| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Medical Service, Veterans Affairs Palo Alto Health Care System, and Department of Medicine, Stanford University, Palo Alto, California 94304
Address all correspondence and requests for reprints to: Andrew Hoffman, Medical Service, Veterans Affairs Palo Alto Health Care System, 3801 Miranda Avenue, Palo Alto, California 94304. E-mail: arhoffman{at}stanford.edu; or Thanh H. Vu, Stanford Medical School, 3801 Miranda Avenue, Palo Alto, California 94304. E-mail: thanhvu{at}stanford.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One of the most intensively studied imprinted genes is the gene that encodes the IGF-II receptor/mannose-6-phosphate receptor (IGF2R). The gene is maternally expressed in marsupials, rodents, and artiodactyls (including cows and pigs), but it is biallelically expressed in primates, including humans (11, 12, 13). The IGF-II receptor binds IGF-II, a potent mitogen, and then internalizes it and transports it to the lysosome for degradation. In the absence of the receptor, IGF-II levels increase, leading to increased growth in the fetus (14) and, potentially, increased cell growth after birth. Numerous malignancies have been associated with loss of heterozygosity of IGF2R (15). Thus, it has been suggested that IGF2R is a tumor suppressor gene (16), especially in liver (17), where IGF-II may play an important role in carcinogenesis (18).
The mouse Igf2r gene is an imprinted gene that encodes two reciprocally imprinted transcripts, each of which is associated with a differentially methylated region (DMR) (19). The first region (DMR1) includes the promoter for the sense Igf2r transcript, whereas DMR2, which includes the promoter for Igf2r antisense (also known as Air), is located within the second intron of the gene. DMR2 is the putative methylated CpG gametic-imprint, because it is inherited from the female gamete, whereas allele-specific region 1 methylation is not completed until postnatal d 4 (20). Dynamic changes in epigenetic modifications in region 2 may function as signals to guide the establishment of genomic imprinting during development. The very large (108 kb) noncoding RNA Air has been shown to suppress the expression of the sense Igf2r as well as the expression of two other nearby paternally imprinted genes, Slc22a2 and Slc22a3 (21). Deletion of Air leads to loss of Igf2r imprinting in peripheral tissues (22). However, regulation of Igf2r expression in the central nervous system (CNS) does not seem to be determined by Air. Although Air is maternally imprinted in the CNS as well as in peripheral tissues, Igf2r sense transcripts are biallelically expressed in brain (23).
The human IGF2R gene also contains a DMR in intron 2 (24), but no antisense transcripts have been detected, and the gene is always biallelically expressed (25, 26). Thus, it seems that, though the DNA is "marked" for imprinting, the putative imprint is never read, and both alleles are transcribed. To understand this lack of epigenetic readout, we were interested in examining the histone code for the mouse Igf2r gene to learn if histone modifications represented the determining epigenetic feature.
The role of histones as active participants in gene regulation has only recently been appreciated. Previously, these proteins, which are assembled into nucleosomes, forming beads around which the DNA is wrapped, were considered to be relatively inert scaffolding for packaging the genetic material. However, the histone amino-terminal tails extend away from the central core and are thus available for posttranslational, reversible acetylation; methylation; phosphorylation; and ubiquitination (27). The array and combination of these specific attachments to the various histones constitute a distinct code that may entail thousands of histone isoforms and which may make it possible to determine how these proteins will interact with their nearby DNA sequences (28). By reading the histone code, it may be possible to predict which gene products will be transcribed, and thus determine a cell’s RNA repertoire and ultimately its proteome. The task of interpreting the histone code is still in its very early stages, because it is not yet clear that we have identified the full gamut of covalent modifications and attachments, or how the various combinations of such changes (e.g. H4 acetylation and H3 K9 methylation) on a single nucleosome or on adjacent nucleosomes interact with one another to modify gene expression (29). We hypothesize that there may be unique histone codes for imprinted (silenced) alleles in primary and secondary imprints, and that various DMRs may attract a set of unique or characteristic histone modifiers. To investigate these hypotheses, we have studied the histone code of Igf2r and Air in CNS, and peripheral tissues in detail, and we have examined how pharmacologic modulators of histone acetylation and DNA methylation alter the histone code and gene expression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Treatment of cells with 5-aza-deoxycytidine (AzaC) and Trichostatin A (TSA)
We and others have reported that treatment of fibroblast cells with AzaC (a DNA demethylating agent) and TSA (a histone deacetylation-blocking reagent) led to reactivation of the silenced alleles of imprinted genes (1, 32, 33, 34, 35). The reactivation has been linked to a change of DNA methylation, i.e. partial demethylation of a CpG island of the normally silenced allele. F1 fibroblasts at passages three to six were seeded in six-well plates at a density of approximately 2 x 105 cells/well. After 24-h seeding, the culture medium was replaced with media containing AzaC (0.2 µM) and/or TSA (0.33 µM). The cells were treated with AzaC for 7 d (three media replacements) or with TSA for 20 h (or the last 20 h of AzaC+TSA treatment). Similar TSA results were obtained using treatment for 16–24 h. We avoided prolonged treatment with TSA (more than 24 h), because this may lead to cell cycle arrest. After treatment, cells were washed twice with PBS and were analyzed by chromatin immunoprecipitation (ChIP) assay or were used for DNA and RNA preparations.
