This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Padjen, K. Articles by Storb, U. Search for Related Content PubMed PubMed Citation Articles by Padjen, K. Articles by Storb, U.

DNA Methylation Precedes Chromatin Modifications under the Influence of the Strain-Specific Modifier Ssm1

Department of Molecular Genetics and Cell Biology, University of Chicago, Chicago, Illinois 60637

Received 9 February 2005/ Returned for modification 23 February 2005/ Accepted 3 March 2005

	ABSTRACT

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ssm1 is responsible for the mouse strain-specific DNA methylation of the transgene HRD. In adult mice of the C57BL/6 (B6) strain, the transgene is methylated at essentially all CpGs. However, when the transgene is bred into the DBA/2 (D2) strain, it is almost completely unmethylated. Strain-specific methylation arises during differentiation of embryonic stem (ES) cells. Here we show that Ssm1 causes striking chromatin changes during the development of the early embryo in both strains. In undifferentiated ES cells of both strains, the transgene is in a chromatin state between active and inactive. These states are still observed 1 week after beginning ES cell differentiation. However, 4 weeks after initiating differentiation, in B6, the transgene has become heterochromatic, and in D2, the transgene has become euchromatic. HRD is always expressed in D2, but in B6, it is expressed only in early embryos. The transgene is already more methylated in B6 ES cells than in D2 ES cells and becomes increasingly methylated during development in B6, until essentially all CpGs in the critical guanosine phosphoribosyl transferase core are methylated. Clearly, DNA methylation of HRD precedes chromatin compaction and loss of expression, suggesting that the B6 form of Ssm1 interacts with DNA to cause strain-specific methylation that ultimately results in inactive chromatin.

	INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

 
Rather than being a passive scaffold for gene expression, chromatin is closely integrated with gene function. Although methylation of DNA and covalent modification of histone proteins are both modifications known to affect gene expression, little is known about how chromatin modification patterns are established during development (23). Ssm1, the strain-specific modifier (of DNA methylation), is a murine locus which controls the CpG methylation of a transgene (HRD) in a strain-specific (7) and developmentally regulated manner (34). While DNA methylation is known to be associated with silenced genes, or heterochromatic loci (3), besides Ssm1, so far there appears to be no other locus known to be involved in determining which genomic regions become methylated during development. Therefore, Ssm1 presents a unique opportunity to study how initial DNA methylation and chromatin patterns are established and how these epigenetic patterns lead to alterations in gene function.

Ssm1 controls the DNA methylation of a defined transgenic target, HRD (Fig. 1). By use of previously characterized recombinant inbred mice, Ssm1 was mapped to the distal end of chromosome 4 (7). Further mapping has placed Ssm1 in a small interval near Nppa (10; P. Engler and U. Storb, unpublished data). Under the influence of Ssm1, the HRD transgene shows a dramatic methylation phenotype that is dependent upon strain background. In 12 independent transgenic lines, HRD is highly methylated when carried in the C57BL/6 strain or the (B6 x D2)F₁ strain, indicating that transgene methylation is both integration site and copy number independent and acts in a dominant manner (7). All the experiments in this text use the (B6 x D2)F₁ (B6) background. In contrast, when HRD is carried in the DBA/2 (D2) background, the transgene is almost free of methylation (7). About half of the laboratory mouse strains tested show the methylating phenotype; the other half do not methylate the HRD transgene (7; Engler and Storb, unpublished). The hypermethylation characteristic of HRD in the B6 background extends throughout the integrated transgene and even some distance into the flanking genomic regions (6). Deletion mutants of the transgene have shown the guanine phosphoribosyl transferase (GPT) region to be critical for strain-specific methylation. Without this region, the transgene shows an intermediate level of methylation regardless of the mouse strain (6).

View larger version (8K): [in this window] [in a new window]   FIG. 1. The HRD transgene (3.6 kb) contains multiple CpGs that can be methylated. Most of these are located within the GPT region. Short vertical bars, HpaII sites; horizontal arrows, approximate locations of PCR primers used for amplification of immunoprecipitated chromatin; Eµ, enhancer (intronic) of Ig heavy chain genes; PMT, promoter of metallothionein gene; V and J, Ig recombination signal sequences; AUG, Kozak translation initiation consensus sequence; GPT, guanine phosphoribosyl transferase of E. coli; splice, poly(A) sites of SV40. BamHI sites B1, B2, and B3 are indicated.

Analysis of the timing of the Ssm1 effect showed that postimplantation murine embryos acquire strain-specific methylation of HRD prior to embryonic day 6.5 (34). The clear strain difference in methylation is restricted to the tissue of the embryo proper; murine trophoblast tissues reveal a hemimethylated phenotype regardless of strain background (34). Also, experiments using ES cells from both D2 and B6 strains showed that all the D2 ES lines have only low, partial methylation of the HRD transgene before and after differentiation (34). However, the undifferentiated B6 ES cell lines showed a variety of DNA methylation patterns, from low (as in the D2 lines) to almost complete, suggesting that DNA methylation under Ssm1 control is initiated during the blastocyst stage. After differentiation, all B6 lines showed almost complete methylation.

