Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

Vesco Mutskov; Gary Felsenfeld

doi:10.1038/sj.emboj.7600013

Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

Mutskov, Vesco; Felsenfeld, Gary 2004-01-14 00:00:00 The EMBO Journal (2004) 23, 138–149 & 2004 European Molecular Biology Organization All Rights Reserved 0261-4189/04 | | THE THE www.embojournal.org EMB EMB EMBO O O JO JOU URN R NAL AL Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9 Vesco Mutskov and Gary Felsenfeld* 2002). Crosstalk between histone modiﬁcations and DNA methylation is thought to exist (Bird and Wolffe, 1999; Ben- Laboratory of Molecular Biology, NIDDK, National Institutes of Health, Porath and Cedar, 2001; Bird, 2002; Richards and Elgin, 2002) Bethesda, MD, USA and can lead to transcription repression, although it is not yet clear who initiates this ‘talk’, or when. In one possible Transgenes stably integrated into cells or animals in many pathway, DNA methylation is the ﬁrst event that initiates a cases are silenced rapidly, probably under the inﬂuence of chain of events resulting in modiﬁcation of the histone amino surrounding endogenous condensed chromatin. This gene terminal tails and creates silenced chromatin. This model silencing correlates with repressed chromatin structure comes from the observation that unmethylated transgenes marked by histone hypoacetylation, loss of methylation at H3 lysine 4, increase of histone H3 lysine 9 methylation as stably integrated into the genome become packaged with acetylated histones, while in vitro CpG-methylated trans- well as CpG DNA methylation at the promoter. However, genes become associated with deacetylated histones (Eden the order and the timing of these modiﬁcations and their et al, 1998; Schubeler et al, 2000). These methyl-CpG- impact on transcription inactivation are less well under- enriched regions may target methyl-CpG-binding proteins, stood. To determine the temporal order of these events, we which in turn recruit repressor complexes containing histone examined a model system consisting of a transgenic cas- sette stably integrated in chicken erythroid cells. We deacetylases (Bird and Wolffe, 1999) as well as histone methyltransferases (Fuks et al, 2003). In a second proposed found that histone H3 and H4 hypoacetylation and loss model, DNA methylation is a secondary event, induced by of methylation at H3 lysine 4 all occurred during the same an already silenced chromatin (Bird, 2002). Recently it was window of time as transgene inactivation in both multi- found that in Neurospora crassa all DNA methylation is copy and low-copy-number lines. These results indicate dependent on H3 lysine 9 (K9) methylation (Tamaru and that these histone modiﬁcations were the primary events Selker, 2001). Further chromatin immunoprecipitation experi- in gene silencing. We show that the kinetics of silencing ments showed that trimethylated H3 K9, but not dimethy- exclude histone H3 K9 and promoter DNA methylation as lated H3 K9, marked chromatin regions for cytosine the primary causative events in our transgene system. methylation in N. crassa (Tamaru et al, 2003). Similarly in The EMBO Journal (2004) 23, 138–149. doi:10.1038/ Arabidopsis, histone H3 K9 methylation is necessary for some sj.emboj.7600013; Published online 11 December 2003 of the CpNpG and asymmetric methylation (Jackson et al, Subject Categories: chromatin and transcription 2002). These observations suggest that DNA methylation acts Keywords: chromatin; DNA methylation; histone acetylation; histone methylation; transgene silencing downstream of H3 K9 methylation, at least in these two organisms. Does DNA methylation cause inactivation of transcription or is it a consequence of it? Which are the critical initiating events in gene repression and what is Introduction responsible for the maintenance of the silenced chromatin? Regulation of gene expression in eukaryotic organisms is a Obviously the answers to these questions may depend on the complex process that includes many levels of chromatin gene system, but these are particularly important questions remodeling and modiﬁcations. The hierarchical order of with respect to transgenes, since they concern their ability to events in the course of gene activation has been well studied maintain their pattern of expression after stable transfection (Cosma, 2002; Emerson, 2002); however, the changes in into cells or animals. In many cases, such transgenes are histone and DNA molecules that accompany transcription silenced rapidly, probably under the inﬂuence of surrounding inactivation have been less thoroughly investigated even endogenous condensed chromatin in which they ﬁnd them- though several repressor complexes with chromatin remodel- selves embedded. In this paper we focus on the silencing of ing activity have been described (Bird and Wolffe, 1999). It is such stably transfected genes. We show among other ﬁndings well established that gene silencing correlates with DNA that the kinetics of silencing speciﬁcally exclude DNA methy- hypermethylation, and the addition of a methyl group to lation as the primary causative event. the cytosine base can inﬂuence transcription by preventing transcription factor binding and/or by forming silent chro- matin structures (Bird and Wolffe, 1999; Jones and Baylin, Results To determine the temporal order of epigenetic chromatin *Corresponding author. NIH-NIDDK LMB, Building 5, Room 212, 5 Center Drive, MSC 0540, Bethesda, MD 20892-0540, USA. modiﬁcations and their impact on the process of transcription Tel.: þ 1 301 496 4173; Fax: þ 1 301 496 0201; inactivation, we employed our previously studied model E-mail: [email protected] system of a transgenic cassette stably integrated in chicken erythroid cells (Pikaart et al, 1998; Mutskov et al, 2002). The Received: 22 May 2003; accepted: 15 October 2003; Published online: 11 December 2003 gene encoding the Tac subunit of the interleukin 2 receptor 138 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld (IL2R) driven by an erythroid-speciﬁc chicken b -globin We previously demonstrated that transgene silencing in promoter and the b/e enhancer (Figure 3 bottom line) was 6C2 chicken cells, due to chromosomal position effects of the integrated into the early erythroid 6C2 cell line. A hygromy- integration site, is accompanied by changes in chromatin cin-resistant gene was cotransfected as a selectable marker structure and DNA methylation (Pikaart et al, 1998; for the transgene-expressing cells. When the hygromycin- Mutskov et al, 2002). Hypermethylation of the promoter as resistant transformed clones were grown in the presence of well as the coding region of the transgene was observed at hygromycin in the medium, all the cells expressed an anti- day 100; however, the critical CpG sites for silencing by DNA genic IL2R cell surface marker (day 0—Figures 1A and 6A), methylation were located over the promoter (Mutskov et al, due to the coordinated activity of the IL2R and hygromycin 2002). This ﬁnding suggested a link between transcription genes (Pikaart et al, 1998). However, after removal of the silencing and epigenetic modiﬁcation, but it did not provide drug, there was a gradual extinction of IL2R activity over a direct evidence as to whether DNA methylation caused IL2R period of 10–80 days in culture (Figures 1A and 6A; Pikaart inactivation or whether it was a consequence of it. For this et al, 1998). We followed the process of transgene silencing in reason we studied the kinetics of transition from the un- a multicopy cell line carrying 13 copies of the IL2R reporter methylated, active state to the densely methylated and in- cassette. FACS analysis of the IL2R surface marker in these active state of the transgene by bisulﬁte genomic sequencing cells at day 0 (still in the presence of hygromycin) and at days of genomic DNA, collected from the cells at different time 2, 4, 6, 8, 11, 13 and 19 after removal of hygromycin revealed points (days 0, 2, 4, 6, 12, 15, 19, 60 and 90) after release an extinction of expression (Figure 1A). At day 11 almost all from selection (Figure 2). We ampliﬁed a fragment covering the cells were IL2R negative. We also performed quantitative the main CpG cluster of the b -globin promoter, including the (TaqMan) RT–PCR to detect the presence of IL2R transcripts transcription start site and part of the IL2R cDNA (Figure 2B). during the time course of our experiment (Figure 1B). For In total, 20 clones from ﬁve individual PCR reactions were each time point, the level of IL2R mRNA was compared to the sequenced in each case to determine the level of methylation level of folate receptor mRNA, which is constantly expressed of individual CpG sites (numbered 1–21). The patterns of CpG in 6C2 cells (Prioleau et al, 1999). This experiment conﬁrmed methylation among the 20 clones were similar from clone to the gradual extinction of IL2R transcription, which was clone (data not shown). At day 0, the 13-copy cell line was abolished around days 10–13 (Figure 1B). hypomethylated at CpG dinucleotides 1–10 in the promoter region (Figure 2B), previously shown to be critical for gene expression (Mutskov et al, 2002). This region contains bind- ing sites for transcription factors that regulate the b -globin promoter in erythroid cells (Figure 2B; Emerson et al, 1985; A Day 0/Hygro Day 2 Day 4 Day 6 Gallarda et al, 1989). The CpGs 19–21, which are downstream of the transcription start site, were partially methylated and M2 M1 they apparently did not interfere with transcription (Mutskov et al, 2002). Following removal of selection, we observed a gradual increase in DNA methylation levels over the entire region, Day 8 Day 11 Day 13 Day 19 including CpGs 1–10, which were methylated in up to 80% of sequenced clones at day 90. Importantly, however, during the time of transcription inactivation (between days 2 and 12) the patterns of promoter methylation were very similar, with quite low levels of modiﬁcation (Figure 2A): only 5–25% of 100 sequenced clones were methylated at sites 1–10 at day 12, even though transcription was abolished. Site 8, lying close to the TATA box, was an exception with a moderate level of methylation (45%). However, this level did not increase beyond 45% until after all expression had been extinguished. In contrast, signiﬁcant DNA methylation (25–100%) over the promoter was observed at each CpG site at much later times—days 90 (Figure 2) and 100 (Mutskov et al, 2002). Based on these observations it appears that promoter DNA 0 5 10 15 20 methylation follows transgene inactivation and suggests a Days role in maintenance, not establishment, of gene silencing. The fact that the transgene seemed to be subject to CpG Figure 1 Extinction of IL2R transgene activity over an extended time in culture. (A) 6C2 cells carrying 13 copies of the IL2R reporter methylation mainly in the inactive state implied that other were grown for extended times without hygromycin selection. For mechanisms were likely involved in the initial steps of IL2R cell surface activity, the cells were analyzed by ﬂow cytometry repressed chromatin formation and silencing of gene expres- at different days. The horizontal bars denote M1 and M2, which sion. To explore this, we employed chromatin immunopreci- deﬁne the ranges of IL2R negative and positive cells, respectively. 6C2 cells untransfected with the IL2R construct were used to pitation (ChIP) assays to assess the kinetics of two major determine the IL2R negative cells zone on the histograms. (B) histone modiﬁcations associated with chromatin transcrip- Quantitative RT–PCR of a 13-copy cell line at days 0, 3, 6, 9, 13, tional states, acetylation of histones H3 and H4 and K9 18 and 20. Each point on the graph represents the amount of IL2R dimethylation of histone H3, during transgene silencing. At mRNA relative to the folate receptor mRNA per copy of the different time points (days 0 and 1, 5, 7, 10, 13, 17, 19 and 74 transgene. IL2R mRNA at day 0 is set to 100. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 139 | | Relative quantity of IL2R mRNA Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Day 0/Hygro Day 2 Day 4 100 100 100 80 80 80 60 60 40 40 40 20 20 20 0 0 Day 6 Day 12 Day 15 100 100 80 80 60 60 40 40 20 20 0 0 Day 19 Day 60 Day 90 100 100 100 80 80 80 60 60 60 40 40 40 20 20 20 0 0 0 CpG dinucleotide sites TATA CpG 1,2,3 4 5,6 7,8,9, 10,11,12,13,14,15,16,17 18 19,20,21 Nuclease hypersensitive β promoter IL2R cDNA Figure 2 Bisulﬁte genomic sequencing analysis of a multicopy cell line. (A) CpG methylation over the promoter and transcription start site proximal region of the IL2R transgene. Genomic DNA collected from the cells at different time points were bisulﬁte-treated, PCR-ampliﬁed and subcloned. In total, 20 individual clones were sequenced to give the methylation pattern of the 21 CpGs. The percentage of methylated clones for each CpG is plotted. (B) Distribution of CpG dinucleotides over the transgene. A primer set was designed to amplify the region corresponding to the main CpG cluster on the promoter, containing CpG dinucleotides 1–21. The nuclease hypersensitive region corresponds to a part of the chicken b -globin promoter where transcription factors (rectangles and ovals) are known to bind. after selection), cells were crosslinked with low concentra- neighborhood on the transgene, and compared these with tions of formaldehyde (up to 0.3% ﬁnal concentration), the RT–PCR curve for transcription activity. Figure 4 shows which allowed shearing of chromatin into small-sized frag- a kinetic view for different regions within the IL2R construct, ments for high resolution. We designed 10 primer sets and where the maximum values for the IL2R activity and for the TaqMan probes (Figure 3, bottom line, primer sets 1–10) histone and DNA modiﬁcations were taken as 100%. Histone spanning the entire transgene including the promoter, the deacetylation was coupled with gene inactivation and its IL2R cDNA, the splice/polyadenylation signal element and kinetics were similar in the 10 studied regions (primer sets the enhancer. A rapid loss of histone acetylation correlated 1–10) across the IL2R transgene (Figure 4A–D, data not with the extinction of IL2R activity (Figure 3). Histone H3 shown). Except for the region spanned by primer set 3, deacetylation in the promoter and the coding region appeared which showed the fastest kinetics of H3 K9 dimethylation to occur perhaps slightly earlier than histone H4 deacetyla- (but still with a relatively low level of modiﬁcation at day 10), tion. During the same time points, a gradual increase in the rest of the characterized regions had between 15 and 40% histone H3 dimethylation at lysine 9 was also observed. of the maximally modiﬁed H3 K9 nucleosomes at day 10. This However, similar to the DNA cytosine methylation, signiﬁ- modiﬁcation attained its maximum value long after the gene cant dimethylation at H3 K9 occurred after the gene was had been inactivated (days 19 and 74, Figures 3 and 4A–D). repressed. The kinetic curves show that CpG methylation of the We built kinetic curves for all the studied processes—DNA promoter region also occurs later than gene inactivation methylation, histone H3 and H4 acetylation and H3 K9 (Figure 4A–D, data not shown). CpG sites 1, 2 and 6 are methylation—with each curve representing an individual presented in Figure 4A, B and C, respectively. Even CpG site 140 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | EKLF NF-E4 CTF/NF-1 NF-E4 PAL AP-2 10 16 11 17 12 18 10 16 11 17 12 18 12 18 13 19 14 20 10 16 12 18 13 19 14 20 10 16 13 19 14 20 1 7 2 8 3 9 1 7 2 8 1 7 3 9 1 7 2 8 % of methylated clones Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Figure 3 ChIP analysis of histone modiﬁcations in different regions of the IL2R transgene in a multicopy cell line. Cells carrying the transgene were ﬁxed with formaldehyde at different time points and immunoprecipitated with antibodies against acetylated histones H3 and H4, dimethylated histone H3 K9 and nonimmune IgG. Primer sets 1–10 with TaqMan probes spanning the transgene were used to amplify the bound and input DNA. In parallel, these DNA samples were ampliﬁed with primers speciﬁc for the chicken b-globin locus as an internal control (primer sets #5.613 or #10.35, described in Litt et al, 2001b). The values from the ChIPs were corrected by subtraction of the nonspeciﬁc signal derived from the normal rabbit IgG ChIP. We compared the relative abundance of transgene sequences to the chicken internal control sequences in each bound fraction versus the input fraction. The presented diagrams show the fold differences of the analyzed proteins, where the highest relative-difference data point set is equal to 1.0. A map of the IL2R transgene is presented with the positions of the different primer sets used. Positions of the various transcription-factor-binding sites over the promoter are shown in Figure 2. 