Nucleic acid preparation
Total nucleic acid (TNA) was prepared using a solution of 4 M guanidinium thiocyanate, 1% 2-mercaptoethanol, and 0.5% Sarkosyl as described previously (36). The homogenate was extracted with phenol/chloroform and then precipitated with 2-propanol. The TNA pellet was dissolved in water and precipitated with ammonium acetate and ethanol. The pellet was washed with 75% ethanol and dissolved in distilled water. RNA was prepared using Tri-reagent and followed the recommended protocol (Sigma, St. Louis, MO).
Mouse Igf2r polymorphisms and DNA sequencing
We sequenced the M. spretus Igf2r by conventional PCR cloning sequencing. We amplified M. spretus genomic DNA by using oligonucleotide primers designed from M. musculus database (GenBank, National Center for Biotechnology Information). The amplified products were verified by agarose gel electrophoresis, and cloned by TOPO-TA cloning kit (Invitrogen). DNA sequencing was performed on an ABI 377 sequencer using Big-Dye terminator chemistry (Perkin-Elmer, Wellesley, MA). Sequences were read on both strands of DNA. Polymorphic sites that were small insertion/deletion or restriction polymorphisms were selected for this study. All polymorphic sites were confirmed by restriction fragment length polymorphism (RFLP)-PCR. Primer sequences for the RFLP-PCR, PCR products, polymorphic sites, and the allelic identification are summarized in Table 1
.
|
Standard PCR and restriction enzyme digestion
Each PCR (3-µl vol under liquid wax) contained approximately 20 ng DNA (or cDNA), 0.1 µM appropriate primers, 50 µM deoxynucleotide triphosphate, 0.2 U KlenTaq I (Ab Peptides, St. Louis, MO), and 0.3 µCi 32P deoxy-CTP (Amersham). PCR conditions were: 95 C for 60 sec, followed by 30–35 cycles of 95 C for 10 sec, and optimal annealing temperature (60–65 C, determined by a gradient-temperature Eppendorff thermal cycler) for 90 sec, and finally 72 C for 10 min. We designed a simple protocol by adding an equal volume (1 µl) of template DNA, equal volume of primers, and equal volume of 3 x master PCR mixture under 15 µl of liquid wax. The master PCR mix, containing KlenTaq I, was added at the annealing step of the first PCR cycle (hot start). To avoid the complication of heteroduplex formation, labeled nucleotide (32P dCTP) or end-labeled primers were added in 1x PCR mixture at the last cycle of amplification. PCR products were digested with appropriate enzymes (New England Biolabs, Beverly, MA; 1 U) in a total vol of 10 µl at 37 C for 12 h under liquid wax. The digested products were separated on 5% polyacrylamide-urea gel, and visualized by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Allelic expression of Igf2r and Air by RT-PCR
Igf2r and Air transcripts from AzaC-and/or TSA-treated cells were amplified by RT-PCR (35 cycles of 95 C, and 65 C). In the case of the Igf2r transcript, we used primers p#6105 in exon 47 and p#971 in exon 48 (Fig. 1) to amplify 164 b cDNA-specific product (Table 1
). A Hae III site common for both parental alleles serves as an internal digestion control, whereas the unique polymorphic Hae III is present only in the paternal Spretus allele. In the case of the Air transcript that extends beyond exon 1 of Igf2r, a 16-bp deletion in M. Spretus (primers p#6458 and p#8–04) was used to identify the two parental alleles (23). Relative levels of the two parental alleles were based on PhosphorImager scanning density.
|
ChIP.
