promoter. CS, chromo shadow do-main; CD, chromo domain; I, LacI DNA binding domain; M,H3K9me2.

976 K. A. Hines et al.

Page 11: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

James, T. C., and S. C. Elgin, 1986 Identification of a nonhistonechromosomal protein associated with heterochromatin in Dro-sophila melanogaster and its gene. Mol. Cell. Biol. 6: 3862–3872.

Kotelnikov, R. N., M. S. Klenov, Y. M. Rozovsky, L. V. Olenina, M.V. Kibanov et al., 2009 Peculiarities of piRNA-mediated post-transcriptional silencing of Stellate repeats in testes of Drosoph-ila melanogaster. Nucleic Acids Res. 37: 3254–3263.

Kristjuhan, A., B. O. Wittschieben, J. Walker, D. Roberts, B. R.Cairns et al., 2003 Spreading of Sir3 protein in cells with severehistone H3 hypoacetylation. Proc. Natl. Acad. Sci. USA 100:7551–7556.

Lechner, M. S., G. E. Begg, D. W. Speicher and F. J. Rauscher, 3rd,2000 Molecular determinants for targeting heterochromatinprotein 1-mediated gene silencing: direct chromoshadowdomain-KAP-1 corepressor interaction is essential. Mol. Cell. Bi-ol. 20: 6449–6465.

Lechner, M. S., D. C. Schultz, D. Negorev, G. G. Maul and F. J.Rauscher, 3rd, 2005 The mammalian heterochromatin pro-tein 1 binds diverse nuclear proteins through a common motifthat targets the chromoshadow domain. Biochem. Biophys.Res. Commun. 331: 929–937.

Li, Y., J. R. Danzer, P. Alvarez, A. S. Belmont and L. L. Wallrath,2003 Effects of tethering HP1 to euchromatic regions of theDrosophila genome. Development 130: 1817–1824.

Li, Y., D. A. Kirschmann and L. L. Wallrath, 2002 Does hetero-chromatin protein 1 always follow code? Proc. Natl. Acad. Sci.USA 99(Suppl 4): 16462–16469.

Locke, J., M. A. Kotarski and K. D. Tartof, 1988 Dosage-dependentmodifiers of position effect variegation in Drosophila and a massaction model that explains their effect. Genetics 120: 181–198.

Maison, C., D. Bailly, A. H. Peters, J. P. Quivy, D. Roche et al.,2002 Higher-order structure in pericentric heterochromatin in-volves a distinct pattern of histone modification and an RNAcomponent. Nat. Genet. 30: 329–334.

Malone, C. D., J. Brennecke, M. Dus, A. Stark, R. W. McCombie

et al., 2009 Specialized piRNA pathways act in germline and so-matic tissues of the Drosophila ovary. Cell 137: 522–535.

Muchardt, C., M. Guilleme, J. S. Seeler, D. Trouche, A. Dejean

et al., 2002 Coordinated methyl and RNA binding is requiredfor heterochromatin localization of mammalian HP1alpha.EMBO Rep. 3: 975–981.

Nielsen, A. L., M. Oulad-Abdelghani, J. A. Ortiz, E. Remboutsika,P. Chambon et al., 2001 Heterochromatin formation in mam-malian cells: interaction between histones and HP1 proteins.Mol. Cell 7: 729–739.

Pak, D. T., M. Pflumm, I. Chesnokov, D. W. Huang, R. Kellum et al.,1997 Association of the origin recognition complex with het-erochromatin and HP1 in higher eukaryotes. Cell 91: 311–323.

Pal-Bhadra, M., B. A. Leibovitch, S. G. Gandhi, M. Rao, U. Bhadra

et al., 2004 Heterochromatic silencing and HP1 localization inDrosophila are dependent on the RNAi machinery. Science 303:669–672.

Perrod, S., and S. M. Gasser, 2003 Long-range silencing and posi-tion effects at telomeres and centromeres: parallels and differen-ces. Cell. Mol. Life Sci. 60: 2303–2318.

