Open Collections

UBC Faculty Research and Publications

H3K9me3-binding proteins are dispensable for SETDB1/H3K9me3-dependent retroviral silencing Maksakova, Irina A; Goyal, Preeti; Bullwinkel, Jörn; Brown, Jeremy P; Bilenky, Misha; Mager, Dixie L; Singh, Prim B; Lorincz, Matthew C Jul 20, 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13072_2011_Article_63.pdf [ 2.12MB ]
JSON: 52383-1.0223344.json
JSON-LD: 52383-1.0223344-ld.json
RDF/XML (Pretty): 52383-1.0223344-rdf.xml
RDF/JSON: 52383-1.0223344-rdf.json
Turtle: 52383-1.0223344-turtle.txt
N-Triples: 52383-1.0223344-rdf-ntriples.txt
Original Record: 52383-1.0223344-source.json
Full Text

Full Text

RESEARCH Open AccessH3K9me3-binding proteins are dispensable forSETDB1/H3K9me3-dependent retroviral silencingIrina A Maksakova1, Preeti Goyal1, Jörn Bullwinkel3, Jeremy P Brown3, Misha Bilenky4, Dixie L Mager1,2,Prim B Singh3 and Matthew C Lorincz1*AbstractBackground: Endogenous retroviruses (ERVs) are parasitic sequences whose derepression is associated with cancerand genomic instability. Many ERV families are silenced in mouse embryonic stem cells (mESCs) via SETDB1-deposited trimethylated lysine 9 of histone 3 (H3K9me3), but the mechanism of H3K9me3-dependent repressionremains unknown. Multiple proteins, including members of the heterochromatin protein 1 (HP1) family, bindH3K9me2/3 and are involved in transcriptional silencing in model organisms. In this work, we address the role ofsuch H3K9me2/3 “readers” in the silencing of ERVs in mESCs.Results: We demonstrate that despite the reported function of HP1 proteins in H3K9me-dependent generepression and the critical role of H3K9me3 in transcriptional silencing of class I and class II ERVs, the depletion ofHP1a, HP1b and HP1g, alone or in combination, is not sufficient for derepression of these elements in mESCs.While loss of HP1a or HP1b leads to modest defects in DNA methylation of ERVs or spreading of H4K20me3 intoflanking genomic sequence, respectively, neither protein affects H3K9me3 or H4K20me3 in ERV bodies.Furthermore, using novel ERV reporter constructs targeted to a specific genomic site, we demonstrate that, relativeto Setdb1, knockdown of the remaining known H3K9me3 readers expressed in mESCs, including Cdyl, Cdyl2, Cbx2,Cbx7, Mpp8, Uhrf1 and Jarid1a-c, leads to only modest proviral reactivation.Conclusion: Taken together, these results reveal that each of the known H3K9me3-binding proteins is dispensablefor SETDB1-mediated ERV silencing. We speculate that H3K9me3 might maintain ERVs in a silent state in mESCs bydirectly inhibiting deposition of active covalent histone marks.Keywords: endogenous retrovirus, ERV, heterochromatin protein 1, HP1, Cbx1, Cbx3, Cbx5, H3K9me3, retroviralrepression, transcriptional silencing, mouse embryonic stem cellsBackgroundEndogenous retroviral sequences (ERVs) are relics ofancient retroviral integration into the germline. Theseparasitic elements are abundant in mammals, occupyingapproximately 8% of the mouse genome and 10% of thehuman genome [1,2]. ERVs are subdivided into threediverse classes based on the similarity of their reversetranscriptase genes or their relationship to different gen-era of exogenous retroviruses. In the mouse, class IERVs, similar to gammaretroviruses, include activefamilies such as murine leukaemia viruses (MLVs) andmurine retroviruses that use tRNAGln (GLN). Class IIERVs are similar to alpha- and betaretroviruses andinclude Mus musculus ERV using tRNALys type 10C(MMERVK10C), the highly retrotranspositionally activeintracisternal A-type particles (IAPEz) and early trans-poson/Mus musculus type D retrovirus (ETn/MusD)families. Class III ERVs, the oldest and most abundantERVs, are most similar to spumaviruses and are repre-sented by mouse endogenous retrovirus type L (MERV-L) and mouse apparent LTR retrotransposons (MaLR)[3,4]. Numerous regulatory motifs in the ERV longterminal repeats (LTRs) can initiate high levels of tran-scription in tissues and cell lines [5], and there is exten-sive evidence of aberrant ERV-driven gene expression incancers [6-11] and tissues of aging mice [12,13]. In an* Correspondence: mlorincz@interchange.ubc.ca1Department of Medical Genetics, Life Sciences Institute, University of BritishColumbia, 2350 Health Sciences Mall, Vancouver, BC, Canada, V6T 1Z3Full list of author information is available at the end of the articleMaksakova et al. Epigenetics & Chromatin 2011, 4:12© 2011 Maksakova et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (, which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.effort to counteract the potentially detrimental effects ofERVs, eukaryotic genomes have evolved multiple linesof defence against active exogenous and endogenous ret-roviruses [14], including DNA methylation and repres-sive histone modifications.DNA methylation was the first epigenetic mark recog-nized to contribute to ERV silencing, with dramaticupregulation of ERVs observed in DNA methylation-deficient somatic cells [15,16]. However, genome-widechromatin immunoprecipitation (ChIP) followed byChIP sequencing (ChIP-seq) [17-19] or ChIP followedby quantitative PCR (qPCR) [20] revealed that in mouseembryonic stem cells (mESCs), class I and class II ERVsare enriched for the repressive histone H3 lysine 9 tri-methylation (H3K9me3) deposited by lysine methyl-transferase (KMTase) SETDB1/ESET/KMT1E [20].SETDB1 is in turn thought to be recruited to ERVs viathe obligatory corepressor KRAB-associated protein 1(KAP-1) [21], presumably through sequence-specificKAP-1-binding zinc finger proteins such as ZFP809 inthe case of MLVs [22]. Moreover, we and others haverecently shown that in mESCs, H3K9me3 and SETDB1play a greater role than DNA methylation in the silen-cing of class I and class II ERVs [20,23]. IAP and ETn/MusD retrotransposons, the two most active class IImouse ERV families and the source of numerous recentgermline mutations [24], are among the families withthe highest H3K9me3 enrichment levels. Intriguingly,these families are dramatically upregulated in SETDB1knockout (SETDB1 KO) mESCs [19,20], confirming thatthey have a high potential for activation in the absenceof H3K9me3. In contrast, the class III MERV-L andMaLR families, which are devoid of the H3K9me3 markin mESCs, are repressed by the histone lysine-specificdemethylase 1 (LSD1/KDM1A) [25], revealing that dif-ferent ERV classes are regulated by distinct epigeneticmodifications in these pluripotent cells.Acetylation of lysine residues on the N-terminal tailsof histones, including H3K9, directly influences the stateof chromatin compaction by reducing the affinity of his-tones for DNA [26,27]. In contrast, methylation per seof such lysine residues is less likely to directly affectchromatin structure, as this modification does not altertheir charge. Rather, the prevailing view is that specificproteins, the so-called “readers,” bind to methylatedlysines and coordinate the biological outcome associatedwith such covalent histone marks. H3K9me3, for exam-ple, which is essential for the establishment and mainte-nance of the silent chromatin state [28-31], is bound bythree isoforms of heterochromatin protein 1 (HP1) inthe mouse genome: HP1a (encoded by Cbx5), HP1b(encoded by Cbx1) and HP1g (encoded by Cbx3) [32].HP1 is a highly conserved family; its members are fre-quently present in several copies in eukaryotic genomesand play both structural and gene regulatory roles[33-35]. The chromodomain of HP1 is responsible forbinding H3K9me2/3 [36,37], and a chromoshadowdomain is required for HP1 homo- and heterodimeriza-tion and the recruitment of other proteins [38,39].Although their exact function in transcriptional regu-lation and cross-talk with histone and DNA methylationvaries between species, the ability of HP1s to modulategene expression via H3K9me2/3 binding has beenreported in multiple systems [33,40-42]. In fission yeast,for example, two HP1 homologues, Swi6 and Chp2, areboth required for assembly of repressive chromatin [43].In mammalian cells, targeting of HP1a, HP1b and HP1gto heterologous loci is sufficient to induce recruitmentof SETDB1 and deposition of H3K9me3 [44], and HP1has been implicated in SUV39H1-mediated silencing ofeuchromatic genes [45].A role for HP1 proteins in silencing of repetitiveand/or transposable elements has been well documen-ted in several model organisms. In Drosophila, twofamilies of transposons are derepressed in larvae withmutant HP1a and, to a lesser extent, mutant HP1c[46]. HP1d/Rhino is required for transposon silencingin the female germline of Drosophila, but this silencingseems to stem from Rhino’s role in Piwi-interactingRNA (piRNA) production rather than establishment ofrepressive chromatin [47]. At transposable elements inNeurospora, DNA methylation is dependent on methy-lated H3K9 bound by HP1 [48,49]. In Arabidopsis,however, H3K9me3-directed DNA methylation appliesonly to CpNpG methylation, not to CpG methylation,of transposons [50,51]. HP1g is a negative regulator ofHIV in human cell lines [52] and of non-LTR LINE1retrotransposons in male mouse germ cells [53]. Onthe contrary, HP1g has also been implicated in activat-ing gene expression through its association with elon-gating RNA polymerase II [54,55]. The latter examplenotwithstanding, HP1 proteins are excellent candidatesfor the role of downstream effectors of H3K9me3-dependent silencing affecting ERVs in mESCs. Indeed,an intact HP1-binding domain of KAP-1 is essentialfor complete restriction of MLV in mouse embryoniccarcinoma cells [56]. Furthermore, direct interaction ofHP1 and KAP-1, as well as binding of HP1 toH3K9me3, is necessary for the full extent of silencingmediated by these factors [57-61]. Moreover, we haverecently demonstrated by ChIP-qPCR that HP1a,HP1b and HP1g are enriched on IAPEz, MusD andMLV ERV sequences in mESCs, albeit at modest levels,and that this binding is partially dependent onSETDB1-deposited H3K9me3 [20]. On the basis ofthese observations, we hypothesized that HP1s mightplay a role in H3K9me3-mediated ERV silencing inmESCs and possibly in early embryos.