1, the fastest to be methylated in the entire promoter, became forms of H3 K4 disappear quite early, together with K9 methylated only after inactivation of expression was nearly acetylation. There is in fact some indication that loss of the complete (Figure 4A). Collectively, these data suggest that the dimethyl H3 K4 modiﬁcation may occur slightly earlier than primary mechanism of inactivation does not involve methy- the other modiﬁcations (Figure 5A-C, data not shown), but lation of the promoter DNA or of histone H3 K9, since these the data do not permit a strong conclusion to be drawn. events follow transgene repression. However, the high level The presence of multiple homologous copies of transgenes of DNA methylation of some CpGs in the gene body is not within an array may result in repeat-induced silencing, the changed during the time of gene inactivation or of histone mechanism of which is still unknown (Dorer and Henikoff, deacetylation/methylation (Figure 4D, CpG site 21), consis- 1997). Copy number reduction in transgenic mouse lines tent with the previous observation that gene body DNA caused a marked increase in expression of the transgenes methylation does not interfere with gene activity (Mutskov and less compaction of the chromatin at the transgene locus et al, 2002). (Garrick et al, 1998). We were interested in knowing if a low- There has been considerable interest recently in the asso- copy-number transgenic cell line would have the same ki- ciation of methylation at lysine 4 of histone H3 with tran- netics of chromatin modiﬁcations during inactivation as scriptionally active chromatin (Litt et al, 2001a; Santos-Rosa already seen for the high-copy line. A single or a low copy et al, 2002), and particularly in the difference between the number of IL2R reporters stably integrated in erythroid 6C2 roles of the dimethylated and trimethylated states (see cells were subjected to position effect silencing (Pikaart et al, Discussion). We therefore repeated our kinetic ChIP analysis 1998; data not shown). However, most of these lines (but not making use of antibodies speciﬁc for these two states. As all of them) lost expression at a slower rate as compared to shown in Figure 5A–C (and data not shown), the earliest the cell lines carrying a high copy number (data not shown). events associated with the disappearance of the IL2R tran- We performed kinetic experiments with a two-copy trans- script once again involve histone modiﬁcation at all the sites genic line, which had a gradual extinction of activity in in the promoter and gene body. Both the di- and trimethylated culture (Figures 4E and 6A). At days 28–29, over 95% of &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 141 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld the cells were IL2R inactive (Figure 6A) and, similar to the 13- cating no increased CpG methylation at the HpaII sites) copy line, at this time point the promoter DNA methylation during days 1–29, the time during which transgene inactiva- was relatively low (Figures 4E and 6B,C). We determined the tion occurred in these lines (Figure 6B). Signiﬁcant increases extent of CpG methylation for the low-copy line by HpaII were observed only at days 100 and later. This result was digestion of four sites in the transgene promoter (Figure 6B), conﬁrmed by bisulﬁte genomic sequencing analysis as well as by bisulﬁte analysis (Fig 6C). We obtained similar (Figure 6C). Finally, we compared the kinetics of histone patterns of low-level methyl-CpG-dependent digestion (indi- deacetylation, histone H3 K9 dimethylation and DNA methy- mRN A H3Ac H4Ac H3Me MeCpG 13 copy cell line ⎯ Primer 3, CpG 1 13 copy cell line ⎯ Primer 3, CpG 2 A B 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Days in culture Days in culture 13 copy cell line ⎯ Primer 4, CpG 6 13 copy cell line ⎯ Primer 6, CpG 21 20 20 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Days in culture Days in culture 2 copy cell line ⎯ Primer 3 0 10 20 30 40 50 60 70 80 90 100 Days in culture 142 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | % Change % Change % Change % Change % Change Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld mRNA H3Ac H3K4DiMe H3K4TiMe A B 13 copy cell line ⎯ Primer 1 13 copy cell line ⎯ Primer 3 0 5 10 15 20 25 0 5 10 15 20 25 Days in culture Days in culture C 13 copy cell line ⎯ Primer 6 0 5 10 15 20 25 Days in culture Figure 5 Combined kinetics of disappearance of mRNA, loss of histone H3 acetylation and loss of histone K4 di- and trimethylation in the multicopy cell line. The relative times of occurrence of histone modiﬁcations when compared to loss of gene expression are similar to those observed in (Figure 4A–D). However, there are some differences in the time in culture at which observable silencing commences, possibly reﬂecting differences in the time of incubation before hygromycin selection is removed. (A) Combined kinetics for histone modiﬁcation at promoter region 1 (primer set 1 in Figure 3, bottom line). (B) Kinetic curves for histone modiﬁcation in promoter region 3 (see Figure 3, bottom line). (C) Combined data from the gene body at region 6 (see Figure 3, bottom line). lation with gene activity (Figure 4E). The relationship methylation and cytosine methylation clearly are closely between these chromatin modiﬁcations and gene repression linked (Bird and Wolffe, 1999; Ben-Porath and Cedar, 2001; for the two-copy IL2R cell line was similar to that already Bird, 2002; Richards and Elgin, 2002), but their role in the seen in the multicopy line (compare Figure 4A–C with dynamic transitions between transcriptionally permissive Figure 4E). and transcriptionally silent chromatin remains to be deﬁned. In this study we performed experiments to monitor the kinetics of the chromatin modiﬁcations that accompany the Discussion gradual silencing of a transgene integrated in chicken ery- The four best-characterized hallmarks of repressed chroma- throid cells. The observed reduction in transcript abundance tin, histone H3 and H4 hypoacetylation, histone H3 K9 with time is not likely to reﬂect an alteration in the stability of Figure 4 Kinetics of transgene inactivation, histone modiﬁcations and DNA methylation for multicopy (A–D) and low–copy-number (E) number cell lines. The kinetic curve for the transgene activity in the timing experiment is built based on the level of IL2R mRNA at different days (pink color line, mRNA). The kinetics of changes in histone H3 acetylation, H4 acetylation and H3 K9 dimethylation are made using the ChIP data, and are indicated as H3Ac (green color line), H4Ac (red color line), and H3Me (blue color line), respectively. The kinetics of methylation of an individual CpG site on the transgene accompanying the extinction of IL2R expression are presented as MeCpG (orange color line). The maximum values of mRNA level, histone H3 and H4 acetylation, histone K9 H3 methylation and individual CpG methylation are set to 100. Multicopy cell line: (A) Combined kinetics for histone modiﬁcation at promoter region 3 (primer set #3 in Figure 3, bottom line) and methylation of CpG site 1 lying in the same region. (B) Combined data from the same region 3 but with the next CpG site: CpG site 2. (C) Kinetics for another region from the transgene promoter: primer set #4 (see Figure 3, bottom line) and CpG site 6. (D) Kinetics of DNA methylation of CpG site 21 from the gene body and the histone modiﬁcations kinetics at this region #6 (see Figure 3, bottom line) . Low-copy- number cell line: (E) Kinetic curves for IL2R activity and histone modiﬁcation in the promoter region #3. Changes in the DNA methylation state of the transgene promoter over time were determined by Southern blotting in Figure 5. Phosphor Imager analysis was performed to quantitate the intensity of the three bands on the blot for each time point. The band ‘a’ intensity is compared to the sum of the intensities of bands ‘a’,‘b’ and ‘c’. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 143 | | % Change % Change % Change Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Day 0/Hygro Day 8 Day 16 Day 23 Day 28 Day 100 β Promoter B IL2R β/ε Enhancer polyA CpG Probe c b a HpaII XbaI MMMM M Southern blotting Days in culture 0 /H, 1, 6, 11 , 21 , 29, 100 ,160 Bisulfite genomic sequencing analysis Day 1 Day 11 Day 21 100 100 80 80 60 60 40 40 20 20 0 0 Day 29 Day 100 Day 160 100 100 80 80 60 60 40 40 20 20 0 0 CpG dinucleotide sites Figure 6 Extinction of IL2R transgene activity and accompanying DNA methylation of a low-copy-number cell line. (A) FACS analysis of IL2R activity at day 0 (in the presence of hygromycin) and at days 8, 16, 23, 28 and 100 after removal of hygromycin. (B) Southern blots at different days in culture to determine the extent of DNA methylation at HpaII restriction sites in the transgene. The top diagram shows the positions of recognition sites for the methylation-sensitive enzyme HpaII within the IL2R transgene. Horizontal black lines with asterisks at the ends indicate the three products expected from complete digestion with the HpaII enzyme. The sites protected from digestion due to CpG methylation are indicated as M. The bottom panel shows DNA isolated at different days, digested with the XbaI/HpaII restriction enzymes and analyzed by Southern blotting after separation on an agarose gel. The membranes were hybridized with a BamHI–XbaI probe from the IL2R cDNA, indicated in the top diagram. (C) Bisulﬁte genomic sequencing analysis of a low-copy-number cell line at days 1, 11, 21, 29, 100 and 160. In total, 20 individual clones were sequenced to give the methylation pattern of the 21 CpGs. The percentage of methylated clones for each individual CpG is plotted. the mRNA in transfected cells. As reported in earlier work periods of time after the removal of hygromycin selection (Pikaart et al, 1998), some transgenic 6C2 cell lines, carrying marker, presumably because of effects of the integration site. the identical reporter used here, maintain expression for long Furthermore, these constructs surrounded by insulators ex- 144 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | 1 7 2 8 3 9 11 17 12 18 14 20 19 1 20 2 19 1 20 2 % of methylated clones Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Figure 7 Schematic representation of the sequence of events leading to silencing of a transgene stably integrated in the genome. Initially, the transgene is expressed and is not affected by the surrounding heterochromatin. The histones over the gene are acetylated (Ac) by HAT and the DNA is not methylated at the promoter region (open circles). After a period of propagation in culture, histone deacetylase activity (HDAC) is recruited to the transgene and the loss of histone acetylation is accompanied by transcription inactivation. Histone H3 that is di- or trimethylated at lysine 4 is also lost at this time. At the next step histone H3 K9 and DNA methyltransferases (HMTand DNMT) methylate their substrates (M for the histones and ﬁlled circles for the CpGs) in the already inactive transgene and form a more stable silenced state. Propagation of the heterochromatic state by heterochromatin protein 1 (HP1) and the binding of methyl-CpG-binding proteins (MBD) may ‘lock’ the silenced transgene. press at high and constant levels for long periods of time gene silencing or is in some cases a secondary modiﬁcation (Pikaart et al, 1998). affecting only the genes that have already been inactivated. Hypermethylation of promoter CpG islands is involved in The relationship of CpG methylation to gene silencing is an the epigenetic silencing of genes on the inactivated X-chro- issue that has been raised by many investigators. In the case mosome, imprinted genes, tumor suppressor genes and exo- of the transgene silencing mechanism we describe here, the genous integrated genes (Bird and Wolffe, 1999; Jones and observation that promoter methylation occurs later than Baylin, 2002). However, most of these observations were transcription inactivation implies that this epigenetic mark made on already repressed genes, leaving open the question is involved in maintaining the inactive state of the chromatin of whether cytosine methylation is always a primary agent of rather than initiating it. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 145 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Other inactivation pathways scriptional activity and DNA methylation over the coding Under other circumstances, for transgenes in which the DNA region. had been methylated in vitro before transfection, methylation It appears from our data that dimethylation of H3 K9 lies has been shown to serve an initiating function in the forma- downstream of histone deacetylation and transgene inactiva- tion of repressed chromatin, leading to H3/H4 hypoacetyla- tion. This is not surprising because H3K9 can be either tion (Eden et al, 1998; Schubeler et al, 2000) and H3 K9 acetylated or methylated and, in the process of inactivation, methylation (Santoro et al, 2002) after their integration into deacetylation must happen ﬁrst. Our observations of the the genome. Both methylation-dependent and -independent timing of modiﬁcations over the gene promoter are not mechanisms of repression have been observed in the case of inconsistent with some recent data on the timing of X- retroviral expression in embryonic stem cells (Cherry et al, chromosome inactivation (Heard et al, 2001). The chromo- 2000; Pannell et al, 2000). In another example where gene some-wide enrichment of K9 H3 dimethylation is thought to silencing is independent of DNA methylation, it has been be an early event, and marks a ‘hotspot’ important for the INK4a found that the tumor suppressor gene p16 can be initial steps in the X-inactivation process. However, a ChIP silenced in cell lines carrying a double knockout of DNMT1 experiment showed that the promoter regions of the X- and DNMT3b (DNMT: DNA methyltransferase). These muta- inactivated genes MeCP2 and G6pd were enriched in di- tions reduce genomic DNA methylation by 495%, and result methyl-K9 H3 later (by day 5 of differentiation) (Heard et al, in demethylation and initial loss of silencing of the gene 2001), indicating a difference between chromosome-wide INK4a (Rhee et al, 2002). However, the p16 gene is re-silenced compared to promoter-speciﬁc methylation on the X-chromo- with continued passage of the cells despite the absence of some. DNA methylation (Bachman et al, 2003). One of the goals of this investigation was to determine The situation with the genes on the inactivated X-chromo- whether an increase in H3 K9 methylation, which is usually some is less clear. Early kinetic studies showed that the associated with silencing, preceded or followed inactivation hypoxanthine phosphoribosyltransferase (Hprt) gene is silent of the transgene. It is possible that H3 K9 methylation is on the inactive X-chromosome before DNA methylation inhibited by methylation of lysine 4 at the same molecule. occurs in intron 1 sequences. This was determined by diges- This interplay has been shown by in vitro experiments to be tion with cytosine methylation-sensitive restriction endonu- speciﬁc for Suv39h1 histone methyltransferase (Wang et al, cleases (Lock et al, 1987). However, the kinetic analyses of 2001; Nishioka et al, 2002), but not for other H3 K9 methyl- the CpG island in the same Hprt gene using the bisulﬁte transferases: G9a (Nishioka et al, 2002) and SETDB1 (Schultz genomic sequencing method revealed that methylation oc- et al, 2002). However, we have also examined in separate curred at the initiation of X-chromosome inactivation (Park experiments the parallel changes in H3 K4 methylation. This and Chapman, 1997). modiﬁcation has been associated with transcriptionally ac- tive regulatory regions (Litt et al, 2001a) and coding regions (Litt et al, 2001a; Bernstein et al, 2002). As shown in Figure 5, loss of the dimethylated and trimethylated K4 modiﬁcations Histone H3 methylation occurs early in the inactivation process. This is certainly Methylation of histone H3 K9 has so far been implicated in consistent with the fact (but does not prove) that K4 must heterochromatin repression, promoter regulation and propa- be demethylated before K9 is methylated. The dimethylated gation of repressed chromatin (Litt et al, 2001a; Nielsen et al, 2001; Fahrner et al, 2002; Nguyen et al, 2002; Saccani and and trimethylated K4 modiﬁcations are lost at similar rates Natoli, 2002; Kondo et al, 2003). Our observation that histone over both the promoter and coding region of the IL2R reporter H3 K9 dimethylation occurs over the IL2R reporter (Figures 3 (Figure 5A–C, data not shown). and 4) also demonstrates that histone methylation is asso- ciated with silencing of the stably integrated transgenic Histone modiﬁcation and gene silencing constructs in eukaryotic cells. The early loss of methyl groups from H3 K4, and of acetyl This raises the question of whether DNA methylation groups from lysines on both H3 and H4 suggest that these in mammalian cells is dependent on the activity of histone modiﬁcations may play an important role in switching off methyltransferases (HMTs), as has been shown in gene transcription. It is important to emphasize that our data Neurospora and Arabidopsis (Tamaru and Selker, 2001; are consistent with such a role, but do not prove it. Our Jackson et al, 2002). It is known that DNMT3 proteins earlier results (Mutskov et al, 2002) showed that silencing of colocalize with HP1 (which binds to methylated histones) the transgenic cell lines carrying the same reporter used here at the heterochromatic loci in embryonic stem cells could be inhibited by growing the cells constantly in the (Bachman et al, 2001); however, a direct site-speciﬁc role presence of Trichostatin A, which blocks histone deacetyla- for histone methylation in the recruitment of DNMTs has not tion. In contrast, the growth of cells in 5-aza-cytidine, an been established. The kinetic data presented here do not inhibitor of DNA methylation, did not affect the rate of allow us to determine whether histone methylation over the inactivation, suggesting that silencing of the transgene was promoter precedes DNA methylation. Both occur, within our not controlled by this modiﬁcation (Mutskov et al, 2002). limits of resolution, at about the same time. We note however Thus, maintenance of acetylation is at least sufﬁcient to that the coding region DNA of the IL2R transgene is largely prevent inactivation. It has been established that speciﬁc methylated at early times despite the fact that H3 K9 di- histone acetylation patterns mediate regulatory factor inter- methylation levels are low over the same region (Figure 4D; actions critical for gene activity (Cosma, 2002; Emerson, Mutskov et al, 2002). This is connected to the fact that the 2002; Turner, 2002). The constant recruitment of proteins IL2R transgene is still active at that point; as shown earlier that carry histone acetyltransferase (HAT) activity and main- (Mutskov et al, 2002), there is no correlation between tran- tain hyperacetylated histones can help prevent switching to 146 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld line, obtained by transformation of chicken bone marrow with a the inactive state. Insulator elements such as the HS4 chicken wild-type avian erythroblastosis virus (AEV). 