Antibodies obtained from Upstate Biotechnology (Waltham, MA) include: H4-Ac (acetyl lysines 5, 8, 12, and 16, catalog no. 06-866), H3-Ac (diacetyl lysines 9, and 14, catalog no. 06-599), H3 K4-Me (methyl lysine 4 of histone H3, catalog no. 07-030), and H3 K9-Me (dimethyl lysine 9, catalog no. 07-212). Antibodies against H3 K9-Ac (acetyl lysine 9 of histone H3, catalog no. 9617) and antibodies against H3-S10-P (phosphorylated serine 10 of histone H3, catalog no. 9701) were from Cell Signaling Technology (Beverly, MA). The specificity of antibodies against H3-S10-P was not affected by the modification of the adjacent lysine 9 (unmodified, acetylated, or methylated lysine 9) (data from Cell Signaling Technology). Antibodies against mono-, and trimethyl lysine 4, and lysine 9 of histone H3 were obtained from Abcam, Cambridge, UK (monomethyl K4, no. ab8895; trimethyl K4, catalog no. ab8580; monomethyl K9, catalog no. ab9045; trimethyl K9, catalog no. ab8898). All antibodies recognize the specific histone modifications (data from the suppliers). ChIP assays were carried out according to the protocol supplied by Upstate Biotechnology (Lake Placid, NY). Briefly, sonicated chromatin was diluted 10-fold in ChIP dilution buffer (200 µl), precleared with 80 µl salmon sperm DNA/protein A agarose for 1 h at 4 C with rotation. A portion of the protein A-purified chromatin (20 µl) was used to prepare DNA as the "input" sample. Antibodies (2–5 µl) were added to the clarified chromatin (180 µl) and incubated overnight with rotation. Sixty microliters of protein A agarose were added to the antibody-histone mix and incubated at 4 C for 1 h with rotation. The protein A-agarose/histone complex was collected by gentle centrifugation and washed three times with ChIP buffers, and the bound chromatin was eluted in 500 µl elution buffer. After adding 20 µl of 5 M NaCl, protein-DNA cross-linking was reversed by heating at 65 C for 4 h. Samples were treated with proteinase K, purified by MiniElute PCR purification kit (QIAGEN, Valencia, CA), and then eluted in 100 µl of low-TE buffer (1 mM Tris, 0.1 mM EDTA).
Analysis of modified histones
The relative enrichment of modified histones associated with DNA, obtained from ChIP, was determined as described previously (37) using ß-actin or Ig-k as an internal control (38). Each sample was quantified in duplicate. All primer sets were tested for the absence of primer-dimer products by conventional PCR using radioisotope and polyacrylamide-urea gel.
Quantitative real-time (Q)-PCR
Enrichments of modified histones in DNA, obtained from some of the ChIP assays, were also determined by (Q)-PCR using the ABI Prism 7900HT sequence detector, following the ABI protocol. The Q-PCR assays were run in triplicate on 384-well plates. We designed allele-specific oligonucleotide primers to amplify specifically each parental allele (Table 1). All primer sets for Q-PCR were free of primer-dimer products. We used SYBR Green in our Q-PCR assays. At the end of the Q-PCR amplification, we ran a melting curve analysis, to confirm the homogeneity of all Q-PCR products. Relative enrichment of a given target sequence by a specific antibody is determined by a delta Ct and delta-delta Ct calculation, by an ABI protocol, with reference to ß-actin or mouse ribosomal L7 protein gene control (Table 1).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To investigate the histone modifications in the chromatin region of the mouse Igf2r gene, by ChIP assay, we made PCR primers to amplify short DNA fragments of approximately 100 bases crossing various regions of the Igf2r gene. These primers were designed to amplify DNA fragments enriched in the chromatin immunoprecipitated fraction; and therefore, they can determine the presence of the histone modification specified by the antibody used in the assay at specific nucleosomes associated with the specific genomic DNA fragments. We used the primers to amplify selected sequences in the DMRs, 3.2 kb upstream of the Igf2r gene, exon 2 (where the sense and antisense transcripts overlap), and exon 48 of the 3' region of Igf2r (Fig. 1A, and Table 1).
We scanned the acetylation of H3 in nucleosomes across the overlapping sense and antisense region using the ChIP assay with a multiplex PCR containing target DNA primers and internal control primers. As shown in Fig. 1B, all target DNAs of predicted sizes (listed in Table 1) were amplified along with the internal controls in the input DNA. After immunoprecipitation using the anti-H3-Ac antibodies, no target DNAs were observed at 3.2 kb upstream (Fig. 1B, amplicon a1) and at exon 48 (Fig. 1B, amplicon e) while consistent amplification of internal controls were evident (Fig. 1B, compare bottom panel to top panel). Histone H3 acetylation was enriched in the DMR1, DMR2, and exon 2 to levels comparable with that of ß-actin (Fig. 1B, amplicons b through d, compare top and bottom panels).