DNA methylation has been shown to recruit proteins that bind to the methylated CpG dinucleotides through specific domains (2). The methylated DNA binding proteins (such as MeCP2) can in turn interact with histone deacetylases (25), as well as cause histone H3 lysine-9 (K9) methylation (13). These interactions link CpG methylation to a repressive chromatin state. Presumably, multiple methylated DNA binding proteins are required for complete repression, since the deletion of MeCP2 causes only a subtle increase in the gene expression level (15, 31). However, CpG methylation may be a secondary event, since methylation of K9 at histone H3 can lead to the methylation of DNA (17, 30). It has been postulated that DNA methylation may reinforce chromatin-induced gene silencing by providing an easy mechanism of propagation of the silencing mark in each cell cycle.

Thus, one explanation for the function of Ssm1 is that it may alter early chromatin structure in a time- and development-specific manner, which ultimately results in DNA methylation. Conversely, Ssm1 would encode some form of methyltransferase or a factor that directly affects methylation patterns. In this study, we compared the chromatin structure, expression, and DNA methylation of the HRD transgene at various stages of mouse development. We found that DNA methylation is strongly enhanced by Ssm1 days before any strain-specific inactivation of chromatin or expression becomes apparent.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

 
Transgenic mice. D2 and B6 strain mice transgenic for the HRD construct were used for all chromatin immunoprecipitations (ChIPs) from adult tissue. The HRD construct (8) consists (from 5' to 3') of the immunoglobulin (Ig) µ heavy chain enhancer, the mouse metallothionein-1 promoter, Vkappa and Jkappa Ig recombination signal sequences flanking a rat preproinsulin initiation codon, the Escherichia coli xanthine-guanine phosphoribosyl transferase (GPT) coding sequence, and simian virus 40 (SV40) splice and polyadenylation sequences (Fig. 1). The transgenic line 342, used in these studies, contains seven head-to-tail tandem integrations of the HRD transgene. The locus of integration is unknown but is the same for the B6 and D2 mice, since the same transgene was propagated in both strains.

Embryonic stem cells. Embryonic stem (ES) cells were derived from the inner cell mass of D2 and B6 HRD transgenic mice (34). In this study, we used the D2 line awD1 and the B6 line awB2. The latter already has extensive methylation of the HRD transgene, but DNA methylation increases with differentiation (34). To prevent differentiation, the cells were grown at 37°C on gamma-irradiated STO feeder cells in the presence of leukemia inhibitory factor at 10³ U/ml (28). To differentiate the cells, leukemia inhibitory factor was removed and the ES cells were replated without STO feeder cells. Cells were allowed to differentiate into embryoid bodies over periods of 1 to 4 weeks.

Chromatin preparation from adult mice. Chromatin was prepared from the livers of adult mice (4, 20, 21). The livers were dissected and nuclei prepared by Dounce homogenization. The nuclei were fixed in 1% formaldehyde for 10 min at 37°C (cross-linking was quenched by the addition of glycine to 125 mM at room temperature). The nuclear slurry was sonicated and centrifuged (SW50.1 rotor at 31,000 rpm for 20 h at 20°C) through a CsCl gradient (three steps: 1.35 g/ml, 1.5 g/ml, and 1.75 g/ml). The gradient was separated into 10 fractions, and those containing the chromatin were pooled. The cross-linked chromatin was dialyzed against Tris-EDTA overnight in the presence of RNase A (25 ng/µl).

Chromatin preparation from ES cells. Chromatin was prepared from both undifferentiated and differentiated ES cells as described by Upstate Biotechnology, Inc. (Lake Placid, NY). Briefly, chromatin was cross-linked by adding formaldehyde directly to cell culture medium (to a final concentration of 1%) for 10 min at 30°C. Cells were washed with phosphate-buffered saline (containing protease inhibitors), removed from the tissue culture plate by trypsinization, and pelleted by centrifugation. Cells were then lysed in sodium dodecyl sulfate lysis buffer and finally dialyzed to Tris-EDTA in the presence of RNase A (25 ng/µl).

Chromatin immunoprecipitation. Before the ChIP assay was performed, chromatin was broken into fragments from 100 to 500 bp by sonication or micrococcal nuclease digestion. Chromatin IP was performed as described by Upstate Biotechnology, Inc. (Lake Placid, NY). Cross-linked chromatin was diluted and precleared with salmon sperm-protein A agarose. One microgram of a specific antibody was added to the precleared chromatin and incubated at 4°C overnight with rotation (all antibodies were from Upstate Biotechnology, Inc., except anti-Brg1 from Santa Cruz [Santa Cruz, CA]; the anti-methylated H3 K9 and H3 K4 antibodies are against the dimethylated forms [catalog no. 07-212 and no. 07-030, respectively]). The antibody complex was recovered by using salmon sperm-protein A agarose. The histone-DNA-antibody-protein A agarose complex was washed several times, and the histone-DNA complex was eluted with 1% sodium dodecyl sulfate and 0.1 M NaHCO₃. Cross-linking was reversed by heating the chromatin complex to 65°C for 4 h. DNA was purified by proteinase K digestion (at a final concentration of 100 µg/ml) for 1 h at 50°C, followed by phenol-chloroform extraction and ethanol precipitation. The resulting DNA was analyzed by PCR amplification. Input DNA was acquired by purifying DNA from chromatin as described above, only without immunoprecipitation.