6C2 cells were grown b-globin insulator also help maintain a high level of histone as described (Boyes and Felsenfeld, 1996) and stably transfected acetylation over the protected region (Pikaart et al, 1998; with the reporters in the presence of DNA fragments encoding the Mutskov et al, 2002). The promoter of the gene adjacent to hygromycin-resistant gene derived from the pREP 7 plasmid (Invitrogen), by electroporation (Boyes and Felsenfeld, 1996), or the insulator is more accessible to transcription factors, and with effectene transfection reagent (Qiagen) according to the the binding of these factors in turn protects the promoter manufacturer’s protocol. Individual hygromycin-resistant colonies DNA from methylation (Mutskov et al, 2002). Immunity to were picked after 2–3 weeks of culture in a-MEM plus 2% methocel DNA methylation caused by active chromatin was also and 2000 U/ml hygromycin (Calbiochem), and expanded in hygromycin-containing medium (1250 U/ml) for a further 6 days. described for locus control regions (Santoso et al, 2000) IL2R expression was monitored by FACS analysis as described and enhancer elements (Schubeler et al, 2000; Mutskov (Pikaart et al, 1998) using a FACSCalibur immunocytometer (Becton et al, 2002), which are tissue-speciﬁc centers for high levels Dickinson). Cells were maintained in log-phase growth for all of histone acetylation. experiments. Our parallel kinetic studies of the changes in transgene Determination of transgene copy number in stably expression, histone acetylation and methylation, and DNA transfected cell lines methylation suggest a sequence of events in which loss of Genomic DNA from the stably transfected 6C2 cell lines was histone acetylation and H3 K4 di- and trimethylation are early isolated by a standard procedure (Sambrook et al, 1989) and series of two-fold dilutions were made. Equal amounts of DNA from each steps in the sequence of events leading to transgene silencing point of the double dilutions were analyzed by real-time PCR, in (Figure 7, step 1). At this stage the gene is repressed, but still parallel with two different primer sets with a very similar efﬁciency reversibly, because transcription can be partially reactivated of ampliﬁcation. The ﬁrst set of primers, including a TaqMan probe, by inhibition of histone deacetylases (TSA) (Pikaart et al, was speciﬁc for the IL2R transgene (primer set IL2R2800 described in Mutskov et al, 2002) and the second set ampliﬁed the 1998). Methylation of histone H3 K9 and of CpG sites on endogenous folate receptor gene (primer set FolateExon 4, promoter DNA are later events (Figure 7, step 2). These two described in Mutskov et al, 2002). The number of integrated chromatin modiﬁcations may target HP1 and MBD-protein transgenes was estimated by comparison of the IL2R signal to the repressor complexes that could ‘lock’ this stable silenced signal from one copy of the endogenous gene (half of the total folate receptor gene signal). As a control, the same experiment was chromatin state (Figure 7, step 3); in any case, the presence performed with genomic DNA isolated from nontransfected 6C2 of promoter CpG methylation almost certainly implies inabil- cells. ity to return to an active state. Moreover, histone and DNA methylation marks may signal to one another to ensure Quantitative TaqMan RT–PCR assay propagation of this more compact form of chromatin. We Total RNA from transfected 6C2 cells was isolated at different time points (see the Results section) and reverse transcribed as described previously observed binding of MBD proteins to the silenced previously (Mutskov et al, 2002). A series of two-fold dilutions of IL2R transgene at late stages (Mutskov et al, 2002), which the cDNA products were made, and equal amounts of DNA were could reinforce the histone modiﬁcation patterns via recruit- ampliﬁed by real-time PCR using speciﬁc primers and TaqMan ment of HDACs (Bird and Wolffe, 1999) and H3 K9 histone probes for the IL2R transgene and the folate receptor gene as described (Mutskov et al, 2002). methylase activity (Fuks et al, 2003). Although our speciﬁc results may serve as a more general Southern blot hybridization model for the silencing of transgenes stably integrated into a Isolation of high molecular weight genomic DNA was performed by cell or animal, we do not believe that they necessarily provide a standard method (Sambrook et al, 1989). In total, 10 mg of each DNA sample was subjected to double digestion with the restriction a universal mechanism for the regulation of endogenous endonucleases XbaI and the DNA methylation-sensitive HpaII. The genes, in which the order of events for gene silencing is digested DNA was resolved on an agarose gel, blotted on to likely to be under much stricter control. For example, the þ Hybond-N nylon membrane (Amersham) and hybridized with an endogenous chicken embryonic r-globin gene remains asso- IL2R probe in QuikHyb solution (Stratagene) according to the manufacturer’s protocol. ciated with hyperacetylated histones at day 15 of differentia- tion, even though this gene starts to be inactivated after day 5 Bisulﬁte genomic sequencing and transcription is barely detected at day 15 (Hebbes et al, Bisulﬁte conversion of DNA was carried out by the method 1992; Litt et al, 2001b). Recent results have indicated that developed by Clark et al (1994) with minor modiﬁcations described previously (Mutskov et al, 2002). Each bisulﬁte-modiﬁed DNA transcription activation in eukaryotes requires the recruit- sample was subjected to PCR reactions using primers speciﬁc for ment of chromatin remodeling and histone modiﬁcation the bisulﬁte-converted sequence of the IL2R transgene (primer set factors, but the order and the recruitment timing are gene- C1–C2, described in Mutskov et al, 2002). Ampliﬁcation, PCR speciﬁc events (Cosma, 2002; Emerson, 2002). Our results product puriﬁcation and subcloning were carried out as described show however that promoter DNA methylation and histone (Mutskov et al, 2002). Individual clones were sequenced using M13 reverse or M13 forward primers and the BigDye terminator cycle H3 K9 dimethylation are not early events in the silencing of a sequencing kit (PE Applied Biosystems) and applied to an ABI transgene. Extinction of expression is instead correlated with PRISM 310 DNA sequencer (Perkin Elmer). loss of histone acetylation and of H3 K4 methylation. It seems reasonable to focus future studies on one or both of these Formaldehyde crosslinking and ChIP Formaldehyde crosslinking in vivo, sonication of the chromatin early events as primary causes of silencing. and chromatin immunoprecipitation assays were performed as described previously (Mutskov et al, 2002). At different time points of our experiment (see the Results section), 6C2 cells stably Materials and methods transfected with the IL2R transgene were ﬁxed with 0.3% of formaldehyde at room temperature for 8 min, followed by incuba- Stable transfection in cell culture and FACS analysis tion at 41C for another 30 min. After isolation of nuclei, sonication The IL2R reporter construct under the control of the chicken b - and lysis with 1% SDS, the soluble chromatin was pre-cleared with globin promoter and b/e enhancer has been described previously salmon sperm DNA/protein A agarose (Upstate Biotechnology) (Pikaart et al, 1998). 6C2 is a CFU-E stage erythroid precursor cell before the immunoprecipitation procedure. Different antibodies &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 147 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld purchased from Upstate Biotechnology were used for ChIP: For each primer set, these fold difference values were corrected by acetylated histone H3 antibody, acetylated histone H4 antibody, subtraction of the nonspeciﬁc signal derived from the nonimmune dimethyl histone H3 (K9) antibody, dimethyl histone H3 (K4) rabbit IgG ChIP (t ): antibody, trimethyl histone H3 (K4) antibody as well as normal rabbit IgG in the control experiments. t t ðIP=InÞ ðIP=InÞ : Quantitative real-time PCR In parallel, DNA samples were ampliﬁed with primers speciﬁc for DNA samples from input (In) and antibody-bound (IP) chromatin the chicken b-globin locus as an internal control (c) (primer sets were analyzed by real-time PCR using the TaqMan Universal PCR #5.613 or #10.35, described in Litt et al, 2001b). Finally, we Master Mix (PE Applied Biosystems) and an ABI Prism 7700 normalized the relative abundance of transgene sequences to the sequence detector according to the manufacturer’s protocols. In chicken internal control sequence for each individual chromatin total, 10 primers and TaqMan probes were selected from the IL2R immunoprecipitation reaction using the following formula: transgene. Their sequences have been described previously (Mutskov et al, 2002): primer set #1 as IL2R1060, primer set #2 as t t c c 0 0 IL2R1420, primer set #3 as IL2R1900, primer set #4 as IL2R2020, ½ðIP=InÞ ðIP=InÞ =½ðIP=InÞ ðIP=InÞ : primer set #5 as IL2R2140, primer set #6 as IL2R2560, primer set #7 as IL2R2800, primer set #8 as IL2R3160, primer set #9 as IL2R3640, and primer set #10 as IL2R3880. Individual PCRs were carried out in triplicate to control for PCR variation and the C values were collected. Data quantiﬁcation was performed by applying the Acknowledgements comparative C method, as described by the manufacturer’s protocols (PE Applied Biosystems), giving the fold difference of a We thank Timur Yusufzai, Catherine Farrell and Adam West for target sequence (t) in the IP fraction versus a ﬁxed amount of In many stimulating discussions and for a critical reading of this DNA as a standard, where manuscript. We also acknowledge members of the Felsenfeld laboratory for useful suggestions, and Michael Krause’s laboratory DDC t ðC ðIPÞC ðInÞÞ t t IP=In ¼ 2 ¼ 2 : for DNA sequencing. References Bachman KE, Park BH, Rhee I, Rajagopalan H, Herman JG, Baylin Gallarda JL, Foley KP, Yang ZY, Engel JD (1989) The beta-globin SB, Kinzler KW, Vogelstein B (2003) Histone modiﬁcations and stage selector element factor is erythroid-speciﬁc promoter/ silencing prior to DNA methylation of a tumor suppressor gene. enhancer binding protein NF-E4. Genes Dev 3: 1845–1859 Cancer Cell 3: 89–95 Garrick D, Fiering S, Martin DI, Whitelaw E (1998) Repeat-induced Bachman KE, Rountree MR, Baylin SB (2001) Dnmt3a and gene silencing in mammals. Nat Genet 18: 56–59 Dnmt3b are transcriptional repressors that exhibit unique Heard E, Rougeulle C, Arnaud D, Avner P, Allis CD, Spector DL localization properties to heterochromatin. J Biol Chem 276: (2001) Methylation of histone H3 at Lys-9 is an early mark on the 32282–32287 X chromosome during X inactivation. Cell 107: 727–738 Ben-Porath I, Cedar H (2001) Epigenetic crosstalk. Mol Cell 8: Hebbes TR, Thorne AW, Clayton AL, Crane-Robinson C (1992) 933–935 Histone acetylation and globin gene switching. Nucleic Acids Bernstein BE, Humphrey EL, Erlich RL, Schneider R, Bouman P, Liu Res 20: 1017–1022 JS, Kouzarides T, Schreiber SL (2002) Methylation of histone H3 Jackson JP, Lindroth AM, Cao X, Jacobsen SE (2002) Control of Lys 4 in coding regions of active genes. Proc Natl Acad Sci USA CpNpG DNA methylation by the KRYPTONITE histone H3 99: 8695–8700 methyltransferase. Nature 416: 556–560 Bird A (2002) DNA methylation patterns and epigenetic memory. Jones PA, Baylin SB (2002) The fundamental role of epigenetic Genes Dev 16: 6–21 events in cancer. Nat Rev Genet 3: 415–428 Bird AP, Wolffe AP (1999) Methylation-induced repression-belts, Kondo Y, Shen L, Issa JP (2003) Critical role of histone methylation braces, and chromatin. Cell 99: 451–454 in tumor suppressor gene silencing in colorectal cancer. Mol Cell Boyes J, Felsenfeld G (1996) Tissue-speciﬁc factors additively Biol 23: 206–215 increase the probability of the all-or-none formation of a hyper- Litt MD, Simpson M, Gaszner M, Allis CD, Felsenfeld G (2001a) sensitive site. EMBO J 15: 2496–2507 Correlation between histone lysine methylation and developmen- Cherry SR, Biniszkiewicz D, van Parijs L, Baltimore D, Jaenisch R tal changes at the chicken beta-globin locus. Science 293: (2000) Retroviral expression in embryonic stem cells and hema- 2453–2455 topoietic stem cells. Mol Cell Biol 20: 7419–7426 Litt MD, Simpson M, Recillas-Targa F, Prioleau M-N, Felsenfeld G Clark SJ, Harrison J, Paul CL, Frommer M (1994) High sensitivity (2001b) Transitions in histone acetylation reveal boundaries mapping of methylated cytosines. Nucleic Acids Res 22: of three separately regulated neighboring loci. EMBO J 20: 2990–2997 2224–2235 Cosma MP (2002) Ordered recruitment: gene-speciﬁc mechanism of Lock LF, Takagi N, Martin GR (1987) Methylation of the Hprt gene transcription activation. Mol Cell 10: 227–236 on the inactive X occurs after chromosome inactivation. Cell 48: Dorer DR, Henikoff S (1997) Transgene repeat arrays interact with 39–46 distant heterochromatin and cause silencing in cis and trans. Mutskov VJ, Farrell CM, Wade PA, Wolffe AP, Felsenfeld G (2002) Genetics 147: 1181–1190 The barrier function of an insulator couples high histone acetyla- Eden S, Hashimshony T, Keshet I, Cedar H, Thorne AW (1998) DNA tion levels with speciﬁc protection of promoter DNA from methy- methylation models histone acetylation. Nature 394: 842 lation. Genes Dev 16: 1540–1554 Emerson BM (2002) Speciﬁcity of gene regulation. Cell 109: 267–270 Nguyen CT, Weisenberger DJ, Velicescu M, Gonzales FA, Lin JC, Emerson BM, Lewis CD, Felsenfeld G (1985) Interaction of speciﬁc Liang G, Jones PA (2002) Histone H3-lysine 9 methylation is nuclear factors with the nuclease-hypersensitive region of the associated with aberrant gene silencing in cancer cells and chicken adult beta-globin gene: nature of the binding domain. is rapidly reversed by 5-Aza-2 -deoxycytidine. Cancer Res 62: Cell 41: 21–30 6456–6461 Fahrner JS, Herman JG, Baylin SB (2002) Dependence of histone Nielsen SJ, Schneider R, Bauer UM, Bannister AJ, Morrison A, modiﬁcations and gene expression on DNA hypermethylation in O’Carroll D, Firestein R, Cleary M, Jenuwein T, Herrera RE, cancer. Cancer Res 62: 7213–7218 Kouzarides T (2001) Rb targets histone H3 methylation and Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T (2003) The HP1 to promoters. Nature 412: 561–565 methyl-CpG-binding protein MeCP2 links DNA methylation to Nishioka K, Chuikov S, Sarma K, Erdjument-Bromage H, Allis CD, histone methylation. J Biol Chem 278: 4035–4040 Tempst P, Reinberg D (2002) Set9, a novel histone H3 methyl- 148 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld transferase that facilitates transcription by precluding histone tail Santoro R, Li J, Grummt I (2002) The nucleolar remodeling complex modiﬁcations required for heterochromatin formation. Genes Dev NoRC mediates heterochromatin formation silencing of ribosomal 16: 479–489 gene transcription. Nat Genet 32: 393–396 Pannell D, Osborne CS, Osborne CS, Yao S, Sukonnik T, Pasceri P, Santoso B, Ortiz BD, Winoto A (2000) Control of organ-speciﬁc Karaiskakis A, Okano M, Li E, Lipshitz HD, Ellis J (2000) demethylation by an element of the T-cell receptor-alpha locus Retrovirus vector silencing is de novo methylase independent control region. J Biol Chem 275: 1952–1958 and marked by a repressive histone code. EMBO J 19: 5884–5894 Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J, Bernstein BE, Park JG, Chapman VM (1997) DNA methylation patterns and Emre NC, Schreiber SL, Mellor J, Kouzarides T (2002) Active sequence transitions of the CpG island of mouse Hprt during genes are tri-methylated at K4 of histone H3. Nature 419: 407–411 early embryogenesis. Mol Cells 7: 259–263 Schubeler D, Lorincz MC, Cimbora DM, Telling A, Feng YQ, Pikaart MJ, Recillas-Targa F, Felsenfeld G (1998) Loss of transcrip- Bouhassira EE, Groudine M (2000) Genomic targeting of methy- tional activity of a transgene is accompanied by DNA methylation lated DNA: inﬂuence of methylation on transcription, replication, and histone deacetylation and is prevented by insulators. Genes chromatin structure, and histone acetylation. Mol Cell Biol 20: Dev 12: 2852–2862 9103–9112 Prioleau MN, Nony P, Simpson M, Felsenfeld G (1999) An insulator Schultz DC, Ayyanathan K, Negorev D, Maul GG, Rauscher III FJ element and condensed chromatin region separate the chicken 2002 SETDB1: a novel KAP-1-associated histone H3, lysine 9- beta-globin locus from an independently regulated erythroid- speciﬁc methyltransferase that contributes to HP1-mediated si- speciﬁc folate receptor gene. EMBO J 18: 4035–4048 lencing of euchromatic genes by KRAB zinc-ﬁnger proteins. Rhee I, Bachman KE, Park BH, Jair KW, Yen RW, Schuebel KE, Cui Genes Dev 16: 919–932 H, Feinberg AP, Lengauer C, Kinzler KW, Baylin SB, Vogelstein B Tamaru H, Selker EU (2001) A histone H3 methyltransferase con- (2002) DNMT1 and DNMT3b cooperate to silence genes in trols DNA methylation in Neurospora crassa. Nature 414: 277–283 human cancer cells. Nature 416: 552–556 Tamaru H, Zhang X, McMillen D, Singh PB, Nakayama J, Grewal SI, Richards EJ, Elgin SC (2002) Epigenetic codes for heterochromatin Allis CD, Cheng X, Selker EU (2003) Trimethylated lysine 9 of formation and silencing: rounding up the usual suspects. Cell 108: histone H3 is a mark for DNA methylation in Neurospora crassa. 489–500 Nat Genet 34: 75–79 Saccani S, Natoli G (2002) Dynamic changes in histone H3 Lys 9 Turner BM (2002) Cellular memory and the histone code. Cell 111: methylation occurring at tightly regulated inducible inﬂammatory 285–291 genes. Genes Dev 16: 2219–2224 Wang H, Cao R, Xia L, Erdjument-Bromage H, Borchers C, Tempst P, Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Zhang Y (2001) Puriﬁcation and functional characterization of Laboratory Manual, 2nd edn. Plainview, NY: Cold Spring Harbor a histone H3-lysine 4-speciﬁc methyltransferase. Mol Cell 8: Laboratory Press 1207–1217 &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 149 | | http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals http://www.deepdyve.com/lp/springer-journals/silencing-of-transgene-transcription-precedes-methylation-of-promoter-3IOEYjXf5y