Though histone acetylation was found in the active chromosomal domain, our results indicate that the acetylation modification is more enriched in the promoter region of the Igf2r (DMR1) and the Air (DMR2). The regions upstream of the Igf2r (amplicon a1) and upstream of Air (amplicon e, exon 48) or downstream of the Igf2r (exon 48) were nearly deficient of the acetylation modification.
Allele-specific histone modification in DMR1 and DMR2
We then investigated allele-specific histone modification using a panel of antibodies against histone acetylation and histone methylation. To differentiate histone modifications in the two parental alleles of the F1 interspecific mice, we sequenced the DMR1 and DMR2 of the wild-derived M. spretus (the paternal allele in the F1 mice) by standard PCR cloning and sequencing. We compared the DMR sequences of the M. spretus with those of the M. musculus (maternal allele, available in the GenBank database) and found six allelic differences between these strains (Table 1, amplicons c1, a, b, c, d, and e). Three of these polymorphisms involve small insertion/deletion differences that are readily quantified by PCR fragment length polymorphism without amplification bias. The remaining three polymorphisms are RFLPs that require restriction digestion. In previous publications, we have used three of these polymorphisms to distinguish the parental alleles (Table 1, amplicons b, d, and e) (34).
We ran PCRs of the ChIP assay along with input DNA (before immunoprecipitation) using ß-actin or Ig-k as internal PCR controls. The active ß-actin gene was used in the ChIP assay with anti H4-Ac, H3-Ac, and H3 K4-Me (methylation) antibodies. The inactive Ig-k gene was used in the ChIP assay with anti H3 K9-Me, because it has been reported that H3 K9-methylation associates with inactive genes (7, 27). Representative data of duplicate PCR from each ChIP assay are shown in Fig. 2, A and B. Calculation of the relative enrichment of specific histone modifications in each parental allele, compared with those of input DNA before immunoprecipitation, was based on the formula in Fig. 2C (40). Relative enrichments as mean values from duplicate PCR are shown at the bottom panels in Fig. 2, A and B and are plotted in Fig. 2D.
|
, in control F1 cells without AzaC and TSA treatment (-/-), H4-Ac, H3-Ac, and H3 K4-Me associated strongly with the expressed alleles across the DMR1 and DMR2 [expressed (white columns) vs. silenced (black columns)]. In the region near the Igf2r transcription site (panel b) and in DMR2 (panels c and d), H3 K9-Me was in low abundance in nucleosomes from both parental alleles.
We further quantified the relative enrichment of histone modifications in each parental allele in DMR1 (amplicon b region) and DMR2 (amplicon d region) by Q-PCR using allele-specific primers (Table 1, ChIP-Q-PCR). As shown in Fig. 2E, acetylated histone (H3-Ac) and methylated lysine 4 (H3 K4-Me) were exclusively enriched in nucleosomes from the expressed alleles (white columns, maternal allele in DMR1 and paternal allele in DMR2). The silenced alleles (black columns, paternal allele in DMR1 and maternal allele in DMR2) were devoid of H3-Ac and H3 K4-Me. Both parental alleles had low levels of lysine 9 methylation (H3 K9-Me). The Q-PCR results are in accordance with the ChIP-PCR data in Fig. 2D and show absolute allele-specific histone modification, reflecting the high specificity of the antibodies employed in the ChIP experiments, and the sensitivity of the real-time PCR assay.
Reactivation of silenced alleles by AzaC and TSA associates with biallelic histone acetylation
We investigated the pharmacological modulation of allele-specific histone modifications in the reactivation of the silenced allele. We treated the F1 fibroblast cells with AzaC (0.2 µM) and/or TSA (0.33 µM) and then quantified Igf2r and Air allelic abundance using RT-PCR, and we assessed the histone modifications in DMR1 and DMR2 using the ChIP assay. Combined treatment with AzaC+TSA led to approximately 1% reactivation of the imprinted Igf2r paternal allele and to approximately 6% of the imprinted maternal Air allele (Fig. 3). Although the magnitudes of the reactivation were low, the detection of reactivation was not artifactual. In the case of Igf2r, the reactivated paternal allele was detected positively as a specific band by restriction enzyme digestion; and in the case of Air, the reactivated maternal allele was detected by insertion/deletion PCR analysis.