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 2 of 18In addition to their reported roles in transcriptionalsilencing, HP1 proteins are required for heterochroma-tin spreading in specific genomic contexts in Drosophila[62,63], yeast [64] and mammals [42,57]. The presenceof both chromodomains and chromoshadow domainssuggests that HP1 proteins may bind H3K9me3 andrecruit additional proteins, such as SUV39H1/2 orSETDB1-bound KAP-1 [61,65,66], to facilitate thespreading of the repressive H3K9me3 mark [67,68].Intriguingly, repetitive elements may act as foci of denovo heterochromatin formation and spreading, asH3K9me3 is enriched at sequences flanking ERVs[18,19]. Conversely, in Neurospora, HP1 is a componentof a histone demethylase-containing complex that pre-vents spreading of heterochromatin [69].In addition to HP1s, many other mouse chromodo-main proteins [70] are reported to bind H3K9me3 invitro, including CDYL, CDYL2, CBX2, CBX4, CBX7 andM-phase phosphoprotein 8 (MPP8) [71-78]. Further-more, nonchromodomain proteins with affinity forH3K9me3 have also been identified [79]. AlthoughMPP8 and CBX7 have been shown to negatively influ-ence transcription of specific genes [71,80], the func-tional and biological significance of the interaction ofmost of these H3K9me3 readers with H3K9me3 remainspoorly understood.To determine what role, if any, H3K9me3 readersplay in silencing of ERVs and spreading of repressivechromatin from these repetitive elements, we firstgenerated Cbx1 (HP1b) knock-out (KO) and Cbx5(HP1a) KO mESCs [40,81]. Surprisingly, we observedno upregulation of ERVs in Cbx5-/- mESCs and onlymodest upregulation of several ERV families inCbx1-/- mESCs compared to that seen in Setdb1 KOmESCs. We found that both HP1a and HP1b are dis-pensable for DNA methylation of the ETnII/MusDfamily of ERVs, although HP1a has a modest influ-ence on DNA methylation of IAP elements. Further-more, we demonstrate that while deposition ofH4K20me3 at major satellite repeats is dependent inpart on HP1a, as reported previously [82], HP1a andHP1b are dispensable for deposition of H4K20me3 atERVs and play only a modest role in spreading ofH4K20me3 into sequences flanking these elements.Finally, employing RNAi and newly derived mESClines harbouring silenced IAP, MusD and exogenousMLV-based reporters, we show that depletion of all ofthe HP1 proteins, alone or in combination, or each ofthe remaining known H3K9me3-binding proteins, hasonly a modest effect on ERV derepression, indicatingthat at classes I and II ERVs, H3K9me3 inhibits tran-scription independently of HP1 and other knownH3K9me3 readers.ResultsCatalytic activity of SETDB1 is largely required for ERVsilencingWe recently showed by ChIP-qPCR [20] and ChIP-seq[19] analyses that numerous class I and class II ERVfamilies are marked by H3K9me3. Furthermore, wedemonstrated the critical role of SETDB1, the KMTasethat deposits this mark, in transcriptional repression ofthese ERVs. Mapping all H3K9me3 ChIP-seq readsalong the span of the consensus sequences of class Iand class II ERVs, including IAPEz, MusD,MMERVK10C, MLV and GLN, confirms a high butnonuniform level of H3K9me3 along these elements inwt mESCs and a significantly lower level of H3K9me3in Setdb1 KO mESCs (Figure 1A and Figure S3 in Addi-tional file 1). Consistent with these data and those pub-lished in a previous report [18], analysis of the uniquelymapped ChIP-seq reads reveals a high level ofH3K9me3 in the regions flanking IAPEz, MusD andMLV ERVs (Figure 1B and Figure S4 in Additional file1).To confirm that the KMTase activity of SETDB1 iscritical for ERV silencing in mESCs [20], we analyzedthe Setdb1 conditional KO mESC line, either unmodi-fied (SETDB1 KO) or stably expressing wild-type (wt)(SETDB1 KO TG+) or KMTase-defective (SETDB1 KOC1243A) SETDB1 transgenes, the latter harbouring asingle amino acid change in the catalytic domain [20].As expected, robust derepression of ERVs is observed inthe SETDB1 KO line (Figure 1C). Despite the fact thatthe SETDB1 C1243A line expresses an approximatelythreefold higher level of Setdb1 than wt cells (Figure 1Dand [20]), derepression of several of these ERVs is alsoobserved in this transgenic line, confirming thatSETDB1 KMTase activity is essential for ERV silencing.Interestingly, the extent of derepression was dependenton the ERV family. The level of upregulation of MusDand IAPEz elements was equivalent in the SETDB1 KOand catalytic mutant lines, suggesting that silencing ofthese elements depends on the KMTase activity ofSETDB1. MMERVK10C and GLN show a lower level ofderepression in the SETDB1 C1243A line than theSETDB1 KO line, and MLV remains completelyrestricted in the SETDB1 C1243A line. Similar resultswere noted previously in Northern blot analyses [20].Taken together, these results indicate that different ERVfamilies are subject to SETDB1-mediated silencing gen-erally dependent on SETDB1 catalytic activity.Depletion of HP1b but not HP1a leads to modestupregulation of select ERV familiesHaving confirmed that the KMTase activity of SETDB1is required for efficient silencing of MMERVK10C,Maksakova et al. Epigenetics & Chromatin 2011, 4:12 3 of 18MusD and IAPEz, we next sought to determine whetherthe archetypal heterochromatic H3K9me2/3 readersHP1a and HP1b [40], both of which are enriched onIAPEz, MusD and MLV ERVs [20], are the effectors oftranscriptional suppression of these elements. We gener-ated Cbx5 (HP1a) KO mESCs (Figure S1 in Additionalfile 1) and Cbx1 (HP1b) KO mESCs (Figure S2 in Addi-tional file 1) and confirmed downregulation of the cor-responding genes at the mRNA level by qRT-PCR andat the protein level by Western blot analysis. Equivalentlevels of expression of the pluripotency factor Nanogwere detected in these lines, indicating that deletion ofHP1 proteins does not stimulate differentiation (Figure2A). Interestingly, while compensatory upregulation ofthe Cbx1 and Cbx3 genes was observed at the mRNAlevel in the Cbx5-/- line, upregulation of these genes wasnot observed at the protein level (Figure 2B).Surprisingly, unlike deletion of Setdb1, deletion ofCbx5 does not lead to upregulation of any members ofthe ERV families analyzed, as determined by qRT-PCR(Figure 2C) or Northern blot analysis (Figure 2D). Simi-larly, deletion of Cbx1 has no effect on MusD, MLV orGLN elements. Although Cbx1 deletion does result inmodest derepression of MMERVK10C (approximately 3-fold) and IAPEz (approximately 1.5-fold) relative to theparental HM1 line, these ERVs show approximately 47-fold and approximately 3-fold upregulation respectively,in the Setdb1 KO line, relative to the parental TT2 line,(see Figure 1). Taken together, these results indicatethat in contrast to SETDB1, HP1a and HP1b play no&5HODWLYH([SUHVVLRQ0XV'00(59.&,$3(]0/9*/177ZW6HWGE .2+.PH53.0[$%0HDQKLVWRQHPDUNGHQVLW\+.PH 77ZW+.PH 6HWGE .2+.PH 9ZW'LVWDQFHIURPWKH(59NE,$3(]NE¶/75 ¶/756HWGE':76(7'%.26(7'%.26(7'%.27*&$,$3(]ƍƍďďP51$Figure 1 Catalytically active SETDB1 is required for endogenous retrovirus silencing. (A) Profiling of trimethylated lysine 9 of histone 3(H3K9me3) along the length of IAPEz endogenous retroviruses (ERVs) in the TT2 wild type (TT2 wt) and Setdb1 knockout (Setdb1 KO) mouseembryonic stem cells (mESCs) (see Figure S3 in Additional file 1 for profiles of murine leukaemia virus (MLV), MusD, MMERVK10C and GLN ERVs).The profile was generated by aligning chromatin immunoprecipitation assay sequencing (ChIP-seq) reads from TT2 wt and Setdb1 KO mESCs [19]to the consensus sequence of IAPEz. H3K9me3 enrichment levels are presented as reads per kilobase per million mapped reads values (RPKM).(B) Profiling of H3K9me3 and H4K20me3 in the genomic regions flanking 599 IAPEz elements in TT2 wt and Setdb1 KO mESCs (see Figure S4 inAdditional file 1 for MusD and MLV profiles). H3K9me3 ChIP-seq reads from TT2 wt (C57BL/6 ± CBA) and Setdb1 KO mESCs [19] were used, alongwith H4K20me3 ChIP-seq from the wt V6.5 mESCs (129SvJae ± C57BL/6) [18]. Reads were aligned to the mouse genome (mm9), and the densityof reads mapping to the 7-kb regions flanking intact IAPEz ERV families was plotted for H3K9me3 in TT2 wt and Setdb1 KO mESCs and forH4K20me3 in V6.5 wt mESCs. Vertical lines indicate the 5’ and 3’ boundaries of the ERV. The average mappability for 50-bp reads was confirmedto be, on average, uniform in the assayed 7 kb region (data not shown), ruling out the possibility of mapping bias. (C) Setdb1 deletion wasinduced with 4-hydroxytamoxifen (4-OHT) in mESCs containing no transgene (KO), a wt transgene (KO TG+) or a transgene with a mutationrendering SETDB1 catalytically inactive (KO C1243A) [20]. Expression is normalized to b-actin relative to wt. Data are presented as means ±standard deviations (SD) for three technical replicates. (D) To establish the expression levels of Setdb1 in the KO and transgenic lines, quantitativeRT-PCR (qRT-PCR) was performed with Setdb1-specific primers, and expression was normalized to b-actin relative to wt. Data are presented asmeans ± SD for three technical replicates.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 4 of 18role or a relatively minor role, respectively, in class IIERV silencing in mESCs.Depletion of HP1a results in a modest reduction of DNAmethylation at IAPEz ERVsWe recently demonstrated that while G9a is dispensablefor silencing of ERVs, this H3K9 KMTase is required forefficient DNA methylation of these elements in mESCs[83]. Similarly, DNA methylation of major satelliterepeats is dependent upon the H3K9 KMTaseSUV39H1/2 in mESCs [84]. Intriguingly, HP1 proteinsare required for DNA methylation of repetitive elementsin Neurospora [48,85], but the role of HP1 proteins inDNA methylation of ERVs in mESCs has not beenexplored. To address this question, ETnII/MusD andIAPEz families, shown previously to be densely DNAmethylated in mESCs [20,83,86], were analyzed bybisulphite sequencing using genomic DNA isolated fromwt, Cbx1-/- and Cbx5-/- mESCs. In wt cells, severalcopies of ETnII and MusD were either completelyunmethylated or hypomethylated specifically at the 5’end of the LTR (Figures 3A and 3C) as observed pre-viously [86]. The number of methylated CpG sites perelement of this family remained similar in either of theCbx KO lines. In contrast, while the level of DNAmethylation was very high at IAPEz elements in wtcells, several IAP molecules showed reduced levels ofDNA methylation in the Cbx5-/- cell line (Figures 3Band 3C), indicating that HP1a plays a role in DNAmethylation of a subset of IAP elements, presumablydependent upon their genomic location. Nevertheless, asdiscussed above, this modest decrease in DNA methyla-tion did not result in derepression of IAP elements inthese cells.Neither HP1a nor HP1b are essential for H4K20me3deposition at ERVsAlthough H4K20me3 is dispensable for proviral silen-cing in mESCs, its deposition by SUV4-20H at ERVsrequires SETDB1-deposited H3K9me3 [20]. On thebasis of the fact that in mouse embryonic fibroblasts(MEFs), H4K20me3 at satellite repeats is dependent onSUV39H1/2-deposited H3K9me3 and subsequent bind-ing of HP1 to this mark [82], we investigated whetherH4K20me3 at ERVs is also dependent upon HP1 pro-teins in mESCs. Native ChIP (N-ChIP) followed byqPCR revealed that H4K20me3 enrichment was reducedby more than 50% at major satellite repeats in theCbx5-/- line (Figure 3D), demonstrating that as in MEFs[82], HP1 proteins are required for efficient H4K20me3deposition at pericentric heterochromatin in mESCs.However, this mark is not entirely lost in either of theKO lines, presumably due to partial redundancy of HP1proteins at major satellites. In line with these findings, itwas recently shown that HP1b is dispensable forH4K20me3 and H3K9me3 deposition and localization inheterochromatin of mouse neurons [81]. Similarly,H4K20me3 levels at IAPEz, ETnII/MusD and MLVERVs in the Cbx1-/- and Cbx5-/- lines remained at levelssimilar to the wt parent line, demonstrating either thatH4K20me3 is deposited independently of HP1 bindingor that these proteins act redundantly to promotedeposition of H4K20me3 at these elements. As expected,H3K9me3 also remained unaltered in the absence ofHP1a or HP1b (Figure 3D).HP1b plays a role in the spreading of H4K20me3 but notH3K9me3 from ERVs into flanking genomic regionsIntriguingly, while HP1 homologs play a positive role inheterochromatin spreading in Drosophila [62,63] andmammals [42,57], HP1 plays a critical role in inhibitingA C     HM1(wt)          Cbx5-/-           Cbx1-/-    HM1(wt)          Cbx5-/-           Cbx1-/-B HP1H3HP1D Relative expressionRelative expressionHM1(wt)     Cbx5-/-       Cbx1-/-   MusDMMERVK10CIAPEzMLVGLN4. gelIAPMusDETnIIH3K9me3Figure 2 Expression of heterochromatin protein 1 genes andERVs in the Cbx5-/- and Cbx1-/- mESCs. (A) qRT-PCR with primersspecific for Cbx5 (encoding HP1a) Cbx1 (encoding HP1b), Cbx3(encoding HP1g) and the pluripotency factor Nanog in the Cbx5-/-and Cbx1-/- mESC lines confirms the KOs and reveals compensatoryupregulation of the genes encoding the remaining HP1 proteins inthe Cbx5-/- line. Expression levels were normalized to b-actin relativeto wt, and the data are presented as means ± SD for threetechnical replicates. (B) Western blot analysis of whole-cell lysatesconfirms the lack of expression of HP1a and HP1b in the Cbx5-/-and Cbx1-/- mESC lines, respectively. (C) Expression of representativeERV families in the HM1 (wt), Cbx5-/- and Cbx1-/- mESCs wasdetermined by qRT-PCR. Expression levels were normalized to b-actin relative to wt. Data are presented as means ± SD of fourindependent experiments, each of which was performed intriplicate. (D) Northern blot analysis of RNA isolated from theparental HM1,Cbx5-/- and Cbx1-/- mESC lines using probes specificfor ETnII, MusD and IAP ERVs are shown. RNA from Dnmt1-/- mESCs,in which IAP elements are upregulated approximately fourfold[20,86] and MusD elements are upregulated approximately 1.5-fold[86], was used as a control.