Loading next page...

References (48)

E. Heard, C. Rougeulle, D. Arnaud, P. Avner, C. Allis, D. Spector (2001)
Methylation of Histone H3 at Lys-9 Is an Early Mark on the X Chromosome during X Inactivation
Cell, 107
D. Garrick, S. Fiering, David Martin, E. Whitelaw (1998)
Repeat-induced gene silencing in mammals
Nature Genetics, 18
D. Dorer, S. Henikoff (1997)
Transgene repeat arrays interact with distant heterochromatin and cause silencing in cis and trans.
Genetics, 147 3
D. Schultz, K. Ayyanathan, D. Negorev, G. Maul, F. Rauscher (2002)
SETDB1: a novel KAP-1-associated histone H3, lysine 9-specific methyltransferase that contributes to HP1-mediated silencing of euchromatic genes by KRAB zinc-finger proteins.
Genes & development, 16 8
Ina Rhee, K. Bachman, B. Park, K. Jair, R. Yen, Kornel Schuebel, H. Cui, A. Feinberg, C. Lengauer, K. Kinzler, S. Baylin, B. Vogelstein (2002)
DNMT1 and DNMT3b cooperate to silence genes in human cancer cells
Nature, 416
A. Bird (2002)
DNA methylation patterns and epigenetic memory.
Genes & development, 16 1
D. Pannell, Cameron Osborne, Shuyuan Yao, Tatyana Sukonnik, P. Pasceri, A. Karaiskakis, M. Okano, E. Li, H. Lipshitz, J. Ellis (2000)
Retrovirus vector silencing is de novo methylase independent and marked by a repressive histone code
The EMBO Journal, 19
L. Lock, N. Takagi, G. Martin (1987)
Methylation of the Hprt gene on the inactive X occurs after chromosome inactivation
Cell, 48
Peter Jones, S. Baylin (2002)
The fundamental role of epigenetic events in cancer
Nature Reviews Genetics, 3
K. Bachman, B. Park, Ina Rhee, H. Rajagopalan, J. Herman, S. Baylin, K. Kinzler, B. Vogelstein (2003)
Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene.
Cancer cell, 3 1
Vesco Mutskov, C. Farrell, P. Wade, A. Wolffe, G. Felsenfeld (2002)
The barrier function of an insulator couples high histone acetylation levels with specific protection of promoter DNA from methylation.
Genes & development, 16 12
E. Richards, S. Elgin (2002)
Epigenetic Codes for Heterochromatin Formation and Silencing Rounding up the Usual Suspects
Cell, 108
B. Turner (2002)
Cellular Memory and the Histone Code
Cell, 111
J. Boyes, G. Felsenfeld (1996)
Tissue‐specific factors additively increase the probability of the all‐or‐none formation of a hypersensitive site.
The EMBO Journal, 15
Jill Fahrner, Sayaka Eguchi, J. Herman, S. Baylin (2002)
Dependence of histone modifications and gene expression on DNA hypermethylation in cancer.
Cancer research, 62 24
Y. Kondo, Lanlan Shen, J. Issa (2003)
Critical Role of Histone Methylation in Tumor Suppressor Gene Silencing in Colorectal Cancer
Molecular and Cellular Biology, 23
Michael Litt, M. Simpson, F. Recillas-Targa, M. Prioleau, G. Felsenfeld (2001)
Transitions in histone acetylation reveal boundaries of three separately regulated neighboring loci
The EMBO Journal, 20
R. Santoro, Junwei Li, I. Grummt (2002)
The nucleolar remodeling complex NoRC mediates heterochromatin formation and silencing of ribosomal gene transcription
Nature Genetics, 32
K. Bachman, M. Rountree, S. Baylin (2001)
Dnmt3a and Dnmt3b Are Transcriptional Repressors That Exhibit Unique Localization Properties to Heterochromatin*
The Journal of Biological Chemistry, 276
M. Prioleau, P. Nony, M. Simpson, G. Felsenfeld (1999)
An insulator element and condensed chromatin region separate the chicken β‐globin locus from an independently regulated erythroid‐specific folate receptor gene
The EMBO Journal, 18
Michael Litt, M. Simpson, M. Gaszner, C. Allis, G. Felsenfeld (2001)
Correlation Between Histone Lysine Methylation and Developmental Changes at the Chicken β-Globin Locus
Science, 293
J. Jackson, Anders Lindroth, Xiaofeng Cao, S. Jacobsen (2002)
Control of CpNpG DNA methylation by the KRYPTONITE histone H3 methyltransferase
Nature, 416
H. Tamaru, E. Selker (2001)
A histone H3 methyltransferase controls DNA methylation in Neurospora crassa
Nature, 414
B. Santoso, B. Ortiz, A. Winoto (2000)
Control of Organ-specific Demethylation by an Element of the T-cell Receptor-α Locus Control Region*
The Journal of Biological Chemistry, 275
A. Bird, A. Wolffe (1999)
Methylation-Induced Repression— Belts, Braces, and Chromatin
Cell, 99
J.CIark Susan, J. Harrison, C. Paul, M. Frommer (1994)
High sensitivity mapping of methylated cytosines.
Nucleic acids research, 22 15
M. Cosma (2002)
Ordered recruitment: gene-specific mechanism of transcription activation.
Molecular cell, 10 2
J. Gallarda, K. Foley, Z. Yang, J. Engel (1989)
The beta-globin stage selector element factor is erythroid-specific promoter/enhancer binding protein NF-E4.
Genes & development, 3 12A
B. Bernstein, Emily Humphrey, R. Erlich, R. Schneider, P. Bouman, J. Liu, T. Kouzarides, S. Schreiber (2002)
Methylation of histone H3 Lys 4 in coding regions of active genes
Proceedings of the National Academy of Sciences of the United States of America, 99
I. Ben-Porath, Howard Cedar (2001)
Epigenetic crosstalk.
Molecular cell, 8 5
J. Sambrook, E. Fritsch, T. Maniatis (2001)
Molecular Cloning: A Laboratory Manual
M. Pikaart, F. Recillas-Targa, G. Felsenfeld (1998)
Loss of transcriptional activity of a transgene is accompanied by DNA methylation and histone deacetylation and is prevented by insulators.
Genes & development, 12 18
S. Nielsen, R. Schneider, U. Bauer, Andrew Bannister, Ashby Morrison, D. O’Carroll, R. Firestein, M. Cleary, T. Jenuwein, R. Herrera, T. Kouzarides (2001)
Rb targets histone H3 methylation and HP1 to promoters
Nature, 412
H. Tamaru, Xing Zhang, D. McMillen, Prim Singh, J. Nakayama, S. Grewal, C. Allis, Xiaodong Cheng, E. Selker (2003)
Trimethylated lysine 9 of histone H3 is a mark for DNA methylation in Neurospora crassa
Nature Genetics, 34
H. Santos‐Rosa, R. Schneider, Andrew Bannister, Julia Sherriff, B. Bernstein, N. Emre, S. Schreiber, J. Mellor, T. Kouzarides (2002)
Active genes are tri-methylated at K4 of histone H3
Nature, 419
Hengbin Wang, R. Cao, L. Xia, H. Erdjument-Bromage, C. Borchers, P. Tempst, Yi Zhang (2001)
Purification and functional characterization of a histone H3-lysine 4-specific methyltransferase.
Molecular cell, 8 6
Carvell Nguyen, D. Weisenberger, M. Velicescu, F. Gonzales, Joy Lin, G. Liang, Peter Jones (2002)
Histone H3-lysine 9 methylation is associated with aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-2'-deoxycytidine.
Cancer research, 62 22
T. Hebbes, A. Thorne, A. Clayton, Colyn Crane-Robinson (1992)
Histone acetylation and globin gene switching.
Nucleic acids research, 20 5
B. Emerson, C. Lewis, G. Felsenfeld (1985)
Interaction of specific nuclear factors with the nuclease-hypersensitive region of the chicken adult β-globin gene: Nature of the binding domain
Cell, 41
R Santoro, J Li, I Grummt (2002)