Robinett, C. C., A. Straight, G. Li, C. Willhelm, G. Sudlow et al.,1996 In vivo localization of DNA sequences and visualization oflarge-scale chromatin organization using lac operator/repressorrecognition. J. Cell Biol. 135: 1685–1700.

Rougvie, A. E., and J. T. Lis, 1990 Postinitiation transcriptional con-trol in Drosophila melanogaster. Mol. Cell. Biol. 10: 6041–6045.

Rubin, G., and A. Spradling, 1982 Genetic transformation of Dro-sophila with transposable element vectors. Science 218: 348–353.

Sakaguchi, A., D. Karachentsev, M. Seth-Pasricha, M. Druzhinina

and R. Steward, 2008 Functional characterization of the Dro-sophila Hmt4–20/Suv4–20 histone methyltransferase. Genetics179: 317–322.

Schotta, G., A. Ebert, V. Krauss, A. Fischer, J. Hoffmann et al.,2002 Central role of Drosophila SU(VAR)3–9 in histone H3–K9 methylation and heterochromatic gene silencing. EMBO J.21: 1121–1131.

Schotta, G., M. Lachner, K. Sarma, A. Ebert, R. Sengupta et al.,2004 A silencing pathway to induce H3–K9 and H4–K20 trimethy-lation at constitutive heterochromatin. Genes Dev. 18: 1251–1262.

Schultz, D. C., K. Ayyanathan, D. Negorev, G. G. Maul and F. J.Rauscher, 3rd, 2002 SETDB1: a novel KAP-1-associated his-tone H3, lysine 9-specific methyltransferase that contributes toHP1-mediated silencing of euchromatic genes by KRAB zinc-fingerproteins. Genes Dev. 16: 919–932.

Schwendemann, A., T. Matkovic, C. Linke, A. Klebes, A. Hofmann

et al., 2008 Hip, an HP1-interacting protein, is a haplo- and trip-lo-suppressor of position effect variegation. Proc. Natl. Acad. Sci.USA 105: 204–209.

Seum, C., E. Reo, H. Peng, F. J. Rauscher, III, P. Spierer et al.,2007 Drosophila SETDB1 Is Required for Chromosome 4 Silenc-ing. PLoS Genet. 3: e76.

Strukov, Y. G., and A. S. Belmont, 2009 Mitotic chromosome struc-ture: reproducibility of folding and symmetry between sisterchromatids. Biophys. J. 96: 1617–1628.

Suka, N., K. Luo and M. Grunstein, 2002 Sir2p and Sas2p oppos-ingly regulate acetylation of yeast histone H4 lysine16 andspreading of heterochromatin. Nat. Genet. 32: 378–383.

Sun, F.-L., M. H. Cuaycong and S. C. R. Elgin, 2001 Long-RangeNucleosome Ordering Is Associated with Gene Silencing in Dro-sophila melanogaster Pericentric Heterochromatin. Mol. Cell.Biol. 21: 2867–2879.

Sun, F.-L., K. Haynes, C. L. Simpson, S. D. Lee, L. Collins et al.,2004 cis-Acting Determinants of Heterochromatin Formationon Drosophila melanogaster Chromosome Four. Mol. Cell. Biol.24: 8210–8220.

Talbert, P. B., and S. Henikoff, 2000 A reexamination of spread-ing of position-effect variegation in the white-roughest region ofDrosophila melanogaster. Genetics 154: 259–272.

Tartof, K. D., C. Bishop, M. Jones, C. A. Hobbs and J. Locke,1989 Towards an understanding of position effect variegation.Dev. Genet. 10: 162–176.

Thiru, A., D. Nietlispach, H. R. Mott, M. Okuwaki, D. Lyon et al.,2004 Structural basis of HP1/PXVXL motif peptide interactionsand HP1 localisation to heterochromatin. EMBO J. 23: 489–499.