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 5 of 18aberrant spreading of heterochromatin in Neurospora[69]. Genome-wide analysis of H3K9me3 in wt mESCsreveals high levels of H3K9me3 in the immediate flanksof ERVs, including IAPEz, MusD and MLV elements,with progressively lower levels of this mark at distancesfarther from the ERV integration site [18,19]. Asexpected, deposition of H3K9me3 in these regions isSETDB1-dependent [19] (see Figure 1B and Figure S4 inAdditional file 1). Notably, the profile of H4K20me3 inthe genomic regions flanking IAP, MusD and MLV ele-ments is similar to that of H3K9me3 and the relativelevels of both marks are consistent with their abundanceA B C D HM1(wt)   Cbx5-/-    Cbx1-/-ETnII/MusD IAPEzIAPEz ETnII/MusDETnII/MusD IAPEz% of moleculesNumber of methylated CpGsMLV0. IP relative to H3 0.300.  IgG     H3K9   H4K20              me3     me3  IgG     H3K9   H4K20              me3     me3  IgG     H3K9   H4K20              me3     me3  IgG     H3K9   H4K20              me3     me3Maj SatHM1(wt)   Cbx5-/-   Cbx1-/-Figure 3 DNA methylation and chromatin marks at ERVs in Cbx5-/- and Cbx1-/- mESCs. (A) An approximately 600-bp fragment of the LTRand downstream region of ETnII/MusD ERVs was analyzed by bisulphite sequencing using primers that detect 105 ETnII/MusD elements,according to in silico PCR analysis (UCSC Genome Browser, (B) An approximately 400-bp fragment of the LTR anddownstream region of IAP ERVs was analyzed by bisulphite sequencing using primers that detect 1,461 IAP elements. (C) Bisulphite-sequencedmolecules were binned into four categories, depending on the number of methylated CpG sites detected, and the data are presented as thepercentage of all clones for each cell line in each bin. While the ETnII/MusD family shows no difference in DNA methylation, several IAPmolecules in the Cbx5-/- cell line exhibit partial demethylation. (D) Native ChIP (N-ChIP) with antibodies against H3K9me3, H4K20me3 and pan-H3 was followed by qPCR using primers specific for major satellite repeats and IAP, MLV and MusD ERVs, and the data are presented as means ±SD for three technical replicates. The level of H4K20me3 was reduced by more than 50% at major satellite repeats in the Cbx5-/- mESCs butremained at the same level at ERVs.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 6 of 18Chr 5 IAPEzChr 2 IAPEz% IP relative to H31. IP relative to H3******** *****HM1(wt)Cbx5-/-Cbx1-/-IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3#1 #2 #3 #4IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3IgGH3K9me3H4K20me3#1 #2 #3 #4HM1(wt)Cbx5-/-Cbx1-/-Figure 4 HP1b plays a role in H4K20me3 but not H3K9me3 spreading from ERVs into flanking genomic DNA. N-ChIP was performedwith H3K9me3-, H4K20me3- and pan-H3-specific antibodies using chromatin isolated from HM1, Cbx5-/- and Cbx1-/- mESCs. The level ofenrichment of these modifications at the flanks of two full-length IAP elements on chromosomes 2 and 5 as well as at positions approximately 1kb, 2 kb and 3.5 kb distal to these flanking regions, was determined by qPCR. Data are presented as means ± SD of three technical replicates,and pairs of control and experimental samples with *P < 0.05 and **P < 0.01 (two-tailed Student’s t-test) are shown. H3K9me3 enrichment levelsacross these genomic regions as determined using our previously published ChIP-seq data sets [19] are also shown for wt and Setdb1 KO mESCs.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 7 of 18in each ERV family (that is, IAP > MusD > MLV [20]).To determine whether spreading of H3K9me3 and/orH4K20me3 is affected in Cbx1-/- and/or Cbx5-/- mESCs,we examined these marks at the flanks of two randomlychosen full-length IAPEz elements and three genomiclocations distal to the integration sites of these ERVs byChIP-qPCR (Figure 4). IAPEz elements were chosenbecause, among the ERVs analyzed, on average, thisfamily showed the highest mean H3K9me3 density inflanking genomic regions (Figure 1B and Figure S4 inAdditional file 1). As expected, in wt cells, the levels ofboth H3K9me3 and H4K20me3 generally declined asthe distance from the IAP increased, dropping substan-tially at approximately 3.5 kb. Depletion of either HP1protein did not show a consistent effect on the spread-ing of H3K9me3 into the flanks of the selected IAP ele-ments, since at the majority of regions surveyedenrichment was not statistically significantly different ineach of the KO lines from the wt control. HP1b may beinvolved in propagation of H3K9me3 beyond 2 kb fromthe IAP assayed on chromosome 2, however, suggestingthat at least at some loci, HP1b may facilitate thespreading of H3K9me3.Analysis of H4K20me3 in the same regions revealedno decrease in this mark in the Cbx5-/- line. In contrast,relative to the HM1 parent line, the Cbx1-/- line showeda consistent, approximately 1.5- to 2-fold decrease (P <0.05, two-tailed Student’s t-test) in H4K20me3 at bothloci in distal regions 2 and 3 (Figure 4). Thus, whileneither HP1 protein is required for deposition ofH4K20me3 at the ERVs themselves (see Figure 3D),HP1b may generally be involved in the spreading of thiscovalent mark into the genomic regions flanking theserepetitive elements.Application of novel ERV reporter lines in a siRNA-basedscreen of H3K9me3-binding proteinsIn addition to HP1 proteins, a number of other chromo-domain proteins, including CDYL2, CBX2, CBX4, CBX7and MPP8, as well as the Tudor domain-containing pro-tein TDRD7, were recently shown to bind H3K9me3 invitro [71-73,75,76]. To address whether any of theH3K9me3 readers expressed in mESCs (all of thosementioned above with the exception of Cbx4 andTdrd7) play a role in SETDB1-dependent silencing, weused recombinase-mediated cassette exchange (RMCE)[87,88] (Figure 5A) to derive novel mESC lines with sin-gle-copy proviral reporters integrated at a specific geno-mic site. Specifically, constructs harbouring the greenfluorescent protein (GFP) gene downstream of theMusD or IAP LTR promoters were generated and intro-duced into the same genomic site in the mESC lineHA36 (a gift from F Lienert and D Schübeler) viaRMCE. In parallel, the MFG-GFP construct [89] derivedfrom the Moloney murine leukaemia virus (MMLV) andefficiently silenced in mESCs and embryonic carcinomacells [20,90-93], and a cytomegalovirus (CMV)-GFP cas-sette were introduced into the same site. Following Cre--ve sortA ES cell line HA36C MFGMusDIAPScramble Setdb1 KDGFPPI% GFP-positive cellsB MFGMusDIAPDays post positive sort60.    3   6   9   12 15 18  21 242.82 55.42.16 57.844.62.6LTR PBS GFPHY+TK-+CreL1L1 1L1LFigure 5 Silencing kinetics and reactivation of ERV reportersintegrated in a specific genomic site. (A) Scheme for targeting ofERV reporters into a specific genomic site in mESCs viarecombinase-mediated cassette exchange (RMCE). The mESC lineHA36 contains a hygromycin B and herpes simplex virus thymidinekinase (HyTK) cassette between inverted Lox sites (L1 and 1L). MFG-green fluorescent protein (GFP), MusD-GFP and IAP-GFP proviralreporter cassettes, which contain the Moloney murine leukaemiavirus, MusD (approximately +130 bp of downstream sequence) andIAP (approximately +450 bp of downstream sequence) LTRs,respectively, flanked by L1 and 1L sites, were cotransfected into theHA36 line with a Cre recombinase expression vector. Negativeselection with ganciclovir eliminated cells with the original HyTKcassette, yielding pools of cells harbouring the proviral reportercassettes predominantly integrated at the same site. (B) The kineticsof silencing of the MFG, MusD and IAP cassettes after reactivation ofthe RMCE pool with siRNA against Setdb1 are shown. (C) GFP-negative cells were sorted at day 12 postinduction with Setdb1siRNA. Robust reactivation of GFP } expression from each of thesepools of cells was observed upon secondary Setdb1 knockdown(KD). Flow cytometry data are presented as contour plots andhistograms of 10,000 viable (propidium iodide (PI)-negative) cells.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 8 of 18mediated recombination and a five-day negative selec-tion with ganciclovir to exclude cells harbouring the ori-ginal hygromycin B-herpes simplex virus thymidinekinase fusion (HyTK) cassette, each of the LTR repor-ters became silenced, while the CMV promoter main-tained expression (data not shown). To select cells thatcontain the ERV-driven GFP gene silenced via theSETDB1 pathway, we transiently transfected the GFP-negative ganciclovir-resistant pools with siRNA specificfor Setdb1. Depending on the cassette, GFP expressionwas induced in 20% to 65% of viable cells and theseGFP-positive cells were isolated by fluorescence-acti-vated cell sorting (FACS). The LTR reporter cassetteswere progressively resilenced over approximately threeweeks in culture (Figure 5B), and the negative popula-tions were sorted at day 12 to be used as reporters.Importantly, GFP was efficiently reactivated in eachpopulation upon subsequent treatment of these poolswith Setdb1 siRNA (Figure 5C), confirming that silen-cing of these LTR reporters is SETDB1-dependent atthis integration site.To determine whether any of the remaining chromo-domain-containing H3K9me3 readers expressed inmESCs are required for SETDB1-mediated silencing, weknocked down Cbx3 (HP1g) as well as Cdyl2, Cbx1,Cbx2, Cbx5, Cbx7 and Mpp8 in the above-describedreporter lines and a previously described pool of mESCsharbouring the silent murine stem cell virus (MSCV)provirus [20]. As expected, treatment of the MFG,MSCV, IAP and MusD reporter lines with Setdb1 andKap1 specific siRNAs induced GFP expression inapproximately 45% and approximately 25% of cells,respectively (Figure 6A, upper