|
, A, B, and D). As shown in Fig. 2D, AzaC alone did not significantly change histone acetylation (both H4-Ac and H3-Ac) or histone methylation at H3 K4 and H3 K9 across the DMR and exon 48 regions [Fig. 2D, compare nontreatment (-/-) with AzaC treatment (+/-) in a, b, c, d, and e]. In the silenced allele, TSA or TSA+AzaC slightly increased H3-Ac in the DMR1 region and H4-Ac in the DMR2 region (Fig. 2D, DMR1 and DMR2, H4-Ac and H3-Ac, black columns).
Tissue-specific reactivation of Igf2r silenced allele in CNS correlates with biallelic acetylation and biallelic H3 K4 methylation
To understand mechanisms underlying the CNS-specific biallelic expression of Igf2r sense transcript despite maintenance of Air imprinting, we analyzed histone modifications in the DMR1 and DMR2 in peripheral tissues, and in CNS. Representative data showing duplicate ChIP-PCR assays, using anti-H3-Ac and anti-H4-Ac antibodies in DMR1 (amplicon b) and DMR2 (amplicon d) from the F1 mice, are shown in Fig. 4A.
|
, DMR1). In contrast, at the same nucleosome location in CNS, where Igf2r was not imprinted, there was no allele-specific histone acetylation; both parental alleles had the relative enrichments of approximately 1.0 (0.85–1.22), and the allelic acetylated histone ratios were approximately1.0 (0.77–0.93) (Fig. 4A).
Despite the loss of Igf2r imprinting, imprinting of the Igf2r antisense (Air) was maintained in CNS. ChIP assays demonstrated predominant allele-specific H3-Ac and H4-Ac in DMR2 in all tissues including CNS (Fig. 4A, DMR2). The ratios of acetylated histone in expressed-paternal/silenced-maternal allele varied from 1.50–5.56.
Methylation of histone 3 at lysine 4 (H3 K4-Me), but not at lysine 9, was predominantly associated with the expressed allele in the DMRs in liver (and skin fibroblasts, see above) where Igf2r was imprinted. The allele-specificity of H3 K4-Me was not present in CNS where there was loss of Igf2r imprinting: both parental alleles had equal enrichment of H3 K4-Me in the DMR1 (Fig. 4B, top panel).
To correlate the histone modifications (acetylation and methylation) of nucleosomes in the DMRs to the DNA methylation status of the two parental alleles, we performed MR-PCR, a modified COBRA (combined bisulfite restriction analysis) method (41, 42), on the bisulfite-treated DNAs from liver, kidney, and CNS of F1 mice. As shown in Fig. 5 the maternal allele was always unmethylated, whereas the paternal allele was predominantly methylated in liver and in kidney, but not in CNS.
|
Most of the published data (including the data reported herein) concerning H3 K4 and H3 K9 methylation refer to dimethyl lysine. To test whether the three states of methylation of lysine 4 and lysine 9, and H3 K4 and H3 K9 correlate with the imprinting and loss of imprinting of Igf2r, we used antibodies directed against mono- and trimethyl H3 K4 and H3 K9 in our ChIP assays. In all tissue tested (liver, kidney, and CNS), antibodies directed against trimethyl lysine yielded results virtually identical to those from antibodies directed against dimethyl lysine: in DMR1, the active maternal allele in liver and kidney was associated predominantly with H3 K4 di- and trimethylation (Fig. 6A). In CNS, both active maternal and active paternal alleles were associated with H3 K4 di-and trimethylation. Low levels of H3 K9 di- and trimethylation were associated with both parental alleles in liver, kidney, and CNS, which may represent, as discussed above, basal levels of association. ChIP assays using anti-monomethylated H3 K4 and H3 K9 antibodies yielded much weaker signals, although more sensitive ChIP-Q-PCR assays revealed patterns of modification similar to those from dimethyl lysine antibodies (Li et al., unpublished). Therefore, it is likely that, in the case of Igf2r, the histone code makes no distinction between the three states of mono-, di-, and trimethylation of H3 K4 and H3 K9.
|
). It is interesting to note that acetylated lysine 9 (H3 K9-Ac) was detected exclusively in the active allele in liver (DMR1 and DMR2) and CNS (DMR2). Inactive or imprinted alleles in liver and in CNS (DMR2) were devoid of H3 K9-Ac. It seems, therefore, that the acetylation status of H3 K9 serves as a marker of allelic activation or silencing in Igf2r sense and antisense (DMR1 and DMR2), at least in liver and in CNS. Biallelic expression of Igf2r sense in CNS correlates with biallelic acetylation of H3 K9 in DMR1.