Tzeng, T.-Y., C.-H. Lee, L.-W. Chan and C.-K. J. Shen,2007 Epigenetic regulation of the Drosophila chromosome 4by the histone H3K9 methyltransferase dSETDB1. Proc. Natl.Acad. Sci. USA 104: 12691–12696.

Vazquez, J., A. S. Belmont and J. W. Sedat, 2002 The Dynamics ofHomologous Chromosome Pairing during Male DrosophilaMeiosis. Curr. Biol. 12: 1473–1483.

Verdel, A., S. Jia, S. Gerber, T. Sugiyama, S. Gygi et al., 2004 RNAi-mediated targeting of heterochromatin by the RITS complex.Science 303: 672–676.

Verschure, P. J., I. van der Kraan, W. de Leeuw, J. van der Vlag, A.E. Carpenter et al., 2005 In vivo HP1 targeting causes large-scale chromatin condensation and enhanced histone lysinemethylation. Mol. Cell. Biol. 25: 4552–4564.

Wallrath, L. L., and S. C. Elgin, 1995 Position effect variegationin Drosophila is associated with an altered chromatin structure.Genes Dev. 9: 1263–1277.

Weiler, K. S., and B. T. Wakimoto, 1995 Heterochromatin andgene expression in Drosophila. Annu. Rev. Genet. 29: 577–605.

Williams, L., and G. Grafi, 2000 The retinoblastoma protein: abridge to heterochromatin. Trends Plant Sci. 5: 239–240.

Yamamoto, K., and M. Sonoda, 2003 Self-interaction of heteroc-hromatin protein 1 is required for direct binding to histonemethyltransferase, SUV39H1. Biochem. Biophys. Res. Commun.301: 287–292.

Yang, H., J. J. Pesavento, T. W. Starnes, D. E. Cryderman, L. L.Wallrath et al., 2008 Preferential dimethylation of histoneH4 lysine 20 by Suv4–20. J. Biol. Chem. 283: 12085–12092.

Yoon, J., K.-S. Lee, J. S. Park, K. Yu, S.-G. Paik et al.,2008 dSETDB1 and SU(VAR)3–9 sequentially function dur-ing germline-stem cell differentiation in Drosophila melanogaster.PLoS ONE 3: e2234.

Zhang, C. L., T. A. McKinsey and E. N. Olson, 2002 Association ofclass II histone deacetylases with heterochromatin protein 1: po-tential role for histone methylation in control of muscle differen-tiation. Mol. Cell. Biol. 22: 7302–7312.

Communicating editor: J. A. Birchler

Drosophila Heterochromatin Spreading 977

Page 12: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

Supporting Information http://www.genetics.org/cgi/content/full/genetics.109.105338/DC1

Domains of HP1 Required for Drosophila melanogaster Heterochromatin Spreading

Karrie A. Hines, Diane E. Cryderman, Kaitlin M. Flannery, Hongbo Yang, Michael W. Vitalini, Tulle Hazelrigg, Craig A. Mizzen and Lori L. Wallrath

Copyright © 2009 by the Genetics Society of America DOI: 10.1534/genetics.109.105338

Page 13: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

K. A. Hines et al. 2 SI

TABLE S1

Summary of ChIP analyses for reporter transgenes inserted at 87C1

0BTether rp49 (negative control) cdc2 (positive control) hsp26-tag hsp70-white

1B87C1 4BGFP 5BHP1 H3K9me 6BGFP 7BHP1 mK9H3 8BGFP 9BHP1 mK9H3 10BGFP 11BHP1 H3K9me

none 0.08U+U0.05 0.03U+U0.02

(0.48)

0.06U+U0.04

(0.73)

0.14U+U0.06

1.03U+U0.33

(0.10)

2.00U+U0.91

(0.18)

0.09U+U0.06 0.18U+U0.05

(0.29)

0.17U+U0.07

(0.41)

0.08U+U0.06

0.12U+U0.05

(0.06)

0.16U+U0.07

(0.43)