panel). In contrast,knockdown (KD) of each of the H3K9me3-binding pro-teins failed to induce GFP expression to the levels seenupon KD of Setdb1 or Kap1, despite efficient depletionof the target mRNAs (Figure 6A, lower panel). KD ofgenes encoding other chromodomain-encoding proteinswith H3K9me-binding properties, such as Cdyl [72] andChd4 [94,95], also did not result in reporter reactivation(Figure S5 in Additional file 1).KD of Cbx3 and Mpp8 did induce GFP expression inabout 10% of treated cells, raising the possibility thatthese H3K9me3 readers act in a redundant manner tomaintain these ERV reporters in a silent state. However,simultaneous KD of Cbx3 in combination with Mpp8(Figure S6 in Additional file 1) or of Cbx3 in combina-tion with Cbx1 and Cbx5 (Cbx1/3/5) (Figure 6A) didnot significantly increase the percentage of GFP-positivecells over that observed with individual KD, despite effi-cient depletion of each mRNA. KD of Cbx1 and Cbx3 inthe Cbx5-/- mESCs and KD of Cbx3 and Cbx5 in theCbx1-/- mESCs showed similar results (data not shown).Thus, none of the assayed chromodomain-encodingproteins with H3K9-binding activity are essential forproviral silencing.The H3K4me2/3 demethylase JARID1C (SMCX,which does not harbour a chromodomain, is capable ofbinding H3K9me3 via its plant homeodomain (PHD)[96], and its yeast homologue, Lid2, interacts directlywith the H3K9 HMTase Clr4 [97]. These interactionssuggest that JARID1C may direct H3K4 demethylationto loci marked by H3K9me3, promoting silencing. How-ever, KD of the Jarid1 genes expressed in mESCs,including Jarid1a, Jarid1b and Jarid1c, either alone or incombination, leads to only modest reactivation of theproviral reporters, indicating that H3K9me3-recognizingH3K4 demethylases are not critical for maintenance ofERV silencing (Figure S7 in Additional file 1). Similarly,KD of Uhrf1 (NP95 in mouse and ICBP90 in human),which was recently shown to bind H3K9me3 via itsPHD or SRA (SET- and RING-associated) domain[79,98-100], and/or KD of a related gene, Uhrf2 yieldsminimal upregulation of the four ERV reporters (FigureS8 in Additional file 1). Based on its pericentric localiza-tion [98], the main function of ICBP90 may lie in repli-cation of heterochromatin and transcriptional regulationof major satellites [101], which show SUV39H1/2-dependent H3K9me3. Taken together, these data revealthat none of the known H3K9me3 readers are essentialfor silencing of ERVs that are repressed by the SETDB1pathway.To determine whether Cbx3 and Mpp8, the H3K9me3readers which showed the highest reactivation of theLTR reporters, are required for silencing of ERVs, weperformed qRT-PCR on cDNA isolated from wt TT2and siRNA-treated mESCs. In Setdb1 KD cells, theMMERVK10C and MusD families showed the highestlevel of derepression, as expected. The same families,however, are only modestly upregulated upon KD ofCbx3 or Mpp8 (Figure 6B, upper panel), despite reduc-tion of the target mRNA to 9% to 30% of wt levels (Fig-ure 6B, lower panel) and dramatic downregulation ofCbx3 at the protein level, as determined by Westernblot analysis (Figure 6C).Finally, to determine whether HP1 proteins act redun-dantly to silence ERVs, we performed simultaneous KDof Cbx1, Cbx3 and Cbx5. Strikingly we observed onlymodest reactivation of each of the ERVs analyzed (Fig-ure 6B, upper panel). Similar levels of upregulation ofMMERVK10C and IAPEz ERVs in the Cbx1-/- mESCs(3.0- and 1.5-fold, respectively) (see Figure 2C), theCbx3 KD (2.3- and 1.8-fold, respectively) and the tripleCbx1/3/5 KD (2.6- and 2.1-fold, respectively) suggestthat Cbx1 and Cbx3 account for most of the HP1-mediated silencing of these ERV families. However,MusD elements, which are upregulated approximatelyfourfold in the triple KD, were not upregulated in anyMaksakova et al. Epigenetics & Chromatin 2011, 4:12 9 of 18of the KOs, suggesting that all three HP1 proteins mustbe depleted to generate the relatively modest level ofderepression observed for this family. Although we can-not rule out the possibility that an insufficient level orduration of HP1 depletion upon KD is responsible forthese negative results, Western blot analysis revealedalmost complete loss of all HP1 proteins in cells simul-taneously depleted of Cbx1, Cbx3 and Cbx5 (Figure 6C).ACB Scram   Setdb1    Kap1     Cbx5     Cbx1     Cbx3   Cbx1/3/5  Cbx2     Cbx7     Cdyl2     Mpp8Setdb1   Kap1     Cbx5      Cbx1     Cbx3   1   3   5     Cbx2     Cbx7    Cdyl2     Mpp8 Scram       Setdb1         Cbx3       Cbx1/3/5       Mpp8Setdb1       Cbx3          1   3   5       Mpp8siRNA% GFP-positive cellsRelative expressionMFGMSCVMusDIAPMusDMMERVK10CIAPEzMLVGLN60. mRNA20. KD% mRNAScram KDCbx1 Cbx5 Cbx3 Cbx1/3/5 Scram Figure 6 Reactivation of ERV reporters and ERVs upon siRNA-mediated KD of H3K9me3-binding proteins. (A) The percentage ofenhanced green fluorescent protein-positive mESCs with reactivated ERV reporters was determined by flow cytometry (upper panel) on day 5after the second transfection with siRNA against specified H3K9me3 readers. At least 10,000 cells were collected for each sample. Data arepresented as means ± SD of three biological replicates. KD efficiency was determined by qRT-PCR at 30 hours after the second siRNAtransfection (lower panel). Data are presented as means ± SD of three technical replicates. (B) Relative expression of ERVs at day 5 after thesecond KD with the indicated siRNA pool (upper panel), along with the efficiency of each KD, as determined by qRT-PCR at 30 hours after thesecond siRNA transfection (lower panel) is presented as means ± SD of three technical replicates. For each amplicon, expression was normalizedto b-actin relative to scramble siRNA KD. (C) Western blot analysis of HP1 proteins in single- and triple-KD cells at day 5 after the second siRNAtransfection is shown. H3 was used as a loading control.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 10 of 18On the basis of the lack of ERV upregulation uponsimultaneous KD of all three HP1 isoforms, we postulatethat redundancy in HP1 function might not be themajor factor preventing broad ERV reactivation. Simi-larly, as the maximum level of ERV reactivation uponKD of the remaining H3K9me3 readers is considerablylower than that observed in Setdb1 KO or Setdb1 KDcells, we conclude that none of these H3K9me3 readersplay a major role in SETDB1-mediated ERV silencing inmESCs.DiscussionRole of H3K9me3 and HP1 in silencing of ERVsWe and others have recently shown that the H3K9me3KMTase SETDB1 is critical for silencing of class I andclass II ERVs in mESCs [20,21]. However, the mechan-ism by which H3K9me3 modification leads to their tran-scriptional repression is currently unclear. In the presentstudy, we have shown that HP1a plays a modest role inmaintaining DNA methylation of IAPEz ERVs, whileHP1b plays a modest role in promoting the spreading ofH4K20me3 into the regions flanking these elements.HP1b also contributes to silencing of select IAPEz andMMERVK10C elements, but has no effect on DNAmethylation of the ERVs analyzed. However, individualdepletion of HP1a, HP1b and all other candidateH3K9me3 readers does not result in upregulation ofERVs or ERV reporters to a level observed in Setdb1KO or Setdb1 KD mESCs, indicating that these factorseither play only a modest role in silencing or act redun-dantly in this process. Strikingly, robust proviral dere-pression was not observed, even after simultaneousdepletion of all three HP1 proteins, ruling out the latter,at least for these readers. Nevertheless, we cannotexclude the possibility that an as yet unidentifiedH3K9me3-binding protein and/or functional redundancybetween H3K9me readers other than HP1 proteins maybe required for H3K9me3-mediated ERV repression.H3K9me3-dependent, H3K9me3 reader-independentproviral silencing?Consistent with our observation that HP1s do not play amajor role in transcriptional silencing of ERVs inmESCs, tethering of HP1 proteins in Drosophila inacti-vates only a limited number of reporter lines [102,103].Indeed, H3K9me3-dependent silencing may occurthrough mechanisms independent of H3K9me3 readers,such as by preventing the binding of transcription fac-tors essential for transcription and/or the recruitment ofthe RNA polymerase II complex itself.Specifically, H3K9me3 may directly or indirectly inhi-bit the deposition of active covalent histone marks.Acetylation of H3 at lysine 9 (H3K9Ac), for example,which is incompatible with methylation at this residue,promotes recruitment of chromatin remodelers andbinding of RNA polymerase II in promoter regions[104-108], and the histone acetyltransferases GCN5 andPCAF, which acetylate H3K9 [109,110], are required forexpression of specific genes [110] and retroviral ele-ments [111]. Furthermore, hyperacetylation of H3 andH4 occurs concomitantly with IAP upregulation inMEFs and early embryos deficient in lymphoid-specifichelicase (LSH) [112], implicating histone acetylation inERV transcriptional activity. Intriguingly, in Xenopusoocytes expressing human H3K9 KMTases and HP1,H3K9me3 mediates transcriptional repression indepen-dently of HP1 recruitment through a mechanism thatinvolves histone deacetylation [59].Alternatively (or in addition), H3K9me3 may blocktranscription by indirectly inhibiting phosphorylation atserine 10 (H3S10ph) in the proviral promoter region.Intriguingly, transcriptional activation of the mousemammary tumour retrovirus is dependent on H3S10phand hyperacetylation of H3, mediated by binding of thenuclear factor 1 (NF-1) transcription factor to the pro-viral LTR [113-115]. Predicted NF-1-binding sites arealso found in the LTRs of other ERVs, including ETnelements [116], implicating a broad role for H3S10ph intranscription of these elements. While experimentsdirectly addressing whether H3K9me3 blocks phosphor-ylation of H3S10ph have not been conducted in mam-malian cells, H3K9me3 severely inhibits H3S10phmediated by the Ipl1/aurora kinase in yeast [117].Finally, H3K4 di- and trimethylation, marks also asso-ciated with the promoter regions of transcriptionallyactive genes, may also be inhibited by the presence ofH3K9me3. Indeed, the H3K4 methyltransferases ASH1L[118] and SET7 [119] are less efficient in depositingH3K4me on histones marked with H3K9me in humancell lines, and H3K9me3 and H3K4me3 are mutuallyexclusive marks in mESCs [18]. While H3K4me2 isdetected in the promoter region of IAP elements in Lsh-/-MEFs concomitant with their upregulation [120], theappearance of such active marks may be a consequenceof, rather than a prerequisite for, transcriptional activa-tion. To directly address the role of H3K4 methylation inretroviral expression, we sought to determine whetherKD of Wdr5, a subunit of MLL/SET1 H3K4 methyltrans-ferase complexes, inhibits Setdb1 KD-induced activationof the ERV reporters. We found that simultaneous KD ofSetdb1 and Wdr5, reduced the level of reactivation of allERV reporters, especially MusD and IAP (Figure S9 inAdditional file 1), indicating that H3K4me3, the catalyticproduct of WDR5-containing complexes [121], is indeedrequired for optimal transcription of ERVs. Thus, thepresence of H3K9me3 may effectively block transcriptionby inhibiting deposition of H3K4me3 and/or the otheractive marks mentioned above.