Nat Genet, 32
S. Eden, T. Hashimshony, I. Keshet, H. Cedar, A. Thorne (1998)
DNA methylation models histone acetylation
Nature, 394
S. Saccani, G. Natoli (2002)
Dynamic changes in histone H3 Lys 9 methylation occurring at tightly regulated inducible inflammatory genes.
Genes & development, 16 17
B. Emerson (2002)
Specificity of Gene Regulation
Cell, 109
S. Cherry, D. Biniszkiewicz, L. Parijs, D. Baltimore, R. Jaenisch (2000)
Retroviral Expression in Embryonic Stem Cells and Hematopoietic Stem Cells
Molecular and Cellular Biology, 20
Jaehong Park, Chapman Vm (1997)
DNA methylation patterns and sequence transitions of the CpG island of mouse Hprt during early embryogenesis.
Molecules and cells, 7 2
F. Fuks, P. Hurd, D. Wolf, X. Nan, A. Bird, T. Kouzarides (2003)
The Methyl-CpG-binding Protein MeCP2 Links DNA Methylation to Histone Methylation*
The Journal of Biological Chemistry, 278
K. Nishioka, S. Chuikov, K. Sarma, H. Erdjument-Bromage, C. Allis, P. Tempst, D. Reinberg (2002)
Set9, a novel histone H3 methyltransferase that facilitates transcription by precluding histone tail modifications required for heterochromatin formation.
Genes & development, 16 4
D. Schübeler, M. Lorincz, D. Cimbora, A. Telling, Yongjie Feng, E. Bouhassira, M. Groudine (2000)
Genomic Targeting of Methylated DNA: Influence of Methylation on Transcription, Replication, Chromatin Structure, and Histone Acetylation
Molecular and Cellular Biology, 20

Publisher: Springer Journals
Copyright: Copyright © European Molecular Biology Organization 2004
ISSN: 0261-4189
eISSN: 1460-2075
DOI: 10.1038/sj.emboj.7600013
Publisher site: See Article on Publisher Site