Phosphorylated serine 10 of H3 (H3 S10-P) has been reported to be associated with active genes (44, 45). To investigate allele-specific modifications of H3 S10-P across the Igf2r gene, we used antibodies against H3 S10-P in our ChIP assays that recognize phosphorylated serine 10 independent of modifications of the adjacent lysine 9 (data from Cell Signaling Technology). As shown in Fig. 6C, H3 S10-P was found only in the DMR2. No H3 S10-P was seen in the DMR1 in liver, kidney, and CNS; whereas consistent amplification of positive control (ß-actin) was seen in all ChIP assays. The results were consistent at polymorphic sites in the DMR1 (amplicons a and b) and in the DMR2 (amplicons c and d). In liver, kidney, and CNS, H3 serine 10 phosphorylation associated exclusively with the active paternal allele in DMR2. There were no H3 S10-P modifications on the silenced maternal allele in DMR2. However, in DMR1, both active and silenced parental alleles (in liver and kidney) and both active parental alleles (in CNS) were devoid of the H3 S10-P modifications. Outside of the DMRs, in exon 48, H3 S10-P was detected equally in both parental alleles (Fig. 6C), indicating that H3 S10-P modifications outside of the DMRs did not correlate with allelic-specific expression.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The epigenetic regulation of Igf2r expression by DNA methylation has been extensively studied. In this report, we have expanded our knowledge concerning the histone modifications regulating gene expression. It has been postulated that a histone code, encompassing numerous posttranslational covalent modifications to these evolutionarily conserved proteins, determines the pattern of gene expression. In general, increases in histone acetylation have been associated with an open chromatin conformation and enhanced gene expression. Histone methylation may induce an active or inactive chromatin state, depending on the particular amino acid that is modified. Studying the Igf2r gene locus allows us to test and refine this histone code hypothesis by studying a gene that has two promoters, one of which is expressed only by the maternal allele, whereas the other is expressed only by the paternal allele. Thus, examination of histone modifications of this gene in various tissues, including CNS, will allow us to determine whether there are consistent epigenetic changes along the gene that correlate with tissue-specific gene expression.
In most tissues, reciprocal imprinting of Igf2r sense and antisense is associated with DNA methylation of the DMRs: methylated DNA is found with the silenced allele, whereas unmethylated DNA is found with the active allele. In CNS, despite the biallelic expression of Igf2r sense, allelic expression of both Igf2r sense and Air is still strictly dictated by allelic DNA methylation (51). The allelic DNA methylation also correlates with a specific set of histone modifications. In general, the methylated DNA of the silenced allele is associated with H3 K9 Me, and the unmethylated DNA of the active allele is associated with H4 Ac, H3 Ac, and H3 K4 Me. This "universal" histone code for silenced and active genes is found in silenced and active alleles in the imprinted Igf2r in this report. Fournier et al. (10), in a recent independent study, also reported the universal histone code in imprinted loci in Snrpn, U2af1-rs1, and Igf2r.
It is interesting to note that H3 and H4 histones in the DMR1 and DMR2 were acetylated differentially after TSA treatment. TSA (with and without AzaC) leads to equal biallelic acetylation in histone H3 (but not in H4-Ac) in DMR1 and equal biallelic acetylation in histone H4 (but not in H3-Ac) in DMR2 (Fig. 2D, H4-Ac and H3-Ac, b and d panels). It is likely that, in the DMR1 promoter region, histone H3 is more susceptible to pharmacologic modulation than histone H4, leading to increased H3-Ac. In the DMR2, however, histone H4 is more susceptible than H3, reflecting differential structure of local nucleosomes in the two DMRs.
Biallelic acetylation of one (H3 or H4, but not both) of these histones may correlate with partial reactivation of the silenced allele to levels of 1–6% of those of the active allele in Fig. 3. The reactivation of the silenced allele by AzaC and/or TSA is therefore associated with (but not completely controlled by) increased H3-, H4-acetylation, and H3 K4 methylation in the DMR1 and DMR2.