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 11 of 18Heterochromatin spreading into sequences flanking ERVHeterochromatin spreading is thought to involve areiterative process of HP1 proteins binding toH3K9me2/3 [36,37] followed by the recruitment of pro-tein complexes with H3K9me2/3 catalytic activity, suchas SUV39H1/2 [65] and SETDB1 [61]. Consistent withthis model, HP1 proteins have been implicated in het-erochromatin spreading in Drosophila [62,63], yeast [64]and mammals [42,57]. Moreover, H3K9me3, a hallmarkof silent chromatin, is abundant in the vicinity of ERVs[18,19]. However, our results indicate that HP1a andHP1b play only a modest role, if any, in the spreadingof H3K9me3 into the sequences flanking ERVs. In con-trast, HP1b is required for efficient spreading ofH4K20me3 at the IAP ERVs analyzed. Although the bio-logical role of H4K20me3 spreading is still unclear,recent studies have indicated that this covalent mark isinvolved in the maintenance of genomic stability[122-124]. Intriguingly, a role for HP1 in the DNAdamage response independent of H3K9me3 has alsobeen reported [125,126]. The availability of HP1b KOembryos will allow for studies aimed at addressingwhether the distribution of H4K20me3 is dependentupon this protein and whether DNA damage repairpathways are perturbed in vivo.ConclusionsIn this work, we demonstrate the surprising finding thatdespite the accepted function of HP1 proteins inH3K9me-dependent gene silencing and the critical roleof H3K9me3 in transcriptional repression of class I andclass II ERVs, HP1a and HP1b are not required forsilencing of these repetitive elements. Furthermore,while neither HP1a nor HP1b is essential for DNAmethylation or the deposition of H3K9me3 orH4K20me3 within ERVs, HP1b plays a role in thespreading of the latter into sequences flanking these ele-ments. Using a RNAi-based screen with newly derivedmESCs harbouring novel ERV reporters, we have shownthat the remaining proteins reported to bind H3K9me3in vitro, including HP1g, CDYL, CDYL2, CBX2, CBX7,MPP8, UHRF1 and JARID1A-C, are also dispensable forERV silencing. The lack of proviral derepression inthese experiments may be explained by functionalredundancy of these or as yet unidentified H3K9me3readers. Alternatively, H3K9me3 may repress ERV tran-scription via inhibiting deposition of covalent histonemodifications required for transcription. The ERVreporter cell lines generated here should be useful infuture screens of factors predicted to play a role in pro-viral expression. Regardless, additional studies aimed atdelineating the functional significance of H3K9 readers,including nuclear processes not directly related to tran-scription, are clearly warranted.Materials and methodsCell culture, constructs and recombinase-mediatedcassette exchangeTo produce the Cbx5-/- and Cbx1 mESC lines, each Cbxallele was targeted sequentially using two different tar-geting vectors. See Figures S1 and S2 in Additional file1 for a detailed description of the ESC-KO derivation.mESCs were cultured in DMEM supplemented with15% fetal bovine serum (HyClone Laboratories, Logan,UT, USA), 20 mM 4-(2-hydroxyethyl)-1-piperazineetha-nesulfonic acid, 0.1 mM nonessential amino acids, 0.1mM 2-mercaptoethanol, 100 U/mL penicillin, 0.05 mMstreptomycin, leukaemia-inhibitory factor and 2 mM L-glutamine on gelatinized plates. For RMCE into HA36cells, CMV was cut out of the L1-CMV-GFP-1L vector[127] by restriction with ClaI and NheI restrictionenzymes. IAP and MusD LTR, together with the down-stream sequence, were cloned into the resulting ClaI-NheI site upstream of the enhanced green fluorescentprotein (EGFP) gene. MusD from the C57BL/6 genomicDNA on chr8:131270355-131277831 (mm9) was cloned,an element similar in sequence to those commonlyexpressed in wt cells [20]. The NheI site at nt 444 pre-vented us from including a longer fragment. However,this sequence still included the 319 bp 5’-LTR and 125bp immediately downstream of it. A fragment contain-ing a LTR and a downstream sequence, approximately800 bp in total, was cloned for an IAP reporter. The ele-ment chosen was the one at the site of a novel insertioninto the A/WySn mouse strain [128,129] and was clonedfrom the DNA of the respective strain. All inserts wereconfirmed by sequencing. The primer sequences aregiven in Table S1 in Additional file 1.Recombinase-mediated cassette exchange, transfectionand transgene selectionFor targeting of the ERV reporter constructs into the gen-ome, Cre RMCE was used [87,130]. The HA36 mESC linecontains a cassette with the HyTK fusion gene at the ran-dom integration site which allows CMV-GFP expressionfor multiple passages (cell line a gift from F Lienert and DSchübeler). This selectable marker allows for positiveselection through resistance to hygromycin B and fornegative selection through sensitivity to ganciclovir. HA36mESCs were cultured in 25 μg/mL hygromycin B for 14days before transfection to select for cells expressing thefusion gene. Cells were transfected with Lipofectamine2000 (Invitrogen, Carlsbad, CA, USA) in a 24-well plateaccording to the manufacturer’s recommendations. Briefly,1.5 μg of a cassette with a MusD, IAP or MFG insert wascotransfected with 0.5 μg of CMV-Cre plasmid using 2 μLof Lipofectamine 2000 per well. After three days, cellswere transferred to medium containing 3 μM ganciclovirto select against cells still expressing the HyTK fusionMaksakova et al. Epigenetics & Chromatin 2011, 4:12 12 of 18gene. Cells were grown in ganciclovir-containing mediumfor five or more days, with subculturing performed whennecessary.siRNA-mediated knockdownFor reporter assays, 10,000 mESCs per well of a 96-wellplate were seeded into antibiotic-free mESC medium theday before transfection. Transfection was performedaccording to the manufacturer’s protocol using 100 nMconcentrations of each siRNA (siGENOME SMARTpoolreagent Dharmacon, Lafayette, CO, USA) and 0.4 μL ofDharmaFECT 1 siRNA transfection reagent (Dharma-con) per well. On the first day after transfection,approximately 1/5th of the cells were transferred intoanother 96-well plate containing antibiotic-free mESCmedium, and the KD was repeated on the third day.The next day, approximately 1/2 of the cells were trans-ferred into a 24-well or 12-well plate, and flow cytome-try was performed on day 4 or 5 after the second KD.For RNA or protein collection, the first KD was per-formed in a 12-well plate and the cells were transferredto two 6-cm dishes the next day. The day after the sec-ond KD in 6-cm dishes, three-fourths of the cells werecollected for RNA for confirmation of KD efficiency,and the rest were plated onto two 10-cm dishes forexpansion and collection for RNA or protein on day 4after the second KD.Preparation of genomic DNA, bisulphite treatment, T/Acloning and sequencingGenomic DNA was extracted using DNAzol reagent(Invitrogen), and bisulphite conversion of DNA was per-formed using the EZ DNA Methylation Kit (ZymoResearch, Orange, CA, USA) according to the manufac-turer’s protocol. The approximately 370 bp of IAP andapproximately 590 bp of ETnII/MusD element sequencecontaining the LTR and the downstream region wereamplified from converted DNA by PCR using PlatinumTaq (Invitrogen). The primer sequences are given inAdditional file 1, Table S1. PCR products from threeseparate PCRs for each sample were cloned using thepGEM-T Easy Vector System kit (Promega, Madison,WI, USA). All sequences had a conversion rate of >98%. QUMA,with some follow-up processing, was used for analysis ofbisulphite data [131].Native chromatin immunoprecipitation assay andquantitative PCRBriefly, 1 × 107 mESCs } for each cell line were resus-pended in douncing buffer and homogenized through a25-gauge 5/8-inchneedle syringe for 20 repetitions. Aquantity of 1.875 μL of 20 U/μL micrococcal nuclease(MNase; Worthington Biochemical Corp., Lakewood,NJ, USA) was added and incubated at 37°C for 7 min-utes. The reaction was quenched with 0.5 M ethylene-diaminetetraacetic acid and incubated on ice for 5minutes; then 1 mL of hypotonic buffer was added andincubated on ice for 1 hour. Cellular debris was pelleted,and the supernatant was recovered. Protein A/G Sephar-ose beads were blocked with single-stranded salmonsperm DNA and BSA, washed and resuspended inimmunoprecipitation buffer. Blocked protein A/GSepharose beads were added to the digested chromatinfractions and rotated at 4°C for 2 hours to preclearchromatin. A quantity of 100 μL of the precleared chro-matin was purified by phenol-chloroform extraction,and DNA fragment sizes were analyzed and confirmedto correspond to one to three nucleosome fragments.Chromatin was subdivided into aliquots for each immu-noprecipitated sample. Antibodies specific for unmodi-fied H3 (H9289; Sigma-Aldrich, St Louis, MO, USA),H3K9me3 (Active Motif 39161, Carlsbad, CA, USA),H4K20me3 (Active Motif 39180) and control immuno-globulin G (I8140; Sigma-Aldrich, St Louis, MO, USA)were added to each tube and rotated at 4°C for 1 hour.The antibody-protein-DNA complex was precipitated byadding 20 μL of the blocked protein A/G Sepharosebeads and rotated at 4°C overnight. The complex waswashed and eluted, and immunoprecipitated materialwas purified using the QIAquick PCR Purification Kit(Qiagen, Germantown, MD, USA). The purified DNAwas analyzed by qPCR with respect to input using Eva-Green dye (Biotium, Hayward, CA, USA) and MaximaHot Start Taq DNA Polymerase (Fermentas, Vilnius,Lithuania). Primers are listed in Table S1 in Additionalfile 1.RNA isolation, reverse transcription and quantitative RT-PCRRNA was isolated using GenElute™ Mammalian TotalRNA Miniprep Kit (Sigma-Aldrich) and reverse-tran-scribed using SuperScript III Reverse Transcriptase(Invitrogen) as per the manufacturers’ instructions.Quantitative RT-PCR was carried out using SsoFAST™EvaGreen Supermix (Bio-Rad Laboratories, Hercules,CA, USA) on StepOne™ version 2.1 software (AppliedBiosystems, Foster City, CA, USA) in a total volume of20 μL. Data are presented as means ± standard devia-tions of three technical replicates. Primer efficiencieswere around 100%. Dissociation curve analysis was per-formed after the end of the PCR to confirm the pre-sence of a single and specific product.Whole-cell protein extracts and Western blot analysisBriefly, cells were resuspended in 2 ± Laemmli bufferand incubated at 100°C for 10 minutes. Cells were thenhomogenized through a 25-gauge needle syringe 10 toMaksakova et al. Epigenetics & Chromatin 2011, 4:12 13 of 1815 times. Extracts were run on SDS-PAGE gels andtransferred onto a membrane. Primary antibodies usedwere a-HP1a (05-684, 1:200 dilution; Upstate Biotech-nology, Lake Placid, NY, USA), a-HP1b (MCA 1946,1:100 dilution; AbD Serotec, Burlington, ON, Canada)and a-H3 (Active Motif 39163, 1:200 dilution, Carlsbad,CA, USA). Secondary antibodies used at 1:10,000 dilu-tions were IRdye 800CW (926-32210) and IRdye 680(926-32221), both from LI-COR Biosciences (Lincoln,NE, USA). Membranes were analyzed using the OdysseyInfrared Imaging System (LI-COR Biosciences).Northern blot analysisFor each lane, 6 mg of RNA were denatured, electro-phoresed in 1% agarose/3.7% formaldehyde gel in 1 ± 3-(N-morpholino)propanesulfonic acid buffer, transferredovernight