Abstract

The EMBO Journal (2004) 23, 138–149 & 2004 European Molecular Biology Organization All Rights Reserved 0261-4189/04 | | THE THE www.embojournal.org EMB EMB EMBO O O JO JOU URN R NAL AL Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9 Vesco Mutskov and Gary Felsenfeld* 2002). Crosstalk between histone modiﬁcations and DNA methylation is thought to exist (Bird and Wolffe, 1999; Ben- Laboratory of Molecular Biology, NIDDK, National Institutes of Health, Porath and Cedar, 2001; Bird, 2002; Richards and Elgin, 2002) Bethesda, MD, USA and can lead to transcription repression, although it is not yet clear who initiates this ‘talk’, or when. In one possible Transgenes stably integrated into cells or animals in many pathway, DNA methylation is the ﬁrst event that initiates a cases are silenced rapidly, probably under the inﬂuence of chain of events resulting in modiﬁcation of the histone amino surrounding endogenous condensed chromatin. This gene terminal tails and creates silenced chromatin. This model silencing correlates with repressed chromatin structure comes from the observation that unmethylated transgenes marked by histone hypoacetylation, loss of methylation at H3 lysine 4, increase of histone H3 lysine 9 methylation as stably integrated into the genome become packaged with acetylated histones, while in vitro CpG-methylated trans- well as CpG DNA methylation at the promoter. However, genes become associated with deacetylated histones (Eden the order and the timing of these modiﬁcations and their et al, 1998; Schubeler et al, 2000). These methyl-CpG- impact on transcription inactivation are less well under- enriched regions may target methyl-CpG-binding proteins, stood. To determine the temporal order of these events, we which in turn recruit repressor complexes containing histone examined a model system consisting of a transgenic cas- sette stably integrated in chicken erythroid cells. We deacetylases (Bird and Wolffe, 1999) as well as histone methyltransferases (Fuks et al, 2003). In a second proposed found that histone H3 and H4 hypoacetylation and loss model, DNA methylation is a secondary event, induced by of methylation at H3 lysine 4 all occurred during the same an already silenced chromatin (Bird, 2002). Recently it was window of time as transgene inactivation in both multi- found that in Neurospora crassa all DNA methylation is copy and low-copy-number lines. These results indicate dependent on H3 lysine 9 (K9) methylation (Tamaru and that these histone modiﬁcations were the primary events Selker, 2001). Further chromatin immunoprecipitation experi- in gene silencing. We show that the kinetics of silencing ments showed that trimethylated H3 K9, but not dimethy- exclude histone H3 K9 and promoter DNA methylation as lated H3 K9, marked chromatin regions for cytosine the primary causative events in our transgene system. methylation in N. crassa (Tamaru et al, 2003). Similarly in The EMBO Journal (2004) 23, 138–149. doi:10.1038/ Arabidopsis, histone H3 K9 methylation is necessary for some sj.emboj.7600013; Published online 11 December 2003 of the CpNpG and asymmetric methylation (Jackson et al, Subject Categories: chromatin and transcription 2002). These observations suggest that DNA methylation acts Keywords: chromatin; DNA methylation; histone acetylation; histone methylation; transgene silencing downstream of H3 K9 methylation, at least in these two organisms. Does DNA methylation cause inactivation of transcription or is it a consequence of it? Which are the critical initiating events in gene repression and what is Introduction responsible for the maintenance of the silenced chromatin? Regulation of gene expression in eukaryotic organisms is a Obviously the answers to these questions may depend on the complex process that includes many levels of chromatin gene system, but these are particularly important questions remodeling and modiﬁcations. The hierarchical order of with respect to transgenes, since they concern their ability to events in the course of gene activation has been well studied maintain their pattern of expression after stable transfection (Cosma, 2002; Emerson, 2002); however, the changes in into cells or animals. In many cases, such transgenes are histone and DNA molecules that accompany transcription silenced rapidly, probably under the inﬂuence of surrounding inactivation have been less thoroughly investigated even endogenous condensed chromatin in which they ﬁnd them- though several repressor complexes with chromatin remodel- selves embedded. In this paper we focus on the silencing of ing activity have been described (Bird and Wolffe, 1999). It is such stably transfected genes. We show among other ﬁndings well established that gene silencing correlates with DNA that the kinetics of silencing speciﬁcally exclude DNA methy- hypermethylation, and the addition of a methyl group to lation as the primary causative event. the cytosine base can inﬂuence transcription by preventing transcription factor binding and/or by forming silent chro- matin structures (Bird and Wolffe, 1999; Jones and Baylin, Results To determine the temporal order of epigenetic chromatin *Corresponding author. NIH-NIDDK LMB, Building 5, Room 212, 5 Center Drive, MSC 0540, Bethesda, MD 20892-0540, USA. modiﬁcations and their impact on the process of transcription Tel.: þ 1 301 496 4173; Fax: þ 1 301 496 0201; inactivation, we employed our previously studied model E-mail: [email protected] system of a transgenic cassette stably integrated in chicken erythroid cells (Pikaart et al, 1998; Mutskov et al, 2002). The Received: 22 May 2003; accepted: 15 October 2003; Published online: 11 December 2003 gene encoding the Tac subunit of the interleukin 2 receptor 138 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld (IL2R) driven by an erythroid-speciﬁc chicken b -globin We previously demonstrated that transgene silencing in promoter and the b/e enhancer (Figure 3 bottom line) was 6C2 chicken cells, due to chromosomal position effects of the integrated into the early erythroid 6C2 cell line. A hygromy- integration site, is accompanied by changes in chromatin cin-resistant gene was cotransfected as a selectable marker structure and DNA methylation (Pikaart et al, 1998; for the transgene-expressing cells. When the hygromycin- Mutskov et al, 2002). Hypermethylation of the promoter as resistant transformed clones were grown in the presence of well as the coding region of the transgene was observed at hygromycin in the medium, all the cells expressed an anti- day 100; however, the critical CpG sites for silencing by DNA genic IL2R cell surface marker (day 0—Figures 1A and 6A), methylation were located over the promoter (Mutskov et al, due to the coordinated activity of the IL2R and hygromycin 2002). This ﬁnding suggested a link between transcription genes (Pikaart et al, 1998). However, after removal of the silencing and epigenetic modiﬁcation, but it did not provide drug, there was a gradual extinction of IL2R activity over a direct evidence as to whether DNA methylation caused IL2R period of 10–80 days in culture (Figures 1A and 6A; Pikaart inactivation or whether it was a consequence of it. For this et al, 1998). We followed the process of transgene silencing in reason we studied the kinetics of transition from the un- a multicopy cell line carrying 13 copies of the IL2R reporter methylated, active state to the densely methylated and in- cassette. FACS analysis of the IL2R surface marker in these active state of the transgene by bisulﬁte genomic sequencing cells at day 0 (still in the presence of hygromycin) and at days of genomic DNA, collected from the cells at different time 2, 4, 6, 8, 11, 13 and 19 after removal of hygromycin revealed points (days 0, 2, 4, 6, 12, 15, 19, 60 and 90) after release an extinction of expression (Figure 1A). At day 11 almost all from selection (Figure 2). We ampliﬁed a fragment covering the cells were IL2R negative. We also performed quantitative the main CpG cluster of the b -globin promoter, including the (TaqMan) RT–PCR to detect the presence of IL2R transcripts transcription start site and part of the IL2R cDNA (Figure 2B). during the time course of our experiment (Figure 1B). For In total, 20 clones from ﬁve individual PCR reactions were each time point, the level of IL2R mRNA was compared to the sequenced in each case to determine the level of methylation level of folate receptor mRNA, which is constantly expressed of individual CpG sites (numbered 1–21). The patterns of CpG in 6C2 cells (Prioleau et al, 1999). This experiment conﬁrmed methylation among the 20 clones were similar from clone to the gradual extinction of IL2R transcription, which was clone (data not shown). At day 0, the 13-copy cell line was abolished around days 10–13 (Figure 1B). hypomethylated at CpG dinucleotides 1–10 in the promoter region (Figure 2B), previously shown to be critical for gene expression (Mutskov et al, 2002). This region contains bind- ing sites for transcription factors that regulate the b -globin promoter in erythroid cells (Figure 2B; Emerson et al, 1985; A Day 0/Hygro Day 2 Day 4 Day 6 Gallarda et al, 1989). The CpGs 19–21, which are downstream of the transcription start site, were partially methylated and M2 M1 they apparently did not interfere with transcription (Mutskov et al, 2002). Following removal of selection, we observed a gradual increase in DNA methylation levels over the entire region, Day 8 Day 11 Day 13 Day 19 including CpGs 1–10, which were methylated in up to 80% of sequenced clones at day 90. Importantly, however, during the time of transcription inactivation (between days 2 and 12) the patterns of promoter methylation were very similar, with quite low levels of modiﬁcation (Figure 2A): only 5–25% of 100 sequenced clones were methylated at sites 1–10 at day 12, even though transcription was abolished. Site 8, lying close to the TATA box, was an exception with a moderate level of methylation (45%). However, this level did not increase beyond 45% until after all expression had been extinguished. In contrast, signiﬁcant DNA methylation (25–100%) over the promoter was observed at each CpG site at much later times—days 90 (Figure 2) and 100 (Mutskov et al, 2002). Based on these observations it appears that promoter DNA 0 5 10 15 20 methylation follows transgene inactivation and suggests a Days role in maintenance, not establishment, of gene silencing. The fact that the transgene seemed to be subject to CpG Figure 1 Extinction of IL2R transgene activity over an extended time in culture. (A) 6C2 cells carrying 13 copies of the IL2R reporter methylation mainly in the inactive state implied that other were grown for extended times without hygromycin selection. For mechanisms were likely involved in the initial steps of IL2R cell surface activity, the cells were analyzed by ﬂow cytometry repressed chromatin formation and silencing of gene expres- at different days. The horizontal bars denote M1 and M2, which sion. To explore this, we employed chromatin immunopreci- deﬁne the ranges of IL2R negative and positive cells, respectively. 6C2 cells untransfected with the IL2R construct were used to pitation (ChIP) assays to assess the kinetics of two major determine the IL2R negative cells zone on the histograms. (B) histone modiﬁcations associated with chromatin transcrip- Quantitative RT–PCR of a 13-copy cell line at days 0, 3, 6, 9, 13, tional states, acetylation of histones H3 and H4 and K9 18 and 20. Each point on the graph represents the amount of IL2R dimethylation of histone H3, during transgene silencing. At mRNA relative to the folate receptor mRNA per copy of the different time points (days 0 and 1, 5, 7, 10, 13, 17, 19 and 74 transgene. IL2R mRNA at day 0 is set to 100. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 139 | | Relative quantity of IL2R mRNA Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Day 0/Hygro Day 2 Day 4 100 100 100 80 80 80 60 60 40 40 40 20 20 20 0 0 Day 6 Day 12 Day 15 100 100 80 80 60 60 40 40 20 20 0 0 Day 19 Day 60 Day 90 100 100 100 80 80 80 60 60 60 40 40 40 20 20 20 0 0 0 CpG dinucleotide sites TATA CpG 1,2,3 4 5,6 7,8,9, 10,11,12,13,14,15,16,17 18 19,20,21 Nuclease hypersensitive β promoter IL2R cDNA Figure 2 Bisulﬁte genomic sequencing analysis of a multicopy cell line. (A) CpG methylation over the promoter and transcription start site proximal region of the IL2R transgene. Genomic DNA collected from the cells at different time points were bisulﬁte-treated, PCR-ampliﬁed and subcloned. In total, 20 individual clones were sequenced to give the methylation pattern of the 21 CpGs. The percentage of methylated clones for each CpG is plotted. (B) Distribution of CpG dinucleotides over the transgene. A primer set was designed to amplify the region corresponding to the main CpG cluster on the promoter, containing CpG dinucleotides 1–21. The nuclease hypersensitive region corresponds to a part of the chicken b -globin promoter where transcription factors (rectangles and ovals) are known to bind. after selection), cells were crosslinked with low concentra- neighborhood on the transgene, and compared these with tions of formaldehyde (up to 0.3% ﬁnal concentration), the RT–PCR curve for transcription activity. Figure 4 shows which allowed shearing of chromatin into small-sized frag- a kinetic view for different regions within the IL2R construct, ments for high resolution. We designed 10 primer sets and where the maximum values for the IL2R activity and for the TaqMan probes (Figure 3, bottom line, primer sets 1–10) histone and DNA modiﬁcations were taken as 100%. Histone spanning the entire transgene including the promoter, the deacetylation was coupled with gene inactivation and its IL2R cDNA, the splice/polyadenylation signal element and kinetics were similar in the 10 studied regions (primer sets the enhancer. A rapid loss of histone acetylation correlated 1–10) across the IL2R transgene (Figure 4A–D, data not with the extinction of IL2R activity (Figure 3). Histone H3 shown). Except for the region spanned by primer set 3, deacetylation in the promoter and the coding region appeared which showed the fastest kinetics of H3 K9 dimethylation to occur perhaps slightly earlier than histone H4 deacetyla- (but still with a relatively low level of modiﬁcation at day 10), tion. During the same time points, a gradual increase in the rest of the characterized regions had between 15 and 40% histone H3 dimethylation at lysine 9 was also observed. of the maximally modiﬁed H3 K9 nucleosomes at day 10. This However, similar to the DNA cytosine methylation, signiﬁ- modiﬁcation attained its maximum value long after the gene cant dimethylation at H3 K9 occurred after the gene was had been inactivated (days 19 and 74, Figures 3 and 4A–D). repressed. The kinetic curves show that CpG methylation of the We built kinetic curves for all the studied processes—DNA promoter region also occurs later than gene inactivation methylation, histone H3 and H4 acetylation and H3 K9 (Figure 4A–D, data not shown). CpG sites 1, 2 and 6 are methylation—with each curve representing an individual presented in Figure 4A, B and C, respectively. Even CpG site 140 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | EKLF NF-E4 CTF/NF-1 NF-E4 PAL AP-2 10 16 11 17 12 18 10 16 11 17 12 18 12 18 13 19 14 20 10 16 12 18 13 19 14 20 10 16 13 19 14 20 1 7 2 8 3 9 1 7 2 8 1 7 3 9 1 7 2 8 % of methylated clones Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Figure 3 ChIP analysis of histone modiﬁcations in different regions of the IL2R transgene in a multicopy cell line. Cells carrying the transgene were ﬁxed with formaldehyde at different time points and immunoprecipitated with antibodies against acetylated histones H3 and H4, dimethylated histone H3 K9 and nonimmune IgG. Primer sets 1–10 with TaqMan probes spanning the transgene were used to amplify the bound and input DNA. In parallel, these DNA samples were ampliﬁed with primers speciﬁc for the chicken b-globin locus as an internal control (primer sets #5.613 or #10.35, described in Litt et al, 2001b). The values from the ChIPs were corrected by subtraction of the nonspeciﬁc signal derived from the normal rabbit IgG ChIP. We compared the relative abundance of transgene sequences to the chicken internal control sequences in each bound fraction versus the input fraction. The presented diagrams show the fold differences of the analyzed proteins, where the highest relative-difference data point set is equal to 1.0. A map of the IL2R transgene is presented with the positions of the different primer sets used. Positions of the various transcription-factor-binding sites over the promoter are shown in Figure 2. 