PCR assay of immunoprecipitated chromatin. DNA recovered from immunoprecipitation was analyzed by PCR amplification. Three primer pairs were used to amplify three different parts of the HRD transgene (Fig. 1). The PMT product comprises the 3' end of the HRD metallothionein promoter and the 5' end of the Vkappa recombination signal sequence. The GPT PCR product comprises a central portion of the HRD E. coli xanthine-guanine phosphoribosyl transferase sequence. The SV40 product is located in the HRD SV40 polyadenylation region. DNA was purified from immunoprecipitated chromatin and then made into four serial fourfold dilutions. Each of these DNA preparations was then amplified by PCR in parallel in order to accurately determine the amount of immunoprecipitated chromatin. Input DNA was amplified by a method identical to that used for DNA obtained by immunoprecipitation.

Quantification of chromatin immunoprecipitation. Immunoprecipitated DNA was amplified by PCR and then electrophoresed through an 8% polyacrylamide gel. Resulting bands were quantified from at least triplicate ChIP analyses with at least two different chromatin preparations, using the program NIH Image 1.63, downloaded from National Center for Biotechnology Information. Quantification results were normalized to the amount of input DNA and expressed as fractions of input.

Southern blotting. DNA from transgenic mice and ES cells was analyzed for CpG methylation by restriction digestion followed by Southern blotting. HpaII and BamHI digestion was used for methylation analysis. Control digests were performed with MspI and BamHI. Restricted DNA was electrophoresed through 1% agarose gels and then transferred to a Hybond N membrane. The DNA was fixed to the membrane by UV irradiation. Radiolabeled [-³²P]dCTP probe to the HRD GPT region was synthesized by random priming.

RT-PCR for HRD transcripts. RNA was purified from livers, ES cell cultures, and embryos by using the RNeasy kit (QIAGEN, Valencia, CA). The cDNA was made by priming with random hexamers and then amplified by using primers specific for the GPT region of HRD (the same primers used to amplify the GPT region in the chromatin immunoprecipitation assays). STO feeder cells (which do not contain the transgene) and water were used as negative controls. ß-Actin expression was used to normalize the amounts of cDNA in each of the samples. Experiments with controls without reverse transcriptase (RT) were carried out to rule out DNA contamination.

Bisulfite sequencing. To study the methylation patterns of the HRD transgene in the undifferentiated ES cells and adult livers obtained from the B6 and D2 strains of transgenic mice, we used the bisulfite genomic sequencing technique (5). We used the plasmid carrying the HRD transgene, which was originally used to create the HRD transgenic line as a control for the bisulfite technique.

For the PCR amplification, nested PCRs were necessary to produce enough product from the genomic DNA. The outer pair of primers used for the adult B6 and DBA/2 livers was 5'GTATTTTTTAGGGTGGGTTT3' and 5'CACCACTACTCCCATTCATC3', and the inner pair of primers was 5'GTGGTGGATGTTTGGTGGAG3' and 5'CAAAACCCACTCATAAATCC3'. For the undifferentiated cells, we could amplify only the GPT region of the transgene in two halves. The first half was amplified with 5'GTATTTTTTAGGGTGGGTTT3' and 5'AAAACTATTATAACCCACCTAAAAT3' as the outer pair of primers and 5'GTGGTGGATGTTTGGTGGAG3' and 5' TGGTTGTTAATTTTTTTAAT3' as the inner pair of primers. The second half of the transgene was amplified with 5'ATTGATGATTTGGTGGATATT3' and 5'CACCACTACTCCCATTCATC3' as the outer pair of primers and 5'TGTATTTTGTTATTATTTTT3' and 5'CAAAACCCACTCATAAATCC3' as the inner pair of primers. The PCRs in the two halves included a central overlap of 162 nucleotides.

PCR products were electrophoresed on a 1% agarose gel and isolated by using a fragment isolation kit (QIAGEN, Valencia, CA). PCR fragments were subcloned into the TOPO-TA pCR2.1 vector (Invitrogen, Carlsbad, CA) and sequenced by using the M13 forward primer.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

 
The HRD transgene is silenced when carried in adult B6 strain mice relative to D2 mice. Earlier results had shown that the HRD transgenes display complete, integration site-independent DNA methylation when carried in a B6 background (7). However, when in a D2 strain background, the transgene is not methylated. Moreover, under the influence of the strain-specific modifier Ssm1, HRD appears to be functionally silenced in B6 mice and active in D2 mice (8, 9, 11, 33). Given the interrelationship of DNA methylation and chromatin (17, 18, 30), we investigated chromatin modifications of the HRD transgene and the temporal relationship of strain-specific DNA methylation and chromatin modifications.