onto a Zeta-Probe nylon membrane (Bio-RadLaboratories, Hercules, CA, USA) and baked at 80°C.ETnII/MusD-, IAP- and Gapdh-specific probes weresynthesized by PCR. Primer sequences are given inAdditional file 1, Table S1. An amplified DNA fragmentwas a-32P-labeled using the Random Primers DNALabeling System (Invitrogen). Membranes were prehy-bridized in ExpressHyb hybridization solution (Clontech,Mountain View, CA, USA) for two to four hours at 68°C, hybridized overnight at the same temperature infresh ExpressHyb solution, washed according to themanufacturer’s instructions and exposed to film.Fluorescence-activated cell sorting and analysis ofcassette integrationFACS analysis was performed using BD FACSAria III cellsorter with BD FACSDiva software (BD Biosciences), andflow analysis was performed using a BD LSR II flow cyt-ometer. Viable cells were gated on the basis of propidiumiodide exclusion. At least 10,000 propidium iodide-nega-tive events were analyzed. Untransfected cells were usedas a control for baseline EGFP fluorescence.H3K9me3 profiling of endogenous retrovirusesTo determine the H3K9me3 status of ERVs in TT2 wtversus Setdb1 KO mESC lines, we generated averageH3K9me3 profiles for representative ERVs upregulatedin the latter [19], including MusD, MMERVK10C,IAPEz, MLV and GLN elements. For each ERV family,all sequenced 50-bp reads from our previously publishedTT2 and Setdb1 KO H3K9me3 native ChIP-seq datasets [19] were aligned to the corresponding consensussequences (including internal regions and correspondingLTRs) from Repbase[132] for all ERVs except MusD. For MusD, a represen-tative active element was used (139824 to 132348 nt ofAC084696, reverse strand). The Burrows-WheelerAligner[133] wasemployed with default parameters (allowing up to twomismatches in the 32-bp seed and one gap). Reads weredirectionally extended by 150 bp, and extended readswere profiled along the element. All mapped reads weretaken into account, and the profiles for each librarywere normalized by the total number of reads uniquelymapped to the mm9 genome. For reads that werealigned into multiple locations (LTRs), we consideredonly one randomly selected alignment location. Theirregular nature of the profile is most likely attributableto SNPs and insertions and/or deletions in the consen-sus vs. genomic reads.H3K9me3 profiling in the sequences flanking endogenousretrovirusesTo compare the average density of H3K9me3 in thegenomic regions flanking ERVs, H3K9me3 N-ChIP-seqdata sets for TT2 wt and Setdb1 KO mESCs [19] wereused. Intact elements were selected for three ERVfamilies: MusD, IAPEz, and MLV. For MusD, IAPEzand MLV, 195, 599 and 51 elements, respectively, satis-fied the length and sequence similarity criteria that weapplied [19]. All H3K9me3 reads that passed the qualityscore threshold above 7 were aligned to the mouse gen-ome (mm9) using the Burrows-Wheeler Aligner [133]and directionally extended by 150 bp [19]. Only readsuniquely aligned to the regions within 7 kb on eitherside of intact elements were considered. If multiplereads were mapped to the same location, only one copyof the read was counted. To generate the profiles shownfor the TT2 wt and Setdb1 KO cell lines, extended readswere first agglomerated for 5’- and 3’-flanks. Subse-quently, the data were normalized to the total numberof included elements and weighted by the total numberof aligned reads to the genome for each sample.Note added in proofWhile this manuscript was under review, an article pub-lished by Shang and colleagues (PNAS 2011, 108(18):7541-7546) revealed that the H3K9me3 demethylaseJMJD2B greatly facilitates H3K4 methylation by purifiedMLL2 in vitro (demonstrating that H3K9 demethylationis required for efficient H3K4 methylation) and isrequired for transcription of MLL2 targets in vivo.Additional materialAdditional file 1: Figure S1. Derivation of Cbx5-/- mESCs via sequentialtargeted disruption of the Cbx5 gene. Figure S2. Derivation of Cbx1-/-mESCs via sequential targeted disruption of the Cbx1 gene. Figure S3.Profiling of trimethylated lysine 9 of histone 3 (H3K9me3) along thelength of endogenous retroviruses (ERVs). Figure S4. Profiling ofH3K9me3 and H4K20me3 in the sequence flanking ERVs in wild-type andSetdb1-knockout mESCs. Figure S5. Knockdown (KD) of Cdyl, Cdyl2, Chd4or Mpp8 does not result in reactivation of proviral reporters. Figure S6.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 14 of 18Simultaneous KD of Mpp8 and Cbx3 does not result in reactivation of theERV reporters. Figure S7. Proviral reporters are modestly reactivatedupon KD of H3K9me3-binding H3K4 demethylases Jarid1a-c. Figure S8.Proviral reporters are modestly reactivated upon KD of H3K9me3-bindingSRA (SET- and RING-associated) domain proteins Uhrf1 and Uhrf2. FigureS9. The level of derepression of the ERV reporters is substantially reducedin the Setdb1-KD cells following KD of the H3K4 methyltransferase Wdr5.Table S1. Primers used in the study.AbbreviationsBSA: bovine serum albumin; DMEM: Dulbecco’s modified Eagle’s medium;qPCR: quantitative polymerase chain reaction; RT: reverse transcriptase;siRNA: small interfering RNA; SNP: single-nucleotide polymorphism.AcknowledgementsWe thank Danny Leung, Sandra Lee, Lucia Lam and the University of BritishColumbia flow cytometry facility for technical support. We also thank FlorianLienert and Dirk Schübeler for providing the HA36 ES cell line, En Li forproviding the Dnmt1-KO line, Yoichi Shinkai for providing the Setdb1 KO lineand Mark Bedford for helpful suggestions. This work was supported by CIHRgrant 77805 (to ML) and CIHR grant 92090 (to ML and DM). This work wasalso supported by Biotechnology and Biological Sciences Research Councilcore strategic grants and Deutsche Forschungsgemeinschaft grant SI 1209/2-1 (to PS). ML is a Scholar of the MSFHR and a CIHR New Investigator.Author details1Department of Medical Genetics, Life Sciences Institute, University of BritishColumbia, 2350 Health Sciences Mall, Vancouver, BC, Canada, V6T 1Z3. 2TerryFox Laboratory, BC Cancer Agency, 675 West 10th Avenue, Vancouver, BC,Canada, V5Z 1L3. 3Division of Immunoepigenetics, Department ofImmunology and Cell Biology, Research Center Borstel, Parkallee 22, D-23845Borstel, Germany. 4Canada’s Michael Smith Genome Sciences Centre, BCCancer Agency, 675 West 10th Avenue, Vancouver, BC, V5Z 1L3, Canada.Authors’ contributionsIM carried out most of the research. PG performed ChIP-qPCR. JB and JPBderived the KO mESCs. MB performed bioinformatics analysis. DM and PScontributed reagents. IM and ML designed the study, analyzed the data andwrote the manuscript. All authors read and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Received: 21 April 2011 Accepted: 20 July 2011 Published: 20 July 2011References1. International Mouse Genome Sequencing Consortium: Initial sequencingand comparative analysis of the mouse genome. Nature 2002,420:520-562.2. International Human Genome Sequencing Consortium: Initial sequencingand analysis of the human genome. Nature 2001, 409:860-921.3. Stocking C, Kozak C: Endogenous retroviruses. Cell Mol Life Sci 2008,65:3383-3398.4. Gifford R, Kabat P, Martin J, Lynch C, Tristem M: Evolution and distributionof class II-related endogenous retroviruses. J Virol 2005, 79:6478-6486.5. Gimenez J, Montgiraud C, Pichon JP, Bonnaud B, Arsac M, Ruel K, Bouton O,Mallet F: Custom human endogenous retroviruses dedicated microarrayidentifies self-induced HERV-W family elements reactivated in testicularcancer upon methylation control. Nucleic Acids Res 2010, 38:2229-2246.6. Lamprecht B, Walter K, Kreher S, Kumar R, Hummel M, Lenze D, Köchert K,Bouhlel MA, Richter J, Soler E, Stadhouders R, Jöhrens K, Wurster KD,Callen DF, Harte MF, Giefing M, Barlow R, Stein H, Anagnostopoulos I,Janz M, Cockerill PN, Siebert R, Dörken B, Bonifer C, Mathas S: Derepressionof an endogenous long terminal repeat activates the CSF1R proto-oncogene in human lymphoma. Nat Med 2010, 16:571-579.7. Moyes D, Griffiths DJ, Venables PJ: Insertional polymorphisms: a new leaseof life for endogenous retroviruses in human disease. Trends Genet 2007,23:326-333.8. McLaughlin-Drubin ME, Munger K: Viruses associated with human cancer.Biochim Biophys Acta 2008, 1782:127-150.9. Howard G, Eiges R, Gaudet F, Jaenisch R, Eden A: Activation andtransposition of endogenous retroviral elements in hypomethylationinduced tumors in mice. Oncogene 2008, 27:404-408.10. Lee JS, Haruna T, Ishimoto A, Honjo T, Yanagawa S: Intracisternal type Aparticle-mediated activation of the Notch4/int3 gene in a mousemammary tumor: generation of truncated Notch4/int3 mRNAs byretroviral splicing events. J Virol 1999, 73:5166-5171.11. Romanish MT, Cohen CJ, Mager DL: Potential mechanisms of endogenousretroviral-mediated genomic instability in human cancer. Semin CancerBiol 2010, 20:246-253.12. Puech A, Dupressoir A, Loireau MP, Mattei MG, Heidmann T:Characterization of two age-induced intracisternal A-particle-relatedtranscripts in the mouse liver: transcriptional read-through into an openreading frame with similarities to the yeast ccr4 transcription factor. JBiol Chem 1997, 272:5995-6003.13. Barbot W, Dupressoir A, Lazar V, Heidmann T: Epigenetic regulation of anIAP retrotransposon in the aging mouse: progressive demethylation andde-silencing of the element by its repetitive induction. Nucleic Acids Res2002, 30:2365-2373.14. Goff SP: Retrovirus restriction factors. Mol Cell 2004, 16:849-859.15. Yoder JA, Walsh CP, Bestor TH: Cytosine methylation and the ecology ofintragenomic parasites. Trends Genet 1997, 13:335-340.16. Walsh CP, Chaillet JR, Bestor TH: Transcription of IAP endogenousretroviruses is constrained by cytosine methylation. Nat Genet 1998,20:116-117.17. Martens JH, O’Sullivan RJ, Braunschweig U, Opravil S, Radolf M, Steinlein P,Jenuwein T: The profile of repeat-associated histone lysine methylationstates in the mouse epigenome. EMBO J 2005, 24:800-812.18. Mikkelsen TS, Ku M, Jaffe DB, Issac B, Lieberman E, Giannoukos G, Alvarez P,Brockman W, Kim TK, Koche RP, Lee W, Mendenhall E, O’Donovan A,Presser A, Russ C, Xie X, Meissner A, Wernig M, Jaenisch R, Nusbaum C,Lander ES, Bernstein BE: Genome-wide maps of chromatin state inpluripotent and lineage-committed cells. Nature 2007, 448:553-560.19. Karimi MM, Goyal P, Maksakova IA, Bilenky M, Leung D, Tang JX, Shinkai Y,Mager DL, Jones S, Hirst M, Lorincz MC: DNA methylation and SETDB1/H3K9me3 regulate predominantly distinct sets of genes, retroelements,and chimeric transcripts in mESCs. Cell Stem Cell 2011, 8:676-687.20. Matsui T, Leung D, Miyashita H, Maksakova IA, Miyachi H, Kimura H,Tachibana M, Lorincz MC, Shinkai Y: Proviral silencing in embryonic stemcells requires the histone methyltransferase ESET. Nature 2010,464:927-931.21. Rowe HM, Jakobsson J, Mesnard D, Rougemont J, Reynard S, Aktas T,Maillard PV, Layard-Liesching H, Verp S, Marquis J, Spitz F, Constam DB,Trono D: KAP1 controls endogenous retroviruses in embryonic stemcells. Nature 2010, 463:237-240.22. Wolf D, Goff SP: Embryonic stem cells use ZFP809 to silence retroviralDNAs. Nature 2009, 458:1201-1204.23. Hutnick LK, Huang X, Loo TC, Ma Z, Fan G: Repression of retrotransposalelements in mouse embryonic stem cells is primarily mediated by aDNA methylation-independent mechanism. J Biol Chem 2010,285:21082-21091.24. Maksakova IA, Romanish MT, Gagnier L, Dunn CA, van de Lagemaat LN,Mager DL: Retroviral elements and their hosts: insertional mutagenesisin the mouse germ line. PLoS Genet 2006, 2:e2.25. Macfarlan TS, Gifford WD, Agarwal S, Driscoll S, Lettieri K, Wang J,Andrews SE, Franco L, Rosenfeld MG, Ren B, Pfaff SL: Endogenousretroviruses and neighboring genes are coordinately repressed by LSD1/KDM1A. Genes Dev 2011, 25:594-607.26. Tse C, Sera T, Wolffe AP, Hansen JC: Disruption of higher-order folding bycore histone acetylation dramatically enhances transcription ofnucleosomal arrays by RNA polymerase III. Mol Cell Biol 1998,18:4629-4638.27. Krajewski WA: Histone hyperacetylation facilitates chromatin remodellingin a Drosophila embryo cell-free system. Mol Gen Genet 2000, 263:38-47.28. Strahl BD, Allis CD: The language of covalent histone modifications.Nature 2000, 403:41-45.29. Jenuwein T, Allis CD: Translating the histone code. Science 2001,293:1074-1080.