Biallelic H3 K9-methylation at DMR1 was observed in CNS that demonstrated biallelic expression of Igf2r. The loss of Igf2r imprinting in CNS, however, was not directly correlated with biallelic H3 K9 methylation, because this was also observed in liver, a tissue that demonstrated maintenance of Igf2r imprinting (Fig. 4B, DMR1). It is likely that the low levels of biallelic H3 K9-methylation observed in liver and CNS reflect the basal levels of the H3 K9-Me in the DMR1 location. (Q)-PCR-ChIP assays also confirmed the basal, low levels of H3 K9-Me in this DMR1 location (Fig. 2E).
In DMR2, H3 K4-Me was found associated predominantly with the expressed paternal allele, whereas H3 K9-Me associated exclusively with the silenced maternal allele (Fig. 4B, DMR2). The association of H3 K9-Me with the silenced allele was more profound in tissues (liver and CNS) than in the cultured fibroblast cells.
In Igf2r imprinting, DMR2, which harbors the gametic methylated CpG pattern, has been referred to as the gametic or primary imprint, whereas DMR1, which is established after implantation, is considered to be a secondary imprint. Because allele-specific H3 S10-P modifications were observed only in DMR2, but not in DMR1 or outside of the DMRs (biallelic modification), we speculate that the allele-specific H3 S10-P modification exclusively marks the primary imprint. Further investigation of the H3 S10 phosphorylation in other imprinting loci having both primary and secondary DMRs (or imprints), such as Igf2-H19 and Nesp55-Gnas, will shed light on this hypothesis.
Are there any unique features of the histone code in imprinted genes? Our present data on Igf2r imprinting indicate the following features of the code (1). Histone modifications occur predominantly in the nucleosomes in the promoter region near the transcription site (2). H3 K9 acetylation is associated exclusively with the active allele, whereas deacetylation of H3 K9 is associated with the silenced allele (3). H3 K9 methylation and H3 S10 phosphorylation are associated exclusively with the silenced and active alleles, respectively, of the primary imprint but not of the secondary imprint. Though data on H3 S10 phosphorylation in imprinted genes are scarce, H3 K9 methylation has been demonstrated to be an epigenetic imprint of the inactive X chromosome in female mammals. H3 K9 methylation also occurs predominantly in the inactive allele in the primary (gametic methylated CpG) DMRs of Snrpn, U2af1-rs1, and Igf2r genes (10).
In Fig. 7, we propose a model of allelic histone modification in the primary (DMR2) and secondary (DMR1) imprints of Igf2r. In regard to the expression of Igf2r, DNA methylation and histone deacetylation in DMR1 are found in the silenced allele, whereas in unmethylated DNA, histone acetylation (including H3 K9-Ac) and H3 K4 methylation are present in the active allele. Air contains similar epigenetic DNA and histone marks; and in addition, it contains specific modifications, which we believe to be the characteristic of the primary imprint: H3 K9 methylation in the silenced allele, and H3 S10 phosphorylation in the active allele.
|
In an analogous situation at the Igf2/H19 imprinted locus, the methylated CpG primary imprint located 2–4 kb upstream of the H19 gene can generate secondary epigenetic DNA modifications in the H19 promoter region (secondary imprint). It is the secondary imprint (not the primary one) that mediates paternal-specific silencing of H19 (52). Interestingly, the secondary methylation imprint is stable through multiple cell divisions, even in the absence of the primary imprint (52), suggesting the functional independence of the secondary imprint once it has been established. Further explication of the histone code will likely provide new clues to specific epigenetic marks that constitute a unique histone code for imprinted genes.
|
Footnotes |
|---|
Y.Y. and T.L. contributed equally.
Abbreviations: AzaC, 5-Aza deoxycytidine; ChIP, chromatin immunoprecipitation; CNS, central nervous system; DMR, differentially methylated region; H4-Ac, acetyl lysines 5, 8, 12, and 16 of histone H4; H3-Ac, acetyl lysines 9 and 14 of histone H3; H3 K9-Ac, acetyl lysine 9 of histone H3; H3 K4-Me, methyl lysine 4 of histone H3; H3 K9-Me, methyl lysine 9 of histone H3; H3 S10-P, phosphorylated serine 10 of histone H3; Igf2, mouse IGF 2; Me, methylation; Q-PCR, quantitative, real-time PCR; RFLP, restriction fragment length polymorphism; RT, reverse transcription; TSA, Trichostatin A.
Received June 26, 2003.