Chromatin from the livers of adult mice was analyzed by ChIP using antibodies specific to various chromatin modifiers. All experiments were done at least in triplicate, and results were compared to results from experiments with input DNA as well as albumin and ß-actin controls (genes that would not be expected to display any strain-specific difference in chromatin structure or DNA methylation). First, we examined histone H3 K9 methylation. Methylation of this residue is strongly correlated with silenced/heterochromatic regions of the genome in many different systems (22, 26). Following ChIP with an antibody specific to H3 dimethyl-K9, the HRD transgene shows much stronger H3 K9 methylation when carried in a B6 background relative to a D2 background at the PMT, GPT, and SV40 regions (Fig. 1 and 2A). This is interesting not only because it shows that the transgene, when carried in B6 mice, is silenced at the level of chromatin structure and DNA methylation, but also because the chromatin changes appear to cover the whole of the transgene, not merely the area near the regulatory regions, Eµ, and PMT (Fig. 1).

View larger version (19K): [in this window] [in a new window]   FIG. 2. ChIP assays with adult liver. Chromatin from B6 and D2 strain mice was immunoprecipitated with the indicated antibody, and DNA was amplified with primers to the PMT, GPT, and SV40 regions of the HRD transgene (see Materials and Methods). Grey bars, B6; white bars, D2. (A) Histone H3 K9 methylation; (B) MeCP2 binding; (C) histone H3 acetylation (K14 and K9); (D) histone H3 acetylation (K14 only); (E) histone H3 K4 methylation; (F) Brg1 binding.

We next investigated activating chromatin modifications that might be carried by the transgene when integrated into the D2 strain genome. Using a polyclonal antibody to histone H3 acetylated at both the K9 and K14 positions of its N-terminal tail, ChIP experiments showed hyperacetylation in a D2 strain (relative to a B6 strain) at three HRD transgene regions (Fig. 2C). However, the PMT region showed acetylation in both D2 and B6 strains. The latter was unexpected, since the PMT region was highly methylated in the B6 strain at H3 K9 in the associated H3 (Fig. 2A). Virtually identical results were obtained with an antibody specific to the N-terminal tail of H3 acetylated at only K14 (Fig. 2D). While the GPT and SV40 regions are both hyperacetylated at K14 only in the D2 strain, the PMT region again appears acetylated in both the B6 and D2 strains. It was possible that the PMT results might be caused by the amplification of endogenous PMT sequence. However, this is unlikely for three reasons. First, while the 5' primer for the PMT product is from the HRD sequence shared with the endogenous murine metallothionein promoter, the 3' primer is rooted in the Vkappa recombination signal sequence of the transgene. Second, with the transgene-specific primer pair, no PMT product is produced when amplification is attempted from nontransgenic mice (data not shown). Finally, when the PMT product is isolated and sequenced, its sequence is exactly as expected for amplification from the HRD transgene (data not shown). Therefore, the PMT results do appear to be genuine. Given that the metallothionein promoter is functional in hepatocytes, it is possible that histone acetylases are recruited to the PMT region when in this environment, despite the overall silenced state of the transgene. This conclusion is supported by the absence of PMT acetylation in 4-week-differentiated ES cells (see Fig. 4). While the PMT results seem to be true outliers, the GPT and SV40 results further support a pattern of HRD silencing in the adult B6 mouse and activation in the adult D2 mouse.

View larger version (20K): [in this window] [in a new window]   FIG. 4. ChIP assays with 4-week- and 1-week-differentiated ES cells. Chromatin from B6 and D2 strain differentiated embryonic stem cells was immunoprecipitated with the indicated antibodies. See the legend to Fig. 2.

Since the MeCP2 binding and covalent histone modifications strongly supported a strain-specific chromatin structure for the HRD transgene, we also examined the binding of chromatin remodeling proteins. With an antibody for Brg1 (an ATPase subunit of the mammalian SWI/SNF complex), strong Brg1 binding was found in all regions of the transgene in the D2 strain. In contrast, binding was weak in all regions of the B6 transgene (Fig. 2F).

As controls for the ChIP assays, the ß-actin and albumin genes were analyzed. Neither of these genes would be expected to show any strain-specific difference in chromatin structure. Amplified regions of both genes showed equivalent levels of H3 K9 methylation and H3 acetylation in both strains (Fig. 3A and B). This supports the conclusion that the HRD transgene displays a true strain-specific difference in chromatin structure.

View larger version (24K): [in this window] [in a new window]   FIG. 3. ChIP controls (ß-actin and albumin genes). ChIP with the indicated antibodies is shown. Diff-28d, 4-week-differentiated ES cells; Undiff, undifferentiated ES cells.