HP1 0.07U+U0.02 0.09U+U0.04

(0.59)

0.06U+U0.02

(0.91)

0.14U+U0.08 1.28U+U0.39

(0.10)

3.44U+U1.16

(0.10)

0.22U+U0.04 2.08U+U0.91

(0.18)

3.37U+U0.04

(0.09)

0.34U+U0.19 2.18U+U0.79

(0.15)

1.74U+U0.06

(0.02)

I191E 0.08U+U0.02 0.02U+U0.01

(0.86)

0.13U+U0.12

(0.46)

0.15U+U0.06 1.20U+U0.12

(0.01)

2.16U+U0.18

(0.01)

0.27U+U0.21 2.69U+U1.00

(0.26)

5.70U+U3.17

(0.32)

0.05U+U0.01 0.13U+U0.07

(0.39)

0.16U+U0.13

(0.50)

W200A 0.19U+U0.01 0.04U+U0.02

(0.21)

0.01U+U0.01

(0.59)

0.30U+U016 1.17U+U0.08

(0.01)

1.36U+U0.19

(0.01)

0.19U+U0.09 3.38U+U1.14

(0.10)

4.00U+U1.5

(0.10)

0.03U+U0.01 0.07U+U0.04

(0.53)

0.13U+U0.01

(0.24)

Antibodies to GFP, HP1 and di-methylH3K9 (H3K9me) were used for ChIP.

p values are listed in parentheses below the relative percent enrichment.

Page 14: Domains of Heterochromatin Protein 1 Required ... - Genetics · *Department of Biochemistry, University of Iowa, Iowa City, Iowa 52242, †Department of Cell and Developmental Biology,

K. A. Hines et al. 3 SI

TABLE S2

Summary of ChIP analyses for reporter transgenes inserted at 4D5

2BTether rp49 (negative control) cdc2 (positive control) hsp26-tag hsp70-white

3B87C1 12BGFP 13BHP1 mK9H3 14BGFP 15BHP1 mK9H3 16BGFP 17BHP1 mK9H3 18BGFP 19BHP1 mK9H3

none 0.16U+U0.03 0.30U+U0.15

(0.47)

0.22U+U0.8

(0.60)

0.25U+U0.02

2.54U+U0.11

(0.01)

3.27U+U2.20

(0.31)

0.20U+U0.10 0.05U+U0.17

(0.23)

0.24U+U0.17

(0.85)

0.23U+U0.08

0.55U+U0.14

(0.49)

0.45U+U0.16

(0.04)

HP1 0.16U+U0.12 0.23U+U0.12

(0.46)

0.10U+U0.04

(0.64)

0.00U+U0.01 4.80U+U2.47

(0.24)

2.30U+U1.30

(0.09)

0.23U+U0.15 2.05U+U0.05

(0.01)

3.55U+U0.96

(0.10)

0.16U+U0.05 2.37U+U0.33

(0.01)

1.58U+U0.42

(0.03)

I191E 0.06U+U0.03 0.13U+U0.08

(0.28)

0.00U+U0.01

(0.84)

0.47U+U0.28

5.26U+U2.20

(0.06)

2.35U+U0.15

(0.01)

0.88U+U0.01 3.57U+U0.68

(0.07)

4.10U+U0.10

(0.28)

0.01U+U0.01 0.55U+U0.08

(0.14)

0.45U+U0.15

(0.31)

W200A 0.04U+U0.01 0.14U+U0.07

(0.29)

0.01U+U0.01

(0.34)

0.07U+U0.02 1.37U+U0.26

(0.04)

1.63U+U0.42

(0.19)

0.36U+U0.05 1.29U+U0.43

(0.16)

6.00U+U2.00

(0.10)

0.35U+U0.16 0.22U+U0.14

(0.57)

0.29U+U0.23

(0.83)

Antibodies to GFP, HP1 and di-methylK9H3 (mK9H3) were used for ChIP.

p values are listed in parentheses below the relative percent enrichment.