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 15 of 1830. Wysocka J: Identifying novel proteins recognizing histone modificationsusing peptide pull-down assay. Methods 2006, 40:339-343.31. Daniel JA, Pray-Grant MG, Grant PA: Effector proteins for methylatedhistones: an expanding family. Cell Cycle 2005, 4:919-926.32. Zeng W, Ball AR Jr, Yokomori K: HP1: heterochromatin binding proteinsworking the genome. Epigenetics 2010, 5:287-292.33. Vermaak D, Malik HS: Multiple roles for heterochromatin protein 1 genesin Drosophila. Annu Rev Genet 2009, 43:467-492.34. Kwon SH, Workman JL: The changing faces of HP1: fromheterochromatin formation and gene silencing to euchromatic geneexpression. HP1 acts as a positive regulator of transcription. Bioessays2011, 33:280-289.35. Singh PB, Georgatos SD: HP1: facts, open questions, and speculation. JStruct Biol 2002, 140:10-16.36. Bannister AJ, Zegerman P, Partridge JF, Miska EA, Thomas JO, Allshire RC,Kouzarides T: Selective recognition of methylated lysine 9 on histone H3by the HP1 chromo domain. Nature 2001, 410:120-124.37. Lachner M, O’Carroll D, Rea S, Mechtler K, Jenuwein T: Methylation ofhistone H3 lysine 9 creates a binding site for HP1 proteins. Nature 2001,410:116-120.38. Nielsen AL, Oulad-Abdelghani M, Ortiz JA, Remboutsika E, Chambon P,Losson R: Heterochromatin formation in mammalian cells: interactionbetween histones and HP1 proteins. Mol Cell 2001, 7:729-739.39. Thiru A, Nietlispach D, Mott HR, Okuwaki M, Lyon D, Nielsen PR,Hirshberg M, Verreault A, Murzina NV, Laue ED: Structural basis of HP1/PXVXL motif peptide interactions and HP1 localisation toheterochromatin. EMBO J 2004, 23:489-499.40. Singh PB: HP1 proteins: what is the essential interaction? Genetika 2010,46:1424-1429.41. Rountree MR, Selker EU: DNA methylation and the formation ofheterochromatin in Neurospora crassa. Heredity 2010, 105:38-44.42. Groner AC, Meylan S, Ciuffi A, Zangger N, Ambrosini G, Dénervaud N,Bucher P, Trono D: KRAB-zinc finger proteins and KAP1 can mediatelong-range transcriptional repression through heterochromatinspreading. PLoS Genet 2010, 6:e1000869.43. Sadaie M, Kawaguchi R, Ohtani Y, Arisaka F, Tanaka K, Shirahige K,Nakayama J: Balance between distinct HP1 family proteins controlsheterochromatin assembly in fission yeast. Mol Cell Biol 2008,28:6973-6988.44. Kourmouli N, Sun YM, van der Sar S, Singh PB, Brown JP: Epigeneticregulation of mammalian pericentric heterochromatin in vivo by HP1.Biochem Biophys Res Commun 2005, 337:901-907.45. Nielsen SJ, Schneider R, Bauer UM, Bannister AJ, Morrison A, O’Carroll D,Firestein R, Cleary M, Jenuwein T, Herrera RE, Kouzarides T: Rb targetshistone H3 methylation and HP1 to promoters. Nature 2001,412:561-565.46. Kwon SH, Florens L, Swanson SK, Washburn MP, Abmayr SM, Workman JL:Heterochromatin protein 1 (HP1) connects the FACT histone chaperonecomplex to the phosphorylated CTD of RNA polymerase II. Genes Dev2010, 24:2133-2145.47. Klattenhoff C, Xi H, Li C, Lee S, Xu J, Khurana JS, Zhang F, Schultz N,Koppetsch BS, Nowosielska A, Seitz H, Zamore PD, Weng Z, Theurkauf WE:The Drosophila HP1 homologue Rhino is required for transposonsilencing and piRNA production by dual-strand clusters. Cell 2009,138:1137-1149.48. Freitag M, Hickey PC, Khlafallah TK, Read ND, Selker EU: HP1 is essential forDNA methylation in Neurospora. Mol Cell 2004, 13:427-434.49. Tamaru H, Selker EU: A histone H3 methyltransferase controls DNAmethylation in Neurospora crassa. Nature 2001, 414:277-283.50. Johnson LM, Cao X, Jacobsen SE: Interplay between two epigeneticmarks: DNA methylation and histone H3 lysine 9 methylation. Curr Biol2002, 12:1360-1367.51. Jackson JP, Lindroth AM, Cao X, Jacobsen SE: Control of CpNpG DNAmethylation by the KRYPTONITE histone H3 methyltransferase. Nature2002, 416:556-560.52. du Chéné I, Basyuk E, Lin YL, Triboulet R, Knezevich A, Chable-Bessia C,Mettling C, Baillat V, Reynes J, Corbeau P, Bertrand E, Marcello A, Emiliani S,Kiernan R, Benkirane M: Suv39H1 and HP1γ are responsible forchromatin-mediated HIV-1 transcriptional silencing and post-integrationlatency. EMBO J 2007, 26:424-435.53. Brown JP, Bullwinkel J, Baron-Lühr B, Billur M, Schneider P, Winking H,Singh PB: HP1γ function is required for male germ cell survival andspermatogenesis. Epigenetics Chromatin 2010, 3:9.54. Vakoc CR, Mandat SA, Olenchock BA, Blobel GA: Histone H3 lysine 9methylation and HP1γ are associated with transcription elongationthrough mammalian chromatin. Mol Cell 2005, 19:381-391.55. Mateescu B, Bourachot B, Rachez C, Ogryzko V, Muchardt C: Regulation ofan inducible promoter by an HP1β-HP1γ switch. EMBO Rep 2008,9:267-272.56. Wolf D, Cammas F, Losson R, Goff SP: Primer binding site-dependentrestriction of murine leukemia virus requires HP1 binding by TRIM28. JVirol 2008, 82:4675-4679.57. Sripathy SP, Stevens J, Schultz DC: The KAP1 corepressor functions tocoordinate the assembly of de novo HP1-demarcatedmicroenvironments of heterochromatin required for KRAB zinc fingerprotein-mediated transcriptional repression. Mol Cell Biol 2006,26:8623-8638.58. Schultz DC, Ayyanathan K, Negorev D, Maul GG, Rauscher FJ: SETDB1: anovel KAP-1-associated histone H3, lysine 9-specific methyltransferasethat contributes to HP1-mediated silencing of euchromatic genes byKRAB zinc-finger proteins. Genes Dev 2002, 16:919-932.59. Stewart MD, Li J, Wong J: Relationship between histone H3 lysine 9methylation, transcription repression, and heterochromatin protein 1recruitment. Mol Cell Biol 2005, 25:2525-2538.60. Nielsen AL, Ortiz JA, You J, Oulad-Abdelghani M, Khechumian R,Gansmuller A, Chambon P, Losson R: Interaction with members of theheterochromatin protein 1 (HP1) family and histone deacetylation aredifferentially involved in transcriptional silencing by members of theTIF1 family. EMBO J 1999, 18:6385-6395.61. Ryan RF, Schultz DC, Ayyanathan K, Singh PB, Friedman JR, Fredericks WJ,Rauscher FJ: KAP-1 corepressor protein interacts and colocalizes withheterochromatic and euchromatic HP1 proteins: a potential role forKrüppel-associated box-zinc finger proteins in heterochromatin-mediated gene silencing. Mol Cell Biol 1999, 19:4366-4378.62. Danzer JR, Wallrath LL: Mechanisms of HP1-mediated gene silencing inDrosophila. Development 2004, 131:3571-3580.63. Hines KA, Cryderman DE, Flannery KM, Yang H, Vitalini MW, Hazelrigg T,Mizzen CA, Wallrath LL: Domains of heterochromatin protein 1 requiredfor Drosophila melanogaster heterochromatin spreading. Genetics 2009,182:967-977.64. Partridge JF, Borgstrom B, Allshire RC: Distinct protein interaction domainsand protein spreading in a complex centromere. Genes Dev 2000,14:783-791.65. Aagaard L, Laible G, Selenko P, Schmid M, Dorn R, Schotta G, Kuhfittig S,Wolf A, Lebersorger A, Singh PB, Reuter G, Jenuwein T: Functionalmammalian homologues of the Drosophila PEV-modifier Su(var)3-9encode centromere-associated proteins which complex with theheterochromatin component M31. EMBO J 1999, 18:1923-1938.66. Li Y, Kirschmann DA, Wallrath LL: Does heterochromatin protein 1 alwaysfollow code? Proc Natl Acad Sci USA 2002, 99(Suppl 4):16462-16469.67. Locke SM, Martienssen RA: Slicing and spreading of heterochromaticsilencing by RNA interference. Cold Spring Harb Symp Quant Biol 2006,71:497-503.68. Talbert PB, Henikoff S: Spreading of silent chromatin: inaction at adistance. Nat Rev Genet 2006, 7:793-803.69. Honda S, Lewis ZA, Huarte M, Cho LY, David LL, Shi Y, Selker EU: The DMMcomplex prevents spreading of DNA methylation from transposons tonearby genes in Neurospora crassa. Genes Dev 2010, 24:443-454.70. Tajul-Arifin K, Teasdale R, Ravasi T, Hume DA, Mattick JS: Identification andanalysis of chromodomain-containing proteins encoded in the mousetranscriptome. Genome Res 2003, 13:1416-1429.71. Kokura K, Sun L, Bedford MT, Fang J: Methyl-H3K9-binding protein MPP8mediates E-cadherin gene silencing and promotes tumour cell motilityand invasion. EMBO J 2010, 29:3673-3687.72. Bua DJ, Kuo AJ, Cheung P, Liu CL, Migliori V, Espejo A, Casadio F, Bassi C,Amati B, Bedford MT, Guccione E, Gozani O: Epigenome microarrayplatform for proteome-wide dissection of chromatin-signaling networks.PLoS One 2009, 4:e6789.73. Liu H, Galka M, Iberg A, Wang Z, Li L, Voss C, Jiang X, Lajoie G, Huang Z,Bedford MT, Li SS: Systematic identification of methyllysine-drivenMaksakova et al. Epigenetics & Chromatin 2011, 4:12 16 of 18interactions for histone and nonhistone targets. J Proteome Res 2010,9:5827-5836.74. Fischle W, Franz H, Jacobs SA, Allis CD, Khorasanizadeh S: Specificity of thechromodomain Y chromosome family of chromodomains for lysine-methylated ARK(S/T) motifs. J Biol Chem 2008, 283:19626-19635.75. Kim J, Daniel J, Espejo A, Lake A, Krishna M, Xia L, Zhang Y, Bedford MT:Tudor, MBT and chromo domains gauge the degree of lysinemethylation. EMBO Rep 2006, 7:397-403.76. Mulligan P, Westbrook TF, Ottinger M, Pavlova N, Chang B, Macia E, Shi YJ,Barretina J, Liu J, Howley PM, Elledge SJ, Shi Y: CDYL bridges REST andhistone methyltransferases for gene repression and suppression ofcellular transformation. Mol Cell 2008, 32:718-726.77. Bernstein E, Duncan EM, Masui O, Gil J, Heard E, Allis CD: Mouse polycombproteins bind differentially to methylated histone H3 and RNA and areenriched in facultative heterochromatin. Mol Cell Biol 2006, 26:2560-2569.78. Quinn AM, Bedford MT, Espejo A, Spannhoff A, Austin CP, Oppermann U,Simeonov A: A homogeneous method for investigation of methylation-dependent protein-protein interactions in epigenetics. Nucleic Acids Res2009, 38:e11.79. Rottach A, Frauer C, Pichler G, Bonapace IM, Spada F, Leonhardt H: Themulti-domain protein Np95 connects DNA methylation and histonemodification. Nucleic Acids Res 2010, 38:1796-1804.80. Bernard D, Martinez-Leal JF, Rizzo S, Martinez D, Hudson D, Visakorpi T,Peters G, Carnero A, Beach D, Gil J: CBX7 controls the growth of normaland tumor-derived prostate cells by repressing the Ink4a/Arf locus.Oncogene 2005, 24:5543-5551.81. Aucott R, Bullwinkel J, Yu Y, Shi W, Billur M, Brown JP, Menzel U, Kioussis D,Wang G, Reisert I, Weimer J, Pandita RK, Sharma GG, Pandita TK, Fundele R,Singh PB: HP1-β is required for development of the cerebral neocortexand