Four-week-differentiated embryonic stem cells recapitulate the strain-specific HRD chromatin structural changes seen in the adult, but 1-week-differentiated cells show essentially no strain-specific difference. We examined the development of the strain-specific chromatin modifications over developmental time. First, differentiated HRD transgenic embryonic stem cells from the D2 or B6 strain were analyzed. Following differentiation into embryoid bodies for either 4 weeks or 1 week, chromatin was extracted from the ES cells and analyzed by ChIP.

Results from ES cells differentiated for 4 weeks recapitulate those seen with adult tissue. H3 lysine-9 methylation is very evident in the PMT, GPT, and SV40 regions of the B6 HRD transgene (Fig. 4A). However, this modification is undetectable when the transgene is carried in D2 cells (Fig. 4A). H3 acetylation (using an antibody specific for histone H3 acetylated at both K9 and K14) is much stronger in all three regions of the D2 transgene (Fig. 4B). In contrast to the adult liver (Fig. 2), in the differentiated B6 ES cells, H3 in PMT is not hyperacetylated. Finally, Brg1 binding is much stronger in all assayed regions of HRD when the transgene is carried in D2 differentiated ES cells (Fig. 4C). However, there is no strain difference for chromatin patterns of ß-actin and albumin genes (Fig. 3C and D). Thus, fully (4-week) differentiated ES cells generally recapitulate the strain-specific chromatin pattern differences seen in the HRD transgene of adult mice.

Because strain-specific structural changes of the HRD transgene appeared to be complete by 4 weeks of differentiation in culture, we investigated an earlier stage of ES cell differentiation. In ES cells differentiated for only 1 week, the central GPT region of the HRD transgene was analyzed. This region is of specific importance because when it is deleted, the HRD transgene no longer shows strain-specific methylation and GPT alone shows the strain difference (6). At the 1-week stage of differentiation, the results differ from those of the adult stage. H3 K9 methylation appears to be identical for B6 and D2 ES cells (Fig. 4D). K4 methylation and acetylation of H3 are also similar (Fig. 4D). In contrast, MeCP2 binding does appear to be slightly (but reproducibly) greater in the B6 differentiated ES cells (Fig. 4D). Therefore, the chromatin structures of the transgenes appear to be similar for the B6 and D2 strains after 1 week of differentiation (with the possible exception of MeCP2 binding). Also of interest is the fact that the transgenes appear to be at an intermediate chromatin structural state in ES cells of both mouse strains. While silencing marks are found (i.e., histone 3 K9 methylation and MeCP2 binding), activating marks are also present (histone 3 K4 methylation and acetylation).

Undifferentiated ES cells show an intermediate HRD chromatin structure that does not differ by strain. As was the case for the adult, H3 lysine-9 methylation was readily evident in the PMT, GPT, and SV40 regions of the HRD transgene when it was carried in B6 undifferentiated ES cells (Fig. 5A). However, in contrast to those of the adult, all regions of D2 strain HRD also showed H3 K9 hypermethylation (Fig. 5A).

View larger version (18K): [in this window] [in a new window]   FIG. 5. ChIP assays with undifferentiated ES cells. Chromatin from B6 and D2 strain undifferentiated embryonic stem cells was immunoprecipitated with the indicated antibodies. See the legend to Fig. 2.

HRD transgene methylations differ in B6 and D2 ES cells before chromatin change is evident. Earlier work has shown that the strain-specific transgene methylation increases between the undifferentiated and differentiated ES cell stages (34). In order to clarify the temporal relationship between DNA methylation and chromatin structural change, the methylation statuses of the HRD transgenes in B6 and D2 mice, and also in ES cells at various stages, were assayed (Fig. 6). Restriction digestion by BamHI followed by either HpaII (a methylation-sensitive enzyme) or MspI (a methylation-insensitive enzyme; blots not shown) gives a characteristic band pattern on a Southern blot. The upper band represents a full-length BamHI/BamHI fragment (Fig. 1) that is not restricted at any internal HpaII sites (and therefore indicates DNA methylation). Two lower bands represent efficient HpaII digestion. Southern blots reveal that both adult tissue and ES cells at various times of differentiation have negligible amounts of complete HRD methylation in the D2 strain (Fig. 6A). In contrast, B6 adult tissue and ES cells at every stage of differentiation show significant complete DNA methylation. However, partially methylated sequences are present in undifferentiated ES cells and up to 12 days after in vitro differentiation. Development in vitro is slow compared to development in vivo, in which complete HRD methylation is already seen in 6.5-day embryos (34). Reprobing the blot with mitochondrial DNA which does not undergo CpG methylation (14) shows that all the DNAs are equally digestible by HpaII (Fig. 6B).

View larger version (123K): [in this window] [in a new window]   FIG. 6. Southern blot of B6 and D2 DNAs to determine DNA methylation. DNA was digested with BamHI and then HpaII. The blots were probed for GPT (Fig. 1) (A) and mitochondrial DNA (no methylated CpG) (B). me, BamHI (B2 and B3 [Fig. 1]) fragment given when all HpaII sites are CpG methylated and therefore not digested by HpaII; un, bands given when most HpaII sites are unmethylated and therefore cut by HpaII. The size markers (left) are 2, 1.65, 1.4, 1, 0.75, 0.5, 0.4, and 0.3 kb (from top to bottom).