Accepted for publication September 3, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Henckel, K. Nakabayashi, L. A. Sanz, R. Feil, K. Hata, and P. Arnaud Histone methylation is mechanistically linked to DNA methylation at imprinting control regions in mammals Hum. Mol. Genet., September 15, 2009; 18(18): 3375 - 3383. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. V. Dindot, R. Person, M. Strivens, R. Garcia, and A. L. Beaudet Epigenetic profiling at mouse imprinted gene clusters reveals novel epigenetic and genetic features at differentially methylated regions Genome Res., August 1, 2009; 19(8): 1374 - 1383. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Wagschal, H. G. Sutherland, K. Woodfine, A. Henckel, K. Chebli, R. Schulz, R. J. Oakey, W. A. Bickmore, and R. Feil G9a Histone Methyltransferase Contributes to Imprinting in the Mouse Placenta Mol. Cell. Biol., February 1, 2008; 28(3): 1104 - 1113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. A. Garcia, S. B. Hake, R. L. Diaz, M. Kauer, S. A. Morris, J. Recht, J. Shabanowitz, N. Mishra, B. D. Strahl, C. D. Allis, et al. Organismal Differences in Post-translational Modifications in Histones H3 and H4 J. Biol. Chem., March 9, 2007; 282(10): 7641 - 7655. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yamasaki-Ishizaki, T. Kayashima, C. K. Mapendano, H. Soejima, T. Ohta, H. Masuzaki, A. Kinoshita, T. Urano, K.-i. Yoshiura, N. Matsumoto, et al. Role of DNA Methylation and Histone H3 Lysine 27 Methylation in Tissue-Specific Imprinting of Mouse Grb10 Mol. Cell. Biol., January 15, 2007; 27(2): 732 - 742. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Monk, P. Arnaud, S. Apostolidou, F. A. Hills, G. Kelsey, P. Stanier, R. Feil, and G. E. Moore From the Cover: Limited evolutionary conservation of imprinting in the human placenta PNAS, April 25, 2006; 103(17): 6623 - 6628. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Tufarelli The silence RNA keeps: cis mechanisms of RNA mediated epigenetic silencing in mammals Phil Trans R Soc B, January 29, 2006; 361(1465): 67 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yamasaki, T. Kayashima, H. Soejima, A. Kinoshita, K.-i. Yoshiura, N. Matsumoto, T. Ohta, T. Urano, H. Masuzaki, T. Ishimaru, et al. Neuron-specific relaxation of Igf2r imprinting is associated with neuron-specific histone modifications and lack of its antisense transcript Air Hum. Mol. Genet., September 1, 2005; 14(17): 2511 - 2520. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Li, T. H. Vu, G. A. Ulaner, E. Littman, J.-Q. Ling, H.-L. Chen, J.-F. Hu, B. Behr, L. Giudice, and A. R. Hoffman IVF results in de novo DNA methylation and histone methylation at an Igf2-H19 imprinting epigenetic switch Mol. Hum. Reprod., September 1, 2005; 11(9): 631 - 640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Sha and A. Fire Imprinting Capacity of Gamete Lineages in Caenorhabditis elegans Genetics, August 1, 2005; 170(4): 1633 - 1652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. H. Vu, T. Li, and A. R. Hoffman Promoter-restricted histone code, not the differentially methylated DNA regions or antisense transcripts, marks the imprinting status of IGF2R in human and mouse Hum. Mol. Genet., October 1, 2004; 13(19): 2233 - 2245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. V. Dindot, P. W. Farin, C. E. Farin, J. Romano, S. Walker, C. Long, and J. A. Piedrahita Epigenetic and Genomic Imprinting Analysis in Nuclear Transfer Derived Bos gaurus/Bos taurus Hybrid Fetuses Biol Reprod, August 1, 2004; 71(2): 470 - 478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Sakamoto, J. Liu, A. Greene, M. Chen, and L. S. Weinstein Tissue-specific imprinting of the G protein Gs{alpha} is associated with tissue-specific differences in histone methylation Hum. Mol. Genet., April 15, 2004; 13(8): 819 - 828. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Li, T. H. Vu, G. A. Ulaner, Y. Yang, J.-F. Hu, and A. R. Hoffman Activating and silencing histone modifications form independent allelic switch regions in the imprinted Gnas gene Hum. Mol. Genet., April 1, 2004; 13(7): 741 - 750. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. BRAIDOTTI, T. BAUBEC, F. PAULER, C. SEIDL, O. SMRZKA, S. STRICKER, I. YOTOVA, and D.P. BARLOW The Air Noncoding RNA: An Imprinted cis-silencing Transcript Cold Spring Harb Symp Quant Biol, January 1, 2004; 69(0): 55 - 66. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