neuromuscular junctions. J Cell Biol 2008, 183:597-606.82. Schotta G, Lachner M, Sarma K, Ebert A, Sengupta R, Reuter G, Reinberg D,Jenuwein T: A silencing pathway to induce H3-K9 and H4-K20trimethylation at constitutive heterochromatin. Genes Dev 2004,18:1251-1262.83. Dong KB, Maksakova IA, Mohn F, Leung D, Appanah R, Lee S, Yang HW,Lam LL, Mager DL, Schübeler D, Tachibana M, Shinkai Y, Lorincz MC: DNAmethylation in ES cells requires the lysine methyltransferase G9a butnot its catalytic activity. EMBO J 2008, 27:2691-2701.84. Lehnertz B, Ueda Y, Derijck AA, Braunschweig U, Perez-Burgos L, Kubicek S,Chen T, Li E, Jenuwein T, Peters AH: Suv39h-mediated histone H3 lysine 9methylation directs DNA methylation to major satellite repeats atpericentric heterochromatin. Curr Biol 2003, 13:1192-1200.85. Honda S, Selker EU: Direct interaction between DNA methyltransferaseDIM-2 and HP1 is required for DNA methylation in Neurospora crassa.Mol Cell Biol 2008, 28:6044-6055.86. Maksakova IA, Zhang Y, Mager DL: Preferential epigenetic suppression ofthe autonomous MusD over the nonautonomous ETn mouseretrotransposons. Mol Cell Biol 2009, 29:2456-2468.87. Feng YQ, Seibler J, Alami R, Eisen A, Westerman KA, Leboulch P, Fiering S,Bouhassira EE: Site-specific chromosomal integration in mammalian cells:highly efficient CRE recombinase-mediated cassette exchange. J Mol Biol1999, 292:779-785.88. Schübeler D, Lorincz MC, Groudine M: Targeting silence: the use of site-specific recombination to introduce in vitro methylated DNA into thegenome. Sci STKE 2001, 2001(83):l1.89. Lorincz MC, Schübeler D, Groudine M: Methylation-mediated proviralsilencing is associated with MeCP2 recruitment and localized histone H3deacetylation. Mol Cell Biol 2001, 21:7913-7922.90. Teich NM, Weiss RA, Martin GR, Lowy DR: Virus infection of murineteratocarcinoma stem cell lines. Cell 1977, 12:973-982.91. Pannell D, Osborne CS, Yao S, Sukonnik T, Pasceri P, Karaiskakis A, Okano M,Li E, Lipshitz HD, Ellis J: Retrovirus vector silencing is de novo methylaseindependent and marked by a repressive histone code. EMBO J 2000,19:5884-5894.92. Niwa O, Yokota Y, Ishida H, Sugahara T: Independent mechanismsinvolved in suppression of the Moloney leukemia virus genome duringdifferentiation of murine teratocarcinoma cells. Cell 1983, 32:1105-1113.93. Cheng L, Du C, Murray D, Tong X, Zhang YA, Chen BP, Hawley RG: A GFPreporter system to assess gene transfer and expression in humanhematopoietic progenitor cells. Gene Ther 1997, 4:1013-1022.94. Musselman CA, Mansfield RE, Garske AL, Davrazou F, Kwan AH, Oliver SS,O’Leary H, Denu JM, Mackay JP, Kutateladze TG: Binding of the CHD4PHD2 finger to histone H3 is modulated by covalent modifications.Biochem J 2009, 423:179-187.95. Mansfield RE, Musselman CA, Kwan AH, Oliver SS, Garske AL, Davrazou F,Denu JM, Kutateladze TG, Mackay JP: The plant homeodomain (PHD)fingers of CHD4 are histone H3-binding modules with preference forunmodified H3K4 and methylated H3K9. J Biol Chem 2011,286:11779-11791.96. Iwase S, Lan F, Bayliss P, de la Torre-Ubieta L, Huarte M, Qi HH,Whetstine JR, Bonni A, Roberts TM, Shi Y: The X-linked mental retardationgene SMCX/JARID1C defines a family of histone H3 lysine 4demethylases. Cell 2007, 128:1077-1088.97. Li F, Huarte M, Zaratiegui M, Vaughn MW, Shi Y, Martienssen R, Cande WZ:Lid2 is required for coordinating H3K4 and H3K9 methylation ofheterochromatin and euchromatin. Cell 2008, 135:272-283.98. Karagianni P, Amazit L, Qin J, Wong J: ICBP90, a novel methyl K9 H3binding protein linking protein ubiquitination with heterochromatinformation. Mol Cell Biol 2008, 28:705-717.99. Bartke T, Vermeulen M, Xhemalce B, Robson SC, Mann M, Kouzarides T:Nucleosome-interacting proteins regulated by DNA and histonemethylation. Cell 2010, 143:470-484.100. Nady N, Lemak A, Walker JR, Avvakumov GV, Kareta MS, Achour M, Xue S,Duan S, Allali-Hassani A, Zuo X, Wang YX, Bronner C, Chédin F,Arrowsmith CH, Dhe-Paganon S: Recognition of multivalent histone statesassociated with heterochromatin by UHRF1 protein. J Biol Chem 2011,286:24300-24311.101. Papait R, Pistore C, Negri D, Pecoraro D, Cantarini L, Bonapace IM: Np95 isimplicated in pericentromeric heterochromatin replication and in majorsatellite silencing. Mol Biol Cell 2007, 18:1098-1106.102. Seum C, Delattre M, Spierer A, Spierer P: Ectopic HP1 promoteschromosome loops and variegated silencing in Drosophila. EMBO J 2001,20:812-818.103. Seum C, Spierer A, Delattre M, Pauli D, Spierer P: A GAL4-HP1 fusionprotein targeted near heterochromatin promotes gene silencing.Chromosoma 2000, 109:453-459.104. Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, Thanos D: Orderedrecruitment of chromatin modifying and general transcription factors tothe IFN-β promoter. Cell 2000, 103:667-678.105. Kasten M, Szerlong H, Erdjument-Bromage H, Tempst P, Werner M,Cairns BR: Tandem bromodomains in the chromatin remodeler RSCrecognize acetylated histone H3 Lys14. EMBO J 2004, 23:1348-1359.106. Vicent GP, Zaurin R, Nacht AS, Li A, Font-Mateu J, Le Dily F, Vermeulen M,Mann M, Beato M: Two chromatin remodeling activities cooperate duringactivation of hormone responsive promoters. PLoS Genet 2009, 5:e1000567.107. Hassan AH, Neely KE, Workman JL: Histone acetyltransferase complexesstabilize SWI/SNF binding to promoter nucleosomes. Cell 2001,104:817-827.108. Wang Z, Zang C, Cui K, Schones DE, Barski A, Peng W, Zhao K: Genome-wide mapping of HATs and HDACs reveals distinct functions in activeand inactive genes. Cell 2009, 138:1019-1031.109. Jin Q, Yu LR, Wang L, Zhang Z, Kasper LH, Lee JE, Wang C, Brindle PK,Dent SY, Ge K: Distinct roles of GCN5/PCAF-mediated H3K9ac and CBP/p300-mediated H3K18/27ac in nuclear receptor transactivation. EMBO J2011, 30:249-262.110. Nagy Z, Riss A, Fujiyama S, Krebs A, Orpinell M, Jansen P, Cohen A,Stunnenberg HG, Kato S, Tora L: The metazoan ATAC and SAGAcoactivator HAT complexes regulate different sets of inducible targetgenes. Cell Mol Life Sci 2010, 67:611-628.111. Nagy Z, Tora L: Distinct GCN5/PCAF-containing complexes function asco-activators and are involved in transcription factor and global histoneacetylation. Oncogene 2007, 26:5341-5357.112. Huang J, Fan T, Yan Q, Zhu H, Fox S, Issaq HJ, Best L, Gangi L, Munroe D,Muegge K: Lsh, an epigenetic guardian of repetitive elements. NucleicAcids Res 2004, 32:5019-5028.113. Vicent GP, Ballaré C, Nacht AS, Clausell J, Subtil-Rodríguez A, Quiles I,Jordan A, Beato M: Induction of progesterone target genes requiresactivation of Erk and Msk kinases and phosphorylation of histone H3.Mol Cell 2006, 24:367-381.Maksakova et al. Epigenetics & Chromatin 2011, 4:12 17 of 18114. Hebbar PB, Archer TK: Nuclear factor 1 is required for both hormone-dependent chromatin remodeling and transcriptional activation of themouse mammary tumor virus promoter. Mol Cell Biol 2003, 23:887-898.115. Vicent GP, Zaurin R, Nacht AS, Font-Mateu J, Le Dily F, Beato M: Nuclearfactor 1 synergizes with progesterone receptor on the mouse mammarytumor virus promoter wrapped around a histone H3/H4 tetramer byfacilitating access to the central hormone-responsive elements. J BiolChem 2010, 285:2622-2631.116. Baust C, Gagnier L, Baillie GJ, Harris MJ, Juriloff DM, Mager DL: Structureand expression of mobile ETnII retroelements and their coding-competent MusD relatives in the mouse. J Virol 2003, 77:11448-11458.117. Rea S, Eisenhaber F, O’Carroll D, Strahl BD, Sun ZW, Schmid M, Opravil S,Mechtler K, Ponting CP, Allis CD, Jenuwein T: Regulation of chromatinstructure by site-specific histone H3 methyltransferases. Nature 2000,406:593-599.118. Gregory GD, Vakoc CR, Rozovskaia T, Zheng X, Patel S, Nakamura T,Canaani E, Blobel GA: Mammalian ASH1L is a histone methyltransferasethat occupies the transcribed region of active genes. Mol Cellular Biol2007, 27:8466-8479.119. Wang H, Cao R, Xia L, Erdjument-Bromage H, Borchers C, Tempst P,Zhang Y: Purification and functional characterization of a histone H3-lysine 4-specific methyltransferase. Mol Cell 2001, 8:1207-1217.120. Yan Q, Huang J, Fan T, Zhu H, Muegge K: Lsh, a modulator of CpGmethylation, is crucial for normal histone methylation. EMBO J 2003,22:5154-5162.121. Wysocka J, Swigut T, Milne TA, Dou Y, Zhang X, Burlingame AL, Roeder RG,Brivanlou AH, Allis CD: WDR5 associates with histone H3 methylated atK4 and is essential for H3 K4 methylation and vertebrate development.Cell 2005, 121:859-872.122. Balakrishnan L, Milavetz B: Decoding the histone H4 lysine 20 methylationmark. Crit Rev Biochem Mol Biol 2010, 45:440-452.123. Gonzalo S, García-Cao M, Fraga MF, Schotta G, Peters AH, Cotter SE,Eguía R, Dean DC, Esteller M, Jenuwein T, Blasco MA: Role of the RB1family in stabilizing histone methylation at constitutive heterochromatin.Nat Cell Biol 2005, 7:420-428.124. Schotta G, Sengupta R, Kubicek S, Malin S, Kauer M, Callén E, Celeste A,Pagani M, Opravil S, De La Rosa-Velazquez IA, Espejo A, Bedford MT,Nussenzweig A, Busslinger M, Jenuwein T: A chromatin-wide transition toH4K20 monomethylation impairs genome integrity and programmedDNA rearrangements in the mouse. Genes Dev 2008, 22:2048-2061.125. Ayoub N, Jeyasekharan AD, Venkitaraman AR: Mobilization andrecruitment of HP1β: a bimodal response to DNA breakage. Cell Cycle2009, 8:2945-2950.126. Luijsterburg MS, Dinant C, Lans H, Stap J, Wiernasz E, Lagerwerf S,Warmerdam DO, Lindh M, Brink MC, Dobrucki JW, Aten JA, Fousteri MI,Jansen G, Dantuma NP, Vermeulen W, Mullenders LH, Houtsmuller AB,Verschure PJ, van Driel R: Heterochromatin protein 1 is recruited tovarious types of DNA damage. J Cell Biol 2009, 185:577-586.127. Feng YQ, Lorincz MC, Fiering S, Greally JM, Bouhassira EE: Position effectsare influenced by the orientation of a transgene with respect toflanking chromatin. Mol Cell Biol 2001, 21:298-309.128. Juriloff DM, Harris MJ, Dewell SL, Brown CJ, Mager DL, Gagnier L, Mah DG:Investigations of the genomic region that contains the clf1 mutation, acausal gene in multifactorial cleft lip and palate in mice. Birth Defects ResA Clin Mol Teratol 2005, 73:103-113.129. Plamondon JA, Harris MJ, Mager DL, Gagnier L, Juriloff DM: The clf2 genehas an epigenetic role in the multifactorial etiology of cleft lip andpalate in the A/WySn mouse strain. Birth Defects Res A Clin Mol Teratol .130. Bouhassira EE, Westerman K, Leboulch P: Transcriptional behavior of LCRenhancer elements integrated at the same chromosomal locus byrecombinase-mediated cassette exchange. Blood 1997, 90:3332-3344.131. Kumaki Y, Oda M, Okano M: QUMA: quantification tool for methylationanalysis. Nucleic Acids Res 2008, , 36 Web Server: W170-W175.132. Jurka J, Kapitonov VV, Pavlicek A, Klonowski P, Kohany O, Walichiewicz J:Repbase Update, a database of eukaryotic repetitive elements. CytogenetGenome Res 2005, 110:462-467.133. Li H, Durbin R: Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009, 25:1754-1760.doi:10.1186/1756-8935-4-12Cite this article as: Maksakova et al.: H3K9me3-binding proteins aredispensable for SETDB1/H3K9me3-dependent retroviral silencing.Epigenetics & Chromatin 2011 4:12.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at et al. Epigenetics & Chromatin 2011, 4:12 18 of 18


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items