View larger version (29K): [in this window] [in a new window]   FIG. 7. Bisulfite analysis of DNA methylation. Each horizontal line represents the sequence from an individual bacterial colony carrying the region of the transgene being analyzed. Each vertical bar represents a methylated CpG dinucleotide, and a lack of methylation at a CpG dinucleotide is depicted by the absence of a bar. All CpG dinucleotides present in the analyzed sequence are shown above the data. (A) Analysis of CpG methylation in the GPT region of the HRD transgene in B6 and D2 undifferentiated ES cells. The analyzed region was PCR amplified in two halves; hence, the GPT region is indicated as two sets of sequences. (B) Analysis of CpG methylation in the GPT region of the HRD transgene in B6 and D2 adult liver. For the adult tissues, it was possible to PCR amplify the entire GPT region. (C) Analysis of CpG methylation in the GPT region of the HRD plasmid.

Transcription of the HRD transgene. In a previous study, no transcripts from the HRD transgene were found in adult B6 mice, whereas HRD was transcribed in adult D2 mice (33). Surprisingly, despite the significant levels of DNA methylation of HRD in the undifferentiated and differentiated B6 ES cells, HRD was transcribed in these cells up to day 7 (Fig. 8). Representing the more mature embryonic stage, RNA from day 10 embryos was analyzed. No expression was seen in B6 embryos, while D2 embryos continued to express HRD (Fig. 8) (reproducing the findings from adult B6 and D2 mice [33]). Thus, it appears that the repression of transcription in B6 may be delayed until the chromatin has adopted an inactive state.

View larger version (51K): [in this window] [in a new window]   FIG. 8. HRD transcription. HRD transcription was assayed by RT-PCR at various developmental stages of B6 and D2 (10-day in vivo embryos, 7- and 3.5-day in vitro differentiated ES cells, and undifferentiated ES cells). STO (feeder cell) and water are negative controls. (A) GPT; (B) ß-actin.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Based on the data reported here, Ssm1 is responsible for strain-specific differences in both DNA methylation and chromatin structure. In other systems, chromatin modifications can precede DNA methylation, with DNA methylation acting to finalize a heterochromatic structure (1, 24, 30). Conversely, in cultured cells, the transfection of methylated DNA causes chromatin inactivation (16, 29). However, in vivo, instances in which DNA methylation is the first or only silencing event (outside of parental imprinting) are less frequently characterized, and in many cases, the initial causative event in gene silencing is nebulous (27). To gain insight into this problem, we analyzed the DNA methylation and chromatin changes of the HRD transgene at several different times of ES cell differentiation and at the adult stage. Interestingly, while D2 cells show little or no DNA methylation at any stage, the B6 cells do appear to have significant levels of DNA methylation not only after induced differentiation but actually at the undifferentiated cell stage. The ES cell line used here (awB2) was apparently more differentiated than another one (awB4) (34) which we were unable to rescue from frozen samples. The ES cell line awB4 showed essentially no complete DNA methylation before differentiation but became highly methylated during 7 days of differentiation (34). Supporting the notion that differentiation was beginning in the awB2 ES cell line, some of the sequences of the undifferentiated stage showed only a little CpG methylation (Fig. 7). It is thus likely that Ssm1 is acting at the late blastocyst stage and during early embryo differentiation, causing HRD methylation. However, it is unlikely that Ssm1 acts beyond the embryonic stage, since HRD transfected into mature cells does not appear to be under Ssm1-dependent DNA methylation (Engler and Storb, unpublished), while it comes under Ssm1 control when transfected into B6 ES cells upon their differentiation (A. Weng and U. Storb, unpublished data).

While DNA methylation is much greater in B6 cells than in D2 cells already at the undifferentiated ES cell stage, the chromatin structures of the HRD transgenes do not differ greatly between the mouse strains until the embryo is highly developed. Chromatin in undifferentiated ES cells shows modifications that suggest both an open (i.e., H3 acetylation) and closed (H3 K9 methylation) structure, regardless of mouse strain (Fig. 5 and Table 2). The similarity in chromatin structure continues through 1 week of ES cell differentiation. At this time, transgenes in both strains continue to retain covalent histone modifications indicative of both inactive and active genes. Also, no strain-specific difference is yet evident, with the exception of a slight increase in MeCP2 binding in the B6 strain. Finally, once cells have fully differentiated into embryoid bodies (approximately day 28), a strong strain-specific difference in chromatin structure is seen. The transgene has adopted a heterochromatic state in the B6 strain, characterized by high H3 K9 methylation and low H3 acetylation and Brg1 binding. The D2 transgene, in contrast, is euchromatic, with low H3 K9 methylation and high H3 acetylation and Brg1 binding. This strain-specific chromatin structure continues to the adult stage (Fig. 2 and Table 2). Furthermore, suppression of the transcription of the transgene in the B6 strain appears to depend on the chromatin state rather than on DNA methylation.