1, the fastest to be methylated in the entire promoter, became forms of H3 K4 disappear quite early, together with K9 methylated only after inactivation of expression was nearly acetylation. There is in fact some indication that loss of the complete (Figure 4A). Collectively, these data suggest that the dimethyl H3 K4 modiﬁcation may occur slightly earlier than primary mechanism of inactivation does not involve methy- the other modiﬁcations (Figure 5A-C, data not shown), but lation of the promoter DNA or of histone H3 K9, since these the data do not permit a strong conclusion to be drawn. events follow transgene repression. However, the high level The presence of multiple homologous copies of transgenes of DNA methylation of some CpGs in the gene body is not within an array may result in repeat-induced silencing, the changed during the time of gene inactivation or of histone mechanism of which is still unknown (Dorer and Henikoff, deacetylation/methylation (Figure 4D, CpG site 21), consis- 1997). Copy number reduction in transgenic mouse lines tent with the previous observation that gene body DNA caused a marked increase in expression of the transgenes methylation does not interfere with gene activity (Mutskov and less compaction of the chromatin at the transgene locus et al, 2002). (Garrick et al, 1998). We were interested in knowing if a low- There has been considerable interest recently in the asso- copy-number transgenic cell line would have the same ki- ciation of methylation at lysine 4 of histone H3 with tran- netics of chromatin modiﬁcations during inactivation as scriptionally active chromatin (Litt et al, 2001a; Santos-Rosa already seen for the high-copy line. A single or a low copy et al, 2002), and particularly in the difference between the number of IL2R reporters stably integrated in erythroid 6C2 roles of the dimethylated and trimethylated states (see cells were subjected to position effect silencing (Pikaart et al, Discussion). We therefore repeated our kinetic ChIP analysis 1998; data not shown). However, most of these lines (but not making use of antibodies speciﬁc for these two states. As all of them) lost expression at a slower rate as compared to shown in Figure 5A–C (and data not shown), the earliest the cell lines carrying a high copy number (data not shown). events associated with the disappearance of the IL2R tran- We performed kinetic experiments with a two-copy trans- script once again involve histone modiﬁcation at all the sites genic line, which had a gradual extinction of activity in in the promoter and gene body. Both the di- and trimethylated culture (Figures 4E and 6A). At days 28–29, over 95% of &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 141 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld the cells were IL2R inactive (Figure 6A) and, similar to the 13- cating no increased CpG methylation at the HpaII sites) copy line, at this time point the promoter DNA methylation during days 1–29, the time during which transgene inactiva- was relatively low (Figures 4E and 6B,C). We determined the tion occurred in these lines (Figure 6B). Signiﬁcant increases extent of CpG methylation for the low-copy line by HpaII were observed only at days 100 and later. This result was digestion of four sites in the transgene promoter (Figure 6B), conﬁrmed by bisulﬁte genomic sequencing analysis as well as by bisulﬁte analysis (Fig 6C). We obtained similar (Figure 6C). Finally, we compared the kinetics of histone patterns of low-level methyl-CpG-dependent digestion (indi- deacetylation, histone H3 K9 dimethylation and DNA methy- mRN A H3Ac H4Ac H3Me MeCpG 13 copy cell line ⎯ Primer 3, CpG 1 13 copy cell line ⎯ Primer 3, CpG 2 A B 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Days in culture Days in culture 13 copy cell line ⎯ Primer 4, CpG 6 13 copy cell line ⎯ Primer 6, CpG 21 20 20 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Days in culture Days in culture 2 copy cell line ⎯ Primer 3 0 10 20 30 40 50 60 70 80 90 100 Days in culture 142 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | % Change % Change % Change % Change % Change Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld mRNA H3Ac H3K4DiMe H3K4TiMe A B 13 copy cell line ⎯ Primer 1 13 copy cell line ⎯ Primer 3 0 5 10 15 20 25 0 5 10 15 20 25 Days in culture Days in culture C 13 copy cell line ⎯ Primer 6 0 5 10 15 20 25 Days in culture Figure 5 Combined kinetics of disappearance of mRNA, loss of histone H3 acetylation and loss of histone K4 di- and trimethylation in the multicopy cell line. The relative times of occurrence of histone modiﬁcations when compared to loss of gene expression are similar to those observed in (Figure 4A–D). However, there are some differences in the time in culture at which observable silencing commences, possibly reﬂecting differences in the time of incubation before hygromycin selection is removed. (A) Combined kinetics for histone modiﬁcation at promoter region 1 (primer set 1 in Figure 3, bottom line). (B) Kinetic curves for histone modiﬁcation in promoter region 3 (see Figure 3, bottom line). (C) Combined data from the gene body at region 6 (see Figure 3, bottom line). lation with gene activity (Figure 4E). The relationship methylation and cytosine methylation clearly are closely between these chromatin modiﬁcations and gene repression linked (Bird and Wolffe, 1999; Ben-Porath and Cedar, 2001; for the two-copy IL2R cell line was similar to that already Bird, 2002; Richards and Elgin, 2002), but their role in the seen in the multicopy line (compare Figure 4A–C with dynamic transitions between transcriptionally permissive Figure 4E). and transcriptionally silent chromatin remains to be deﬁned. In this study we performed experiments to monitor the kinetics of the chromatin modiﬁcations that accompany the Discussion gradual silencing of a transgene integrated in chicken ery- The four best-characterized hallmarks of repressed chroma- throid cells. The observed reduction in transcript abundance tin, histone H3 and H4 hypoacetylation, histone H3 K9 with time is not likely to reﬂect an alteration in the stability of Figure 4 Kinetics of transgene inactivation, histone modiﬁcations and DNA methylation for multicopy (A–D) and low–copy-number (E) number cell lines. The kinetic curve for the transgene activity in the timing experiment is built based on the level of IL2R mRNA at different days (pink color line, mRNA). The kinetics of changes in histone H3 acetylation, H4 acetylation and H3 K9 dimethylation are made using the ChIP data, and are indicated as H3Ac (green color line), H4Ac (red color line), and H3Me (blue color line), respectively. The kinetics of methylation of an individual CpG site on the transgene accompanying the extinction of IL2R expression are presented as MeCpG (orange color line). The maximum values of mRNA level, histone H3 and H4 acetylation, histone K9 H3 methylation and individual CpG methylation are set to 100. Multicopy cell line: (A) Combined kinetics for histone modiﬁcation at promoter region 3 (primer set #3 in Figure 3, bottom line) and methylation of CpG site 1 lying in the same region. (B) Combined data from the same region 3 but with the next CpG site: CpG site 2. (C) Kinetics for another region from the transgene promoter: primer set #4 (see Figure 3, bottom line) and CpG site 6. (D) Kinetics of DNA methylation of CpG site 21 from the gene body and the histone modiﬁcations kinetics at this region #6 (see Figure 3, bottom line) . Low-copy- number cell line: (E) Kinetic curves for IL2R activity and histone modiﬁcation in the promoter region #3. Changes in the DNA methylation state of the transgene promoter over time were determined by Southern blotting in Figure 5. Phosphor Imager analysis was performed to quantitate the intensity of the three bands on the blot for each time point. The band ‘a’ intensity is compared to the sum of the intensities of bands ‘a’,‘b’ and ‘c’. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 143 | | % Change % Change % Change Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Day 0/Hygro Day 8 Day 16 Day 23 Day 28 Day 100 β Promoter B IL2R β/ε Enhancer polyA CpG Probe c b a HpaII XbaI MMMM M Southern blotting Days in culture 0 /H, 1, 6, 11 , 21 , 29, 100 ,160 Bisulfite genomic sequencing analysis Day 1 Day 11 Day 21 100 100 80 80 60 60 40 40 20 20 0 0 Day 29 Day 100 Day 160 100 100 80 80 60 60 40 40 20 20 0 0 CpG dinucleotide sites Figure 6 Extinction of IL2R transgene activity and accompanying DNA methylation of a low-copy-number cell line. (A) FACS analysis of IL2R activity at day 0 (in the presence of hygromycin) and at days 8, 16, 23, 28 and 100 after removal of hygromycin. (B) Southern blots at different days in culture to determine the extent of DNA methylation at HpaII restriction sites in the transgene. The top diagram shows the positions of recognition sites for the methylation-sensitive enzyme HpaII within the IL2R transgene. Horizontal black lines with asterisks at the ends indicate the three products expected from complete digestion with the HpaII enzyme. The sites protected from digestion due to CpG methylation are indicated as M. The bottom panel shows DNA isolated at different days, digested with the XbaI/HpaII restriction enzymes and analyzed by Southern blotting after separation on an agarose gel. The membranes were hybridized with a BamHI–XbaI probe from the IL2R cDNA, indicated in the top diagram. (C) Bisulﬁte genomic sequencing analysis of a low-copy-number cell line at days 1, 11, 21, 29, 100 and 160. In total, 20 individual clones were sequenced to give the methylation pattern of the 21 CpGs. The percentage of methylated clones for each individual CpG is plotted. the mRNA in transfected cells. As reported in earlier work periods of time after the removal of hygromycin selection (Pikaart et al, 1998), some transgenic 6C2 cell lines, carrying marker, presumably because of effects of the integration site. the identical reporter used here, maintain expression for long Furthermore, these constructs surrounded by insulators ex- 144 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | 1 7 2 8 3 9 11 17 12 18 14 20 19 1 20 2 19 1 20 2 % of methylated clones Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Figure 7 Schematic representation of the sequence of events leading to silencing of a transgene stably integrated in the genome. Initially, the transgene is expressed and is not affected by the surrounding heterochromatin. The histones over the gene are acetylated (Ac) by HAT and the DNA is not methylated at the promoter region (open circles). After a period of propagation in culture, histone deacetylase activity (HDAC) is recruited to the transgene and the loss of histone acetylation is accompanied by transcription inactivation. Histone H3 that is di- or trimethylated at lysine 4 is also lost at this time. At the next step histone H3 K9 and DNA methyltransferases (HMTand DNMT) methylate their substrates (M for the histones and ﬁlled circles for the CpGs) in the already inactive transgene and form a more stable silenced state. Propagation of the heterochromatic state by heterochromatin protein 1 (HP1) and the binding of methyl-CpG-binding proteins (MBD) may ‘lock’ the silenced transgene. press at high and constant levels for long periods of time gene silencing or is in some cases a secondary modiﬁcation (Pikaart et al, 1998). affecting only the genes that have already been inactivated. Hypermethylation of promoter CpG islands is involved in The relationship of CpG methylation to gene silencing is an the epigenetic silencing of genes on the inactivated X-chro- issue that has been raised by many investigators. In the case mosome, imprinted genes, tumor suppressor genes and exo- of the transgene silencing mechanism we describe here, the genous integrated genes (Bird and Wolffe, 1999; Jones and observation that promoter methylation occurs later than Baylin, 2002). However, most of these observations were transcription inactivation implies that this epigenetic mark made on already repressed genes, leaving open the question is involved in maintaining the inactive state of the chromatin of whether cytosine methylation is always a primary agent of rather than initiating it. &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 145 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld Other inactivation pathways scriptional activity and DNA methylation over the coding Under other circumstances, for transgenes in which the DNA region. had been methylated in vitro before transfection, methylation It appears from our data that dimethylation of H3 K9 lies has been shown to serve an initiating function in the forma- downstream of histone deacetylation and transgene inactiva- tion of repressed chromatin, leading to H3/H4 hypoacetyla- tion. This is not surprising because H3K9 can be either tion (Eden et al, 1998; Schubeler et al, 2000) and H3 K9 acetylated or methylated and, in the process of inactivation, methylation (Santoro et al, 2002) after their integration into deacetylation must happen ﬁrst. Our observations of the the genome. Both methylation-dependent and -independent timing of modiﬁcations over the gene promoter are not mechanisms of repression have been observed in the case of inconsistent with some recent data on the timing of X- retroviral expression in embryonic stem cells (Cherry et al, chromosome inactivation (Heard et al, 2001). The chromo- 2000; Pannell et al, 2000). In another example where gene some-wide enrichment of K9 H3 dimethylation is thought to silencing is independent of DNA methylation, it has been be an early event, and marks a ‘hotspot’ important for the INK4a found that the tumor suppressor gene p16 can be initial steps in the X-inactivation process. However, a ChIP silenced in cell lines carrying a double knockout of DNMT1 experiment showed that the promoter regions of the X- and DNMT3b (DNMT: DNA methyltransferase). These muta- inactivated genes MeCP2 and G6pd were enriched in di- tions reduce genomic DNA methylation by 495%, and result methyl-K9 H3 later (by day 5 of differentiation) (Heard et al, in demethylation and initial loss of silencing of the gene 2001), indicating a difference between chromosome-wide INK4a (Rhee et al, 2002). However, the p16 gene is re-silenced compared to promoter-speciﬁc methylation on the X-chromo- with continued passage of the cells despite the absence of some. DNA methylation (Bachman et al, 2003). One of the goals of this investigation was to determine The situation with the genes on the inactivated X-chromo- whether an increase in H3 K9 methylation, which is usually some is less clear. Early kinetic studies showed that the associated with silencing, preceded or followed inactivation hypoxanthine phosphoribosyltransferase (Hprt) gene is silent of the transgene. It is possible that H3 K9 methylation is on the inactive X-chromosome before DNA methylation inhibited by methylation of lysine 4 at the same molecule. occurs in intron 1 sequences. This was determined by diges- This interplay has been shown by in vitro experiments to be tion with cytosine methylation-sensitive restriction endonu- speciﬁc for Suv39h1 histone methyltransferase (Wang et al, cleases (Lock et al, 1987). However, the kinetic analyses of 2001; Nishioka et al, 2002), but not for other H3 K9 methyl- the CpG island in the same Hprt gene using the bisulﬁte transferases: G9a (Nishioka et al, 2002) and SETDB1 (Schultz genomic sequencing method revealed that methylation oc- et al, 2002). However, we have also examined in separate curred at the initiation of X-chromosome inactivation (Park experiments the parallel changes in H3 K4 methylation. This and Chapman, 1997). modiﬁcation has been associated with transcriptionally ac- tive regulatory regions (Litt et al, 2001a) and coding regions (Litt et al, 2001a; Bernstein et al, 2002). As shown in Figure 5, loss of the dimethylated and trimethylated K4 modiﬁcations Histone H3 methylation occurs early in the inactivation process. This is certainly Methylation of histone H3 K9 has so far been implicated in consistent with the fact (but does not prove) that K4 must heterochromatin repression, promoter regulation and propa- be demethylated before K9 is methylated. The dimethylated gation of repressed chromatin (Litt et al, 2001a; Nielsen et al, 2001; Fahrner et al, 2002; Nguyen et al, 2002; Saccani and and trimethylated K4 modiﬁcations are lost at similar rates Natoli, 2002; Kondo et al, 2003). Our observation that histone over both the promoter and coding region of the IL2R reporter H3 K9 dimethylation occurs over the IL2R reporter (Figures 3 (Figure 5A–C, data not shown). and 4) also demonstrates that histone methylation is asso- ciated with silencing of the stably integrated transgenic Histone modiﬁcation and gene silencing constructs in eukaryotic cells. The early loss of methyl groups from H3 K4, and of acetyl This raises the question of whether DNA methylation groups from lysines on both H3 and H4 suggest that these in mammalian cells is dependent on the activity of histone modiﬁcations may play an important role in switching off methyltransferases (HMTs), as has been shown in gene transcription. It is important to emphasize that our data Neurospora and Arabidopsis (Tamaru and Selker, 2001; are consistent with such a role, but do not prove it. Our Jackson et al, 2002). It is known that DNMT3 proteins earlier results (Mutskov et al, 2002) showed that silencing of colocalize with HP1 (which binds to methylated histones) the transgenic cell lines carrying the same reporter used here at the heterochromatic loci in embryonic stem cells could be inhibited by growing the cells constantly in the (Bachman et al, 2001); however, a direct site-speciﬁc role presence of Trichostatin A, which blocks histone deacetyla- for histone methylation in the recruitment of DNMTs has not tion. In contrast, the growth of cells in 5-aza-cytidine, an been established. The kinetic data presented here do not inhibitor of DNA methylation, did not affect the rate of allow us to determine whether histone methylation over the inactivation, suggesting that silencing of the transgene was promoter precedes DNA methylation. Both occur, within our not controlled by this modiﬁcation (Mutskov et al, 2002). limits of resolution, at about the same time. We note however Thus, maintenance of acetylation is at least sufﬁcient to that the coding region DNA of the IL2R transgene is largely prevent inactivation. It has been established that speciﬁc methylated at early times despite the fact that H3 K9 di- histone acetylation patterns mediate regulatory factor inter- methylation levels are low over the same region (Figure 4D; actions critical for gene activity (Cosma, 2002; Emerson, Mutskov et al, 2002). This is connected to the fact that the 2002; Turner, 2002). The constant recruitment of proteins IL2R transgene is still active at that point; as shown earlier that carry histone acetyltransferase (HAT) activity and main- (Mutskov et al, 2002), there is no correlation between tran- tain hyperacetylated histones can help prevent switching to 146 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld line, obtained by transformation of chicken bone marrow with a the inactive state. Insulator elements such as the HS4 chicken wild-type avian erythroblastosis virus (AEV). 