How the initial DNA methylation is directed is still an open question and will likely have to await the cloning of Ssm1 itself. The Ssm1 locus may be involved not only in methylating the transgene but also in its further inactivation. While Ssm1 has been tightly mapped to a single locus, there are numerous genes in the candidate interval on chromosome 4. Thus, more than one gene may be involved in the Ssm1 effect.

The endogenous targets of Ssm1 are currently unknown, but they must be rather specific, since no overall DNA methylation defect has been seen in the D2 strain. One possible scenario is that Ssm1 plays a role in controlling methylation and subsequent inactivation of repetitive DNA sequences. Their methylation during early differentiation is essential for successful embryogenesis (32). All mouse strains (and vertebrates?) may possess Ssm1-like genes that recognize slightly different sequences or conformations within very similar repetitive elements. The endogenous targets of Ssm1 as well as its physiological function will be defined once Ssm1 has been cloned. The striking effects of Ssm1 on DNA methylation and chromatin structure represent an attractive system for the investigation of developmental epigenetic changes in mammals.

	ACKNOWLEDGMENTS

 
We are very grateful to Peter Engler for advice and reagents and to Grazyna Bozek for breeding mice and technical help. We thank Ken Zaret for methods and Peter Engler for critical reading of the paper.

K.P. was supported by National Institutes of Health training grants GM07281 and GM07283. S.R. is supported by a postdoctoral fellowship from the Cancer Research Institute. This work was supported by National Institutes of Health grant AI39535.

	FOOTNOTES

 
* Corresponding author. Mailing address: University of Chicago, Department of Molecular Genetics and Cell Biology, 920 E. 58th St., Chicago, IL 60637. Phone: (773) 702-4440. Fax: (773) 702-3172. E-mail: stor{at}midway.uchicago.edu.

	REFERENCES

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Bachman, K. E., B. H. Park, I. Rhee, H. Rajagopalan, J. G. Herman, S. B. Baylin, K. W. Kinzler, and B. Vogelstein. 2003. Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene. Cancer Cell 3:89-95.[CrossRef][Medline]

2. Ballestar, E., and A. P. Wolffe. 2001. Methyl-CpG-binding proteins. Targeting specific gene repression. Eur. J. Biochem. 268:1-6.[Medline]

3. Bird, A., and A. Wolffe. 1999. Methylation-induced repression-belts, braces, and chromatin. Cell 99:451-454.[CrossRef][Medline]

4. Chaya, D., T. Hayamizu, M. Bustin, and K. S. Zaret. 2001. Transcription factor FoxA (HNF3) on a nucleosome at an enhancer complex in liver chromatin. J. Biol. Chem. 276:44385-44389.[Abstract/Free Full Text]

5. Clark, S. J., J. Harrison, C. L. Paul, and M. Frommer. 1994. High sensitivity mapping of methylated cytosines. Nucleic Acids Res. 22:2990-2997.[Abstract/Free Full Text]

6. Engler, P., L. T. Doglio, G. Bozek, and U. Storb. 1998. A cis-acting element that directs the activity of the murine methylation modifier locus Ssm1. Proc. Natl. Acad. Sci. USA 95:10763-10768.[Abstract/Free Full Text]

7. Engler, P., D. Haasch, C. Pinkert, L. Doglio, M. Glymour, R. Brinster, and U. Storb. 1991. A strain-specific modifier on mouse chromosome 4 controls the methylation of independent transgene loci. Cell 65:939-947.[CrossRef][Medline]

8. Engler, P., P. Roth, J. Kim, and U. Storb. 1991. Factors affecting the rearrangement efficiency of an Ig test gene. J. Immunol. 146:2826-2835.[Abstract]

9. Engler, P., and U. Storb. 1987. High-frequency deletional rearrangement of immunoglobulin kappa gene segments introduced into a pre-B-cell line. Proc. Natl. Acad. Sci. USA 84:4949-4953.[Abstract/Free Full Text]

10. Engler, P., and U. Storb. 2000. A linkage map of distal mouse chromosome 4 in the vicinity of Ssm1, a strain-specific modifier of methylation. Mamm. Genome 11:694-695.[CrossRef][Medline]

11. Engler, P., A. Weng, and U. Storb. 1993. Influence of CpG methylation and target spacing on V(D)J recombination in a transgenic substrate. Mol. Cell. Biol. 13:571-577.[Abstract/Free Full Text]

12. Fujita, N., S. Watanabe, T. Ichimura, S. Tsuruzoe, Y. Shinkai, M. Tachibana, T. Chiba, and M. Nakao. 2003. Methyl-CpG binding domain 1 (MBD1) interacts with the Suv39h1-HP1 heterochromatic complex for DNA methylation-based transcriptional repression. J. Biol. Chem. 278:24132-24138.[Abstract/Free Full Text]

13. Fuks, F., P. J. Hurd, D. Wolf, X. Nan, A. P. Bird, and T. Kouzarides. 2003. The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J. Biol. Chem. 278:4035-4040.[Abstract/Free Full Text]

14. Groot, G. S., and A. M. Kroon. 1979. Mitochondrial DNA from various organisms does not contain internally methylated cytosine in -CCGG- sequences. Biochim. Biophys. Acta 564:355-357.[Medline]

15. Guy, J., B. Hendrich, M. Holmes, J. Martin, and A. Bird. 2001. A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat. Genet. 27:322-326.[CrossRef][Medline]

16. Hashimshony, T., J. Zhang, M. Bustin, and H. Cedar. 2003. The role of DNA methylation in setting up chromatin structure during development. Nat. Genet. 34:187-192.[CrossRef][Medline]

17. Jackson, J. P., A. M. Lindroth, X. Cao, and S. E. Jacobsen. 2002. Control of CpNpG DNA methylation by the kryptonite histone H3 methyltransferase. Nature 416:556-560.[CrossRef][Medline]

18. Johnson, L. M., X. Cao, and S. E. Jacobsen. 2002. Interplay between two epigenetic marks: DNA methylation and histone H3 lysine 9 methylation. Curr. Biol. 12:1360-1367.[CrossRef][Medline]

19. Kimura, H., and K. Shiota. 2003. Methyl-CpG-binding protein, MeCP2, is a target molecule for maintenance DNA methyltransferase, Dnmt1. J. Biol. Chem. 278:4806-4812.[Abstract/Free Full Text]

20. Kornberg, R., J. LaPointe, and Y. Lorch. 1989. Preparation of nucleosomes and chromatin. Methods Enzymol. 170:3-14.[Medline]

21. Kuo, M.-H., and C. D. Allis. 1999. In vivo cross-linking and immunoprecipitation for studying dynamic protein:DNA associations in a chromatin environment. Methods 19:425-433.[CrossRef][Medline]

22. Lachner, M., and T. Jenuwein. 2002. The many faces of histone lysine methylation. Curr. Opin. Cell Biol. 14:286-298.[CrossRef][Medline]

23. Muller, C., and A. Leutz. 2001. Chromatin remodeling in development and differentiation. Curr. Opin. Genet. Dev. 11:167-174.[CrossRef][Medline]

24. Mutskov, V., and G. Felsenfeld. 2004. Silencing of transgene transcription precedes methylation of promoter DNA and histone H3 lysine 9. EMBO J. 23:138-149.[CrossRef][Medline]

25. Nan, X., H. Ng, C. Johnson, C. Laherty, B. Turner, R. N. Eisenman, and A. Bird. 1998. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393:386-389.[CrossRef][Medline]

26. Peters, A., J. Mermoud, D. O'Carroll, M. Pagani, D. Schweizer, N. Brockdorff, and T. Jenuwein. 2001. Histone H3 lysine 9 methylation is an epigenetic imprint of facultative heterochromatin. Nat. Genet. 30:77-80.

27. Richards, E. J. 2002. Chromatin methylation: who's on first? Curr. Biol. 12:R694-R695.[CrossRef][Medline]

28. Rossant, J. 2001. Stem cells from the mammalian blastocyst. Stem Cells 19:477-482.[Abstract/Free Full Text]

29. Schuebeler, D., M. C. Lorincz, D. M. Cimbora, A. Telling, F. Yong-Quing, E. E. Bouhassira, and M. Groudine. 2000. Genomic targeting of methylated DNA: influence of methylation on transcription, replication, chromatin structure, and histone acetylation. Mol. Cell. Biol. 20:9103-9112.[Abstract/Free Full Text]

30. Tamaru, H., and E. U. Selker. 2001. A histone H3 methyltransferase controls DNA methylation in Neurospora crassa. Nature 414:277-283.[CrossRef][Medline]

31. Tudor, M., S. Akbarian, R. Chen, and R. Jaenisch. 2002. Transcriptional profiling of a mouse model for Rett syndrome reveals subtle transcriptional changes in the brain. Proc. Natl. Acad. Sci. USA 99:15536-15541.[Abstract/Free Full Text]

32. Walsh, C., and T. Bestor. 1999. Cytosine methylation and mammalian development. Genes Dev. 13:26-34.[Abstract/Free Full Text]

33. Weng, A., P. Engler, and U. Storb. 1995. The bulk chromatin structure of a murine transgene does not vary with its transcriptional or DNA methylation status. Mol. Cell. Biol. 15:572-579.[Abstract]

34. Weng, A., T. Magnuson, and U. Storb. 1995. Strain-specific transgene methylation occurs early in mouse development and can be recapitulated in embryonic stem cells. Development 121:2853-2859.[Abstract]

This article has been cited by other articles:

• Shen, H. M., Ratnam, S., Storb, U. (2005). Targeting of the Activation-Induced Cytosine Deaminase Is Strongly Influenced by the Sequence and Structure of the Targeted DNA. Mol. Cell. Biol. 25: 10815-10821 [Abstract] [Full Text]  

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Padjen, K. Articles by Storb, U. Search for Related Content PubMed PubMed Citation Articles by Padjen, K. Articles by Storb, U.