6C2 cells were grown b-globin insulator also help maintain a high level of histone as described (Boyes and Felsenfeld, 1996) and stably transfected acetylation over the protected region (Pikaart et al, 1998; with the reporters in the presence of DNA fragments encoding the Mutskov et al, 2002). The promoter of the gene adjacent to hygromycin-resistant gene derived from the pREP 7 plasmid (Invitrogen), by electroporation (Boyes and Felsenfeld, 1996), or the insulator is more accessible to transcription factors, and with effectene transfection reagent (Qiagen) according to the the binding of these factors in turn protects the promoter manufacturer’s protocol. Individual hygromycin-resistant colonies DNA from methylation (Mutskov et al, 2002). Immunity to were picked after 2–3 weeks of culture in a-MEM plus 2% methocel DNA methylation caused by active chromatin was also and 2000 U/ml hygromycin (Calbiochem), and expanded in hygromycin-containing medium (1250 U/ml) for a further 6 days. described for locus control regions (Santoso et al, 2000) IL2R expression was monitored by FACS analysis as described and enhancer elements (Schubeler et al, 2000; Mutskov (Pikaart et al, 1998) using a FACSCalibur immunocytometer (Becton et al, 2002), which are tissue-speciﬁc centers for high levels Dickinson). Cells were maintained in log-phase growth for all of histone acetylation. experiments. Our parallel kinetic studies of the changes in transgene Determination of transgene copy number in stably expression, histone acetylation and methylation, and DNA transfected cell lines methylation suggest a sequence of events in which loss of Genomic DNA from the stably transfected 6C2 cell lines was histone acetylation and H3 K4 di- and trimethylation are early isolated by a standard procedure (Sambrook et al, 1989) and series of two-fold dilutions were made. Equal amounts of DNA from each steps in the sequence of events leading to transgene silencing point of the double dilutions were analyzed by real-time PCR, in (Figure 7, step 1). At this stage the gene is repressed, but still parallel with two different primer sets with a very similar efﬁciency reversibly, because transcription can be partially reactivated of ampliﬁcation. The ﬁrst set of primers, including a TaqMan probe, by inhibition of histone deacetylases (TSA) (Pikaart et al, was speciﬁc for the IL2R transgene (primer set IL2R2800 described in Mutskov et al, 2002) and the second set ampliﬁed the 1998). Methylation of histone H3 K9 and of CpG sites on endogenous folate receptor gene (primer set FolateExon 4, promoter DNA are later events (Figure 7, step 2). These two described in Mutskov et al, 2002). The number of integrated chromatin modiﬁcations may target HP1 and MBD-protein transgenes was estimated by comparison of the IL2R signal to the repressor complexes that could ‘lock’ this stable silenced signal from one copy of the endogenous gene (half of the total folate receptor gene signal). As a control, the same experiment was chromatin state (Figure 7, step 3); in any case, the presence performed with genomic DNA isolated from nontransfected 6C2 of promoter CpG methylation almost certainly implies inabil- cells. ity to return to an active state. Moreover, histone and DNA methylation marks may signal to one another to ensure Quantitative TaqMan RT–PCR assay propagation of this more compact form of chromatin. We Total RNA from transfected 6C2 cells was isolated at different time points (see the Results section) and reverse transcribed as described previously observed binding of MBD proteins to the silenced previously (Mutskov et al, 2002). A series of two-fold dilutions of IL2R transgene at late stages (Mutskov et al, 2002), which the cDNA products were made, and equal amounts of DNA were could reinforce the histone modiﬁcation patterns via recruit- ampliﬁed by real-time PCR using speciﬁc primers and TaqMan ment of HDACs (Bird and Wolffe, 1999) and H3 K9 histone probes for the IL2R transgene and the folate receptor gene as described (Mutskov et al, 2002). methylase activity (Fuks et al, 2003). Although our speciﬁc results may serve as a more general Southern blot hybridization model for the silencing of transgenes stably integrated into a Isolation of high molecular weight genomic DNA was performed by cell or animal, we do not believe that they necessarily provide a standard method (Sambrook et al, 1989). In total, 10 mg of each DNA sample was subjected to double digestion with the restriction a universal mechanism for the regulation of endogenous endonucleases XbaI and the DNA methylation-sensitive HpaII. The genes, in which the order of events for gene silencing is digested DNA was resolved on an agarose gel, blotted on to likely to be under much stricter control. For example, the þ Hybond-N nylon membrane (Amersham) and hybridized with an endogenous chicken embryonic r-globin gene remains asso- IL2R probe in QuikHyb solution (Stratagene) according to the manufacturer’s protocol. ciated with hyperacetylated histones at day 15 of differentia- tion, even though this gene starts to be inactivated after day 5 Bisulﬁte genomic sequencing and transcription is barely detected at day 15 (Hebbes et al, Bisulﬁte conversion of DNA was carried out by the method 1992; Litt et al, 2001b). Recent results have indicated that developed by Clark et al (1994) with minor modiﬁcations described previously (Mutskov et al, 2002). Each bisulﬁte-modiﬁed DNA transcription activation in eukaryotes requires the recruit- sample was subjected to PCR reactions using primers speciﬁc for ment of chromatin remodeling and histone modiﬁcation the bisulﬁte-converted sequence of the IL2R transgene (primer set factors, but the order and the recruitment timing are gene- C1–C2, described in Mutskov et al, 2002). Ampliﬁcation, PCR speciﬁc events (Cosma, 2002; Emerson, 2002). Our results product puriﬁcation and subcloning were carried out as described show however that promoter DNA methylation and histone (Mutskov et al, 2002). Individual clones were sequenced using M13 reverse or M13 forward primers and the BigDye terminator cycle H3 K9 dimethylation are not early events in the silencing of a sequencing kit (PE Applied Biosystems) and applied to an ABI transgene. Extinction of expression is instead correlated with PRISM 310 DNA sequencer (Perkin Elmer). loss of histone acetylation and of H3 K4 methylation. It seems reasonable to focus future studies on one or both of these Formaldehyde crosslinking and ChIP Formaldehyde crosslinking in vivo, sonication of the chromatin early events as primary causes of silencing. and chromatin immunoprecipitation assays were performed as described previously (Mutskov et al, 2002). At different time points of our experiment (see the Results section), 6C2 cells stably Materials and methods transfected with the IL2R transgene were ﬁxed with 0.3% of formaldehyde at room temperature for 8 min, followed by incuba- Stable transfection in cell culture and FACS analysis tion at 41C for another 30 min. After isolation of nuclei, sonication The IL2R reporter construct under the control of the chicken b - and lysis with 1% SDS, the soluble chromatin was pre-cleared with globin promoter and b/e enhancer has been described previously salmon sperm DNA/protein A agarose (Upstate Biotechnology) (Pikaart et al, 1998). 6C2 is a CFU-E stage erythroid precursor cell before the immunoprecipitation procedure. Different antibodies &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 147 | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld purchased from Upstate Biotechnology were used for ChIP: For each primer set, these fold difference values were corrected by acetylated histone H3 antibody, acetylated histone H4 antibody, subtraction of the nonspeciﬁc signal derived from the nonimmune dimethyl histone H3 (K9) antibody, dimethyl histone H3 (K4) rabbit IgG ChIP (t ): antibody, trimethyl histone H3 (K4) antibody as well as normal rabbit IgG in the control experiments. t t ðIP=InÞ ðIP=InÞ : Quantitative real-time PCR In parallel, DNA samples were ampliﬁed with primers speciﬁc for DNA samples from input (In) and antibody-bound (IP) chromatin the chicken b-globin locus as an internal control (c) (primer sets were analyzed by real-time PCR using the TaqMan Universal PCR #5.613 or #10.35, described in Litt et al, 2001b). Finally, we Master Mix (PE Applied Biosystems) and an ABI Prism 7700 normalized the relative abundance of transgene sequences to the sequence detector according to the manufacturer’s protocols. In chicken internal control sequence for each individual chromatin total, 10 primers and TaqMan probes were selected from the IL2R immunoprecipitation reaction using the following formula: transgene. Their sequences have been described previously (Mutskov et al, 2002): primer set #1 as IL2R1060, primer set #2 as t t c c 0 0 IL2R1420, primer set #3 as IL2R1900, primer set #4 as IL2R2020, ½ðIP=InÞ ðIP=InÞ =½ðIP=InÞ ðIP=InÞ : primer set #5 as IL2R2140, primer set #6 as IL2R2560, primer set #7 as IL2R2800, primer set #8 as IL2R3160, primer set #9 as IL2R3640, and primer set #10 as IL2R3880. Individual PCRs were carried out in triplicate to control for PCR variation and the C values were collected. Data quantiﬁcation was performed by applying the Acknowledgements comparative C method, as described by the manufacturer’s protocols (PE Applied Biosystems), giving the fold difference of a We thank Timur Yusufzai, Catherine Farrell and Adam West for target sequence (t) in the IP fraction versus a ﬁxed amount of In many stimulating discussions and for a critical reading of this DNA as a standard, where manuscript. We also acknowledge members of the Felsenfeld laboratory for useful suggestions, and Michael Krause’s laboratory DDC t ðC ðIPÞC ðInÞÞ t t IP=In ¼ 2 ¼ 2 : for DNA sequencing. References Bachman KE, Park BH, Rhee I, Rajagopalan H, Herman JG, Baylin Gallarda JL, Foley KP, Yang ZY, Engel JD (1989) The beta-globin SB, Kinzler KW, Vogelstein B (2003) Histone modiﬁcations and stage selector element factor is erythroid-speciﬁc promoter/ silencing prior to DNA methylation of a tumor suppressor gene. enhancer binding protein NF-E4. Genes Dev 3: 1845–1859 Cancer Cell 3: 89–95 Garrick D, Fiering S, Martin DI, Whitelaw E (1998) Repeat-induced Bachman KE, Rountree MR, Baylin SB (2001) Dnmt3a and gene silencing in mammals. Nat Genet 18: 56–59 Dnmt3b are transcriptional repressors that exhibit unique Heard E, Rougeulle C, Arnaud D, Avner P, Allis CD, Spector DL localization properties to heterochromatin. J Biol Chem 276: (2001) Methylation of histone H3 at Lys-9 is an early mark on the 32282–32287 X chromosome during X inactivation. Cell 107: 727–738 Ben-Porath I, Cedar H (2001) Epigenetic crosstalk. Mol Cell 8: Hebbes TR, Thorne AW, Clayton AL, Crane-Robinson C (1992) 933–935 Histone acetylation and globin gene switching. Nucleic Acids Bernstein BE, Humphrey EL, Erlich RL, Schneider R, Bouman P, Liu Res 20: 1017–1022 JS, Kouzarides T, Schreiber SL (2002) Methylation of histone H3 Jackson JP, Lindroth AM, Cao X, Jacobsen SE (2002) Control of Lys 4 in coding regions of active genes. Proc Natl Acad Sci USA CpNpG DNA methylation by the KRYPTONITE histone H3 99: 8695–8700 methyltransferase. Nature 416: 556–560 Bird A (2002) DNA methylation patterns and epigenetic memory. Jones PA, Baylin SB (2002) The fundamental role of epigenetic Genes Dev 16: 6–21 events in cancer. Nat Rev Genet 3: 415–428 Bird AP, Wolffe AP (1999) Methylation-induced repression-belts, Kondo Y, Shen L, Issa JP (2003) Critical role of histone methylation braces, and chromatin. Cell 99: 451–454 in tumor suppressor gene silencing in colorectal cancer. Mol Cell Boyes J, Felsenfeld G (1996) Tissue-speciﬁc factors additively Biol 23: 206–215 increase the probability of the all-or-none formation of a hyper- Litt MD, Simpson M, Gaszner M, Allis CD, Felsenfeld G (2001a) sensitive site. EMBO J 15: 2496–2507 Correlation between histone lysine methylation and developmen- Cherry SR, Biniszkiewicz D, van Parijs L, Baltimore D, Jaenisch R tal changes at the chicken beta-globin locus. Science 293: (2000) Retroviral expression in embryonic stem cells and hema- 2453–2455 topoietic stem cells. Mol Cell Biol 20: 7419–7426 Litt MD, Simpson M, Recillas-Targa F, Prioleau M-N, Felsenfeld G Clark SJ, Harrison J, Paul CL, Frommer M (1994) High sensitivity (2001b) Transitions in histone acetylation reveal boundaries mapping of methylated cytosines. Nucleic Acids Res 22: of three separately regulated neighboring loci. EMBO J 20: 2990–2997 2224–2235 Cosma MP (2002) Ordered recruitment: gene-speciﬁc mechanism of Lock LF, Takagi N, Martin GR (1987) Methylation of the Hprt gene transcription activation. Mol Cell 10: 227–236 on the inactive X occurs after chromosome inactivation. Cell 48: Dorer DR, Henikoff S (1997) Transgene repeat arrays interact with 39–46 distant heterochromatin and cause silencing in cis and trans. Mutskov VJ, Farrell CM, Wade PA, Wolffe AP, Felsenfeld G (2002) Genetics 147: 1181–1190 The barrier function of an insulator couples high histone acetyla- Eden S, Hashimshony T, Keshet I, Cedar H, Thorne AW (1998) DNA tion levels with speciﬁc protection of promoter DNA from methy- methylation models histone acetylation. Nature 394: 842 lation. Genes Dev 16: 1540–1554 Emerson BM (2002) Speciﬁcity of gene regulation. Cell 109: 267–270 Nguyen CT, Weisenberger DJ, Velicescu M, Gonzales FA, Lin JC, Emerson BM, Lewis CD, Felsenfeld G (1985) Interaction of speciﬁc Liang G, Jones PA (2002) Histone H3-lysine 9 methylation is nuclear factors with the nuclease-hypersensitive region of the associated with aberrant gene silencing in cancer cells and chicken adult beta-globin gene: nature of the binding domain. is rapidly reversed by 5-Aza-2 -deoxycytidine. Cancer Res 62: Cell 41: 21–30 6456–6461 Fahrner JS, Herman JG, Baylin SB (2002) Dependence of histone Nielsen SJ, Schneider R, Bauer UM, Bannister AJ, Morrison A, modiﬁcations and gene expression on DNA hypermethylation in O’Carroll D, Firestein R, Cleary M, Jenuwein T, Herrera RE, cancer. Cancer Res 62: 7213–7218 Kouzarides T (2001) Rb targets histone H3 methylation and Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T (2003) The HP1 to promoters. Nature 412: 561–565 methyl-CpG-binding protein MeCP2 links DNA methylation to Nishioka K, Chuikov S, Sarma K, Erdjument-Bromage H, Allis CD, histone methylation. J Biol Chem 278: 4035–4040 Tempst P, Reinberg D (2002) Set9, a novel histone H3 methyl- 148 The EMBO Journal VOL 23 NO 1 2004 &2004 European Molecular Biology Organization | | Timing of chromatin modiﬁcations in transgene silencing V Mutskov and G Felsenfeld transferase that facilitates transcription by precluding histone tail Santoro R, Li J, Grummt I (2002) The nucleolar remodeling complex modiﬁcations required for heterochromatin formation. Genes Dev NoRC mediates heterochromatin formation silencing of ribosomal 16: 479–489 gene transcription. Nat Genet 32: 393–396 Pannell D, Osborne CS, Osborne CS, Yao S, Sukonnik T, Pasceri P, Santoso B, Ortiz BD, Winoto A (2000) Control of organ-speciﬁc Karaiskakis A, Okano M, Li E, Lipshitz HD, Ellis J (2000) demethylation by an element of the T-cell receptor-alpha locus Retrovirus vector silencing is de novo methylase independent control region. J Biol Chem 275: 1952–1958 and marked by a repressive histone code. EMBO J 19: 5884–5894 Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J, Bernstein BE, Park JG, Chapman VM (1997) DNA methylation patterns and Emre NC, Schreiber SL, Mellor J, Kouzarides T (2002) Active sequence transitions of the CpG island of mouse Hprt during genes are tri-methylated at K4 of histone H3. Nature 419: 407–411 early embryogenesis. Mol Cells 7: 259–263 Schubeler D, Lorincz MC, Cimbora DM, Telling A, Feng YQ, Pikaart MJ, Recillas-Targa F, Felsenfeld G (1998) Loss of transcrip- Bouhassira EE, Groudine M (2000) Genomic targeting of methy- tional activity of a transgene is accompanied by DNA methylation lated DNA: inﬂuence of methylation on transcription, replication, and histone deacetylation and is prevented by insulators. Genes chromatin structure, and histone acetylation. Mol Cell Biol 20: Dev 12: 2852–2862 9103–9112 Prioleau MN, Nony P, Simpson M, Felsenfeld G (1999) An insulator Schultz DC, Ayyanathan K, Negorev D, Maul GG, Rauscher III FJ element and condensed chromatin region separate the chicken 2002 SETDB1: a novel KAP-1-associated histone H3, lysine 9- beta-globin locus from an independently regulated erythroid- speciﬁc methyltransferase that contributes to HP1-mediated si- speciﬁc folate receptor gene. EMBO J 18: 4035–4048 lencing of euchromatic genes by KRAB zinc-ﬁnger proteins. Rhee I, Bachman KE, Park BH, Jair KW, Yen RW, Schuebel KE, Cui Genes Dev 16: 919–932 H, Feinberg AP, Lengauer C, Kinzler KW, Baylin SB, Vogelstein B Tamaru H, Selker EU (2001) A histone H3 methyltransferase con- (2002) DNMT1 and DNMT3b cooperate to silence genes in trols DNA methylation in Neurospora crassa. Nature 414: 277–283 human cancer cells. Nature 416: 552–556 Tamaru H, Zhang X, McMillen D, Singh PB, Nakayama J, Grewal SI, Richards EJ, Elgin SC (2002) Epigenetic codes for heterochromatin Allis CD, Cheng X, Selker EU (2003) Trimethylated lysine 9 of formation and silencing: rounding up the usual suspects. Cell 108: histone H3 is a mark for DNA methylation in Neurospora crassa. 489–500 Nat Genet 34: 75–79 Saccani S, Natoli G (2002) Dynamic changes in histone H3 Lys 9 Turner BM (2002) Cellular memory and the histone code. Cell 111: methylation occurring at tightly regulated inducible inﬂammatory 285–291 genes. Genes Dev 16: 2219–2224 Wang H, Cao R, Xia L, Erdjument-Bromage H, Borchers C, Tempst P, Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Zhang Y (2001) Puriﬁcation and functional characterization of Laboratory Manual, 2nd edn. Plainview, NY: Cold Spring Harbor a histone H3-lysine 4-speciﬁc methyltransferase. Mol Cell 8: Laboratory Press 1207–1217 &2004 European Molecular Biology Organization The EMBO Journal VOL 23 NO 1 2004 149 | |

Journal

The EMBO Journal – Springer Journals

Published: Jan 14, 2004

Keywords: chromatin; DNA methylation; histone acetylation; histone methylation; transgene silencing

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9

References (48)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies